Diagnosis of pancreatic allograft rejection by measurement of urinary radioimmunoreactive insulin

The accurate, early, and sensitive diagnosis of pancreatic allograft rejection is one of the major problems in clinical pancreas transplantation today. Pancreati-cocystotomy is a popular technique for pancreas transplantation that permits simple and frequent urinary chemistry examinations. In this experiment, 20 mongrel dogs with a bladder-drained pancreatic transplant had serial monitoring of urinary amylase (UA) and urinary insulin (UI). The mean UI in the nonrejection state was 9.6+12 mlU/L. In eight dogs varying degrees of rejection were documented by histopathology. All three animals having severe acute rejection had high levels of IU (all>300-800 mlU/L). Of the five animals with mild-to-moderate rejection, all had significant UI elevations to >100 mlU/L but none had elevations above 200 mlU/L (P<0.05 for all groups). Ten animals were treated with prednisone, Imuran, and cyclosporine (CsA), and five of these dogs had good graft function for >14 days, during which the mean UI was extremely low (11±6.4 mlU/L, P<0.05). These values were not significantly different from the 0—14-day values for three pancreas autotrans-plants with bladder drainage (8.9±7.2 mlU/L, P<0.05). All rejections were preceded by significant rises of UI occurring two to five days prior to rejection. In seven animals, early graft dysfunction (1—4 days) developed, with total graft necrosis by five days. This graft injury was presumably caused by preservation damage or early vascular thrombosis and was associated with early (1— 4 days) marked elevations of the UI (>300 mlU/L). None of the animals with grafts surviving to rejection at seven days or more had these early severe elevations, and thus these early UI rises are pathognomonic of graft damage. In contrast, UA and lipase showed inconsistent association with rejection or early damage, although falls in UA generally occurred at or following the time of rejection. Marked daily variations in UA measurements were the most difficult aspect of UA monitoring. Serial electromagnetic flow probe studies of blood flow to the pancreas graft showed a good correlation between loss of blood flow and rises in UI associated with early graft injury. These results suggested that the UI assay gives a sensitive, early, accurate, and specific differential prediction of pancreas graft dysfunction. Specifically, the UI assay appears to be of value not only in the early differential diagnosis of graft injury and graft rejection, but also in the assessment of the severity of rejection. Finally, preliminary studies suggest that the treatment of pancreas graft rejection following a diagnosis by UI permits reversal of rejection in 75% of cases—thus suggesting a potential utility for this test in pancreatic transplantation.