D. Ross Camidge, MD, PhD, discusses the use of immunotherapy in patients with non–small cell lung cancer.
D. Ross Camidge, MD, PhD
The use of single-agent immunotherapy and chemoimmunotherapy in patients with both nonsquamous and squamous non—small cell lung cancer (NSCLC) has been addressed in several clinical trials. However, the decision of whether to use a monotherapy or combination approach requires further investigation, said D. Ross Camidge, MD, PhD.
“We are entering a renaissance in terms of immunotherapy for NSCLC,” said Camidge. “This is not necessarily just because we have these drugs, but we have a better understanding that they don’t work in everybody and how we should use them.”
Data from the KEYNOTE-024 trial of frontline pembrolizumab (Keytruda), for example, indicated a 37% reduction in the risk of death in patients with metastatic NSCLC (HR, 0.63; 95% CI, 0.47-0.86; P = .002). Median overall survival (OS) was 30.2 months with pembrolizumab versus 14.2 months with chemotherapy.1
In KEYNOTE-189, the frontline combination of pembrolizumab and chemotherapy was found to reduce the risk of death by more than 50% in patients with nonsquamous NSCLC. The estimated 12-month OS rate at a median follow-up of 10.5 months was 69.2% (95% CI, 64.1-73.8) and 49.4% (95% CI, 42.1-56.2) in the pembrolizumab/chemotherapy arm and control arm, respectively (HR, 0.49; 95% CI, 0.38-0.64; P <.001).2,3
Following the results of KEYNOTE-024 and KEYNOTE-189, Camidge said that combinatorial biomarkers and corresponding enrichment strategies will better define patients who will benefit most from immunotherapeutic approaches, as not everyone should receive PD-1/PD-L1 inhibitors.
In an interview during the 2018 OncLive® State of the Science Summit™ on Advanced Non—Small Cell Lung Cancer, Camidge, professor, Division of Medical Oncology, Joyce Zeff Chair in Lung Cancer Research, School of Medicine, Division of Medical Oncology, University of Colorado, discussed the use of immunotherapy in patients with NSCLC.Camidge: We’re entering a renaissance in terms of immunotherapy for NSCLC. This is not necessarily just because we have these drugs, but we have a better understanding that they don’t work in everybody and [of] how we should use them. The whole “battleground” has moved to the first-line setting. We have the idea that by enriching for PD-L1, we can give pembrolizumab monotherapy. That was the KEYNOTE-024 study. In that study, enrichment meant a tumor proportion score (TPS) of 50% or greater, which is about 30% of lung cancer across all histologies.
Then we had the KEYNOTE-189 trial, which shook up [the landscape] a little bit. That trial compared carboplatin/ pemetrexed with or without pembrolizumab in the nonsquamous population. You could see that, regardless of PD-L1 level, patients seem to get benefit—though not [as much as they do] from the addition of pembrolizumab.
Now, everybody with nonsquamous [NSCLC] has a chemoimmunotherapy option. If you have a high TPS, you have both a chemoimmunotherapy option and an immunotherapy monotherapy option. One of the research questions is in that high TPS group. Which is the right one? Do [those patients] need chemotherapy? That will be answered in a study called INSIGNIA. That will be running over the next few years.
The other burning research questions are in KEYNOTE-024, in which they had enriched up to 30% of the population to say [that these are the patients] you can give pembrolizumab monotherapy to. Was that bar too high, or could there be benefit given we know that occasionally you can get responders at lower TPS? Is there a larger group that can receive an immunotherapy option alone? KEYNOTE-024 allowed people with 1% TPS or greater.
The headline was that the OS hazard ratio was better, but that is misleading. When you show greater than or equal to 1% TPS, you’re still including those people with high TPS. When you take those out and look at just the patients with 1% to 49% TPS, the OS hazard ratio was not statistically significant. We’re seeing that most of the benefit was driven by the patients with high TPS. For me, if a patient doesn’t have a high TPS, pembrolizumab monotherapy shouldn’t be an option unless they’re unfit for anything else. In that case, patients should probably have chemoimmunotherapy.
The other area that we’re starting to see is whether patients with squamous NSCLC can join the party. We already know that if you have a high TPS and squamous NSCLC, you have pembrolizumab as a monotherapy option. What about chemoimmunotherapy for squamous NSCLC? We saw 2 studies at the 2018 ASCO Annual Meeting: one with atezolizumab (Tecentriq) and one with pembrolizumab added to chemotherapy. Both suggested benefit in terms of progression-free survival (PFS) and OS. A lot of people are expecting that we’re going to have a chemoimmunotherapy option for squamous [NSCLC] fairly soon.Biomarkers for immunotherapy is a work in progress. In the first-line setting, when you’re combining with chemotherapy, you could say that it’s a done deal. However, there are certainly people who are not deriving benefit.
The assumptions that we make for immunotherapy are so much more exaggerated than for targeted therapy. If I find the mutation, I can be very confident that it is going to turn into an abnormal protein and drive the cell and be sensitive to the tyrosine kinase inhibitor.
For immunotherapy, the assumptions that I make are much higher risk. I’m assuming if I have a high tumor mutation burden (TMB) that that means I’ve got more neoantigens. I’m assuming that those are going to be presented on that patient’s major histocompatibility complex (MHC).
I’m assuming that they will be presented in a permissive immune environment and that the relevant checkpoints I’m going to use are involved. Each one of those assumptions is high risk. Therefore, the predictive benefit of any of these biomarkers is much less than an EGFR mutation or ALK rearrangement.
How do we get around that? We have to combine biomarkers at multiple steps in that process to reduce the assumptive risks that we’re making. TMB doesn’t guarantee benefit. It’s just enriching [for benefit] at the top end of the immune cascade. PD-L1 expression is at the bottom end, at least for PD-1 monotherapy.
Then we’re going to get biomarkers along the way. Maybe we’ll get better bioinformatics. Are the mutations changing the immune-acid sequence? Are they going to be presented on that patient’s MHC? Are all mutations equally immunogenic? We know that insertions are very immunogenic and point mutations less so. Do we have other biomarkers for what’s going on in that tumor microenvironment?
Multiple biomarkers will be a way of enriching for benefit. How will we use them? I don’t know if we’ll ever get to something that says you have a 100% chance of responding to immunotherapy. Perhaps we will get to some markers that say you have a 0% chance or close to that.
Maybe we’ll use these as a means of enriching for the population who will benefit from any kind of immunotherapy. We’re going to see them in combination studies going forward [and in] the immune-oncology—experienced population [for whom] we don’t know what to do next. That is not to say that we’ve identified the group it works in, but we have gotten rid of part of the group that is less likely to benefit.There are a whole bunch out there. My worry is that the whole bunch of immune combination studies are not going to give us a definitive result because they’re not enriching for the right population; they’re just randomly guessing. In terms of immunotherapy research opportunities, the INSIGNIA trial is a very interesting study. Asking questions about the duration of immunotherapy in people who are benefiting is an interesting question.
In the targeted therapy setting, there are a bunch of studies that are going to be interesting. The first-line ALK inhibitors are going to be interesting to read out. We have alectinib (Alecensa). That’s kind of grabbed hold of the first-line setting, but we have a number of other drugs that have better outcomes in that setting after crizotinib. Will they transition that ranking to the second-line setting? Will brigatinib (Alunbrig) turn out to be a better drug? Will lorlatinib turn out to be a better drug? The good thing is these drugs are working so well. It’s taking years for the data to mature, which is great for the patients and frustrating for the scientists. We’re just going to have to be patient to get that result.