Emerging Biomarkers in Metastatic Gastric Cancer

Kei Muro, MD: Until now, many multitargeted agents have been developed in the gastric cancer field. However, almost all multitargeted agents failed to produce an overall survival benefit. For example, MET; EGFR; HER2, except for trastuzumab; anti-angiogenesis, except for ramucirumab; STAT3 inhibitors, and so on. With FGFR, similar trials failed in overall survival. However, now we are focusing on patients with FGFR abnormality. In clinical trials, FGFR is very promising. Claudin 18.2 inhibition is also promising in its first trial. It was a randomized phase 2 trial, very good data. A newer agent, DS-8201a, also called Enhertu, is very promising for HER2-positive gastric cancer. It is also active in HER2-low expressing gastric cancer. HER3 is attractive; however, it is now under evaluation.

At this moment, it has very limited impact on the gastric cancer field. But now there are many attractive agents being developed in the gastric cancer field. Later, I hope NGS [next-generation sequencing] or the many-biomarkers direction becomes impactful on the gastric cancer field.

Daniel V. Catenacci, MD: In addition to some of the predictive biomarkers that we’ve discussed with respect to immunotherapy—such as MSI [microsatellite instability]-high status, PD-L1 status, and others—there are a number of emerging biomarkers that may render some predictive utility, both for immunotherapy or other targeted therapies. Speaking with respect to immunotherapy, EBV [Epstein–Barr virus]–positive tumors, which have extremely high expression of PD-L1, may derive some benefit. I think it’s very rare and it’s hard to get a robust data set around that, but there’s potential.

Another potential marker is tumor mutation burden. We know that MSI-high tumors have extremely high tumor mutation burdens, but other tumors that are microsatellite stable have high tumor mutation burden. I think that data are emerging and more studies are needed to really define whether tumor mutation burden independently will help to predict benefit from immunotherapy. For example, in microsatellite-stable patients who are PD-L1 negative, is that something that might be useful and help predict patients who might respond?

Outside of immunotherapy, there are other genomic biomarkers that are promising and have some data around them. MET gene amplification, EGFR gene amplification, and FGFR2 gene amplification are all receptor tyrosine kinases akin to HER2 gene amplification. We already know that in the first-line setting, anti-HER2 monoclonal antibody therapy with trastuzumab leads to improved outcomes when combined with chemotherapy. There are a number of smaller phase 2 studies that show consistent response rates—not all of them, but in a subgroup of each of those biomarker groups—that are worth mentioning.

In terms of biomarkers of prediction for anti-angiogenesis therapy, this has been elusive not only in this tumor type but across tumor types, where efforts over the last few decades have tried to find a predictive biomarker to help enhance or enrich benefit among those patients getting anti-angiogenesis therapy. But a number of biomarkers have been very elusive, and there’s no routine biomarker implemented to date to help enrich for patients who might benefit and spare others.

In fact, my sense is that there is a modest benefit across all patients. It’s not in 1 subgroup that you see a remarkable benefit and in others, not. It’s different than other targeted therapies where you don’t have the biomarkers. For example, without HER2 amplification, anti-HER2 therapy is not going to be effective. If you do have it, then you will substantially enrich for benefit. Anti-angiogenesis drugs work differently. They modify the tumor microenvironment, and so they work generally across all tumor types to a modest degree.

Transcript edited for clairty.