Benjamin Levy, MD: Similar to the ALK story there’s a lot of traction and movement in the ROS-rearranged lung cancers. Bob, do you want to talk about how you currently treat ROS-rearranged lung cancer? And then some of the new data that have come out including the repotrectinib data that we saw, I believe on Friday.

Robert Doebele, MD, PhD: For ROS1-rearranged lung cancer, this is a smaller population than ALK, it’s 1% to 2%. The only FDA approved option right now is crizotinib. I’ve been fortunate to have the entrectinib trials at my institution, so all of our frontline patients with ROS1-rearranged lung cancer have been going on entrectinib, and now we’ve presented data, and the study has actually completed accrual, and we’ll hear from the FDA soon.

The big difference between entrectinib and crizotinib is the CNS [central nervous system] penetration. We’ve seen really good activity in the ROS1-positive patients with brain metastasis; a 55% response rate in patients with brain metastasis. It’s a little bit hard to compare the overall PFS [progression-free survival] in the population because there are so many different crizotinib trials. We’re talking about the heterogeneity in trials. There’s an original phase I Pfizer trial upon which it was FDA approved for crizotinib, but there are multiple other studies with a range of PFS anywhere from 9 to 19 months.

One of the things that we just presented yesterday was a real-world trial, a virtual trial of entrectinib from the STARTRK trials and other trials, versus crizotinib-treated patients from the Flatiron Health database. And there we saw a doubling of the time to treatment discontinuation and a significant overall survival advantage. And the data looked like what we would expect in terms of why patients with crizotinib were failing therapy.

There was a steep drop-off at about 8 months for crizotinib, and when we looked into that data, it was all CNS progression. I think it kind of mirrors what we think is improved with entrectinib, and that mirrors what we’ve seen with alectinib, brigatinib, and osimertinib in terms of having better upfront brain control, whether you have brain metastases at the beginning or whether you’re just at risk of developing them later on. And that would really be potentially a new frontline option for our patients.

In terms of what we do after these drugs, crizotinib or presumably entrectinib; repotrectinib, presented data this week, and I think it’s a very exciting drug. It has CNS penetration, 1. Two, they demonstrated, at least preliminary activity against the G2032R mutation in ROS1, which is one of the common mutations that we see upon progression. I don’t think we know how common it is yet. My guess is it’s probably going to be about 25% to 30% of patients with ROS1 G2032R, but hopefully it will work in other patients as well. I think this may represent a future option because we really need to be able to sequence our ROS1 patients the same that we do in ALK.

Benjamin Levy, MD: It is a nice summary. Currently crizotinib is still your treatment of choice for patients who are ROS1, off of a trial?

Robert Doebele, MD, PhD: Until this summer hopefully. I think, again, the CNS penetrant thing is so important, I think it’s critical. And my crizotinib-treated patients who couldn’t go on the trial for one reason or another, fall off at about a year, just as you would expect for CNS progression.

Benjamin Levy, MD: Are there data on entrectinib in crizotinib-refractory patients, I think you mentioned?

Robert Doebele, MD, PhD: No. They explored that early on and they really weren’t seeing a lot of activity, except in the situation where there was CNS-only progression, although they really didn’t pursue that either. In terms of managing resistance, it’s probably not very good there, so it really needs to be used as a frontline option to control the brain metastases.

Benjamin Levy, MD: And any comment on lorlatinib? There was a cohort of ROS-rearranged in that study, correct?

Robert Doebele, MD, PhD: Yes. It was a small cohort. There were about 13 naïve patients I think with great data but very small numbers. Again, it’s very good CNS penetration. And there were some data in the ROS1 resistance, but the response rate wasn’t all that fantastic, and it’s unclear where that drug is going; although it is available as an FDA-approved agent if you can get it off label.

D. Ross Camidge, MD: On that point, the lorlatinib in the post-crizotinib ROS1 cohort has got a 30%-ish response rate. But it’s got a 56% response rate intracranially. And what has not been shown is what contributions that intracranial response rate had to that 30% response rate. Because if you pull some of those out, if they’re just CNS lesions responding, that number is even lower, and it would be great to see that.

Benjamin Levy, MD: Yes, interesting. We’ve got a lot, we’ve got to be a little patient, but I think we’ll have some.

D. Ross Camidge, MD: We have asked.

Benjamin Levy, MD: Incredible movement in that genotype as well.

Transcript Edited for Clarity