Richard S. Finn, MD: I think IMbrave validated for the first time in a phase 3 study the importance of immuno-oncology [IO] agents in the liver cancer space. It’s taken a long time to get there. But we have one of the pioneers with us today, Dr El-Khoueiry. You saw a vision for this maybe early on with nivolumab in the CheckMate-040 study. CheckMate-040 has evolved over time, initially single-agent nivolumab and now recently another accelerated approval in second line of nivolumab/ipilimumab. There are a lot of data out there about new combinations. Can we increase the response rate of single-agent IO with combinations, the efficacy? IMbrave looked at VEGF antibody and PD-L1. Tell us about the exciting work going on here.
Anthony El-Khoueiry, MD: Thank you, Richard. As Tony alluded to earlier, there are different rationales for the different combinations. I would put them into 3 groups: anti–PD-1 agents with VEGF antibodies like atezolizumab/bevacizumab, anti–PD-1 agents with TKIs [tyrosine kinase inhibitors] that target the VEGF axis plus other targets, and then the PD-1/CTLA-4 combinations. And we’ve seen data for both nivolumab/ipilimumab, as well as durvalumab and tremelimumab.
There are different rationales for these 3 categories of combinations. So maybe spend 2 seconds on the nivolumab/ipilimumab or durvalumab/tremelimumab. The idea here, the way I explain it to patients is that we are removing 2 brakes off of the T cells, not just the PD-1 inhibition, but the CTLA-4 as well. And in CheckMate-040, there was a separate cohort that evaluated this combination of nivolumab and ipilimumab with different doses and different schedules of ipilimumab. And the same thing you will see with durvalumab/tremelimumab as well, because the dose and schedule of CTLA-4 may affect toxicity and efficacy. Briefly, with nivolumab and ipilimumab, the combination got the accelerated approval in second line post-sorafenib. The combo that received the approval is nivolumab 1 mg/kg and ipilimumab 3 mg/kg every 3 weeks for 4 doses, followed by nivolumab maintenance as a single agent. And there, we saw responses in the 30% range, 31% to 32%, 8% complete response, and intriguing median overall survival in an uncontrolled setting. This is a single-arm study of 22 months.
Now, of course, this comes at some cost of toxicity. Certainly, the grade 3 and 4 treatment-related adverse events were slightly above 50%, and half the patients required some steroids for immune-mediated adverse events, including 20% immune-mediated hepatitis-type events. So this is definitely a therapy that has to be chosen for the right patient and managed carefully. At the virtual ASCO [American Society of Clinical Oncology annual meeting] 2020, Katie, who’s with us today, presented data on the durvalumab/tremelimumab. Again, same concept of PD-1/CTLA-4 inhibition where we saw there were 2 different doses of tremelimumab evaluated:75 mg for 4 doses, kind of similar to the nivolumab/ipilimumab, versus tremelimumab 300 mg given once up front and then continuing with durvalumab alone. There were 2 other arms in that study. It’s a bit of a complicated study, how it’s evolved. But there was a durvalumab single arm and tremelimumab single arm. And the punchline is that with the tremelimumab 300 mg combination, the high-dose CTLA-4, even given once, we saw a promising median overall survival of 18 to 19 months, certainly within the range of what we saw with nivolumab and ipilimumab, with a manageable safety profile. Again, certainly, the grade 3 events were higher than with single agents, but manageable. And intriguingly enough, even with single-agent tremelimumab, there was a promising median OS [overall survival], in my opinion, that’s above 15 months or so. Certainly a repeating theme here with these combinations.
Richard S. Finn, MD: Yes, Katie, I’ve always been impressed with this 1 dose of tremelimumab lighting a fuse. Can you comment on the science on that a little bit? It’s impressive to me.
R. Kate Kelley, MD: It’s quite interesting. I think the evolution of that arm, which as Anthony mentioned, evolved along the way of the trial, came from pharmacodynamic data from both durvalumab/tremelimumab in lung cancer, as well as some similar data from nivolumab/ipilimumab in melanoma, which showed that a population of both proliferative, as well as a different population of activated T cells, could arise in a burst after the first dose of tremelimumab and that the extent of this T-cell proliferation and activation was dose-dependent. Intriguingly, this burst of T cells—this is in peripheral blood, by the way—this proliferation of activation of T-cells really occurred just after the first dose. It wasn’t recapitulated with sequential dosing. So really raising the hypothesis that you’re activating T cells in the periphery and potentially in lymph nodes and allowing them to become activated after the first dose, but perhaps there isn’t the same measurable benefit with sequential dosing. That led to the premise of the durvalumab/tremelimumab priming dose arm.
Richard S. Finn, MD: Which speaks to the idea of understanding something and then trying to optimize it. And I think CheckMate-040 and the study presented with durvalumab/tremelimumab are these studies that are trying to optimize how to dose and to maximize efficacy.
Transcript Edited for Clarity