Key Takeaways in the Treatment of Advanced Colorectal Cancer - Episode 6

Key Takeaway 6: I/O: Combination vs Monotherapy


Tanios S. Bekaii-Saab, MD: One subgroup of patients with metastatic colorectal cancer [mCRC] that essentially used to have fewer less options and worse prognosis than most patients is the MSI [microsatellite instability]—high. So it’s interesting. MSI-high colorectal cancer in the metastatic setting confers a poor prognosis, and in the early-stage setting, a good prognosis. Theoretically, the reason for that is what MSI-high dose is. It builds those mutations into tumors so you have a hypermutated tumors, which are very inflammatory, and the immune system keeps it in check. That’s why most of the MSI-high patients with early-stage cancers are actually stage II.

For the stage IV—I mean for diseases to metastasize when they are MSI-high—they have to really be aggressive enough to overcome the check of the immune system. For that, they may be actually more aggressive. Interestingly, those respond incredibly well to these immune checkpoint inhibitors. And we’ve seen that first with pembrolizumab, a study that essentially looked at MMR [mismatch repair] proficient versus deficient colorectal cancer, and found 0 response in the proficient. In the deficient, the response rate was north of 35% to 40%. In fact, with long-term follow-up, more and more patients actually went from a PR [partial response] to a CR [complete response], and more from stable disease to a PR. So that response continued beyond 20, 24 weeks.

In the first study, actually, the PFS [progression-free survival] was not reached and the overall survival [OS] was also not reached, and that led to the approval of pembrolizumab. And this wasn’t a study that included just colon cancer. It also included noncolon cancers with very similar results. But let’s focus on colon cancer. Then came another study, CheckMate with nivolumab or nivolumab plus ipilimumab. And I want to emphasize the OR [overall response]. There was a choice; there was not randomization. Oftentimes I think the study in the way it was published can be very deceiving to the reader because the curves were placed on the same graphs and figures, which essentially makes it very difficult. For someone who just looks at them without reading the context, they may think these were randomized. They were not, so they were picking essentially who goes on each arm.

Those studies did confirm again that nivolumab ends up having very similar value to pembrolizumab, historically at least in this group of patients. The ipilimumab plus nivolumab seems historically at least to have a higher response rate than nivolumab alone and a better PFS. We don’t know about the survival as much. Interestingly, Heinz-Josef Lenz of the University of Southern California Norris Comprehensive Cancer Center presented data with nivolumab and ipilimumab in the first line that also suggested that you can reach a response rate up to 40% to 50% with a PFS and an OS that were not reached.

What we know today is that we have the option of a PD-1 [programmed cell death protein 1] alone or a PD-1 inhibitor plus CTLA4 inhibitor. How do you choose the right 1 for your patient? We don’t have any randomized data. All we have are historical data. The assumption is always that more is better, which is not completely true. In fact, for about 30% to 50% of patients with MSI-high metastatic colorectal cancer, all they need is the PD-1 inhibitor alone.

There is a small percentage of patients who may benefit from the doublet, meaning the PD-L1 [programmed death-ligand 1] plus CTLA4. And we have to keep in mind that although ipilimumab was at the lower dose, it still confers added toxicities. But there’s also the significant addition of cost. We really have to be very judicious about who those patients are. In my practice, most patients will get the PD-L1 inhibitor alone, and typically it’s pembrolizumab because that’s the 1 I’ve used the most and the longest. That doesn’t mean nivolumab couldn’t be the one. They probably have equal or similar activity. They’ve never been compared, so we can’t say equal. But they have similar activity historically. Pembrolizumab—historically at least, when we look at the separate studies—seems to have a higher response rate. But I’m not sure that’s a fair comparison.

For some patients, I would choose nivo [nivolumab] plus ipi [ipilimumab]. Who are those patients for whom I would choose nivo [nivolumab] and ipi [ipilimumab]. Indeed, if response rate is indeed higher, at least historically again, and the PFS is more prolonged for some of these patients who are younger and have more symptoms, I may actually choose to start them with nivolumab plus ipilimumab. For all other patients, I start with pembrolizumab, and that’s going to be the majority of the patients. I start with pembrolizumab, and then if they have a good response, that’s all they’re getting. If they don’t have an optimal response, I may switch them to nivo [nivolumab] plus ipi [ipilimumab]. That’s going to be a small percentage of patients.

But for those patients frankly who will not respond to a PD-1 inhibitor, it’s very difficult to think that combination of the 2 will switch a nonresponder to a responder. Again, I hope these studies will be done. I know in melanoma we have some hints about the fact that we essentially don’t need to start with that intensified regimen. We can introduce that in the nonresponders later. But we need to do that in colorectal cancer to make sure we indeed would improve outcome if we use 2 versus 1. Today we don’t have that, and as I said, I’ll go with 1 and for a select few, we’ll go with 2.

Transcript Edited for Clarity