<p>Although ROS1 alterations are present in only a small subset of patients with non–small cell lung cancer (NSCLC), this molecular subtype can present unique challenges, such as the risk of peridiagnostic blood clots and the development of acquired resistance to ROS1 inhibitors through on-target mutations and secondary driver mutations.</p>

ROS1 alterations are rare, occurring in approximately 2% of patients with NSCLC, and are enriched in those with adenocarcinoma, in those who are never smokers, and in somewhat younger age groups. ROS1-altered NSCLC stands out as being associated with a high thrombotic potential during the peridiagnostic period. Within 12 months before and after diagnosis, a study of people with ROS1-altered NSCLC found that 34.7% developed a venous or arterial clot. Those with ALK alterations had a fairly high clot rate too, at 22.3%, and the rate was 13.7% and 18.4% for those with EGFR and KRAS alterations, respectively. No one knows the reason for the association between ROS1 and thrombotic potential, but it could be related to downstream signaling from ROS1.

<br>

The mechanisms of acquired resistance to ROS1 TKIs fall into 2 broad categories. The first is an on-target resistance, which is an additional ROS1 mutation in the kinase domain that can alter the binding kinetics of the drug. The second is an off-target resistance, classically through driver alterations in another non-ROS1 pathway. There is nothing licensed in this setting, but many academic centers will combine TKIs targeting the other pathway with an ROS1 inhibitor off-label if they identify an actionable second driver.

<br>

In some patients with NSCLC, you do not initially know the mechanism of acquired resistance. In an ideal world, in these cases, you would rebiopsy, reanalyze, find the mechanism of resistance, and then make your next choice. This might be a combination of TKIs or a next-generation ROS1 inhibitor that might work on that mutation. The alternative is using ROS1 inhibitors such as repotrectinib and taletrectinib, which are licensed in the second line. Some people give their patients these drugs and see a 30% to 40% response rate, which tells you the frequency of the on-target resistance on which these drugs work. Although 40% might sound great in the second-line setting, you could potentially increase the response rate even more if you know what the mechanism of resistance is and use the right drug straight away.

<br>

If a patient with NSCLC develops resistance while receiving crizotinib or entrectinib in the first line, there are a lot of potential causes. Progression in the brain may be due to low drug penetration. Progression in the body is more obviously from a change in the biology of the cancer. The spectrum of potential on-target ROS1 resistance mutations depends on which drug the patient was on. Even if new ROS1 drugs are better at suppressing known on-target resistance mechanisms, we may still have a few on-target mechanisms that the drugs cannot suppress or secondary driver alterations leading to resistance. Using the best possible drugs early on in treatment might mean that we may have to deal with new mechanisms of resistance that we have not seen before, but that is what research is for, and this does not mean that we should shy away from doing so.