Pan-cancer analysis of ROS1 genomic aberrations

Abstract

The ROS proto-oncogene 1 (ROS1) encodes the ROS1 receptor kinase. ROS1 rearrangements are known to be oncogenic in glioblastoma, non–small-cell lung carcinoma (NSCLC) and cholangiocarcinoma. The clinical relevance of ROS1 genomic aberrations in other human cancers is largely unexamined. Here, we performed a pan-cancer analysis of ROS1 genomic aberrations across 20 cancer sites by interrogating the whole-exome sequencing data of the Cancer Genome Atlas (TCGA) via the cBioportal (www.cbioportal.org). We found that ROS1 point mutations occur at a noticeable average frequency of 4.09%, while ROS1 gene rearrangement is uncommon, except for lung cancer. As ROS1 is potentially druggable by small molecule kinase inhibitors, it remains to be determined if any of these ROS1 mutations are associated with potential clinical benefits in patients. In our study, ROS1 gene amplification was noted in a small subset of patient tumors. For reasons unclear, ROS1 gene copy loss was found to be common in some cancer types. We also found that ROS1 mutations extend over the entire gene with three emerging hotspot mutation sites (R245I, P1539S and E2308) and several other recurrent clusters that frequently occur at the interdomain 2, 3, 4 regions, kinase and C’-terminal kinase domains by domain mapping. Similar mutation pattern (except for the C’-terminal kinase domain) was found in ROS1-mutated cancer cell lines from the Cell Line Encyclopedia (CCLE) and the NCI-60 panel. Lastly, we summarized all published ROS1 mutants related to tumorigenesis and secondary mutations of drug resistance in order to facilitate our understanding of the potential contributions or impacts of these pan-cancer ROS1 mutation spectra in multiple cancer types.