MSI Status and Immunotherapy in Advanced Colorectal Cancer

Transcript:John Marshall, MD: So, the consensus molecular groupings, the 4 groups that have emerged, is this something that doctors should be sending off some place or can you get this result back?

Johanna Bendell, MD: So far, there is no way to get the result back on these different CMSs (consensus molecular subtypes). I think for right now, and until we hear Alan’s data, which might clarify things a little bit more, they are not used in treatment decision making yet. But they certainly could potentially identify groups of patients that maybe we need to do specific clinical trials for, that these patients might need a different treatment than the others.

John Marshall, MD: Okay. So, for right now, right versus left, stay away from EGFRs in the right. Bevacizumab works in both, has some impact, so it’s been our default anyway, not changing a lot of practice. We’ve got to get the right ICD-10 coding in. We’re going to send off RAS and RAF testing. But you had mentioned, Johanna, earlier that you almost don’t start a new patient without knowing microsatellite instability, and I agree with that. So, tell us a little bit about this. Why do we, all of a sudden in every patient with colon cancer, need MSI (microsatellite instability) testing?

Johanna Bendell, MD: We know that patients who have microsatellite instability are more likely to have mutations than patients who are microsatellite-stable. And we know that the number of mutations that a tumor has can be related to their responsiveness to immunotherapy, the checkpoint inhibitors. Certainly, we’ve seen in colon cancer that when you give a checkpoint inhibitor to a patient who is microsatellite-instable, you get a very nice response rate and a very durable response rate. And so, these guys have been the focus of research, mostly to this point, with checkpoint inhibitors in colon cancer. There are different ways to test microsatellite instability, both by IHC (immunohistochemistry) and by PCR (polymerase chain reaction), but MSI is a very reasonable and relatively less expensive assay to check, where you could make a major treatment decision for your patients based on the result.

John Marshall, MD: Well, to be fair, let’s say I do this and I get back a positive result for MSI. I can just go get a checkpoint drug? What are you talking about?

Johanna Bendell, MD: So, interestingly enough, both pembrolizumab and nivolumab are compendia listed for patients with microsatellite-instable colorectal cancer. So, in some senses, yes, you can. I would actually recommend getting these folks on clinical trials.

John Marshall, MD: Should I give a frontline patient a checkpoint inhibitor right from the get-go?

Johanna Bendell, MD: That’s a huge question that we don’t know the answer to yet. What we think is that they do have responses. We’ve all had anecdotal experiences with people having amazing responses to these agents, but we’ve also had experiences where patients don’t have the responses. So, do we give up so fast on the chemotherapy in first-line that’s relatively effective? I think it really depends on the clinical situation. If you have a patient that’s asymptomatic, relatively low burden of disease in the first-line, maybe that’s the person that you would give a checkpoint inhibitor to right away. So, you have a little time to see if they’ll work. But if the person is symptomatic, maybe you should do chemotherapy first. We don’t know the appropriate answer to that yet.

John Marshall, MD: Tony, I’m going to put you on the spot. You have a patient with metastatic colon cancer, it’s MSI-high, they’re a brand new diagnosis. What’s going to be your frontline treatment?

Tanios Bekaii-Saab, MD: I actually would use a PD-1 inhibitor. I’ve had more than just anecdotal experiences with patients who were very symptomatic, high tumor burden, who actually had significant responses to PD-1 inhibitors. Those are patients that do moderately well with chemotherapy, and we always have the opportunity to put them back on chemotherapy. I don’t think there’s a right or wrong answer. I agree with Johanna. We need to study this further, and this is being studied. But I’m compelled with many of these patients to actually start them on a PD-1 inhibitor.

John Marshall, MD: You mentioned the chemotherapy. I get a lot of calls and e-mails about how their MSI-high chemotherapy doesn’t work.

Tanios Bekaii-Saab, MD: Not true. Fluoropyrimidines alone may not have a benefit, but chemotherapy does work. The question is whether it works as well. That’s less understood.

John Marshall, MD: And that’s fluoropyrimidine specifically in the adjuvant setting and older studies?

Tanios Bekaii-Saab, MD: In 5-FU, absolutely, but in older studies, it may be or may not be. Perhaps these patients in the adjuvant setting do so well that they don’t even care about chemotherapy anyways.

John Marshall, MD: So, Alan, let me pick on you a little bit. I think that although we’ve gotten a lot better at this, testing for MSI and interpreting the results, I can’t tell you how many times people will come and see “present” for their IHC testing and think they have a mutation. Maybe talk us through a little bit about testing?

Alan P. Venook, MD: That’s how it’s reported. It’s stunningly common that the community oncologist doesn’t quite understand that.

John Marshall, MD: It’s bad language. I don’t blame them.

Alan P. Venook, MD: It is. It reports, “present, present, present,” but then if you read, it says there’s no evidence of the absence of any or there are no mutations. You’re looking for the protein in the IHC. You’re looking for the product of the genes. In our multigene platform, for example, we do MSI testing and actually look for microsatellite instability. And there you get a clear-cut answer. So, the nomenclature needs to be fixed, and I do think we also get it on every patient right off the bat. I think whether to use it immediately or not, of course, is a good debate; that’s being studied. But I agree with everybody else, that these drugs clearly have a play. Pembrolizumab, I believe, is before the FDA now already, and, as Johanna said, though, the NCCN guidelines include both nivolumab and pembrolizumab, recommending MSI testing right off the bat. In fact, we’ve had no trouble getting either of those drugs for patients with metastatic colon cancer.

John Marshall, MD: And the point you’re making, Tony, is if they work, they work really well and they’re relatively nontoxic. If they don’t, you’ll have a shot at chemotherapy.

Tanios Bekaii-Saab, MD: We’ll have a shot at chemotherapy, absolutely. John Marshall, MD: And the tumor control rate is very high and refractory, right, 80% or 90%?

Tanios Bekaii-Saab, MD: Yes, close to 80%. That’s the control rate, with a good 20% to 30% response in the very refractory setting.

John Marshall, MD: Johanna, does it matter if you’re inherited versus somatic?

Johanna Bendell, MD: So far, we don’t know. They’ve looked at and tried to break down, for the data that we have now for the patients who have Lynch syndrome, where it’s inherited or sporadic mutations. And so far, we don’t see a difference in outcomes no matter which way you are.

John Marshall, MD: Okay. We were all crisp and clean about MSS, MSI. It doesn’t work for MSS, and then you published a paper that basically said, “Well, wait, if we combine it, maybe we’ll see some benefit.” Could you just give us a little summary of where we are in developing immunotherapy for microsatellite colon cancer?

Johanna Bendell, MD: Microsatellite-instable colon cancer is only about 5% of patients with metastatic disease. So, what do we do for the other 95% of patients? Is there a way to convert these patients into responders for immunotherapy? That’s the big subject of research right now. I presented probably just the first crack at it, and certainly, there are a lot of studies that are coming that are looking for different ways to try to rev up the immune system to have more of an immune presence within the cancer cells, to bring T cells into the cancers to get that immune therapy going and working for those patients as well. This particular study was using a MEK inhibitor, cobimetinib, to do that, and we know that when you give a MEK inhibitor in tumor models, it increases CD8-positive T cells within the tumor, it increases expression of class 1 MHC. So, when we gave that in combination with a PD-L1 inhibitor, atezolizumab, we saw some responses in patients with microsatellite-stable colon cancer. It’s just the first step along the way, and I think there’s a lot more development to come.

John Marshall, MD: I’ve been in this business a long time, and as a drug developer, I’ve never seen a gold rush like this. There are so many studies, they’re so quick. The combinations are coming fast and furious.

Alan P. Venook, MD: Some of it is the Internet, of course. Patients come in asking for these drugs no matter what their diagnosis. I would add one other nuance, which is important to comment on, and that is the coexistence of MSI-high and BRAF mutations. This is a subset of patients, and actually, I’m not exactly sure what to do for them. If you look at all patients with BRAF mutations, about 25% or 30% will be MSI-high. And so, is that a patient on who you use a checkpoint inhibitor? We probably should, but, of course, there’s a big effort now to identify BRAF mutations early on in the all-RAS wild-type patients, and these are patients who do correlate on average. Really, there are data at the meeting today, at the GI ASCO, that will show at least a combination of molecules will improve outcome in those patients. So, that’s a nuance of what do you do with a coexistence of those mutations. I think it’s something also to be aware of, not to get lost over that. That is not a mistake that they coexist.

John Marshall, MD: So, there is a lot to learn. Certainly, for the MSI-highs, the mismatch repair, we want to get them somewhere on treatment, on trial. And for the others, there are lots of study opportunities, so that’s great.

Transcript Edited for Clarity