Surgical approach to insulinomas. Assessing the need for preoperative localization.

The purpose of this study was to examine our experience with the diagnosis, surgical approach, and outcomes of surgery for organic hyperinsulinemia in the era of transhepatic venous sampling. During the period from 1978 to 1991, 50 patients were evaluated and treated for hyperinsulinemia at the University of Michigan Medical Center, all of whom underwent preoperative localization. Forty-one patients (82%) had solitary, benign tumors; four (8%) had either multiple tumors or islet cell dysplasia; and five (10%) had metastatic disease. Forty-seven patients underwent laparotomy, and the source of the hyperinsulinemia was found in all patients. In three patients (6%) preoperative localization did not accurately locate the tumor due to technical difficulties with the completion of the studies. Overall, computed tomography localized nine (26%) of 35 tumors. Angiography accurately localized 18 (44%) of 41 tumors. Transhepatic venous sampling localized 34 (94%) of 36 tumors, and was essential to successful surgical treatment in 15 patients. Compared with angiography, transhepatic venous sampling was a more accurate method of localization and should be performed in all patients in whom computed tomography and visceral angiography do not clearly identify the site of disease. Preoperative localization plays a critical role in the surgical treatment of patients with organic hyperinsulinemia and eliminates the need for blind pancreatic resection.