Management of Newly Diagnosed mCRC Driven by Molecular Markers

Wells A. Messersmith, MD

The average survival of patients with metastatic colorectal cancer (mCRC) hovers around 3 years, but by understanding where the tumor originated and the increasing array of molecular markers, clinicians have a better chance of optimizing outcomes, explained Wells A. Messersmith, MD.

“There are a whole host of molecular markers including KRAS, NRAS, and BRAF that go into making these treatment decisions,” said Messersmith. “There is also microsatellite status and HER2 down the line. In order to take care of a patient with CRC these days, you need to know the side [of the tumor], what line of therapy you're looking at, and the host of molecular markers to inform treatment decisions.”

In an interview during the 2019 OncLive^® State of the Science Summit™ on Gastrointestinal Malignancies, Messersmith, co-director of the Developmental Therapeutics Program, and director, GI Cancer Program, at the University of Colorado School of Medicine, discussed the evolving standards of care in CRC and the importance of molecular testing in guiding treatment decisions.

OncLive^®: How has the treatment of mCRC changed in recent years?

Messersmith: The biggest change has been the number of options we have to treat patients. When I started working in this field about 17 years ago, there were a very limited number of agents available. There were hardly any targeted agents and just a couple of cytotoxic agents. Now we have this huge armamentarium of different agents we can use in terms of oral and intravenous cytotoxics as well as a bunch of different targeted therapies.

The second change is how much better patients are doing overall. For certain subtypes, even in the advanced disease setting, survival is almost in the 40-month range. Obviously, we want to cure people, but if you think about where we were 15 to 20 years ago where most patients were living 6 to 12 months, we have come a long way.

What are some key trials that led to these changes?

The key trials that I discussed [during my presentation] were the first-line trials with bevacizumab (Avastin), which is an antibody against VEGF. There have been several different trials with various combinations, but overall, there has been incremental benefit across the board in the first-line setting. We also talked about some of the angiogenesis inhibitors in the second-line setting. In addition to bevacizumab, you can also look at ziv-aflibercept (Zaltrap) and ramucirumab (Cyramza).

Then we discussed the use of EGFR inhibitors in the first-line setting. I highlighted the CRYSTAL trial, but there are a whole bunch of different studies that could be discussed, including the PRIME trial with panitumumab (Vectibix). Then, we discussed how one would pick one agent versus the other, and discussed the information we need to make treatment decisions in the first-line setting.

One of the things that has come up in the last few years that was kind of unexpected is that right-sided and left-sided tumors have to be treated differently. We don't know exactly why that is, although we do know they have a different embryonic origin. The right side comes from the midgut whereas the left-sided colon comes from the hindgut. There is probably a very different microbiome between the ascending colon and the descending colon.

Could you discuss the importance of molecular testing in CRC and the markers that should be accounted for?

There's a whole host of molecular markers we need to test for in CRC. One of the first ones is microsatellite stability (MSS) or mismatch repair proficiency or deficiency. There, we're looking for the editing function in DNA to make sure that patients don't have an inherited defect in that editing [function, known as] Lynch syndrome. Patients with Lynch syndrome are at risk for developing multiple tumor types. We suggest that this should be tested for reflexively. Every patient with CRC should be tested for Lynch syndrome because if you pick up a case, you can treat an entire family. Here, you're looking at siblings, children, parents, cousins, second cousins, etc. You can really help an entire family by putting people who are at high risk for these cancers into a screening program.

For MSS, you're doing a polymerase chain reaction—based test and for mismatch repair, you're doing immunohistochemistry. Both of those are kind of 2 sides of the same coin that are trying to pick up either inherited or acquired mismatch repair. If patients have that, whether it's inherited or acquired, you're going to be thinking about immunotherapy. Unfortunately, this occurs in only about 3% to 15% of patients with CRC depending on stage.

The second thing we talked about was RAS status. If patients have a KRAS or an NRAS mutation, they don’t tend to respond to EGFR inhibitors, such as panitumumab or cetuximab (Erbitux). That's very important because the last thing you want to do is give a possibly harmful drug to patients who aren't going to benefit from it. Unfortunately, we don't have drugs that target RAS, but we know what not to do if someone has a RAS mutation, especially in the first-line setting. We discussed BRAF, which is a mutation in roughly 15% to 18% of patients with CRC, the most common mutation being BRAF V600E; those patients tend to have a worse prognosis overall. Increasingly, we have BRAF-directed combination therapies that are now in the National Comprehensive Cancer Network (NCCN) guidelines and other guidelines, so this is a special set of patients you want to pick up. Then there's some overlap between BRAF and microsatellite instability-high (MSI-H) patients. Many patients with BRAF mutations have a hypermethylation of MLH1, and, therefore, will have MSI even though it's not inherited.

HER2 is coming, but there's not as much data available as we would like to have. It would be good to know if a patient is HER2-amplified or -overexpressing because there are some interesting therapeutic strategies that can be used in that case. Larotrectinib (Vitrakvi) was recently approved for patients with NTRK fusions; they're very rare, but we can pick those up. There is a whole battery of tests that need to be done in CRC to figure out how to treat patients, especially as you get into the second- and third-line settings.

Are most providers aware of the tests that need to be ordered for newly diagnosed patients?

I've seen a shift for most providers where many of the tests are done. However, not all tests that need to be done are done. When I used to do second opinions a couple years ago, sometimes hardly anything was done, or if it was done, it was very outdated. For example, people would test for KRAS but not NRAS. You really need to test for both unless you have a KRAS mutation, as they’re probably not going to present together. Providers tested only for codon 12 and 13, whereas we now know we need to look at codon 61 and 146. They would often not test for BRAF or microsatellite status as well.

Now, more of these tests are being done. It is important to go back and look, especially if someone was tested a few years ago. There, you'll often find that the testing that was done a couple years ago really isn't state of the art now. You have to look at what targets, codons, and mutations were tested to be able to interpret the results appropriately. Sometimes, you have to send the test again for further testing to figure out what's going on.

Are there emerging biomarkers that could help distinguish between VEGF inhibitors?

Biomarkers for angiogenesis inhibitors are incredibly lacking at this point. We have drugs that hit VEGF, such as bevacizumab, or PlGF, such as aflibercept. We have agents that hit VEGFR-2, such as ramucirumab. All these agents can be used in the second-line setting in CRC. The problem is, with angiogenesis inhibitors, we just do not have markers to predict who is going to benefit and who is not going to benefit. Right now, it's completely empirical. There has been a lot of effort looking at that in terms of the soluble levels of these markers, the microvessel density, and the expression of endothelial cells. None of those trials ended up being positive, so at this point we're basically treating everyone with the understanding that there is probably some subset that benefits. We just don't know who they are.

Could you shed light on some emerging agents or approaches?

The main thing in 2019 is the various BRAF inhibitors. We know that using BRAF inhibitors alone results in several resistance mechanisms in which there is either no benefit or a very short-lived benefit. There's increasing interest in combining BRAF inhibitors, MEK inhibitors, and then a feedback loop with EGFR inhibitors. These are now in the NCCN guidelines and can be effective in these patient populations, but are not curative. That's a pretty big change in the guidelines. The other approach is combining CTLA-4 with PD-1 checkpoint inhibitors. That approach has been reported and has shown a pretty good response rate; that'll be important to keep up with. Those would probably be the main two, although I also mentioned NTRK, which is extremely rare at approximately 0.5%. However, if you don't look, you're never going to find it.