Evolution of MSI Testing for mCRC

Transcript: Scott Kopetz, MD, PhD: As I mentioned before, MSI [microsatellite instability] testing has to be part of the initial testing for patients with metastatic colorectal cancer, and I would argue any patient with colorectal cancer. This has important implications for treatment, but also for screening for hereditary disease as well. We do know that the prior approach of only testing MSI in those patients who meet some familial criteria really did a poor job of screening for HNPCC [hereditary nonpolyposis colorectal cancer] or Lynch syndrome. Therefore, we do recommend universal MSI testing both in early stage as well as metastatic disease.

MSI testing can be several different varieties. The most common and probably the cheapest is the use of the MSI immunohistochemistry looking at 4 mismatch repair [MMR] proteins, the loss of any of which defines an MSI-high or deficient mismatch repair phenotype. It’s important to understand the terminology and the field has used these interchangeably: deficient mismatch repair is equivalent, at least clinically, to a microsatellite-high or MSI-high phenotype. There are other methodologies that can be used, including a PCR [polymerase chain reaction]-based test to really look for those microsatellites and quantify the instability that’s present. That requires a germline blood sample and so that’s a little extra effort to do, and not all laboratories are equipped for that testing.

Increasingly we’re seeing MSI calls on NGS [next-generation sequencing] panels as well. And I think that’s a reasonable first pass screen, if there are specific loci for MSI. I would not recommend trying to use tissue tumor mutation burden [TMB] as the primary methodology for screening for MSI, although that’s something that was previously suggested in other settings. But it doesn’t look like it’s sufficiently sensitive for that. I think the key for all of this is to make sure that at least some testing is done in all patients, and then if positive, that should be followed up by germline testing to rule out a hereditary syndrome, acknowledging most MSI-high patients will be sporadic and not have a hereditary form of the disease.

Richard Kim, MD: We definitely have biomarkers for immunotherapy, and that’s the MMR status or MSI status. We know that patients who are MMR deficient and MSI-high patients, those are the patients who tend to respond. However, the response rate is not 100%. Depending on the series you look at, it could be anywhere from 40% to 60%. So the question is, can we do better than that? They’ve looked at the PD-L1 [programmed death-ligand 1] status, and I think there’s no role for PD-L1 status, unlike in lung cancer, in colon cancer. And other thing that’s now slowly coming up is what we call tumor mutational burden. It is something that we typically do as a part of a big panel. It comes with a panel along with the MSI status.

Mohamed Salem, MD, et al looked at some of the large series of molecular profiling, and what they found in colon cancer is that the patients who have high TMB are most of the patients who are MSI-high patients. So I think that’s been established. They also found that the patients who were MSI-high patients tend to have more high expression of PD-L1. And the TMB seems to be a little bit higher on the right-sided versus the left-sided tumors.

Having said that, the question is, can TMB play a role as a predictive biomarker in colon cancer? There’s a small series that was presented last year from City of Hope National Medical Center. This is a very interesting study, about 22 patients only. However, this was a study where they looked at patients who got pembrolizumab for MSI-high tumors. They also collected the TMB data as well. And what they found was that the patients who were MSI-high but had high TMB burden are the patients who really responded, had longer PFS [progression free survival], and longer overall survival.

However, in the patients with MSI-high who had low TMB, their PFS was only 2 months, telling us that in addition to MSI status, adding TMB as a part of it can possibly narrow down the patients who benefit the most from immunotherapy. The biggest problem I think with TMB is, what is high TMB? That’s a cutoff that’s a little bit arbitrary. Dr Salem, in this study, used 17 mutations per megabase as a cutoff. But other study from the City of Hope used a higher cutoff of 37 to 41 mutations per megabase. So that still has yet to be worked out. But clearly, there seems to be a role of TMB in conjunction with MSI and MMR status to determine who the best candidate for immunotherapy is.

Transcript Edited for Clarity