Impact of Coalterations on KRAS-Mutant NSCLC

Benjamin P. Levy, MD: We’ve learned more recently about coalterations with KRAS and how they may impact how patients do specifically with chemotherapy or I/O [immuno-oncology]. But we learned at the AACR [American Association for Cancer Research Annual] Meeting, at least on a virtual presentation this year, that it impacts single-agent I/O. How do we use other alterations to make treatment decisions? I’ll start by saying that we’ve learned that potentially, if you have a coalteration with an STK11 mutation or a KEAP1 mutation with KRAS, those patients may not benefit by adding immunotherapy to chemotherapy. We’re still learning. Maybe it’s different if you have a TP53 mutation with KRAS. Do those patients do better? What are your thoughts on this? Do you use this routinely to make decisions?

David R. Gandara, MD: I don’t think it’s something I would use in standard-of-care fashion. But oncologists are used to using all the data they have in front of them: the clinical data, the biochemical data, and chemistries plus the renal function. If you have data in front of you that might be helpful, very often there’s not a right and a wrong answer. There are several options. I do take them into account, but it’s harder to deselect. “I’m not going to offer Mrs Jones this treatment, because she has an STK11 mutation”—I don’t think we’re there yet. In fact, some of the LOCs [leaders in oncology care] of large trials have been contradictory, saying that KEAP1 or STK11 was not predictive but prognostic.

Another gene of emerging importance here is ARID1A, the chromatin remodeling gene, and we’ll be presenting quite a large series at the ASCO [American Society of Clinical Oncology Annual] Meeting this year on the impact of that biomarker, which was reported to be predictive of outcome where STK11 and KEAP1 were not.

I have to make a confession, Ben. I got my time zones mixed up in my calendar, and when I tuned in to the virtual AACR meeting yesterday, I was 3 hours off and the lung session was ending. Tell me what you heard at AACR that I should be interested in and that I may have seen on Twitter.

Benjamin P. Levy, MD: That’s why we’re doing this, so we can catch each other up. I haven’t read the full abstract yet. I saw the Twitter feed looking at KEYNOTE-024 and a small subset of patients, specifically those who had STK11 mutation, to see if that blunted the effect of immunotherapy over chemotherapy for this patient population with greater than 50% PD-L1 expression. It didn’t. The results seemed to hold up. I’m sorry if I’m butchering these data. I’ll look at them again, but that’s my understanding based on a very small subset. STK11 mutation didn’t seem to knock out the effect for patients who were PD-L1 greater than 50% treated with pembrolizumab versus chemotherapy. I think we just need more data to better understand the relevance of these coalterations. You’ve mentioned others that are emerging, which speaks to the importance of doing comprehensive genomic profiling so that you can get the full story and not just chapter 1 of a tumor.

David R. Gandara, MD: I agree. Nothing is ever as simple as it seems at the beginning.

Transcript edited for clarity.