Overview of Immune Checkpoint Inhibitors in Metastatic Colorectal Cancer

Transcript:

Kristen K. Ciombor, MD: We kind of started talking a little bit about immunotherapy but, in general, in metastatic colorectal cancer [mCRC]. We’ve known now for a while with many studies over the last 10 years that immunotherapy is very active in microsatellite instability [MSI]-high patients. I think 2 of our biggest studies that have really informed our practice in the first line are KEYNOTE-177, which randomized patients to pembrolizumab [Keytruda] vs investigator choice chemotherapy in patients with untreated MSI-high mCRC. What we saw there was that pembrolizumab more than doubled the PFS [progression-free survival] about 16 months vs 8 months (with chemotherapy). Even though we didn’t see an overall survival benefit, largely because most patients in the chemotherapy arm crossed over, we had about a 60% crossover rate. The lack of survival benefit doesn’t really impact the importance of those findings. At the same time, CheckMate 142, which is a nonrandomized phase 2 as opposed to the phase 3 KEYNOTE study. CheckMate 142 looked at the combination of nivo [nivolumab (Opdivo)] and ipi [ipilimumab (Yervoy)], NIVO-3, IPI-1. The response rates there were really great, approaching 70% in the first line with pretty manageable toxicities, and some of the median PFS and overall survival not yet reached. I think that’s really exciting. This is a relatively small subset of patients. We’re kind of whittling down the pie and figuring out how to treat each individual patient. One of the takeaways that I have from immunotherapy and MSI-high patients is that it can be curative without surgery. Patients have really durable responses when they respond and some people can be functionally cured, which is really exciting. MSI is clearly a good biomarker, probably our best biomarker for response to immunotherapy. Dr Kasi, we hear all different tumor types are looking at PD-L1. What do we think about PD-L1 in colorectal cancer? Is it helpful?

Pashtoon M. Kasi, MD: Well, if you had asked me a year ago, the answer would’ve been absolutely no. That it’s not helpful at all and should not be tested in CRC under any circumstance. That’s where knowledge and understanding are changing. Where some of the recent data in the last 6 months or a year in terms of looking at IO [immunotherapy] doublets in trials, there is a signal to the point that some of the prospective studies ongoing are using this as an integral biomarker for enrollment. Now, again, in the clinical practice, PD-L1 CPS [combined positive score] in colorectal cancer has no role and should not be tested, but if you are looking for a trial that’s enrolling this as an integral biomarker response or enrollment criteria, it is something that I never thought this would be relevant but it is surprisingly.

Kristen K. Ciombor, MD: My next question is what about high TMB [tumor mutational burden]? Do you think the cutoff of 10 is the right cutoff, or something controversial?

Pashtoon M. Kasi, MD: That’s more of a controversial question, especially in the context of GI [gastrointestional] malignancies, and bringing back to add to the conundrum even more is the fact that the liquid TMB that’s reported is, I would say, 5 to 10 points even more inflated, so we have to even be more careful. So the use of TMB-high, at least in colorectal cancer, even though we have an agnostic approval of pembrolizumab similar to the agonistic approval MSI-high, in CRC the only MSS [microsatellite stable]-subset that really benefits are these poli or polyposis ultra-hypermutated subset. But these intermediate-range double-digits TMB and especially in the context of pembrolizumab being reported, I think those are opportunities of caution because...

Joel R. Hecht, MD: I think it’s actually worse where there are people who are being treated, I agree with you, who don’t have poly or polti. And then someone says they have a TMB of 11, you’re right, on some liquid biopsy and they’re exposed to the toxicities, not to mention the expense. I agree with you. I do think there is this group and I don’t know how to characterize it. If you look at the MSS, the tail of the curve, there’s a little bit of something there even though the response rate is incredibly low in patients with checkpoint inhibitors. But I don’t know in your practice, but I do see patients who were treated with IO who clearly do not have and then we have NGS [next-generation sequencing] on tissue. They don’t have a poli. They don’t have a polti. They don’t have MSI. And they have toxicity.

Pashtoon M. Kasi, MD: In the context of young-onset CRC patients, you’re also robbing them the opportunity of any combination trials and then exclude any prior anti-PD-1, PD-L1 exposure so…

Joleen M. Hubbard, MD: That’s a really great point. I’ve had multiple trials now come out with immune checkpoint inhibitor combinations, and yes, using when it’s not indicated excludes them from these trials. Very important to know.

Kanwal P. Raghav, MD: Of note, PD-L1 expression and TMB are linked to MSI-high states in CRC. And I don’t really agree with the cutoff of 10, but I think that TMB is not just a quantitative issue. It’s also a qualitative issue and I think there is, hopefully, there is more research that is coming out which can help us refine exactly what TMB means.

Joel R. Hecht, MD: I think though the histology-agnostic idea of TMB-10 as being that all tumors are the same. Clearly, a lot of that comes from the lung data, which is an immune-sensitive disease. And we take care of MSS colon, which is a not particular immune-sensitive disease, and they’re just different.

Pashtoon M. Kasi, MD: And it’s sobering even in non-small cell lung cancer. TMB-enriched studies have not been able to show that as only a predictive marker response. So I would say for both those markers, at least the context CRC, caution is advised.

Transcript edited for clarity.