ASCO Data Establish Chemo-IO as Mainstay Frontline for NSCLC

H. Jack West, MD; Joshua Bauml, MD; Sandip Patel, MD


June 21, 2018

H. Jack West, MD: Hello. I am Jack West, medical director of the thoracic oncology program at Swedish Cancer Institute in Seattle, Washington. Welcome to Medscape Oncology Insights. Joining me today is Dr Josh Bauml, assistant professor of medicine at the University of Pennsylvania in Philadelphia, and Dr Sandip Patel, assistant professor of medicine at University of California San Diego.

At this year's American Society of Clinical Oncology (ASCO) Annual Meeting, we again saw a strong showing for immunotherapy and a lot of work in the field of lung cancer. In fact, for the first time in many years, we had a plenary presentation on immunotherapy for lung cancer. This session reported KEYNOTE-042,[1] which was presented by Dr Gilberto Lopez from the University of Miami in Florida. KEYNOTE-042 was a global study of first-line therapy for patients who did not have an epidermal growth factor receptor (EGFR) mutation or anaplastic large-cell lymphoma kinase (ALK) rearrangement, had either squamous or nonsquamous cell histology, and a PD-L1 expression level in their tumor of 1% or higher—about two thirds of patients.

We knew the trial was positive for an overall survival benefit, but I would say that overall the results were a bit disappointing. Although there was this improvement in survival, Dr Leena Gandhi, the discussant for the presentation at the plenary session, highlighted that no crossover was allowed and only about 20% of the patients on the chemo arm had the opportunity to receive any form of immunotherapy afterwards, which we would consider to be a standard of care.

At the same time, the results presented separately for patients with a PD-L1 greater than 50% showed a marked difference in that group favoring pembrolizumab, [a treatment approach] which we were already doing.

In contrast, when you look at [survival in those with] the lower PD-L1 of 1%-49%, it is not worse, but it is not clearly better. There is better tolerability, as we would expect.

Where does this leave you in thinking about how this changes practice, especially when we also have seen the results of the KEYNOTE-407 trial,[2] which used a different design looking at squamous cell histology, and demonstrating results that look a lot like the previously reported KEYNOTE-189 trial?[3] Both of these studies looked at chemoimmunotherapy versus chemo.

In KEYNOTE-189, platinum and pemetrexed with pembrolizumab looked better than chemo alone in patients with nonsquamous non–small cell lung cancer (NSCLC). Then KEYNOTE-407 showed a strong result for chemo (carboplatin/paclitaxel or nab-paclitaxel) plus pembrolizumab compared with the same chemo. Where does this leave us in terms of thinking about how to prioritize chemo alone or pembrolizumab monotherapy versus chemo-pembrolizumab?

Joshua Bauml, MD: We have to view KEYNOTE-042 within the lens of our knowledge that 189 and 407 exist. In 189 and 407, we saw an improvement in overall survival with chemoimmunotherapy regardless of PD-L1 status, and the hazard ratio for those was very, very impressive. There was a marked improvement in overall survival, essentially showing that for both squamous and for nonsquamous NSCLC, chemoimmunotherapy was better than chemotherapy.

In KEYNOTE-042, we saw that, overall, there was an improvement [with pembrolizumab], but that was entirely driven by those patients with greater than 50% PD-L1. For those patients who were 1%-49% PD-L1, toxicity was improved with pembrolizumab, but overall survival was no better. But we have another trial that says, "We have something better."

For me, for any patient who has greater than 50% PD-L1 staining, I plan to use pembrolizumab monotherapy. For patients who have less than 50% PD-L1 staining and can tolerate it, I plan on using chemoimmunotherapy even if PD-L1 staining is less than 1% or otherwise. These regimens are relatively well tolerated. The majority of my patients can tolerate platinum-pemetrexed and pembrolizumab. The tolerability was reasonable in this trial, and similarly with the taxane and the [squamous histology]. I think it will be a very small group of patients with less than 50% PD-L1 to whom I would give pembrolizumab monotherapy.

Sandip Patel, MD: I absolutely agree with Josh. For me, KEYNOTE-042 is practice-reinforcing to the current practice paradigm, and KEYNOTE-407 is much more likely to be practice-changing. The take-away I get from the majority of the front-line NSCLC studies I have seen is that chemotherapy seems to be the best partner for PD-1 in front-line NSCLC. This is a bit different from in renal cell carcinoma with VEGF inhibition, or melanoma and intermediate- and poor-risk renal cell carcinoma with CTLA-4 inhibition with PD-1.

I think the point is very well taken that we haven't quite sorted out the biomarker story, but that from KEYNOTE-042 [the survival benefit] was driven by those patients with PD-L1 immunohistochemistry greater than 50%. It reinforces that the patient with PD-L1 greater than 50% is the patient who may benefit from pembrolizumab; or for those in visceral crisis, because the time to response is faster, chemo-pembrolizumab perhaps in that unique scenario. But for the vast majority of patients, chemo/immuno-oncology (IO), I think, will be the mainstay in front-line treatment of NSCLC for the foreseeable future.

West: We should clarify that both of the chemoimmunotherapy-versus-chemo trials, KEYNOTE-189 and KEYNOTE-407, involved the broadest population of patients with no driver mutation, but any degree of PD-L1—including less than 1%. If there is one [potential practice-changing finding], it is that pembrolizumab monotherapy wasn't worse; the efficacy was acceptable.

I agree that most of our patients who have a performance status in the fit range should be able to tolerate chemo, including a carboplatin doublet. But for those patients who are highly disinclined to do it, you are at least now permitted to think about that more liberally.

Bauml: But this is where we have to consider the crossover. Crossover makes the pembrolizumab seem better than it would otherwise seem. Pembrolizumab was not worse in the setting of no crossover allowed. Could it be worse? It could be worse, actually. That is something we have to consider.

Patel: In the context of the CheckMate 026 study[4]—that was the front-line negative study of nivolumab versus chemotherapy—the curves look awfully similar to those of the 1%-49% PD-L1-positive patients in KEYNOTE-042. In fact, there is more crossover in the nivolumab study than there was in the pembrolizumab study.

Moreover, if you look at the initial part of the curves, you see that the patients on chemotherapy do better for the first 6 months or so. You really are selecting for particular patients. I believe that is what we're getting from these chemo/IO strategies.

The other point that is particularly interesting to me about KEYNOTE-407 is that even in the PD-L1 zero patient population, the hazard ratio is about the same (0.6) as in the PD-L1-positive patient. These chemo/IO combos seem to be of benefit to these PD-L1-negative patients.

One point that was made in KEYNOTE-189 was the unique toxicity-potential signal of nephritis. The signal in KEYNOTE-407 may have been related to pneumonitis with the taxane-based combination.

It's a little atypical and may just be related to the reporting of pneumonitis, where we're all more aware of it and it may mimic the colitis reporting in the days of melanoma. I believe we will learn whether this is a real signal or not in the real-world utilization of these agents. We really don't expect to see pneumonitis with taxanes as monotherapy, as opposed to renal dysfunction, which you can see rarely with pemetrexed.

West: We also heard some information about other options for chemoimmunotherapy, including the IMpower150 trial[5] in the nonsquamous population of patients, with carbo/paclitaxel with bevacizumab in the control arm versus adding atezolizumab to that. That shows a benefit for progression-free, as well as overall, survival.

We've also got some results coming in about IMpower131[6] for the squamous population, getting carboplatin and a taxane—specifically here nab-paclitaxel being the focus of benefit in terms of progression-free survival, if not overall survival, for the atezolizumab arm.

Where do these fit in? I would say the comparator today is not the platinum doublet alone. [The comparator] is: Is this better than what I would otherwise do, KEYNOTE-189 for a nonsquamous NSCLC and KEYNOTE-407 for a squamous NSCLC? Do you see it differently? Are we going to just have a cat fight—[that is], whatever you feel like doing—or do you see a consolidation with most people getting one regimen with pembrolizumab or not, and the leading competitor being pembrolizumab monotherapy for the folks with the higher PD-L1 percentages?

Patel: In IMpower150 at least, the cohort that was particularly provocative was the EGFR-ALK cohort. That is unique compared with the other studies. There are multiple studies ongoing to focus on this patient population that received carboplatin—in that study, paclitaxel, bevacizumab, and atezolizumab, with benefit and a hazard ratio on the order of 0.6, which recapitulated the overall population. I believe that there may be some biology there that we will hear more and more about.

Otherwise, the quadruplet is even more expensive than the triplet from a financial standpoint. If we are not adding durability to survival for these patients, we are adding toxicity, both financial and medical. I believe it will be a niche population.

The IMpower131 study, which we'll hear more about, is a larger study than KEYNOTE-407. There has not been a reported overall survival benefit compared with the control population. I think there, as we get more patients on [trial], whether we see a reversion to the mean—that's a question that remains.

Bauml: One thing that should be emphasized when we're looking at this EGFR mutant and ALK translocation subgroup is that tyrosine kinase inhibition (TKI) failed in these patients. It is not appropriate to be giving PD-1 inhibition prior to TKI to patients who have EGFR mutations or ALK translocations. In my opinion, one of the most important studies presented here at ASCO was the study by Dr Lisberg and Dr Garon and colleagues,[7] where they looked at patients with an EGFR mutation, gave them pembrolizumab monotherapy, and saw no responses but significant toxicity. That was an important question that had to be answered.

West: Those were even in patients who had high PD-L1.

Bauml: That's right. The issue was that the toxicity occurred after they stopped the pembrolizumab. There was a grade 5 pneumonitis from TKI. That is not what we would expect in any reasonable frequency with TKI alone.

West: We have seen a lot of treatment evolution at this meeting. I am sure we will have a lot more to talk about in the upcoming months, with new presentations and press releases every few weeks. It has been an exciting meeting with a lot to change how we approach treatment, starting now.

Thanks, Josh and Sandip, and our online audience for joining me today. This is Jack West for Medscape Oncology, looking forward to our next discussion.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: