ROS1 immunohistochemistry as a screening tool for detection of ROS1 rearrangement in non-small cell lung cancer: Comparison between SP384 and D4D6 clones.

Abstract

e21733 Background: ROS1 rearrangement occurs in approximately 1%-2% of non-small cell lung cancer, and it is predictive of treatment response to tyrosine kinase inhibitor. The gold standard method to detect ROS1 rearrangement is break-apart fluorescence in situ hybridization (FISH). ROS1 immunohistochemistry has been proposed as a cost-effective screening test for detection of ROS1 rearrangement. Most studies about the correlation between ROS1 FISH and immunohistochemistry derived from the D4D6 clone anti-ROS1 antibody. Recently, a novel anti-ROS1 antibody (SP384 clone) is available. However, data regarding the performance of this new anti-ROS1 antibody is still limited. In this study, we evaluated the performance of these two anti-ROS1 antibody as a screening tool to detect ROS1 rearrangement. Methods: Two hundred non-small cell lung cancers, including 18 ROS1 FISH-positive and 182 ROS1 FISH-negative samples, were studied. All specimens were stained with D4D6 clone and SP384 clone anti-ROS1 antibodies. The stained slides were scored by a pathologist using the H-score method, which was defined as the sum of products of multiplying intensity (0, 1, 2, and 3) by extent of each staining intensity (%). Receiver Operating Characteristics (ROC) curves were used to determine the optimal cut-off value that discriminates between ROS1-rearranged and non-rearranged tumors. Results: There was high correlation between the H-score of D4D6 clone and SP384 clone (Spearman's rho correlation coefficient = 0.862, P-value < 0.001). The SP384 clone showed a higher H-score than D4D6 clone (Wilcoxon signed rank test, P-value < 0.001). ROC analysis showed H-score of ≥ 150 was the optimal cut-off value for both D4D6 clone and SP384 clone to discriminate between ROS1-rearranged and non-rearranged tumors. Using this cut-off value, the D4D6 clone showed 100% sensitivity and 96.7% specificity; the SP384 clone showed 100% sensitivity and 95.1% specificity. The areas under ROC curve for D4D6 clone and SP384 clone were 0.992 and 0.990, respectively. Conclusions: The performance of D4D6 clone or SP384 clone is comparable. ROS1 IHC with either D4D6 clone or SP384 clone is a cost-effective screening tool for the presence of ROS1 rearrangements. [Table: see text]