191P ROS1 fusions in resected stage I-III adenocarcinoma (ADC): A Lungscape ETOP study

ROS1 fusion is a relatively low prevalence (1–2%) but targetable driver in non-small cell carcinoma (NSCLC). Therefore, robust and low-cost tests are desirable in order to identify patients harboring this fusion. ROS1 immunohistochemistry (IHC) using monoclonal antibody clone D4D6 is widely used for screening patients for a potential ROS1 fusion. However, due to relatively low specificity, confirmation using an orthogonal test (FISH or NGS) has been reported as necessary. We investigated the performance of a novel IHC clone SP384 in a retrospective, stage I-III, lung ADC cohort. FFPE sections of resected ADC from the ETOP Lungscape multicenter cohort, constructed in tissue microarrays, were prospectively stained for ROS1 protein expression using the SP384 clone in a ready-to-use kit and Ventana immunostainers. After passing an external quality control, scoring was performed locally by trained pathologists using the H-score. Staining intensity of 2+ (any %) was considered IHC positive (IHC 2/3+). Subsequently, IHC 2/3+ cases were 1:1:1 matched with IHC 0/1+ cases, and subjected to FISH and NGS analyses. Valid immunostaining results were available in 866 pts with 35 (4%) IHC 2/3. Nineteen had an average H-score of <100, eleven of 100–199 and five of 200–300. Seven IHC 2/3+ cases were not further tested (5 with an EGFR mutation (1 with H-score >200), 1 with an ALK fusion, and 1 with no material left). Of the remaining 28 IHC 2/3+ cases, only 2 (50% of 4 tested with H-score ≥200) were confirmed to have a ROS1 rearrangement, leading to an overall prevalence of 0.23%. Of the ROS1 IHC 2/3+ cases not confirmed by FISH or NGS, 7 harbored a KRAS mutation and 1 a MET mutation. All matched IHC 0/1+ cases were negative by FISH/NGS. Thus, ROS1 positivity and negativity by FISH-NGS were correctly predicted 100% by IHC 2/3+, and 96% by IHC 0/1+, respectively. The prevalence of ROS1 fusion in an ADC cohort screened by IHC using SP384 clone was relatively low compared to literature. No additional ROS1 fusion was identified by FISH or NGS among IHC 0/1+. Thus, this method is useful for preselection of patients with a potential ROS1 fusion, followed by confirmatory FISH and/or NGS.