Pancreas resection and islet autotransplantation for end-stage chronic pancreatitis

Abstract

Objective: To assess the safety and efficacy of islet autotransplantation (IAT) combined with total pancreatectomy (TP) to prevent diabetes.

Summary background data: There have been recent concerns regarding the safety of TP and IAT. This is thought to be related to the infusion of large volumes of unpurified pancreatic digest into the portal vein. Minimizing the volume of islet tissue by purifying the pancreatic digest has not been previously evaluated in terms of the postoperative rate of death and complications, pain relief, and insulin independence.

Method: During a 54-month period, 24 patients underwent pancreas resection with IAT. Islets were isolated using collagenase and a semiautomated method of pancreas digestion. Where possible, islets were purified on a density gradient and COBE processor. Islets were embolized into the portal vein, within the spleen and portal vein, or within the spleen alone. The total median volume of digest was 9.9 mL.

Results: The median number of islets transplanted was 140,419 international islet equivalents per kilogram. The median increase in portal pressure was 8 mmHg. Early complications included duodenal ischemia, a wedge splenic infarct, partial portal vein thrombosis, and splenic vein thrombosis. Intraabdominal adhesions were the main source of long-term problems. Eight patients developed transient insulin independence. Three patients were insulin-independent as of this writing. Patients had significantly decreased insulin requirements and glycosylated hemoglobin levels compared with patients undergoing TP alone. Of the patients alive and well as of this writing, four had failed to gain relief of their abdominal pain and were still opiate-dependent.

Conclusion: Combined TP and IAT can be a safe surgical procedure. Unfortunately, almost all patients were still insulin-dependent, but they had decreased daily insulin requirements and glycosylated hemoglobin levels compared with patients undergoing TP alone. A prospective randomized study is therefore needed to assess the long-term benefit of TP and IAT on diabetic complications.

Figure 1. Temporal secretion of basal C-peptide after islet cell autotransplantation. Box plots represent median, 95% confidence intervals, and range values. Kruskall-Wallis one-way analysis of variance test, P > .05.

Figure 2. Comparison of hemoglobin A1c (%) levels. Box plots represent median and 95% confidence intervals (Mann-Whitney test, P = .007). During follow-up (15 months to 5 years), the mean hemoglobin A1c level was calculated for each patient, and the mean value is plotted for islet autotransplant recipients (n = 21) and patients undergoing total pancreatectomy alone (n = 10).

Figure 3. Comparison of 24-hour insulin requirements (IU/24 hours) in islet autotransplant recipients and those undergoing total pancreatectomy alone. Plots represent mean values and standard error.