Systematic review with meta-analysis: endoscopic retrograde cholangiopancreatography-based modalities for the diagnosis of cholangiocarcinoma in primary sclerosing cholangitis

Background: The accuracy of current endoscopic modalities for diagnosing cholangiocarcinoma in primary sclerosing cholangitis (PSC) is suboptimal. Aim: To evaluate the comparative effectiveness of endoscopic retrograde cholangiopancreatography (ERCP)-based modalities, independently or in combination, for the diagnosis of cholangiocarcinoma in patients with PSC-induced biliary strictures. Methods: Searches of PubMed, EMBASE, Web of Science and the Cochrane Library databases were performed through December 2015. Measured outcomes included sensitivity and specificity of all diagnostic modalities independently or in combination. A bivariate model was used to compute the pooled sensitivity and specificity, and to plot the summary receiver operating characteristics curve with summary point and corresponding 95% confidence interval (95% CI). A logistic regression model was used to impute the incremental performance of combining two diagnostic tests. Results: Twenty-one studies met inclusion criteria: 13 on bile duct brushing for cytology, 7 on fluorescence in situ hybridisation (FISH), 2 on probe-based confocal laser endomicroscopy, and 4 on single-operator cholangioscopy with targeted biopsies. Single-operator cholangioscopy with targeted biopsies was the most accurate diagnostic modality at 96% (95% CI, 94–97%). The pooled sensitivity and specificity of single-operator cholangioscopy for diagnosis of cholangiocarcinoma in patients with PSC was 65% (95% CI, 35–87%) and 97% (95% CI, 87–99%), respectively. The pooled diagnostic odds ratio to detect cholangiocarcinoma was 59 (95% CI, 10–341). Conclusions: Single-operator cholangioscopy with targeted biopsies appears to be the most accurate ERCP-based modality for diagnosing cholangiocarcinoma in primary sclerosing cholangitis. However, future large, well-designed comparative diagnostic studies are warranted to validate these findings.