Background: Most modalities for tissue confirmation during endoscopic retrograde cholangiopancreatography (ERCP) suffer from low sensitivity and poor diagnostic accuracy. Probe-based confocal laser endomicroscopy (pCLE) was prospectively evaluated in a multicenter registry including 102 patients with indeterminate strictures and demonstrated excellent sensitivity (98 %). Yet, several false-positive cases were induced by benign inflammatory conditions resulting in a specificity of 67 %.
Aims: To evaluate the diagnostic performance of pCLE for the diagnosis of indeterminate biliary stricture; and to propose interpretation criteria for benign inflammatory conditions to reduce the number of false positives.
Methods: Sixty cases from the prospective registry were reviewed retrospectively (27 malignant, 33 benign) by a panel of three biliary endoscopists. Each case's clinical history, ERCP impression, and corresponding pCLE sequences was used to score image quality, propose presumptive diagnoses, and rate level of diagnostic confidence.
Results: Using the Miami Classification (MC), the overall accuracy in retrospectively diagnosing malignancy in those 60 cases was 85 versus 78 % for the prospective analysis, reducing the number of false positives from 12 to 8. A second review of the false-positive cases' pCLE sequences (benign inflammatory) helped refine the existing classification by identifying four descriptive criteria specific to benign inflammatory conditions (Paris Classification): Vascular congestion, dark granular patterns with scales, increased inter-glandular space, and thickened reticular structure.
Conclusions: The Paris Classification is a refinement of the existing Miami Classification to improve the accuracy of pCLE for diagnosing benign inflammatory strictures. Prospective multicenter studies are needed to further validate this refined classification criteria.