Relapsed or Refractory Metastatic CRC - Episode 8

Relapsed/Refractory mCRC: Immunotherapy in MSI-High Tumors

February 25, 2020


Axel Grothey, MD: Immunotherapies are approved in metastatic colorectal cancer for patients with mismatch repair—deficient and/or MSI [microsatellite instability]–high tumors. There are different ways to assess the mismatch repair deficiency or MSI status of patients. One of the most commonly used screening methodologies is immunohistochemistry for the 4 critical mismatch repair proteins, which really flows very nicely in the work-up of tumor specimens. Actually, right now, independent of stage and age, every colorectal cancer is being tested for mismatch repair deficiency. We have the result, including actually virtually with the pathology report right up front.

MSI testing requires a PCR [polymerase chain reaction] reaction, so we look for the length of the microsatellite. It’s a consequence of the presence or absence of the mismatch repair proteins that we test for with immunohistochemistry. Next-generation sequencing [NGS] testing also gives us microsatellite instability based on algorithms that have been established from various companies, including liquid biopsy companies. They can now also read out the MSI status in a lot of these tests.

For all practical purpose, however we slice and dice it, whatever tests we use, we identify patients for the usefulness of immunotherapies. There might not be a 100% overlap, but the concordance is good enough that whether we identify patients with immunohistochemistry, PCR reaction, or NGS, it doesn’t matter because we can use immunotherapy in these patients.

Marwan Fakih, MD: Every patient with colon cancer should have microsatellite instability testing, whether that’s a patient with a stage I, stage II, stage III, or stage IV disease. The implications of MSI testing can vary based on the stage. For stage IV disease, it’s more for treatment decisions. For earlier-stage disease, it’s more to define if the patient, for example, has Lynch syndrome or should be screened for Lynch syndrome. In our institute, every single patient with stage IV disease will be tested up front for microsatellite instability. We use immunohistochemistry looking for MLH2, MSH2, MSH6, and PMS2 as our screening tool.

However, in addition to that, every single patient in our institute undergoes next-generation sequencing testing. Our NGS panel using Ashion, also known as GEM ExTra, or when we use FoundationOne and there are other panels such as Caris, all these NGS panels include microsatellite instability within their panel. Practically speaking, some of our patients are tested in 2 different ways rather than 1. But it doesn’t matter how you test it; it’s important to test it. Certainly, there may be discordances sometimes between immunohistochemistry, PCR, and NGS panels. It appears to me that when you couple the data regarding tumor mutational burden [TMB] and the microsatellite instability status, that’s when you have even more confidence that your immunotherapy is going to be a good option for your patients.

Scott Kopetz, MD: As was mentioned, in my practice, the primary decisions on the basis of MSI high and its presence or absence—I don’t explicitly order TMB, although I’ll use the NGS panel-testing results to really look for ultra-mutators that may not be captured on the MSI results. There is a mutation in POLE most commonly. There is some POLD, but those are fairly rare. But the POLEs, which are also uncommon, can result in a much higher tumor mutational burden, even greater than what we see in MSI high. As a result of that, sometimes you can get secondary MSI loss with your testing by proteins.

In that setting, those are patients we think benefit from I/O [immuno-oncology] therapy. But we also have to be really reticent, and I don’t see evidence of activity of these borderline elevated TMBs. Sometimes on the report they’ll say TMB moderate high or TMB high when they’re MSS [microsatellite stable]. For those patients, there’s really no evidence that a large number of patients will benefit from a PD-1 [programmed cell death protein 1] inhibition. I don’t necessarily order the TMB, because in those settings the TMB doesn’t add anything for those borderline patients.

We’re certainly always exploring about what is it in MSI high that may play a role in predicting response. I think we’re more comfortable saying what doesn’t predict response to a PD-1 inhibition in an MSI-high tumor. We know that BRAF mutation, which is commonly seen in MSI patients, does not associate with response or resistance. We also know that the presence of a hereditary syndrome, Lynch syndrome, is not associated with a greater or lower response rate compared with somatic MSI. There have been data that have demonstrated that PD-L1 [programmed death-ligand 1]—obviously an area of interest in many other tumor types—is not at all predictive in the setting of MSI, which is intriguing of why this may be different. But there’s really no signal that that’s relevant.

In my practice, we’re using PD-L1 inhibitors commonly in combination with a CTLA4 inhibitor in second line or earlier. I think we’re increasingly hopeful that administering this in patients earlier, such as in the first-line setting, will provide benefit. It should be noted, however, that this has not been shown. We’re hopeful that we’ll have data about this, but we don’t know what the impact of immunotherapy in first line for that is. There’s hope that as in the COMMIT study that’s ongoing, we’re asking the question about whether the combination of a PD-L1 with chemotherapy may be better than chemotherapy alone. There’s also a completed study looking at PD-1 inhibition versus chemotherapy that’s completed enrollment, and we’re awaiting results.

Axel Grothey, MD: When we talk about immunotherapy for colorectal cancer, MSI-high, mismatch repair—deficient colorectal cancer, there are 2 main options: pembrolizumab single-agent therapy and nivolumab plus or minus ipilimumab. This is where we have data. The initial data that we saw very early on were pembrolizumab based, and that generated a lot of excitement. A lot of people now use pembrolizumab as 1 of the defaults in the later-line setting in MSI-high colorectal cancer.

We have seen data here and at the ASCO GI conference [Gastrointestinal Cancers Symposium], at ASCO [American Society of Clinical Oncology Annual Meeting], and at other conferences that the combination of nivolumab plus ipilimumab in a first-line setting in 45 patients actually, data presented by Heinz-Josef Lenz, are very intriguing. There was high response rate, high durability of response. The median overall survival, the median progression-free survival isn’t reached, after more than a year and a half, 2 years, so that is intriguing. And I believe the trend will go toward using immunotherapy in first-line treatment.

In my clinical practice, I always try to treat patients the way I would like to be treated. If I had an MSI-high colorectal cancer, I would potentially like to see nivolumab-ipilimumab. The combination with ipilimumab is always a little looked at for a higher adverse-effect rate, and so forth. But the combination they’ll use is actually a low-dose ipilimumab every 6 weeks, not every 3 weeks as in other standard dosing. So, I think this is very well tolerated and highly effective. Pending the results of randomized trials, I think this is going to be the way forward.

Transcript Edited for Clarity