Improved classification of indeterminate biliary strictures by probe-based confocal laser endomicroscopy using the Paris Criteria following biliary stenting.

Abstract

Probe-based confocal laser endomicroscopy (pCLE) using the Miami Criteria has improved classification of indeterminate biliary strictures. However, previous biliary stenting may result in their misclassification as malignant strictures. Inflammatory criteria were added to form the Paris Classification to prevent this misclassification and reduce false positives. The aim of this study was to assess if the Paris Classification was more accurate than the Miami Classification in classifying indeterminate biliary strictures after biliary stenting. This was a retrospective observational study involving 21 patients with indeterminate biliary strictures from whom 27 pCLE video sequences were obtained (20 benign and seven malignant). Patients with and without prior biliary stenting underwent pCLE. Two investigators classified the strictures as malignant or benign using the Miami and Paris Classifications. Diagnostic accuracy, sensitivity (Se), and specificity (Sp) of endoscopic retrograde-guided pCLE were compared with final histopathology. In those without biliary stenting, the Miami Criteria resulted in Se 88%, Sp 75%, positive predictive value (PPV) 64%, negative predictive value (NPV) 92%, and accuracy 79%, while the Paris Classification resulted in Se 63%, Sp 88%, PPV 71%, NPV 82%, and accuracy 79%. In those with prior biliary stenting, the Miami Criteria resulted in Se 88%, Sp 36%, PPV 23%, NPV 93%, and accuracy 45%, while the Paris Classification resulted in Se 63%, Sp 73%, PPV 31%, NPV 91%, and accuracy 71%. The kappa statistic was 0.56. The Paris Classification improved specificity and accuracy of biliary stricture classification in those who had been previously stented and decreased the rate of misclassification of benign strictures as malignant.