Japan Pancreatic Disease Research Foundation

ABSTRACTAimTraditional actuarial graft survival estimates are useful in pancreas transplantation (PTx) cases. From the perspective of patients, however, a more accurate estimate is necessary to predict long‐term prognosis with consideration of their specific circumstances. Graft conditional survival analysis after transplantation is useful, but no relevant studies have included PTx. This study addressed pancreas graft conditional survival (pCS) and associated factors in PTx recipients in Japan.MethodsA total of 526 PTx cases (2000–2023) were analyzed from the Japan Pancreas Transplant Registry. In addition to unconditional actuarial survival, pCS was calculated with the Kaplan–Meier method based on graft survival to landmark times, and factors affecting pCS were investigated using a Cox proportional hazards regression model.ResultsUnconditional actuarial 1/3/5/10‐year graft survival rates were 86.6%/80.8%/76.7%/68.3% after PTx. Donor sex (male vs. female), recipient age (≤ 49 vs. ≥ 50 years), and type of PTx procedure (simultaneous pancreas‐kidney transplantation vs. pancreas after kidney transplantation + pancreas transplantation alone) were significantly associated with survival. pCS increased with landmark time, indicating that posttransplant prognosis in surviving patients was more favorable than predicted at PTx. Donor sex and type of PTx procedure affected unconditional actuarial pancreas graft survival but not pCS, and recipient age influenced both pCS and unconditional actuarial pancreas graft survival.ConclusionpCS provided a dynamic perspective on long‐term pancreas graft survival, demonstrating improved prognosis with extended posttransplant time. Donor sex and PTx procedure type influenced unconditional actuarial survival, but their impact diminished with time, while recipient age remained a consistent determinant of pCS.

Donor-recipient (D/R) size mismatch has been evaluated for a number of organs but not for pancreas transplantation. We retrospectively evaluated 438 patients who had undergone pancreas transplantation. The D/R body surface area (BSA) ratio was calculated, and the relationship between the ratio and graft prognosis was evaluated. We divided the patients into two groups and evaluated graft survival. The incidence of pancreas graft thrombosis resulting in graft failure within 14days and 1-year graft survival were compared using Kaplan-Meier curves, and the prognostic factors associated with graft thrombosis were identified by univariate and multivariate analyses. The mean/median donor and recipient BSAs were 1.63m2 /1.65m2 , and 1.57m2 /1.55m2 , respectively; the mean and median D/R BSAs were both 1.05. The receiver operating characteristic curve cutoff for the D/R BSA ratio was 1.09, and significant differences were identified between patients with ratios of ≥1.09 (high group) versus <1.09 (low group). The incidence of graft thrombosis resulting in pancreas graft failure within 14days was significantly higher in the high group than in the low group (p<.01). One-year overall and death-censored pancreas graft survival were significantly higher in the low group than in the high group (p<.01). Multivariate analysis identified recipient height, donor BSA, and donor hemoglobin A1c as significant independent factors for graft thrombosis. Cubic spline curve analysis indicated an increased risk of graft thrombosis with increasing D/R BSA ratio. D/R size mismatch is associated with graft thrombosis after pancreas transplantation.

In cases after pancreas transplantation (PTx), the pancreas graft might be lost for various reasons, including a pancreatoduodenal graft-related complication or patient death with a functioning graft. Although the causes seem to have distinct characteristics, the causes of pancreas graft loss have not been characterized. This study aimed to characterize the causes of pancreas graft loss by analyzing data from a Japanese nationwide registry. This study included 391 patients who received simultaneous pancreas-kidney transplantation among 461 patients with PTx during the study period in approved institutions in Japan. We characterized each cause in terms of the timing of the cause-specific graft loss and preoperative factors associated with graft loss based on nationwide data from a Japanese nationwide registry. Among the 391 cases, 113 lost the pancreas graft due to patient death with a functioning graft (DWFG; n = 34, 44.2%), graft thrombus (n = 22, 28.6%), or chronic rejection (CR; n = 7, 9.1%). Average (± standard deviation) time from PTx to graft loss due to DWFG, graft thrombus, and CR was 3.70 ± 4.36, 0.02 ± 0.01, and 2.37 ± 2.08 years, respectively. Duration of type 1 diabetes mellitus and donor body mass index were significantly associated with pancreas graft loss due to DWFG and graft thrombus, respectively. This characterization showed that the timing and preoperative factors associated with pancreas graft loss were significantly different for different causes of graft loss. These results may inform PTx follow-up protocols to ensure that appropriate care is based on the cause of graft loss.

The feasibility of pancreas transplantation (PT) in older recipients remains a matter of debate. We examined the influence of recipient age on PT outcomes and identified the prognostic factors for older recipients. We compared the outcomes of PT in recipients aged < 50years (younger group; n = 285) with those in recipients aged ≥ 50years (older group; n = 94). Prognostic factors in the older group were analyzed by a logistic regression model and the influence of recipient age on survival outcomes were analyzed using propensity score matching. The patient survival rate was significantly worse in the older group (P < 0.001). Patient death from infection or/and multiple organ failure or cardiac/cerebrovascular events was also more frequent in the older group than in the younger group (P = 0.012 and P = 0.045, respectively). A longer duration of diabetes was an independent risk factor of 1-year mortality in the older group. In a propensity score-matched comparison, the older recipients (n = 77) had significantly poorer survival than the younger recipients (n = 77) (P = 0.026). PT should be considered with appropriate caution, especially for older recipients with a long duration of diabetes.

Go, Vay Liang W. MD; Shimosegawa, Tooru MD, PhD; Saluja, Ashok K. PhD; Fuchs, Anna R.; Takeuchi, Tadashi MD, PhD Author Information