Systematic Evaluation of Pancreas Allograft Quality, Outcomes, and Geographic Variation in Utilization
Axelrod DA, Sung RS, Meyer KH, Wolfe RA, Kaufman DB
Am J Transplant. 2010;10:837-845
Study Summary
Scientific Registry of Transplant Recipients (SRTR) data were reviewed retrospectively for all patients who underwent simultaneous pancreas-kidney (SPK, n = 6248) or sequential pancreas after kidney (PAK, n = 2373) transplantation between January 1, 2000, and January 31, 2006. Using a multivariable Cox regression analysis and controlling for donor and recipient characteristics, 9 donor variables (age, gender, black race, Asian race, body mass index [BMI], height, stroke as cause of death, donation after cardiac death [DCD], and serum creatinine) and 2 transplant factors (cold ischemia time and an interaction between stroke as a cause of death and PAK transplant) were used to construct a pancreas donor risk index (DRI).
Among the donor factors examined, increasing donor age (relative risk [RR] 1.56) and DCD donor status (RR 1.39) appeared to have the greatest effect on risk for early graft loss. Increasing pancreas DRI was associated with a significant increase in graft loss in the first year following transplantation for all 3 categories of pancreas transplantation (SPK, PAK, pancreas alone transplantation [PAT]). Nationally, a significant difference between the pancreas DRIs of pancreata recovered but not transplanted and pancreata that were transplanted was observed. Although the mean pancreas DRIs of organs used for SPK did not differ significantly from those used for PAK/PAT, the 1-year graft survival rate (77%) in recipients of PAK/PAT from donors with an elevated pancreas DRI was lower than the survival rate for SPK transplants (83%) from donors with an elevated pancreas DRI. The mean pancreas DRI decreased nationally over time, suggesting a recent decline in the use of marginal donors for pancreas transplantation. Pancreas allograft acceptance varied widely by region when controlling for pancreas DRI, particularly for PAK/PATs. Not surprisingly, the use of higher DRI pancreata was correlated with center activity, with higher volume centers having a mean higher pancreas DRI.
Viewpoint
In this study, the investigators developed a pancreas DRI as a measure of donor quality. This risk index is similar to the expanded criteria donor (ECD) definition in kidney transplantation. However, unlike the ECD definition, which is a dichotomous variable, the pancreas DRI is a continuous scale that attempts to determine the relative risk of graft loss at 1 year following pancreas transplantation, on the basis of a number of variables. This is a timely study of the role of donor factors in determining pancreas transplant outcomes, particularly because the number of pancreas transplant donors and recipients in the United States has declined in recent years. In 2008, the overall recovery rate of pancreata from deceased donors was 23%, and the transplant rate of recovered pancreata was only 72%. The pancreas discard rate has doubled over the past decade, whereas the pancreas recovery rate has remained static. Only 51% of recovered pancreata are transplanted locally, and much of the decline in transplanted pancreata can be attributed to a reduction in the number of sequential PAK transplants. The reasons for these ominous trends in national pancreas transplantation are believed to be multifactorial and include an uneven distribution of pancreas transplant programs across the United States, difficulty in sharing kidney-pancreas transplants between different Donor Service Areas (DSA), caution by many programs in acceptance of pancreata recovered outside their own DSA, improvements in the medical therapy of diabetic kidney disease, and age and obesity trends in the donor and recipient populations. The pancreas DRI can be used to assess the differential effect of organ quality on pancreas transplant outcomes as well as account for interactions between various factors.
Abstract