Marina Chiara Garassino, MD: The IMpower results are really important because they suggested that the combination of chemotherapy/angiogenics/immunotherapy is important. And we have the results only on the progression-free survival, so this is an important piece of information. But the combination of carboplatin/Taxol plus bevacizumab is not used in Europe compared with the United States, for example, because in Europe, in general, people prefer to use the combination of cisplatin/pemetrexed—because, for example, you have less alopecia; sometimes you have less side effects, and sometimes it is not always easy to use bevacizumab in the metastatic setting. But anyway, there is a stronger rationale to combine chemotherapy but also use the angiogenic fix to promote the penetration of the T cells into the tumor for allowing the use of immunotherapy.
From the rationale comes an important piece of knowledge, because we know that adding the antiangiogenics plus chemotherapy to atezolizumab works. And this is important, because we know that antiangiogenics can enhance the penetration of the T cells in the tumor, and with the chemotherapy, you can promote the immunogenic death, so that can be synergistic with immunotherapy.
I don’t know a scenario in Europe where carboplatin/Taxol plus bevacizumab is not the most commonly used regimen. It will become the standard of care—something like the combination, for example, of cisplatin/pemetrexed/pembrolizumab that will come soon in the next congresses. I think that, again, the toxicity profile will be important to make the decisions for the patient—also because all the regimens have a very different toxicity profile.
Solange Peters, MD, PhD: The field of advanced disease is now characterized by, I would say, a very large number of trials trying to question frontline therapy—what to do when your patient is naïve, never received any treatment, and trying to have immunotherapy be part of this game. IMpower is a trial, interestingly, addressing a complex combination of drugs, as compared with the United States regimen. The United States regimen—there are many, but 1 of the classical ones is paclitaxel/carboplatin/bevacizumab, based on an important randomized trial in the past. So, this is a skeleton, the standard of care: chemotherapy plus bevacizumab. And this trial is evaluating chemotherapy/immunotherapy/an anti-PD-1/atezolizumab versus the skeleton, or chemotherapy plus atezolizumab plus bevacizumab. It’s a 4-drug regimen versus the skeleton standard of care. And during this meeting, we were able to see that the results of the main endpoint of the trial is to have the 4 drugs being compared with the standard of care. The intermediate one without bevacizumab is only to read out if all the other endpoints come out positive for the 4 drugs—so, PFS and OS. And if PFS and OS of the 4 drugs are positive, then the atezolizumab chemotherapy will be compared with the standard of care, chemotherapy/bevacizumab.
So, today, we can only speak about the 4 drugs versus the chemotherapy/bevacizumab standard. The rationale to do so is not just adding drugs. It could look like having lots of drug together, but the rationale existed since the ’90s. Since the ’90s, it has been shown that VEGF, or vascular endothelial growth factor, is impacting the immune system. When you have lots of VEGF, the T cells are not trafficking well to the tumor. They’re trafficking if the tumor is inefficient, so they cannot reach the tumor site. It has also been shown that VEGF per se, because of being in the blood, is inhibiting the T-cell function and is inhibiting the presentation by dendritic cells of some antigens to the immune system. So, many steps of the immune activation are blocked by VEGF. It makes sense when you give immunotherapy to also give an VEGF blocker, which is bevacizumab. So, biologically, it’s a very good model and has a very strong rationale. That was the question of IMpower: If you give 4 drugs—bevacizumab/atezolizumab/chemotherapy—is it better than giving chemotherapy plus bevacizumab in terms of PFS, in term of OS, and if everything is positive, without bevacizumab? But that’s where we are.
So, at this [ESMO Immunotherapy 2017] meeting, we could see the PFS of 4 drugs versus the standard of care. To make a long story short, PFS is statistically significantly improved. Like with immunotherapy, it’s not so much median PFS, which is interesting, but it’s to see the curves separating over time, more and more—we call it nonproportional hazard ratio. The benefit is growing over time—seeing that the curves separate to a point that at 1 year, you will double the number of patients who are without relapse when you add the atezolizumab and bevacizumab or you add the atezolizumab to the regimen, as compared with the chemotherapy/bevacizumab. So, it’s a very interesting improvement in terms of preventing the progression of the disease.
The main question of IMpower is to know if you have to give 4 drugs as frontline treatment. How are you going to impact overall survival? Because it means you decide not to sequence so much. You give everything at the same time. The second thing is, are there subgroups of patients who benefit more from the strategy? And we could see the subgroups; it is broken down by PD-L1 in the trial. They also had a surrogate signature for PD-L1, but let’s say broken down by PD-L1. You can see that, basically, this looked like it would be pretty independent of PD-L1. The benefit is observed across various categories of PD-L1 expression.
The weaker is probably the PD-L1—negative. They still had a significant hazard ratio, but what we find, it was a hazard ratio of 0.77, so it’s just on the limit of what we would consider. I don’t want to make a conclusion out of it, but I’ll just say that due to this nonproportional hazard ratio, we have to wait a little bit more to see if the difference becomes bigger. And what about survival? So, for this negative PD-L1, let’s wait a little bit more. For the very strong PD-L1, we know we have pembrolizumab, and the results in the PFS are more or less the same. So, again, we have to wait a little bit more, see the overall survival. But maybe in very high PD-L1, 1 drug, pembrolizumab, might be preferred to 4 drugs for cost but also for toxicity and quality of life, so this may be the niche where pembrolizumab might stay as being the standard. So, we have all these patients in between 49% of PD-L1 up to 1% of PD-L1 where, really, IMpower is showing that it’s a new potential treatment option in these patients, being better than chemotherapy, and should be considered in that setting. And for the other 2 extremes, we need more data and more time to see the results.
Well, I think we have to start to think about the landscape of frontline immunotherapy, because after IMpower, there are many trials to come next year. And the complexity—it’s just the beginning, because we will have so many different results coming from the trials that we have to start to know what we want to adopt or not. The same thing for the regulatory authorities—they have to know what they are going to accept and based on which endpoint, and for IMpower, it remains a mystery. Are they going to potentially make it available, maybe in the United States, already based on PFS or not? But what I would say is if I were now in Europe and I had access—because my authority has given me access to atezolizumab, bevacizumab, and chemotherapy—I would use it in this category of patients, between 1% of PD-L1 and 49% of PD-L1 expression, because it’s better in term of PFS and because the shape of the curve showing you the potential for a long-term benefit is something that is unprecedented in non—small cell lung cancer. So, if you have the opportunity to act on it, you probably should do it. And keep in mind, in lung cancer, you are never sure that you will have a second line, a third line, a fourth line, because these patients die from the disease on the chemotherapy. So, if you have the opportunity to test a strategy of immunotherapy, because you are reimbursed, you have the right to do it. And if you have a positive trial, you should do it frontline.
Transcript Edited for Clarity