Nanoparticulates As Drug Carriers
EDITOR VLADIMIR P TORCHILIN Northeastern University, USA Contents 2 Nanoparticle Engineering 30 2.1 Drug release mechanisms 32 3 Site-specific Targeting with Nanoparticles: Importance of Size and Surface Properties 33 4 Conclusions 37 References 38 4. Genetic Vaccines: A Role for Liposomes 43 Gregory Gregoriadis, Andrew Bacon, Brenda McCormack and Peter Laing 1 Introduction 43 2 The DNA Vaccine 44 3 DNA Vaccination via Liposomes 45 3.1 Procedure for the entrapment of plasmid DNA into liposomes 46 3.2 DNA immunization studies 47 3.3 Induction of a cytotoxic T lymphocyte (CTL) response by liposome-entrapped plasmid DNA 50 4 The Co-delivery Concept 51 References 53 5. Polymer Micelles as Drug Carriers 57 Elena V. Batrakova, Tatiana K. Bronich, Joseph A. Vetro and Alexander V. Kabanov 1 Introduction 57 2 Polymer Micelle Structures 58 2.1 Self-assembled micelles 58 2.2 Unimolecular micelles 61 2.3 Cross-linked micelles 62 3 Drug Loading and Release 63 3.1 Chemical conjugation 63 3.2 Physical entrapment 64 3.3 Polyionic complexation 66 4 Pharmacokinetics and Biodistribution 68 5 Drug Delivery Applications 72 5.1 Chemotherapy of cancer 72 5.2 Drug delivery to the brain 76 5.3 Formulations of antifungal agents 77 5.4 Delivery of imaging agents 77 5.5 Delivery of polynucleotides 78 Contents xv 6 Clinical Trials 79 7 Conclusions 79 References 80 6. Vesicles Prepared from Synthetic Amphiphiles — Polymeric Vesicles and Niosomes 95 Ijeoma Florence Uchegbu and Andreas G. Schatzlein 1 Introduction 95 2 Polymeric Vesicles 96 2.1 Polymer self assembly 97 2.2 Polymers bearing hydrophobic pendant groups 98 2.3 Block copolymers 101 2.4 Preparing vesicles from self-assembling polymers 102 2.5 Self assembling polymerizable monomers 103 3 Polymeric Vesicle Drug Delivery Applications 104 3.1 Drug targeting 104 3.2 Gene delivery 105 3.3 Responsive release 106 3.3.1 pH 106 3.3.2 Enzymatic 106 3.3.3 Magnetic 107 3.3.4 Oxygen 108 4 Non-ionic Surfactant Vesicles (Niosomes) 108 4.1 Self assembly 108 4.2 Polyhedral vesicles and giant vesicles (Discomes) Ill 4.3 Vesicle preparation 113 5 Niosome Delivery Applications 113 5.1 Drug targeting 113 5.1.1 Anti cancer drugs 113 5.1.2 Anti infectives 115 5.1.3 Delivery to the brain 115 5.2 Topical use of niosomes 116 5.2.1 Transdermal 116 5.2.2 Ocular 116 5.3 Niosomal vaccines 116 5.4 Niosomes as imaging agents 117 6 Conclusions 117 References 117 xv i Contents 7. Recent Advances in Microemulsions as Drug Delivery Vehicles 125 M Jayne Lawrence and Warankanga Warisnoicharoen 1 Definition 125 1.1 Microemulsion versus an emulsion 125 1.2 Microemulsion versus a nanoemulsion 126 1.3 Microemulsions 128 1.4 Microemulsions, swollen micelles, micelles 129 1.5 Microemulsions and cosolvent systems 130 2 Microemulsions as Drug Delivery Systems 130 2.1 Self-emulsifying drug delivery systems (SEDDS) 131 2.2 Related systems 133 2.2.1 Microemulsion gels 133 2.2.2 Double or multiple microemulsions 134 2.3 Processed microemulsion formulations 134 2.3.1 Solid state or dry emulsions 134 3 Formulation 135 3.1 Surfactants and cosurfactants 136 3.2 Oils 138 3.3 Characterization 139 4 Routes of Administration 139 4.1 Oral 139 4.1.1 Proteins and peptides 140 4.1.2 Other hydrophilic molecules 141 4.1.3 Hydrophobic drugs 142 4.2 Buccal 144 4.3 Parenteral 144 4.3.1 Long circulating microemulsions 147 4.3.2 Targeted delivery 148 4.4 Topical delivery 148 4.4.1 Dermal and transdermal delivery 148 4.5 Ophthalmic 154 4.6 Vaginal 156 4.7 Nasal 157 4.8 Pulmonary 158 4.8.1 Antibacterials 159 5 Conclusion 160 References 160 Contents xvii 8. Lipoproteins as Pharmaceutical Carriers 173 Suwen Liu, Shining Wang and D. Robert Lu 1 Introduction 173 2 The Structure of Lipoproteins 174 3 Chylomicron as Pharmaceutical Carrier 175 4 VLDL as Pharmaceutical Carrier 176 5 LDL as Pharmaceutical Carrier 177 5.1 LDL as anticancer drug carriers 178 5.2 LDL as carriers for other types of bioactive compounds . . . .179 5.3 LDL for gene delivery 179 6 HDL as Pharmaceutical Carriers 179 7 Cholesterol-rich Emulsions (LDE) as Pharmaceutical Carriers . . . .180 8 Concluding Remark 181 References 182 9. Solid Lipid Nanoparticles as Drug Carriers 187 Karsten Mader 1 Introduction: History and Concept of SLN 187 2 Solid Lipid Nanoparticles (SLN) Ingredients and Production . . . .188 2.1 General ingredients 188 2.2 SLN preparation 189 2.2.1 High shear homogenization and ultrasound 189 2.3 High pressure homogenization (HPH) 189 2.4 Hot homogenization 190 2.5 Cold homogenization 190 2.5.1 SLN prepared by solvent emulsification / evaporation 191 2.5.2 SLN preparations by solvent injection 191 2.5.3 SLN preparations by dilution of microemulsions or liquid crystalline phases 192 2.6 Further processing 193 2.6.1 Sterilization 193 2.6.2 Drying by lyophilization, nitrogen purging and spray drying 194 3 SLN Structure and Characterization 196 4 The "Frozen Emulsion Model" and Alternative SLN Models . . . . 200 5 Nanostructured Lipid Carriers (NLC) 201 6 Drug Localization and Release 202 xviii Contents 7 Administration Routes and In Vivo Data 203 8 Summary and Outlook 205 References 205 10. Lipidic Core Nanocapsules as New Drug Delivery Systems 213 Patrick Saulnier and Jean-Pierre Benoit 1 Introduction 213 2 Lipidic Nanocapsule Formulation and Structure 215 2.1 Process 215 2.2 Influence of the medium composition 216 2.3 Structure and purification of the LNC by dialysis 217 2.4 Imagery techniques 218 3 Electrical and Biological Properties 219 3.1 Electro kinetic comportment 219 3.2 Evaluation of complement system activation 220 4 Pharmacokinetic Studies and Biodistribution 220 5 Drug Encapsulation and Release 222 5.1 Ibuprofene 222 5.2 Amiodarone 223 6 Conclusions 223 References 224 11. Lipid-Coated Submicron-Sized Particles as Drug Carriers 225 Evan C. linger, Reena Zutshi, Terry O. Matsunaga and Rajan Ramaswami 1 Technology 225 2 Ultrasound Contrast Agents 228 3 Sonothrombolysis ^r_._ 232 4 Clinical Studies 237 5 Blood Brain Barrier 239 6 Drug Delivery 242 6.1 Targeted bubbles 242 6.2 Targeted submicron-sized droplets 244 7 Gene Delivery 245 8 Oxygen Delivery 247 9 Pulmonary Delivery 248 10 Conclusion 249 References 250 Contents xix Nanocapsules: Preparation, Characterization and Therapeutic Applications 255 Ruxandra Grefand Patrick Couvreur 1 Introduction 255 2 Preparation 257 2.1 Nanocapsules obtained by interfacial polymerization 257 2.1.1 Oil-containing nanocapsules 257 2.1.2 Nanocapsules containing an acqueous core 259 2.2 Nanocapsules obtained from preformed polymers 261 3 Characterization 263 4 Drug Release 265 5 Applications 266 5.1 Oral route 266 5.2 Parenteral route 267 5.3 Ocular delivery 269 6 Conclusion 270 References 271 Dendrimers as Nanoparticulate Drug Carriers 277 Sbnke Svenson and Donald A. Tomalia 1 Introduction 277 2 Nanoscale Containers — Micelles, Dendritic Boxes, Dendrophanes, and Dendroclefts 279 2.1 Dendritic micelles 279 2.2 Dendritic box (Nano container) 280 2.3 Dendrophanes and dendroclefts 282 3 Dendrimers in Drug Delivery 282 3.1 Cisplatin 283 3.2 Silver salts 285 3.3 Adriamycin, methotrexate, and 5-fluorouracil 285 3.4 Etoposide, mefenamic acid, diclofenac, and venlafaxine . . . . 286 3.5 Ibuprofen, indomethacin, nifedipine, naproxen, paclitaxel, and methylprednisolone 287 3.6 Doxorubicin and camptothecin — self-immolative dendritic prodrugs 289 3.7 Photodynamic therapy (PDT) and boron neutron capture therapy (BNCT) 291 Contents 4 Nano-Scaffolds for Targeting Ligands 292 4.1 Folic acid 292 4.2 Carbohydrates 293 4.3 Antibodies and biotin-avidin binding 294 4.4 Penicillins 295 5 Dendrimers as Nano-Drugs 295 6 Routes of Application 296 7 Biocompatibility of Dendrimers 297 8 Conclusions 299 References 299 Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 307 Rainer H. Muller and Jens-Uive A. H. Junghanns 1 Introduction 307 2 Definitions 308 3 Physicochemical Properties of Drug Nanocrystals 309 3.1 Change of dissolution velocity 309 3.2 Saturation solubility 309 3.3 Does size really matter? 311 3.4 Effect of amorphous particle state 312 4 Production Methods 313 4.1 Precipitation methods 313 4.1.1 Hydrosols 313 4.1.2 Amorphous drug nanoparticles (NanoMorph®) . . . .313 4.2 Homogenization methods 314 4.2.1 Microfluidizer technology 314 4.2.2 Piston-gap homogenization in water (Dissocubes®) . . 314 4.2.3 Nanopure technology 315 4.3 Combination Technologies 315 4.3.1 Microprecipitation™ and High Shear Forces (NANOEDGE™) 315 4.3.2 Nanopure® XP technology 316 5 Application Routes and Final Formulations 317 5.1 Oral administration 317 5.2 Parenteral administration 319 5.3 Miscellaneous administration routes 321 6 Nanosuspensions as Intermediate Products 322 Contents xxi 7 Perspectives 324 References 324 Cells and Cell Ghosts as Drug Carriers 329 Jose M. Lanao and M. Luisa Sayalero 1 Introduction 329 2 Bacterial Ghosts 329 2.1 Application of bacterial ghosts as a delivery system 331 3 Erythrocyte Ghosts 333 3.1 Applications of erythrocyte ghosts as a delivery system . . . .335 4 Stem Cells 338 5 Polymorphonuclear Leucocytes 340 6 Apoptopic Cells 340 7 Tumor Cells 340 8 Dendritic Cells 341 9 Conclusions 341 References 342 Cochleates as Nanoparticular Drug Carriers 349 Leila Zarif 1 Introduction 349 2 Cochleates Nanoparticles in Oral Delivery 350 2.1 Cochleate structure 350 2.2 Cochleate preparation 350 2.2.1 Which phospholipid and which cation to use? 350 2.2.2 Which molecules can be entrapped in cochleates nanoparticles 352 2.2.3 Multiple ways of preparing cochleates 353 2.3 Cochleates as oral delivery system for antifungal agent, amphotericin B 355 2.3.1 In candidiasis animal model 355 2.3.2 In aspergillosis animal model 355 2.3.3 In cryptococcal meningitis animal model 357 2.3.4 Toxicity of amphotericin B cochleates 357 2.3.5 Pharmacokinetics of amphotericin B cochleates . . . . 357 2.4 Other potential applications for cochleates 359 2.4.1 Cochleate for the delivery of antibiotics 359 2.4.2 Delivery of clofazimine 360 xxii Contents 2.4.3 Delivery of tobramycin 360 2.4.4 Cochleate for the delivery of anti-inflammatory drugs 361 2.5 Other uses of cochleates 361 3 Conclusion 361 References 362 17. Aerosols as Drug Carriers 367 N. Renee Labiris, Andrew P. Bosco and Myrna B. Dolovich 1 Introduction 367 2 Pulmonary Drug Delivery Devices 369 2.1 Nebulizers 369 2.2 Metered-dose inhalers 371 2.3 Dry powder inhalers 373 3 Aerosol Particle Size 373 4 Targeting Drug Delivery in the Lung 376 5 Clearance of Particles from the Lung 378 5.1 Airway geometry and humidity 378 5.2 Lung clearance mechanisms 379 6 Nanoparticle Formulations for Inhalation 381 6.1 Diagnostic imaging 382 6.2 Vaccine delivery 383 6.3 Anti Tuberculosis therapy 385 6.4 Gene therapy 386 7 Conclusion 388 References 388 18. Magnetic Nanoparticles as Drug Carriers 397 Urs O. Hafeli and Mathieu Chastellain 1 Introduction 397 2 Definitions 398 2.1 Properties of magnetic materials 398 2.2 Nanoparticles 400 3 Magnetic Nanoparticles 401 3.1 Iron oxide based magnetic nanoparticles 401 3.2 Cobalt based magnetic nanoparticles 402 3.3 Iron based magnetic particles 402 3.4 Encapsulated magnetic nanoparticles 403 3.5 Biocompatibility issues of magnetic nanoparticles 403 Contents xxiii 4 Application of Magnetic Nanoparticles as Drug Carriers 404 4.1 Magnetic hyperthermia 405 4.2 Magnetic chemotherapy 406 4.3 Other magnetic treatment approaches 408 4.4 Magnetic gene transfer 409 5 Conclusions 410 References 411 19. DQAsomes as Mitochondria-Specific Drug and DNA Carriers 419 Volkmar Weissig 1 Introduction 419 2 The Self Assembly Behavior of Bis Quinolinium Derivatives 420 2.1 Monte Carlo computer simulations 420 2.2 Physico-chemical characterization 421 2.3 Structure activity relationship studies 422 3 DQAsomes as Mitochondrial Transfection Vector 424 4 DQAsomes as Carriers of Pro-apoptotic Drugs 429 5 Summary 432 References 432 20. Liposomal Drug Carriers in Cancer Therapy 437 Alberto A. Gabizon 1 Introduction 437 2 The Challenge of Cancer Therapy 439 3 The Rationale for the Use of Liposomal Drug Carriers in Cancer . . 442 4 Liposome Formulation and Pharmacokinetics — Stealth Liposomes 445 5 Preclinical Observations with Liposomal Drug Carriers in Tumor Models 448 6 Liposomal Anthracyclines in the Clinic 449 6.1 Doxil 450 6.2 Myocet 454 6.3 Daunoxome 454 7 Clinical Development of Other Liposome-entrapped Cytotoxic Agents 455 8 The Future of Liposomal Nanocarriers 456 References 457 xxiv Contents 21. Nanoparticulate Drug Delivery to the Reticuloendothelial System and to Associated Disorders 463 Mukul Kumar Basu and Sanchaita Lala 1 Introduction 463 2 Reticuloendothelial System and Associated Disorders 464 3 Uptake of Nanoparticles by the Reticuloendothelial System 464 3.1 Sites of uptake 464 3.2 Mechanism of uptake 465 3.3 Factors influencing uptake 468 3.4 Role of surface modifications on uptake 469 4 Active Targeting of Nanoparticles by Receptor Mediated Endocytosis 471 5 Application in Chemotherapy 473 6 Summary 475 References 477 22. Delivery of Nanoparticles to the Cardiovascular System 481 Ban-An Khazv 1 Introduction 481 2 Targeting the Myocardium with Immunoliposomes 481 3 Other Nanoparticle-Targeting of the Cardiovascular System 484 4 Novel Application of Nano-Immunoliposomes 485 5 CSIL as Targeted Gene or Drug Delivery 492 6 Conclusion 495 References 496 23. Nanocarriers for the Vascular Delivery of Drugs to the Lungs 499 Thomas Dziubla and Vladimir Muzykantov 1 Introduction 500 2 Biomedical Aspects of Drug Delivery to Pulmonary Vasculature . . 500 2.1 Routes for pulmonary drug delivery: Intratracheal vs vascular 501 2.2 Pulmonary vasculature as a target for drug delivery 501 3 Pulmonary Targeting of Nanocarriers 503 3.1 Effects of carrier size on circulation and tissue distribution . .503 Contents xxv 3.2 Passive targeting 505 3.2.1 Mechanical retention 505 3.2.2 Charge-mediated retention and non-viral gene delivery 506 3.2.3 Pulmonary enhanced permeation-retention (EPR) effect 507 3.3 Active targeting 507 4 Carrier Design 509 4.1 Biocompatibility 509 4.2 Material selection (by application) 510 4.2.1 Imaging 510 4.2.2 Gene delivery 510 4.2.3 Delivery of therapeutic enzymes 511 4.2.4 Small molecule drugs 512 4.3 Types of nanocarriers 512 4.4 Mechanisms of drug loading 512 4.5 Drug release mechanisms 515 4.6 Nanocarriers for active targeting 516 5 Conclusion: Safety Issues, Limitations and Perspectives 517 References 518 24. Nanoparticulate Carriers for Drug Delivery to the Brain 527 Jorg Kreuter 1 Introduction 527 2 Nanoparticles 528 3 Biodistribution 530 3.1 Influence of surfactants on the biodistribution of nanoparticles 530 3.2 Influence of PEGylation on the biodistribution of nanoparticles 532 4 Pharmacology 534 5 Brain Tumors 536 6 Toxicology 538 7 Mechanism of the Delivery of Drug Across the Blood-Brain Barrier with Nanoparticles 539 8 Summary 541 9 Conclusions 542 References 542 Contents Nanoparticles for Targeting Lymphatics 549 William Phillips 1 Introduction 549 1.1 The lymphatic vessels 550 1.2 Lymph nodes 551 2 Potential for Nanoparticles for Drug Delivery to Lymphatics . . . . 553 3 Importance of Lymph Nodes for Disease Spread and Potential Applications of Lymph Node Drug Delivery 554 3.1 Cancer 554 3.2 HIV 555 3.3 Filaria 555 3.4 Anthrax 556 3.5 Tuberculosis 556 3.6 Importance of lymph node antigen delivery for development of an immune response 557 4 Factors Influencing Nanoparticle Delivery to Lymph Nodes 559 4.1 Nanoparticle size 559 4.2 Nanoparticle surface 559 4.3 Effect of massage on lymphatic clearance of subcutaneously injected liposomes 560 4.4 Macrophage phagocytosis 561 4.5 Fate of nanoparticles in lymph nodes 561 5 Nanoparticle Diagnostic Imaging Agents for Determining Cancer Status of Lymph Nodes 561 5.1 Subcutaneous injection of iodinated nanoparticles for computed tomography imaging 561 5.2 Subcutaneous and intraorgan injection of magnetic resonance (MRI) contrast agents 563 5.3 Intravenous injection of magnetic nanoparticles for MRI imaging 563 5.4 Nanoparticle diagnostic agents for localizing the sentinel lymph node 565 5.5 Radiolabeled nanoparticles for sentinel lymph node identification 566 5.6 99mTc-Colloidal nanoparticles for sentinel node identification . 566 5.7 Optical 568 5.8 Ultrasound nanobubbles 569 6 Recently Introduced Medical Imaging Devices for Monitoring Lymph Node Delivery and Therapeutic Response 569 Contents xxvii 7 Nanoparticle Lymph Node Drug Delivery 571 7.1 Confusion in reporting lymph node delivery 571 7.2 Calculation of lymph node retention efficiency 573 8 Specific Types Nanoparticles for Lymph Node Targeting 573 8.1 PLGA nanoparticles 573 8.2 Micelles 574 8.3 Liposomes 574 9 Avidin Biotin-Liposome Lymph Node Targeting Method 577 10 Massage and the Avidin-Biotin Liposome Targeting Method 578 11 Nanoparticles for Lymph Node Anti-Infectious Agent Delivery . . . 580 12 Liposomes for Intraperitoneal Lymph Node Drug Delivery 581 12.1 Intraperitoneal liposome encapsulated drugs 582 12.2 Effect of liposome size on intraperitoneal clearance 583 12.3 Avidin/Biotin-liposome system for intraperitoneal and lymph node drug delivery 584 12.4 Mediastinal lymph node drug delivery with avidin-biotin system by intrapleural injection 585 12.5 Avidin biotin for diaphragm and mediastinal lymph node targeting 586 13 Nanoparticles for Cancer Therapy 587 13.1 Intralymphatic drug delivery to lymph nodes 587 13.2 Nanoparticles for treatment of metastatic lymph nodes of upper GI malignacies 589 13.3 Lessons from endolymphatic radioisotope therapy 591 14 Advantages of Nanoparticles for Lymphatic Radiotherapy 592 15 Intraoperative Radiotherapy for Positive Tumor Margins and for Treatment of Lymph Nodes 593 16 Potential of Using Radiolabeled Nanoparticles for Intra tumoral Radionuclide Therapy 593 17 Liposome Pharmacokinetics after Intra tumoral Administration . . .595 18 Rhenium-Labeled Liposomes for Tumor Therapy 595 19 Nanoparticles for Immune Modulation 597 20 Conclusions 598 References 598 26. Polymeric Nanoparticles for Delivery in the Gastro-Intestinal Tract 609 Mayank D. Bhavsar, Dinesh B. Shenoy and Mansoor M. Amiji 1 Oral Drug Delivery 609 Contents 2 Anatomical and Physiological Considerations of Gastro-intestinal Tract (GIT) for Delivery 610 3 Introduction to Polymeric Nanoparticles as Carriers 614 4 Preparation of Polymeric Nanoparticles 615 5 Design Consideration for Nanoparticle-based Delivery Systems . . 619 5.1 Polymer characteristics 619 5.2 Drug characteristics 620 5.3 Application characteristics 621 6 Nanoparticles in Experimental and Clinical Medicine 621 6.1 Drug delivery in the oral cavity 621 6.2 Gastric mucosa as a target for oral nanoparticle-mediated therapy 625 6.3 Nanoparticles for delivery of drugs and vaccines in the small intestine 626 6.4 Nanoparticles for colon-specific delivery 632 7 Integrating Polymeric Nanoparticles and Dosage Forms 634 8 Toxicology and Regulatory Aspects 636 8.1 Safety 637 8.2 Quality of material/characterization 638 8.3 Environmental considerations 638 9 Conclusion and Outlook 639 References 640 Nanoparticular Carriers for Ocular Drug Delivery 649 Alejandro Sanchez and Maria J. Alonso 1 Biopharmaceutical Barriers in Ocular Drug Delivery. Classification of Nanoparticulate Carriers for Ocular Drug Delivery 650 2 Nanoparticulate Polymer Compositions as Topical Ocular Drug Delivery Systems 651 2.1 First generation: Polymer nanoparticles and nanocapsules for topical ocular drug delivery 652 2.1.1 Acrylic polymers-based nanoparticles 654 2.1.2 Polyester-based nanoparticles and nanocapsules . . .655 2.1.3 Polysaccharide-based nanoparticles 657 2.2 Second nanoparticles generation: The coating approach . . . . 659 2.2.1 Polyacrylic coating 659 2.2.2 Polysaccharide coating 660 2.2.3 Polyethyleneglycol (PEG) coating 662 Contents xxix 2.3 Third nanoparticles generation: Towards functionalized nanocarriers 663 3 Nanoparticulate Polymer Compositions as Subconjuctival Drug Delivery Systems 665 4 Nanoparticulate Polymer Compositions as Intravitreal Drug Delivery Systems 665 5 Conclusions and Outlook 667 References 668 Nanoparticles and Microparticles as Vaccine Adjuvants 675 Janet R. Wendorf, Manmohan Singh and Derek T. O'Hagan 1 Introduction 675 2 Nanoparticle and Microparticle Preparation Methods 678 2.1 Nanoparticles and microparticles made from polyesters . . . . 678 2.2 Nanoparticles and microparticles made with chitosan 681 2.3 Other nanoparticles and microparticles 681 3 Adjuvant Effect of Nanoparticles and Microparticles 681 3.1 Nanoparticles and microparticles as mucosal adjuvants . . . . 682 3.2 Nanoparticles and microparticles as systemic adjuvants . . . . 686 4 Delivery of DNA Using Nanoparticles and Microparticles 688 5 Conclusions 690 References 691 Pharmaceutical Nanocarriers in Treatment and Imaging of Infection 697 Raymond M. Schijfelers, Gert Storm and Irma A. J. M. Bakker-Woudenberg 1 Introduction 697 2 Carriers that are Easily Recognized as Foreign Materials 698 3 Carriers that Avoid Recognition as Foreign Materials 701 4 Local Application of Carriers 705 5 Concluding Remarks 706 References 707 713 1 Introduction. Nanocarriers for Drug Delivery: Needs and Requirements Vladimir Torchilin Fast developing nanotechnology, among other areas, is expected to have a dramatic impact on medicine. The application of nanotechnology for treatment, diagnosis, monitoring, and control of biological systems has recently been determined by the NIH as nanomedicine. Among the approaches for exploiting nanotechnology developments in medicine, various nanoparticulates offer some unique advantages as pharmaceutical delivery systems and image enhancement agents.1,2 Several varieties of nanoparticles are available3: different polymeric and metal nanoparticles, liposomes, micelles, quantum dots, dendrimers, microcapsules, cells, cell ghosts, lipoproteins, and many different nanoassemblies. All of these nanoparticles can play a major role in diagnosis and therapy. This book is attempting to present the broad overview of different nanoparticulate drug delivery systems with all their advantages and limitations, as well as potential areas of their clinical applications. The paradigm of using nanoparticulate pharmaceutical carriers to enhance the in vivo efficiency of many drugs, anti-cancer drugs, first of all, well established itself over the past decade both in pharmaceutical research and clinical setting, and does not need any additional proofs. Numerous nanoparticle-based drug delivery and drug targeting systems are currently developed or under development.4,5 Their use aims to minimize drug degradation upon administration, prevent undesirable side effects, and increase drug bioavailability and the fraction of the drug accumulated in the pathological area. Pharmaceutical drug carriers, especially the 1 2 Torchilin ones for parenteral administration, are expected to be easy and reasonably cheap to prepare, biodegradable, have small particle size, possess high loading capacity, demonstrate prolonged circulation, and, ideally, specifically or non-specifically accumulate in required pathological sites in the body.6 High molecular weight (40 kDa or higher), long-circulating macromolecules, including proteins and peptides, conjugated with water-soluble polymers, are capable of spontaneous accumulations in various pathological sites such as solid tumors, infarcts, and inflammations via the enhanced permeability and retention effect (EPR).7'8 This effect is based on the fact that pathological (tumor, infarct) vasculature, unlike vasculature of healthy tissues, is "leaky", i.e. penetrable for macromolecules and nanoparticulates which allows for macromolecules to accumulate in the pathological tissue (such as interstitial tumor space). In the case of tumors, such accumulation is also facilitated by the fact that lymphatic system, responsible for the drainage of macromolecules from normal tissues, is virtually not working in case of many tumors as the result of the disease.8 It has been found that the effective pore size of most peripheral human tumors range from 200 nm to 600 nm in diameter, with a mean of about 400 nm. The EPR effect allows for passive targeting to tumors and other pathological sites based on the cut-off size of the leaky vasculature.9 Among particulate drug carriers, liposomes, micelles and polymeric nanoparticles are the most extensively studied and possess the most suitable characteristics for encapsulation of many drugs and diagnostic (imaging) agents. Many other systems meeting certain more specific requirements (and reviewed in this book) are also suggested and currently under development. Making these nanocarriers multifunctional and stimuli-responsive can dramatically enhance the efficiency of various drugs carried by these carriers. These functionalities are expected to provide: (a) prolonged circulation in the blood10'11 and the ability to accumulate in various pathological areas (such as solid tumors) via the EPR effect (protective polymeric coating with PEG is used for this purpose)12,13; (b) ability to specifically recognize and bind target tissues or cells via the surface-attached specific ligand (monoclonal antibodies as well as their Fab fragments and some other molecules are used for this purpose)14; (c) ability to respond local stimuli characteristic of the pathological site by, for example, releasing an entrapped drug or specifically acting on cellular membranes under the abnormal pH or temperature in disease sites (this property could be provided by surface-attached pH- or temperature-sensitive coatings); (d) ability to penetrate inside cells bypassing the lysosomal degradation for efficient targeting of intracellular drug targets (for this purpose, the surface of nanocarriers may be decorated by cell-penetrating peptides). Those are just the most evident examples. Some other specific properties can also be listed, such as an attachment of diagnostic moieties. Even the use of individual functionalities is already associated with highly Introduction. Nanocarriers for Drug Delivery: Needs and Requirements 3 positive clinical outcome — the success of Doxil®, doxorubicin in long-circulating PEG-coated liposome, represents a good example.15 In addition, there are numerous engineered constructs, assemblies, architectures, and particulate systems, whose unifying feature is the nanometer scale size range (from a few to 250 nm). Together with already listed systems, these include cyclodextrins, niosomes, emulsion particles, solid lipid particles, drug nanocrystals, metal and ceramic nanoparticles, protein cage architectures, viral-derived capsid nanoparticles, polyplexes, cochleates, and microbubbles.4,5,16-19 Therapeutic and diagnostic agents can be encapsulated, covalently attached, or adsorbed on to such nanocarriers. These approaches can easily overcome drug solubility issues, particularly with the view that large proportions of new drug candidates emerging from high-throughput drug screening initiatives are water insoluble. Yet, some carriers have a low capacity to incorporate active compounds (e.g. dendrimers, whose size is in the order of 5-10 nm). There are alternative nanoscale approaches for solubilization of water insoluble drugs too.20-23 One approach is to mill the substance and then stabilize smaller particles with a coating; this forms nanocrystals in size ranges suitable for oral delivery, as well as for intravenous injection.24,25 Pharmacokinetic profiles of injectable nanocrystals may vary from rapidly soluble to slowly dissolving in the blood. In general, the development of drug nanocarriers for poorly soluble pharmaceuticals represents a special task and still faces some unresolved issues. The therapeutic application of hydrophobic, poorly water-soluble agents is associated with some serious problems, since low water-solubility results in poor absorption and low bioavailability.26 In addition, drug aggregation upon intravenous administration of poorly soluble drugs might lead to such complications as embolism27 and local toxicity.28 On the other hand, the hydrophobicity and low solubility in water appear to be intrinsic properties of many drugs,29 since it helps a drug molecule to penetrate a cell membrane and reach important intracellular targets.30,31 To overcome the poor solubility of certain drugs, clinically acceptable organic solvents are used in their formulations,28 as well as liposomes32 and cyclodextrins.16 Another alternative is associated with the use of various micelle-forming surfactants in formulations of insoluble drugs. By virtue of their small size and by functionalizing their surface with synthetic polymers and appropriate ligands, nanoparticulate carriers can be targeted to specific cells and locations within the body after intravenous and subcutaneous routes of injection. Such approaches may enhance detection sensitivity in medical imaging, improve therapeutic effectiveness, and decrease side effects. Some of the carriers can be engineered in such a way that they can be activated by changes in the environmental pH, chemical stimuli, by the application of a rapidly oscillating magnetic field, or by application of an external heat source.19,33-35 Such modifications 4 Torchilin offer control over particle integrity, drug delivery rates, and the location of drug release, for example, within specific organelles. Some are being designed with the focus on multifunctionality; these carriers target cell receptors and delivers drugs and biological sensors simultaneously. Some include the incorporation of one or more nanosystems within other carriers, as in the micellar encapsulation of quantum dots; this delineates their inherent nonspecific adsorption and aggregation in biological environments.36 The use of nanoparticulate drug carriers seems to be especially important for developing efficient anticancer therapies. Although significant advances have occurred in our understanding of tumor origin, growth, metastasis, and many different types of pharmacological agents have been developed over the years to treat tumors, the problem of optimum delivery remains a formidable challenge. For any of the drug therapy strategies to be effective, the agent must be able to reach the tumor mass in sufficient concentration, traverse through the tumor microcirculation, diffuse into the interstitium, and remain at the site for the duration to induce tumoricidal effect. As was already mentioned, due to the porosity of the tumor vasculature and the lack of lymphatic drainage, blood-borne macromolecules and nanoparticles are preferentially distributed in the tumor via the EPR effect. However, nanoparticles can also be actively targeted to tumors by modifying their surface with certain cell-specific ligands for receptor-mediated uptake. The use of specific "vector" molecules can further enhance tumor targeting of nanocarries or make them the EPR-effect independent. The latter is especially important for the cases of tumors with immature vasculature, such as tumors in the early stages of their development, and for delocalized tumors. Vector molecules (those having affinity toward ligands characteristic for target tissues), capable of recognizing tumors were found among antibodies, peptides, lectins, saccharides, hormones, transferrin and some low molecular weight compounds (riboflavin, folate). From this list, antibodies and their fragments provide the most universal opportunity to target various for targeting and have the highest potential specificity. Vector molecules can be used for the targeting of nanoreservoir delivery systems as well. PEG-modified long-circulating doxorubicin-containing immunoliposomes targeted with anti-HER-2/neu monoclonal antibody fragments represent a recent example of increased efficiency of targeted delivery systems.37 In all studied HER2- overexpressing models, immunoliposomes showed potent anticancer activity superior to that of control non-targeted liposomes. In part, this superior activity was attributed to the ability of the immunoliposomes to deliver their load inside the target cells via the receptor-mediated endocytosis, which is obviously important if the drug's site of action sites locates inside the cell. An important problem is associated with the clearance of drug carriers from the circulation. Nanoparticular pharmaceutical carriers administered into the systemic Introduction. Nanocarriers for Drug Delivery: Needs and Requirements 5 circulation will be essentially removed within an hour of administration by the macrophages of the reticulo-endothelial system. To prolong the circulation of nanoparticles by evading the macrophages, their surface is modified with watersoluble polymers. Poly(ethylene glycol) (PEG) is very popular for surface modification of nanoparticulate drug delivery systems, since it has a long history of safe usage in biological and pharmaceutical products. Surface-bound PEG chains extend into the aqueous physiological environment and repel proteins, decrease antibody formation, and increase the circulation of the formulation in the plasma for extended period of time by the steric repulsion mechanism.38 With rapid advances in molecular biology and genetic engineering, there is an unprecedented opportunity for delivery of drugs and genes to intracellular targets.39 In the case of cancer, for instance, the effectiveness of many anticancer drugs is limited due to its inability to reach the target site in sufficient concentrations and to exert the pharmacological effect. Current gene delivery systems are classified as being either viral or non-viral in origin. Viruses are efficient in delivery of genes; however, they suffer from poor safety profile. Non-viral gene delivery systems, albeit not as efficient as viruses, have promise of safety and reproducibility in manufacturing. To enhance delivery of drugs to intracellular targets and gene transfection efficiency using non-viral delivery systems, it is necessary to identify ways of overcoming the cellular barriers, for example, by using various cell-penetrating proteins and peptides.40,41 Self-assembled nanosystems (nanoassemblies) for targeting subcellular organelles, such as the mitochondria, are also developed.42 It has become increasingly evident that mitochondrial dysfunction contributes to a variety of human disorders. Moreover, since the middle of 1990s, mitochondria, the "power houses" of the cell, have also become accepted as the cell's "arsenals", which reflects their increasingly acknowledged key role during apoptosis. Based on these recent developments in mitochondrial research, increased pharmacological and pharmaceutical efforts have led to the emergence of "Mitochondrial Medicine" as a whole new field of biomedical research. Nanoparticulate drug delivery systems are very important for the delivery of peptide and protein drugs and may represent a valid alternative to soluble polymeric carriers used earlier. The use of this type of carriers allows achieving much higher active moiety/carrier material ratio compared with "direct" molecular conjugates. They also provide better protection of protein and peptide drugs against enzymatic degradation and other destructive factors upon parenteral administration, because the carrier wall completely isolates drug molecules from the environment. All nanoparticulate carriers have the size, which excludes a possibility of renal filtration. Among particulate drug carriers, liposomes are the most extensively studied and poses the most suitable characteristics for protein (peptide) encapsulation. 6 Torch i I in Similar to macromolecules, protein and peptide drug-bearing liposomes are capable of accumulating in tumors of various origins via the EPR effect.6-8 In some cases, however, the liposome size is too large to provide an efficient accumulation via the EPR effect presumably due to relatively small tumor vasculature cut off size.43,44 In such cases, alternative delivery systems with smaller sizes, such as micelles (prepared, for example, from PEG-phospholipid conjugates) can be used. These particles lack the internal aqueous space and are smaller than liposomes. Protein or peptide pharmaceutical agent can be covalently attached to the surface of these particles or incorporated into them via chemically attached hydrophobic group ("anchor"). In conclusion, even a brief listing of some key problems of nanocarrier-mediated drug delivery shows how broad and intense this area is. In addition to this, nanoscale-based delivery strategies are beginning to make a significant impact on global pharmaceutical planning and marketing. The leading experts in the area of nanparticulate-mediated drug delivery attempted to address these and many other topics in this book. We strongly believe that every reader will find the book useful and stimulating. References 1. West JL and Halas NJ (2000) Applications of nanotechnology to biotechnology commentary. Curr Opin Biotechnol 11:215. 2. La Van DA, Lynn DM and Langer R (2002) Moving smaller in drug discovery and delivery. Nat Rev Drug Discov 1:77. 3. Sahoo SK and Labhasetwar V (2003) Nanotech approaches to drug delivery and imaging. Drug Discov Today 8:1112. 4. Miiller, RH (1991) Colloidal Carriers for Controlled Drug Delivery and Targeting. Wissenschaftliche Verlagsgesellschaft: Stuttgart, Germany and CRC Press: Boca Raton. 5. Cohen S and Bernstein H (eds.) (1996). Microparticulate Systems for the Delivery of Proteins and Vaccines. Marcel Dekker, New York. 6. Gref R, Minamitake Y, Peracchia MT, Trubetskoy V, Torchilin VP and Langer R (1994) Biodegradable long-circulating polymeric nanospheres. Science 263:1600. 7. Maeda H (2001) SMANCS and polymer-conjugated macromolecular drugs: Advantages in cancer chemotherapy. Adv Drug Deliv Rev 46:169. 8. Maeda H, Sawa T and Konno T (2001) Mechanism of tumor-targeted delivery of macromolecular drugs, including the EPR effect in solid tumor and clinical overview of the prototype polymeric drug SMANCS. / Control Rel 74:47. 9. Yuan F, Dellian M, Fukumura M, Leunig M, Berk BD, Torchilin VP and Jain RK (1995) Vascular permeability in a human tumor xenograft, Molecular size dependence and cutoff size. Cancer Res 55:3752. 10. Lasic DD and Martin F (eds.) (1995) Stealth Liposomes. CRC Press: Boca Raton. Introduction. Nanocarriers for Drug Delivery: Needs and Requirements 7 11. Torchilin VP and Trubetskoy VS (1995) Which polymers can make nanoparticulate drug carriers long-circulating? Adv Drug Del Rev 16:141. 12. Lukyanov AN, Hartner WC and Torchilin VP (2004) Increased accumulation of PEGPE micelles in the area of experimental myocardial infarction in rabbits. / Control Rel 8, 94:187. 13. Maeda H, Wu J, Sawa T, Matsumura Y and Hori K (2001) Tumor vascular permeability and the EPR effect in macromolecular therapeutics: A review. / Control Rel 65:271. 14. Torchilin VP (1998) Polymer-coated long-circulating microparticular pharmaceuticals. ] Microencapsulation 15:1. 15. O'Shaughnessy JA (2003) Pegylated liposomal doxorubicin in the treatment of breast cancer. Clin Breast Cancer 4,318. 16. Thompson D and Chaubal MV (2000) Cyclodextrins (CDS) — excipients by definition, drug delivery systems by function (Part I: Injectable applications). Drug Del Technol 2:34. 17. Zhang L and Eisenberg A (1995) Multiple morphologies of "crew-cut" aggregates of polystyrene-b-poly(acrylic acid) block copolymers. Science 268:1728. 18. Gref R, Domb A, Quellec P, Blunk T, Muller RH, Verbavatz JM and Langer R (1995) The controlled intravenous delivery of drugs using PEG-coated sterically stabilized nanospheres. Adv Drug Del Rev 16:215. 19. Cammas S, Suzuki K, Sone C, Sakurai Y, Kataoka K and Okano T (1997) Thermorespensive polymer nanoparticles with a core-shell micelle structure as site specific drug carriers. / Control Rel 48:157. 20. Kabanov AV, Batrakova EV and Alakhov VY (2002) Pluronic block copolymers as novel polymer therapeutics for drug and gene delivery. / Control Rel 82:189. 21. Kwon GS (2003) Polymeric micelles for delivery of poorly water-soluble compounds. Crit Rev Ther Drug Can Syst 20:357. 22. Jones M and Leroux J (1999) Polymeric micelles — a new generation of colloidal drug carriers. Eur J Pharm Biopharm 48:101. 23. Torchilin VP (2001) Structure and design of polymeric surfactant-based drug delivery systems. / Control Rel 73:137. 24. Muller RH and Keck CM (2004) Challenges and solutions for the delivery of biotech drugs — a review of drug nanocrystal technology and lipid nanoparticles. / Biotechnol 113:151. 25. Kraft WK, Steiger B, Beussink D, Quiring JN, Fitzgerald N, Greenberg HE and Waldman SA (2004) The pharmacokinetics of nebulized nanocrystal budesonide suspension in healthy volunteers. / Clin Pharmacol 44:67. 26. Lipinski CA, Lombardo F, Dominy BW and Feeney PJ (2001) Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Del Rev 46:3. 27. Fernandez AM, Van Derpoorten K, Dasnois L, Lebtahi K, Dubois V, Lobl TJ, Gangwar S, Oliyai C, Lewis ER, Shochat D and Trouet A (2001) N-Succinyl-(beta-alanyl-L-leucyl- L-alanyl-L-leucyl) doxorubicin: An extracellularly tumor-activated prodrug devoid of intravenous acute toxicity. / Med Chem 44:3750. 8 Torchilin 28. Yalkowsky SH (ed.) (1981) Techniques of Solubilization of Drugs. Marcel Dekker: New York and Basel. 29. Shabner BA and Collings JM (eds.) (1990) Cancer Chemotherapy: Principles and Practice. J. B. Lippincott Co: Philadelphia. 30. Yokogawa K, Nakashima E, Ishizaki J, Maeda H, Nagano T and Ichimura F (1990) Relationships in the structure-tissue distribution of basic drugs in the rabbit. Pharm Res 7:691. 31. Hageluken A, Grunbaum L, Nurnberg B, Harhammer R, Schunack W and Seifert R (1994) Lipophilic beta-adrenoceptor antagonists and local anesthetics are effective direct activators of G-proteins. Biochem Pharmacol 47:1789. 32. Lasic DD and Papahadjopoulos (eds.) (1998) Medical Applications of Liposomes. Elsevier: New York. 33. Le Garrec D, Taillefer J, VanLier JE, Lenaerts V and Leroux JC (2002) Optimizing pH-responsive polymeric micelles for drug delivery in a cancer photodynamic therapy model. / Drug Targ 10:429. 34. Meyer O, Papahadjopoulos D and Leroux JC (1998) Copolymers of N-isopropylacrylamide can trigger pH sensitivity to stable liposomes. FEBS Lett 41:61. 35. Chung JE, Yokoyama M, Yamato M, Aoyagi T, Sakurai Y and Okano T (1999) Thermoresponsive drug delivery from polymeric micelles constructed using block copolymers of poly(N-isopropylacrylamide) and poly(butylmethacrylate). / Control Rel 62:115. 36. Stroh M, Zimmer JP, Duda DG, Levchenko TS, Cohen KS, Brown EB, Scadden DT, Torchilin VP, Bawendi MG, Fukumura D and Jain RK (2005) Quantum dots spectrally distinguish multiple species within the tumor milieu in vivo. Nat Med 11:678. 37. Park JW, Kirpotin DB, Hong K, Shalaby R, Shao Y, Nielsen UB, Marks JD, Papahadjopoulos D and Benz CC (2001) Tumor targeting using anti-her2 immunoliposomes. / Control Rel 74:95. 38. Veronese FM and Harris JM (2002) Introduction and overview of peptide and protein pegylation. Adv Drug Del Rev 54:453. 39. Torchilin VP and Lukyanov AN (2003) Peptide and protein drug delivery to and into tumors: Challenges and solutions. Drug Discov Today 8:259. 40. Schwarze SR, Ho A, Vocero-Akbani A and Dowdy SF (1999) In vivo protein transduction: Delivery of a biologically active protein into the mouse. Science 285:1569. 41. Gupta B, Levchenko TS and Torchilin VP (2005) VP: Intracellular delivery of large molecules and small particles by cell-penetrating proteins and peptides. Adv Drug Del Rev 57:637. 42. Weissig V (2003) Mitochondrial-targeted drug and DNA delivery. Crit Rev Ther Drug CarrSyst 20:1. 43. Weissig V, Whiteman KR and Torchilin VP (1998) Accumulation of protein-loaded longcirculating micelles and liposomes in subcutaneous Lewis lung carcinoma in mice. Pharm Res 15:1552. 44. Hobbs SK, Monsky WL, Yuan F, Roberts WG, Griffith L, Torchilin VP and Jain RK (1998) Regulation of transport pathways in tumor vessels: Role of tumor type and microenvironment. Proc Natl Acad Sci USA 95:4607. 2 Nanoparticle Flow: Implications for Drug Delivery Alexander T. Florence 1. Introduction While the experimental study of nanoparticle flow in vivo proves to be difficult, a variety of theoretical and practical techniques are becoming available to allow some understanding of the phenomena involved. These processes include (a) convective flow induced by the flow of blood, lymph or interstitial fluid, (b) the influence of the interaction of nanoparticles with themselves or with biological components and the effect of this on their transport, and (c) the effect of fluid flow and hence shear forces on particle access to, interaction with and removal from receptors. Diffusion and movement of particle suspensions in complex media such as interstitial tissue must also be considered. Much of the theoretical work which is relevant to this exploration of nanoparticle flow has not been directed towards biological endpoints, but this body of knowledge, and the analogous literature on the dynamic behavior of bacteria, erythrocytes and platelets provides the basis of a more rigorous analysis of the factors involved in drug carrier nanoparticle flow. As discussed in this book, nanoparticles are of value in drug, vaccine and gene delivery because their small dimension compared with microparticles allows them to interact more effectively with cells, be safely injected, and amongst other characteristics, diffuse further into tissues, and into and through individual cells. The flow of nanoparticles in capillaries, lymphatics, tumor vessels, their extravasation1 and movement in the cytoplasm of cells are all aspects of the topic covered in 9 10 Florence this overview, albeit from a phenomenological viewpoint. It is clear that particle diameter is a key parameter in the characterization and behavior of nanoparticle suspensions. In several of the analyses here, it becomes clear that another advantage of nanoparticles may be the relative lack of effect of shear stress, once particles adhere to surfaces as a prelude to uptake; this is contrasted with targeted microspheres whose residence on receptors and surfaces is size dependent, the larger particles being more vulnerable to detachment by shear forces. This chapter considers questions relating to the flow of nanoparticles in vivo, but which has often been simulated in vitro by chemical engineers and physicists interested in particle behavior in flow conditions. Spherical particles are the norm, but not all nanosystems are spherical. The influence of asymmetry on the transport of nanoparticles in vivo is largely unknown, although the rheological characteristics of asymmetric particle suspensions have been understood for a long time. Flow behavior of nanoparticles in complex networks of narrow capillaries has relevance in the design and operation of microfluidic devices, as well as in drug delivery and targeting, and in toxicology2; the extent to which it is relevant for delivery and targeting is explored here. Figure 1 illustrates diagrammatically some of the areas of interest. In physical terms, the following situations could be considered: (i) particle flow in rapidly flowing blood, including segregation and deposition of particles, and the behavior of particles at bifurcations in the capillary supply; (ii) the effect of shear on adhesion of particles of different size and shape; (iii) particle flow in more static conditions, for example, in the tumor interstitium or the lymphatics; (iv) flow of particles in tissues, including the flow of particles into narrow pores and (v) flow or diffusion within the anisotropic interior of cells. Added complications arise where bioadhesive or ligand-decorated particles are involved. In the latter case, binding and flow are linked. The enquiry can be divided into a discussion of the flow and movement of nanoparticles as a function of their size and route of administration, the influence of convection and blood or lymph flow, the influence of flow dynamics on the interaction of nanoparticles with target tissues and receptors, and the movement of nanoparticles once they have been absorbed and are making their way through individual cells and tissues towards a target. The routes of administration where flow is important potentially include all, even the oral route where particle flow and dynamics in the gut lumen and in the vicinity of both villi and microvilli is important. Particles below a critical size are taken up by the M-cells of Peyer's patches and by normal enterocytes, albeit in small quantities and find themselves in the lymph vessels, lymph nodes, blood, liver and spleen.3'4 If the flow of nanoparticles away from their site of absorption is restricted due to the flow in lymph or blood being slow, this will reduce Nanoparticle Flow: Implications for Drug Delivery 11 * II Fig. 1. A diagrammatic representation (not to scale) of some of the areas where flow and transport of nanoparticles is key. I: flow in the GI tract after oral administration; II: access to and adhesion to M-cells of Peyer 's patches or to enterocy tes; III: passage into the mesenteric lymph; IV: flow in the lymph vessels and entrapment in the lymph nodes (not shown); V: transport between lymph and blood. A: blood flow; B: adhesion to capillary walls; C: extravasation and flow in tissue; D: flow and deposition at vessel bifurcations; and E: movement into tumours. Each route (the subcutaneous route is also indicated) will involve a complex sequence of nanoparticle pathways, most involving lymph, blood and intestinal fluid. bioavailability and distribution. Rapid flow provides superior sink conditions and hence the use of everted gut sacs and cell monolayers in vitro can give unrealistic results for nanoparticle transit. The influence of flow dynamics on extravasation and perhaps on the enhanced permeability and retention (EPR) effect for nanocarriers has perhaps not been fully addressed. Both involve consideration of particle size, the diffusion and flow of nanoparticles through narrow channels, as well as navigation of tortuous environments. The availability of "extreme" nanoparticles in the form of dendrimers5 and quantum dots6 makes this topic a vital one in understanding the fate, toxicity7 or accumulation of what is metrically a wide range of systems. 2. Background Our own interest in this field has resulted in part from studies on the size dependency of nanoparticle uptake after oral administration, where mesenteric lymphatic transport of 500 nm nanoparticles post absorption is determined by the flow of particles in a single file in the smallest mesenteric lymph vessels (Fig. 2). In addition, 12 Florence Fig. 2. 500 nm polystyrene latex particles flowing in the mesenteric lymph vessels mostly in single-file mode, from Jani et al. studies on the flow of liposomes and niosomes, and the extrusion of flexible vesicles from glass capillaries under pressure, converting polyhedral vesicles into multilayer tubules, led us to consider the influence of stress forces on carrier integrity. Clearly, the elasticity of vesicles is important in their negotiation of capillaries when their diameter exceeds that of the capillary. The flow of particles in fabricated capillaries which have a radius close to the particle radius is a challenge that has been tackled theoretically.8'1* We have suggested that multi-bilayer tubules (Fig. 3) can act as models for such flow experiments.11 Rheological examination of nanoparticle-blood mixtures and nanoparticle suspensions of mixed radius has also illustrated the potential complexity of particle Fig. 3. A flexible non-ionic surfactant based multi-bilayer tube, around 1 /xm in diameter, extruded from a suspension of polyhedral niosomes, which might be adapted for use as a model for the study of capillary nanoparticle flow.11 Nanoparticle Flow: Implications for Drug Delivery 13 movement in blood (unpublished data). In addition, erythrocyte blockage at bifurcations, or narrowing of vessels can lead to slowing down of blood flow and a change in rheology as the haematocrit increases. More recently, investigation of the transport of nanoparticles across cell monolayers12 and intracellular transport of dendrimers13 has assisted in defining some of the issues involved in targeting within cells. 3. Studies on Nanoparticle Flow The work of Fokin and colleagues14 on the transport of viral-sized colloids, following intravenous or intra-lymphatic injection, is relevant to drug delivery even if their objectives were different. 100-200 nm diameter sulphur colloid particles reach the lymph after IV injection in around 25 minutes; after intra-lymphatic injection particles appear in the venous blood only after 4 seconds. Following subcutaneous injection, similar particles14-16 reach the lymph after 2-9 min, although 95% of the particles remain at the injection site for at least 45 minutes. Here, the nanoparticles are being used as indicators of blood and lymph flow. What is also relevant to drug delivery is the influence of fluid flow on the movement and fate of nanoparticles. Ilium et al.17 observed uptake rates of 1.27 jxm and 15.8 /zm polystyrene particles in the lung and liver after IV injection. The sequestration in the lung was size dependent, but possibly affected because the smaller particles were taken up by the Kupffer cells of the liver, leaving the larger particles free to be taken up by the lung tissue. The rapidity of this suggests that, in effect, flow of the microparticles is solely determined by blood flow. 4. Convection and Diffusion Blood flow drives the convective flow of suspended particles. Diffusional transport occurs in static conditions or conditions of low fluid velocity. In a tube of flowing liquid, convective dynamics propel the particles in the direction of flow, but at the walls of the tube, there is the possibility of particle diffusion resulting in deposition. Blood velocity (mm s_1) in arterioles and venules is a function of vessel diameter, as shown in Fig. 4. In venules, the maximum velocity according to Jain18 is approximately 12 mm s_1, while in arterioles, it can reach about 30 mm s^1. Fluid velocity in tubes is not constant throughout the diameter of the tube as Fig. 5 indicates, a feature that is important when the interaction of nanoparticles with epithelial cells or capillary walls is considered. The radial variation of shear is a factor that must be considered in polydisperse nanoparticulate systems and where nanoparticles adhere to erythrocytes, causing two distinct size distributions. If nanoparticles adhere to erythrocytes20 or other 14 Florence Arterioles o 0 o o °° s o I" I 0 25 Vmax(mm/s) o s o © r-35 Venules - 30 -25 " 20 o 5 g »|5i d^^OD0 a ) 25 50 Vessel Diameter (nm) Fig. 4. Maximum blood velocity (mms 1) in arterioles and venules as a function of vessel diameter, redrawn from R. K. Jain.18 Fig. 5. Diagram showing the velocity pattern in a tube of flowing liquid. Particles of different size separate according to their diameter. The large particles, being unable to approach close to the capillary wall, experience the faster fluid streamlines toward the centre; hence, they move more rapidly, as described by Silebi and DosRamos.18,19 This is the basis of the field flow fractionation. blood elements, the translocation of the particles is controlled by the particular element to which it adheres. The rheology of suspensions of mixed particles is complex: viscosity reduces first with an increase in the fraction of larger particles in a suspension, and as the volume fraction increases, so does the viscosity.21 Ding et al.22 formulated a theoretical model examining particle migration in nanoparticle suspensions flowing through a pipe. "The model considers particle migration due to spatial gradients in viscosity and shear rate as well as Brownian motion. Particle migration due to these effects can result in significant non-uniformity in particle concentration over the cross section of the pipe" in particular for larger particles. Three mechanisms were proposed by the authors for migration in such non-uniform shear flow: (i) shear induced migration where Nanoparticle Flow: Implications for Drug Delivery 15 particles move from regions of higher shear rate to regions of lower shear rate; (ii) viscosity gradient induced migration — particles move from regions of higher viscosity to regions of lower viscosity and (iii) self-diffusion due to Brownian motion. Diffusion inside microtubules has been studied to understand taxol binding to tubulin structures.23 The dimension of the tubulin lumen is of the order of 17nm, approaching macromolecular dimensions, leading to friction between the inner walls and the moving macromolecules. This "hindrance" will also be an issue in the movement of nanoparticles in the smallest capillaries. With dendrimers whose diameters may be as small as 6 nm, the application of hindered theory to their movement could be relevant. No vessels are of this small radius, but the key parameter is the ratio of particle diameter to capillary diameter. The approach may well be important in cellular networks. It is not only capillary vessels that are the conduits of particle movement, but after extravasation, there is passage through cellular networks. The process could be considered to be akin to diffusion in porous networks. Binding of the moving particle (or macromolecule) to the luminal surface of the vessel will also hinder free flow or movement, a positive event in the case of specific ligand targeting of "decorated" systems. Polydisperse nanosystems can segregate during flow or migrate differentially leading to concentration differences.22 Particle-image velocimetry (PIV)24 has been used to track the flow characteristics of microparticles. The effects of flow on adhesion of monocytes to endothelial cells25 is relevant to the influence of flow and shear in particle interactions and uptake. The significance of flow can be demonstrated by the use of pharmacological agents which change normal vessel patency, so that by the concomitant use of noradrenalin26 or angiotensin26,27 which constrict only normal vessels, the ratio of tumor to normal tissue blood flow can be optimized. 5. Bifurcations Many theoretical studies of nanoparticle flow deal with linear tubes, whereas in vivo movement occurs through complex vessel architectures with bifurcations28,29 (Fig. 6). Behavior at bifurcations in a vascular or capillary system is dependent not only on particle diameter, but also on the rigidity or flexibility of the particle concerned. Colloid transport in a bifurcating structure has been the subject of one recent paper.30 It is a process which depends on the orientation of the bifurcations, especially with particles whose density is greater than that of the medium, as well as on the different flow rates in the individual branches which are likely to be of different sizes. If nanoparticles are trapped or associate at bifurcations or indeed other obstacles in capillaries, then it is likely that they might associate more permanently, thereby 16 Florence A •& - Q„.30IAnin HB M ,.., 2.»..«s.» M ae'0.3 Fig. 6. Three-dimensional distributions of nanoparticles in a bifurcation airway model of Zhang et ah, Aerosol Sci., 2005, 36, 211-233. DEF is the deposition enhancement factor, the representations shown here being for a steady inhalation. While these data are for air-flow, not dissimilar patterns of deposition might be estimated to occur in liquid flows. Deposition in these models occurs primarily by Brownian diffusion; deposition efficiencies increase with decreasing nanoparticle size and lower inlet Reynolds numbers. changing their intrinsic rheological behavior. Flexible particles do not of course suffer the same constraints in movement and progress, but their flexibility can lead to slow negotiation of movement around obstacles (Fig. 7). 6. Interaction with Blood Constituents and Endogenous Molecules Nanoparticles may interact with blood constituents31: the adsorption of albumin, IgG and fibrinogen from blood onto hydrophobic particles is well known, but the Nanoparticle Flow: Implications for Drug Delivery 17 Fig. 7. Two captured pictures from a video of a large vesicle moving in a flowing stream of smaller vesicles. The stills show a flexible vesicle approaching an obstacle, and rolling around the obstacle while adhering to it, a process encouraged by its elasticity. effect of nanoparticles on blood has been less well studied. Kim's31 data indicate that the interaction of nanoparticles with erythrocytes changes the dynamics of flow of both erythrocytes and particles. Chambers and Mitragotri20 found that nanoparticles as large as 450 nm adhered to erythrocytes, and thus remained in the circulation for several weeks. The percentage of latex nanospheres in the circulation over a period of 6 hrs was highly dependent on particle size, retention times decreasing with increasing diameter from 220 nm to 1100 nm. These data are difficult to interpret on the basis of flow, as the erythrocytes with attached nanoparticles are eliminated somewhat faster than the native erythrocytes. Gorodetsky and colleagues32 explored interactions of carboplatin (CPt) nanoparticles (formed by CPt interaction with fibrinogen) with the fibrin mesh caused by the induction of clot formation. 18 Florence 7. Nanoparticles with Surface Ligands There appear to be no rheological studies comparing surface protein decorated nanoparticles with the unadorned forms. Certainly, it is possible that aggregation may be caused by the change in surface properties and that this will in turn change flow patterns and perhaps masking of ligands33 as posited in Fig. 8. Nanoparticles are of course sensitive to the medium in which they are placed34 even in vitro when cell media can cause significant increases in diameter because of particle flocculation. We have suggested that the interaction with surface receptors of nanoparticles decorated with ligands is more complex than intimated in discussions of targeting generally.33 Figure 8 represents some of the factors: the aggregation of particles, the masking of ligands by this process, the detachment of ligands and the shearinduced removal of attached particles as discussed above. The instability of plant lectins, frequently used as surface proteins on nanosystems, is discussed by Gabor et al.35 The processes illustrated in Fig. 8 might explain some of the lack of complete success of targeted drug delivery. 8. Deposition on Surfaces and Attachment to Receptors in Flow Conditions Nanoparticles in vivo flow in blood, lymph or tissue fluid at greater or lesser velocities, as discussed above. Deposition of particles which might occur in a static situation is itself a complex process, and will depend on the rugosity of the Aggregation and loss of ligand accessibility Repulsion Blocking by cleaved ligands n B n Fig. 8. Diagram illustrating variations from the ideal of a single ligand-decorated nanoparticle interacting with receptors spaced at an appropriate distance from the particles. The diagram shows the loss of ligand accessibility which would follow from the aggregation of the particles before interaction with the desired surface, repulsion between a particle attached to the receptor surface, and an approaching particle and blockage of the receptors due to interaction of cleaved ligands with the receptors. Nanoparticle Flow: Implications for Drug Delivery 19 receiving surface.36 Particle deposition from flowing suspensions has been the subject of research37 which has considered not only diffusion, convection, geometrical interception and migration under gravity, but also the influence of tangential interactions. Patil et al.39 examined the rate of attachment of 5, 10, 15 and 20 /zm particles with a reconstituted P-selectin glycoprotein ligand-1 construct 19.ek.Fc. The rate of attachment was not affected by particle diameter. However, the shear stress required to set the adherent particles in motion (Sc) decreased with increasing particle diameter, and the rolling velocity of the 19.ek.Fc microspheres increased with increasing diameter. From their data, if we extrapolate the critical shear (a plot of 1 / S c is linear with diameter over the range 5-20 jxm), it suggests that particles below one micron in diameter will not be removed by shear forces. Usually we consider the flow of many particles in collective diffusion. The diffusion coefficient of a single particle and the collective diffusion coefficient coincides at infinite dilution, but can differ at higher concentrations.40 Cell adhesion mediated by not one but two receptors has been considered by Bhatia et al.41; the analysis would also apply to decorated nanoparticles. In their study, the two receptors were selectin and integrin ICAM; "the state diagram" evolved shows the area of firm adhesion as opposed to rolling adhesion for leukocytes as a function of receptor densities and association rate constants. The fate of transport initial adhesion attachment uptake 104- Distance (nm) Adhesion time short range interactions ' or specific ligand e receptor interactions Fig. 9. Processes occurring in the deposition of nanoparticles in flow conditions as a function of the range of interaction forces (nm) and adhesion times. At the start, mass transport to the surface occurs, initial adhesion following through electrostatic attraction and van der Waals' forces. Hydrophobic interactions can play their part as well as specific receptorligand interactions which are short-range interactions. Drawn after Vacheethasanee and Marchant.38 20 Florence nanoparticles in flowing blood, their adhesion, extravasation and permeation into tumors, thus depends on a complex of factors such as diameter, surface ligand density and orientation, shape, capillary diameter and rugosity, bifurcations, viscosity and flow gradients. 9. Does Shape Matter? Nanosystems can be prepared in a variety of shapes. Nanocrystals42 are often irregular; there are asymmetric carbon nanotubes, and surfactant and lipid vesicles can be produced as discs, polyhedral structures,40,43 toroids and tubes.21,44 The vesicle constructs often have dimensions larger than 500 nm; it must be assumed that vesicles in the nanometer size range will be less affected. In these systems, shape is less important than membrane properties in controlling the release of encapsulated drug, but the flow properties of vesicular suspensions are clearly determined by shape and elasticity As most particulate delivery vectors have been spherical, little attention has been paid to the influence of shape on fate; yet it is known that the shape of environmental particles and fibres, for example, influences their fate and toxicity.45 As discussed above, there are two different but related effects of particle flow: the effect of particle shape and size and characteristics on flow, as well as the effect of flow on flexible particles, as discussed by Bruinsma.46 With elastic vesicles, we have argued44 that shape matters because it affects flow and potential fate in vivo through extravasation for instance; elasticity also allows vesicles to be transported in vessels which would be blocked by solid particles. The elasticity and visco-elasticity of such systems may be important in differentiating them from solid nanoparticles. Much of the debate on whether the shape of vesicles matters, is dependent on the knowledge of the nature of the capillary blood supply and the forces exerted on, and the damage done to, vesicles as they move in capillaries.44 In studies conducted in our laboratories with doxorubicin loaded niosomes, 60% of the drug remained in the vesicles 8 hrs after intravenous administration.47 The extent to which the drug loss was due to diffusion or to damage is not known, but vesicles subjected to deliberate stress can lose considerable amounts of their payload, simply by extrusion of the vesicles through capillaries of reducing diameter.48 Reduction in diameter of systems below 1 micron will clearly reduce such stresses and allow flexible systems to retain their loads intact. Vasanthi et al.49 treated the anisotropic diffusion of oblate spheroids, explaining that because non-spherical molecules rotate as they translate, their motion differs significantly from that of a sphere. For rods, theory predicts that the diffusion coefficient in the direction parallel to the major axis of the rod (Dn) is twice that in the perpendicular direction (Di.). Nanoparticle Flow: Implications for Drug Delivery 21 B IJ. • I T' ' • • • -l I 3*M x/2 *!* 0 Platelet angle a Fig. 10. The non-dimensional bond force as a function of the angle of an ellipsoidal platelet passing through zero when the platelet is 90° to the surface. From Mody et a/.50 There are few studies which have considered the motion of ellipsoidal particles near a plane wall, although this is relevant to platelet flow and adhesion to the walls of vessels. Mody and colleagues50 have addressed the issue, observing the effects of shear stress on platelet adhesion. Platelets, unlike leukocytes, do not roll but display a flipping motion in the direction of flow, due to their flattened ellipsoidal structure. The bond force between the ellipse and the surface is dependent on the platelet angle as defined in Fig. 10. Flexible systems such as vesicles have been widely studied, while being forced under pressure in capillaries smaller than the vesicle diameter. The elasticity of the membranes can be estimated from the extent of deformation. Vesicle flow in linearly forced motion has been followed. Flexible vesicles adjust their shape to equilibrate the applied force51; locally in some cases, two-dimensional flow of lipids in the vesicle membrane occurs,52 clearly influencing the position of the embedded surface ligands. There are many nanoparticulates which are produced in non-spherical forms, hence the transport properties of asymmetric particles is important.53 10. Speculations on Flow and the EPR Effect Erythrocyte velocity in normal vessels depends on vessel diameter (see Fig. 4 above), but there is no such dependence in tumors (Fig. 11), even though flow may be an order of magnitude slower. According to Jain,18,52 "to reach cancer cells in a tumor, a blood-borne therapeutic molecule, particle or cell must make its way 22 Florence 0.8 0.7 •t 0.5 0.4 > i 0.2 o.i H 0.0 MCalV J U»7 i r*—I—1 r——i 1 I i ""—> r 1 1 T" 0 10 20 30 40 SO 60 70 0 10 20 30 40 50 60 70 Tumor Vessel Diameter Qua) Fig. 11. Diagram from Jain18 showing the lack of a clear relationship between erythrocyte velocity and tumor vessel diameter in two tumor types, MCalV and U87. The low and variable velocities compared to those shown in Fig. 4 are evident. into blood vessels of the tumor and cross the vessel wall into the interstitium and finally migrate through the interstitium". While blood flow is reduced in tumor vessels, nonetheless cancer cells have been reported to compress tumor vessels and this will have consequences on fluid flow.54 This is highly relevant to the enhanced permeation and retention effect (EPR) which allows entry of macromolecules into tumors from spaces in the ill-formed tumor vasculature.55 Access of nanoparticles to tumors is equally important and must be critically size-dependent. In convective flow, stable colloidal particles may be captured by the process of hydrodynamic bridging,52,56 events which may be relevant to the first process in the enhanced permeation and retention (EPR) effect. At high velocities but in the low Re regime, hydrodynamic forces acting on the particles at an entrance to a pore (or a defect in a tumor vessel) may overcome colloidal repulsive forces and result in flocculation of the particles and the plugging of the pore. The effects of velocity, particle concentration, and the ratio of pore size to particle size (the aspect ratio) on retention by hydrodynamic bridging have been studied. The effect of velocity on retention by bridging is opposite to that of retention by deposition. There is a critical flow velocity necessary for particle bridging to occur, a function of the net colloidal interparticle and particle—porous medium repulsion that must be overcome by the hydrodynamic forces for bridging to occur. Figure 12 demonstrates the effect for an aspect ratio of 3.7 (220 nm particles) 11. Intra-tumoral Injection Direct injection of delivery systems into tumors has both been a mode of experimental and clinical drug delivery. Solutions allow the drugs to diffuse or leach out Nanoparticle Flow: Implications for Drug Delivery 23 f l v^ o- • f • * Fig. 12. Particle behavior prior to entry to a pore of radius, rp: (a) a discrete nanoparticle, (b) aggregate, (c) individual particles converging on the pore opening demonstrating hydrodynamic bridging, as discussed by Ramachandran.56 We speculate that events such as bridging might occur during entry of nanoparticles into tumors through fenestrations in the tumor capillary blood supply, aspects of the enhanced permeation and retention effect. of the tumor, especially through the needle track, whereas suspensions might allow some greater residence time. Viral vectors have been administered by intra-tumoral injection.57 To decrease the extent of viral dissemination into the systemic circulation, a viscous alginate solution was used as the viral vehicle. However, transgene expression was not increased perhaps because, as the authors speculate, the diffusion of the virus is reduced by the viscous medium once in situ. The transport of particles of viral dimensions requires, according to Higuchi et al.,16 convective rather than diffusional transport. "The early transport of colloids into the vascular and lymphatic vessels relies largely on an extracellular pathway which depends on convective transport (i.e. solvent drag)". "Thus the particle uptake in the period immediately after injection is relatively insensitive to particle size; it is expected that viruses will be carried in the tissue towards lymphatics and microvessels with great efficacy leading to enhanced escape compared with the relatively low levels" for 1 and 0.4 /xm particles.16 The question of how resistance to convective transport in the interstitial space (the interstitial fluid plus the extracellular matrix) has been considered at least for molecules.58 Clearly, the spacing between the cells or between fibres will be a significant factor in determining the size cut-off for transport. 12. Conclusions This phenomenological survey of possible factors affecting the flow and hence the mass transport of nanoparticles has explored a range of scenarios. It is by no means a comprehensive survey, but there is sufficient in the literature to stimulate further analyses to provide a better overall prediction of the influence of particle characteristics, particularly, diameter and surface nature, shape and flexibility on delivery and targeting to remote sites in the body. Conf ocal microscopy and other techniques 24 Florence will allow experimental study of nanoparticles so that their movement and fate can be studied in a variety of tissues. Atomic force microscopy allows measurement of forces of interaction of particles with cells and receptors to aid a more quantitative approach. However, it is wrong to underestimate the challenges ahead if nanoparticulate carriers are to be designed to overcome the various biological barriers and survive transit in the conduits of capillary blood or lymph, extravasation and tissue, and subsequently intracellular transport.59 One cannot help but conclude that as many properties including flow are dictated by particle diameter, one of the most important strategies is to ensure the maintenance of particle stability in vivo. References 1. El-Sayed M, Kiani MF, Naimark MD, Hikal AH and Ghandehari H (2001) Extravasation of poly(amidoannine) (PAMAM) dendrimers across microvascular network endothelium. Pharm Res 18:23-28. 2. Health and Safety Executive, Health Effects of particles produced for nanotechnologies. Sudbury, UK, pp. 1-37. 3. Hussain N, Jaitley V and Florence AT (2001) Recent advances in the understanding of uptake of microparticulates across the gastrointestinal lymphatics. Adv Drug Del Rev 50:107-142. 4. Florence AT (1997) The oral absorption of micro- and nanoparticulates: Neither exceptional nor unusual. Pharm Res 14:259-266. 5. Florence AT and Hussain N (2001) Transcytosis of nanoparticle and dendrimer delivery systems: Evolving vistas. Adv Drug Del Rev 50 (Suppl 1):S69-S89. 6. Fortina P, Kricka LJ, Surrey S and Grodzinski P (2005) Nanobiotechnology: The promise and reality of new approaches to molecular recognition. Trends Biotechnol 23: 168-173. 7. Warheit DB, Laurence BR, Reed KL, Roach DH, Reynolds GA and Webb TR (2004) Comparative pulmonary toxicity assessment of single-wall carbon nanotubes in rats. Toxicol Sci 77:117-125. 8. Sugihara-Seki M and Skalak R (1997) Asymmetric flows of spherical particles in a cylindrical tube. Biorheology 34:155-159. 9. Wang H and Skalak R (1969) Viscous flow in a cylindrical tube containing a line of spherical particles. / Fluid Mech 38:75-96. 10. Jani P, Halbert GW, Langridge J and Florence AT (1989) The uptake and translocation of latex nanospheres and microspheres after oral-administration to rats.}Pharm Pharmacol 41:809-812. 11. Nasseri B and Florence AT (2003) Microtubules formed by capillary extrusion and fusion of surfactant vesicles. Int} Pharm 266:91-98. 12. Rowland RES, Taylor PW and Florence AT (2005) / Drug Del Sci Tech 13. Ruenraroengsak P, Hartell N and Florence AT (2005) unpublished. 14. Fokin AA, Robicsek F and Masters TN (2000) Transport of viral-size particulate matter after intravenous versus intralymphatic entry. Microcirculation 7:357-365. Nanoparticle Flow: Implications for Drug Delivery 25 15. Fokin AA, Robicsek F, Masters TN, Schmid-Schonbein GW and Jenkins SH (2000) Propagation of viral-size particles in lymph and blood after subcutaneous inoculation. Microcirculation 7:193-200. 16. Higuchi M, Fokin A, Masters TN, Robicsek F and Schmid-Schonbein GW (1999) Transport of colloidal particles in lymphatics and vasculature after subcutaneous injection. JAppl Physiol 86:1381-1387. 17. Ilium L, Davis SS, Wilson CG, Thomas NW, Frier M and Hardy JG (1982) Blood clearance and organ deposition of intravenously administered colloidal particles. The effects of particle size, nature and shape. Int ] Pharm 12:135-146. 18. Jain RK (2001) Delivery of molecular medicine to solid tumors: Lessons from in vivo imaging of gene expression and function. / Control Rel 74:7-25. 19. Silebi CA and DosRamos JG (1989) Separation of submicrometer particles by capillary hydrodynamic fractionation (CHDF). / Coll Interf Sci 130:14-24. 20. Chambers E and Mitragotri S (2004) Prolonged circulation of large polymeric nanoparticles by non-covalent adsorption on erythrocytes. / Control Rel 100:111-119. 21. Nunez ADR, Pinto R and Paredes VME (2002) Viscosity minimum in bimodal concentrated suspensions under shear. Eur Phys } E 9:327-334. 22. Ding Y and Wen D (2005) Particle migration in a flow of nanoparticle suspensions. Powder Technol 149:84-92. 23. Odde D (1998) Diffusion inside microtubules. Eur Biophys J 27:514-520. 24. Sinton D (2004) Microscale flow visualization. Microfluid Nanofluid 1:2-21. 25. Chiu J-J, Chen C-N, Lee P-L, Yang CT, Chuang HS and Chien SUS (2003) Analysis of the effect of disturbed flow in monocytic adhesion to endothelial cells. / Biomech 26:1883-1895. 26. Shankar A, Loizidou M, Burnstock G and Taylor I (1999) Noradrenaline improves the tumour to normal blood flow ratio and drug delivery in a model of liver metastases. Br } Surgery 86:453^57. 27. Goldberg JA, Murray T, Kerr DJ, Willmott N, Bessent RG, McKillop JH and McCardle CS (1991) The use of angiotensin II as a potential method of targeting cytotoxic microspheres in patients with intrahepatic tumours. Br J Cancer 63:308-310. 28. Zhang Z, Kleinstreuer C, Donohue JF and Kim CS (2005) Comparison of micro- and nano-size particle depositions in a human upper airway model. Aerosol Sci 36:211-233. 29. Shi HKC, Zhang Z and Kim CS (2004) Nanoparticle transport and deposition in bifurcating tubes with different inlet conditions. Phys Fluids 16:2199-2213. 30. James SC and Chrysikopoulos CV (2004) Dense colloid transport in a bifurcating fracture. / Coll Interf Sci 270:250-254. 31. Kim D, El-Shall H, Dennis D and Morey T (2005) Interaction of PLGA nanoparticles with human blood constituents. Coll SurfB 40:83-91. 32. Gorodetsky R, Peylan-Ramu N, Reshef A, Gaberman E, Levdansky L and Marx G (2005) Interactions of carboplatin with fibrin(ogen), implications for local slow release chemotherapy. / Control Rel 102:235-245. 33. Florence AT (2005) Issues in oral nanoparticle drug carrier uptake and targeting. / Drug Targ 12:65-70. 26 Florence 34. Singh B, Hussain N, Sakthivel T and Florence AT (2003) Effect of physiological media on the stability of surface-adsorbed DNA-dendron-gold nanoparticles. / Pharm Pharmacol 55:1635-1640. 35. Gabor F, Bogner E, Weissenboeck A and Wirth M (2004) The lectin-cell interaction and its implications to intestinal lectin-mediated drug delivery. Adv Drug Del Rev 56:459-480. 36. Adamczyk Z, Siwek B, Jaszczolt K and Weronski P (2004) Deposition of latex particles at heterogeneous surfaces. Colloids Surface A: Physicochem Eng Aspects 249:95-98. 37. Adamczyk Z (1989) Particle transfer and deposition from flowing colloid suspensions. Coll Surf 35:283-308. 38. Vacheethasanee K and Marchant RE (2000) Non-specific staphylococcus epidermidis adhesion: Contribtuions of biomaterial hydrophobicity and charge, in An, YH, Friedman RJ (eds.) Handbook of Bacterial Adhesion: Principles, Methods and Applications. Humana Press, Totowa, NJ, pp. 73-90. 39. Patil VRS, Campbell CJ, Yun YH, Slack SM and Goettz DJ (2001) Particle diameter influences adhesion under flow. Biophys J 80:1733-1743. 40. Bowen WR and Mongruel A (1998) Calculation of the collective diffusion coefficient of electrostatically stabilised colloidal particles. Coll Surface A 138:161-172. 41. Bhatia SK, King MR and Hammer DA (2003) The state diagram for cell adhesion mediated by two receptors. Biophys J 84:2671-2690. 42. Akerman ME, Chan WC, Laakkonen P, Bhatia SN and Ruoslahti E (2002) Nanocrystal targeting in vivo. Proc Natl Acad Sci USA 99:12617-12621. 43. Uchegbu IF, Schatzlein A, Vanlerberghe GMN and Florence AT (1997) Polyhedral nonionic surfactant vesicles. J Pharm Pharmacol 49:606-610. 44. Florence AT, Nasseri B and Arunothyanun P (2004) Does shape matter? Spherical, polyhedral and tubular vesicles, in Sonke S (ed.) Carrier-based Drug Delivery. American Chemical Society, Washington, pp. 75-84. 45. Schins RP (2002) Mechanisms of genotoxicity of particles and fibers. Inhal Toxicol 14: 57-78. 46. Bruinsma R (2005) Rheology and shape transitions of vesicles under capillary flow. Physica A 234:249-270. 47. Uchegbu IF, Double JA, Turton JA and Florence AT (1995) Distibution, metabolism and tumoricidal activity of doxorubicin administered in sorbitan monostearate (Span 60) niosomes in the mouse. Pharm Res 12:1019-1024. 48. Nasseri B and Florence AT (2003) Some properties of extruded non-ionic surfactant micro-tubes. Int f Pharm 254:11-16. 49. Vasanthi R and Bhattacharyya S (2005) Anisotropic diffusion of spheroids in liquids: Slow orientational relaxation of the oblates. / Chem Phys 116:1092-1096. 50. Mody NA, Lomakin O, Doggett TADTG and King MR (2005) Mechanics of transient platelet adhesion to von Willebrand factor under flow. Biophys J 88:1432-1443. 51. Kern N and Fourcade B (1999) Vesicles in linearly forced motion. Europhys Lett 46:262-267. 52. Nasseri B and Florence AT (2005) The relative flow of the walls of phospholipid tethers. Int J Pharm 298:372-377. Nanoparticle Flow: Implications for Drug Delivery 27 53. Naess SN and Elgsaeter A (2005) Transport properties of non-spherical nanoparticles studied by Brownian dynamics: Theory and numerical simulations. Energy 30:831-844. 54. Padera TP, Stoll BR, Tooredman JB, Capen D, di Tomaso E and Jain RK (2004) Cancer cells compress intratumour vessels. Nature 427:695. 55. Maeda H (2001) The enhanced permeability and retention (EPR) effect in tumor vasculature: the key role of tumor-selective macromolecular drug targeting. Adv Enzyme Regul 41:189-207. 56. Ramachandran VV, Venkatesan R, Tryggvason G and Scott FH (2000) Low Reynolds Number Interactions between Colloidal Particles near the Entrance to a Cylindrical Pore. / Coll Interf Sci 229:311-322. 57. Wang Y, Hu JK, Krol A, Li YP, Li CY and Yuan F (2003) Systemic dissemination of viral vectors during intratumoral injection. Mol Cancer Ther 2:1233-1242. 58. McGuire S and Yuan F (2001) Quantitative analysis of intratumoral infusion of color molecules. Am J Physiol Heart Circ Physiol 281:H715-H721. 59. Jones AT, Gumbleton M and Duncan R (2003) Understanding endocytic pathways and intracellular trafficking; a prerequisite for effective design of advanced drug delivery systems. Adv Drug Del Rev 55:1353-1357. This page is intentionally left blank 3 Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices SM Moghimi, E Vega, ML Garcia, OAR Al-Hanbali and KJ Rutt 1. Introduction Polymeric nanoparticles are submicron size entities, often ranging from 10-1000 nm in diameter, and are assembled from a wide variety of biodegradable (e.g. albumin, chitosan, alginate) and non-biodegradable polymers (Tables 1 and 2). The most active area of research using polymeric nanoparticles is in controlled delivery of pharmaceuticals following parenteral, oral, pulmonary, nasal, and topical routes of administration.1-6 Indeed, therapeutic agents can be encapsulated, covalently attached, or adsorbed onto such nanocarriers. These approaches can easily overcome drug solubility issues; this is particularly important as a significant proportion of new drug candidates arising from high-throughput screening initiatives are water insoluble. Polymeric nanoparticles, however, differ from nanosuspensions of drugs which are sub-micron colloidal dispersions of pure particles of drug that are stabilized by surfactants.7 By virtue of their small size and by functionalizing their surface with polymers and appropriate ligands, polymeric nanoparticles can also be targeted to specific cells and locations in the body.1,3'5'8-10 Thus, polymeric nanoparticles may overcome stability issues for certain drugs and minimize druginduced side effects. The extent of drug encapsulation/incorporation, as well as 29 30 Moghimi etal. the release profile from polymeric nanocarriers, however, depends on the polymer type and its physicochemical properties, the particle size and its morphology (e.g. solid nanospheres as opposed to polymeric nanocapsules).4 In addition, depending on the polymer characteristics, polymeric nanocarriers can also be engineered in such a way that they can be activated by changes in the environmental pH, chemical stimuli, or temperature.1112 Such modifications offer control over particle integrity, drug delivery rates, and the location of drug release, for example, within specific organelles. For instance, nanoparticles made from poly(lactide-coglycolide), PLGA, can escape the endo-lysosomal compartment within minutes of internalization in intact cells and reach the cytosol.12 This is due to the selective reversal of the surface charge of nanoparticles from the anionic to the cationic state in endo-lysosomes, resulting in a local particle-membrane interaction with subsequent cytoplasmic release. This is an excellent approach for channelling antigens into the highly polymorphic MHC class-I molecules of macrophages and dendritic cells for subsequent presentation to CD8+ T lymphocytes. Other applications include cytoplasmic release of plasmid vectors and therapeutic agents (e.g. for combating cytoplasmic infections and for slow cytoplasmic release of drugs that act on nuclear receptors). Polymeric nanoparticles are also beginning to make a significant impact on global pharmaceutical planning (life-cycle management) and market intelligence. For example, due to imminent expiration of patents, pharmaceutical companies may launch follow-up or nano-formulated versions of a product to minimize generic threats to best-selling medicines. This could lead to an extension of as much as 20 years from a new patent on the nanoparticulate formulation of the drug. By coalescing certain polymeric nanoparticles carefully from an aqueous suspension, shape retentive hydrogels can be formed to erode partially or completely.1113 Drugs and macromolecules may be trapped within interstitial spaces between particles during aggregate formation. Thus, hydrogel nanoparticles have potential as controlled release implant devices following local administration or implantation, and may also serve as tissue engineering scaffolds with concurrent morphogenic protein release. This article will briefly review some of the most commonly used laboratory scale methods for the production of polymeric nanoparticles and drug encapsulation procedures. The importance of the nanometre scale size range and surface engineering strategies for site-specific targeting of polymeric nanoparticles, following different routes of administration, are also discussed. 2. Nanoparticle Engineering Polymeric nanoparticles are usually prepared either directly from preformed polymers such as aliphatic polyesters (Table 1) and block copolymers (Table 2), Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices 31 Table 1 Chemical properties of some commonly used aliphatic polyesters in nanoparticle engineering. Polymer Type Melting Point (°C) Glass Transition Resorption Time Temperature (°C) (Months) DL-PLA Amorphous 50-60 12-16 PGA 220-230 35^0 6-12 DL-PLGA (50/50) Amorphous 45-50 1-2 DL-PLGA (75/25) Amorphous 45-50 4-5 PCL 55-65 (-65)-(-60) >24 DL-PLA: poly(L-lactide); PGA: poly(glycolide); DL-PLGA: poly(DL-lactide-co-glycolide); PCL: poly- .-caprolactone. Table 2 Selected examples of block copolymers for production of biodegradable nanospheres. PLA-poly(ethyleneglycol),PLA-PEG MonomethoxyPEG-poly(alkylcyanoacrylate) Poly(poly(ethyleneglycol)cyanoacrylate-co-hexadecylcyanoacrylate) Poly(ethyleneoxide-b-sebacicacid) Poly(phosphazene)-poly(ethyleneoxide) poly(2-methyloxazoline)-b-poly(dimethylsiloxane)-b-poly(2-methyloxazoline) or by polymerization of monomers.4 Commonly used methodologies include the solvent evaporation,14-15 the spontaneous emulsification/ solvent diffusion,16 nanoprecipitation or solvent displacement17'18 and emulsion polymerization techniques.19-21 The method of choice depends on the polymer and the drug type, as well as the required particle size distribution and polydispersity indices. However, some polymers, such as comb-like polyesters, the di-block copolymer poly(ethylene oxide-b-sebacic acid) and tri-block copolymer poly(2- methyloxazoline)-fr-poly(dimethylsiloxane)-fr-poly(2-methyloxazoline) can spontaneously form stable nanoparticles (core-shell type nanospheres).22-24 In the solvent evaporation method, the polymer is simply dissolved together with the drug in an organic solvent and the mixture is then emulsified to form either an oil-in-water nanoemulsion (for encapsulation of hydrophobic drugs) or waterin- oil nanoemulsion (for encapsulation of hydrophilic drugs) using suitable surfactants. Nanoparticles are then obtained following evaporation of the solvent and can be concentrated by filtration, centrifugation or lyophilization. The spontaneous emulsification/solvent diffusion method is a modified version of the solvent evaporation technique, which utilizes a water-soluble solvent (e.g. methanol or acetone) along with a water-insoluble one such as chloroform. As a result of the spontaneous 32 Moghimi etal. diffusion of the water-soluble solvent into the water-insoluble phase, an interfacial turbulence is created leading to the formation of nanoparticles. Nanoprecipitation, however, is a versatile and simple method. This is based on spontaneous formation of nanoparticles during phase separation (the Marangoni effect), which is induced by slow addition of the diffusing phase (polymer-drug solution) to the dispersing phase (a non-solvent of the polymers, which is miscible with the solvent that solubilizes the polymer). The dispersing phase may contain surfactants. Depending on the solvent choice and solvent/non-solvent volume ratio, this method is suitable for encapsulation of both water-soluble and hydrophobic drugs, as well as protein-based pharmaceuticals.17'18 In emulsion polymerization, the monomer is dispersed into an aqueous phase using an emulsifying agent. The initiator radicals are generated in the aqueous phase and they diffuse into the monomer-swollen micelles. Anionic polymerization in the micelles is then initiated by the hydroxyl ions of water. Chain transfer agents are abundant and termination occurs by radical combination. The size and molecular masses of nanoparticles are dependent on the initial pH of the polymerization medium.20 Drugs are incorporated during the polymerization step or can be adsorbed into the nanosphere surface afterwards. The addition of cyclodextrins to the polymerization medium can promote the encapsulation of poorly watersoluble drugs.25 Depending on the monomer used, some drugs can also initiate the polymerization step, resulting in the covalent attachment of drug molecules to the nanospheres. For instance, photosensitizers such as naphthalocyanines, can initiate the polymerization of alkylcyanoacrylates.26 A number of specialized approaches (e.g. dialysis, salting-out, supercritical fluid technology, denaturation, ionic interaction, ionic gelation, and interfacial polymerization) have also been described for the preparation of polymeric nanoparticles, based on the choice of the starting material and the biological needs.4'27-32 2.1. Drug release mechanisms The release profile of drugs from nanoparticles depends on the physicochemical nature of the drug molecules as well as the matrix.4'16'28,33-36 Factors include mode of drug attachment and/or encapsulation (e.g. surface adsorption, dispersion homogeneity of drug molecules in the polymer matrix, covalent conjugation), the physical state of the drug within the matrix (such as crystal form), and parameters controlling matrix hydration and/or degradation. Generally, rapid release occurs by desorption, where the drug is weakly bound to the nanosphere surface. If the drug is uniformly distributed in the polymer matrix, the release occurs either by diffusion (if the encapsulated drug is in crystalline form, the drug is first dissolved locally Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices 33 and then diffuses out) or erosion of the matrix, or a combination of both mechanisms. Erosion can be further subdivided into either homogeneous (with uniform degradation rates throughout the matrix) or heterogeneous (where degradation is confined at the surface) processes. Parameters such as polymer molecular weight distribution, crystallinity, hydrophobicity/hydrophilicity, melting and glass transition temperature, polymer blends and prior polymer treatment (e.g. oxygen-plasma treatment) all control the extent of matrix hydration and degradation. For instance, in the case of aliphatic polyesters, their degradation time is shorter for low molecular weight polymers, more hydrophilic polymers, more amorphous polymers and copolymers with high glycolide content (Table 1). 3. Site-specific Targeting with Nanoparticles: Importance of Size and Surface Properties Numerous articles have recently discussed the importance of nanoparticle size and surface characteristics in controlling their biodistribution, following different routes of administration.1 ~3/5 Only a brief overview is provided here. Following intravenous injection, liver (Kupffer cells) and spleen (marginal zone and red pulp) macrophages clear polymeric nanoparticles rapidly from the blood circulation.1 Opsonization, which is surface deposition of blood opsonic factors such as fibronectin, immunoglobulins, C-reactive and certain complement proteins, often aid particle recognition by these macrophages. Indeed, the propensity of macrophages of the reticuloendothelial system for rapid recognition and clearance of particulate matter has provided a rational approach to macrophagespecific targeting with nanoparticles (e.g. for the treatment of obligate intracellular microorganisms, delivery of toxins for macrophage killing, and diagnostic agents).1 However, the rapid sequestration of nanoparticles by macrophages in contact with blood is problematic for the efficient targeting of polymeric nanoparticles to nonmacrophage sites. Thus, inherent in nanoparticle design is the precision surface manipulation and engineering with synthetic polymers; this affords control over nanoparticle interaction and fate within biological systems. There are numerous examples where the surface of nanocarriers is carefully assembled with projected "macromolecular hairs" made from poly(ethyleneglycol), PEG, or its derivatives (e.g. methoxyPEG-albumin, PLA-PEG) or other related polymers [e.g. block copolymers such as selected poloxamers and poloxamines, poly(phosphazene)- poly(ethyleneoxide)].3,5 This is achieved either during the particle assembly procedures or polymerization step, or post particle manufacturing. This strategy suppresses macrophage recognition by an array of complex mechanisms, which collectively achieve reduced protein adsorption and surface opsonization. Therefore, such entities, provided that they are below 150 nm in size, exhibit prolonged 34 Moghimi et al. residency time in the circulation, and are referred to as "stealth" or "macrophageevading" nanoparticles.1,5 The efficiency of the "macrophage-evading" process is dependent on polymer type and its surface stability, reactivity, and physics (e.g. surface density and assumed conformation).5 Prolonged circulation properties are ideal for slow or controlled release of therapeutic agents in the blood to treat vascular disorders. Long circulating polymeric nanoparticles may have application in vascular imaging too (e.g. detection of vascular bleeding or abnormalities). Long-circulating nanoparticles can also escape from vasculature and this is normally restricted to sites where the capillaries have open fenestration or when the integrity of the endothelial barrier is perturbed by inflammatory processes or by tumor growth.5 However, extravasated nanoparticles, as in tumour interstitium, distribute heterogeneously in perivascular clusters that do not move significantly; these particles may therefore act as depot systems, particularly for the sustained release of antiangiogenic agents, and to some extent, for drug delivery to multidrug resistant tumors (e.g. by co-encapsulation of both anticancer drugs and the competitive inhibitors of active drug efflux pumps).1 The surface of long-circulating nanoparticles is also amenable for modification with targeting ligands. Such entities can navigate capillaries and escape routes in search of signature molecules expressed by the target; this process is often referred to as "active targeting".1-5 For example, certain cancer cells express folate receptors and these receptors have the ability to endocytose stealth nanoparticles that are decorated with folic acid. Delivery of anti-cancer agents to tumor cells by such means could overcome the possibility of multi-drug resistance.1,37 Non-deformable "stealth" nanoparticles, however, are prone to splenic filtration at interendothelial cell slits, if their size exceeds that of the width of the cell slits (200-250 nm).38,39 Indeed, these "splenotropic" vehicles can deliver their cargo efficiently to the red-pulp regions of the sinusoidal spleen. Activated or stimulated macrophages are also known to rapidly phagocytose stealth nanoparticles; stealth nanospheres may therefore have applications as diagnostic/imaging tools for the identification of stimulated or newly recruited hepatic macrophages.40 Such diagnostic procedures may prove useful for patient selection or for monitoring the progress of treatment with long-circulating nanoparticles carrying anti-cancer agents, thus minimizing damage to hepatic macrophages.41 Polymeric nanospheres can also target endothelial cells on the bloodbrain barrier. For instance, following intravenous injection polysorbate 80-coated poly(alkylcyanoacrylate), PACA, nanospheres attract apolipoprotein E from the blood, thus mimicking low density lipoprotein (LDL) and become recognizable by LDL receptors expressed by the blood-brain barrier endothelial cells.10 Another related example is PEG-coated PACA nanoparticles, with the ability to localize mainly in the ependymal cells of the choroid plexus and the epithelial cells of pia Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices 35 region and the ventricles of the mouse and the rat brain.42 The molecular basis of this deposition pattern remains to be unravelled. Others have administered nanoparticles directly to pathological sites for optimal biological performance.43 One example is intramurally delivered PLGA nanoparticles to an injured artery following angioplasty, using a cardiac infusion catheter. Here, nanoparticles penetrate the dilated arterial wall under pressure and once the pressure is released, the artery returns to its normal state resulting in particle immobilization in the arterial wall, where they may act as a sustained release system for drugs and genetic materials.43 Again, particle size is an important parameter; the smaller the size, the greater the arterial deposition and cellular entry, as well as lower inflammatory responses. Polymeric nanospheres also provide intriguing opportunities for lymphatic drug delivery, as well as for diagnostic imaging of the lymphatic vessels and their associated lymph nodes when injected interstitially.44 The extent of lymphatic delivery and lymph node localization of nanospheres depends on their size and surface characteristics. For instance, hydrophilic nanoparticles, in the size range of 30-100 nm, as opposed to their hydrophobic counterparts, repulse each other and interact poorly with the ground substance of the interstitium and drain rapidly into the initial lymphatics through patent junctions in the lymphatic capillaries.45,46 The drained particles are conveyed to the nodes via the afferent lymph. Macrophages of medullary sinuses and paracortex are mainly responsible for particle capture from the lymph, but this also depends on nanoparticle surface properties. Larger nanospheres (>150nm), however, are retained at interstitial sites for prolonged periods of time and may therefore act as sustained release systems for drugs and antigens.47,48 For example, large-sized PLGA particles can provide antigen release over weeks and months following continuous or pulsatile kinetics. By mixing particle types with different degradation and pulsatile release kinetics, multiple discrete booster doses of encapsulated antigens can be provided after a single administration of the formulation (e.g. 1-2 and 6-12 months).48 An alternative approach is the use of nanoparticle hydrogels for slow and local antigen release. For example, by controlling the ionic strength of the dispersion medium, monodisperse nanoparticles of poly-2-hydroxyethylmethacrylate, poly(HEMA), and poly[HEMA-co-methacrylic acid] coalesce together to form a shape retentive hydrogel suitable for interstitial implantation.13 Macromolecules may be trapped between the particle aggregates and their release is controlled by a combination of diffusion (larger particles packed together have larger spaces in the lattice, and this allows for faster diffusion) and erosion (arising from aggregates that contain particles with methacrylic acid).13 Nanoparticles that erode from the aggregate are drained into the lymphatic system and may be retained by the regional nodes. Similarly, by controlling the inherent physical attractive forces between model polystyrene nanoparticles, ordered lattices 36 Moghimi et al. Fig. 1. Scanning electron micrographs of uncoated and surface-modified polystyrene nanoparticles. Due to surface hydrophobicity uncoated nanospheres (A), 350 nm in size, tend to aggregate. By controlling the physical attractive forces between the nanoparticles (by surface coating with an appropriate concentration of a block copolymer), ordered structures are formed and these can be deposited onto the surface of large microspheres (B). can be deposited on the surface of very large microspheres (Fig. 1). Following subcutaneous localization, surface adsorbed nanospheres may gradually detach from the parent microsphere and gain entry into the lumen of the lymphatic capillaries. Polymeric nanoparticles also have numerous applications following oral delivery. Evidence suggests that the adsorption of particulates in the intestine following oral administration take place at the Peyer's patches.49-50 The epithelial cell layer overlying the Peyer's patches contains specialized M cells. These cells can Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices 37 sample particles from the lumen and transport them to the underlying macrophages and dendritic cells. Indeed, numerous studies have confirmed protective immunity induced by mucosal immunization with PACA, PLGA and chitosan based particulate systems.3,32,48'50-53 Part of the success is due to the encapsulation of antigens in polymeric particulate systems, which provides better protection for the antigen during intestinal transit. The immune outcomes have included mucosal (secretory IgA) and serum antibody (IgG and IgM) responses, as well as systemic cytotoxic T lymphocyte responses in splenocytes. Induction of an appropriate immune response following oral administration depends primarily on factors that affect uptake and particle translocation by M cells. These include particle size, dose, composition, and surface chemistry, as well as the region of the intestine where particles are taken up, membrane recycling from intracellular sources and the species.50 Tolerance to orally administered microparticulate encapsulated antigens is another potential outcome, but it has received little attention. The bioavailability of some drugs can be improved after oral administration by means of polymeric nanoparticles.54-57 This is a reflection of drug protection by the nanoparticle against hostile conditions of the gastrointestinal tract, as well as the mode of nanoparticle interaction with mucosal layers. However, the bioadhesive properties of nanoparticles may vary with their size and surface characteristics (e.g. surface charge, surface polymer density and conformation), as well as the location and type of the mucosal surface in the gastrointestinal tract. Similarly, improved drug bioavailability has also been reported following ocular administration with PLA, PACA, poly(butylcyanoacrylate) and Eudragit nanoparticles.6,58-61 For example, loading of tamoxifen in PEGylated nanoparticles proved successful in the treatment of autoimmune uveortinitis following intraocular injection.59 Interaction of surface-modified polymeric nanoparticles with nasal associated lymphoid tissue and their transport across nasal mucosa have also received attention, particularly with respect to peptide-based pharmaceuticals and antigen delivery.53,62 4. Conclusions Polymeric nanoparticles are promising vehicles for site-specific and controlled delivery of therapeutic agents, following different routes of administration and these trends seem to continue with advances in materials and polymer chemistry and pharmaceutical nanotechnology. However, nanoparticles do not behave similarly; their encapsulation capacity, drug release profile, biodistribution and stability vary with their chemical makeup, morphology and size. Inherently, nanosphere design and targeting strategies may vary according to physiological and therapeutic needs, as well as in relation to the type, developmental stage and location of 38 Moghimietal. the disease. Attention should also be paid to toxicity issues that may arise from nanoparticle administration and the release of their polymeric contents and degradation products. These issues are discussed elsewhere.1,63~66 References 1. Moghimi SM, Hunter AC and Murray JC (2005) Nanomedicine: Current status and future prospects. FASEB ] 19:311-330. 2. Panyam J and Labhasetwar V (2003) Biodegradable nanoparticles for drug and gene delivery to cells and tissue. Adv Drug Del Rev 55:329-347. 3. Vauthier C, Dubernet C, Fattal E, Pinto-Alphandary H and Couvreur P (2003) Poly (alkyleyanoacrylates) as biodegradable materials for biomedical applications. Adv Drug Del Rev 55:519-548. 4. Soppimath KB, Aminabhavi TM, Kulkami AR and Rudzinski WE (2001) Biodegradable polymeric nanoparticles as drug delivery devices. / Control Rel 70:1-20. 5. Moghimi SM, Hunter AC and Murray JC (2001) Long-circulating and target-specific nanoparticles: Theory to practice. Pharmacol Rev 53:283-318. 6. Salgueiro A, Egea MA, Espina M, Vails O and Garcia ML (2004) Stability and ocular tolerance of cyclophosphamide-loaded nanospheres. J Microencapsul 21:213-223. 7. Rabinow BE (2004) Nanosuspensions in drug delivery. Nat Rev Drug Discov 3: 785-796. 8. Moghimi SM (2002) Chemical camouflage of nanospheres with a poorly reactive surface: Towards development of stealth and target-specific nanocarriers. Biochim Biophys Acta (Mol Cell Res) 1590:131-139. 9. Porter CJH, Moghimi SM, Ilium L and Davis SS (1992) The polyoxyethylene/ polyoxypropylene block co-polymer poloxamer-407 selectively redirects intravenously injected microspheres to sinusoidal endothelial cells of rabbit bone marrow. FEBS Lett 305:62-66. 10. Kreuter J, Ramge P, Petrov V, Hamm S, Gelperina SE, Engelhardt B, Alyautdin R, von Briesen H and Begley DJ (2003) Direct evidence that polysorbate-80-coated poly(butylcyanoacrylate) nanoparticles deliver drugs to the CNS via specific mechanisms requiring prior binding of drugs to the nanoparticles. Pharm Res 20:409-416. 11. Huang G, Gao J, Hu Z, St. John JV, Ponder BC and Moro D (2004) Controlled drug release from hydrogel nanoparticle network. / Control Rel 94:303-311. 12. Panyam J, Zhou WZ, Prabha S, Sahoo SK and Labhasetwar V (2002) Rapid endolysosomal escape of poly(DL-lactide-co-glycolide) nanoparticles: Implications for drug and gene delivery. FASEB J 16:1217-1226. 13. St. John JV, Moro DG, Russell-Jones GJ and McDougall F (2004) Protein release from and cellular infiltration into hydrogel nanoparticle scaffolds. 31st Annual Meeting of the Controlled Release Society, Honolulu, Hawaii, June 12-16. 14. Scholes PD, Coombes AGA, Ilium L, Davis SS, Vert M and Davies MC (1993) The preparation of sub-500 nm poly(lactide-co-glycolide) microspheres for site-specific drug delivery. / Control Rel 25:145-153. Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices 39 15. Desgouilles S, Vauthier C, Bazile D, Vacus J, Grossiord JL, Veillard M and Couvreur P (2003) The design of nanoparticles obtained by solvent evaporation: A comparative study. Langmuir 19:9504-9510. 16. Niwa T, Takeuchi H, Hino T, Kunou N and Kawashima Y (1993) Preparations of biodegradable nanospheres of water-soluble and insoluble drugs with D,Llactide/ glycolide copolymer by a novel spontaneous emulsification solvent diffusion method and the drug release behavior. / Control Rel 25:89-98. 17. Quintanar-Guerrero D, Allemann E, Fessi H and Doelker E (1998) Preparation techniques and mechanisms of formation of biodegradable nanoparticles from preformed polymers. Drug Dev Ind Pharm 24:1113-1128. 18. Bilati U, Allemann E and Doelker E (2005) Development of a nanoprecipitation method intended for the entrapment of hydrophilic drugs into nanoparticles. Eur } Pharm Sci 24: 67-75. 19. Couvreur P, Kante M, Roland M, Guiot P, Bauduin P and Speiser P (1979) Polycyanoacrylate nanocapsules as potential lysosomotropic carriers: Preparation, morphological and sorptive properties. / Pharm Pharmacol 31:331-332. 20. Lescure F, Zimmer C, Roy D and Couvreur P (1992) Optimization of polycyanoacrylate nanoparticle preparation: Influence of sulfur dioxide and pH on nanoparticle characteristics. / Coll Interf Sci 154:77-86. 21. De Keyser JL, Poupaert JH and Dumont P (1991) Poly(diethylmethylidenemalonate) nanoparticles as a potential drug carrier: Preparation, distribution and elimination after intravenous and peroral administration to mice. / Pharm Sci 80:67-70. 22. Jung T, Breitenbach A and Kissel T (2000) Sulfobutylated poly(vinylalcohol)-graftedpoly( lactide-co-glycolide) facilitate the preparation of small negatively charged biodegradable nanospheres for protein delivery. / Control Rel 67:157-169. 23. Wu C, Fu J and Zhao Y (2000) Novel nanoparticles formed via self-assembly of poly(ethylene glycol-b-sebacic anhydride) and their degradation in water. Macromolecules 33:9040-9043. 24. Broz P, Benito SM, Saw CL, Burger P, Heider H, Pfisterer M, Marsch S, Meier W and Hunziker P (2005) Cell targeting by a generic receptor-targeted polymer nanocontainer platform. / Control Rel 102:475^88. 25. Boudad H, Legrand P, Lebas G, Cheron M, Duchene D and Ponchel GG (2001) Combined hydroxylpropyl-beta-cyclodextrin and poly(alkylcyanoacrylate) nanoparticles intended for oral administration of saquinavir. Int} Pharm 218:113-124. 26. Labib A, Lenaerts V, Chouinard F, Leroux JC, Ouellet R and van Lier JE (1991) Biodegradable nanospheres containing phthalocyanines and naphthalocyanines for targeted photodynamic tumor therapy. Pharm Res 8:1027-1031. 27. Jeong YI, Cho CS, Kim SH, Ko KS, Kim SI, Shim YH and Nah JW (2001) Preparation of poly(DL-lactide-co-glycolide) nanoparticles without surfactant. / Appl Polym Sci 80: 2228-2236. 28. Allemann E, Leroux JC, Gurnay R and Doelker E (1993) In vitro extended-release properties of drug-loaded poly(D,L-lactic) acid nanoparticles produced by a salting-out procedure. Pharm Res 10:1732-1737. 40 Moghimi et al. 29. Randolph TW, Randolph AD, Mebes M and Yeung S (1993) Submicron-sized biodegrad-. able particles of poly(L-lactic acid) via the gas antisolvent spray precipitation process. Biotechnol Prog 9:429-435. 30. Tokumitsu H, Ichikawa H and Fuukumori Y (1999) Chitosan-gadopenteic acid complex nanoparticles for gadolinium neutron-capture therapy of cancer: Preparation by novel emulsion-droplet coalscence technique and charaterization. Pharm Res 16:1830-1835. 31. Prokop A, Kozlov E, Newman GW and Newman MJ (2002) Water-based nanoparticulate polymeric system for protein delivery: Permeability control and vaccine application. Biotechnol Bioeng 78:459^166. 32. Calvo P, Remunan-Lopez C, Vila-Jato JL and Alonso MJ (1997) Chitosan and chitosan/ ethylene oxide-propylene oxide block copolymer nanoparticles as novel carriers for proteins and vaccines. Pharm Res 14:1431-1436. 33. Liu H, Finn N and Yates MZ (2005) Encapsulation and sustained release of a model drug, indomethacin, using C02-based microencapsulation. Langmuir 21:379-385. 34. Panyam J, Williams D, Dash A, Leslie-Pelecky D and Labhasetwar V (2004) Solid-state solubility influences encapsulation and release of hydrophobic drugs from PLGA/PLA nanoparticles. f Pharm Sci 93:1804-1814. 35. Polakovic M, Gorner T, Gref R and Dellacherie E (1999) Lidocaine loaded biodegradable nanospheres. II. Modeling of drug release. / Control Rel 60:169-177. 36. Tamber H, Johansen P, Merkle HP and Gander B (2005) Formulation aspects of biodegradable polymeric microspheres for antigen delivery. Adv Drug Del Rev 57: 357-376. 37. Stella B, Arpicco S, Peracchia MT, Desmaele D, Hoebeke J, Renoir M, D'Angelo J, Cattel L and Couvreur P (2000) Design of folic acid-conjugated nanoparticles for drug targeting. / Pharm Sci 89:1452-1464. 38. Moghimi SM, Porter CJH, Muir IS, Ilium L and Davis SS (1991) Non-phagocytic uptake of intravenously injected microspheres in rat spleen: Influence of particle size and hydrophilic coating. Biochem Biophys Res Commun 177:861-866. 39. Moghimi SM, Hedeman H, Ilium L and Davis SS (1993) Effect of splenic congestion associated with haemolytic anaemia on filtration of "spleen-homing" microspheres. Clin Sci 84:605-609. 40. Moghimi SM, Hedeman H, Christy NM, Ilium L and Davis SS (1993) Enhanced hepatic clearance of intravenously administered sterically stabilized microspheres in zymosanstimulated rats. / Leukoc Biol 54:513-517. 41. Laverman P, Carstens MG, Storm G and Moghimi SM (2001) Recognition and clearance of methoxypoly(ethyleneglycol) 2000-grafted liposomes by macrophages with enhanced phagocytic capacity. Implications in experimental and clinical oncology. Biochim Biophys Acta (General Subjects) 1526:227-229. 42. Calvo P, Gouritin B, Villarroya H, Eclancher F, Giannavola C, Klein C, Andreux JP and Couvreur P (2002) Quantification and localization of PEGylated polycyanoacrylate nanoparticles in brain and spinal cord during experimental allergic encephalomyelitis in the rat. Eur } Neurosci 15:1317-1326. Polymeric Nanoparticles as Drug Carriers and Controlled Release Implant Devices 41 43. Song C, Labhasetwar V, Cui X, Underwood T and Levy RJ (1998) Arterial uptake of biodegradable nanoparticles for intravascular local drug delivery: Results with an acute dog model. / Control Rel 54:201-211. 44. Moghimi SM and Bonnemain B (1999) Subcutaneous and intravenous delivery of diagnostic agents to the lymphatic system: Applications in lymphoscintigraphy and indirect lymphography. Adv Drug Del Rev 37:295-312. 45. Hawley AE, Ilium L and Davis SS (1997) Lymph node localisation of biodegradable nanospheres surface modified with poloxamer and poloxamine block co-polymers. FEBS Lett 400:319-323. 46. Moghimi SM (2003) Modulation of lymphatic distribution of subcutaneously injected poloxamer 407-coated nanospheres: The effect of the ethylene oxide chain configuration. FEBS Lett 540:241-244. 47. Moghimi SM and Rajabi-Siahboomi AR (1996) Advanced colloid-based systems for efficient delivery of drugs and diagnostic agents to the lymphatic tissues. Prog Biophys Mol Biol 65:221-249. 48. Jiang W, Gupta RK, Deshpande MC and Schwendeman SP (2005) Biodegradable poly (lactic-co-glycolic acid) microparticles for injectable delivery of vaccine antigens. Adv Drug Del Rev 57:391^10. 49. Simecka JW (1998) Mucosal immunity of the gastrointestinal tract and oral tolerance. Adv Drug Del Rev 34:235-259. 50. Ermak TH and Giannasca PJ (1988) Microparticle targeting to M cells. Adv Drug Deliv Rev 34:261-283. 51. O'Hagan DT and Valiante NM (2003) Recent advances in the discovery and delivery of vaccine adjuvants. Nat Rev Drug Discov 2:727-735. 52. O'Hagan DT, Singh M and Ulmer JB (2004) Microparticles for delivery of DNA vaccines. Immunol Rev 199:191-200. 53. van der Lubben IM, Kersten G, Fretz MM, Beuvery C, Verhoef JC and Junginger HE (2003) Chitosan microparticles for mucosal vaccination against diphteria: Oral and nasal efficacy studies in mice. Vaccine 21:1400-1408. 54. Takeuchi H, Yamamoto H and Kawashima YY (2001) Mucoadhesive nanoparticulate systems for peptide drug delivery. Adv Drug Del Rev 47:39-54. 55. Carino GP, Jacob JS and Mathiowitz E (2000) Nanosphere based oral insulin delivery. / Control Rel 65:261-269. 56. Arbos P, Campanero MA, Arangoa MA, Renedo MJ and Irache JM (2003) Influence of the surface characteristics of PVM/MA nanoparticles on their bioadhesive properties. / Control Rel 89:19-30. 57. Tobio M, Sanchez A, Vila A, Soriano I, Evora C, Vila-Jato JL and Alonso MJ (2000) The role of PEG on the stability in digestive fluids and in vivo fate of PEG-PLA nanoparticles following oral administration. Coll Surf (B-Biointerface) 18:315-323. 58. Fresta M, Fontana C, Bucolo G, Cavallaro G, Giammona G and Puglisi G (2001) Ocular tolerability and in vivo bioavailability of poly(ethylene glycol) (PEG)-coated polyethyl- 2-cyanoacrylate nanospheres-encapsulated acyclovir. / Pharm Sci 90:288-297. 42 Moghimi et al. 59. de Kozak Y, Andrieux K, Villarroya H, Klein C, Thillaye-Goldenberg B, Naud MC, Garcia E and Couvreur P (2004) Intraocular injection of tamoxifen-loaded nanoparticles: Anew treatment of experimental autoimmune uveoretinitis. Eur J Immunol 34:3702-3712. 60. Giannavola C, Bucolo C, Maltese A, Paolino D, Vandelli MA, Puglisi G, Lee VHL and Fresta M (2003) Influence of preparation conditions on acyclovir-loaded poly-d,l-lactic acid nanospheres and effect of PEG coating on occular drug bioavailability. Pharm Res 20:584-590. 61. Bucolo C, Maltese A, Maugeri F, Busa B, Puglisi G and Pignatello R (2004) Eudragit RL100 nanoparticle system for the opthalmic delivery of cloricromene. / Pharm Pharmacol 56:841-846. 62. Vila A, Gill H, Mccallion O and Alonso MJ (2004) Transport of PLA-PEG particles across the nasal mucosa: Effect of particle size and PEG coating density. / Control Rel 98:231-244. 63. Moghimi SM, Hunter AC, Murray JC and Szewczyk A (2004) Cellular distribution of nonionic micelles. Science 303:626-627. 64. Hunter AC and Moghimi SM (2002) Therapeutic synthetic polymers: A game of Russian roulette? Drug Discov Today 7:998-1001. 65. Moghimi SM, Symonds P, Murray JC, Hunter AC, Debska G and Szewczyk A (2005) A two-stage poly(ethylenimine)-mediated cytotoxicity: Implications for genetransfer /therapy. Mol Ther 11:990-995. 66. Gbadamosi JK, Hunter AC and Moghimi SM (2002) PEGylation of microspheres generates a heterogeneous population of particles with differential surface characteristics and biological performance. FEBS Lett 532:338-344. 4 Genetic Vaccines: A Role for Liposomes Gregory Gregoriadis, Andrew Bacon, Brenda McCormack and Peter Laing 1. Introduction Prevention of microbial infections by the use of vaccines is a preferred alternative to treatment. Vaccines have been applied successfully, for example, in the eradication of smallpox as well as against tetanus, diphtheria, whooping cough, polio and measles, thus preventing millions of deaths each year. However, vaccines made of attenuated organisms, which mimick natural infections usually without the disease, can be potentially unsafe. For instance, there is a risk of reversion during replication of live viruses or even mutation to a more pathogenic state. Furthermore, with immunocompromised individuals, some of the attenuated viruses may still provoke disease. On the other hand, with killed virus vaccines, their extracellular localization and subsequent phagocytosis by professional antigen presenting cells (APC) or antigen-specific B cells, lead to MHC-II class restricted presentation and to T helper cell and humoural immunity. However, they do not elicit significant cytotoxic T cell (CTL) responses. Moreover, subunit vaccines produced from biological fluids may not be entirely free of infectious agents. Even with subunit and peptide vaccines produced recombinantly or synthetically (and thus considered safe), immune responses are weak and often not of the appropriate kind. The great variety of immunological adjuvants1'2 that are now available go a long way in rendering subunit and peptide vaccines stronger and more efficient. However, more than seventy years after the introduction of aluminium salts as an adjuvant, only two other adjuvants, liposomes3 and MF59,1 have been approved for use in humans.4 Thus, 43 44 Gregoriadis et al. inspite of considerable progress, the road to the ideal vaccine appears as elusive as ever, until recently. Recent developments have led to a novel and exciting concept, namely de novo production of the required vaccine antigen by the host's cells in vivo, which promises to revolutionize vaccination especially where vaccines are either ineffective or unavailable. The concept entails the direct injection of antigen-encoding plasmid DNA which, on uptake by cells, localizes to some extent into the nucleus where it transfects the cells episomally. The produced antigen is recognized as foreign by the host and is thus subjected to pathways similar to those observed for antigens of internalized viruses (but without their disadvantages), leading to protective humoural and cell mediated immunity.5-9 A series of publications since 1992 first established the ability of plasmid DNA to induce an immune (antibody) response to the encoded foreign protein10; in experiments with DNA encoding influenza nucleoprotein, immunity was both humoural and cell-mediated, and also protective in mice challenged with the virus.11,12 This was the first demonstration of an experimental DNA vaccine. Another observation was the induction of humoural and cell-mediated immunity against HIV-1 using plasmids encoding the HIV rev and env proteins.13 Similar results were obtained with a gene for the hepatitis B surface antigen (HBsAg).14 DNA immunization was also found to apply in cancer treatment. For instance, injection of plasmids encoding tumor antigens promoted immune responses15,16 which were protective in an animal model.6 The concept of DNA immunization has now been adopted by vaccinologists worldwide using an ever increasing number of plasmids encoding immunogens from bacterial, viral and parasitic pathogens, and a variety of tumors.8,9 In many of these studies, genetic immunization has led to the protection of animals from infection.5-9 A number of clinical trials for the therapy of, or prophylaxis against, a variety of infections are in progress.8,9 2. The DNA Vaccine A plasmid DNA vaccine is usually6 supercoiled and consists of the gene encoding the vaccine antigen (the section of the target pathogen which elicits protective immunity), a promoter sequence which is often derived from cytomegalovirus (CMV) or Rous sarcoma virus (RSV)), an mRNA stability polyadenylation region at the 3' end of the insert, and the plasminogen activator gene which controls the secretion of the recombinant product. In addition, there are an origin of replication for the amplification of the plasmid in bacteria, and a gene for antibiotic resistance to select the transformed bacteria. Immunization procedures with DNA vaccines are carried out by the intramuscular and, to a lesser extent, the intraepidermal route. Other routes include the Genetic Vaccines: A Role for Liposomes 45 oral, nasal, vaginal, intravenous, intraperitoneal and subcutaneous routes.8,9 Intramuscular injection of DNA vaccines leads to such types of immunity as CTL.5'9,11'12 This was unexpected because antigen presentation requires the function of professional APC.17 However, myocytes which were shown5 to take up the plasmid only to a small extent and with only a fraction of cells participating in the uptake, are not professional APCs. Although myocytes carry MHC class I molecules and can present endogenously produced viral peptides to the CD8+ cells to induce CTLs, they do so inefficiently18 as they lack vital costimulatory molecules (e.g. the B7-1 molecule). It is thus difficult to accept that antigen presentation, leading to a CTL response, occurs via myocytes. Instead, it was reported18 that CTL responses occur as a result of the transfer of antigenic material between the myocytes and professional APC to some extent. In parallel, it could also be that plasmid secreted by the myocytes or as such, is taken up directly by APC infiltrating the injected site. Such APC would include dendritic cells which will express and present peptides to CD8+ cells following transport to the lymph nodes or spleen. On the other hand, CD4+ cells may be activated by APCs via MHC class II presentation of antigen secreted by the myocytes (or released from them after their destruction via a Tc response) and captured by the cells. Such events would lead to both cellular (Th 1) and humoural (Th 2) immunity. Indeed, it has been shown6 that dendritic cells are the essential APC involved in immune responses elicited by intramuscularly given DNA vaccines. 3. DNA Vaccination via Liposomes Vaccination with naked DNA by the intramuscular route is dependent on the ability of myocytes to take up the plasmid. However, some of the DNA may also be engulfed by APC infiltrating the site of injection, or in the lymph nodes following migration of the DNA to the lymphatics. The extent of DNA degradation by extracellular deoxyribonucleases is unknown, but depending on the time of its residence interstitially, degradation could be considerable. Therefore, approaches that protect DNA from the extracellular nucleases and promote DNA uptake by cells more efficiently, or target it to APC, should contribute to the optimal design of DNA vaccines. It has been suggested19 that as APC are a preferred alternative to muscle cells for DNA vaccine uptake and expression, liposomes (known3 to be taken up avidly by APC infiltrating the site of injection or in the lymphatics, an event that has been implicated3 in their immunoadjuvant activity) would be a suitable means of delivery of entrapped DNA to such cells. Liposomes would also protect20 their DNA content from deoxyrubonuclease attack. Moreover, the structural versatility21 of the system would ensure that its tranfection efficiency is further improved 46 Cregoriadis ct al. by the judicial choice of its structural characteristics or by the co-entrapment of cytokine genes, other adjuvants (e.g. immunostimulatory sequences), or indeed protein antigens (see later) together with the plasmid vaccine. As a number of injectable liposome-based drug formulations, including vaccines against hepatitis A and influenza, have been already licensed for clinical use,21 acceptance of the system clinically would be less problematic than with other systems that are still at an experimental stage. 3.1. Procedure for the entrapment of plasmid DNA into liposomes A variety8,22,23 of plasmid DNAs have been quantitatively entrapped into liposomes by a mild dehydration-rehydration procedure.20'22,23 The procedure (Fig. 1) consists of mixing preformed small unilamellar vesicles (SUV) with a solution of the DNA destined for entrapment, freeze-drying of the mixture, followed by controlled rehydration of the formed powder, and centrifugation to remove nonentrapped material. Formed liposomes are multilamellar.20 However, when an appropriate amount of sucrose is added to the SUV and DNA mixture prior to dehydration,24 the resulting liposomes are much smaller (about 100-160 nm in diameter). As expected, DNA incorporation values8'23-26 were higher (up to 90% of the amount used) when a cationic lipid was present in the bilayers. No apparent relationship was observed between amount of DNA used (10-500//g) and the values of incorporation for the compositions and lipid mass used.8,23,26 The possibility that DNA was not entrapped within the bilayers of cationic liposomes, but was rather complexed with their surface (as suggested by the high Fig. 1. Entrapment of DNA and/or protein into cationic liposomes. The procedure entails mixing up empty SUV with the solute(s) destined for entrapment and subsequent dehydration. On rehydration, most of the solute(s) is recovered entrapped within the generated multilamellar liposomes. Genetic Vaccines: A Role for Liposomes 47 Naked DNA IMBXM™ (DNA) § m "Complexed** DNA Naked DNA taraXeB™ (DNA) g "Complexed" DNA 4* W o Fig. 2. Gel electrophoresis of a mixture of cationic SUV and pRc/CMV HBS before (complexed DNA) and after (entrapped DNA) dehydration-rehydration of the mixture. "incorporation" values obtained on mixing)20 was examined by treating liposomeentrapped and liposome-complexed DNA with deoxyribonuclease. Substantially, more liposome-entrapped DNA remained intact than when it was complexed,20 presumably because of the inability of the enzyme to reach its substrate in the former case. The significant resistance of complexed DNA (despite its accessibility) to the enzyme could be attributed to its condensed state.25 Additional evidence that the DNA was entrapped within liposomes was obtained by gel electrophoresis of a mixture of cationic SUV and plasmid DNA before (complexed DNA) and after dehydration-rehydration of the mixture (entrapped DNA). When the anionic sodium dodecylsulphate (SDS) was incorporated in the gel, complexed DNA was dissociated from the SUV, presumably because of ionic competition for the cationic charges. As expected, "entrapped" DNAretained its association with the liposomes, suggesting its unavailability to the competing SDS anions26 (Fig. 2). 3.2. DNA immunization studies Previously,20 liposome-entrapped plasmid found to transfect cells in vitro regardless of the vesicle surface charge was tested in immunization experiments,19,27 using a plasmid (pRc/CMV HBS) encoding the S region of the hepatitis B surface antigen (HBsAg; subtype ayw). Mice (Balb/c) that are repeatedly injected intramuscularly with 5 or 10/ig plasmid entrapped in cationic liposomes, exhibited at all times much greater (up to 100-fold) antibody (IgGi) responses (Fig. 3) against the 48 Gregoriadis et al. © I O | E c 5 a. I a. 8 I D JBfcei_ 26 34 44 Days after first injection Fig. 3. Immune responses in mice injected with naked, or liposome-entrapped pRc/CMV HBS. Balb/c mice were injected intramuscularly on days 0, 10, 20, 27 and 37 with 5 /xg of DNA entrapped in cationic liposomes composed of PC, DOPE and DOTAP (A), DC-Chol (B) or SA (C) (molar ratios 1:0.5:0.25), or in the naked form (D). Animals were bled 7, 15, 26, 34 and 44 days after the first injection and sera tested by ELISA for IgGT (black bars), IgG2a (white bars) or IgG2b (grey bars) responses against the encoded hepatitis B surface antigen (HBsAg; S region, ayw subtype). Values are means ±SD of log10 of reciprocal end point serum dilutions required for OD to reach readings of about 0.2. Sera from untreated mice gave log10 values of less than 2.0. IgGj responses were mounted by all mice injected with liposomal DNA but became measurable only at 26 days. Differences in log10 values (all IgG subclasses at all time intervals) in mice immunized with liposomal DNA and mice immunized with naked DNA were statistically significant (P < 0.0001-0.002). (Reproduced with permission from Ref. 19.) Genetic Vaccines: A Role for Liposomes 49 encoded antigen than animals immunized with the naked plasmid. Values of other subclasses (IgG2a and IgG2b) were also greater (up to 10-fold) (Fig. 3). Moreover, IgGj responses for the liposome-entrapped plasmid DNA were higher (up to 10- fold) than those obtained with DNA complexed with similar cationic liposomes.19 This was also true for IFN-y and IL-4 levels in the spleens of immunized mice.19 In other experiments,8 the effect of the route of injection of the pRc/CMV HBS plasmid was examined with respect to both humoural and cell-mediated immunity, using Balb/c mice and an outbred mouse strain (T.O.). Results8 comparing responses for liposome-entrapped and naked plasmid DNA showed greater antibody (IgGi) responses for the entrapped DNA, not only by the intramuscular route, but also the subcutaneous and the intravenous routes. As there were no significant differences in the titers between the two strains,8 it was concluded that immunization with liposomal pRc/CMV HBS is not MHC restricted. Results obtained on the testing of IFN-y and IL-4 in the spleens (not shown) exhibited a similar pattern. Involvement of muscle cells in the mechanism by which liposomes promote greater immune responses to the encoded antigen than seen with the naked plasmid, is rather unlikely. Although, cationic liposomes could in theory bind to and be taken up by the negatively charged myocytes, the negatively charged proteins present in the interstitial fluid would neutralize21 the cationic liposomal surface and thus interfere with such binding. In addition, vesicle size (about 600-700 nm average diameter; Ref. 26) would render access to the cells difficult, if not impossible. It is therefore more likely that cationic liposomes are endocytosed by APC, including dendritic cells, in the lymphatics where liposomes are expected to end up.28 Uptake of liposomal plasmid DNA is supported in studies where mice were injected intramuscularly or subcutaneously with liposomes entrapping the plasmid (pCMV- EFGP), encoding the enhanced fluorescent green protein or with the naked plasmid. Fluorescence microscopy of sections of the lymph nodes draining the injected site revealed (Fig. 4) much more green fluorescence when the plasmid was administered in the liposomal form.27 It appears8'19 that the key ingredient of the DNA-containing liposomes as used in Fig. 3, contributing to enhanced immune responses, is the cationic lipid. The mechanism by which liposomal DNAreaches the nucleus for episomal transfection is poorly understood. It is conceivable, however, that some of the endocytosed liposomal DNA escapes the endocytic vacuoles prior to their fusion with lysosomes (in a way similar to that proposed29 for vesicle-DNA complexes) to enter the cytosol for eventual episomal transfection and presentation of the encoded antigen. It is perhaps at this stage of intracellular trafficking of DNA, spanning its putative escape from endosomes and access to the nucleus, that the cationic lipid, possibly together within the fusogenic phosphatidylethanolamine (PE) component, plays a significant role. 50 Gregoriadis et al. Kttmam w»* ii'm<|»* IWt •ff!f b ((lit tyfent Fig. 4. Fluorescence images of muscle and lymph node sections from mice injected intramuscularly with 10/xg liposome-entrapped or naked pCMVEGFP and killed 48h later. Sections from untreated animals were used as controls. (Reproduced with permission from Ref. 27.) 3.3. Induction of a cytotoxic T lymphocyte (CTL) response by liposome-entrapped plasmid DNA Immunization studies with liposome-entrapped DNA vaccines were expanded30 to include the cytotoxic T lymphocyte (CTL) component of the immune response. This was measured by the specific killing of syngeneic target cells pulsed with a recognized CTL epitope peptide derived from the antigen tested. To that end, the type and degree of immune response induced following subcutaneous injection of DNA in cationic liposomes was monitored and compared with that obtained with DNA alone injected by the same route. 6-8 week old, female C57/BL6 (H-2d) mice were injected subcutaneously with one or two doses of 2.5 or 10 ^g ovalbumin (OVA)-encoding plasmid DNA (pCI-OVA), either alone or entrapped in liposomes. Animals immunized subcutaneously with 100 /xg of OVA protein complexed with 1 /xg of cholera toxin (CT) served as a positive control. Blood samples and spleens were collected from all animals one week after the last injection and tested for anti-OVA total IgG (serum), CTL activity and cytokine release (spleen). After a single dose of antigen, only animals immunized with either protein or 10/xg of liposomal DNA showed significant anti-OVA antibody titres by ELBA. After two doses of antigen, only animals immunized with either protein or liposomal DNA (both 2.5 and 10 ttg DNA) showed significant levels of seroconversion and serum antibody titres against OVA by ELBA.30 Similarly, no anti-OVA CTL activity was detected in animals immunized with DNA alone. However, animals immunized with two doses of 10 /xg of liposomal DNA displayed a CTL response higher (60% cell killing vs 50%) than that obtained in the positive control group immunized Genetic Vaccines: A Role for Liposomes 51 with OVA protein and adjuvant (CT).30 Thus, delivery of a small dose of liposomal plasmid DNA subcutaneously, a route of immunization not normally inducing significant plasmid DNA mediated immune activation,9 results in a strong antigenspecific cellular response which is greater than that achieved by higher doses of a conventional protein antigen together with a powerful adjuvant (CT). 4. The Co-delivery Concept Proteins that are synthesized within a cell (e.g. from plasmid DNA having a mammalian-active promoter) are continuously sampled as peptides by the proteosome / class-I MHC antigen presenting pathway. Conversely, proteins that are acquired exogenously by antigen-presenting cells are sampled in an analogous way by the endosomal/MHC-class-II pathway. It follows that the delivery of both protein and plasmid-DNA-encoded forms of a protein antigen to the same individual antigen-presenting cell would result in the simultaneous presentation of the antigen via both class-I and class-II pathways, thereby providing an opportunity for synergy in the resulting immune response to the antigen. Several appropriate liposomal formulations were designed to test the "co-delivery" hypothesis, exploiting the advantages of the dehydration-rehydration liposome technology that entraps both DNA and protein immunogens efficiently. The formulations, described in Table 1, comprise various test and control permutations of plasmid DNA and protein, either free or entrapped (together or separately) in the liposomal vehicle. Immunization with DNA encoding the influenza haemagglutinin protein has been explored previously with naked31 or liposomally formulated DNA.32 Although immune responses elicited by DNA alone were adequate to achieve protective efficacy against influenza virus challenge in preclinical studies, only weak anti-HA antibody responses were elicited.31 The present "co-delivery" concept was designed to rectify this deficiency of DNA-based influenza vaccines. In a series of experiments, plasmid DNA encoding the haemagglutinin (HA) antigen [referred to in Table 1 and Fig. 5 as DNA(ha)] of the influenza virus (A/Sichuan/87 or A/PR/8 strains) was co-entrapped with the corresponding whole inactivated virus (referred to as HA) within the same liposomes using the dehydration-rehydration method (for details on lipid composition and method see Refs. 26 and 27). A variety of control preparations including liposomes co-entrapping irrelevant DNA (i.e. plasmid DNA encoding ovalbumin) with HA or irrelevant protein (i.e. ova) with DNA (ha), entrapping DNA(ha) or HA alone, a mixture of the latter two preparations, and a mixture of the naked DNA(ha) and HA were used to immunize mice. Results shown in Fig. 5 demonstrate that the "co-delivery" hypothesis formulation (comprising both HA and its corresponding DNA in the same liposomes), elicited a greater response than all other formulations at each time point in the series, and it 52 Gregoriadis et al. Table 1 Liposomal formulations of DNA and protein used in immunization experiments. Sample 1.1 2.1 3.1 4.1 5.1 6.1 7 8 9 10 11 12 Dose (/ig/animal (0.2 ml S/O) Formulation Liposomes (co-delivery) Liposomes (co-delivery) Liposomes (co-delivery) Liposomes Liposomes Liposomes (samples 4.1 & 5.1) DNA and protein (mixed) DNA and protein (mixed) DNA and protein (mixed) DNA alone Protein alone Control (PBS) DNA ha (10) ova (11) ha (10) ha (10) Nil ha (10) ha (10) ova (11) ha (10) ha (10) Nil Nil Protein HA (0.6) HA (0.6) OVA (0.76) Nil HA (0.6) HA (0.6) OVA (0.76) HA (0.6) HA (0.6) Nil HA (10) Nil Plasmid DNA encoding the HA antigen [DNA(ha)] and the HA antigen (HA) were entrapped in liposomes either together (co-entrapped; sample 1.1) or separately in different formulations (sample 6.1) mixed before injection. In some formulations, DNA(ha) and HA were entrapped alone (samples 4.1 and 5.1 respectively). In others, ovalbumin (OVA) and plasmid DNA encoding ha fDNA(ha)] (sample 7) or HA and plasmid DNA encoding OVA (sample 8) were entrapped separately and then mixed. Mice were injected subcutaneously on days 0 and 28 and blood samples analyzed by ELISA for Ig responses. 1OD0O - 1000 100- A/Sichuan/87 Ig response -p=o.oca ;r***OM burton) DNA (10 (ig) / Protein (0.6 ng) - • - Up(DNA(HA)/HA) - * - Lip(ONA{OVA)/HA) - * - Up(DNA(HA)/OVA) - • - Up (DNA (HA)/no protein) - • - Up(noDNA/HA) - * - Up(DNA(H;)) + Up(HA) •••••• DNA {HA ) • OVA • DNA (OVA)* HA • DNA(HA)+HA - * - DNA ( n » ) no protein - * - HP (protein alone) • control (negative J 20 A 3 0 boost 40 50 Day post 1st dose Fig. 5. Serum Ig endpoint titres in Balb/c mice immunized on days 0 and 28 with DNA and/or antigen formulations as described in Table 1 a nd bled at time intervals. Genetic Vaccines: A Role for Liposomes 53 is by far the strongest response after a single dose. Notably, the formulation "Lip (OVA/ha)", which is a control for the CpG adjuvant effect of plasmid DNA,33 gave a response which was much lower than that of "co-delivery" with the appropriate homologous pair of HA DNA and protein. Likewise, Lip (HA/ova) (an inappropriate pairing according to the hypothesis), gave a markedly weaker response. Figure 5 also demonstrates that separately entrapped HA DNA and protein (in neighbouring vesicles) gave rise to an inferior response, supporting the hypothesis that delivery of both payloads to the same cell (which is best achieved by co-entrapment in the same liposome) is important in achieving the optimal antibody response. It is also remarkable that, inspite the modest DNA dose (10 /xg) and small number (2) of immunizations used, several formulations completely failed to generate an anti-HA response. These included HA DNA alone, and liposomally entrapped HA DNA. These findings serve to emphasize the striking degree of superiority of "codelivery" over previous methods of DNA-based immunization against influenza virus. In conclusion, the present studies demonstrate that very small doses of protein as an additive in DNA immunization can dramatically improve the antibody response to the target protein, provided that the protein and DNA are homologous to one-another (i.e. that the DNA can express the protein), and that the payloads are delivered in the same individual liposomal vehicle. The simplest hypothesis to explain our observation is that the synergy observed between the appropriately delivered "homologous pair" of protein and DNA involves delivery of both payloads to the same antigen-presenting cell. The application of the co-delievery concept to alternative delivery systems, e.g. niosomes, dendimers, PLA/PLGA, chitosans, alginates and other microparticles awaits investigation. It is anticipated that the "co-delivery" approach will lead to better DNA-based vaccines for prophylactic and therapeutic use, particularly where vaccines require the elicitation of antibody responses (e.g. influenza vaccines). References 1. Powel MF and Newman MJ (eds.) (1995) Vaccine Design: The Subunit and Adjuvant Approach. Plenum Press: New York. 2. Gregoriadis G, McCormack B, Allison AC and Poste G (eds.) (1993) New Generation Vaccines: The Role of Basic Immunology. Plenum Press: New York. 3. Gregoriadis G (1990) Immunological adjuvants: A role for liposomes. Immunol Today. 11:89-97. 4. Gluck R, Mischler R, Brantschen S, Just M, Althans B and Cryz SJ, Jr (1992) Immunopotentiating reconstituted influenza virome vaccine delivery system for immunization against hepatitis A. / Clin Invest 90:2491-2495. 54 Gregoriadis et al. 5. Davis HL, Whalen RG and Demeneix BA (1993) Direct gene transfer in skeletal muscle in vivo: Factors influencing efficiency of transfer and stability of expression. Hum Gene Ther 4:151-156. 6. Manickan E, Karem KL and Rouse BT (1997) DNA vaccines — A modern gimmick or a boon to vaccinology? Crit Rev Immunol 17:139-154. 7. Chattergoon M, Boyer J and Weiner DB (1997) Genetic immunization: A new era in vaccines and immune therapeutics. FASEB 11:754-763. 8. Gregoriadis G (1998) Genetic vaccines: Strategies for optimization. Pharm Res 15:661-670. 9. Lewis PJ and Babiuk LA (1999) DNA vaccines: A review. Adv Virus Res 54:129-188. 10. Tang DC, Devit M and Johnston SA (1992) Genetic immunization is a simple method for eliciting an immune response. Nature 356:152-154. 11. Ulmer JB, Donnelly J, Parker SE, et al. (1993) Heterologous protection against influenza by injection of DNA encoding a viral protein. Science 259:1745-1749. 12. Fynan EF> Webster RG, Fuller DH and Haynes JR (1993) DNA vaccines: Protective immunizations by parenteral, mucosal and gene-gun inoculations. Proc Natl Acad Sci USA 90:11478-11482. 13. Wang B, Ugen K, Srikantan V, et al. (1993) Gene inoculation generates immune responses against HIV-I. Proc Natl Acad Sci USA 90:4156^160. 14. Davis HL, Michel ML, Mancini M, Schleef M and Whalen RG (1994) Direct gene transfer in skeletal muscle: Plasmid DNA based immunization against the hepatitis B virus surface antigen. Vaccine 12:1503-1509. 15. Conry R, LoBuglio A, Loechel F, et al. (1995) A carcinoembryonic antigen polynucleotide vaccine for human clinical use. Cancer Gene Ther 2:33-38. 16. Bright RK, Beames B, Shearer MH and Kennedy RC (1996) Protection against lethal tumor challenge with SV40-transformed cells by the direct injection of DNA encoding SV-40 large tumor antigen. Cancer Res 56:1126-1130. 17. Matzinger P (1994) Tolerance, danger and the extended family. Annu Rev Immunol 12: 991-1045. 18. Spier E (1996) Meeting Report: International meeting on the nucleic acid vaccines for the prevention of infectious disease and regulatory nuclear acid (DNA) vaccines. Vaccine 14:1285-1288. 19. Gregoriadis G, Saffie R and de Souza B (1997) Liposome-mediated DNA vaccination. FEES Lett 402:107-110. 20. Gregoriadis G, Saffie R and Hart SL (1996) High yield incorporation of plasmid DNA within liposomes: Effect on DNA integrity and transfection efficiency. / Drug Targ 3: 469-475. 21. Gregoriadis G (1995) Engineering targeted liposomes: Progress and problems. Trends Biotechnol 13:527-537. 22. Gregoriadis G, McCormack B, Obrenovic M and Perrie Y (1999) Entrapment of plasmid DNA vaccines into liposomes by dehydration/rehydration, in Lowrie DB and Whalen R. (eds.) Methods in Molecular Medicine, DNA Vaccines: Methods and Protocols. Humana Press Inc.: Totowa, NJ. pp. 305-312. Genetic Vaccines: A Role for Liposomes 55 23. Gregoriadis G, McCormack B, Obrenovic M, Saffie R, Zadi B and Perrie Y (1999) Liposomes as immunological adjuvants and vaccine carriers. Methods 19:156-162. 24. Zadi B and Gregoriadis G (2000) A novel method for high-yield entrapment of solutes into small liposomes. J Lipos Res 10:73-80. 25. Feigner PL and Rhodes G (1991) Gene therapeutics. Nature 349:351-352. 26. Perrie Y and Gregoriadis G (2000) Liposome-entrapped plasmid DNA: Characterization studies. Biochim Biphys Acta 1475:125-132. 27. Perrie Y and Gregoriadis G (2001) Liposome mediated DNA vaccination: The effect of vesicle composition. Vaccine 19:3301-3310. 28. Velinova M, Read N, Kirby C and Gregoriadis G (1996) Morphological observations on the fate of liposomes in the regional lymphs nodes after footpad injection into rats. Biochim Biophys Acta 1299:207-215. 29. Szoka FC, Xu Y and Zelpati O (1996) How are nucleic acids released in cells from cationic lipid-nucleic acid-complexes? / Lipos Res 6:567-587. 30. Bacon A, Caparros-Wanderley W, Zadi B and Gregoriadis G (2002) Induction of a cytotoxic T lymphocyte (CTL) response to plasmid DNA delivered by Lipodine™. / Lipos Res 12:173-183. 31. Johnson PA, Conwey MA, Daly J, Nicolson C, Robertson J and Mills KH (2000) Plasmid DNA encoding influenza virus haemagglutinin induces Th 1 cells and protection against respiratory infection despite its limited ability to generate antibody responses. / Gen Virol 81:1737-1745. 32. Sha Z, Vincent MJ and Compans RW (1999) (Title) Lmmunobiology 200:21-30. 33. Gursel M, Tunca S, Ozkan M, Ozcengiz G and Alaeddinoglu G (1999) Immunoadjuvant action of plasmid DNA in liposomes. Vaccine 17:1376-1383. This page is intentionally left blank 5 Polymer Micelles as Drug Carriers Elena V. Batrakova, Tatiana K. Bronich, Joseph A. Vetro and Alexander V. Kabanov 1. Introduction It has long been recognized that improving one or more of the intrinsic adsorption, distribution, metabolism, and excretion (ADME) properties of a drug is a critical step in developing more effective drug therapies. As early as 1906, Paul Ehrlich proposed altering drug distribution by conjugating toxic drugs to "magic bullets" (antibodies) having high affinity for cancer cell-specific antigens, in order to both improve the therapeutic efficacy of cancer while decreasing its toxicity.1 Since then, it has become clear that directly improving intrinsic ADME through modifications of the drug is limited or precluded by structural requirements for activity. In other words, low molecular mass drugs are too small and have only limited number of atomic groups that can be altered to improve ADME, which often adversely affects drug pharmacological activity. In turn, the modifications of many low molecular mass drugs, aimed to increase their pharmacological activity, often adversely affect their ADME properties. For example, the potency and specificity of drugs can be improved by the addition of hydrophobic groups.2 The associated decrease in water solubility, however, increases the likelihood of drug aggregation, leading to poor absorption and bioavailability during oral administration2 or lowered systemic bioavailability, high local toxicity, and possible pulmonary embolism during intravenous administration.3 Although there have been considerable difficulties for improving some existing drugs through chemical modifications, the problem became even more obvious 57 58 Batrakova et al. with the development of high-throughput drug discovery technologies. Almost half of lead drug candidates identified by high-throughput screening have poor solubility in water, and are abandoned before the formulation development stage.4 In addition, newly synthesized drug candidates often fail due to poor bioavailability, metabolism and/or undesirable side effects, which together decrease the therapeutic index of the molecules. Furthermore, a new generation of biopharmaceuticals and gene therapy agents are emerging based on novel biomacromolecules, such as DNA and proteins. The use of these biotechnology-derived drugs is completely dependent on efficient delivery to the critical site of the action in the body. Therefore, drug delivery research is essential in the translation of newly discovered molecules into potent drug candidates and can significantly improve therapies of existing drugs. Polymer-based drugs and drug delivery systems emerged from the laboratory bench in the 1990s as a promising therapeutic strategy for the treatment of certain devastating human diseases.5'6 A number of polymer therapeutics are presently on the market or undergoing clinical evaluation to treat cancer and other diseases. Most of them are low molecular weight drug molecules or therapeutic proteins that are chemically linked to water-soluble polymers to increase drug solubility, drug stability, or enable targeting to tumors. Recently, as a result of rapid development of novel nanotechnology-derived materials, a new generation of polymer therapeutics has emerged, using materials and devices of nanoscale size for the delivery of drugs, genes, and imaging molecules.7-12 These materials include polymer micelles, polymer-DNA complexes ("polyplexes"), liposomes, and other nanostructured materials for medical use that are collectively known as nanomedicines. Compared with first generation polymer therapeutics, the new generation nanomedicines are more advanced. They entrap small drugs or biopharmaceutical agents such as therapeutic proteins and DNA, and can be designed to trigger the release of these agents at the target site. Many nanomedicines are constructed using self-assembly principles such as the spontaneous formation of micelles or interpolyelectrolyte complexes, driven by diverse molecular interactions (hydrophobic, electrostatic, etc.). This chapter considers polymeric micelles as an important example of the new generation of nanomedicines, which is also perhaps among the most advanced approach toward clinical applications in diagnostics and the treatment of human diseases. 2. Polymer Micelle Structures 2.1. Self-assembled micelles Self-assembled polymer micelles are created from amphiphilic polymers that spontaneously form nanosized aggregates when the individual polymer chains Polymer Micelles as Drug Carriers 59 Single polymer chains Polymeric micelle ("Unimers") Fig. 1. Self-assembly of block copolymer micelles. ("unimers") are directly dissolved in aqueous solution (dissolution method)13 above a threshold concentration (critical micelle concentration or CMC) and solution temperature (critical micelle temperature or CMT) (Fig. 1). Amphiphilic polymers with very low water solubility can alternatively be dissolved in a volatile organic solvent, then dialyzed against an aqueous buffer (dialysis method).14 Amphiphilic di-block (hydrophilic-hydrophobic) or tri-block (hydrophilichydrophobic- hydrophilic) copolymers are most commonly used to prepare selfassembled polymer micelles for drug delivery,9'15,16 although the use of graft copolymers has been reported.17-19 For drug delivery purposes, the individual unimers are designed to be biodegradable20,21 and/or have a low enough molecular mass (< ~40 kDa) to be eliminated by renal clearance, in order to avoid polymer buildup within the body that can potentially lead to toxicity.22 The most developed amphiphilic block copolymers assemble into spherical core-shell micelles approximately 10 to 80 nm in diameter,23 consisting of a hydrophobic core for drug loading and a hydrophilic shell that acts as a physical ("steric") barrier to both micelle aggregation in solution, and to protein binding and opsonization during systemic administration (Fig. 2). The most common hydrophilic block used to form the hydrophilic shell is the FDA-approved excipient poly(ethylene glycol) (PEG) or poly(ethylene oxide) (PEO).24 PEG or PEO consists of the same repeating monomer subunit CH2-CH2-O, and may have different terminal end groups, depending on the synthesis procedure, e.g. hydroxyl group HO-(CH2-CH2-0)n-H; methoxy group CH30-(CH2-CH2-0)n-H, etc. PEG/PEO blocks typically range from 1 to 15 kDa.16,24 In addition to its FDA approval, PEG is extremely soluble and has a large excluded volume. This makes it especially suitable for physically interfering with intra-micelle interactions and subsequent micelle aggregation. PEG also blocks protein and cell surface interactions, which greatly decreases nanoparticle uptake by the reticuloendothelial system (RES), and consequently increases the plasma 60 Batrakova etal. Self-Assembled No self assembly Homopolymer A n n n Di-block copolymer Tri-block copolymer Graft copolymer i n n n***** + ^ ^ , "w>>2^i?^s /^' Charged copolymer / ? S Covalentlv-Assembled (unimolecular micelles) Star Dendritic hydrophilic block hydrophobic block cation ic block anionic block annn ^ ^ H +++++ i i i i . Fig. 2. Polymer micelle structures. half life of the polymer micelle.25 The degree of steric protection by the hydrophilic shell is a function of both the density and length of the hydrophilic PEG blocks.25 Unlike the hydrophilic block, which is typically PEG or PEO, different types of hydrophobic blocks have been sufficiently developed as hydrophobic drug loading cores.16 Examples of diblock copolymers include (a) poly(L-amino acids), (b) biodegradable poly(esters), which includes poly(glycolic acid), poly(D lactic acid), poly(D,L-lactic acid), copolymers of lactide/glycolide, and poly(ecaprolactone), (c) phospholipids/long chain fatty acids26; and for tri-block copolymers, (d) polypropylene oxide (in Pluronics/poloxamers).9 The choice of hydrophobic block is largely dictated by drug compatibility with the hydrophobic core (when drug is physically loaded, as described later) and the kinetic stability of the micelle. The self-assembly of amphiphilic copolymers is a thermodynamic and, consequently, a reversible process that is entropically driven by the release of ordered water from hydrophobic blocks; it is either stabilized or destabilized by solvent interactions with the hydrophilic shell. As such, the structural potential of amphiphilic copolymer unimers to form micelles is determined by the mass ratio of hydrophilic to hydrophobic blocks, which also affects the subsequent morphology if aggregates are formed.14 If the mass of the hydrophilic block is too great, the copolymers exist in aqueous solution as unimers, whereas, if the mass of the hydrophobic block is too great, unimer aggregates with non-micellar morphology are formed.27 If the mass of the hydrophilic block is similar or slightly greater than the hydrophobic block, then conventional core shell micelles are formed. An important consideration for drug delivery is the relative thermodynamic (potential for disassembly) and kinetic (rate of disassembly) stability of the polymer Polymer Micelles as Drug Carriers 61 micelle complexes, after intravenous injection and subsequent extreme dilution in the vascular compartment.28 This is because the polymer micelles must be stable enough to avoid burst release of the drug cargo, as in the case of a physically loaded drug, upon systemic administration and remain as nanoparticles long enough to accumulate in sufficient concentrations at the target site. The relative thermodynamic stability of polymer micelles (which is inversely related to the CMC) is primarily controled by the length of the hydrophobic block.13 An increase in the length of the hydrophobic block alone significantly decreases the CMC of the unimer construct (i.e. increases the thermodynamic stability of the polymer micelle), whereas an increase in the hydrophilic block alone slightly increases the CMC (i.e. decrease the thermodynamic stability of a polymer micelle).14 Although the CMC indicates the unimer concentration below which polymer micelles will begin to disassemble, the kinetic stability determines the rate at which polymer micelle disassembly occurs. Many diblock copolymer micelles possess good kinetic stability and only slowly dissociate into unimers after extreme dilution.29 Thus, although polymer micelles are diluted well below typical unimer CMCs29 (10~6-10-7M) after intravenous injection, their relative kinetic stability might still be suitable for drug delivery. The kinetic stability depends on several factors, including the size of a hydrophobic block, the mass ratio of hydrophilic to hydrophobic blocks, and the physical state of the micelle core.14 The incorporation of hydrophobic drugs may also further enhance micelle stability. 2.2. Unimolecular micelles Unimolecular micelles are topologically similar to self-assembled micelles, but consist of single polymer molecules with covalently linked amphiphile chains. For example, copolymers with star-like or dendritic architecture, depending on their structure and composition, can either aggregate into multimolecular micelles,30-32 or exist as unimolecular micelles.33 Dendrimers are widely used as building blocks to prepare unimolecular micelles, because they are highly-branched, have welldefined globular shape and controled surface functionality.34-40 For example, unimolecular micelles were prepared by coupling dendritic hypercores of different generations with PEO chains.40'41 The dendritic cores can entrap various drug molecules. However, due to the structural limitations involved in the synthesis of dendrimers of higher generation, and relatively compact structure of the dendrimers, the loading capacity of such micelles is limited. Thus, to increase the loading capacity, the dendrimer core can be modified with hydrophobic block, followed by the attachment of the PEO chains. For example, Wang et al. recently synthesized an amphiphilic 16-arm star polymer with a polyamidoamine dendrimer core and arms composed of inner lipophilic poly(e-caprolactone) block and outer PEO 62 Batrakova et al. block.42 These unimolecular micelles were shown to encapsulate a hydrophobic drug, etoposide, with high loading capacity. Multiarm star-like block copolymers represent another type of unimolecular micelles.42-46 Star polymers are generally synthesized by either the arm-first or core-first methods. In the arm-first method, monofunctional living linear macromolecules are synthesized and then cross-linked either through propagation, using a bifunctional comonomer,47 or by adding a multifunctional terminating agent to connect precise number of arms to one center.45 Conversely, in the core-first method, polymer chains are grown from a multifunctional initiator.43'44'46'48 One of the first reported examples of unimolecular micelles, suitable for drug delivery, was a three-arm star polymer, composed of mucic acid substituted with fatty acids as a lipophilic inner block and PEO as a hydrophilic outer block.44 These polymers were directly dispersible in aqueous solutions and formed unimolecular micelles. The size and solubilizing capacity of the micelles were varied by changing the ratio of the hydrophilic and lipophilic moieties. In addition, star-copolymers with polyelectrolyte arms can be prepared to develop pH-sensitive unimolecular micelles as drug carriers.46 2.3. Cross-linked micelles The multimolecular micelles structure can be reinforced by the formation of crosslinks between the polymer chains. These resulting cross-linked micelles are, in essence, single molecules of nanoscale size that are stabile upon dilution, shear forces and environmental variations (e.g. changes in pH, ionic strength, solvents etc.). There are several reports on the stabilization of the polymer micelles by crosslinking either within the core domain49-53 or throughout the shell layer.54-56 In these cases, the cross-linked micelles maintained small size and core-shell morphology, while their dissociation was permanently suppressed. Stable nanospheres from the PEO-b-polylactide micelles were prepared by using polymerizable group at the core segment.49 In addition to stabilization, the core polymerized micelles readily solubilized rather large molecules such as paclitaxel, and retained high loading capacity even upon dilution.50 Formation of interpenetrating network of a temperature-sensitive polymer (poly-N-isopropylacrilomide) inside the core was also employed for the stabilization of the Pluronic micelles.53 The resulting micelle structures were stable against dilution, exhibited temperature-responsive swelling behavior, and showed higher drug loading capacity than regular Pluronic micelles. Recently, a novel type of polymer micelles with cross-linked ionic cores was prepared by using block ionomer complexes as templates.57 The nanofabrication of these micelles involved condensation of PEO-b-poly(sodium methacrylate) diblock Polymer Micelles as Drug Carriers 63 copolymers by divalent metal cations into spherical micelles of core-shell morphology. The core of the micelle was further chemically cross-linked and cations removed by dialysis. Resulting micelles represent hydrophilic nanospheres of coreshell morphology. The core comprises a network of the cross-linked polyanions and can encapsulate oppositely charged therapeutic and diagnostic agents, while a hydrophilic PEO shell provides for increased solubility. Furthermore, these micelles displayed the pH- and ionic strength-responsive hydrogel-like behavior, due to the effect of the cross-linked ionic core. Such behavior is instrumental for the design of drug carriers with controled loading and release characteristics. 3. Drug Loading and Release In general, there are three major methods for loading drugs into polymer micelle cores: (1) chemical conjugation, (2) physical entrapment or solubilization, and (3) polyionic complexation (e.g. ionic binding). 3.1. Chemical conjuga tion Drug incorporation into polymer micelles via chemical conjugation was first proposed by Ringsdorf's group58 in 1984. According to this approach, a drug is chemically conjugated to the core-forming block of the copolymer via a carefully designed pH- or enzyme-sensitive linker, that can be cleaved to release a drug in its active form within a cell.59,60 The polymer-drug conjugate then acts as a polymer prodrug which self assembles into a core-shell structure. The appropriate choice of conjugating bond depends on specific applications. The nature of the polymer-drug linkage and the stability of the drug conjugate linkage can be controled to influence the rate of drug release, and therefore, the effectiveness of the prodrug.61-63 For instance, recent work by Kataoka's group proposed pH-sensitive polymer micelles of PEO-b-poly(aspartate hydrazone doxorubicin), in which doxorubicin was conjugated to the hydrophobic segments through acid-sensitive hydrazone linkers that are stable at extracellular pH 7.4, but degrade and release the free drug at acidic pH 5.0 to 6.0 in endosomes and lysosomes.63,64 The original approach developed by this group used doxorubicin conjugated to the poly(aspartic acid) chain of PEO-b-poly(aspartic acid) block copolymer through an amide bond.65 Adjusting both the composition of the block copolymer and the concentration of the conjugated doxorubicin, led to improved efficacy, as evidenced by a complete elimination of solid tumors implanted in mice.66 It was later determined that doxorubicin physically encapsulated within the micellar core was responsible for antitumor activity. This finding led to the use of PEO-b-poly(aspartate doxorubicin) conjugates as nanocontainers for physically entrapped doxorubicin.67 64 Batrakova et al. 3.2. Physical entrapment The physical incorporation or solublization of drugs within block copolymer micelles is generally preferred over micelle-forming polymer-drug conjugates, especially for hydrophobic drug molecules. Indeed, many polymers and drug molecules do not contain reactive functional groups for chemical conjugation, and therefore, specific block copolymers have to be designed for a given type of drug. In contrast, a variety of drugs can be physically incorporated into the core of the micelles, by engineering the structure of the core-forming segment. In addition, molecular characteristics (i.e. molecular weight, composition, presence of functional groups for active targeting) within a homologous copolymer series can be designed to optimize the performance of a drug for a given drug delivery situation.9,14 This concept was introduced by our group in the late 1980s and was initially termed "micellar microcontainer",68 but is now widely known as a "micellar nanocontainer".9,10 Haloperidol was encapsulated in Pluronic block copolymer micelles,68 the micelles were targeted to the brain using brain-specific antibodies or insulin, and enhancement of neuroleptic activity by the solubilized drug was observed. During the last 25 years, a large variety of amphiphilic block copolymers have been explored as nanocontainers for various drugs. Different loading methods can be used for physical entrapment of the drug into the micelles, including but not limited to dialysis,69-72 oil in water emulsification,69 direct dissolution,42,73,74 or solvent evaporation techniques.75,76 Depending on the method, drug solubilization may occur during or after micelle assembly. The loading capacity of the polymer micelles, which is frequently expressed in terms of the micelle-water partition coefficient, is influenced by several factors, including both the structure of core-forming block and a drug, molecular characteristics of the copolymer such as composition, molecular weight, and the solution temperature.13 Many studies indicate that the most important factor related to the drug solubilization capacity of a polymer micelle is the compatibility between the drug and the core-forming block.9,14,77-80 For this reason, the choice of the core-forming block is most critical. One parameter that can be used to assess the compatibility between the polymer and a drug is the Flory-Huggins interaction parameter, Xsp/ defined as Xsp= (Ss - <5p)2Vs/kT; where Ss and <5p are Scatchard-Hildebrand solubility parameters, and Vs is the molecular volume of the solubilizate. It was successfully used as a correlation parameter for the solubilization of aliphatic and aromatic hydrocarbons in block copolymer micelles.80,81 Recently, Allen's group82 elegantly demonstrated that the calculation and comparison of partial solubility parameters of polymers and drugs could be used as a reliable means to predict polymer-drug compatibility and to guide formulation development. Polymer micelles, possessing core-forming blocks predicted to be compatible with the drug of interest (Ellipticine), were able Polymer Micelles as Drug Carriers 65 to increase the solubility of the drug up to 30,000 times, compared with its saturation solubility in water.82 The degree of compatibility between the drug and the core-forming block has also been shown to influence the release rate of the drug from the micelles. When the environment within the core of the micelle becomes more compatible with the drug, it results in a considerable decrease in the rate of drug release. For a given drug, the extent of incorporation is a function of factors that also control the micelle size and/or aggregation number. Such factors include the ratio of hydrophobic to hydrophilic block length and the copolymer molecular weight. For example, the loading capacity of Pluronic micelles was found to increase with the increase in the hydrophobic PPO block length. This effect is attributed to a decrease in CMC, and therefore, an increase in aggregation number and micelle core size. Also, but to a lesser extent, the hydrophilic block length affects the extent of solubilization, such that an increase in percentage of PEO in Pluronic block copolymers results in a decrease in the loading capacity of the micelles.80,83-85 For a given ratio of PPO-to-PEO, higher molecular weight polymers form larger micelles, and therefore, show a higher drug loading capacity. Therefore, the total amount of loaded drug can be adjusted as a function of the micellar characteristics as clearly was demonstrated by Nagaradjan83 and Kozlov et al.85 Several studies indicate that both the copolymer concentration as well as the drug to polymer ratio upon loading, have a complex effect on the loading capacity of polymer micelles.79,84,86 In general, more polymer chains provide more absorption sites. As a result, solubilization is increased with polymer concentration.82 However, the solubilization capacity was found to reach a saturation level with an increase of polymer concentration.79 The maximum loading level is largely influenced by the interaction between the solubilizate and core-forming block, and stronger interactions enable saturation to be reached at lower polymer concentration. It was also demonstrated in the studies by Hurter and Hatton84'86 that the loading capacity of micelles formed from copolymers with high hydrophobic content was independent of the polymer concentration. In addition, the location of the incorporated molecules within polymer micelles (micelle core or the core-shell interface) determines the extent of solubilization, as well as the rate of drug release.87,88 It has been found that more soluble compounds are localized at the core-shell interface or even in the inner shell, whereas more hydrophobic molecules have a tendency to solubilize in the micelle core.85,87,88 The release rate of drug localized in the shell or at the interface appears to account for the "burst release" from the micelles.87 In general, for drugs physically incorporated in polymer micelles, release is controled by the rate of diffusion of the drug from the micellar core, stability of the micelles, and the rate of biodegradation of the copolymer. If the micelle is stable and the rate of polymer biodegradation is slow, the diffusion rate of the drug will be mainly determined by the abovementioned factors, 66 Batrakova et al. i.e. the compatibility between the drug and core forming block of copolymer,69,82 the amount of drug loaded, the molecular volume of drug, and the length of the core forming block.89 In addition, the physical state of the micelle core and drug has a large influence on release characteristics. It was demonstrated that the diffusion of incorporated molecules from the block copolymer micelles with glassy cores is slower, in comparison to the diffusion out of the cores that are more mobile.87 3.3. Poly ionic complexation Charged therapeutic agents can be incorporated into block copolymer micelles, through electrostatic interactions with an oppositely charged ionic segment of block copolymer. Since it was being proposed independently by Kabanov and Kataoka in 1995,90,91 this approach is now widely used for the incorporation of various polynucleic acids into block ionomer complexes, for developing non-viral gene delivery systems. Ionic block lengths, charge density, and ionic strength of the solution affect the formation of stable block ionomer complexes, and therefore, control the amount of drug that can be incorporated within the micelles.8'92 The pHand salt-sensitivity of such block ionomer micelles provide a unique opportunity to control the triggered release of the active therapeutic agent.1563,93-96 Furthermore, block ionomer complexes can participate in the polyion interchange reactions which are believed to account for the release of the therapeutic agent and DNA in an active form inside cells.7 Several comprehensive reviews can be found in the literature that focus on block ionomer micelles as drug and gene delivery systems.8,92 In addition, physicochemical aspects of the DNA complexes with cationic block copolymers have also been recently reviewed.97 As an example, the metal-complex formation of ionic block copolymer, PEOb- poly(L-aspartic acid), was explored to prepare polymer micelles incorporating cz's-dichlorodiamminoplatinum (II) (CDDP);98,99 a potent chemotherapeutic agent widely used in the treatment of a variety of solid tumors, particularly, testicular, ovarian, head and neck, and lung tumors.100,101 The CDDP-loaded micelles had a size of approximately 20 nm. These micelles showed remarkable stability upon dilution in distilled water, while in physiological saline, they displayed sustained release of the regenerated Pt complex over 50hrs, due to inverse ligand exchange from carboxylate to chloride. The release rate was inversely correlated with the chain length of poly(L-aspartic acid) segments in the block copolymer. The stability of CDDP-loaded micelle against salt was shown to be improved by the addition of homopolymer, poly(L-aspartic acid), in the micelles.102 Recently, CDDP-loaded micelles were newly prepared using another block copolymer, PEO-b-poly(glutamic acid) to improve and optimize the micellar stability, as well Polymer Micelles as Drug Carriers 67 as the drug release profile.103 The drug loading in the micelles was as high as 39% (w/w), and these micelles released the platinum in physiological saline at 37°C in sustained manner > 150 hrs, without initial burst of the drug. The principle of polyionic complexation can also be used to design new photosensitizers for photodynamic therapy of cancer. The group of Kataoka reported formation of micelles, as a result of mixing of oppositely charged dendrimer porphyrin and block ionomer, based on electrostatic assembly104 or combination of electrostatic and hydrogen bonding interactions.95'105 The micelles were stabile at physiological conditions and released the entrapped dendrimers in the acidic pH environment (pH 5.0), suggesting a possibility of pH-triggered drug release in the intracellular endosomal compartments. Overall, the photodynamic efficacy of the dendrimer porphyrins was dramatically improved by inclusion into micelles. This process resulted in more than two orders of magnitude increase in the photocytotoxicity, compared with that of the free dendrimer porphyrins. In addition, the polyionic complexation has been used to immobilize charged enzymes such as egg white lysozyme106 or trypsin,107 which were incorporated in the core of polyion micelles, after mixing with oppositely charged ionic block copolymer. A remarkable enhancement of enzymatic activity was observed in the core of the micelles. Furthermore, the on-off switching of the enzyme activity was achieved through the destabilization of the core domain by applying a pulse electric field.108 These unique features of the polyion micelles are relevant for their use as smart nanoreactors in the diverse fields of medical and biological engineering. Last, but not the least, a special class of polyion complexes has been synthesized by reacting block ionomers with surfactants of opposite charge, resulting in the formation of environmentally responsive nanomaterials, which differ in sizes and morphologies, and include micelles and vesicles.109-113 These materials contain a hydrophobic core formed by the surfactant tail groups, and a hydrophilic shell formed, for example, by PEO chains of the block ionomer. These block ionomer complexes can incorporate charged surfactant drugs such as retinoic acid, as well as other drugs via solubilization in the hydrophobic domains formed by surfactant molecules.114 They display transitions induced by changes in pH, salt concentration, chemical nature of low molecular mass counterions, as well as temperature. They can also be fine tuned to respond to environmental changes occurring in a very wide range of conditions that could realize during delivery of biological and imaging agents.94115 The unique self-assembly behavior, the simplicity of the preparation, and the wide variety of available surfactant components that can easily produce polymer micelles with a very broad range of core properties, make this type of materials extremely promising for developing vehicles for the delivery of diagnostic and therapeutic modalities. 68 Batrakova et al. 4. Pharmacokinetics and Biodistribution Incorporation of a low molecular mass drug into polymer micelles drastically alters pharmacokinetics and biodistribution of the drug in the body, which is crucial for the drug action. Low molecular mass drugs, after administration in the body, rapidly extravasate to various tissues affecting them almost indiscriminately, and then are rapidly eliminated from the body via renal clearance, often causing toxicity to kidneys.116 Furthermore, many drugs display low stability and are degraded in the body, often forming toxic metabolites. An example is doxorubicinol, a major metabolite of doxorubicin, which causes cardiac toxicity.117 These impediments to the therapeutic use of low molecular mass drugs can be mitigated by encapsulating drugs in polymer micelles. Within the micelles, the drug molecules are protected from enzymatic degradation by the micelle shell. The pharmacokinetics and biodistribution of the micelle-incorporated drugs are mainly determined by the surface properties, size, and stability of the micelles, and are less affected by the properties of the loaded drug. The surface properties of the micelles are determined by the micelle shell. The shell from PEO effectively masks drug molecules and prevents interactions with serum proteins and cells, which contributes to prolonged circulation of the micelles in the body.16 From the size standpoint, polymer micelles fit an ideal range of sizes for systemic drug delivery. On the one hand, micelles are sufficiently large, usually exceeding 10 nm in diameter, which hinders their extravasation in nontarget tissues and prevents renal glomerular excretion. On the other hand, the micelles are not considered large, since their size usually does not exceed 100 nm. As a result, micelles avoid scavenging by the mononuclear phagocytes system (MPS) in the liver and spleen. To this end, "stealth" particles whose surface is decorated with PEO are known to be less visible to macrophages and have prolonged half-lives in the blood.64,118,119 The contribution of the micelle stability to pharmacokintetics and biodistribution is much less understood, although it is clear that micelle degradation should result in a decrease of the size and drug release, perhaps, prematurely. Degradation of the micelles, resulting in the formation of block copolymer unimers, could also be a principal route for the removal of the polymer material from the body. The molecular mass of the unimers of most block copolymers is below the renal excretion limit, i.e. less than ~ 20 to 40 kDa,22,120121 while the molecular mass of the micelles, which usually contain several dozen or even hundreds of unimers molecules, is above this limit. Thus, the unimers are sufficiently small and can be removed via renal excretion, while the micelles cannot. A recent study by Batrakova et al. determined pharmacokinetic parameters of an amphiphilic block copolymer, Pluronic P85, and perhaps provided first evidence that the pharmacokinetic behavior of a block copolymer can be a function of its aggregation state.119 Specifically, the formation Polymer Micelles as Drug Carriers 69 of micelles increased the half-life of the block copolymer in plasma and decreased the uptake of the block copolymer in the liver. However, it had no effect on the total clearance, indicating that the elimination of Pluronic P85 was controled by the renal tubular transport of unimers, but not by the rate of micelles disposition or disintegration. Furthermore, the values of the clearance suggested that a significant portion of the block copolymer was reabsorbed back into the blood, probably, through the kidney's tubular membranes. Chemical degradation of the polymers comprising the micelles, followed by renal excretion of the relatively low molecular mass products of degradation, may be another route for the removal of the micelle polymer material from the body. This route could be particularly important in the case of the cross-linked or unimolecular micelles, micelles displaying very high stability, and / or micelles composed from very hydrophobic polymer molecules that can bind and retain considerably biological membranes and other cellular components. The delivery of chemotherapeutic drugs to treat tumors is one of the most advanced areas of research using polymer micelles. Two approaches have been explored to enhance delivery of drug-loaded polymer micelles to the tumor sites: (1) passive targeting and (2) vectorized targeting. The passive targeting involves enhanced permeability and retention (EPR) effect.122,123 It is based on the fact that solid tumors display increased vascular density and permeability caused by angiogenesis, impaired lymphatic recovery, and lack of a smooth muscle layer in solid tumor vessels. As a result, micellar drugs can penetrate and retain in the sites of tumor lesions. At the same time, extravasation of micellar drugs in normal tissues is decreased, compared with low molecular drug molecules. Among normal organs, spleen and liver can accumulate polymer drugs, but the drugs are eventually cleared via the lymphatic system. The increased circulation time of the micellar drugs should further enhance exposure of the tumors to the micellar drug, compared with the low molecular mass drugs. Along with passive targeting, the delivery of micellar drugs to tumors can potentially be enhanced by the modification of the surface of the polymer micelles with the targeting molecules, vectors that can selectively bind to the surface of the tumor cells. Potential vectors include antibodies, aptamers and peptides, capable of binding tumor-specific antigens and other molecules diplayed at the surfaces of the tumors.124-126 Altered biodistribution of a common antineoplastic agent was demonstrated for CDDP encapsulated in polyionic micelles with PEO-b-poly(glutamic acid) block copolymers.103 Free CDDP is rapidly distributed to each organ, where its levels peak at about one hr after i.v. administration. In contrast, in the case of the CDDPincorporated micelles, due to their remarkably prolonged blood circulation time, the drug level in the liver, spleen and tumor continued to increase up to at least 24 hrs after injection. Consequently, the CDDP-incorporated micelle exhibited 4-, 39-, and 20-fold higher accumulation in the liver, spleen and tumor respectively, 70 Batrakova et al. than the free CDDP. At the same time, the encapsulation of CDDP into the micelles significantly decreased drug accumulation in the kidney, especially during first hr after administration. This suggested potential for the decrease of severe nephrotoxicity observed with the free drug, which is excreted through the glomerular filtration, thus affecting the kidney.127 Promising results were also demonstrated for doxorubicin incorporated into styrene-maleic acid micelles.128 In this case, as a result of drug entrapment into micelles, the drug was redirected from the heart to the tumor, and the doxorubicin cardiotoxicity was diminished. Complete blood counts and cardiac histology for the micellar drug showed no serious side effects for i.v. doses as high as 100 mg/kg doxorubicin equivalent in mice. Similar results were reported for doxorubicin incorporated in mixed micelles of PEO-b-poly(L-histidine) and PEO-b-poly(L-lactic acid) block copolymers.129 Tissue levels of doxorubicin administered in the micellar formulation were decreased in the blood and the liver, and considerably increased in the solid tumor, compared with the free drug. Further increase in the tumor delivery was achieved by modifying the surface of the micelles with the folate molecules. The accumulated doxorubicin levels observed using folate-modified micelles was 20 times higher than those for free doxorubicin, and 3 times higher than those for unmodified micelles. The first micellar formulation of doxorubicin to reach clinical evaluation stage, used the micelles composed of triblock copolymer, PEO-b-poly (propylene oxide)-b- PEO, Pluronic.130 Analysis of pharmacokinetics and biodistribution of doxorubicin incorporated into mixed micelles of Pluronics L61 and F127, SP1049C, demonstrated more efficient accumulation of the micellar drug in the tumors, compared with the free drug. Specifically, the areas under the curves (AUC) in the Lewis lung carcinoma 3LL M-27 solid tumors in C57B1 / 6 mice were increased about two fold using SP1049, compared with the free doxorubicin. Furthermore, this study indicated that the peak levels of doxorubicin formulated with SP1049 in the tumor were delayed and the drug residence time was increased, in comparison with the free doxorubicin.130 A clear visualization of drug delivery to the tumor site was shown for doxorubicin covalently incorporated through the pH-sensitive link into polymer micelles of PEO-poly(aspartate hydrazone doxorubicin).64 A phase-contrast image showed that the tumor blood vessels containing the micelles leaked into extra vascular compartments of the tumors, resulting in the infiltration of the micelles into tumor sites. The micelles circulated in the blood for a prolonged time, and the AUC for micellar doxorubicin was 15-fold greater than the AUC for the free doxorubicin. Furthermore, the AUC values of the micellar doxorubicin in the heart and kidney decreased, compared with the free drug. Thus, the selectivity of drug delivery to the tumor, compared with heart and kidney (AUCtumor/AUC0rgan) was increased by 6- and 5-folds respectively. This may result in the reduction of side effects of Polymer Micelles as Drug Carriers 71 doxorubicin such as cardiotoxicity and nephrotoxicity. Moreover, the micellar doxorubicin showed relatively low uptake in the liver and spleen, despite very long residence time in the blood. Biodistribution of paclitaxel incorporated into biodegradable polymer micelles of monomethoxy-PEO-b-poly(D,L-lactide) block copolymer, Genexol-PM, was compared with the regular formulation of the drug in Cremophor EL.131 Two to three-fold increases in drug levels were demonstrated in most tissues including liver, spleen, kidneys, lungs, heart and tumor, after i.v. administration of Genexol- PM, compared with paclitaxel. Nevertheless, acute dose toxicity of Genexol-PM was about 25 times lower than that of the conventional drug formulation, which appears to be a result of the reformulation avoiding the use of Chremophor EL and dehydrated ethanol that are toxic. Selective tumor targeting with paclitaxel encapsulated in micelles, modified with tumor-specific antibodies 2C5 ("immunomicelles"), was reported using Lewis lung carcinoma solid tumor model in C57B1/6J mice.26 These micelles were prepared from PEO-distearyl phosphatidylethanolamine conjugates with the free PEO end activated with the p-nitrophenylcarbonyl group for the antibody attachment. The amount of micellar drug accumulated in the tumor exceeded that in the nontarget tissue (muscles) by more than ten times. It is worth noting that the highest accumulation in the tumor was demonstrated in the micelles containing the longest PEO chains, which also had the longest circulation time in the blood. Furthermore, the immunomicelles displayed the highest amount of tumor-accumulated drug, compared with either free paclitaxel or non-vectorized micelles. It was demonstrated that paclitaxel delivered by plain micelles in the interstitial space of the tumor was eventually cleared after gradual micellar degradation. In contrast, paclitaxel-loaded 2C5 immunomicelles were internalized by cancer cells and the retention of the drug inside the tumor was enhanced.132 Unexpected results were found using pH-sensitive polymer micelles of Nisopropylacrylamide and methacrylic acid copolymers randomly or terminally alkylated with octadecyl groups.64,133 It was demonstrated that aluminium chloride phthalocyanine (AlClPc) incorporated in such micelles was cleared more rapidly and less accumulated in the tumor, than the AlClPc formulated with Cremophor EL. Furthermore, significant accumulation in the liver and spleen (and lungs for most hydrophobic copolymers) was observed, compared with Cremophor EL formulation. The enhanced uptake of such polymer micelles by the cells of mononuclear phagocyte system (MPS) could be due to micelle aggregation in the blood and embolism in the capillaries. Thus, it attempted to reduce the uptake of the micelles in MPS by incorporating water soluble monomers, N-vinyl-2-pyrrolidone in the copolymer structure.134 The modified formulation displayed same levels of tumor accumulation and somewhat higher antitumor activity than the Cremophor 72 Batrakova et al. EL formulation. This work serves as an example reinforcing the need of proper adjustment of the polymer micelle structure, and perhaps the need of using block copolymers to produce a defined protective hydrophilic shell to facilitate evasion of the polymer micelles from MPS. 5. Drug Delivery Applications The studies on the application of polymer micelles in drug delivery have mostly focused on the following areas that are considered below: (1) delivery of anticancer agents to treat tumors; (2) drug delivery to the brain to treat neurodegenerative diseases; (3) delivery of antifungal agents; (4) delivery of imaging agents for diagnostic applications; and (5) delivery of polynucleotide therapeutics. 5.1. Chemotherapy of cancer To enhance chemotherapy of tumors using polymer micelles, four major approaches were employed: (1) passive targeting of polymer micelles to tumors due to EPR effect; (2) targeting of polymer micelles to specific antigens overexpressed at the surface of tumor cells; (3) enhanced drug release at the tumor sites having low pH; and (4) sensitization of drug resistant tumors by block copolymers. A series of pioneering studies by Kataoka's group used polymer micelles for passive targeting of various anticancer agents and chemotherapy of tumors.102,103'135 One notable recent example reported by this group involves polymer micelles of PEO-b-poly(L-aspartic acid) incorporating CDDP. Evaluation of anticancer activity using murine colon adenocarcinoma C26 as an in vivo tumor model, demonstrated that CDDP in polymer micelles had significantly higher activity than the free CDDP, resulting in complete eradication of the tumor.103 A formulation of paclitaxel in biodegradable polymer micelles of monomethoxy-PEO-b-poly(D,L-lactide) block copolymer, Genexol-PM, also displayed elevated activity in vivo against human ovarian carcinoma OVCAR-3 and human breast carcinoma MCF7, compared with a regular formulation of the drug in Cremophor EL.131 In addition, anthracycline antibiotics, doxorubicin and pirarubicin, incorporated in styrene-maleic acid micelles each revealed potent anticancer effects in vivo against mouse sarcoma S-180, resulting in complete eradication of tumors in 100% of tested animals.128 Notably, animals survived for more than one year, after treatment with the micelleincorporated pirarubicin at doses as high as lOOmg/kg of pirarubicin equivalent. Complete blood counts, liver function test, and cardiac histology showed no sign of adverse effects for intravenous doses of the micellar formulation. In contrast, animals receiving free pirarubicin had a much reduced survival and showed serious side effects.136 Collectively, these studies suggested that various micelleincorporated drugs display improved therapeutic index in solid tumors, which Polymer Micelles as Drug Carriers 73 correlates with enhanced passive targeting of the drug to the tumor sites, as well as decreased side effects, compared with conventional formulations of these drugs. Tumor-specific targeting of polymer micelles to molecular markers expressed at the surface of the cancer cells has also been explored to eradicate tumor cells. For example, a recent study by Gao's group developed a polymer micelle carrier to deliver doxorubicin to the tumor endothelial cells with overexpressed Xvfi3 integrins.137 A cyclic pentapeptide, cRGD was used as a targeting ligand that is capable of selective and high affinity binding to the Xvfio, integrin. Micelles of PEOb- poly(e-caprolactone) loaded with doxorubicin were covalently bound with cRGD. As a result of such modification, the uptake of doxorubicin-containing micelles in in vitro human endothelial cell model derived from Kaposi's sarcoma, was profoundly increased. In addition, folate receptor often overexpressed in cancer cells has been evaluated for targeting various drug carriers to tumors.138 This strategy has also been evaluated to target polymer micelles. For example, mixed micelles of PEO-b-poly(L-histidine) and PEO-b-poly(L-lactic acid) block copolymers with solubilized doxorubicin129 or micelles of PEO-b-poly(DL-lactic-co-glycolic acid) block copolymer with covalently attached doxorubicin,139 were each surface modified by conjugating folate molecules to the free PEO ends. In both cases, in vitro and in vivo studies demonstrated increased antitumor activity of the micelleincorporated drug resulting from such modification. The enhanced delivery of the micellar drugs through the folate receptor, and the enhanced retention of the modified micelles at the tumor sites are possible explanations for the effects of these folate modifications. Micelles conjugated with antibodies or antibody fragments capable to recognize tumor antigens were shown to improve therapeutic efficacy in vivo over non-modified micelles.23 This approach can result in high selectivity of binding, internalization, and effective retention of the micelles in the tumor cells. In addition, recent advances in antibody engineering allow for the production of humanized antibody fragments, reducing problems with immune response against mouse antibodies.140 For example, micelles of PEO-distearyl phosphatidylethanolamine were covalently modified with the monoclonal antibody 2C5 that binds to microsomes, displayed at the surface of many tumor cells. The micelles were then used for incorporating various poorly soluble anticancer drugs including tamoxifen, paclitaxel, dequalinium, and chlorine e6 trimethyl ester.26'132'141 It was shown that paclitaxel-loaded 2C5-immunomicelles could specifically recognize a variety of tumor types. The binding of these immunomicelles was observed for all cancer cell lines tested, i.e. murine Lewis lung carcinoma, T-lymphoma EL4, and human breast adenocarcinomas, BT-20 and MCF7.141 Moreover, paclitaxel-loaded 2C5 immunomicelles demonstrated highest anticancer activity in Lewis lung carcinoma tumor model in mice, compared with plain paclitaxel-loaded micelles and 74 Batrakova et al. the free drug.132 The increased antitumor effect of immunomicelles in vivo correlated with the enhanced retention of the drug delivered with the immunomicelles inside the tumor. Tumors often display low pH of interstitial fluid, which is mainly attributed to higher rates of aerobic and anaerobic glycolysis in cancer cells than in normal cells.142,143 This phenomenon has been employed in the design of various pH-sensitive polymer micelle systems for the delivery of anticancer drugs to the tumors. One approach consisted in the chemical conjugation of anticancer drugs to the block copolymers through pH sensitive cleavable links that are stable at neutral pH, but are cleavable and release the drug in the mildly acidic pH. For example, several groups used hydrasone-based linking groups, to covalently attach doxorubicin to PEO-b-poly(DL-lactic-co-glycolic acid) block copolymer,21,144 PEOb- block-poly(allyl glycidyl ether)145 or PEO-b-poly(aspartate hydrazone) block copolymer.63,64 It was suggested that doxorubicin will remain in the micelles in the blood stream, and will be released at tumor sites at lower pH. For example, in vitro and in vivo studies using PEO-b-poly(aspartate hydrazone doxorubicin) micelles demonstrated that the micelles display an intracellular pH-triggered drug release capability, tumor-infiltrating permeability, and effective antitumor activity with extremely low toxicity.63,64 Overall, the animal studies suggested that such polymer micelle drug has a wide therapeutic window due to increased efficacy and decreased toxicity, compared with free doxorubicin.64 An alternative mechanism for pH-induced triggering of drug release at the tumor sites consists of using pH sensitive polyacids or polybases as building blocks for polymer micelles.94,146,147 For example, mixed micelles of PEO-bpoly( L-histidine) and PEO-b-poly(L-lactic acid) block copolymers incorporate pHsensitive poly-base, poly(L-histidine) in the hydrophobic core.147 The core can also solubilize hydrophobic drugs such as doxorubicin. The protonation of the polybase at acidic conditions resulted in the destabilization of the core and triggered release of the drug. This system was also targeted to the tumors through the folate molecules as described earlier and has shown significant in vivo antitumor activity and less side effects, compared with the free drug.129 Notably, it was also effective in vitro and in vivo against multidrug resistant (MDR) human breast carcinoma MCF7/ADR that overexpresses P-glycoprotein (Pgp). Pgp is a drug efflux transport protein that serves to eliminate drugs from the cancer cells and significantly decreases the anticancer activity of the drugs. The micelle incorporated drug was released inside the cells, and thus avoided the contact with Pgp localized at the cell plasma membrane, which perhaps contributed to the increased activity of pH sensitive doxorubicin micelles in the MDR cells. A different approach using Pluronic block copolymer micelles to overcome MDR in tumors has been developed by our group.130,148-151 Studies by Alakhov Polymer Micelles as Drug Carriers 75 et al. demonstrated that Pluronic block copolymers can sensitize MDR cells, resulting in an increased cytotoxic activity of doxorubicin, paclitaxel, and other drugs by 2,3 orders of magnitude.148'149 Remarkably, Pluronic can enhance drug effects in MDR cells through multiple effects including (1) inhibiting drug efflux transporters, such as Pgp149-152 and multidrug resistance proteins (MRPs),153'154 (2) abolishing drug sequestration within cytoplasmic vesicles,149'153 (3) inhibiting the glutathione/glutathione S-transferase detoxification system,154 and (4) enhancing proapoptotic signaling in MDR cells.155 Similar effects of Pluronics have also been reported using in vivo tumor models.130,150 In these studies, mice bearing drug-sensitive and drug-resistant tumors were treated with doxorubicin alone and with doxorubicin in Pluronic compositions. The tumor panel included i.p. murine leukemias (P388, P388-Dox), s.c. murine myelomas (Sp2/0, Sp2/0-Dnr), i.v. and s.c. Lewis lung carcinoma (3LL-M27), s.c. human breast carcinomas (MCF7, MCF7/ADR), and s.c. human oral epidermoid carcinoma (KBv).130 Using the NCI criteria for tumor inhibition and increased lifespan, Pluronic/doxorubicin has met the efficiency criteria in all models (9 of 9), while doxorubicin alone was only effective in selected tumors (2 of 9) .130 Results showed that the tumors were more responsive in the Pluronic /doxorubicin treatment groups than in doxorubicin alone. These studies demonstrated improved treatment of drug resistant cancers with Pluronics. The mechanisms of effects of Pluronic on Pgp have been studied in great detail.151 In particular, exposure of MDR cells to Pluronics has resulted in the inhibition of Pgp-mediated efflux,149 and this overcomes defects in intracellular accumulation of Pgp-dependent drugs,148,149,152 and abolishes the directionality difference in the flux of these drugs across polarized cell monolayers.156-158 The lack of changes in membrane permeability with Pluronics to (1) non-Pgp compounds in MDR cells,158,159 and (2) Pgp probes in non-MDR cells,149,153 suggested that Pluronic effects were specific to the Pgp efflux system. These effects were observed at Pluronic concentrations less than or equal to the critical micelle concentration (CMC).152,159 Thus, Pluronic unimers rather than the micelles were responsible for these effects. Specifically, Pluronic molecules displayed a dual function in MDR cells.160-162 Firstly, they incorporated into the cell membranes and decreased the membrane microviscosity. This was accompanied by the inhibition of Pgp ATPase activity. Secondly, they translocated into cells and reached intracellular compartments. This was accompanied by the inhibition of respiration,163 presumably due to Pluronic interactions with the mitochondria membranes. As a result, within 15 min after exposure to select Pluronics, intracellular levels of ATP in MDR cells were drastically decreased.160-162 Remarkably, such ATP depletion was not observed in non-MDR cells, suggesting that the Pluronic was "selective", with respect to the MDR phenotype.160'164 Combining these two effects, Pgp ATPase inhibition and ATP depletion, resulted in the shut-down of the efflux system in MDR 76 Batrakova et al. cells.160-162 The Pgp remained functionally active when (1) ATP was restored using an ATP supplementation system in the presence of a Pluronic, or (2) when ATP was depleted, but there was no direct contact between the Pluronic and Pgp (and no ATPase inhibition). Overall, these detailed studies which resulted in the development of a micellar formulation of doxorubicin that is evaluated clinically, reinforce the fact that block copolymers, comprising the micelles, can serve as biological response modifying agents that can have beneficial effects in the chemotherapy of tumors. 5.2. Drug delivery to the brain By restricting drug transport to the brain, the blood brain barrier (BBB) represents a formidable impediment for the treatment of brain tumors and neurodegenerative diseases such as HIV-associated dementia, stroke, Parkinson's and Alzheimer's diseases. Two strategies using polymer micelles have been evaluated to enhance delivery of biologically active agents to the brain. The first strategy is based on the modification of polymer micelles with antibodies or ligand molecules capable of transcytosis across brain microvessel endothelial cells, comprising the BBB. The second strategy uses Pluronic block copolymers to inhibit drug efflux systems, particularly, Pgp, and selectively increase the permeability of BBB to Pgp substrates. An earlier study used micelles of Pluronic block copolymers for the delivery of the CNS drugs to the brain.68'73 These micelles were surface-modified by attaching to the free PEO ends, either polyclonal antibodies against brain-specific antigen, a2-glycoprotein, or insulin to target the receptor at the lumenal side of BBB. The modified micelles were used to solubilize fluorescent dye or neuroleptic drug, haloperidol, and these formulations were administered intravenously in mice. Both the antibody and insulin modification of the micelles resulted in enhanced delivery of the fluorescent dye to the brain and drastic increases in neuroleptic effect of haloperidol in the animals. Subsequent studies using in vitro BBB models demonstrated that the micelles, vectorized by insulin, undergo receptor-mediated transport across brain microvessel endothelial cells.156 Based on one of these observations, one should expect development of novel polymer micelles that target specific receptors at the surface of the BBB to enhance transport of the incorporated drugs to the brain. The studies by our group have also demonstrated that selected Pluronic block copolymers, such as Pluronic P85, are potent inhibitors of Pgp, and they have the increased entry of the Pgp-substrates to the brain across BBB.156'158'159'165 Pluronic did not induce toxic effect in BBB, as revealed by the lack of alteration in paracellular permeability of the barrier,156'158 and in histological studies, using specific markers for brain endothelial cells.166 Overall, this strategy has potential in developing Polymer Micelles as Drug Carriers 77 novel modalities for the delivery of various drugs to the brain, including selective anti cancer agents to treat metastatic brain tumors, as well as HIV protease inhibitors to eradicate HIV virus in the brain.167'168 5.3. Formulations of antifungal agents The need for safe and effective modalities for the delivery of chemotherapeutic agents to treat systemic fungal infections in immunocompromised AIDS, surgery, transplant and cancer patients is very high. The challenges to the delivery of antifungal agents include low solubility and sometimes high toxicity of these agents. These agents, such as amphotericin B, have low compatibility with hydrophobic cores of polymer micelles formed by many conventional block copolymers. Thus, to increase solubilization of amphotericin B, the core-forming blocks of methoxy-PEOb- poly(L-aspartate) were derivatized with stearate side chains.169-172 The resulting block copolymers formed micelles. Amphotericin B interacted strongly with the stearate side chains in the core of the micelles, resulting in an efficient entrapment of the drug in the micelles, as well as subsequent sustained release in the external environment. As a result of solubilization of amphotericin B in the micelles, the onset of hemolytic activity of this drug toward bovine erythrocytes was delayed, relative to that of the free drug.171 Using a neutropenic murine model of disseminated Candidas, it was shown that micelle-incorporated amphotericin B retained potent in vivo activity. Pluronic block copolymers were used by the same group for incapsulation of another poorly soluble antifungal agent, nystatin.172 This is a commercially available drug that has shown potential for systemic administration, but has never been approved for that purpose, due to toxicity issues. The possibility to use Pluronic block copolymers to overcome resistance to certain antifungal agents has also been demonstrated.173-176 Overall, one should expect further scientific developments using polymer micelle delivery systems for the treatment of fungal infection. 5.4. Delivery of imaging agents Efficient delivery of imaging agents to the site of disease in the body can improve early diagnostics of cancer and other diseases. The studies in this area using polymer micelles as carriers for imaging agents were initiated by Torchilin.177 For example, micelles of amphiphilic PEO-lipid conjugates were loaded with i n In and gadolinium diethylenetriamine pentaacetic acid-phosphatidylethanolamine (Gd- DTPA-PE) and then used for visualization of local lymphatic chain after subcutaneous injection into the rabbit's paw.178 The images of local lymphatics were acquired using a gamma camera and a magnetic resonance (MR) imager. The 78 Batrakova et al. injected micelles stayed within the lymph fluid, thus serving as lymphangiographic agents for indirect MR or gamma lymphography. Another polymer micelle system composed of amphiphilic methoxy-PEO-b-poly[epsilon,N-(triiodobenzoyl)-Llysine] block copolymers, labeled with iodine, was administered systemically in rabbits and visualized by X-ray computed tomography.179 The labeled micelles displayed exceptional 24 hrs half-life in the blood, which is likely due to the coreshell architecture of the micelle carriers that protected the iodine-containing core. Notably, small polymer micelles (<20nm) may be advantageous for bioimaging of tumors, compared with PEG-modified long-circulating liposomes (ca. lOOnm). In particular, the micelles from PEO-distearoyl phosphatidyl ethanolamine conjugates containing m In-labeled model protein were more efficacious in the delivery of protein to Lewis lung carcinoma than larger long-circulating liposomes.180 Overall, polymer micelles loaded with various agents for gamma, magnetic resonance, and computed tomography imaging represent promising modalities for non-invasive diagnostics of various diseases. 5.5. Delivery of polynucleotides To improve the stability of polycation-based DNA, delivery complexes in dispersion block and graft copolymers containing segments from polycations and nonionic water-soluble polymers, such as PEO, were developed.90,181,182 Binding of these copolymers with DNA results in the formation of micelle-like block ionomer complexes ("polyion complex micelles"), containing hydrophobic sites formed by the polycation-neutralized DNA and hydrophilic sites formed by the PEO chains. Despite neutralization of charge, complexes remain stable in aqueous dispersion due to the effect of the PEO chains.183 Overall, the PEO modified polycation-DNA complexes form stable dispersions and do not interact with serum proteins.183,184 These systems were successfully used for intravitreal delivery of an antisense oligonucleotide and the suppression of gene expression in retina in rats.185 Furthermore, they displayed extended plasma clearance kinetics and were shown to transfect liver and tumor cells, after systemic administration in the body.186-188 In addition, there is a possibility targeting such polyplexes to the specific receptors at the surface of the cell, for example, by modifying the free ends of PEO chains with specific targeting ligands.189-191 Alternatively, to increase the binding of the complexes with the cell membrane and the transport of the polynucleotides inside the cells, the polycations were modified with amphiphilic Pluronic molecules.192,193 One recent study has shown a potential of Pluronic-polyethyleneimine-based micelles for in vivo delivery of antisense oligonucleotides to tumors, and have demonstrated sensitization of the tumors to radiotherapy as a result of systemic administration of the oligonucleotide-loaded micelles.194 Polymer Micelles as Drug Carriers 79 6. Clinical Trials Three polymer micelle formulations of anticancer drugs have been reported to reach clinical trials. The doxorubicin-conjugated polymer micelles developed by Kataoka's group195 have progressed recently to Phase I clinical trial at the National Cancer Center Hospital, Tokyo, Japan. The micelle carrier NK911 is based on PEO-bpoly( aspartic acid) block copolymers, in which the aspartic acid units were partially (ca. 45%) substituted with doxorubicin to form hydrophobic block. The resulting substituted block copolymer forms micelles that are further noncovalently loaded with free doxorubicin. Preclinical studies in mice demonstrated higher NK911 activity against Colon 26, M5076, and P388, compared with the free drug. Moreover, NK911 has less side effects, resulting in less animal body and toxic death than the free drug.196 Clinically, the Pluronic micelle formulation of doxorubicin has been most advanced. Based on the in vivo efficacy evaluation, Pluronic L61 was selected for clinical development for the treatment of MDR cancers. The final block copolymer formulation is a mixture of 0.25% Pluronic L61 and 2% Pluronic F127, formulated in isotonic buffered saline.130 This system contains mixed micelles of L61 and F127, with an effective diameter of ca. 22 to 27 nm and is stable in the serum. Prior to administration, doxorubicin is mixed with this system, which results in spontaneous incorporation of the drug in the micelles. The drug is easily released by diffusion after dilution of the micelles. The formulation of doxorubicin with Pluronic, SP1049C, is safe, following systemic administration based on toxicity studies in animals.130 A two-site Phase I clinical trial of SP1049C has been completed.197 Based on its results, the dose-limiting toxicity of SP1049C was myelosuppression, reached at 90mg/m2 (maximum tolerated dose was 70mg/m2). Phase II study of this formulation to treat inoperable metastatic adenocarcinoma of the esophagus is near completion as well.198 Finally, Phase I studies were reported for Genexol-PM, a Cremophor-free polymer micelle-formulated paclitaxel.199 Twenty-one patient entered into this study with lung, colorectal, breast, ovary, and esophagus cancers. No hypersensitivity reaction was observed in any patient. Neuropathy and myalgia were the most common toxicities. There were 14% partial responses. The paclitaxel area under the curve and peak of the drug concentration in the blood were increased with the escalating dose, suggesting linear pharmacokinetics for Genexol-PM.199 7. Conclusions Approximately two decades have passed since the conception of the polymer micelle conjugates and nanocontainers for drug delivery. During the first decade, 80 Batrakova et al. only a few studies were published; however, more recently, the number of publications in this field has increased tremendously. During this period, novel biocompatible and/or biodegradable block copolymer chemistries have been researched, the block ionomer complexes capable of incorporating DNA and other charged molecules have been discovered, the pH and other chemical signal sensitive micelles have been developed. Many studies focused on the use of polymer micelles for delivery of poorly soluble and toxic chemotherapeutic agents to the tumors to treat cancer. There has been considerable advancement in understanding the processes of polymer micelle delivery into the tumors, including passive and vectorized targeting of the polymer micelles. Notable achievements also include the studies demonstrating the possibilities for overcoming multidrug resistance in cancer, and enhancing drug delivery to the brain using block copolymer micelles systems. Overall, it is clear that this area has reached a mature stage, reinforced by the fact that several human clinical trials using polymer micelles for cancer drug delivery have been initiated. At the same time, it is obvious that the possibilities for delivery of the diagnostic and therapeutic agents using polymer micelles are extremely broad, and one should expect further increase in the laboratory and clinical research in this field during the next decade. Targeting polymer micelles to cancer sites within the body will address an urgent need to greatly improve the early diagnosis and treatment of cancer. Capabilities for the discovery and use of targeting molecules will support the development of multifunctional therapeutics that can carry and retain antineoplastic agents within tumors. This will also be instrumental in developing novel biosensing and imaging modalities for the early detection of cancer and other devastating human diseases. Acknowledgment The authors acknowledge the support of the research using polymer micelles by grants from the National Institutes of Health CA89225, NS36229 and EB000551, as well as the National Science Foundation DMR0071682, DMR0513699 and BES- 9907281. We also acknowledge financial support of Supratek Pharma, Inc. (Montreal, Canada). AVK and EVB are shareholders and AVK serves as a consultant to this Company. References 1. Ehrlich P (1956) The relationship existing between chemical constitution, distribution, and pharmacological action, in Himmelweite F, Marquardt M and Dale H (eds.) The Collected Papers of Paul Ehrlich. Pergamon, Elmsford, New York, Vol. 1, pp. 596-618. 2. Lipinski CA, Lombardo F, Dominy BW and Feeney PJ (2001) Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Del Rev 46:3-26. Polymer Micelles as Drug Carriers 81 3. Fernandez AM, Van Derpoorten K, Dasnois L, Lebtahi K, Dubois V, Lobl TJ, Gangwar S, Oliyai C, Lewis ER, Shochat D and Trouet A (2001) N-Succinyl-(beta-alanyl-L-leucyl- L-alanyl-L-leucyl)doxorubicin: An extracellularly tumor-activated prodrug devoid of intravenous acute toxicity. / Med Chem 44:3750-3753. 4. Thompson TN (2001) Optimization of metabolic stability as a goal of modern drug design. Med Res Rev 21:412-449. 5. Langer R (2001) Drag delivery. Drags on target. Science 293:58-59. 6. Duncan R (2003) The dawning era of polymer therapeutics. Nat Rev Drug Discov 2:347-360. 7. Kabanov AV and Kabanov VA (1995) DNA complexes with polycations for the delivery of genetic material into cells. Bioconjug Chem 6:7-20. 8. Kakizawa Y and Kataoka K (2002) Block copolymer micelles for delivery of gene and related compounds. Adv Drug Del Rev 54:203-222. 9. Kabanov AV and Alakhov VY (2002) Pluronic block copolymers in drug delivery: From micellar nanocontainers to biological response modifiers. Crit Rev Ther Drug Can Syst 19:1-72. 10. Savic R, Luo L, Eisenberg A and Maysinger D (2003) Micellar nanocontainers distribute to defined cytoplasmic organelles. Science 300:615-618. 11. Hubbell JA (2003) Materials science. Enhancing drug function. Science 300:595-596. 12. Salem AK, Searson PC and Leong KW (2003) Multifunctional nanorods for gene delivery. Nat Mater 2:668-671. 13. Kabanov AV, Nazarova IR, Astafieva IV, Batrakova EV, Alakhov VY, Yaroslavov AA and Kabanov VA (1995) Micelle formation and solubilization of fluorescent probes in poly(oxyethylene-b-oxypropylene-b-oxyethylene) solutions. Macromolecules 28: 2303-2314. 14. Allen C, Maysinger D and Eisenberg A (1999) Nano-engineering block copolymer aggregates for drug delivery. Coll Surf B: Biointerf 16:3-27. 15. Kataoka K, Harada A and Nagasaki Y (2001) Block copolymer micelles for drug delivery: Design, characterization and biological significance. Adv Drug Del Rev 47:113-131. 16. Adams ML, Lavasanifar A and Kwon GS (2003) Amphiphilic block copolymers for drug delivery. J Pharm Sci 92:1343-1355. 17. Sakuma S, Hayashi M and Akashi M (2001) Design of nanoparticles composed of graft copolymers for oral peptide delivery. Adv Drug Del Rev 47:21-37. 18. Francis MF, Lavoie L, Winnik FM and Leroux JC (2003) Solubilization of cyclosporin A in dextran-g-polyethyleneglycolalkyl ether polymeric micelles. Eur } Pharm Biopharm 56:337-346. 19. Francis MF, Piredda M and Winnik FM (2003) Solubilization of poorly water soluble drugs in micelles of hydrophobically modified hydroxypropylcellulose copolymers. / Control Rel 93:59-68. 20. Jeong B, Bae YH and Kim SW (2000) Drug release from biodegradable injectable thermosensitive hydrogel of PEG-PLGA-PEG triblock copolymers. / Control Rel 63:155-163. 21. Yoo HS and Park TG (2001) Biodegradable polymeric micelles composed of doxorubicin conjugated PLGA-PEG block copolymer. / Control Rel 70:63-70. 82 Batrakova et al. 22. Duncan R and Kopecek J (1984) Soluble synthetic polymers as potential drug carriers. AdvPolym Sci 57:51-101. 23. Torchilin VP (2004) Targeted polymeric micelles for delivery of poorly soluble drugs. Cell Mol Life Sci 61:2549-2559. 24. Torchilin VP (2001) Structure and design of polymeric surfactant-based drug delivery systems. / Control Rel 73:137-172. 25. Moghimi SM, Hunter AC and Murray JC (2001) Long-circulating and target-specific nanoparticles: Theory to practice. Pharmacol Rev 53:283-318. 26. Lukyanov AN, Gao Z and Torchilin VP (2003) Micelles from polyethylene glycol/ phosphatidylethanolamine conjugates for tumor drug delivery. / Control Rel 91:97-102. 27. Zhang L, Yu K and Eisenberg A (1996) Ion-Induced morphological changes in "crewcut" aggregates of amphiphilic block copolymers. Science 272:1777-1779. 28. Kabanov AV and Alakhov VY (2000) Micelles of amphilphilic block copolymers as vehicles for drug delivery, in, Alexandridis P and Lindman B (eds.) Amphiphilic Block Copolymers: Self-Assembly and Applications: Elsevier: Amsterdam, Lausanne, New York, Oxford, Shannon, Singapore, Tokyo, pp. 347-376. 29. Kwon GS and Okano T (1999) Soluble self-assembled block copolymers for drug delivery. Pharm Res 16:597-600. 30. Kim KH, Guo HC, Lim HJ, Huh J, Ahn C-H and Jo WH (2004) Synthesis and micellization of star-shaped poly (ethylene glycol)-block-poly(e-caprolactone). Macromolec Chem Phys 205:1684-1692. 31. Gitsov I and Frechet JMJ (1993) Solution and solid-state properties of hybrid lineardendritic block copolymers. Macromolecules 26:6536-6546. 32. Gitsov I, Lambrych KR, Remnant VA and Pracitto R (2000) Micelles with highly branched nanoporous interior: Solution properties and binding capabilities of amphiphilic copolymers with linear dendritic architecture. / Polym Sci Part A: Polym Chem 38:2711-2727. 33. Hawker CJ, Wooley KL and Frechet JMJ (1993) Unimolecular micelles and globular amphiphiles: Dendritic macromolecules as novel recyclable solubilization agents. / Chem Soc Perkin Trans 1: Org Bio-Org Chem (1972-1999):1287-1297. 34. Tomalia DA, Berry V, Hall M and Hedstrand DM (1987) Starburst dendrimers. 4. Covalently fixed unimolecular assemblages reiminiscent of spheroidal micelles. Macromolecules 20:1164-1167. 35. Stevelmans S, Hest JCMv, Jansen JFGA, Van Boxtel DAFJ, de Berg EMM and Meijer EW (1996) Synthesis, characterization, and guest-host properties of inverted unimolecular dendritic micelles. J Am Chem Soc 118:7398-7399. 36. van Hest JCM, Delnoye DAP, Baars MWPL, van Genderen MHP and Meijer EW (1995) Polystyrene-dendrimer amphiphilic block copolymers with a generation-dependent aggregation. Science 268:1592-1595. 37. van Hest JCM, Elissen-Roman C, Baars MWPL, Delnoye DAP, Van Genderen MHP and Meijer EW (1995) Polystyrene-poly(propylene imine) dendrimer block copolymers: A new class of amphiphiles. Polym Mater Sci Eng 73:281-282. Polymer Micelles as Drug Carriers 83 38. Lorenz K, Muelhaupt R, Frey H, Rapp U and Mayer-Posner FJ (1995) Carbosilane-Based Dendritic Polyols. Macromolecules 28:6657-6661. 39. Gitsov I and Frechet JMJ (1996) Stimuli-responsive hybrid macromolecules: novel amphiphilic star copolymers with dendritic groups at the periphery. / Am Chem Soc 118:3785-3786. 40. Liu M, Kono K and Frechet JM (2000) Water-soluble dendritic unimolecular micelles: Their potential as drug delivery agents. / Control Rel 65:121-131. 41. Kojima C, Kono K, Maruyama K and Takagishi T (2000) Synthesis of polyamidoamine dendrimers having poly(ethylene glycol) grafts and their ability to encapsulate anticancer drugs. Bioconjug Chem 11:910-917. 42. Wang F, Bronich TK, Kabanov AV, Rauh RD and Roovers J (2005) Synthesis and evaluation of a star amphiphilic block copolymer from poly(epsilon-caprolactone) and poly(ethylene glycol) as a potential drug delivery carrier. Bioconjug Chem 16:397-405. 43. Heise A, Hedrick JL, Frank CW and Miller RD (1999) Starlike block copolymers with amphiphilic arms as models for unimolecular micelles. / Am Chem Soc 121:8647-8648. 44. Liu H, Jiang A, Guo J and Uhrich KE (1999) Unimolecular micelles: synthesis and characterization of amphiphilic polymer systems. / Polym Sci Part A: Polym Chem 37: 703-711. 45. Antoun S, Gohy JF and Jerome R (2001) Micellization of quaternized poly(2- (dimethylamino)ethyl methacrylate)-block-poly(methyl methacrylate) copolymers in water. Polym 42:3641-3648. 46. Jones M-C, Ranger M and Leroux J-C (2003) pH-sensitive unimolecular polymeric micelles: Synthesis a novel drug carrier. Bioconjug Chem 14:774-781. 47. Morton M, Helminiak TE, Gadkary SD and Bueche F (1962) Preparation and properties of monodisperse branched polystyrene. / Polym Sci 57:471^82. 48. Gauthier M, Li J and Dockendorff J (2003) Arborescent polystyrene-graft-poly(2- vinylpyridine) copolymers as unimolecular micelles. Synthesis from acetylated substrates. Macromolecules 36:2642-2648. 49. Iijima M, Nagasaki Y, Okada T, Kato M and Kataoka K (1999) Core-polymerized reactive micelles from heterotelechelic amphiphilic block copolymers. Macromolecules 32:1140-1146. 50. Kim J-H, Emoto K, Iijima M, Nagasaki Y, Aoyagi T, Okano T, Sakurai Y and Kataoka K (1999) Core-stabilized polymeric micelle as potential drug carrier: increased solubilization of taxol. Polym Adv Technol 10:647-654. 51. Guo A, Liu G and Tao J (1996) Star polymers and nanospheres from cross-linkable diblock copolymers. Macromolecules 29:2487-2493. 52. Won Y-Y, Davis HT and Bates FS (1999) Giant wormlike rubber micelles. Science 283:960-963. 53. Rapoport N (1999) Stabilization and activation of Pluronic micelles for tumor-targeted drug delivery. Coll SurfB: Biointerf16:93-111. 54. Thurmond KB, II, Huang H, Clark CG, Jr., Kowalewski T and Wooley KL (1999) Shell crosslinked polymer micelles: Stabilized assemblies with great versatility and potential. Coll SurfB: Biointerf16:45-54. 84 Batrakova et al. 55. Zhang Q, Remsen EE and Wooley KL (2000) Shell cross-linked nanoparticles containing hydrolytically degradable, crystalline core domains. / Am Chetn Soc 122: 3642-3651. 56. Buetuen V, Lowe AB, Billingham NC and Armes SP (1999) Synthesis of zwitterionic shell cross-linked micelles. / Am Chem Soc 121:4288-4289. 57. Bronich TK and Kabanov AV (2004) Novel block ionomer micelles with cross-linked ionic cores. Polym Prepr 45:384-385. 58. Bader H, Ringsdorf H and Schmidt B (1984) Water-soluble polymers in medicine. Angew Makromol Chem 123/124:457-485. 59. Duncan R (2003) The dawning era of polymer therapeutics. Nat Rev Drug Discov 2:347-360. 60. Bulmus V, Woodward M, Lin L, Murthy N, Stayton P and Hoffman A (2003) A new pH-responsive and glutathione-reactive, endosomal membrane-disruptive polymeric carrier for intracellular delivery of biomolecular drugs. / Control Rel 93: 105-120. 61. Veronese FM and Morpurgo M (1999) Bioconjugation in pharmaceutical chemistry. Farmaco 54:497-516. 62. D'Souza AJ and Topp EM (2004) Release from polymeric prodrugs: Linkages and their degradation. / Pharm Sci 93:1962-1979. 63. Bae Y, Fukushima S, Harada A and Kataoka K (2003) Design of environment-sensitive supramolecular assemblies for intracellular drug delivery: Polymeric micelles that are responsive to intracellular pH change. Angewandte Chemie, Int Edn 42:4640^643, S4640/1-S4640/11. 64. Bae Y, Nishiyama N, Fukushima S, Koyama H, Yasuhiro M and Kataoka K (2005) Preparation and biological characterization of polymeric micelle drug carriers with intracellular pH-triggered drug release property: Tumor permeability, controlled subcellular drug distribution, and enhanced in vivo antitumor efficacy. Bioconjug Chem 16:122-130. 65. Yokoyama M (1992) Block copolymers as drug carriers. Crit Rev Ther Drug Can Syst 9:213-248. 66. Yokoyama M, Sugiyama T, Okano T, Sakurai Y, Naito M and Kataoka K (1993) Analysis of micelle formation of an adriamycin-conjugated polyethylene glycol-poly(aspartic acid) block copolymer by gel permeation chromatography. Pharm Res 10:895-899. 67. Yokoyama M, Fukushima S, Uehara R, Okamoto K, Kataoka K, Sakurai Y and Okano T (1998) Characterization of physical entrapment and chemical conjugation of adriamycin in polymeric micelles and their design for in vivo delivery to a solid tumor. / Control Rel 50:79-92. 68. Kabanov AV, Chekhonin VP, Alakhov VY, Batrakova EV, Lebedev AS, Melik-Nubarov NS, Arzhakov SA, Levashov AV, Morozov GV, Severin ES and Kabanov VA (1989) The neuroleptic activity of haloperidol increases after its solubilization in surfactant micelles. Micelles as microcontainers for drug targeting. FEBS Lett 258:343-345. 69. La SB, Okano T and Kataoka K (1996) Preparation and characterization of the micelleforming polymeric drug indomethacin-incorporated poly(ethylene oxide)-poly(betabenzyl L-aspartate) block copolymer micelles. / Pharm Sci 85:85-90. Polymer Micelles as Drug Carriers 85 70. Yokoyama M, Satoh A, Sakurai Y, Okano T, Matsumura Y, Kakizoe T and Kataoka K (1998) Incorporation of water-insoluble anticancer drug into polymeric micelles and control of their particle size. / Control Rel 55:219-229. 71. Inoue T, Chen G, Nakamae K and Hoffman AS (1998) An AB block copolymer of oligo(methyl methacrylate) and poly(acrylic acid) for micellar delivery of hydrophobic drugs. / Control Rel 51:221-229. 72. Allen C, Han J, Yu Y, Maysinger D and Eisenberg A (2000) Polycaprolactone-bpoly( ethylene oxide) copolymer micelles as a delivery vehicle for dihydrotestosterone. / Control Rel 63:275-286. 73. Kabanov AV, Batrakova EV, Melik-Nubarov NS, Fedoseev NA, Dorodnich TY, Alakhov VY, Chekhonin VP, Nazarova IR and Kabanov VA (1992) A new class of drug carriers: Micelles of poly(oxyethylene)-poly(oxypropylene) block copolymers as microcontainers for drug targeting from blood in brain. / Control Rel 22:141-157. 74. Burt HM, Zhang X, Toleikis P, Embree L and Hunter WL (1999) Development of copolymers of poly(DL-lactide) and methoxypolyethylene glycol as micellar carriers of paclitaxel. Coll Surf, B: Biointerf 16:161-171. 75. Lavasanifar A, Samuel J and Kwon GS (2001) Micelles self-assembled from poly(ethylene oxide)-block-poly(N-hexyl stearate L-aspartamide) by a solvent evaporation method: Effect on the solubilization and haemolytic activity of amphotericin B. / Control Rel 77:155-160. 76. Lavasanifar A, Samuel J, Sattari S and Kwon GS (2002) Block copolymer micelles for the encapsulation and delivery of amphotericin B. Pharm Res 19:418-^22. 77. Hurter PN, Scheutjens JMHM and Hatton TA (1993) Molecular modeling of micelle formation and solubilization in block copolymer micelles. 1. A self-consistent meanfield lattice theory. Macromolecules 26:5592-5601. 78. Nagarajan R and Ganesh K (1996) Comparison of solubilization of hydrocarbons in (PEO-PPO) diblock versus (PEO-PPO-PEO) triblock copolymer micelles. / Coll Interf Sci 184:489-499. 79. Xing L and Mattice WL (1997) Strong solubilization of small molecules by triblockcopolymer micelles in selective solvents. Macromolecules 30:1711-1717. 80. Gadelle F, Koros WJ and Schechter RS (1995) Solubilization of aromatic solutes in block copolymers. Macromolecules 28:4883-4892. 81. Nagarajan R, Barry M and Ruckenstein E (1986) Unusual selectivity in solubilization by block copolymer micelles. Langmuir 2:210-215. 82. Liu J, Xiao Y and Allen C (2004) Polymer-drug compatibility: A guide to the development of delivery systems for the anticancer agent, ellipticine. / Pharm Sci 93:132-143. 83. Nagarajan R and Ganesh K (1989) Block copolymer self-assembly in selective solvents: Theory of solubilization in spherical micelles. Macromolecules 22:4312-4325. 84. Hurter PN and Hatton TA (1992) Solubilization of polycyclic aromatic hydrocarbons by poly(ethylene oxide-propylene oxide) block copolymer micelles: Effects of polymer structure. Langmuir 8:1291-1299. 85. Kozlov MY, Melik-Nubarov NS, Batrakova EV and Kabanov AV (2000) Relationship between pluronic block copolymer structure, critical micellization concentration 86 Batrakova etal. and partitioning coefficients of low molecular mass solutes. Macromolecules 33: 3305-3313. 86. Hurter PN, Scheutjens JMHM and Hatton TA (1993) Molecular modeling of micelle formation and solubilization in block copolymer micelles. 2. Lattice theory for monomers with internal degrees of freedom. Macromolecules 26:5030-5040. 87. Teng Y, Morrison ME, Munk P, Webber SE and Prochazka K (1998) Release kinetics studies of aromatic molecules into water from block polymer micelles. Macromolecules 31:3578-3587. 88. Choucair A and Eisenberg A (2003) Interfacial solubilization of model amphiphilic molecules in block copolymer micelles. J Am Chem Soc 125:11993-12000. 89. Kim SY, Shin IG, Lee YM, Cho CS and Sung YK (1998) Methoxy polyethylene glycol) and epsilon-caprolactone amphiphilic block copolymeric micelle containing indomethacin. II. Micelle formation and drug release behaviours. / Control Rel 51:13-22. 90. Kabanov AV, Vinogradov SV, Suzdaltseva YG and Alakhov VY (1995) Water-soluble block polycations as carriers for oligonucleotide delivery. Bioconjug Chem 6:639-643. 91. Harada A and Kataoka K (1995) Formation of polyion complex micelles in an aqueous milieu from a pair of oppositely-charged block copolymers with poly(ethylene glycol) segments. Macromolecules 28:5294-5299. 92. Kabanov VA and Kabanov AV (1998) Interpolyelectrolyte and block ionomer complexes for gene delivery: Physico-chemical aspects. Adv Drug Del Rev 30:49-60. 93. Solomatin SV, Bronich TK, Kabanov VA, Eisenberg A and Kabanov AV (2001) Block ionomer complexes: Novel environmentally responsive materials. Polym Prepr 42: 107-108. 94. Solomatin SV, Bronich TK, Eisenberg A, Kabanov VA and Kabanov AV (2003) Environmentally responsive nanoparticles from block ionomer complexes: Effects of pH and ionic strength. Langmuir 19:8069-8076. 95. Zhang G-D, Harada A, Nishiyama N, Jiang D-L, Koyama H, Aida T and Kataoka K (2003) Polyion complex micelles entrapping cationic dendrimer porphyrin: Effective photosensitizer for photodynamic therapy of cancer. / Control Rel 93:141-150. 96. Itaka K, Kanayama N, Nishiyama N, Jang W-D, Yamasaki Y, Nakamura K, Kawaguchi H and Kataoka K (2004) Supramolecular Nanocarrier of siRNA from PEGBased Block Catiomer Carrying Diamine Side Chain with Distinctive pKa Directed To Enhance Intracellular Gene Silencing. / Am Chem Soc 126:13612-13613. 97. Kabanov AV and Bronich TK (2002) Structure, dispersion stability and dynamics of DNA and polycation complexes, in Kim SW and Mahato R (eds.) Pharmaceutical Perspectives of Nucleic Acid-Based Therapeutics. Taylor & Francis; London, New York, pp. 164-189. 98. Yokoyama M, Okano T, Sakurai Y, Suwa S and Kataoka K (1996) Introduction of cisplatin into polymeric micelle. / Control Rel 39:351-356. 99. Nishiyama N, Yokoyama M, Aoyagi T, Okano T, Sakurai Y and Kataoka K (1999) Preparation and characterization of self-assembled polymer-metal complex micelle Polymer Micelles as Drug Carriers 87 from cis-dichlorodiammineplatinum(II) and poly(ethylene glycol)-poly(alpha.,.beta.- aspartic acid) block copolymer in an aqueous medium. Langmuir 15:377-383. 100. Holleb AI, Fink DJ and Murphy GP (eds.) (1991) American Cancer Society Textbook of Clinical Oncology. American Cancer Society: Atlanta, GA. 101. Sherman SE and Lippard SJ (1987) Structural aspects of platinum anticancer drug interactions with DNA. Chem Rev 87:1153-1181. 102. Nishiyama N and Kataoka K (2001) Preparation and characterization of size-controlled polymeric micelle containing cis-dichlorodiammineplatinum(II) in the core. / Control Rel 74:83-94. 103. Nishiyama N, Okazaki S, Cabral H, Miyamoto M, Kato Y, Sugiyama Y, Nishio K, Matsumura Y and Kataoka K (2003) Novel cisplatin-incorporated polymeric micelles can eradicate solid tumors in mice. Cancer Res 63:8977-8983. 104. Jang W-D, Nishiyama N, Zhang G-D, Harada A, Jiang D-L, Kawauchi S, Morimoto Y, Kikuchi M, Koyama H, Aida T and Kataoka K (2005) Supramolecular nanocarrier of anionic dendrimer porphyrins with cationic block copolymers modified with polyethylene glycol to enhance intracellular photodynamic efficacy. Angewandte Chemie, Int Edn 44:419^23. 105. Stapert HR, Nishiyama N, Jiang D-L, Aida T and Kataoka K (2000) Polyion complex micelles encapsulating light-harvesting ionic dendrimer zinc porphyrins. Langmuir 16:8182-8188. 106. Harada A and Kataoka K (1998) Novel polyion complex micelles entrapping enzyme molecules in the core: Preparation of narrowly-distributed micelles from lysozyme and poly(ethylene glycol)-poly(aspartic acid) block copolymer in aqueous medium. Macromolecules 31:288-294. 107. Kawamura A, Yoshioka Y, Harada A and Kono K (2005) Acceleration of enzymatic reaction of trypsin through the formation of water-soluble complexes with poly(ethylene glycol)-block-poly(a,b-aspartic acid). Biomacromolecules 6:627-631. 108. Harada A and Kataoka K (2003) Switching by pulse electric field of the elevated enzymatic reaction in the core of polyion complex micelles.} Am Chem Soc 125:15306-15307. 109. Bronich TK, Kabanov AV, Kabanov VA, Yu K and Eisenberg A (1997) Soluble complexes from poly(ethylene oxide)-block-polymethacrylate anions and N-alkylpyridinium cations. Macromolecules 30:3519-3525. 110. Bronich TK, Cherry T, Vinogradov SV, Eisenberg A, Kabanov VA and Kabanov AV (1998) Self-assmbly in mixtures of poly(ethylene oxide)-graft-poly(ethyleneimine) and alkyl sulfates. Langmuir 14:6101-6106. 111. Kabanov AV, Bronich TK, Kabanov VA, Yu K and Eisenberg A (1998) Spontaneous formation of vesicles from complexes of block ionomers and surfactants. / Am Chem Soc 120:9941-9942. 112. Bronich TK, Popov AM, Eisenberg A, Kabanov VA and Kabanov AV (2000) Effects of block length and structure of surfactant on self-assembly and solution behavior of block ionomer complexes. Langmuir 16:481^89. 113. Bronich TK, Ouyang M, Kabanov VA, Eisenberg A, Szoka FC, Jr. and Kabanov AV (2002) Synthesis of vesicles on polymer template. / Am Chem Soc 124:11872-11873. 88 Batrakova et al. 114. Bronich TK, Nehls A, Eisenberg A, Kabanov VA and Kabanov AV (1999) Novel drug delivery systems based on the complexes of block ionomers and surfactants of opposite charge. Col Surf B 16:243-251. 115. Solomatin SV, Bronich TK, Eisenberg A, Kabanov VA and Kabanov AV (2004) Colloidal stability of aqueous dispersions of block ionomer complexes: Effects of temperature and salt. Langmuir 20: 2066-2068. 116. Pinzani V, Bressolle F, Haug IJ, Galtier M, Blayac JP and Balmes P (1994) Cisplatininduced renal toxicity and toxicity-modulating strategies: A review. Cancer Chemother Pharmacol 35:1-9. 117. Weinstein DM, Mihm MJ and Bauer JA (2000) Cardiac peroxynitrite formation and left ventricular dysfunction following doxorubicin treatment in mice. / Pharmacol Exp Ther 294:396-401. 118. Kwon GS and Kataoka K (1995) Block copolymer micelles as long-circulating drug vehicles. Adv Drug Del Rev 16:295-309. 119. Batrakova EV, Li S, Li Y, Alakhov VY, Elmquist WF and Kabanov AV (2004) Distribution kinetics of a micelle-forming block copolymer Pluronic P85. / Control Rel 100:389-397. 120. Fraser JR, Laurent TC, Pertoft H and Baxter E (1981) Plasma clearance, tissue distribution and metabolism of hyaluronic acid injected intravenously in the rabbit. Biochem J 200:415^124. 121. Kissel M, Peschke P, Subr V, Ulbrich K, Schuhmacher J, Debus J and Friedrich E (2001) Synthetic macromolecular drug carriers: Biodistribution of poly[(N-2- hydroxypropyl)methacrylamide] copolymers and their accumulation in solid rat tumors. PDAJPharm Sci Technol 55:191-201. 122. Matsumura Y and Maeda H (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: Mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res 46:6387-6392. 123. Maeda H (2001) The enhanced permeability and retention (EPR) effect in tumor vasculature: The key role of tumor-selective macromolecular drug targeting. Adv Enzyme Regul 41:189-207. 124. Giblin MF, Veerendra B and Smith CJ (2005) Radiometallation of receptor-specific peptides for diagnosis and treatment of human cancer. In Vivo 19:9-29. 125. Rini BI (2005) VEGF-targeted therapy in metastatic renal cell carcinoma. Oncologist 10:191-197. 126. Lin MZ, Teitell MA and Schiller GJ (2005) The evolution of antibodies into versatile tumor-targeting agents. Clin Cancer Res 11:129-138. 127. Levi FA, Hrushesky WJ, Halberg F, Langevin TR, Haus E and Kennedy BJ (1982) Lethal nephrotoxicity and hematologic toxicity of cis-diamminedichloroplatinum ameliorated by optimal circadian timing and hydration. Eur } Cancer Clin Oncol 18:471-477. 128. Greish K, Sawa T, Fang J, Akaike T and Maeda H (2004) SMA-doxorubicin, a new polymeric micellar drug for effective targeting to solid tumours. / Control Rel 97: 219-230. 129. Lee ES, Na K and Bae YH (2005) Doxorubicin loaded pH-sensitive polymeric micelles for reversal of resistant MCF-7 tumor. / Control Rel 103:405-418. Polymer Micelles as Drug Carriers 89 130. Alakhov V, Klinski E, Li S, Pietrzynski G, Venne A, Batrakova E, Bronitch T and Kabanov AV (1999) Block copolymer-based formulation of doxorubicin. From cell screen to clinical trials. Coll SurfB: Biointerf'16:113-134. 131. Kim SC, Kim DW, Shim YH, Bang JS, Oh HS, Wan Kim S and Seo MH (2001) In vivo evaluation of polymeric micellar paclitaxel formulation: Toxicity and efficacy. / Control Rel 72:191-202. 132. Torchilin VP, Lukyanov AN, Gao Z and Papahadjopoulos-Sternberg B (2003) Immunomicelles: Targeted pharmaceutical carriers for poorly soluble drugs. Proc Natl Acad Sci USA 100:6039-6044. 133. Taillefer J, Brasseur N, van Lier JE, Lenaerts V, Le Garrec D and Leroux JC (2001) In vitro and in vivo evaluation of pH-responsive polymeric micelles in a photodynamic cancer therapy model. / Pharm Pharmacol 53:155-166. 134. Le Garrec D, Taillefer J, Van Lier JE, Lenaerts V and Leroux JC (2002) Optimizing pHresponsive polymeric micelles for drug delivery in a cancer photodynamic therapy model. / Drug Targ 10:429-437. 135. Yokoyama M, Okano T, Sakurai Y, Fukushima S, Okamoto K and Kataoka K (1999) Selective delivery of adriamycin to a solid tumor using a polymeric micelle carrier system. / Drug Targ 7:171-186. 136. Greish K, Nagamitsu A, Fang J and Maeda H (2005) Copoly(styrene-maleic acid)- pirarubicin micelles: High tumor-targeting efficiency with little toxicity. Bioconjug Chem 16:230-236. 137. Nasongkla N, Shuai X, Ai H, Weinberg BD, Pink J, Boothman DA and Gao J (2004) cRGD-functionalized polymer micelles for targeted doxorubicin delivery. Angew Chem Int Ed Engl 43:6323-6327. 138. Paulos CM, Turk MJ, Breur GJ and Low PS (2004) Folate receptor-mediated targeting of therapeutic and imaging agents to activated macrophages in rheumatoid arthritis. Adv Drug Del Rev 56:1205-1217. 139. Yoo HS and Park TG (2004) Folate receptor targeted biodegradable polymeric doxorubicin micelles. / Control Rel 96:273-283. 140. Allen TM (2002) Ligand-targeted therapeutics in anticancer therapy. Nat Rev Cancer 2:750-763. 141. Gao Z, Lukyanov AN, Chakilam AR and Torchilin VP (2003) PEG-PE/ phosphatidylcholine mixed immunomicelles specifically deliver encapsulated taxol to tumor cells of different origin and promote their efficient killing. / Drug Targ 11:87-92. 142. Tannock IF and Rotin D (1989) Acid pH in tumors and its potential for therapeutic exploitation. Cancer Res 49:4373^1384. 143. Kataoka K, Matsumoto T, Yokoyama M, Okano T, Sakurai Y, Fukushima S, Okamoto K and Kwon GS (2000) Doxorubicin-loaded poly(ethylene glycol)-poly(beta-benzyl-Laspartate) copolymer micelles: Their pharmaceutical characteristics and biological significance. / Control Rel 64:143-153. 144. Yoo HS, Lee EA and Park TG (2002) Doxorubicin-conjugated biodegradable polymeric micelles having acid-cleavable linkages. / Control Rel 82:17-27. 90 Batrakova et al. 145. Hruby M, Konak C and Ulbrich K (2005) Polymeric micellar pH-sensitive drug delivery system for doxorubicin. / Control Rel 103:137-148. 146. Kabanov AV, Bronich TK, Kabanov VA, Yu K and Eisenberg A (1996) Soluble stoichiometric complexes from poly(N-ethyl-4-vinylpyridinium) cations and poly(ethylene oxide)-Wocfc-polymethacrylate anions. Macromolecules 29:6797-6802. 147. Lee ES, Shin HJ, Na K and Bae YH (2003) Poly(L-histidine)-PEG block copolymer micelles and pH-induced destabilization. / Control Rel 90:363-374. 148. Alakhov VY, Moskaleva EY, Batrakova EV and Kabanov AV (1996) Hypersensitization of multidrug resistant human ovarian carcinoma cells by pluronic P85 block copolymer. Bioconjug Chem 7:209-216. 149. Venne A, Li S, Mandeville R, Kabanov Aand Alakhov V (1996) Hypersensitizing effect of pluronic L61 on cytotoxic activity, transport, and subcellular distribution of doxorubicin in multiple drug- resistant cells. Cancer Res 56:3626-3629. 150. Batrakova EV, Dorodnych TY, Klinskii EY, Kliushnenkova EN, Shemchukova OB, Goncharova ON, Arjakov SA, Alakhov VY and Kabanov AV (1996) Anthracycline antibiotics non-covalently incorporated into the block copolymer micelles: In vivo evaluation of anti cancer activity. Br } Cancer 74:1545-1552. 151. Kabanov AV, Batrakova EV and Alakhov VY (2002) Pluronic block copolymers for overcoming drug resistance in cancer. Adv Drug Deliv Rev 54:759-779. 152. Batrakova EV, Lee S, Li S, Venne A, Alakhov V and Kabanov A (1999) Fundamental relationships between the composition of pluronic block copolymers and their hypersensitization effect in MDR cancer cells. Pharm Res 16:1373-1379. 153. Miller DW, Batrakova EV and Kabanov AV (1999) Inhibition of multidrug resistanceassociated protein (MRP) functional activity with pluronic block copolymers. Pharm Res 16:396-401. 154. Batrakova EV, Li S, Alakhov VY, Elmquist WF, Miller DW and Kabanov AV (2003) Sensitization of cells overexpressing multidrug-resistant proteins by pluronic P85. Pharm Res 20:1581-1590. 155. Minko T, Batrakova E, Li S, Li Y, Pakunlu R, Alakhov V and Kabanov A (2005) Pluronic block copolymers alter apoptotic signal transduction of doxorubicin in drug-resistant cancer cells. / Control Rel. 156. Batrakova EV, Han HY, Miller DW and Kabanov AV (1998) Effects of pluronic P85 unimers and micelles on drug permeability in polarized BBMEC and Caco-2 cells. Pharm Res 15:1525-1532. 157. Evers R, Kool M, Smith AJ, van Deemter L, de Haas M and Borst P (2000) Inhibitory effect of the reversal agents V-104, GF120918 and Pluronic L61 on MDR1 Pgp-, MRP1- and MRP2-mediated transport. Br J Cancer 83:366-374. 158. Batrakova EV, Miller DW, Li S, Alakhov VY, Kabanov AV and Elmquist WF (2001) Pluronic P85 enhances the delivery of digoxin to the brain: In vitro and in vivo studies. / Pharmacol Exp Ther 296:551-557. 159. Miller DW, Batrakova EV, Waltner TO, Alakhov V and Kabanov AV (1997) Interactions of pluronic block copolymers with brain microvessel endothelial cells: Evidence of two potential pathways for drug absorption. Bioconjug Chem 8:649-657. Polymer Micelles as Drug Carriers 91 160. Batrakova EV, Li S, Elmquist WF, Miller DW, Alakhov VY and Kabanov AV (2001) Mechanism of sensitization of MDR cancer cells by Pluronic block copolymers: Selective energy depletion. Br J Cancer 85:1987-1997. 161. Batrakova EV, Li S, Vinogradov SV, Alakhov VY, Miller DW and Kabanov AV (2001) Mechanism of pluronic effect on P-glycoprotein efflux system in blood-brain barrier: Contributions of energy depletion and membrane fluidization. / Pharmacol Exp Ther 299:483-493. 162. Batrakova EV, Li S, Alakhov VY, Miller DW and Kabanov AV (2003) Optimal structure requirements for Pluronic block copolymers in modifying P-glycoprotein drug efflux transporter activity in bovine brain microvessel endothelial cells. / Pharmacol Exp Ther 304:845-854. 163. Rapoport N, Marin AP and Timoshin AA (2000) Effect of a polymeric surfactant on electron transport in HL-60 cells. Arch Biochem Biophys 384:100-108. 164. Kabanov AV, Batrakova EV and Alakhov VY (2003) An essential relationship between ATP depletion and chemosensitizing activity of Pluronic block copolymers. / Control Rel 91:75-83. 165. Batrakova EV, Li S, Miller DW and Kabanov AV (1999) Pluronic P85 increases permeability of a broad spectrum of drugs in polarized BBMEC and Caco-2 cell monolayers. Pharm Res 16:1366-1372. 166. Batrakova EV, Zhang Y, Li Y, Li S, Vinogradov SV, Persidsky Y, Alakhov V, Miller DW and Kabanov AV (2004) Effects of Pluronic P85 on GLUT1 and MCT1 transporters in the blood brain barrier. Pharm Res in press. 167. Kabanov AV, Batrakova EV and Miller DW (2003) Pluronic((R)) block copolymers as modulators of drug efflux transporter activity in the blood-brain barrier. Adv Drug Del Rev 55:151-164. 168. Kabanov AV and Batrakova EV (2004) New technologies for drug delivery across the blood brain barrier. Curr Pharm Des 10:1355-1363. 169. Kwon GS (2003) Polymeric micelles for delivery of poorly water-soluble compounds. Crit Rev Ther Drug Carr Syst 20:357-403. 170. Adams ML and Kwon GS (2003) Relative aggregation state and hemolytic activity of amphotericin B encapsulated by poly(ethylene oxide)-block-poly(N-hexyl- L-aspartamide)-acyl conjugate micelles: Effects of acyl chain length. / Control Rel 87:23-32. 171. Adams ML, Andes DR and Kwon GS (2003) Amphotericin B encapsulated in micelles based on poly(ethylene oxide)-block-poly(L-amino acid) derivatives exerts reduced in vitro hemolysis but maintains potent in vivo antifungal activity. Biomacromolecules 4:750-757. 172. Croy SR and Kwon GS (2004) The effects of Pluronic block copolymers on the aggregation state of nystatin. / Control Rel 95:161-171. 173. Jagannath C, Sepulveda E, Actor JK, Luxem F, Emanuele MR and Hunter RL (2000) Effect of poloxamer CRL-1072 on drug uptake and nitric-oxide-mediated killing of Mycobacterium avium by macrophages. Immunopharmacology 48: 185-197. 92 Batrakova et al. 174. Jagannath C, Emanuele MR and Hunter RL (2000) Activity of poloxamer CRL- 1072 against drug-sensitive and resistant strains of Mycobacterium tuberculosis in macrophages and in mice. Int J Antimicrob Agents 15:55-63. 175. Jagannath C, Emanuele MR and Hunter RL (1999) Activities of poloxamer CRL-1072 against Mycobacterium avium in macrophage culture and in mice. Antimicrob Agents Chemother 43:2898-2903. 176. Jagannath C, Wells A, Mshvildadze M, Olsen M, Sepulveda E, Emanuele M, Hunter RL, Jr. and Dasgupta A (1999) Significantly improved oral uptake of amikacin in FVB mice in the presence of CRL-1605 copolymer. Life Sci 64:1733-1738. 177. Torchilin VP (2002) PEG-based micelles as carriers of contrast agents for different imaging modalities. Adv Drug Del Rev 54:235-252. 178. Trubetskoy VS, Frank-Kamenetsky MD, Whiteman KR, Wolf GL and Torchilin VP (1996) Stable polymeric micelles: Lymphangiographic contrast media for gamma scintigraphy and magnetic resonance imaging. Acad Radiol 3:232-238. 179. Trubetskoy VS, Gazelle GS, Wolf GL and Torchilin VP (1997) Block-copolymer of polyethylene glycol and polylysine as a carrier of organic iodine: Design of longcirculating particulate contrast medium for X-ray computed tomography. / Drug Targ 4:381-388. 180. Weissig V, Whiteman KR and Torchilin VP (1998) Accumulation of protein-loaded longcirculating micelles and liposomes in subcutaneous Lewis lung carcinoma in mice. Pharm Res 15:1552-1556. 181. Katayose S and Kataoka K (1997) Water-soluble polyion complex associates of DNA and poly(ethylene glycol)-poly(L-lysine) block copolymer. Bioconj Chem 8:702-707. 182. Wolfert MA, Schacht EH, Toncheva V, Ulbrich K, Nazarova O and Seymour LW (1996) Characterization of vectors for gene therapy formed by self-assembly of DNA with synthetic block co-polymers. Hum Gene Ther 7:2123-2133. 183. Vinogradov SV, Bronich TK and Kabanov AV (1998) Self-assembly of polyaminepoly( ethylene glycol) copolymers with phosphorothioate oligonucleotides. Bioconjug Chem 9:805-812. 184. Itaka K, Harada A, Nakamura K, Kawaguchi H and Kataoka K (2002) Evaluation by fluorescence resonance energy transfer of the stability of nonviral gene delivery vectors under physiological conditions. Biomacromolecules 3:841-845. 185. Roy S, Zhang K, Roth T, Vinogradov S, Kao RS and Kabanov A (1999) Reduction of fibronectin expression by intravitreal administration of antisense oligonucleotides. Nat Biotechnol 17:476-479. 186. Ogris M, Steinlein P, Kursa M, Mechtler K, Kircheis R and Wagner E (1998) The size of DNA/transferrin-PEI complexes is an important factor for gene expression in cultured cells. Gene Ther 5:1425-1433. 187. Oupicky D, Ogris M, Howard KA, Dash PR, Ulbrich K and Seymour LW (2002) Importance of lateral and steric stabilization of polyelectrolyte gene delivery vectors for extended systemic circulation. Mol Ther 5:463-472. Polymer Micelles as Drug Carriers 93 188. Harada-Shiba M, Yamauchi K, Harada A, Takamisawa I, Shimokado K and Kataoka K (2002) Polyion complex micelles as vectors in gene therapy-pharmacokinetics and in vivo gene transfer. Gene Ther 9:407-414. 189. Choi YH, Liu F> ParkJS and Kim SW (1998) Lactose-poly(ethylene glycol)-grafted poly- L-lysine as hepatoma cell- tapgeted gene carrier. Bioconjug Chem 9:708-718. 190. Vinogradov S, Batrakova E, Li S and Kabanov A (1999) Polyion complex micelles with protein-modified corona for receptor-mediated delivery of oligonucleotides into cells. Bioconjug Chem 10:851-860. 191. Ward CM, Pechar M, Oupicky D, Ulbrich K and Seymour LW (2002) Modification of pLL/DNA complexes with a multivalent hydrophilic polymer permits folate-mediated targeting in vitro and prolonged plasma circulation in vivo. } Gene Med 4:536-547. 192. Nguyen HK, Lemieux P, Vinogradov SV, Gebhart CL, Guerin N, Paradis G, Bronich TK, Alakhov VY and Kabanov AV (2000) Evaluation of polyether-polyethyleneimine graft copolymers as gene transfer agents. Gene Ther 7:126-138. 193. Gebhart CL, Sriadibhatla S, Vinogradov S, Lemieux P, Alakhov V and Kabanov AV (2002) Design and formulation of polyplexes based on pluronic-polyethyleneimine conjugates for gene transfer. Bioconjug Chem 13:937-944. 194. Belenkov AI, Alakhov VY, Kabanov AV, Vinogradov SV, Panasci LC, Monia BP and Chow TY (2004) Polyethyleneimine grafted with pluronic P85 enhances Ku86 antisense delivery and the ionizing radiation treatment efficacy in vivo. Gene Ther 11:1665-1672. 195. Yokoyama M, Miyauchi M, Yamada N, Okano T, Sakurai Y, Kataoka K and Inoue S (1990) Characterization and anticancer activity of the micelle-forming polymeric anticancer drug adriamycin-conjugated poly(ethylene glycol)-poly(aspartic acid) block copolymer. Cancer Res 50:1693-1700. 196. Nakanishi T, Fukushima S, Okamoto K, Suzuki M, Matsumura Y, Yokoyama M, Okano T, Sakurai Y and Kataoka K (2001) Development of the polymer micelle carrier system for doxorubicin. / Control Rel 74:295-302. 197. Danson S, Ferry D, Alakhov V, Margison J, Kerr D, Jowle D, Brampton M, Halbert G and Ranson M (2004) Phase I dose escalation and pharmacokinetic study of pluronic polymer-bound doxorubicin (SP1049C) in patients with advanced cancer. Br } Cancer 90:2085-2091. 198. Valle JW, Lawrance J, Brewer J, Clayton A, Corrie P, Alakhov V and Ranson M (2004) A phase II, window study of SP1049C as first-line therapy in inoperable metastatic adenocarcinoma of the oesophagus. 2004 ASCO Annual Meeting Vol. Abstract No: 4195. 199. Kim TY, Kim DW, Chung JY, Shin SG, Kim SC, Heo DS, Kim NK and Bang YJ (2004) Phase I and pharmacokinetic study of Genexol-PM, a cremophor-free, polymeric micelle-formulated paclitaxel, in patients with advanced malignancies. Clin Cancer Res 10:3708-3716. This page is intentionally left blank 6 Vesicles Prepared from Synthetic Amphiphiles — Polymeric Vesicles and Niosomes Ijeoma Florence Uchegbu and Andreas G. Schatzlein 1. Introduction This chapter will examine what is known about vesicles prepared from synthetic amphiphiles and will encompass a review of the data published on polymeric vesicles and non-ionic surfactant vesicles (niosomes). Schematic representations of the molecular arrangements in these systems are as depicted in Fig. 1. Examples of drug delivery applications will also be presented. Vesicular systems arise when amphiphilic molecules self assemble in aqueous media in an effort to reduce the high energy interaction between the hydrophobic portion of the amphiphile and the aqueous disperse phase, and maximize the low energy interaction between the hydrophilic head group and the disperse phase (Fig. 1). These self assemblies reside in the nanometre to micrometre size domain. Excellent reviews exist on the self assembly of amphiphiles/16 and hence this topic will not be dealt with in great detail here. Vesicles are important pharmaceutical systems, especially as liposomes, the result of phospholipid self assembly,19 are licensed for the clinical delivery of anti cancer drugs.21 It is thus possible that the vesicles described here may be incorporated into licensed medicines at some point in future. 95 96 Uchegbu & Schatzlein t % If 111 II mnme&m Self assembling polymerisable monomers Polymerisation! (a) Polymerised vesicles mm isfi? M# V * » » * ^ *^*^* i«s m it (b) Self assembling amphiphilic polymers Q U80% favors dense nanoparticles, while a polydactic acid) fraction of 58-80% favors bilayer vesicle assemblies, and a polydactic acid) fraction of less than 50% favors the production of micellar self assemblies.31 The sizes of the vesicle and dense nanoparticle assemblies formed from amphiphilic poly(ethylenimines) are also dependent on polymer levels of hydrophobic modification (mole % cetylation) and the relationships shown in Eqs. (1) and (2) have been developed,18 dv = 1.95Ct + 139 (1) dn = 2.31Ct + 5.6 (2) where dv = vesicle z-average mean hydrodynamic diameter, Ct = mole% cetylation (number of cetyl groups per 100 monomer units), and dn — nanoparticle z-average mean hydrodynamic diameter. The molecular weight of the polymer is also an important factor to consider when choosing vesicle forming polymers. The importance of this parameter has been demonstrated with the poly(L-lysine) vesicle system20 [e.g. Compound 6, Fig. 3(a)]. With these amphiphiles a vesicle formation index (F') has been computed: F' = - ^ = (3) LVDP where H = mole% unreacted L-lysine units, L — mole% L-lysine units substituted with palmitic acid and DP = the degree of polymerisation of the polymer. An F' value in excess of 0.168 is necessary for vesicle formation.20 Additionally, not only does the molecular weight of the polymer impact on vesicle formation, but it is also a direct controller of the vesicle mean size; the relationship shown in Eq. (4) has been developed for the palmitoyl glycol chitosan system,11 VMW = 0.782dv + 107 (4) Vesicles Prepared from Synthetic Amphiphiles 101 where MW = polymer molecular weight, and dv = vesicle z-average mean hydrodynamic diameter. 2.3. Block copolymers Block copolymer vesicles, termed "polymersomes" are fairly new discoveries, being first reported in the 1990s.32 Polymersomes have been prepared from a variety of block copolymers, some examples of which are given in Fig. 4. There is a clear relationship between the hydrophobic content of polymers and self assembly. Low levels of hydrophobicity (less than 50% of the polymer consisting of hydrophobic HO, X N^- J5"H HN' 9 Fig. 4. Examples of some vesicle forming block copolymers Compound 7,1 Compound 8,7 and Compound 9.13 102 Uchegbu & Schatzlein moieties) favors the formation of micelles33 and intermediate levels of hydrophobicity (50-80%) favors the formation of bilayer vesicles.31,33'34 For the self assembly of block copolymers, it has been established that generally the critical packing parameter (CPP): CPP = ^ (5) al should approach unity for vesicular self assemblies to prevail,24 where v = volume of the hydrophobic block, 1 — length of the hydrophobic block and a = the area of the hydrophilic block. Vesicle sizes are varied and range from tens of nanometres35 to tens of microns.36 Polymersome membranes are 8-21 nm thick; 2-5 times thicker than the 4nm membrane thickness displayed by conventional low molecular weight amphiphiles.16'27,31'34,35 The thickness of the membrane is determined by the degree of polymerization in the hydrophobic block34 and these extra thick membranes confer, on the vesicle, exceptional stability to soluble surfactantS24 and mechanical stress.24'27,37'38 With these vesicles, there is an asymmetric distribution of the polymers in the inner and outer leaflets of the bilayer and polymers with a large hydrophilic chain length are preferentially localized to the exterior leaflet and vice versa.39 Preferred residence in the outer leaflet is favored by the more hydrophilic polymers, because the greater repulsion between the longer hydrophilic corona molecules on the outer leaflet stabilize the vesicle curvature.39 Vesicle stability is a desirable characteristic for pharmaceutical vesicles and as such, a great deal of effort has been expended on producing stable systems. As the drive for nanomedicines (medicines incorporating functional nanoparticles) grows, stability issues will need to be adequately addressed to ensure the widespread adoption of such systems. In actual fact, the early workers in the polymeric vesicle field were primarily driven by this need to produce stable drug carriers. Extremely stable systems are possible on polymerization of block copolymers subsequent to self assembly. Poly(ethylene oxide)-WocA:-poly[3-(trimethoxysilyl)propyl methacrylate] copolymer vesicles in water, methanol, triethylamine mixtures produced polymerized polymersomes that are stable for up to one year.40 Triethylamine hydrolyzes the trimethoxysilyl groups and then catalyzes their polycondensation to yield an extremely stable hydrophobic polysilsesquioxane core.40,41 Additionally, poly(ethylene oxide)-Wocfc-poly(butadiene) vesicles on cross linking produce vesicles which are organic solvent resistant.42 2.4. Preparing vesicles from self-assembling polymers Polymeric vesicles are relatively simple to prepare. The input of energy is achieved in the laboratory by probe sonication of the amphiphilic polymer in the disperse Vesicles Prepared from Synthetic Amphiphiles 103 phase.1120 However, clearly the energy required for self assembly is not trivial as vesicles are not easily formed by hand shaking, unlike low molecular weight surfactant formulations.4 Vesicles once formed are morphologically stable for months11 and may be loaded with hydrophilic43-45 and hydrophobic [see Fig. 6(b) below] solutes, by probe sonicating in the presence of such solutes. Commercially, it is envisaged that polymeric vesicles may be fabricated by microfluidization and high pressure homogenization techniques. 2.5. Self assembling polymerizable monomers Polymerized vesicles may also be prepared by utilizing self assembling polymerizable amphiphiles, followed by the polymerization of the resulting vesicular self assembly (Fig. 1). Examples of some polymerizable vesicle forming monomers are shown in Fig. 5. This method of producing polymerised vesicles is the oldest form of polymeric vesicle technology.12,46 HO-P-OH l O HO-P-OH O 13 Fig. 5. Polymerizable vesicle forming monomers used to make polymerized vesicles by Jung and others (Compound 10),5 Cho and others (Compound ll),8 Hub and others (Compound 12)12 and Bader and others (Compound 13).15 104 Uchegbu & Schatzlein Polymerized vesicles prepared using polymerized self assembling monomers are essentially polymer shells and it is unclear how much of the bilayer assembly actually survives the polymerization step. The advantage, however, is that they are extremely stable, resisting degradation by detergents47-49 or organic solvents.8'48,50,51 They are also less leaky,50 thermostable,52 and because the vesicle forming components are kinetically trapped by the polymerization process, they have improved colloidal stability.8 A major advantage of these nanosystems is that they may be isolated as dry powders which are readily dispersible in water to give 50-100 nm particles;48 thus potentially allowing the formulation of solid vesicle dosage forms. Polymerization involves fairly reactive species and hence vesicles are best prepared prior to drug loading, which may be a limitation. 3. Polymeric Vesicle Drug Delivery Applications Polymeric vesicles, which are the focus of this chapter, exist in two main varieties as illustrated in Fig. 1. These technologies are suitable candidates for the development of robust, controllable and responsive nanomedicine drug carriers. 3.1. Drug targeting Poly(oxyethylene) amphiphiles, when incorporated into liposomal26 and niosomal6 bilayers, prolong vesicle circulation and facilitate tumor targeting,6'53 due to the leaky nature of the poorly developed tumor vascular endothelium.54 Only 10 mole % poly(ethylene oxide) — lipid amphiphiles may be incorporated into liposomes55 or niosomes,56'57 without a loss of vesicle integrity due to the preferred tendency of the hydrophilic poly(oxyethylene) amphiphiles to form micelles. Polymersomes composed of poly(ethylene oxide)-Wocfc-polybutadiene or poly(ethylene oxide)-Wocfc-poly(ethylethylene), in which the entire vesicle surface is covered with the poly(ethylene oxide) coat, have been studied as long circulating nanocarriers for drug delivery58 The circulation time of poly(ethylene oxide) polymersomes is directly dependent on the length of the poly(ethylene oxide) block and polymersome half lives of up to 28 hrs have been recorded in rats with a poly(ethylene oxide) degree of polymerization of 50.58 This half life compares favorably with a half life of 14 hrs recorded for poly(oxyethylene) coated liposomes.59 It is assumed that the 100% surface coverage of the polymeric vesicles is responsible for the reduced clearance of these polymersomes from the blood.38 The long half life of these polymersomes makes them excellent candidates for the development of anti tumor medicines. Furthermore, drug release may be controlled in the polymersomes by controlling the hydrolysis rate of the hydrophobic blocks.31 This has been demonstrated Vesicles Prepared from Synthetic Amphiphiles 105 with poly(L-lactic acid)-fr/ocfc-poly(ethylene glycol) and poly(caprolactone)-Wod> poly(ethylene glycol) vesicles.31 Hydrolysis of the hydrophobic block causes the polymer to move from a vesicular to a micellar assembly, as the overall level of hydrophobic content diminishes, and this in turn leads to drug release.31 Hydrolysis rates and implicitly release rates may be controlled by varying the relative level of the hydrophobic blocks. Carbohydrate polymeric vesicles may also be used as drug targeting agents. Vesicles prepared from glycol chitosan vesicles improve the intracellular delivery of hydrophilic macromolecules44 and anti cancer drugs,45 the latter is achieved with the help of a transferrin ligand attached to the surface of the vesicle. 3.2. Gene delivery Poly(L-lysine) based vesicles, prepared from Compound 6 [Fig. 3(a)] have been used for gene delivery,29,60 as these vesicles are less toxic than unmodified poly(L-lysine) and produce higher levels of gene transfer (Table l).29 The production of polymeric vesicles and the resultant reduction in cytotoxicity enables poly(L-lysine) to be used in in vivo gene, as the unmodified polymer is too toxic for in vivo use. When the targeting ligand, galactose, was bound to the distal ends of the poly(oxyethylene) chains, gene expression was increased in HepG2 cells in vitro.60 However, in vivo targeting to the liver hepatocytes was not achieved with these systems.60 A similar procedure with the poly(ethylenimine) vesicles prepared using Compound 5 [Fig. 3(a)] also resulted in a reduction in the cytotoxicity of the polymer (Table l),17 although in this case, the poly(ethylenimine) vesicles were not as efficient gene transfer agents as the free polymer. Table 1 Biological Activity of poly(ethylenimine)17 and poly(L-lysine)29 Vesicles. Polymer A431 cells A549 IC50 Gene Transfer IC50 Gene Transfer (AtgmL-1) Relative to Parent (jiigmL-1) Relative to Parent Polymer Polymer Poly(ethylenimine) 1.9 1 5.2 1 Polymer 5 (Fig. 6(a)) 16.9 0.2 12.6 0.08 Polymer 5, cholesterol 15.9 0.2 11 0.08 vesicles 2:1 (gg_1) Poly(L-lysine) 7 1 7 1 Polymer 6 (Fig. 6(a)) 74 7.8 63 2.3 106 Uchegbu & Schatzlein 3.3. Responsive release The ultimate goal of all drug delivery efforts is the simple fabrication of responsive systems that are capable of delivering precise quantities of their pay load in response to physiological or more commonly pathological stimuli. Pre-programmable pills, implants and injectables are so far merely the unobtainable ideal, however, polymeric systems have been fabricated with responsive capability and it is possible that in the future, these may be fine tuned to produce truly intelligent and dynamic drug delivery devices or systems. The various environmental stimuli that may be used to trigger the release of encapsulated drug are outlined below and examples are given of existing developments in the area. However, in addition to the areas covered below, it may be possible in future for pathology specific molecules to interact with polymeric vesicles to trigger release. 3.3.1. pH Diblock polypeptides, in which the hydrophilic block consists of ethylene glycol derivatised amino acids (L-lysine), and the hydrophobic block consists of poly (L-leucine), form pH responsive vesicles which disaggregate at low pH, providing the level of L-leucine and polymer chain length is maintained within defined limits of about 12-25 mole% and the polymer has a degree of polymerization of less than 200.13 These L-lysine based systems may be applied to facilitate endosome specific release. 3.3.2. Enzymatic Vesicles which release their contents in the presence of an enzyme may be formed by loading polymeric vesicles with an enzyme activated prodrug (Fig. 6). The particulate nature of the drug delivery system should allow the drug to accumulate in tumors, for example, where it may then be activated by an externally applied enzyme in a similar manner to the antibody directed enzyme prodrug therapeutic strategy. The antibody directed enzyme prodrug therapeutic strategy enables an enzyme to be homed to tumors using antibodies followed by the application of an enzyme activated prodrug.61 Alternatively, a membrane bound enzyme may be used to control and ultimately prolong the activity of either an entrapped hydrophilic drug (entrapped in the vesicle aqueous core) or an entrapped hydrophobic drug (entrapped in the vesicle membrane) as illustrated in Fig. 6. It is possible that the enzyme may be chosen such that it is activated in the presence of pathology specific molecules, thus achieving pathology responsive and localized drug activity. Vesicles Prepared from Synthetic Amphiphiles 107 2.5 2.0 ii CD 1.5 1.0' <2 0.5- 0.0- I 1 NM/ -O— vesicle bound enzyme + external substrate - •— external enzyme + vesicle loaded substrate -A— control solution + substrate --? ^""V • • • • • W»>* 0 20 40 60 80 A • Enzy m e Time(min) ^ W A Water soluble Substrate Fig. 6(a). Enzyme activated polymeric vesicles. Vesicles bearing membrane bound enzyme (i) were formed by probe sonicating Compound 2 [Fig. 3(a)], cholesterol, N-biotinylated dipalmitoyl phosphatidyl ethanolamine (8: 4: 1 gg"1) in neutral phosphate buffer (2mL), isolation of the vesicles by ultracentrifugation (150,000 g), redispersion in a similar volume of neutral phosphate buffer and incubation of the vesicles with /S-galactosidase streptavidin (3 U). Membrane bound enzyme (0.2 mL) was then incubated with o-nitrophenyl-/J-Dgalactoside (2.1 mM, 2 mL) and the absorbance monitored (X = 410 ran). The control solution contained similar levels of substrate (o-nitrophenyl-jS-D-galactoside) but no enzyme. Vesicles encapsulating O-nitrophenyl-jS-D-galactoside (ii) were prepared by probe sonicating Compound 2, cholesterol (8: 4gg_1) in the presence of o-nitrophenyl-jS-D-galactoside solution (34 mM, 2 mL) and isolation of the vesicles by ultracentrifugation and redispersion in neutral phosphate buffer. These latter vesicles (0.4 mL) were then incubated with /J-D-galactosidase (2UmL_1, 0.1 mL) and the absorbance once again monitored. 3.3.3. Magnetic Magnetically responsive polymerized liposomes composed of 1,2-di (2,4- octadecadienoyl)-sn-glycerol-3-phosphorylcholine, loaded with ferric oxide and subsequently polymerized may be localized by an external magnetic field to the small intestine, and specifically the Payer's patches.47 These polymerized vesicles are stable to the degradative influence of solubilizing surfactants such as triton-X 100,47 and hence should not suffer excessive bile salt mediated degradation during gut transit. These magnetically responsive polymeric vesicles improve the absorption of drugs via the oral route.47 108 Uchegbu & Schatzlein 0.0 10 20 30 40 50.0 MIN ^k Membrane bound enzyme ^ ^ Hydrophobic substrate Fig. 6(b). Enzyme activated polymeric vesicles. Vesicles bearing membrane bound enzyme and containing the hydrophobic substrate fluorescein di-/S-D-galactospyranoside were formed by probe sonicating Compound 2 [Fig. 3(a)], cholesterol, N-biotinylated dipalmitoyl phosphatidyl ethanolamine, fluorescein di-^-D-galactospyranoside (8: 4: 1: 0.0005 g g_1) in neutral phosphate buffer (2 mL) and incubation of the resulting vesicles with b-galactosidase streptavidin (0.3 U). The fluorescence of the enzyme hydrolysed substrate was then monitored (Excitation wavelength = 490 nm, Emission wavelength = 514 nm). 3.3.4. Oxygen Block copolymer vesicles which are destabilized by oxidative mechanisms have been constructed from poly(oxyethylene)-Wocfc-poly(propylene sulphide)-fr/ocfcpoly( oxyethylene) ABA block copolymers.62 These polymeric vesicles are destabilized on the oxidation of the central sulphide block to give sulphoxides and ultimately sulphones.62 On oxidation, vesicles are transformed to worm-like micelles and finally to spherical micelles, eventually releasing their contents. 4. Non-ionic Surfactant Vesicles (Niosomes) 4.1. Self assembly The self assembly of non-ionic surfactants into niosomes is dependent on the hydrophilic — hydrophobic balance of the surfactant and a CPP (Eq. 1) of between 0.5-1016 enables niosomal self assembly. Some examples of niosome forming molecules are given in Fig. 7. Further molecular specifics that govern niosome Vesicles Prepared from Synthetic Amphiphiles 109 15 16 17 OH HO 0 /k0J - O — v ^ ^ O - OH Fig. 7(a). Examples of some niosome forming surfactants: Compound 14,2 Compound 15,6 Compound 16,9 and Compound 17.14 18 19 Fig. 7(b). Niosomal membrane additives, Compound 18 = cholesterol, Compound 19 Solulan C24.4 110 Uchegbu & Schatzlein formation by non-ionic surfactants may be found in published reviews.4'63 Compounds such as Compounds 15 (from the sorbitan surfactant class) are established pharmaceutical excipients,64 and hence formulation scientists looking to prepare a niosome formulation for speedy transition to the clinic will do well looking at this class of molecules for exploitable materials. Most niosomes will not only contain the non-ionic surfactant, but will also contain other molecules such as the membrane stabilizer cholesterol [Fig. 7(b)].4 The bilayer membrane is an ordered structure which may exist in the gel or liquid crystal state. Essentially, molecules are more mobile in the liquid crystalline state, enjoying lateral diffusion within the bilayer that is denied them in the gel state. For any system, the liquid crystal state exists at a higher temperature (T) than the gel state. An increase in temperature favors the transition from the gel to the liquid state because of the entropy gain (AS) associated with this transition, ultimately leading to a lowering of the free energy (AG) of the system. Cholesterol abolishes this membrane phase transition, thus fluidizing the gel state.65 Niosomes are 30 nm to 120 JJLVSX in size4 and often their surfaces must be stabilized against aggregation. Molecules such as the cholesteryl poly(oxyethylene ether) — Solulan C246 (Compound 19, Fig. 7b) or the ionic molecule dicetyl phosphate66 have been used to confer steric and electrostatic stabilization on these vesicles respectively. The reader should be aware that the inclusion of minor quantities (<10% by actual weight or molar content) of ionic surfactant does not prevent these structures from being discussed in this chapter under the niosome heading. Niosomes are often formulated with minor quantities of cationic and other surfactants.4 It can be said that the formulation of liposomes with poly(ethylene oxide) amphiphiles such as distearolyphosphatidylethanolamine-poly(ethylene glycol)26 was the crucial step that allowed liposomes to become clinically relevant drug delivery systems. The resulting liposomes possess a hydrophilic polymer surface, which prevents recognition and clearance of the particles from the blood by the liver and spleen macrophages,26,67 thus increasing the liposomes' circulation time and allowing tumor targeting.68 Niosomes (non-ionic surfactant vesicles), when formulated with a water soluble poly(oxyethylene) cholesteryl ether — (Solulan C24), also circulate for prolonged periods in the blood, accumulate in the tumor tissue and improve tumoricidal activity.6 As well as stabilizing vesicles in the blood, poly(oxyethylene) amphiphiles also stabilize vesicles against aggregation, thus promoting vesicle colloidal stability.56 Poly(oxyethylene) amphiphiles, such as Solulan C24, have a large hydrophilic head group [Fig. 7(b)], and are thus more hydrophilic than the vesicle forming amphiphiles, and hence the level of the former must be kept low to avoid solubilization of the membrane and the formation of mixed micelles.57 In actual fact, Vesicles Prepared from Synthetic Amphiphiles 111 unusual morphologies57 result from the incorporation of non-micellizing quantities of Solulan C24 in vesicles as discussed below. 4.2. Polyhedral vesicles and giant vesicles (Discomes) A series of unusual morphologies have been isolated from the hexadecyl diglycerol ether, Solulan C24, cholesterol phase diagram [Fig. 8(a)]. The addition of Solulan C24 to hexadecyl diglycerol ether [Compound 16, Fig. 7(a)] niosomes eventually results in the formation of mixed micelles.57 At sub-micellar concentrations of Solulan C24 (20-40 mole%), however, giant vesicles (discomes) of 25-100 pm in size are formed.57 Discomes are thermoresponsive vesicles, which become more leaky as the temperature is increased from room temperature to 37°C. These vesicles may thus be used to construct thermoresponsive controlled release systems. In cholesterol low regions of the hexadecyl diglycerol ether, cholesterol, Solulan C24 phase diagram, polyhedral vesicles [Figs. 8(a) and 8(b)] are found.9 These polyhedral vesicles are able to entrap water soluble solutes and the membrane, which is in the gel state contains areas of high and low curvature as shown in © Polyhedral Vesicles (2 -10 jim) ® Spherical, helical, tubular Vesicles (0.5 -10 urn) •3 Discomes (10 - 30 jim) + small spherical & helical vesicles (0.5 -10 jim) @ Discomes (12-60 fun) + mixed micelles \ Reverse Micelles ••• \ \ V \ Oil Fig. 1. A hypothetical pseudo-ternary phase diagram of an oil/surfactant/water system with emphasis on microemulsion and emulsion phases. Within the phase diagram, existence fields are shown where micelles, reverse micelles or water-in-oil (w/o) microemulsions and oil-in-water microemulsions are formed along with the bicontinuous microemulsions. At very low surfactant concentrations two phase systems are observed (taken from Ref. 107). Recent Advances in Microemulsions as Drug Delivery Vehicles 129 fractions, microemulsions are generally considered to be a dispersion of either oil or water droplets stabilized by an interfacial film of surfactant and where appropriate, cosurf actant. These droplet structures are probably the most commonly encountered type of microemulsion microstructure. It is worth noting that both an emulsion and a nanoemulsion can only occur in the form of a droplet, either as an oil-in-water or water-in-oil droplet. At intermediate oil and water compositions, it is obviously not possible for the microstructure to be composed of droplets of one phase dispersed in the other. In these cases, it is thought that a bicontinuous structure exists, in which the water and oil domains are separated by a regular or topologically chaotic continuous amphiphile-rich interfacial layer. A bicontinuous microemulsion is often the intermediate microstructure between an oil-in-water and a water-in-oil microemulsion, although a number of other microstructures such as cylinders and worm-like microemulsions have been reported to exist. In terms of its microstructure, a microemulsion is therefore a very complex system, and in instances where a microemulsion exists over a wide range of compositions, several different types of microstructure may be present.73 It is also important to remember that whatever the microstructure, a microemulsion is a dynamic system in which the interface is continuously and spontaneously fluctuating.104 For this reason, microemulsions stabilized by polymeric surfactants may be the most long lived. 1.4. Microemulsions, swollen micelles, micelles There is much debate in the literature as to what exactly differentiates a microemulsion from a micelle at low volume fractions of disperse phase. Some investigators have perceived a difference between microemulsions and micellar systems containing solubilized oil or water, and have used the terms "swollen" micellar solutions or solubilized micellar solutions to describe such systems. These investigators argue that the term microemulsion should be restricted to systems in which the droplets are of large enough size such that the physical properties of the dispersed oil or water phase are indistinguishable from those of the corresponding oil or water phase, thereby theoretically making it possible to distinguish between oil-in-water (or water-in-oil) microemulsions and micellar solutions containing small amounts of solubilized oil (water). However, in most cases, the transformation between micelles progressively swollen with oil (water) and a microemulsion containing an isotropic core of oil (water) appears to be gradual with no obvious transition point. As a consequence, there is no simple method available for determining the oil (water) content at which the core of the swollen micelle becomes identical to that of a bulk phase. Many researchers therefore use the term microemulsion to include 130 Lawrence & Warisnoicharoen swollen micelles, but not micelles containing no oil (or water).34'107 In biotechnological applications, water-in-oil microemulsions are frequently known as reverse micelles and or even as nonaqueous media. 1.5. Microemulsions and cosolvent systems The above broad definition does not require a microemulsion to contain any microstructure. In other words, it includes systems that are co-solvents, i.e. systems in which the constituent components are molecularly dispersed. Most researchers in the field agree, however, that for a microemulsion to be formed, it is important that the system contains some definite microstructure. In other words, there is a definite boundary between the oil and water phases, and at which the amphiphilic molecules are located and that a co-solvent is not a type of microemulsion. The only way to distinguish a microemulsion from a co-solvent unambiguously is to perform either a scattering study (light, X-rays or neutrons) or PFG-NMR measurements. Regions of co-solvent formation generally appear at low concentrations of oil or water. 2. Microemulsions as Drug Delivery Systems It is clear from its description that microemulsions possess a number of properties that make their use as drug delivery vehicles particularly attractive. Indeed, microemulsions were first studied with the view of using them as potential vehicles for poorly-water soluble drugs, in the mid 1970s by Elworthy and Attwood.17 However, it was not until the mid to late 1980s that they were widely investigated as drug delivery systems; this interest being largely the result of the arrival on the market of the cyclosporin A microemulsion preconcentrate, Neoral. Among the physical properties that make microemulsions attractive as drug delivery vehicle is their transparent nature, which means that the product is not only aesthetically pleasing, but allows easy visualization of any contamination. The small size of the domains present means that a microemulsion can be sterilized by terminal filtration.84 Furthermore, depending on the composition of the microemulsion, it may be possible to heat sterilize the microemulsions.39 Since oilin- water microemulsions are able to incorporate lipophilic substances, they can be used to facilitate the administration of water-insoluble drugs.24 Significantly, the small droplet size provides a large interfacial area for rapid drug release, and so the drug should exhibit an enhanced bioavailability, enabling a reduction in dose, more consistent temporal profiles of drug absorption, and the protection of drug(s) from the hostile environment of the body. In addition to increasing the rate of drug release, microemulsions can also be used as a reservoir and actually slow the release of drug and prolong its effect, thereby avoiding high concentrations in Recent Advances in Microemulsions as Drug Delivery Vehicles 131 the blood.64'142 Whether a drug is rapidly or slowly released from a microemulsion depends very much on the affinity of the drug for the microemulsion. Since microemulsions contain surfactants (cosurfactants) and other excipients, they may serve to increase the membrane penetration of drug.163'189 A number of reviews have been presented, describing the pharmaceutical use of microemulsions.16'19-50'105"107'176 Since the last major review in the area was writen in 2001, the present review will mainly deal with developments henceforth, although important work prior to this will be discussed when appropriate. 2.1. Self-emulsifying drug delivery systems (SEDDS) Before discussing how microemulsions are being exploited in drug delivery, it is worth making one more distinction, namely the difference between a selfemulsifying drug delivery system (SEDDS) and a microemulsion. A SEDDS is a mixture of oil(s), and surf actant(s), ideally isotropic, sometimes containing cosolvent(s), which when introduced into aqueous phase under gentle agitation, spontaneously emulsifies to produce a fine oil-in-water dispersion.36'146 Typically, the size of the droplets produced by dilution of a SEDDS is in the range of 100 and 300 nm, while, upon dispersal in water, a SMEDDS formulation (a sub-group of the SEDDS) forms a transparent microemulsion with particle sizes <100 nm. ASMEEDS is also known as a pre-microemulsion concentrate.97 It is worth noticing that this method of producing a fine oil-in-water emulsion using a S(M)EEDS is identical to the low energy emulsification method for producing oil-in water nanoemulsions.173 It is therefore likely that a diluted S(M)EDDS and nanoemulsion are identically the same. The technique of low-energy or self-emulsification has been commercially exploited for many years in the agrochemical industry, in the form of emulsifiable concentrates of lipophilic herbicides and pesticides.146 However, it has only recently been introduced in the pharmaceutical industry as a tool to improve the delivery of lipophilic drugs by incorporating the drug into a S(M)EDDS formulation which is then filled into capsules.65 Once the capsule has been swallowed and its contents come into contact with the GI fluid, the drug containing (micro)emulsion should be spontaneously formed. Once the drug containing (micro)emulsion is formed, there should be little difference between the fate of the drug thus administered and the same drug administered in a (pre-formulated) microemulsion, although the droplets formed from the S(M)EDDS tend to be of a larger size. One advantage of administering a drug in a SMEEDS as opposed to a pre-formulated microemulsion, is its relatively small volume which can be incorporated into soft or hard gelatin capsules, convenient for oral delivery. To date, there has been a good amount of commercial success for the first selfmicroemulsifying drug delivery systems (SMEDDS) on the market, namely Neoral (cyclosporin A). In addition, the recent commercialization of two self-emulsifying 132 Lawrence & Warisnoicharoen formulations, namely Norvir (ritonavir) and Fortovase (saquinavir), has undoubtedly increased the interest in SEDDS and other emulsion-based delivery systems to improve the delivery of a range of drugs of varying physico-chemical properties. However, there are a number of reasons why S(M)EDDS are not in greater widespread use, but the main reason is probably the stability of the diluted SEDDS, which is in fact a thermodynamically unstable emulsion (although it may exhibit some limited kinetic or "meta" stability). It should be noted however that as a SEDDS is either diluted just prior to administration or else in the body, the required droplet stability is less than 6 hrs (i.e. the transit time of materials down the small intestine). Although most studies of SEEDS have utilized isotropic liquids, the earliest reports of these self-emulsifying systems using pharmaceutical materials are in fact related to pastes based on waxy polyoxyethylene n-alkyl ethers.67 In the context of drug delivery via self-emulsifying systems, isotropic liquids are generally preferred to waxy pastes because if one or more excipient(s) crystallize(s) on cooling to form a waxy mixture, it is very difficult to determine the morphology of the materials. Despite this, there is currently a general move towards formulating semi-solid SEDDS. For example, attempts have been made to transform SEDDS into solid dosage forms by addition of large amounts of solidifying excipients such as adsorbents and polymers15,134 Unfortunately, as the ratio of SEDDS to solidifying excipients required for this approach is very high, this leads to problems in formulating drugs having limited solubility in the oil phase. Recent attempts have been made to reduce the amount of solidifying excipients by gelling the SEDDS with colloidal silicon dioxide.141 Khoo et al.93 have recently reported the preparation of a halofantrine-containing lipid-based solid self-emulsifying system using either Vitamin E TPGS or a blend of Gelucire 44:14:Vitamin E TPGS as the base. Upon dispersal, these systems produced dispersions that the authors described as microemulsions. Studies in fasted dogs showed that these solid dispersions exhibited a five- to seven-fold improvement in absolute oral bioavailability, when compared with the commercially available tablet formulation. In a different approach, Nazzal et alP2 have determined the potential of a reversibly induced re-crystallized semi-solid self-nanoemulsifying drug delivery system, based on a eutectic interaction between the drug and the carrying agent, as an alternative to a conventional SEDDS. In these eutectic-based self-nanoemulsified systems, the melting point depression method allows the oil phase containing the drug itself to melt at body temperature from its semisolid consistency, and disperse to form emulsion droplets in the nanometer size range. Emulsion systems based on a eutectic mixture of lidocaine-prilocaine,135 and lidocaine-menthol87 have been Recent Advances in Microemulsions as Drug Delivery Vehicles 133 used in the preparation of topical formulations. However, little is known of the use of eutectic mixtures for the preparation of self-(micro) emulsified formulations. 2.2. Related systems There are a number of other putative delivery systems that are closely related to, or are prepared from, a microemulsion. These systems include a variety of gel formulations (including microemulsion-based gels, ringing gels, microemulsion gels) and double microemulsions. 2.2.1. Microemulsion gels Oil-in-water microemulsions can be readily gelled or thickened by the addition of a non-interacting, water-soluble polymer such as polyHEMA,158 Carbopol 94044 or carrageenan179 to form clear "microemulsion gels". In these cases, it is the external aqueous phase that is gelled, while the microemulsion droplets are unperturbed. The structure of the resulting "microemulsion gel" is quite different, if it is prepared using an interacting polymer, such as stearate-polyethylene oxide-stearate. In this instance, the hydrophilic mid-block of the polymer is located in the continuous aqueous phase, while the hydrophobic end blocks are dissolved in the oil droplets, thereby connecting the various microemulsion droplets and resulting in the formation of a transient gel network.159 Clear, "microemulsion gels" are also sometimes obtained at surfactant and/or oil concentrations just outside the oilin- water microemulsion region.180 Sometimes, the resultant gel "rings" or vibrates when tapped.180 The ringing is due to the resonance of shear modes within the gel body.167 Neither of these "microemulsion gels", which are water continuous, are true microemulsions, which are fluid by definition. Clear gels can also be formed in oil continuous systems. For example, a gel can be formed when water is added to reverse micellar solutions of lecithin-in-oil.5,12'161 Here, the water causes the worm-like lecithin reverse micelles to intertwine and form a gel. In addition, gels, widely known as microemulsion-based gels, can be formed from water-in-oil microemulsions stabilized (predominately) by the dichain surfactant sodium bis (2-ethylhexyl) sulfosuccinate (AOT), when gelatin, the natural amphiphilic polymer is added.70,148 Microemulsion-based gels have now been prepared in systems in which a large amount of the AOT has been replaced by nonionic surfactant;88'89 or more recently, using in place of AOT, the single chain surfactant, cetyltrimethylammonium bromide in combination with pentanol.116 In these gels, the gelatin is thought to form water-continuous channels between the microemulsion-droplets. These microemulsion-based gels are very unusual in that, although they are oil continuous, they are electrically conducting. In addition, the 134 Lawrence & Warisnoicharoen continuous oil behaves as if it were still a fluid, even though placing a gel in a solution of the oil does not dissolve it. All of these "microemulsion gels" have potential, or are being explored for use as drug delivery systems. Of particular interest is the fact that the gels possess the properties of being transparent, infinitely stable and readily prepared using only the mildest of mixing. In addition the wide range of microemulsion gels available means that it is possible to select the gel of the required consistency for application to large areas of skin, the nasal membrane, vaginal and buccal membranes and for permeation enhancement. Microemulsion-based gels have been explored as vehicles for the iontophorectic delivery of drugs.88 2.2.2. Double or multiple microemulsions Double (or multiple) emulsions have attracted much interest as potential drug delivery vehicles. For example, adding a water-soluble drug into the internal aqueous phase of a water-in-oil-in-water emulsion may allow the sustained release of the water-soluble drug.59 A double microemulsion should offer similar advantages over the rate of drug release of entrapped solutes. Double emulsions are notoriously difficult to formulate due to the requirement to have one surfactant (or mixture of surfactants) to stabilize the first (internal) emulsion and a second surfactant (or mixture of surfactants) of quite different physico-chemical properties to stabilize the second emulsion. Although a few papers have detailed the production of nanoparticles from systems they described as double microemulsions,35,56'184 the term "double microemulsion" in this context is very misleading, as it refers to the mixing of two water-in-oil microemulsions of comparable composition, but containing different solute in the aqueous phase. There are however two papers which describe the preparation of double (oilin- water-in-oil) microemulsions. In the first, Castro et al.,30 report spectroscopic studies of nifedipine in Brij 96 based oil/water/oil multiple microemulsions. In the second, Carli et al.,29 detail the preparation of an oil-in-water-in-oil microemulsion from an oily phase of either polyglycolized glycerides or a mixture of mono-, diand tri-glycerides, which is microemulsified using a mixture of water and surfactant (soy lecithin and Tween 80). The resultant o /w microemulsion is subsequently redispersed in an oily phase to produce the double (o/w/o) microemulsion.29 2.3. Processed microemulsion formula tions 2.3.1. Solid state or dry emulsions In practical terms, a solid dosage form is preferable to a liquid dosage form in respect of convenience, ease of handling and accurate dosing. Consequently, a number of Recent Advances in Microemulsions as Drug Delivery Vehicles 135 researchers have attempted to develop powdered, re-dispersible emulsion-derived formulations, known as solid state or dry emulsions. Such solid-state emulsions can be used to modulate the release rates of emulsified compound.128 Dry emulsions have been variously prepared by removing water from an oil-in-water emulsion, using water-soluble182 or -insoluble150 solid carriers or indeed a mixture of both water-soluble and water-insoluble carriers,143 by rotary evaporation,128 lyophilization,115 or spray drying.55'127 Attempts have also been recently made to prepare dry microemulsions using similar methodology. For example, Moreno et al.126 have lyophilized an amphotericin B-containing lecithin-based oil-in-water microemulsion in the presence of 5wt% mannitol. The lyophilized product was an oily cake from which the microemulsion could be easily reconstituted over several months. The rationale for developing the water-free formulation was to avoid the hydrolysis of lecithin, which occurs upon its dispersal in water, thereby preventing any deterioration of the formulation upon storage. Overall, the lyophilized lecithin based oil-in-water microemulsions appear to be valuable systems for the delivery of amphotericin B, with regard to ease and low-cost of manufacturing and their stability and safety, compared with other formulations already in the market. In a recent paper, Carli et al.29 reported an alternative approach to prepare a "dry" formulation known as a nanoemulsified composite of the coenzyme Q10, ubidecarenone. This composite is prepared by incorporating the ubidecarenone into the inner phase of the double microemulsion, which is then deposited onto a solid microporous carrier such as cross-linked polyvinylpyrrolidone. Among the advantages offered by this approach are good processing and storage properties, easy re-dispersibility in water, high bioavailability and maintenance of the submicron size of the released droplets. Kim et al.,97 have prepared entric-coated solid state premicroemulsion concentrates by first preparing a pre-microemulsion concentrate containing 10 wt% of the drug cyclosporin A, 18.5 wt% of a medium chain triglyceride, 51 wt% of surfactant and 20 wt% of cosurfactant. The pre-microemulsion concentrates were then enteric-coated as films using polymers, such as sodium alginate, Eudragit L 100 and cellulose acetate phthalate, and the resulting films were pulverized to produce powdered, dry, enteric coated premicroemulsion concentrates. Using this approach, the authors successfully prepared a once-a-day oral dose form of cyclosporin A. 3. Formulation Microemulsions are far more difficult to formulate than emulsions because their formation is a highly specific process involving spontaneous interactions among their constituent molecules. In addition, in a number of cases, effects due to the order of 136 Lawrence & Warisnoicharoen the mixing of the component molecules have been observed. Since no adequate theory currently exists to predict from which molecules microemulsions can be formed, mainly because of the requirement to determine a number of unknown parameters, microemulsion formulations are generally developed empirically, although some useful practical guidance as to the choice of the constituent components can be found in the literature.105,107 A recognized and classical approach to microemulsion formulation is to undertake a systematic study of the phase behavior of the systems understudying utilizing of phase diagrams. A major drawback of this approach is the considerable time it takes to develop the phase diagram, especially considering the combination of possible oil, surfactant and cosurfactant, and the fact that time may be necessary for a system to equilibrate. Heat and sonication are therefore often used, particularly with systems containing nonionic surfactants, to speed up the formation process. While there are now commercially available automated systems to prepare phase diagrams,78 the chief drawback of these systems is their cost. A number of attempts have been recently made to use modeling to predict microemulsion formation, thereby aiding in the formulation of microemulsions. A range of modeling techniques have been used including artificial neural networks,3'4'8'149 genetic algorithms3,174 and a combination of data mining, computer-aided molecular modeling, descriptor calculation and multiple linear regression techniques.174,175 Unfortunately, however, all of these techniques require a considerable amount of work prior to prediction, thereby restricting their potential usefulness. Furthermore, the amount of work required for the predictions increases as the number of components of the microemulsion increase; microemulsions formulated from five components (i.e. oil, water, surfactant, cosurfactant, electrolyte and drug) are not uncommon in pharmaceutical use. To the authors' knowledge, to date, no work has been performed predicting how much drug can be incorporated into a microemulsion and whether the presence of drug has any effect on microemulsion phase behavior. This is an important ommision as microemulsions cannot be considered to be inert, since the presence of drug in some instances (greatly) influences phase behavior (see for example Ref. 138). 3.1. Surfactants and cosurfactants The selection of components for the preparation of microemulsions suitable for pharmaceutical use involves a consideration of their toxicity, and if the systems are to be used topically, their irritancy and sensitizing properties as well. There are a number of surfactant and cosurfactants that are considered acceptable for use as excipients in pharmaceutical formulation.153 Strickley172 has recently reviewed Recent Advances in Microemulsions as Drug Delivery Vehicles 137 those surfactants and cosurfactants currently used in commercially available oral and intravenous formulations. In the general scientific literature, by far the most widely used surfactant to prepare a microemulsion is the double chain, ionic surfactant, sodium bis (2-ethylhexyl) sulfosuccinate (AOT), although a large number of studies have used the single chain, nonionic surfactants of the type QEj, where i is the number of carbons in the alkyl chain, C, and j is the number of ethylene oxide units in the polyoxyethylene chain, E. Both AOT and the QEj surfactants possess the important advantage of being able to form microemulsions in the absence of a cosurfactant,121'185 unlike most other types of surfactant such as the widely studied single chain, ionic surfactant, sodium dodecyl sulphate (SDS), which will only form microemulsions in the presence of an alcohol cosurfactant. Neither AOT nor SDS would be considered to be apprpropriate for the preparation of pharmaceutically acceptable microemulsions, even though they are listed in the Pharmaceutical Excipient Handbook, Rowe et al.l5i As a general rule, nonionic and zwitterionic surfactants tend to be less toxic than ionic surfactants and are therefore more widely used as pharmaceutical excipients.154 Assuming that the surfactants do not degrade into toxic materials, surfactants that posses biodegradable/chemically unstable linkers tend to exhibit less chronic toxicity than those that are chemically stable. For example, as a group, the polyoxyethylene n-acyl surfactants exhibit ~ten times less chronic toxicity than their n-alkyl counterparts, mainly due to their quicker degradation time of days as opposed to weeks. When it comes to comparing acute toxicity, the two groups of surfactants exhibit comparable toxicity. Perhaps the most widely used nonionic surfactants in pharmaceutical formulations are the polyoxyethylene sorbitan n-acyl esters, i.e. the Tweens and in particular, Tween 20 and Tween 80, both of which are used parenterally and orally. In addition, polyoxyethylene derivatives of the triglyceride, castor oil have acceptability for intravenous administration. Other pharmaceutically acceptable surfactants are the polyoxyethylene n-alkyl ethers and n-acyl esters, although both these groups of surfactant tend to be restricted for topical use.154 Other nonionic surfactants that are currently attracting much pharmaceutical interest, although they do not yet have acceptability, are the polyglycerol n-acyl esters, the n-alkyl amine N-oxides and the w-alkyl polyglycosides (or sugar surfactants). The n-alkyl polyglycosides have attracted much pharmaceutical interest, not because of their excellent biodegradability, but because they can be manufactured from renewable resources. All of the aforementioned surfactants have been used to prepare microemulsions, generally as sole surfactant, the only exception being the w-alkyl polyglycosides, which tend to require the presence of a cosurfactant. Pluronics (or poloxamers) of the type poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO) are another class of pharmaceutically 138 Lawrence & Warisnoicharoen attractive surfactant. Interestingly, most reports detailing the use of polymeric surfactants to stabilize a microemulsion describe the preparation of water-in-oil, generally in conjunction with a second surfactant.96,181 Siebenbrodt and Keipert162 have reported the formation of a triacetin-in-water microemulsion using Pluronic L 64 as sole surfactant. Lettow et al.,m have used Pluronic PI23 as sole surfactant to prepare oil-in-water microemulsions, incorporating a 1:1 oil:P123 weight ratio of either 1,3,5-trimethylbenzene or 1,2-dichlorobenzene. Finally, the pharmaceutically acceptable zwitterionic lecithin has been extensively used as a surfactant, however, with very few exceptions, it is not possible to prepare a microemulsion using lecithin as sole surfactant. Generally, lecithin is combined with another surfactant such as Tween 80, or a cosurfactant such as ethanol, when formulating microemulsions. Although ethanol is considered to be pharmaceutically acceptable, typical cosurf actants such as propanol and butanol are not. In addition to toxicity issues, the use of such cosurfactants, which may possess partial oil and water-solubility, can lead to problems with the dilutability of the microemulsion. This is a particular issue if the microemulsion is to be administered orally or parenterally. Consequently, a number of researchers have explored the use of a second surfactant as cosurfactant when formulating a microemulsion. Microemulsions thus prepared tend to be very stable against dilution, as the "cosurfactant" generally has little solubility in either the oil or aqueous phase. Alternately (pharmaceutically acceptable), short chain mono- and di-glycerides have been used in place of a short chain alcohol to successfully prepare microemulsions. In a number of instances, short chain fatty acids such as sodium caprylate have been used as cosurfactants, primarily for the formation of microemulsions for oral delivery; sodium caprylate is known to enhance absorption of drugs across the gastrointestinal tract. A number of researchers have also used cosolvents such as the polyhydric alcohols, sorbitol, glycerol and propylene glycol to aid microemulsion formation. In a number of instances, these materials have been described as "cosurfactants", which quite clearly do not sit in the interfacial surfactant monolayer. Rather, they tend to exert their effect by altering the solvent properties of the polar phase. 3.2. Oils Most reports in the chemical literature detail the preparation of microemulsions using aromatic oils such as benzene and short chain alkanes such as hexane. "Pharmaceutical" oils, unlike those used in the chemical and agrochemical industries, tend to be large in terms of molecular weight and therefore volume, and are relatively polar. Both of these properties tend to work against the oil when it comes to formulating it in a microemulsion, as it is well established that smaller molecular volume oils are easier to solubilize and are solubilized to a greater extent than larger Recent Advances in Microemulsions as Drug Delivery Vehicles 139 oils.2 Although there are reports that in some systems, particularly those containing surfactants with long, unsaturated hydrophobes such as polyoxyethylene (10) oleyl ether, the largest molecular volume oil is solubilized to a greater extent than some of the smaller molecular volume oils.122 The most commonly used "pharmaceutical" oils are medium and long chain triglycerides, and esters of fatty acids such as ethyl oleate, isopropyl myristate are popular. It has become common practice for researchers to screen the solubility of drug in the various components of the microemulsion, in order to predict the optimal composition of the final formulation. However, extreme care has to be exercised when using this approach, as very often, the solubility in the final microemulsion formulation does not correlate well with that seen in the various components. 3.3. Characterization It is noticeable that in contrast to their ease of preparation, it is very difficult to establish the microstructure of a microemulsion. Yet, such information is important as it may influence the drug behavior of the microemulsion in use. For example, it is known that the microstructure of the microemulsion may alter the release rate of any incorporated solute.1'95 Currently, a range of physico-chemical techniques are used to characterize microemulsions. These techniques are often used in tandem to obtain a better picture of the system, as it is unlikely that any one technique alone will give sufficient information.144 Scattering techniques (light, neutron and X-ray) and pulsed field gradient NMR are generally used to determine the microstructure of the microemulsion. One serious limitation with characterizing microemulsions is that most techniques rely on the concentration of disperse phase being low enough to avoid particle-particle interactions, as an estimated volume fraction of 1 vol% is suitable.123 The requirement is a particular problem with microemulsions that contain cosurfactants that partition between the oil and water phases, because these systems frequently undergo a change upon dilution. 4. Routes of Administration Although most of the original work exploring microemulsions as drug delivery vehicles examined their potential for oral drug delivery, microemulsions have now been explored as vehicles for most routes of administration. Currently, they are probably most widely studied for their potential as transdermal delivery vehicles. 4.1. Oral Microemulsions (and SMEDDS) have been widely studied as oral drug delivery vehicles. Indeed, the first commercially available "microemulsion" formulation was 140 Lawrence & Warisnoicharoen a premicroemulsion concentrate of the lipophilic peptide, cyclosporin A. This formulation, known commercially as Neoral, was introduced onto the market in the late 1980s and immediately attracted much attention, mainly because of the high and reproducible bioavailability it produced, but also because developments in biotechnology at that time meant that it had never been easier to produce on a large scale therapeutically-relevant protein and peptides. Unfortunately, because of their physico-chemical properties, in particular their large size and poor stability, proteins and peptides are very difficult to formulate. Microemulsions offered an attractive solution to this problem, and consequently, most of the original exploratory studies on microemulsions as drug delivery vehicles were spent developing oral protein/peptide microemulsion formulations. 4.1.1. Proteins and peptides As the majority of therapeutic proteins and peptides are hydrophilic and watersoluble, most studies utilizing microemulsions as vehicles for such molecules have exploited water-in-oil microemulsions. After cyclosporin A, which is unusually highly lipophilic, for a therapeutic peptide, the most widely studied peptide is insulin, with much of the early work in this area being performed by Ritschel.152 For example, Kraeling and Ritschel101 compared the peroral microemulsion formulation of insulin and capsule forms and determined that the microemulsion formulation increased the bioavailability of the insulin. Recently, more complex microemulsion-based systems have been developed in an attempt to improve the extent of insulin absorption. For example, a recent study performed by Natnasirichaiku et al.186 showed a significant improvement in the oral bioavailability of insulin (in diabetic rats) when administered in nanocapsules dispersed in a water-in-oil microemulsion. Santiago et al.155 have developed a new, enteric oral dosage form of insulin, in which an association of insulin and cyclodextrin contained within a microemulsion is processed into granules. In the most recent study aimed at developing an oral formulation of insulin, Iek et al.77 used a conventional lecithin-based water-in-oil microemulsion formulation prepared from 21.6 wt% water, 37.6 wt% Labrafil M 1944 CS as oil and stabilized by 40.8 wt% of a 1:1 weight ratio of lecithin (Phospholipon 90G) and ethanol. In addition to insulin (21.6IU/g water), some of the microemulsions contained the enzyme inhibitor aprotinin (2500KlU/g water). Although it is the first time that a microemulsion formulation has contained both a protein/peptide and an enzyme inhibitor, the concept of adding an enzyme inhibitor, to a formulation containing a peptide in an attempt to reduce its degradation is not new.188 The plasma glucose and insulin levels of the rats after intragastric administration of the formulations to both diabetic and non-diabetic rats were significantly different from those obtained after oral Recent Advances in Microemulsions as Drug Delivery Vehicles 141 administration of an aqueous insulin solution. Although the addition of aprotinin to the microemulsion containing insulin increased bioavailability when compared with those not containing it, the difference was insignificant. Other peptides formulated as water-in-oil microemulsions in an attempt to improve their oral absorption include RGB peptides,37'38 and more recently, Nacetylglucosaminyl- N-acetylmuramyl dipeptide (GMDP).119 The poor bioavailability of GMDP has been attributed to both its poor stability in the lumen of the gastrointestinal tract and its poor intestinal permeability. When GMDP was administered intraduodenally in a water-in-medium-chain trigylceride microemulsion, a ten-fold increase in bioavailability was observed, i.e. a bioavailability of 80.2% was achieved as opposed to 8.4%, seen after administration of an aqueous solution of GMDP. This increase in bioavailability is consistent with the work of Constantinides et al.37,3S who utilized a similar medium chain triglyceride based microemulsion to increase the oral bioavailability of the water-soluble peptide SK&F 106760, after intraduodenal administration to rats. Ke et al.92 have recently reported an attempt to develop water-in-oil microemulsions suitable for the incorporation of therapeutic proteins and peptides using a medium chain triglyceride, water and tocopheryl polyethylene glycol 1000 succinate (TPGS) as the primary surfactant. However, as TPGS could not form microemulsions when used as sole surfactant, it was mixed with a second surfactant, either Tween 20,40,60 or 80, at a weight ratio in the range of 4:1 to 1:4. A range of glycols and polyols were examined as cosurfactants. Although stable, transparent microemulsion and gel regions were identified, the extent of these regions was influenced by the precise nature and the amount of the secondary surfactant and cosurfactant. For example, Tween 80, which is an ester of the unsaturated CI 8 fatty acid, oleic acid, was more effective in forming a microemulsion than Tween 60, which is an ester of the saturated C18 fatty acid, stearic acid. In this study, although the microemulsions were ultimately intended for use as delivery vehicles for protein or peptide drugs, they were not examined for this purpose. 4.1.2. Other hydrophilic molecules Other water-soluble therapeutic molecules that have been administered in microemulsions include the aminoglycoside antibiotic, gentamicin74 and the biologically active polysaccharide, heparin." In common with all aminoglycosides, gentamicin is highly polar and is therefore considered unlikely to be absorbed from the gastrointestinal tract via simple diffusion. In order to facilitate the transmucosal delivery of the drug, Hu et al74 prepared a SMEDDS formulation of gentamicin using a range of surfactants. When Labrasol was used as surfactant, a 54.2% bioavailability of gentamicin was obtained, compared with only 8.4 and 3.4% when 142 Lawrence & Warisnoicharoen Tween 80 and Transcutol P were respectively used. Labrasol was also found to inhibit intestinal secretory transport from the intestinal enterocytes, providing the formulation with the additional benefit of inhibiting the efflux of gentamicin out of the enterocytes into the GI lumen. Due to its low bioavailability, heparin is generally administered by injection. In an attempt to formulate an orally active version of heparin, Kim et al." synthesized a low molecular weight heparin (LMWH)-deoxycholic acid (DOCA) conjugate (termed LMWH-DOCA) and formulated it in a water-in-oil microemulsion using as oil, the medium chain trigylceride, tricaprylin, a mixture of Tween 80 and Span 20 surfactants, LMWH-DOCA and water (volume ratios of 5:3:1:1 respectively). Oral administration of LMWH-DOCA in the water-in-tricaprylin microemulsion to mice resulted in a bioavailability of 1.5%. Toxicity studies suggested that the enhancement in bioavailability, observed with the DOCA-conjugated LMWH, was administered in a microemulsion not due any local toxicity such as disruption or damaging of the intestinal membrane. 4.1.3. Hydrophobic drugs A number of poorly water-soluble, low molecular weight, lipophilic drugs have also been formulated in microemulsions (or SMEEDS) for oral delivery including nitrendipine,90 danzol145 halofantrine94 and biphenyl dimethyl dicarboxylate.98 These studies serve as an illustration of how important it is to understand the influence on microemulsion formation of the various formulation components. It is worth commenting that the main use of SMEEDS formulations is for the oral administration of lipophilic drugs. Formulating nitrendipine in a SMEEDS formulation, composed of a 1:1 (w/w) mixture of glycerol monocaprylic ester (MCG) and propyleneglycol dicaprylic ester (DCPG) and nonionic surfactant (various), was observed to significantly enhance its absorption when compared with a suspension or an oil solution,90,91 and served to reduce the effect of the presence of food on its absorption. However, the absorption profile of nitrendipine was seen to vary with the type of surfactant used; absorption was rapid from the Tween 80-stabilized formulation, while the HCO-60- based formulation gave a prolonged plasma concentration profile. No absorption of nitrendipine was observed from the formulation containing BL-9EX (polyoxyethylene alkyl ether, C12E9). Damage to the gastrointestinal mucosa also differed with the type of surfactant employed. HCO-60 and Tween 80-based formulations were mild to the organs, while BL-9EX-based formulation caused serious damage. The study of Porter et al.145 appropriately demonstrates the effect of changing the nature of the trigylceride involved in the formulation on drug absorption. These workers studied three lipid-based danazol formulations; namely a long-chain triglyceride solution (LCT-solution), a SMEDDS based on long (C18) chain lipids Recent Advances in Microemulsions as Drug Delivery Vehicles 143 (LC-SMEDDS) and a SMEEDS formulation containing medium (C8-C10) chain lipids (MC-SMEDDS). These formulations were administered to fasted beagle dogs and their absorption, compared with that obtained with a micronized danazol formulation administered postprandially and in the fasted state. Although both the LCT-solution and LC-SMEDDS formulations were found to significantly enhance the oral bioavailability of danazol, when compared with fasted administration of the micronized formulation, the MC-SMEDDS produced little improvement in danazol bioavailability. This result was partly attributed to the fact that upon digestion of the medium-chain formulation, significant drug precipitation was observed. Khoo et al.94 also considered the effect of formulating halofantrine as a pre-microemulsion concentrate in a formulation based on either a medium- or longchain triglyceride. Both formulations, which were administered as soft-gelatin capsules, contained the same amount of medium or long chain trigylceride and were stabilized by the same surfactant/cosurfactant mixture, consisting of Cremophor EL and ethanol. Although the plasma levels of the drug were not significantly different between the two formulations, the amount of drug absorbed lymphatically varied in that 28.3% of the dose administered in the long-chain trigylceride formulation was transported lymphatically, as opposed to only 5.0% of the dose administered in the medium-chain formulation. Kim et al.9S attempted to improve the solubility and bioavailability of biphenyl dimethyl dicarboxylate, a drug used in treating liver diseases, by formulating it as a premicroemulsion concentrate. In order to optimize drug loading in the formulation, these workers screened drug solubility in a range of surfactants and oils, and on the basis of these results selected: Tween 80 and Neobee M-5. However, care must be taken when using this approach to optimize the formulation with respect to drug loading, as it has shown that solubility of drug in the bulk components is not a reliable indicator of solubility, in the final microemulsion formulation.120,122 The danger of predicting drug solubility in the final formulation, on the basis of bulk solubility, can be seen in the study of Kim et al.98 where the solubility of the drug in a formulation consisting of a 2:1 weight ratio of Tween 80 to Neobee M-5 was 7 times that of the formulation containing a Tween 80:Neobee M-5 weight ratio of 1:4, despite the solubility of the drug in Neobee M-5 being 10 times that seen in Tween 80. The final formulation, which consisted of 35 wt% triacetin and 65 wt% Tween 80 and Neobee M-5 at a weight ratio of 2:1, greatly enhanced the oral bioavailability of BDD, possibly due to the increased solubility of the drug and its immediate dispersion in the gastrointestinal tract. Itoh et al.79 optimized the formulation of the poorly water-soluble drug N-4472, N-[2-(3,5-di-tert-butyl-4-hydroxyphenethyl)-4,6-difluorophenyl]-N- [4-(Nbenzylpiperidyl)] urea, by complexing it with L-ascorbic acid and incorporating the complex into a SMEEDS comprising Gelucire 44/14, HCO-60 and sodium dodecyl sulfate. Upon dilution with water, the SMEEDS formulation produce a fine 144 Lawrence & Warisnoicharoen dispersion of 18 nm droplets which were stable over the pH range of 2.0 to 7.0. The oral bioavailability of the drug was between 2-4 times that which was obtained with an aqueous solution of the complex. 4.2. Buccal To date, very little work has been performed on investigating the use of microemulsions as vehicles for buccal delivery. In 1988, Ceschel et a\?x showed that the penetration of the essential oil, Salvia sclarea L. through porcine buccal mucosa in vitro was increased when formulated as a microemulsion, as opposed to the pure essential oil. Scherlund et al.5S investigated the potential of lidocaine and prilocaine thermosetting microemulsions and mixed micellar solutions as drug delivery systems for anesthesia of the periodonlal pocket. The formulations contained between 2-10 wt% of a eutectic mixture of lidocaine or prilocaine (melting point 18°C), while the block copolymer surfactants, Pluronic F127 and F68, were present at between 13 and 17 wt% for F127, and between 2 and 6 wt% for F68. F127 was chosen, as it is known to gel at body temperature and it is important that the formulation is easy to apply, remain at the application site, have a fast onset time, be non-irritant, and stable under normal storage conditions. The pH of the formulations was varied between 5 and 10. Most of the combinations were found to result in clear solutions, presumably oil-in-water microemulsions or mixed micellar solutions, depending on the pH of the system. At low pH, lidocaine and prilocaine are positively charged, and they could be expected to behave largely as water-soluble cationic surfactants, hence possibly forming mixed micelles. On the other hand, at high pH, the drug substances are poorly soluble and could be expected to act largely as hydrophobic solutes and form the core of the microemulsion droplets. 4.3. Parenteral In recent years, considerable emphasis has been given to the development of injectable microemulsions (o/w) for the intravenous delivery of drug, in order to increase the solubility of the drug39'138'139 to reduce drug toxicity,25'26'126 to reduce hypersensitivity,72 and to improve drug solubility and reduce pain upon injection.109 A very recent development is the formulation of microemulsions as long circulating vehicles, and more recently, as drug tageting agents. In addition, water-in-oil microemulsions have been investigated as depot vehicles for the intramuscular delivery of drugs.22'64 The first published study which established the potential of microemulsions for use in intravenous delivery was probably that of von Corswant et al. in Ref. 39. These researchers prepared a pharmaceutically acceptable, bicontinuous microemulsion from a medium-chain triglyceride oil, poly(ethylene Recent Advances in Microemulsions as Drug Delivery Vehicles 145 glycol) 400 and ethanol cosolvents and stabilized by soybean phosphatidylcholine and poly(ethylene glycol)(660)-12-hydroxystearate. Prior to administration, the microemulsion required dilution with a suitable aqueous phase. Upon dilution, the microemulsion formed an oil-in-water microemulsion with droplets of size between 60 and 200 nm, smaller than the size of the droplets in a commercial intravenous emulsion, namely Intralipid. From their animal studies, the authors concluded that the microemulsion they developed was suitable for administion by intravenous infusion to conscious rats. Unfortunately, although the researchers did determine drug solubility in the bicontinuous microemulsions, they did not report this. Park and Kim138 also investigated the formulation of poorly water-soluble flurbiprofen at ~8 times its aqueous solubility into an oil-in-water microemulsion suitable for intravenous administration. The microemulsions were prepared using varying weight ratios of oil (ethyl oleate) to surfactant (Tween 20), and contained a range of isotonic solutions as the polar (aqueous) phase. Unfortunately, insufficient information was supplied regarding the precise compositions of the microemulsions, in particular, how much oil and surfactant were present, so as to draw conclusions about the formulation; (perhaps surprisingly) the ratio of oil to surfactant used did not seem to have any effect on the amount of drug solubilized and that the presence of too much drug had a destabilizing effect on the microemulsion. Disappointingly, the pharmacokinetic parameters of flurbiprofen, after intravenous administration of flurbiprofen-loaded microemulsion to rats, were also not significantly different from those of flurbiprofen in phosphate buffered saline solution. In a later publication, Park et a/.138 overcame the problem of stability seen in their earlier study by replacing the surfactant Tween 20 with lecithin and distearoylphosphatidyl- ethanolamine-N-poly(ethyleneglycol) 2000 (DSPE-PEG) and using ethanol as a cosolvent. Due to the presence of the long chain polyoxyethylene groups on the exterior surface of the microemulsion droplets, it was perhaps unsurprising that the biodistribution of flurbiprofen administered in this microemulsion was quite different. In particular, reticuloendothelial uptake of flurbiprofen decreased, suggesting that it may ultimately be possible to target drugs incorporated in this microemulsion to different sites of the body. As part of a series of papers, Brime et al.25,26 and Moreno et al.126 prepared a novel amphotericin B lecithin-based oil-in-water microemulsion, in an attempt to produce a formulation with less toxic effects than the currently available commercial formulation, Fungizone. The microemulsion which contained as oil isopropyl mystriate and a mixture of either Tween 80 or Brij 96 with lecithin as surfactant. In some instances, formulation was lyophilized in an attempt to increase its stability. The overall results of the toxicity studies were encouraging as the amphotericin B-containing microemulsions exhibited a low toxicity, suggesting a potential therapeutic application. 146 Lawrence & Warisnoicharoen Zhang et al.m prepared a lecithin-based SMEDDS formulation of the drug norcantharidin. Upon dilution, the release rate of norcantharidin contained in the SMEEDS formulation was found to be dependent on the size of the disperse phase and the type of lecithin used. Interestingly, although norcantharidin was poorly soluble in the ethyl oleate and only slightly soluble in water, microemulsions containing ethyl oleate oil exhibited a significant increase in solubilization over the corresponding aqueous solution. Clonixic acid is currently marketed in salt form because of its poor watersolubility. However, the commercial dosage form causes severe pain after intramuscular or intravenous injection. To improve the apparent aqueous solubility of clonixic acid and to reduce the pain it causes on injection, Lee et al. (2000) incorporated 3 mg/mL clonixic acid into oil-in-water microemulsions (size 120 nm) prepared from pre-microemulsion concentrate of castor oil, and a mixture of Tween 20 and Tween 85 surfactants (present in a weight ratio of 5:12:18). Although the microemulsion formulation significantly reduced the number of rats licking their paws as well as the total licking time, suggesting less pain induction by the microemulsion formulation; the pharmacokinetic parameters of clonixic acid after intravenous administration were not significantly different from those of the commercial formulation, lysine clonixinate. The results of the study suggested that a microemulsion formulation is an alternative vehicle for clonixic acid. Paclilaxel (Taxol) injection is known to cause hypersensitivity reactions. Consequently, He et al.72 explored whether it was possible to prepare a non-sensitizing paclitaxel microemulsion using egg phosphatidylcholine, Piyronic F68 ancl Cremophor EL as surfactants, and ethanol as cosurfactant. Note that there was no mention of the presence of a specific oil. The study showed that for an equivalent dose, the paclitaxel microemulsion did not cause any hypersensitivity reaction, whereas Taxol did. In addition, the bioavailability of the paclitaxel in the new microemulsion was significantly higher and the elimination rate slower than that achieved with Taxol. The authors suggested that the drug molecules, trapped in the oil droplets, diffused into the systemic circulation slowly. Furthermore, the small particle size of the droplets (10-50 nm) meant that the microemulsion droplets could escape from uptake and phagocytosis of RES. Infact it was previously suggested that it should be possible to modify the surface of the microemulsion droplets, with polyoxyethylene chains, to significantly improve circulation time.57'118'190 Kanga et al.S6 have recently explored the possibility of optimizing the release of paclitaxel from a SEEDS formulation using the polymer, PLGA. The SEEDS formulation, which was a mixture of drug, tetraglycol, Cremophor ELP, and Labrafil 1944 also contained PLGA of varying molecular weight. The droplet size of the microemulsions was in the range of 45-270 nm, with the systems without PLGA exhibiting the smaller size. The release rate of paclitaxel decreased in the order of Recent Advances in Microemulsions as Drug Delivery Vehicles 147 PLGA, PLGA 8 K, PLGA 33 K, and PLGA 90 Kg/mol, suggesting that the molecular weight of PLGA in microemulsion could control the release rate of paclitaxel from microemulsion. 4.3.1. Long circulating microemulsions Long circulating microemulsions have been suggested as an alternative formulation to long circulating vesicles on the basis of their small size, thus avoiding uptake by the RES, their stability and their ability to solubilize lipophilic compounds more effectively than vesicles, and their ease of preparation. Wang et al.,S3 and Junping et al.,m have determined the potential of intravenous delivery systems of emulsion/microemulsion systems based on vitamin E, cholesterol and PEG2ooo-lipid. In their first study, Wang et a/.,183 prepared emulsions containing 1 part drug, 3 parts vitamin E, 3 parts cholesterol and 3 parts PEG2000-DSPE with the final formulation containing 5mg of drug in 10 mL of saline solution. Although the emulsion was reported to form spontaneously on the addition of the required amount of saline, the formulation was homogenized to produce a more uniform particle size distribution of 123.0 ±1.2 nm; no information was given as to the size of the droplets prior to homogenization. The zeta potential and drug loading efficiency of the sub-micron emulsion were -12.67 + 1.35 mv and 96.3 + 0.3. Although the size and loading efficiency of the formulation remained uncharged when stored at 7 to 8°C for a year, ~6.5% decomposition of the drug was observed. The plasma area under the curve (AUC) of the drug in the sub-micron emulsion was significantly greater than that of free drug. Overall, the drug in the emulsion had a lower acute toxicity and greater potential antitumor effects than the free drug, suggesting that the formulation is a useful tumor-targeting sub-micron emulsion drug delivery system. In a follow-up study, Junping et al.84 prepared microemulsions of vincristine suitable for injection using vitamin E, PEG2000-DSPE and cholesterol, adding oleic acid to it. The weight ratio of components used was I part drug, 5 parts oleic acid, 5 parts vitamin E, 5 parts cholesterol and 5 parts PEG2000-DSPE. No homogenization was used in the preparation of the microemulsion which yielded microemulsion droplets of 138.1 ± 1.2 nm, when prepared using saline at pH 7.4. Note that 10 mL of microemulsion solution contained 1 mg of drug. The adjustment the pH of the aqueous phase pH and the presence of oleic acid was essential for a high drug loading (94.3 ± 0.3%), while the vitamin E was required for long-term storage of the formulation at 7 to 8°C. The formulation was stable, with respect to particle size, when stored at 78°C in the dark for 1 year, while the loading efficiency of drug decreased by approximately 3%, and 7.4% decomposition of the drug was observed. The plasma AUC of the vincristine in the microemulsion was significantly 148 Lawrence & Warisnoicharoen greater than that of free drug. As with the previous formulation, the drug in the microemulsion exhibited a low acute toxicity and a high potential antitumor effect. 4.3.2. Targeted delivery Shiokawa et al.M recently reported the development of a novel, tumor targeted microemulsion formulation suitable for delivery of the lipophilic antitumor antibiotic, aclacinomycin A. Tumor targeting was achieved via folate linked to the exterior surface of long circulating (pegylated) microemulsions. Folate was selected because the folate receptor is abundantly expressed in a large percentage of human tumors, but it is only minimally distributed in normal tissues. The basic composition of the microemulsion was PEG2ooo-DSPE/cholesterol/vitarnin E/drug (present at a 3:3:3:1 weight ratio or 7:48.3:43,3:1.5 molar ratio). In one microemulsion, 0.24 mol% of folate linked PEG2000-DSPE was present, another contained 0.24 mol% of folate linked PEG5000-DSPE. In a third, the folate was linked directly to the DSPC and in the final one, no folate was present. The association of the folate- PEGsooo-linked microemulsion and folate-PEGaooo-lhiked microemulsion with the target cells was 200-and 4-fold higher, whereas their cytotoxicity was 90- and 3.5- fold higher than those of nonfolate microemulsion respectively. The folate-PEGsooolinked microemulsions showed 2.6-fold higher accumulation in solid tumors 24 hrs after i.v. injection and greater tumor growth inhibition than free drug. These findings suggest that a folate-linked microemulsion is a feasible means for tumortargeted delivery of lipophilic drug. This study shows that folate modification with a sufficiently long PEG chain on the exterior of a microemulsions is an effective way of targeting the carrier to tumor cells. 4.4. Topical delivery AAA. Dermal and transdermal delivery The dermal and transdermal routes of administration offer several advantages compared with other routes of administration. However, the poor permeability of the stratum corneum often limits the possibilities for choosing the topical administration route. Therefore, novel innovative formulations such as microemulsions that have the potential to facilitate skin permeation are of great interest. The investigation of microemulsions as vehicles for cutaneous drug delivery is increasingly common as their potential is realized. Indeed, the cutaneous route is currently the most popular route of adminstration for a microemulsion. Microemulsions offer significant potentials as transdermal delivery vehicles, since they are robust, frequently stable to the addition of significant amounts of soluble enhancers, excipients and depending on their molecular architecture. Kreilgaard has reviewed the use of Recent Advances in Microemulsions as Drug Delivery Vehicles 149 microemulsions as cutaneous drug delivery vehicles in 2002. In the present review, work prior to 2002 will not be dealt with in any detail. In addition, due to the large amount of research in the area, the review is not exhaustive. Proteins and peptides Recently, the transdermal route has received attention as a promising means to enhance the delivery of drug molecules, particularly peptides, across the skin, using harsh physical enhancement techniques such as iontophoresis and sonophoresis. Very little research has been performed, investigating microemulsions as vehicles for peptide delivery. Getie et al.66 examined the skin penetration profiles of 0.75 wt% desmopressin acetate released from a water-in-oil microemulsion comprising 5 wt% water, 20wt% Tagot 02:Span 80 3:2 and 74.25 wt% isopropyl myristate. However, the profile was comparable to that obtained using a standard amphiphilic cream. Although the amount of drug that penetrated the upper layers of the skin was significantly higher from the cream than from the microemulsion at all time intervals, within 6 hrs 6% of the applied dose reached the acceptor compartment from the microemulsion instead of 2% from the cream within 300 min, suggesting that the water-in-oil microemulsion has potential for the systemic administration of the drug. Hydrophilic drugs Water-in-oil microemulsions have been used to enhance the penetration of watersoluble drugs. For example, Alvarez-Figueroa and Blanco-Mendez9 reported the in vitro delivery of water-soluble methotrexate from hydrogels using iontophoresis, and passively from oil-in-water and water-in-oil microemulsions prepared using either a 3:1 v:v Labrasol: Plurol Isostearique mixture or a 3:1:1.2 v:v:v Tween 80:Span 80:l,2-octanediol mixture as surfactant/cosurfactant, and either ethyl oleate or isopropyl myristate as oil. All microemulsion formulations studied were more effective than passive delivery from aqueous solution of the hydrophilic drug, although for the microemulsions, delivery was greater from the oil-in-water systems. However, delivery from the microemulsions was less than that using iontophoresis, probably because of the lower solubility of drug in microemulsions than in simple aqueous solution. Escribano et al.53 attempted to improve the transdermal permeation of sodium diclofenac. Four formulations were studied. One was an oil-in-water microemulsion based on transcutol (19wt%), plurol oleique (19.5 wt%), water (30.6 wt%), isostearyl isostearate (10.9 wt%) and Labrasol (19wt%). The other three formulations were "co-solvent" systems prepared from various of the ingredients used for 150 Lawrence & Warisnoicharoen the microemulsion formulation. In this study, the microemulsion performed less well than the various co-solvent formulations and in a similar manner to an aqueous solution of the drug. This observation is perhaps not surprising as various enhancers were involved in the microemulsion droplets and were not available to improve drug penetration. Also, as it is likely that the drug was predominately in the continuous phase of the microemulsion, it is not surprising that the formulation behaved in a similar manner to an aqueous solution. The in vitro transdermal permeation of the antineoplastic, 5-fluorouracil, incorporated at I.25mg/mL in water-in-oil microemulsions prepared using AOT/water/isopropylmyristate has been studied by Gupta et al.69 These researchers found that as the water content increased from 0.9, 1.8, 2.7 and 3.6% w/w, microemulsions prepared with a surfactant to oil ratio of 5:95 showed 1.68, 2.36, 3.58 and 3.77-fold increases respectively in the skin flux of 5-fluorouracil, compared with an aqueous solution of drug. Increasing the surfactant: oil weight ratio from 5:95 through 9:91 to 13:87, at fixed water:surfactant content of 15, gave 3.58-, 5.04- and 6.3-fold enhancements of drug flux. In their study69 used attenuated total reflectance-Fourier transform infrared spectroscopy to determine that the microemulsions exerted their enhancement by interacting and perturbing the architecture of the statun corneum. The extent of this perturbation was dependent upon the concentrations of water and AOT in the microemulsion. Preliminary toxicity studies suggested that the microemulsions were a suitable vehicle for transdermal delivery. Amphiphilic drugs Jurkovic et al.85 have investigated the formulation of the amphiphilic antioxidant ascorbyl palmitate in a microemulsion, with a view to using the formulation as a protectant against free radical formation due to UV irradiation. Both oil-in-water and water-in-oil microemulsions were prepared using a medium chain triglyceride as oil, and PEG-8 caprylic/capric glycerides (Labrasol) and polyglyceryl-6-dioleate, (Plurol oleique) as surfactant and cosurfactant. The ascorbyl palmitate was incorporated into the microemulsions at various concentrations between 0.5-5.0 wt%. The microemulsions were gelled using either xanthan gum (water-in-oil) or Aerosil 200 (water-in-oil). The effectiveness of the ascorbyl palmitate in the microemulsions depended on both the concentration and type of microemulsion. Regardless of the type of microemulsion, efficacy was significantly higher at the higher ascorbyl palmitate concentrations. Overall, the oil-in-water microemulsions were more effective at protecting against UV irradiation, although they delivered ascorbyl palmitate to the skin at a slower rate than the water-in-oil microemulsions. The effect of formulation composition on the in vitro release rate of the amphiphilic drug, diclofenac diethylamine, from a range of microemulsion vehicles Recent Advances in Microemulsions as Drug Delivery Vehicles 151 containing PEG-8 caprylic/capric glycerides (surfactant), polyglyceryl-6 dioleate (cosurfactant), isopropyl myristate and water was determined by Djordjevic.49 The phase behavior of the microemulsions was determined in the absence of drug. In the microemulsions selected for further study, the level of water present ranged from 10 to 60 wt% while the amount of oil varied from 8 to 46.6 wt%. The physico-chemical characterization studies indicated the microstructure to be either bicontinuous or non-spherical, and despite its amphiphilic nature, the drug was partitioned mainly in the water phase. The non-linearity of the drug release profile from the bicontinuous microemulsions was thought to be due to a complex distribution of drug within the microemulsion. The flux of the drug increased by >4 times, from a waterin- oil to an oil-in-water microemulsion, the release of drug from the bicontinuous microemulsion, suggesting that the microstructure hampers the release of the drug. Hydrophobic drugs Dalmora and Oliveria43 and Dalmora et al.,u investigated the release of piroxicam encapsulated in /8-CD in cationic oil-in-water microemulsions, in an attempt to optimize the drug's delivery. The results demonstrated the potential of the reservoir in vivo system following the use of a microemulsion. The high degree of retention of the active substance can provide a means for modulating the anti-inflammatory effect, by greatly extending the release period relative to those formulations where the piroxicam is only dissolved or dispersed in a homogeneous aqueous medium. In conclusion, both microemulsions and ^-CD-containing microemulsions can offer many promising features for their use as topical vehicles for piroxicam delivery. Some of the microemulsions gelled using carbopol 940. Paolino et alP7 examined the potential of oil-in-water microemulsions as topical drug vehicles for the percutaneous delivery of ketoprofen. Microemulsions were prepared using triglycerides as oil, and were stabilized by a mixture of lecithin and n-butanol as a surfactant/ co-surfactant system. The percutaneous enhancer, oleic acid, was added to some of the microemulsions. Physicochemical characterization of the microemulsions yielded a mean droplet size of 35 nm and a negative zeta potential of -19.7 mV in the absence of oleic acid and — 39.5 mV in its presence. The ketoprofen-loaded microemulsions showed an enhanced permeation through excised human skin with respect to conventional formulations, although no significant percutaneous enhancer effect was observed in the presence of oleic acid. Microemulsions showed a good human skin tolerability on volunteers. Shukla et al.165 have investigated the potential of oil-in-water (o/w) microemulsions as vehicles for the dermal delivery of a eutectic mixture of lidocaine (lignocaine) and prilocaine, which acted as the oil phase. The microemulsion was stabilized by a blend of a 2:3 ratio Tween 80 and Poloxamer 331, a mixture of water 152 Lawrence & Warisnoicharoen and propylene glycol were used as the hydrophilic phase. These microemulsions were able to solubilize up to 20 wt% of the eutectic mixture. In an attempt to enhance the transdermal delivery of the poorly water soluble drug, triptolide, and to reduce the toxicity problems associated with its usage, a water-in-oil microemulsion was compared with that of solid lipid nanopartides.124 The microemulsion which was formulated using 40wt% isopropyl myristate, 50 wt% Tween-80:l,2-propylene glycol (5:1, v/v) and water and contained 0.025 wt% of triptolide, gave a steady-state flux (for over 12 hours) and a permeability coefficient of triptolide of 6.4 ± 0.7 mg/cm2 per h and 0.0256 ± 0.002 cm/h; a value which was approximately double that of the solid liquid nanoparticles and 7 times higher than that of triptolide solution of the same concentration. In another study, Chen etal.33 also studied the incorporation of the drug, into a similar microemulsion using oleic acid as oil. Oleic acid was added because it is a known penetration enhancer, although there was no evidence of it acting as such in the present formulation. The addition, however, of 1 wt% menthol to the formulation slightly increased penetration from 1.58 ± 0.04 to 2.08 ± 0.06 \ig/cm2 per h (p < 0.05). Encouragingly, no obvious skin irritation was observed for the formulation studied, suggesting that microemulsions are promising vehicles for the transdermal delivery of triptolide. Ross et al.153 examined the transdermal penetration, across full thickness hairless mouse skin, of the insect repellant, N,N-diethyl-m-toluamide (DEET), contained in either a 1:1 v / v ethanohwater solution (containing 20 wt% DEET) or one of two commercially available microemulsion formulations (3M Ultrathon Insect Repellant (containing 31.6 wt% DEET; 3M, St. Paul, MN), and Sawyer Controlled Release DEET Formula (19.0%; Sawyer Products, Safety Harbor, FL). Both formulations were of interest because they were marketed as retarding the absorption of DEET due to being microemulsions. All of the DEET preparations exhibited considerable penetration, e.g., the ethanolic DEET formulation had a time to detection of approximately 30 min with steady stale at 85 min. The penetration obtained with the Sawyer was no different from that obtained from the ethanolic solution. The other microemulsion formulation (3M) demonstrated a different profile; despite being a higher concentration of DEET (30wt% versus 20wt%) and a comparable time to detection (40 min), the time to reach steady state was delayed, although there was still substantial absorption at steady state. Sintov and Shapiro168 prepared a high surfactant lidocaine microemulsion, containing as surfactant a mixture of glyceryl oleate and either PEG-40 stearate or PEG-40 hydrogenated castor oil, isopropyl myristate as oil, tetraglycol as cosurfactant, water, and up to 10wt% of drug, although 2.5 wt% was generally used. The microstructure of the microemulsion went from oil-in-water, through bicontinuous to water-in-oil. The penetration of the drug from the various formulations showed that the surfactant mixture containing PEG-40 stearate was best, while the Recent Advances in Microemulsions as Drug Delivery Vehicles 153 water and surfactant/cosurfactant concentration was also important. Significantly, the lag time for penetration was reduced, suggesting that these microemulsions loaded with drug would provide rapid local analgesia. Priano et tzl.U7 investigated the delivery from a water-in-oil microemulsion, of apomorphine present as ion-pair complex with octanoate to increase its lipophilicity and to diminish its dissociation. As the drug was present at a high concentration, the dispersed phase acted as a reservoir, making it possible to maintain an almost constant concentration in the continuous phase and therefore achieving pseudozero- order release kinetics. The composition of the microemulsion was complex, containing 18.2 wt% water, 42.1 wt% of oily phase of isopropyl miristate-decanol 1:1.5 v/v, 3.9 wt% R-apomorphine hydrochloride, 7.3 wt% Epikuron 200, 7.1 wt% benzyl alcohol, 4.6 wt% octanoic acid 3.5 wt% sodium octanoate, 5.7 wt% sodium taurocholate, 7.6 wt% 1,2-propanediol. The microemulsion was thickened by the addition of 5.9 wt% Aerosil 2000. The microemulsion was able to provide in vitro, through hairless mouse skin, a flux of 88g/h per cm2 for 24hrs, with a kinetic release of pseudo-zero-order, and was chosen for in vivo study; all the components were biocompatible and safe. The flux gave a first approximation of the feasibility of the transdermal administration in man. The pain and discomfort caused by the injection of local anesthetics has stimulated research into developing topical anesthetics. However, the currently available formulations, such as Ametop®gel, (4 wt% amethocaine base preparation) have a number of disadvantages, in particular a long delay of typically 40-60 min between application and anesthetic effect and the requirement for a plastic occlusive dressing. Arevalo et alP have recently developed a decane-in-water microemulsion stabilized by lauromacrogol 300 and containing 4 wt% of amethocaine in an attempt to achieve faster drug permeation, thus reducing the time to reach optimum anesthetic effect. The amethocaine microemulsion proved to be a promising fast-acting analgesic in experimental preclinical studies. Mixtures of hydrophilic and hydrophobic drugs Although microemulsions have long been suggested as suitable formulations for the co-adminstration of drugs of very varying physico-chemical properties, it is only very recently that anyone has reported doing so. Lee et al.lw have developed a novel microemulsion enhancer formulation for the co-administration of hydrophilic (lidocaine HC1, diltiazem HC1) and lipophilic (lidocaine free base, estradiol) drugs. The microemulsions composed of isopropyl myristate and water, and were stabilized by the nonionic surfactant, Tween 80. Transdermal enhancers such as w-methyl pyrrolidone (NMP) and oleyl alcohol were incorporated into all systems without apparent disruption of the system. Unfortunately, the authors did not give the precise, 154 Lawrence & Warisnoicharoen composition of the microemulsions tested; it was only mentioned that they contained a 1:1 v:v mixture of water and ethanol, isopropylmyristate as oil and Tween 80 as surfactant, and were either oil-in-water or water-in-oil. Interestingly, regardless of the physico-chemical nature of the drug, the oil-in-water microemulsions provided significantly better flux for all drugs studied (p < 0.025). Enhancement of drug permeability from the oil-in-water systems was 17-fold for lidocaine base, 30-fold for lidocaine HC1,58-fold for estradiol, and 520-fold for diltiazem HC1. Significantly, the simultaneous delivery of estradiol with diltiazem hydrochloride did not affect the transport of either drug (p > 0.5). Immunization Traditionally, vaccines have been administered by injection using needles, although the concept of topical immunization through intact skin has attracted much attention. Cui et al.42 recently hypothesized that a fluorocarbon-based microemulsion system could be one possible way to deliver plasmid DNA across the skin. Cui et al.42 screened a range of fluorosurfactants for their ability to form ethanolin- perfluorooctyl bromide microemulsions. Note that the authors provided no evidence of a microemulsion being formed. The stability of plasmid DNA in the formulations was also examined. From the surfactant screen, the commercially available Zonyl® FSN-100, an ethoxylated nonionic fluorosurfactant, was selected for further study. Significant enhancements in luciferase expression and antibody and T-helper type-1 based immune responses, relative to those of "naked" pDNA in saline or ethanol, were observed after topical application of plasmid DNA in ethanol-in-perfluorooctyl bromide microemulsion system. From these studies, it can be concluded that fluorocarbon-based microemulsions are suitable for DNA vaccine delivery, although the mechanism(s) of the immune response induction is not known. It is possible that the transport of the molecules across the skin is via the hair-follicles, because DNA is too large and highly charged to cross intact stratum corneum. 4.5. Ophthalmic Conventional ophthalmic dosage forms tend to be either simple solutions of watersoluble drugs or suspension or ointment formulations of water-insoluble drugs. Unfortunately, as these delivery vehicles generally result in poor levels of drug absorption across the cornea, most of the applied drug does not reach its intended site of action. However, because of the relative safety and convenience of topical application in ophthalmology, as well as the relatively low risk (compared with other routes of administration) of systemic side-effects, topical administration Recent Advances in Microemulsions as Drug Delivery Vehicles 1 55 of ophthalmic agents is the preferred route of delivery. Microemulsions and submicroemulsions should offer a possible solution to the problem of poor delivery to the cornea, by sustaining the release of the drug, as well as by providing a higher penetration of drug into the deeper layers of the eye. In addition, they offer the potential of increasing the solubility of the drug in the ophthalmic delivery vehicle.162 Gallarate et al.6} were probably the first to examine the potential of microemulsions as vehicles for ophthalmic delivery. Since then, a number of groups have successfully demonstrated the ability of microemulsions (sub-microemulsions) to prolong the ocular delivery of drug. In their study, Gallarte et al.a were able to further prolong the release of timolol by forming an ion pair with octanoic acid. Garty and Lusky63 demonstrated that the delivery of pilocarpine from an oil-inwater microemulsion was delayed to such an extent that the instillations of the microemulsion formulation twice daily were equivalent to four times daily the applications of conventional eye drops. A similar result was reported by Muchtar et alP° who determined in vitro that the corneal penetration of indomethacin formulated in a sub-microemulsion was more than three times that obtained using commercially available drops. A number of researcher have investigated the potential of positively charged microemulsions to retain the delivery vehicle in the eye, thereby sustaining delivery23,52 To date, a range of drugs have been formulated in a microemulsion in an attempt to sustain release including adaprolol maleate,11'125 timolol,61 levobunolol,62 chloramphenicol162 tepoxalin,54 piroxicam,100 delta-8-tetrahydrocannabinol,129 pilocarpine,21'52'63'71,133 indomethacin,130 antibodies20 and dietary iso-flavonoids and flavonoids.83 In general, these studies showed that it was possible to delay the effect of drug incorporated in a microemulsion, thereby improving bioavailability. The proposed mechanism of the delayed action is that microemulsion droplets are not eliminated by the lachrymal drainage, thereby acting as drug reservoirs. The first studies conducted on man with microemulsions containing adaprolol maleate and pilocarpine, confirmed the results of the earlier studies performed mainly using rabbits.18,178 Vandamme178 has recently reviewed the use of microemulsions as ocular delivery system, and thus only studies since then will be considered in the present review. Fialho and da Silva-Cunha58 recently investigated the long term application of a microemulsion system in rabbits intended for the topical ocular administration of dexamethasone. The formulation contained 5 wt% isopropyl myristate as oil, 15 wt% Cremophor EL as surfactant, and a polar phase of water and 15 wt% propylene glycol, with dexamethasone present at a concentration of 0.1 wt%. Significantly, ocular irritation tests in rabbits suggested that the microemulsion did not provide significant alteration to eyelids, conjunctiva, cornea and iris over a Fe 156 Lawrence & Warisnoicharoen 3-month period. In addition, the formulation exhibited greater penetration of dexamethasone in the anterior segment of the eye and longer release of the drug when compared with a conventional preparation. The area under the curve obtained for the microemulsion system was more than two-fold that of the conventional preparation (p < 0.05). Gulsen and Chauhan68 have recently developed a disposable soft contact lens of a drug-containing microemulsion dispersed in a poly 2-hydroxyethyl methacrylate (HEMA) hydrogel, suitable for ophthalmic delivery, in an attempt to reduce drug loss and side-effects. Upon insertion into the eye, the lens will slowly release the drug into the pre lens (the film between the air and the lens) and the post-lens (the film between the cornea and the lens) tear films, thus providing a sustained delivery of drug. Assuming the size and drug loading of the microemulsions is low, the lenses should be transparent. It was found using these microemulsion-containing lenses, with and without a stabilizing silica shell, that drug could be released for a period of >8 days. By altering droplet size and loading, it is possible to tailor release. 4.6. Vaginal In their 2001 review, D'Cruz and Uckun proposed that microemulsion gel formulations had great potential as intra vaginal/ rectal drug delivery vehicles for lipophilic drugs, such as microbicides, steroids, and hormones, because of their high drug solubilization capacity, increased absorption, and improved clinical potency, as long as a non toxic formulation could be prepared. In their review, D'Cruz and Uckun reported the formulation of two microemulsion-based gels using commonally available pharmaceutical excipients. Repeated intravaginal applications of formulations to rabbits and mice were found to be safe and did not cause local, systemic, or reproductive toxicity. D'Cruz and Uckun investigated the potential of the microemulsion-based gels as delivery vehicles of two lipophilic drugs, WHI- 05 and WHI-07, which exhibit potent anti-HIV and contraceptive activity. As AIDS is spread largely through sexual intercourse, the development of a dual action vaginal spermicidal microbicide to curb mucosal viral transmission, as well as to provide fertility control would have a tremendous impact world wide. D'Cruz and Uckun46"48 investigated the formulation of 2 wt% of the lipophilic drugs in a microemulsion-based gel, composed of Phospholipon 90G and Captex 300 as the oil phase, with Pluronic F68 and Cremophor EL as surfactants, and seaspan carragennan and Xantral as gelling agents. The microemulsions were gelled to obtain the necessary viscosity for the gel-microemulsion formulation. Under the conditions of their intended use, intravaginal application of the gel-microemulsions containing 2 wt% of drug in a rabbit model resulted in marked contraceptive activity, as well as exhibiting a lack of toxicity. Therefore, as a result of its dual anti-HIV and Recent Advances in Microemulsions as Drug Delivery Vehicles 157 spermicidal activities, the drug-containing gels shows unique clinical potential as a vaginal prophylactic contraceptive for women who are at a high risk of acquiring HIV by heterosexual transmission. 4.7. Nasal Nasal route has been demonstrated as being a possible alternative to the intravenous route for the systemic delivery of drugs. In addition to rapid absorption and avoidance of hepatic first-pass metabolism, the nasal route allows the preferential delivery of drug to the brain via the olfactory region, and is therefore a promising approach for the rapid-onset delivery of CMS medications. The solution-like feature of microemulsions could provide advantages over emulsions in terms of the sprayability, dose uniformity and formulation physical stability. Li et al.lu developed a diazepam-containing ethyl laurate-in-water microemulsion, stabilized by Tween 80 and containing propylene glycol and ethanol as cosolvents for the rapid-onset intranasal delivery of diazepam. A single isotropic region, which was considered to be a bicontinuous microemulsion, was seen at high surfactant concentrations but at various Tween 80: propylene glycol: ethanol ratios. Increasing Tween 80 concentration increased the microemulsion area, microemulsion viscosity, and the amount of water and oil solubilized. In contrast, increasing ethanol concentration produced the opposite effect. A microemulsion consisting of 15 wt% ethyl laurate, 15 wt% water and 70 wt% Tween 80:propylene glycohethanol at a 1:1:1 weight ratio contained 41 mg/mL of the poorly-water soluble diazepam. The nasal absorption of diazepam from the formulation was fairly rapid with a maximum drug plasma concentration being obtained within 2 to 3 min, while bioavailability at 2hrs post-administration was ~50% of that obtained with intravenous injection. Zhang et al.192 attempted to prepare an oil-in-water microemulsion, containing a high concentration of nimodipine, suitable for brain uptake via the intranasal route of delivery. Three microemulsion systems stabilized by either Cremophor RH 40 or Labrasol, and containing a variety of oils, namely isopropyl myristate, Labrafil M 1944CS and Maisine 35-1, were developed and characterized. The nasal absorption of the drug from the three microemulsions was studied in rats. The formulation composed of 8 wt% Labrafil M 1944CS, 30 wt% Cremophor RH 40/ethanol (3:1 weight ratio) and water solubilized up to 6.4 mg/mL of drug and exhibited no ciliotoxicity. After intranasal administration, the peak plasma concentration was obtained of 1 hr, while the absolute bioavailability was ~32%. Significantly, uptake of the drug in the olfactory bulb after nasal administration was three times that which was obtained from intravenous injection. In addition, the ratios of the AUC in brain tissues and cerebrospinal fluid to that in plasma obtained after nasal administration 1 58 Lawrence & Warisnoicharoen were significantly higher than those seen after administration. In conclusion, the microemulsion system appears to be a promising approach for the intranasal delivery of nimodipine. Richter and Keipert51 investigated the in vitro permeability of the highly lipophilic material, androstenedione, across excised bovine nasal mucosa, porcine cornea and an artificial cellulose membrane. In order to control release, the two microemulsion formulations studied contained either hydroxypropyl-yScyclodextrin or propylene glycol. Both microemulsions were prepared from 5 wt% isopropyl myristate, 20 wt% Cremophor EL and water. The permeation of the drug through the three tissues was influenced by the microemulsion. For example, the apparent permeability coefficients (Papp) of the drug from the microemulsions across nasal mucosa did not differ from the Papp of the drug contained in solution. In the case of the other two membranes, release from both of the microemulsion formulations exhibited extended time lags, so no Papp could be calculated. It seems that the composition of the microemulsion had a greater impact on the Papp of cornea than on the Papp of the other tissues. The structure of the different membranes is probably responsible for the observed differences in permeation. 4.8. Pulmonary Emulsions and (to a far lesser extent) microemulsions have been investigated as vehicles for pulmonary delivery. By far, the most widely studied systems are those containing fluorocarbon oil and are stabilized by a (predominately) fluorinated surfactant. Fluorocarbon oils are of pharmaceutical interest because of their biological inertness and their high (and unique) ability to dissolve gas, which means they can support the exchange of the respiratory gases in the lungs. In addition, a fluorocarbon oil, namely perfluorooctylbromide, is in Phase 11:111 clinical trials in the United States, for the treatment of acute respiratory distress by liquid ventilation. It should be noted that en-large hydrocarbon surfactants are ineffective solubilizers in fluorocarbon-based systems. Instead, fluorocarbon surfactants are required. To date, fluorocarbon-based (micro)emulsions have been investigated for use as oil-in-water systems for in vivo oxygen delivery (blood substitutes), targeted systems for diagnosis and therapy, and water-in-fluorocarbon systems for pulmonary drug delivery.40'102 Water-in-perfluorooctylbromide microemulsions have been shown to deliver homogeneous and reproducible doses of a tracer (caffeine) using metered-dose inhalers (pMDI) pressurized with hydrofluoroalkanes (HFAs).27 Lecithin-based reverse microemulsions have also been investigated as a means of pulmonary drug delivery.170'171 In these studies, dimethylethyleneglycol (DMEG) and hexane were used as models for the propellants, dimethyl ether (DME) and Recent Advances in Microemulsions as Drug Delivery Vehicles 159 propane respectively. A combination of equilibrium analysis and component diffusion rate determination (by pulsed-field gradient [PFG]-NMR) and iodine solubilization experiments were used to confirm the formation of a microemulsion. Water soluble solutes, including selected peptides and fluorescently labeled polya„ 6-[N-(2-hydroxyethyl) D,L-aspartamide] were dissolved in the microemulsions in a lecithin- and water-dependent manner. Experiments with DME/lecithin demonstrated microemulsion characteristics similar to those in the model propellant and produced a droplet size and a fine particle fraction suitable for pulmonary drug delivery. Patel et al.uo have prepared water-in-hydrofluorocarbon (specifically 134a) microemulsions using a combination of fluorinated polyoxyethylene ether surfactants and a short chain hydrocarbon alcohol such as ethanol. In the absence of a hydrocarbon alcohol, only cosolvent systems, but not microemulsions, were formed. Due to the high molecular weight of the fluorocarbon surfactant, large concentrations of fluorocarbon surfactant are required to solubilize relatively small amounts of water compared with comparable hydrocarbon-based surfactants. This has obvious implications for the pharmaceutical application of such systems. To date, very little on the potential of oil-in-water microemulsions for pulmonary drug delivery has been investigated, yet they are attractive vehicles because of their ability to solubilize high amounts of drug.157 4.8.1. Antibacterials Al-Adham et al.6 demonstrated that microemulsion formulations have a significant antimicrobial action against planktonic populations of both Pseudomonas aeruginosa and Staphylococcus aureus (i.e. greater than a 6 log cycle loss in viability over a period as short as 60s). Transmission electron microscopy studies indicated that this activity may in part be due to significant losses in outer membrane structural integrity. Nevertheless, these results have implications for the potential use of microemulsions as antimicrobial agents against this normally intransigent microorganism. More recently, the same group6 have determined the antibiofilm activity of an oil-in-water microemulsion, prepared from 15wt% Tween 80, 6wt% pentanol and 3wt% ethyl oleate, by incubating the microemulsion with an established biofilm culture of Ps. aeruginosa PA01 for a period of 4hrs. The planktonic MIC of sodium pyrithione and the planktonic and biofilm MICs of cetrimide were used as positive controls and a biofilm was exposed to a volume of normal sterile saline as a treatment (negative) control. The results showed that exposure to the microemulsion resulted in a three log-cycle reduction in biofilm viability, as compared to a one long-cycle reduction in viability observed with the positive 1 60 Lawrence & Warisnoicharoen control treatments, suggesting that microemulsions are highly effective antibiofilm agents. 5. Conclusion As can be seen, microemulsions are attractive d r u g delivery vehicles that offer much scope for improving drug delivery. Although microemulsions have been seriously studied as a delivery vehicle in the last >20 years, there are few microemulsion products currently on the market. Comparing microemulsions with vesicular drug delivery systems, it is pertinent to note that it took >25 years before vesicles were commercially exploited as drug delivery vehicles, and this was with the immense research effort expended in their study. Microemulsions have by contrast been much less widely studied. It is only a matter of time before more microemulsion-based formulations appear on the market. References 1. Aboofazeli R, Mortazavi SA and Khoshnevis P (2003) In vitro release study of sodium salicylate from lecithin based phospholipid microemulsions. Iran } Pharm Res 95-101. 2. Aboofazeli R, PatelN, Thomas M and Lawrence MJ (1995) Investigations into the formation and characterization of phospholipid microemulsions. 4. Pseudo-ternary phasediagrams of systems containing water-lecithin-alcohol and oil — the influence of oil. Int] Pharm 125:107-116. 3. Agatonovic-Kustrin S and Alany RG (2001) Role of genetic algorithms and artificial neural networks in predicting the phase behavior of colloidal delivery systems. Pharm Res 18:1049-1055. 4. Agatonovic-Kustrin S, Glass BD, Wisch MH and Alany RG (2003) Prediction of a stable microemulsion formulation for the oral delivery of a combination of antitubercular drugs using ANN methodology. Pharm Res 20:1760-1765. 5. Agrawal GP, Juneja M, Agrawal S, Iain SK and Pancholi SS (2005) Preparation and characterization of reverse micelle based organogels of piroxicam. Pharmazie 59: 191-193. 6. Al-Adham ISI, Khalil E, Al-Hmoud ND, Kierans M and Collier PI (2000) Microemulsions are membrane-active, antimicrobial, self-preserving systems. / Appl Microbiol 89:32-39. 7. Al-Adham ISI, Al-Hmoud ND, Khalil E, Kierans M and Collier PI (2003) Microemulsions are highly effective anti-biofilm agents. Lett Appl Microbiol 36:97-100. 8. Alany RG, Agatonovic-Kustrin S, Rades T and Tucker IG (1999) Use of artificial neural networks to predict quaternery phase systems from limited experimental data. / Pharm Biomed Anal 19:443-452. 9. Alvarez-Figueroa MI and Blanco-Mendez } (2001) Transdermal delivery of methotrexate: Iontophoretic delivery from hydrogels and passive delivery from microemulsions. Int J Pharm 215:57-65. Recent Advances in Microemulsions as Drug Delivery Vehicles 1 61 10. Amselem S and Friedman D (1998) Submicron emulsions in drug targeting and delivery. Benita S. (eds) Harwood Academic: Amsterdam. 11. Anselem S, Beilin M and Garty N (1993) Submicron emulsion as ocular delivery system for adaprolol maleate, a soft b-blocker. Pharm Res 10(suppl):S025. 12. Angelico R, Ceglie A, Colafemmina G, Lopez F, Murgia S, Olsson U and Palazzo G (2005) Biocompatible lecithin organogels: Structure and phase equilibria. Langmuir 21: 140-148. 13. Arevalo MI, Escribano E, Calpena A, Domenech J and Queralt J (2004) Rapid skin anesthesia using a new topical amethocaine formulation: a preclinical study. Anesth Anal 98:1407-1412. 14. Atay NZ and Robinson BH (1999) Kinetic studies of metal ion complexation in glycerolin- oil microemulsions. Langmuir 15:5056-5064. 15. Attama AA, Nzekwe IT, Nnamani PO, Adikwu MU and Onugu CO (2003) The use of solid self-emulsifying systems in the delivery of diclofenac. Int J Pharm 262:23-28. 16. Attwood D (1994) Microemulsions, in Colloidal Drug Delivery Systems. Kreuter J (ed.) Dekker: New York. 17. Attwood D, Currie LRJ and Elworthy PH (1974) Studies of solubilized micellar solutions. 1. Phase studies and particle-size analysis of solutions formed with nonionic surfactants. / Coll Interf Sci 46:249-254. 18. Aviv H, Friedman D, Bar-Ilan A and Vered M (1996) Submicron emulsions as ocular drug delivery vehicles. US Patent 5496811. 19. Bagwe RP, Kanicky JR, Palla BJ, Patanjali PK and Shah DO (2001) Improved drug delivery using microemulsions: Rationale, recent progress, and new horizons. Crit Rev Ther Drug 18:77-140. 20. Becker MD, Kruse FE, Azzam L, Nobiling R, Reichling J and Volcker HE (1999) In vivo significance of ICAM-1 dependent leukocyte adhesion in early corneal angiogenesis. Invest Ophthalmol Vis Sci 40:612-618. 21. Beilin M, Bar-Ilan A and Amselem S (1995) Ocular retention time of submicron emulsion (SME) and the miotic response to pilocarpine delivered in SME. Invest Ophthalmol Vis Sci 36:S166. 22. Bello M, Colangelo D, Gasco MR, Maranetto F, Morel S, Podio V, Turco GL and Viano I (1994) Pertechnetate release from a water oil microemulsion and an aqueous-solution after subcutaneous injection in rabbits. ] Pharm Pharmacol 46:508-510. 23. Benita S and Levy MY (1993) Submicron emulsions as colloidal drug carriers for intravenous administration — comprehensive physicochemical characterization. / Pharm Sci 82:1069-1079. 24. Bhargava HN, Narurkar A and Lieb LM (1987) Using microemulsions for drug delivery. Pharmaceut Technol 11:46-54. 25. Brime B, Moreno MA, Frutos G, Ballesteros MP and Frutos P (2002) Amphotericin B in oil-water lecithin-based microemulsions: Formulation and toxicity evaluation. / Pharm Sci 91:1178-1185. 26. Brime B, Molero G, Frutos P and Frutos G (2004) Comparative therapeutic efficacy of a novel lyophilized amphotericin B lecithin-based oil-water microemulsion and 1 62 Lawrence & Warisnoicharoen deoxycholate-amphotericin B in immunocompetent and neutropenic mice infected with Candida albicans. Eur J Pharm Sci 22:451-458. 27. Butz N, Porte C, Courrier HM, Krafft MP and Vandamme TF (2002) Reverse waterin- fluorocarbon emulsions for use in pressurized metered-dose inhalers containing hydrofluoroalkane propellants. hit} Pharm 238:257-269. 28. Calvo P, Alonso MJ and Vila-Jato J (1996) Comparative in vitro evaluation of several colloidal systems, nanoparticles, nanocapsules, and nanoemulsions, as ocular drug carriers. } Pharm Sci 85:530-536. 29. Carli F, Chiellini EE, Bellich B, Macchiavelli S and Cadelli G (2005) Ubidecarenone nanoemulsified composite systems, hit J Pharm 291:113-118. 30. Castro D, Moreno MA and Lastres JL (1999) First-derivative spectrophotometric and LC determination of nifedipine in Brij 96 based oil/water/oil multiple microemulsions on stability studies. / Pharm Biomed Anal 26:563-572. 31. Ceschel GC, Maffei P, Moretti MDL, Peana AT and Demontis S (1998) In vitro permeation through porcine buccal mucosa of Salvia sclarea L. essential oil from topical formulations. STP Pharm Sci 8:103-106. 32. Chantrapornchai W, Clydesdale FM and McClements DJ (2001) Influence of relative refractive index on optical properties of emulsions. Food Res hit 34:827-835. 33. Chen H, Chang X, Weng T, Zhao X, Gao Z, Yang Y, Xu H and Yang X (2004) A study of microemulsion systems for transdermal delivery of triptolide. / Control Rel 98:427-436. 34. Chevalier Y and Zemb T (1990) The structure of micelles and microemulsions. Rep Prog Phy's 53:279-371. 35. Chu L-W, Hsiue G-H and Lin I-N (2005) Ultra-fine Ba2Ti902o powders synthesized by inverse microemulsion processing and their microwave dielectric properties, / Am Ceram Soc. 36. Constantinides PP (1995) Lipid microemulsions for improving drug dissolution and oral absorption: Physical and biopharmaceutical aspects. Pharm Res 12:1561-1572. 37. Constantinides PP, Scalart JP, Marcello LJ, Marks G, Ellens H and Smith PL (1994) Formulation and intestinal absorption enhancement evaluation of wateri-in-oil microemulsions incorporating medium-chain glycerides. Pharm Res 11:1385-1390. 38. Constantinides PP, Lancaster CM, Marcello J, Chiosone DC, Orner D, Hidalgo I, Smith PL, Sarkahian AB, Yiv SH and Owen AJ (1995) Enhanced intestinal absorption of an RGD peptide from water-in-oil nanoemulsions of different composition and size. J Control Rel 34:109-116. 39. von Corswant C, Thorean P and Engstrom S (1998) Triglyceride-based microemulsion for intravenous administration of sparingly soluble substances. ] Pharm Sci 87:200-207. 40. Courrier HM, Butz N and Vandamme TF (2002) Pulmonary drug delivery systems: Recent developments and prospects. Crit Rev Ther Drug 19:425-498. 41. Courrier HA, Vandamme TF and Krafft MP (2004) Reverse water-in-fluorocarbon emulsions and microemulsions obtained with a fluorinated surfactant. Coll Surf Phy sicochem Eng Asp 244:141-148. 42. Cui Z, Fountain W, Clark M, Jay M and Mumper RJ (2003) Novel ethanol-influorocarbon microemulsions for topical genetic immunization. Pharm Res 20:16-23. Recent Advances in Microemulsions as Drug Delivery Vehicles 163 43. Dalmora MED and Oliveira AG (1999) Inclusion complex of piroxicam with (ficyclodextrin and incorporation in hexadecyltnmethylammonium bromide based microemulsion. Int J Pharm 184:157-164. 44. Dalmora ME, Dalmora SL and Oliveira AG (2001) Inclusion complex of piroxicam with /S-cyclodextrin and incorporation in cationic microemulsion. In vitro drug release and in vivo topical anti-inflammatory effect. Int f Pharm 222:45-55. 45. Danielsson I and Lindman B (1981) The definition of a microemulsion. Coll Surf 3:391-392. 46. D'Cruz OJ and Uckun MH (2001) Gel-microemulsions as vaginal spermicides and intravaginal drug delivery vehicles. Contraception 64:113-123. 47. D'Cruz OJ and Uckun FM (2002) Pre-clinical safety evaluation of novel nucleoside analogue-based dual-function microbicides (WHI-05 and WHI-07). } Antimicrob Chem 50:793-803. 48. D'Cruz OJ and Uckun FM (2003) Contraceptive activity of a spermicidal aryl phosphate derivative of bromo-methoxyzidovudine (compound WHI-07) in rabbits. Fertil Sterio 9:864-872. 49. Djordjevic L, Primorac M and Stupar M (2005) In vitro release of diclofenac diethylamine from caprylocaproyl macrogolglycerides based microemulsions. Int} Pharm 296:73-79. 50. Eccleston J (1994) Microemulsions, in Encyclopedia of Pharmaceutical Technology, Vol 9, Swarbrick J and Boylan JC (eds.) Marcel Dekker: New York. 51. El-Aasser MS (1997) Polymeric Dispersions. Asua JM (ed.) Kluwer Academic Publications: The Netherlands. 52. Elbaz E, Zeevi A, Klang S and Benita S (1993) Positively charged submicron emulsions, a new type of colloidal drug carrier. Int f Pharm 96:R1-R6. 53. Escribano E, Calpena AC, Queralt J, Obach R and Domenech J (2003) Assessment of diclofenac permeation with different formulations: Anti-inflammatory study of a selected formula. Eur f Pharm Sci 19:203-210. 54. Evitts DP, Olejnik O, Musson DG and Bilgood AM (1991) Aqueous ophtalmic microemulsions of tepoxalin. European patent application 0 480 690 Al. 55. Faldt P and Bergenstahl B (1995) Fat encapsulation in spray dried food powders. / Assoc Offic Anal Chem 72:171-176. 56. Fang J, Wang J, Ng S-C, Chew C-H and Gan L-M (1997) Ultrafine zirconia powders via microemulsion processing route. Nanostruct Mater 8:499-505. 57. Feng L and Dexi LL (1995) Circulating emulsions (oil-in-water) as carriers for lipophilic drugs. Pharm Res 12:1060-1064. 58. Fialho SL and da Silva-Cunha A (2004) New vehicle based on a microemulsion for topical ocular administration of dexamethasone. Clin Exp Ophthalmol 32:626-632. 59. Florence AT and Attwood D (1998) Physicochemical Pharmacy. 3rd ed., Macmilliam Press, Ltd. 60. Forgiarini A, Esquena J, Gonzalez C and Solans C (2001) Formation of nanoemulsions by low-energy emulsification methods at constant temperature. Eangmuir 17:2076-2083. 1 64 Lawrence & Warisnoicharoen 61. Gallarate M, Gasco MR and Trotta M (1988) Influence of octanoic acid on membrane permeability of timolol from solutions and from microemulsions. Acta Pharm Technol 34:102-105. 62. Gallarate M, Gasco MR, Trotta M, Chetoni P and Saettone MF (1993) Preparation and evaluation in vitro of solutions and o/w microemulsions containing levobunolol as ion-pair. Int J Pharm 100:219-225. 63. Garty N and Lusky M (1994) Pilocarpine in submicron emulsion formulation for treatment of ocular hypertension: A phase II clinical trial. Invest Ophth Vis Sci 35:2175-2179. 64. Gasco MR, Pattarino F and Lattanzi IF (1990) Long-acting delivery systems for peptides: reduced plasma testosterone levels in male rats after a single injection. Int ] Pharm 62:119-123. 65. Gershanik T and Benita S (2000) Self-dispersing lipid formulations for improving oral absorption of lipophilic drugs. Eur J Pharm Biopharm 50:179-188. 66. Getie M, Wohlrab J and Neubert RRH (2005) Dermal delivery of desmopressin acetate using colloidal carrier systems. / Pharm Pharmacol 57:423^127. 67. Groves MJ and de Galindez DA (1976) The self-emulsifying action of mixed surfactants in oil. Acta Pharm Suet 13:361-372. 68. Gulsen D and Chauhan A (2005) Dispersion of microemulsion drops in HEMA hydrogel: A potential ophthalmic drug delivery vehicle. Int f Pharm 292:95-117. 69. Gupta RR, Jain SK and Varshney M (2005) AOT water-in-oil microemulsions as a penetration enhancer in transdermal drug delivery of 5-fluorouracil. Coll Surf Biointerf 41:25-32. 70. Haering G and Luisi PL (1986) Hydrocarbon gels from water-in-oil microemulsions. JPhys Chem 90:5892-5895. 71. Hasse A and Kiepert S (1997) Development and characterization of microemulsions for ocular application. Eur } Pharm Biopharm 43:170-183. 72. He L, Wang G and Zhang Q (2003) An alternative paclitaxel microemulsion formulation: Hypersensitivity evaluation and pharmacokinetic profile. Int} Pharm 250:45-50. 73. Hellweg T (2002) Phase structures of microemulsions. Curr Opin Coll Interf Sci 7:50-56. 74. Hu Z, Tawa R, Konishi T, Shibata N and Takada K (2001) A novel emulsifier, Labrasol, enhances gastrointestinal absorption of gentamicin. Life Sci 69:2899-2910. 75. Husband FA, Garrood MJ, Mackie AR, Burnett GR and Wilde PJ (2001) Adsorbed protein secondary and tertiary structures by circular dichroism and infrared spectroscopy with refractive index matched emulsions. / Agri Food Chem 49:859-866. 76. Hwang SR, Lim S-J, Park J-S and Kim C-K (2004) Phospholipid-based microemulsion formulation of alMrans-retinoic acid for parenteral administration. Int ] Pharm 276:175-183. 77. Iek AC, Eebi NC, Tirnaksiz F and Tay A (2005) A lecithin-based microemulsion of rhinsulin with aprotinin for oral administration: Investigation of hypoglycemic effects in non-diabetic and STZ- induced diabetic rats. Int J Pharm 298:176-185. 78. Imberg A and Engstrom S (2003) An increased throughput method for determination of phase diagrams/method development and validation. Coll Surf Physicochem EngAsp 221:109-117. Recent Advances in Microemulsions as Drug Delivery Vehicles 165 79. Itoh K, Matsui S, Tozuka Y, Oguchi T and Yamamoto K (2002) Improvement of physicochemical properties of N-4472 Part II: characterization of N-4472 microemulsion and the enhanced oral absorption. Int} Pharm 246:75-83. 80. Izquierdo P, Esquena J, Tadros TF, Dederen JC, Feng J, Garcia-Celma MJ, Azemar N and Solans C (2004) Phase behavior and nano-emulsion formation by the phase inversion temperature method. Langmuir 20:6594-6598. 81. Izquierdo P, Feng J, Esquena J, Tadros TF, Dederen JC, Garcia MJ, Azemar N and Solans C (2005) The influence of surfactant mixing ratio on nano-emulsion formation by the pit method. / Coll Interf Sci 285:388-394. 82. Jaitely V, Sakthivel T, Magee G and Florence AT (2004) Formulation of oil in oil emulsions: Potential drug reservoirs for slow release. / Drug Del Sci Tech 14:113-117. 83. Joussen AM, Rohrschneider K, Reichling J, Kirchhof B and Kruse FE (2000) Treatment of corneal neovascularization with dietary iso-avonoids and flavonoids. Exp Eye Res 71:483-487. 84. Junping WJ, Takayama K, Nagai T and Maitani Y (2003) Pharmacokinetics and antitumor effects of vincristine carried by microemulsions composed of PEG-lipid, oleic acid, vitamin E and cholesterol. Int} Pharm 251:13-21. 85. Jurkovic P, Sentjurc M, Gasperlin M, Kristl J and Pecar S (2003) Skin protection against ultraviolet induced free radicals with ascorbyl palmitate in microemulsions. Eur J Pharm Biopharm 56:59-66. 86. Kanga BK, Chonb SK, Kimb SH, Jeongc SY, Kimc MS, Choc SH, Leec HB and Khanga G (2004) Controlled release of paclitaxel from microemulsion containing PLGA and evaluation of anti-tumor activity in vitro and in vivo. Int} Pharm 286:147-156. 87. Kang LS, Jun HW and McCall JW (2000) Physicochemical studies of lidocaine-menthol binary systems for enhanced membrane transport. Int J Pharm 206:35-42. 88. Kantaria S, Rees GD and Lawrence MJ (1999) Gelatin-stabilised microemulsion-based organogels: Rheology and application in iontophoretic transdermal drug delivery. / Control Rel 60:355-365. 89. Kantaria S, Rees GD and Lawrence MJ (2003) Formulation of electrically conducting microemulsion-based organogels. Int ] Pharm 250:65-83. 90. Kawakami K, Yoshikawa T, Moroto Y, Kanaoka E, Takahashi K, Nishihara Y and Masuda K (2002a) Microemulsion formulation for enhanced absorption of poorly soluble drugs I. Prescription design. / Control Rel 81:65-74. 91. Kawakami K, Yoshikawa T, Hayashi T, Nishihara Y and Masuda K (2002b) Microemulsion formulation for enhanced absorption of poorly soluble drugs II. In vivo study. / Control Rel 81:75-82. 92. Ke W-T, Lin S-Y, Ho HO and Sheu MT (2005) Physical characterizations of microemulsion systems using tocopheryl polyethylene glycol 1000 succinate (TPGS) as a surfactant for the oral delivery of protein drugs. / Control Rel 102:489-507. 93. Khoo S-M, Porter CJH and Charman WN (2000) The formulation of halofantrine as either non-solubilising PEG 6000 or solubilising lipid based solid dispersions: Physical, stability and absolute bioavailability assessment. Int} Pharm 205:65-78. 166 Lawrence & Warisnoicharoen 94. Khoo S-M, Shackleford DM, Porter CJH, Edwards GA and Charman WN (2003) Intestinal lymphatic transport of halofantrine occurs after oral administration of a unit-dose lipid-based formulation to fasted dogs. Pharm Res 20:1460-1465. 95. Khoshenvis P, Mortazavi SA, Lawrence MJ and Aboofazeli R (1997) In vitro release of sodium salicylate from water-in-oil microemulsions. / Pharm Pharmacol 49(S4):47. 96. Kim C-K, Ryuu S-A, Park K-M, Lim S-J and Hwang S-J (1997) Preparation and physicochemical characterization of phase inverted water/oil microemulsion containing cyclosporin A. Int J Pharm 147:131-134. 97. Kim C-K, Shin H-J, Yang S-G, Kim J-H and Oh Y-K (2001a) Once-a-day oral dosing regimen of cyclosporin A: Combined therapy of cyclosporin A premicroemulsion concentrates and enteric coated solid-state premicroemulsion concentrates. Pharm Res 18:454-459. 98. Kim C-K, Cho Y-J and Gao Z-G (2001b) Preparation and evaluation of biphenyl dimethyl dicarboxylate microemulsions for oral delivery. / Control Rel 70:149-155. 99. Kim SK, Lee EH, Vaishali B, Lee S, Lee YK, Kim CY, Moon HT and Byun Y (2005) Tricaprylin microemulsion for oral delivery of low molecular weight heparin conjugates. / Control Rel 105:32-42. 100. Klang SH, Baszkin A and Benita S (1996) The stability of piroxicam incorporated in a positively charged submicron emulsion for ocular administration. Int } Pharm 132:33-44. 101. Kraeling MEK and Ritschel WA (1992) Development of a colonic release capsule dosage form and the absorption of insulin. Meth Find Exp Clin Pharmacol 14:199-209. 102. Krafft MP, Chittofrati A and Riess JG (2003) Emulsions and microemulsions with a fluorocarbon phase. Curr Opin Coll Interf Sci 8:251-258. 103. Kreilgaard M (2002) Influence of microemulsions on cutaneous drug delivery. Adv Drug Del Rev 54:S77-S98. 104. Lam AC and Schechter RS (1987) The theory of diffusion in microemulsion. / Coll Interf Sci 120:56-63. 105. Lawrence MJ (1994) Surfactant systems: Microemulsions and vesicles as vehicles for drug delivery. Eur } Drug Metab Pharmacokinet 3:257-269. 106. Lawrence MJ (1996) Microemulsions as drug delivery vehicles. Curr Opin Coll Inter Sci 1:826-832. 107. Lawrence MJ and Rees GD (2000) Microemulsion-based media as novel drug delivery systems. Adv Drug Del Rev 45:89-121. 108. Lee M-J, Lee M-H and Shim C-K (1995) Inverse targeting of drugs to reticuloendothelial system-rich organs by e:pid microemulsion emulsified with poloxamer 338. Int J Pharm 113:175-187. 109. Lee J-M, Park K-M, Lim S-J, Lee M-K and Kim C-K (2002) Microemulsion formulation of clonixic acid: Solubility enhancement and pain reduction. / Pharm Pharmacol 54:43-49. 110. Lee PJ, Langer R and Shastri VP (2003) Novel Microemulsion Enhancer Formulation for Simultaneous Transdermal Delivery of Hydrophilic and Hydrophobic Drugs. Pharm Res 20:264-269. Recent Advances in Microemulsions as Drug Delivery Vehicles 167 111. Lettow JS, Lancaster TM, Glinka CJ and Ying JY (2005) Small-angle neutron scattering and theoretical investigation of poly(ethylene oxide)-poly(propylene oxide)- poly(ethylene oxide) stabilized oil-in-water microemulsions. Langmuir 21:5738-5746. 112. Levy MY and Benita S (1989) Design and characterization of a submicronized o/w emulsion of diazepam for parenteral use. Int ] Pharm 54:103-112. 113. Levy MY and Benita S (1991) Short-term and long-term stability assessment of a new injectable diazepam submicron emulsion. } Parent Sci Tech 45:101-107. 114. Li L, Nandi I and Kim KH (2002) Development of an ethyl laurate-based microemulsion for rapid-onset intranasal delivery of diazepam. Int J Pharm 237:77-85. 115. Lladser M, Medrano C and Arancibia A (1968) The use of supports in the lyophilization of oil-in-water emulsions. / Pharm Pharmacol 20:450-455. 116. Lopez F, Venditti F, Ambrosone L, Colafemmina G, Ceglie A and Palazzo G (2004) Gelatin microemulsion-based gels with the cationic surfactant cetyltrimethylammonium bromide: A self-diffusion and conductivity study. Langmuir 20:9449-9452. 117. Lu YY, Xia Q, Xia Y Ma QH and Gu N (2005) Studies on the phase behaviors of drugloading microemulsions. Acta Physicochimica Sinica 21:98-101. 118. Lundberg BB (1997) A submicron lipid emulsion coated with amphipathic polyethylene glycol for parenteral administration of paclitaxel (Taxol). / Pharm Pharmacol 49:16-21. 119. Lyons KC, Charman WN, Miller R and Porter CJH (2000) Factors limiting the oral bioavailability of N-acetylglucosaminyl-N-acetylmuramyl dipeptide (GMDP) and enhancement of absorption in rats by delivery in a water-in-oil microemulsion. Int } Pharm 199:17-28. 120. Malcolmson C and Lawrence MJ (1993) A comparison of the incorporation of model steroids into nonionic micellar and microemulsion systems. / Pharm Pharmacol 45:141-143. 121. Malcolmson C and Lawrence MJ (1995) Three-component non-ionic oil-in-water microemulsions using polyoxyethylene ether surfactants. Coll Surf B Biointerf 4:97-109. 122. Malcolmson C, Satra C, Kantaria S, Sidhu A and Lawrence MJ (1998) Effect of oil on the level of solubilization of testosterone propionate into nonionic oil-in-water microemulsions. / Pharm Sci 87:109-116. 123. Malcolmson C, Barlow DJ and Lawrence MJ (2002) Light-scattering studies of testosterone enanthate containing soybean oil/Ci8:iEio /water oil-in-water microemulsions. } Pharm Sci 91:2317-2331. 124. Mei Z, Chen H, Weng T, Yang Y and Yang X (2003) Solid lipid nanoparticle and microemulsions for topical delivery of triptolide. Eur } Pharm Biopharm 56:189-196. 125. Melamed S, Kurtz S, Greenbaum A, Haves JF, Neumann R and Garty N (1994) Adaprolol maleate in submicron emulsion, a novel soft /S-blocking agent, is safe and effective in human studies. Invest Ophthalmol Vis Sci 35:1387. 126. Moreno MA, Ballesteros MP and Frutos P (2003) Lecithin-based oil-in-water microemulsions for parenteral use: Pseudoternary phase diagrams, characterization and toxicity studies. / Pharm Sci 92:1428-1437. 168 Lawrence & Warisnoicharoen 127. Mistry VV, Hassan HN and Robison DJ (1992) Effect of lactose and protein on the microstructure of dried milk. Food Struct 11:73-82. 128. Myers SL and Shively ML (1992) Preparation and characterization of emulsifiable glasses: Oil-in-water and water-in-oil-in-water emulsions, / Coll Interf Sci 149:271-278. 129. Muchtar S, Almong S, Torraca MT, Saettone MF and Benita S (1992) A submicron emulsion as ocular vehicle for delta-8- tetrahydrocannabinol: Effect on intraocular pressure in rabbits. Ophtalmic Res 24:142-149. 130. Muchtar S, Abdulrazik M, Benita S (1997) Ex vivo permeation study of indomethacin from a submicron emulsion through albino rabbit cornea. / Control Rel 44:55-64. 131. Nakajima H, Tomomasa S and Okabe M (1993) Preparation of Nanoemulsions, in Proceedings of First Emulsion Conference, Paris, EDS: Paris, Vol. 1. 132. Nazzal S, Guven N, Reddy IK and Khan MA (2002) Preparation and characterization of coenzyme QlO-Eudragit solid dispersion. Drug Dev Ind Pharm 28:49-57. 133. Naveh N, Muchtar S and Benita S (1994) Pilocarpine incorporated into a submicron emulsion vehicle causes an unexpectedly prolonged ocular hypotensive effect in rabbits. / Ocul Pharmacol 10:509-520. 134. Newton M, Petersson J, Podczeck F, Clarke A and Booth S (2001) The influence of formulation variables on the properties of pellets containing a self-emulsifying mixture. J Pharm Sci 90:987-995. 135. Nyqvist-Mayer AA, Brodin AF, Frank SG (1985) Phase distribution studies on an oilwater emulsion based on a eutectic mixture of lidocaine and prilocaine as the dispersed phase. / Pharm Sci 74:1192-1195. 136. Osborne DW, Ward AJ and O'Neill KJ (1991) Microemulsions as topical drug delivery vehicles: In vitro transdermal studies of a model hydroplilic drug. / Pharm Pharmacol 43:451-454. 137. Paolino D, Ventura CA, Nistico S, Puglisi G and Fresta M (2002) Lecithin microemulsions for the topical administration of ketoprofen: Percutaneous adsorption through human skin and in vivo human skin tolerability. Int ] Pharm 244:21-31. 138. Park K-M and Kim C-K (1999) Preparation and evaluation of flurbiprofen-loaded microemulsion for parenteral delivery. Int J Pharm 181:173-179. 139. Park K-M, Lee M-K, Hwang K-J and Kim C-K (1999) Phospholipid-based microemulsions of flurbiprofen by the spontaneous emulsification process. Int J Pharm 183:145-154. 140. Patel P, Marlow M and Lawrence MJ (2003) Formation of fluorinated nonionic surfactant microemulsions in hydrofluorocarbon 134a (HFC 134a). J Coll Interf Sci 258:345-353. 141. Patil P, Joshi P and Paradkar A (2004) Effect of formulation variables on preparation and evaluation of gelled self-emulsifying drug delivery system (SEDDS) of ketoprofen. AAPS Pharm Sci Tech 5: Article 42. 142. Pattarino F, Marengo E, Gasco MR and Carpignano R (1993) Experimental-design and partial least-squares in the study of complex-mixtures — microemulsions as drug carriers. Int} Pharm 91:157-165. 143. Pedersen GP, Faldt P, Bergenstahl B and Kristensen HG (1998) Solid state characterisation of a dry emulsion: A potential drug delivery system. Int J Pharm 171:257-270. Recent Advances in Microemulsions as Drug Delivery Vehicles 169 144. Podlogar F, Gasperlin M, Tomsk M, Jamnik A and Bester Rogac M (2004) Structural characterisation of water-Tween 40/Imwitor 308-isopropyl myristate microemulsions using different experimental methods. Int J Pharm 276:115-128. 145. Porter CJH, Kaukonen AM, Boyd BJ, Edwards GA and Charman WN (2004) Susceptibility to lipase-mediated digestion reduces the oral bioavailability of danazol after administration as a medium-chain lipid-based microemulsion formulation. Pharm Res 21:1405-1412. 146. Pouton CW (1997) Formulation of self-emulsifying drag delivery systems. Adv Drug Del Rev 25:47-58. 147. Priano L, Albani G, Brioschi A, Calderoni S, Lopiano L, Rizzone M, Cavalli R, Gasco MR, Scaglione F, Fraschini F, Bergamasco B and Mauro A (2004) Transdermal apomorphine permeation from microemulsions: A new treatment in parkinson's disease. Movement Disorders 19:937-942. 148. Quellet C and Eicke H-F (1986) Mutual gelation of gelatin and water-in-oil microemulsions. Chimia 40:233-238. 149. Richardson CJ, Mbanefo A, Aboofazeli R, Lawrence MJ and Barlow DJ (1997) Prediction of phase behavior in microemulsion systems using artificial neural networks. / Coll Inter'/ Sci 187:296-303. 150. Richter A and Steiger-Trippi K (1961) Untersuchungen uber die zerstaubungstrocknung von emulgierten arzneizubereitungen. Pharm Acta Helv 36:322-337. 151. Richter T and Keipert S (2004) In vitro permeation studies comparing bovine nasal mucosa, porcine cornea and artificial membrane: Androstenedione in microemulsions and their components. Eur J Pharm Biopharm 58:137-143. 152. Ritschel WA (1991) Microemulsions for improved peptide absorption from the gastrointestinal-tract. Meth Findings Exp Clin Pharmacol 13:205-220. 153. Ross EA, Savage KA, Utley LJ and Tebbett IR (2004) Insect repellant interactions: Sunscreens enhance deet(w, n-diethyl-m-toluamide) absorption. Drug Metabolism Dispos 32:783-785. 154. Rowe RC, Sheskey PJ and Weller PJ (2003) Pharmaceutical Excipients Handbook, Pharmaceutical Press and American Pharmaceutical Association, Dundee. 155. Santiago RM, Fialho SL and Silva-Cunha A (2003) Design and characterization of an oral delivery system for insulin administration. STP Pharma Sci 13:377-380. 156. Santos Magalhaes NS, Cave G, Seiller M and Benita S (1991) The stability and in vitro, release kinetics of a clofibride emulsion. Int J Pharm 76:225-237. 157. Satra C, Thomas M and Lawrence MJ (1995) Formulating oil-in-water microemulsions for pulmonary drug delivery, in Drug Delivery to the Lungs IV, The Aerosol Society (Bristol), London. 158. Scherlund M, Malmsten M, Holmqvist P and Brodin A (2000) Thermosetting microemulsions and mixed micellar solutions as drug delivery systems for periodontal anesthesia. Int} Pharm 194:103-116. 159. Schwab M and Stuhn B (2000) Relaxation phenomena and development of structure in a physically crosslinked nonionic microemulsion studied by photon correlation spectroscopy and small angle x-ray scattering. / Chem Phys 112:6461-6471. 170 Lawrence & Warisnoicharoen 160. Schulman JH, Stoechenius W and Prince LM (1959) Mechanism of formation and structure of microemulsions by electron microscopy. / Phys Chem 63:1677-1680. 161. Schurtenberger P and Cavaco C (1994) The static and dynamic structure factor of polymer-like lecithin reverse micelles. / Physique II 4:305-317. 162. Siebenbrodt I and Keipert S (1993) Poloxamer-systems as potential ophthalmic microemulsions. Eur J Pharm Biopharm 39:25-30. 163. Sha X, Yan G, Wu Y, Li J and Fang X (2005) Effect of self-microemulsifying drug delivery systems containing Labrasol on tight junctions in Caco-2 cells. Eur J Pharm Sci 24:477-486. 164. Shiokawa T, Hattori Y, Kawano K, Ohguchi Y, Kawakami H, Toma K and Maitani Y (2005) Effect of polyethylene glycol linker chain length of folate-linked microemulsions loading aclacinomycin A on targeting ability and antitumor effect in vitro and in vivo. Clin Cancer Res 11:2018-2025. 165. Shukla A, Krause A and Neubert RRH (2003) Microemulsions as colloidal vehicle systems for dermal drug delivery. Part IV: Investigation of microemulsion systems based on a eutectic mixture of lidocaine and prilocaine as the colloidal phase by dynamic light scattering. / Pharm Pharmacol 55:741-748. 166. Siebenbrodt I and Keipert S (1993) Poloxamer-systems as potential ophtalmic microemulsions. Eur J Pharm Biopharm 39:25-30. 167. Sinn C (2004) When jelly gets the blues — audible sound generation with gels and its origin. / Non-Cryst Sol 347:11-17. 168. Sintov AC and Shapiro L (2004) New microemulsion vehicle facilitates percutaneous penetration in vitro and cutaneous drug bioavailability in vivo. J Control Rel 95:173-183. 169. Solans C, Esquena J, Forgiarini A, Uson N, Morales D, Izquierdo P, Azemar N and Garcia MJ (2002) Adsorption and Aggregation of Surfactants in Solution. Mittal KL and Shah DO, (eds.) Marcel Dekker: New York. 170. Sommerville ML, Cain JB, Johnson CS and Hickey AJ (2000) Lecithin inverse microemulsions for the pulmonary delivery of polar compounds utilizing dimethylether and propane as propellants. Pharmaceut Dev Technol 5:219-230. 171. Sommerville ML, Johnson CS, Cain JB, Rypacek F and Hickey AJ (2002) Lecithin microemulsions in dimethyl ether and propane for the generation of pharmaceutical aerosols containing polar solutes. Pharmaceut Dev Technol 7:273-288. 172. Strickley RG (2004) Solubilizing excipients in oral and injectable formulations. Pharm Res 21:201-230. 173. Tadros TF, Izquierdo P, Esquena J and Solans C (2004) Formation and stability of nanoemulsions. Adv Coll Interf Sci 108-109:303-318. 174. Taha MO, Al-Ghazawi M, Abu-Amara H and Khalil E (2002) Development of quantitative structure-property relationship models for pseudoternary microemulsions formulated with nonionic surfactants and cosurfactants: Application of data mining and molecular modeling. Eur } Pharm Sci 15:461^178. 175. Taha MO, Abu-Amara H, Al-Ghazawi M and Khalil E (2005) QSPR modeling of pseudoternary microemulsions formulated employing lecithin surfactants: Application of data mining, molecular and statistical modeling. Int} Pharm 295:135-155. Recent Advances in Microemulsions as Drug Delivery Vehicles 171 176. Tenjarla S (1999) Microemulsions: An overview and pharmaceutical applications. Crit Rev Ther Drug Can Sys 16:461-521. 177. Ugelstadt J, El-Aassar MS and Vanderhoff JW (1973) Emulsion polymerization — initiation of polymerization in monomer droplets. / Polym Sci 11:503-513. 178. Vandamme TF (2002) Microemulsions as ocular drug delivery systems: Recent developments and future challenges. Prog Retin Eye Res 21:15-34. 179. Valduga CJ, Fernandes DC, Lo Prete AC, Azevedo, CHM, Rodrigues DG and Maranhao RC (2003) Use of a cholesterol-rich microemulsion that binds to low-density lipoprotein receptors as vehicle for etoposide. / Pharm Pharmacol 55:1615-1622. 180. Valenta C and Schultz K (2004) Influence of carrageenan on the rheology and skin permeation of microemulsion formulations. / Control Rel 95:257-265. 181. Varshney M, Morey TE, Shah DO, Flint JA, Moudgil BM, Seubert CN and Dennis DM (2004) Pluronic microemulsions as nanoreservoirs for extraction of bupivacaine from normal saline. J Am Chem Soc 126:5108-5112. 182. Vyas SP, Jain CP, Kaushik A and Dixit VK (1992) Preparation and characterisation of griseofulvin dry emulsion. Pharmazie 47:463-464. 183. Wang J, Maitani Y and Takayma K (2002) Antitumor effects and pharmacokinetics of aclacinomycin A carried by injectable emulsions composed of vitamin E, cholesterol and PEG-lipid. / Pharm Sci 91:1128-1134. 184. Wang J, Chong PP, Ng SC and Gan LM (1997) Microemulsion processing of manganese zinc ferrites. Mat Lett 30:217-221. 185. Warisnoicharoen W, Lansley AB and Lawrence MJ (2000) Nonionic oil-in-water microemulsions: The effect of oil type on phase behaviour. Int} Pharm 98:7-27. 186. Watnasirichaikul S, Rades T, Tucker IG and Da vies NM (2002) In vitro release and oral bioactivity of insulin in diabetic rats using nanocapsules,dispersed in biocompatible microemulsion. / Pharm Pharmacol 54:473^180. 187. Xu QY, Nakajima M, Nabetani H, Ichikawa S and Liu XQ (2002) Factors affecting the properties of ethanol-in-oil emulsions. Food Sci Tech Res 8:36-41. 188. Yamamoto A, Taniguchi T, Rikyuu K, Tsuji T, Fujita T, Murakami M and Muranishi S (1994) Effects of various protease inhibitors on the intestinal absorption and degradation of insulin in rats. Pharm Res 11:1496-1500. 189. Yang S, Gursoy RN, Lambert G and Benita S (2004) Enhanced oral absorption of paclitaxel in a novel self-microemulsifying drug delivery system with or without concomitant use of p-glycoprotein inhibitors. Pharm Res 21:261-270. 190. Zhang Z-Q and Lu B (2001) Advances in microemulsions as a vehicle of drug delivery system. Chin J Pharm 32:139-142. 191. Zhang L, Sun X, Xiang D and Zhang ZR (2004a) Formulation and physicochemical characterization of norcantharidin microemulsion containing lecithin-based surfactants. / Drug Del Sci Tech 14:461^69. 192. Zhang Q, Jiang X, Jiang W, Lu, W, Su L and Shi Z (2004b) Preparation of nimodipineloaded microemulsion for intranasal delivery and evaluation on the targeting efficiency to the brain. Int J Pharm 275:85-96. This page is intentionally left blank 8 Lipoproteins as Pharmaceutical Carriers Suwen Liu, Shining Wang and D. Robert Lu 1. Introduction Large protein structures (in nanometer range) may be utilized as pharmaceutical carriers of drugs and DNA for targeted and other specialized delivery in biological systems. Lipoproteins are such structures which function as natural biological carriers and transport various types of lipids in blood circulation. There are many studies suggesting that lipoproteins can serve as efficient carriers for anticancer drugs, gene or other type of compounds.1-4 Previous results showed that hydrophobic cytotoxic drugs could be incorporated into lipoproteins, without changing the integrity of native lipoprotein structure. Lipoproteins as drug carriers offer several advantages.5-6 Firstly, they are endogenous components and do not trigger immunological response. They have a relatively long half-life in the circulation. Secondly, they have small particle size in the nanometer range, allowing the diffusion from vascular to extravascular compartments. Thirdly, lipoproteins can potentially serve as the carriers for targeted drug delivery through specific cellular receptors. For example, low density lipoprotein (LDL)-drug complexes may target cancer cells which, in many cases, have higher LDL-receptor expression than normal cells. Fourthly, the lipid core of lipoprotein provides a suitable compartment for carrying hydrophobic drugs. As a result of these advantages, lipoproteins have received wide attentions in recent years in the development of drug-targeting strategies to use them as specialized delivery vehicles. This review intends to provide an overview of the development and the specialized utilization of lipoproteins for drug delivery purpose. After 173 1 74 Liu, Wang & Lu briefly introducing the structure and the basic biological functions of lipoproteins, we will focus on four classes of lipoproteins, namely, chylomicron, very low-density lipoprotein (VLDL), low-density lipoprotein (LDL), and high-density lipoprotein (HDL), as the carriers for various drug compounds. Cholesterol-rich emulsions (LDE) and artificial lipoproteins as drug carriers will also be discussed. 2. The Structure of Lipoproteins Lipoproteins, as implied by their names, are biological protein-lipid complexes. Lipoproteins serve the functions of carrying hydrophobic substances in blood circulation and transporting them to various biological sites through the protein-receptor interactions.6,7 The size of lipoproteins is in the nanometer range and they have a spherical shape with complex physicochemical properties. Figure 1 illustrates the general structure of lipoprotein. The hydrophobic core contains water-insoluble substances and is surrounded by a polar shell. The polar shell consists of phospholipids, unesterified cholesterol and different types of apolipoproteins, which bind to various cellular receptors for specific biological functions. Therefore, based on their physicochemical properties, lipoproteins are nanoemulsions with targeting functions provided by the apolipoproteins. Owing to the unique structure of lipoproteins, they can serve a two-mode function of solubilizing hydrophobic substances, including triglycerides and cholesteryl esters, within the nanoemulsion core and allow themselves to float in blood circulation. Lipoproteins can be classified into five major classes, based on their densities from gradient ultracentrifugation experiments. The lipoprotein classification includes chylomicron, very low-density lipoprotein (VLDL), intermediate-density lipoprotein (IDL), low-density lipoprotein (LDL), and high-density lipoprotein (HDL). These classes of lipoproteins have different sizes, different protein to lipid ratios and different types of apolipoproteins. In general, chylomicrons act on transporting dietary triacylglycerols and cholesterol to the adipose tissue and liver, following the absorption of dietary hydrophobic substances from the intestines. Very Fig. 1. General structure of lipoproteins. Lipoproteins as Pharmaceutical Carriers 1 75 Table 1 Physicochemical properties of lipoproteins. Lipoprotein Transport Route Size(nm) Protein (%) Total lipids (%) Chylomicron Intestines to Liver 75-1200 1.5-2.5 97-99 VLDL Liver to tissues 30-80 5-10 90-95 IDL Liver to tissues 25-35 15-20 80-85 LDL Liver to tissues 18-25 20-25 75-80 HDL Tissues to liver 5-12 40-55 45-60 low density lipoprotein, intermediate density lipoprotein and low density lipoprotein work at different stages to transport triacylglycerols and cholesterol from the liver to various tissues. High density lipoprotein brings endogenous cholesterol from the tissues back to the liver. The general physicochemical properties of lipoproteins can be seen in Table 1. 3. Chylomicron as Pharmaceutical Carrier Chylomicrons are assembled in the intestine from the absorbed dietary lipids and transported by lymphatic system. Although most of the drugs administered orally are absorbed directly into the portal blood to reach the systemic circulation, an alternative absorption route through the intestinal lymphatics may be available for hydrophobic drugs. It is estimated that a high hydrophobicity (log o / w partition co-efficient > 5) of drug molecules is required for intestinal lymphatic transport.8 Chylomicrons can thus potentially serve as an important natural carrier for hydrophobic drugs to transport through lymphatic system.9 It is known that targeted drug delivery through the lymphatics is important for anti-viral drug molecules for the protection of B- and T-lymphocytes, which maintain relatively higher concentrations through the lymphatics than the systemic circulation. Chylomicrons have a much larger size than other lipoproteins, and thus can carry more drug molecules from the absorption site. With the presence of food, chylomicrons are the predominant lipoprotein produced by the small intestine to carry dietary lipids efficiently because of its large size. Various types of bioactive molecules have been incorporated into reconstituted chylomicron structure for delivery purposes. In gene delivery, Hara et a/.10,11 developed reconstituted chylomicron which incorporated a hydrophobic DNA complex and used it as an in vivo gene transfer vector. They found that the DNA-incorporated chylomicrons induced a high gene expression in mouse liver after the reconstituted chylomicron was administered through portal vain injection. Furthermore, it was also reported that artificial, protein-free lipid emulsions could be utilized to model the metabolism of lymph chylomicron in rats, not only in the initial partial 176 Liu, Wang & Lu hydrolysis by lipoprotein lipase, but also in the delivery of a remnant-like particle to the liver.12 As a targeted therapeutic approach to hepatitis B, anti-viral iododeocyuridine was incorporated into recombinant chylomicrons, resulting in the drug molecules being selectively targeted to the liver parenchymal cells.13 It has been suggested that chylomicron can serve as a special carrier for liver cell targeting.14 Due to the targetability, this approach could be further developed as an effective therapy for hepatitis B patients. 4. VLDL as Pharmaceutical Carrier VLDL particles have a size range of 30-80 nm. They are assembled in the endoplasmic reticulum (ER) and matured in Golgi apparatus of hepatocytes before secretion.15 After entering into the plasma, VLDL particles are catabolized by a series of biochemical actions, including apolipoprotein exchange of apoC-I, apoCII, apoC-III, and apoE; lipolysis by triglyceride lipase; and cell-surface receptormediated uptake. As lipolysis proceeds, VLDL particles become smaller and are eventually converted to IDL. Some of the IDL particles are rapidly taken up by hepatocytes via a receptor-mediated mechanism while others undergo further hydrolysis before being converted to LDL. The catabolism route of VLDL suggests the possibility of using VLDL as a drug carrier for targeted delivery. ApoE is a protein ligand present on the surface of VLDL and it is well known that the receptor of apoE is overexpressed on some types of cancer cells. Therefore, VLDL can potentially serve as an antineoplastic drug carrier. As a drug carrier, VLDL is an interesting candidate because it contains a relatively small amount of proteins (about 5-10 % protein) and a large amount of triglycerides (about 50-65% within the emulsion core) which can be used to solubilize hydrophobic substances sufficiently. By mimicking the compositions and structure of VLDL, Shawer et al. developed a VLDL-resembling phospholipid nanoemulsion system that carried a new anti-tumor boron compound for targeted delivery to cancer cells.16 The nanoemulsion demonstrated sufficient capability to solubilize the hydrophobic compound. The structure of the phospholipid nanoemulsion was verified based on the changes in the molecular surface area and the molecular volume of each component of the nanoemulsion when the particle size is changed (from different size fractions). If certain molecules are located at the core of nanoemulsion, their numbers per overall volume should not be changed when the particle size is increased. If certain molecules are located at the surface of nanoemulsion, their numbers per overall volume should decrease when particle size is increased. This is because the overall surface area decreases when particle size is increased. Similar to the natural lipoproteins, it was demonstrated that phospholipid was predominately Lipoproteins as Pharmaceutical Carriers 177 located at the surface and the hydrophobic substances, triolein and cholesteryl oleate, were mainly located in the core of the phospholipid nanoemulsion. Recently, a similar nanoemulsion formulation was used to encapsulated quantum dots (QD) as a new bioimaging carrier.17 Quantum dots (QDs) are semiconductor nanocrystals that are emerging as unique fluorescence probes in biomedicine.18-21 When manufactured, most of the quantum dots have organic ligand coating on their surface and are extremely hydrophobic. The research goal was to encapsulate QDs in phospholipid nanoemulsion and to examine the physical stability, size distribution and their interactions with cancer cells. It was found that CdSe QDs can be efficiently encapsulated in the phospholipid nanoemulsion. The QD-encapsulated phospholipid nanoemulsion are stable and interact well with cultured cells to deliver the QDs inside the cells for fluorescence imaging.17 In other studies, it has been demonstrated that cytotoxic drugs such as 5-fluorouracil (5-FU), 5-iododeoxyuridine (IudR), doxorubicin (Dox), and vindesine can be effectively incorporated into VLDL, and the resultant complexes showed effective cytotoxicity to human carcinoma cells.22 5. LDL as Pharmaceutical Carrier LDL (18-25 nm) is not directly synthesized in human body. Instead, most of them are formed through the VLDL pathway. LDL is the major circulatory lipoprotein for the transport of cholesterol and cholesteryl esters, and it can be internalized by cells via LDL receptor-mediated endocytosis. The internalization process of LDL has been well characterized and the understanding of the mechanism can potentially help the designing of the drug targeting strategy through the LDL receptor (Fig. 2). The binding of dephosphorylated adaptor protein to the plasma membrane LDL Receptors (. ( . X l B l O l f c . H K . * ^ . , . ^ Cell , HMGCoA t ACAT T Cholesterol \ LDL Receptors mug •> ^-.t, ^ v f i ' * o„o -* LDL Binding —• Internalization —•Drug Release —^Regulation Fig. 2. LDL receptor pathway and targeted drug delivery. 1 78 Liu, Wang & Lu initiates the formation of coated pits which are covered by the protein clathrin. The receptors from the surrounding regions of the plasma membrane shift towards the binding site for internalization. Apolipoproteins including apo B-100 and apo E are recognized and bound by the LDL receptor on the cell surface to form a complex which is internalized into the coated pits. After internalization of the LDL, the coated pits are pinched off and within a very short time, they shed off their clathrin coating. The internalized LDL particle is transferred to endocytotic vesicles or endosomes. Due to the acidic pH within the endosomes, LDL dissociates from its receptor. This is followed by the fusion of the endosomes with lysosomes which contain hydrolases. The protein component of LDL is broken into free amino acids, while the cholesteryl ester component is cleaved by lysosomal lipase. The free cholesterol is released and incorporated into the cell membrane. Excess cholesterol is re-esterified by the action of acyl-CoA:cholesterol acyltransferase (ACAT). Among various lipoproteins, LDL has been widely studied as a drug carrier for targeted and other specialized deliveries, because many types of cancer cells show elevated expression of LDL receptors than the corresponding normal cells.23-26 Comparing with chylomicron, VLDL, and IDL, LDL also has a longer serum halflife of 2-4 days,27 making it a desirable drug carrier. Low density lipoprotein was found to be suitable as carriers for cytotoxic drugs to target cancer cells. LDLdrug complexes can be formed through various processes without changing the lipoprotein integrity.28-31 5.1. LDL as anticancer drug carriers Doxorubicin (Dox) is widely used in treating different tumors. Its main side effects are cadiotoxicity and multidrug resistance, especially during prolonged treatment in the patients. LDL has been studied as a target carrier for Dox in nude mice, bearing human hepatoma HepG2 cells.32 Both in vitro and in vivo studies indicated that when Dox was incorporated into LDL, the multidrug resistance could be circumvented and the cardiotoxicity could be reduced as well.33 Kader and Pater22 used VLDL, LDL and HDL as carriers to deliver four cytotoxic drugs, 5-fluorouracil (5-FU), 5-iododeoxyuridine (IUdR), doxorubicin (Dox) and vindesine. They found that significant drug loading was achieved in all three classes of lipoproteins, consistent with the sizes and hydrophobicity of the drug. Experiments were carried out to examine the changes in drug cytotoxicity against HeLa cervical and MCF-7 breast carcinoma cells, after the incorporation into lipoprotein. The results demonstrated that VLDL-drug complex did not affect their IC50 on both HeLa and MCF-7 cell lines, when compared with free drugs. However, the IC50 values of LDL- and HDLdrug complexes were significantly lower compared with free drugs. Their studies further indicated that drugs were incorporated into lipoproteins without disrupting Lipoproteins as Pharmaceutical Carriers 1 79 their integrity; drugs remained in their stable forms inside lipoproteins; and human LDL and HDL could be particularly useful in the delivery of antineoplastic drugs. 5.2. LDL as carriers for other types ofbioactive compounds Although LDL has been widely studied as a carrier to deliver anticancer compounds, it may also be useful to deliver other types of bioactive compounds. LDL may serve as a carrier for site-specific delivery of drugs to atherosclerotic lesions.34 When dexamethasone palmitate (DP), a steroidal anti-inflammatory drug, was incorporated in LDL, an inhibitory effect of this complex on foam cell formations was demonstrated. The study indicated that LDL could potentially carry DP to atherosclerotic lesions.34 Fluorophore-labeled LDL was also used for optical imaging in tumors diagnosis. For example, carbocyanine dyes can be used as near infrared (NIR) optical imaging probes with long wavelength absorption, high extinction coefficients and high fluorescence quantum yield. In vitro confocal microscopic study and ex vivo low-temperature fluorescent scanning demonstrated that carbocynine-labled LDL probes, Dil-LDL, could be selectively delivered to B16/HepG2 tumor cells and the corresponding animal tumors via the LDL receptor pathway.35 It was also proposed that Dil is located and oriented in the phospholipid monolayer when it binds to LDL. 5.3. LDL for gene delivery LDL has also been investigated as gene delivery carriers. Comparing with viral gene-delivery vectors and some other types of non-viral gene delivery vectors, the LDL system shows certain advantages in transfection efficiency and safety considerations.5 Several LDL based gene delivery systems have been reported. Kim's group developed a terplex system which comprises LDL, lipidized poly(Llysine) and plasmid DNA. The complex had a diameter of about 100 nm. The studies showed high efficiency to deliver plasmid DNA to smooth muscle cells and fibroblast cells.36,37 In addition, a novel LDL-DNA complex was formulated by Khan et al.38 LDL was cationized using carbodiimide and the modified lipoprotein complex significantly increased the DNA binding capacity with improved stability. The novel delivery system also demonstrated the ability to target cells through LDL receptor.38 6. HDL as Pharmaceutical Carriers Among various lipoproteins, HDL has the smallest size with a diameter of 5-12 nm. It shares common structural characteristics with other lipoproteins. However, its 180 Liu, Wang & Lu polar shell contributes more than 80% of the total mass. Newly synthesized HDL hardly contains any cholesteryl ester molecules. Cholesteryl esters are gradually added to the particles by lecithin via enzymatic reaction: cholesterol acyltransferase (LCAT), which is a 59-kD glycoprotein associated with HDL. The interaction of HDL with cells appears similar to that of LDL.39 Although the function of HDL in the human body is not well-defined, it generally transports excess cholesterol and cholesteryl esters from various tissue cells back to the liver. Comparing with other types of lipoproteins, small size and fast internalization by tumor cells are the major advantages of utilizing HDL for drug delivery and targeting. HDL has mainly been utilized for the delivery of water insoluble anticancer drugs through the targeting function.40-41 When the anticancer drug, Taxol, was incorporated into HDL, stable complexes were formed and they were examined for cancer-cell targeting.41 Reconstituted HDL was explored as a drug carrier system for a lipophilic prodrug, IDU-OI2.42 The studies indicated that the lipophilic prodrug could be efficiently incorporated into reconstituted HDL particles. This approach may also be useful to encapsulate other lipophilic derivatives of water-soluble drugs. The utilization of HDL for drug targeting may lead to a more effective therapy for infectious diseases, such as hepatitis B, since the HDL-drug complexes were demonstrated to be selectively taken by parenchymal liver cells.42 Comparing with free drugs in cytotoxicity assays, the IC values of HDL-drug complexes were significantly decreased, about 2.5 to 23-fold lower.22 Interestingly, it was observed that HDL-drug complex specifically increased the cytotoxicity to carcinoma cells. Earlier studies showed that HDL could increase the sensitivity of HeLa cells to the cytotoxic effects of Dox.43 Similar to LDL-drug complex, the lipoprotein receptor pathway appears to be involved in the interactions between HDL-drug complex and cancer cells. 7. Cholesterol-rich Emulsions (LDE) as Pharmaceutical Carriers LDE is a lipid based formulation, an emulsion with a lipid structure resembling LDL particle and it is made without protein incorporation. Essentially, it is composed of a cholesteryl ester core surrounded by a monolayer of phospholipids. Comparing with native LDL, LDE is removed from the blood circulation more rapidly.44 It appears possible that LDE can acquire apoE and other apolipoproteins from native lipoproteins in plasma. ApoE can be recognized by LDL receptors, thus allowing the binding of LDE to the receptors. However, it is known that LDE binds to receptors through apoE, but not through apoBlOO. The interaction between apoE and the receptor appears stronger than that of apoBlOO.45 Lipoproteins as Pharmaceutical Carriers 181 LDE is considered as a potential carrier for anticancer drugs to deliver chemotherapeutic agents to neoplastic cells. Although there is no protein in the LDE formulations, previous studies showed that the LDL receptor could still play an important role in the cellular uptake of these lipid complexes.46-56 LDE binds to low-density lipoprotein receptors which are upregulated in cancer cells, leading to a higher concentration in neoplastic tissues.24-57 LDE-carmustine complex was studied with a neoplastic cell line and its biodistribution was studied in mice. An exploratory clinical study was also conducted. The result showed that the uptake of LDE-carmustine complex by tumor was several fold greater than the uptake by the corresponding normal tissue. The association of carmustine with LDE preserves the tumor-cytotoxicity of carmustine with reduced side effects.58 Preliminary clinical study59 was also carried out using LDE-carmustine complex to treat patients with advanced cancers. The results demonstrated that the systemic toxicity of the drug was significantly reduced. Rodrigues et ah investigated the formulation of LDE containing antineoplastic compound paclitaxel.55 The experiments revealed a 75% incorporation efficiency and the stable complex of the drug molecules incorporated in LDE emulsion. Its LD50 was ten-fold greater than that of a commercial formulation of paclitaxel. It was suggested by the authors that the cellular uptake and the cytotoxic activity of LDEpaclitaxel complex might be mediated by the LDL receptors due to the cholesterol moiety in the LDE formulation.55 In addition to LDE, artificial lipoproteins have been constructed. Several research groups have developed various types of artificial lipoproteins.44-60-62 Most of them constructed the artificial lipoproteins by incorporating natural apoB protein into lipid microemulsion for the purpose of examining the lipoprotein metabolism. Artificial lipoproteins containing poly-lysine has also been investigated as the DNA carrier for cellular transfection, with the potential to reduce the cytotoxicity and to improve the transfection efficiency.63-64 8. Concluding Remark Lipoproteins are natural nanostructures in biological systems. They have unique physicochemical properties which may be utilized as pharmaceutical carriers for drug compounds and other bioactive substances. Owing to the structural diversity of lipoproteins, including chylomicron, VLDL, LDL and HDL, various specialized delivery systems may be developed to fully utilize their delivery potentials. New nanostructures, such as LDE and artificial lipoproteins, can also be constructed to mimic the structure of natural lipoproteins. As these new nanostructures are built from scratch, they may be more efficient in encapsulating drug and other bioactive molecules, and more effective for specialize drug delivery. 1 82 Liu, Wang & Lu References 1. Smidt PC and van Berkel TJC (1990) LDL-mediated drug targeting. Crit Rev Ther Drug Carrier Syst 7:99-119. 2. Filipowska D, Filipowski T, Morelowska B, Kazanowska W, Laudanski T, Lapinjoki S, Akerland M and Breeze A (1992) Treatment of cancer patients with a low density lipoprotein delivery vehicle containing a cytotoxic drug. Cancer Chemother Pharmacol 29: 396-400. 3. Firestone RA (1994) Low density lipoprotein as a vehicle for targeting antitumor compounds to cancer cells. Bioconjug Chem 5:105-113. 4. van Berkel TJC (1993) Drug targeting: Application of endogenous carriers for site specific delivery of drug. / Control Rel 24:145-155. 5. Pan G, 0ie S and Lu DR (2003) Biological protein nanostructures and targeted drug delivery. Lu DR and 0ie S (eds.) in Cellular Drug Delivery: Principles and Practice, pp. 217-234. 6. Chung NS and Wasan KM (2004) Potential role of the low-density lipoprotein receptor family as mediators of cellular drug uptake. Adv Drug Del Rev 56:1315-1334. 7. Sarkar R, Halpern DS, Jacobs SK and Lu DR (2002) LDL-receptor mediated drug targeting to malignant tumors. Muzykantov VR and Torchilin VP (eds.) in Biomedical Aspects of Drug Targeting. (Kluwer Academic Publisher), pp. 327-345. 8. Charman WN and Stella VJ (1986) Estimating the maximal potential for intestinal lymphatic transport of lipophilic drug molecules, hit} Pharm 34:175-178. 9. Shen H, Howies P and Tso P (2001) From interaction of lipidic vehicles with intestinal epithelial cell membranes to the formation and secretion of chylomicrons. Adv Drug Del Rev 50:S103-S125. 10. Hara T, Liu F, Liu DX and Huang L (1997) Emulsion formulations as a vector for gene delivery in vitro and in vivo. Adv Drug Del Rev 24:265-271. 11. Hara T, Tan Y and Huang L (1997) In vivo gene delivery to the liver using reconstituted chylomicron remnants as a novel nonviral vector. Proc Natl Acad Sci USA 94: 14547-14552. 12. Redgrave TG and Maranhao RC (1985) Metabolism of protein-free lipid emulsion models of chylomicrons in rats. Biochem Biophys Acta 835:104-112. 13. Rensen PCN, De Vrueh RLA, van Berkel TJC (1996) Targeting hepatitis B therapy to the liver: Clinical pharmacokinetic considerations. Clin Pharmacokinet 31:131-155. 14. Rensen PC, van Dijk MC, Havenaar EC, Bijsterbosch MK, Kruijt JK and van Berkel TJ (1995) Selective liver targeting of antivirals by recombinant chylomicrons — a new therapeutic approach to hepatitis B. Nat Med l(3):221-5. 15. Olofsson SO, Bjursell G, Bostrom K, Carlsson P, Elovson J, Protter AA, Reuben MA and Bondjers G (1987) Apolipoprotein B: Structure, biosynthesis and role in the lipoprotein assembly process. Atherosclerosis 68:1-17. 16. Shawer M, Greenspan P, 0ie S and Lu DR (2002) VLDL-resembling phospholipidsubmicron emulsion for cholesterol-based drug targeting. / Pharm Sci 91:1405-1413. Lipoproteins as Pharmaceutical Carriers 183 17. Liu S, Lee CM, Wang S and Lu DR (2006) A new bioimaging carrier for quantum dot nanocrystals — phospholipid nanoemulsion mimicking natural lipoprotein core. Drug Del 13:159-164. 18. Dubertret B, Skourides P, Norris DJ, Noireaux V, Brivanlou AH and Libchaber A (2002) In vivo imaging of quantum dots encapsulated in phospholipid micelles. Science 298: 1759-1762. 19. Gao X, Cui Y, Levenson RM, Chung LW and Nie S (2004) In vivo cancer targeting and imaging with semiconductor quantum dots. Nat Biotechnol 22:969-976. 20. Bruchez M, Moronne M, Gin P, Weiss S and Alivisatos AP (1998) Semiconductor nanocrystals as fluorescent biological labs. Science 281:2013-2016. 21. Chan WCW and Nie S (1998) Quantum dot biocojugates for ultrosensitive nonisotopic detection. Science 281:2016-2018. 22. Kader A and Pater A (2002) Loading anticancer drugs into HDL as well as LDL has little affect on properties of complexes and enhances cytotoxicity to human carcinoma cell. / Control Rel 80:29^4. 23. Alexopoulos CG, Blatsios B and Avgerinos A (1987) Serum lipids and lipoprotein disorders in cancer patients. Cancer 3065-3070. 24. Ho YK, Smith RG, Brown MS and Goldstein JL (1978) Low density lipoprotein (LDL) receptor activity in human acute myelogenous leukemia cells. Blood 52:1099-1114. 25. Klock JC and Pieprzyk JK (1979) Cholesterol, phospholipids, and fatty acids of normal immature neutrophils: Comparison with acute myeloblastic leukemia cells and normal neutrophils. / Lipid Res 20:908-911. 26. Nakagawa T, Ueyama Y, Nozaki S, Yamashita S, Menju M, Funahashi T, Takemura KK, Kubo M, Tokunaga K, Tanaka T, Yagi M and Matsuzawa Y (1994) Marked hypocholesterolemia in a case with adrenal adenoma — Enhanced Catabolism of low density lipoprotein (LDL) via the LDL receptors of tumor cells. / Clin Endocrinol Metabol 80: 92-96. 27. Kader A, Davis PJ, Kara M and Liu H (1998) Drug targeting using low density lipoprotein (LDL): Physicochemical factors affecting drug loading into LDL particles. / Control Rel 55:231-243. 28. Firestone RA (1994) Low density lipoprotein as a vehicle for targeting antitumor compounds to cancer cells. Bioconjug Chem 5:105-113. 29. Filipowska D, Filipowski T, Morelowska B, Kazanowska W, T. Laudanski T, Lapinjoki S, Akerland M and Breeze A (1992) Treatment of cancer patients with a low density lipoprotein delivery vehicle containing a cytotoxic drug. Cancer Chemother Pharmacol 29:396^00. 30. de Smidt PC and van Berkel TJC (1990) LDL-mediated drug targeting. Crit Rev Ther Drug Can Syst 7:99-119. 31. van Berkel TJC (1993) Drug targeting: Application of endogenous carriers for site specific delivery of drug. / Control Rel 24:145-155. 32. Chu ACY, Tsang SY, Lo EHK and Fung KP (2001) Low density lipoprotein as a targeted carrier for doxorubicin in nude mice bearing tumor hepatoma HepG2 cells. Life Sci 70:591-601. 184 Liu, Wang & Lu 33. Lo EHK, Ooib VEL and Fung KP (2002) Circumvention of multidrug resistance and reduction of cardiotoxicity of doxorubicin in vivo by coupling it with low density lipoprotein. Life Sci 72:677-687. 34. Tauchi Y, Takase M, Zushida I, Chono S, Sato J, Ito K and Morimoto K (1999) Preparation of a complex of dexamethasone palmitate-low density lipoprotein and its effect on foam cell formation of murine peritoneal macrophages. / Pharma Sci 88:709-714. 35. Li H, Zhang Z, Blessington D, Nelson DS, Zhou R, Lund-Katz S, Chance B, Glickson JD and Zheng G (2004) Carbocyanine labeled LDL for optical imaging of tumors. Acad Radiol 11:669-677. 36. Kim JS, Maruyama A, Akaike T and Kim SW (1997) Ln vitro gene expression on smooth muscle cells using a terplex delivery system. / Control Rel 47:51-59. 37. Kim JS, Kim BI, Maruyama A, Akaike T and Kim SW (1998) A new non-viral DNA delivery vector: The terplex system. / Control Rel 53:175-182. 38. Khan Z, O. Hawtrey A and Ariatti M (2003) New cationized LDL-DNA complexes: Their targeted delivery to fibroblasts in culture. Drug Del 10:213-220. 39. Steinberg D (1996) A docking receptor for HDL cholesterol esters. Science 271: 46CM61. 40. Rensen PC, de Vrueh RL, Kuiper J, Bijsterbosch MK, Biessen EA and van Berkel TJ (2001) Recombinant lipoproteins: Lipoprotein-like lipid particles for drug targeting. Adv Drug Del Rev 47(2-3):251-276. 41. Lacko AG, Nair M, Paranjape S, Johnso S and McConathy WJ (2002) High density lipoprotein complexes as delivery vehicles for anticancer drugs. Anticancer Res 22: 2045-2049. 42. Bijsterbosch MK, Schouten D and van Berkel TJ (1994) Synthesis of the dioleoyl derivative of iododeoxyuridine and its incorporation into reconstituted high density lipoprotein particles. Biochemistry 33:14073-14080. 43. Chassany O, Urien S, Claudepierre P, Bastian G and Tillement JP (1996) Comparative serum protein binding of anthra- cycline derivatives. Cancer Chemother Pharmacol 38: 571-573. 44. Hirata RDC, Hirata MH, Mesquita CH, Cesar TB and Maranhao RC (1999) Effects of apolipoprotein B-100 on the metabolism of a lipid microemulsion model in rats. Biochim Biophys Acta 1437:53-62. 45. Innerarity TL and Mahley RW (1978) Enhanced binding by cultured human broblasts of apo-E-containing lipoproteins as compared with low density lipoproteins. Biochemistry 17:1440. 46. Versluis AJ, Rump ET, Rensen PC, van Berkel TJ and Bijsterbosch MK (1998) Synthesis of a lipophilic daunorubicin derivative and its incorporation into lipidic carriers developed for LDL receptor-mediated tumor therapy. Pharm Res 15:531-537. 47. Versluis AJ, Rensen PC, Rump ET, van Berkel TJ and Bijsterbosch MK (1998) Lowdensity lipoprotein receptor-mediated delivery of a lipophilic daunorubicin derivative to B16 tumours in mice using apolipoprotein E-enriched liposomes. Br J Cancer 78: 1607-1614. Lipoproteins as Pharmaceutical Carriers 185 48. Amin K, Wasan KM, Albrecht RM and Heath TD (2002) Cell association of liposomes with high fluid anionic phospholipids content is mediated specifically by the LDL and its receptor. / Pharm Sci 91:1233-1244. 49. Amin K, Ng K, Brown CS, Bruno MS and Heath TD (2001) LDL induced association of anionic liposomes with cells and delivery of contents as shown by the increase in potency of liposome dependent drugs. Pharm Res 18:914-921. 50. Amin K and Heath TD (2001) LDL-induced association of anionic liposomes with cells and delivery of contents: II. Interaction of liposomes with cells in serum-containing medium. / Control Rel 73:49-57. 51. Lakkaraju A, Rahman Y and Dubinsky JM (2002) Low-density lipoprotein-related protein mediates the endocytosis of anionic liposomes in neurons. / Biol Chem 277: 15085-15092. 52. Rensen PC, Schiffelers RM, Versluis AJ, Bijsterbosch MK, Van Kuijk-Meuwissen ME and van Berkel TJ (1997) Human recombinant apolipoprotein E-enriched liposomes can mimic low-density lipoproteins as carriers for site-specific delivery of antitumor agents. Mol Pharmacol 52:445^55. 53. Koller-Lucae SKM, Schott H and Schwendener RA (1997) Interactions with human blood in vitro and pharmacokinetic properties in mice of liposomal N4-octadecyl-l-h- D-arabinofuranosylcytosine, a new anticancer drag. J Pharmacol Exp Ther 282:1572-1580. 54. Koller-Lucae SKM, Schott H and Schwendener RA (1999) Lowdensity lipoprotein and liposome mediated uptake and cytotoxic effect of N4-octadecyl-l-h-Darabinofuranosylcytosine in Daudi lymphoma cells. Br ] Cancer 80:1542-1549. 55. Rodrigues DG, Covolan CC, Coradi ST, Barboza R and Maranhao RC (2002) Use of a cholesterol-rich emulsion that binds to low-density lipoprotein receptors as a vehicle for paclitaxel. / Pharm Pharmacol 54:765-772. 56. Versluis AJ, Rump ET, Rensen PC, van Berkel TJ and Bijsterbosch MK (1999) Stable incorporation of lipophilic daunorubicin prodrug into apolipoprotein E-exposing liposomes induces uptake of prodrug via low-density lipoprotein receptor in vivo. J Pharmacol Exp Ther 289:1-7. 57. Maranhao RC, Roland IA, Toffoletto O, Ramires JA, Gone, alves RP, Mesquita CH and Pileggi P (1997) Plasma kinetic behavior in hyperlipidemic subjects of a lipidic microemulsion that binds to LDL receptors. Lipids 32:627-633. 58. Maranhao RC, Graziani SR, Yamaguchi N, Melo RF, Latrilha MC, Rodrigues DG, Couto RD, Schreier S and Buzaid AC (2002) Association of carmustine with a lipid emulsion: In vitro, in vivo and preliminary studies in cancer patients. Cancer Chemother Pharmacol 49:487-^198. 59. Hungria VTM, Latrilha MC, Rodrigues DG, Bydlowski SP, Chiattone CS and Maranhao RC (2004) Metabolism of a cholesterol-rich microemulsion (LDE) in patients with multiple myeloma and a preliminary clinical study of LDE as a drug vehicle for the treatment of the disease. Cancer Chemother Pharmacol 53:51-60. 60. Reisinger RE and Atkinson D (1990) Phospholipid/cholesteryl ester microemulsion containing unesterified cholesterol: Model systems for low density lipoproteins. / Lipid Res 31:849-858. 186 Liu, Wang & Lu 61. Chun PW, Brumbauge EE and Shiremann RB (1986) Interaction of human low density lipoprotein and apolipoprotein B with ternary lipid microemulsion. Biophys Chem 25:223-241. 62. Maranhao RC, Cesar TB, Pedroso-Mariani SR, Hirata MH and Mesquita CH (1993) Metabolic behavior in rats of a nonprotein microemulsion resembling low-density lipoprotein. Lipids 28:691-695. 63. Pan G, Shawer M, 0 i e S and Lu DR (2003) In vitro gene transfection to glioma cells using a novel and less cytotoxic artificial lipoprotein delivery system. Pharm Res 20:738-745. 64. Alanazi F, Fu ZF and Lu DR (2004) Effective transfection of rabies DNA vaccine in cell culture using an artificial lipoprotein carrier system. Pharm Res 21:676-683. 9 Solid Lipid Nanoparticles as Drug Carriers Karsten Mader 1. Introduction: History and Concept of SLN Nanosized drug delivery systems have been developed to overcome one or several of the following problems: (i) low and highly variable drug concentrations after peroral administration due to poor absorption, rapid metabolism and elimination (ii) poor drug solubility which excludes i.v. injection of an aqueous drug solution (iii) drug distribution to other tissues combined with high toxicity (e.g. cancer drugs). Several systems, including micelles, liposomes, polymer nanoparticles, nanoemulsions and nanocapsules have been developed. During the last few years, solid lipid nanodispersions (SLN) have attracted increased attention. It is the aim of this chapter to discuss the general features of these systems with respect to manufacturing and performance. In the past, solid lipids have been mainly used for rectal and dermal applications. In the beginning of the 80s, Speiser and coworkers developed solid lipid microparticles (by spray drying)1 and "Nanopellets for peroral administration".2 These Nanopellets were produced by dispersion of melted lipids with high speed mixers or ultrasound. The manufacturing process was unable to reduce all particles to the submicron size. A considerable amount of microparticles was present in the samples. This might not be a serious problem for peroral administration, but it excludes an intravenous injection. "Lipospheres", described by Domb, are 187 188 Mader close related systems.3-5 They are also produced by means of high shear mixing or ultrasound and also often contain considerable amounts of microparticles. The quality of the SLN has been significant improved by the use of high pressure homogenization (HPH) in the early 90s.6-8 Higher shear forces and a better distribution of the energy force more effective particle disruption, compared with high shear mixing and ultrasound. Dispersions obtained by this HPH are called Solid Lipid Nanoparticles (SLN™). Most SLN dispersions produced by high pressure homogenization (HPH) are characterized by an average particle size below 500 nm and a low microparticle content. Other production procedures are based on the use of organic solvents HPH/solvent evaporation9 or on dilution of microemulsions.10'11 The ease and efficacy of manufacturing lead to an increased interest in SLN. Furthermore, it has been claimed that SLN combine the advantages yet without inheriting the disadvantages of other colloidal carriers.12,13 Proposed advantages include: • Possibility of controlled drug release and drug targeting • Increased drug stability • High drug pay load • Feasibility to incorporate lipophilic and hydrophilic drugs • No biotoxicity of the carrier • Avoidance of organic solvents • No problems with respect to large scale production and sterilization. However, during the last years, some of these claims have been questioned and it became evident that SLN are rather complex systems which possess not only advantages but also serious limitations. 2. Solid Lipid Nanoparticles (SLN) Ingredients and Production 2.1. General ingredien ts General ingredients include solid lipid(s), emulsifier(s) and water. The term lipid is used generally in a very broad sense and includes triglycerides (e.g. tristearine, hard fat), partial glycerides (e.g. Imwitor), pegylated lipids, fatty acids (stearic acid), steroids (e.g. cholesterol) and waxes (e.g. cetylpalmitate). All classes of emulsifiers (with respect to charge and molecular weight) have been used to stabilize the lipid dispersion. The most frequently used compounds include different kinds of poloxamer, polysorbates, lecithin and bile acids. It has been found that the combination of emulsifiers might prevent particle agglomeration more efficiently. Solid Lipid Nanoparticles as Drug Carriers 189 Unfortunately, poor attention has been given by most investigators to the physicochemical properties of the lipid. Fatty acids, partial glycerides and other polar lipids are able to interact with water to much a greater extent, compared with a long chain triglyceride (e.g. they might form liquid crystalline phases). Polar lipids will have much more interaction with stabilizers (e.g. formation of mixed micelles), while more lipophilic lipids will show phase segregation. The author strongly suggests to follow the proposal by Small and to classify lipids according to their interactions with water.14 2.2. SLN preparation 2.2.1. High shear homogenization and ultrasound High shear homogenization and ultrasound are dispersion techniques which were initially used for the production of solid lipid nanodispersions.1-3 Both methods are widespread and easy to handle. However, dispersion quality is often poor due to the presence of microparticles. Furthermore, metal contamination has to be considered if ultrasound is used. Ahlin et al. used a rotor-stator homogenizer to produce SLN from different lipids, including trimyristin, tripalmitin, tristearin, partial glycerides (Witepsol®W35, Witepsol®H35) and glycerol tribehenate (Compritol®888) by meltemulsification. 15 They investigated the influence of different process parameters, including emulsification time, stirring rate and cooling conditions on the particle size and the zeta potential. Poloxamer 188 was used as steric stabilizer (0,5%w/w). For Witepsol®W35 dispersions, the following parameters were found to produce the best SLN quality: stirring 8min at 20000rpm, the optimum cooling conditions lOmin at 5000 rpm at room temperature. In contrary, the best conditions for Dynasan® 116 dispersions were 10 min emulsification at 25 000 rpm and 5 min of cooling at 5000 rpm in cool water (T = 16°C). An increased stirring rate did not significantly decrease the particle size, but improved the polydispersity index slightly. No general rule can be derived from differences in the established optimum emulsification and cooling conditions. In most cases, average particle sizes in the range of 100-200 nm were obtained in this study. 2.3. High pressure homogenization (HPH) HPH has emerged as a very reliable and probably the most powerful technique for the preparation of SLN. HPH has been used for many years for the production of nanoemulsions for parenteral nutrition. In most cases, scaling up represents zero or limited problems. High pressure homogenizers push a liquid with high pressure (100-2000 bar) through a narrow gap (in the range of few microns). The 190 Mader fluid accelerates on a very short distance to very high velocities. The high shear stress disrupts the particles down to the submicron range. Typical lipid contents are in the range of 5 to 10%. Even higher lipid concentrations (up to 40%) have been homogenized to lipid nanodispersions.16 Two general approaches of the homogenization step, the hot and the cold homogenization techniques, can be used for the production of SLN.17,18 In both cases, a preparatory step involves the drug incorporation into the bulk lipid by dissolving the drug in the lipid melt. 2.4. Hot homogenization The hot homogenization is carried out at temperatures above the melting point of lipid. Therefore, it is in fact the homogenization of an emulsion. A preemulsion of the drug loaded lipid melt and the aqueous emulsifier phase (same temperature) is obtained by high-shear mixing device (Ultraturrax). The quality of the preemulsion is very important for the final product quality. In general, higher temperatures result in lower particle sizes due to the decrease of the viscosity of the inner phase.19 However, high temperatures may also increase the degradation rate of the drug and the carrier. The homogenization step can be repeated several times. It should be kept in mind however, that HPH increases the temperature of the sample (approximately 10°C for 500 bar20). In most cases, 3 to 5 homogenization cycles at 500 to 1500 bar are sufficient. Increasing the homogenization pressure or the number of cycles often results in an increase of the particle size due to particle coalescence, which occurs as a result of the high kinetic energy of the particles.21 It is important to note that the primary product of the hot homogenization is a nanoemulsion due to the liquid state of the lipid. Solid particles are expected to be formed by the following cooling of the sample to room temperature, or to temperatures below. Due to the small particle size and the presence of emulsifiers, lipid crystallization may be highly retarded and the sample may remain as a supercooled melt for several months.22 2.5. Cold homogeniza Hon Cold homogenization has been developed to overcome the following three problems of the hot homogenization technique: (1) Temperature induced drug degradation (2) Drug distribution into the aqueous phase during homogenization (3) Complexity of the crystallization step of the nanoemulsion, leading to several modifications and/or supercooled melts The first preparatory step for cold homogenization is the same as in the hot homogenization procedure and includes the solubilization of the drug in the melt of the Solid Lipid Nanoparticles as Drug Carriers 191 bulk lipid. However, the following steps differ. The drug containing melt is rapidly cooled. The high cooling rate favors a homogenous distribution of the drug within the lipid matrix. The solid, drug containing lipid is milled to microparticles. Typical particle sizes obtained by means of ball or mortar milling are in the range of 50 to 100 microns. Low temperatures increase the fragility of the lipid, and therefore favor particle disruption. The solid lipid microparticles are suspended in a chilled emulsifier solution. The preemulsion is subjected to HPH at or below room temperature. An effective temperature control and regulation is needed in order to ensure the unmolten state of the lipid due to the increase in temperature during homogenization.20 In general, compared with hot homogenization, larger particle sizes and a broader size distribution are observed in cold homogenized samples.23 A modified version of this technique has been recently published by the group of Miiller-Goymann. They dispersed a solid 1:1 lecithin-hardfat mixture (described as solid reversed micelles) in Tween containing water using high pressure homogenization. 24 2.5.1. SLN prepared by solvent emulsification/evaporation The solvent emulsification/evaporation processes adapts techniques which have been previously used for the production of polymeric micro- and nanoparticles. The solid lipid is dissolved in a water-immiscible organic solvent (e.g. cyclohexane, or chloroform) that is emulsified in an aqueous phase. Upon evaporation of the solvent, a nanoparticle dispersion is formed by precipitation of the lipid in the aqueous medium. Westesen prepared nanoparticles of tripalmitate by dissolving the triglyceride in chloroform.25 This solution was emulsified into an aqueous phase by high pressure homogenization. The organic solvent was removed from the emulsion by evaporation under reduced pressure. The mean particle size ranges from approximately 30 to lOOnm depending on the lecithin/co-surfactant blend. Particles with very small diameters (30 nm) were obtained by using bile salts as co-surfactants. Comparable small particle size distributions were not achievable by melt emulsification of similar composition. The mean particle size depends on the concentration of the lipid in the organic phase. Very small particles could only be obtained with low fat loads (5 w%) related to the organic solvent. With increasing lipid content, the efficacy of the homogenization declines due to the higher viscosity of the dispersed phase. 2.5.2. SLN preparations by solvent injection The solvent injection method has been developed by Fessi to produce polymer nanoparticles.26 Nanoparticles were only produced with solvents which distribute very rapidly into the aqueous phase (e.g. ethanol, acetone, DMSO), while larger 192 Mader particle sizes were obtained with more lipophilic solvents. According to Fessi, the particle size is critically determined by the velocity of the distribution processes and only water miscible solvents can be used. The solvent injection method can also be used for the production of solid lipid nanoparticles.27'28 However, the method is limited to lipids which dissolve in the polar organic solvent. Advantages of the method are the avoidance of elevated temperatures and high shear stress. However, the lipid concentration in the primary suspension will be less compared with High-Pressure-Homogenization. Furthermore, the use of organic solvents clearly represents a drawback of the method. 2.5.3. SLN preparations by dilution of microemulsions or liquid crystalline phases SLN preparation techniques which are based on the dilution of microemulsions have been developed by Gasco and coworkers. Unfortunately, there is no common agreement within the scientific community about the definition of a microemulsion. One part of the scientific community understands under microemulsions high fluctuating systems which can be regarded as a critical solution, and therefore do not contain an inner and outer phase. This model has been confirmed by self-diffusion NMR studies of Lindman.29 In contrast, Gasco and other scientists understand microemulsions as two systems composed of an inner and outer phase (e.g. O/W-microemulsions). They are made by stirring an optical transparent mixture at 65-70°C, typically composed of a low melting lipid fatty acid (e.g. stearic acid), emulsifier (e.g. polysorbate 20, polysorbate 60, soy phosphatidylcholin, taurodeoxycholic acid sodium salt), co-emulsifiers (e.g. Butanol, Na-monooctylphosphate), and water. The hot microemulsion is dispersed in cold water (2-3°C) under stirring. Typical volume ratios of the hot microemulsion to the cold water are in the range of 1:25 to 1:50. The dilution process is critically determined by the composition of the microemulsion. According to the literature, the droplet structure is already contained in the microemulsion, and therefore, no energy is required to achieve submicron particle sizes.30,31 The temperature gradient and the pH-value determine the product quality in addition to the composition of the microemulsion. High temperature gradients facilitate rapid lipid crystallization and prevent aggregation.32'33 Due to the dilution step, lipid contents which are achievable are considerably lower, compared with the HPH based formulations. Another disadvantage includes the use of organic solvents. Recent work describes a similar approach to produce SLN. A hot liquid crystalline phase (instead of a microemulsion) is diluted in cold water to yield a solid lipid nanodispersion.34 This approach avoids the use of high pressure homogenization and organic solvents, and therefore represents an interesting opportunity. Solid Lipid Nanoparticles as Drug Carriers 193 2.6. Further processing 2.6A. Sterilization Sterility is required for parenteral formulations. Dry or wet heat, filtration, y-irradiation, chemical sterilization and aseptic production are general, opportunities to achieve sterility. The sterilization should not change the properties of the sample with respect to physical and chemical stability and the drug release kinetics. Sterilization by heat is a reliable procedure which is most commonly used. It was also applied for Liposomes.35,36 Steam sterilization will cause the formation of an oil in water emulsion, due to the melting of the lipid particles. The formation of SLN requires recrystallization of the lipids. Concerns are related to temperature induced changes of the physical and chemical stability. The correct choice of the emulsifier is of significant importance for the physical stability of the sample at high temperatures. Increased temperatures will affect the mobility and the hydrophilicity of all emulsifiers, but to a different extent. Schwarz found that Lecithin is preferable to Poloxamer for steam sterilization, as only a minor increase in the particle size and the number of microparticles was observed after steam sterilization.37'38 An increase in particle size for Poloxamer 188 stabilized Compritol-SLN was observed after steam sterilization. It was found that a decrease of the sterilization temperature from 121°C to 110°C can reduce sterilization induced particle aggregation to a large extent. This destabilization can be attributed to the decreased steric destabilization of the Poloxamer. It is well known for PEG-based emulsifiers that increased temperatures lead to dehydration of the ethylenoxide chains, pointing to a decrease of the thickness of the protecting layer. It has been demonstrated by 1H-NMR spectroscopy on Poloxamer stabilized lipid nanoparticles, that even a moderate temperature increase from RT to 37°C decreases the mobility of the ethylenoxide chains on the particle surface.39 Results of Freitas et dl. indicate that the lowering of the lipid content (to 2%), and the surface modification of the glass vials and nitrogen purging might prevent the particle growth to a large extent and avoid gelation.40 Further studies of Cavalli et al.4* and Heiati42 demonstrate the possibility of steam sterilization of drug loaded SLN. Filtration sterilization of dispersed systems requires very high pressure and is not applicable to particles >0, 2 /nm. As most SLN particles are close to this size, filtration is of no practical use, due to the clocking of the filters. Few studies investigated the possibility of y-sterilization. It must be kept in mind that free radicals are formed during y-sterilization in all samples, due to the high energy of the yrays. These radicals may recombine with no modification of the sample or undergo secondary reactions which might lead to chemical modifications of the sample. The degree of sample degradation depends on the general chemical reactivity and the molecular mobility and the presence of oxygen. It is therefore not surprising 194 Mader that chemical changes of the lipid bilayer components of liposomes were observed after y-irradiation.43 Schwarz investigated the impact of different sterilization techniques [steam sterilization at 121°C (15min) and 110°C (15min); y-sterilization] on SLN characteristics.37'38 In comparison to lecithin stabilized systems, Poloxamer stabilized SLN were less stable than steam sterilization. However, this difference was not detected for y-sterilized samples. Compared with steam sterilization at 121 °C, the increase in particle size after y-irradiation was lower, but comparable to that at 110°C. Unfortunately, most investigators did not search for steam sterilization or irradiation induced chemical degradation. It should be kept in mind that degradation does not always cause increased particle sizes. In contrast, the formation of species like lysophosphatides or free fatty acids could even preserve small particle sizes, but might cause toxicological problems. Further studies with more focus on chemical degradation products are clearly necessary to permit valid statements of the possibilities of SLN sterilization. 2.6.2. Drying by lyophilization, nitrogen purging and spray drying SLN are thermodynamic unstable systems, and therefore, particle growth has to be minimized. Furthermore, SLN ingredients and incorporated drugs are often unstable, hydrolyzing or oxidizing. The transformation of the aqueous SLN-suspension in a dry, redispersible powder is therefore often a necessary step to ensure storage stability of the samples. Lyophilization is widely used and is a promising way to increase chemical and physical SLN stability over extended periods of time. Lyophilization also offers principle possibilities for SLN incorporation into pellets, tablets or capsules. Two additional transformations are necessary which might be the source of additional stability problems. The first transformation, from aqueous dispersion to powder, involves the freezing of the sample and the evaporation of water under vacuum. Freezing of the sample might cause stability problems due to the freezing out effect which results in the changes of the osmolarity and the pH. The second transformation, resolubilization, involves situations at least in its initial stages which favor particle aggregation (i.e. low water and high particle content, high osmotic pressure). The protective effect of the surfactant can be compromised by lyophilization.44 It has been found that the lipid content of the SLN dispersion should not exceed 5%, so as to prevent an increase in the particle size. Direct contact of lipid particles are decreased in diluted samples. Furthermore, diluted SLN dispersions will also have higher sublimation velocities and a higher specific surface area.45 The addition of cryoprotectors (e.g. Sorbitol, Mannose, Trehalose, Glucose, and Solid Lipid Nanoparticles as Drug Carriers 195 Polyvinylpyrrolidon) will be necessary to decrease SLN aggregation and to obtain a better redispersion of the dry product. Schwarz et al. investigated the lyophilization of SLN in detail.46 Best results were obtained with the cryoprotectors, Glucose, Mannose, Maltose and Trehalose, in the concentration range between 10% and 15%. The observations come into line with the results of the studies on liposome lyophilization, which indicated that Trehalose was the most sufficient substance to prevent liposome fusion and the leakage of the incorporated drug.47 Encouraging results obtained with unloaded SLN cannot predict the quality of drug loaded lyophilizates. Even low concentrations of 1% Tetracain or Etomidat caused a significant increase in particle size, excluding an intravenous administration.46 Westesen investigated the lyophilization of tripalmitate-SLN using glucose, sucrose, maltose and trehalose as cryoprotective agents.48 Handshaking of redispersed samples was an insufficient method, but bath sonification produced better results. Average particle sizes of all lyophilized samples with cryoprotective agents were 1.5 to 2.4 times higher than the original dispersions. One year storage caused increased particle sizes of 4 to 6.5 times compared with the original dispersion. In contrast to the lyophilizates, the aqueous dispersions of tyloxapol/phospholid stabilized tripalmitate SLN exhibited remarkable storage stability. The instability of the SLN lyophilizates can be explained by the sintering of the particles. TEM pictures of tripalmitate SLN show an anisometrical, platelet-like shape of the particles. Lyophilization changes the properties of the surfactant layer due to the removal of water, and increases the particle concentration which favors particle aggregation. Increased particle sizes after lyophilization (2.1 to 4.9 times) were also reported by Cavalli.41 Heiati compared the influence of four cryoprotectors (i.e. trehalose, glucose, lactose and mannitol) on the particle size of azidothymidine palmitate loaded SLN lyophilizates.42 In agreement to other reports, Trehalose was found to be the most effective cryoprotectant. The freezing procedure will affect the crystal structure and the properties of the lyophilizate. Literature data suggest that the freezing process needs to be optimized to a particular sample size. Schwarz recommended rapid freezing in liquid nitrogen.46 In contrast, other researchers observed the best results after a slow freezing process.49 Again, best results were obtained with samples of low lipid content and with the cryoprotector trehalose. Slow freezing in a deep freeze (—70°C) was superior to rapid cooling in liquid nitrogen. Furthermore, introduction of an additional thermal treatment of the frozen SLN dispersion (2 hr at —22°C; followed by 2 hr temperature decrease to — 40°C) was found to improve the quality of the lyophilizate. Lately, lyophilization has been used to stabilize retinoic acid loaded SLN.50 An interesting alternative to lyophilization has been recently suggested by Gasco's group. Drying with a nitrogen stream at low temperatures of 3 to 10°C has been found to be superior.51 Compared with lyophilization, the advantages of 196 Mader this process are the avoidance of freezing and the energy efficiency resulting from the higher vapor pressure of water. Spray drying has been scarcely for SLN drying, although it is cheaper compared with lyophilization. Freitas obtained a redispersable powder with this method, which meets the general requirements of i.v.-injections, with regard to the particle size and the selection of the ingredients.52 Spray drying might potentially cause particle aggregation due to high temperatures, shear forces and partial melting of the particles. Freitas recommends the use of lipids with high melting points >70°C to avoid sticking and aggregation problems. Furthermore, the addition of carbohydrates and low lipid contents favor the preservation of the colloidal particle size in spray drying. 3. SLN Structure and Characterization The characterization of SLN is a necessity and a great challenge. Lipid characterization itself is not trivial as the statement by Laggner shows53: "Lipids and fats, as soft condensed material in general, are very complex systems, which not only in their static structures but also with respect to their kinetics of supramolecular formation, Hysteresis phenomena or supercooling can gravely complicate the task of defining the underlying structures and boundaries in a phase diagram". This is especially true for lipids in the colloidal size range. Therefore, possible artifacts caused by sample preparation (removal of emulsifier from particle surface by dilution, induction of crystallization processes, changes of lipid modifications) should be kept in mind. For example, the contact of the SLN dispersion with new surfaces (e.g. a syringe needle) might induce lipid crystallization or modification, and sometimes result in the spontaneous transformation of the low viscous SLN-dispersion into a viscous gel. The most important parameters of SLN include particle size and shape, the kind of lipid modification and the degree of crystallization, and the surface charge. Photon correlation spectroscopy (PCS) and Laser Diffraction (LD) are the most powerful techniques for routine measurements of particle size. It should be kept in mind that both methods are not "measuring" particle sizes. Rather, they detect light scattering effects which are used to calculate particle sizes. For example, uncertainties may result from nonspherical particle shapes. Platelet structures commonly occur during lipid crystallization54 and are very often described in the SLN literature.55-59 The influence of the particle shape on the measured size is discussed by Sjostrom.55 Further difficulties arise both in PCS and LD measurements for samples which contain several populations of different size. Therefore, additional techniques might be useful. For example, light microscopy is recommended although it is not sensitive to the nanometer size range. It gives a fast indication about the Solid Lipid Nanoparticles as Drug Carriers 197 presence and the character of microparticles. Electron Microscopy provides, in contrast to PCS and LD, direct information on the particle shape.57'58 Atomic force microscopy (AFM) has attracted increasing attention. A cautionary note applies to the use of AFM in the field of nanoparticles, as an immobilization of the SLN by solvent removal is required to assess their shape by the AFM tip. This procedure is likely to cause substantial changes of the molecular structure of the particles. Zur Miihlen demonstrated the ability of AFM to image the morphological structure of SLN.60 The sizes of the visualized particles are of the same magnitude, compared with the results of PCS measurements. The AFM investigations revealed the disklike structure of the particles. Dingier investigated cetylpalmitate SLN (stabilized by polyglycerol methylglucose distearate, Tego Care 450) by electron microscopy and AFM and found an almost spherical form of the particles.61 The usefulness of cross flow Field-Flow-Fractionation (FFF) for the characterization of colloidal lipid nanodispersions has been recently demonstrated.58 Lipid nanodispersions with constant lipid content, but different ratios of liquid and solid lipids did show similar particle sizes in dynamic light scattering. However, retention times in FFF were remarkably dissimilar due to the different particle shapes (i.e. spheres vs. platelets). Anisotropic particles such as platelets will be constrained by the cross flow much more heavily compared with the spheres of similar size. The very high anisometry of the SLN particles has been confirmed by electron microscopy, where very thin particles of 15 nm thickness and the length of several hundred nanometers became visible. The measurement of the zeta potential allows predictions about the storage stability of colloidal dispersions.62 In general, particle aggregation is less likely to occur for charged particles (i.e. high zeta potential) due to electric repulsion. However, this rule cannot strictly apply to systems which contain steric stabilizers, because the adsorption of steric stabilizer will decrease the zeta potential due to the shift in the shear plane of the particle. Particle size analysis is just one aspect of SLN quality. The same attention has to be paid on the characterization of lipid crystallinity and modification, because these parameters are strongly correlated with drug incorporation and release rates. Thermodynamic stability and lipid packing density increase, and drug incorporation rates decrease in the following order: supercooled melt < a-modification < B'-modification < 6-modification In general, it has been found that melting and crystallization processes of nanoscaled material can differ considerable from that of the bulk material.63 The thermodynamic properties of material having small nanometer dimensions can be considerably different, compared with the material in bulk form (e.g. the reduction 198 Mader of melting point). This occurs because of the tremendous influence of the surface energy. This statement is also valid for SLN, where lipid crystallization and modification changes might be highly retarded,64 due to the small size of the particles and the presence of emulsifiers. Moreover, crystallization might not occur at all and has been shown that samples which were previously described as SLN (solid lipid particles) were in fact supercooled melts (liquid lipid droplets).65 The impact of the emulsifier on SLN lipid crystallization has been shown by Bunjes.66 The same group demonstrated also a size dependent melting of SLN.67 Differential Scanning Calorimetry (DSC) and X-ray scattering are most commonly applied to asses the status of the lipid. DSC uses the fact that different lipid modifications possess different melting points and melting enthalpies. By means of X-ray scattering, it is possible to assess the length of the long and short spacings of the lipid lattice. It is highly recommended to measure the SLN dispersion themselves, because solvent removal will lead to modification changes. Sensitivity problems and long measurement times of convential X-ray sources might be overcome by synchrotron irradiation.64 In addition, this method permits to conduct time resolved experiments and allows the detection of intermediate states of colloidal systems which will be non detectable by convential X-ray methods.53 Recent work shows that SLN might form superstructures by parallel alignment of SLN platelets. These reversible particle self-assemblies were observed by Illing et al. in tripalmitin dispersions when the lipid concentration exceeds 40mg/g. Higher lipid concentrations did enhance particle self-assembly. The tendency to form self-assemblies has been found to depend on the particle shape, the lipid and the surfactant concentration.68 Infrared and Raman Spectroscopy are useful tools to investigate structural properties of lipids and they might give complentary information to X-ray and DSC.54 Raman measurements on SLN show that the arrangement of lipid chains of SLN dispersions changes with storage.69 Rheometry might be particularly useful for the characterization of the viscoelastic properties of SLN dispersions. The rheological properties are important with respect to the dermatological use of SLN, but they also provide useful information about the structural features of SLN dispersions and their storage dependency. Studies of Lippacher show that the SLN dispersion posses higher elastic properties than emulsions of comparable lipid content.70-72 Furthermore, a sharp increase of the elastic module is observed at a certain lipid content. This point indicates the transformation from a low viscous lipid dispersion to an elastic system, with a continuous network of lipid nanocrystals. Illing and Unruh did compare the rheological properties of trimyristic, tripalmitic and tristearic SLN suspensions. The results indicate that the viscosity of triglyceride suspensions increases with the lipid chain length and an increased anisotropy of the particles.73 Souto et al. used Solid Lipid Nanoparticles as Drug Carriers 199 rheology to study the influence of SLN addition on the rheological properties of hydrogels.74 The co-existence of additional colloidal structures (micelles, liposomes, mixed micelles, nanodispersed liquid crystalline phases, supercooled melts, drugnanoparticles) has to be taken into account for all SLN dispersions. Unfortunately, many investigators neglect this aspect, although the total amount of surface active compounds is often comparable to the total amount of the lipid. The characterization and quantification are serious challenges due to the similarities in size. In addition, the sample preparation will modify the equilibrium of the complex colloidal system. Dilution of the original SLN dispersion with water might cause the removal of surfactant molecules from the particle surface and induce further changes such as crystallization or the changes of the lipid modifications. It is therefore highly desirable to use methods which are sensitive to the simultaneous detection of different colloidal species, which do not require preparatory steps such as Raman, NMR and ESR spectroscopy. NMR active nuclei of interest are 1H, 13C, 19F and 35P. Due to the different chemical shifts, it is possible to attribute the NMR signals to particular molecules or their segments. For example, lipid methyl protons give signals at 0.9 ppm, while protons of the polyethylenglycole chains give signals at 3.7 ppm. Simple ^-spectroscopy permits an easy and rapid detection of supercooled melts, due to the low linewidths of the lipid protons69,75-77. This method is based on the different proton relaxation times in the liquid and semisolid/solid state. Protons in the liquid state give sharp signals with high signal amplitudes, while semisolid/solid protons give very broad or invisible NMR signals under these circumstances. NMR has been used to characterize calixarene SLN78 and hybrid lipid particles (NLC), which are composed of liquid and solid lipids.59 Protons from solid lipids are not detected by standard NMR, but they can be visualized by solid state NMR. A drawback of solid state NMR is the rapid spinning of the sample that might cause artifacts. A recent paper describes the use of this method to monitor the distribution of Q10 in lipid matrices.79 Unfortunately, the authors did use "drying of the sample to constant weight" as a preparatory step, which will cause significant changes of the sample characteristics. ESR requires the addition of paramagnetic spin probes to investigate SLN dispersions. A large variety of spin probes is commercially available. The corresponding ESR spectra give information about the microviscosity and micropolarity. ESR permits the direct, repeatable and non-invasive characterization of the distribution of the spin probe between the aqueous and the lipid phase.80 Experimental results demonstrate that storage induced crystallization of SLN leads to an expulsion of the probe out of the lipid into the aqueous phase.81 Furthermore, using an ascorbic acid reduction assay, it is possible to monitor the time scale of the exchange between 200 Mader the aqueous and the lipid phase.59 The transfer rates of molecules between SLN and liposomes or cells have been determined by ESR.82 4. The "Frozen Emulsion Model" and Alternative SLN Models Lipid nanoemulsions are composed of a liquid oily core and a surfactant layer (lecithin). They are widely used for the parenteral delivery of poorly soluble drugs.83-85 The original idea of SLN was to achieve a controlled release of incorporated drugs by increasing the viscosity of the lipid matrix. Therefore it is not surprising that in original model, SLN is being described as "frozen emulsions" (see Fig. 1, left and middle).8687 However, lipids are known to crystallize very frequently in anisotropic platelet shapes54 and anisotropic. Sjostrom et al. described in 1995 that the particle shape of Cholesterylacetate SLN did strongly depend on the emulsifier.55 Platelet shaped particles have been detected for lecithin stabilized particles, while PEG-20-sorbitanmonolaurate stabilized particles preserved their spherical shape. Anisotropic particles have been found in numerous other SLN dispersions.56-59 Based on the experimental results, a platelet shaped SLN model can be proposed as an alternative (see Fig. 1, right). In the year 2000, Westesen questioned the frozen emulsion droplet model with the following statement88: "Careful physicochemical characterization has demonstrated that these lipid-based nanosuspensions (solid lipid nanoparticles) are not just emulsions with solidified droplets. During the development process of these systems, interesting phenomena have been observed, such as gel formation on solidification and upon storage, unexpected dynamics of polymorphic transitions, extensive annealing of nanocrystals over significant periods of time, stepwise melting of particle fractions in the Nanoemulsion SLN: "Frozen emulsion droplet" SLN: Platelet shaped particles o o — Core: liquid lipid (oil) S Core: solid lipid H Shell: stabilizer Fig. 1. General structure of a nanoemulsion (left), and proposed models for SLN: Frozen emulsion droplet model (middle) and platelet shaped SLN model (right). Solid Lipid Nanoparticles as Drug Carriers 201 lower-nanometer-size range, drug expulsion from the carrier particles on crystallization and upon storage, and extensive supercooling." Her comment highlights the complex behavior and changes of SLN dispersions. In addition, the presence of competing colloidal structures (e.g. micelles, liposomes, mixed micelles, nanodispersed liquid crystalline phases, supercooled melts and drug-nanoparticles) should be considered. Additional colloids might have an impact on very different aspects, including the correct measurement of particle size, drug incorporation and toxicity. A recent study shows that the cell toxicity of the SLN dispersion was reduced by dialysis due to the removal of water soluble components.89 5. Nanostructured Lipid Carriers (NLC) Nanostructured lipid carriers (NLC) have been recently proposed as a new SLN generation with improved characteristics.90 The general idea behind the system is to improve the poor drug loading capacity of SLN by "mixing solid lipids with spatially incompatible lipids leading to special structures of the lipid matrix",91 while still preserving controlled release features of the particles. Three different types of NLC have been proposed (NLC I: The imperfect structured type, NLC II: The structureless type and NLC III: The multiple type). Unfortunately, these structural proposals have not been supported by experimental data. They assume a spherical shape and they are not compatible with lipid platelet structures. For example, NLC III structures should contain small oily droplets in a solid lipid sphere (Fig. 2, left). Detailed analytical examination of NLC systems by Jores et al. demonstrate that "nanospoon" structures are formed, in which the liquid oil adheres on the solid surface of a lipid platelet (Fig. 2, right). Jores et al. did conclude that "Neither SLN nor NLC lipid nanoparticles showed any advantage with respect to incorporation rate or retarded accessibility to the drug, compared with conventional nanoemulsions. The experimental data concludes that NLCs are not spherical solid lipid particles with embedded liquid liquid lipid (oil) J A solid lipid • stabilizer Fig. 2. Proposed NLC III structure (modified after91) and experimental determined "nanospoon" structure described by Jores et al. (side view of particle).58'59 202 Mader droplets, but rather, they are solid platelets with oil present between the solid platelet and the surfactant layer". Very similar structures have been found on Q10 loaded SLN by Bunjes et til.92 6. Drug Localization and Release Proposed advantages of SLN, compared with nanoemulsions, include increased protection capacity against drug degradation and controlled release possibilities due to the solid lipid matrix. The general low capacity of crystalline structures to accommodate foreign molecules is a strong argument against the proposed rewards. It is therefore necessary to distinguish between drug association and drug incorporation. Drug association means that the drug is associated with the lipid, but it might be localized in the surfactant layer or between the solid lipid and the surfactant layer (similar to the oil in Fig. 2, right). Drug incorporation would mean the distribution of the drug within the lipid matrix. Another limiting aspect comes from the fact that the platelet structure of SLN, which is found in many systems, leads to a tremendous increase in surface area and the shortening of the diffusion lengths. Furthermore, additional colloid structures present in the sample are alternatives for drug localization the SLN for drug incorporation as it was pointed out by Westesen88: "The estimation of drug distribution is difficult for dispersions consisting of more than one type of colloidal particle. Depending on the type of stabilizer and on the concentration ratio of stabilizer to matrix material significant numbers of particles such as liposomes and/or (mixed) micelles may coexist with the expected type of particles". The detailed investigation of drug localization is very difficult and only a few studies exist. Parelectric spectroscopy has been used to investigate the localization of glucocorticoids. The results indicate that the drug molecules are attached to the particle surface, but not incorporated into the lipid matrix. With Betamethasonvalerate, the loading capacity of the particle surface was clearly below the usual concentration of 0.1%.93 Lukowski used Energy Dispersive X-ray Analysis and found that the drug Triamcinolone, Dexamethasone and Chloramphenicol are partially stored at the surface of the individual nanoparticles.94 The importance of the emulsifier is reflected in a study from Danish scientists.95 They produced gamma-cyhalothrin (GCH) loaded lipid micro- and nanoparticles. GCH had only limited solubility in the solid lipid and was expulsed during storage. The appearance of GCH crystals was strongly dependent from the solubility of the GCH in the emulsifier solutions. Emulsifier with high GCH solubility provoked rapid crystal growth. This observation is in accordance with a mechanism of crystal growth according to Ostwald ripening. Slovenian scientist found that ascorbylpalmitate was more resistant against oxidation in non-hydrogenated soybean Solid Lipid Nanoparticles as Drug Carriers 203 lecithin liposomes, compared with SLN.96 It shows that liposomes might have a higher protection capacity compared with SLN. Fluorescence and ESR studies have been used by Jores et al. to monitor the microenvironment and the mobility of model drugs. The results indicate that even highly lipophilic compounds are pushed into a polar environment during lipid crystallization. Therefore, the incorporation capacity of SLN is very poor for most molecules.69 A nitroxide reduction assay gave results in accordance with the results of the distribution. Compared with nanoemulsions, nitroxides were more accessible in SLN and NLC to ascorbic acid, localized in the aqueous environment. Therefore, nanoemulsions were more protective than SLN and NLC systems. Drug release from SLN and NLC could be either controlled by the diffusion of the drug or the erosion of the matrix. The original idea was to achieve a controlled release of SLN due to the slowing down of drug diffusion to the particle surface. This idea is, however, questionable due to drug expulsion during lipid crystallization. In addition, very short diffusion lengths in nanoscaled delivery systems lead to short diffusion times, even in highly viscous or solid matrices. In most cases, the delivery of the drug will be controlled by the slow dissolution rate in the aqueous environment. Drug release rate will be highly dependent on the presence of further solubilizing colloids (e.g. micelles), which are able to work as a shuttle for the drug and the presence or absence of a suitable acceptor compartment. Many investigators studied only the release in buffer media. A controlled release pattern under such conditions is not surprising, as it is caused by low solubilization kinetics due to the poor solubility of the drug. In vivo, acceptor compartments will be present (e.g. lipoproteins, membranes) and will speed up release processes significantly. Whenever possible, drug loaded SLN should be compared with nanosuspensions to separate the general features of the drug and the influence of the lipid matrix. Results by Kristl et al. indicate that lipophilic nitroxides diffuse between SLN and liposomes. The diffusion kinetics was strongly dependent on the nitroxide structure. In contrast, uptake of nitroxides in cells was similar between lipophilic nitroxides, suggesting endocytosis as the main mechanism.82 The detailed mechanisms of drug release in vivo are poorly understood. In vitro data by Olbrich demonstrate that SLN are degraded by lipases.97,98 Degradation by lipase depends on the lipid and strongly on the surfactant. Steric stabilization (e.g. by poloxamer) of SLN and NLC are less accessible because lipase needs an interface for activation. It is also known that highly crystalline lipids are poorly degraded by lipase. 7. Administration Routes and In Vivo Data SLN and NLC can be administrated at different routes, including peroral, dermal, intravenously and pulmonal. Peroral administration of SLN could enhance the drug 204 Mader absorption and modify the absorption kinetics. Despite the fact that in most of the SLN, the drug will be associated but not incorporated in the lipid, SLN might have advantages due to enhanced lymphatic uptake, enhanced bioadhesion or increased drug solubilization by SLN lipolysis products such as fatty acids and monoglycerides. A serious challenge represents the preservation of the colloidal particle in the stomach, where low pH values and high ionic strengths favor agglomeration and particle growth. Zimmermann and Muller studied the stability of different SLN formulations in artificial gastric juice." The main findings of this study are that (i) some SLN dispersions preserve their particle size under acidic conditions, and (ii) there is no general lipid and surfactant which are superior to others. The particular interactions between lipid and stabilizer are determining the robustness of the formulation. Therefore, the suitable combination of ingredients has to be determined on a case by case basis. Several animal studies show increased absorption of poorly soluble drugs. The efficacy of orally administrated Triptolide free drug and Triptolide loaded SLN have compared in the carrageenan-induced rat paw edema by Mei et al.wo Their results suggest that SLN can enhance the anti inflammatory activity of triptolide and decrease triptolide-induced hepatotoxicity. The usefulness of SLN to increase the absorption of the poorly soluble drug all-trans retinoic acid has been shown by Hu et al. on rats.101 Gascos group investigated the uptake and distribution of Tobramycin loaded SLN in rats.102'103 They observed an increased uptake into the lymph, which causes prolonged drug residence times in the body of the animals. Furthermore, AUC and clearance rates did depend on the drug load. The same group described also enhanced absorption of Idarubicin-loaded solid lipid nanoparticles (IDA-SLN), in comparison to the drug solution. Furthermore, the authors described that SLN were able to pass the blood-brain barrier and concluded that duodenal administration of IDA-SLN modifies the pharmacokinetics and tissue distribution of idarubicin.104 Parenteral administration of SLN is of great interest too. To avoid the rapid uptake of the SLN by the RES system, stealth SLN particles have been developed by the adoption of the stealth concept from liposomes and polymer nanoparticles. Reports indicate that Doxorubicin loaded stealth SLN circulate for long period of time in the blood and change the tissue distribution.105 Therefore, SLN could be alternatives to marketed stealth-liposomes, which can decrease the heart toxicity of this drug due to changed biodistribution. Long circulation times have also been observed for Poloxamer stabilized SLN with Paclitaxel.106 The dermal application is of particular interest and it might become the main application of SLN.107 SLN pose occlusive properties which are related to the solid structure of the lipid.108 Human in vivo results of the group of Muller demonstrate that SLN can improve skin hydration and viscoelasticity.109 SLN have also Solid Lipid Nanoparticles as Drug Carriers 205 UV protection capacity due to their reflection of UV light.110 Furthermore, data by Schafer-Korting suggest SLN can be used to decrease drug side effects due to SLN mediated drug targeting to particular skin layers.111 Further reports describe additional applications of SLN as well as gene delivery,112 delivery to the eye,113 pulmonary delivery,114 and drug targeting of anticancer drugs.115 Studies of the different groups also propose the use of SLN for brain targeting to deliver MRI contrast agents116 or antitumour drugs.117'118 8. Summary and Outlook SLN and NLC are now investigated by many scientists worldwide. In contradiction to early proposals, they certainly do not combine all the advantages of the other colloidal drug carriers and avoid the disadvantages of them. SLN are complex colloidal dispersions, not just "frozen emulsions". SLN dispersions are very susceptible to the sample history and storage conditions. Disadvantages of SLN include gel formation on solidification and upon storage, unexpected dynamics of polymorphic transitions, extensive annealing of nanocrystals over significant periods of time, stepwise melting of particle fractions in the lower-nanometer-size range, drug expulsion from the carrier particles on crystallization and upon storage, and extensive supercooling. The anisotropic shape of many SLN dispersions increases the surface area significantly, decreases the diffusion lengths to the surface and changes the rheological behavior dramatically (e.g. gel formation). Furthermore, the presence of alternative colloidal structures (micelles, liposomes) has to be considered to contribute to drug localization. In most cases, the drug will be associated with the lipid and not incorporated. Studies demonstrate that SLN and NLC might have no advantages compared with submicron emulsions, in regard to protection from the aqueous environment. On the other side, animal data suggest that SLN can change the pharmacokinetics and the toxicity of drugs. In many cases, drug incorporation might not be required and drug association with the lipid can be sufficient for lymphatic uptake. Clearly, more detailed studies are necessary to get a deeper understanding of the in vivo fate of these carriers. Whenever possible, SLN and NLC systems should be compared directly with alternative colloidal carriers (e.g. liposomes, nanoemulsions, nanosuspensions) to evaluate their true potential. References 1. Eldem T, Speiser P and Hincal A (1991) Optimization of spray-dried and congealed lipid micropellets and characterization of their surface morphology by scanning electron microscopy. Pharm Res 8:47-54. 206 Mader 2. Speiser P (1990) Lipidnanopellets als Tragersystem fur Arzneimittel zur peroralen Anwendung, European Patent EP 0167825. 3. Domb AJ (1993) Lipospheres for controlled delivery of substances. United States Patent No. 5188837. 4. Domb AJ (1995) Long acting injectable oxytetracycline-liposphere formulation. Int } Pharm 124:271-278. 5. Domb AJ (1993) Liposphere parenteral delivery system. Proc Intl Symp Control Rel Bioact Mater 20:346-347. 6. Siekmann B and Westesen K (1992) Submicron-sized parenetral carrier systems based on solid lipids, Pharm. Pharmacol Lett 1:123-126. 7. Miiller RH, Lucks JS (1996) Arzneistofftrager aus festen Lipidteilchen, Feste Lipidnanospharen (SLN). European Patent No. 0605497. 8. Miiller RH, Mehnert W, Lucks JS, Schwarz C, zur Miihlen A, Weyhers H, Freitas C and Riihl D (1995) Solid Lipid Nanoparticles (SLN) —An Alternative Colloidal Carrier System for Controolled Drug Delivery. Eur J Pharm Biopharm 41:62-69. 9. Sjostrom B and Bergenstahl B (1992) Preparation of submicron drug particles in lecithinstabilized o /w emulsions. I. Model studies of the precipitation of cholesteryl acetate. Int} Pharm 88:53-62. 10. Cavalli R, Caputo O and Gasco MR (1993) Solid lipospheres of doxorubicin and idarubicin. Int]Pharm 89:R9-R12. 11. Gasco MR (1993) Method for producing solid lipid microspheres having a narrow size distribution. United States Patent No. 5250236. 12. Miiller RH (1997) Pharmazeutische Technologie, Moderne Arzneiformen, Wiss. Verlagsges. Stuttgart. 13. Miiller RH and Runge SA, Solid Lipid Nanoparticles (SLN) for controlled drug delivery, in Submicron Emulsions in Drug Targeting and Delivery, Benita S (ed.), Harwood Academic Publishers. 14. Small D (1986) Handbook of lipids. The physical chemistry of lipids: From alkanes to phospholipids, Plenum Press, New York. 15. Ahlin P, Kristl J and Smid-Kobar J (1998) Optimization of procedure parameters and physical stability of solid lipid nanoparticles in dispersions. Acta Pharm 48:257-267. 16. Lippacher A, Miiller RH and Mader K (2000) Investigation on the viscoelastic properties of lipid based colloidal drug carriers. Int ] Pharm 196:227-230. 17. zur Miihlen A and Mehnert W (1998) Drug release and release mechanism of prednisolone loaded solid lipid nanoparticles. Pharmazie 53:552-555. 18. zur Miihlen A, Schwarz C and Mehnert W (1998) Solid lipid nanoparticles (SLN) for controlled drug delivery — Drug release and release mechanism. 19. Lander R, Manger W, Scouloudis M, Ku A, Davis C and Lee A (2000) Gaulin homogenization: A mechanistic study. Biotechnol Prog 16:80-85. 20. Jahnke S (1998) The theory of high pressure homogenization, in Emulsions and Nanosuspensions for the Formulation of Poorly Soluble Drugs, Miiller RH, Benita S and Bohm B (eds.), Medpharm Scientific Publishers: Stuttgart, pp. 177-200. Solid Lipid Nanoparticles as Drug Carriers 207 21. Siekmann B and Westesen K (1994) Melt-homogenized solid lipid nanoparticles stabilized by the nonionic surfactant tyloxapol, I. Preparation and particle size determination. Pharm Pharmacol Lett 3:194-197. 22. Bunjes H, Siekmann B and Westesen K (1998), Emulsions of supercooled melts — a novel drug delivery system, in Submicron Emulsions in Drug Targeting and Delivery, Benita S (ed.), Harwood Academic Publishers. 23. zur Miihlen A (1996) Feste Lipid-Nanopartikel mit prolongierter Wirkstoffliberation: Herstellung, Langzeitstabilitat, Charakterisierung, Freisetzungsverhalten und - machanismen, PhD thesis, Free University of Berlin. 24. Friedrich I and Miiller-Goymann CC (2003) Characterization of solidified reverse micellar solutions (SRMS) and production development of SRMS-based nanosuspensions. Eur J Pharm Biopharm 56:111-119. 25. Siekmann B and Westesen K (1996) Investigations on solid lipid nanoparticles prepared by precipitation in o /w emulsions. Eur } Pharm Biopharm 43:104-109. 26. Fessi H, Puisieux F, Ammoury N and Benita S (1989) Nanocapsule formation by interfacial polymer deposition following solvent displacement. Int J Pharm 55:R1-R4. 27. Hu FQ, Yuan H, Zhang HH and Fang M (2002) Preparation of solid lipid nanoparticles with clobetasol propionate by a novel solvent diffusion method in aqueous system and physicochemical characterization. Int J Pharm 239:121-128. 28. Schubert MA and Miiller-Goymann CC (2003) Solvent injection as a new approach for manufacturing lipid nanoparticles—evaluation of the method and process parameters. Eur } Pharm Biopharm 55:125-131. 29. Danielsson I and Lindman B (1981) The definition of microemulsion. Coll Surf B 3: 391-392. 30. Gasco MR (1997) Solid lipid nanospheres from warm micro-emulsions. Pharma Technol Eur 52-58. 31. Boltri L, Canal T, Esposito PA and Carli F (1993) Lipid nanoparticles: Evaluation of some critical formulation parameters. Proc Intl Symp Control Rel Bioact Mater 20:346-347. 32. Cavalli R, Marengo E, Rodriguez L and Gasco MR (1996) Effect of some experimental factors on the production process of solid lipid nanoparticles. Eur J Pharm Biopharm 43:110-115. 33. Gasco MR, Morel S and Carpigno R (1992) Optimization of the incorporation of desoxycortisone acetate in lipospheres. Eur } Pharm Biopharm 38:7-10. 34. Dahms G and Seidel H (2004) Method for the preparation of solid-lipid nanoparticles (SLNs) without high pressure homogenizer for pharmaceutical, cosmetic and food applications. German Patant application DE 2003-10312763 20030321. 35. Zuidam NJ, Lee SS, L and Crommelin DJA (1992) Sterilization of liposomes by heat treatment. Pharm Res 10:1591-1596. 36. Lukyanov AN and Torchilin VP (1994) Autoclaving of liposomes. / Microencap 11: 669-672. 37. Schwarz C and Mehnert W (1995) Sterilization of drug-free and tetracaine-loaded solid lipid nanoparticles (SLN). Proc 1st World Meeting APGI/APV, Budapest, 485^86. 208 Mader 38. Schwarz C, Freitas C, Mehnert W and Muller RH (1995) Sterilization and physical stability of drug-free and etomidate-loaded solid lipid nanoparticles. Proc Intl Symp Control Rel Bioct Mater 22:766-767. 39. Liedtke S, Jores K, Mehnert W and Mader K (2000) Possibilities of non-invasive physicochemical characterisation of colloidal drug carriers, 27th Intl Symp Control Rel Bioact Mater Vol. 27, Controlled Release Society, Paris, 1088-1089. 40. Freitas C (1998) Feste Lipid-Nanopartikel (SLN): Mechanismen der physikalischen Destabilisierung und Stabilisierung. PhD thesis, Free University of Berlin. 41. Cavalli R, Caputo O, Carlotti ME, Trotta M, Scarnecchia and Gasco MR (1997) Sterilization and freeze-drying of drug-free and drug-loaded solid lipid nanoparticles. Int} Pharm 148:47-54. 42. Heiati H, Tawashi R and Phillips NC (1998) Drug retention and stability of solid lipid nanoparticles containuing azidothymidine palmitate after autoclaving, storage and lyophilisation.} Microencap 15:173-184. 43. Sculier JP, Coune A, Brassine C, Laduron C, Atassi G, Ruysschert GM and Fruhling J (1986) Intravenous infusion of high doses of liposomes containing NSC 251635, a water insoluble cytostatic agent. A pilot sudy with pharmacokinetic data. / Clin Oncol 4:789-797. 44. Rupprecht H (1993) Physikalisch-chemische Grundlagen der Gefriertrocknung, in Essig D and Oschmann R (eds.), Lyophilization. Paperback APV, Band 35, Wissenschaftliche Verlagsgesellschaft mbH, Stuttgart, pp. 13-38. 45. Pikal MJ, Shah S, Roy ML and Putman R (1990) The secondary drying stage of freeze drying: drying kinetics as a function of temperature and chamber pressure. Int} Pharm 60:203-217. 46. Schwarz C and Mehnert W (1997) Freeze-drying of drug-free and drug-loaded solid lipid nanoparicles. Int J Pharm 157:171-179. 47. Crowe LM, Crowe JH, Rudolph A, Womersley C and Appel L (1985) Preservation of Freeze-dried Liposomes by Trehalose. Arch Biochem Biophys 242:240-247. 48. Siekmann B and Westesen K (1994) Melt-homogenized solid lipid nanoparticles stabilized by the nonionic surfactant tyloxapol, II. Physicochemical characterization and lyophilisation. Pharm Pharmacol Lett 3:225-228. 49. Zimmermann E, Muller RH and Mader K (2000) Influence of different parameters on reconstitution of lyophilized SLN. Int J Pharm 196:211-213. 50. Lim SJ, Lee MK and Kim CK (2004) Altered chemical and biological activities of all-trans retinoic acid incorporated in solid lipid nanoparticle powders. / Control Rel 100:53-61. 51. Marengo E, Cavalli R, Rovero G and Gasco MR (2003) Scale-up and optimization of an evaporative drying process applied to aqueous dispersions of solid lipid nanoparticles. Pharm Dev Techn 8:299-309. 52. Freitas C and Muller RH (1998) Spray-drying of solid lipid nanoparticles (SLN™). Eur J Pharm Biopharm 46:145-151. 53. Laggner P (1999) X-ray diffraction of lipids, in Spectral Properties of Lipids, Hamilton RJ and Cast J (eds.), Sheffield Academic Press. Solid Lipid Nanoparticles as Drug Carriers 209 54. Garti N and Sato K (eds.), (1998) Crystallization and Polymorphism of Fats and Fatty Acids, Marcel Dekker; New York and Basel. 55. Sjostrom B, Kaplun A, Talmon Y and Cabane B (1995) Structures of nanoparticles prepared from oil in water emulsions. Pharm Res 12:39^8. 56. Siekmann B and Westesen K (1992) Sub-micron sized parenteral carrier systems based on solid lipid, Pharma Pharmacol Lett 1:123-126. 57. Illing A, Unruh T and Koch MHJ (2004) Investigation on particle self-assembly in solid lipid-based colloidal drug carrier systems. Pharm Res 21:592-597. 58. Jores K, Mehnert W, Drechsler M, Bunjes H, Johann C and Mader K (2004) Investigations on the structure of solid lipid nanoparticles (SLN) and oil-loaded solid lipid nanoparticles by photon correlation spectroscopy, field-flow fractionation and transmission electron microscopy. / Control Rel 95:217-227. 59. Jores K, Mehnert W and Mader K (2003) Physicochemical investigations on solid lipid nanoparticles (SLN) and on oil-loaded solid lipid nanoparticles: A NMR- and ESRstudy. Pharm Res 20:1274-1283. 60. zur Muhlen A, zur Miihlen E, Niehus H and Mehnert W (1996) Atomic force microscopy studies of solid lipid nanoparticles. Pharm Res 13:1411-1416. 61. Dingier A, Blum RP, Niehus H, Muller RH and Gohla S (1999) Solid lipid nanoparticles (SLN™/Lipopearls™) — a pharmaceutical and cosmetic carrier for the application of vitamin E in dermal products.} Microencap 16:751-767. 62. Muller RH (1996) Zetapotential und Partikelladung - Kurze Theorie, praktische Mefidurchfuhrung, Dateninterpretation, Wissenschaftliche Verlagsgesellschaft Stuttgart. 63. Lai SL, Guo JY, Petrova V, Ramanath G and Allen LH (1996) Size-Dependent Melting Properties of Small Tin Particles: Nanocalorimetric Measurements. Phys Rev Lett 77:99-103. 64. Westesen K, Siekmann B and Koch MHJ (1993) Investigations on the physical state of lipid nanoparticles by synchrotron radiation X-ray diffraction. Int} Pharm 93:189-199. 65. Westesen K and Bunjes H (1995) Do nanoparticles prepared from lipids solid at room temperature always possess a solid matrix? Int ] Pharm 115:129-131. 66. Bunjes H, Koch MHJ and Westesen K (2003) Influence of emulsifiers on the crystallization of solid lipid nanoparticles. / Pharm Sci 92:1509-1520. 67. Bunjes H, Koch MHJ and Westesen K (2000) Effect of particle size on colloidal solid triglycerides. Langmuir 16:5234-5241. 68. Illing A, Unruh T and Koch MHJ (2004) Investigation on Particle Self-Assembly in Solid Lipid-Based Colloidal Drug Carrier Systems. Pharm Res 21:592-597. 69. Jores K (2004) Lipid nanodispersions as drug carrier systems — a physicochemical characterization, Thesis, University of Halle (http://sundoc.bibliothek.uni-halle.de/dissonline / 04 / 04H310 / prom.pdf). 70. Lippacher A, Muller RH and Mader K (2004) Liquid and semisolid SLN dispersions for topical application: Rheological characterization. Eur J Pharm Biopharm 58:561-567. 71. Lippacher A, Muller RH and Mader K (2001) Preparation of semisolid drug carriers for topical application based on solid lipid nanoparticles. Int} Pharm 214:9-12. 210 Mader 72. Lippacher A, Miiller RH and Mader K (2002) Semisolid SLN dispersions for topical application: Influence of formulation and production parameters on viscoelastic properties. Eur J Pharm Biopharm 53:155-160. 73. Illing A and Unruh T (2004) Investigation on the flow behavior of dispersions of solid triglyceride nanoparticles. Int} Pharm 284:123-131. 74. Souto EB, Wissing SA, Barbosa CM and Miiller RH (2004) Evaluation of the physical stability of SLN and NLC before and after incorporation into hydrogel formulations. Eur J Pharm Biopharm 58:83-90. 75. Westesen K and Siekmann B (1997) Investigation of the gel formation of phospholipids stabilized solid lipid nanoparticles. Int J Pharm 151:35^5. 76. Bunjes H, Westesen K and Koch MHJ (1996) Crystallization tendency and polymorphic transitions in triglyceride nanoparticles. Int J Pharm 129:159-173. 77. Zimmermann E, Liedtke S, Miiller RH and Mader K (1999) H-NMR as a method to characterize colloidal carrier systems, Proc Inc Symp Control Rel Bioact Mater 26: 591-592. 78. Shahgaldian P, Da Silva E, Coleman AW, Rather Beth and Zaworotko MJ (2003) Paraacyl- calix-arene based solid lipid nanoparticles (SLNs): A detailed study of preparation and stability parameters. Int J Pharm 253:23-38. 79. Wissing, S A, Miiller RH, Manthei L and Mayer C (2004) Structural Characterization of QlO-Loaded Solid Lipid Nanoparticles by NMR Spectroscopy. Pharm Res 21:400^05. 80. Ahlin P, Kristl J, Pecar S, Strancar J and Sentjurc M (2003) The effect of lipophihcity of spin-labeled compounds on their distribution in solid lipid nanoparticle dispersions studied by electron paramagnetic resonance. / Pharm Sci 92:58-66. 81. S Liedtke, E Zimmermann, RH Miiller and K Mader (1999) Physical characterisation of solid lipid nanoparticles (SLN™), Proc Intl Symp Control Rel Bioact Mater 26:595-596. 82. Kristl J, Volk B, Ahlin P, Gombac K and Sentjurc M (2003) Interactions of solid lipid nanoparticles with model membranes and leukocytes studied by EPR. Int } Pharm 256:133-140. 83. Miiller RH and Heinemann S (1994) Fat emulsions for parenteral nutrition IV: Lipofundin MCT/LCT regimens for total parenteral nutrition (TPN) with high electrolyte load. Int J Pharm 107:121-132. 84. Klang SH, Parnas M and Benita S (1998) Emulsions as drug carriers — possibilities, limitations and future perspectives, in Emulsions and Nanosuspensions for the Formulation of Poorly Soluble Drugs, RH Miiller, S Benita and BBohm (eds.), Medpharm Scientific Publishers, Stuttgart, pp. 31-65. 85. Davis SS, Washington C, West P and Ilium L (1987) Lipid emulsions as drug delivery systems. Ann N Y Acad Sci 507:75-88. 86. Miiller RH and Runge SA (1998) Solid lipid nanoparticles (SLN) for controlled drug delivery, in Submicron Emulsions in Drug Targeting and Delivery, Benita S (ed.), Harwood Academic Publishers; Amsterdam, pp. 219-234. 87. Mehnert W and Mader K (2001) Solid lipid nanoparticles: Production, characterization and applications. Adv Drug Del Rev 47:165-196. Solid Lipid Nanoparticles as Drug Carriers 211 88. Westesen K (2000) Novel lipid-based colloidal dispersions as potential drug administration systems. Expectations and reality. Coll Polym Sci 278:608-618. 89. Heydenreich AV, Westmeier R, Pedersen N, Poulsen HS and Kristensen HG (2003) Preparation and purification of cationic solid lipid nanospheres — Effects on particle size, physical stability, and cell toxicity. Int ] Pharm 254:83-87. 90. Wissing SA, Kayser O and Miiller RH (2004) Solid lipid nanoparticles for parenteral drug delivery. Mv Drug Del Rev 56:1257-1272. 91. Miiller RH, Radtke M and Wissing SA (2002) Nanostructured lipid matrices for improved microencapsulation of drugs. Int J Pharm 242:121-128. 92. Bunjes H, Drechsler M, Koch MHJ and Westesen K (2001) Incorporation of the model drug ubidecarenone into solid lipid nanoparticles. Pharm Res 18:287-293. 93. Sivaramakrishnan R, Nakamura C, Mehnert W, Korting HC, Kramer KD and Schafer- Korting M (2004) Glucocorticoid entrapment into lipid carriers — characterization by parelectric spectroscopy and influence on dermal uptake. / Control Rel 97:493-502. 94. Lukowski G and Kasbohm (2001) Energy Dispersive X-ray Analysis of loaded solid lipid nanoparticles. J Proceedings — 28th International Symposium on Controlled Release of Bioactive Materials and 4th Consumer & Diversified Products Conference, San Diego, CA, United States, 516-517. 95. Frederiksen HK, Kristensen HG and Pedersen M (2003) Solid lipid microparticle formulations of the pyrethroid gamma-cyhalothrin-incompatibility of the lipid and the pyrethroid and biological properties of the formulations. / Control Rel 86:243-252. 96. Kristl J, Volk B, Gasperlin M, Sentjurc M and Jurkovic P (2003) Effect of colloidal carriers on ascorbyl palmitate stability. Eur J Pharm Sci 19:181-189. 97. Olbrich C, Kayser O and Miiller RH (2002) Lipase degradation of Dynasan 114 and 116 solid lipid nanoparticles (SLN) — effect of surfactants, storage time and crystallinity. Int J Pharm 237:119-128. 98. Olbrich C, Kayser O and Miiller RH (2002) Enzymatic Degradation of Dynasan 114 SLN — Effect of Surfactants and Particle Size. / Nanopar Res 4:121-129. 99. Zimmermann E and Miiller RH (2001) Electrolyte- and pH-stabilities of aqueous solid lipid nanoparticle (SLN) dispersions in artificial gastrointestinal media. Eur J Pharm Biopharm 52:203-210. 100. Mei Z, Li X, Wu Q, Hu S and Yang X (2005) The research on the anti-inflammatory activity and hepatotoxicity of triptolide-loaded solid lipid nanoparticle. Pharmacol Res 51:345-351. 101. Hu LD, Tang X and Cui FD (2004) Solid lipid nanoparticles (SLNs) to improve oral bioavailability of poorly soluble drugs. / Pharm Pharmacol 56:1527-1535. 102. Bargoni A, Cavalli R, Zara GP, Fundaro A, Caputo O and Gasco MR (2001) Transmucosal transport of tobramycin incorporated in solid lipid nanoparticles (SLN) after duodenal administration to rats. Part II. Tissue distribution. Pharmacol Res 43:497-502. 103. Cavalli R, Bargoni A, Podio V, Muntoni E, Zara GP and Gasco MR (2003) Duodenal administration of solid lipid nanoparticles loaded with different percentages of tobramycin. / Pharm Sci 92:1085-1094. 212 Mader 104. Zara GP, Bargoni A, Cavalli R, Fundaro A, Vighetto D and Gasco MR (2002) Pharmacokinetics and tissue distribution of idarubicin-loaded solid lipid nanoparticles after duodenal administration to rats. J Pharm Sci 91:1324-1333. 105. Zara GP, Cavalli R, Bargoni A, Fundaro A, Vighetto D and Gasco MR (2002) Intravenous administration to rabbits of non-stealth and stealth doxorubicin-loaded solid lipid nanoparticles at increasing concentrations of stealth agent: Pharmacokinetics and distribution of doxorubicin in brain and other tissues. / Drug Targ 10:327-335. 106. Chen D, Lu W, Yang T, Li J and Zhang Q (2002) Preparation and characterization of long-circulating solid lipid nanoparticles containing paclitaxel. Yixueban 34:57-60. 107. Miiller RH, Radtke M and Wissing SA (2002) Solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC) in cosmetic and dermatological preparations. Adv Drug Del Rev 54(Suppl 1):S131-S155. 108. Wissing S A and Miiller RH (2002) The influence of the cry stallinity of lipid nanoparticles on their occlusive properties. Int} Pharm 242:377-379. 109. Wissing SA and Miiller RH (2003) The influence of solid lipid nanoparticles on skin hydration and viscoelasticity — in vivo study. Eur J Pharm Biopharm 56:67-72. 110. Wissing SA and Miiller RH (2001) Solid lipid nanoparticles (SLN) — a novel carrier for UV blockers. Pharmazie 56:783-786. 111. Maia CS, Mehnert W, Schaller M, Korting HC, Gysler A, Haberland A and Schafer- Korting M (2002) Drug targeting by solid lipid nanoparticles for dermal use. / Drug Targ 10:489-495. 112. Rudolph C, Schillinger U, Ortiz, A, Tabatt K, Plank C, Miiller RH and Rosenecker J (2004) Application of Novel Solid Lipid Nanoparticle (SLN)-Gene Vector Formulations Based on a Dimeric HIV-1 TAT-Peptide in vitro and in vivo. Pharm Res 21:1662-1669. 113. Gasco MR, Zara GP, Saettone MF and PCT Int. Appl. (2004) Pharmaceutical compositions suitable for the treatment of ophthalmic diseases. Patent application WO 2004039351. 114. Videira MA, Botelho MF, Santos AC, Gouveia LF, Pedroso De Lima JJ and Almeida AJ (2002) Lymphatic uptake of pulmonary delivered radiolabeled solid lipid nanoparticles. / Drug Targ 10:607-613. 115. Stevens PJ, Sekido M and Lee RJ (2004) Synthesis and evaluation of a hematoporphyrin derivative in a folate receptor-targeted solid-lipid nanoparticle formulation. Anticancer Res 24:161-165. 116. Peira E, Marzola P, Podio V, Aime S, Sbarbati A and Gasco MR (2003) In vitro and in vivo study of solid lipid nanoparticles loaded with superparamagnetic iron oxide. / Drug Targ 11:19-24. 117. Wang JX, Sun X and Zhang ZR (2002) Enhanced brain targeting by synthesis of 3',5'- dioctanoyl-5-fluoro-2'-deoxyuridine and incorporation into solid lipid nanoparticles. Eur } Pharm Biopharm 54:285-290. 118. Zara GP, Cavalli R, Bargoni A, Fundaro A, Vighetto D and Gasco MR (2002) Intravenous administration to rabbits of non-stealth and stealth doxorubicin-loaded solid lipid nanoparticles at increasing concentrations of stealth agent: Pharmacokinetics and distribution of doxorubicin in brain and other tissues. / Drug Targ 10:327-335. 10 Lipidic Core Nanocapsules as New Drug Delivery Systems Patrick Saulnier and Jean-Pierre Benoit A new generation of controlled size Lipidic core NanoCapsules (LNC) is presented with respect to their simple formulation, interfacial characteristics, pharmacokinetic and biodistribution properties. We describe their ability to load and release hydrophobic drugs. 1. Introduction The ultimate goal of therapeutics is to deliver any drug at the right time in a safe and reproducible manner to a specific target at the required level. A great deal of effort is currently made to develop novel drug delivery systems that are able to fulfil these specifications. Among them, nanoscale drug carriers appear to be promising candidates. Colloidal carriers are particularly useful because they can provide protection of a drug from degradation in biological fluids and promote its penetration into cells. However, because the body is so well equipped to reject any intruding object, for the materials to stand any chance of success within this hostile yet sensitive environment, they must be chosen very carefully. In particular, attention has to be turned to the composition of the surface of colloidal drug carriers.1 Indeed, their clearance rate from the circulatory system is determined by their uptake by the mononuclear phagocytic system (MPS), which in turn depends on their physico chemical surface characteristics. In order to enhance circulation time, steric protection of various nanoparticulate drug carriers can be achieved by the presence of hydrophilic and flexible polymers to their surface. In the search 213 214 Saulnier & Benoit for injectable, biocompatible and long-circulating systems, many colloidal systems have been evaluated. Different kinds of vectors can be used. For example, molecular vectors where the drug is complexed or associated to a transport molecule are currently used. Many vectors are also constituted by viruses or hybrid viruses, following the modification of their genomes in order to avoid the possibility of replication. In this way, they are used as gene delivery systems. However, we will focus on non viral vectors in this chapter. They are always formulated using soft physico chemical methods, by taking advantage of macromolecular self-assembly properties at the colloidal state in order to produce well-controlled particles. The number of required biological and physico chemical properties of these systems is high in order to formulate operant vectors. One of the most important specifications of these systems is the biocompatibility and biodegradability of each component that needs to be chosen carefully from a restricted list of molecules. Secondly, they need to be well constructed in terms of size and interf acial properties, in order to constitute stealthy systems that will not be phagocyted by the MPS and consequently will have the longest residence time in blood. We should not forget that such vectors exist biologically. Low density lipoproteins (LDL) are interesting systems possessing many of the required specifications. Unfortunately, their extraction, purification or reconstitution is still a challenge with strong physico chemical problems to solve. No convenient common solvent of proteins and lipids exists in order to reconstitute a similar supra-molecular framework. Consequently, we have to keep in mind a formulation of nanoparticles with biomimetic properties to those related to LDL as close as possible. We would now like to describe a novel class of nanoparticles (Lipidic core NanoCapsules:LNC) formulated without organic solvents with biocompatible and biodegradable molecules.2 We will see that after modification of the composition, we can control their size without difficulty in the 10-200 nm range, with a monomodal and narrow size distribution. Initially, we suggest describing the LNC formulation following some particular auto-organizational properties of Poly Ethylene Glycol (PEG)-like surfactants, induced by several emulsion-phase inversions in which they are incorporated. We will particularly emphasize the different physical methods that determine the characterization of the final structure of LNC, as well as their stability in suspensions. Then, we will describe strong correlations between their stealthy properties in blood and structural characteristics, mainly size and interfacial properties. In specific, we have evaluated the activation of the complement system in an original in vitro model. These nanocapsules are devoted to the encapsulation of drugs that need to be dispersed in their oily core. As a proof that the concept works, we will describe the ability of LNC to encapsulate and release simple lipophilic molecules, ibuprofene and amiodarone, in the last paragraph. Lipidic Core Nanocapsules as New Drug Delivery Systems 215 2. Lipidic Nanocapsule Formulation and Structure 2.1. Process The first step of the process consists of the formulation of a stable emulsion characterized by its oily phase (O), aqueous phase (W) and finally its surfactants mixture (S). Due to the complexity of the mixture, the brand names will be used throughout the following text. It is important to note that no organic solvent or mediumchain alcohols are used in the formulation. All these molecules are known to be biocompatible and biodegradable. This indicates that the lack of residual toxicity can guarantee the safe use of LNC for human administration. Solutol® is mainly comprised of 12-hydroxystearate of PEG 660 that corresponds to a hydrophilic surfactant (HLB = 11). The lecithin used is a mixture of hydrophobic phospholipids. The main compounds of each phase are reported in Table 1. The beginning of the formulation (see Fig. 1) corresponds to a magnetic stirring of all the components for which the proportions will be defined later, with a gradual rise in temperature from room temperature to 80°C at a rate of 4°C/min, leading to an W/O emulsion characterized by low conductivity. The system is Table 1 Compounds used in the LNC formulation. S • Solutol® HS-15:12-hydroxysterarate of PEG 660 and PEG 660 (low content) • Lipoid®: lecithin O • Labrafac®: triglycerides (C8-C10) W • Purified water • NaCl Fig. 1. Emulsion-phase inversion induced by temperature changes and the principle of LNC formulation. 216 Saulnier & Benoft cooled from 80 to 55°C (4°C/min), leading to an O/W emulsion characterized by its high conductivity. Between these two kinds of emulsion, a transition zone called the Phase Inversion Zone (PIZ) is defined where the system is known to be in bicontinuous states.23 In order to provide appropriate and optimal interfacial properties to the wateroil interfaces, the formulation typically requires three temperature cycles across the PIZ. The system is stopped at a temperature corresponding to the beginning of the PIZ, just before performing a final, fast-cooling dilution process in cold water (2°C). This second step of the formulation leads to LNC in suspension in an aqueous phase. The interfacial rheology method developed in several papers demonstrates that the interfacial association of all the implicated molecules of the process is different from other commoner systems.4 Cohesion energy at the interface, as well as the interaction of the interfacial molecules with the adjacent phases, reaches a minimum for the concentrations used. We think that this particularity can explain why the system can be broken down in an ideal way during final dilution. The surfactants involved in the stabilization of the bicontinuous systems can easily leave the microemulsion in order to constitute the colloidal structures (LNC). It might be noted that temperatures corresponding to the PIZ are much too high to decline this method to the simple encapsulation of thermo-sensitive molecules. Fortunately, we have shown that the electrolyte concentration (NaCl) strongly influences the location of PIZ on the temperature scale. When we increase the electrolyte concentration, we decrease the PIZ temperature to reach acceptable levels. 2.2. Influence of the medium composition Obviously, the presence or not of LNC strongly depends on the composition of the system reported in Fig. 2(a) as a pseudo-ternary diagram.5 Each point corresponds to strictly similar formulation processes and the entire diagram describes the appropriate feasibility zone. It should be noticed that the optimal formulation corresponds to w /w concentration of around 20% for the oil phase, 60% for the water phase and 20% for Solutol®. In the zone corresponding to the LNC formulation, a statistical model is applied in order to approximate the influence of the composition on the size distribution measured by the dynamic light scattering method. Polynomial interpolations between well-controlled points are performed. The corresponding results are reported in Fig. 2(b) where different iso-size curves are presented. The same procedure was applied to the size variation coefficients. These two curve beams are powerful tools, allowing an optimized formulation to be found, once a given and reproducible size distribution is elaborated just by tuning the composition. Lipidic Core Nanocapsules as New Drug Delivery Systems 21 7 (a) (b) Fig. 2. Feasibility diagram of LNC. a: zone of favorable formulation; b: iso-size curves in the favorable zone. Fig. 3. Schematic representation of LNC. It is important to note that LNC have demonstrated very good freeze-drying and stability characteristics in storage conditions for several months, as determined by DSC measurements,6 confirming the structure presented in Fig. 3. LNCs are constituted of a lipidic core surrounded by a surfactant shell, where lecithin is located in the inner part of the shell and the Solutol® in the outer part. 2.3. Structure and purification of the LNC by dialysis Considering that in the biological environment of the blood stream, the particles interact strongly with various interfaces, one possible model for studying the interfacial behavior of these particles is their spreading at the air-water interface. Classically, the Langmuir balance was used to describe interfaces composed by simple 21 8 Saulnier & Beno?t mixtures. The basic technique was the measurement of the surface pressure (7r)-area (A) isotherm, by determining the decrease in surface tension as a function of the area available for each molecule on the aqueous sub phase. This included the study of the monolayer formation, the compressibility of the interface, the mutual interactions of molecules in the monolayer, but also interactions with the sub-phase molecules across interfacial rheological measurements.7 Following this, these suspension spreading results were compared with zeta potential measurements. These studies8,9 clearly indicate that the mother suspension, just after dilution in cold water, is composed of • Stable nanocapsules as described before; these objects diffuse strongly in the aqueous phase after spreading at the air/water interface. • Unstable nanocapsules with similar size, but with a lower amount of phospholipids (Lipoid®) in the inner part of their shell. These capsules are not sufficiently robust to support the interfacial energies during spreading. Consequently, the components or fragments of the initial particles can be detected at the air-water interface. • Free PEG (minor component of the Solutol®) released from the outer part of the shell. It is obvious that the excess of PEG, as well as an important fraction of the unstable particles could be limited by dialysis. We will see in the next chapter an original investigation of these dialysis effects. 2.4. Imagery techniques AFM images [Fig. 4(a)] were obtained after spreading the initial suspension of 50 nm (±10 nm) LNC on a fresh mica plate, and then allowing a complete evaporation of the water at room temperature. A contact mode was applied with a contact force of 10 nN, as well as a non contact mode without modification of the related images. The particle shape looked like a cylinder, 2nm high and 275 nm wide, corresponding to a total volume similar to a 60 nm sphere. We demonstrate the deformation of LNC after water evaporation, but without fusion of the particles, something that often occurs with liposomes. Classical TEM images were taken of the covered copper grids, following staining with a 2% phosphotungstic acid aqueous solution. It is noted on Fig. 4(b) that the lateral diameters are relatively polydispersed in a 20-70 nm range. Fig. 4(c) corresponds to a cryo-TEM image (kindly provided by Olivier Lambert, IECB-UBS UMR CNRS 5471) where individualized LNC are detectable. It is important to note that this image was performed after a dialysis, followed by an appropriate dilution of the mother suspension. Lipidic Core Nanocapsules as New Drug Delivery Systems 219 (b) TEM (c) Cryo-TEM <>,J 'HUE ^ > * "s ' * * *o * If Fig. 4. Visualization of LNC by (a) AFM, (b) TEM and (c) cryo-TEM. 3. Electrical and Biological Properties 3.1. Electro kinetic comportment The stable Lipid NanoCapsules (LNC) contain pegylated 12-hydroxy stearate, as well as free PEG in the outer part of the shell, which can be an important biological specification that we will describe latter. The distribution of PEG chains at the surface was determined by their electrokinetic properties. Thus, electrophoretic mobility was measured as a function of ionic strength and pH, for particles differing in sizes, dialysis effects, and the presence or not of lecithin in their shell. The study enabled us to find the isoelectric point (IEP) as well as the charge density (ZN) in relation to the dipolar distribution in the polyelectrolyte accessible layer (thickness 1 A), by using soft particle electrophoresis analysis10 (see Fig. 5). This study showed that LNC presented electrophoretic properties conferred by PEG groups at the surface constituting dipoles that are able to interact with counter ions (H+, Na+) or water dipoles. The levels of IEP, ZN and 1/1 changed after dialysis, due to the removal of molecules that were poorly linked (mainly free PEG) at the outer part of the surface, allowing accessibility to the inner adjacent part of the shell. Water shell Fig. 5. Accessible layer to counter ions characterized by its thickness (1 A ) and its dipolar charge density (ZN). (a) AFM * 1 m ! it i • '• :&&&* • • ' . , . ) A • •••: . : - ? .' •••-ViCS. . ' . i f , . I - " • • V o ^ -7 L—» J*^ •, ft". - •' n •:•%*. '.••-.•:?»••&* H -':' •''"•"' v ' • •' • .•.:.->.;.:- •• .. m • « " • • > • • •, .) urn 220 Saulnier & Beno?t 100 nm LNC presented the best-organized and the accessible part of the shell, compared with other sizes of LNC, before and after dialysis. Lecithin was found to be present in the inner part of the polyelectrolyte layer and was found to play a role in the disorganization of the outer part. Dialyzing LNC formulated with lecithin led to stable and well structured nanocapsules, ready for an in vivo use as a drug delivery system.11 3.2. Evaluation of complement system activation Generally, after intravenous administration, nanoparticles (NP) are rapidly removed from the blood stream because they are recognized by cells of the MPS such as Kiipffer cells in the liver, or spleen and bone-marrow macrophages. However, a brush of PEG chains grafted on the surface is known to decrease the recognition of nanoparticles by the immune system after intravenous administration.12 One has demonstrated that a strong correlation prevails between the complement activation and the stealthy properties of LNC. Therefore, these properties were evaluated by measuring the degree of complement activation11 [CH50 technique and crossed Immunoelectrophoresis (C3 cleavage)] and the level of macrophage uptake, in relation to the organization of PEG chains, according to the electrokinetic properties of the LNC surface. These experiments were performed on 20, 50 and 100 nm LNC before and after dialysis. The CH50 technique is presented in Fig. 6. Nanoparticles are dispersed in human serum with sensitized erythrocytes. After incubation, lysis is evaluated by a classical spectrophotometric method. The measured absorbance is related to the consumption of complement proteins by particles. The main conclusions are that whatever the in vitro test, all LNC were not recognized by the non specific components of the immune system. It was probably due to the strong density of PEG chains at their surface. Furthermore, dialysis maintains a sufficiently high density of PEG and had no incidence on the complement consumption. 4. Pharmacokinetic Studies and Biodistribution At first, the biodistribution of radiolabeled nanocapsules was studied by scintigraphy and y counting, after intravenous administration in rat whereby the 99mTc-oxine was incorporated in the lipid core and 125I labelled the shell of the nanocapsules.13 Dynamic scintigraphic acquisition was carried out 3 hrs after administration and y activity in blood and tissues was followed for more than 24 hrs (see Fig. 7). An early half-disappearance time of about 47 ± 6 min was found for 125I and 41 ± 11 min for 99mTc. These ranges of residence times were interesting for specific Lipidic Core Nanocapsules as New Drug Delivery Systems 221 «—car"" Lysis of erythrocytes M, **CSf—. ^B Sheep erythrocyte • Complement proteins M Amibody anti-sheep eryihrocyie No lysis of ervlhrocvtes Fig. 6. CH50 method for the evaluation of complement system activation. 200 300 Time (min) 500 600 Fig. 7. Evolution of radioactivity blood repartition after the intravenous administration of LNC expressed as a percentage of the injected dose. 222 Saulnier & Benoit site delivery. Meanwhile, it appears that the length of the PEG chain (in this case, 15 ethylene oxyde groups per molecule) should be increased to extend the vascular residence time. Recently, it has been shown that adding different DSPE-PEG to the system enhances the t1/2 values to several hours, depending on the concentration and the PEG length.14 t1/2 (half-life), MRT (Mean Residence Time) in blood and AUC (Area Under Curve) were evaluated by using [3H]-cholesteryl hexadecyl ether mixed with the lipid and the surfactant at the beginning of the formulation. The main conclusion was that the LNC formulated in this study compared advantageously with other nanoparticulate systems, particularly for their residence time in blood. Nanocapsule uptake by the different organs of rat was evaluated 24 hrs after intravenous administration. It was shown that LNC deposited mainly in the liver and the spleen, but also in the heart, and the results were comparable to a liposome reference. 5. Drug Encapsulation and Release 5.1. Ibuprofene LNC were characterized for their suitability as an ibuprofene delivery device for pain treatments.15 After in vitro investigations, ibuprofene- loaded LNC were evaluated after intravenous and oral administration in rats. For each system, the carrier was evaluated through its potential antinociceptive efficiency. We present in Fig. 8, the release of ibuprofene in a phosphate buffer after its incorporation in LNC during formulation. For each case, LNC provide high ibuprofene loadings (95%). The main feature is an initial burst followed by a 100 CP" 80 o^ W ) 8 60 J 40 2 Q. fi 20 0 0 4 8 12 16 20 24 time [h] Fig. 8. Ibuprofene release from three batches of drug-loaded LNC in phosphate buffer (pH = 7.4). Lipidic Core Nanocapsules as New Drug Delivery Systems 223 sustained release, where the time for 50% drug released values are 0.84,1.78 and 2.29 hrs for 2,10 and 20 mg/ml loaded amounts respectively. Furthermore, after oral administration, these nanocarriers offered a better bioavailability as well as prolonged antinociceptive effects than other nanoparticulate systems. 5.2. Amiodarone Amiodarone is widely used because of its anti-anginal and anti-arrythmic properties. Unfortunately, this molecule can provoke severe adverse effects due to its accumulation in other tissues, after classical intravenous or intraperitoneal administration. In that manner, the use of LNC was evaluated in in vitro conditions in order to incorporate and release amiodarone from their lipidic core.16 It was found that sustained drug release was achieved over a range of significant period between 25 hrs and 263 hrs depending on the pH of the release medium. 6. Conclusions A new kind of colloidal drug carrier, the LNC, was formulated without organic solvent or toxic surfactants via a rapid and easy protocol. These nanoparticulate systems were designed in order to have biomimetic properties and can be considered as pseudo-lipoproteins. A lipidic core is surrounded by a surfactant shell, stabilized by phospholipids in the inner part of the shell and by stearate of PEG in the external part of the shell. The structural characteristics of these carriers allow for the incorporation in their core of different lipophilic drugs, initially dispersed in the oily phase at the beginning of the formulation. The narrow size distribution can be selected anywhere in a 10-200 nm range. One of their most important features is the presence of PEG groups on the surface ideally presented, providing very low recognition. These surface properties are crucial in order to hide the LNC from the MPS system. In addition, the presence of hydroxyl groups should allow the functionalization of the LNC surface to attach ligands of interest and to improve the specificity of drug targeting. These PEG groups could participate to the inhibition of the efflux pumps involved in multidrug resistance. In this context, preliminary in vitro studies are very promising. Different strategies of incorporation or attachment of a specific ligand of the BBB and also of a glioblastoma tumor are studied. The linkage of a grafted monoclonal antibody is currently evaluated. It could provide an interesting way of accumulating LNC in the brain after intravenous administration. The release of simple lipophilic drugs taken as models were analyzed and have shown sustained release over several days. Furthermore, since we are able to elaborate stable lipidic complexes (hydrophilic molecules associated to a cationic lipid 224 Saulnier & Benoit for example), these vectors could provide an interesting alternative to liposomes for the delivery of hydrophilic compounds like DNA or proteins. References 1. Heurtault B, Saulnier P, Pech B, Proust JE and Benoit JP (2003) Physico-chemical stability of colloidal lipid particles, Biomaterials 24:4283. 2. Heurtault B, Saulnier P, Pech B, Proust JE, Richard J and Benoit JP (2001) Lipidic nanocapsules, formulation process and use as a drug delivery system, Patent No. W001 /64328. 3. Heurtault B, Saulnier P, Pech B, Proust JE and Benoit JP (2002) A novel phase inversionbased process for the preparation of lipid nanocarriers, Pharm Res 19(6):875. 4. Heurtault B, Saulnier P, Pech B, Proust JE and Benoit JP (2002) Properties of polyethylene glycol 660 12-hydroxy stearate at a triglyceride/water interface, Int J Pharm 242:167. 5. Heurtault B, Saulnier P, Pech B, Venier-Julienne MC, Proust JE, Phan-Tan-Luu R and Benoit JP (2003) The influence of lipid nanocapsule composition on their size distribution, Eur J Pharm Sci 18:55. 6. Dulieu C and Bazile D (2005) Influence of lipid nanocapsules composition on their aptness to freeze-drying, Pharm Res 22(2):285. 7. Heurtault B, Saulnier P, Pech B, Proust JE and Benoit JP (2003) Interfacial stability of lipid nanocapsules, Coll SurfB: Biointerf 30:225. 8. Minkov I, Ivanova Tz, Panaiotov I, Proust JE and Saulnier P (2005) Reorganization of lipid nanocapsules at air-water interface: I. Kinetics of surface film formation, Coll Surf B45(l):14. 9. Minkov I, Proust JE, Saulnier P, Ivanova Tz and Panaiotov I (2005) Reorganization of lipid nanocapsules at air-water interface: Part 2. Properties of the formed surface film, Coll SurfB 44(4): 197. 10. Vonarbourg A, Saulnier P, Passirani C and Benoit JP (2005) Electrokinetic properties of noncharged lipid nanocapsules: Influence of the dipolar distribution at the interface, Electrophoresis 26(11 ):2066. 11. Vonarbourg A, Passirani C, Saulnier C, Simard P, Leroux JC and Benoit JP, Evaluation of pegylated lipid nanocapsules versus complement system activation and macrophage uptake, / Biomed Mater Res A , in press. 12. Passirani C and Benoit JP (2005) Biomaterials for delivery and targeting of proteins and nucleic Acids, Ed. Ram and Mahato I, CRC Press: Boca Raton London New York Washington, D.C., 187. 13. Cahouet A, Denizot B, Hindre F, Passirani C, Heurtault B, Moreau M, Le Jeune JJ and Benoit JP (2002) Biodistribution of dual radiolabeled lipidic nanocapsules in the rat using scintigraphy and y counting, Int J Pharm 242:367. 14. Hoarau D, Delmas P, David S, Roux E and Leroux JC (2004) Novel long-circulating lipid nanocapsules, Pharm Res 21(10):1783. 15. Lamprecht A, Saumet JL, Roux J and Benoit JP (2004) Lipid nanocarriers as drug delivery system for ibuprofen in pain treatment, Int ] Pharm 278:407. 16. Lamprecht A, Bouligand Y and Benoit JP (2002) New lipid nanocapsules exhibit sustained release properties for amiodarone, / Control Rel 84:59. 11 Lipid-Coated Submicron-Sized Particles as Drug Carriers Evan C. Unger, Reena Zutshi, Terry O. Matsunaga and Rajan Ramaswami Lipid-coated submicron-sized particles afford a new platform for drug delivery and therapy. In this chapter, we will discuss the characteristics and some of the potential clinical applications of submicron-sized particles. 1. Technology In general, bubbles present a hydrophilic exterior and hydrophobic interior stabilized by detergents. Detergents are characterized by their polar head group and a hydrophobic domain consisting of long chain fatty acids, alcohols, ethers, etc. In bubbles, (Fig. 1) detergents aggregate by orienting the hydrophilic polar groups on the outside in contact with aqueous environment and stacking their hydrophobic sections of alkyl chains on the inside, away from the water.1 This results in an energy minimized spherical structure that can incorporate gas and/or other hydrophobic materials inside. Phospholipids are specialized surfactants with characteristics similar to that of detergents and can stabilize micron- and submicron-sized gas bubbles, especially when perfluoropropane is used. Perfluorocarbon (PFC) gases have low solubility in aqueous media, relatively high molecular weight and can be used to prepare stable microbubbles or submicron-sized bubbles (SMBs) that are less than 1 micron in diameter. Fully 225 226 Ungeretal. e n v i r o n m e n y ^ p P I l O T ^ Hydrophobic tail "e^" ^~ Hydrophilic head \ Fig. 1. The structure of bubbles. halogenated PFCs are inert and generally not metabolized in the body. The table below lists some of the different PFCs and other gases useful in making lipid-coated microbubbles and PFC emulsions. Compound Nitrogen Sulfur hexafluoride (SF6) Perfluoropropane (C3F8) Perfluorobutane (C4F10) Perfluoropentane (C5F12) Perfluorohexane (C6F14) Molecular 28 147.07 188.2 238.04 288.05 338.06 Weight Boiling Point -195.8 Sublimes -36.7 - 2 29.5 57.11 Solubility in Water Sparingly Sparingly Insoluble Insoluble Insoluble Insoluble Lipid-coated submicron-sized particles may be prepared by agitating an aqueous mixture of lipids with a selected gas (as in the approved product Definity® microbubbles ultrasound contrast agent, marketed by Bristol-Meyers Squibb), by lyophilizing the material and storing with a head space of the pre-selected gas, spray-drying or by creating an emulsion of the gas in the bubbles, e.g. when a PFC material is used to formulate the particles below the boiling point of the gas, while the PFC is in its liquid state. The properties of the lipid-coated bubble will vary in part, depending upon the gas or material that is encapsulated in the bubble. Lipid-coated microbubbles of air or nitrogen will be relatively short lived. Following intravascular injection, bubbles composed of air or nitrogen may be stable enough to pass from the right heart through the pulmonary circulation and into the left heart, but will likely be unstable to undergo multiple passes through the circulation. When bubbles are prepared from air or nitrogen, the gas is relatively water-soluble and diffuses rapidly Lipid-Coated Submicron-Sized Particles as Drug Carriers 227 across the lipid membrane into the blood. PFC gases have much lower solubility in the blood and therefore make more stable bubbles. The solubility of the PFC in part reflects the molecular weight of the compound with the higher molecular weight materials generally having lower solubility, and the boiling point of the materials increases with increasing molecular weight. PFCs with 4 or less carbon atoms will be gases at room temperature. Dodecafluoropentane has a boiling point of about 28.5°C. Liquid perfluoropentane filled phospholipid-coated submicron droplets (SMDs) may volatilize in vivo to form gas bubbles after intravenous injection. Perfluorohexane will be a liquid at physiological temperature, but because of its vapor pressure, a small fraction of the material may be in gaseous state at physiological temperature. A wide variety of different lipids can potentially be employed to make the lipid-coated microbubbles. Experiments were performed using agitation to prepare the microbubbles. Microbubbles were more stable when prepared when the lipids were at gel state and when the same chain length of lipids was used in the formulation. The product Definity was developed using a blend of lipids: dipalmitoylphosphatidylcholine (DPPC), dipalmitoyl-phosphatidic acid (DPPA) and dipalmitoylphosphatidylethanolamine-PEG5000 (DPPE-PEG5000, polyethyleneglycol, MW = 5000). The product is primarily composed of the neutral lipid DPPC. The anionic lipid in the formulation may aid in the electrostatic repulsion of the bubbles and the PEG may form a steric barrier to further prevent aggregation or fusion of the bubbles. The microbubbles have various potential medical applications: (a) They can be used as the active drug product. (b) They can be coadministered with other biologically active drug substances. (c) Biologically active drug materials can be incorporated into the hydrophobic domain of the microbubbles. (d) Biologically active gases can also be entrapped inside the bubbles and used for delivery. In (a), the bubbles themselves can be used as contrast agents with diagnostic ultrasound, or as therapeutic agents with therapeutic ultrasound. An unusual feature of lipid-coated microbubbles compared with other delivery systems is that these agents can be activated by energy, particularly using ultrasound for localized therapy. The ultrasound energy can be targeted precisely to small regions in the body. Localized therapy and drug delivery can be accomplished using ultrasound to activate the bubbles to disperse a clot, increase local capillary permeability or to release drugs from the bubbles [as in (c) & (d) above]. Figure 2 below shows some of the ways that bubbles may be used to deliver therapeutic agents. The following sections will review some of the specific biomedical applications. 228 Ungeretal. Fig. 2. Microbubbles can be used to transport materials in a variety of methods. In (a), the drug is injected in conjunction with the bubbles and driven into the target tissue by the acoustic activation of the bubbles. To create drug carrying bubbles, the agent may be: (b) attached to the outside of the lipid, (c) embedded in the lipid layer, (d) associated to the membrane by electrostatic interactions, (e) encapsulated directly in the bubble, or (f) dissolved in an oil or other compatible liquids and then encapsulated within the bubble. Also, smaller bubbles or spheres (e.g. delivering gene products) may encapsulate the agent and then associate with larger bubbles as in (g). 2. Ultrasound Contrast Agents In biological fluids and tissues, microbubbles are efficient reflectors of sound. Ultrasound is the most common biomedical imaging modality and ultrasound contrast agents are used to increase the reflectivity or backscatter of blood and tissues. In ultrasound imaging the reflectivity of the bubbles is proportional to r6 where r = the radius of the microbubble.2,3 This implies that the larger the bubble, the more efficient it is as a reflector of sound. For intravascular applications, however, the bubbles must be smaller than the diameter of a red blood cell to safely pass through the capillaries without causing vascular blockade. The relationships between reflectivity and effectiveness as an ultrasound contrast agent are far more complex than the prediction based upon mere size or diameter of the microbubbles. Biomedical ultrasound is commonly performed over a range of ultrasound frequency from 1.5 MHz to 20 MHz. The resolution of ultrasound increases as the frequency increases (shorter wavelengths of ultrasound), but penetration in tissues also decreases linearly with frequency. The highest frequencies of ultrasound (e.g. 20 MHz) are mainly used for imaging with catheter-based ultrasound Lipid-Coated Submicron-Sized Particles as Drug Carriers 229 (e.g. for looking within vessels), or for imaging very superficial tissues such as the skin. The lower frequencies (e.g. 3 to 7 MHz) will penetrate abdominal tissues and other structures for general purpose imaging. Bubbles have resonant properties and reflect sound most efficiently at their resonant frequencies. For example, the resonant frequency of a 1 micron bubble is approximately 9.5 MHz and the resonant frequency of a 5 micron bubble is approximately 1.3 MHz.4 When insonated at their resonant frequencies, microbubbles will emit harmonic signals at higher frequencies. For example, when microbubbles are insonated with a fundamental insonation frequency = B0, the bubbles will reflect signals at the B0 frequency as well as signaling powerfully at 2xB0. The images below (Fig. 3) show fundamental and harmonic images of the heart pre- and post-contrast, using the lipid-coated microbubble contrast agent, Definity®. As shown in the images, the harmonic image, obtained from the signal twice the insonation frequency, has higher contrast and greater suppression of background signal. Harmonic imaging (sampling the 2x B0 signal) as well as other techniques enable ultrasound to suppress signal from the background and to enhance the signal from blood. Fundamental 2nd Harmonic Fig. 3. Comparison of fundamental and second harmonic imaging. The second precontrast harmonic image is clearer. The contrast was unclear post-contrast, when the sonographer switched back to fundamental imaging. The infusion rate was then increased, resulting in excessive shadowing, thereby further obscuring the details. Courtesy of Kevin Wei, MD, University of Virginia.5 230 linger et al. Another factor contributing to the effectiveness of the ultrasound contrast agent is the elasticity of the shell surrounding the microbubbles. The more elastic the shell, the more efficient the bubble may be as a reflector of sound.6 A less elastic shell will not only decrease the efficiency with which sound is reflected, but will also raise the frequency at which the bubble resonates. Lipid coatings surrounding microbubbles are thin and are most likely monolayers or bilayers of lipid, and as such, are relatively elastic compared with other materials that may be used to coat bubbles such as cross-linked synthetic polymers. The lipid materials used in coating the microbubbles, enable production of highly elastic and efficient reflectors of ultrasound. In preparing bubbles for drug delivery, as materials are added to the bubbles, the bubbles may become less elastic and require higher amounts of ultrasound energy for activation. In certain imaging regimes, bubbles can be ruptured from the ultrasound energy. As shown in Fig. 4 below, microbubbles lower the threshold of ultrasound energy for cavitation to occur.7'8 Cavitation creates an acoustic signal that can be detected. Cavitation can occur as a stable inertial cavitation of a bubble, where the bubble expands and collapses in concert with the phase of the waves of ultrasound. At higher energy, cavitation can lead to localized energy deposition analogous to a local explosion on the microscopic scale. Cavitation can be used for therapeutic purposes as described below, or it can be used to create a strong ultrasonic signal Initial radius (|im) Fig. 4. Plot of the cavitation threshold in water as a function of initial nucleus radius for three frequencies of insonification: 1,5, and 10 MHz. Nuclei consist of air bubbles initially at 300 K that undergo growth in a single cycle of ultrasound and collapse adiabatically to a temperature of 5000 K. Surface tension, viscosity, and inertia of the host fluid are included in this analytical model (Holland and Apfel, 1989). For 5 MHz, the optimal nucleus radius is 0.3 [im with a corresponding cavitation threshold, P0pt, of 0.58 MPa peak negative pressure. Note that at a pressure P' greater than P0pt, a broader size range of nuclei cavitate. (Reproduced by permission of Apfel and Holland, Ultrasound Med Biol.) Lipid-Coated Submicron-Sized Articles as Drug Carriers 231 for diagnosis and detection of diseases. Calculations indicate that with cavitation imaging, it is possible to detect a single microbubble.9 Bubbles as ultrasound contrast agents open the field of diagnostic ultrasound to molecular imaging. In terms of sensitivity to concentration of material, ultrasound imaging using bubbles rivals the most sensitive imaging techniques such as nuclear medicine. In addition to cavitation-based imaging which may detect a single bubble, when bubbles are targeted to a certain structure and accumulate to present an interface of several or more bubbles, they may form a so-called specular reflector or highly efficient interface for the reflection of ultrasound. Our group has created targeting ligands for incorporation into the microbubbles and performed imaging of models of disease with these contrast agents.10,11 The images below show a thrombus in the left atrial appendage in a dog, pre- and postcontrast. The thrombus is not visible on the ultrasound images precontrast, but is readily detected when it is postcontrast. Figure 6 is a depiction of the bioconjugate that was synthesized to develop the thrombus targeted ultrasound contrast agent. The lipid in the bioconjugate serves as a hydrophobic anchor to bind the bioconjugate to the surface of the bubble. A small number of bubbles bound to the surface of a target such as a thrombus, appear to be sufficient for contrast enhancement and detection on ultrasound. A number of different molecular targets have been imaged with ultrasound using targeted bubbles. Some of the different diseases that have been imaged include vulnerable plaque, inflammation, angiogenesis and ischemia. Figure 7 shows P-selectin targeted imaging in a model of ischemia with myocardial contrast echocardiography (MCE). Fig. 5. Ultrasound images of left atrial appendage (LAA) clot enhancement in the canine model using an intravenous infusion of targeted microbubbles at a dose of 0.01 cm3 /kg: precontrast image (left); postcontrast infusion image (right), highlighting clear enhancement of the clot in the left atrial appendage. AO, aorta; LA, left atrium; PA, pulmonary artery. [Reproduced with permission of Unger et ah, in Ultrasound Contrast Agents, 2nd Edition, Goldberg, Raichlen & Forsberg (eds.).] 232 Ungeretal. Anchor Tether Ligand Fig. 6. Microbubble with bioconjugates attached with enlarged view showing the anchor, tether and ligand. Reproduced by permission Unger et ah, EjR. Risk Area P-selectin Fig. 7. The left panel shows an area of hypoperfusion imaging with MCE during myocardial ischemia of the left circumflex artery. The right panel shows enhancement 60 min after reflow from P-selectin-targeted imaging in the risk area. Courtesy of Jonathan R. Lindner, MD, University of Virginia.13 3. Sonothrombolysis Bubble-assisted sonothrombolysis is the term we use to describe the ultrasoundmediated cavitation of bubbles to aid in the lysis of venous and arterial thrombi. MRX-815 is the designation for ImaRx Therapeutic Inc.'s (ImaRx) manufactured phospholipid-coated submicron-sized bubble product that will be used in clinical trials. MRX-815 is the next generation bubble, developed based on the Definity bubble. The bubbles in MRX-815 exhibit a size profile where 70% of the particle Lipid-Coated Submicron-Sized Particles as Drug Carriers 233 •v ,." "'"• 1 nm >atex beads l : < r> % V - J1 - • • • © • • • • • V - : • • . • • ; ' . . • • , , . « : WOK Fig. 8. Sizing studies of submicron bubbles. distribution is less than one micron and the mean size is less than 1 micron in diameter.14'15 The images below depict a photomicrograph of MRX-815 bubbles alongside a photomicrograph of one-micron size latex microbeads. The bubbles are one micron in diameter and smaller. We found in our lab that the smallest bubbles are not well shown on the light microscopy due to limitations of the imaging technique. The sizing profile shows that there are bubbles up to approximately two microns in diameter, but more than 70% of the bubbles are smaller than one micron in size. 16,17 Investigators have demonstrated that ultrasound can be used to generate cavitation in an aqueous medium.18 Cavitation research has led to studies involving ultrasound-mediated clot lysis at a variety of frequencies.19"23 Furthermore, microbubbles and submicron-sized bubbles provide a nucleus at which cavitation can occur, thereby lowering the ultrasound energy requirements.24 While intravenous administration with local application of ultrasound appears to be effective for sonothrombolysis in both pre-clinical and clinical models, applications using an infusion catheter are also being investigated.25 It is believed that submicron-sized bubbles and ultrasound-mediated cavitation are able to affect the thrombus architecture by increasing permeability through the thrombus matrix, thereby improving accessibility and the penetration of thrombolytic enzymes to more efficiently lyse clots. Studies by Francis et a/.,26'27 demonstrated that ultrasound alone increased the spacing between fibrin strands in clots, presumably improving the penetration of lytic enzymes, such as t-PA, into the clot. By way of explanation, when bubbles are insonified, these bubbles can oscillate in response to the acoustic pressure wave. If driven with a sufficient acoustic pressure, the rapid expansion and contraction of the bubble will result in local 234 Ungeretal. velocities at the bubble surface on the order of hundreds of meters per second. If the expansion of the bubble is large enough, the bubble will become unstable, resulting in the destruction of the bubble into smaller fragments.28 The rapid oscillation of the bubble in response to an acoustic pulse is referred to as "cavitation". Bubbles undergoing this violent expansion and contraction produce liquid jets, local shock waves, and free radicals. Although the exact mechanism is still being studied, the effect of cavitating bubbles has been demonstrated to have several effects on the surrounding tissues, including the poration of cell membranes resulting in enhanced membrane permeability (sonoporation) or the disruption of local thrombus. Thus, the combination of ultrasound with microbubbles has potential applications in blood clot dispersion and local drug delivery to treat cardiovascular disease, cancer, and diseases of the central nervous system. The figure below shows individual images from ultra-high speed videomicroscopy of a single bubble. The bubble is shown in the resting state on the far left hand side of the figure. The bubble expands after the application of the ultrasound pulse, then collapses and fragments. The daughter bubbles expand and collapse again, leaving behind small nano-sized fragments.29 Localized activation of bubbles with ultrasound can be used for a number of different medical applications including SonoLysis. Whereas in diagnostic ultrasound contrast imaging where there is an r6 dependence between size and ultrasound reflection for therapy, it is advantageous to have much smaller bubbles. As shown in Fig. 10, when bubbles are cavitated by ultrasound, they may undergo a relatively greater increase in the expansion ratio ri/ror where r^ is the maximum size for the radius of the bubble after insonation, and r0 — the initial resting radius.31 The relative expansion with insonation is greatest for the smallest diameter submicron-sized bubbles. This conceivably results in a more effective cavitational force, and hence more efficient lysis of thrombi. Another effect of ultrasound on microbubbles which has the potential to be utilized therapeutically is the use of acoustic radiation force to selectively concentrate microbubbles at a target site.32'33,34 Microbubbles driven with ultrasound, experience radiation force in the direction of ultrasound wave propagation.35 Pulses of t % '2 IV \ ,' V y. Fig. 9. In the images above, a single 3 (im bubble is shown (far left) in the resting state. Insonation with a single pulse of ultrasound energy causes the bubble to expand, collapse, and fragment, yielding nanometer-sized fragments. As the bubbles expand and collapse, they generate a local Shockwave that can be used therapeutically. Reproduced with permission from Chomas et (A., Appl Phys Lett, 2000.30 Lipid-Coated Submicron-Sized Particles as Drug Carriers 235 Fig. 10. The relationship between nanobubbles' size at resting state and expansion ratio under insonation. Reproduced with permission of D. Patel et a\., IEEE Ultrasonics, Ferroelectrics, and Frequency Control. In press. many cycles can deflect resonant microbubbles over distances in the order of millimeters. Thus, it may be possible to bring microbubbles circulating in the blood pool into contact with targeting sites on a blood vessel wall, in a region selected by the positioning of the ultrasound beam. This effect has been demonstrated to increase the retention of microbubbles at a target site over an order of magnitude.3 6 In addition to favorable acoustic characteristics, submicron-sized bubbles have other potential advantages for therapy, compared with larger-sized microbubbles. The smaller bubbles may penetrate a clot more easily and may have better biodistribution characteristics for targeting. The pictorial representation below (Fig. 11) is the hypothetical mechanism of action for MRX-815 bubbles flowing through the vasculature in association with Fig. 11. It is hypothesized that when submicron-sized bubbles are injected systemically, some will aggregate on the thrombus, and due to their small size, work into the clot. When the bubbles cavitate, the kinetic energy disperses the clot, both from its periphery, and due to the fact that bubbles are able to penetrate the clot from within. 236 Ungeretal. a thrombus. Ultrasound could cause cavitation of the bubbles, transferring their dispersive energy to the clot and dispersing the clot safely and painlessly. Particle sizing studies of the effluent from in vitro studies of SMB-assisted sonothrombolysis have shown that the particles are submicron in size.37 The figures below show the experimental set-up used in our lab for a flow through phantom for testing sonothrombolysis, and then treatment of a clot in the phantom. The clot was exposed to 1 MHz ultrasound and tissue plasminogen activator (t-PA), followed by an infusion of MRX-815 microbubbles. As shown in the figures, after 40 min of treatment there is near complete resolution of the clot. The graph below shows the results from a series of clots exposed to t-PA, t-PA + ultrasound and t-PA + ultrasound + MRX-815 bubbles in our lab. Note that the greatest reduction of thrombi was in the group exposed to bubbles. , - C Fig. 12. Above a schematic of the experimental set-up: (A) the clot pre-treatment, (B) after 32 min of treatment, (C) after 40 min of treatment. The clot was 96% dissolved. 80.00 70.00 60.00 ••2 50.00 Z 40.00 2. « 30.00 s? 20.00 10,00 0.00 Saline US t-PA t-PA, SMB, SMB, US US US. t-PA Fig. 13. SMB = Bubbles. Lipid-Coated Submicron-Sized Particles as Drug Carriers 237 4. Clinical Studies Vascular thrombosis is a major cause of death in industrialized countries, responsible for myocardial infarction, stroke and peripheral arterial occlusions.38 In addition, deep vein thrombosis (DVT), which afflicts one in twenty Americans during their lifetime,39'40 may also be an application for sonothrombolysis. ImaRx completed a Phase I/II clinical trial in thrombosed dialysis grafts for the purpose of preliminary feasibility and safety for sonothrombolysis treatment of clotted grafts. Initial studies in thrombosed dialysis grafts provided a venue to evaluate the principle of sonothrombolysis in vascular thrombosis. As such, clinical trial efforts will move forward to address the treatment of stroke, peripheral arterial occlusions (PAO) and deep vein thrombosis (DVT). Below are examples shown from clinical trials for sonothrombolysis in dialysis grafts and DVT. The examples are not an indication that all sonothrombolysis treatments will have similar outcomes. Images from a venogram in a patient with DVT showed that the patient was administered bubbles via infusion catheter into the popliteal vein over a period of 1 hr, while ultrasound was applied across the skin. No thrombolytic drug such as t-PA was administered. Clinically, this particular patient had marked reduction in pain post-treatment with sonothrombolysis. Stroke is the third most common cause of death, after heart disease and cancer in North America. It incurs far more expenses than any other diseases due to its long term disability.41 In the US, stroke accounts for over $50 billion each year to the health care system.42 The only approved pharmacologic therapy to help restore blood flow in stroke patients is t-PA (Activase®). Less than 5% of patients are treated with t-PA due to concerns over bleeding and the risk relative to the benefit.43 Encouraging results have been obtained, however, in human studies with ultrasound and t-PA, and most recently, with ultrasound + t-PA + microbubbles. Fig. 14. The j:iyio.i a:n on the left is of a clotted dialysis graft. Very little contrast enters the graft as it is filled with clot. The image on the right, post-bubble treatment, shows complete opacification of the graft due to successful dissolution of thrombosis by sonothrombolysis. 238 Ungeretal. Fig. 15. On the pre-treatment image (left), there is complete occlusion of the superficial femoral vein (SFV). Collateral veins are seen carrying the blood flow that would normally be carried by the SFV. Post-treatment, there is good flow in the SFV and much less flow is seen in the collateral vessels due to the increased flow in the SFV. Dr. Andrei Alexandrov from the University of Texas in Houston led a study of ultrasound + t-PA in acute ischemic stroke.44,45 In this study, 126 patients were randomized prospectively to receive either a 1 hr infusion of t-PA at a dose of 0.9 mg/kg alone, or t-PA plus 2 hrs of continuous trans-cranial Doppler (TCD) ultrasound applied through the temporal window where the skull is thinnest and most easily penetrated by ultrasound. Of the 63 patients treated with t-PA alone, there was a 13% recanalization rate of the intra-cranial circulation at 2 hrs.46,47 In the same number of patients receiving t-PA + ultrasound, there was a highly significant increase in recanalization to 38% at two hours, indicating that ultrasound-mediated therapy aided in thrombus dispersion. Dr. Carlos Molina, from Barcelona, Spain, conducted a similar study but with microbubbles.48 The addition of microbubbles enhances the cavitational nuclei with a decrease in power requirements. Dr. Molina's study demonstrated that the recanalization rate increased impressively to 55%.49 In this study, Dr. Molina administered three doses of Levovist®, a microbubble agent comprised of air-filled galactose microparticles. Dr. Molina's pioneering work has demonstrated the utility of using bubbles in conjunction with ultrasound to improve the clinical outcome of acute stroke. ImaRx is currently moving MRX-815 into stroke treatment trials. Lipid-Coated Submicron-Sized Particles as Drug Carriers 239 SMB-assisted sonothrombolysis therapy could move beyond the current clinical regimens by eliminating the thrombolytic agent. Pre-clinical trials in both canine and porcine models have been encouraging.50'51-52 Human studies will be conducted to determine if lipid-coated bubbles will improve recanalization rates in patients treated with this new ultrasound-mediated paradigm. 5. Blood Brain Barrier Poor transport into the CNS is an obstacle to effectively treat diseases including brain tumors, Alzheimer's and other neuro-degenerative diseases. There are two principal barriers to drug transport into the CNS: (a) the blood brain barrier (BBB) and (b) the ABC transporters, ABCC1 and ABCB1. Unlike the rest of the body, the capillary foot processes of the cerebral endothelial cells are tight, preventing peptides and macromolecules from leaking through to the brain.53 Although the BBB may be permeable to selected ions and small molecules, ABCB1, also known as the P-glycoprotein, acts to remove the molecules by a drug-efflux system before they enter the brain. Several different strategies have been developed to overcome these limitations.54 One approach to drug delivery to the brain is by the transient opening of the BBB. Hypertonic solutions containing mannitol, which act by shrinking the endothelial cells when co-administered with drugs, have been shown to result in enhanced cerebral drug uptake.55,56 However, to cause minimum side effects, it is essential for the therapy to be regional and localized. Recently, Hynynen et al.57 have shown that the BBB can be transiently opened using ultrasound and microbubbles (Illustrated in Fig. 16). When bubbles were administered intravenously and focused ultrasound was applied across the intact skull, the BBB could be reversibly opened, permitting passage of hydrophilic low molecular weight molecules such as gadolinium-DTPA, and macromolecules such as fluorescently labeled albumin (Fig. 17) into the CNS.58 The permeability resolved over a period of hours without damage to the neurons. Similar studies have been performed in a porcine model showing that nonfocused ultrasound with microbubbles can be used to open the BBB.59 Figure 18 shows increased dye deposition in the cerebral tissue. Introduction of microbubbles as the cavitation nucleus prior to the application of ultrasound, lowered the energy needed to open the BBB, thereby lowering the bioeffects of ultrasound.60 Using this technique, large biomolecules such as horseradish peroxidase (a 40 kDa protein) have been shown to pass through the BBB with minimal damage to the brain tissues.61 It can be envisaged that drugs (small or macromolecules) bound to the microbubbles would function as a more efficient drug delivery vehicle, since these 240 Unger et al. A / , 0 <(S / / / MlLI'JCUDblO Nii'iadrofltil B UltMSOUll'l *@ c ' < * # • « Fig. 16. Cartoon representation of hypothesized ultrasound mediated drug delivery to the brain. (A) Cerebral capillaries with tight endothelial junctions prevent passage of molecules (including microbubbles and nanoparticles) into the brain. (B) Ultrasound is applied to the skull through the temporal window where the skull is thinnest (inset), cavitating the microbubbles and opening up the endothelial junctions. (C) Therapeutic agents may now pass through the opened junctions. Location 1 2 3 4 Pressure amplitude values 4.7 MPa 2.3 MPa 3.3 MPa 1.0 MPa Fig. 17. Tl-weighted MR images of rabbit brain after treatment shows contrast enhancement at 4 locations (arrows), coronal image across focal plane. Reproduced with permission from Hynynen et ah, Radiology. would provide the cavitation nuclei and the drug payload in one entity, circumventing the co-administration of drug and microbubble. In such instances, the drug could be (a) bound to the lipid membrane (hydrophobic drugs), (b) bound to the charged lipids on the surface (gene delivery), or (c) buried in the interior in an oily layer of a droplet (hydrophobic drugs) (Fig. 19). Furthermore, (d) these drug loaded bubbles or droplets may have the potential to be targeted to a specific site in the brain by surface ligands. Lipid-Coated Submicron-Sized Particles as Drug Carriers 241 ug/g tissue 30 25 20 15 10 P=0.83 30 25 20 15 10 P=0.006 Untreated Ultrasound Untreated Ultrasound + MB Fig. 18. Control pigs and pigs treated with ultrasound alone showed no difference in Evan's blue uptake. There was a significant difference in uptake when microbubbles were used in conjunction with ultrasound. Adapted from Porter et ah,} Am Soc Echocardiogr. Fig. 19. Different ways that bubbles or droplets may be able to transport drugs. Drugs may be (a) bound or embedded in the lipid membrane, (b) bound to the surface charges of the phospholipid membrance (c) buried in the oil in a droplet (d) targeting ligands can be incorporated onto the membrance. This technology of activation with ultrasound and microbubbles has the potential to also be used in the drug discovery process. By exposing cultured neurons to drugs, ultrasound and bubbles, high concentrations of the drug may be able to deliver to the cells without damaging them. This can potentially be used to screen neurons for new therapeutic compounds. 242 Unger et al. Potential CNS diseases amenable to treatment with submicron bubble delivery and classes of drugs Disease Drugs Alzheimer's Disease and other neurodegenerative disease, seizures and psychiatric disorders Primary and Secondary (metastases) Brain Tumors Stroke, brain ischemia Infection, e.g., AIDS Low molecular weight therapeutics with poor delivery to CNS, proteins, gene-based therapeutics. Low molecular weight therapeutics with poor delivery to CNS, proteins, genetic drugs. Radiation sensitizers. Cavitation nuclei to augment sonothrombolysis, either with or without use of thrombolytic agent. Delivery of oxygen with microbubbles. Improvement of cerebral perfusion with microbubble-enhanced sonication. Delivery of anti-oxidants and growth factors. Delivery of anti-infectives, anti-retrovirals to CNS. 6. Drug Delivery In the foregoing sections, we discussed activating the bubbles or using them in conjunction with ultrasound-mediated processes (e.g. microbubble mediated sonothrombolysis to enhance the local activity of the drug such as t-PA), or that the availability of a drug may be increased, e.g. by opening the blood brain barrier. In this section, we will discuss evaluating drug-carrying microbubbles for drug delivery. 6.1. Targeted bubbles As preliminary studies to demonstrate feasibility of using targeted bubbles as potential drug delivery agents, two different targeted bubbles were prepared using a mixture of DPPC, DPPE-PEG5000 and DPPA, as well as different oils and perfluorocarbons using a mixture of DDFP and n-perfluorohexane. In one study, a bioconjugate ligand targeted to the am, An integrin was synthesized by solid phase peptide methodology.62 Briefly, the bioconjugate, lipids, biocompatible drug, perfluoropropane were combined into a mixture and bubbles prepared by shaking the vials at approximately 4200 rpm. The size of the targeted bubbles Lipid-Coated Submicron-Sized Particles as Drug Carriers 243 Fig. 20. Intravital microscopy demonstrating adherence of targeted microbubble to thrombus. Picture on the right is a graphic representation outlining the location of bound microbubbles on thrombus. Reproduced with permission from Schumann et ah, Investi Radiol. was approximately 2 fim, as measured by light obscuration measurements on a Particle Sizing Systems Model 470 sizer (Particle Sizing Systems, Santa Barbara, Calif.). Bubbles were injected into a mouse model where thrombi were previously formed in the cremasteric arterioles and venules. Fluorescent imaging revealed binding of the targeted bubbles to the thrombi in both arterioles and venules. Figure 20 demonstrates the utility of a targeted bubble. Similarly, targeted bubbles were used in a HUVEC cell culture model. Briefly, bubbles with a targeting ligand directed to a^ft receptors on HUVEC cells were (a) O 1 /xm), which allows them to be administered intravenously without any risk of embolization. According to the process and the composition used in the preparation of nanoparticles, nanospheres 255 256 Gref & Couvreur Fig. 1. (A) Schematic representation of the nanocapsule structure; (B) Morphological appearance of a nanocapsule with an oily core (transmission electron microscopy after freeze fracture). or nanocapsules can be obtained. Nanospheres are matrix systems in which the drug is dispersed within the polymer throughout the particle. Contrarily, nanocapsules are vesicular or "reservoir" (heterogenous) systems, in which the drug is essentially confined to a cavity surrounded by a tiny polymeric membrane (Fig. 1). As in the case of nanospheres, depending on their physicochemical properties and composition, the drug may adsorb onto the surface as well as being included in the central core of nanocapsules. Therefore, drug localization is an important parameter in the characterization of nanocapsule preparations. The nanocapsule core may be acqueous or composed of a lipophilic solvent, usually an oil. In order to achieve good drug loading, the core materials are chosen among the good solvents for the drug.1 Expected advantages of confining the drug within a central cavity are: (a) burst effect may be avoided; (b) the drug is not in direct contact with tissues and therefore irritation at the site of administration could be reduced, and (c) the drug may be better protected from degradation both during storage and after administration. One of the advantages of nanocapsules over nanospheres is their low polymer content and a high loading capacity for lipophilic drugs. Nanocapsules can either be obtained by interfacial polymerization of monomers or from preformed polymers. In the former, the molar mass of the coating polymer will depend on the preparation conditions and even on the drug used, whereas in the latter, it is determined at the outset. Polymerization of monomers may lead to a covalent linkage between the polymer and the drug. To date, all the methodologies described for preparing nanocapsules involve the preparation of emulsions. Oil-in-water (O/W) emulsions lead to the formation of nanocapsules with an oily core, suspended in water. Water-in-oil (W/O) emulsions lead to the Nanocapsules: Preparation, Characterization and Therapeutic Applications 257 obtention of nanocapsules with an acqueous core, suspended in oil. More recently, nanocapsules with an acqueous core suspended in an acqueous medium were also obtained. Nanocapsule technology and their pharmaceutical applications will be further discussed according to the method of obtaining the polymeric wall (polymerization in situ or preformed polymer) and whether the core is acqueous or oily. 2. Preparation 2.1. Nanocapsules obtained by interfacial polymerization The advantage of obtaining nanocapsules by interfacial polymerization is that the polymer is formed in situ, allowing the polymer membrane to follow the contours of the inner phase of an O/W or W/O emulsion, thus entrapping drugs with high loadings. However, because reactive monomers are used, unwanted chemical reactions may occur between the drug and the monomer, before or during the polymerization process. The preparation of nanocapsules by polymerization requires a fast polymerization of the monomers at the interface between the organic and the acqueous phase of the emulsions. Alkylcyanoacrylates, which polymerize within seconds, have been proposed for the preparation of both oil- and water-containing nanocapsules. Their polymerization is initiated by hydroxyl ions either from the equilibrium dissociation of water or by nucleophilic groups of any compound in the polymerization medium.2 2.1.1. Oil-containing nanocapsules The oil-containing nanocapsules are suitable for the encapsulation of the lipophilic and oil-soluble compounds. They are generally obtained by interfacial polymerization of alkylcyanoacrylates, after preparing a very fine oil-in-water emulsion with an additional water-miscible organic solvent such as ethanol or acetone.3'4 These solvents serve as vehicles for the monomers, and also help to disperse the oil as very small droplets in the acqueous phase, which contains a hydrophilic surfactant. Indeed, as pointed out by Gallardo et al.,5 the organic solvents must be completely water-miscible, so that the formation of small enough oil droplets occurs spontaneously, while the solvent is diffusing towards the acqueous phase and the water is diffusing toward the organic phase. Meanwhile, the polymerization of the monomer induced by the contact with hydroxyl ions from the water phase must be swift to allow efficient formation of the polymer envelope around the oil droplet, thus achieving effective encapsulation of drugs. Generally, particles with 258 Gref & Couvreur sizes ranging between 250 and 300 nm, depending on the experimental conditions, were obtained.5,6 In a general procedure of nanocapsule preparation, the oil, the monomer, and the biologically active compound are dissolved together in the water-miscible organic solvent to prepare the organic phase.3-9 This organic phase is then injected via a cannula, under strong stirring, into the acqueous phase containing water and a hydrophilic surfactant. The nanocapsules are formed to give a milky suspension immediately. The organic phase is then removed under reduced pressure and the nanocapsules are purified by ultracentrifugation. Depending on the density of the oil forming the core, nanocapsules will concentrate either as a pellet at the bottom of the ultracentrifuge tubes or as a floating layer at the top of the tubes. A wide range of oils is suitable for the preparation of nanocapsules, including vegetable or mineral oils and pure compounds such as ethyl oleate and benzyl benzoate. The criteria for selection are the absence of toxicity, lack of affinity for the coating polymer, the absence of risk of degradation of the polymer, and a high capacity to dissolve the drug that is entrapped. Generally, Miglyol® is used to form the core of the nanocapsules.3-7,9,10 Lipiodol® and benzyl benzoate have also been successfully used to form nanocapsules.4 Soluble surfactants were chosen among Poloxamers,3-9 Triton X1009 and Tween 80.9 In some cases, nanospheres formation together with nanocapsules were observed. Aprotic, fully water-soluble solvents such as acetone and acetonitrile lead to high-quality nanocapsule preparations, whereas protic water-miscible solvents including ethanol, n-butanol, and isopropanol promoted the formation of nanospheres during nanocapsule preparation.5,9 It has been hypothesized that alcohols potentially initiate the polymerization reaction of alkylcyanoacrylates to form polymer nuclei or preformed polymers that may precipitate as nanospheres, when the organic phase is added to the acqueous phase.5 Lowering the pH in the organic phase was shown to inhibit polymerization in this medium.6 Oil-containing nanocapsules have been used to encapsulate several types of biologically active compounds including both lipophilic molecules such as carbamazepine, indomethacin, lomustine, ethosuccimide, phenytoin,1,10-14 and hydrophilic drugs such as peptides.15-18 The lipophilic drugs were solubilized in the organic phase and were encapsulated during the preparation of the nanocapsules, usually using ethanol as the water-miscible organic solvent.4,17 The encapsulation efficiency of lipophilic drugs was found to be related to their solubility in the encapsulated oil.1 Quite surprisingly, hydrophilic compounds such as peptides have also been successfully encapsulated in oil-containing nanocapsules. Indeed, these highly water-soluble compounds do not tend to dissolve in oil. It has been suggested that the extremely rapid polymerization of the alkylcyanoacrylate occurring at the surface of the oil droplet limits the diffusion of the peptide towards the acqueous Nanocapsules: Preparation, Characterization and Therapeutic Applications 259 phase, therefore leading to its entrapment in nanocapsules.15 Another explanation is that surfactants may form inverse micelles in the oily phase, allowing some dissolution of hydrophilic compounds in this phase. Interestingly, in contrast to what has been observed with poly(alkylcyanoacrylate) nanospheres,19 peptides do not react chemically with the alkylcyanoacrylate monomer during the preparation of nanocapsules when ethanol is used. The presence of a large excess of alcohol seems to prevent the hydroxyl and amino groups of the peptides from reacting with the monomer, thus retaining the biological activity of the entrapped peptides.16-18,20'21 For example, encapsulated insulin was still recognized by the insulin receptor of hepatocytes after nanoencapsulation.15,22 2.1.2. Nanocapsules containing an acqueous core Nanocapsules with an acqueous core are a recent technology developed for the efficient encapsulation of water-soluble compounds, which are generally difficult to include within nanospheres. They were obtained by interfacial polymerization, where the alkylcyanoacrylates monomers were added to a W/O emulsion.23 Anionic polymerization of the cyanoacrylate in the oily phase was initiated at the interface by nucleophiles such as hydroxyl ions in the acqueous phase, leading to the formation of nanocapsules with an acqueous core. In a typical procedure (Fig. 2), an acqueous phase at pH 7.4, consisted of ethanol and water, was prepared.23 This solution was emulsified in an organic phase containing Miglyol® and Montane® 80. The slow addition (4hrs) of the isobutylcyanoacrylate monomer in the organic phase under mechanical stirring allowed the polymerization to occur. This typical procedure leads to water droplets that are surrounded by a polymer core. The * " \ 9^ ' = Monomer CH2=CH COOR oo OILY PHASE Fig. 2. Schematic representation of the interfacial polymerization of cyanoacrylic monomers leading to the formation of nanocapsules with an acqueous core. 260 Gref & Couvreur resuspension of the nanocapsules with a mean diameter approximately 350 nm in a water phase has been achieved by the ultracentrifugation of the oily suspension, with an excess of demineralized water containing a surfactant. After removal of the upper oily phase, the nanocapsules pellet was resuspended in water. These nanocapsules are very useful for the encapsulation of hydrophilic compounds such as oligonucleotides and peptides. In this case, these macromolecules are dissolved in the acqueous phase before the interfacial polymerization process takes place. For example, encapsulation efficiencies of 50% with an oligothymidylate (phosphodiester) and of 81% with a full phosphorothioate oligonucleotide (directed against EWS Fli-chimeric RNA) were obtained.23'24 These entrapment differences were attributed to possible interactions of the oligonucleotides with the oily phase, Montane® 80, or to the possible location of the oligonucleotide at the water-oil interface which could become saturated.24 The localization of the oligonucleotide (within the acqueous core or adsorbed on the surface) has been investigated through fluorescence quenching experiments using fluorescein-labeled oligonucleotide and potassium iodine as an external quencher.23 It has been shown that fluorescent oligonucleotides were located in the acqueous core of the nanocapsules, surrounded by a polymeric wall, inaccessible to the quencher. On the contrary, when the fluorescent-oligonucleotides were free in solution, the fluorophores were highly accessible and strong quenching occurred. Similar quenching could be obtained with nanoencapsulated oligonucleotides only after the hydrolysis of the polymer wall, thus releasing the oligonucleotides. Zeta potential experiments have confirmed the localization of oligonucleotide in the acqueous core of the capsule.25 Moreover, nanoencapsulated oligonucleotides were protected against degradation by serum nucleases.25,26 Phosphorothioate oligonucleotides directed against EWS Fli-1 chimeric RNA encapsulated within poly(alkylcyanoacrylate) nanocapsules were tested in vivo for their efficacy against the experimental Ewing sarcoma in mice after intratumoral administration.24 Intratumoral injection of antisense-loaded nanocapsules led to a significant inhibition of tumor growth, whereas no antisense effect could be detected with the free oligonucleotide. These results were explained on the basis of a good protection of the oligonucleotide in the nanocapsules, which may act as a controled release system of oligonucleotide within the tumor. Salmon calcitonin was also successfully entrapped within poly (butylcyanoacrylate) nanocapsules of 300 nm in diameter.27 When the diameter was reduced to 50 nm, the encapsulation efficiencies decreased from 50 to 30%. After storage at room temperature or at 4°C, the nanocapsules retained their size for at least 34 months. The encapsulated calcitonin remained stable at 4°C for one year. Polyalkylcyanoacrylate nanocapsules were also prepared by interfacial polymerization, using a microemulsion instead of an emulsion as the template. Nanocapsules: Preparation, Characterization and Therapeutic Applications 261 Microemulsions are spontaneously forming, thermodynamically stable dispersed systems having a uniform droplet size of less than 200 nm. As such, they represent an interesting system that may be exploited for the preparation of nanocapsules too. Practically, a pseudo-ternary phase diagram of a mixture of medium chain glycerides (caprylic/capric triglycerides and mono-, diglycerides), a mixture of surfactants (polysorbate 80 and sorbitan monooleate) and water was constructed. Microemulsion domains were characterized by conductivity and viscosity to select systems suitable for the interfacial polymerization of ethyl-2-cyanoacrylate. Nanocapsules of 150 nm were obtained in those conditions and they were found to be able to encapsulate significant amounts of insulin.28 Size of the capsules may be controled, depending on different formulation variables.29 Factors influencing the encapsulation of hydrophilic compounds have been identified too.30 2.2. Nanocapsules obtained from preformed polymers The preparation of nanocapsules from preformed polymers avoids some drawbacks of the interfacial polymerization process, such as the lack of control of the polymer molar masses and polydispersity, the presence of residual monomer in the preparation, and the possibility of drug inactivation.31 An interfacial deposition process to prepare nanocapsules, also known as nanoprecipitation, has been developed.32,33 In this simple and reproducible method, a water-miscible organic phase such as an alcohol or a ketone containing oil (with or without lipophilic surfactant) is mixed with an acqueous phase containing a hydrophilic surfactant. The preformed polymer, insoluble in both the oily and the acqueous phase, is solubilized in the organic phase. After the addition of the organic phase to the acqueous phase, the polymer diffuses with the organic solvent towards the acqueous phase and is stranded at the interface between oil and water. The driving force for nanocapsule formation is the rapid diffusion of the organic solvent in the acqueous phase, inducing interfacial nanoprecipitation of the polymer surrounding the droplets of the oily phase. Synthetic polymers such as poly(D,Llactide), poly(e-caprolactone) and poly(alkylcyanoacrylate) are most frequently employed for nanocapsule formation.32 Arabic gum, gelatin, ethylcellulose or hydroxypropylmethylcellulose phthalate were also successfully used.32 The size of nanocapsules is usually found between 100 and 500 nm, and it depends on several factors, namely, the chemical nature and the concentration of the polymer and the encapsulated drug, the amount of surfactants, the ratio of organic solvent to water, the concentration of oil in the organic solution, and the speed of diffusion of the organic phase in the acqueous phase. In general, the lower the interfacial tension and the viscosity of the oil, the smaller the nanocapsules are formed.34 262 Gref & Couvreur Both lipophilic and hydrophilic surfactants are used in the preparation of nanocapsules by this technique. However, not all the surfactants that are technically suitable are acceptable for parenteral administration; as such, the choice has to be made with the administration route in mind. Generally, the lipophilic surfactant is a natural lecithin of relatively low phosphatidylcholine content, whereas the hydrophilic one is ionic (i.e. lauryl sulphate, quaternary ammonium), or more commonly nonionic (i.e. poly(oxyethylene)-poly(propropylene) glycol). Poly(ethylene glycol)-coated nanocapsules were also prepared by nanoprecipitation, using preformed diblock poly(lactide)-poly(ethylene glycol) copolymers or blends of these copolymers with the homopolymer poly(lactide.)35-38 However, the most physically stable nanocapsules were those prepared with poly(lactide)-poly(ethylene glycol) copolymer alone. RU 58668, a promising pure antiestrogen, was entrapped into poly(ethylene glycol)-coated nanospheres and into nanocapsules with a similar coating.37 A series of preformed diblock polyesterpolyethylene glycol) copolymers were used for the design of these nanoparticles, both the molar masses of the poly(ethylene glycol) blocks and the nature of the hydrophobic polyester blocks being varied. Nanospheres which had a smaller size (~110nm), compared with nanocapsules (~250nm), were however able to incorporate larger amounts of the antioestrogen than the nanocapsules counterpart. In an alternative method named solvent displacement method, an O/W emulsion was formed.39 The organic phase contained the polymer, the oil and the drug, and the acqueous solution contained a stabilizing agent. In this procedure, the organic solvent was displaced into the external phase by the addition of an excess of water. This technique has several advantages such as the small quantities of solvents used, the good control of the size of the nanocapsules (80-900 nm), and the control of the thickness of the polymeric wall by monitoring the polymer concentrations.40 However, large amounts of water have to be removed at the end of the process. Two formulation processes which bring lipids into play should also be mentioned. The first methodology is based on the inversion phase of an emulsion to prepare original lipidic nanocapsules. These capsules, interestingly obtained as a suspension in saline water, were constituted by medium chain triglycerides and hydrophilic /lipophilic surfactants. According to the authors, the formulation method has been developed to avoid the use of organic solvent or the high quantity of surfactants and co-surfactants, due to the potential toxicity of their residues after human administration. Their original structure was found to be a hybrid between polymeric nanocapsules and liposomes as their oily core is being surrounded by a tensioactive rigid membrane.41-43 Nanocapsules: Preparation, Characterization and Therapeutic Applications 263 In another process, cisplatin lipid-based nanocapsules have been prepared by the repeated freezing and thawing of an equimolar dispersion of phosphatidylserine (PS) and phosphatidylcholine (PC) in a concentrated acqueous solution of cisplatin. Here, the molecular architecture of these novel nanostructures was elucidated by solid-state NMR techniques.15N NMR and 2H NMR spectra of nanocapsules containing 15N- and 2H-labeled cisplatin respectively, demonstrated that the core of the nanocapsules consists of solid cisplatin devoid of free water. Magicangle spinning 15N NMR showed that approximately 90% of the cisplatin in the core is present as the dichloro species. The remaining 10% was accounted for by a newly discovered dinuclear Pt compound that was identified as the positively charged chloride-bridged dimer of cisplatin. NMR techniques, sensitive to lipid organization 31P NMR and 2H NMR, revealed that the cisplatin core is coated by phospholipids in a bilayer configuration and that the interaction between solid core and bilayer coat exerts a strong ordering effect on the phospholipid molecules. Compared with phospholipids in liposomal membranes, the motion of the phospholipid headgroups is restricted and the ordering of the acyl chains is increased, particularly in PS.44 Analysis of the mechanism of the nanocapsule formation suggests that the method may be generalized to include other drugs showing low water solubility and lipophilicity.45 3. Characterization Size evaluation of nanocapsules is most frequently done by photon correlation spectroscopy, transmission electron microscopy, and scanning electron microscopy, without or after freeze-fracture.33,39,46 At present, transmission electron microscopy performed after freeze-fracture has given the most useful information about nanocapsule structure, highlighting the polymer envelope and the inner cavity, and allowing the wall thickness to be estimated.1'7,47 Thus, polymer coatings were estimated to be around 5 ran, depending on the monomer concentration.47 Freezefracture (Fig. 1) has also allowed the visualization of different possible organizations of lipophilic surfactant, which can form vesicles, micelles, bilayers, or monolayers, depending on its concentration.33 The spherical shape of the nanocapsules was confirmed by atomic force microscopy.39 Most images of nanocapsules have been obtained by transmission electron microscopy performed on negatively stained preparations, allowing to gain information about nanocapsule morphology and integrity1,47 (Fig. 3A). Nanocapsules embedded in a suitable resin were cut into thin slices.48 They were observed using electron microscopy, the contrast being created by encapsulation of a colloidal gold-labeled molecule during nanocapsule preparation. In this manner, both polymer envelope and the internal cavity were distinguished easily (Fig. 3B). 264 Gref& Couvreur 50nm 100 nm 100 nm B 3 l Fig. 3. (A) Morphological appearance of polydactic acid-co-glycolic) nanocapsules using the transmission electron microscopy. (B) Labeling insulin with gold allows to distinguish the localization of this molecule into the internal core of poly(isobutyl cyanoacrylate) nanocapsules; Transmission Electron Microscopy. Zeta potential measurements are also very useful for the chraracterization of the nanocapsules. Surfactants and polymer are the major components that can affect this parameter. Many polymers such as poly (D,Llactide), poly(e-caprolactone) and lecithins impart a negative charge to the surface, whereas nonionic surfactants such as Poloxamer tend to reduce the absolute value of zeta potential.34 Calvo et alP described nanocapsules coated with positively charged polysaccharide chitosan. Their surface charge depended mainly on the viscosity of the chitosan solution used for coating. Positive values up to 46 mV were also observed with diethylaminoethyldextran coated nanocapsules.8 Generally, Zeta potential values above 30 mV (positive or negative values) lead to more stable nanocapsule suspensions, because repulsion between the particles prevented their aggregation. In contrast to observations with nanospheres, the negative Zeta potential of the nanocapsules was not completely masked by the presence of neutral poly(ethylene glycol) chains at the surface.63 This was due to the presence of lecithin in the polyethylene glycol) "brush", which remained necessary for nanocapsule stability. It was further highlighted that the presence of such a "brush" could reduce complement activation, an important step in the recognition of particles by macrophages.50'51 Nanocapsules: Preparation, Characterization and Therapeutic Applications 265 Centrifugation in a density gradient was used to confirm the existence of nanocapsules by comparing with the colloidal carriers prepared without polymer or oil. For example, isopycnic centrifugation in a density gradient of Percoll was used in the case of nanocapsules with a Miglyol core and a coating of poly (alky lcyanoacry late) or poly(D,L lactide).39 The density of the nanocapsules was found to be intermediate between that of nanospheres and that of emulsions. These studies also demonstrated that the density of nanocapsules and the band thickness increased when the quantity of polymer increased. No contamination of nanocapsules with nanospheres was observed. However, Mosqueira et al.3i performed similar experiments and observed that nanocapsule preparations obtained by nanoprecipitation contained small amounts of nanospheres, as it has previously been described by Gallardo et al.5 for nanocapsules prepared by interfacial polymerization. When lecithin was present in excess as lipophilic surfactant, liposomes were also detected in the nanocapsule preparations. Liposomes could not be distinguished from nanocapsules on the basis of density differences, but have been detected by electron microscopy52 and by the encapsulation of an acqueous tracer.34 4. Drug Release Release of encapsulated drugs from nanocapsules made of preformed polymers, appears only to be controled by the partition coefficient of the drug between the oily core and the acqueous external medium, and the relative volumes of these two phases. Except for macromolecules, the rate of diffusion of the drug through the thin polymeric coating does not seem to be a limiting factor, nor does the nature of the polymeric wall. This clearly suggests that the polymer membrane may be porous rather than a continuous film barrier to diffusional release. The nature of the external acqueous phase is of prime importance in the release. For example, indomethacin release was faster and more complete in the presence of albumin, which acts as an acceptor in the acqueous phase.11,52 Similarly, release of halofantrine, a highly lipophilic drug, was only observed in the presence of serum, because the drug has a high affinity for lipoproteins.36 The presence of a hydrophilic poly(ethylene glycol) "brush" at the nanocapsule surface was also shown to play a role in drug release. Release of halofantrine and primaquine from such surface-modified nanocapsules was reduced, compared with conventional nanocapsules.36,53 In conclusion, it may be considered a challenge to develop nanocapsule systems with release profiles, which may be controled not only by the partitioning coefficient, but also by the nature or morphology (i.e. thickness or porosity) of the surrounding membrane. 266 Gref & Couvreur 5. Applications Nanocapsules have been proposed as drug delivery systems for several drugs by different routes of administration such as oral, ocular or parenteral. Drug-loaded nanocapsules were used to improve the stability of the drug either in biological fluids, or simply in the formulation. Another goal was to reduce the toxicity of some drugs known for their undesirable side effects. 5.1. Oral route Challenging aspects related to oral administration deal with the entrapment of unstable molecules, such as peptides or that of anti-inflammatory compounds that cause local side effects on the mucosae. Pioneering studies in the mid 1980s dealt with indomethacin and insulin entrapment. Indomethacin, an anti inflammatory drug, has been successfully encapsulated in the polyalkylcyanoacrylate nanocapsules with the aim of reducing its side effects on the gastric and intestinal mucosa.11 The drug retained its biological activity after nanoencapsulation. Moreover, nanoencapsulated formulations allowed a dramatic reduction of the ulcerative side effects usually induced by indomethacin on the mucosae.54 This protection was attributed to the combined effect of the sustained release of indomethacin from the nanocapsules, with a significant reduction of the direct contact between drug and the mucosae. In the case of nanocapsules obtained by nanoprecipitation using polyesters, the release kinetics in media mimicking pH of the gut were more sensitive to changes in drug partitioning related to the change of pH, than to the type of polymer used.55,56 Drug release from nanocapsules was accelerated in the presence of digestive enzymes such as proteases and esterases. This was correlated with a decrease in polymer molecular weight.55'56 Diclofenac and indomethacin, two major nonsteroidal anti inflammatory agents, have been encapsulated in polyQactic acid) nanocapsules obtained by nanoprecipitation, with the aim of reducing their side effects on the gastric mucosa.54,57,58 The side effects of both drugs were completely modified and reduced by the encapsulation in nanocapsules.54 As in the case of nanocapsules produced by interfacial polymerization, a marked protective effect on the gastrointestinal mucosa, as compared with the ulcerative effect observed with the drug solutions, was observed. Insulin-loaded nanocapsules yielded promising pharmacological results.16,21 When given orally to diabetic rats and dogs, single administration produced a reduction in glycemia after an unusually long lag of several days, and this hypoglycemia was sustained for up to 20 days.16,20,21,59 It was suggested that nanocapsules could release insulin slowly from a depot within the body. The nanocapsules seemed to be involved in carrying the insulin near the intestinal Nanocapsules: Preparation, Characterization and Therapeutic Applications 267 epithelium where they were absorbed and translocated as intact nanocapsules to the blood vessels.48,59-61 However, Lowe and Temple16 reported that insulin adsorption from orally administered nanocapsules reached a maximum of absorption, 15 min after administration and any trace of insulin in blood was detected after a few hours. Sai et al.62 have proposed the use of insulin-loaded nanocapsules as a new prophylactic tool to prevent diabetes. They showed in a model of non-obese diabetic mice that prophylactic injection of such nanocapsules reduced the incidence of diabetes. Anti infectious agents such as atovaquone and rifabutin, two compounds active against the opportunistic parasite Toxoplasma gondii, were successfully entrapped in poly(lactide) nanocapsules formed by nanoprecipitation. These drugs have a poor bioavailability because of their insolubility in water. Nanoencapsulation is allowed to decrease in the brain parasitic burden in a higher extent than the same amount of free drug.63 Chitosan-coated nanocapsules were particularly interesting for oral administration, probably because their positive charge allow them to stick efficiently along the gastro-intestinal mucosa, with a further possible diffusion through the epithelium , thus providing a continuous drug delivery into the blood stream.64,65 When the peptide salmon calcitonin was entrapped into these nanocapsules, long-lasting hypocalcemia effects were observed, following oral administration to rats.66 In contrast, calcitonin control emulsions led to negligible responses. 5.2. Parenteral route As far as the parenteral route is concerned, nanocapsules could be useful for the formulation of poorly soluble drugs, and for controling the drug biodistribution according to the properties of the carrier. In this view, indomethacin and diclofenac were entrapped in nanocapsules, but diclofenac in solution or in nanocapsules showed similar plasma concentration profiles. After intravenous administration, encapsulated indomethacin showed even lower plasma concentrations than the free drug because of enhanced hepatic uptake of loaded nanocapsules.57 One possible explanation for the absence of the modification of the pharmacokinetics and biodistribution profiles of the encapsulated drugs probably results from the rapid rate of release of these drugs into the circulation, due to the high blood dilution and/or the presence of plasma proteins. Subcutaneous injection did not lead to a slow release of the drug either. Nevertheless, after intramuscular administration, the nanocapsules containing diclofenac showed a significantly reduced inflammation at the site of injection, compared with the free drug in solution.67 Similarly, darodipine nanocapsules provided a prolonged antihypertensive effect compared with free drug which lasted for at least 24 hrs.68 268 Gref & Couvreur Nanocapsules prepared by interf acial polymerization of the isobutylcyanoacrylate monomers were retained longer at the injection site after intramuscular administration than the other types of carriers such as emulsions or liposomes.69 Moreover, they were taken up to a significant extent by the regional lymph nodes, likely owing to the phagocytosis by macrophages. These observations open up the possibility of delivering cytostatic drugs and immunomodulators to the lymph node metastases. When administered intravenously, nanocapsules made by interfacial polymerization or by nanoprecipitation were taken up rapidly by organs of the mononuclear phagocyte system, mainly the liver.70 To take advantage of this particular tissue distribution, nanocapsules containing muramyltripeptide cholesterol (MTPChol) were designed.71,72 This immunostimulating agent, able to activate the macrophages and to stimulate their innate defense functions against tumor cells, is a useful agent to treat metastatic cancer. In vitro studies with rat alveolar macrophages have shown that nanocapsules prepared from poly(D,Llactic acid) containing MTPChol were more efficient activators than the free drug. This was attributed to the intracellular delivery of the nanoencapsulated immunomodulator after cell phagocytosis; an intermediate transfer of the drug to serum proteins was another suggested mechanism.73 In vivo, this type of nanocapsules is allowed to obtain significant antimetastatic effects in a model of liver metastases.74 For other types of applications, to avoid the rapid clearance by the mononuclear phagocyte system, nanocapsules coated with poly(ethylene glycol) with a molar mass of 20,000 g/mole were developed. An antimalarial drug, halofantrine, was entrapped with the aim of obtaining a well-tolerated injectable form for the treatment of this severe intravascular disease.36 In mice, at an advanced stage of infection with Plasmodium berghei, the area under the curve for plasma halofantrine was increased six-fold, compared with the free drug when the molecule was presented as nanocapsules. Moreover, the toxicity of halofantrine was reduced by incorporation into the nanocapsules. Up to 100 mg/kg could be administered intravenously without toxicity, yet all mice injected with this dose of free halofantrine died instantaneously. However, in vivo, only small differences were observed in terms of the therapeutic activity between poly(ethylene glycol) coated nanocapsules and the uncoated ones. This was explained by the possible saturation of the phagocytic capacity of the liver in severely infected mice, as a result of the uptake of parasitized erythrocytes.75 Moreover, it was emphasized that the amount of serum lipoproteins, which acted as acceptors for halofantrine released from nanocapsules, is reduced during the disease. Poly(ethylene glycol) coated nanocapsules were also used to deliver lipophilic drugs to the solid tumors. In this case, the vascular endothelium is known to be more permeable, thus allowing the extravasation of small-sized colloidal particles. This specific distribution of colloids into tumoral sites is known as the enhanced Nanocapsules: Preparation, Characterization and Therapeutic Applications 269 permeability and retention effect (EPR effect). The efficacy of this strategy has been demonstrated using a photosensitizer, meta-tetra (hydroxyphenyl) chlorine, encapsulated in nanocapsules designed from diblock poly(D,L lactide)-poly(ethylene glycol) copolymers.35 5.3. Ocular delivery The major problems encountered when delivering drugs to the eyes are the poor permeability of the corneal epithelium and the rapid clearance because of tear turnover and lacrimal drainage. Nanocapsule formulations were developed with the aim of improving drug efficacy by retaining it at the level of the ocular tissue, thus reducing the number of administrations.7677 Betaxolol-loaded poly(isobutylcyanoacrylate) nanocapsules made by interracial polymerization were prepared for the treatment of glaucoma. Only a marginal decrease in the intraocular pressure was observed with this type of formulation, compared with the activity obtained with the commercial form (single solution) or by other carriers.13 More promising results have been obtained with pilocarpine.14 In this case, sustained drug release was obtained when incorporating the pilocarpine loaded nanocapsules into a Pluronic gel. Thus, a significant increase in the bioavailability of the drug was achieved. Ganciclovir is an antiviral drug used for the treatment of cytomegalovirus infections. In the clinical practice, two to three intravitreal injections per week are needed to overcome the rapid clearance of the drug from the eyes. Ganciclovir encapsulation in poly(ethylcyanoacrylate) nanocapsules made by interfacial polymerization provided a sustained release of the drug over four days.10 Moreover, after intravitreal injection of the nanocapsules, the drug could still be detected in the eyes at a therapeutic level after ten days. Significant amounts of ganciclovir were found in the retina and in the vitreous humor which is considered as beneficial in the treatment of cytomegalovirus retinitis. On the contrary, after administration of single solutions of the drug free, the maximum concentration of ganciclovir was reached in less than one day and no drug could be detected later. However, despite these beneficial results, some toxicity (opacification of the lens and vitreous humor turbidity) was found as a result of the nanocapsules. Antiglaucomatous agents such as carteolol and betaxolol were also encapsulated in nanocapsules prepared from preformed polymers, but they only showed a reduction of the noncorneal absorption (systemic circulation), leading to lesser side effects as compared with the free drug.13'78'79 Encapsulation in nanocapsules produced an improved pharmacological effect characterized by a more important reduction of the intraocular pressure, compared with the free drug treatment, as well as with the same treatment but delivered by nanospheres; reduced 270 Gref & Couvreur cardiovascular systemic side effects were also observed with the nanocapsules. ' In the case of betaxolol, the nature of the polymer making up the nanocapsule wall was found to play a major role in the pharmacological responses.78'80 Thus, poly(e-caprolactone) walls were more efficient than poly(isobutylcyanocrylate) or poly(lactide-co-glycolide) ones. Indeed, as shown by the confocal microscopy, poly(e-caprolactone) nanocapsules could specifically penetrate the corneal epithelium by an endocytic process, without causing any damage to the cells. In contrast, poly(isobutylcyanoacrylate) nanoparticles produced a cellular lysis.81 As no differences in penetration were observed between nanospheres and nanocapsules, the presence of an oily core did not seem to influence activity of the formulation. Coating the negatively charged surface of poly(e-caprolactone) nanocapsules with chitosan, a cationic polymer, provided the best corneal drug penetration, together with preventing the degradation caused by the adsorption of lysozyme, a positively charged enzyme found in tear fluid.82 This was explained by the higher penetration of the nanocapsules into the corneal epithelial cells and by the mucoadhesion of these positively charged particles onto the negatively charged membranes. Additionally, a specific effect of chitosan on the tight junctions has been mentioned.83 Encouraging results were also obtained with nanocapsules containing the immunosuppressive peptide cyclosporin A.84 This drug was efficiently entrapped in poly(e-caprolactone) nanocapsules, leading to a five-fold increase of the cyclosporin A corneal concentrations, compared with an oily solution of the drug. Again, chitosan-overcoated nanocapsules were able to provide a selective and prolonged delivery of cyclosporine A to the ocular mucosae, without compromising the inner ocular tissues and avoiding systemic absorption.84 The mechanism that explains the increased ocular penetration was understood as the combination of an improved interaction with the corneal epithelium, followed by the penetration of the particles into the corneal epithelium.85 In the case of indomethacin associated with chitosane-coated nanocapsules, the use of confocal microscopy established the fact that the nanocapsules penetrated through the corneal epithelium following a transcellular pathway.85,86 6. Conclusion As discussed in this chapter, there are now various technologies for the preparation of nanocapsules. These methods which obey a wide variety of principles may either start from a monomer or from a preformed polymer. They employ macromolecular materials of synthetic or natural origin and they allow the design of nanocapsules with either an acqueous or an oily core. Thus, they can efficiently entrap almost every molecule. The most significant advantage of nanocapsules over nanospheres is that the drug to polymer ratio is generally much higher, which allows the use of Nanocapsules: Preparation, Characterization and Therapeutic Applications 271 lesser polymer to deliver the same amount of drug to the cells and tissues. This is, from a toxicological point of view, a substantial advantage of this type of technology. On the contrary, drug release from nanocapsules is mainly dependent on the partitioning coefficient of the biologically active compound between the nanocapsule core and the biological receptor medium. If the nanocapsule thin polymer membrane may be a barrier for the diffusion of macromolecules, it is not the case for small organic molecules. Thus, to control the drug release kinetic from nanocapsules, it is likely to remain the primary challenge to be resolved with this kind of technology in the next few years. References 1. Fresta M, Cavallaro G, Giammona G, Wehrli E and Puglisi G (1996) Preparation and characterization of polyethyl-2-cyanoacrylate nanocapsules containing antiepileptic drugs. Biomaterials 17:751. 2. Fattal E, Peracchia MT and Couvreur P (1997) Poly(alkylcyanoacrylate), in: Domb AJ, Kost J, Wiseman DM (eds.) Handbook of Biodegradable Polymers. Harwood Academic Publishers: Amsterdam. 3. Al Khouri N, Fessi H, Roblot-Treupel L, Devissaguet JP and Puisieux F (1986) An original procedure for preparing nanocapsules of polyalkyl-cyanoacrylates for interfacial polymerization. Pharm Acta Helv 61:274. 4. Al Khouri Fallouh N, Roblot-Treupel L, Fessi H, Devissaguet JP and Puisieux F (1986) Development of a new process for the manufacture of polyisobutylcyanoacrylate nanocapsules. Int J Pharm 28:125. 5. Gallardo MM, Couarraze G, Denizot B, Treupel L, Couvreur P and Puisieux F (1993) Preparation and purification of isohexylcyanoacrylate nanocapsules. Int ] Pharm 100:55. 6. Wohlgemuth M, Machtle W and Mayer C (2000) Improved preparation and physical studies of polybutylcyanoacrylate nanocapsules. / Microencapsulation 17:437. 7. Chouinard F, Kan FW, Leroux JC, Foucher C and Lenaerts V (1991) Preparation and purification of polyisohexylcyanoacrylate nanocapsules. Int} Pharm 72:211. 8. Chouinard F, Buczkowski S and Lenaerts V (1994) Poly(alkylcyanoacrylate) nanocapsules: Physicochemical characterization and mechanism of formation. Pharm Res 11:869. 9. Puglisi G, Fresta M, Giammona G and Venture CA (1995) Influence of the preparation conditions in poly(ethylcyanoacrylate) nanocapsules formation. Int J Pharm 125:283. 10. El-Samaligy MS, Rojanasakul Y, Charlton JF, Weinstein GW and Lim JK (1996) Ocular disposition of nanoencapsulated acyclovir and ganciclovir via intravitreal injection in rabbit's eye. Drug Del 3:93. 11. Ammoury N, Fessi H, Devissaguet JP, Puisieux F and Benita S (1989) Physicochemical characterization of polymeric nanocapsules and in vitro release evaluation of indomethacin as a drug model. STP Pharm 5:642. 12. Giirsoy A, Eroglu L, Ulutin S, Tasyiirek M, Fessi H, Puisieux F and Devissaguet JP (1989) Evaluation of indomethacin nanocapsules for their physical stability and inhibitory activity on inflammation and platelet aggregation. Int J Pharm 52:101. 272 Gref& Couvreur 13. Marchal-Heussler L, Fessi H, Devissaguet JP, Hoffman M and Maincent P (1992) Colloidal drug delivery systems for the eye. A comparison of the efficacy of three different polymers: Polyisobutylcyanoacrylate, polylactic-co-glycolic acid, poly-epsiloncaprolactone. STP Pharm Sci 2:98. 14. Desai SD and Blanchard J (2000) Pluronic® F127-based ocular delivery systems containing biodegradable polyisobutylcyanoacrylate nanocapsules of pilocarpine. Drug Del J 7:201. 15. Aboubakar M, Puisieux F, Couvreur P, Deyme M and Vauthier C (1999) Study of the mechanism of insulin encapsulation in poly(isobutylcyanoacrylate) nanocapsules obtained by interfacial polymerization. / Biomed Mater Res 47:568. 16. Damge C, Michel C, Aprahamiam M and Couvreur P (1988) New approach for oral administration of insulin with polyalkylcyanoacrylate nanocapsules as drug carrier. Diabetes 37:246. 17. Lowe PJ and Temple CS (1994) Calcitonin and insulin in isobutylcyanoacrylate nanocapsules: Protection against proteases and effect in intestinal absorption in rats. / Pharm Pharmacol 46:547. 18. Damge C, Vonderscher J, Marbach P and Pinget M (1997) Poly(alkylcyanoacrylate) nanocapsules as a delivery system in the rat for octreotide, a long-acting somatostatin analogue. / Pharm Pharmacol 49:949. 19. Damge C, Vranckx H, Baldschmidt P and Couvreur P (1997) Poly(alkylcyanoacrylate) nanospheres for oral administration of insulin. / Pharm Sci 86:1403. 20. Michel C, Aprahamiam M, Defontaine L, Couvreur P and Damge C (1991) The effect of site of administration in the gastrointestinal tract on the absorption of insulin from nanocapsules in diabetic rats. / Pharm Pharmacol 43:1. 21. Damge C, Hillaire-Buys D, Puech R, Hoelizel A, Michel C and Ribes G (1995) Effects of orally administered insulin nanocapsules to normal and diabetic dogs. Diabetes Nutr Metab 8:3. 22. Roques M, Damge C, Michel C, Staedel C, Cremel G and Hubert P (1992) Encapsulation of insulin for oral administration preserves interaction of the hormone with its receptor in vitro. Diabetes 41:451. 23. Lambert G, Fattal E, Pinto-Alphandary H, Gulik A and Couvreur P (2000) Polyisobutylcyanoacrylate nanocapsules containing an acqueous core as a novel colloidal carrier for the delivery of oligonucleotides. Pharm Res 17:707. 24. Lambert G, Bertrand JR, Fattal E, Subra F, Pinto-Alphandary H, Malvy C, Auclair C and Couvreur P (2000) EWS Fli-1 antisense nanocapsules inhibits Ewing sarcoma-related tumor in mice. BBRC 279:401. 25. Chavany C, Le Doan T, Couvreur P, Puisieux F and Helene C (1992) Polyalkylcyanoacrylate nanoparticles as polymeric carriers for antisense oligonucleotides. Pharm Res 9:441. 26. Nakada Y, Fattal E, Foulquier M and Couvreur P (1996) Pharmacokinetics and biodistribution of oligonucleotide adsorbed onto poly(isobutylcyanoacrylate) nanoparticles after intravenous administration in mice. Pharm Res 13:38. Nanocapsules: Preparation, Characterization and Therapeutic Applications 273 27. Vrankx H, Demoustier M and Deleers M (1996) A new nanocapsule formulation with hydrophilic core: Application to the oral administration of salmon calcitonin in rats. Eur J Pharm Biopharm 42:345. 28. Watnasirichaikul S, Davies NM, Rades T and Tucker IG (2000) Preparation of biodegradable insulin nanocapsules from biocompatible microemulsions. Pharm Res 17:684-689. 29. Watnasirichaikul S, Rades T, Tucker IG and Davies NM (2002) Effects of formulation variables on characteristics of poly (ethylcyanoacrylate) nanocapsules prepared from w/o microemulsions. Int J Pharm 235:237-246. 30. Pitaksuteepong T, Davies NM, Tucker IG and Rades T (2002) Factors influencing the entrapment of hydrophilic compounds in nanocapsules prepared by interfacial polymerisation of water-in-oil microemulsions. Eur } Pharm Biopharm 53:335-342. 31. Grangier JL, Puygrenier M, Gauthier JC and Couvreur P (1991) Nanoparticles as carriers for growth hormone releasing factor (GRF). / Control Rel 15:3. 32. Fessi H, Devissaguet JP and Puisieux F (1986) Procede de preparation des systemes collo'idaux dispersibles d'une substance sous forme de nanocapsules. French Patent Application No. 8618444. 33. Fessi H, Puisieux F, Devissaguet JP, Ammoury N and Benita S (1989) Nanocapsule formation by interfacial deposition following solvent displacement. Int J Pharm 55:R1-R4. 34. Mosqueira VCF, Legrand P, Pinto-Alphandary H, Puisieux F and Barratt G (2000) Poly(D,L-lactide) nanocapsules prepared by a solvent displacement process: Influence of the composition on physicochemical and structural properties. / Pharm Sci 89:614. 35. Bourdon O, Mosqueira V, Legrand P and Blais J (2000) A comparative study of the cellular uptake, localization and phototoxicity of meta-tetra(hydroxyphenyl) chlorine encapsulated in surfacemodified submicronic oil/water carriers in HT29 tumor cells. / Photochem Photobiol B 55:164. 36. Mosqueira VCF, Legrand P, Gref R and Barratt G (1999) In vitro release kinetic studies of PEG-modified nanocapsules and nanospheres loaded with a lipophilic drug: Halofantrine base. Proceedings of the 26th International Symposium on Controlled Release ofBioactive Materials, 20-23 June, Boston, MA, 1074-1075. 37. Ameller T, Marsaud V, Legrand P, Gref R and Barratt G et Renoir JM (2003) Polyester — poly(ethylene glycol) nanospheres and nanocapsules loaded with the pure antioestrogen RU 58668: Physico-chemical and opsonisation properties. Pharm Res 20(7):1063-1070. 38. Ameller T, Marsaud V, Legrand P, Gref R and Barratt G et Renoir JM (2003) Polyester — poly(ethylene glycol) nanospheres and nanocapsules loaded with the pure antioestrogen RU 58668: Physico-chemical and opsonisation properties. Pharm Res 20(7):1063-1070. 39. Quintanar-Guerrero D, Allemann E, Doelker E and Fessi H (1998) Preparation and characterization of nanocapsules from preformed polymers by a new process based on emulsification-diffusion technique. Pharm Res 15:1056. 40. Quintanar-Guerrero D, Fessi H, Doelker E and Allemann E (1997) Procede de preparation de nanocapsules de type vesiculate utilisable notamment comme vecteurs collo'idaux de principes actifs pharmaceutiques ou autres, French Patent 97.09.672. 41. Heurtault B, Saulnier P, Pech B, Proust JE and Benoit JP (2002) A novel phase inversionbased process for the preparation of lipid nanocarriers. Pharm Res 19:875-880. 274 Gref & Couvreur 42. Heurtault B, Saulnier P, Pech B, Proust JE and Benoit JP (2002) Properties of polyethyleneglycol 660 12 hydroxystearate at the triglyceride/water interface. Int J Pharm 242:176-170. 43. Heurtault B, Saulnier P, Pech B, Venier-Julienne MC, Proust JE, Phan-Tan-Luu R and Benoit JP (2003) The influence of lipid nanocapsule composition on their size distribution. Eur Pharm Sci 18:55-61. 44. Chupin V, de Kroon AI and de Kruijff B (2004) Molecular architecture of nanocapsules, bilayer-enclosed solid particles of Cisplatin. / Am Chem Soc 126:13816-13821. 45. Burger KN, Staffhorst RW, de Vijlder HC, Velinova MJ, Bomans PH, Frederik PM and de Kruijff B (2002) Nanocapsules: Lipid-coated aggregates of cisplatin with high cytotoxicity. Nat Med 81-84. 46. Magalhaes NS, Fessi H, Puisieux F, Benita S and Seiller M (1995) An in vitro release kinetic examination and comparative evaluation between submicron emulsion and polylactic acid nanocapsules of clofibride. / Microencapsul 12:195. 47. Rollot JM, Couvreur P, Roblot-Treupel L and Puisieux F (1986) Physicochemical and morphological characterization of polyisobutyl cyanoacrylate nanocapsules. / Pharm Sci 75:361. 48. Vauthier C, Dubernet C, Fattal E, Pinto-Alphandary H and Couvreur P (2003) Poly(alkylcyanoacrylates) as biodegradable materials for biomedical applications. Adv Drug Del Rev 55(4):519-548. 49. Calvo P, Remunan-Lopez C, Vila-Jato JL and Alonso MJ (1997) Development of positively charged colloidal drug carriers: Chitosan-coated polyester nanocapsules and submicronemulsions. Coll Polym Sci 275:46. 50. Mosqueira VCF, Legrand P, Gulik A, Bourdon O, Gref R, Labarre D and Barratt G (2001) Relationship between complement activation, cellular uptake and surface physicochemical aspects of novel PEG-modified nanocapsules. Biomaterials 22:2967. 51. Mosqueira VCF, Legrand P, Gref R, Heurtault B, Appel M and Barratt G (1999) Interactions between a macrophage cell line (J774A1) and surface-modified poly(D,L-lactide) nanocapsules bearing poly(ethylene glycol). / Drug Targ 7:65. 52. Ammoury N, Fessi H, Devissaguet JP, Puisieux F and Benita S (1990) In vitro release kinetic pattern of indomethacin from poly(D,L-lactide) nanocapsules. / Pharm Sci 79:763. 53. Heurtault B, Legrand P, Mosqueira V, Devissaguet JP, Barratt G and Bories C (2001) In vitro antileishmanial properties of surface modified and primaquine loaded nanocapsules in Leishmania donovani intramacrophagic amastigotes. Ann Trop Med Parasitol 95:529. 54. Ammoury N, Fessi H, Devissaguet JP, Dubrasquet M and Benita S (1991) Jejunal absorption, pharmacological activity, and pharmacokinetic evaluation of indomethacin-loaded poly(D,L-lactide) and poly(isobutyl-cyanoacrylate) nanocapsules in rats. Pharm Res 8:101. 55. Marchais H, Benali S, Irache JM, Tharasse-Bloch C, Lafont O and Orecchioni AM (1998) Entrapment efficiency and initial release of phenylbutazone from nanocapsules prepared from different polyesters. Drug Dev Ind Pharm 24:883. Nanocapsules: Preparation, Characterization and Therapeutic Applications 275 56. Kedzierewicz F, Thouvenot P, Monot I, Hoffman M and Maincent P (1998) Influence of different physicochemical conditions on the release of indium oxine from nanocapsules. J Biomed Mater Res 39:588. 57. Guterres SS, Fessi H, Barratt G, Puisieux F and Devissaguet JP (1995) Poly(D,L-lactide) nanocapsules containing non-steroidal anti-inflammatory drugs: Gastrointestinal tolerance following intravenous and oral administration. Pharm Res 12:1545. 58. Andrieu V, Fessi H, Dubrasquet M, Devissaguet JP, Puisieux F and Benita S (1989) Pharmacokinetic evaluation of indomethacin nanocapsules. Drug Des Del 4:295. 59. Damge C, Michel C, Aprahamiam M, Couvreur P and Devissaguet JP (1990) Nanocapsules as carriers for oral peptide delivery. / Control Rel 13:233. 60. Aboubakar M, Couvreur P, Pinto-Alphandary H, Gouritin B, Lacour B, Farinotti R, et al. (2000) Insulin-loaded nanocapsules for oral administration: In vitro and in vivo investigation. Drug Dev Res 49:109. 61. Aprahamiam M, Michel C, Humbert W, Devissaguet JP and Damge C (1987) Transmucosal passage of polyalkylcyanoacrylate nanocapsules as a new drug carrier in the small intestine. Biol Cell 61:69. 62. Sai P, Damge C, Rivereau AS, Hoeltzel A and Gouin E (1996) Prophylactic oral administration of metabolically active insulin entrapped in isobutylcyanoacrylate nanocapsules reduces the incidence of diabetes in nonobese diabetic mice. ] Autoimmunity 9:713. 63. Dalengon F, Amjaud Y, Lafforgue C, Derouin F and Fessi H (1998) Atovaquone and rifabutine-loaded nanocapsules: Formulation studies, hit J Pharm 153:127. 64. Calvo P, Remufian C, Vila Jato JL and Alonso MJ (1997) Development of positively charged colloidal drug carriers: Chitosan-coated polyester nanocapsules and submicron emulsions. Coll Polym Sci 275:46-53 65. Vila A, Sanchez A, Evora C, Soriano I, Vila Jato JL and Alonso MJ (2004) PEG-PLA nanoparticles as carriers for nasal protein/vaccine delivery. / Aerosol Med. 66. Prego C, Fernandez-Megia E, Novoa-Carballal R, Quinoa E, Torres D and Alonso MJ (2003) Chitosan and Chitosan-PEG nanocapsules: New carriers for improving the oral absorption of calcitonin. 67. Guterres SS, Fessi H, Barratt G, Puisieux F and Devissaguet JP (2000) Poly(D,L-lactide) nanocapsules containing diclofenac: Protection from muscular damage in rats. / Biomater Sci Polym Ed 11:1347. 68. Hubert B, Atkinson J, Guerret M, Hoffman M, Devissaguet JP and Maincent P (1991) The preparation and acute antihypertensive effects of a nanocapsular form of darodipine, a dihydropyridine calcium entry blocker. Pharm Res 8:734. 69. Nishioka Y and Yoshino H (2001) Lymphatic targeting with nanoparticulate system. Adv Drug Del Rev 47:55. 70. Marchal-Heussler L, Thouvenot P, Hoffman M and Maincent P (1999) Comparison of the biodistribution in mice of Ill-indium oxine encapsulated into poly(lactic-co-glycolic)- D,L-85/15 and poly(epsilon-caprolactone) nanocapsules. ] Pharm Sci 88:450. 71. Morin C, Barratt G, Fessi H, Devissaguet JP and Puisieux F (1994) Improved intracellular delivery of a muramyl dipeptide analog by means of nanocapsules. Int J Immunopharmacol 16:461. 276 Gref& Couvreur 72. Seyler I, Appel M, Devissaguet JP Legrand P and Barratt G (1996) Relationship between NO-synthase activity and TNF-alpha secretion in mouse macrophage lines stimulated by a muramyl peptide entrapped in nanocapsules. Int J Immunopharmacol 18:385. 73. Seyler I, Appel M, Devissaguet JP, Legrand P and Barratt G (1999) Macrophage activation by a lipophilic derivative of muramyldipeptide within nanocapsules: Investigation of the mechanism of drug delivery. / Nanoparticle Res 1:91. 74. Barratt G, Puisieux F, Yu WP, Foucher C, Fessi H and Devissaguet JP (1994) Antimetastatic activity of MDP-L-alanyl-cholesterol incorporated into various types of nanocapsules. Int J Immunopharmacol 16:457. 75. Mosqueira VCF, Legrand P, Bories C, Devissaguet JP and Barratt G (2000) Comparative pharmacokinetics and in vivo efficacy of an intravenous formulation of halofantrine in long-circulating nanocapsules in Plasmodium berghei-iniected mice. Proceedings of the 27th International Symposium on Controlled Release of Bioactive Materials, 7-13 July Paris, France; 490-491. 76. Langer K, Zimmer A and Kreuter J (1999) Acrylic nanoparticles for ocular drug delivery. STP Pharm Sci 7:445. 77. Legrand P, Barratt G, Mosqueira V, Fessi H and Devissaguet JP (1999) Polymeric nanocapsules as drug delivery systems, a review. STP Pharm Sci 9:411. 78. Marchal-Heussler L, Sirbat D, Hoffman M and Maincent P (1993) Poly(epsiloncaprolactone) nanocapsules in carteolol ophthalmic delivery. Pharm Res 10:386. 79. Losa C, Marchal-Heussler L, Orallo F, Vila Jato JL and Alonso MJ (1993) Design of new formulations for topical ocular administration: Polymeric nanocapsules containing metipranolol. Pharm Res 10:80. 80. Calvo P, Thomas C, Alonso MJ, Vila Jato JL and Robinson JR (1994) Study of the mechanism of interaction of poly(e-caprolactone) nanocapsules with the cornea by confocal laser scanning microscopy. Int} Pharm 103:283. 81. Zimmer A, Kreuter J and Robinson JR (1991) Studies on the transport pathway of PBCA nanoparticles in ocular tissues. / Microencapsulation 8:497. 82. Calvo P, Vila-Jato JL and Alonso MJ (1997) Effect of lysozyme on the stability of polyester nanocapsules and nanoparticles: Stabilization approaches. Biomaterials 18:1305. 83. Calvo P, Vila-Jato JL and Alonso MJ (1997) Evaluation of cationic polymer-coated nanocapsules as ocular drug carriers. Int f Pharm 153:41. 84. De Campos A, Sanchez A and Alonso MJ (2001) Chitosan nanoparticles: A new vehicle for the improvement of the ocular retention of drugs. Application to cyclosporin A. Int J Pharm 224:159-168. 85. De Campos A, Sanchez A and Alonso MJ (2003) The effect of a PEG vs a chitosan coating on the interaction of drug colloidal carriers with the ocular mucosa. Eur J Pharm Sci 20:73-81. 86. Calvo P, Vila-Jato JL and Alonso MJ (1997) Evaluation of cationic polymer-coated nanocapsules as ocular drug carriers. Int} Pharm 153:41-50. 13 Dendrimers as Nanoparticulate Drug Carriers SSnke Svenson and Donald A. Tomalia 1. Introduction The development of molecular nanostructures with well-defined particle size and shape is of eminent interest in biomedical applications such as the delivery of active pharmaceuticals, imaging agents, or gene transfection. For example, constructs utilized as carriers in drug delivery generally should be in the nanometer range and uniform in size to enhance their ability to cross cell membranes and reduce the risk of undesired clearance from the body through the liver or spleen. Two traditional routes to produce particles that will meet some of these requirements have been widely investigated. The first route takes advantage of the ability of amphiphilic molecules (i.e. molecules consisting of a hydrophilic and hydrophobic moiety) to self-assemble in water above a system-specific critical micelle concentration (CMC) to form micelles. Size and shape of these micelles depend on the geometry of the constituent monomers, intermolecular interactions, and conditions of the bulk solution (i.e. concentration, ionic strength, pH, and temperature). Spherical micelles are monodisperse in size; however, they are highly dynamic in nature with monomer exchange rates in millisecond to microsecond time ranges. Micelles have the ability to encapsulate and carry lipophilic actives within their hydrocarbon cores. Depending on the specific system, some micelles either spontaneously rearrange to form liposomes after a minor change of solution conditions, or when they are exposed to external energy input such as agitation, sonication, or extrusion through a filter 277 278 Svenson &Tomalia membrane. Liposomes consist of bilayer lipid membranes (BLM) enclosing an aqueous core, which can be utilized to carry hydrophilic actives. Furthermore, liposomes with multilamellar membranes provide cargo space for lipophilic actives as well. However, most liposomes are considered energetically metastable, and will eventually rearrange to form planar bilayers.1'2 The second route relies on engineering the well-defined particles through processing protocols. Examples for this approach include (i) shearing or homogenization of oil-in-water (o/w) emulsions or w / o / w double emulsions to produce stable and monodisperse droplets, (ii) extrusion of polymer strands or viscous gels through nozzles of defined size to manufacture stable and monodisperse micro and nanospheres, (iii) layer-by-layer (LbL) deposition of polyelectrolytes and other polymeric molecules around colloidal cores, resulting in the formation of monodisperse nanocapsules after the removal of the templating core, and (iv) controlled precipitation from a solution into an anti-solvent, including supercritical fluids. Size, degree of monodispersity, and stability of these structures depend on the systems that are being used in these applications.3 These systems and their utilization in drug delivery are being discussed in detail in other chapters of this book. Currently, a new third route to create very well-defined, monodisperse, stable molecular level nanostructures is being studied based on the "dendritic state" architecture.4 Dendritic architecture is undoubtedly one of the most pervasive topologies observed throughout biological systems at virtually all dimensional length scales. This architecture is found at the meter scale in tree branching and roots, on the centimeter and millimeter scales in circulatory topologies in the human anatomy such as lungs, kidney, liver, and spleen, and on the micrometer scale in cerebral neurons. On the nanometer level, key examples of dendritic structures include glycogen, amylopectin, and proteoglycans. Amylopectins and glycogen are critical molecular level constructs involved in energy storage in plants and animals, while proteoglycans are an important constituent of connective tissue, determining its viscoelastic properties. Upon the analysis of these ubiquitous dendritic patterns, it is evident that these highly branched architectures offer unique interfacial and functional performance advantages. The objective of this review is to study the use of dendrimers in drug delivery applications. Four main properties of dendrimers will be discussed: (i) nanoscale container properties (i.e. encapsulation and transport of a drug), (ii) nano-scaffolding properties (i.e. surface adsorption or attachment of a drug and/or targeting ligand), (iii) dendrimers as drugs, and (iv) biocompatibility of dendrimers. In addition, routes of application currently investigated will be presented. Particular emphasis will be placed on poly (amidoamine) (PAMAM) dendrimers, the first and most extensively studied family of dendrimers.4c'5 Dendrimers as Nanoparticulate Drug Carriers 279 2. Nanoscale Containers — Micelles, Dendritic Boxes, Dendrophanes, and Dendroclefts Dendrimers may be visualized as consisting of three critical architectural domains: (i) the multivalent surface, containing a larger number of potentially reactive/ passive sites (nano-scaffolding), (ii) the interior shells (i.e. branch cell layers defined by dendrons) surrounding the core, and (iii) the core to which the dendrons are attached. The two latter domains represent well-defined nano-environments, which are protected from the outside by the dendrimer surface (nanoscale containers) in the case of higher generation dendrimers. These domains can be tailored for a specific purpose. The interior is well-suited for host-guest interaction and the encapsulation of guest molecules. 2.1. Dendritic micelles Tomalia and coworkers demonstrated by electron microscopy observation that sodium carboxylated PAMAM dendrimers possess topologies reminiscent of regular classical micelles.4 It was also noted from electron micrographs that a large population of individual dendrimers possessed a hollow core. Supporting these observations, Turro and colleagues designed a hydrophobic 12-carbon atom alkylene chain into the core of a homologous series of PAMAM dendrimers (G = 2, 3, and 4) to mimic the hydrophobic and hydrophilic core-shell topology of a regular micelle. The hosting properties of this series towards a hydrophobic dye as a guest molecule were then compared with a PAMAM dendrimer series possessing non-hydrophobic cores (e.g. NH3 and ethylenediamine). Dramatically enhanced emission of the hydrophobic dye was noted in aqueous solution in the presence of hydrophobic versus hydrophilic cored dendrimers.6a Less polar dendrimers (i.e. dendrimers containing aryl groups or other hydrophobic moieties as building elements), behave as inverse micelles.6b A critical property difference relative to micelles is the increased density of surface groups with higher generations. At some generational level, the surface groups will reach the so-called "de Gennes dense packing" limit and seal the interior from the bulk solution (Fig. I ) . 7 - 9 The limit depends on the strength of intramolecular interactions between adjacent surface groups, and therefore, on the condition of the bulk solution (i.e. pH, polarity and temperature). This nanoscale container feature, originally noted for PAMAM dendrimers by Tomalia et al. and referred to as "unimolecular encapsulation", can be utilized to tailor the encapsulation and release properties of dendrimers in drug delivery applications.910 For example, adding up to a limiting amount of Xmmol of either 2,4-dichlorophenoxyacetic acid or aspirin (acetylsalicyhc acid) to 1 mmol of 280 Svenson & Tomalia ,--oV..> jfc-iiw-'i.. J&:5Sgtfhv ;#88c?pi „•;*: 7.:;^- *#.# p f e $$8§? • • J • ^ • ' ' ^ ,s&$p?* '%$$? 4 5 6 7 8 9 10 Fig. 1. Periodic properties of PAMAM dendrimers generations G = 4-10, depicting the decreasing distances between surface charges (Z-Z). The "de Gennes dense packing" appears atG = 8. Dendrimers G = 4-6 display "nanoscale container" properties, the larger analogues G = 7-10 display "nano-scaffolding" properties. STARBURST® carbomethoxy-terminated PAMAM dendrimers generations 0.5-5.5 produced spin-lattice relaxation times (Tj) much lower than the values of these guest molecules in solvent without dendrimer. The new relaxation times decreased for generations 0.5-3.5, but remained constant for generations 3.5 to 5.5. The maximum concentration X varied uniformly from 12 (generation 0.5) to 68 (generation 5.5). On the basis of these maximum concentrations, the guest-to-host ratios were shown to be ~ 4:1 by weight and ~ 3:1 based on a molar comparison of dendrimer guest carboxylic acid-to-interior tertiary nitrogen moieties for generations 2.5-5.5. Exceeding the maximum concentration X resulted in the appearance of a second relaxation time, Tv, characteristic of the guest molecules in bulk solvent phase.10 2.2. Dendritic box (Nano container) Surface-modification of G = 5 poly(propyleneimine) (PPI) dendrimers with Boc-protected amino acids induced dendrimer encapsulation properties by the formation of dense, hydrogen-bonded surface shells with solid-state character ("dendritic box").8 Small guest-molecules were captured in such dendrimer interiors and were unable to escape even after extensive dialysis. The maximum amount of entrapped guest molecules was directly proportional to the shape and size of the guest molecules, as well as to the amount, shape and size of the available internal dendrimer cavities. Four large guest-molecules (i.e. Rose Bengal) and 8-10 small guest-molecules (i.e. p-nitrobenzoic acid) could be simultaneously encapsulated within PPI dendrimers containing four large and twelve smaller cavities. Remarkably, this dendritic box could be opened under controlled conditions to release either some or all of the entrapped guest molecules. For example, partial hydrolysis of the hydrogen-bonded Boc-shell liberated only small guest-molecules, whereas total hydrolysis released all sizes of entrapped molecules.8'11-12 Although the "dendritic box" concept demonstrates the unique shapedependent cargo space that can be found in certain dendrimers, other parameters have to be considered as well for delivering and releasing therapeutic drugs Dendrimers as Nanoparticulate Drug Carriers 281 under physiological conditions. From a thermodynamic perspective, free guestmolecules (i.e. drugs) can be distinguished from those encapsulated or bound in a complex by finite energy barriers related to the ease of entry and departure to the dendrimer cavities. If the drug molecule is incompatible with either the dimension or hydrophilic/lipophilic character of the dendrimer cavity, a complex might not form, or the guest might only be partially encapsulated within the dendrimer host. A hydrophobic drug would be expected to associate with a dendrimer core to achieve maximum contact with its hydrophobic domain. In addition, the hydrophobic character of this guest molecule would be expected to isolate itself from the dendrimer surface and the interface to the bulk solution to afford minimum contact with polar and aqueous domains (i.e. physiological media). Notably, the hydrophobic and hydrophilic properties, as well as other non-covalent binding properties of these spatial binding-sites are expected to strongly influence these guest-host relationships. Analysis of a typical symmetrically branched dendrimer makes it apparent that there are other subtle and yet important parameters that could control the interior space of a dendrimer and influence the guest-host interactions. These include components such as branching angles, branching symmetry rotational angles, and the length of a repeat-unit segment.13 Of equal importance are the properties of the core. Within a homologous PAMAM dendrimer series, the effect of changing the length scale of the core on dendrimer guest-host properties was studied. Specifically, a series of polyhydroxy-surfaced PAMAM dendrimers with core molecules differing in length by one carbon atom (NH2-Cn-NH2 with n = 2-6) were synthesized. Three aromatic carboxylic acids, differing systematically by one aromatic ring (benzoic acid, 1-naphthoic acid, 9-anthracene carboxylic acid), were examined as guest-molecule probes. Two sets of dendrimers, possessing 24 and 48 surface hydroxy groups, were investigated.14 The observed trends can be summarized as follows: (i) in general, all dendritic hosts accommodated larger amounts of the smaller guest-molecule (i.e. molar uptake benzoic > 1-naphthoic > 9-anthracene carboxylic acid). This observation was particularly significant for the more congested dendrimer surface having 48 surface OH-groups. (ii) Uptake maxima values specific to both the core size and the specific guest-probe were noted. This observation might be related to the combination of shape and lipophilicity manifested by the guest probe, (iii) A decrease in the molar uptake was measured for all probes as the core was enhanced beyond an ideal dimension (i.e. 5-6 carbons). It is therefore obvious that both core size and surface congestion dramatically affect the cargo-space of the dendrimer host. Furthermore, it is apparent that size and shape of the guest probe can significantly affect the maximum loading as a function of core size. Finally, it should also be noted that for the dendrimers G = 2 (24-OH) and G = 3 (48-OH), the guest probes had desirable release properties from the host as a function of time, when re-dissolved in water. Performing these same experiments using a dendrimer 282 Svenson & Tomalia with more densely packed surface groups (i.e. G = 4 with 96 surface OH-groups) appeared to produce dendritic box behavior. Although guest molecules could be encapsulated within the core, the release from the host was delayed as determined by analysis after extensive dialysis.14 Structure-property relationships in dendritic encapsulation have been studied extensively, mainly using photoactive and redoxactive model dendrimers to gain a better understanding of the structural effects that cores and branches have on encapsulation.15-17 2.3. Dendrophanes and dendroclefts Specific binding of guest molecules to the dendrimer core can affect the loading capacity by enhancing specific interactions between the core and guest (i.e. hydrophobic and polar interactions). Dendrimers specifically tailored to bind hydrophobic guests to the core have been created by Diederich and coworkers and coined "dendrophanes". These water-soluble dendrophanes are built around a cyclophane core, and can bind aromatic compounds, presumably via p -p interactions. Dendrophanes were shown to be excellent carriers of steroids.18'19 The same group synthesized dendrimers tailored to bind more polar bioactive compounds to the core, coined "dendroclefts".20'21 In another approach, the surface amines of PAMAM dendrimers were modified with tris(hydroxymethyl)aminomethane (TRIS) to create water-soluble dendrimers capable of binding carboxylic aromatic, antibacterial compounds, which could be released by lowering the pH.14 An alternative approach to creating dendritic hosts with highly selective guest recognition utilized the principle of "molecular imprinting".22 A dendrimer consisting of a porphyrin core and a surface containing terminal double bonds was polymerized into a polydendritic network. Subsequently, the base-labile ester bonds between cores and dendritic wedges were cleaved, releasing the porphyrin core from the dendritic polymer. This polymer was capable of selectively binding porphyrins with association constants of 1.4 x 105 M_1. Very recently, an impressive approach has been presented, using tandem mass spectrometry, i.e. the combination of electrospray ionization (ESI) and collision-induced dissociation (CID) mass spectrometers connected in series, to investigate the dynamic behavior of host-guest dendrimer complexes.23 This approach offers the potential to provide better insights into these constructs. 3. Dendrimers in Drug Delivery Dendrimers have been utilized to carry a variety of small molecule pharmaceuticals with the purpose to enhance their solubility and therefore bioavailability, and to utilize the passive and active targeting properties of dendrimers, either through the Dendrimers as Nanoparticulate Drug Carriers 283 "Enhanced Permeability and Retention" (EPR)24 effect or specific targeting ligands. Some aspects of dendrimers in drug delivery have been reviewed recently.13,25-27 In the following, selected examples of important drug delivery aspects will be presented. 3.1. Cisplatin Encapsulation of the well-known anticancer drug cisplatin within PAMAM dendrimers gives complexes that exhibit slower release, higher accumulation in solid tumors, and lower toxicity compared with free cisplatin.28'29 Cisplatin is an antitumor drug that exerts its effects by forming stable DNA-cisplatin complexes through intrastrand cross-links, resulting in an alteration of the DNA structure that prevents replication and activates cell repair mechanisms. The cell detects defective DNA and initiates apoptosis. Cisplatin is effective in treating several cancers such as ovarian, head and neck, and lung cancers, as well as melanomas, lymphomas, osteosarcomas, bladder, cervical, bronchogenic, and oropharyngeal carcinomas. Unfortunately, cisplatin has many adverse side effects to the body, the most important being nephrotoxicity and cytotoxicity to non-cancerous tissue, because of the non-selective interaction between cisplatin and DNA. In addition, the therapeutic effect of cisplatin is limited by its poor water solubility (1 mg/mL), low lipophilicity, and the development of resistance to cisplatin drugs. Although numerous cisplatin derivatives have undergone preclinical and clinical testing, only cisplatin and its derivatives carboplatin and oxaliplatin have been approved for routine clinical use (Fig. 2).30 Preliminary studies gave cisplatin loadings of 15-25 wt% for PAMAM dendrimers generation 3.5 (size ~ 3.5 nm; MW ~ 13 kDa). In comparison, the cisplatin loading of linear poly(amidoamines) and linear N-(2-hydroxypropyl) methacrylamide (HPMA; MW 25-31 kDa) was found to be 5-10 wt% and 3-8 wt%, respectively. HPMA-cisplatin complexes are currently in clinical trials.31 The cisplatin-dendrimer complex could be visualized by Atomic Force Microscopy (AFM; carbon nanotip) as shown in Fig. 3. H3N, CI H3NT \ H3N- \ l ^ V O Fig. 2. Chemical structures of the platinum drugs cisplatin (PLATINOL®), carboplatin (PARAPLATIN®), and oxaliplatin (ELOXATIN™). 284 Svenson & Tomalia Fig. 3. AFM images of cisplatin-dendrimer complexes at 120 (left) and 4nm (right) magnification. Table 1 AUC value (/xg Pt/mLblood or /xg Pt/organ) over 48 hours; 5 mice/data point. Organ Cisplatin Cisplatin-dendrimer Complex Tumor 5.3 25.4 Blood 9.4 10.7 Liver 51.6 17.0 Kidney 57.6 138.1 The tumor activity of the cisplatin-dendrimer formulation was studied using B16F10 cells. These cells were injected into C57 mice subcutaneously (s.c.) to provide a solid tumor model. After approximately 12 days, when the tumors had developed to a mean area of 50-100 mm2, the animals were injected i.v. with a single dose of either cisplatin or cisplatin-dendrimer complex (1 mg/kg cisplatin for both formulations). At certain time points within 48 hours, animals were culled and blood and tissue samples were taken. Compared with cisplatin alone, the cisplatin-dendrimer complex was found to accumulate preferentially in the tumor site relatively quickly after the injection. The tumor area under the curve (AUC) for the complex was 5 times higher than that of free cisplatin, while that in the kidney only increased 2.4 times, and accumulation in the liver was reduced (Table I).29 Another recent study revealed a sufficient stability of cisplatin-dendrimer complexes, with a 20% release of cisplatin over the first 8 hours, and an additional 60% release within 150 hours. In vivo animal efficacy of the platinate was demonstrated using B16F10 tumor cells that are subcutaneous implanted into mice. The tumor was allowed to grow for 7 days prior to treatment with two doses of drug on day 7 and day 14, providing equal cisplatin (5 mg/kg) doses in both the dendrimercisplatin complex and free cisplatin. A tumor weight reduction of ~ 40% above that observed for the free drug was found in this study. Dendrimers as Nanoparticulate Drug Carriers 285 3.2. Silver salts The encapsulation of silver salts within PAMAM dendrimers produced conjugates, exhibiting slow silver release rates and antimicrobial activity against various Gram positive bacteria.32 PAMAM dendrimers, generation four with ethylenediamine (EDA) core and tris(2-hydroxymethyl)amidomethane (TRIS) OH-surface and generation five, EDA core with carboxylate COO~ surface, were used. Silver containing PAMAM complexes were prepared by adding aqueous solutions of the dendrimers to the calculated amount of silver acetate powder. Although CHaCOOAg is hardly soluble in water, it quickly dissolved in the PAMAM solutions. This enhancement is due to the combined action of the silver carboxylate salt formation and/or to the complex formation with the internal dendrimer nitrogens. This procedure resulted in slightly yellow dendrimer-complex/salt solutions that very slowly photolyzed when exposed to light, into dark brown, metallic silver, containing dendrimersilver nanocomposite solutions. Final sample concentrations were confirmed by atomic absorption spectroscopy. For antimicrobial testing, the standard agar overlay method was used. In this test, dendrimer-silver compounds were examined for diffusible antimicrobial activity by placing a lO-^L sample of each solution onto a 6-mm filter paper disk and applying the disk to a dilute population of the test organisms, Staphylococcus aureus, Pseudomonas aeruginosa, and Escherichia coli. The silver-dendrimer complexes displayed antimicrobial activity, comparable to or better than those of silver nitrate solutions. Interestingly, increased antimicrobial activity was observed with dendrimer carboxylate salts, which was attributed to the very high local concentration (256 carboxylate groups around a 5.4 nm diameter sphere) of nanoscopic size silver composite particles that are accessible for microorganisms. The antimicrobial activity was smaller when internal silver complexes were applied instead of silver adducts to the surface, indicating that the accessibility of the silver is an important factor. 3.3. Adriamycin, methotrexate, and 5-fluorouracil The anticancer drugs, adriamycin and methotrexate, were encapsulated into generations 3 and 4 PAMAM dendrimers which had poly(ethylene glycol) monomethyl ether chains with molecular weights of 550 and 2000 Da attached to their surfaces via urethane bonds (Fig. 4). The encapsulation efficiency was dependent on the PEG chain length and the size of the dendrimer, with the highest encapsulation efficiencies (on average, 6.5 adriamycin molecules and 26 methotrexate molecules per dendrimer) found for the G = 4 PAMAM terminated with PEG2000 chains. The drug release from this dendrimer was sustained at low ionic strength, again reflecting PEG chain length and dendrimer size, but fast in isotonic solution.33 In a related study, it was reported that the surface coverage of PAMAM dendrimers with 286 Svenson & Tomalia Fig. 4. Above: Structures of anticancer drugs adriamycin (left) and methotrexate (right). Below: Schematic presentations of the encapsulation of methotrexate (left) and 5-fluorouracil (right) into PAMAM dendrimers. PEG2000 chains had little influence on the encapsulation efficiency of methotrexate, but affected the release rate.34 A similar construct between PEG chains and PAMAM was utilized to deliver the anticancer drug 5-fluorouracil. Encapsulation of 5-fluorouracil into G — 4 PAMAM dendrimers with carboxymethyl PEG5000 surface chains revealed reasonable drug loading, a reduced release rate, and reduced hemolytic toxicity compared to the non-PEGylated dendrimer (Fig. 4).35 3.4. Etoposide, mefenamic acid, diclofenac, and venlafaxine The combination between dendrimers and hydrophilic and/or hydrophobic polymer chains has recently been extended to solubilize the hydrophobic anticancer drug etoposide. A star polymer composed of amphiphilic block copolymer arms has been synthesized and characterized. The core of the star polymer was a generation two PAMAM-OH dendrimer, the inner block of the arm a lipophilic poly(e-caprolactone) (PCL) and the outer block of the arm a hydrophilic PEG500o- The star-PCL polymer was synthesized first by ring-opening polymerization of e-caprolactone with the PAMAM-OH dendrimer as initiator. The PEG polymer was then attached to the PCL terminus by an ester-forming reaction. Characterization with SEC, 1-H NMR, FTIR, TGA, and DSC confirmed the star structure of the polymers. A loading capacity of up to 22% (w/w) was achieved with etoposide. Dendrimers as Nanoparticulate Drug Carriers 287 A cytotoxicity assay demonstrated that the star-PCL-PEG copolymer was nontoxic in cell culture.36 Citric acid-poly(ethylene glycol)-citric acid (CPEGC) triblock dendrimers generations 1-3 were applied to encapsulate small molecule drugs such as mefenamic acid and diclofenac. The formulations were stored at room temperature for up to ten months and remained stable with no reported release of the drugs.37 The attachment of the novel third-generation antidepressant venlafaxine onto anionic PAMAM dendrimers (G = 2.5) via a hydrolyzable ester bond and the incorporation of this drug-dendrimer complex into a semi-interpenetrating network of an acrylamide hydrogel has been studied as a novel drug delivery formulation to avoid the currently necessary multiple daily administration of the antidepressant. The effect of PEG concentration and molecular weight was studied to find optimal release conditions.38 3.5. Ibuprofen, indomethacin, nifedipine, naproxen, paclitaxel, and methylprednisolone The anti-inflammatory drug ibuprofen was used as a model compound to study its complexation and encapsulation into generations 3 and 4 PAMAM dendrimers and a hyperbranched polyester, having approximately 128 surface OH-groups. It was found that up to 78 ibuprofen molecules were complexed by the PAMAM dendrimers through electrostatic interactions between the dendrimer amines and the carboxyl group of the drug. In contrast, up to 24 drug molecules were encapsulated into the hyperbranched polyol.39 The drug was successfully transported into A549 human lung epithelial carcinoma cells by the dendrimers. The PAMAM dendrimers with either amino or hydroxy surfaces entered the cells faster (in approximately 1 hr) than the hyperbranched polyol (approximately 2 hrs). However, both entries were faster than the pure drug. The anti-inflammatory effect of ibuprofen-dendrimer complexes was demonstrated by more rapid suppression of COX-2 mRNA levels than that achieved by the pure drug.40 The non-steroidal anti-inflammatory drug (NSAID) indomethacin is practically insoluble in water and only sparingly soluble in alcohol. Encapsulation of indomethacin into generation 4 PAMAM dendrimers with amino, hydroxy, and carboxylate surfaces remarkably enhanced the drug solubility in water, and therefore, its bioavailability (Fig. 5).41 The encapsulation efficiency of indomethacin into PAMAM dendrimers is dependent on the dendrimer size (G6 > G5 > G4 > G3) and the surface functionalization, (NH2 > PEG = PYR > AE) (Fig. 6).42 The effect of PAMAM dendrimer generation size and surface functional group on the aqueous solubility, and therefore, bioavailability of the calcium channel blocking agent nifedipine has been studied using PAMAM dendrimers with EDA 288 Svenson & Tomalia CH, COOH E 800^ 600- g 400- 200 0.1 0.2 Dendrimerconc. (%v/w) Fig. 5. Molecular structure of indomethacin and its solubility profiles in the presence of differing concentrations of G4-NH2/ (•) G4-OH (•), and G4.5-COOH (A) PAM AM dcndrimers at pH 7 (« = 3, R.S.D. < 5%). Encapsulation efficiency of EDA core PAMAM dendrimer « E 80 - 60- g 40 3 20- I ft ,rf sJl • G3 • G4 • G5 • G6 —n .r-dl ^ NH2 PEG PYR AE COONa Surface Functionality sue TRIS Fig. 6. Encapsulation efficiency into PAMAM dendrimers generations 3-6 with amino (NH2), poly(ethylene glycol) (PEG), carbomethoxypyrrolidinone (PYR), amidoethanol (AE), sodium carboxylate (COONa), succinamic acid (SUC), and tris(hydroxymethyl)- aminomethane (TRIS) surface groups. core and amino surface (G = 0,1,2,3) or ester surface (G — 0.5,1.5,2.5) at pH 4, 7 and 10. The solubility enhancement of nifedipine was higher in the presence of ester-terminated dendrimers than their amino-terminated analogues, possessing the same number of surface groups. The nifedipine solubility expectedly increased with the size of the dendrimers. For pH 7, the sequence G2.5 > G3 > G1.5 > G2 > G0.5 > Gl > GO was reported.43 In another approach, the non-steroidal anti-inflammatory drug naproxen was covalently attached to unsymmetrical poly(arylester) dendrimers to prepare a complex with enhanced water solubility of the drug and access for hydrolytic cleavage Dendrimers as Nanoparticulate Drug Carriers 289 1E-3 0.01 0.1 1 0 2k 4k 6k 8k Concentration (M) Molecular weight Fig. 7. Aqueous paclitaxel solubility as a function of the polyglycerol dendrimer concentration (mean ± SD, n = 3); G5 (circle), G4 (triangle), G3 (square), and PEG400 (diamond) (left). Molecular weight dependency of dendrimers (closed circle) and PEG (open circle) on the aqueous paclitaxel solubility. The concentration of dendrimers and PEG was 10wt%. (Reproduced with permission from Ref. 45. Copyright 2004 American Chemical Society.) of the bond between drug and carrier. Detailed results on the biological evaluation of these complexes have not been reported.44 The anticancer drug paclitaxel, which is being used to treat metastatic breast and ovarian cancers and Kaposi's sarcoma, has poor water solubility. To enhance its bioavailability, paclitaxel has been encapsulated into polyglycerol dendrimers, resulting in a 10,000-fold improved water solubility compared with the pure drug, which is much higher than that found for PEG400, a commonly used linear chain cosolvent or hydrotropic agent (Fig. 7). The drug release rate was a function of the dendrimer generation.45 Generation 4 PAMAM dendrimers with hydroxy surface have been utilized to improve the bioavailability of the corticosteroid methylprednisolone, which decreases inflammation by stabilizing leukocyte lysosomal membrane. By connecting the drug to the dendrimer using glutaric acid as the spacer, a payload of 32 wt% was achieved. The drug-dendrimer complex was taken up by A549 human lung epithelial carcinoma cells and mostly localized in the cytosol. The complex showed a pharmacological activity comparable to the free drug as measured by the inhibition of the prostaglandin secretion.46 3.6. Doxorubicin and camptothecin — self-immolative dendritic prodrugs An exciting new approach to dendritic drug delivery involves the utilization of a drug as a part of the dendritic molecule. Self-immolative dendrimers have recently been developed and introduced as a potential platform for a multi-prodrug. These unique structural dendrimers can release all of their outer branch units through 10 •jj 0.01- I 1E-3, 290 Svenson & Tomalia co3 >~q vr HsO CegS^-'y «Ht Fig. 8. Mechanism of dimeric prodrug activation by a single enzymatic cleavage. (Reproduced with permission from Ref. 47. Copyright 2004 American Chemical Society.) a self-immolative chain fragmentation, initiated by a single cleavage at the dendrimer's core. Incorporation of drug molecules as these outer branch units and an enzyme substrate as the trigger can generate a multi-prodrug unit that will be activated with a single enzymatic cleavage (Fig. 8). The first generation of dendritic prodrugs with doxorubicin and camptothecin as branch units and retro-Michael focal trigger, which can be cleaved by the catalytic antibody 38C2, has been reported. Bioactivation of the dendritic prodrugs was evaluated in cell-growth inhibition assay with the Molt-3 leukemia cell line in the presence and absence of antibody 38C2. A remarkable increase in toxicity was observed. Dependent on the linker molecule, different numbers of drug molecules can be released in one single activation step.47'48 In a more "classical" approach to deliver doxorubicin, two polyester-based dendrimers (generation 4 with trisphenolic core) were synthesized, one carrying a hydroxy surface, the other a tri(ethylene glycol) monomethyl ether surface. These dendrimers were compared with a 3-arm poly(ethylene oxide) star polymer, carrying G = 2 dendritic polyester units at the surface. The star polymer gave the most promising results regarding cytotoxicity and systemic circulatory half-life (72hrs). Therefore, the anticancer drug doxorubicin was covalently bound to this carrier via an acid-labile hydrazone linkage. The cytotoxicity of doxorubicin was significantly reduced (80-98%) and the drug was successfully taken up by several cancer cell lines.49 Dendrimers as Nanoparticulate Drug Carriers 291 3.7. Photodynamic therapy (PDT) and boron neutron capture therapy (BNCT) Dendrimers have been used to optimize the antitumor effect in photodynamic therapy (PDT) and boron neutron capture therapy (BNCT). One of the newest developments in the dendrimer field is their application to photodynamic therapy (PDT). This cancer treatment involves the administration of a light-activated photosensitizing moiety that selectively concentrates in diseased tissue. Subsequent activation of the photosensitizer leads to the generation of reactive oxygen, primarily singlet oxygen, that damages intracellular components such as lipids and amino acid residues through oxidation, ultimately leading to cell death by apoptosis. Disadvantages of currently used photosensitizers include skin phototoxicity, poor selectivity for tumor tissue, poor water solubility, and difficulties in the treatment of solid tumors because of the impermeability of the skin and tissues to the visible light required to excite the chromophores. In one set of studies, dendrimers have been constructed around a light harvesting core (i.e. a porphyrin).50 To reduce the toxicity under non-irradiative conditions (dark toxicity) and to prevent aggregation, and consequently, self-quenching of the porphyrin cores, these dendrimers have been further encapsulated into micelles. For example, poly(ethylene glycol)-b-poly(aspartic acid) and PEG-b-poly(L-lysine) micelles have been studied in this regard. These micelles are stable under physiological conditions pH 6.2 to 7.4. However, they disintegrate in the acidic intracellular endosomal compartment (pH ~ 5.0).51/52 Alternatively, the photosensitizer 5-aminolevulinic acid has been attached to the surface of dendrimers and studied as an agent for PDT of tumorigenic keratinocytes.53 Photosensitive dyes have been incorporated into dendrimers and utilized in PDT devices. For example, uptake, toxicity, and the mechanism of photosensitization of the dye pheophorbide a (pheo) was compared with its complex with diaminobutane poly(propylene imine) (DAB) dendrimers in human leukemia cells in vitro.5i The second therapy, boron neutron capture therapy, is a cancer treatment based on a nuclear capture reaction. When 10B is irradiated with low energy or thermal neutrons, highly energetic a-particles and 7Li ions are produced, that are toxic to tumor cells. To achieve the desired effects, it is necessary to deliver 10B to tumor cells at a concentration of at least 109 atoms per cell. High levels of boron accumulation in tumor tissue can be achieved by using boronated antibodies that are targeted towards tumor antigens. However, this approach can impair the solubility and targeting efficiency of the antibodies. One study, involving intratumoral injection of a conjugation between a generation 5 PAMAM dendrimer carrying 1100 boron atoms and cetuximab, a monoclonal antibody specific for the EGF receptor, showed that the conjugate was present 292 Svenson & Tomalia Fig. 9. Schematic presentation of an EDA core G = 3 PAMAM dendrimer (1), the boron carrier Na(CH3)3NB10H8NCO (2), and the targeting ligand folic acid (3). (Reproduced with permission from Ref. 56. Copyright 2003 American Chemical Society.) at an almost 10-fold higher concentration in brain tumors than in normal brain tissue.55 To reduce the liver uptake observed for boronated PAMAM dendrimer conjugates, PEG chains were attached onto the dendrimer surface, in addition to the borane clusters, to provide steric shielding. As compared with a dendrimer without PEG chains, the amount of liver uptake was found to be less for PEG-conjugated dendrimers with an average of 1.0-1.5 chains of PEG2000/ but higher for dendrimers with 11 chains of PEG550. Folic acid moieties were also conjugated to the ends of the PEG chains to enhance the uptake of the dendrimers by tumors overexpressing folate receptors. Although this strategy was successful in enhancing localization of the molecules to tumors in mice bearing 24JK-FBP tumors expressing the folate receptor, it also led to an increase in the uptake of the dendrimers by the liver and kidneys.56 4. Nano-Scaffolds for Targeting Ligands The surface of dendrimers provides an excellent platform for the attachment of cellspecific ligands, solubility modifiers, stealth molecules, reducing the interaction with macromolecules from the body defense system, and imaging tags. The ability to attach any or all of these molecules in a well-defined and controllable manner onto a robust dendritic surface, clearly differentiates dendrimers from other carriers such as micelles, liposomes, emulsion droplets, and engineered particles. 4.1. Folic acid One example of cell-specific dendritic carriers is a dendrimer modified with folic acid. The membrane-associated high affinity folate receptor (hFR) is a folate binding protein that is overexpressed on the surface of a variety of cancer cells, and Dendrimers as Nanoparticulate Drug Carriers 293 therefore, folate-modified dendrimers would be expected to internalize into these cells preferentially over normal cells via receptor-mediated endocytosis. Folatedendrimer conjugates have been shown to be well-suited for targeted, cancerspecific drug delivery of cytotoxic substances.56-59 In a very recent study, branched poly(L-glutamic acid) chains were centered around PAMAM dendrimers generations 2 and 3 and poly(ethylene imine) (PEI) cores to create new biodegradable polymers with improved biodistribution and targeting ability. These constructs were surface-terminated with poly(ethylene glycol) chains to enhance their biocompatibility, and folic acid ligands to introduce cellspecific targeting. Cell binding studies have been performed using the epidermal carcinoma cell line, KB.60 4.2. Carbohydrates In addition to folates, carbohydrates constitute another important class of biological recognition molecules, displaying a wide variety of spatial structures due to their branching possibility and anomericity. To achieve sufficiently high binding affinities between simple mono- and oligosaccharide ligands and cell membrane receptors, these ligands have to be presented to the receptors in a multivalent or cluster fashion.61,62 The highly functionalized surface of dendrimers provides an excellent platform for such presentations. The design, synthesis, and biomedical use of glycodendrimers, as well as their application in diagnostic and for vaccinations, have been thoroughly reviewed recently.63-69 For example, the Thomsen- Friedenreich carbohydrate antigen (T-antigen), /J-Gal-(l-3)-a-GalNAc, which has been well documented as an important antigen for the detection and immunotherapy of carcinomas, especially relevant to breast cancer, has been attached to the surface of PAMAM and other dendrimers.70-72 An enhanced binding affinity was observed for all glycodendrimers. These constructs could have potential in blocking the metastatic sites of invasive tumor cells. A series of dendritic ,6-cyclodextrin derivatives, bearing multivalent mannosyl ligands, has been prepared and their binding efficiency towards the plant lectin concanavalin A (Con A) and a mammalian mannose-specific cell surface receptor from macrophages has been studied. The effects of glycodendritic architecture on binding efficiency, molecular inclusion, lectin-binding properties, and the consequence of complex formation using the anticancer drug docetaxel on biological recognition were investigated.73 Di- to tetravalent dendritic galabiosides, carrying (Galal-4Gal) moieties on their surfaces, were studied as inhibitors of pathogens based on bacterial species such as E. colt and Streptococcus suis. Attachment of dendritic galabiosides onto cell surfaces would be expected to inhibit the attachment of bacteria using the same sugar ligand-receptor interactions. The study revealed a clear enhancement of the binding affinity between 294 Svenson & Tomalia glycodendrons and cell surfaces, with an increasing number of sugar moieties.74 In a similar approach, glycodendrons carrying two to four /i-D-galactose moieties on their surface, while the dendron core was connected to a protein-degrading enzyme, were synthesized. These glycodendriproteins are expected to attach to the surface of bacteria, allowing the enzyme to degrade the bacterial adhesin, hence rendering the bacteria incapable of attaching to the cell surfaces.75 Anionic PAMAM dendrimers (G = 3.5) were conjugated to D(+)-glucosamine and D(+)-glucoseamine 6-sulfate. These water-soluble conjugates not only revealed immuno-modulatory and antiangiogenic properties, but synergistically prevented scar tissue formation after glaucoma filtration surgery. In a validated and clinically relevant rabbit study, the longterm success rate was increased from 30 to 80% using these dendrimer-conjugates.76 4.3. Antibodies and biotin-avidin binding Generation 5 PAMAM dendrimerendrimers with amino surface were conjugated to fluorescein isothiocyanate as a means to analyze cell binding and internalization. Two different antibodies, 60bca and J591, which bind to CD14 and prostate-specific membrane antigen (PSMA) respectively, were used as model targeting molecules. The binding of the antibody-conjugated dendrimers to antigen-expressing cells was evaluated by flow cytometry and confocal microscopy. The conjugates specifically bound to the antigen-expressing cells in a time- and dose-dependent fashion, with affinity similar to that of the free antibody (Fig. 10). Confocal microscopic analysis suggested at least some cellular internalization of the dendrimer conjugate. Dendrimer-antibody conjugates are, therefore, a suitable platform for targeted molecule delivery into antigen-expressing cells.77 Monolayers formed by generation 4 PAMAM dendrimers on a gold surface were functionalized with biotin and produced a biomolecular interface that was Fig. 10. Confocal microscopic analysis of HL60 cells, which were incubated (1 h at 4°C) with 12.5nM G5 PAMAM carrying fluorescence dye and 60bca antibody on the surface. The cells were rinsed and confocal images were taken. The left and right panels represent the FITC fluorescence and light images taken in the same cell. The arrow indicates the binding of the conjugate on the cell surface at 4°C. (Reproduced with permission from Ref. 77. Copyright 2004 American Chemical Society.) Dendrimers as Nanoparticulate Drug Carriers 295 capable of binding high levels of avidin. Avidin binding as high as 88% coverage of the surface was observed despite conditions that should cause serious steric hindrance. These dendritic monolayers were utilized as a model to study proteinligand interactions.78 4.4. Penicillins The surfaces of PAMAM dendrimers, generations 0 to 3, were decorated with benzylpenicillin in an attempt to develop a new in vitro test to quantify IgE antibodies to specific ^-lactam conjugates, with the goal of improving the existing methods for diagnosing allergy to this type of antibiotic. The monodispersity of dendrimers is advantageous over conventional peptide carrier conjugates such as human serum albumin (non-precise density of haptens in their structure) and poly-L-lysine (mixture of heterogeneous molecular weight peptides). Preliminary radioallergosorbent tests (RAST), using sera from patients allergic to penicillin, have confirmed the usefulness of penicilloylated dendrimers.79 Penicillin V was used as a model drug containing a carboxylic group and attached to the surface of PAMAM dendrimers generations 2.5 and 3, both containing 32 surface functionalities. The drug was complexed to the dendrimers via amide or ester bonds. It was found in tests using a single-strain bacterium, Staphylococcus aureus, that the bioavailability of the penicillin was unaltered after the drug was released from the complex through ester bond hydrolysis.80 5. Dendrimers as Nano-Drugs Dendrimers have been studied as antitumor, antiviral and antibacterial drugs.25 The most prominent and advanced example is the use of poly(lysine) dendrimers, modified with sulfonated naphthyl groups, as antiviral drugs against the herpes simplex virus.81 Such a conjugate based on dendritic poly(lysine) scaffolding is VivaGel™, a topical agent currently under development by Starpharma Ltd., Melbourne, Australia, that can potentially prevent/reduce transmission of HIV and other sexually transmitted diseases (STDs). VivaGel™ (SPL 7013) is being offered as a water-based gel, with the purpose to prevent HIV from binding to cells in the body. The gel differs from physical barriers to STDs such as condoms, by exhibiting inhibitory activity against HIV and other STDs. In July 2003, following submission of an Investigational New Drug (IND) application, Starpharma gained clearance under U.S. FDA regulations to proceed with a Phase I clinical study to assess the safety of VivaGel™ in healthy human subjects. This phase 1 study, representing for the first time a dendrimer pharmaceutical tested in humans, compared 36 women who received either various intra-vaginal doses of VivaGel™ or a placebo gel daily for one week. The trial was double blinded so that the volunteers, principal 296 Svenson & Tomalia investigator and Starpharma did not know who was receiving placebo or VivaGel™. Study participants were assessed for possible irritant effects of the gel. Additionally, the women were assessed for any possible effect upon vaginal microflora (natural micro-organisms in the vagina) or absorption into the blood of the active ingredient of VivaGel™. A thorough review of the complete data revealed no evidence of irritation or inflammation. Preclinical development studies had demonstrated that VivaGel™ was 100% effective at preventing infection of primates exposed to a humanized strain of simian immunodeficiency virus (SHIV).82 In earlier studies, it was found that PAMAM dendrimers covalently modified with naphthyl sulfonate residues on the surface, also exhibited antiviral activity against HIV. This dendrimer-based nano-drug inhibited early stage virus/cell adsorption and later stage viral replication, by interfering with reverse transcriptase and/or integrase enzyme activities.83,84 The general mode of action of antibacterial dendrimers is to adhere to and damage the anionic bacterial membrane, causing bacterial lysis.25,85 PPI dendrimers with tertiary alkyl ammonium groups attached to the surface have been shown to be potent antibacterial biocides against Gram positive and Gram negative bacteria. The nature of the counterion is important, as tetraalkylammonium bromides were found to be more potent antibacterials over the corresponding chlorides.86 Poly(lysine) dendrimers with mannosyl surface groups are effective inhibitors of the adhesion of E. coli to horse blood cells in a haemagglutination assay, making these structures promising antibacterial agents.87 Chitosan-dendrimer hybrids have been found to be useful as antibacterial agents, carriers in drug delivery systems, and in other biomedical applications. Their behavior have been reviewed very recently88 Triazine-based antibiotics were loaded into dendrimer beads at high yields. The release of the antibiotic compounds from a single bead was sufficient to give a clear inhibition effect.89 In many cases, dendritic constructs were more potent than analogous systems based on hyperbranched polymers. The anti-prion activity of cationic phosphorus-containing dendrimers with tertiary amine surface groups has been evaluated. These molecules had a strong anti prion activity at non-toxic doses. They have been found to decrease the amount of pre-existing PrPSc from several prion starins, including the BSE strain. In addition, these dendrimers were able to reduce PrPSc accumulation in the spleen by more than 80%.90 6. Routes of Application Most commonly, dendrimers are applied as parenteral injections, either directly into the tumor tissue or intravenous for systemic delivery. However, recent oral drug delivery studies using the human colon adenocarcinoma cell line, Caco-2, Dendrimers as Nanoparticulate Drug Carriers 297 have indicated that low generation PAMAM dendrimers cross cell membranes, presumably through a combination of two processes, i.e. paracellular transport and adsorptive endocytosis.91 The P-glycoprotein (P-gp) efflux transporter does not effect dendrimers, and therefore, drug-dendrimer complexes are able to bypass the efflux transporter.92 Furthermore, recent work has shown that PAMAM dendrimers enhanced the bioavailability of indomethacin in transdermal delivery applications.93 Similarly, the drug tamsulosin was used as a model to study transdermal delivery utilizing PAMAM dendrimers. The dendrimers were found to be weak penetration enhancers.94 However, no dendrimer-driven effect was observed for the drugs ketoprofen and clonidine. As an explanation, dendrimer-triggered drug crystallization within the transdermal delivery matrix was discussed, allowing the formation of drug polymorphs that can or cannot facilitate transdermal delivery95 Several PAMAM dendrimers (generations 1.5, 2-3.5 and 4) with amine, carboxylate and hydroxyl surface groups were studied for controlled ocular drug delivery. The duration of residence time was evaluated after solubilization of these dendrimers in buffered phosphate solutions containing 2 parts per thousand (w/v) of fluorescein. The New Zealand albino rabbit was used as an in vivo model for qualitative and quantitative assessment of ocular tolerance and retention time, after a single application of 25 /xL of dendrimer solution to the eye. The same model was also used to determine the prolonged miotic or mydriatic activities of dendrimer solutions, some containing pilocarpine nitrate and some tropicamide, respectively. Residence time was longer for the solutions containing dendrimers with carboxylic and hydroxyl surface groups. No prolongation of remanence time was observed when dendrimer concentration (0.25-2%) increased. The remanence time of PAMAM dendrimer solutions on the cornea showed size and molecular weight dependency. This study allowed novel macromolecular carriers to be designed with prolonged drug residence time for the ophthalmic route.96 7. Biocompatibility of Dendrimers Dendrimers have to exhibit low toxicity and be non-immunogenic in order to be widely used in biomedical applications. To date, the cytotoxicity of dendrimers has been primarily studied in vitro, however, a few in vivo studies have been published.25 As observed for other cationic macromolecules, including liposomes and micelles, dendrimers with positively charged surface groups are prone to destabilize cell membranes and cause cell lysis. For example, in vitro cytotoxicity IC50 measurements (i.e. the concentration where 50% of cell lysis is observed) for aminoterminated PAMAM dendrimers revealed significant cytotoxicity on human intestinal adenocarcinoma Caco-2 cells.97,98 Furthermore, the cytotoxicity was found to be 298 Svenson & Tomalia generation-dependent, with higher generation dendrimers being the most toxic. ' A similar generation dependence of amino-terminated PAMAM dendrimers was observed for the haemolytic effect, studied on a solution of rat blood cells.100 However, some recent studies have shown that amino-terminated PAMAM dendrimers exhibit lower toxicity than more flexible amino-functionalized linear polymers perhaps due to lower adherence of the rigid globular dendrimers to cellular surfaces. The degree of substitution, as well as the type of amine functionality, is important, with primary amines being more toxic than secondary or tertiary amines." Amino-terminated PPI and PAMAM dendrimers behave similarly with regard to cytotoxicity and haemolytic effects, including the generation-dependent increase of both.100'101 Comparative toxicity studies on anionic (carboxylate-terminated) and cationic (amino-terminated) PAMAM dendrimers using Caco-2 cells have shown a significantly lower cytotoxicity of the anionic compounds.97 In fact, lower generation PAMAM dendrimers possessing carboxylate surface groups show neither haematotoxicity nor cytotoxicity at concentrations up to 2 mg/ml.100 The biocompatability of dendrimers is not solely determined by the surface groups. Dendrimers containing an aromatic polyether core and anionic carboxylate surface groups have shown to be haemolytic on a solution of rat blood cells after 24hrs. It is suggested that the aromatic interior of the dendrimer may cause haemolysis through hydrophobic membrane contact.100 One way to reduce the cytotoxicity of cationic dendrimers may reside in partial surface derivatization with chemically inert functionalities such as PEG or fatty acids. The cytotoxicity towards Caco-2 cells can be reduced significantly (from IC50 ~ 0.13 mM to >lmM) after such a modification. This observation can be explained by the reduced overall positive charge of these surface-modified dendrimers. Apartial derivatization with as few as six lipid chains or four PEG chains on a G4-PAMAM, respectively, was sufficient to lower the cytotoxicity substantially.98 In studies conducted at Dendritic Nano Technologies, Inc. using Caco-2 and two other cell lines, it was found that besides (partial) PEGylation of the surface, surface modification with pyrrolidone, another biocompatible compound, can significantly reduce cytotoxicity to levels far better than those of currently available products.102 In some cases, the cytotoxicity of PAMAM dendrimers could be reduced by additives such as fetal calf serum.103 Only a few systematic studies on the in vivo toxicity of dendrimers have been reported so far. Upon injection into mice, doses of 10 mg/kg of PAMAM dendrimers (up to G = 5), displaying either unmodified or modified amino-terminated surfaces, did not appear to be toxic.81-104 Hydroxy- or methoxy-terminated dendrimers based on a polyester dendrimer scaffold have been shown to be of low toxicity both in vitro and in vivo. At very high concentrations (40 mg/ml), these polyester Dendrimers as Nanoparticulate Drug Carriers 299 dendrimers induced some inhibition of cell growth in vitro, but no increase in cell death was observed. Upon injection into mice, no acute or long-term toxicity problems were observed. The non-toxic properties make these new dendritic motifs very promising candidates for drug delivery devices.49 Initial immunogenicity studies performed on unmodified amino-terminated PAMAM dendrimers showed no or weak immunogenicity of the G3-G7 dendrimers. However, later studies indicated some immunogenicity of these dendrimers, which could be reduced by surface-modification utilizing PEG chains.105 8. Conclusions The high level of control over the architecture of dendrimers, their size, shape, branching length and density, and their surface functionality, makes these compounds ideal carriers in drug delivery applications. The bioactive agents may either be encapsulated into the interior of the dendrimers or they may be chemically attached or physically adsorbed onto the dendrimer surface, with the option to tailor the properties of the carrier to the specific needs of the active material and its therapeutic applications. Furthermore, the high density of surface groups allows attachment of targeting groups as well as groups that modify the solution behavior or toxicity of dendrimers. Surface-modified dendrimers themselves may act as nano-drugs against tumors, bacteria and viruses. This review of drug delivery applications of dendrimers clearly illustrates the potential of this new "fourth architectural class of polymers"106 and substantiates the high optimism for the future of dendrimers in this important field. Acknowledgments The authors wish to thank all contributors to this fascinating field of research, as well as the funding agents that have supported this work over the years. In particular, DNT would like to acknowledge current funding by the US Army Research Laboratory (ARL) (Contract # W911NF-04-2-0030). References 1. Svenson S (2004) Controlling surfactant self-assembly. Curr Opin Coll Interj Sci 9: 201-212. 2. Svenson S (2004) Self-assembly and self-organization: Important processes — but can we predict them? / Dispersion Sci Technol 25:101-118. 3. Svenson S (ed.) (2004) Carrier-based Drug Delivery Vol. 879. ACS Symposium Series, American Chemical Society, Washington, DC. 4. (a) Tomalia DA (2004) Birth of a new macromolecular architecture: Dendrimers as quantized building blocks for nanoscale synthetic organic chemistry. Aldrichimica Acta 37: 300 Svenson & Tomalia 39-57; (b) Tomalia DA (2005) Birth of a new macromolecular architecture: Dendrimers as quantized building blocks for nanoscale synthetic polymer chemistry. Prog Polym Sci 30:294-324; (c) Tomalia DA (2005) The dendritic state. Materials Today March: 34-46; (d) Tomalia DA (2005) Dendrimeric supramolecular and supramacromolecular assemblies, in Supramolecular Polymers, 2nd Ed., CRC Press, Taylor & Francis, Boca Raton, FL. 5. Tomalia DA and Frechet JMJ (eds.) (2001) Dendrimers and Other Dendritic Polymers. J. Wiley & Sons Ltd., Chichester. 6. (a) Watkins DM, Sayed-Sweet Y, Klimash JW, Turro NJ and Tomalia DA (1997) Dendrimers with hydrophobic cores and the formation of supramolecular dendrimer — surfactant assemblies. Langmuir 13:3136-3141; (b) Sayed-Sweet Y, Hedstrand DM, Spinder R and Tomalia DA (1997) Hydrophobically modified poly(amidoamine) (PAMAM) dendrimers: Their properties at the air-water interface and use as nanoscopic container molecules. / Mater Chem 7:1199-1205. 7. Recker J, Tomcik DJ and Parquette JR (2000) Folding dendrons: The development of solvent-, temperature-, and generation-dependent chiral conformational order in intramolecularly hydrogen-bonded dendrons. J Am Chem Soc 122:10298-10307. 8. Boas U, Karlsson AJ, de Waal BFM and Meijer EW (2001) Synthesis and properties of new thiourea-functionalized poly(propylene imine) dendrimers and their role as hosts for urea functionalized guests. / Org Chem 66:2136-2145. 9. Tomalia DA, Naylor AM and Goddard III WA (1990) Starburst dendrimers: Molecular level control of size, shape, surface chemistry, topology and flexibility from atoms to macroscopic matter. Angew Chem Int Ed Engl 29:138-175. 10. Naylor AM, Goddard III WA, Kiefer GE and Tomalia DA (1989) Starburst dendrimers 5. Molecular shape control. } Am Chem Soc 111:2339-2341. 11. Jansen JFGA, Debrabandervandenberg EMM and Meijer EW (1994) Encapsulation of guest molecules into a dendritic box. Science 266:1226-1229. 12. Jansen JFGA, Meijer EW and Debrabandervandenberg EMM (1995) The dendritic box: Shape-selective liberation of encapsulated guests. J Am Chem Soc 117:4417-^1418. 13. (a) Esfand R and Tomalia DA (2001) Poly(amidoamine) (PAMAM) dendrimers: From biomimicry to drug delivery and biomedical applications. Drug Disc Today 6:427-436; (b) Tomalia DA, Hall M, Hedstrand (1987) STARBURST® Dendrimers III. The importance of branch junction symmetry in the development of topological shell molecules. J Am Chem Soc 109:1601-1603. 14. Twyman LJ, Beezer AE, Esfand R, Hardy MJ and Mitchell JC (1999) The synthesis of water-soluble dendrimers, and their application as possible drug delivery systems. Tetrahedron Lett 40:1743-1746. 15. Gorman CB and Smith JC (2001) Structure-property relationships in dendritic encapsulation. Ace Chem Res 34:60-71. 16. Chasse TL, Yohannan JC, Kim N, Li Q, Li Z and Gorman CB (2003) Dendritic encapsulation-roles of cores and branches. Tetrahedron 59:3853-3861. 17. Chasse TL, Sachdeva R, Li Q, Li Z, Petrie RJ and Gorman CB (2003) Structural effects on encapsulation as probed in redox-active core dendrimer isomers. / Am Chem Soc 125:8250-8254. Dendrimers as Nanoparticulate Drug Carriers 301 18. Wallimann P, Marti T, Furer A and Diederich F (1997) Steroids in molecular recognition. Chem Rev 97:1567-1608. 19. Wallimann P, Seiler P and Diederich F (1996) Dendrophanes: Novel steroid-recognizing dendritic receptors. Helv Chim Acta 79:779-788. 20. Smith DK, Zingg A and Diederich F (1999) Dendroclefts: Optically active dendritic receptors for the selective recognition and chiroptical sensing of monosaccharide guests. Helv Chim Acta 82:1225-1241. 21. Smith DK and Diederich F (1998) Dendritic hydrogen bonding receptors: Enantiomerically pure dendroclefts for the selective recognition of monosaccharides. Chem Commun 2501-2502. 22. Zimmerman SC, Wendland MS, Rakow NA, Zharov I and Suslick KS (2002) Synthetic hosts by monomolecular imprinting inside dendrimers. Nature 418: 399-403. 23. Broeren MAC, van Dongen JLJ, Pittelkow M, Christensen JB, van Genderen MHP and Meijer EW (2004) Multivalency in the gas phase: The study of dendritic aggregates by mass spectrometry. Angew Chem hit Ed 43:3557-3562. 24. Maeda H and Matsumura Y (1986) A new concept in macromolecular therapeutics in cancer chemotherapy: Mechanism of tumoritropic accumulation of proteins and antitumor agent SMANCS. Cancer Res 46:6387-9392. 25. Boas U and Heegaard PMH (2004) Dendrimers in drug research. Chem Soc Rev 33: 43-63. 26. Frechet JMJ (2000) Designing dendrimers for drug delivery. Pharm Sci Technol Today 2:393-401. 27. Gillies ER and Frechet JMJ (2005) Dendrimers and dendritic polymers in drug delivery. Drug Disc Today 10:35^2. 28. Malik N, Evagorou EG and Duncan R (1999) Dendrimer-platinate: A novel approach to cancer chemotherapy. Anticancer Drugs 10:767-776. 29. (a) Malik N and Duncan R (2003) Dendritic-platinate drug delivery system. US 6585956; (b) Malik N and Duncan R (2004) Method of treating cancerous tumors with a dendritic-platinate drug delivery system. US 6 790 437; (c) Malik N, Duncan R, TomaliaDA, Esfand R (2006) Dendritic-antineoplastic drug delivery system. US 7005124. Patents owned by Dendritic Nanotechnologies, Inc. (DNT). 30. Kelland LR and Farrell NP (eds.) (2000) Platinum-Based Drugs in Cancer Chemotherapy. Humana Press: Totowa, NJ. 31. Gianasi E, Wasil M, Evagorou EG, Keddle A, Wilson G and Duncan R (1999) HPMA copolymer platinates as novel antitumor agents: In vitro properties, pharmacokinetics and antitumor activity in vivo. Eur J Cancer 35:994-1002. 32. Balogh L, Swanson DR, Tomalia DA, Hagnauer GL and McManus AT (2001) Dendrimersilver complexes and nanocomposites as antimicrobial agents. Nano Lett 1:18-21. 33. Kojima C, Kono K, Maruyama K and Takagishi T (2000) Synthesis of polyamidoamine dendrimers having poly(ethylene glycol) grafts and their ability to encapsulate anticancer drugs. Bioconjug Chem 11:910-917. 34. Pan GF, Lemmouchi Y, Akala EO and Bakare O (2005) Studies on PEGylated and drugloaded PAMAM dendrimers. / Bioactive Compat Polym 20:113-128. 302 Svenson & Tom all a 35. Bhadra D, Bhadra S, Jain S and Jain NK (2003) A PEGylated dendritic nanoparticulate carrier of fluorouracil. Int J Pharm 257:111-124. 36. Wang F, Bronich TK, Kabanov AV, Rauh RD and Roovers J (2005) Synthesis and evaluation of a star amphiphilic block copolymer from poly (e-caprolactone) and poly(ethylene glycol) as a potential drug delivery carrier. Bioconjug Chem 16:397-405. 37. Namazi H and Adell M (2005) Dendrimers of citric acid and poly(ethylene glycol) as the new drug delivery agents. Biomaterials 26:1175-1183. 38. Yang H and Lopina ST (2005) Extended release of a novel antidepressant, venlafaxine, based on anionic poly(amidoamine) dendrimers and poly(ethylene glycol)-containing semi-interpenetrating networks. / Biomed Mater Res Part A 72A:107-114. 39. Kolhe P, Misra E, Kannan RM, Kannan S and Lieh-Lai M (2003) Drug complexation, in vitro release and cellular entry of dendrimers and hyperbranched polymers. Int J Pharm 259:143-160. 40. Kannan S, Kolhe P, Raykova V, Glibatec M, Kannan RM, Lieh-Lai M and Bassett D (2004) Dynamics of cellular entry and drug delivery by dendritic polymers into human lung epithelial carcinoma cells. / Biomater Sci Polym Ed 15:311-330. 41. Chauhan AS, Jain NK, Diwan PV and Khopade AJ (2004) Solubility enhancement of indomethacin with poly(amidoamine) dendrimers and targeting to inflammatory regions of arthritic rats. / Drug Targ 12:575-583. 42. Kubasiak LA, Chauhan AS and Tomalia DA (2005) Manuscript in preparation. 43. Devarakonda B, Hill RA and de Villiers MM (2004) The effect of PAMAM dendrimer generation size and surface functional group on the aqueous solubility of nifedipine. Int]Pharm 284:133-140. 44. Potluri SK, Ramulu AR and Pardhasaradhi M (2004) Synthesis of new unsymmetrical optically active (s)-(+)-naproxen. Tetrahedron 60:10915-10920. 45. Ooya T, Lee J and Park K (2004) Hydrotropic dendrimers of generations 4 and 5: Synthesis, characterization, and hydrotropic solubilization of paclitaxel. Bioconjug Chem 15:1221-1229. 46. Khandare J, Kolhe P, Pillai O, Kannan S, Lieh-Lai M and Kannan RM (2005) Synthesis, cellular transport, and activity of poly(amidoamine) dendrimer-methylprednisolone conjugates. Bioconjug Chem 16:330-337. 47. Shamis M, Lode HN and Shabat D (2004) Bioactivation of self-immolative dendritic prodrugs by catalytic antibody 38C2. / Am Chem Soc 126:1726-1731. 48. Haba K, Popkov M, Shamis M, Lerner RA, Barbas III CF and Shabat D (2005) Singletriggered trimeric prodrugs. Angew Chem Int Ed 44:716-720. 49. Padilla De Jesus OL, Ihre HR, Gagne L, Frechet JMJ and Szoka Jr. FC (2002) Polyester dendritic systems for drug delivery applications: In vitro and in vivo evaluation. Bioconjug Chem 13:453-461. 50. Nishiyama N, Stapert HR, Zhang GD, Takasu D, Jiang DL, Nagano T, Aida T and Kataoka K (2003) Light-harvesting ionic dendrimer porphyrins as new photosensitizers for photodynamic therapy. Bioconjug Chem 14:58-66. 51. Zhang GD, Harada A, Nishiyama N, Jiang DL, Koyama H, Aida T and Kataoka K (2003) Polyion complex micelles entrapping cationic dendrimer porphyrin: Effective photosensitizer for photodynamic therapy of cancer. / Control Rel 93:141-150. Denclrimers as Nanoparticulate Drug Carriers 303 52. Jang WD, Nishiyama N, Zhang GD, Harada A, Jiang DL, Kawauchi S, Morimoto Y, Kikuchi M, Koyana H, Aida T and Kataoka K (2005) Supramolecular nanocarrier of anionic dendrimer porphyrins with cationic block copolymers modified with poly(ethylene glycol) to enhance intracellular photodynamic efficacy. Angew Chem Int Ed 44:419^123. 53. Battah SH, Chee CE, akanishi H, Gerscher S, MacRobert AJ and Edwards C (2001) Synthesis and biological studies of 5-aminolevulinic acid-containing dendrimers for photodynamic therapy. Bioconjug Chem 12:980-988. 54. Paul A, Hackbarth S, Molich A, Luban C, Oelckers S, Bohm F and Roder B (2003) Comparative study of the photosensitization of Jurkat cells in vitro by pheophorbide a and a pheophorbide a-diaminobutane poly(propylene imine) dendrimer complex. Laser Phys 13:22-29. 55. Wu G, Barth RF, Yang WL, Chatterjee M, Tjarks W, Ciesielski MJ and Fenstermaker RA (2004) Site-specific conjugation of boron-containing dendrimers to anti-EGF receptor monoclonal antibody cetuximab (IMC-C225) and its evaluation as a potential delivery agent for neutron capture therapy. Bioconjug Chem 15:185-194. 56. Shukla S, Wu G, Chatterjee M, Yang WL, Sekido M, Diop LA, Muller R, Sudimack JJ, Lee RJ, Barth RF and Tjarks W (2003) Synthesis and biological evaluation of folate receptor-targeted boronated PAMAM dendrimers as potential agents for neutron capture therapy. Bioconjug Chem 14:158-167. 57. Kono K, Liu M and Frechet JMJ (1999) Design of dendritic macromolecules containing folate or methotrexate residues. Bioconjug Chem 10:1115-1121. 58. Quintana A, Raczka E, Piehler L, Lee I, Myc A, Majoros I, Patri AK, Thomas T, Mule J and Baker Jr. JR (2002) Design and function of a dendrimer-based therapeutic nanodevice targeted to tumor cells through the folate receptor. Pharm Res 19:1310-1316. 59. Ross JF, Chaudhuri PK and Ratnam M (1994) Differential regulation of folate receptor isoforms in normal and malignant tissues in vivo and established cell lines. Physiologic and clinical implications. Cancer 73:2432-2443. 60. Tansey W, Ke S, Cao XY, Pasuelo MJ, Wallace S and Li C (2004) Synthesis and characterization of branched poly(L-glutamic acid) as a biodegradable drug carrier. / Control Rel 94:39-51. 61. Lundquist JJ and Toone EJ (2002) The cluster glycoside effect. Chem Rev 102:555-578. 62. Zanini D and Roy R (1997) Synthesis of new a-thiosialodendrimers and their binding properties to the sialic acid specific lectin from Limax flavus. / Am Chem Soc 119:2088- 2095. 63. Bezouska K, Pospisil MF, Vannucci LF, Fiserova AF, Krausova KF, Horvath OF, Kren VF, Mosca FF, Lindhorst TK, Sadalapure KF and Bezouska K (2002) Design, functional evaluation and biomedical applications of carbohydrate dendrimers (glycodendrimers). Rev Mol Biotechnol 90:269-290. 64. Roy R (1996) Syntheses and some applications of chemically defined multivalent glycoconjugates. Curr Ovin Struct Biol 6:692-702. 65. Lindhorst TK (2002) Artificial multivalent sugar ligands to understand and manipulate carbohydrate-protein interactions. Top Curr Chem Host-Guest Chem 218:201-235. 304 Svenson & Tomalia 66. Rockendorf N and Lindhorst TK (2001) Glycodendrimers. Top Curr Chem Dend IV 217:201-238. 67. Veprek P and Jezek J (1999) Peptide and glycopeptide dendrimers. Part II. / Pept Sci 5:203-220. 68. Andre S, Pieters RJ, Vrasidas I, Kaltner H, Kuwabara I, Liu FT, Liskamp RM and Gabius HJ (2001) Wedgelike glycodendrimers as inhibitors of binding of mammalian galectins to glycoproteins, lactose maxiclusters, and cell surface glycoconjugates. ChemBioChem 2:822-830. 69. Pieters RJ (2004) Interference with lectin binding and bacterial adhesion by multivalent carbohydrates and peptidic carbohydrate mimics. Trends Glycosci Glycotechnol 16: 243-254. 70. Baek MG and Roy R (2002) Synthesis and protein binding properties of T-antigen containing GlycoPAMAM dendrimers. Bioorg Med Chem 10:11-17. 71. Roy R, Baek MG and Rittenhouse-Olson K (2001) Synthesis of N,N'- bis(acrylamido)acetic acid-based T-antigen glycodendrimers and their mouse monoclonal IgG antibody binding properties. / Am Chem Soc 123:1809-1816. 72. Roy R and Baek MG (2002) Glycodendrimers: Novel glycotope isosteres unmasking sugar coating. Case study with T-antigen markers from breast cancer MUC1 glycoprotein. Rev Mol Biotechnol 90:291-309. 73. Benito JM, Gomez-Garcia M, Mellet CO, Baussanne I, Defaye J and Fernandez JMG (2004) Optimizing saccharide-directed molecular delivery to biological receptors: Design, synthesis, and biological evaluation of glycodendrimer-cyclodextrin conjugates. J Am Chem Soc 126:10355-10363. 74. Hansen HC, Haataja S, Finne J and Magnusson G (1997) Di-, tri-, and tetravalent dendritic galabiosides that inhibit hemagglutination by Streptococcus suis at nanomolar concentration. / Am Chem Soc 119:6974-6979. 75. Rendle PM, Seger A, Rodrigues J, Oldham NJ, Bott RR, Jones JB, Cowan MM and Davies BG (2004) Glycodendriproteins: A synthetic glycoprotein mimic enzyme with branched sugar-display potently inhibits bacterial aggregation. / Am Chem Soc 126:4750^1751. 76. Shaunak S, Thomas S, Gianasi E, Godwin A, Jones E, Teo I, Mireskandari K, Luthert P, Duncan R, Patterson S, Khaw P and Brocchini S (2004) Polyvalent dendrimer glucosamine conjugates prevent scar tissue formation. Nat Biotech 22:977-984. 77. Thomas TP, Patri AK, Myc A, Myaing MT, Ye JY, Norris TB and Baker Jr JR (2004) In vitro targeting of synthesized antibody-conjugated dendrimer nanoparticles. Biomacromol 5:2269-2274. 78. Hong MY, Yoon HC and Kim HS (2003) Protein-ligand interactions at poly (amidoamine) dendrimer monolayers on gold. Langmuir 19:416-421. 79. Sanchez-Sancho F, Perez-Inestrosa E, Suau R, Mayorga C, Torres MJ and Blanca M (2002) Dendrimers as carrier protein mimetics for IgE antibody recognition. Synthesis and characterization of densely penicilloylated dendrimers. Bioconjug Chem 13:647-653. 80. Yang H and Lopina ST (2003) Penicillin V-conjugated PEG-PAMAM star polymers. / Biomater Sci-Polym Ed 14:1043-1056. Dendrimers as Nanoparticulate Drug Carriers 305 81. Bourne N, Stanberry LR, Kern ER, Holan G, Matthews B and Bernstein DI (2000) Dendrimers, a new class of candidate topical microbicides with activity against herpes simplex virus infection. Antimicrob Agents Chemother 44:2471-2474. 82. Product Focus: VivaGel™, Starpharma Limited, Melbourne, Australia. 83. Gong Y, Matthews B, Cheung D, Tarn T, Gadawski I, Leung D, Holan G, Raff J and Sacks S (2002) Evidence of dual sites of action of dendrimers: SPL-2999 inhibits both virus entry and late stages of herpes simplex virus replication. Antiviral Res 55:319-329. 84. Witvrouw M, Fikkert V, Pluymers W, Matthews B, Mardel K, Schols D, Raff J, Debyser Z, DeClercq E, Holan G and Pannecouque C (2000) Polyanionic (i.e. polysulfonate) dendrimers can inhibit the replication of human immunodeficiency virus by interfering with both virus adsorption and later steps (Reverse transcriptase/integrase) in the virus replicative cycle. Mol Pharmacol 58:1100-1108. 85. Chen CZ and Cooper SL (2002) Interactions between dendrimer biocides and bacterial membranes. Biomaterials 23:3359-3368. 86. Chen CZ, Beck-Tan NC, Dhurjati P, van Dyk TK, LaRossa Ra and Cooper SL (2000) Quaternary ammonium functionalized poly(propylene imine) dendrimers as effective antimicrobials: Structure-activity studies. Biomacromol 1:473^80. 87. Nagahori N, Lee RT, Nishimura S, Page D, Roy R and Lee YC (2002) Inhibition of adhesion of type 1 fimbriated Escherichia coli to highly mannosylated ligands. Chem- BioChem 3:836-844. 88. Sashiwa H and Aiba SI (2004) Chemically modified chitin and chitosan as biomaterials. Prog Polymer Sci 29:887-908. 89. Lebreton S, Newcombe N and Bradley M (2003) Antibacterial single-bead screening. Tetrahedron 59:10213-10222. 90. Solassol J, Crozet C, Perrier V, Leclaire J, Beranger F, Caminade AM, Meunier B, Dormont D, Majoral JP and Lehmann S (2004) Cationic phosphorus-containing dendrimers reduce prion replication both in cell culture and in mice infected with scapie. / Gen Virol 85:1791-1799. 91. El-Sayed M, Rhodes CA, Ginski M and Ghandehari H (2003) Transport mechanism(s) of poly(amidoamine) dendrimers across Caco-2 cell monolayers. IntJ Pharm 265:151-157. 92. D'Emanuele A, Jevprasesphant R, Penny J and Attwood D (2004) The use of a dendrimer-propanolol prodrug to bypass efflux transporters and enhance oral bioavailability. / Control Rel 95:447-453. 93. Chauhan AS, Sridevi S, Chalasani KB, Jain AK, Jain SK, Jain NK and Diwan PV (2003) Dendrimer-mediated transdermal delivery: Enhanced bioavailability of indomethacin. / Control Rel 90:335-343. 94. Wang ZX, Itoh YS, Hosaka Y, Kobayashi I, Nakano Y, Maeda I, Umeda F, Yamakawa J, Kawase M and Yagi K (2003) Novel transdermal drug delivery system with polyhydroxyalkanoate and starburst poly(amidoamine) dendrimer. / Biosci Bioeng 95:541-543. 95. Wang ZX, Itoh YS, Hosaka Y, Kobayashi I, Nakano Y, Maeda I, Umeda F, Yamakawa J, Nishimine M, Suenobu T, Fukuzumi S, Kawase M and Yagi K (2003) Mechanism of 306 Svenson & Tomalia enhancement effect of dendrimer on transdermal drug permeation through polyhydroxyalkanoate matrix. / Biosci Bioeng 96:537-540. 96. Vandamme TF and Brobeck L (2005) Poly(amidoamine) dendrimers as ophthalmic vehicles for ocular delivery of pilocarpine nitrate and tropicamide. / Control Rel 102: 23-38. 97. Jevprasesphant R, Penny J, Jalal R, Attwood D, McKeown NB and D'Emanuele A (2003) The influence of surface modification on the cytotoxicity of PAMAM dendrimers. Int J Pharm 252:263-266. 98. El-Sayed M, Ginski M, Rhodes C and Ghandehari H (2002) Transepithelial transport of poly(amidoamine) dendrimers across Caco-2 cell monolayers. / Control Rel 81:355-365. 99. Fischer D, Li Y, Ahlemeyer B, Krieglstein J and Kissel T (2003) In vitro cytotoxicity testing of polycations: Influence of polymer structure on cell viability and hemolysis. Biomaterials 24:1121-1131. 100. Malik N, Wiwattanapatapee R, Klopsch R, Lorenz K, Frey H, Weener JW, Meijer EW, Paulus W and Duncan R (2000) Dendrimers: Relationship between structure and biocompatibility in vitro, and preliminary studies on the biodistribution of I-125-labelled poly(amidoamine) dendrimers in vivo.} Control Rel 65:133-148. 101. Zinselmeyer BH, Mackay SP, Schatzlein AG and Uchegbu IF (2002) The lowergeneration polypropylenimine dendrimers are effective gene-transfer agents. Pharm Res 19:960-967. 102. Kubasiak LA and Tomalia DA (2005) Manuscript in preparation. 103. Yoo H and Juliano RL (2000) Enhanced delivery of antisense oligonucleotides with fluorophore-conjugated PAMAM dendrimers. Nucleic Acids Res 28:4225-4231. 104. Roberts JC, Bhalgat MK and Zera RT (1996) Preliminary biological evaluation of poly(amidoamine) (PAMAM) starburst dendrimers. / Biomed Mater Res 30:53-65. 105. Kobayashi H, Kawamoto S, Saga T, Sato N, Hiraga A, Ishimori T, KonishiJ, Togashi K and Brechbiel MW (2001) Positive effects of polyethylene glycol conjugation to generation-4 polyamidoamine dendrimers as macromolecular MR contrast agents. Magn Reson Med 46:781-788. 106. Tomalia DA and Frechet JMJ (2002) Discovery of dendrimers and dendritic polymers: A brief historical perspective. / Polym Sci Part A: Polym Chem 40:2719-2728. 14 Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs Rainer H. Muller and Jens-Uwe A. H. Junghanns 1. Introduction Since the last ten years, the number of poorly soluble drugs is steadily increasing. According to estimates, about 40% of the drugs in the pipelines have solubility problems.1 The increased use of high throughput screening methods leads to the discovery of more drugs being poorly water soluble. In the literature, figures are quoted that about 60 percent of the drugs coming directly from synthesis are nowadays poorly soluble.2 Poor solubility is not only a problem for the formulation development and clinical testing, it is also an obstacle at the very beginning when screening new compounds for pharmacological activity. From this, there is a definite need for smart technological formulation approaches to make such poorly soluble drugs bioavailable. Making such drugs bioavailable means that they show sufficiently high absorption after oral administration, or they can alternatively be injected intravenously. There is quite a number of formulation approaches for poorly soluble drugs which can be specified as "specific approaches". These approaches are suitable for molecules having special properties with regard to their chemistry (e.g. solubility in certain organic media) or to the molecular size or conformation (e.g. molecules to be incorporated into the cyclodextrin ring structure). Of course it would be much smarter to have a "universal formulation approach" applicable to any molecule. Such a universal formulation approach to increase the oral 307 308 Muller & Junghanns bioavailability is micronization, meaning the transfer of drug powders into the size range between typically 1-10 /j-va. However, nowadays many drugs are so poorly soluble that micronization is not sufficient. The increase in surface area, and thus consequently in dissolution velocity, is not sufficient to overcome the bioavailability problems of very poorly soluble drugs of the biopharmaceutical specification class II. A consequent next step was to move from micronization to nanonization. Since the beginning of the 90s, the company Nanosystems propagated the use of nanocrystals (instead of microcrystals) for oral bioavailability enhancement, and also to use nanocrystals suspended in water (nanosuspensions) for intravenous or pulmonary drug delivery. The solution was simple; in general, simple solutions possess the smartness that they can be realized easier than complex systems and introduction to the market is faster. Nevertheless, it took about ten years before the first nanocrystals in a tablet appeared on the market, the product Rapamune® by the company Wyeth in 2000. Compared with liposomes developed in 19683 with the first products on the market around 1990 (e.g. Alveofact®, a lung surfactant), this was still relatively fast. What were the reasons that it took about one decade for nanocrystals to enter the market? From our point of view, pharmaceutical companies prefer to use formulation technology already established with know how available in the company. In addition, if formulation technologies are established, a company also has the possibility for production of the final product. Therefore, all the traditional formulation approaches were exploited to solve a formulation problem. In addition, formulation approaches were preferred, being even simpler than nanocrystals. For example, production of drug-containing microemulsions administered in a capsule is, in many cases, even simpler. Another reason for the reluctance of pharmaceutical companies at the beginning was the lack of large scale production methods. These were not available at the very beginning of the development of the nanocrystal technology. Meanwhile, this has changed and the major pharmaceutical companies try to secure or have already secured their access to nanocrystal technology. Access to nanocrystal technology is possible either by licencing in or alternatively by the attempt to develop one's own production technologies for the nanocrystals, which do not depend on already existing intelectual property (IP). This chapter discusses the physicochemical properties of nanocrystals which make them interesting for drug delivery, reviews and discusses briefly the various production methods available and highlights the opportunities for improved drug delivery using different application routes. 2. Definitions Drug nanocrystals are crystals with a size in the nanometer range, meaning that they are nanoparticles with a crystalline character. There are discussions about the Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 309 definition of a nanoparticle, referring to the size of a particle to be classified as a nanoparticle. Depending on the discipline, e.g. in colloid chemistry, particles are only considered as nanoparticles when they are in sizes below 100 nm or even below 20 nm. Based on the size unit, in the pharmaceutical area, nanoparticles should be defined as having a size between a few nanometers and 1000 nm (1 /im); thus, microparticles possess consequently a size 1-1000 micrometer. A further characteristic is that drug nanocrystals are composed of 100% drug; there is no carrier material as in polymeric nanoparticles. Dispersion of drug nanocrystals in liquid media leads to "nanosuspensions", in contrast to "microsuspensions" or "macrosuspensions". In general, the dispersed particles need to be stabilized, e.g. by surfactants or polymeric stabilizers. Dispersion media can be water, aqueous solutions or non-aqueous media [e.g. liquid polyethylene glycol (PEG), oils]. Depending on the production technology, processing of drug microcrystals to drug nanoparticles can lead to either a crystalline or to an amorphous product, especially when applying precipitation. In the strict sense, such an amorphous drug nanoparticle should not be called nanocrystal. However, one often refers to "nanocrystals in the amorphous state". 3. Physicochemical Properties of Drug Nanocrystals 3.1. Change of dissolution velocity The reason for micronization is to increase the surface area, thus consequently according to the Noyes-Whitney equation, increasing the dissolution velocity. Therefore, micronization can be succesfuUy employed if the dissolution velocity is the rate-limiting step for oral absorption (drugs of BSC II). Of course, by moving one dimension further to smaller particles, the surface area is further enlarged and consequently, the dissolution velocity is further enhanced. In most cases, a low dissolution velocity is correlated with a low saturation solubility. 3.2. Saturation solubility The general textbook statement is that the saturation solubility cs is a constant depending on the compound, the dissolution medium and the temperature. This is valid for powders of daily life with a size in the micrometer range or above. However, below a critical size of 1-2 /zm, the saturation solubility is also a function of the particle size. It increases with decreasing particle size below 1000 nm. Therefore, drug nanocrystals possess an increased saturation solubility. This has two advantages: 1. According to Noyes-Whitney, the dissolution velocity is further enhanced because dc/dt is proportional to the concentration gradient (cs — cx)/h (cx — bulk concentration, h — diffusional distance). 310 Muller & Junghanns 2. Due to the increased saturation solubility, the concentration gradient between gut lumen and blood is increased, consequently, the absorption by passive diffusion. The interesting question very often asked is "How manyfold is the increased saturation solubility?". Data published in the literature or available to us from discussions range from 2-14 fold. What are the factors affecting the increase in saturation solubility? The factors can be identified when looking at the theoretical background. The Kelvin equation describes the increase in the vapor pressure of droplets in a gas medium as a function of their particle size, i.e. as a function of their curvature: , P -y*VL*cos6 [pj~ rK*RT Fig. 1. The Kelvin equation. P = vapor pressure Po = equilibrium pressure of a flat liquid surface y = surface tension VL = molar volume cos(#) = contact angle rK = radius of droplet R = universal gas constant T = absolute temperature (K) The vapor pressure increases with increasing curvature of the surface, that means decreasing particle size. Each liquid has its compound specific vapor pressure, thus the increase in vapor pressure will be influenced by the available compoundspecific vapor pressure. The situation of a transfer of molecules from a liquid phase (droplet) to a qas phase is in principal identical to the transfer of molecules from a solid phase (nanocrystal) to a liquid phase (dispersion medium). The vapor pressure is equivalent to the dissolution pressure. In the state of saturation solubility, there is an equilibrium of molecules dissolving and molecules recrystallizing. This equilibrium can be shifted in case the dissolution pressure increases, thus increasing the saturation solubility. Identical to liquids with different vapor pressures under normal conditions (micrometer droplet size), each drug crystal has a specific dissolution pressure in micrometer size. Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 311 Relative vapor pressure at 293 K CD Droplet size [u,m] Fig. 2. Comparison of the relative increase in vapor pressure between water, ether and oleic acid (calculated using the Kelvin equation) as a function of the droplet size (with permission after4). The important question is how the dissolution pressure changes, depending on the specific dissolution pressure of each compound and on the particle size. Model calculations were performed applying the Kelvin equation to compounds with different vapor pressures (droplets) as a function of droplet size (Fig. 2). Liquids with low medium and high vapor pressure were selected, such as oleic acid as an oil, water and ether. The important result for a drug formulation was: 1. The increase in vapor pressure is more pronounced for compounds having a priori a low vapor pressure. Applied to solid compounds, increase in dissolution pressure will be more pronounced for compounds having a priori a low dissolution pressure, i.e. the relative increase is highest for poorly soluble drugs. 2. The increase in vapor pressure is exponential, with a very pronounced increase occurring at droplet sizes below 100 nm. Figure 3 shows a calculated increase for barium sulfate as solid model compound. 3.3. Does size really matter? Transferring this to drug nanocrystals means that really smart crystals with highest increase in saturation solubility should have a size of e.g. 50 nm or 20-30 nm. From this, it can be concluded that the slogan "size matters" is correct regarding the increase in saturation solubility, and consequently, the increase in dissolution 312 Muller & Junghanns Saturation solubility of BaS04 in water at 293 K 1,2-, C^ 1,0- 1 0,8- o en c 0,6- CO jjj 0,4- o c § 0,2- +3 J2 K 0,0- - 0 , 2 - . BaS04 properties: \ M = 233.40 g/mol \ p = 4.50 g/cm3 \ cs = 2.22 mg/L \ o = 26.7mN/m -> i- i i i •-! n - ' i i i . . . i i i . | • i 1 '—'— I ' '— ' '—' ' ' ' 111 0,1 1 10 100 Drug size [jam] Fig. 3. Increase in saturation solubility of BaS04 in water as a function of the particle size calculated using the Kelvin equation (with permission after4). velocity caused by a higher cs. It needs to be kept in mind which blood profile is anticipated with a certain drug. In many cases, too fast a dissolution is not desired (creation of high plasma peaks, reduction of tmax). There is the request to combine drug nanocrystals with traditional controlled release technology (e.g. coated pellets) to avoid too fast a dissolution, too high plasma peaks, too early a tmax and to reach prolonged blood levels. To summarize, the optimal drug nanocrystal size will depend on: 1. Required blood profile 2. Administration route In the case of i.v. injected nanocrystals, the size should be as small as possible in case the pharmacokinetics of a solution should be mimicked. In the event that a targeting is the aim (e.g. to the brain by PathFinder technology,5 the drug nanocrystals should possess a certain size to delay dissolution and to give them the chance to reach the blood-brain barrier (BBB) for internalization by the endothelial cells of the BBB.6 3.4. Effect of amorphous particle state It is well known that amorphous drugs possess a higher saturation solubility, compared with crystalline drug material. A classical example from the literature Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 313 is chloramphenicol palmitate. The polymorphic modification I has a solubility of 0.13, the high energy modification II a solubility of 0.43 and the amorphous material of 1.6mg/mL.7'8 The same is valid for drug nanoparticles, amorphous drug nanoparticles possess a higher saturation solubility, compared with equally sized drug nanocrystals in the crystalline state. Therefore, to reach highest saturation solubility increase, a combination of nanometer size and amorphous state is ideal. However, prerequisite for exploitation in pharmaceutical products is that the amorphous state can be maintained for the shelf life of the product. 4. Production Methods 4.1. Precipitation methods 4.1.1. Hydrosols The Hydrosol technology was developed by Sucker and the intellectual property owned by the company Sandoz, now known as Novartis.9,10 It is basically a classical precipitation process known to pharmacists under the term "via humida paratum" (v.h.p.). This v.h.p. process was employed to prepare ointments containing finely dispersed, precipitated drugs. The drug is dissolved in a solvent, the solvent added to a non-solvent leading to the precipitation of finely dispersed drug nanocrystals. A problem associated with this technology is that the formed nanoparticles need to be stabilized to avoid growth in micrometer crystals. In addition, the drug needs to be soluble at least in one solvent. This creates problems for the newly synthesized or discovered drugs, being poorly soluble in water and simultaneously in organic media. Lyophilization is recommended to preserve the particle size.1 To our knowledge, this technology has not been applied to a product to date. 4.1.2. Amorphous drug nanoparticles (NanoMorph®) Depending on the precipitation methodology, drug nanoparticles can be generated which are in the amorphous state. A nice example are carotine nanoparticles in food industry.11 A solution of the carotinoid, together with a surfactant and a digestible oil, are admixed into an appropriate solvent at a specific temperature. The solution is mixed with a protective colloid. This tranforms the hydrophilic solvent components into the water phase and the hydrophobic phase of the carotinoid forms a monodisperse phase. X-ray analysis after subsequent lyophilization shows that approximately 90% of the carotinoid is in the amorphous state.11 Amorphous precipitation technology is used by the company Soliqs and the technology is advertised under the tradename NanoMorph®. The preservation of 314 Miiller & Junghanns the amorphous state could be achieved successfully for food products. To exploit the amorphous technology for pharmaceutical products, the stricter requirements for pharmaceuticals need to be met. 4.2. Homogenization methods 4.2A. Microfluidizertechnology The previous Canadian company RTP (Montreal, now Skyepharma Canada Inc.) employed the microfluidizer to homogenize drug suspensions. The microfluidizer is a jet stream homogenizer of two fluid streams collied frontally with high velocity (up to 1000m/sec)12 under pressures up to 4000 bar. There is a turbulant flow, high shear forces, particles collied leading to particle diminution to the nanometer range.13-15 The high pressure applied and the high streaming velocity of the lipid can also lead to cavitation additionally, contributing to size diminution. The patent describes examples requiring up to 50 passes through the microfluidizer to obtain a nanosuspension.16 Sometimes, up to 100 cycles are required when applying the microfluidizer technology. This does not pose any problem on the small lab scale, but it is not production friendly for larger lab scale. The dispersion medium is water. 4.2.2. Piston-gap homogenization in water (Dissocubes®) In 1994, Mueller et al.17-18 developed a high pressure homogenization method based on piston-gap homogenizers for drug nanosuspension production. Dispersion medium of the suspensions was water. A piston in a large bore cylinder creates pressure up to 2000 bar. The suspension is pressed through a very narrow ring gap. The gap width is typically in the range of 3-15 micrometer at pressures between 1500-150 bar. There is a high streaming velocity in the gap according to the Bernouli equation.19 Due to the reduction in diameter from the large bore cylinder (e.g. 3 cm) to the homogenization gap, the dynamic pressure (streaming velocity) increases and simultaneously decreases the static pressure on the liquid. The liquid starts boiling, and gas bubbles occur which subsequently implode, when the suspension leaves the gap and is again under normal pressure (cavitation). Gas bubble formation and implosion lead to shock waves which cause particle diminution. The patent describes cavitation as the reason for the achieved size diminution.17,20 Piston-gap homogenizers which can be used for the production of nanosuspensions are e.g. from the companies APV Gaulin, Avestin or Niro Soavi. The technology was aquired by Skyepharma PLC at the end of the 90s and employed in its formulation development.21-23 Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 315 4.2.3. Nanopure technology For oral administration, the drug nanosuspensions themselves are, in most cases, not the final products. For patient's convenience, the drug nanocrystals should be incorporated in traditional dry dosage form, e.g. tablets, pellets and capsules. An elegant method to obtain a final formulation directly is the production of nanocrystals in non-aqueous homogenization media. Drug nanocrystals dispersed in liquid polyethylene glycol (PEG) or oils can be directly filled as drug suspensions into gelatine or HPMC capsules. The non-aqueous homogenization technology was established against the teaching that cavitation is the major diminution force in high pressure homogenization. Efficient particle diminution could also be obtained in non-aqueous media.24-30 To prepare tablets or pellets, the dispersion medium of the nanosuspension needs to be removed, i.e. in general, evaporated. Evaporation is faster and possible under milder conditions when mixtures of water with water miscible liquids are used, e.g. water-ethanol. To obtain isotonic nanosuspensions for intravenous injection, it is beneficial to homogenize in water-glycerol mixtures. The IP owned by Pharmasol covers, therefore, water-free dispersion media (e.g. PEG, oils) and also water mixtures. 4.3. Combination Technologies 4.3.1. Microprecipitation™ and High Shear Forces (NANOEDGE™) The Nanoedge technology by the company Baxter covers a combination of precipitation and subsequent application of high energy shear forces, preferentially high pressure homogenization with piston-gap homogenizers.31 As outlined in Sec. 4.1.1, the precipitated particles have a tendancy to grow. According to the patent by Kipp et ah, treatment of a precipitated suspension with energy (e.g. high shear forces) avoids particle growth in precipitated suspensions (= annealing process). The relative complex patent description can be summarized in a simplified way that the subsequent annealing stabilizes the obtained particle size by precipitation. As described in Sec. 4.1.2, precipitated particles can be amorphous or partially amorphous. This implies the risk that during the shelf life of a product, the amorphous particles can recrystalize, leading subsequently to a reduction in oral bioavailability or a change in pharmacokinetics after intravenous injection. The annealing process by Baxter converts amorphous or partially amorphous particles to completely crystalline material.31 31 6 Muller & Junghanns 4.3.2. Nanopure® XP technology An important criteria for a technology is its scaling up ability and the possibility to produce on large scale, applying "normal" production conditions. The number of 50-100 passes through a homogenizer as partially required for the microfluidizer technology16 is not production friendly. Piston-gap homogenizers (Sec. 4.2.2) proved to be more efficient, typically between 10-20 homogenization cycles are sufficient to obtain a nanosuspension. However, it would of course be desirable to apply even less homogenization cycles, reducing production time, potential product contamination by wearing of the machine and production costs. Pharmasol developed a new combination process, Nanopure XP (Xtended Performance)32 leading to: 1. Identical particle sizes compared with high pressure homogenization in water (Sec. 4.2.2), but at half the cycle numbers or less. 2. Lower particle sizes at identical cycle numbers. The process is again a combination technology, a pre-treatment step is followd by a high pressure homogenization step, typically performed with a piston-gap homogenizer.34'35 The code for this homogenization technology is H42. Figure 4 [Mm] LD - volume size distribution • 50% • 90% H99% cycle 15 new (H42) cycle 40 old technology Fig. 4. Comparison of the old homogenization technology (homogenization in water, piston-gap homogenizer) on the right side to the new technology on the left side, presented are the laser diffractometry (LD) diameters 50%, 90% and 99% (volume distribution, Coulter LS230, Beckman-Coulter/Germany) (with permission after33). Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 31 7 demonstrates the efficiency of method processing a very hard drug material. Applying the novel H42 technology leads to distinctly smaller crystals after just 15 cycles, compared with the "old" technology of applying 40 cycles. 5. Application Routes and Final Formulations 5.1. Oral administra tion Most attractive regarding regulatory and commercial aspects is the oral administration route. Compared with parenteral administration, the regulatory hurdles are much lower. In addition, the patient prefers an oral dosage form, that is why oral products possess the largest percentage of the pharmaceutical market. However, for the oral administration route, it is generally necessary to transfer the liquid nanosuspension into a solid dosage form. Aqueous nanosuspensions can be used as a granulation fluid for producing tablets or as a wetting agent for pellet production. In addition, spray drying can be performed in order to obtain a product which can subsequently be processed to oral products. The first nanosuspension product in the market was Rapamune®, introduced in 2000 by the company Wyeth. Rapamune® is available on the market as oral solution, and alternatively as tablet. The tablet is more user-friendly. Comparing the oral bioavailabilities of solution and nanocrystal tablet, the bioavailability of the nanocrystals is 21 % higher compared with the solution. The oral single dose of Rapamune® is 1 or 2 mg, the total tablet weight being 364 mg for 1 mg formulation and 372 mg for the 2 mg formulation, meaning that it contains a very low percentage of its total weight as nanocrystals. An important point is that the drug nanocrystals are released from the tablet as ultrafine suspension. In the event that crystal aggregation takes place to a pronounced extent, the dissolution velocity, and subsequently, the oral bioavailability of the BSCII drugs will be reduced. Therefore, there is an upper limit to load tablets with nanocrystals. In case the limit is exceeded and nanocrystals get in contact with each other within the excipient mixture of the tablet, the nanocrystals might fuse to larger crystals under the compression pressure during tablet production. For drugs with a low oral single dose such as Sirolimus in Rapamune®, incorporation into tablets causes little issues. A total nanoparticle load of less than 1% is well below the percentage being critical.36 The second product on the market was Emend®, introduced in 2001 by the Company Merck. The drug Aprepiptant is for the treatment of emesis (single dose is either 80 or 125 mg). Aprepiptant will only be absorbed in the upper gastrointestinal tract. Bearing this in mind, nanoparticles proved to be ideal in overcoming this narrow absorption window. The large increase in surface area due to nanonization leads to rapid in vivo dissolution, fast absorption and increased bioavailability.37,38 The formulation of a tablet from micronized bulk powder made higher doses necessary, 31 8 Muller & Junghanns leading to increased side effects.39 The drug nanocrystals are contained within the hard gelatin capsules as pellets. Aprepiptant was formulated as capsules for it to be user friendly by healthcare providers and patients, and on the other hand, to make it applicable as pellets via a stomach tube. Currently, studies are being undertaken to evaluate the change in pharmacokinectics (if any) between the pellets and the capsules. All nanocrystals in these first two products were produced using the pearl mill technology by Nanosystems/Elan. The prerequiste was the bioavailability of sufficient large scale production facilities for the respective product. In general, the candidates of first choice for nanosuspension technology are drugs with a relatively low dose. It is interesting that drugs such as Naproxen are formulated as nanosuspension (e.g. for fast action onset and reduced gastric irritancy),40 however, it requires more sophisticated formulation technology to ensure the release of the drug nanocrystals as fine suspension when incorporated in a tablet in a relatively high concentration of a single dose of 250 mg. The tablet size (weight) has to be acceptable for the patient and that a dosing with two tablets should be avoided, for reasons of patient's compliance and marketing purposes. Alternatively, to aqueous nanosuspensions, nanosuspensions in nonaqueous media can be produced by the Nanopure technology (Pharmasol). Nanocrystals dispersed in liquid PEG or oil can be directly filled into gelatine or HPMC capsules.25 It saves the step of water removal and subsequent dispersion of the powder in a liquid capsule filling medium. The Nanopure technology also allows production of nanocrystals in melted PEG (at 60° C). After solidification of the PEG nanosuspension, the drug nanocrystals are fixed (and kept seperated) in the solid PEG matrix. The solidified drug nanocrystal containing PEG can either be milled into powders and filled into the capsules, or alternatively, the hot liquid PEG nanosuspension can be directly filled into the capsules (Fig. 5, upper). Instead of using aqueous nanosuspensions as fluids for the wet granulation process or extrusion of pellet mass, the nanosuspensions can be converted into a dry powder which is subsequently further processed into a tablet or a capsule. It also appears attractive to package such powders in sachets for redispersion in water or soft drinks prior to oral administration. Spray-drying is the only feasable costeffective way to produce such powders. An attractive approach is the production of so-called "compounds" as described in the direct compress technology.41 The term "compound" does not mean a chemical compound; in powder technology, "compounds" are defined as freely flowable granulate powders. In the direct compress technology, water-insoluble polymeric particles (e.g. Eudragit RSPO, ethylcellulose) are dispersed in the aqueous drug suspension, and lactose is dissolved. The mixture is a freely flowable compound yielded by spray-drying. The lactose Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 319 Fig. 5. Gelatin capsules filled directly with hot liquid PEG nanosuspension, solidification takes place in the capsules (upper) or filling of the capsules with milled solidified PEG nanosuspension (lower). (From Ref. 36 with permissions.) is responsible for the good flowing properties. The water-insoluble polymeric particles also contribute to the formation of flowable granules, while at the same time allowing the compound to be compressed in a direct compaction process into tablets. The polymers form the matrix structure of the tablet. Depending on the percentage of the insoluble polymeric particles added, the resulting tablets may disintegrate fast or present a prolonged release system. A prolonged release of dissolving nanocrystal is desired in the case of high plasma that peaks at very early times (short tmax) and a targeted sustained blood level. Alternatively, the drug nanocrystal compound can be filled into hardgelatine capsules. Due to the presence of lactose and surfactant from the original nanosuspension, the compounds disperse relatively fast in liquids. Figure 6 shows the dispersion process of a compound after layering it on the surface of water in a beaker. As outlined above, efficient release and redispersion of the drug nanocrystals in a fine, nonaggregated state is a prerequisite for benefiting fully from the drug nanocrystal features.42 5.2. Parenteral administration Intravenous administration is the second frequently investigated route. The company Baxter, with its technology NANOEDGE™, is presently focusing on intravenous nanosuspensions. They investigated Itraconazole nanosuspensions intensive.44 It could be nicely shown that the side effects of the commercial product Sporanox® could be distinctly reduced by the administration of a nanosuspension. The nephrotoxicity of Sporanox® is not caused by the drug, but by the excipient 320 Miiller &/unghanns 0 sec 15 sec 30 sec 60 sec 120 sec Fig. 6. Dispersion of a drug nanocrystal compound as a function of time after layering it on the surface of water in a beaker (with permission after43) (Compound: Aquacoat 40%, Lactose, 60%.) used for solubilizing the drug, the hydroxypropyl-^-cyclodextrin.45'46 The itraconzole nanosuspension was stabilized with Tween 80 surfactant being well tolerated intravenously.44 Administration of nanosuspsensions into body cavities is also of great interest, e.g. to increase the tolerability of the drug, to achieve a local treatment or to have a depot with slow release (e.g. into the blood). It could be shown that intraperitonal administration of a nanosuspension was well tolerated, whereas administration of a macrosuspension leads to irritancy [azodicarbonamide (ADA), unpublished data]. Intraperitonal administration can be used for local treatment or to obtain a depot with prolonged release into the blood. Interesting therapeutic targets include local inflammations, e.g. in joints. For instance, arthritic joint inflammations are caused by secretion products of activated macrophages. An interesting approach is therefore the administration of a corticoid nanosuspension directly into the joint capsule. The drug particles will be phagocytosed, the drug dissolves and reduces the hyperactivity of the macrophages. This concept is not new, being adopted by the company Boots in the 80s in an attempt to incorporate the corticoid prednisolone into polymeric nanoparticles made from PLA-GA-copolymer.47 The particle load (polymer load) required to achieve a therapeutic drug level was being calculated. However, incubating macrophage cell cultures with the required particle concentration lead to cytotoxicity. The concept could not be realized, as it cannot occur with drug nanocrystals since no carrier polymer to required and present. Producing parenteral products with drug nanocrystals has to meet higher regulatory hurdles and product quality standards distinctly. The produced drug nanosuspensions need to be terminally sterilized or alternatively produced in an aseptic process. In principal, sterilization is possible by autoclaving. However, the increase in temperature can reduce hydration of steric stabilizers, thus leading to some aggregation during the sterilization process. Gamma irradiation is a priori a Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 321 non-preferred process by the industry due to the necessary analytics (i.e. proof of absence of toxic irradiation products). In addition, it was also observed that irradiation can cause aggregation not by directly interacting with the drug nanocrystals, but with the stabilizing surfactant. Irradiation of Tarazepide nanosuspension leads to aggregation; simultaneously, a decrease in zeta potential also occurred during the irradiation process. A decrease in zeta potential, i.e. electrostatic repulsion, was considered as the cause for the aggregation process. It can be concluded that the production of drug nanosuspensions in an aseptic, controlled process has to be preferred, compared with the terminal sterilization by irradiation. The aseptic production process can be validated and documented relatively easy, therefore, being simpler to handle as an irradiation sterilization with accompanied analytics. 5.3. Miscellaneous administration routes Oral and parenteral/intravenous routes are the ones in which developments are focusing, clearly due to the commercial background and the relation between the development costs for a market product versus its potential annual sales. However, drug delivery could also be improved when using drug nanocrystals for pulmonary and ophthalmic adminstration or dermal application. Poorly soluble drugs could be inhaled as drug nanosuspension. The drug nanosuspension can be nebulized using commercially available nebulizers.48,49 Disposition in the lungs can be controlled via the size distribution of the generated aerosol droplets. Compared with microcrystals, the drug is more evenly distributed in the droplets when using a nanosuspension. The number of crystals are higher, consequently, the possibility that one or more drug crystals are present in each droplet is higher. It could be shown that nanoparticles possess a prolonged retention time in the eye, most likely due to their adhesive properties.50-52 From this, poorly soluble drugs could be administered as a nanosuspension. However, the major obstacles are the commercial considerations. In many cases, the sales volume do not justify the costs for the development of a new market product. This is especially the case when a company has already a drug formulation which might be less efficient, but is already a product on the market. The price achievable with an improved product is not sufficiently high to cover the development costs of this new product. An additional major obstacle for the development of such improved products is the cost reduction policy of the healthcare systems worldwide. A longer treatment time with a less efficient product might still be less expensive for the healthcare system than a shorter treatment time with a more efficient, but distinctly more expensive product. The same is valid for dermal products. Sales per product are lower compared with e.g. oral products, as the dermal market is smaller. Dermal nanosuspensions 322 MiJller&Junghanns are mainly of interest if conventional formulation technology fails or if it is distinctly less efficient. Dermal drug nanosuspensions lead to a supersaturated system because of their increased saturation solubility. The higher concentration gradient between topical formulation and skin can improve drug penetration into the skin. In addition, because of their small size, drug nanocrystals could target the hair follicle by protruding into the gap around the hairs. This was illustrated in solid lipid nanoparticles of a similar size.53 Adhesive properties of drug nanocrystals are also an area of interest. Adherence to the skin reduces the "loss" of drug to the environment/ third persons. This is especially so in the event that highly active compounds are applied, e.g. hormones. For this reason, the drug estradiole was incorporated into solid lipid nanoparticles to better localize it on the skin.54 6. Nanosuspensions as Intermediate Products As described above, nanosuspensions can be produced such that nanocrystals appear in final products. Alternatively, drug nanosuspensions can be used as intermediate product, i.e. the drug nanocrystals do not appear in the final product. Recently, the SolEmuls® technology was developed to produce drug-loaded emulsions for intravenous injection, i.e. localizing poorly soluble drugs in the interfacial layer of lecithin emulsions.55-57 The applicability of the technology has been proven for several drugs including amphotericin B,58 itraconazole,59,60 ketoconazole,61 and carbamazepine,62'63 among others. The drug Amphotericin B is on the market as a solution (Fungizone®), but also in liposomes (Ambisome®); the latter having the benefit of reduced nephrotoxicity.64 Liposomes are relatively expensive (daily treatment costs approximately EUR 1000-200064,65), therefore Amphotericin B was incorporated into parenteral emulsions. These emulsions can also reduce nephrotoxicity,66 but for their production, it was necessary to use organic solvents. Egg lecithin and amphotericin B were dissolved in an organic solvent, the solvent evaporated and the obtained drug-lecithin mixture was used to produce an o /w emulsion. In these emulsions, amphotericin B was located in the interfacial lecithin layer as Amphotericin B is simultaneously poorly soluble in water and in oils.67 There were also attempts to incorporate amphotericin B in the emulsion by simply adding Amphotericin B powder to the emulsion and subsequently shaking it. However, even shaking for 18 hours with 1800rph was unable to completely dissolve the Amphotericin B. The reason was simply due to its low solubility in the water, and the dissolution velocity was also extremely low, i.e. the process of dissolution and redistribution into the lecithin layer takes too long for it to be used in pharmaceutical production. The problem was solved by the SolEmuls technology, i.e. simple co-homogenization of oil droplets and microcrystals. For a de novo production, a coarse pre-emulsion of lecithin stabilized Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 323 oil droplets in water is prepared, the drug powder is admixed under stirring and the obtained hybrid suspension subsequently homogenized at 600 bar (pressure being in the range to be used in pharmaceutical production lines). The high streaming velocities in the homogenization process lead to fast dissolution of the drug microcrystals and the re-distribution into the interfacial lecithin layer (Fig. 7). Depending on the size of the drug crystals, 5-10 homogenization cycles are required. The number of homogenization cycles can be reduced when adding the . drug not as microcrystals, but as nanocrystals in the form of a nanosuspension. A concentrated nanosuspension is prepared (e.g. 20-30% solid content) and added to the pre-emulsion. Ideally the nanosuspension is also stabilized by lecithin, i.e. the same emulsifier for the suspension and the emulsion. Alternatively, intravenously accepted stabilizers such as Tween 80 or Poloxamer 188 can be used. They are accepted intravenously without posing any regulatory issues. In addition, mixing the emulsion and nanosuspension at a ratio of 10:1 or higher will dilute the stabilizer concentration used in the nanosuspension by at least a factor of 10, meaning that in the final product, the nanosuspension surfactant concentration is typically 0.1 or 0.01%. The question might arise as to why an emulsion should be prepared using a nanosuspension as an intermediate product, when it can administer the nanosuspension itself intravenously? simple shaking r ~\ iecithin / drug mixture evaporation O rf-Q O ' o T direct production through highpressure homogenization dc/dt-co organic solution t lecithin + drug drug crystal or suspension Fig. 7. Drug incorporation through various methods in comparison. Left: traditional attempt of shaking or alternatively use of organic solvent; Right: SolEmuls® process. 324 Muller & Junghanns The reason is that drug-loaded parenteral emulsions are already products on the market (e.g. Diazepam-Lipuro, Etomidate-Lipuro, etc.), i.e. in a dosage form with which the regulatory authorities are already familiar with. Applying the SolEmuls technology and using lecithin-stabilized nanosuspension, the final product will only contain the excipients of an o /w emulsion for parenteral nutrition, without additional excipient plus the drug. It is an accepted known system with regard to the excipient status and its perfomance after intravenous injection. In contrast, drug nanosuspensions represent a new dosage form not yet present as intravenous formulations on the market. Registration of a completely new dosage form for a certain administration route is just more complicated and timely than registration of a product based on an established, known technology. 7. Perspectives There was a "delayed" acceptance of the nanocrystal technology in the 90s. Pharmaceutical companies tried to solve their formulation problems with the traditional formulation approaches. However, the increasing number of drugs having a very low solubility, and not able to be formulated with these traditional formulation approaches, lead to a broad acceptance of the drug nanocrystal technology. This is clearly reflected in the increasing number of licensing agreements between companies holding nanocrystal IP and a number of medium and large pharmaceutical companies. The smartness of the technology is that it can be universally applied to practically any drug. Identical to micronization, it is a universal formulation principle, but limited to BSC drugs class II. The time between the beginning of intensive research in the drug nanocrystal technology and the first products on the market was relatively short, about one decade. The value of a formulation principle or technology can be clearly judged by looking at the number of products on the market, in the clinical phases, and/or the time of entry into the market. Based on these criteria, the drug nanocrystal technology is a successful emerging technology. Meanwhile, "Big Pharma" also realized the drug nanocrystal value. In combination with the further increasing number of poorly soluble drugs, a distinct increase in drug nanocrystal-based products on the market can be expected. In many cases, oral products will dominate because of the market share, higher sales volumes and less regulatory hurdles and quality requirements, compared with parenteral products. References 1. Speiser PP (1998) Poorly soluble drugs, a challenge in drug delivery, in Muller RH, Benita S and Bohm B (eds.), Emulsions and Nanosuspensions for the Formulation of Poorly Soluble Drugs, Medpharm Scientific Publishers: Stuttgart. Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 325 2. Merisko-Liversidge E (2002) Nanocrystals: Resolving Pharmaceutical Formulation Issues Associated with Poorly Water-soluble Compounds in Particles, Marcel Dekker: Orlando. 3. Bangham AD and Haydon DA (1968) infrastructure of membranes: Biomolecular organization. Brit Med Bull 24(2):124-6. 4. Anger S (2005) PhD Thesis, in PhD Thesis Pharmaceutical Technology. Freie Universitat: Berlin. 5. Miiller RH, Luck M and Kreuter J (1998) Arzneistofftragerpartikel fur die gewebsspezifische Arzneistoffapplikation. Europaische Patentschrift, PCT/EP98/06429. 6. Kreuter J, Kharkevich RN and Ivanov DA (1995) Passage of peptides through the bloodbrain barrier with colloidal polymer particles (nanoparticles). Brain Res 674(1):171-174. 7. Gu Chong-Hui GDJW (2001) Estimating the relative stability of polymorphs and hydrates from heats of solution and solubility data. / Pharm Sci 90(9):1277-1287. 8. Hancock C and Bruno P (2000) What is the True Solubility Advantage for Amorphous Pharmaceuticals? Pharm Res 17(4):397-404. 9. Gassmann P, List M, Schweitzer A and Sucker H (1994) Hydrosols — Alternatives for the Parenteral Applikation of Poorly Water Soluble Drugs. Eur ] Pharm Biopharm 40:64-72. 10. List M and Sucker H (1988) Pat.No. GB 2200048. 11. Auweter HB, Haberkorn C, Horn H, Lueddecke D and Rauschenberger V, Patent No. DE19637517A1. 12. Tunick MHVH, Diane L, Cooke PH and Malin EL (2002) Transmission electron microscopy of Mozzarella cheeses made from microfluidized milk. / Agri Food Chem 50(1):99-103. 13. Bruno RPM (1999) Microfluidizer Processor Technology for High Performance Particle Size Reduction, Mixing and Dispersion. Microfluidizer Processor Technology. 14. Sunstrom JEM and Marshik-Guerts B (1996) General Route to Nanocrystalline Oxids by Hydrodynamic Cavitation. Chem. Mater. 15. Gruverman IJ and Thum JR, Production of Nanostructures Under Turbulent Collision Reaction Conditions — Application to Catalysts, Superconductors, CMP Abrasives, Ceramics and other Nanoparticles. Microfluidics Research. 16. Dearns R (2000), Atovaquone pharmaceutical compositions. US patent US 6 018 080. 17. Miiller RH, Peters K, Becker R and Kruss B (1995) Nanosuspensions — A Novel Formulation for the i.v. Administration of Poorly Soluble Drugs, in 1st World Meeting of the International Meeting on Pharmaceutics, Biopharmaceutics and Pharmaceutical Technology hosted by APGI/APV. Budapest. 18. Miiller RH, Becker R, Kruss B and Peters K (1999) Pharmaceutical Nanosuspensions for Medicament Administration as Systems with Increased Saturation Solubility and Rate of Solution, in United States Patent 5,858,410. USA. 19. Miiller RH, Jacobs C and Kayser O (2000) Nanosuspensions for the formulation of poorly soluble drugs, in Nielloud F and Marti-Mestres G (eds.) Pharmaceutical Emulsions and Suspensions, Marcel Dekker. 20. Miiller RH, Becker R, Kruss B and Peters K (1994) Pharmazeutische Nanosuspensionen zur Arzneistoffapplikation als Systeme mit erhohter Sattigungsloslichkeit und Losungsgeschwindigkeit. German patent 4440337.2, US Patent 5.858.410 (1999). 326 Muller & Junghanns 21. Muller RH, Dingier A, Schneppe T and Gohla S (2000) Large scale production of solid lipid nanoparticles (SLNTM) and nanosuspensions (DissoCubes™). in Wise D (ed.), Handbook of Pharmaceutical Controlled Release Technology. 22. Muller RH, Jacobs C and Kayser O (2003) DissoCubes — A novel formulation for poorly soluble and poorly bioavailable drugs, in Rathbone MJ, Hadgraft}, Roberts MS (eds.), Modified- Release Drug Delivery Systems, Marcel Dekker. 23. Rabinow BE (2004) Nanosuspensions in Drug Delivery. Nat Rev 3:785-796. 24. Muller RH (2002) Nanopure Technology for the Production of Drug Nanocrystals and Polymeric Particles, in 4th World Meeting ADRITELF/APV/APGI. Florence. 25. Bushrab NF and Muller RH (2003) Nanocrystals of Poorly Soluble Drags for Oral Administration. New Drugs 5: 20-22. 26. Radtke M (2001) Nanopure™ pure drug nanoparticles for the formulation of poorly soluble Drugs. New Drugs 3: 62-68. 27. Fichera MA, Keck CM and Muller RH (2004) Nanopure Technology — Drug Nanocrystals for the Delivery of Poorly Soluble Drugs, in Particles. Orlando. 28. Fichera MA, Wissing SA and Muller RH (2004) Effect of 4000 Bar Homogenisation Pressure on Particle Diminution in Drug Suspensions, in APV. Niirnberg. 29. Keck CM, Bushrab NF and Muller RH (2004) Nanopure® Nanocrystals for Oral Delivery of Poorly Soluble Drugs, in Particles. Orlando. 30. Muller RH, Mader K and Krause K (2000) Verfahren zur schonenden Herstellung von hochfeinen Micro-/Nanopartikeln, in PCT Application PCT/EP00/06535: Germany. 31. Kipp JE, Wong JCT, Doty MJ and Rebbeck CL (2003) Microprecipitation Method For Preparing Submicron Suspensions, in United States Patent 6,607,784. Baxter International Inc. (Deerfield, IL): USA. 32. Moschwitzer J and Muller RH (2005) Method for the Production of Ultrafine Submicron Nanosuspensions (Pat. Application). 33. Moschwitzer J (2005) PhD Thesis in Preparation, in PhD Thesis Pharmaceutical Technology. Freie Universitat: Berlin. 34. Moschwitzer J and Muller RH (2005) Effective production of ibuprofen drug nanocrystals by high pressure homogenization using new two-step process, in AAPS. submitted. Nashville. 35. Moschwitzer J and Muller RH (2005) Development of a new two-step process for the effective production drug nanocrystals by high pressure homogenization. in AAPS. submitted. Nashville. 36. Bushrab NF (2005) PhD Thesis in preparation, in PhD Thesis Pharmaceutical Technology. Freie Universitat: Berlin. 37. Moschwitzer J and Muller RH (2004) From the Drug Nanocrystal to the Final Mucoadhesive oral Dosage Form, in International Meeting on Pharmaceutics, Biopharmaceutics & Pharmaceutical Technology. Niirnberg. 38. Moschwitzer J and Muller RH (2004) Nanosuspensions as Formulation Principle for Chemical Stabilization of Chemically Labile Drugs, in International Meeting on Pharmaceutics, Biopharmaceutics & Pharmaceutical Technology. Niirnberg. 39. Wua YL, Landisb A, Hettricka E, Novaka L, Lynna L, Chenc K, Thompson A, Higgins R, Batrad U, Shelukard S, Kweia G and Storeye G (2004) The role of biopharmaceutics in Drug Nanocrystals/Nanosuspensions for the Delivery of Poorly Soluble Drugs 327 the development of a clinical nanoparticle formulation of MK-0869: A Beagle dog model predicts improved bioavailability and diminished food effect on absorption in human. Int} Pharm 285(1-2):135-146. 40. Liversidge GGCP (1995) Drug particle size reduction for decreasing gastric irritancy and enhancing absorption of naproxen in rats. Int J Pharm 125:309-313. 41. Miiller RH (1997) Preparation in Form of a Matrix Material-Auxiliary Agent Compound Containing Optionally an Active Substance: Europe. 42. Keck CM et al. (2004) Production and Optimisation of Oral Cyclosporine Nanocrystals, in AAPS. Baltimore. 43. Krause K (2004) Herstellung hochfeiner Polymer- und Arzneistoffdispersionen und deren Spruhtrocknung, in PhD Thesis Pharmaceutical Technology, Freie Universitat: Berlin. 44. Khar A (2002) Nanoedge Technologies. Baxter Company Booklet. 45. Yamaguchi H and Hachioji I (2001) New antifungal agents currently under clinical development. Nippon Kagaku Ryoho Gakkai Zasshi 9(49):535-545. 46. Slain DR, Cleary PD and Chapman SW (2001) Intravenous itraconazole. Annals of pharmacotherapy 35(6):720-729. 47. Smith A and Hunneyball LM (1986) Evaluation of poly(lactic acid) as a biodegradable drug delivery system for parenteral administration. Int} Pharm 30(2-3):215-220. 48. Hernandez-Trejo N, Kayser O, Miiller RH and Steckel H (2004) Physical Stability ofBuparvaquone Nanosuspensions Following Nebulization with Jet and Ultrasonic Nebulizers. Proceedings of the International Meeting on Pharmaceutics, Biopharmaceutics and Pharmaceutical Technology, Nuremberg Germany. 49. Hernandez-Trejo N, Kayser O, Miiller RH and Steckel H (2004) Characterization of nebulized buparvaquone nanosuspensions — Effect of nebulization technology. Pharm Res. submitted. 50. Patravale VBD, Abhijit A and Kulkarni RM (2004) Nanosuspensions: A promising drug delivery strategy. / Phar Pharmacol 56(7):827-840. 51. Pignatello R and Puglisi G (2002) Ocular tolerability of Eudragit RS100 and RL100 nanosuspensions as carriers for ophthalmic controlled drug delivery. / Pharm Sci 91(12):2636-41. 52. Bucolo CM, Puglisi G and Pignatello R (2002) Enhanced ocular anti-inflammatory activity of ibuprofen carried by an Eudragit RSI 00 nanoparticle suspension. Ophfhal Res 34(5):319-323. 53. Miinster UN, Haberland C, Jores A, Mehnert W, Rummel S, Schaller K, Korting M, Zouboulis Ch, Blume-Peytavi C and Schafer-Korting M (2005) RU 58841-myristateprodrug development for topical treatment of acne and androgenetic alopecia. Die Pharmazie 60(1):8-12. 54. Maia C, Mehnert W and Schafer-Korting M (2000) Solid lipid nanoparticles as drug carriers for topical glucocorticoids. Int} Pharm 196:165-167. 55. Miiller RH (2001) Dispersions for the Formulation of Slightly or Poorly Soluble Drugs, in PCT/EP01/08726. PharmasSol GmbH Berlin. 56. Miiller RH et al. (2004) SolEmuls — A novel technology for the formulation of i.v. emulsions with poorly soluble drugs. Int} Pharm 269:293-302. 328 Miiller & Junghanns 57. Muller RH et al. (2004) SolEmuls-novel technology for the formulation of i.v. emulsions with poorly soluble drugs. Int J Pharm 269(2):293-302. 58. Buttle I (2004) O/W-Emulsionen fiir die intravenose Applikation von Arzneistoffen, in PhD Thesis Pharmaceutical Technology, Freie Universitat: Berlin. 59. Akkar A and Muller RH (2004) Solubilisation by Emulsification. Pharm. Ind. 66(12): 1537-1544. 60. Akkar A and Muller RH (2003) Intravenous itraconazole emulsions produced by SolEmuls technology. Eur J Pharm Biopharm 56(l):29-36. 61. Akkar A et al. (2004) Solubilising Poorly Soluble Antimycotic Agents by Emulsification via a Solvent-Free Process. AAPS PharmSciTechpending, submitted. 62. Akkar A and Muller RH (2003) Formulation of intravenous Carbamazepine emulsions by SolEmuls technology. Eur J Pharm Biopharm 55(3):305-12. 63. Akkar A (2004) Poorly Soluble Drugs: Formulation by Nanocrystals and SolEmuls Technologies, in PhD Thesis Pharmaceutical Technology. Freie Universitat: Berlin. 64. Hann IM and Prentice HG (2001) Lipid-based amphotericin B: A review of the last 10 years of use. Int J Antimicrob Agents (17): 161-169. 65. Lewis R (2003) Antifungal therapy cost analysis (Patterson T. F. and McGinis M. R., ed.). www.doctorfungus.org. 66. Janoff A et al. (1993) Amphotericin b lipidcomplex (ABLC (TM)): A molecular rationale for the attenuation of amphotericin B related toxicities. / Liposome Res 3:451-471. 67. Davis SS and Washington C (1988) EP 0 296 845 Al. 15 Cells and Cell Ghosts as Drug Carriers Jose M. Lanao and M. Luisa Sayalero 1. Introduction Microparticle and nanoparticle polymeric systems currently occupy an important place in the field of drug delivery and targeting.1 Nevertheless, there are biological drug carriers that offer an efficient alternative to such systems. Within the different systems of biological carriers, of great importance are cells and cell ghosts, which are both efficient and highly compatible systems from the biological point of view, capable of providing the sustained release and specific delivery to tissues, organs and cells of drugs, enzymatic systems and genetic material. Cell systems such as bacterial ghosts, erythrocyte ghosts, polymorphonuclear leukocytes, apoptotic cells, tumor cells, dendritic cells, and more recently, genetically engineered stem cells, are all examples of how cell systems of very diverse nature can be suitably manipulated and loaded with drugs and other substances, to permit specific drug delivery in vivo with important therapeutic applications.2-8 Cell carriers for drug delivery are used in very different applications such as cancer therapy, cardiovascular disease, Parkinson's, AIDS, gene therapy, etc. Table 1 shows the classification of biological carriers for drug delivery based on the use of cells and cell ghosts. 2. Bacterial Ghosts Bacterial ghosts are intact, non-living, non-denatured bacterial cell envelopes devoid of cytoplasmic contents. They are created by lysis of bacteria, but maintain 329 330 Lanao & Sayalero Table 1 Kinds of cells and cell ghosts used for drug and gene delivery. Cell carrier Bacterial ghost Erythrocyte ghost Engineered stem cells Polymorphonuclear leucocytes Apoptopic cells Tumor cells Denditric cells Target Tissues, macrophages, cells RES, macrophages Tumor cells, T cells, macrophages Tissues Tumor cells Tumor cells T cells Encapsulated substance Drugs, vaccines, genetic material Drugs, enzymes, peptides Genetic material Drugs Drugs Drugs Drugs their cellular morphology and native surface antigenic structures, including their bioadhesive properties.3,9 Bacterial ghosts allow the encapsulation of drugs and other substances, and their specific attachment to mammalian tissues and cells. This kind of cell carrier acts as a true drug delivery system, allowing the permanency of drugs in the systemic circulation to be increased together with tissue-specific targeting. They are thus a promising alternative to conventional drug delivery systems such as liposomes or nanoparticles. The main advantages of bacterial ghosts as delivery systems are the fact that they are non-living, i.e. they can act as delivery systems of drugs, antigens or DNA; allow specific delivery to different tissues and cell types; and are well captured by phagocytic cells and antigen-presenting cells as dendritic cells. Among the drawback of bacterial ghosts is the possibility that they might revert to being virulent, the possibility of horizontal gene transfer, the stability of the recombinant phenotype, and pre-existing immunity against the carrier used.10 Usually, bacterial ghosts are produced by protein E-mediated lysis of Gramnegative bacteria.11 The production of bacterial ghosts is based on the controlled expression of the bacteriophage PhiX174-derived lysis gene E. Expression of this gene from a plasmid in Gram-negative bacteria leads to the formation of a transmembrane lysis tunnel structure that penetrates the inner and outer membranes, and is formed by protein E with border values fluctuating between 40-200 nm in diameter. Protein E is a hydrophobic protein localized exclusively in the cell envelope.12 E-mediated lysis has been achieved in many Gram-negative bacteria.13 Scanning electron micrographs of E-lysed cells reveal that bacterial ghosts contain only one E hole in a bacterial ghost, although in a few cases, there are two holes. The cytoplasm is expelled as a consequence of the high osmotic pressure inside the Cells and Cell Ghosts as Drug Carriers 331 cell. The collapse of membrane potential and the release of cytoplasmic components such as proteins, nucleic acids, etc occur simultaneously.14 In the case of strains of E. coli, this effect occurs within a period of 10 min after the induction of expression.15 The resulting empty bacterial cell envelope is considered a bacterial ghost. Bacterial ghosts show all the morphological, structural and immunogenic properties of a living cell.9-15-17 Since bacterial ghosts are derived from Gram-negative bacteria that are able to adhere to structures such as fimbriae and lipopolysaccharide, they are used for specific binding to human tissue.18 Bacterial ghost drug-loading is accomplished by the suspension of lyophilised bacterial ghosts in a buffered medium containing the drug. The ghosts are then subjected to an incubation process varying from 5 to 30 min at 24°C. They are then washed to remove excess drug.18'19 In order to prevent rapid leakage of loaded water-soluble drugs or other substances, the bacterial ghosts are sealed by fusion of the cell membrane with membrane vesicles at the edges of the lysis pore. For the sealing step, the bacterial ghosts suspension is incubated in the fusion buffer at 28°C for 10 min.18 Figure 1 shows a scheme of the production of bacterial ghosts by protein E-mediated bacterial lysis. The in vitro release of drugs from loaded bacterial ghosts is performed from a suspension of drug-loaded bacterial ghosts that is dialysed through a membrane suitable for excluding the ghosts. Dialysis is performed at 28°C in PBS buffer.19 The concentrations of drug released into the buffer at preset times are quantified using an appropriate analytical technique. In studies addressing the adherence and capture of loaded bacterial ghosts by target cells such as macrophages, human colorectal adenocarcinoma cells (Caco-2) or dendritic cells, fluorescent markers such as fluorescein isothiocyanate (FITC) are used. These allow adherence to be assessed using fluorescence microscopy and flow cytometry techniques.18,19 Macrophages internalize bacterial ghosts to a greater extent than Caco-2 cells.18,19 Studies carried out using confocal laser scanning microscopy with M. haemolytica ghosts loaded with Doxorubicin have shown that the drug was associated with the ghosts membranes and the inner lumen.19 Denditric cells that are professional phagocytic cells displaying the phagocytic capacity of antigens also have a good capacity for capturing bacterial ghosts, allowing the latter to be used as a vehicle for immunization and immunotherapy.20 2.1. Application of bacterial ghosts as a delivery system Bacterial ghosts have important therapeutic applications. They can be loaded with drugs, proteins and other substances, and can be targeted selectively to macrophages, tumors or endothelial cells.10,19 332 Lanao & Sayalero Cytoplasmic content AAAAAAA Inner Membrane Outer Membrane GRAM-NEGATIVE BACTERIA Cytoplasmic Membrane Protein E-mediated lysis E hole (40-200 nm) DRUG-LOADING BACTERIAL GHOST DRUG-LOADED BACTERIAL GHOST Fig. 1. Production and drug loading of bacterial ghosts. Bacterial ghosts have been used as efficient drug delivery systems10 in the field of anti-cancer drugs.18 Bacterial ghosts obtained have been used as a delivery system of doxorubicin to human colorectal carcinoma cells. Cytotoxicity assays revealed that doxorubicin-loaded ghosts show better antiproliferative capacity in Caco-2 cells than when free doxorubicin is used at the same concentration.18 Experiments have also been carried out using E.coli ghosts containing streptavidin, in order to increase the affinity of streptavidin for biotinylated compounds. Streptavidinloaded ghosts permit specific targeting to mucosal surfaces of the gastrointestinal and respiratory tracts, and also to phagocytic cells.3 Bacterial ghosts have been used as veterinary vaccines for the immunization of different animal species.9 Pasteurella multocida is a pathogen that causes morbidity and mortality in rabbits and its importance as a human pathogen has also been recognized. P. multocida Cells and Cell Ghosts as Drug Carriers 333 ghosts have been used to immunize rabbits and mice.17 Similar results have been obtained in the immunization of cattle against pasteurellosis using Pasteurella haemolytica ghosts.11 Actinobacillus pleuropneumoniae is a highly contagious microorganism and is the cause of porcine pleuropneumonia, infecting 30-50% of pig populations. However, Actinobacillus pleuropneumoniae vaccines provide limited protection, since they decrease mortality but not morbidity in swine. Comparative studies have been carried out on immunization using a aerosol infection model for pigs vaccinated with loaded-ghosts or formalin inactivated Actinobacillus pleuropneumoniae bacterins. The results obtained showed that immunization with bacterial ghosts is more efficient in protecting pigs than bacteria.21,22 Bacterial ghosts can also be used as carriers of therapeutic DNA or RNA.3/13 The use of nucleic acid vaccines currently offers a technique for the development of prophylactic or therapeutic vaccines, based on the use of DNA plasmids to induce immune responses by direct administration of DNA-encoding antigenic proteins into animals, and this is also suitable for the induction of cytotoxic T cells.23,24 Bacterial ghosts loaded with DNA produce a high level of gene expression. They can be used to enhance the mucosal immune response to target antigens expressed in the bacterial ghost system. They can also be used for the specific targeting of DNA-encoded antibodies to primary antigens located in cells.13 Ghosts of Vibrium cholerae have been tested as antigen carriers of Chlamidia trachomatis as potential vaccines for the control of genital infections produced by this bacteria. Recombinant Vibrium cholerae ghosts, previously cloned with a major outer membrane protein of C. trachomatis, afforded a high level of protective immunity against Chlamydia in a murine model.25,26 Mannheimia haemolytica is a pathogen that causes ovine mastitis. M. haemolytica ghosts loaded with plasmid DNA stimulate the elicitation of efficient immune responses in mice, with no symptoms of acute or subacute toxicity during the experiment.27 3. Erythrocyte Ghosts Erythrocytes constitute the largest population of blood cells and are produced in the bone marrow. They are mature blood cells that produce haemoglobin and carry out the exchange of oxygen and carbon dioxide between the lungs and the body tissues. The term "erythrocyte ghost" attempts to define the resulting cell-like structure when erythrocytes are subjected to a reversible process of osmotic lysis.28 For more than 30 years, many studies, both in vivo and in vitro, have been carried out to explore the use of erythrocyte ghosts as delivery systems of drugs and other substances.2 334 Lanao & Sayalero Erythrocyte ghosts are obtained from fresh erythrocytes coming from human blood or the blood of different animal species such as the rat, mouse, rabbit, etc, and are loaded with different types of substance, mainly drugs, peptides and enzymes, using different encapsulation methods. The most frequent methods for collecting erythrocyte ghosts are osmosis-based methods such as hypotonic dialysis.2,29 Autologous erythrocyte ghosts offer a drug delivery system that can act as a reservoir of the drug or substance encapsulated, providing the sustained release of the drug into the organism together with selective targeting of the drugs to the reticuloendothelial system (RES) of the liver, spleen and bone marrow.2 The main advantages of carrier erythrocytes as drug delivery systems are their high degree of biocompatibility, the possibility of encapsulating the drug in a small amount of cells, the sustained release of the encapsulated drug or substance into the body, the selective targeting to the RES, and the possibility of encapsulating substances of high molecular weight such as peptides. Among the drawbacks of these systems are the rapid leakage of some drugs out of the loaded erythrocytes and other problems related to their standardized preparation, storage and potential contamination.2 Erythrocyte ghosts can be obtained by diverse procedures such as hypotonic dilution, hypotonic pre-swelling, osmotic pulse, hypotonic hemolysis, hypotonic dialysis, electroporation, drug-induced endocytosis and chemical methods.2,30 Of the different ways of obtaining carrier erythrocytes, hypotonic dialysis is undoubtedly the most frequently used encapsulation method. The reasons why it is so popular are its simplicity, its ease of application for a large number of drugs, enzymes and other substances, and because it is the method that best conserves the morphological and haematological properties of the erythrocyte ghosts obtained. Hypotonic dialysis is based on the exposure of red cells to the action of a hypotonic buffer, inducing cell swelling and the formation of pores that permit the drug to enter erythrocytes by means of a passive mechanism. Figure 2 shows a scheme of the production of erythrocyte ghosts using a hypotonic dialysis method. Morphological inspection of erythrocyte ghosts is usually performed using transmission (TEM) or scanning (SEM) electron microscopy.2 Some physical parameters of red cell membranes can also be studied from the diffusion of haemoglobin.28 The haemolytic methods employed in the production of erythrocyte ghosts normally affect the haemolytic volume, surface area and tension.28 Figure 3 shows the morphological changes observed by SEM that occur in amikacin-loaded erythrocytes due to hypotonic dialysis.31 Haematological parameters determine the effects of the procedure used to collect erythrocyte ghosts on their haematological properties. Among others, parameters such as reduced glutathione (GSH), mean corpuscular volume (MCV) or red cell distribution width (RDW), may be evaluated using a haematology analyzer. Cells and Cell Ghosts as Drug Carriers 335 Dialysis bag SG ERYTHROCYTES Erythrocytes suspension T \ Drug ANNEALING (Isotonic buffer) RESEALFNG (Hypertonic buffer) (10 min, 37°C, pH 7.4) (30 min, 37°C, pH 7.4) Hypotonic buffer HYPOTONIC DIALYSIS (45 min, 4"C, pH 7.4) DRUG-LOADED GHOST ERYTHROCYTES Fig. 2. Production and drug loading of erythrocyte ghosts using a hypotonic dialysis method. o V CONTROL ERYTHROCYTES V'^C AMIKACIN LOADED ERYTHROCYTES Fig. 3. SEM micrographs of amikacin carrier erythrocytes obtained by hypotonic dialysis31 (Copyright 2005 from Encapsulation and in vitro Evaluation of Amikacin-Loaded Erythrocytes by C. Gutierrez Millan. Reproduced by permission of Taylor & Francis Group, LLC, http: / / www. taylorandfrancis .com). Erythrocyte ghosts obtained by hypotonic dialysis show a decrease in the mean corpuscular volume and an increase in size dispersion.28'29 Erythrocyte ghosts show a greater degree of haemolysis than normal erythrocytes.29 3.1. Applications of erythrocyte ghosts as a delivery system Erythrocyte ghosts can be used as potential drug delivery systems for enzymes, proteins and peptides, allowing sustained release into the systemic circulation and the delivery of these substances into the RES.2 In vitro release of drugs from loaded erythrocyte ghosts is usually tested using autologous plasma or an isoosmotic buffer at 37°C; alternatively, a dialysis bag may be used.32 The in vitro release of drugs and substances from loaded erythrocytes is usually a first-order process, suggesting that the drug crosses the plasma membrane through a passive diffusion mechanism.33 However, zero-order release 336 Lanao & Sayalero kinetics from loaded erythrocytes has also been described.34 In vitro studies about the release kinetics of different drugs, enzymes and peptides from loaded erythrocytes have shown a slow release of the encapsulated substance.2 When loaded erythrocyte ghosts are administered in vivo, changes in the pharmacokinetics of the encapsulated drugs occur, involving a systemic drug clearance related to the biological half-life of the erythrocytes.35 Increased serum half-lives and the areas under the curve of drugs encapsulated in loaded erythrocyte ghosts, in comparison with the free drug, have been observed in animals and humans.36,37 At the same time, erythrocyte ghosts show a greater accumulation in tissues such as liver and spleen.38,39 Surface treatment of erythrocyte ghosts with substances such as glutaraldehyde, ascorbate, Fe(+2), diamide, band 3-cross-linking reagents, trypsin, phenylhydrazine and the N-hydroxysuccinimide ester of biotin (NHS-biotin), enhances the recognition of erythrocyte ghosts by macrophages in vitro and liver targeting in vivo.40~i3 Red cells may be used as carriers for some drugs such as antineoplastics, antiinfective agents, antihypertensives, corticosteroids, etc.2 Thus, carrier erythrocytes have been widely studied as delivery systems of antineoplastic drugs for targeting the RES located in organs such as liver and spleen. Different antineoplastic drugs have been encapsulated in erythrocyte ghosts in both in vitro and in vivo experiments.2 Increases have been obtained in average survival times in the treatment of mice bearing hepatomas, using methotrexateloaded carrier erythrocytes.44 Better recognition and capture of erythrocyte ghosts by macrophages have been obtained by using biotinylated erythrocytes containing methotrexate,45 by alterations to the membrane using band-3 cross-linkers of erythrocyte ghosts containing etoposide,46 or by treatment of erythrocytes containing doxorubicin with glutaraldehyde.47 Anti-infective agents such as gentamicin, metronidazole, primaquine or imizol have also been encapsulated in erythrocytes.2 Human erythrocytes containing gentamicin have proven to act as an efficient slow-release system in ofco.48,49 Erythrocyte ghosts containing dexamethasone have been used in vivo in rabbits and humans. A sustained release of dexamethasone in vivo in animals and humans was observed using carrier erythrocytes. An increased anti-inflammatory effect of the drug using carrier erythrocytes was observed in rabbits.50,51 Moreover, new prodrugs of anti-opioid drugs such as naltrexone and naloxone have been encapsulated in erythrocytes to solve stability problems of the primary drug within the erythrocyte. The encapsulated prodrugs are transformed into the active compound, following their release from erythrocyte ghosts.52 In the fields of biochemistry and enzymatic therapeutics, the encapsulation of enzymes in erythrocytes has been studied in some depth. Enzymatic Cells and Cell Ghosts as Drug Carriers 337 deficiencies or the treatment of specific illnesses may be approached using carrier erythrocytes loaded with enzymes. The encapsulation of enzymes in erythrocytes solves some of the problems associated with enzyme therapy, such as the short half-life deriving from the action of plasma proteases, intolerant reactions, and the immunological disorders or allergic problems associated with the use of enzymes in therapeutics. In vitro or in vivo studies with enzyme carrier erythrocytes have been performed using L-asparaginase,53 hexokinase,54 alcohol dehydrogenase,55 aldehyde dehydrogenase,56 alcohol oxidase,57 glutamate dehydrogenase,58 uricase,59 urokinase,60 lactate 2-mono oxigenase,61 arginase,62 rhodanase,63 recombinant phosphotriestearase,64 delta-aminolevulinate dehydratase,65 urease,66 pegademase,67 brinase68 and alglucerase.69 One of the best examples of the use in therapeutics of carrier erythrocytes containing enzymes, is that of L-asparaginase encapsulated in human erythrocytes. This has been successfully used in the treatment of acute lymphoblastic leukaemia in paediatrics.70 Erythrocyte ghosts may act as carrier systems for the delivery of peptides and proteins. One of the main therapeutic applications of carrier erythrocytes in this field is that of anti-HIV peptides. Nucleoside analogues successfully inhibit the replication of immunodeficiency virases. In view of the importance of the monocyte-macrophage system in infection by HIV-1, it would be of maximum therapeutic interest to have available, the specific delivery of these therapeutic peptides into macrophages, which act as an important reservoir for the virus. Carrier erythrocytes containing anti-HIV peptides such as azidothimidine (AZT) and didanosine (DDI), significantly reduced the pro-viral DNA content in comparison with the administration of free peptides in a murine AIDS model.71 Similar results have been obtained with 2',3'-dideoxycytidine 5'-triphosphate'(ddCTP),72 2',3'-dideoxycytidine (ddCyd)73 and AZT prodrugs74 encapsulated in erythrocytes. Anti-neoplastic peptides such as 2-fluoro-ara-AMP (fludarabine) and 5'- fluoro-2'-deoxyuridine 5'-monophosphate (FdUMP), a pro-drug of 5-fluro-2'- deoxyuridine (FdUrd), have been encapsulated in human carrier erythrocytes, behaving as a slow-release delivery system.75,76 Macrophage uptake in vitro of antisense oligonucleotides may be increased by using carrier erythrocytes.77,78 Other peptides, such as erythropoietin,79 heparin,80 dermaseptin S3,81 interleukin-382 or vaccines,83 have also been encapsulated in erythrocytes to increase their stability,84 acting as a slow release system with a prolonged half-life,80,84 or for their specific targeting to bacterial membranes.85 Erythrocyte ghost derivatives can also be used as drug delivery systems. Nanoerythrosomes are erythrocyte membrane derivatives formed by spheroid vesicles, obtained by consecutive extrusion under nitrogen pressure through a polycarbonate filter membrane of a erythrocyte ghost suspension to produce small vesicles having an average diameter of 100 nm. In vitro and in vivo studies, carried out with 338 Lanao & Sayalero nanoerythrosomes loaded with daunorubicin, have shown that when linked covalently to nanoerythrosomes, the drug produces slow release of daunorubicin to the organism over a prolonged period of time and also that, in comparison with the free drug, cytotoxicity is greater.86 The advantage of nanoerythrosomes, as compared with erythrocyte ghosts as drug delivery system, is that the former are able to escape from the reticuloendothelial system faster.86,87 In vitro studies have shown that the nanoerythrosome-daunorubicin complex is rapidly adsorbed and phagocytosed by macrophages.88 Liver, spleen and lungs are the organs in which nanoerythrosomes show the greatest capacity of accumulation.89 Another derivative of erythrocyte ghosts are reverse biomembrane vesicles loaded with drugs.90 Reverse biomembrane vesicles are produced by spontaneous vesiculation of the ghost erythrocyte membrane by endocytosis, using an appropriate vesiculating medium, producing small vesicles containing the drug within the parent ghost. In vivo studies carried out using reverse biomembrane vesicles from erythrocyte ghosts loaded with doxorubicin in rats have revealed increases in the half-life and bioavailability of the drug, the liver and spleen, being the main organs for the clearance of this drug delivery system.90 4. Stem Cells In gene therapy, a therapeutic transgene is introduced into the patient with a view of supplementing the functions of an abnormal gene. To achieve the delivery of genetic material into the target cell, it is necessary to have a suitable carrier. One important aim in the field of gene therapy is the design and development of gene carriers that encapsulate and protect the nucleic acid, and selectively release the vector/nucleic acid complex to the target tissue, so that the genetic material will be released at the cellular level later. In practice, there are several ways to achieve this. The first is through the use of modified viruses containing the genetic material of interest. The use of viruses for gene delivery has some drawbacks since it is limited to specific cells susceptible to being infected by the virus, and also the administration itself of the virus, has some immunological problems among others.91-93 The second alternative is to use living cells modified genetically, such as stem cells, to deliver transgenic material into the body.8,94 Stem cell therapy is a new form of treatment, in which cells that have died or lost their function are replaced by healthy adult stem cells. One advantage of this kind of cell is that it is possible to use samples from adult tissues or cells from the actual patient, for culture and subsequent implantation. Within the framework of stem cell research, the use of stem cells as delivery systems is a novel and attractive technique in the field of gene therapy, in which the cells of the patients themselves are genetically engineered, in order to introduce a therapeutic transgene used to deliver the genetic material. A promising therapeutic Cells and Cell Ghosts as Drug Carriers 339 strategy is the use of stem cells such as lymphocytes or fibroblasts as drug delivery systems. Experimental studies using stem cells as such systems have been tested in different therapeutic applications, especially in the field of cancer therapy. Considering the affinity of stem cells for tumor tissue, engineered stem cells have been successfully used for direct drug delivery to cancer cells.8'94 In vitro cultures have been made of human mesenchymal stem cells from bone marrow that are transduced with an adenovirus vector carrying the human interferon beta-gene, which exerts therapeutic action against cancer. Engineered stem cells administered in vivo allow the delivery of the genetic material to cancer cells. This new drug delivery system has proven to be efficient in the treatment of experimental neoplasms, such as lung cancer, in mice.94 Figure 4 shows a scheme of the application of stem cells as carriers for gene delivery in experimental cancer therapy. In vivo studies have also been carried out with neural stem cells engineered using adenoviral vectors to express interleukm-12, an oncolytic gene, whose efficiency has been demonstrated in the treatment of intracranial malignant gliomas in mice.8,95 The used of haematopoietic stem cells has allowed antiviral genes to be introduced in both T cells and macrophages for the treatment of AIDS.96 The use of stem cells as vehicles for gene therapy has also been suggested for the treatment of ischaemic heart disease,97 Stem cells have also been employed in the field of antiepileptic therapy. Glial precursor cells, which release adenosine, have been derived from adenosine kinase embryonic stem cells. In these experiments, the fibroblasts were engineered to release adenosine by inactivating adenosine metabolising enzymes. After encapsulation within polyethersulfone hollow-fibre capsules, and the introduction into Mesenchymal Stem cells Interferon Beta Recombinant adenovirus Engineered i„ vitro Stem cells expansion '. Ficoll ; ', Bone marrow Fig. 4. Application of stem cells as carriers for gene delivery in experimental cancer. 340 Lanao & Sayalero the brain ventricles in a rat epilepsy model, the local release of adenosine allows drug-resistant focal epilepsy to be treated. These engineered cells were shown to suppress seizure activity.98-99 5. Polymorphonuclear Leucocytes Polymorphonuclear leucocytes (PMN) can be used as carriers of antibiotics in view of their selective targeting to sites of infection. Simply incubating PMN in the presence of high concentrations of antibiotic for 1 hr at 37° C guarantees cell loading with the antibiotic. PMN loaded with the macrolide azithromycin have been found to be efficient in an in vitro model that permits the delivery of the antibiotic in a bioactive form to Chlamydia inclusions in polarized human endometrial epithelial (HEC-1B) cells infected with Chlamydia trachomatis. PMN carriers allow the accumulation of large amounts of antibiotic in endometrial epithelial cells and its retention over long periods of time.4 6. Apoptopic Cells Programmed cellular death or apoptosis is a process that is controlled genetically in which the cells induce their own death in response to different types of stimulus such as the binding of death-inducing ligands to cell surface receptors. A new strategy for drug delivery, called apoptopic induced drug delivery (AIDD), allows drug delivery to tumor cells upon the initiation of apoptosis by using a biological mechanism to achieve drug delivery.5 This new system is based on the fact that apoptosis produces many changes in cell morphology that can be taken advantage of to achieve drug delivery. Apoptosis is reflected in enhanced membrane permeability, which favors the release of the encapsulated drug from the apoptotic cells to the tissue. Phagocytosis of drug- loaded apoptotic carrier cells by tumor cells facilitates the localization of the drug within the tumor cell. One advantage of the apoptotic induced drug delivery system (AIDD) is that the drug carrier cells may be genetically engineered to modify their properties. In vitro studies have been performed using S49 mouse lymphoma cells in which apoptosis is produced by exposure to dexamethasone. The cytotoxicity of RG-2 cells caused by temazolamide-loaded-S49 apoptotic cells was from 4 to 7 times higher than that of control temazolamide-loaded S49 cells.5 7. Tumor Cells A novel strategy for drug delivery based on the use of cell systems is the drugloaded tumor cell system (DLTC), developed for drug delivery and targeting in Cells and Cell Ghosts as Drug Carriers 341 lung metastasis.6'100 The tumor cells as drug carriers permit drug targeting to the blood-borne cancerous cells and the lungs as potential metastatic organs. In practice, there is affinity between the plasma membrane of malignant tumor cells and the metastatic addressins expressed by the endothelial cells of the targeted organ.6101 In vivo studies have been carried out with DLTC based on Doxorubicin-loaded B16-F10 murine melanoma cells. Doxorubicin accumulation in the mouse lung was several times higher than that seen after administering free Doxorubicin.6 8. Dendritic Cells Dendritic cells (DC) are antigen-presenting cells. They ingest antigen by phagocytosis, degrade it, and present fragments of the antigen at their surface. Dendritic cells have huge potential for immunization against a broad variety of diseases, because they travel throughout the body in search of pathogens indicative of infection or disease. They are very important for the induction of T cell responses, which result in cell-mediated immunity. Selective targeting of drugs incorporated in dendritic cells to T cells allows the response of these cells to be manipulated in vivo. It has been shown that when incorporated into dendritic cells, the drug O-galactosylceramide improves their anti-tumor activity.7 9. Conclusions This chapter has focused on the use of cells and cell ghosts as delivery systems of drugs, enzymes or therapeutic genes. The use of carrier cells such as bacterial ghosts, erythrocyte ghosts and engineered stem cells, for drug delivery and targeting are reviewed among others. Their high biocompatibility, together with their capacity for selective delivery and targeting in cells and specific tissues mean that these types of carrier are promising alternatives to the use of nano- and microparticle systems, with applications in the fields of interest such as cancer therapy, cardiovascular therapy, AIDS, gene therapy, etc. As an alternative to the use of cell carriers, modified viruses can also be used as drug delivery systems, especially in the field of gene therapy. Despite their potential interest, clinical studies with these types of carrier are still very limited, although in the near future, increase in the use and therapeutic applications of cell delivery systems is expected. Acknowledgments This chapter was supported in part by a project of the National Research and Development Plan (Project: SAF 2001-0740). 342 Lanao & Sayalero References 1. Mainardes RM and Silva LP (2004) Drug delivery systems: Past, present, and future. Curr Drug Targets 5:449-455. 2. Gutierrez-Millan C, Sayalero ML, Castaneda AZ and Lanao JM (2004) Drug, enzyme and peptide delivery using erythrocytes as carriers. / Control Rel 95:27-49. 3. Huter V, Szostak MP, Gampfer J, Prethaler S, Wanner G, Gabor F and Lubitz W (1999) Bacterial ghosts as drug carrier and targeting vehicles. / Control Rel 61: 51-63. 4. Paul TR, Knight ST, Raulston JE and Wyrick PB (1997) Delivery of azithromycin to Chlamydia trachomatis-infected polarized human endometrial epithelial cells by polymorphonuclear leucocytes. / Antimicrob Chemother 39:623-630. 5. Man J and Gallo JM (1998) Delivery of cytotoxic drugs from carrier cells to tumour cells by apoptosis. Apoptosis 3:195-202. 6. Shao J, DeHaven J, Lamm D, Weissman DN, Runyan K, Malanga CJ, Rojanasakul Y and Ma JK (2001) A cell-based drug delivery system for lung targeting: II. Therapeutic activities on B16-F10 melanoma in mouse lungs. Drug Del 8:61-69. 7. Fujii S, Shimizu K, Kronenberg M and Steinman RM (2002) Prolonged IFN-gammaproducing NKT response induced with alpha-galactosylceramide-loaded DCs. Nat Immunol 3:867-874. 8. Yang SY, Liu H and Zhang JM (2004) Gene therapy of rat malignant gliomas using neural stem cells expressing IL-12. Cell Biol 23:381-389. 9. Jalava K, Hensel A, Szostak M, Resch S and Lubitz W (2002) Bacterial ghosts as vaccine candidates for veterinary applications. / Control Rel 85:17-25. 10. Tabrizi CA, Walcher P, Mayr UB, Stiedl T, Binder M, McGrath J and Lubitz W (2004) Bacterial ghosts-biological particles as delivery systems for antigens, nucleic acids and drugs. Curr Opin Biotechnol 15:530-537. 11. Marchart J, Dropmann G, Lechleitner S, Schlapp T, Wanner G, Szostak MP and Lubitz W (2003) Pasteurella multocida- and Pasteurella haemolytica-ghosts: New vaccine candidates. Vaccine 21:3988-3997. 12. Altman E, Young KD, Garrett J, Altman R and Young R (1985) Subcellular localization of lethal lysis proteins of bacteriophages X and X174. / Virol 53:1008-1011. 13. Jalava K, Eko FO, Riedmann E and Lubitz W (2003) Bacterial ghosts as carrier and targeting systems for mucosal antigen delivery. Exp Rev Vaccines 2:45-51. 14. Witte A and Lubitz W (1989) Dynamics of PhiX174 protein E-mediated lysis of Escherichia coli. Eur } Biochem 393. 15. Witte A, Wanner G, Blasi U, Halfmann G, Szostak M and Lubitz W (1990) Endogenous transmembrane tunnel formation mediated by phi XI74 lysis protein E. / Bacteriol 172:4109-4114. 16. Witte A, Wanner G, Sulzner M and Lubitz W (1992) Dynamics of PhiX174 protein Emediated lysis of Escherichia coli. Protective immunity against pasteurellosis in cattle, induced by Pasteurella haemolytica ghosts. Arch Microbiol 381. Cells and Cell Ghosts as Drug Carriers 343 17. Marchart J, Rehagen M, Dropmann G, Szostak MP, Alldinger S, Lechleitner S, Schlapp T, Resch S and Lubitz W (2003) Protective immunity against pasteurellosis in cattle, induced by Pasteurella haemolytica ghosts. Vaccine 21:1415- 1422. 18. Paukner S, Kohl G, Jalava K and Lubitz W (2003) Sealed bacterial ghosts-novel targeting vehicles for advanced drug delivery of water-soluble substances. / Drug Targ 11:151-161. 19. Paukner S, Kohl G and Lubitz W (2004) Bacterial ghosts as novel advanced drug delivery systems: Antiproliferative activity of loaded doxorubicin in human Caco-2 cells. / Control Rel 94:63-74. 20. Halsberger AG, Khol G, Felnerova D, Mayr UB, Furst-Ladani S and Lubitz W (2000) Activation, stimulation and uptake of bacterial ghosts in antigen presenting cells. / Biotechnol 83:57-66. 21. Huter V, Hensel A, Brand E and Lubitz W (2000) Improved protection against lung colonization by Actinobacillus pleuropneumoniae ghosts: Characterization of a genetically inactivated vaccine. / Biotechnol 83:161-172. 22. Hensel A, Huter V, Katinger A, Raza P, Strnistschie C, Roesler U, Brand E and Lubitz W (2000) Intramuscular immunization with genetically inactivated (ghosts) Actinobacillus pleuropneumoniae serotype 9 protects pigs against homologous aerosol challenge and prevents carrier state. Vaccine 18:2945-2955. 23. Donnelly JJ, Ulmer JB, Shiver JW and Liu MA (1997) DNA vaccines. Annu Rev Immunol 15:617-648. 24. Felnerova D, Kudela P, Bizik J, Haslberger A, Hensel A, SaalmuUer A and Lubitz W (2004) T cell-specific immune response induced by bacterial ghosts. Med Sci Monit 10:BR362-370. 25. Eko FO, Lubitz W, McMillan L, Ramey K, Moore TT, Ananaba GA, Lyn D, Black CM and Igietseme JU (2003) Recombinant Vibrio cholerae ghosts as a delivery vehicle for vaccinating against Chlamydia trachomatis. Vaccine 21:1694-1703. 26. Eko FO, He Q, Brown T, McMillan L, Ifere GO, Ananaba GA, Lyn D, Lubitz W, Kellar KL, Black CM and Igietseme JU (2004) A novel recombinant multisubunit vaccine against Chlamydia. ] Immunol 173:3375-3382. 27. Ebensen T, Paukner S, Link C, Kudela P, de Domenico C, Lubitz W and Guzman CA (2004) Bacterial ghosts are an efficient delivery system for DNA vaccines. / Immunol 172:6858-6865. 28. Hoffman J (1992) Magnani M and DeLoach JR (eds.) The Use of Reseated Erythrocytes as Carriers and Bioreactors. Advances in Experimental Medicine and Biology 326, Plenum Press: New York, pp. 1. 29. Gutierrez Millan C, Zarzuelo Castaneda A, Sayalero Marinero ML and Lanao JM (2004) Factors associated with the performance of carrier erythrocytes obtained by hypotonic dialysis. Blood Cells Mol Dis 33:132-140. 30. Hamidi M and Tajerzadeh H (2003) Carrier erythrocytes: An overview. Drug Deliv 10:9-20. 344 Lanao & Sayalero 31. Gutierrez-Millan C, Arevalo M, Zarzuelo A, Gonzalez F> Sayalero ML and Lanao JM (2005) Encapsulation and in vitro evaluation of amikacin-loaded erythrocytes. Drug Del 12:409^16. 32. Pitt E, Johnson CM and Lewis DA (1983) Encapsulation of drugs in intact erythrocytes: An intravenous delivery system. Biochem Pharmacol 32:3359-3368. 33. Ogiso T, Iwaki M and Ohtori A (1985) Encapsulation of dexamethasone in rabbit erythrocytes, the disposition in circulation and anti-inflammatory effect. / Pharmacobio- Dyn 8:1032-1040. 34. Hamidi M, Tajerzadeh H, Dehpour AR, Rouini MR and Ejtemaee-Mehr S (2001) In vitro characterization of human intact erythrocytes loaded by enalaprilat. Drug Del 8: 223-230. 35. Bax BE, Bain MD, Talbot PJ, Parker-Williams EJ and Chalmers RA (1999) Survival of human carrier erythrocytes in vivo. Clin Sci 96:171-178. 36. Magnani M, Rossi L, Fraternale A, Bianchi M, Antonelli A, Crinelli R and Chiarantini L (2002) Erythrocyte-mediated delivery of drugs, peptides and modified oligonucleotides. Gene Titer 9:749-751. 37. Tonetti M, Astroff AB, Satterfield W, De Flora A, Benatti U and DeLoach JR (1991) DeLoach, Pharmacokinetic properties of doxorubicin encapsulated in glutaraldehydetreated canine erythrocytes. Am } Vet Res 52:1630-1635. 38. Lizano C, Perez MT and Pinilla M (2001) Mouse erythrocytes as carriers for coencapsulated alcohol and aldehyde dehydrogenase obtained by electroporation in vivo survival rate in circulation, organ distribution and ethanol degradation. Life Sci 68:2001-2016. 39. Alvarez FJ, Herraez A, Murciano JC, Jordan JA, Diez JC and Tejedor MC (1996) In vivo survival and organ uptake of loaded carrier rat erythrocytes. / Biochem 120:286-291. 40. Muzykantov VR, Zaltsman AB, Smirnov MD, Samokhin GP and Morgan BP (1996) Target-sensitive immunoerythrocytes: Interaction of biotinylated red blood cells with immobilized avidin induces their lysis by complement. Biochim Biophys Acta 1279:137-143. 41. Jordan JA, Alvarez FJ, Lotero LA, Herraez A, Diez JC and Tejedor MC (2001) In vitro phagocytosis of carrier mouse red blood cells is increased by Band 3 cross-linking or diamide treatment. Biotechnol Appl Biochem 34:143-149. 42. Lotero LA, Jordan JA, Olmos G, Alvarez FJ, Tejedor MC and Diez JC (2001) Differential in vitro and in vivo behaviour of mouse ascorbate/Fe3+ and diamide oxidized erythrocytes. Biosci 21:857-871. 43. Mishra PR and Jain NK (2000) Biotinylated methotrexate loaded erythrocytes for enhanced liver uptake. "A study on the rat". / Drug Targ 217. 44. Kruse CA, Freehauf CL, Patel KR and Baldeschwieler JD (1987) Mouse erythrocyte carriers osmotically loaded with methotrexate. Biotechnol Appl Biochem 9:123-140. 45. Mishra PR and Jain NK (2002) Biotinylated methotrexate loaded erythrocytes for enhanced liver uptake. "A study on the rat". Int} Pharm 145. 46. Lotero LA, Olmos G and Diez JC (2003) Delivery to macrophages and toxic action of etoposide carried in mouse red blood cells. Biochim Biophys Acta 1620:160-166. Cells and Cell Ghosts as Drug Carriers 345 47. Zocchi E, Tonetti M, Polvani C, Guida L, Benatti U and De Flora A (1988) In vivo liver and lung targeting of adriamycin encapsulated in glutaraldehyde-treated murine erythrocytes. Biotechnol Appl Biochem 10:555-562. 48. Eichler HG, Gasic S, Bauer K, Korn A and Bacher S (1986a) In vivo clearance of antibody-sensitized human drug carrier erythrocytes. Clin Pharmacol Ther 40: 300-303. 49. Eichler HG, Rameis H, Bauer K, Korn A, Bacher S and Gasic S (1986b) Survival of gentamicin-loaded carrier erythrocytes in healthy human volunteers. Eur J Clin Invest! 16:39-42. 50. Ogiso T, Iwaki M and Ohtori A (1985) Encapsulation of dexamethasone in rabbit erythrocytes, the disposition in circulation and anti-inflammatory effect. / Pharmacobiodyn 8:1032-1040. 51. Rossi L, Serafini S, Cenerini L, Picardi F, Bigi L, Panzani I and Magnani M (2001) Encapsulation of dexamethasone in rabbit erythrocytes, the disposition in circulation and anti-inflammatory effect. Biotechnol Appl Biochem 85. 52. Noel-Hocquet S, Jabbouri S, Lazar S, Maunier JC, Guillaumet G and Ropars C (1992) Magnani M and DeLoach JR (eds.), The Use of Resealed Erythrocytes as Carriers and BioreactorsAdvances in Experimental Medicine and Biology 326, Plenum Press, New York, pp. 215-221. 53. Kravtzoff R, Desbois I, Lamagnere JP, Muh JP, Valat C, Chassaigne M, Colomba P and Ropars M (1996) Improved pharmacodynamics of L-asparaginase-loaded in human red blood cells. Eur J Clin Pharmacol 49:465^70. 54. Rossi L, Bianchi M and Magnani M (1992) Increased glucose metabolism by enzymeloaded erythrocytes in vitro and in vivo normalization of hyperglycemia in diabetic mice. Biotechnol Appl Biochem 15:207-216. 55. Lizano C, Sanz S, Luque J and Pinilla M (1998) In vitro study of alcohol dehydrogenase and acetaldehyde dehydrogenase encapsulated into human erythrocytes by an electroporation procedure. Biochim Biophys Acta 1425:328-336. 56. Magnani M, Laguerre M, Rossi L, Bianchi M, Ninfali P, Mangani F and Ropars C (1989) Acetaldehyde dehydrogenase-loaded erythrocytes as bioreactors for the removal of blood acetaldehyde. Alcohol Clin Exp Res 13:849-859. 57. Magnani M, Fazi A, Magnani F, Rossi L and Mancini U (1993) Methanol detoxification by enzyme-loaded erythrocytes. Biotechnol Biochem Appl 18:217-226. 58. Sanz S, Lizano C, Luque J and Pinilla M (1999) In vitro and in vivo study of glutamate dehydrogenase encapsulated into mouse erythrocytes by hypotonic dialysis procedure. Life Sci 65:2781-2789. 59. Magnani M, Mancini U, Bianchi M and Fazi A (1992) Magnani M and DeLoach JR (eds.), The Use of Resealed Erythrocytes as Carriers and Bioreactors. Advances in Experimental Medicine and Biology 326, Plenum Press: New York, pp. 189. 60. Ito Y, Ogiso T, Iwaki M and Atago H (1987) Encapsulation of human urokinase in rabbit erythrocytes and its disposition in the circulation system in rabbits. / Pharmacobiodyn 10:550-556. 346 Lanao & Sayalero 61. Garin M, Rossi L, Luque J and Magnani M (1995) Lactate catabolism by enzyme-loaded red blood cells. Biotechnol Appl Biochem 22:295-303. 62. Adriaenssens K, Karcher D, Marescau B, Van Broeckhoven A and Terheggen HC (1984) Hyperargininemia: The rat as a model for the human disease and the comparative response to enzyme replacement therapy with free arginase and arginase loaded erythrocytes in vivo. Int J Biochem 16:779-786. 63. Petrikovics I, Pei L, McGuinn WD, Cannon EP and Way JL (1994) Encapsulation of rhodanese and organic thiosulfonates by mouse erythrocytes. Fundam Appl Toxicol 23:70-75. 64. Pei L, Omburo G, McGuinn WD, Petrikovics I, Dave K, Raushel FM, Wild JR, DeLoach JR and Way JL (1994) Encapsulation of phosphotriesterase within murine erythrocytes. Toxicol Appl Pharmacol 124:296-301. 65. Bustos NL and Batlle AM (1989) Enzyme replacement therapy in porphyrias: V. In vivo correction of delta-aminolaevulinate dehydratase defective in erythrocytes in lead intoxicated animals by enzyme-loaded red blood cell ghosts. Drug Des Del 5: 125-131. 66. Hamarat Baysal S and Uslan AH (2002) In vitro study of urease/ AlaDH enzyme system encapsulated into human erythrocytes and research into its medical applications. Artif Cells Blood Substit Immobil Biotechnol 30:71-77. 67. Bax BE, Bain MD, Fairbanks LD, Simmonds HA, Webster AD and Chalmers RA (2000) Carrier erythrocyte entrapped adenosine deaminase therapy in adenosine deaminase deficiency. Adv Exp Med Biol 486:47-50. 68. Flynn G, Hackett TJ, McHale L and McHale AP (1994) Encapsulation of the thrombolytic enzyme, brinase, in photosensitized erythrocytes: A novel thrombolytic system based on photodynamic activation. J Photochem Photobiol B Biol 76:193-196. 69. Bax BE, Bain MD, Ward CP, Fensom AH and Chalmers RA (1996) The entrapment of mannose-terminated glucocerebrosidase (alglucerase) in human carrier erythrocytes. Biochem Soc Trans 24:442S. 70. Oettgen HF, Old LJ, Boyse EA, Campbell HA, Philips FS, Clarkson BD, Tallal L, Leeper RD, Schwartz MK and Kim JH (1967) Inhibition of leukaemia in man by L-asparaginase. Cancer Res 27:2619-2631. 71. Fraternale A, Casabianca A, Tonelli A, Chiarantini L, Brandi G and Magnani M (2001) New drug combinations for the treatment of murine AIDS and macrophage protection. Eur J Clin Investig 31:248-252. 72. Magnani M, Rossi L, Fraternale A, Silvotti L, Quintavalla F, Piedimonte G, Matteucci D, Baldinotti F and Bendinelli M (1995) FIV infection of macrophages: In vitro and in vivo inhibition by dideoxycytidine 5'-triphosphate. Vet Immunol Immunopathol 46:151-158. 73. Magnani M, Bianchi M, Rossi L and Stocchi V (1989) Human red blood cells as bioreactors for the release of 2',3'-dideoxycytidine, an inhibitor of HIV infectivity. Biochem Biophys Res Commun 164:446-457. 74. Fraternale A, Casabianca A, Rossi L, Chiarantini L, Schiavano GF, Palamara AT, Garaci E and Magnani M (2003) Erythrocytes as carriers of reduced glutathione (GSH) in the treatment of retroviral infections. / Antimicrob Chemother 52:551-554. Cells and Cell Ghosts as Drug Carriers 347 75. Fraternale A, Rossi L and Magnani M (1996) Encapsulation, metabolism and release of 2-fluoro-ara-AMP from human erythrocytes. Biochim Biophys Acta 1291:149-154. 76. Al-Achi A, Greenwood R and Walker B (1994) Buccal administration of erythrocyteghosts- insulin in rats. / Control Rel 267. 77. Nielsen PE (2000) Antisense peptide nucleic acids. Curr Opin Mol Ther 2:282-287. 78. Chiarantini L, Cerasi A, Fraternale A, Andreoni F, Scari S, Giovine M, Clavarino E and Magnani M (2002) Inhibition of macrophage iNOS by selective targeting of antisense PNA. Biochemistry 41:8471-8477. 79. Garin MI, Lopez RM and Luque J (1997) Pharmacokinetic properties and in vivo biological activity of recombinant human erythropoietin encapsulated in red blood cells. Cytokine 9:66-71. 80. Eichler HG, Schneider W, Raberger G, Bacher S and Pabinger I (1986) Erythrocytes as carriers for heparin. Preliminary in vitro and animal studies. Res Exp Med 186:407-412. 81. Feder R, Nehushtai R and Mor A (2001) Affinity driven molecular transfer from erythrocyte membrane to target cells. Peptides 22:1683-1690. 82. Olmos G, Lotero LA, Tejedor MC and Diez JC (2000) Delivery to macrophages of interleukin 3 loaded in mouse erythrocytes. Biosci Rep 70:399-410. 83. Polvani C, Gasparini A, Benatti U, De Flora A, Silvestri S, Volpini G and Nencioni L (1991) Murine red blood cells as efficient carriers of three bacterial antigens for the production of specific and neutralizing antibodies. Biotechnol Appl Biochem 14:347-356. 84. Garin MI, Lopez RM, Sanz S, Pinilla M and Luque J (1996) Erythrocytes as carriers for recombinant human erythropoietin. Pharm Res 13:869-874. 85. Feder R, Nehushtai R and Mor A (2001) Affinity driven molecular transfer from erythrocyte membrane to target cells. Peptides 22:1683-1690. 86. Lejeune A, Moorjani M, Gicquaud C, Lacroix J, Poyet P and Gaudreault R (1994) Nanoerythrosome, a new derivative of erythrocyte ghost: Preparation and antineoplastic potential as drug carrier for daunorubicin. Anticancer Res 14:915-919. 87. Moorjani M, Lejeune A, Gicquaud C, Lacroix J, Poyet P and Gaudreault RC (1996) Nanoerythrosomes, a new derivative of erythrocyte ghost II: Identification of the mechanism of action. Anticancer Res 2831. 88. Lejeune A, Poyet P, Gaudreault RC and Gicquaud C (1997) Nanoerythrosomes, a new derivative of erythrocyte ghost: III. Is phagocytosis involved in the mechanism of action? Anticancer Res 3599. 89. Desilets J, Lejeune A, Mercer J and Gicquaud C (2001) Nanoerythrosomes, a new derivative of erythrocyte ghost: IV. Fate of reinjected nanoerythrosomes. Anticancer Res 1741. 90. Mishra PR and Jain NK (2000) Reverse biomembrane vesicles for effective controlled delivery of doxorubicin HC1. Drug Del 7:155-159. 91. El-Aneed A (2004) An overview of current delivery systems in cancer gene therapy. / Control Rel 94:1-14. 92. Klink D, Schindelhauer D, Laner A, Tucker T, Bebok Z, Schwiebert EM, Boyd AC and Scholte BJ (2004) Gene delivery systems-gene therapy vectors for cystic fibrosis. / Cyst Fibros 203. 348 Lanao & Sayalero 93. Tomanin R and Scarpa M (2004) Why do we need new gene therapy viral vectors? Characteristics, limitations and future perspectives of viral vector transduction. Curr Gene Ther 4:357-372. 94. Studeny M, Marini FC, Dembinski JL, Zompetta C, Cabreira-Hansen M, Bekele BN, Champlin RE and Andreeff M (2004) Mesenchymal stem cells: Potential precursors for tumor stroma and targeted-delivery vehicles for anticancer agents. / Natl Cancer Inst 96:1593-1603. 95. Ehtesham M, Kabos P, Kabosova A, Neuman T, Black KL and Yu JS (2002) The use of interleukin 12-secreting neural stem cells for the treatment of intracranial glioma. Cancer Res 62:5657-5663. 96. Su L, Lee R, Bonyhadi M, Matsuzaki H, Forestell S, Escaich S, Bohnlein E and Kaneshima H (1997) Hematopoietic stem cell-based gene therapy for acquired immunodeficiency syndrome: Efficient transduction and expression of RevMlO in myeloid cells in vivo and in vitro. Blood 89:2283-2290. 97. Sunkomat JN and Gaballa MA (2003) Stem cell therapy in ischemic heart disease. Cardiovasc Drug Rev 21:327-342. 98. Huber A, Padrun V, Deglon N, Aebischer P, Mohler H and Boison D (2001) Grafts of adenosine-releasing cells suppress seizures in kindling epilepsy. Proc Natl Acad Sci USA 98:7611-7616. 99. Boison D, Huber A, Padrun V, Deglon N, Aebischer P and Mohler H (2002) Seizure suppression by adenosine-releasing cells is independent of seizure frequency. Epilepsia 43:788-796. 100. Shao J, DeHaven J, Lamm D, Weissman DN, Malanga CJ, Rojanasakul Y and Ma JK (2001b) A cell-based drug delivery system for lung targeting: II. Therapeutic activities on B16-F10 melanoma in mouse lungs. Drug Del 8:71-76. 101. Alby L and Auerbach R (1984) Differential adhesion of tumor cells to capillary endothelial cells in vitro. Proc Natl Acad Sci USA 81:5739-5743. 16 Cochleates as Nanoparticular Drug Carriers Leila Zarif 1. Introduction In spite of the availability of many non-traditional novel dosage forms, oral route remains the most attractive way for administration of therapeutical materials. However, many therapeutic agents, especially the increasing number of biological molecules cannot be taken up by intestine due to their intrinsic impermeability to tissue membranes and the enzymatic degradation through the wall of the GI tract. Carrier systems that facilitate intestine uptake of these molecules are of major interests in the drug delivery arena. Moreover, drug delivery systems that provide a route of administration that does not involve injection can improve patient compliance and expand the market for existing, injectable, drugs. The factors which are important for the oral efficiency of a vehicle system have been repeatedly summarized in the literature.1,2 Small particle size, appropriate surface properties, mucoadhesive and targeting moieties, stability, as well as dose are the major factors imparting the efficiency of oral uptake. Producing formulations of poorly soluble drugs with high bioavailability is an even higher challenge. Known technologies are nanocrystals and nanoparticles which use the approach of enhancing the bioavailability by a decrease in particle size, resulting in an increase of surface area and subsequently a faster dissolution. Other technologies such as solid dispersions, polymeric micelles and selfemulsifying systems were developed to increase the drug solubility. 349 350 Zarif Many lipid-based systems were developed to enhance oral bioavailability3,4 Examples are lipid-based emulsions & microemulsions5-7; Solid lipid nanoparticles (SLN), a high melting point lipids enclosed in a surfactant layer8,9 adequate to enhance the oral bioavailability of poorly absorbed drugs; Lipid nanocapsules (LNC) for oral, injectable use10 and improved bioavailability11; Lipid nanospheres prepared from egg lecithin and soybean, described for their low toxicity12 and higher efficacy, compared with other delivery systems when incorporating amphotericin B,13 due to their smaller particle size and lower uptake by reticuloendothelial system.14,15 Recently, solid lipid microparticles, prepared by the solvent-in-wateremulsion- diffusion technique, were described for the encapsulation and oral delivery of insulin.16 In particular, lipid-based cochleate delivery system appears to provide answers to oral delivery challenges by (1) formulating different kind of molecules, especially hydrophobic ones17,18 and (2) protecting the sensitive and biologically active molecules from harsh environmental conditions. In this review, we will focus on cochleates nanoparticular drug carrier and will present the main features and the state of the art of this delivery technology. 2. Cochleates Nanoparticles in Oral Delivery 2.1. Cochleate structure Cochleates were first described by Dimitrious Papahadjopoulos and his co-workers in 1975 as precipitates formed by the interaction of negatively charged phosphatidylserine and calcium.19-21 He named these cylindrical structures "cochleate", meaning shell in the Greek language because of their rolled-up form, and explained the mechanism of cochleates formation by the fusion of negatively charged vesicles induced by the calcium cation22 (Fig. 1). These cigar-like structures have gained interest as antigen delivery system for vaccine applications.23 More recently cochleates were studied as tools to deliver small molecule drugs.17,18,24 A cochleate lipid formulation of amphotericin B has been developed as an oral composition to treat systemic fungal infections.24-26 Other medical and non-medical applications are also under investigation.27 2.2. Cochleate preparation 2.2 A. Which phospholipid and which cation to use? Cochleates are a phospholipid-ion precipitates. Does that mean that cochleate is a structure obtained from precipitation of any phospholipid with any ion as presented in some litterature?,28 i.e. a complex of negatively charged phospholipid with any cation or a complex made from a positively charged lipid with any anion? Cochleates as Nanoparticular Drug Carriers 3 51 f) Ca 0 fusion ( \ I 1 ///EPTA. A B C D E F Fig. 1. Cochleate cylindrical structure and mechanism of formation (adapted from Refs. 19 and 69 with permission). Papahadjopoulos has given in 1975 this appellation to a rolled phospholipid structure. So far, to our knowledge no physico-chemical evidence on the obtention of such cigar-like structure from positively charged phospholipid with an anion had been described; on the contrary, extensive litterature is available on obtaining these cigar-like structure when negatively charged phospholipid such as phosphosphatidylserine (PS) had been precipitated with a cation such as calcium.17'18'20-22'29-30 Other negatively charged phospholipids, such as phosphatidic acid (PA) or phosphatidyl glycerol derivatives, have been studied as well. Mixture of negatively charged phospholipids with other lipids can lead to cochleate formation. In this case, the cochleate formation depends on the negatively charged lipid/other lipid ratio and depends on the nature of the negatively charged lipid in the mixed lipid system. For example, PA derivatives form cochleate domains after the addition of calcium cation. However, when mixed with the corresponding diacylphosphatidylcholine (PC) and diacylphosphatidylethanolamine (PE), it was found that up to 20 mole% of PC or PE can be introduced into the cochleate phase of PA(Ca2+), above which a distinct PC rich or PE-rich phase appears.31 Other phospholipid derivatives such as galactosphingolipid hydroxy fatty acid cerebroside were reported to form cochleate cylinders by thermal mechanical treatment of glycol suspensions.32 However, the addition of conjugated lipid, such as 352 Zarif poly(ethylene glycol)-lipid conjugates to PS vesicles, inhibited the calcium-induced fusion.33 In general, an additional desired feature of an oral drug delivery system is that the excipient permitting this transport to be classified is generally regarded as safe (GRAS). Soy phosphatidylserine fits this criteria. Furthermore, Soy PS has been used as a nutrient supplement since early 1980s. Clinical trials showed that PS may play a role in supporting mental functions in aging brains such as enhancing the memory, improving learning ability,34-41 reducing the stress42'43 and anxiety.44 Cochleates can be made from purified soy phosphatidylserine, which represents an affordable source of raw material.45 A study comparing the purified soy phosphatidylserine (PSPS) to non-purified soy PS (NPSPS) has been disclosed in this patent, showing that PS should be present in an amount of at least 75% of the total lipid in order to allow the formation of cochleates. The other 25% phospholipids present can be selected either from the anionic group such as phosphatidic acid, phosphatidylglycerol, phosphatidyl inositol or phosphatidylcholine. PSPS cochleates can be loaded with different bioactive materials such as nutritional supplement, vitamins, antiviral, antifungal, small peptides. Proof of principle of the use of purified soy PS has been achieved using a polyene antifungal agent, amphotericin B. The preparation method for amphotericin B cochleates can be either via High pH-trapping or film method18 or by hydrogel method;29 the latter leading to nanocochleates formation. The nature of the cation is an important factor in cochleate formation. In the precipitation process, divalent cations are preferred to monovalent cations. Monovalent cations such as Na+ were described to prevent the cochleate formation.46 Increases concentration of Na+ ions was shown to interfere with the destabilization effect of Ca2+. A critical Ca/PS ratio is necessary for the destabilization effect of divalent cations and the formation of cochleate phases.46 The formation of cochleate is easier from small unilamellar vesicles (SUV). However, multilamellar vesicles (MLV) can also lead to cochleate formation. In this case, the first mechanism is a destabilization of the outer bilayer of PS by Ca2+ which causes its collapse, leading to a higher access of Ca2+ to inner PS bilayers and so forth. 2.2.2. Which molecules can be entrapped in cochleates nanoparticles Due to the intrinsic nature of the lipid-contained cochleates, these nanoparticles can encapsulate a variety of molecules of all shapes and sizes. Preference is given, however, to hydrophobic molecules, for which a need to enhance chemical stability or bioavailability is desired [Fig. 2(a)]. Amphiphatic molecules which can easily Cochleates as Nanoparticular Drug Carriers 353 Fig. 2. Type of molecules which can be encapsulated into lipid based cochleate (adapted from Ref. 18 with permission). insert in the membrane bilayers [Fig. 2(b)], negatively charged moiety [Fig. 2(c)] or positively charged moiety [Fig. 2(d)] could be encapsulated in the cochleate nanoparticle structure. The nature of the drug influence the percentage of encapsulation. Hydrophobic drag shows a quantitative encapsulation, whereas less was seen for amphiphatic molecules. For instance, doxorubicin which presents hydrophobic regions is a water-soluble drug, has a partition between the bilayers and the external aqueous phase [Fig. 2(b)]. As calcium induces dehydration of the interbilayer domains, the amount of water in this region is low,47 therefore, small hydrophilic molecules will not be suitable for cochleate system. 2.2.3. Multiple ways of preparing cochleates Several processes were developed to obtain cochleates with a nanosize range, with the objective to allow oral delivery.24,29'48-59 Particle size is process dependent. When a small nanosized particle is desired, the "hydrogel method" can be used, based on the use of an aqueous-aqueous emulsion system.29 Briefly, this method consists of 2 steps: The preparation of small size liposomes either by high pH method18'25 or by film method,18 then the liposomes are mixed with a high viscosity polymer 354 Zarif such as dextran. The dextran/liposome phase is then injected into a second, nonmiscible, polymer (i.e. PEG). The calcium was then added and diffused slowly from one phase to another, resulting in the formation of nanocochleates. The final step is the washing of the gel. These nanosized cochleates showed potential in the oral delivery of drugs.18,29,48,59 Electron microscopy and X-Ray crystallography of the nanoparticles show a unique multilayered structure consisting of continuous, solid lipid bilayer sheets, rolled up in a spiral with no internal aqueous space and the localization of AmB in the lipid bilayer.25 Other preparation techniques are known, e.g. the trapping method, useful for the encapsulation of hydrophilic and hydrophobic molecules,17'18 which consist in the preparation of the liposomal suspension containing the drug either in the aqueous space of liposome (when hydrophilic) or intercalated in between the bilayers (when hydrophobic). A step of addition of calcium follows, and an aggregate of cochleates are formed. The cochleates made by the Trapping method present higher aggregation compared with other methods. This has been demonstrated using Electron microscopy after Freeze-fracture.25 Another method was developed for hydrophobic drugs,61 known as "the solvent drip method" which consists of preparing a liposomal suspension separately based on soy PS and a hydrophobic or amphipathic cargo moiety solution. Solvent for hydrophobic drug can be selected from DMSO, DMF. The solution is then added to liposomal suspension. Since the solvent is miscible in water, a decrease of the solubility of the cargo moiety is observed, which associates at least in part with the lipid-hydrophobic liposomal bilayers. The cochleates are then obtained by addition of calcium and the excess solvent is being washed. Usually, the cochleate formation can be characterized by optical microscopy when they are present in needle form in the micrometer size range. In this case, direct observation using a higher magnification can be used.25 When nanocochleate are obtained, optical microscope can be used as an indirect method to assess the formation of cochleate, i.e. observation of the liposome formation after chelation of the calcium present, by addition of EDTA (ethylene diamine tetraacetate) to nanocochleate. A more sophisticated method is the electron microscopy after freezefracture18' 25 which allows the observation of the tighted packed bilayers. Recently, other methods were described using Laurdan (6-dodecanoyl-2-dimethylamino naphtalene) to monitor the cochleate phase formation.62 In this case, the lipid vesicles are labeled with Laurdan and the addition of calcium to the laurdan labeled vesicles resulted in a shift in the emission peak maximum of Laurdan. Due to dipolar relaxation, excitation and emission, generalized polarization (GPgx and GPEm) indicates the transition from a LC to a rigid and dehydrated cochleate phase. Cochleates as Nanoparticular Drug Carriers 355 2.3. Cochleates as oral delivery system for antifungal agent, amphotericin B Among the drug of choice using nanocochleate delivery system, amphotericin B (AmB) presented all aspects of a good candidate. Amphotericin B is a hydrophobic drug with poor oral bioavailability. This drug had been used for decades in injectable form to treat systemic fungal infections of Candida, cryptococcus and aspergillosis species.63-65 Lipid formulations of Amphotericin B such as liposomes, lipid complexes, lipid emulsions and colloidal dispersions, were developed with the aim to achieve a higher therapeutic index.26-66 These formulations indeed showed enhanced therapeutic index, even though none of these formulations showed ability to deliver AmB orally. Cocheate technology seems to offer the advantage over other delivery systems in providing the possibility for the oral delivery of AmB. Oral administration of amphotericin B cochleates (CAMB) to healthy mice achieved potentially therapeutic concentrations in key target tissues.51 Preclinical studies demonstrate a promising activity of CAMB in murine models of clinically relevant invasive fungal infections such as disseminated candidiasis,25'48,67 disseminated aspergillosis17,18-58'59 and central nervous system cryptococcosis.68 2.3.1. In candidiasis animal model In Candida albicans infected murine animal model, AmB cochleates showed potential either after intraperitoneal (i.p.) or oral (p.o.) administration.17,18,48,49,54,55,57,60,66-68 After i.p. administration CAMB provided protection against C. albicans at doses as low as 0.1 mg/kg/day, kidney tissues burden showed that CAMB was more potent than Fungizone® at 1 mg/kg/day and was equivalent to AmBisome® at 10 mg/kg/day18,25,60 (Fig. 3). CAMB was also effective after oral administration. Complete eradication of C. albicans from the lungs was noticed after p.o. administration at 2.5 mg/kg/day. These results were comparable to i.p. Fungizone® at 2.0 mg/kg/day.48,54-56 2.3.2. In aspergillosis animal model Oral administration of CAMB was shown to be protective in a dosedependent manner against systemic infection of Aspergillus fumigatus in animals immunosusppressed with cyclophosphamide.58,59 In this mouse model, intragastric administration of CAMB at 40 mg/kg/day for 15 days resulted in 80% survival, while Fungizone at 4 mg/kg/day (i.p.) resulted in 20% survival; higher doses of Fungizone were lethal to animals. 356 Zarif 0) •J t/1 UJ 0) ~-~ :-> Uo 1071 106 - 105 - 104 - mJ- 102- 1 0 ' • 10°- 1 U 1 1 . 1 1 1 1 1 • > Control 0.1 1.0 10.0 0.1 1.0 10.0 0.1 1.0 AmB Dose Concentration (mg/kg) Fig. 3. Kidneys tissue burden of infected mice treated with either CAMB (•), Fungizone (•) or AmBisome (•), compared with controls (T) (from Ref. 18 with permission) 2 trt • 1 1 I J ffl -1 u * 1 i* * JL ' l 5 ! U JJ M • M. , , • Liver • Kidney • 1.urn's 1 control DAMB Smg/kg lOmg/kg 20mg/kg 30mg/kg 40mg/kg 5- a a c; & <=> K 3 <5 -J (3 CIO s? ~ — • < Sfl 40 30 20 10 Concentration of Drug Fig. 4. Tissue burden for mice infected in a model of invasive aspergillosis after oral administration of CAMB (from Ref. 58 with permission). The tissue fungal burden for target organs, kidneys, liver and lungs, demonstrated the benefic effect of CAMB (Fig. 4 ). CAMB showed a pronounced dosedependent reduction in the fungal burden in all organs. The near eradication of Aspergillus was observed above a concentration of 20mg/kg/day. CAMB at 30 mg/kg (PO) was as effective as CAMB at 20 mg/kg (PO) in reducing fungal tissue burden.58 n? o<> 00° ... . t I Cochleates as Nanoparticular Drug Carriers 357 2.3.3. In cryptococcal meningitis animal model Oral amphotericin B cochleates were effective in a murine cryptococcal meningitis model with an 80% survival after 17 days, obtained after oral treatment with CAMB (lOmg/kg) to mice having intracerebral infection with cryptococcus neoformans.68 2.3.4. Toxicity of amphotericin B cochleates In vitro, Amphotericin B cochleates (CAMB) showed a low toxicity on red blood cells when compared with Fungizone (DAMB). CAMB showed no hemoglobin release and therefore no hemolysis of red blood cells when incubated at 500 ^g/ml. In contrast, DAMB was hemolytic at 10 /xg/ml due to the presence of the detergent, sodium desoxycholate.25 In vivo, CAMB was non toxic to mice when administered orally at 50mg/kg/day for 14 days. No nephrotoxicity was observed as demonstrated by the normal BUN level, and the histopathology of kidneys, lungs, liver, spleen and GI tract showed that animals dosed with CAMB were comparable to controls.18 2.3.5. Pharmacokinetics of amphotericin B cochleates Oral pharmacokinetics?)^ Pharmacokinetic studies have shown that after oral administration of CAMB, AmB is distributed into the target tissues (e.g. brain, liver, lung, spleen and kidneys)18,50'52 in healthy mice and AmB tissue level suggests a zero-order uptake process for all tissues. When CAMB was administered po to C57BL/6 mice at lOmg/kg (n = 5), and blood and tissues collected and AmB level measured by HPLC, blood shows a plateau-shaped profile with Tmax = 6h and Cmax = 0.05mg/ml. Noncompartmental (NCA) analysis showed blood AUC0-oo = 1.20/xg*h/ml, ti/2 = 12.8 h, MRTo_oo = 21.1 h, Cl/F = 139.2ml/min/kg, Vz /F = 153.91 L/kg. AmB tissue exposure (AUCo-oo, .ig*h/g) evaluated using NCA was greater for lungs (23.11), followed by liver (16.91), spleen (15.40) kidneys (14.97) and heart (3.34). Tissue elution ti/2(h): kidneys 9.3, lungs 5.6, heart 5.3, liver 4.9 and spleen 4.3. For all tissues, Tmax = 12 h and Cmax ranged between 0.23/zg/ml for heart and 1.58/xg/ml for lungs.52 The delivery of AmB by cochleates after multiple oral doses (10) was assessed in the same mouse model and was compared with AmBisome. It was found that cochleate provides therapeutic levels in tissue and presents better delivery and transfer efficiency of AmB to the target tissue, as well as better tissue penetration.53 358 Zarif The ability of cochleate vehicles to deliver systemic AmB after single or multiple oral dosing suggest the potential of CAMB formulations to treat and prevent systemic fungal infections. Pharmacokinetics AmB given intraveneously (IV) to mice showed a two-phase pharmacokinetic profile.69,70 Pharmacokinetic analysis in target tissues (liver, spleen, kidney and lungs) shows a multi-peak profile, large AUC and MRT. After IV administration of 0.625 mg/kg, AMB presented a two-phase blood concentration time course [Fig. 5(A)]. This profile is characterized by a very fast distribution phase and an elimination phase with t1/2 = 11.68 hrs. The AUCo-oo w a s 1.006 A<,g*h/ml, CI = 10.36 ml/min/kg, MRT0_oo = 15.41 hrs and Vs s = 9.587 L/kg. This pharmacokinetic profile indicates that CAMB is removed fast from blood. In addition, the large Vss also indicates a large distribution into the tissues. The results obtained in target tissues showed this extensive distribution and penetration [Fig. 5(B)]. Calculation of pharmacokinetic parameters showed that the main target tissues have a large AMB exposure reflected in the AUC and CMAX values (Table 1), as well as the tissue to blood AUC ratio. The large AMB exposure in liver and spleen suggests involvement of the mononuclear phagocyte system (MPS) in the removal of CAMB. Cochleates are particulates that can be quickly cleared from the circulation by the macrophages of the reticular endothelial system (RES) related to the liver and the spleen. In addition, "physical retention" seems to play a role in the kinetic profile of the lungs due to its capillary nature. Time (hrs) "* 0 10 20 30 40 50 Time (hours) Fig. 5. (A) AMB profile in blood after a single dose (B) IV PK profile of AMB in target tissues, (from Ref. 69, with permission). Cochleates as Nanoparticular Drug Carriers 359 Table 1 Pharmacokinetics parameters for CAMB in different target organs after IV administration to C57BL/6 mice (n = 5 per time point) (From Ref. 69, with permission). Tissue3 Liver Spleen Lung Kidney Heart Intestine Stomach AUCo-oo (/ig*h/g) 474.519 116.388 39.707 12.564 0.970 9.173 8.184 T max (min) 10 2 2 5 5 20 20 *~ max (Mg/g) 8.559 6.633 16.408 1.032 0.478 0.609 0.343 t l/2>-z b (hrs) 75.03 66.71 22.34 21.86 2.82 13.88 20.77 This phenomenon and the mobility of the macrophages seem to cause certain redistribution of cochleates that gives a multi-peak and plateau shape profiles in liver and spleen. Finally, AMB was also detected in bile and intestine contents, suggesting that bile excretion may be an additional elimination route. 2.4. Other potential applications for cochleates 2.4.1. Cochieate for the delivery of antibiotics As cochieate has shown a high affinity to be engulfed by macrophages [Fig. 6(A)] probably due to a dual mechanism, the cochieate essential particulate feature71 and possibly a PS receptor mediated internalization of the cochieate into macrophage.72 Fig. 6. Uptake of amphotericin B cochleates by J774 macrophages as seen by (A) fluorescence microscopy, (B) confocal microscopy (from Ref. 17, with permission). 360 Zarif This particulate system would have potential for the delivery of antibacterial agents such as aminoglycosides and vancomycin.17 Illustration is given by the encapsulation of clofazimine, an anti-TB drug, and tobramycin, an aminoglycoside antibiotic used in treating bacterial infections, both given intraveneously thus far. The cochleate system may possibly offer a new oral way of delivery. 2.4.2. Delivery of clofazimine Clofazimine cochleates were prepared by the Trapping method.18 Clofazimine is a known hydrophobic anti-TB drug, the efficacy of Clofazimine cochleate was assessed by measuring the IC50 in Vero Cells and in bone marrow derived macrophage (BM-M).73 Clofazimine cochleates exhibit a greater decrease in toxicity versus free clofazimine and had a higher efficacy in killing intracellular M. Tuberculosis than free clofazimine:2 Log reduction (CE99) was achieved at 20.9 /xg/ml for cochleates, while free clofazimine was toxic at this concentration. This shows that encapsulation of clofazimine in cochleates potentiates the antimicrobial efficacy of the drug, i.e. when higher concentration of drug can be used because of less toxicity, bactericidal levels of the drug could be attained. 2.4.3. Delivery of tobramycin A recent research work has been published on the possible use of nanocochleates as an oral delivery system for Tobramycin.74 Tobramycin is a well known aminoglycoside antibiotic used in treating bacterial infections, and is usually administered by intravenous (i.v.) infusion, intramuscular (i.m.) injection, or inhalation. This aminogycoside drug is known for its side effects such as mineral depletion (i.e. calcium, magnesium, potassium) after i.v. administration.75,76 In this work, the author described that tobramycin which is positively charged at low pH, will be encapsulated in the inter-bilayer space of cochleates. The fusion of unilamellar liposomes is no longer induced by a metal cation such as Ca2+, but by the organic molecule to be encapsulated. The cochleate cylinders formation has been described by Papahadjoupolos as resulting partly from the intrinsic properties of the calcium cation. Indeed, phosphatidylserine shows considerable selectivity for calcium due to the propensity of calcium to lose part of its hydration shell, and to displace water upon complex formation.19'77 In the cochleate solid crystalline structures formation, calcium plays a crucial role in bringing bilayers together closely through partial dehydration of the membrane surface and the crosslinking of opposing molecules of phosphatidylserine. In our opinion, in this recent work where formation of cochleate is claimed with no calcium present, additional Cochleates as Nanoparticular Drug Carriers 361 relevant physico-chemical evidence on cochleate formation and the localization of the drug in the interbilayer space will be needed. 2.4.4. Cochleate for the delivery of anti-inflammatory drugs As a result of the deep embedding of the molecules in the cochleates structures, drug molecules are hidden from the outside environment. This should have two beneficial effects: one is to hide and protect the molecule from the degradation due to environment; the other is to protect, the environment when needed, from the active molecule when such molecule presents side effects. This is the case of anti-inflammatory drugs, which associates cure to the disturbance of GI tract (stomach for instance). Cochleates were described to act beneficially in this area, reducing the stomach irritation when anti-inflammatory drugs such as aspirin is hidden in the cochleate structure, and administered to a carrageenan rat model for acute inflammation.27,61 2.5. Othet uses of cochlea tes Cochleates were also described as vehicles for nutrients27 as an improved drug and contrast agent delivery system,28 as well as intermediate in the preparation of special liposomes such as Large Unilamellar Vesicles (LUV) and proteoliposomes. In fact, the discovery of the cochleate structures was a result of the desire to prepare LUV by Pr papahadjoupoulos,19'20 which were developed for the delivery of hydrophilic drugs. Proteoliposomes prepared from cochleates intermediates were described for vaccine applications in general,78 and more recently, when containing lipopolysaccharide as a novel adjuvant.79 3. Conclusion Cochleates lipid-based nanocarrier appears to have potential for the oral delivery of bioactive molecules. Future work should be directed towards more fundamental science, as many research aspects of the cochleate drug carrier system are still hardly known (e.g. localization of the drug in lipid bilayers, impact of multivalent cations on the cochleate formation, mechanism of action of cochleate after oral uptake). In addition, the development of friendly analytical assays to monitor the drug localization and loading percentage in cochleates will be desired. This nano drug carrier is currently under development by Biodelivery Sciences International.27 Having the first drug-cochleate in the market place represents a big challenge. For instance, when oral amphotericin B cochleates are ultimately available for patients, thus will provide a new opening in the treatment of systemic fungal infections. 362 Zarif References 1. Chien YW (1992) Novel drug delivery systems. Drugs and the Pharmaceutical Sciences, Vol. 50. Marcel Dekker: New York, NY. 2. Rathbone MJ, Hadgraft } and Michael SR (2003) Modified-release drug delivery technology. Drugs and the Pharmaceutical Sciences, Vol. 126. Marcel Dekker: New York, NY. 3. Charman WN (2000) Lipids, lipophilic drugs and oral delivery-some emerging concepts. JPharmSci 89:967-978. 4. Bowtle W (2000) Lipid formulations for oral drug delivery. Pharm Technol Eur 12(9):20-30. 5. Attwood D (1994) Microemulsions, in Kreutrer J (ed.) Colloidal Drug Delivery Systems. Marcel Dekker: New York, pp. 31-71. 6. Lawrence MJ (1996) Microemulsions as drug delivery vehicles. Curr Opin Colloid Interface Sci 1:826-832. 7. Pouton CW and Charman WN (1997) The potential of oily formulations for drug delivery to the gastrointestinal tract. Adv Drug Del Rev 25:1-2. 8. Muller RH, Mader K and Gohla S (2000) Solid lipid nanoparticles (SLN) for controlled drug delivery. A review of the state of the art. Eur J Pharm Biopharm 50:161-177. 9. Westesen K (2000) Novel lipid-based colloidal dispersions as potential drug administration systems, expectations and reality. Colloid Polym Sci 278:608-618. 10. Lamprecht A, Bouligand Y and Benoit JP (2002) New lipid nanocapsules exhibit sustained release properties for amiodarone. / Control Rel 84:59-68. 11. Lamprecht A, Saumet JL, Roux J and Benoit JP (2004) Lipid nanocarriers as drug delivery systems for Ibuprofen in pain treatment. Intl} Pharm 278(2):407-414. 12. Razzaque MS, Koji T, Kumatori A and Tagushi T (1999) Cisplatin-induced apoptosis in human proximal tubular epithelial cells is associated with the activation of the Fas/Fas ligand system. Histochem Cell Biol 111:359-365. 13. Razzaque MS, Hossain MA, Ahsan N and Tagushi T (2001) Lipid formulations of polyene antifungal drugs and attenuation of associated nephrotoxicity. Nephron 89:251-254. 14. Hossain MA, Maesaki S, Kakeya H, Noda T, Yanagihara K, Sasaki E, Hirakata Y, Tomono K, Tashiro T and Kohno S (1998) Efficacy of NS-718, a novel lipid nanosphereencapsulated amphotericin B, against cryptococcus neoformans. Antimicrob Agents Chemother 42:1722-1725. 15. Otsubo T, Maesaki S, Yamamoto Y, Tomono K, Tashiro T, Seki J, Tomii Y, Sonoke S and Kohno S (1999) In vitro and in vivo activities of NS-718, a new lipid nanosphere incorporating amphotericin B, against Aspergillus Fumigatus. Antimicrob Agents Chemother 43:471-475. 16. Trotta M, Cavalli R, Carlotti ME, Battaglia L and Debemardi F (2005) Solid lipid microparticles carrying Insulin formed by solvent-in-water emulsion-diffusion technique. Int J Pharm 288:281-288. 17. Zarif L (2002) Elongated supramolecular assemblies in drug delivery. / Control Rel Rev 81:7-23. Cochleates as Nanoparticular Drug Carriers 363 18. Zarif L, Graybill JR, Perlin D and Mannino RJ (2000) Cochleates: New lipid-based drug delivery system. / Liposome Res 10(4)523-538. 19. Papahadjopoulos D, Vail WJ, Jacobson K and Poste G (1975) Cochleate lipid cylinders: Formation by fusion of unilamellar lipid vesicles. Biochim Biophys Acta 394(3) :483^91. 20. Papahadjopoulos D (1978) Large unilamellar vesicles (LUV) and method of preparing the same. US Patent 4078052. 21. Papahadjopoulos D, Nir S and Duzgunes N (1990) Molecular mechanisms of calciuminduced membrane fusion. / Bioenerg Biomembr 22(2):157-179. 22. Wilschut J and Papahadjopoulos D (1979) Ca2+ induced fusion of phospholipid vesicles monitored by mixing of aqueous contents. Nature 281(5733):690-692. 23. Mannino RJ and Gould-Fogerite S (1997) Antigen cochleate formulations for oral and systemic vaccination, in Levine MM (ed.) New Generation Vaccines. Marcel Dekker: New York, pp. 229-239. 24. Zarif L and Perlin D (2002) Amphotericin B nanocochleates: From formulation to oral efficacy. Drug Del Technol 2(3):34-37. 25. Zarif L, Graybill JR, Perlin D, Navjar L, Bocanegra R and Mannino RJ (2000) Antifungal activity of amphotericin B cochleates against Candida albicans in a mouse model. Antimicrob Agents Chemother 44(6):1463-1469. 26. Walsh TJ, Viviani MA, Arathoon E, Chiou C, Ghannoum M, Groll AH and Odds FC (2000) New targets and delivery systems for antifungal therapy. MedMycol 38(Supp l):335-347. 27. Biodelivery Sciences: www.biodeliverysciences.com 28. Unger E (2002) Method for delivering bioactive agents using cochleates. US 6403056 Bl. 29. Zarif L, Jin T, Segarra I and Mannino RJ (2001) New cochleate formulations, process of preparation and their use for the delivery of biologically relevant molecules. PCT application WO01 /52817 A2. 30. Zarif L and Mannino RJ (2000) Cochleates: lipid-based vehicles for gene deliveryconcept, achievements and future development, in Habib N (ed.) Cancer Gene Therapy: Past Achievements and Future Challenges. Kluwer Academic/Plenum Publishers: New York, pp. 83-94. 31. Graham I, Gagne J and Silvius JR (1985) Kinetics and thermodynamics of calciuminduced lateral phase separations in phosphatidic acid containing bilayers. Biochemistry 24(25):7123-7131. 32. Archibald DD and Mann S (1994) Self-assembled microstructures from 1,2-ethanediol suspensions of pure and binary mixtures of neutral and acidic biological galactosylceramides. Chem Phys Lipids 69(l):51-64. 33. Holland JW, Hui C, Cullis PR and Madden TD (1996) Polyethylene glycol)-lipid conjugates regulate the calcium-induced fusion of liposomes composed of phosphatidylethanolamine and phosphatidylserine. Biochemistry 35(8):2618-2624. 34. Villardita C, Grioli S, Salmeri G, Nicoletti F and Pennisi G (1987) Multicenter clinical trial of brain phosphatidylserine in elderly patients with intellectual deterioration, Clin Trials ] 24:84-93. 35. Crook TH, Tinklenberg J, Yesavage J, Petrie W, Nunzi MG and Massari DC (1991) Effects of phosphatidylserine in age-associated memory impairment. Neurology 41:644-649. 364 Zarif 36. Engle RR, Satzger W and Gunther W (1992) Double-blind cross-over study of phosphatidylserine vs. placebo in patients with early dementia of the Alzheimer type. Eur Neuropsychopharmacol 2:149-155. 37. Amaducci L (1988) Phosphatidylserine in the treatment of Alzheimer's disease: Results of a multicenter study. Psychopharmacol Bull 24(1):130-134. 38. Cennachi T, Bertoldin T, Farina C, Fiori MG and Crepaldi G (1993) Cognitive decline in the elderly: A double blind, placebo-controlled multicenter study on efficacy of phosphatidylserine administration. Aging 5(2):123-133. 39. Guidin J et al. (1995) Effect of soy lecithin phosphatidylserine (PS) complex on memory impairment and mood in the functioning elderly. Dept Geriatrics, Kaplan Hospital, Rehovot, Israel. 40. Maggioni M, Picotti GB, Bondiolotti GP et al. (1990) Effects of phosphatidylserine therapy in geriatric patients with depressive disorders. Acta Psychiatr Scand 81:265-270. 41. Nerozzi D, Aceti F, Melia E, Magnani A, Marino R, Genovesi G, Amalfitano M, Cozza G, Murgiano S, De Giorgis G, et al. (1987) Phosphatidylserine and Memory Disorders in the Aged. Clin Ter 120(5):399^04. 42. Monteleone P, Beinat L, Tanzillo C, Maj M and Kemali D (1990) Effects of phosphatidylserine on the neuroendocrine response to physical stress in humans. Neuroendocrinology 52:243-248. 43. Monteleone P et al. (1992) Blunting by chronic phosphatidylserine administration of the stress-induced activation of the hypothalamo-pituitary-adrenal axis in healthy men. Eur ] Clin Pharmacol 41:385-388. 44. Funfgeld EW, Baggen M, Nedwidek P, Richstein B and Mistlberger G (1989) Doubleblind study with phosphatidylserine (PS) in parkinsonian patients with senile dementia of Alzheimer's type (SDAT). Prog Clin Res 317:1235-1246. 45. Zarif L and Tan F (2003) Cochleates made with purified soy phosphatidylserine. US2003 / 0219473 Al. 46. Duzgunes N, Nir S, Wischut J, Bentz J, Newton C, Portis A and Papahadjopoulos D (1981) Calcium- and magnesium induced fusion of mixed phosphatidylserine/ phosphatidylcholine vesicles: Effect of ion-binding, f Membr Biol 59:115-125. 47. Portis A, Newton C, Pangborn W and Papahadjopoulos D (1979). Studies on the mechanism of membrane fusion: Evidence for an intermembrane Ca2+-phospholipid complex, synergism with Mg2+, and inhibition by spectrin. Biochemistry 18:780-790. 48. Santangelo R, Paderu P, Delmas G, Chen ZW Mannino R, Zarif L and Perlin D (2000) Efficacy of oral cochleates amphotericin b in a mouse model of systemic candidiasis. Antimicrob Agents Chemother 44(9):2356-2360. 49. Zarif L, Segarra I, Jin T, Scolpino A, Hyra D, Daublin P, Krause S, Perlin DS, Lambros C, Graybill JR and Mannino RJ (1999) Lipid-based cochleate system for oral and systemic delivery of drugs. AAPS Eastern Regional Meeting and Exposition. 50. Segarra I, Hyra-Movshin DA, Chen ZW, Santangelo R, Perlin D, Paderu P, Mannino RJ and Zarif L (2000) AmB Cochleates, a new lipid-based formulation for amphotericin Cochleates as Nanoparticular Drug Carriers 365 B: From IV pharmacokinetics to oral efficacy. Millenial World Congress of Pharmaceutical Sciences, San Franscisco, CA, April, pp. 124. 51. Segarra I, Jin T, Hyra D, Mannino RJ and Zarif L (1999) Oral administration of amphotericin B with a new AmB-cochleate formulation: Tissue distribution after single and multiple oral dose. 1CAAC 39:Abs 1940. 52. Segarra I, Movshin D, Mannino RJ and Zarif L (2000) Pharmacokinetics and tissue distribution of amphotericin B in Mice after oral administration of AmB cochleates, a new effective lipid-based formulation for the oral treatment of systemic fungal infections. ICAAC 40:Abs 861. 53. Segarra I, Chen ZW, Movshin DA, Tan F, Mannino RJ and Zarif L (2000) Tissue distribution of oral amphotericin B lipid-based cochleate formulation: Comparison with AmBisome. 27th International Symposium on Controlled Release ofBioactive Materials, Paris France, pp. 67-68. 54. Zarif L, Segarra I, Jin T, Hyra D and Mannino RJ (1999) Amphotericin B cochleates as a novel oral delivery system for the treatment of fungal infections. 26th International Symposium on Controlled Release ofBioactive Materials. Boston, MA, June 20-23. 55. Perlin D, Santangelo R, Mannino R and Zarif L (2000) Oral delivery of cochleates containing amphotericin B (CAMB) is highly effective in a candidiasis murine model, Focus Fungal Infect. 56. Zarif L, Segarra I, Jin T, Hyra D, Perlin D, Graybill JR and Mannino JR (1999) Oral and systemic delivery of amphotericin B mediated by cochleates. AAPS Annual Meeting and Exposition, November. 57. Zarif L, Jin T, Scolpino A and Mannino RJ (1999) Are cochleates the new lipid-based carrier for oral drug delivery? 39th ICAAC, San Francisco, CA, September 26-29. 58. Delmas G, Perlin D, Chen ZW and Zarif L (2001) Amphotericin B cochleates: Evaluation for the oral treatment of aspergillosis in murine model, The 28th International Symposium of Controlled Release of Bioactive Materials, San Diego, CA, June 23-29, pp. 433^34. 59. Delmas G, Park S, Chen ZW, Tan F, Kashiwazaki R, Zarif L and Perlin DS (2002) Efficacy of orally delivered cochleates containing amphotericin B in a murine model of aspergillosis. Antimicrob Agents Chemother 46(8):2704-2707. 60. Graybill JR, Navjar L, Bocanegra R, Scolpino A, Mannino RJ and Zarif L (2000) A new lipid vehicle for amphotericin B, Abstract, 39th ICAAC, San Franscisco, CA, September, Abs 583. 61. Delmarre D, Lu R, Taton N, Krause-Elsmore S, Gould-Fogerite S and Mannino RJ (2004) Cochleate-mediated delivery: Formulation of hydrophobic drugs into cochleate delivery vehicles: A simplified protocol & bioral formulation kit. Drug Del Techno 4(l):64-69. 62. Ramani K and Balasubramanian S (2003) Fluorescence properties of Laurdan in cochleate phases. Biochim Biophys Acta 1618(l):67-78. 63. Rex JH, Walsh TJ, Sobel JD, Filler SG, Pappas PG, Dismukes WE and Edwards JE (2000) Practice guidelines for the management of candidiasis. Infectious Diseases Society of America. Clin Infect Dis 30(4):662-678. 64. Saag MS, Graybill RJ, Larsen RA, Pappas PG, Perfect JR, Powderly WG, Sobel JD and Dismukes WE (2000) Practice guidelines for the management of cryptococcal disease. Infectious Diseases Society of America. Clin Infect Dis 30(4):710-718. 366 Zarif 65. Stevens DA, Kan VL, Judson MA, Morrison VA, Dummer S, Dening DW, Bennett JE, Walsh TJ, Patterson TF and Pankay GA (2000) Practice guidelines for diseases caused by Aspergillus. Infectious Diseases Society of America. Clin Infect Dis 30(4):696-709. 66. Hiemenz JW and Walsh TJ (1996) Lipid formulations of amphotericin B: Recent progress and future directions. Clin Infect Dis 22(Suppl 2):133-144. 67. Graybill JR, Najvar LK, Bocanegra R, Scolpino A, Mannino RJ and Zarif L (1999). Cochleate: A new lipid vehicle for amphotericin B. ICAAC 39:Abs 2009. 68. Zarif L, Graybill J, Najvar L, Perlin D and Mannino RJ (2000) Amphotericin B cochleates: Novel lipid-based drug delivery system for the treatment of systemic fungal infections., 14th ISHAM World Congress, May 8-12, Buenos Aires, Argentina. 69. Segarra I, Movshin DA and Zarif L (2002) Extensive tissue distribution of amphotericin B after intravenous administration in cochleate vehicle to mice. 29th International Symposium on Controlled Release of Bioactive Materials, Seoul, Korea. 70. Segarra I, Movshin D and Zarif L (2002) Pharmacokinetics and tissue distribution after intravenous administration of a single dose of amphotericin B cochleates, a new lipidbased delivery system. / Pharm Sci 91(8):1827-1837. 71. Legrand P, Vertut-Doi A and Bolard J (1996) Comparative internalization and recycling of different amphotericin B formulations by a macrophage-like cell line. / Antimicrob Chemother 37:519-533. 72. Bratosin D, Mazurier J, Tissier JP, Slomianny C, Estaquier J, Russo-Marie F, Huart JJ, Freyssinet JM, Aminoff D, Ameisen JC and Montreuil J (1997) Molecular mechanism of erythrophagocytosis. Characterization of the senescent erythrocytes that are phagocytized by macrophages. CR Acad Sci Paris Sciences de la Vie/Life Sci 320:811-818. 73. Popescu C, Adams L, Franzblau S and Zarif L (2001) Cochleates potentiate the efficacy of the antimycobacterial drug, clofazimine. ICAAC 41:Abs 2278. 74. Jin T (2003) Cochleates without metal cations as bridging agents. US Patent application 10/636,522. 75. Slayton W, Anstine D, Lakhdir F, Sleasman J and Neiberger R (1996) Tetany in a child with AIDS receiving intravenous tobramycin. South Med J 89:1108-1110. 76. Keating MJ, Sethi MR, Bodey GP and Samaan NA (1977) Hypocalcemia with hypopara thyroidism and renal tubular dysfunction associated with aminoglycoside therapy. Cancer 39:1410-1414. 77. RRC (1990) New (Ed.), Liposomes, a practical approach, IRL Press, Oxford University Press, New York. 78. Gould-Fogerite S, Mazurkiewicz JE, Raska K Jr, Voelkerding K, Lehman JM and Mannino RJ (1989) Gene 84(2):429-438. 79. Perez O, Brach G, Lastre M, Mora N, Del Campo J, Gil D, Zayas C, Acevedo R, Gonzales D, Lopez J, Taboada C and Solis RL (2004) Novel adjuvant based on a proteoliposome-derived cochleate structure containing native polysaccharide as a pathogen-associated molecular pattern. Immunol Cell Biol 82(6):603-610. 17 Aerosols as Drug Carriers N. Renee Labiris, Andrew P. Bosco and My ma B. Dolovich 1. Introduction As the end organ for the treatment of local diseases or as the route of administration for systemic therapies, the lung is a very attractive target for drug delivery (Table 1). The lung provides direct access to the site of disease for the treatment of respiratory illness, without the inefficiencies and unwanted effects of systemic drug delivery. In addition, it provides an enormous surface area and a relatively low enzymatic environment for the absorption of drugs to treat systemic diseases (Table 1). Inhaled medications have been available for many years for the treatment of lung diseases. Inhalational delivery has been widely accepted as being the optimal route of administration of first line therapy for asthmatic and chronic obstructive pulmonary diseases. Drug formulation plays an important role in producing an effective inhalable medication. In addition to being pharmacologically active, it is important that a drug be efficiently delivered into the lungs, to the appropriate site of action and remain in the lungs until the desired pharmacological effect occurs. A drug designed to treat a systemic disease, such as insulin for diabetes, must be deposited in the lung periphery to ensure maximum systemic bioavailability. For gene therapy, anti cancer or anti infective treatment, cellular uptake and prolonged residence in the lungs of the drug may be required to obtain the optimal therapeutic effect. Thus, a formulation that is retained in the lungs for the desired length of time and avoids the clearance mechanisms of the lung may be necessary. The human lung contains airways and approximately 300 million alveoli with a surface area of 140 m2, equivalent to that of a tennis court.1 As a major port of 367 368 Labiris, Bosco & Dolovich Table 1 disease. Advantages of pulmonary delivery of drugs to treat respiratory and systemic Treatment of respiratory diseases Treatment of systemic diseases Deliver high drug concentrations directly to the disease site Minimizes risk of systemic side effects Rapid clinical response Bypass the barriers to therapeutic efficacy, such as poor gastrointestinal absorption and first-pass metabolism in the liver Achieve a similar or superior therapeutic effect at a fraction of the systemic dose. For example, oral salbutamol 2-4 mg is therapeutically equivalent to 100-200 /xg byMDI A non-invasive Needle-free delivery system. Suitable for a wide range of substances from small molecules to very large proteins Enormous absorptive surface area (140 m2) and a highly permeable membrane (0.2 to 0.7 /xm thickness) in the alveolar region. Large molecules with very low absorption rates can be absorbed in significant quantities; the slow mucociliary clearance in the lung periphery results in prolonged residency in the lung. A less harsh, low enzymatic environment Avoids first-pass metabolism. Reproducible absorption kinetics. Pulmonary delivery is independent of dietary complications, extracellular enzymes and inter-patient metabolic differences that affect gastrointestinal absorption. entry, the lung has evolved to prevent the invasion of unwanted airborne particles from entering into the body. Airway geometry, humidity, mucociliary clearance and alveolar macrophages play a vital role in maintaining the sterility of the lung, and consequently, they can be barriers to the therapeutic effectiveness of inhaled medications. The size of the drug particle can play an important role in avoiding the physiological barriers of the lung and targeting to the appropriate lung region (Fig. 1). Nanoparticles are solid colloidal particles ranging in size from 10 to 1000 nm.2 Studies have demonstrated that they are taken up by macrophages, cancer cells, and epithelial cells.3-6 Their small size ensures the particles containing the active pharmacological ingredient will reach the alveolar regions. However, the use of an aerosol delivery system that generates nano-sized particles for inhalation, places these particles at risk of being exhaled, leaving very few drug particles to be deposited in the periphery of the lung. Residence time is not long enough for the particles to be deposited by sedimentation or diffusion.7 DIFFUSION Aerosols as Drug Carriers 369 SEDIMENTATION INITIAL IMPACTION 05 ID 2.0 5.0 AERODYNAMIC DIAMETER pm (Mkrara) Fig. 1. Relationship between particle size and lung deposition. SOB 105 2. Pulmonary Drug Delivery Devices The origin of inhaled therapies can be traced back 4000 years ago to India, where people smoked the leaves of the Atropa belladonna plant to suppress cough. In the 19th and early 20th centuries, asthmatics smoked asthma cigarettes that contained stramonium powder mixed with tobacco to treat the symptoms of their disease. Modern inhalation devices can be divided into three different categories (Fig. 2), the refinement categories (Fig. 2), the refinement of the nebulizer and the of compact portable devices, the pressurized metered dose inhaler (pMDI), and the dry powder inhaler (DPI). The advantages and disadvantages of each are summarized in Table 2. 2.1. Nebulizers Nebulizers have been used for many years to treat asthma and other respiratory diseases. There are 2 basic types of nebulizers, jet and ultrasonic nebulizers. The jet nebulizer functions by the Bernoulli principle by which compressed gas (air or oxygen) passes through a narrow orifice, creating an area of low pressure at the outlet of the adjacent liquid feed tube. This results in the drug solution being drawn up from the fluid reservoir and shatter into droplets in the gas stream. The ultrasonic nebulizer uses a piezoelectric crystal, vibrating at a high frequency (usually 1 to 3 MHz), to generate a fountain of liquid in the nebulizer chamber; the higher the frequency, the smaller the droplets produced. Nebulizers can aerosolize 3 70 Labiris, Bosco & Dolovich Glass Nebulizer (Late 19* century) Hand Bulb Nebulizer (1938) Adaptive Aerosol Delivery Metered Dose Inhalers (MDI) (1956, CFC prcmellant) Metered Dose Liquid Inhalers Breath-Actuated MDI Add-On Devices CFC-Free MDI Dry Powder Inhaler (DPI) Passive Active Fig. 2. Evolution of pulmonary delivery devices. most drug solutions and provide large doses with very little patient coordination or skill. However, treatments using these nebulizers can be time consuming and inefficient, with large amounts of drug wastage e.g. 50% loss with continuously operated nebulizers.8 Most of the prescribed drug never reaches the lung with nebulization. The majority of the drug is either retained within the nebulizer (referred to as residual or dead volume) or released into the environment during expiration. On average, only 10% of the dose placed in a continuous output jet nebulizer is actually deposited in the lungs.8 Advances in technology have led to the development of novel nebulizers that reduce drug wastage and improve delivery efficiency. Breath-enhanced jet nebulizers such as the Pari LC Star, (PARI, Germany) increase aerosol output by directing auxiliary air, entrained during inspiration, through the nebulizer, causing more of the generated aerosol to be swept out of the nebulizer and available for inhalation. Drug wastage during exhalation is reduced to the amount of aerosol produced by the jet airflow rate that exceeds the storage volume of the nebulizer. Adaptive aerosol delivery (Halolite, Medic-Aid, Bognor Regis, UK) monitors a patient's breathing pattern in the first 3 breaths and then targets the aerosol delivery into the first 50% of each inhalation. This ensures that the aerosol is delivered to the patient during inspiration only, thereby eliminating drug loss during expiration that occurs with continuous output nebulizers.9 A number of metered dose liquid inhalers, including AERx (Aradigm, Hayward, CA), Aero- Dose (AeroGen, Sunnyvale, CA) and Respimat (Boehringer Ingelheim, Ingelheim Rhein, Germany), have been developed to produce a fine aerosol in the respirable Aerosols as Drug Carriers 371 Table 2 Advantages and disadvantages of inhalation devices. Inhalation device Advantages Disadvantages Nebulizers (jet, ultrasonic) no specific inhalation technique or coordination required aerosolizes most drug solutions delivers large doses suitable for infants and people too sick or physically unable to use other devices time consuming bulky non-portable contents easily contaminated relatively expensive poor delivery efficiency drug wastage wide performance variation between models and operating conditions pressurized Metered Dose Inhalers (pMDI) Dry Powder Inhalers (DPI) compact portable multi-dose (-200 doses) inexpensive sealed environment (no degradation of drug) reproducible dosing compact portable breath actuated easy to use no hand-mouth coordination required inhalation technique and patient coordination required high oral deposition maximum dose of 5 mg limited range of drugs available respirable dose dependent on IFR* humidity may cause powders to aggregate and capsules to soften dose lost if patient inadvertently exhales into the DPI most DPIs contain lactose *IFR = Inspiratory Flow Rate range by forcing the drug solution through an array of nozzles, using vibrating mesh or electronic micropump platforms with 30 to 75% of the emitted dose being deposited in the lungs.10,11 2.2. Metered-dose inhalers The pressurized metered-dose inhaler (pMDI) was a revolutionary invention that overcame the problems of the hand-bulb nebulizer, and it is the most widely used aerosol delivery device today. The pMDI emits a drug aerosol driven by 372 Labiris, Bosco & Dolovich propellants, such as chlorofluorocarbons (CFC) and more recently, hydrofluoroalkanes (HFAs) through a nozzle at high velocity (>30m/sec). pMDIs deliver only a small fraction of the drug dose to the lung. Typically, only 10 to 20% of the emitted dose is deposited in the lung.12 The high velocity and large particle size of the spray causes approximately 50% to 80% of the drug aerosol to impact in the oropharygeal region.13 Hand-mouth discoordination is another obstacle in the optimal use of the pMDI. Crompton and colleagues14 found 51% of patients experienced problems coordinating the actuation of the device with inhalation, 24% of patients halted inspiration upon firing the aerosol into the mouth, and 12% inspired through the nose instead of the mouth when the aerosol was actuated into the mouth. The delivery efficiency of a pMDI depends on a patient's breathing pattern, inspiratory flow rate and hand-mouth coordination. The studies by Bennett15 and Dolovich16 demonstrated that for any particle size between 1 to 5 /tm mass median aerodynamic diameter (MMAD), deposition was more dependent on inspiratory flow rate than any other variable. Fast inhalations (>60 L/min) result in a reduced peripheral deposition because the aerosol is more readily deposited by inertial impaction in the conducting airway and oropharyngeal regions. When aerosols are inhaled slowly, deposition by gravitational sedimentation in peripheral lung regions are enhanced.17 Peripheral deposition has also been shown to increase with an increase in tidal volume and a decrease in respiratory frequency. As the inhaled volume is increased, aerosols are able to penetrate more distally into the lungs.18 A period of breath holding on completion of inhalation enhances deposition of particles in the periphery, thus preventing the particles from being exhaled during the expiratory phase. Thus, the optimal conditions for inhaling pMDI aerosols are from a starting volume equivalent to the functional residual capacity, the actuation of the device at the start of inhalation, inspiratory flow rate of <60 L/min, followed by a 10 second breath-hold at the end of inspiration.17,19 Spacer tubes, valved holding chambers and mouthpiece extensions have been developed to eliminate coordination requirements and reduce the amount of drug deposited in the oropharynx, by decreasing the particle size distribution and slowing the aerosol's velocity. Spacer geometry and materials of manufacture influence the quality and quantity of aerosol available. The aerosols from a pMDI and the holding chamber are finer than that with the pMDI alone, with an approximate 25% decrease in the mass median aerodynamic diameter (MMAD), compared with the original aerosol.20,21 This finer aerosol is more uniformly distributed in the normal lung, with increased delivery to the peripheral airway. However, in patients with airway obstructions, the addition of a holding chamber to the pMDI may not change the distribution of the aerosol.22 Aerosols as Drug Carriers 373 2.3. Dry powder inhalers Dry powder inhalers (DPIs) were designed to eliminate the coordination difficulties associated with the pMDI. There are a wide range of DPI devices on the market from single-dose devices loaded by the patient (e.g. Aerolizer from Novartis, Rotahaler from GSK, Ware UK) to multi unit dose devices provided in a blister pack (e.g. Diskhaler, GSK, Ware UK), multiple unit doses sealed in blisters on a strip that moves through the inhaler (e.g. Diskus, GSK, Ware UK) or reservoir-type (bulk powder) systems (e.g. Turbuhaler, AstraZeneca, Lund Sweden). Lung deposition varies among the different DPIs. Approximately 12% to 40% of the emitted dose is delivered to the lungs with 20 to 25% of the drug being retained within the device.10,23,24 Poor drug deposition with DPIs can be attributed to inefficient deaggregation of the fine drug particles from coarser carrier lactose particles or drug pellets. Slow inspiratory flow rate, high humidity and rapid, large changes in temperature are known to affect drug deaggregation and hence the efficiency of pulmonary drug delivery with DPIs.25,26 With most DPIs, drug delivery to the lungs is augmented by fast inhalation. Borgstrom and colleagues27 demonstrated that increasing inspiratory flow from 35L/min to 60L/min through the Turbuhaler7, increased the total lung dose of terbutaline from 14.8% of nominal dose to 27.7%. This is in contrast to the MDI which requires slow inhalation and breath holding to enhance lung deposition of the drug. Each DPI has a different air flow resistance that governs the required inspiratory effort.28,29 The higher the resistance of the device, the more difficult it is to generate an inspiratory flow great enough to achieve the maximum dose from the inhaler.30-32 However, deposition in the lung tends to increase when using high resistance inhalers.32-36 Active DPIs are being investigated to reduce the importance of a patient's inspiratory effort. By adding either a battery driven propeller that aids in the dispersion of the powder (Spiros, Elan Pharmaceuticals, San Diego, CA), or using compressed air to aerosolize the powder and converting it into a standing cloud in a holding chamber, the generation of a respirable aerosol becomes independent of a patient's inspiratory effort (Inhance Pulmonary Delivery System, Nektar Therapeutic, San Carlos, CA). 3. Aerosol Particle Size Aerosol particle size is one of the most important variables in defining the dose deposited and the distribution of drug aerosol in the lung (Fig. 3). Fine aerosols are distributed on peripheral airways, but deposit less drug per unit surface area than larger particle aerosols which deposit more drug per unit surface area, but on 3 74 Labiris, Bosco & Dolovich (a) JOO*. FREQUENCY DISTRIBUTION % 50 _ NUMBERVOLUME {MASS] 0.1 i )0 100 AERODYNAMfC DIAMETER JJm (b) CUMULATIVE DISTRIBUTION 100 | - / > % 50 NUMBER-^ /--VOLUME CMAS5) / MMAD = 2.25 ftm _L 0.1 i 10 AERODYNAl-flC DIAMETER Jtm 100 Fig. 3. Frequency (a) and cumulative (b) distribution curves for Beclovent MDI used with an Aerochamber, in terms of number of particles and volume (mass) of particles vs. particle aerodynamic diameter. The volume distribution curves are displaced to the right of the number distribution curves. The smaller number of large particles within the aerosol carry the greater mass of the drug; this is reflected in the larger, second peak of the volume distribution curve, which corresponds to the smaller second peak of the number distribution curve. MMAD is read from the cumulative distribution curve at the 50% point and if the distribution is log-normal, the GSD can be calculated as the ration of the diameter at the 84.1% point to the MMAD. Particle distribution was measured using the Anderson Cascade Impactor.105 the larger, more central airways.37 Most therapeutic aerosols are nearly always heterodisperse, consisting of a wide range of particle sizes. These aerosols are described by the log-normal distribution, with the log of the particle diameters plotted against particle number, surface area or volume (mass) on a linear or probability scale and expressed as absolute values or cumulative %. Since delivered dose is very important when studying medical aerosols, particle number may be misleading as smaller particles contain less drug than larger ones. Particle size is defined from this distribution by several parameters. Mass median diameter of an aerosol refers to the Aerosols as Drug Carriers 375 particle diameter that has 50% of the aerosol mass residing above and 50% of its mass below it. The aerodynamic diameter relates the particle to the diameter of a sphere of unit density that has the same settling velocity as the particle of interest, regardless of its shape or density. MMAD is read from the cumulative distribution curve at the 50% point (Fig. 3). Geometric standard deviation (GSD) is a measure of the variability of the particle diameters within the aerosol, and is calculated from the ratio of the particle diameter at the 84.1% point on the cumulative distribution curve to the MMAD. For a log-normal distribution, the GSD is the same for the number, surface area or mass distributions. A GSD of 1 indicates a monodispersed aerosol, while a GSD of > 1.2 indicates a heterodispersed aerosol. Particles can be deposited by inertial impaction, gravitational sedimentation or diffusion (Brownian motion), depending on their size. While deposition occurs throughout the airways, inertial impaction usually occurs in the first 10 generations of the lung, where air velocity is high and airflow is turbulent.38 Most particles above 10 /xm are deposited in the oropharyngeal region with a large amount impacting on the larynx, particularly when the drug is inhaled from devices requiring a high inspiratory flow rate (DPIs) or when the drug is dispensed from a device at a high forward velocity (MDIs).39,40 The large particles are subsequently swallowed and contributed minimally, if at all, to the therapeutic response. In the tracheobronchial region, inertial impaction also plays a significant role in the deposition of particles, particularly at bends and airway bifurcations. Deposition by gravitational sedimentation predominates in the last 5 to 6 generation of airways (smaller bronchi and bronchioles), where air velocity is low.38 In the alveolar region, air velocity is negligible and thus the contribution to deposition by inertial impaction is also negligible. Particles in this region have a longer residence time and are deposited by both sedimentation and diffusion. Particles not deposited during inhalation are exhaled. Deposition due to sedimentation affects particles down to 0.5 ^tm in diameter, whereas below 0.5 /xm, the main mechanism for deposition is by diffusion. Targeting the aerosol to conducting or peripheral airways can be accomplished by altering the particle size of the aerosol. It is difficult to predict the actual site of deposition, since airway calibre and anatomy differ among people. However, in general, aerosols with a MMAD of 5 to 10 /xm are mainly deposited in the large conducting airways and the oropharyngeal region.41 Particles 1 to 5 /xm in diameter are deposited in the small airways and alveoli with greater than 50% of the 3 /tm diameter particles being deposited in the alveolar region. In the case of pulmonary drug delivery for systemic absorption, aerosols with a small particle size would be required to ensure peripheral penetration of the drug.42 Particles <3 /xm have approximately 80% chance of reaching the lower airways, with 50 to 60% being deposited in the alveoli.43'44 Nanoparticles <100nm are deposited mainly in the alveolar region. 376 Labiris, Bosco & Dolovich 4. Targeting Drug Delivery in the Lung The therapeutic effect of aerosolized therapies is dependent on the dose deposited and its distribution within the lung. If a drug aerosol is delivered at a suboptimal dose or to a part of the lung, devoid of the targeted disease or receptors, the effectiveness of therapy may be compromised. For example, the receptors for the fc agonist, salbutamol and the muscarine (M3) agonist, ipratropium bromide, are not uniformly distributed throughout the lung. Autoradiographic studies have shown P2 adrenergic receptors are present in high density in the airway epithelium from the large bronchi to the terminal bronchioles. Airway smooth muscle has a lower /S-receptor density, greater in the bronchioles than bronchi.45 However, greater than 90% of all /3 receptors are located in the alveolar wall, a region where no smooth muscle exists and whose functional significance is unknown. Another autoradiographic study has shown a high density of M3 receptors in submucosal glands and airway ganglia, and a moderate density in smooth muscles throughout the airways, nerves in intrapulmonary bronchi and in alveolar walls.46 The location of these receptors in the lung suggests that ipratropium bromide needs to be delivered to the conducting airways, while salbutamol requires a more peripheral delivery to the medium and small airways to produce a therapeutic effect. Since particle size affects the lung deposition of an aerosol, it can also influence the clinical effectiveness of a drug. Rees et al. reported the varying clinical effect of 250 /xg of aerosolized terbutaline from a pMDI, given in three different particle sizes of <5 /xm, 5 to 10 /xm, and 10 to 15 /xm.47 In asthmatics, the greatest increase in forced expiratory volume in one second (FEVi) was found with the smallest particle size (<5/xm), suggesting that the smaller particle aerosol was considerably more effective than larger particle size aerosols in producing bronchodilation, since it has the best penetration and retention in the lungs in the presence of airway narrowing. Using three monodisperse salbutamol aerosols (MMAD of 1.5 /xm, 2.8 /xm, 5 Aim), Zanen and colleagues demonstrated in patients with mild to moderate asthma that the 2.8 /xm particle size aerosol produced a superior bronchodilation, compared with the other two aerosols.48 In patients with severe airflow obstruction (FEVi < 40%), Zanen et al. demonstrated that the optimal particle size for /J2 agonist or anticholinergic aerosols is approximately 3 /xm.49 They examined the effect on lung function of equal doses of three different sizes of monodisperse aerosols, 1.5 /xm, 2.8 /xm and 5 /xm, of salbutamol and ipratropium bromide. Their findings suggest that small particles penetrate more deeply into the lung and more effectively dilate the small airways than larger particles, which are filtered out in the upper airways. The 1.5 /xm aerosol induced significantly less bronchodilation than the 2.8 /xm aerosol, suggesting that this fine aerosol may be deposited too peripherally to be effective, since smooth muscle is not present in the alveolar region. Aerosols as Drug Carriers 377 The optimal site of deposition in the respiratory tract for aerosolized antibiotics depends on the infection being treated. Pneumonias represent a mixture of purulent tracheobronchitis and alveolar infection. Successful therapy would theoretically require the antibiotic to be evenly distributed throughout the lungs. However, those confined to the alveolar region would most likely benefit from a greater peripheral deposition. Pneumocystis carinii pneumonia, the most common life-threatening infection among patients infected with HIV, is found predominately within the alveolar spaces, with relapses occurring in the apical region of the lung after treatment with inhaled pentamidine given as a 1 fim MMAD aerosol.50 The mechanism suggested for this atypical relapse is the poorer apical deposition of the aerosol. Regional changes in intrapleural pressure result in the lower lung regions receiving relatively more of the inspired volume than the upper lung, when sitting in an upright position or standing. This influence on deposition has been shown to occur in an experimental lung model, analyzing sites of aerosol deposition in a normal lung. The experiment showed a 2:1 ratio in the overall deposition for a 4 /xm aerodynamic diameter aerosol between the lower and upper lobes when in the upright position.51 Chronic lung infection with Pseudomonas aeruginosa, in patients with cystic fibrosis or non-CF bronchiectasis, resides in the airway lumen with limited invasion of the lung parenchyma.52'53 Infection starts in the smaller airways, the bronchioles, and moves into the larger airways. The optimal site of deposition for inhaled antimicrobial therapy would, therefore, be a uniform distribution on the conducting airways. Mucus plugs in the bronchi and bronchioles may prevent deposition of even small particle aerosols in regions distal to the airway obstruction, possibly the regions of highest infection, and thereby limiting the therapeutic effectiveness of the aerosolized antibiotic.54-56 Until recently, aerosol drug delivery has been limited to topical therapy for the lung and nose. The major contributing factor to this restriction was the inefficiencies of available inhalation devices that deposit only 10% to 15% of the emitted dose in the lungs. While appropriate lung doses of steroids and bronchodilators can be achieved with these devices, for systemic therapies, large amounts of the drug are necessary to achieve therapeutic drug levels systemically. Recent advances in aerosol and formulation technologies have led to the development of delivery systems that are more efficient and that which produce small particle aerosols, allowing higher drug doses to be deposited in the alveolar region of the lungs, where they are available for systemic absorption. Most macromolecules cannot be administered orally because proteins are digested before they are absorbed into the bloodstream. In addition, their large size prevents them from naturally passing through the skin or nasal membrane; therefore, they cannot be administered intranasally or transdermally without the 378 Labiris, Bosco & Dolovich use of penetration enhancers. Thus, the easiest route of administration for proteins has been through intravenous or intramuscular/subcutaneous injection. It has been known for many years that proteins can be absorbed from the lung as demonstrated with insulin in 1925.57 Macromolecules < 40 kiloDaltons (kDa) (<5-6nm in diameter) appear rapidly in the blood following inhalation into the airways. Insulin which has a molecular weight (mw) of 5.7 kDa and a diameter of 2.2 nm peaks in the blood 15 to 60 min after inhalation.58-62 Macromolecules >40 kDa (>5- 6 nm in diameter) are slowly absorbed over many hours; inhaled albumin (68 kDa) and alphai-antitrypsin (45-51 kDa) have a Tmax of 20hrs and between 12 to 48hrs respectively.63 The lung is the only organ through which the entire cardiac output passes. Before the inhaled drug can be absorbed into the blood from the lung periphery, it has several barriers to overcome such as lung surfactant, surface lining fluid, epithelium, interstitium and basement membrane, and the endothelium. Drug absorption in the lung periphery is regulated by a thin alveolar-vascular permeable barrier. An enormous alveolar surface area with epithelium, consisting of a thin single cellular layer (0.2 to 0.7 /xm thickness), promotes efficient gas exchange through passive transport, but also provides a mechanism for efficient drug delivery into the bloodstream.64 Although the mechanism of absorption is unknown, it has been hypothesized that macromolecules either pass through the cells via absorptive transcytosis (adsorptive or receptor mediated), paracellular transport between bijunctions or trijunctions or through large transitory pores in the epithelium caused by cell injury or apoptosis.65 Thus, the high bioavailability of macromolecules deposited in the lung (10 to 200 times greater than nasal and gastrointestinal values) may be due to its enormous surface area, very thin diffusion layer, slow surface clearance and anti-protease defense system. 5. Clearance of Particles from the Lung Like all major points of contact with the external environment, the lung has evolved to prevent the invasion of unwanted airborne particles from entering into the body. Airway geometry, humidity and clearance mechanisms contribute to this filtration process. The challenge in developing therapeutic aerosols is to produce an aerosol that eludes the lung's various lines of defense. 5.1. Airway geometry and humidity Progressive branching and narrowing of the airways encourages impaction of particles. The larger the particle size, the greater the velocity of incoming air, while the greater the bend angle of bifurcations and the smaller the airway radius, the Aerosols as Drug Carriers 379 greater the probability of deposition by impaction.66 Drug particles are known to be hygroscopic and grow in size in high humidity environments, such as the lung which has a relative humidity of approximately 99.5%. The addition and removal of water can significantly affect the particle size and thus deposition of a hygroscopic aerosol.67 A hygroscopic aerosol that is delivered at relatively low temperature and humidity into one of high humidity and temperature would be expected to increase in size when inhaled into the lung. The rate of growth is a function of the initial diameter of the particle, with the potential for the diameter of fine particles less than 1 /xm to increase 5-fold, compared with 2 to 3-fold for particles greater than 2 /xm.68 The increase in particle size above the initial size should affect the amount of drug deposited, and particularly, the distribution of the aerosolized drug within the lung. Ferron and colleagues have predicted that for initial sizes between 0.7 /xm and 10 /xm, total deposition of hygroscopic aerosols increases by a factor of 2.69 For particles with an initial size of 1 /xm, Xu and Yu were able to predict changes in the distribution pattern due to particle growth.70 The calculations showed a shift from deposition due to sedimentation to primarily impaction on more central airways.69 5.2. Lung clearance mechanisms Once deposited in the lungs, inhaled drugs are either cleared from the lungs, absorbed into the circulatory or lymphatic systems, or metabolized. Drug particles deposited in the conducting airways are primarily removed through mucociliary clearance, and to a lesser extent, are absorbed through the airway epithelium into the blood or lymphatic system. Ciliated epithelium extends from the trachea to the terminal bronchioles. The airway epithelial goblet cells and submucosal glands secrete mucus forming a two-layer mucus blanket over the ciliated epithelium: a low-viscosity periciliary or sol layer covered by a high-viscosity gel layer. Insoluble particles are trapped in the gel layer and moved towards the pharynx (and ultimately to the gastrointestinal tract) by the upward movement of mucus generated by the metachronous beating of cilia. In the normal lung, the rate of mucus movement varies with the airway region and is determined by the number of ciliated cells and their beat frequency. Movement is faster in the trachea than in the small airways, and is affected by factors influencing ciliary functioning and the quantity and quality of the mucus.40'71 For normal mucociliary clearance to occur, airway epithelial cells must be intact, ciliary structure and activity normal, the depth and chemical composition of the sol layer optimal, and the rheology of the mucus within the physiological range. Mucociliary clearance is impaired in lung diseases such as immotile cilia syndrome, bronchiectasis, cystic fibrosis and asthma.72 In immotile cilia syndrome and bronchiectasis, the ciliary function can be 380 Labiris, Bosco & Dolovich either impaired or nonexistent. In cystic fibrosis, the ciliary structure and function are normal, however, the copious amounts of thick, tenacious mucus present in the airways impairs their ability to clear the mucus effectively73 In these diseases, clearance of aerosolized drugs deposited in the conducting airways is generally decreased and secretions are cleared from the lung by cough.74-76 In addition to mucociliary clearance, soluble particles can also be removed by absorptive mechanisms in the conducting airways.77 Lipophilic molecules pass easily through the airway epithelium via passive transport. Hydrophilic molecules cross via extracellular pathways such as tight junctions or by active transport via endocytosis and exocytosis.78 From the submucosal region, particles are absorbed either into systemic circulation, bronchial circulation or lymphatic systems. Drugs deposited in the alveolar region may be phagocytosed and cleared by alveolar macrophages or absorbed into the pulmonary circulation. Alveolar macrophages are the predominant phagocytic cell for the lung defense against inhaled microorganisms, particles and other toxic agents. There are approximately 5 to 7 alveolar macrophages per alveolus in the lungs of healthy, non-smokers.79 Macrophages phagocytose insoluble particles deposited in the alveolar region are either cleared by the lymphatic system or moved into the ciliated airways along currents in alveolar fluid and then cleared via the mucociliary escalator.65 This process can take weeks or months to complete.7 As discussed above, soluble drug particles deposited in the alveolar region can be absorbed into the systemic circulation. The pulmonary epithelium appears to be more resistant to soluble particle transport than to the endothelium or the interstitium.42 The lung-blood barrier may behave as a molecular sieve, allowing the passage of small solutes but restricting the passage of macromolecules. Conhaim and colleagues proposed that the lung barrier was best fitted to a three pore size model, including a small number (2%) of large-sized pores (400 nm pore radius), 30% of medium-sized pores (40 nm radius) and 68% of small-sized pores (1.3 nm).80 The rate of protein absorption from the alveoli is size dependent. Effros and Mason demonstrated an inverse relationship between alveolar permeability and molecular weight.42 In rats, after intratracheal instillation of DDAVP (1-desamino- 8-D-arginine vasopressin) (raw = 1.1 kDa), peak serum DDAVP levels occurred at 1 hr compared with 16 to 24hrs after the intratracheal instillation of albumin (mw = 67 kDa).43 However, some proteins are cleared from the lung more rapidly than expected for their size. After intratracheal instillation or aerosolization of human growth hormone (mw = 22 kDa), peak serum levels were observed between 0.5 to 4 hrs, indicating a rapid, saturable clearance from the lung that is suggestive of receptor-mediated endocytosis.65 Vasoactive intestinal polypeptide (VIP) is believed to be completely degraded during the passage across the pulmonary epithelium and into the bloodstream.81 Aerosols as Drug Carriers 381 Nanoparticles can pass rapidly into the systemic circulation. The distribution of radioactivity, after the inhalation of a 99mTechnetium (Tc)-labeled ultrafine carbon particles (5 to 10 nm), was detected in the blood one min post-inhalation and peaked between 10 and 20 min. This blood radioactivity level was sustained up to 60 min. 8% of the initial lung radioactivity was measured in the liver 5 min postadministration and remained stable over time. The rapidity of the appearance of radioactivity systemically makes the translocation from the lung unlikely due to phagocytosis, by macrophages or endocytosis by epithelial and endothelial cells, but by passive diffusion.82 6. Nanoparticle Formulations for Inhalation Delivery of nano-sized aerosols to the lung may result in very little drug being deposited in the lung. The majority of particles <500nm inhaled will not have enough residence time in the lung to deposit, and therefore will be exhaled (Fig. 1). However, if the nanoparticles were delivered in larger carrier particles, they could be sufficiently deposited in the lung. The carrier particle would dissolve after contact with the lung surface fluid, releasing the nanoparticle at the target tissue or cells. Sham and colleagues demonstrated that nanoparticles (173 to 242 nm) could be delivered into the lung in larger respirable lactose carrier particles produced by spray-drying.83 The dry powder containing the nanoparticles had a MMAD of 3.0 /xm. pMDI formulations are typically micronized drugs in the 2 to 3 /xm range suspended in a hydrofluoroalkane (HFA) propellant. Solution pMDI such as QVAR produce smaller drug particles on propellant evaporation, resulting in better deposition and distribution than a micronized formulation.84 However, for insoluble drug particles in the propellant, the efficiency of pMDI is limited. A study by Dickinson et al. proposed the use of nanoparticles suspended in propellant as a method of increasing the delivery efficiency of insoluble drugs in pMDIs.85 They produced hydrophilic nanoparticles using a reverse phase microemulsion technique that captures nanoparticles by snap freezing, followed by freeze-drying. The nanoparticles of pure drug (salbutamol) and the drug in a non-polymer matrix (lecithin-based), with and without lactose, were dispersed in HFA-227 and in aerosol performance assessed by cascade impaction. The size of the salbutamol nanoparticles ranged from 34 to 216 nm. Dispersion of the nanoparticles in a HFA-227:hexane (95:5 v/v) blend resulted in a homogeneous fine suspension that showed no signs of sedimentation or creaming over several months. Rapid release of salbutamol from the nanoparticle was observed (approximately 4 min) as expected from the large surface area of the particles and the high water solubility of the drug. A high fine particle fraction (ex-device, % < 5.8 /xm) of 58.3% to 65.5% and a low MMAD 382 Labiris, Bosco & Dolovich (1.2 to 1.5 /u.m) were observed with the nanoparticle formulations. This data suggests that a high fraction of the nanoparticles would be distributed in the alveolar region of the lung and represents the best aerosol that can be produced using a pMDI. Budesonide is a potent corticosteroid used as an inhaled anti inflammatory agent to treat asthma. It is available as a dry powder inhaler and as a suspension for inhalation with a nebulizer. A new formulation for nebulization has been developed that contains nanocrystals of budesonide that give the suspension solution-like qualities.86 The particles are 75 to 300 nm in diameter, compared with 4400 nm for the marketed budesonide suspension (Pulmicort Respules, AstraZeneca). In a randomized crossover study, 16 healthy volunteers were given the nanocrystal budesonide formulation (0.5 mg and 1.0 mg doses), Pulmicort respules and placebo via nebulization using a Pari LC jet nebulizer. Nebulization times were shorter for the nanocrystal formulation, compared with Pulmicort respules (~7.1 min vs. 8.7 min). Similar AUCs were observed with the formulations, suggesting similar pulmonary absorption. However, a higher Cmax (1212pg/mL vs. 662pg/mL) and shorter Tmax (8.4 min vs. 14.4 min) for nanocrystal budesonide compared with the same dose of Pulmicort, suggests a more rapid drug delivery or absorption with the nanocrystal formulation. 6.1. Diagnostic imaging Radiolabeled nanoparticles have been used for many years in pulmonary ventilation studies.87 Ultrafine 99mTc labeled carbon particles (Technegas) is a relatively new advance in ventilation scintigraphy.88 Technegas (Vita Medical Ltd., Sydney Australia) consists of nanoparticles of carbon with a diameter of approximately 5 nm, that behaves more like a 0.2 /tm particle.89 Technegas is generated by the electrostatic heating of a graphite crucible to 2500°C in which a saline solution of 99mTc-pertechnetate had been placed and dried. The aerosol is dispersed in a lead-lined chamber in an atmosphere of 100% argon gas that is then inhaled by the patient. It is deposited in alveoli by inhalation and distributes similarly as the inert gas radioisotopes. Once they are inhaled, the particles adhere to the alveolar structures without appreciable movement for at least 40 min.88 Pulmonary delivery of nanoparticles is also being investigated for lymphoscintigraphy to assess the spread of or the staging of lung cancer. Lung cancer usually exhibits metastasis proliferation, spreading through the lymphatic system and the blood circulation. Lymphatic drainage is responsible for the alveolar clearance of the deposited particulates and drugs up to a certain particle diameter (500 nm).90 Thus, radiolabeled nanoparticles could be used to visualize the lymph nodes to determine the presence of tumors. Aerosols as Drug Carriers 383 The lymphatic uptake of solid lipid nanoparticles has also been studied as an imaging method to stage lung cancer. The lipid nanoparticles were radiolabeled with the lipophilic tracer, D,L-hexamethylpropylene amine oxime (HMPAO), tagged with 99 m-Tc. The lipid nanoparticles were prepared by the melted homogenization method and had a mean diameter of 200 nm.90 The radiolabeled nanoparticles were aerosolized using an ultrasonic nebulizer and delivered to rats until 200,000 cpm was achieved over the lung. After inhalation, the total activity in the lung was observed, followed by a fast clearance rate (ti/2 = lOmin) that decreases activity in the lung to 25% of the total dose. Asignificant uptake (16.7%) was detected in the regional lymph nodes during the first 45 to 60 min, suggesting that aerosol delivery to the lungs of solid lipid nanoparticles could be used as an effective colloidal carrier for lymphoscintigraphy. Drainage into the lymph nodes following the lung instillation of nanoparticles of insoluble iodinated CT x-ray contrast agents was studied in beagle dogs.91 Nanoparticles of the contrast agent were prepared by microfluidization. A particle size of 150 to 200 nm was achieved. The nanoparticles were suspended in 2 different surfactant solutions. 1.5 mL of the suspension was instilled using a fiber optic bronchoscope at specific sites in the small airways and alveoli. The nanoparticles were transported from the lung to the draining lymph nodes, 6 to 9 days post instillation as visible on the CT radiographs. No adverse clinical signs were observed in the dogs. However, microscopic lung lesions were observed at the instillation sites for both formulations and vehicle. The lesions consisted of inflammatory infiltrates, mainly macrophages, in intra-alveolar, interstitial and perivascular locations. A few small sites had fibrosis and granulomatous nodules with the destruction of the lung parenchyma. The presence of foamy macrophages was observed in the lymph nodes. The microscopic findings suggest that instillation of these nanoparticles of contrast agent may be harmful to the lung. The authors suggested that administering the nanoparticles as an aerosol, rather than by instillation, would prevent high concentrations in focal areas believed to be responsible for these lesions. 6.2. Vaccine delivery Mucosal vaccine administration is an attractive method of inducing an immune response, since many pathogens invade the body through mucosal surfaces in the nose, lung and gut. As it is the first contact point, the mucosa has developed barriers to protect the body. The mucosa associated lymphoid tissue (MALT) is one of these barriers. It contributes 80% of the immunocytes and secretes more immunoglobulins than any other organs in the body.92 Antigens are delivered locally in the respiratory tract to nasal-associated and bronchus-associated lymphoid tissues (NALT and BALT, respectively) and a mucosal immunity is induced. Using nanoparticles, 384 Labiris, Bosco & Dolovich systemic immunity may also be induced. Several studies have investigated the use of nanoparticles as carriers for the nasal delivery of vaccines. Using tetanus toxoid as a model antigen, Vila and colleagues have studied the use of chitosan nanoparticles as well as polyethyleneglycol and polylactic acid (PEG-PLA) nanoparticles as nasal vaccine carriers.93,94 They compared PEG-PLA nanoparticles with PLA alone.94 Tetanus toxoid was entrapped in the hydrophobic PLA core and protected from interacting with enzymes such as lysozymes, by a hydrophilic PEG coating. Upon incubation with lysozymes in vitro, PLA particles aggregate and do not reach the epithelium, whereas PEG-PLA nanoparticles remain stable and size unmodified. The nanoparticles were produced using a double emulsion technique. PEG-PLA tetanus toxoid nanoparticles had a similar diameter to the PLA particles (196 nm vs. 188 nm), but had a lower loading efficiency of 33.4% compared with 48.1 % with PLA. The IgG antibody response induced by PEG-PLA was superior at weeks 2 to 24, after intranasal instillation of 30 fig of tetanus toxoid (10 fil per nostril) on days 1, 8 and 15 in male BALB/c mice. In a similar study, the same group compared radiolabeled PEG-PLA, PEG-PLA with gelatin stabilizer to radiolabeled PLA encapsulated tetanus toxoid. They reported that 1 hr after intranasal administration, PEG-PLA nanoparticles produced a radioactivity level 10-fold higher in the blood than PLA which remained constant for 24 hrs. The radioactivity detected in the lymph nodes, lungs, liver and spleen was 3 to 6 fold higher for PEG-PLA than PLA nanoparticles 24 hrs post instillation. The results of this work suggest that the PEG-PLA nanoparticles are partially taken up by the M cells of the NALT, as well as being transported to the submucosa and drained into the lymphatic system and blood stream.95 Recent work by the same group has investigated the potential use of chitosan nanoparticles for nasal administration of vaccines.93 Chitosan is a hydrophilic natural polysaccharide that is biodegradable and has mucoadhesive properties. The nanoparticles are formed spontaneously by adding the counter anion sodium TPP into the chitosan solution, without the use of energy sources or organic solvents required for the production of PEG-PLA nanoparticles. Again, using tetanus toxoid as the model antigen, the investigators studied the effect of chitosan dose (200 fig and 70 fig) and molecular weight (23,38 or 70 kDa) on the efficacy of the nanoparticles. The nanoparticles produced were 300 to 350 nm and had a positive surface charge (+40 mV). The loading efficiency of tetanus toxoid was 50 to 60%, irrespective of the molecular weight of chitosan. In vitro, the formulations exhibited a rapid release over the initial 2 hrs followed by a slow release for 16 days, with the greater initial release at lower molecular weights of chitosan. 30 or 10 fig of antigen (associated with 200 and 70 fig of chitosan) was given intranasally to BALB/c mice on days 1, 8 and 15. The IgG levels induced by the nanoparticles were significantly higher than those elicited by free tetanus toxoid. The response lasted for the 24 weeks studied with the IgG titres increasing over time. Anti-tetanus IgA titers were detected in the saliva, bronchoalveolar and intestinal lavage fluids 24 weeks post Aerosols as Drug Carriers 385 administration. The results were independent of the administered dose and were significantly higher for the nanoparticle than the free tetanus toxoid. Jung and colleagues evaluated tetanus toxoid-loaded polymer nanoparticles as potential nasal vaccine carriers in mice.96 The nanoparticles were produced with various diameters (100 nm, 500 nm) using a novel polyester, sulfobutylated poly (vinyl alcohol)-graft-poly(lactide-co-glycolide), SB(43)-PVAL-g-PLGA. The surface charge was —43 to 59 mV. Mice were immunized with tetanus toxoid nanoparticles or free toxoid in solution at weeks 1, 2 and 3, either by oral, intranasal or intraperitoneal administration. Four weeks after the first intranasal immunization, IgG and IgA titers were significantly higher than baseline. Oral immunization with the nanoparticles produced a weak IgG antibody response. Only 10% of the oral dose was administered to the nose (2.89 vs. 28.9 /u,g), however, intranasal immunization appeared to be more effective in inducting an immune response. Particle size had an effect on the titer levels. Particles > 1 /i.m did not induce an immune response, but no difference was observed between the 500 nm and 100 nm nanoparticles which both induced significantly levels of IgG and IgA. These studies suggest that nasal delivery of vaccines using biodegradable nanoparticles are a promising method of inducing mucosal and systemic immunity. 6.3. Anti Tuberculosis therapy Intracellular bacterial infections caused by pathogens such as Mycobacterium tuberculosis are difficult to eradicate because they are generally inaccessible to free antibiotics. By loading antibiotics into nanoparticles, it is expected that delivery to the infected cells would improve since nanoparticles have been shown to localize preferentially in organs with high phagocytic activity and in circulating macrophages as well.97 The encapsulation of antibiotics has several advantages: (1) It modifies their pharmacokinetic characteristics by prolonging the antibiotics half-life and increasing the area under the concentration time curve (AUC), while decreasing its apparent volume of distribution. (2) It improves the targeting of the drug to the phagocytic cells. (3) It reduces toxicity of the antibiotics, such as the hepatotoxicity of anti tuberculosis drugs and the nephrotoxicity of aminoglycosides. Antibiotics encapsulated in nanoparticles have been shown to be superior at treating intracellular infections when administered intravenously. However, the pulmonary delivery of these nanoparticles have only been investigated recently. Although effective therapy for tuberculosis is available, treatment failure and drug resistance is typically the result of patient's noncompliance. To improve compliance, investigators have been studying ways to reduce the dosing frequency of the drugs. Poly (lactide-co-glycolide) (PLG) nanoparticles as an aerosolized sustained release formulation for anti tuberculosis drugs, isoniazid, rifampicin and pyrazinamide, has been investigated since pulmonary tuberculosis is the most 386 Labiris, Bosco & Dolovich common form of the infection.98 The majority of the nanoparticles were 186 to 290 nm in diameter. Drug encapsulation efficiency was 56.9% to 68%. Aerosolized nanoparticles had a MMAD of 1.88 //.m, with 96% of the particles in the respirable range (<6 /u,m). A single nebulization to guinea pigs resulted in sustained plasma drug concentrations for 6 to 8 days and in the lung for 11 days. The half-life and mean residence time of the drugs was significantly prolonged, compared with the oral free drugs. Nebulizing the nanoparticles every 10 days to guinea pigs infected with Mycobacterium tuberculosis resulted in no detectable bacilli in the lung after 5 doses of treatment, compared with 46 daily doses of orally administered drug to achieve the equivalent efficacy. The use of lectin-based PLG nanoparticles as an aerosolized sustained release formulation of isoniazid, rifampicin and pyrazinamide has also been studied in guinea pigs." Mucoadhesive drug delivery systems such as chitosan have been previously investigated as a method of prolonging residence at a site of absorption. The main drawback of mucoadhesive systems is that its residence time is limited by the turnover time of the mucous gel layer, which is only a few hrs. Attaching the polymeric nanoparticles to cytoadhesive ligands such as lectins could prolong the duration of adhesion, thereby prolonging residence time. Lectins bind to epithelial surfaces via specific receptors. Wheat germ agglutinin (WGA) is the least immunogenic lectin and has known receptors on the alveolar epithelium as well as the intestinal wall. WGA lectin-PLG nanoparticles were prepared by a two-step carbodiimide procedure. Their size ranged from 350 to 400 nm with drug encapsulation efficiency between 54% and 66%. The nanoparticles were delivered via nebulization to guinea pigs. 88% of the aerosol was in the respirable range (<6/U,m) with a MMAD of 2.8/zm (GSD of 2.1). Three doses of nanoparticles were administered every 15 days for 45 days. The WGA-PLG nanoparticles resulted in a prolonged Tmax/ increased AUC and mean residence time after inhaled delivery. All three drugs were present in the lungs, liver and spleen at concentrations above the minimum inhibitory concentration 15 days post dosing, compared with orally-administered free drug. Chemotherapeutic studies in guinea pigs infected with Mycobacterium tuberculosis showed that 3 doses administered every 15 days for 45 days yielded undetectable mycobacterial colony forming units, which was only achievable with 45 doses of the oral free drugs. The study results suggest that WGA-based PLG nanoparticles could be potential drug carriers for anti tuberculosis through aerosol delivery, reducing the drug dosing frequency. 6.4. Gene therapy Pulmonary gene delivery and DNA vaccinations are attractive therapies for a variety of lung diseases such as cystic fibrosis, asthma, chronic obstructive pulmonary Aerosols as Drug Carriers 387 disease, lung cancer and infections caused by Mycobacterium tuberculosis, influenza or SARS-associated coronavirus. Gene delivery requires carriers to transfer DNA into the nuclei of cells. There are two approaches for delivery: viral and non viral carriers. Viral delivery systems, although very efficient at transfection, are problematic due to their inherent immunogenicity. Non viral are safer but their transfection efficiency is low. Recently, biodegradable polymer-based nanoparticles have been investigated as a non viral pulmonary gene delivery system, taking advantage of their prolonged residence time in the lung and ability to be taken up by macrophages and dendritic cells, and to escape degradation by lysosomes. Asthma is characterized by elevated eosinophilic inflammation in the airway and increased airway hyperresponsiveness. Chronic inflammation can lead to structural damage and airway remodeling. IFN-y is a cytokine that promotes T-helper type 1 (Thl) responses which down regulates the Th2 immune responses present in asthma. Recombinant IFN-y has been shown to reverse inflammation in murine models of asthma. However, its short half-life and severe adverse effects at high doses have prevented its therapeutic use.100 An intra-nasal IFN-y gene therapy had been developed as an attempt to circumvent the drawbacks to its use. Kumar and colleagues studied the effects of a chitosan-IFN-y plasmid DNA nanoparticle in a BALB/c mouse model of allergic asthma (using ovalbumin-sensitization).101 Mice treated with the chitosan nanoparticles exhibited a significantly lower airway hyperresponsiveness (to methacholine challenge), reduced number of eosinophils and a significant decrease in epithelial denudation, mucus cell hyperplasia and cellular infiltration. Production of IFN-y was increased post-treatment while IL-5 and IL-4 and ovalbumin-specific IgE were reduced. Chitosan IFN-y nanoparticles induced IFN-y gene expression predominately in epithelial cells and worked within 3 to 6 hrs after intranasal administration. Poly (D,L-lactide-co-glycolide (PLGA)-polyethyleneimine (PEI) nanoparticles are also being investigated for pulmonary gene delivery. PLGA had been extensively evaluated for its sustained-release profile and ability to be taken up by macrophages. PEI is a cationic polymer. Its high positive charge density suggests that it would be a promising candidate as a non viral vector.102 PLGA nanoparticles with PEI on their surface had a mean particle diameter between 207 and 231 nm, surface charge > 30 mV and a DNA loading efficiency of >99%. Internalization of the nanoparticles in the human airway submucosal epithelial cell line, Calu-3, was observed and DNA detected 6 hrs after administration. However, in vivo efficiency of this system still needs to be studied. Respiratory syncytial virus (RSV) infection is a major cause of respiratory tract infections and is associated with approximately 17000 deaths annually on a worldwide basis, with no anti viral therapy or vaccine available.103 RSV NS1 protein appears to antagonize the host type 1 interferon-mediated response. Zhang and 388 Labiris, Bosco & Dolovich colleagues hypothesized that blocking the NS gene expression might inhibit RSV replication and thus provide effective antiviral therapy.104 Small interfering RNA (siRNA) targeting the NS1 gene (siNSl) were encapsulated in chitosan nanoparticles. BALB/c mice were intranasally treated with siNSl chitosan-nanoparticles before or after RSV infection. A significant decrease in virus titers in the lung was observed, in addition to a decrease in inflammation and airway hyperresponsiveness, compared with controls. The effect of siNSl lasted at least 4 days. The data show that siNSl nanoparticles may be a promising anti viral therapy against RSV infection. 7. Conclusion Innovations in the biotechnology and pharmaceutical industries have led to novel approaches for delivering drugs more efficiently and to specific targets in the lung and the body. One of the growth areas is the development of nanoparticles as carriers of active pharmaceutical agents for diagnosis and treatment. Aerosol delivery systems, discussed at the beginning of this chapter, are the current technologies for delivering therapies to treat respiratory diseases and some systemic diseases. The accepted philosophy, and one based on sound in vitro and in vivo clinical data, is that the optimal size of aerosol needed to target the distal lung is of the order of 3 /u,m. This size is 10-100 times greater than the nanoparticles being considered in the design of agents including antibiotics, vaccines and gene therapies for inhaled delivery. Novel techniques and formulations are being studied to produce successful vehicles for delivering these types of products in vivo. Positive outcomes using animal models to test these new aerosol formulations have been reported. However, clinical studies still need to be conducted to determine their efficacy in humans. As with any new technology, there will be benefits and risks associated with its use. The use of nanotechnology to provide improved targeting of drugs via the inhaled route is an exciting development that has the potential to yield novel treatments for many diseases in the near future. References 1. Horsfield K, Dart G, Olson DE, Filley GF and Cumming G (1971) Models of the human bronchial tree. / Appl Physiol 31:207-217. 2. Kreuter J (1994) Drug targeting with nanoparticles. Eur J Drug Met Pharmacol 19:253-256. 3. Ghirardelli R, Bonasoro F, Porta C and Cremaschi D (1999) Identification of particular epithelial areas and cells that transport polypeptide-coated nanoparticles in the nasal respiratory mucosa of the rabbit. Biochim Biophys Acta 12:1-2. Aerosols as Drug Carriers 389 4. Huang M, Ma Z, Khor E and Lim LY (2002) Uptake of FITC-chitosan nanoparticles in A549 cells. Pharm Res 19:1488-1494. 5. Russell-Jones GJ, Veitch H and Arthur L (1999) Lectin-mediated transport of nanoparticles across Caco-2 and OK cells. Int f Pharm 190:165-174. 6. Brigger I, Dubernet C and Couvreur P (2002) Nanoparticles in cancer therapy and diagnosis. Adv Drug Del Rev 54:631-651. 7. Martonen TB (1993) Mathematical model for the selective deposition of inhaled pharmaceuticals. / Pharm Sci 82:1191-1199. 8. O'Callaghan C and Barry PW (1997) The science of nebulized drug delivery. Thorax 52 (Suppl 2):s31-s44. 9. Denyer J, Dyche T, Nikander K, Newman SP, Richards J and Dean A (1997) Halolite a novel liquid drug aerosol delivery system. Thorax 52(suppl 6):208. 10. Dolovich M (1999) New propellant-free technologies under investigation. / Aerosol Med 12(Suppl I):s9-sl7. 11. Smaldone GC, Agosti J, Castillo R, Cipolla D and Blanchard JD (1999) Deposition of radiolabelled protein from AERx in patients with asthma. / Aerosol Med 12:98. 12. Newman SP and Clarke SW (1992) Inhalation devices and techniques, in Clark TJH, Godfrey S and Lee TH (eds.), Asthma. Chapman & Hall, London, pp. 469-^55. 13. Newman SP, Pavia D, Moren F, Sheahan NF and Clarke SW (1981) Deposition of pressurized aerosols in the human respiratory tract. Thorax 36:52-55. 14. Crompton GK (1982) Problems patients have using pressurized aerosol inhalers. European Journal of Respiratory Diseases — Supplement 119:101-104. 15. Bennett WD and Smaldone GC (1987) Human ventilation in the peripheral air-space deposition of inhaled particles. Am } Physiol 62:1603-1610. 16. Dolovich M, Ryan G and Newhouse MT (1981) Aerosol penetration into the lung. Influence on airway responses. Chest 80 (suppl)(6):834-836. 17. Newman SP, Pavia D, Garland N and Clarke SW (1982) Effects of various inhalation modes on the deposition of radioactive pressurized aerosols. Eur } Respir Dis 63(Suppl. 119):57-65. 18. Pavia D, Thomson ML, Clarke SW and Shannon HS (1977) Effects of lung function and mode of inhalation on penetration of aerosol into the human lung. Thorax 32:194-197. 19. Dolovich M, Ruffin RE, Roberts R and Newhouse MT (1981) Optimal delivery of aerosols from metered dose inhalers. Chest 80:911-915. 20. Dolovich MB (1995) Characterization of medical aerosols: Physical and clinical requirements for new inhalers. Aerosol Sci Technol 22:392-399. 21. Dolovich M, Chambers C, Mazza M and Newhouse MT (1992) Relative efficiency of four metered dose inhaler (MDI) holding chambers (HC) compared to albuterol MDI. J Aerosol Med 5:307. 22. Dolovich M, Ruffin R, Corr D and Newhouse MT (1983) Clinical evaluation of the Aerochamber: A simple demand inhalation MDI delivery device. Chest 84:36-41. 23. Newman SP, Moren F, Trofast E, Talaee N and Clarke SW (1989) Deposition and clinical efficacy of terbutaline sulphate from Turbohaler, a new multi-dose inhaler. Eur Respir } 2:247-252. 390 Labiris, Bosco & Dolovich 24. Pedersen S (1996) Inhalers and nebulizers: Which to choose and why. Respiratory Medicine 90:69-77. 25. Newhouse MT and Kennedy A (2000) Condensation due to rapid, large temperature (t) changes impairs aerosol dispersion from Turbuhaler(T). Am } Respir Cell Mol Biol 161(3):A35. 26. Newhouse MT and Kennedy A (2000) Inspiryl Turbuhaler (ITH) DPI vs. Ventolin MDI + Aerochamber (AC): Aerosol dispersion at high and low flow and relative humidity/ temperature (RH/T) in vitro. Am } Respir Crit Care Med 161(3):A35. 27. Borgstrom L, Bondesson E, Moren F, Trofast E and Newman SP (1994) Lung deposition of budesonide inhaled via Turbuhaler: A comparison with terbutaline sulphate in normal subjects. Euro Respir } 7:(1) 69-73. 28. Richards R and Saunders M (1993) Need for a comparative performance standard for dry powder inhalers. Thorax 48:1186-1187. 29. Everard ML, Devadason SG and Le Souef PN (1997) Flow early in the inspiratory manoeuvre affects the aerosol particle size distribution from a Turbuhaler. Respir Med 91:624-628. 30. Newman SP, Hollingworth AandClarkAR(1994)Effectof different modes of inhalation on drug delivery from a dry powder inhaler. Int} Pharma 102:127-132. 31. Newman SP, Moren F, Trofast E, Talaee N and Clarke SW (1991) Terbutaline sulphate Turbuhaler: Effect of inhaled flow rate on drug deposition and efficacy. Int J Pharma 74:209-213. 32. Clark AR and Hollingworth AM (1993) The relationship between powder inhaler resistance and peak inspiratory conditions in healthy volunteers — Implications for in vitro testing.} Aerosol Med 6(2):99-110. 33. Svartengren K, Lindestad PA, Svartengren M, Philipson K, Bilin G and Camner P (1995) Added external resistance reduces oropharyngeal deposition and increases lung deposition of aerosol particles in asthmatics. Am J Respir Crit Care Med 152:32-37. 34. Melchor R, Biddiscombe MF, Mak VH, Short MD and Spiro SG (1993) Lung deposition patterns of directly labelled salbutamol in normal subjects and in patients with reversible airflow obstruction. Thorax 48(5)506-511. 35. Vid gren M, Paronen P, Vidgren P, Vainio P and Nuutinen J (1990) In vivo evaluation of the new multiple dose