Mutations May Explain Outcomes Based on Sidedness in Colorectal Cancer

Colorectal cancer (CRC) outcomes can be affected by the side that a person’s tumor is on, with research showing that patients with left-sided CRC have superior results, explains Arturo Loaiza-Bonilla, M.D.

An analysis of the phase 3 80405 trial comparing survival outcomes of left-sided versus right-sided CRC discovered that the median overall survival was nearly 14 months longer in patients with left-sided tumors.

Researchers are now trying to determine biological and genetic explanations for the sidedness phenomenon. Researchers at the University of Pennsylvania (UPENN) conducted a retrospective review of next-generation sequencing data of CRC samples. The results showed that BRAF and CTNNB1 mutations were more prevalent in right-sided CRC, potentially contributing to the variation in survival outcomes.

What was the rationale behind this study of left- versus right-side CRC?

In an interview with CURE, Loaiza-Bonilla, a medical oncologist specializing in gastrointestinal malignancies and an assistant professor of Clinical Medicine at the University of Pennsylvania, discussed the UPENN analysis and the potential role of BRAF and CTNNB1 mutations in the survival difference between CRC patients with left- and right-sided tumors.The most recent data that has been presented regarding the sidedness of CRC has been a subject of much investigation. We know that a patient with right-sided CRC is now, regardless of the kind of chemotherapy you give to those patients, going to have a different survival.

The survival for patients with right-sided CRC is going to be worse compared to the patients with the left-sided CRC. That led to the question, “Is there anything beyond just sidedness that [causes] that potential outcome?” To answer that question, or at least to be hypothesis generating, we decided to use the samples that we had at the University of Pennsylvania Center for Personalized Diagnostics. We have a panel of about 38 genes that are targetable or assessable in samples from patients who had advanced CRC. And we stratified them between right-sided and left-sided CRC.

What were the findings of this study?

When we look at that data we're looking for specific mutations or genomic alterations that may justify why those patients are not doing as well, not just because of the location of the tumor, but also if there was any molecular underpinning that may explain the reason as to why these patients had a different outcome. We have over 250 samples from our center. We found that among the different genomic alterations, there were two that were the most common ones and were statistically significant. The first one was the BRAF mutation and the second one was the CTNNB1 mutation, which is a ubiquitin mutation. What we know is the BRAF mutation typically tends to have a worse prognosis compared to BRAF wild-type CRCs.

This is hypothesis generating because it may explain to some degree, why these patients having differences in locations may have different outcomes when treated with chemotherapy. Most of the phase 2/3 clinical trials that had been done, haven't done extensive next-generation sequencing as part of the biomarker testing. They don't do a complete analysis for a number of reasons.

First, is because those technologies were more costly at the time of the introduction of the clinical trial. Second, was because we're not sure which biomarkers were appropriate to test. Now that we have the capability of doing massive next-generation sequencing of a lot of our patients and there's a lot of in-house panels and also commercially available ones, we can help to prognosticate those patients in a different way.

Are there any next steps planned for this study?

What do you think the oncology community should take away from this study?

In the field of CRC, are there any other studies that you're particularly excited about?

This information that we have found in our patients may help to assist in prognostication and also for allocation in clinical trials. For example, if we've found BRAF mutations we can consider using targeted therapies for that subset of patients and it may also justify the reason why it's not only about side, it's also about DNA of the cancer and the genetics. I think we're not the only group that is doing this. There are several groups that are trying to understand the rationale of why these differences in survival are so prevalent. There are registries of tumors that have been in store for many patients undergoing clinical trials that could be used, as well, for next-gen sequencing. We’re hoping that what we describe in our patients correlates with future findings and that we can then have not only a prognostic but also a potentially predictive biomarker for those CRC patients. What I think this study shows is that it is not only about the side of the cancer, but we also have to look deeper. We have to look into the DNA of the cancer and realize that sometimes those molecular alterations make a whole difference for the treatment of those patients and how they're going to respond to different treatments, including a standard of care. For example, cetuximab or Vectibix (panitumumab), which are EGFR inhibitors, may have different responses depending on the BRAF division or lack thereof because of the mutation. I am very interested in the combinations of targeted therapies, along with immunotherapy. There is some recent data on the use of the PD-L1 inhibitor Tecentriq (atezolizumab), along with a MEK inhibitor in KRAS-mutated patients. We know there's a lot of evidence now that patients with microsatellite instability-high (MSI-H) tumors respond nicely to immunotherapy. But the patients who do not have that MSI-H profile do not respond well to immunotherapy. Now the combination of targeted therapy along with immune checkpoint inhibitors may change the treatment paradigm and provide better options for our patients. In those specific cases, it's very important to do next-gen sequencing, not only for KRAS but also for extended RAS and looking for BRAF mutations as they may also be targetable by these approaches.