Journal Review in Surgical Oncology: Gastric Cancer

Hi, everyone. Welcome back to another journal article review with your surgical oncology team. Let's get started with quick introductions. I'm Joe Broderick. I am a general surgery resident from Brooke Army Medical Center, currently in my research years I'm Galen Gift. I'm a third-year general surgery resident at Brooke Army Medical Center. I'm Lexi Adams. I'm now a staff surgical oncologist at Brooke Army Medical Center. Congratulations. I'm Dan Nelson. I'm an associate professor of surgery at the University of Tennessee Health Science Center, Chattanooga. Practice mostly surgical oncology. Tim Bruhn. I'm a surgical oncologist at Brooke Army Medical Center. Happy to have Lexi joining me as my partner fresh out of fellowship. Congratulations, Lexi. Today we're discussing a landmark trial in gastric cancer, the Matterhorn trial, published in New England Journal of Medicine in twenty twenty-five. Treatment for resectable gastric adenocarcinoma has followed a similar evolution as other surgical oncology diseases. Patients

with locally advanced disease would undergo upfront major life-altering resections, followed by intense cytotoxic chemotherapy regimens, which showed marginal benefit and high rates of recurrence. This treatment paradigm began to change with the Magic trial in the early two thousands, and then Flaunt trial in two thousand tens, establishing perioperative chemotherapy as the standard of care for these patients. We've discussed these trials in a previous BTK episode from twenty twenty-four, but let's do a quick recap of what they are so we can understand these going to the Matterhorn trial. Dr. Nelson, do you think you could get us started? Yeah, sure. So at the time of the Magic trial, the standard chemotherapy regimen for G junctional and gastric cancers was epirubicin, cisplatin, and 5-FU, or ECF, and had demonstrated improved response rates in advanced gastric or G junctional cancer compared to older regimens. But there was non-conclusive evidence that adjuvant chemo regimens improved survival compared to surgery alone at the time. Furthermore, the regimen was associated with significant toxicity. Five-year survivals were really

only about twenty to thirty percent at the time, with most patients succumbing to either local regional disease recurrence or distant metastatic disease. So the Magic trial was a first phase three randomized controlled trial designed to investigate whether ECF given before and after surgery, or what we call perioperative, would improve outcomes of operable distal esophagus, G junctional, and gastric cancer, as opposed to surgery alone. The theorized benefits of administering ECF perioperatively included possibly increasing the likelihood of curative resection by down-staging the tumor, treating occult metastatic disease, and decreasing distant metastatic recurrence. Galen, can you tell us what they found? Yeah, so the five-year survival rates were thirty-six percent for perioperative chemo versus twenty-three percent for upfront surgery. This study ultimately concluded that anyone with stage two or higher disease should get operative chemotherapy. Thanks, Galen. So this is really the first study in gastric cancer that showed the benefits of perioperative or neoadjuvant approach,

as only about fifty to sixty percent of locally advanced gastric cancer patients were able to receive adjuvant chemotherapy after surgery. It really highlighted how difficult it was to get therapy after you've had curative resection. But 90% of patients in the study were able to complete their neoadjuvant therapy, which is really kind of talking about dose density. These patients are able to receive more of their planned therapy. And obviously, it showed that there was improvement in systemic control and long-term outcomes. But in reality, the survival rate achieved even by this study still left room for improvement, and now we know that ECF is no longer really the first-line therapy we use today. So what's changed, Joe? So then we had another trial, the FLOT4 phase 3 RCT, which introduced perioperative fluorouracil, leucovorin, oxaliplatin, and docetaxel, and compared it to perioperative ECF or ECX, where the fluorouracil was replaced with capecitabine. And this was in resectable stage two or greater

gastric or GI junction adenocarcinoma. And FLOT won pretty clearly. The median overall survival went from about thirty-five months in the ECF/ECX group up to about 50 months with FLOT. They also had higher rates of complete or R0 resection and higher rates of pathological complete response. So with this one, FLOT really became, and still is, the gold standard chemo backbone for these patients. Thanks, Joe. So that's where we've landed. Perioperative FLOT w- was, until recently, the standard of care for resectable gastric and GI junctional cancer. But here's the problem that set up today's trial. Even with FLOT, about half of these patients would still recur. So the question becomes, is there anything else that we can do or impr-improve on with FLOT therapy? Of course, the natural place to start in 2025 and 2026 is looking at the role of immunotherapy, because checkpoint inhibitors are already a part of how we treat metastatic gastric cancer, and that's really the question that drove the Matterhorn trial that set out to

answer. Lexi, can you take us through this trial? Yes, absolutely. But before we get into the trial itself, Joe, can you remind us how these drugs work, and specifically, what is durvalumab? Yeah. So durvalumab is a checkpoint inhibitor, specifically an anti-PD-L1 antibody. The short version is that these tumors can put the brakes on the immune system by engaging these checkpoints, which basically switches off the T-cells that would otherwise attack the cancer. Durvalumab and other checkpoint inhibitors block these brakes, so the immune system can recognize and go after the tumor. And like we talked about in our melanoma podcast, checkpoint inhibitors can be given in the neoadjuvant setting, the theory being that there's more antigen before resection to prime a stronger and more durable immune response. Yeah, exactly. So, Dr. Nelson, how do you frame this move, immunotherapy coming into the resectable curative setting? So I would say that the Matterhorn trial that we're discussing today is an example of a broad, well-established trend

across oncology over the last several years, in which immune checkpoint inhibitors first approved in the meta-metastatic setting have been moved earlier and earlier into the treatment sequencing, such as neoadjuvant or perioperative setting, where the goal is really still curative intent and no longer palliative. Melanoma is a clear example that you all have discussed previously. Lung cancer, colorectal cancer, biliary tract cancer, and HCC are some other examples. But the results aren't uniformly impressive. In MSI-high tumors, immunotherapy has been truly transformative with producing dramatic durable responses. But in the unresected gastric and GI junctional cancer, that really hasn't necessarily been the case. Here, immunotherapy is really acting as an additive boost on top of the chemotherapy. It's not replacing FLOT, it's adding to it. And we have to remember that FLOT is an intense cytotoxic regimen with less than 50% of patients being able to complete all allocated cycles.

So any benefit from disease control that we're gonna achieve with the use of immunotherapy has to be weighed against the added toxicities that are, are gonna come with it I would just pipe in and say that it's interesting because I think in the early days of immunotherapy, the thought was that it should never be combined with chemotherapy because chemotherapy suppresses the immune system. So why would you wanna give something that's knocking the immune system down at the same time that you're giving immunotherapy? But that paradigm has shifted, particularly in GI cancers, and that's, you know, could be for a couple different reasons that we probably won't get into. It's just interesting that now more and more, particularly in the world of GI cancers, you're seeing a combination of immunotherapy and chemotherapy at the same time. Part of that may be that we're not ready to give immunotherapy alone to a GI cancer that's not MSI high and doesn't have a high PD-L1 score, and we're just afraid that in the neoadjuvant setting, we're not gonna be giving effective therapy. So we're trying to give effective chemotherapy with the immunotherapy boost so that those patients who

do respond to immunotherapy will come out in the trials, and down the road maybe we'll have better ways of figuring out who benefits from what. Potentially, the cell death from the immunotherapy causes some antigens to be released, and then that stimulates an immune response. People have tried that with radiation therapy. People are trying with histotripsy now. Anything that is killing cells, you wanna add that, the immunotherapy, with the hope that you're causing a bunch of antigens to be released and causing a big immune response. So it's just an interesting thing that now we're doing that, whereas, you know, just for the residents understanding, in a cancer that is clearly already immunogenic, such as melanoma, we're not gonna give chemo and immunotherapy. Also, chemo just doesn't seem to work very well for immunotherapy. So where you're seeing immunotherapy alone is mainly gonna be in cancers that are already immunogenic, whether they be MSI high or have high PD-L1 scores or just cancers that are known to be immunogenic. Where you're seeing the combination is these cancers where

immunotherapy alone has not been all that impressive, and so I think people are a little afraid to give immunotherapy alone in the neoadjuvant setting, and there's also this kind of additional theoretical benefit of giving both. Thanks for those points, Dr. Nelson, Dr. Breland. So now with all of those points in mind, let's go into the Matterhorn trial itself. Joe, do you wanna s- give, give us the design? Yeah, sure. So Matterhorn was a phase three international double-blind placebo-controlled trial of patients with resectable gastric or GE junction adenocarcinoma stage II through IIIA comparing perioperative nivolumab and FOLF to FOLF and placebo. So in other words, does adding nivolumab to the standard of care improve outcomes for patients with stage II through IIIA gastric adenocarcinoma? All patients received standard perioperative FOLF and were randomized to either perioperative nivolumab or placebo. Thanks, Joe. Patients were included if they had a good performance status and no prior treatment, and

excluded if they had peritoneal or distant metastatic spread. Galen, what were their endpoints? Yeah. So the primary endpoint was event-free survival, and then secondary endpoints were overall survival and pathologic complete response, i.e., the rate of no residual tumor in the specimen at surgery. And as a reminder, event-free survival is defined as the time from randomization as opposed to disease-free survival, which is typically defined from the time of surgery. So Galen, can you start walking us through the results? The final analysis population was about 950 patients across 20 countries, split evenly between the two arms. About two-thirds of tumors were in the stomach and one-third at the GE junction. Roughly 95% of patients in both arms completed their neoadjuvant therapy, and about 75% of patients in both arms started their adjuvant therapy, but only half the patients in each arm completed it. Surgery was delayed in about 10% of patients in each arm and canceled in less than 1% of patients.

So almost all patients completed their neoadjuvant therapy, but only half of the patients completed their adjuvant therapy, which is consistent with prior gastric cancer trials despite adding durvalumab. So what were the findings from the study? The conclusions from the initial trial results was that adding durvalumab clearly improved event-free survival. So at two years, about 67% of patients on durvalumab were event-free versus about 58% on FLOT and placebo alone. So a meaningful durable separation in favor of adding immunotherapy. The pathological complete response rate backed that up. It nearly tripled, so about 19% of patients on durvalumab plus FLOT had a PCR versus 7% receiving placebo and FLOT. So more patients were going to the OR and had no residual invasive tumor cells in the surgical specimen. So for the primary endpoint, that's about a nine percentage point absolute improvement in event-free survival at two years, and one in five patients had a complete pathological response

as compared to fewer than one in ten in the placebo arm. And overall survival did trend in favor of durvalumab, roughly 76 versus 70% at two years. But this was a secondary endpoint interim analysis, and that difference was not yet statistically significant. The survival data just wasn't mature enough to draw a definitive conclusion, so all we have at this point is that it delays recurrence. This past October, however, they did announce at the ESMO meeting, the European Society for Medical Oncology, their final trial results, and this demonstrated a three-year overall survival of 69% with durvalumab versus 62% in placebo. Yeah, and, and the hazard ratio for death was.78, so a 22% reduction in the risk of death at a median follow-up of 43 months. But there's something about the survival curves worth pointing out. They don't separate right away, so they sit on top of each other for the first year and only start to diverge after about 13 months. In the original New England Journal of Medicine paper, they

even stated that the proportional hazard assumption for the overall survival was violated. The hazard ratio was.99 for the first 12 months, meaning no difference, and then.67 after about 12 months. And- Yeah, so I think that's probably the most interesting thing about this trial so far, and I'm not allowed to speculate too much into data that's not well-published and everything yet. But the cool thing about that, and the thing you hope comes out over time, is that chemo's helping everyone for about a year or two, and that's kind of what we see in gastric cancer and pancreas cancer. Like, you give good chemo, and patients will be disease-free for somewhere on the order of a couple years. But then often, you know, Dan, Alexi, I'm sure you've seen this, like, you do a good surgery, you have what you think is a reasonable pathologic response, and you're kind of high-fiving, and then a year later the patient comes back with recurrence, and it's like, "Ugh." Then they got their adjuvant chemo, it probably helped, but then that sort of

wore off. But, you know, a real response to immunotherapy is much more durable because the immune response sticks around. And so maybe this separation is the subset of patients who are really responding to the immunotherapy. And so the question over time will be trying to help pick out and predict which of those patients are gonna have that durable response from immunotherapy rather than the more short-lived response from chemotherapy. So that's the question is, like- Are we just giving chemo and immunotherapy and everybody's responding to chemo for some time but not forever, and then some subset of patients are gonna have really durable response to immunotherapy and probably didn't need the chemo in the first place? Or is there some additive effect of giving these two classes of drugs? You know, I think more data is needed to sort of figure that stuff out, but the cool thing is that hopefully that extra, you know, however many percent of patients, I'd have to do the math, but whatever that subset of patients that seem to respond

specifically to the immunotherapy, that those responses may be very durable. So we've-- now, now we know immunotherapy has demonstrated a benefit in these resectable patients, and, like, we also have talked about melanoma. So Galen, was there anything else from that ESMO update that is specific to us surgeons? So two things. First, most patients in the durvalumab arm were node-negative at the time of surgery. It's fifty-eight percent compared to forty-five percent, an odds ratio of one point seven two. And then second, the benefit of adding durvalumab held up regardless of partial versus complete pathologic response. Hazard ratios range from point two nine to point six zero but with overlapping confidence intervals, so there's no real gradient there, but we do see that benefit of durvalumab. So it sounds like the durvalumab may be down-staging nodal disease, and that benefit does hold up even in the patients who already had a complete path response, which kind of comes back to what Dr. Vreeland was saying about the curves. If

the tumor is already gone from the specimen and the durvalumab is still helping, that would fit with the immune system doing work on micro-metastatic disease rather than just shrinking the primary. Can you tell us about the safety data? Is this going to come up at the expense of increased adverse events? So durvalumab didn't add much in terms of severe toxicity compared to plot alone. The rate of grade three or four adverse events and deaths from adverse events were about the same in each group, seventy percent and five percent respectively. And the addition of durvalumab did not delay surgery or adjuvant treatment. The main difference is more immune-related adverse events with durvalumab, which is exactly what you would expect from a checkpoint inhibitor. So all pretty genuinely promising. So what are the limitations we should keep in mind? So everyone got durvalumab both before and after surgery. So the trial can't tell us whether the benefit comes from the neoadjuvant part, the adjuvant part, or both. There's also the biomarker question. PD-L1 and MSI status

didn't clearly identify who benefits, so we don't yet have a great way to select out these patients. And like a lot of these trials, there are generalizability questions. Only 19% of the trial was enrolled in Asia, you know, and, um, where the disease occurs the most commonly in Eastern Asia, and Black patients were underrepresented. I mean, I think that there's a number of questions. It's an exciting trial. It changed the standard of care. A bunch of questions come up. First of all, I've always wanted somebody to do a trial of total neoadjuvant in gastric cancer, particularly for those patients that need total gastrectomy, right? So I think we talked about that with FLOT. Like, if you need a total gastrectomy, my preference is to try to get you through all of your chemotherapy before surgery, and I think the same is true here. Now, if all they were getting were immunotherapy, would that be as big a deal? Maybe not. But it's just really hard to tolerate the back end of FLOT after a total gastrectomy. And so 50% of the patients in this trial, and this is kind of true across these trials, didn't get adjuvant chemo, and that's a mixed group of total and

distals, right? So after a distal gastrectomy, I don't worry so much. Total gastrectomy, I just think it's the rare patient who bounces back so quick that they can get back to their adjuvant therapy quickly. So- I've never seen it, honestly. Yeah. I, I- You know, being able to tolerate FLOT after a total gastrectomy is- And so with a medical oncologist, what they do, right, the patient comes in and they look terrible, and they've lost 50 pounds, and they're like, "I'll give you capecitabine." And there's, like, no data for that. It's just they're afraid to give FLOT, so they give some kind of 5-FU-based something. And so, yeah, I typically try to get the medical oncologist to give more FLOT before we go if the patient's doing well. You know, I know the data says four, but try to give, get a little more. So I think that's always an open question in my mind. The other thing that this has brought up that nobody knows really what to do in these situations, and there are a couple, like, small phase two things, but now that we have 19% past CR What if we do an endoscopy after their FLOT four and durvalumab,

and like rectal cancer, there's nothing there but a scar? Am I really doing a total gastrectomy on somebody who has no residual disease, no, you know, that I can find? And so, you know, the standard of care is still to do the operation, but I would be hard-pressed to not watch that patient. You know, it's kind of like a low rectal cancer, you know. Do you have to do an APR? Now we have really good data for watch-and-wait. Now, that's mixing things. I get it, chemo rads versus this and that, you know. But i-if somebody's getting the immunotherapy response specifically, and they're potentially cured of their disease, and I do a life-altering total gastrectomy, it doesn't feel good. I think that's a really important question that needs to be answered is, is are these patients who have a complete clinical response, have a biopsy of their scar, and it's negative, do they still need a total gastrectomy? If it's a distal, I would definitely still do it. If it's a total, I would, you know, it's a hard conversation to have. I think those are important things

that will come up. And then with all the immunotherapies, we went over MEDINA in melanoma, and the interesting questions are how much more therapy do they need? If you've had ninety percent of your cancer cells killed, do you need more immunotherapy? If you had a PACR, are you done? Can you just go on with your life and not get anything else? These are all open questions for sure, but I think they're very interesting and exciting questions to ask. I mean, gastric cancer not too long ago was one where we were just sort of threw our hands up and knew that the patients weren't gonna do well, so just the fact that we're developing new therapies is really cool. And the really interesting thing in gastric cancer is how fast things are changing. We have done an episode on magic versus FLOT in our short time as podcasters, and now we're doing one on the next first-line therapy for gastric cancer. So as much as people probably think we talk about medical therapy too much on a surgery podcast, you know, this is exciting stuff if you're a surgical oncologist because the standard of

care is changing and giving patients real hope, and it's really changing people's care and keeping them alive a lot longer. Thanks, Dr. Vreeland. Do you have any other thoughts about how this all fits in, Dr. Nelson? Yeah. I mean, Tim made some great points, and I agree with him. The Matterhorn isn't the first attempt to add immunotherapy into a perioperative treatment protocol. Some other trials that we'll include in the show notes include the Keynote five eight five and Attraction five. So those are the two examples. Matterhorn stands out because it did combine immunotherapy with the standard of care chemo, in this case, FLOT, and that's really what kinda broke some thresholds with event-free and overall survival. And like Tim said, is now the new standard of care. To pick up on some of Tim's points and also what Galen brought up regarding biomarkers, in this trial, it probably didn't sort things out cleanly because-- which is worth understanding because ninety percent of the patients in this trial were

PD-L1 positive, but only five percent were MSI high, with another thirty percent whose MSI status was unevaluable or missing. But to that end, one question I have for Tim that you kinda brought up is, you know, there's this recent phase two trial out of France that looked at perioperative nivo and ipi-- epilimumab in, uh, MMR-deficient GE junction gastric cancer patients, and they had a pathologic complete response rate of sixty-two percent. So how are you incorporating that into your practice? Yeah. Yeah. I mean, I think if I have a, a MMR-deficient or MSI high patient, I don't even think they should get Matterhorn. They should just get immunotherapy. And so I think in those cases, push the medical oncologist to, to start with just immunotherapy, and then, you know, they're gonna have a, a response, most of them. I shouldn't say everybody. But-- And that PCR rate of almost two-thirds, I think, is real. And so, you know,

then the question comes up of, what do you do with that, with that? Uh, well, well, I shouldn't say PCR, but if they have a clinical response, and you don't operate, you know, there are some small trials that support the idea of not operating and watching closely and just like rectal cancer. So I think we-- It's a little bit difficult when you're not at a place that's doing this stuff on protocol to feel good about doing it because it's not standard of care yet. But I think it's very interesting. I mean, Lexi, you're a brand-new grad. You're gonna be seeing patients in clinic as soon as your EMR gets opened up. But, but yeah, I mean, let's say you have a patient who they're MSI high, they have gastric cancer, they get, you know, whatever, pick your poison, nivo/ipi. They come back three months later, scope shows nothing but a scar, biopsy is negative. What are you gonna do? I think they very reasonably can go under a watch and wait protocol with scopes every three months, and I think there is some Japanese data to support that, uh, right now as well.

Yeah, I, I would do the same. And I think particularly in that situation, I think you have a leg to stand on whether MSI high or MMR deficient because it's borne out across all these different cancers. I think the Matterhorn question is more difficult. If they're not M- MSI high, and they get Matterhorn, and then they have a complete clinical response, then what do you do? Is it the chemo that gave you that clinical response? Do you trust that 7% or the other 12% that's from the immunotherapy? Which one is it? And if it's just from the chemo, do you feel comfortable watchful waiting? Or do you only feel comfortable if it's from the immunotherapy? How would you possibly parse that out? I don't know, but total gastrectomy is a really hard operation to tolerate, and so I think being able to avoid it in patients is a good thing if we can do it. We never wanna study our way out of a job, but I think, you know, I think helping, you know, if we can avoid it in those patients, I think it'd be good. Thanks, Dr. Vreeland and Dr. Nelson.

So let's do a quick recap of the papers we touched on today just to remind you of what we talked about. So MAGIC back in 2006 established that perioperative chemotherapy beats surgery alone. FLOT4 in 2019 gave us our current chemo backbone, FLOT. And then today's trial, Matterhorn from 2025, showed that adding perioperative immunotherapy with durvalumab improves event-free survival and overall survival without any increase in severe toxicity. That's all we got for the day. Thanks for joining us on Behind the Knife. Until next time, dominate the day