Tests for detection of ROS1 gene alterations in people with non-small cell lung cancer (NSCLC): A Health Technology Assessment

Key messages: The Norwegian Institute of Public Health has been commissioned to evaluate molecular tests for the identification of somatic ROS1 gene alterations in people with locally advanced or metastatic non-small cell lung cancer (NSCLC). People with tumours harbouring ROS1 gene alterations probably make up 1-2% of NSCLC cases. Accurate and reliable detection of ROS1 gene alterations is important for identification of people who may benefit from treatment, as well as ROS1 negative patients, to avoid provision of unnecessary and costly treatment. We included one systematic review, six narrative reviews, a survey of Norwegian Hospital trusts, and two reviews on the preferences of patients related to molecular testing. Experts were contacted for cost information. The results of this HTA show that: • There is scarce, incomplete and low-quality evidence on the sensitivity and specificity of tests for the detection of ROS1 gene alterations in people with advanced or metastasised NSCLC • Positive IHC ROS1 results needs confirmation with FISH or other methods, due to a tendency for false positive staining. • While the different tests had different pros and cons, single gene testing may be unfeasible, since people with NSCLC typically are tested for more than one type of actionable gene alteration. • NGS due to its capacity to analyse multiple genes simultaneously, may have the potential to reduce the risk of repeat biopsies. • The cost for ROS1 using IHC as pre-test with FISH confirmation, is possibly less than for the other methods. • The cost associated with NGS testing will significantly decrease when parallel tests are to be performed for several biomarkers (i.e. gene panels) from multiple patients. However, at present, the capital and infrastructure as well as maintenance costs are higher for NGS than the other diagnostic methods. • Future research should focus on conducting larger cohort studies with welldefined patient populations, that follows the patients from testing (or no testing), through treatment and final outcomes.