Brush Cytology, Forceps Biopsy, or Endoscopic Ultrasound-Guided Sampling for Diagnosis of Bile Duct Cancer: A Meta-Analysis.

Endoscopic sampling is essential for tissue diagnosis of cholangiocarcinoma (CCA). To evaluate and compare the diagnostic sensitivities of endoscopic retrograde cholangiopancreatography-guided brush cytology biopsy, and endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) in patients with CCA. A comprehensive literature search through multiple databases was conducted for articles published between January 1995 and August 2020. The pooled rates of sensitivity for the diagnosis of CCA and of adverse events were compared among brushing, biopsy, brushing & biopsy, and EUS-FNA. In total, 1123 patients with CCA (32 studies), 719 patients (20 studies), 358 patients (13 studies), and 422 patients (17 studies) were tested by brushing, biopsy, brushing & biopsy, and EUS-FNA, respectively. The pooled diagnostic sensitivity was 56.0% (95% confidence interval (CI) 48.8-63.1%, I2 = 83.0%) with brushing, 67.0% (95% CI 60.2-73.5%, I2 = 72.5%) with biopsy, 70.7% (95% CI 64.1-76.8%, I2 = 42.7%) with brushing & biopsy, and 73.6% (95% CI 64.7-81.5%, I2 = 74.7%) with EUS-FNA. The diagnostic sensitivity was significantly lower for brushing than for biopsy, brushing & biopsy, or EUS-FNA. No significant difference was noted in diagnostic sensitivities among biopsy, brushing & biopsy, and EUS-FNA. Adverse events were comparable between the groups. Intraductal biopsy, brushing & biopsy, and EUS-FNA had comparable efficacy and safety for the diagnosis of CCA. Brushing was the least sensitive diagnostic tool compared with intraductal biopsy or EUS-FNA. Given the modest diagnostic sensitivities of intraductal biopsy and EUS-FNA in the diagnosis of CCA, further studies for complementing these techniques with biomarkers may be needed.