Performance of fluorescence in situ hybridization in biliary brushings with equivocal cytology: an institutional experience.

INTRODUCTION: Biliary brushing (BB) cytology has a sensitivity of 15%-65% and specificity approaching 100% for detecting malignancy. Fluorescence in-situ hybridization (FISH) using the UroVysion probe set has been advocated to enhance the detection of malignancies with reported sensitivity of 43%-84%. We sought to evaluate the performance of FISH in BB with equivocal cytology at our institution.

MATERIALS AND METHODS: Patients with atypical and suspicious BB with concurrent diagnostic FISH performed at our institution from 2014 to 2021 were identified through a query of our pathology database. FISH (using UroVysion probe set containing centromere enumeration probes to chromosomes 3, 7, and 17) was positive if at least 5 cells demonstrated polysomy. Electronic medical records were reviewed for pathology results and outcomes. Patients were classified malignant if they had positive pathology or documented clinical impression of malignancy and benign if they had negative pathology and/or documented benign clinical course for at least 12 months.

RESULTS: We identified 254 equivocal BB (238 atypical/16 suspicious) with concurrent FISH results from 191 patients (105 benign, 86 malignant). 12% (22/191) of patients were FISH positive. Twenty-four percent (21/86) of patients with malignancy had positive FISH but were nonspecific for pancreaticobiliary/ampullary adenocarcinomas. Almost all positive FISH were associated with malignancy (21/22; 95%). There was 1 positive FISH in a patient with primary sclerosing cholangitis who had a benign outcome.

CONCLUSIONS: The small number of positive FISH results in BB with equivocal cytology raises the question of the optimal criteria for malignancy. Using only polysomy could result in lower sensitivity.