Second-Line after I-O: Where Do We Go from There?

Panelists: David R. Gandara, MD, UC Davis Comprehensive Cancer Center; Everett E. Vokes, MD, University of Chicago Medicine


Everett E. Vokes, MD:
So, David, now that we use so much of our ammunition up-front with either single-agent immunotherapy or combination chemotherapy and immunotherapy, eventually unfortunately the majority of patients will still progress. And so now we’ve used our best guns, so what do we do next?

David R. Gandara, MD: Right. That’s a great question. I was giving a presentation in another country recently where there is no immunotherapy yet approved for first-line [therapy]. And so, they wanted me to talk about immunotherapy as second-line [therapy], and I had to think to myself, “Who in the United States is not going to get immunotherapy [in] first-line [therapy]?” So that was the first thing I thought of. And I said, “Well, if it’s a patient who has an underlying autoimmune disease like, let’s say, rheumatoid arthritis or lupus, they wouldn’t get it first-line [therapy]. If it’s a patient maybe who has a bad performance status, since the trials have largely been performance status 0 to 1, they wouldn’t get it first-line [therapy].”

And then I thought about what about if we go back to this issue about PD-L1 [programed cell death ligand 1]and TMB. A group of patients who had “low-low”—low PD-L1, low TMB [tumor mutational burden]—they might not get it. But they might be considered for its checkout line based on the data that we have. And then it could be some other factor such as somebody with interstitial pulmonary fibrosis or some other problem where you thought the risk of complications.

But otherwise in the United States at least, you could rationalize giving immunotherapy to almost every category of patients. Now that being said, assuming somebody is a candidate [for] second-line [therapy], we have 4 positive trials. The CheckMate trials—017 and 057—being in squamous and nonsquamous cell lung cancer, and then we have the OAK trial for atezolizumab [Tecentriq] and we have the KEYNOTE trial for pembrolizumab [Keytruda], and they were all designed a little differently.

The CheckMate trials were biomarker agnostic. And what I mean by that is that although they collected tissue, they did not base the design of the trial on having a certain PD-L1 level. Where the KEYNOTE trial and the OAK trial did, but they included patients with PD-L1 negative. All 4 of these trials were positive for the primary endpoint of overall survival, but they were positive in different ways. And I come back to the issue about squamous versus nonsquamous because the CheckMate trials, the 017, there was improved survival in squamous cell lung cancer of nivolumab over docetaxel regardless of PD-L1 level. My impression is that’s because it was squamous cell lung cancer.

In the CheckMate057 trial in nonsquamous cell lung cancer, there was benefit across PD-L1 status, but the greatest benefit was in patients with very high PD-L1. In this case they used 10, a score of 10. So these were retrospective analyses which means from a regulatory standpoint it didn’t get in the label. Wherein the other trials, of course pembrolizumab was approved at greater than 1% PD-L1.

Atezolizumab was approved regardless of PD-L1 levels, although there again, the greatest benefit was in the patients that had the highest levels of PD-L1 expression using that assay that accounts for both tumor and immune infiltrating cells. So from a practicing oncology standpoint you have lots of options.

Everett E. Vokes, MD: The PD-L1 was always more of an enrichment factor than an absolute cutoff, but then for regulatory purposes it is a cutoff. But patients will eventually fail immunotherapy—first-line or second-line [therapy]. And at that point what do you advise your colleagues you know how to treat the patients?

David R. Gandara, MD: Well I think it’s a little bit the same situation that we have with EGFR TKIs [tyrosine kinase inhibitors], which is, if we give our best drug, or in this case maybe a combination up-front, what do you do afterwards? So in a patient that receives let’s say pembrolizumab monotherapy first-line, and they don’t have a driver mutation, it’s easy to say, well, second-line [therapy] I would go to platinum chemotherapy. If that same patient has received a platinum-based chemotherapy plus an immunotherapy, let’s say pembrolizumab or atezolizumab first-line [therapy], then you’re stuck saying, “What do I do after that?”

Everett E. Vokes, MD: Right.

David R. Gandara, MD: And here increasingly I’ve been using the combination of docetaxel and ramucirumab [Cyramza]. We have a positive phase III trial, the REVEL trial, which shows improved survival for both histologies. And up until now that’s kind of been a regimen that maybe has been third-line or something where you wouldn’t consider it, but if the more you do [in] first-line [therapy], then things move up. So I’ve actually done that in several recent patients and we’ve had some very good responses.

Everett E. Vokes, MD: Yeah, I agree. I think it is now the regimen of choice, certainly for the patients who failed the triplets up-front or the 4-drug regimen. And it is then what’s left prior to going to other investigational agents or combination immunotherapy, which is, of course investigational.

But you’re right. For the single-agent immune oncology agents to go [to] second-line[therapy] is the exception. We’ve even seen PS2 [performance score 2] data this morning that would be compatible with giving the drugs first.

Up until now I’ve been reluctant to do that. Again, not many studies have included PS2, but 2 of them, 1 from France and from the UK [United Kingdom] suggested that there was much increased toxicity in patients with PS2 and reduced efficacy.

Transcript Edited for Clarity
Printer Printing...