PTU-124 Increased diagnostic sensitivity of pancreaticobiliay malignancy by modifying ERCP brushing practice: a single centre experience

Introduction Brushings at ERCP play a crucial role in establishing a diagnosis in cases of biliary strictures. In this study, we examined if brushing practice can make a difference to the diagnostic yield resulting in less interventions. We analysed ERCP data over the last two and a half years at the Royal Berkshire Hospital, where inclusion of the brush head for cytology and performing three or more passages across the stricture has been standard practice. Methods We looked at results of brushings taken by 2 ERCPists in a single centre between September 2016 to February 2018. The initial brushing result from each individual patient was included. Brushing results were classified as non-diagnostic, negative, atypical, suspicious or positive for malignancy. Brushings classified as suspicious and identified as malignant were categorised into the positive group; those classified as non-diagnositc, non-malignant and atypical were categorised as negative. For comparison purposes we considered the corresponding histology and/or radiological findings and/or positive immunohistochemistry in the follow-up period. Data was analysed using the statistical package SPSS v.25. Results In total, 71 individual brushing outcomes were identified over a 2.5-year period. In 61 of 71 cases (85.9%) the final diagnosis was cancer, with the majority being pancreatic cancer (36/61; 59%). The sensitivity was estimated at 77% (47/61 true positive), the specificity and the positive predictive value (PPV) were both 100% (0 false negative, 10 true negative), while the negative predictive value (NPV) was 38%. When we looked at only ‘positive for malignancy’ results, the sensitivity stood at 60%. In 15/61 cases (24.6%) a supplementary report was needed to confirm the result and it changed the outcome in only 3 cases (20%). Conclusions Biliary brushings are a very useful means of providing a diagnosis during ERCP and are characterised by a high PPV and low NPV. Negative results should not be interpreted as absence of malignancy but if the yield is at least suspicious, the specificity approaches 100%. This study suggests that sending the brush head for cytology and performing three or more passages across the stricture increases diagnostic sensitivity when compared to published data reporting on sensitivities up to 64% (Burnett AS, Calvert TJ, Chokshi RJ. Sensitivity of endoscopic retrograde cholangiopancreatography standard cytology: 10-y review of the literature. J Surg Res. 2013;184:304–11.)