Findings from clinical trials evaluating HER2-directed and immunotherapeutic approaches for treating patients with gastric cancer are among recent data attracting interest from experts in the gastrointestinal cancer field.
Yelena Y. Janjigian, MD
Findings from clinical trials evaluating HER2-directed and immunotherapeutic approaches for treating patients with gastric cancer are among recent data attracting interest from experts in the gastrointestinal cancer field. The results of 4 studies presented at the 2020 American Society of Clinical Oncology (ASCO) Virtual Scientific Program were the topic of discussion when a panel of specialists got together for The Talk, a new OncLive® video program that features lively exchanges about practical oncology issues in a virtual format.
Moderator Yelena Y. Janjigian, MD, was joined for the roundtable by Jaffer A. Ajani, MD; Daniel Catenacci, MD; and Zev A. Wainberg, MD. The conversation revolved around the results of the following studies:
Janjigian launched the discussion by delving into the RTOG 1010 trial, which attempted unsuccessfully to establish a disease-free survival (DFS) benefit for adding trastuzumab to trimodality treatment.
Janjigian: Several important findings presented at ASCO involved targeting HER2 both in the locally advanced and the metastatic setting. Why do you think RTOG 1010 failed [Table 11]?
Catenacci: I think it was multifactorial: 1 aspect is just an inherent issue with the perioperative studies in this disease. It’s difficult to accrue, and now you’re limiting it to 10% to 15% of the patients who have HER2 amplification. So, inevitably, it’s going to be difficult to accrue to such a study. The target hazard ratio, I would say, is potentially unrealistic—a hazard ratio of 0.64 DFS.
Even in breast cancer, where trastuzumab was used in the adjuvant setting, the hazard ratio for DFS in a positive study was 0.77, in a disease where it works and works better than here. So, I think that to all of a sudden have a [statistically significant] hazard ratio of 0.60 was a little bit unrealistic. But again, that was just a numbers thing, and ultimately, this was a 200-patient study, which is small for a phase 3 study and more like a phase 2 type of study.
Wainberg: This was a complicated study; underpowered, perhaps, for what they were trying to achieve and overtly complex in some ways. But I don’t know if this means that adjuvant or perioperative trastuzumab should be abandoned.
Ajani: Another thing to consider: I don’t think we need to pursue this group of patients right now, because as everybody said, it takes forever to accrue. But I think a molecule like ZW25 [zanidatamab] or even, you know, DS-8201a [trastuzumab deruxtecan], could have been even more effective with a reduced dose.
Janjigian: Do you think there are enough data based on DESTINY-Gastric01 [Table 22,3] to consider trastuzumab deruxtecan in the third-line setting patients in the United States or Europe?
Wainberg: I think the study is compelling. It was designed cleverly, in such a way that it allowed patients in Asia to be randomized only to either single-agent irinotecan or paclitaxel, not ramucirumab; they had already progressed on that. I think before we introduce trastuzumab deruxtecan to patients in the United States, we’re going to want to see the ongoing data from this country, or the Western world, I should say.
I think the DESTINY trial had a couple of interesting things about it. One of them, obviously, is that they did not biopsy patients from enrollment.
They relied on HER2 archived tissue and yet ended up with a favorable result compared with chemotherapy. In the study being done now in the United States and in Europe, there is a requirement to do a fresh tissue biopsy and confirm HER2 status. So, if anything, we should have an absolutely HER2-positive patient population in our sample size.
So, I think we want to see at least some response rate or similar PFS [progression-free survival] data, recognizing that it may not be as good as the Asians’, before we use this agent, or until it gets FDA approval. That being said, the FDA these days is a little bit unpredictable, so we don’t quite know what will happen.
Janjigian: If you have a HER2positive patient in the third-line setting and you somehow magically have access to trastuzumab deruxtecan, would you give them irinotecan in third line in the United States, as opposed to giving them trastuzumab deruxtecan? You can’t say you’re going to put them on the trial. Let’s say there’s no trial available.
Ajani: This is a very active drug. There is no question about it. In the third line, you’re getting a 40% response rate. That doesn’t happen with any other treatment that we know; maybe ZW25 comes close. So that’s 1 really important thing.
But if you look at the toxicity table, in every important category, this drug has more events than chemotherapy does.
Catenacci: I would really echo Dr Wainberg’s point that we need to see the performance of this drug in the Western population, where there is more proximal disease, such as junction esophageal cancer compared with a majority distal gastric cancer: different cancers, different biology, different heterogeneity.
Janjigian: To subject our patients in the third-line setting to nivolumab [Opdivo] or pembrolizumab for PD-L1, where the responses are in single digits, or punish them with cytotoxic chemotherapy, when the responses and survival are measured in months, is not an ideal option.
I think realistically, there is such a disparity between Asia and the West that we will have to wait another 2 years for data. I’m not suggesting we use trastuzumab deruxtecan in first line or second line. But the truth is, in third line, you really don’t have anything to offer to your patients.
Janjigian: Are you sort of heartbroken and over targeting HER2 in the neoadjuvant or adjuvant setting? What would you think if you were to design a trial now?
Wainberg: For HER2 studies in the adjuvant setting, asking this question a little simpler might be better because of the challenges of enrolling this patient population and also recognizing that trastuzumab is going to be used a lot in the first line and in second line—or not just trastuzumab but also other anti-HER2 agents they’ll be exposed to over the course of their life. So, it’s going to require some degree of change in study design, I think, from the ones that we’ve been using to date.
Janjigian: I know that there have been strong feelings about approval of pembrolizumab for TMB [tumor mutational burden]–high population. What are your thoughts on the recent tumor-agnostic approval for pembrolizumab?
Wainberg: I think the FDA is sending us all a nice message that they like tumor-agnostic approvals, and that’s very clear. And that’s a very interesting and important point to keep in mind as we continue to do new drug development. These tumor-agnostic approvals are here to stay, I think. However, TMB is very confusing, and I would say that there is not still a standard, absolute best way to do this.
We saw at ASCO how, even within the same studies, different measurements of TMB are used, and they’re not absolutely the same. The way the FDA defined it, patients with a TMB of 10 or more [≥ 10 mutations/megab-ase] are eligible.6
If I have a patient in the office who has noncolorectal cancer, gastrointestinal, or gastric primary cancer and has a high TMB score, independent of CPS, I think this is something to consider now, whereas before we weren’t checking it. It’ll be interesting to see how quickly this gets adopted, because it’s not necessarily standardized just yet.
Janjigian: It certainly highlights the importance of NGS [next-generation sequencing] in these diseases.
Wainberg: We have to consider the fact that in the community practices, the majority of patients aren’t getting NGS testing at this point. So, are clinicians to begin ordering TMB on every single patient just because we have an FDA approval for second line? This is a particularly difficult question to answer, actually.
Janjigian: Most referrals that come to me from the community already have Foundation Medicine or 1 of the other commercial platforms done. Is NGS typically available to your patients outside of your academic center?
Catenacci: I’ve noticed, at least in the Chicago area, that over the years—and certainly now with FDA approval of the Foundation One platform and Medicare payments and many other commercial companies offering these assays—the majority of our patients we’re seeing as consults already have it from the local oncologists.
And so, I don’t think that that’s going to be a barrier there. Along the points that were mentioned earlier, it was sort of the same question as to when the MSI [microsatellite instability]–high pan tumor approval came out. There was a similar question: Do we have to check all our upper-GI cancers for MSI high? And the answer should be yes. I’m not convinced yet about TMB, and I’d like to see more data after excluding MSI high, which are all TMB high and with high CPS.
Is this still a residual benefit of those other TMB high cancers? I want to see a little bit more information before I jump at that one.
Janjigian: Could you discuss the subgroup of KEYNOTE-061 data (Table 34) and the Mayo data suggesting that pembrolizumab may work better in second- or third-line therapy even after pembrolizumab failures? How does that fit into your thoughts on sequencing of immunotherapy once you administer it?
Ajani: I think we have seen that phenomenon from very early on, but I don’t think the data are all that convincing at the moment. I’m not aware of this phenomenon in other tumor types where immunotherapy is used up front.
As you mentioned, pembrolizumab is not necessarily a very good drug in gastric cancer because most patients don’t respond, and so we should start selecting those who have high CPS or high TMB, and we’re not doing that.
I think it is fantastic because you’ve got TMB, you’ve got high CPS score, you have algorithms now with NGS that can figure out the neoantigen, whether they’re immunogenic or not. These things are coming our way, and it’s exciting that we will be able to select drugs based on that.
Janjigian: Do you use ramucirumab with any other regimens? What are your thoughts about AIO [RAMIRIS] data [Table 45] that were presented?
Ajani: I think it’s an important study because some people are using taxane up front, which they shouldn’t. Even the FLOT authors7 have recently published, saying, “Don’t do that in the metastatic setting.” A lot of times you have a patient coming off first-line therapy with significant neuropathy, and then you can’t use paclitaxel, so I think this study probably will become important. And I think I agree with the conclusion to conduct a phase 3 study.
Wainberg: Some clinicians use FOLFIRI plus ramucirumab in gastric cancer a lot sooner than we tend to do in the United States. There are already good safety data with this regimen in many cancers, including colorectal cancer and gastric cancer. I think it’s certainly a reasonable option for someone who’s been on platinum-based therapy and a taxane to use this regimen.
It’s an active regimen, but I agree with Dr Ajani—it’s going to be a challenge, and we need to see what the control arm is going to be as they initiate the randomized phase 3 trial.
Janjigian: Do we think this is a worthy phase 3 trial to run? It seems like we should be answering more innovative questions.
Catenacci: I completely agree. In my practice, I’ve been using FOLFIRI plus ramucirumab since, essentially, ramucirumab was approved years ago. And in fact, Dr Wainberg and I, and also Dr Sam Klempner, put our data together, and we already did a retrospective analysis. It should be noted that this regimen was added to the NCCN [National Comprehensive Cancer Network] guidelines as an option for second line in January.
So, this is an option. We have now randomized phase 2 data showing it’s at least equivalent, if not trending to better PFS. I agree with you: I don’t think it’s necessary to waste resources on this in the second line. I think it should already be something that is an option for patients. It’s common to have neuropathy after first-line platinum. And what are you going to do? It just makes a lot of sense to use FOLFIRI plus ramucirumab in that setting. So, for me, it’s a no-brainer, and it’s already part of my practice.