Pancreatic Insulinomas

OBJECTIVE: To describe our experience in the management of patients with pancreatic insulinomas, emphasizing the need for preoperative localization and the outcome of surgical treatment. DESIGN: A case series. SETTING: A university hospital in Hong Kong. PATIENTS: From 1981 to 1995, 27 patients with pancreatic insulinomas were surgically treated; the mean follow-up was 25.4 months. MAIN OUTCOME MEASURES: Postoperative morbidity and euglycemia during the follow-up period. RESULTS: The accuracy of tumor localization by ultrasonography, computed tomography, and angiography was 33%, 44%, and 52%, respectively. Venous sampling for an insulin assay regionalized 90% of the tumors. In 24 patients with solitary tumors, most lesions detected or missed by preoperative localization could be either seen (n = 14) or palpated (n = 22). Intraoperative ultrasonography (n = 17) has been routinely performed since 1987; nonpalpable tumors were imaged in 2 of 15 patients with solitary tumors. Eight solitary occult tumors were detected by palpation alone or a combination of palpation and intraoperative ultrasonography. Operative mortality occurred in 1 (3.7%) of the 27 patients, while major morbidity developed in 9 (33%) of the patients. Euglycemia was achieved in 25 patients. Surgery cured all patients with benign insulinomas, whereas the cure rate for patients with malignant neoplasms was only 33%. The type of surgical treatment or correct preoperative localization did not affect the outcome of surgery. CONCLUSIONS: Pancreatic insulinomas can be readily localized intraoperatively despite failed preoperative localization studies. Surgical treatment cured benign adenomas but was associated with notable morbidity.