A major problem in vascularized pancreas transplantation is the lack of reliable methods for the early diagnosis and effective treatment of allograft rejection. Over a 2-year period, 54 rejection episodes occurred in 31 patients (13 isolated pancreas, 18 simultaneous pancreas-kidney recipients) with pancreaticoduodenocystostomy. A total of 253 radionuclide pancreas examinations were performed (mean 8.4 per patient) utilizing 99mtechnetium-DTPA. Computer analysis generated a quantitative measure of blood flow to the allograft caused the technetium index (TI). Rejection episodes were characterized as isolated pancreas (22), combined pancreas-kidney (16), or isolated renal (16) allograft rejection in combined engraftments. The majority of rejection episodes occurred early (within 3 months of transplant, N = 47) and were more responsive than late rejection to anti-rejection therapy (89.4% vs 42.9%, P = 0.01). Mean urinary amylase (UA) levels and TI during normal allograft function were 29,398 U/l and 0.55%, while levels heralding rejection were 6,528 U/l and 0.40%, respectively (P less than 0.05). The treatment of rejection based upon renal dysfunction or combined renal and pancreas dysfunction resulted in significantly higher graft salvage with a lower incidence of hyperglycemia when compared to isolated pancreas allograft rejection. Of the 11 patients who developed hyperglycemia, 8 (72.7%) ultimately lost their pancreas grafts (P less than 0.001). Following therapy, a TI above 0.3% was associated with 97.4% graft survival, while levels below 0.3% resulted in a 70% rate of graft loss (P less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)