Value of abdominal sonography and MR imaging at 0.5 T in preoperative detection of pancreatic insulinoma: a comparison with dynamic CT and angiography.

Abstract

Abdominal sonography, computed tomography (CT), angiography, and magnetic resonance (MR) imaging are the most widely used modalities for preoperative localization of insulinomas. CT and angiography are generally considered the techniques of reference, and the role of sonography and MR imaging in these patients is controversial. The purpose of this study was to compare these four modalities in a group of patients with pancreatic insulinoma and determine an effective radiological approach to this disease. Twenty-eight patients with clinical and biochemical signs of pancreatic insulinoma underwent abdominal sonography, MR imaging at 0.5 T (spin echo technique), bolus dynamic CT, and digital subtraction angiography. Examinations were evaluated independently for the presence, size, and location of the lesions; preoperative diagnoses were compared with surgical findings based on palpation and intraoperative sonography. Tumoral vascularity was histologically graded. Sensitivities of the four imaging techniques were calculated and compared with the size, location, and vascularity of the tumors. Detection rates of combined techniques were finally determined. At surgery, 29 lesions in the 28 patients were found (range = 0.8-4.3 cm, average = 1.65 cm). Sensitivities of abdominal sonography. MR imaging, CT, and angiography were 79.3%, 65.5%, 44.8%, and 69% respectively. Correct localization of tumor was achieved in 96.6% of cases by a combination of sonography and MR imaging and in 72.4% of cases by using CT with angiography. In our experience, sonography and MR imaging performed well in the preoperative detection of pancreatic insulinoma. Therefore, we believe that the combination of abdominal sonography and MR imaging may represent the first radiological approach in clinically suspected insulinomas and that CT and angiography should be reserved for negative and/or doubtful cases.