Pancreas After Islet Transplantation: A First Report of the International Pancreas Transplant Registry

Transcription

1 American Journal of Transplantation 2016; 16: Wiley Periodicals Inc. Brief Communication Copyright 2015 The American Society of Transplantation and the American Society of Transplant Surgeons doi: /ajt Pancreas After Islet Transplantation: A First Report of the International Pancreas Transplant Registry R. W. G. Gruessner 1,2, * and A. C. Gruessner 2,3 1 University of Arizona, Tucson, AZ 2 SUNY Upstate Medical University, Syracuse, NY 3 College of Public Health, University of Arizona, Tucson, AZ Corresponding author: Rainer W. G. Gruessner, rgruessner@surgery.arizona.edu Pancreas after islet (PAI) transplantation is a treatment option for patients seeking insulin independence through a whole-organ transplant after a failed cellular transplant. This report from the International Pancreas Transplant Registry (IPTR) and the United Network for Organ Sharing (UNOS) studied PAI transplant outcomes over a 10-year time period. Forty recipients of a failed alloislet transplant subsequently underwent pancreas transplant alone (50%), pancreas after previous kidney transplant (22.5%), or simultaneous pancreas and kidney (SPK) transplant (27.5%). Graft and patient survival rates were not statistically significantly different compared with matched primary pancreas transplants. Regardless of the recipient category, overall 1- and 5-year PAI patient survival rates for all 40 cases were 97% and 83%, respectively; graft survival rates were 84% and 65%, respectively. A failed previous islet transplant had no negative impact on kidney graft survival in the SPK category: It was the same as for primary SPK transplants. According to this IPTR/UNOS analysis, a PAI transplant is a safe procedure with low recipient mortality, high graft-function rates in both the short and long term and excellent kidney graft outcomes. Patients with a failed islet transplant should know about this alternative in their quest for insulin independence through transplantation. Abbreviations: CITR, Collaborative Islet Transplant Registry; DM, diabetes mellitus; IDDM, insulin-dependent diabetes mellitus; IPTR, International Pancreas Transplant Registry; PAI, pancreas after alloislet transplant; PAK, pancreas after kidney transplant; PRA, panel reactive antibody; PTA, pancreas transplant alone; SPK, simultaneous pancreas kidney transplant; Tx, transplantation; UNOS, United Network for Organ Sharing Received 13 May 2015, revised 19 July 2015 and accepted for publication 21 July 2015 Introduction Beta cell replacement therapy through pancreas or islet transplantation is a viable treatment option for many uremic patients with insulin-dependent diabetes mellitus (IDDM) and end-stage renal disease. It is also considered an alternative for selected nonuremic or posturemic (ie, after a previous kidney transplant) patients with labile IDDM. Since the first pancreas transplant by Kelly and Lillehei in 1966 (1), > pancreas transplants in the United States have been reported to the United Network for Organ Sharing (UNOS) and the International Pancreas Transplant Registry (IPTR) (2). Over time, outcomes after pancreas transplantation have gradually improved, and 5-year patient survival rates regardless of recipient category have reached 90%; graft survival rates by recipient category range between 57% and 75% at 5 years (2). Since the first clinical islet transplant by Najarian and Sutherland in 1974 (3), the number of islet transplants increased significantly only after the introduction of the socalled Edmonton protocol in 2000 by Shapiro et al (4). In 2001, the Collaborative Islet Transplant Registry (CITR) was founded to collect and analyze information on islet transplants in North America (5); however, reporting of islet transplants to either UNOS or CITR is not mandatory (6). A comparison of graft outcomes between pancreas and islet transplantation is hampered by the fact that the definition of pancreas graft function is different than that of islet graft function. Pancreas and islet transplantation should not be considered competing but rather complementary transplant options: Pancreas transplantation requires major surgery with better graft outcome, whereas islet transplantation is minimally invasive (6). Undergoing one or the other is not mutually exclusive in case the graft fails and retransplant is contemplated. Due to the lack of mandatory reporting, the number of islet transplants after a previously failed pancreas transplant is unknown. In contrast, the number of pancreas transplants after a previously failed islet transplant is known and allows for analysis of patient and graft outcomes for pancreas after alloislet (PAI) transplants. Using the IPTR/UNOS database, we studied outcomes for PAI transplants that were performed in the United States 688

2 Pancreas After Islet Transplantation over the past 10 years and analyzed the impact on patient and graft survival. Materials and Methods We analyzed all PAI transplants that were reported to the IPTR and UNOS over a 10-year time period, from January 1, 2004, to December 31, Over this time period, 40 patients received a primary whole-pancreas transplant after a failed alloislet transplant (ie, PAI). Data regarding the islet transplants (eg, number of islet transplants, number of islets transplanted) were largely incomplete because islet transplant reporting to UNOS is not required. All PAI transplant patients were matched by transplant category, recipient and donor age, donor cause of death, diabetes type, race, transplant year and class 1 panel reactive antibody (PRA) levels to create a comparable control group for comparisons. We matched the PAI transplant patients with only primary pancreas transplants in a first analysis and with pancreas retransplants in a second analysis. The greedy matching algorithm was used. Overall, 36 PAI recipients could be matched. For three recipients with type 2 diabetes mellitus, no appropriate match could be found because of extremely high PRA levels. For a fourth patient, no information about sensitization was available. Pancreas graft function was used as reported to UNOS by the individual transplant centers. Patients with reported partial function or who died with a functioning graft were counted as graft failures. Patient characteristics are given as mean and standard deviation or median and range, according to the underlying distribution. For the comparison of the PAI transplants with the primary control group, a Wilcoxon test was used. Qualitative variables were compared using a chi-square test. Graft and patient survival was calculated using the Kaplan-Meier method. Outcomes were compared using the Wilcoxon and log-rank tests. The p values for subsequent pairwise comparisons were adjusted according to Sidak. Results In all, 40 PAI transplants were performed at 21 institutions. All pancreas transplants were performed at institutions known for both their pancreas and islet transplant programs. Of the 40 recipients, 27 (68%) were female and 13 (32%) were male. Based on ethnicity, 31 (78%) recipients were white; five were black; and four were Asian, Hispanic, or other. The median age at transplant was 45 years (range: years). Type 1 diabetes mellitus was diagnosed in 36 patients (90%), with type 2 diagnosed in the remaining four patients. Half of the recipients (n ¼ 20) underwent a previous or simultaneous kidney transplant: In the SPK category, one patient had a second and one patient had a third kidney transplant simultaneously with their pancreas transplant. The time intervals between (last) kidney and pancreas transplant are shown in Table 1. According to pancreas transplant recipient category, 20 (50%) patients underwent a pancreas transplant alone (PTA), 9 (22.5%) underwent a pancreas after kidney (PAK) transplant, and 11 (27.5%) underwent a simultaneous pancreas and kidney (SPK) transplant. In all, 36 (90%) of all pancreas transplants were enteric, and 4 (10%) were bladder drained. Systemic vein drainage was used in 34 (85%) transplants, and portal vein drainage was used in 6 (15%). Because all patients had undergone previous islet transplants and 27.5% of the patients underwent previous kidney transplant, only 59% of patients had 0% Class 1 and 75% patients 0% Class 2 antibodies at the time of the pancreas transplant. The median PRA class 1 level was 0% (range: 0 98%), and median PRA class 2 level was 0% (range: 0 100%). Table 1 shows the comparison between PAI recipients and the matched primary control group. Recipient and donor age as well as pancreas preservation time are almost identical due to the matching process. No differences were Table 1: Patient characteristics for pancreas transplant patients with a previous alloislet transplant and matched primary pancreas transplants as control PAI Primary control p Transplant type PTA PAK 9 9 SPK Age at transplant (y) 42.0 (8.6) 42.1 (8.5) 0.95 Donor age (y) 22.4 (6.9) 22.3 (7.1) 0.96 Pancreas preservation time (h) 12.1 (5.5) 11.9 (4.6) 0.80 HLA mismatch 4.0 (1.4) 4.1 (1.5) 0.55 PRA class 1 Percentage >20% 20% 20% 1.00 PRA class 2 Percentage >20% 12% 8% 0.57 Age at DM onset (y) 15 (1 35) 16.0 (1 37) 0.66 Duration of DM (y) 26.8 (8.4) 27.4 (8.2) 0.74 Time between kidney and PAK 18.8 (4 116) 23.2 (10 223) Pre-Tx BMI (kg/m 2 ) 23.2 (3.3) 25.0 (4.0) 0.10 Pre-Tx insulin requirement 21 (10 35) 34 (12 90) 0.01 DM, diabetes mellitus; PAI, pancreas after alloislet transplant; PAK, pancreas after kidney transplant; PRA, panel reactive antibody; PTA, pancreas transplant alone; SPK, simultaneous pancreas kidney transplant; Tx, transplantation. American Journal of Transplantation 2016; 16:

3 Gruessner and Gruessner found for HLA matching, age at diabetes onset, duration of diabetes, and the time interval between the previous kidney and the current pancreas transplant. The class 1 and 2 PRA levels were comparable due to the matching process. The only significant difference between the two groups was a lower average amount of insulin pretransplant in PAI recipients. The initial BMI also tended to be slightly lower in the PAI recipients. Patient survival is shown in Table 2 for all 40 PAI transplant patients and for the matched PAI recipients. One- and 5-year PAI patient survival rates were 94% and 85%, respectively, in the PTA category; 100% and 83%, respectively, in the PAK category; and 100% and 80%, respectively, in the SPK category. The difference among the three PAI recipient categories was not statistically significant (p ¼ 0.88). Of the four deaths (one PAK, two PTA, one SPK) during this time period, one was due to a fungal infection; in the remaining three cases, the cause of death was unknown. When compared with patient survival rates for the primary matched controls (Figure 1), no significant difference in patient survival was noted (p ¼ 0.31). In a separate analysis, the PAI recipients were matched with pancreas retransplants. No differences in patient survival between the two groups could be detected (p ¼ 0.95). Graft survival is shown in Table 2 for all 40 PAI transplant patients and matched recipients. One- and 5-year PAI pancreas graft survival rates were 78% and 53%, respectively, in the PTA category; 100% and 83%, respectively, in the PAK category; and 81% and 65%, respectively, in the SPK category. The difference among the three PAI recipient categories was statistically not significant (p ¼ 0.39). During this time period, a total of 10 pancreases failed. One patient with a PAK transplant and two patients with PTA died with a functioning pancreas graft, one patient Table 2: Patient and graft survival of all 40 pancreas after alloislet transplant recipients Overall PTA SPK PAK p n 40 (36) 20 (17) 11 (10) 9 (9) Patient survival 1 year 97 (97) 94 (93) 100 (100) 100 (100) years 83 (88) 85 (83) 80 (100) 83 (83) Pancreas survival 1 year 84 (88) 78 (80) 81 (89) 100 (100) years 65 (67) 53 (47) 65 (71) 83 (83) Kidney survival 1 year 100 (100) 5 years 60 (75) In parenthesis are the results of the matched PAI recipients. Data are shown as percentages except as specified. PAI, pancreas after alloislet transplant; PAK, pancreas after kidney transplant; PTA, pancreas transplant alone; SPK, simultaneous pancreas kidney transplant. Figure 1: Patient survival for pancreas transplant patients with a previous alloislet transplant (PAI yes) and matched primary pancreas transplants as control (PAI no). PAI, pancreas after alloislet transplant. 690 American Journal of Transplantation 2016; 16:

4 Pancreas After Islet Transplantation with PTA and one with SPK transplant were lost due to graft thrombosis, one patient with PTA was lost due to pancreatitis, and one patient with PTA was lost due to acute rejection. Pancreases for one patient with PTA and one patient with SPK transplant were only partially functioning, and one SPK pancreas failed for unknown causes. Compared with graft survival rates of the matched primary control group (Figure 2), no significant differences were noted (p ¼ 0.86). Compared with outcomes of the matched pancreas retransplants, PAI pancreas graft survival was significantly better (p ¼ 0.04). At 5 years, 71% of the PAI pancreases were functioning compared with only 44% of the matched retransplants (Figure 3). One- and 5-year PAI kidney graft survival rates in the SPK category were 100% and 60%, respectively. One patient died with a functioning kidney graft, and one patient lost the kidney graft at 53 months after transplant due to chronic rejection. Compared with kidney graft survival rates of the matched control group of primary (p ¼ 0.92) or retransplanted SPK patients (p ¼ 0.21), no differences in outcome of kidney function were noted. For the nine PAI recipients in the PAK category who had undergone a previous kidney transplant, 1- and 5-year kidney graft survival rates from the time of pancreas transplant were 100% and 51%, respectively. Of the nine PAI patients with a previous kidney transplant, three lost their kidney graft. One patient died with a functioning kidney graft 27 months after the PAI transplant, the kidney graft of the second patient failed 2 years after the PAI transplant for unknown reason, and the kidney graft of the third patient failed 37 months later due to chronic rejection. It is unknown whether the patients received the kidney transplant before, simultaneously with, or after the islet transplant. Discussion Given the high patient survival rates after pancreas transplantation in all three recipient categories and the significant decrease in major surgical complications over time, many patients with a failed whole-organ graft opt for retransplant. Consequently, pancreas retransplantation has become a reasonable alternative for many recipients. Acceptance for pancreas retransplantation is also supported by previous IPTR analyses demonstrating that outcomes in the different recipient categories (PAK retransplants being most common) are similar for primary and retransplants (7,8). For these reasons, funding for pancreas retransplantation is provided by most insurance companies in the United States. Islet transplant recipients undergo a varied number of islet infusions, and it is not uncommon for them to receive two or more islet infusions from different donors (9). Because funding for islet transplantation remains an unresolved issue in the United States, patients with a previous but failed islet transplant cannot easily undergo an islet retransplant. Consequently, some patients consider a pancreas transplant to be a reasonable alternative. Figure 2: Pancreas graft survival for pancreas transplant patients with a previous alloislet transplant (PAI yes) and matched primary pancreas transplants as control (PAI no). PAI, pancreas after alloislet transplant. American Journal of Transplantation 2016; 16:

5 Gruessner and Gruessner Figure 3: Pancreas graft survival for pancreas transplant patients with a previous alloislet transplant (PAI yes) and the matched pancreas retransplant as control (PAI no). PAI, pancreas after alloislet transplant. PAI transplants are rare procedures. They accounted for only 0.4% of all pancreas transplants over the past decade. The purpose of this IPTR analysis was twofold: (1) to determine whether pancreas transplant after a previously failed islet transplant is safe and (2) to study graft survival in comparison with a matched cohort of primary deceaseddonor pancreas transplants and a matched cohort of pancreas retransplants. We noted two interesting findings regarding the patient demographics between the two groups: Both patient BMI and overall insulin requirements before the pancreas transplant were significantly lower in PAI recipients compared with the matched cohort of primary deceased-donor pancreas transplant recipients. These two findings are in line with reports showing that islet transplant candidates are selected to increase graft survival odds and that restrictions for pancreas transplant candidates have been increasingly lifted over time due to very favorable long-term outcomes for both patient and graft survival (9,10). Of note, these differences are seen not only in IPTR and CITR registry reports but also in the few single-center comparisons between pancreas and islet transplantation (11,12). The two most recently published single-center studies comparing pancreas with islet transplantation also revealed higher rates of insulin independence with solid organ versus cellular transplantation (11,12). Our analysis showed that a PAI transplant is a safe procedure: Overall patient survival at 1 year after transplant was >94% (100% in the PAK and SPK categories). At 5 years, 83% of the recipients were still alive. There was no difference in patient survival when PAI recipients were compared with the matched cohort of primary deceaseddonor pancreas transplants or the matched cohort of pancreas retransplants. Consequently, the previous islet transplant had no negative impact on patient mortality. Overall graft function at 1 year was 84% (highest in the PAK category at 100%). At 5 years, overall graft survival was 65% (also highest in the PAK category at 83%). The difference in graft survival rates among the three recipient categories was not statistically significant. There was also no significant difference compared with the matched cohort of primary deceased-donor pancreas transplants, both overall and category specific. Consequently, previous islet transplant had no detrimental effect on graft function of the subsequent pancreas transplant. This result is also in line with the observation that graft survival rates for primary versus pancreas retransplants are similar (2,8). When graft survival rates of PAI recipients were compared with the matched cohort of pancreas retransplants, the result was somewhat surprising: PAI pancreas graft survival was significantly better, which may indicate that a primary pancreas transplant may be more of a surgical and immunological challenge to the diabetic transplant recipient than a primary islet transplant. In this paper, we presented quite favorable findings for PAI recipients, given the fact that sensitization after islet 692 American Journal of Transplantation 2016; 16:

6 Pancreas After Islet Transplantation transplantation was not uncommon. In fact, a recent CITR report demonstrated that the development of HLA class I antibodies while on immunosuppression was associated with subsequent islet graft failure (13). The authors speculated that the risk of sensitization may be reduced in the future by minimizing the number of islet donors used per recipient. With the exception of one recipient in our patient series (who had an 80% class 1 PRA level), it appeared that prior sensitization of the islet recipients had no impact on pancreas graft survival. Finally, kidney graft function in the SPK category after a previously failed islet transplant was not different from primary or retransplanted deceased-donor kidney transplants in the matched SPK category. Consequently, the previous islet transplant did not negatively influence kidney or pancreas graft outcome. Registry reports are not without limitations. First, this analysis covers a time period with follow-up of >10 years. Second, during this time period, the definition of pancreas graft failure evolved regarding the amount and the duration of insulin administration. Based on the available information, we used a conservative approach by defining any type of partial function and death with a functioning graft as graft failure; therefore, we may have underestimated pancreas graft function. Third, despite our best efforts, information on the islet transplants was not available, and factors that may have had an impact on the subsequent pancreas transplant could not be investigated. Despite these limitations, this first IPTR/UNOS analysis demonstrated clearly that a PAI transplant is a safe procedure with very low mortality and provides a high chance of graft function both short and long term. Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation. References 1. Kelly WD, Lillehei RC, Merkel FK, Idezuki Y, Goetz FC. Allotransplantation of the pancreas and duodenum along with the kidney in diabetic nephropathy. Surgery 1967; 61: Gruessner RW, Gruessner AC. The current state of pancreas transplantation. Nat Rev Endocrinol 2013; 9: Najarian JS, Sutherland DE, Matas AJ, Steffes MW, Simmons RL, Goetz FC. Human islet transplantation: A preliminary report. Transplant Proc 1977; 9: Shapiro AM, Lakey JR, Ryan EA, et al. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoidfree immunosuppressive regimen. N Engl J Med 2000; 343: Close N, Anand R, Hering B, Eggerman T. NIH-supported national islet transplantationregistry. Cell Biochem Biophys 2004; 40 (Suppl): Gruessner RW, Gruessner AC. What defines success in pancreas and islet transplantation-insulin independence or prevention of hypoglycemia? A review. Transplant Proc 2014; 46: Gruessner AC, Sutherland DE, Gruessner RW. Long-term outcome after pancreas transplantation. Curr Opin Organ Transplant 2012; 17: Rizzari M, Gruessner A, Redfield R, et al. Pancreas retransplantation: A 10-year report of International Pancreas Transplant Registry outcomes. Am J Transplant 2014; 14(Suppl): abstract Barton FB, Rickels MR, Alejandro R, et al. Improvement in outcomes of clinical islet transplantation: Diabetes Care 2012; 35: Gruessner AC, Gruessner RW. Pancreas transplant outcomes for United States and non-united States cases as reported to the United Network for Organ Sharing and the International Pancreas Transplant Registry as of December Clin Transpl 2012: Maffi P, Scavini M, Socci C, et al. Risks and benefits of transplantation in the cure of type 1 diabetes: Whole pancreas versus islet transplantation. A single center study. Rev Diabet Stud 2011; 8: Lehmann R, Graziano J, Brockmann J, et al. Glycemic control in simultaneous islet-kidney versus pancreas-kidney transplantation in type 1 diabetes: A prospective 13-year follow-up. Diabetes Care 2015; 38: Naziruddin B, Wease S, Stablein D, et al. HLA class I sensitization in islet transplant recipients: Report from the Collaborative Islet Transplant Registry. Cell Transplant 2012; 21: American Journal of Transplantation 2016; 16: