Results of pancreas transplantation in the United States for 1987-90 from the United Network for Organ Sharing (UNOS) Registry with comparison to 1984-87 results

An outcome analysis was performed on pancreas transplants in the United States reported to the United Network for Organ Sharing (UNOS) Registry from its inception on 1 October 1987 to 21 October 1990 (n = 1021). These cases comprise nearly one-third of the 3082 pancreas transplants reported to the International Pancreas Transplant Registry (1819 U.S., 1263 non-U.S.) from 1 December 1966 to 31 December 1990, including 619 in 1990 (528 U.S., 91 non-U.S.). Nearly all pancreas transplants in the U.S. during the 1987-90 period were by the bladder-drainage (BD) technique (92%). The overall patient and pancreas graft actuarial survival rates were 92% and 72% at 1 year. Patient survival rates were similar in all recipient categories, but pancreas graft survival rates were significantly higher (p < 0.001) in recipients of a simultaneous pancreas and kidney (SPK) transplant (n = 883) than in recipients of a pancreas after a kidney (PAK, n = 112) or a pancreas transplant alone (PTA, n = 71), being 77%, 52%, and 54%, respectively, at 1 yr. Kidney graft survival at 1 yr in U.S. SPK recipients was 86%. Most grafts (81%) were preserved in University of Wisconsin (UW) solution, and more than half were stored > 12 hours, with no difference in outcome with increasing duration of storage. At 1 yr, functional survival rates were 72% for U.S. pancreas grafts stored for either < 12 (n = 439), 12-24 (n = 422), or 24-30 h (n = 42). For grafts stored > 30 h (n = 8), the 1-yr functional survival rate was 50% (p = ns versus the other storage times). On univariate analysis, no effect of HLA antigen mismatching on outcome for 1987-90 U.S. cases could be discerned. The results in the UNOS Registry were compared to the results for U.S. cases in the International Pancreas Transplant Registry performed between 1 January 1984 and 30 September 1987. In all recipient categories the pancreas graft functional survival rates were significantly higher in the 1987-90 (UNOS) than in the 1984-87 (pre-UNOS) era. A Cox multivariate analysis of 1984-90 cases showed the relative risk for pancreas graft loss to be significantly less (p < 0.05) with bladder-drainage, with simultaneous transplantation of the kidney, with use of UW solution for preservation, and with 0-1 HLA-A, B, DR or 0 HLA-DR mismatches. However, transplantation in the 1987-90 UNOS era was an independent variable associated witha lower risk of graft loss, indicating that experience may also be a factor in the improvement in results. This interpretation is supported by the fact that there was a significant decrease in the technical failure rate for 1987-90 versus 1984-87 cases, and by the fact that, in the Cox analysis restricted to technically successful cases, era was not a significant independent variable. A separate Cox multivariate analysis restricted to 1987-90 UNOS cases showed only recipient category to have a significant effect on outcome, the risk for pancreas graft loss being lowest with SPK transplants. The results of pancreas transplantation have continuously improved in the U.S. and graft survival rates are now in the range of those for other solid organ transplants.