Anti-Human Leukocyte Antigen Antibodies After Islet Transplantation: What do They Really Mean?

Abstract

Human islet allotransplantation is one possible way to achieve insulin independence in type-1 diabetic patients. Current results, emerging from few selected centers worldwide are much more encouraging than in the past. Nevertheless, although the number of insulin-free recipients has significantly increased with modern isolation procedures (including multiple donors) and new immunosuppressive drug associations, the percentage of long-term success is far from that observed after solid pancreatic transplantation (1). As a result of such disappointing results, islet transplantation cannot yet be considered as an alternative therapy to whole pancreatic transplantation. In the absence of histology and reliable markers of rejection, islet transplant failure is commonly attributed to insufficient tissue mass or to acute allo- or autoimmunity aggression. Because definitive islet failure is mostly diagnosed after the first year of transplantation (1), at a time when de novo antibodies are detected in the sera of some patients, one could hypothesize that humoral immunity may participate in the process culminating in graft failure. The real impact of these newly diagnosed anti-human leukocyte antigen (HLA) antibodies on islet graft survival is unfortunately unknown because of the low number of patients evaluated to date. Human leukocyte antigen typing (A, B, DR, and DQ) and screening for anti-HLA antibodies is systematically performed in all candidates awaiting organ or cellular transplantation. T- and B-cell cross-matches are routinely carried out before transplantation and a positive T-cell crossmatch is a contraindication to transplantation. Although complement dependent cytotoxicity can still be used to detect any known anti-HLA antibody, these antibodies are now usually detected and identified by an ELISA or Luminex test for class I and class II. These two latter tests are sensitive and HLA-specific, allowing a comparison of results between different centers. Flow cytometry (Flow PRA) is used in few cases. In this issue of Transplantation, Ferrari-Lacraz et al. (2) described a rate of de novo anti-HLA immunization of 10.8% in type-1 diabetic recipients of combined kidney and islet grafts. This percentage is similar to that published by our group after kidney transplantation alone (3), but lower than the 31% reported by Campbell et al. (4) after islet transplantation alone. Many confounding factors may, however, explain this difference. All 98 individuals in the Edmonton group received isolated islet transplants (one or more transplants, i.e., free from immunosuppression before transplantation) whereas only 8 (24%) transplantations in the Geneva group were isolated islet transplants, the remaining patients received a combined kidney and islet transplant (n=16) and an islet transplant several months after a successful kidney transplant (n=13). More patients in Geneva were under immunosuppression at the time of islet transplantation, suggesting that “preemptive” immunosuppression may decrease allosensitization. Of the 34 patients reported in this issue by the Geneva group, 43% were women, compared with 59.2% in Campbell’s cohort. It is well known that women develop anti-HLA antibodies more frequently than men. None of the patients in Geneva were immunized before transplantation as compared with 28.6% in the Edmonton group. This point is not devoid of interest because pretransplant anti-HLA sensitization is a well-known high-risk factor for rejection and transplant failure, mainly in fully HLA-mismatched recipients. Taken together, these demographic patient characteristics and dissimilar indications for transplantation may partially explain the differences in anti-HLA sensitization observed between Geneva and Edmonton. The type of immunosuppression can have a major effect on anti-HLA antibody production. Different immunosuppressive regimens can affect the development, as well as the type and level of posttransplant antibodies. More particularly, are recipients included in steroid-free regimens more susceptible to develop posttransplant de novo anti-HLA antibodies? Neither of these two recent studies (2, 4) adequately address these issues and certainly prospective studies, both after cellular and organ transplantations, are required if definitive conclusions are to be made. The Geneva and Edmonton groups used current immunosuppressors including tacrolimus and sirolimus. A steroid-free regimen was systematically applied in Edmonton, whereas steroid use was much more frequent in Geneva (45% of patients). Mycophenolic mofetil (MMF) was, on the contrary, used more frequently in Geneva than in Edmonton, although the precise use of this drug was not reported in detail. Additionally, the majority of patients in Edmonton received an anti-CD25 monoclonal antibody as induction therapy whereas a minority of patients in Geneva received an induction prophylaxis. Regardless of all types and combinations of immunosuppressors used (as well as timing, doses, etc.), it seems obvious that human islet allotransplantation, which consists of a pool of tissues from multiple major histocompatibility complex incompatible donors, infused into the recipient’s portal vein, can induce specific anticlass I and class-II HLA alloantibodies, with a mixture ranging from 10% to 30% of panel reactive antibody (mostly donor-specific), according to identified and well-known pretransplant donor and recipient factors. This percentage can significantly increase in the case of graft failure and immunosuppression withdrawal (4). Similar evidence of rapid islet failure was described in the context of xenogenic islet transplantation in a nonhuman primate model (5). This finding, which mainly concerned patients receiving islet transplants alone, should be seriously analyzed before indicating this experimental procedure. In fact, in the case of islet failure and high-levels of anti-HLA sensitization, the ability to be retransplanted (with islets, a pancreas, or a kidney graft) is strongly compromised. A pancreatic graft, which includes exocrine and endocrine tissues, a large segment of duodenum, and multiple lymph nodes, is antigenic and immunogenic. In fact, the incidence of acute rejection is significantly higher and graft survival is significantly lower (because of immune-mediated graft failure) when a pancreatic graft is transplanted alone as compared with a simultaneous pancreas-kidney (SPK) transplantation. In our institution, among 106 non-anti-HLA- immunized (i.e., panel reactive antibody <5%) recipients of a primary SPK with more than 1 year posttransplant follow-up, 39 (36.8%) were tested positive for HLA antibodies with Luminex screening; 22 of the 39 patients (21%) were confirmed to have developed de novo anti-HLA class I or class II antibodies with Luminex (64% antidonor specific). All 106 patients received antithymocyte globulin induction and MMF, whereas 40 (38%) were included in a calcineurin inhibitor (CNI)-based/steroid-free regimen, 56 (53%) received the same regimen plus steroids for 3 months, and 10 (9.4%) received sirolimus, MMF and steroids without CNI. The incidence of de novo anti-HLA antibodies was 17.5%, 15.4%, and 60%, respectively (CNI vs. sirolimus P=0.001). The acute rejection rate (pancreas and kidney combined) among patients with anti-HLA antibodies versus patients without anti-HLA antibodies was statistically significant: 45% vs. 8% (P <0.0001). This de novo anti-HLA sensitization negatively and significantly impact pancreas graft survival, censored for death and technical failure, after the first year posttransplantation (96% vs. 68% at 5 years; P=0.0003; unpublished data). Type-1 diabetic patients with severe medical and surgical diabetic complications can be candidates for pancreatic or islet allotransplantation. These procedures can be performed in nonuraemic patients or in patients with chronic renal failure requiring concomitant kidney transplantation. Sustained insulin independence is achieved with a pancreatic graft in 60% to 80% of patients at 5 years in the case of pancreas transplant alone. This percentage is greater in the case of SPK. Islet transplantation can achieve insulin independence in an almost similar percentage of patients. However, most of them restart on exogenous insulin after the first posttransplant year. The finding that de novo anti-HLA antibodies (against donor incompatible antigens or not) could be found in some patients suggests that humoral immunity may be partly responsible for graft failure. As compared with pancreatic grafts, islet grafts may be more sensitive and “fragile” and probably much more rapidly destroyed by both nonimmune and immune aggressions (5). Anti-HLA antibodies could accelerate dysfunction and failure of the islet transplant as soon as a few months after the serological diagnosis, differing from a solid organ that may require much longer. The main objective today among the transplant community is to avoid late graft failure of any cellular or solid organ transplant. This chronic process (that may start immediately after transplantation) represents the leading cause of failure among all type of transplants, principally islet, kidney, and whole pancreas. The presence of anti-HLA antibodies (mostly de novo and donor specific) significantly correlates with lower graft survival (3). Until today and to our knowledge, no randomized study has focused on the removal of de novo anti-HLA antibodies diagnosed after systematic screening in recipients with normal transplant function. Guidelines for regular screening, prevention, and treatment of this crucial endpoint urgently need to be established by international workshops. Until then, medical doctors in charge of diabetic patients requiring islet transplantation will continue to deal with a “dark field” that could seriously compromise the future of the patient.