Use of glucose disappearance rates (k(G)) to monitor endocrine function of pancreas allografts

We have reported that a decline in glucose disappearance rate (k(G)) in pancreas transplant recipients is associated with pancreatic rejection. The purpose of this study was to determine test-retest reliability of k(G), monitoring and to establish the k(G) criteria for diagnosing abnormal graft function. Six healthy non-diabetic volunteers and 14 stable pancreas recipients underwent 2 intravenous glucose tolerance tests 7 d apart. All k(G) values in non-diabetic volunteers had < 15% variation between the two determinations (r = 0.96, P ≤ 0.0006). Similarly, 13/14 recipients experienced < 20% variation in k(G) with one patients experiencing a 23% variation (r = 0.90, P ≤ 0.0001). Using a 20% change from baseline as the reference value, we monitored 28 pancreas recipients for 2-36 months post-transplant. Of 253 k(G) values, 160 (64%) did not exceed the 20% change from baseline, 65 (26%) declined > 20% and 28 (11%) increased > 20%. Of 160 stable k(G) values, 154 (96%) were associated with normal graft function while 6 (4%) occurred in the context of rejection. Of 65 k(G), values declining by ≤ 20%, 47 (72%) accurately identified acute rejections diagnosed by biopsy (70%) or response to treatment (30%), 12 (19%) were associated with identifiable causes of increased insulin resistance and only in 6 (9%) cases a cause for the decline was unidentifiable. The k(G) values with > 20% rise from baseline were, in 15%, associated with identifiable causes of decreased insulin resistance. The sensitivity of the k(G) as a marker for rejection was 88.7%, specificity 91%, positive predictive value 72.3%, negative predictive value 96.8%, and accuracy 90.5%. These data confirm that a > 20% deterioration of glucose disappearance rate is associated with pancreas allograft rejection, and confirms the utility of k(G) monitoring in clinical follow-up of pancreas transplant recipients.