Recommendations for immunocytochemistry in lung cancer typing: An update on the resource-efficient approach with large-scale comparative Bayesian analysis

Objectives The majority of lung cancer cases are of advanced stage and diagnosis is usually made using minimally invasive small biopsies and cytological specimens. The WHO'2015 classification recommends limiting immunocytochemistry(ICC) to lung cancer typing and molecular testing drives for personalized therapies. An algorithm using Bayes' theorem could be useful for defining antibody profiles. This study aims to assess the impact of using different antibody profiles for cytological samples on the accuracy of lung cancer typing with a large scale Bayesian analysis. Methods Between 2011-2016 a total of 3,419 consecutive smears and/or cytospins were diagnosed. There were 1,960 primary lung cancer (972 adenocarcinomas (ADC), 256 squamous carcinomas(SQC), 268 neuroendocrine tumors[(NET);166 SCLC; 85 large cell neuroendocrine carcinoma(LCNEC),7 atypical carcinoid(AC),10 typical carcinoid(TC)], and 464 tumors without any ADC, SQC or neuroendocrine growth features(NSCC-NOS). Before and after the use of ICC with single/combination of antibodies, the a priori and a posteriori probabilities for different lung cancer types were calculated. Results TTF-1 or CK7 alone improves the a posteriori probabilities of correct cytological typing of ADC to 86.5% and 95.8% respectively. In SQC, using p40(∆Np63) or CK5/6, CK5/14 led to comparable results (78.3% and 90.3%). With synaptophysin or CD56 alone a posteriori probabilities for the correct recognition of NET of 87.5 and 90.3% could be achieved. Conclusions Based on morphological and clinical data, the use of two antibodies appears to be sufficient for the reliable detection of the different lung cancer types. This applies to both clinical-cytological and histological-only validated final diagnoses.