The Hunt for Predictive Markers in Metastatic Colorectal Cancer - Episode 8
John L. Marshall, MD: The Holy Grail for immuno-oncology is figuring out if there’s a biomarker. The drugs are expensive. We know that. They have some toxicities. We know that. They work really well in some people, a little bit in others, and not at all in some people. We need to figure out who should get them and who should not.
There are 3 leading contenders for biomarkers for immuno-oncology checkpoint inhibition. One that everybody knows about is PD-L1 [programmed death-ligand 1]. Do you have the receptor? Is the ligand present? That’s wishy-washy. Some studies say it matters and some studies say it doesn’t. We know that microsatellite instability [MSI], mismatch repair-deficiency, predicts for responsiveness to immuno-oncology therapy. And so, for all patients with metastatic disease, not just with colon cancer, we need to know their MSI status. Even though it’s fairly rare in some cancers, the drugs seem to work.
The third is something called tumor mutational burden [TMB]. The only way you’re going to figure this out is by doing broad next-generation sequencing. How many mutations does that cancer have? There’s pretty good prediction that the more mutations there are, the more neoantigens that are produced; and, therefore, the more likely that you’ll respond to immuno-oncology.
We’ve studied this. We’ve looked at the frequencies of this in colon cancer and other diseases and know that almost all mismatch repair-deficient MSI patients have high tumor mutational burden. But there are some patients with high tumor mutational burden that aren’t mismatch repair-deficient, or MSI. PD-L1 is actually a third piece of that. So doing broad next-generation sequencing is the way to go because you get both MSI and tumor mutational burden. They’re confusing. These are new biomarkers that we are not yet comfortable with. We do a lot of MSI testing using immunohistochemistry, which can be confusing to a lot of doctors. We’re now shifting to doing most of this testing using next-generation sequencing, as I’ve said, which will get you both MSI status and tumor mutational burden.
We know that right-sided colon cancer is different than left-sided colon cancer. We also know that when we look across the spectrum there are different percentages of gene abnormalities. A lot of our MSI tumors are on the right side, but that doesn’t mean that we shouldn’t check on the left side. It’s not exclusive to the right side, it’s just more frequent over there. It can happen anywhere throughout the colon, so check all of your colon cancer patients, whether they have left-sided or right-sided tumors.
Alan Venook, MD: There are a couple of important issues with MSI testing. One of them, of course, is determining whether the cancer’s a genetically informed cancer. That has relevance to the siblings or family members of the person who’s in your office. In particular, that’s important to determine in young people. There has been a stunning increase in the incidence of colon cancers in young people, which is very, very disconcerting.
The real value for MSI testing is to identify the subset of patients who have the greatest likelihood of benefiting from a checkpoint inhibitor. Because the benefit can be off the charts, it’s imperative to make sure we know the MSI status of every patient when they come through the door, whether they are being seen in the adjuvant setting with stage II or III disease, or in the advanced disease setting.
There are studies in each of these settings—a couple, actually—looking at employing the checkpoint inhibitor in the upfront setting with chemotherapy. This is worth considering seriously, as a study. Clearly, for patients with MSI-high disease, you have to make sure you offer them the opportunity to use the checkpoint inhibitor. The effects can be so dramatic, and you want to make sure you don’t miss that opportunity.
We check MSI immediately and we’re ready to pounce. If patients have toxicity or complications from conventional therapy, we immediately go to the checkpoint inhibitor. Interestingly enough, most patients are not willing to go on the trials because they’re wary about not being randomized to the checkpoint inhibitor and are concerned about the ramifications of that. So the studies have had a bit of trouble accruing because of that.
For some diseases, tumor mutational burden may carry some weight. In colon cancer, I’m not convinced that it really matters. In my experience, you either have a sky-high tumor mutational burden—like off the charts—or you don’t respond to the checkpoint inhibitor. So I think intermediate load or some sort of continuum—I don’t find that very useful. I may be proven wrong with more data. Or, if we can tweak the checkpoint inhibitors and get them to work in a different way, perhaps I will get renewed interest in that. Right now we do calculate the tumor mutational burden, but, for colorectal cancer, I’m not sure that it’s very helpful. It’s either very high or I don’t think patients get any benefit. That’s at least where it stands today, I think.
Transcript Edited for Clarity