Combination Therapy Updates from ASCO 2019

Transcript:Axel Grothey, MD: One of the most exciting data that actually I’ve seen here at ASCO [American Society of Clinical Oncology Annual Meeting] 2019 in colorectal cancer was a poster presentation hidden in the drug development section, which really didn’t get a lot of attention by the GI [gastrointestinal] community. It was for a combination of regorafenib plus nivolumab in the treatment-refractory patients with gastric cancer and colorectal cancer, and they were all, except for 1 patient, MSS [microsatellite stable] tumors. One patient with colon cancer had a MSI [microsatellite instability]—high cancer. And the response rate of 25 patients with metastatic colorectal cancer, MSS, was 33% with a combination of regorafenib-nivolumab, which by themselves have very limited activity, actually probably low single-digit activity. And the waterfall plots that it showed for the gastric cancer cohort, and the colon cancer cohort, and the durability of response was quite intriguing.

Now the regorafenib dose that was tolerable in this setting, which was a Japanese phase I study, actually, was 80 mg in combination with full-dose nivolumab. It seems to be that for immune modulation, if we believe its affect, regorafenib can actually be dose reduced, and it’s only limited toxicities at that point in time. And I find these data quite intriguing. The question is, is it passing the threshold for, let’s say, starting a phase III study that would eventually be the gold standard for drug development? Or should we kind of expand this cohort now—let’s say in a Western population, in the United States, etc—and look for a response rate as primary endpoint? The data are definitely intriguing, and I think they need to be pursued further.

Tanios Bekaii-Saab, MD: Yeah. The REGONIVO regorafenib—immune study was very interesting, although we have to keep in mind 2 elements that limit the excitement. 1) It was a Japanese-only study, so we don’t know how this applies to the rest of the world. 2) This was a small study and a reminder to all of us that the promise for the atezolizumab and cobimetinib study was a phase Ib study that started with a 20%-plus response rate, and by the end it ended up with a less than 10% response rate. But the phase III was already on its way.

We always have to take a step back, and this study included both gastric and colorectal cancer patients, and gastric makes a lot of sense because both agents have activity that’s meaningful. Regorafenib has activity, nivolumab has activity. In colorectal, regorafenib has some activity, but nivolumab—in the MSS at least—has 0 activity. Unlike gastric cancer, where both work well and you put them together and you hope to see synergism, which was seen, in colorectal I would remain a little bit more cautious. Because again, we have a drug in this setting that works OK, and we have another drug that doesn’t work at all. And we put them together, and suddenly it’s magic. That does not hold the scientific test at this point in time. We need to understand a little better.

Now, I do think that regorafenib offers a great partner to a PD-1 [programmed cell death protein 1] inhibitor, and the reasons are it hits VEGF, and we know VEGF and PD-1 seem to synergize well. But it also hits CSF1R, which I think is very important. It’s 1 of the mechanisms that seem to drive, even further, the activity of PD-1 inhibitors. But it also hits a lot of other targets that may have relevance. Some we understand, others I would say we don’t understand, but they may have a role. So regorafenib and NIVO [nivolumab], or regorafenib and PEMBRO [pembrolizumab], or regorafenib or whatever PD-1, PD-L1 [programmed death-ligand 1] inhibitor is probably a rational combination to think moving forward. In microsatellite stable colorectal cancer, I think it’s a reasonable strategy to explore. In gastric cancer I think it needs to be explored. In colorectal cancer, I think it should be explored cautiously.

I don’t frankly believe that if you take really cold tumor—and I’ll use just very simplistic terms, a really cold tumor that has no evidence of much activity, inflammatory activity—that suddenly you’re going to make it warm or hot to have a response. I don’t think those will ever respond, at least with what we have today. We have evidence that the intermediate tumor mutational burden, or perhaps even those high that are non—MSI-high or perhaps the CMS4 or some variation of the CMS4 patients, those may actually fit a good bucket of, I’d say, about 20% of the patients who may not respond really well to single-agent PD-1. Those are the patients where you could actually, perhaps, move them from warm or warmish to hot, and then you can get the benefit in those.

I think when we look at the landscape in colorectal cancer in the microsatellite stable, we have to perhaps have a cautious approach, start with those a little more targeted fashion. And then for the other patients, perhaps explore as well, separately. But I wouldn’t really jump into a large phase II, even large phase II randomized study without having a little more data. Because if the results of this Japanese study hold out, that certainly will transform how we treat patients with microsatellite stable colon cancer, even consider moving it up the line. So these are certainly exciting times. Caution needs to be applied, and I think if the past tell us anything, it is to be even more extra cautious over this group of patients and hold the excitement until we see more data.

Transcript Edited for Clarity