LAUSR

OBJECTIVE To assess the value of 18F-FDG PET/CT in differentiating between benign and malignant intraductal papillary mucinous neoplasms (IPMN) of the pancreas. SUMMARY BACKGROUND DATA Malignant or high-risk IPMN require surgical resection but surgery should be avoided in patients with IPMN carrying a low risk of malignancy. 18F-FDG PET has been studied mostly in small, single center, retrospective series. METHODS Prospective, non-comparative, multicenter French study. The primary endpoint was the specificity of PET/CT for identifying malignant IPMN (in situ or invasive carcinoma). Final diagnosis was obtained from pathological examination of the resected specimen. RESULTS Among 120 patients analyzed, 99 had confirmed IPMN, including 24 with malignant lesions, namely 9 with carcinoma in situ and 15 with invasive carcinoma. The 18F-FDG PET/CT was positive in 44 and 31 patients in the overall and IPMN populations respectively. In the 99 IPMN patients, PET/CT showed 13 true positive, 18 false positive, 57 true negative and 11 false negative results. The sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) for the diagnosis of malignancy were 54.2%, 76.0%, 83.8% and 41.9% respectively, versus 64.9%, 75.9%, 82.9% and 54.5% in the overall population. We could not identify a cut-off value for SUVmax to distinguish benign from malignant lesions. Conventional imaging included computed tomography, magnetic resonance cholangiopancreatography and endoscopic ultrasound. In IPMN patients who underwent the 3 techniques, sensitivity, specificity, NPV and PPV were 66.7%, 84.4%, 84.4% and 66.7% respectively. CONCLUSIONS In this study, 18F-FDG PET/CT did not perform better than conventional imaging to differentiate malignant from benign IPMN.