Navigating Adjuvant Therapies in Early-Stage Esophageal Cancer: The Role of IO and MRD Testing

A roundtable discussion, moderated by Nataliya Uboha, MD, PhD, discussed the current treatment considerations for gastric cancer, GEJC, and ESCC, as well as relevant clinical trial data from the 2025 American Society of Clinical Oncology Gastrointestinal Cancers Symposium. Dr. Uboha was joined by Zev Wainberg, MD, MSc; Suneel Kamath, MD; and Sunnie Kim, MD.

In the third segment of the roundtable series, the panel covers ctDNA’s emerging role in surveillance for early-stage esophageal and gastric cancers and updates on first-line treatments for metastatic gastric adenocarcinoma, including biomarker-driven therapies and the recent approval of tislelizumab.

View the next segment on Exploring Tislelizumab and Zolbetuximab: Immunotherapy and Targeted Therapy Options in Gastric Cancer.

—

Dr. Uboha: Do any of you incorporate ctDNA in the monitoring of patients with early-stage disease in your practice?

Dr. Kim: I’ve been using it for patients who actually had a pathologic CR, patients with esophageal adenocarcinoma who got the chemo-radiation pathologic CR. Because there’s no approval to give adjuvant nivo there, I get really nervous. I’ve had people recur in that situation. So I’ve actually done ctDNA monitoring in that setting, MRD testing. I haven’t had one yet to become positive, but I probably would have a discussion about the adjuvant nivo because I think it’s a data-free zone, unless you all know any newer data that’s come out about that population?

Dr. Uboha: I think it’s a data-free zone. Although we’re not going to be using chemo radiation much anymore in adenocarcinomas, so it’s probably more applicable to squamous cell tumors going forward.

What about you, Dr. Kamath?

Dr. Kamath: I haven’t been much, but I think recent things that I’ve seen in the last, honestly, weeks to a couple of months have convinced me that I should start. Especially tumor-informed assays here are critically important, especially when you’re in a space where there’s not a lot of data about it. At least having things rooted in a tissue-based signature is important if you don’t have a lot of evidence to support that. But yes, I definitely think it has a role. The few series that I have seen definitely seem to be quite predictive. Maybe not quite as good as the colorectal data, but it still seems like if you’re negative, you’re likely not to recur; if you’re positive, a large number of those patients are going to recur.

Dr. Uboha: Right. I’ve incorporated it as a surveillance tool because of its prognostic value. I don’t make any treatment decisions based on the ctDNA test, but I will move up surveillance scans if somebody’s ctDNA turns positive. I wonder whether you use it the same way as that one, whether you’re not using it.

Dr. Wainberg: Yeah, I probably use it the same way you’ve described, as an additional surveillance thing.

Dr. Uboha: A tumor marker, but better.

Dr. Wainberg: Yeah, it’s like a super expensive tumor marker.

Dr. Uboha: Tumor marker. That’s good. Well, let’s move into the advanced disease stage. We started to talk a little bit about how our treatment changes in early stage may affect first-line treatments. But that first-line treatment of metastatic gastric adenocarcinoma is a busy space with a lot of updates, a couple of recent approvals. Maybe we can summarize just the kinds of agents we use in advanced disease, and then talk about some of the more recent approvals in 2024.

Dr. Kamath: Starting with HER2-positive disease, if the PD-L and CPS is one or higher, we incorporate trastuzumab and pembrolizumab with doublet chemotherapy. And then with the PD-L1 story, if the CPS is positive, I know this is an ongoing debate. I think there’s a decent argument for adding a PD-1 inhibitor to doublet chemotherapy. I tend to use it with CPS 5 and higher. The 1 to 4 range is controversial. Of course, there’s that PD-L1 heterogeneity, not just within the primary versus the metastasis but also temporally; the PD-L1 CPS can change too.

But the majority of my patients do get a PD-1 inhibitor with doublet chemo. And then with zolbetuximab, the newest FDA approval, if the Claudin 18.2 expression is greater than or equal to 75% of the cells, then I’ll add zolbetuximab as long as the PD-L1 score is low, definitely with zero, but ongoing conversation about that too.

Dr. Uboha: Again, it brings up the importance of having these biomarkers available at hand so you can make the best treatment decisions.

We have a new anti-PD-1 agent recently approved, tislelizumab. Zev, would you like to maybe summarize some of the data with tislelizumab and maybe some factors that distinguish this anti-PD-1?

Dr. Wainberg: I think that the study—it’s been published now for a year or two—and it’s Rationale 305, randomized patients to chemotherapy alone or chemotherapy plus tislelizumab. It really looked at PD-L1 scoring slightly differently, which is TAP scoring, which is more of a numerical rather than counting or more of a, I should say—

Dr. Uboha: Chemo area positivity visual.

Dr. Wainberg: Thank you—there you go. “Visual” is the term. So maybe easier for the pathologists in some ways, but it does correlate very well with CPS score across the board. If you look at all of these studies, it lines up pretty well, so five is five, and one is one. So I think the study met its endpoints and appropriately got FDA-approved just in the last few weeks of 2024 with chemotherapy. So just another option that I think lines up very, very nicely.

If you look at these three studies across the board with CheckMate 649, Keynote 859, and now Rationale 305, you see the hazard ratios are almost perfectly identical. The long-term survivorship is very similar. So there’s a nice reassurance to the field when you have reproducible data sets that are really showing the same results. So I think there’s a pretty nice expectation of benefit from what you might achieve with tislelizumab.

Dr. Uboha: Tislelizumab was approved for patients that have tumors with a PD-L1 score of one or greater. I think it was actually nice that it was approved regardless of the testing method, but it was more restricted for patients with PD-L1 positive tumors.

Dr. Wainberg: Dovetailed with the ODAC, right?

Dr. Uboha: Right.

Dr. Wainberg: So when you had the ODAC and the tislelizumab approval, it corresponded with what I think most people thought was the direction the field was going in anyway. The ODAC reaffirmed that and aligned with most guideline recommendations, that the PD-L1 score should be at least 1%.

Dr. Uboha: Taken into account, right, and I wonder whether we’ll see some updated labels for other anti-PD-1 agents as well.

Dr. Wainberg: Yes, why hasn’t it happened yet? It’s interesting because it’s been a few months, but we’ll see.