Diagnostic accuracy of cross-sectional and endoscopic imaging in ampullary tumours: systematic review

Abstract Background Differentiation between adenomas and carcinomas of the ampulla of Vater is crucial for therapy and prognosis. This was a systematic review of the literature on the accuracy of diagnostic modalities used to differentiate between benign and malignant ampullary tumours. Methods A literature search was conducted in PubMed, Embase, CINAHL, and the Cochrane Library. Studies were included if they reported diagnostic test accuracy information among benign and malignant ampullary tumours, and used pathological diagnosis as the reference standard. Risk of bias was assessed using Quality Assessment on Diagnostic Accuracy Studies (QUADAS) 2 and QUADAS-C. Results Ten studies comprising 397 patients were included. Frequently studied modalities were (CT; 2 studies), endoscopic ultrasonography (EUS; 3 studies), intraductal ultrasonography (IDUS; 2 studies), and endoscopic forceps biopsy (3 studies). For CT, the reported sensitivity for detecting ampullary carcinoma was 44 and 95%, and the specificity 58 and 60%. For EUS, the sensitivity ranged from 63 to 89% and the specificity between 50 and 100%. A sensitivity of 88 and 100% was reported for IDUS, with a specificity of 75 and 93%. For forceps biopsy, the sensitivity ranged from 20 to 91%, and the specificity from 75 to 86%. The overall risk of bias was scored as moderate to poor. Data were insufficient for meta-analysis. Conclusion To differentiate benign from malignant ampullary tumours, EUS and IDUS seem to be the best diagnostic modalities. Sufficient high-quality evidence, however, is lacking.