Clinical Management of ROS1+ Advanced NSCLC

Benjamin Levy, MD: ROS1 gene rearrangements occur in roughly 1-2% of all lung adenocarcinomas. There’s been some data on the association of the clinical phenotype. Is there a patient that is more likely to harbor ROS1 rearrangements? We know that they're potentially more common in females, in patients without a smoking history, in patients as I said with adenocarcinoma and probably more common in patients with advanced stage disease, at least based on some MET analyses. That said I want to make a point here that ROS1 rearrangements, yes, they occur in only 1-2% of the time and maybe more common in female, never smoked, or adenocarcinoma patients, but every patient who walks through the door independent of their clinical phenotype, needs to be tested for these genes. Why? If we identify a ROS1 gene rearrangement we know that there are available therapies at the time of diagnosis. We also know there are available therapies coming down the pike that may be good at the time of resistance. For all these reasons we need to test. But generally, are seen in never smoked, more common in female, even though I’ve seen them in male, and certainly more common in adenocarcinoma with advanced stage disease.

Stephen Liu, MD: There are currently 2 FDA-approved agents for ROS1 fusion-positive non-small cell lung cancer, crizotinib, and entrectinib. Crizotinib was approved in 2016 based on the PROFILE 1001 trial. That study added a ROS1 cohort in 2009. You'll recall it, started off as a MET inhibitor, added a cohort for ALK, and then followed by ROS1. In that study, we saw a response rate of 72% progression-free survival [PFS] of 19 months. Very impressive data upfront leading to its approval. A lot of other series in other areas of the world showed the high response rates but a lot of variation in that progression-free survival. Some as low as 5 and 61/2 months, others as high as 20 months. That likely reflects differences in the populations enrolled. A big difference is involvement of the CNS [central nervous system] or brain metastases. We know when those are present, the progression-free survival is a bit worse and there might have been a lot of imbalances between these studies. Not really sure because in the early studies wasn't reported. We weren't regularly doing MRI [magnetic resonance imaging] at routine intervals. CNS involvement is important because a lot of reports say half of progression like crizotinib is due to CNS progression and our newer ROS1 inhibitor is entrectinib. This is designed to have better CNS penetration. It's highly effective in the brain. When we look at a large study now, in modern times, we see the response rate of 67%. We see a PFS [progression free survival] very robust at about 16 months. This was approved in August 2019. Importantly, if we look at patients with measurable CNS metastases, the intracranial response rate of entrectinib is 79%. Overall, if you put the amount in table, the numbers look similar but, in my opinion, crizotinib at an era when we were only detecting ROS1 in very select patients that survived long enough to get that therapy. Today, we're detecting ROS1 upfront at the time of diagnosis and that includes ROS1 with maybe a slightly different biology, maybe a more aggressive subset. And to see comparable PFS and response rate numbers, that gives us a slight edge. Based on that, based on the CNS efficacy, my preference now is probably entrectinib, both drugs approved, both reasonable options in the front line setting for ROS1 lung cancer.

Benjamin Levy, MD: Importantly, a first drug approved was crizotinib and this data came out years ago looking at TKI-[tyrosine kinase inhibitor] naïve patients that were ROS1 rearranged with advanced stage disease. The drug showed a response rate of 72%. The median progression-free survival of roughly 19 months, and incredibly an updated analysis showing a media overall survival of 51.4 months. We don’t see these types of overall survivals if we’re 10, 15 years ago. Again it is important to identify the gene rearrangement and wed that to a targeted therapy. The data I shared with you with crizotinib was 53 patients. I only say that because we have data on entrectinib for the exact number of patients, 53. And this data came from the STARTRK trials. It was a pooled analysis of some phase 1,2 trials, 53 patients with locally advanced or metastatic non-small cell lung cancer that were positive for ROS1 rearrangement. Response rate a little bit higher of 77% compared to the crizotinib. Identical median progression-free survival of 19 months. The median overall survival hasn’t been reported yet. It’s a little bit too early I think and hopefully we will have that data in the next year or so. I think the differentiator between entrectinib and crizotinib is entrectinib was designed to cross the blood-brain barrier and remain within the CNS. It’s also a more potent ROS1 inhibitor than crizotinib at least in pre-clinical studies. We have data from that pooled STARTRK analysis trials that shows that this drug has meaningful activity in the brain. And we don’t have that data from the crizotinib literature. We know that the intracranial response rate from the patients with brain metastases from this pooled analysis was north of 50%. This is the differentiator here that there is proclivity across the blood-brain barrier and get in and would remain in the CNS. And that’s the differentiator here. In summary, we’ve got two drug that are FDA-approved both eliciting response rates north of 70%, PFS both of 19 months. We have a median overall survival north of 15 months with crizotinib but we do have this intracranial response rate reported from the STARTRK analysis, pooled analysis of north of 50%, 55%. And because of that if patients do have ROS-rearranged lung cancer with brain metastases I would potentially favor giving them entrectinib and you could make the argument even if they don’t – now we don’t have the data for this – that entrectinib may provide a CNS protective effect. But both are approved. I’ve used both. And both elicit meaningful activity for patients with ROS-rearranged lung cancer.

Mark A. Socinski, MD: All of non-small cell lung cancer, particularly the driver mutations in drivers like EGFR, ALK, ROS1, and others, tend to have a higher propensity for brain metastases. An MRI [magnetic resonance imaging] of the brain, in my opinion, is the standard part of the staging workup that we do for these patients whether or not you had a driver or not. Because brain mets [metastases] do change the therapeutic approach for potential therapeutic approach, depending upon the molecular characteristics of the tumor. We've mentioned before, from the BFAST trial, we saw about 40% asymptomatic brain mets in the ALK population. It's similar in the ROS1 population. That means that most people don't have brain mets, but one should be cognizant of the fact that they are at risk of developing brain mets over time. And if patients are diagnosed without brain mets at the time of their initial diagnosis, are treated appropriately, then the brain needs to have a surveillance strategy to make sure that you detect early brain mets and keep them asymptomatic with – however you may choose to do that.

Specifically in ROS1, we've talked about the much less activity of crizotinib in the CNS in the ALK-positive populations, the same thing is true in the ROS1 population. Entrectinib, another approved drug in the space, has good CNS [central nervous system] penetration and good CNS activity, so that would be my preferred agent. I've got another point to make about this. As mentioned, crizotinib has less CNS activity. That was our standard agent for ROS1-positive advanced non-small cell lung cancer. But with the entrectinib approval, I've changed my standard of care. Entrectinib does have more CNS penetration, more CNS activity than crizotinib. It is preferred, in my opinion, for those patients who initially are diagnosed with brain metastases. And even though I don't know that we have robust data,what we've seen with a number of these active TKIs that have good CNS penetration is that they prevent progression in the brain to a greater degree than some of the less active CNS drugs. The ability to keep patients from having CNS progression for longer periods of time is important in this population.

Benjamin Levy, MD: This is important that both entrectinib and crizotinib are well-tolerated drugs for patients with ROS-rearranged lung cancer. We’re all familiar with crizotinib. It’s been leveraged initially for ALK-rearranged lung cancer. We’ve got experience with this. In terms of toxicities with crizotinib you're really looking at potentially some LFT abnormalities, some nausea, some potential diarrhea, maybe rash. Those are the things that we see from crizotinib but a generally well-tolerated drug. Entrectinib has a similar toxicity profile. You get the GI [gastrointestinal] toxicities. It has potentially a more peripheral edema and swelling and that’s something that we need to follow. Both drugs can cause fatigue, although it’s tough to sort out if that’s drug-related or cancer-related. But both are well-tolerated. This is really proven by the fact that roughly eight to nine percent of patients in each of the trials, the crizotinib trial and the entrectinib trials had an AE-[adverse event] related discontinuation rate. 8% of the patients on crizotinib had to discontinue the drug due to AEs. And 9% of the patients receiving entrectinib had to discontinue due to AE-related events. That tells you right there these drugs are reasonably well-tolerated. 90% could stay on some dose of these drugs. And a heck of a lot better than chemotherapy. Again, important to test for these rearrangements so that you can wed them to the right targeted therapy.

Transcript edited for clarity.