We thank Drs. Charles and Selam for their thoughtful comments (1) regarding our study describing the National Institutes of Health (NIH) islet transplantation experience (2). We agree that the question remains open whether islet transplantation and subsequent antirejection therapy decrease the morbidity and mortality associated with diabetes. In fact, one of our article’s main messages is the need to develop criteria for identifying patients with “end-stage” diabetes for whom a therapy’s known and unknown risks are most likely to favor an appropriate risk/benefit ratio. We have recently published similar concerns regarding the more mature pancreas transplantation therapy (3). Our islet transplant study, formally approved by our institutional review board, was designed to test only whether the NIH team could reproduce the exciting successes then recently reported from Edmonton, using a steroid-free immunosuppressive strategy with defined islet isolation, graft characterization, transplantation, and posttransplant patient monitoring (4), not to formally address the risk/benefit ratio of this still experimental technique or to compare the therapy with other treatment approaches. As an inclusion criterion (borrowed largely from criteria to determine eligibility for a pancreas transplant), all of our patients had severe hypoglycemia (defined as requiring the assistance of others more than once in the preceding 20 months and not explained by a clear precipitating event) or hypoglycemia unawareness (defined as an inability to sense hypoglycemia with blood glucose levels <54 mg/dl). During the fairly detailed evaluation to determine protocol eligibility, we advised all potential enrollees on the use of modern insulin regimens designed to more closely mimic the normal physiologic insulin pattern, i.e., basal insulin levels with meal-associated boluses. In fact, several candidates achieved improved glycemia control sufficient to withdraw from further protocol participation. Three of our transplanted patients were treated (before transplant) with continuous insulin infusion systems, and the other three were treated with multiple daily insulin injections.
Our patients were well informed that islet transplantation remained an experimental procedure and were made aware of the multiple known and hypothetical risks of the procedure as well as the risks associated with long-term immunosuppression. We did not offer our patients the option of intraperitoneal insulin infusion through an implantable pump (IPII). Although promising, this procedure also remains experimental. Of note, one study (5) recently reported that in a head-to-head comparison of IPII and islet transplantation, the transplant recipients had less hypoglycemia. That being said, which treatment approach is the superior one with regard to more global end points (e.g., quality of life, prognosis for survival, etc.) has yet to be critically addressed.
Finally, we agree that part of the selection process for all future islet transplant trials should include attention to all the latest in accepted diagnostic and treatment approaches for diabetes. We hope to soon initiate a clinical research protocol that will incorporate these principles.