ROS1 immunohistochemistry as a potential predictive biomarker for ROS1-targeted therapy in breast cancer: impact of antibody clone selection

Abstract Aims Invasive lobular carcinoma (ILC) may show targetable vulnerabilities secondary to the characteristic loss of the cell adhesion protein E-cadherin. Specifically, a synthetic lethal interaction was identified between E-cadherin loss and ROS1 inhibition. Several clinical trials are currently underway to assess the efficacy of ROS1 inhibitors in ILC, however, ROS1 expression has not been confirmed in ILC tumours and ROS1 has not been validated as a biomarker in the breast cancer setting. This study aimed to (i) examine ROS1 expression in a large cohort of breast cancer cases and (ii) investigate the biology and clinical significance of ROS1 positivity in breast cancer. Methods ROS1 immunohistochemistry was performed on a large cohort of ILC (n=274) and invasive carcinoma of no special type (NST; n=431) cases with extensive clinicopathological data. The staining performance of four ROS1 antibody clones was compared. Results There was marked variation in ROS1 status according to antibody clone. D4D6 and SP384 were negative in almost all breast cancer cases, whereas EP282 and EPMGHR2 were positive in 37% and 47% of ILC cases, and 49% and 74% of NST cases, respectively. Only data from clones D4D6 and SP384 were highly concordant, while EP282 and EPMGHR2 were positive in distinct breast cancer subtypes. Conclusions Assessment of ROS1 status in breast cancer appears to be highly antibody clone dependent. ROS1 antibody clone selection will be an important consideration in the design of clinical trials investigating the clinical validity of ROS1 as a predictive biomarker in breast cancer.