Insights From: Richard Kim, MD, Moffitt Cancer Center; Scott Kopetz, MD, PhD MD Anderson Cancer Center
Scott Kopetz, MD, PhD: We are still learning a lot about how tumors respond to immunotherapy [I/O] when they have MSI [microsatellite instability]-high disease. The key findings are that the majority of patients will have durable benefit with PD-1 [programmed cell death protein 1] inhibition, or PD-1 plus CTLA-4 [cytotoxic T-lymphocyte–associated protein 4] inhibition, in metastatic colorectal cancer when they have that MSI-high disease. But we do know that some patients don’t. We’re still trying to sort out who those patients are and why that may be. There are a lot of theories about co-mutations, antigen presentation machinery. Perhaps the microbiome is playing a role in this. What we can say is that it is not an association with whether or not it’s a hereditary syndrome, such as Lynch syndrome. Whether it’s Lynch syndrome or sporadic doesn’t appear to play a role. Similarly, whether or not they have a concurrent BRAF mutation does not appear to play a major role in sorting out the responders or nonresponders.
I may be a bit of an outlier in this setting, but I am a utilizer of the doublet of nivolumab and ipilimumab commonly for an MSI-high patient. The reason is that increasing data have suggested that a patient with MSI-high colorectal cancer treated with immunotherapy can be cured of their disease. And we don’t use that term lightly. This is metastatic disease that we’re treating, but indeed there is a proportion of patients who, even if they have some radiographic evidence of something that remains on their scan, really are cured of their disease. In those settings where we go in and resect what remains, we see a lot of just fibrosis and complete pathological response.
When we can talk about curing patients of their metastatic disease, we then also have to ask questions about, what’s going to give us that highest response rate and the greatest tail on that curve? We don’t have randomized data. What we do have are data from the BMS [Bristol-Myers Squibb] study that enrolled first a cohort of patients treated with nivolumab, and then a second cohort later—same study, same centers, just a subsequent population—treated with nivolumab and ipilimumab.
Ipilimumab is low dose. It’s just a few doses that are given, so this is reasonably well tolerated. And the cross-cohort comparison suggests a higher response rate and a higher tail for these patients. I wish we had level 1 data that could compare the 2. But really when we’re at the point of talking about cure for these patients, I err on the side of overtreating with the I/O combination in order to really derive the maximum benefit for the patients.
Richard Kim, MD: In patients who progress on checkpoint inhibitors after second- or third-line therapy, unfortunately at this moment there are no good options for those patients other than on a trial basis. There are some data that if you fail a checkpoint inhibitor by itself, can you add a CTLA-4 on top of it to overcome resistance? I think that’s yet to be answered. I think that’s still a question whether it works. Unfortunately, at best, we try to enroll them on different studies, on another study that allows patients who received prior checkpoint inhibitors.
The biggest question that we have right now is, do checkpoint inhibitors work in microsatellite stable [MSS] tumors? Eric Chen, MD, PhD, presented data last year that combined durvalumab plus tremelimumab, which is once again PD-1 plus CTLA-4 in microsatellite stable patients. This was a study where they were randomized to best supportive care versus the combination of I/O. And interestingly enough, this is the first study that actually showed there could possibly be an overall survival benefit in MSS patients by giving a combination of those 2 drugs. Now, the difference was only 2 months. And the disease control rate was much higher in the combination arm versus the best supportive care arm. However, once again, we have to take it with a caveat.
Obviously, we’re still waiting for a final biomarker. There are some issues with the study because it was not a placebo-controlled study. It is unclear what they got afterward, and we don’t know the correlation between TMB [tumor mutational burden] versus the response or stable disease in those patients as well. So I don’t think using I/O therapy in MSS disease is ready for prime time. I would like to see more data, especially other molecular profiling from the study, trying to correlate which patient had longer stable disease that led to overall survival difference. I think I would like to take a look at that. So, until some of those data come out, I would not be using I/O therapy in patients with microsatellite stable tumors in colon cancer.