ROS1 is a receptor tyrosine kinase (encoded by the gene ROS1) with structural similarity to the anaplastic lymphoma kinase (ALK) protein; it is encoded by the c-ros oncogene, and it was first identified in 1986.1,2,3,4 The exact role of the ROS1 protein in normal development, as well as its normal physiologic ligand, have not been defined.2 Nonetheless, as gene rearrangement events involving ROS1 have been described in lung and other cancers, and since such tumors have been found to be remarkably responsive to small molecule tyrosine kinase inhibitors, interest in identifying ROS1 rearrangements as a therapeutic target in cancer has been increasing.1,5 Recently, the small molecule tyrosine kinase inhibitor, crizotinib, was approved for the treatment of patients with metastatic NSCLC whose tumors are ROS1-positive.1
ROS1 Preclinical Findings
In 2007, Rikova and colleagues conducted a large-scale survey of tyrosine kinase activity in non-small cell lung cancer (NSCLC), and they identified more than 50 distinct tyrosine kinases and over 2500 downstream substrates, with the goal of identifying candidate oncogenes.6 In a sampling of 96 tissue samples from NSCLC patients, approximately 30% displayed high levels of phosphotyrosine expression; further analysis was conducted to identify highly phosphorylated tyrosine kinases in NSCLC from a panel of 41 NSCLC cell lines and 150 patient samples.6 Among the top 20 receptor tyrosine kinases identified in this analysis, 15 were identified in both cell lines and tumors, and among these were both ALK and ROS1.6 These initial findings paved the way for more expansive analyses of ROS1 kinase fusions in NSCLC and other cancers.
ROS1 Fusion Clinical Prevalence
In 2012, a panel of 1,073 patients with NSCLC was screened for ROS1 rearrangements; a total of 18 patients (1.7%) were positive for a ROS1 gene rearrangement, and these rearrangements were mutually exclusive of ALK rearrangement.5 Importantly, this study also described the clinical characteristics of patients with ROS1 fusion-positive NSCLC tumors in this large sample, which defined a new and clinically relevant genetic subtype of NSCLC; specifically, patients whose tumors tested positive for ROS1 fusions tended to be younger, with a median age of 49.8 years, had never smoked, and had a diagnosis of adenocarcinoma. In addition, although based on a smaller number of patients, there was a higher representation of Asian ethnicity (P<0.01) and patients with Stage IV disease.5 ROS1 rearrangements were estimated to be roughly half as common as ALK-rearranged NSCLCs; other clinical similarities with ALK-rearranged NSCLC were noted in this study, including the younger age of onset and a non-smoking history.5 The study was also the first to show a benefit of a small molecule ALK, ROS1, and cMET inhibitor, crizotinib, in this patient group.
In a study to identify gene rearrangements in a Chinese population of NSCLC patients (N=556), ROS1 expression was found in 1.6% of patients, and subsequent analysis showed that its expression was limited to those patients with ROS1 gene fusions.7 Similar findings were reported in a separate analysis of 447 NSCLC samples, of which 1.2% were found to be positive for ROS1 rearrangement. This study also confirmed the activity of the ALK/ROS1/cMET inhibitor crizotinib in ROS1-positive tumors.2 ROS1 fusions were also identified in 5 of 518 lung adenocarcinomas (1.6%), as well as 1 of 157 glioblastoma samples (0.6%) in a comprehensive assessment of kinase fusions across different cancers.8
One of the most commonly observed kinase fusions in ROS1 gene-rearranged NSCLC tumors is CD74–ROS1, and preclinical findings have demonstrated a mechanism whereby intracellular signals generated by this fusion kinase confer an invasive and metastatic phenotype.9 It was shown in this study that the expression of either the CD74–ROS1 or another ROS1 rearrangement, FIG–ROS1, was oncogenic, capable of causing tumor formation in immunocompromised mice.9 Accordingly, the inhibition of the ROS1 fusion kinases with small molecule inhibitors reversed the transformed phenotype of these cells, validating these ROS1 fusion kinases as therapeutic targets.9
ROS1 Fusions in Melanoma and Lung Cancer
An evaluation of 236 colorectal cancer specimens found ROS1 rearrangements in 2 cases (0.8%).10 In a survey of 140 Spitzoid neoplasms, including Spitz nevi (n=75), atypical Spitz tumors (n=32) and Spitzoid melanoma (n=33) ROS1 rearrangements were found in 17.1% of the neoplasms overall, all of which retained the intact kinase coding sequence for ROS1.11 The range of ROS1 fusions that have been identified across NSCLC and other cancers (Table 1) and its effective targeting with available small molecule inhibitors demonstrate that ROS1 fusion is a highly relevant molecular alteration to screen for across a range of cancer types, particularly in subsets of patients who display unique clinical characteristics (e.g., younger age, non-smokers with NSCLC) and those without evidence of other targetable mutations (Table 1).