The Bionic Pancreas

Whole organ pancreas transplantation was established by Richard Lillehei and William D. Kelly at the University of Minnesota in 1966.1 However, graft survival was limited, and only one tenth of pancreas grafts survived for more than a year at that time. Although early research on islet cell transplantation yielded equally poor long-term results, pancreas transplantation gained momentum with the introduction of cyclosporine. A uniform surgical technique was not established because of the problem of exocrine pancreatic secretions and various modifications were undertaken to reduce complications from this source. In parallel, progress had also been observed in islet cell transplantation with promising results in experimental studies, although early clinical applications did not yield euglycemia. The Edmonton protocol became, at least temporarily, a game changer, although follow-up reports were less optimistic with only 25% of patients remaining insulin-free by 5 years.2,3 Recent developments including refined isolation techniques and improved immunosuppression continue to provide promise.4 Research on continuous insulin delivery by extracorporeal devices is not entirely new and has been developed in parallel to the advancements in pancreas transplantation. The first closed-loop system that measured real-time glucose levels to control insulin infusion was introduced in 1964. By the 1970s, machines of considerable size with monitors and pumps that calculated insulin infusions based on intravenous glucose monitoring were developed.5 The goal was to create an artificial pancreas to monitor and infuse insulin to control glucose levels in the most physiological like manner to meet both long-term blood glucose control and prevent hypoglycemic and hyperglycemic episodes. In their recent publication in the New England Journal of Medicine (NEJM), Russel and co-workers6 have demonstrated glycemic control with an artificial, bihormonal pancreas demonstrating significant clinical benefits when compared to a conventional insulin pump. Both adults and adolescents with type 1 diabetes had near physiological glucose levels with less hypoglycemic readings during an observation period of 1 week. Of note, the frequency of hypoglycemic interventions was lower in adolescents with an artificial pancreas. The system proved to be user friendly because it was controlled wirelessly via an iPhone and patients rapidly became accustomed to the device. Moreover, patients’ meals and physical activities were not limited. Safety of the trial was achieved through a continuous supervision by medical staff. Closed-loop systems have been shown to effectively reduce nocturnal hypoglycemia. However, previous studies did not include fully automated pumps and medical personnel were in charge of transmitting data from the monitor to the pump.7 Another study has previously demonstrated the feasibility to treat outpatients with a closed-loop system, yet with some technical issues and hurdles.8 In a similar study, recently published in NEJM, an artificial pancreas had shown superior results in adolescents with type 1 diabetes in diminishing hypoglycemic episodes, compared to a sensor-augmented pump.9 These findings are supported by a meta-analysis showing that sensor-augmented pumps and real-time glucose monitoring are superior in treating type 1 diabetes without increasing the risk of hypoglycemia.10 “…those advances raise the question of whether an artificial and safe bi-hormonal pancreas may supplement or even replace pancreas transplantation…” Clearly, those advances raise the question of whether an artificial and safe bihormonal pancreas may supplement or even replace pancreas transplantation, particularly given the continued shortcomings of pancreas transplantation: restricted long-term survival, organ scarcity, and complications linked to surgery and immunosuppression. Indeed, in the absence of surgical risks and lifelong immunosuppression, an artificial and user-friendly pancreas seems to offer, at least in the short term, satisfactory glycemic control. Limitations of the closed-loop device are mentioned by the authors and include a potential bias by supervising staff, by allowing an immediate response to hypoglycemic episodes. Moreover, although the device depends on wireless connectivity, subcutaneous insulin delivery is linked to a slower rate of absorption while glucagon needs to be replaced daily due to the instability of the hormone. Constant glucose monitoring is performed in the interstitial fluid, and therefore associated with a time delay implicit from plasma concentrations. In addition, patients need to be somewhat technologically astute because the device control depends on a patient’s ability to handle a smartphone and constant wireless connectivity. The current study did not explore long-term reliability of glucose control nor safety of the device. The authors do need to be congratulated for the clinical introduction of this device in an experimental approach until technical concerns are resolved. Once the technology of the artificial pancreas has advanced to the stage where patients are completely independent of constant medical supervision and technical issues have been optimized, the artificial pancreas may become a relevant competitor for pancreas transplantation alone. For type 1 diabetes patients progressing to renal failure, simultaneous pancreas-kidney transplantation or pancreas after kidney transplantation remains, for the time being, the treatment of choice. At the same time, as these advances are being reported, alternative insulin replacement strategies are appearing on the horizon: in adults, β cells have been shown to self-replicate, representing a major mechanism of β-cell expansion.11 In addition, pancreatic multipotent precursor cells expressing insulin have been identified that generated mature β-cells in vivo.12,13 Intensive research into these mechanisms is currently underway and identification of factors controlling β-cell proliferation may provide novel therapeutic options. Insulin replacement will, in the near future, probably include artificial devices, β-replacement strategies, and pancreatic transplantation. These modalities will, together, allow us to design optimal individualized treatment for patients with insulin dependence.