Implications for treating uveal melanoma with tebentafusp based on responses demonstrated by the IMCgp100-102 and IMCgp100-202 clinical studies.
Richard D. Carvajal, MD: We've had 2 dedicated clinical trials now for the uveal melanoma patient population, the 102 trial and then the 202 trial. The 102 trial was a dose escalation followed by an expansion trial for patients who were previously treated, and the 202 trial was the pivotal randomized trial where patients who were A0201-positive and untreated were randomized on a 2:1 ratio to either tebentafusp or best alternative care. And the best alternative care in this trial was either single-agent PD-1 [programmed cell death protein 1], single-agent CTLA-4 [cytotoxic T-lymphocyte-associated protein 4], or chemotherapy. And in this trial, it turned out that the majority of the patients who were randomized to best alternative care did in fact get single-agent PD-1. But Marlana, I don't know if you want to share some of the high-level big news—the survival advantage for this study.
Marlana M. Orloff, MD: Yes, even just going back real quick to 102, the patients that were enrolled on that trial were patients who were heavily pretreated, a lot of whom had either prior liver-directed therapy or even prior immune checkpoint inhibitor therapy or chemotherapy. Early on in that setting, there was a 1-year overall survival of potentially 70%-plus in patients who were heavily pretreated. As far as 202, certainly our expectations were high, but we obviously needed to see it in the randomized study. The study survival data had been presented at AACR [American Association for Cancer Research meeting], and the big take-home there was understanding that the IC or investigator choice was a PD-1 dominant group. We're looking at a 1-year overall survival for tebe [tebentafusp.] of 73.2% vs investigator choice, 58.5% with a hazard ratio of 0.51. With that kind of hazard ratio, it was obvious that there was a clear overall survival benefit for tebentafusp. Since that presentation at AACR, there have been a number of presentations teasing out some of the unique aspects of this drug and some of the other trends that we have seen. Rich, do you have any other thoughts about the general 202?
Richard D. Carvajal, MD: We can’t underestimate how phenomenal these results were. The overall survival of nearly 22 months is just far beyond what I think I would have expected, because based off historical data sets, we would have said the median overall survival's typically about a year, give or take. That 21, 22 months is absolutely phenomenal. And that's in the setting of the hazard ratio of .51, which is incredible, but that's also in the case where the control arm did better than I would have expected.
Marlana M. Orloff, MD: Yes, especially because PD-1 in uveal, if you look at all the studies, both retrospective and there have been a few prospective series, of a response rate of only about 3% to 5%, mostly in the PD-1 studies. A 1-year OS [overall survival] of 58.5 for basically a PD-1 dominant control group is decent. And there wasn't crossover in these patients, at least at the time of this data, so you ask yourself, OK, what did that patient population ultimately get after the fact? But it wasn't tebentafusp. The bar was high, kind of in both groups, but tebe still outperformed, which is impressive.
Richard D. Carvajal, MD: And I think the other important thing to highlight is the trial did hit some of its secondary endpoints—for instance, progression free survival, right? The hazard ratio in terms of PFS [progression-free survival] was .73 in favor of tebentafusp, but the medians were not that good. We're talking median PFS was still 3.3 months.
Marlana M. Orloff, MD: Yes.
Richard D. Carvajal, MD: And the response rate is 9%.
Marlana M. Orloff, MD: Yes. I remember when ipilimumab came out, and I think that might have been our first inkling that especially with immunotherapies, response rates may not tell us everything. And even PFS may not tell us everything, and I think we're still trying to tease this out. Is the way that we traditionally look at efficacy appropriate for especially unique types of immunotherapies? Is it a problem with imaging? Is response how we describe it on recessed or immune recessed, the appropriate way to be looking at this? Obviously, overall survival is overall survival, and that's kind of the holy grail, but we really use all these other proto-surrogate kind of earlier markers to try to help us predict, but maybe we're not doing such a great job at that, just by looking at things like response rate and PFS.
Transcript Edited for Clarity