Experts within the lung cancer community are eagerly anticipating the phase III findings of the KEYNOTE-189 trial, and they are hopeful that the combination of pembrolizumab (Keytruda) plus carboplatin and pemetrexed (Alimta) will demonstrate an overall survival advantage in patients with nonsquamous non–small cell lung cancer.
Hossein Borghaei, DO, MS
Experts within the lung cancer community are eagerly anticipating the phase III findings of the KEYNOTE-189 trial, and they are hopeful that the combination of pembrolizumab (Keytruda) plus carboplatin and pemetrexed (Alimta) will demonstrate an overall survival (OS) advantage in patients with nonsquamous nonsmall cell lung cancer (NSCLC).
In the phase III study, the PD-1 inhibitor pembrolizumab plus carboplatin/pemetrexed is being compared with carboplatin/pemetrexed alone in the frontline setting for patients with nonsquamous NSCLC (NCT02578680). It will serve as the confirmatory trial for the FDA’s accelerated approval of the regimen, which occurred in May 2017.
The decision of the immunotherapy/carboplatin and pemetrexed approval was based on the findings of cohort G in the phase II KEYNOTE-021 study, in which pembrolizumab combined with carboplatin/pemetrexed demonstrated a 57% objective response rate versus 32% with chemotherapy alone (P = .0029).
Following a median 18.7 months of follow-up, the median progression-free survival (PFS) in the pembrolizumab arm was 19.0 months (95% CI, 8.5-not reached) versus 8.9 months with carboplatin/pemetrexed alone (95% CI, 6.2-11.8). Additionally, the median OS was not reached in the pembrolizumab and carboplatin/pemetrexed arm (22.8-not reached) and was 20.9 months (14.9-not reached) in the carboplatin/pemetrexed arm. The 18-month OS rate with pembrolizumab plus carboplatin/pemetrexed was 70% compared with 56% for those treated with carboplatin/pemetrexed.
“It is a really exciting time. Although it appears to be confusing right now, it won’t be long before [we have the data of] multiple phase III studies with chemotherapy combinations with [immunotherapy] and we'll know exactly which direction we’re going to be able to go in squamous patients or nonsquamous patients,” said Hossein Borghaei, DO, MS. “A little more patience is required, but data are going to come out soon and we’ll know how to guide patients [through] all of the choices they have for treatment. What we want to see are better patient outcomes and it looks like we're definitely heading in that direction. It is welcome news for everybody.”
In an interview withTargeted Oncology, Borghaei, who is chief, Division of Thoracic Medical Oncology and associate professor, Department of Hematology/Oncology, at Fox Chase Cancer Center in Philadelphia, Pennsylvania, expanded on the importance of the pembrolizumab plus carboplatin/pemetrexed approval, the ongoing KEYNOTE-189 trial, and how the phase III findings could dramatically change clinical practice for patients with nonmutated, nonsquamous NSCLC.
TARGETED ONCOLOGY: What are your thoughts on the milestone FDA approval of pembrolizumab in combination with carboplatin and pemetrexed?
BORGHAEI: To start with, I have to point out that this is only a randomized phase II study with 60 patients per arm, so the current approval is based on limited experience. The phase III version of the protocol has completed accrual, but we are waiting for the results so we don’t have anything [yet] about the phase III [study]. It is important to keep in mind that this is only a phase II clinical trial. It is significant because, for the first time, we are showing that a combination of immunotherapy [IO] drugs with a standard platinum doublet can, in fact, improve PFS, response, and OS. It is [being explored] in an all-comer patient population, regardless of the level of PD-L1 expression. Up to now, IO, specifically with pembrolizumab only, has shown clinical efficacy in only patients with high PD-L1 expression at greater than 50% as a single agent compared head-to-head with chemotherapy. This study would [show] that, if you add carboplatin/pemetrexed to pembrolizumab, then you can apply it to an all-comers patient population. In an intent-to-treat patient population, it improves the clinical outcomes. It is significant because of [those reasons].
It is also a well-tolerated regimen based on all the data that have been presented, suggesting that the addition of carboplatin/pemetrexed to an IO agent does not seem to increase toxicity of either regimen. Granted, every time you add a third drug to a combination of 2 different chemotherapies, there are slightly more adverse events [AEs]. But overall, if you look at the treatment discontinuation rate and all the immune-related AEs, it just did not seem to be a significant increase compared with either therapy alone. Altogether, it appears to be a very good, effective regimen.
The difficulties here are that it is important to try to figure outto tease out—if patients with no PD-L1 expression or those with high expression are [having] that much benefit from the addition of carboplatin/pemetrexed. With the final analysis, when we have phase III data with larger patient sets, we can with confidence say that even in patients with greater than 50% [PD-L1 expression], they reap the benefit that much more from the addition of chemotherapy. Or, is it going to be that the patients with lower expression of PD-L1, who right now are not able to get pembrolizumab alone, when carboplatin/pemetrexed is added, can have a better response and survival? We don’t have any of that information, in terms of the subgroup analysis of patients with different levels of PD-L1 expression.
There is also this idea that perhaps tumor mutational burden can be a better predictor of response to IO. A couple of studies looked at that and we don’t have those data here either. I will be curious to see if there is going to be anything like that in terms of identifying patients who could really draw benefit from this » treatment based on tumor mutational load, though that is sort of a hypothesis.
That is why the study is significant, with the caveat that it is only a phase II study and we're awaiting the results of the phase III trial.
TARGETED ONCOLOGY: How could the phase III results of KEYNOTE-189 affect this approval and treatment for patients? What questions are we hoping to have answered?
BORGHAEI:The phase III study is going to have to show an OS advantage with the combination. That is how the approval will need to be sustained, honestly, but I don’t know what the regulatory agencies are going to say if there is only a PFS and response benefit and not an OS advantage. For a lot of clinicians and patients, the biggest question is, “How long am I going to get to live?” With any particular treatment…OS data are going to be significant. That is what all of us want to see in the phase III study and, again, what we would also like to see is the subgroup analysis based on the level of PD-L1 expression, tumor mutational burden, or both, to have both to guide for the management of our patients.
TARGETED ONCOLOGY: Will there continue to be a role for chemotherapy in these nonmutated patients?
Depending on the results of the study, chemotherapy could be an integral part of the care of patients who don’t have driver mutations. If the phase III study is positive in all-comers, then it would suggest that we should keep the chemotherapy for the frontline setting with a platinum doublet and add an IO agent to it. So, we will continue to use doublet chemotherapy.
But if the results are less than ideal, meaning they don’t show an OS advantage, then we are going to be in the same place we are nowmeaning we will offer single-agent pembrolizumab to anyone with greater than 50% PD-L1 expression and, for everybody else, we'll revert back to just chemotherapy for now until there are additional studies.
TARGETED ONCOLOGY: Can you describe the synergy between PD-1 inhibitors and chemotherapy?
BORGHAEI: In my view, the data are a little bit conflicting. Some data suggest that some chemotherapy drugs can be immune suppressive. Some data suggest that most chemotherapy drugs can be immune stimulatory. There is this idea of immunogenic cell death caused by chemotherapy, suggesting that even if you use chemotherapy then you cause cell death, which exposes potential neoantigens in the immune system that can be activated. With PD-1/PD-L1 inhibitors you activate T cells in the microenvironment that are already primed for chemotherapy, so you can have better responses.
I wouldn’t necessarily call it synergism or an additive effect. [With] all of that in mind, it is not quite clear yet. We have to look at cohort G of the KEYNOTE-021 study to say that, based on the response and PFS, and now survival data, there is definitely some sort of interaction. But what it is and how exactly this is happening is still a little bit unknown, and it requires more research [to determine] how chemotherapy is making immunotherapy work better. That is, again, as far as I’m concerned, immune suppression.
TARGETED ONCOLOGY: How can we continue to utilize PD-L1 expression going forward?
BORGHAEI: It all depends on what these [studies] show. If we have convincing evidence that PD-L1 expression is required…[and patients] get fantastic responses and survival [with the combination], then there is a likelihood that [we] won’t be testing anybody for PD-L1. However, if the data are really presented the way we want them tobased on the level of PD-L1 expression—and we look at the data and we say, “Those who have greater than 50% PD-L1 expression did not have a significantly better response with the addition of chemotherapy,” then we will continue to use PD-L1 to give only single-agent IO and then everybody [else] will get the combination. But this is hypothetical and we have to wait and see what the data show.
TARGETED ONCOLOGY: Currently, what patient characteristics are the deciding factors in whether patients will receive pembrolizumab plus chemotherapy or another regimen?
BORGHAEI: At this point, I would still use the PD-L1 marker. If [expression is] 50% or higher, I will only offer pembrolizumab alone to my patients. For everybody else, who is less than 50% and if I have someone who has a significant disease burden and needs a better reduction in disease burden, then I’m going to consider the combination. Based on the phase II data, we have close to a 55% response rate versus close to 30% with chemotherapy. That would tell me that a patient who has a heavy burden of disease requires a rapid reduction and, since the response rates are better with the combination, then I would use the combination. If there was any question about patient tolerability or potential toxicity, then I’ll just stick with a platinum-doublet chemotherapy.
Remember, we don’t have any data right now to tell us if the sequence matters. [This means], is it better to start with chemotherapy and then, at progression, bring in the IO agent? Or, is it simply better to start with that whole “kitchen sink” approach right off the bat? We don’t have any information like that. These are the types of questions that subsequent studies will shed some light on and help us guide the treatment for the patients. Basically, what I described is what I do in my practice.
TARGETED ONCOLOGY: How do you feel these decision-making factors will be different following confirmatory results from the KEYNOTE-189 trial?
BORGHAEI: It would be different if again, in all subgroups, chemotherapy with IO is shown to be better. For a phase III study, we must have phase III data to sort of change our practice. Simply having a response or PFS should not be the criteria for changing this kind of care. I feel strongly that survival should be looked at very carefully and should be the determining factor on whether practice should change or not. What about the utility of maintenance therapy with pemetrexed for these types of patients? That is the standard of care right now. We do 4 cycles of a platinum doublet if they have a nonsquamous histology and, for those who are responding and tolerating, you give maintenance pemetrexed. The way cohort G of the KEYNOTE-021 trial was done is that everybody would continue on pemetrexed, and that was a sign to give pembrolizumab. As it stands now, the data would suggest that we should continue to use pemetrexed in a maintenance setting in this particular patient population, because that’s where we have the best data. If the phase III study shows the same thing and follows the phase II design, then we are going to have to stick with that.
Borghaei H, Langer CJ, Papadimitrakopoulou V, et al. Updated results from KEYNOTE-021 cohort G: A randomized, phase 2 study of pemetrexed and carboplatin (PC) with or without pembrolizumab (pembro) as first-line therapy for advanced nonsquamous NSCLC [published online ahead of print September 18, 2017]. Ann Oncol. 2017;28(5): doi: 10.1093/annonc/mdx440.052.