COMMENTARY In cell encapsulation, transplanted cells are protected from immune rejection by an artificial, semipermeable membrane, potentially allowing transplantation (allo- or xenotransplantation) without the need for immunosuppression. Yet, despite some promising results in animal studies, the field has not lived up to expectations, and clinical products based on encapsulated cell technology continue to elude the scientific community. This commentary discusses the reasons for this, summarizes recent progress in the field and outlines what is needed to bring this technology closer to clinical application. Cell encapsulation: Promise and progress In 1964, T.M.S. Chang1 proposed the idea Some of the important considerations of using ultrathin polymer membrane GORKA ORIVE1, for consistent clinical success of cell en- microcapsules for the immunoprotection ROSA MARÍA HERNÁNDEZ1, capsulation include a source of func- of transplanted cells and introduced the ALICIA R. GASCÓN1, tional cells; a biocompatible, as well as term ‘artificial cells’ to define the concept RICCARDO CALAFIORE2, mechanically and chemically stable, of bioencapsulation, which was success- THOMAS M.S. CHANG3, membrane of a suitable permeability fully implemented 20 years later to im- PAUL DE VOS4, cut-off value that provides immune pro- mobilize xenograft islet cells. When GONZALO HORTELANO5, tection to the implant; functional per- implanted into rats, the microencapsu- DAVID HUNKELER6, IGOR LACÍK7, formance; biosafety; and long-term lated islets corrected the diabetic state for A.M. JAMES SHAPIRO8 & survival of the graft. several weeks2. Since then, there has been JOSÉ LUIS PEDRAZ1 considerable progress toward understand- Lack of clinical-grade polymers. In http://www.nature.com/naturemedicine ing the biological and technological requirements for successful the quest for a better microencapsulation design, many transplantation of encapsulated cells in experimental animal types of natural and synthetic polymers are being explored. models, including rodents and non-human primates. A substantial challenge related to the biomaterials used in Bioencapsulation has provided a range of promising therapeutic cell encapsulation has been the lack of clinical-grade poly- treatments for diabetes3, hemophilia4, cancer5 and renal failure6. mers. Although its intrinsic properties make alginate the Additionally, the functional applicability of cell encapsulation current encapsulation material of choice, batches of algi- in humans has also been reported in several clinical trials7,8. nate need to be standardized to minimize endotoxin and Despite considerable interest, however, the field has not lived protein content, both of which can affect biocompatibility. up to expectations9. A lack of reproducibility, or uncertainty sur- This requires standardized protocols to eliminate such im- rounding the reproducibility of prior studies has been a major purities11. In this regard, an international task force should problem. For example, in the treatment of diabetes, researchers be instituted to set up a ‘central alginate factory’ that would have been plagued by the inability to reproduce one excellent prepare standardized prototypes for use by participating lab- trial in dogs undertaken by Metabolex of Hayward, California oratories in their transplant studies. The European © Group 2003 Nature Publishing (http://www.metabolex.com), one important primate study on Community and Norway now support the formation of such rhesus monkeys3 and a single patient study7. To put it simply, de- dedicated centers. spite the Edmonton protocol10, in which islet transplantation in patients with type I diabetes resulted in insulin independence Production of uniform capsules. Another challenge in- for a year, very few groups have isolated sufficiently good islets volves the production of uniform capsules with excellent re- and developed a suitable biocompatible encapsulation material. peatability and reproducibility both within and between Indeed, the consensus is that microencapsulation still repre- batches. The adoption of automated machines for microen- sents a sort of ‘in-house’ procedure, administered by a small capsulation could result in improved reproducibility in number of laboratory groups who are reluctant, or who are un- terms of shape, size and morphology. As such, the recent de- able because the technology is proprietary, to share complete velopment of an automated chemical reactor12 may help to information and protocols. Also retarding progress of this field address this problem. is a lack of standardized technology, including optimized tissue and clinically proven materials for membrane manufacturing Use of polycations. The discovery of suitable immune-com- produced in reproducible batches. These limitations have be- patible polycations represents another principal area of deviled universities and startup companies alike. University- study. After 20 years of research, the clinical application of based laboratories have had the added difficulty of scaling up polycations for microcapsule formulation remains contro- technologies to produce materials in sufficient quantities to versial. Although some believe that poly-L-lysine (PLL) poly- permit duplicate studies. As a consequence, many essential re- cation has a low probability of success as a result of its poor search questions, such as the exact selection of membrane ma- biocompatibility13, others have obtained promising in vivo terials, their final properties, site of transplantation, cell source results replacing PLL with poly-L-ornithine14 or poly(meth- and choice of purification methods, remain unanswered. ylene-co-guanidine) hydrochloride15. It may be that ulti- mately the simplest system of them all, the alginate bead Current challenges with its non-uniform density, will be sufficient to provide Technological and biological limitations, as well as ethical, po- immune protection in the case of allogeneic models (in- litical and regulatory obstacles, must be overcome if the traspecific human), whereas the development of microcap- promise of cell encapsulation technology is to be realized. sule materials for xenogeneic models (interspecific 104 NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003 COMMENTARY ized solid supports are being studied to a b improve the nutrition of the encapsu- lated cells25 (Fig. 1). However, the question of vascularization remains open, because other groups have ob- tained promising results by transplant- ing capsules into the peritoneal cavity of large animals without direct contact with blood vessels14. Regulatory and ethical issues With advances in the science of encap- sulated cell therapies, regulatory au- thorities have been gradually adjusting Fig. 1 Cell microencapsulation. a, Nutrients, oxygen and stimuli diffuse across the membrane, their policies to accommodate these whereas antibodies and immune cells are excluded. b, Pre-vascularized solid support system to facil- new therapeutic approaches. In the itate optimal nutrition of the enclosed cells. United States, for example, all islet transplant studies (and presumably all future encapsulation-type clinical nonhuman) will remain a challenge. This is because algi- studies) will be regulated by the US Food and Drug nates are too porous to offer suitable immune protection to Administration (FDA) under an investigational new-device the implant. submission. For now, Europe will likely rely on FDA guide- lines, because specific regulations in this field are presently http://www.nature.com/naturemedicine Considerations before transplantation. Still other chal- lacking. Recently, the US Pharmacopeia and National lenges involve the assessment of the exact dosage10 and mol- Formulary included a new therapeutic category for cell- ecular-weight cutoff value, as well as the overall based products26, which constitutes a significant step toward biocompatibility of the system. X-ray photoelectron spec- accepting this technology and encouraging clinical trials. troscopy and Fourier-transform infrared spectroscopy tech- A major ethical concern surrounding the use of microen- niques could shed light on the latter, helping to identify the capsulated cells is to ensure that patients are treated with a chemical groups causing bioincompatibility on the surface technology that demonstrates a clearly proven biosafety of a capsule and predicting the biosafety of the devices be- based on standardized protocols and procedures. In this re- fore implantation16. gard, it is important to avoid poorly conducted studies that put individuals at unnecessary risk and unfairly raise hopes Selecting suitable cell types for immobilization. A num- and expectations. For example, the recent trial of pig islet ber of issues should be carefully evaluated when selecting xenotransplantation in children by Valdes’ group at the suitable cell types for immobilization. Indeed, encapsula- National University of Mexico has sparked a fresh round of 27 © Group 2003 Nature Publishing tion requires an appropriate source of functional cells. In debate , because it is in direct contravention of the Helsinki this regard, the potential use of allogeneic versus xeno- Agreement on the ethical performance of clinical trials. geneic cells has provoked important social and ethical de- Moreover, reports of this kind run the risk of hampering fu- bates17. The principal controversy surrounds the potential ture progress of the entire field. risk of inadvertent transfer to humans of animal viruses pre- sent in