GI Oncology at ASCO 2022: Tumor Sidedness, ctDNA, and MRD

John L. Marshall, MD; Mark A. Lewis, MD


June 10, 2022

This transcript has been edited for clarity.

John L. Marshall, MD: Welcome to Medscape Oncology Insights. I'm Dr John Marshall, director of the Ruesch Center for the Cure of Gastrointestinal Cancers at Lombardi Comprehensive Cancer Center at Georgetown University.

Joining me is not only a very good friend, but also an amazing person in the world of GI cancer on so many levels — not only the data, but also the human side — Dr Mark Lewis. He's the director of GI Oncology at Intermountain Healthcare in Utah. Mark, how are you doing?

Mark A. Lewis, MD: I'm doing great, John. It's so exciting to be here with you. It's our first in-person ASCO meeting in several years and the excitement is palpable.

Marshall: I think the best part about being in person is you can sneak around the booths and steal stuff and not have to talk to the people at the booths. [Laughter] We'll talk about that later. We have quite a bit to talk about and we'll just focus on colorectal cancer.

When I first saw the list of abstracts, I thought, Oh my gosh, colon cancer has the number-one abstract in the plenary session. I thought there must have been some new drug I didn't know about. Something has happened. But then I looked at that title and I thought, This study has been done before, about 10 years ago, so why would it be number one? Please walk us through the study and what it showed.

Lewis: What did Yogi Berra say? It's déjà vu all over again. It's such a fascinating study because it was essentially asking the question — John, it's this paradigm that you're referring to — in the first-line setting in metastatic colon cancer, what is the best biologic agent? It was a Japanese trial.

We can get into specifics in a second, but I think the reason it evokes so many memories for us is that CALGB 80405, FIRE-3, and PEAK, all in their slightly different ways, attempted to answer the same question. It's striking that you would point out the lack of novelty of the agents because to me, I think the takeaway from today was that "sidedness" matters very much. I think now we're getting a hint about sequencing as well.

In this trial, the Japanese investigators took roughly 400 patients in each group, and they received either FOLFOX with panitumumab or FOLFOX and bevacizumab. The results were fascinating.

Marshall: RAS wild-type?

Lewis: All RAS wild-type. Exactly right. There was an interesting focus, and an understandable one, on left-sided tumors. There was a little bit of a change in schema over time where you could tell initially that the focus was going to be on all-comers. Then there was a focus on the left-sided tumors, which understandably enriched some of the response and survival data, and then an expansion to the entire group. The right-sided tumors were represented but, understandably, were a minority.

Marshall: Let me go back because I think with those original studies that were done, there were some parts of the world that believed the left and right; the US believed it but Europe did not. There were some parts of the world that didn't really embrace EGFR. We didn't, but the Europeans did. In some ways, we both were wrong, and we both were right in the end.

Lewis: I think that's extremely fair. For all the sophistication of our science, John, I'm constantly humbled to remember that one of the greatest advances in GI oncology in the past decade was this realization that maybe the right and the left colon are different. Embryologically, they're different. The junction between the mid-gut and the hind-gut is somewhere in the transverse colon.

It's really important to think about this in the stage IV setting. What I'm getting at is that a breast cancer oncologist would never treat one of their patients without knowing, at the very least, ER, PR, and HER2. As GI oncologists, I think in the metastatic setting, we have to think about sidedness. It's so interesting, and thank you for refining that this was the RAS wild-type population.

Dr Yoshino, in his conclusions, was advancing that for left-sided RAS wild-type, FOLFOX/panitumumab is now the first line. Dr Cremolini, in her refinement of that, said, well, listen, let's not forget about microsatellite stability or BRAF. I thought that was really important and gets down to the fact that we are looking now at taking a very common cancer, and we're not lumping — we're splitting.

Marshall: I went exactly where you're going. In breast cancer, they even have different rooms for ER, PR, HER2, and triple negative where we've all been in the same mosh pit and sorted it later. To me, today, this paper says we've got some distinct rooms now. We have the RAS wild-type left-sided tumors. We have all RAS mutations. They're their own space and actually the largest group at present, maybe 50%-60%. We then have the others: MSI, HER2, and BRAF, each in their own, having their own pathways, if you will. I think now we have to prospectively pick patients for trials based on what kind of molecular profile they have.

Lewis: Precisely, because again, a breast cancer trial would never be histologically agnostic, upfront, or seldom. The design would include this stratification, and I think we're going to have to do the same. Someone on Twitter was joking after today's paradigm presentation that, from now on, it's going to have to be left-sided colon cancer subspecialists and right-sided colon cancer subspecialists. I don't think we're there.

Marshall: We are there in Washington with left and right. Clearly, you're either one or the other. There is no transverse colon in Washington, DC, for sure.

This brings me to the fact that we have to do broad profiling in everybody in the front line because we need to know what pathway they're on.

For the second study, let's go to neoadjuvant therapy for what I thought was an incredibly brave study. It was exactly on point with a remarkable result that made it into The New York Times this morning. It was from Andrea Cercek's group in Manhattan, New York. They originally reported 11 patients. I don't remember how many they reported here.

Lewis: I think it was a dozen here, actually.

Marshall: Now, one more. Can you talk us through why the report on 12 patients is such a big deal?

Lewis: To see this already escalate to the level of mainstream press coverage in The New York Times is remarkable. I think it speaks to why even such a small sample size is reverberating.

Here, we were looking at single-agent anti-PD-1 therapy with dostarlimab in MSI-high, locally advanced rectal cancer, so stage II/stage III. The whole point, as you know, John, in rectal cancer has been how much can we do upfront that might avoid trimodality therapy. What I'm getting at — and I'm wading into very controversial waters here — is watch and wait.

What was so fascinating about this study is that by administering the checkpoint inhibitor upfront, they saw so many, like you say, 100% clinical complete responses. They actually did not advance to chemoradiation. They did not advance to surgery. We should be honest that, with still relatively short follow-up, all these patients remain in that remission.

It's absolutely remarkable, and especially to think about the MSI-high population. I gave a talk at this ASCO about the special burden that patients with germline cancer syndromes carry, including myself, if I'm perfectly honest. To see patients with, say, Lynch syndrome potentially be spared the rigors of trimodality therapy is just incredible.

Marshall: There are many things that are exciting about this. The investigators point out that there was a higher proportion of Lynch syndrome patients in these rectal tumors compared with, say, up in the colon, where you see more sporadic MSI-high, so there's that. In fact, we know that there are probably different proteins tracking with Lynch than sporadic, so there may be something there.

In the metastatic disease, we are not seeing 100%. We're seeing more like 50%. With this enriched, localized disease cohort, they went 12-for-12. What's the next study? To me, you're not going to randomize. You're just going to keep going and keep following and set the standard. Is that what you would do?

Lewis: That's exactly what happened. I actually had many people talk to me today and say, well, how do you do the randomized study now? I mean, essentially, have we destroyed equipoise? The fact that this has already reached the attention of the national news media, patients are so savvy that they're going to be aware of this.

Someone said on Twitter that it's going to be difficult to sit down in clinic this week and not have a conversation about an MSI-high locally advanced rectal cancer that doesn't involve this immunotherapy approach. We'll see what happens, but very watchful waiting. To me, the key is ongoing vigilance.

Already, this area of rectal cancer observation was fraught with tension. I think the tradeoff, what you are exchanging for, is potentially losing the surgery but really committing to very, very rigorous follow-up.

Marshall: These patients are going to be able to sit down more comfortably in clinic with this new approach if they have Lynch syndrome or MSI-high rectal cancer.

One of the tools I've been using to follow in watchful waiting is ctDNA (circulating tumor DNA).

We've had one main major publication out of that same Japanese group that just wowed us today. They wowed us in San Francisco back in January with some remarkable data in the adjuvant space. A different group with a different test, actually, came forward with their version of impact of MRD, or minimal residual disease, positivity or negativity in the adjuvant setting. Could you give us the high-level overview on that?

Lewis: I trained as a hematologist/oncologist, but for the past several years I've done only GI oncology. I have to admit some sour grapes when I saw my colleagues who were in the hematology space having all these MRD assays at their disposal. Now I think we have, increasingly, evidence that MRD can work in solid tumor oncology, and specifically, it can work in GI oncology.

Here, we were facing the very vexing clinical quandary of stage II colon cancer and adjuvant therapy decisions. John, I've often framed this through the very basic medical ethics problem of the trolley problem. This is how I first came to grips with adjuvant chemotherapy.

When I was first taught about adjuvant chemo as a fellow, I thought, Hold on a second. You want me to give chemotherapy to someone who has no visible signs of cancer? How can I do that? Then it was explained to me about micrometastases and I said, Okay, I can now do this and sleep at night.

We've known, or maybe not known, from QUASAR and MOSAIC that to benefit a small number of stage II patients, we had to treat a large number. That, of course, risks overtreatment. Again, thinking about pulling that lever and diverting away from a huge catastrophe, but you still know you're inflicting some harm.

What I'm getting at is using ctDNA negativity as a way not to pull the lever at all and potentially avoid the prescription of adjunct chemotherapy. It's a fascinating study. I'd be curious to get your take.

Yes, we still need clinicopathologic classification of these cancers. But I doubt that they're entirely sufficient for fully informed adjuvant therapy decisions. So I think this trial gets us one step closer to using ctDNA, especially ctDNA positivity, as a sign for who needs adjuvant, whereas negativity may be a signal that a patient does not require adjuvant chemotherapy.

Marshall: Yeah, that's my exact take. We are doing this increasingly. There are now some prospective studies being done across the world using different assay methodologies and different decision-making based on this. It's not a one-size-fits-all.

The whole goal, as you laid out, is to make us more efficient and quit giving treatment to people who won't benefit because they don't have cancer. Treat the ones that we're going to fix and have a reasonable idea without having to wait 5 years to find out if we were right or not, and instead to know early on that things are going well.

I think it's equally important to identify those patients that current treatments aren't fixing. Those are the ones who need to be on a clinical trial because — let's face it — we cured all the other ones. We just didn't know how we were doing it. Now we can really pull out that group.

Everything becomes more efficient because of the very high positive predictive value of a positive test. It may be that soon we'll be better at the negative, so that's cleaner; it's pretty good now, but it's not perfect. We know we're on the soft underbelly, but we'll get there one day.

It's been an exciting meeting. We need to get back out there and learn what's coming next. Thank you very much for joining today. It was a great discussion as always. Mark Lewis is, as I said and promised, just amazing, knows what he's talking about, and puts it in a context that's useful for everyone.

Thank you for joining us for this episode of Medscape Oncology Insights.

John L. Marshall, MD, is director of the Otto J. Ruesch Center for the Cure of Gastrointestinal Cancer; associate director for clinical care at the Lombardi Comprehensive Cancer Center; and professor and chief of the Division of Hematology/Oncology at Georgetown University, Washington, DC. A global leader in research for colon cancer and other GI cancers, he is the principal investigator of over 150 clinical trials.

Mark A. Lewis, MD, is director of gastrointestinal oncology at Intermountain Healthcare in Salt Lake City, Utah. He has an interest in neuroendocrine tumors, hereditary cancer syndromes, and patient-physician communication.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.