Practice Changing Trials in Advanced Non-Small Cell Lung Cancer - Episode 8
Benjamin P. Levy, MD: Let’s move on to the other trial combining immunotherapy to chemotherapy using the ECOG 4599 backbone, carboplatin/paclitaxel/bevacizumab, and adding atezolizumab to that regimen in a 3-arm, phase III large study with 1200 patients. Hoss, you want to walk us down the design?
Hossein Borghaei, DO, MS: Sure. So, very interesting chemotherapy backbone—carboplatin/Taxol/bevacizumab. And, personally, I can’t remember the last time I gave that regimen to a nonsquamous non—small cell lung cancer, but it is definitely used. And I think there are market data to suggest that the combination is being used. But interesting design, as you said. Arm A was ECOG 4599: carboplatin/Taxol plus bevacizumab. One arm was that, one arm was the quadruple combination as opposed to triple, and the other arm was just carboplatin/Taxol plus atezolizumab without having the bevacizumab added. So, the presentation by Dr. Reck at ESMO, the immuno-oncology meeting in 2017, concentrated on only 2 arms of the study. That was the quadruple arm—carboplatin/paclitaxel/bevacizumab/atezolizumab—versus carboplatin/Taxol/bevacizumab. The other arm is chemotherapy plus atezolizumab. We don’t have the full data set, so we’ll have to wait for that. As you said, 1200 patients regardless of the level of PD-L1 expression, a coprimary endpoint, and intent-to-treat population, and a new biomarker was introduced, T-effector gene signature. And that’s a combination of PD-L1, interferon gamma signature, and another protein, CXCL5. And the idea here was that perhaps a T-effector signature would be a better biomarker. Mind you that we don’t have the full publication. We’re basing this on a 15-minute abstract presentation, so there’s a lot that still is not answered. But the overall data clearly show that the quadruple combination has a better PFS compared with the triple combination.
And so, the question then is, is this a new standard of care potentially for patients? I think the details of this study are very, very interesting, and there were multiple tables that were presented. For instance, regardless of the level of PD-L1 expression, we seem to have a benefit for the PFS. Again, as we’ve seen consistently throughout the history of PD-L1 development, higher level of expression definitely correlates with better clinical efficacy. I don’t think anybody can argue that PD-L1 can predict, if you have high levels, that you get better clinical efficacy out of it. But in this subgroup analysis, even patients who were TC0, tumor cell 0, and immune cell 0 still seem to benefit, although the hazard ratio was a little bit less. So, that’s one. T-effector signature didn’t seem to pan out as much as far as I’m concerned. Be curious to see what the panel thinks. It wasn’t any more predictive or less predictive than the level of PD-L1 expression. I don’t think that’s adding anything.
But I think the biggest question here for us is, Are we going to start using a quadruple combination? Now, the toxicity, as you would expect, was a little bit more with the quadruple combination. Although immune-related adverse events were not significantly worse; that was not a major issue here. You get the typical side effects from the bevacizumab and Taxol combination that you might not otherwise have if you didn’t use that particular backbone. But, again, another interesting study suggests that there is something to this chemotherapy. You can also look at the data and say, “Well, this really is a positive study suggesting that VEGF inhibition can augment immunomodulation.”
Benjamin P. Levy, MD: And we have a lot of preclinical data on that.
Hossein Borghaei, DO, MS: We do, definitely. So, this one is a little bit of a mix until we get the other arm of the study evaluated, and we have the data without bevacizumab to see whether chemotherapy plus atezolizumab alone gives us better clinical efficacy or not.
Benjamin P. Levy, MD: We may be forced to do cross-trial comparisons, and I’ll just throw this out to the group. This trial also showed a trend toward overall survival, as well, 19.2 months in the quadruple arm, coupled with a biological rationale of antiangiogenesis with immunotherapy and all the things we know about that at least preclinically. If we see a numerically higher OS in this trial than we see with 189—and, again, this is playing a game—would we be compelled to use the quadruple therapy over a triplet therapy if the median overall survival in IMpower is 19 to 20 and in KEYNOTE-189 it’s only 15 or 16? Not that I think we’re going to see that, but can we do this, compare apples to oranges?
Hossein Borghaei, DO, MS: I think that’s really a problematic kind of a comparison, different patient populations even though when we do the PD-L1 expression, all the tests are very similar. You’re still having variability in the different testing platforms and all of that. So I think it’s going to be a little bit difficult to do cross-trial comparisons like that and say one quadruple is superior to triple because you have a couple of months’ improvement in survival in one arm versus the other. I don’t think I would go there. I think, to me, if both studies are that positive, it just says you now have an option, you can use that or not. But I’d be curious to see what Roy thinks about the combination, I-O plus VEGF, because I think this has been one of your focuses recently. You had a presentation with such a combination.
Roy S. Herbst, MD, PhD: Well, without chemotherapy, there’s a suggestion that ramucirumab, an anti-VEGFR2 antibody, plus pembrolizumab looks reasonably interesting. That was only in a single-arm trial; it needs to be validated in a randomized setting. That’s being designed right now. But clearly, the preclinical data would suggest that antiangiogenic agents do have an effect on the tumor microenvironment, on antigen presentation, on T-Regs (regulatory T cells), and so forth. So, this makes a lot of sense. That said, I don’t know how we’re going to make any sense out of comparing the 2 trials.
Thomas E. Stinchcombe, MD: Sure.
Roy S. Herbst, MD, PhD: What we might get from both might make us feel more comfortable as a group here saying that a chemotherapy combination makes sense. Maybe adding in the bevacizumab makes sense, but we’re not going to be able to say much about the 2. But I think with this trial now, with the KEYNOTE trial, I think there’s no doubt that we’re going to be adding these agents to chemotherapy in the future in certain patients. And now we are looking forward to learning more about VEGF.
Naiyer A. Rizvi, MD: In terms of cross-study comparisons, what people are doing, though, is saying that the PFS for single-agent pembrolizumab in PD-L1—positive disease is 10 months. And then they look at the chemotherapy combinations, and they go, “Well, it’s a little bit better, but I’m not sure it’s so much better that I would displace pembrolizumab as my first-line treatment of choice in PD-L1–positive disease.” So, I don’t know about the others. To me, I think we need a lot more information around the chemotherapy combinations to make any conclusive statements. To me, I think that in the PD-L1–positive disease, single-agent pembrolizumab is still the standard of care.
Roy S. Herbst, MD, PhD: I think you make a good point. I spent some time last weekend with doctors who treat melanoma. And listening to them, they’re not thinking in 1 or 2 years; they’re thinking at 3, 4, 5 years. So, I think what we also have to be cautious about is, is it the short-term PFS or is it the long-term survival, and how are we affecting that by adding it to these combinations and the effect on the immune system? There’s really still a lot that we don’t know. But I agree with you, Naiyer, that is one way to look at it when you compare the PFS.
Thomas E. Stinchcombe, MD: I think people are going to look at the toxicity of the 2 chemotherapy platforms and probably make some decisions based on the tolerability and their perceptions of that.
Naiyer A. Rizvi, MD: I think that’s going to be important.
Thomas E. Stinchcombe, MD: And the other piece is I’m a little bit obsessed with complete response rates, and with single-agent pembrolizumab and PD-L1—positive diseases, CR rate was almost 10%. So, it was much higher than any of the chemotherapy trials that I’ve seen so far. And to me, I think that there may be some added benefit with the 2 together, but is there real synergy? We’re also trying to define more abbreviated courses of chemotherapy where we’re maybe getting sort of maximal androgen release without the myelosuppression, and there are a number of trials looking at 2 cycles of chemotherapy. So, I don’t think we necessarily found the sweet spot of how to combine chemotherapy with I-O.
Hossein Borghaei, DO, MS: I absolutely agree.
Transcript Edited for Clarity