IDEA pooled analysis: Duration of adjuvant oxaliplatin-based chemotherapy for Stage III colon cancer

DR VENOOK: So this was called the IDEA, I-D-E-A — it was a pooled analysis of 6 trials done across the world in 6 different countries or parts of the world. Fundamentally, each of them looked at 6 versus 3 months of an oxaliplatin-based regimen. And as you said, the real key is, if you can get away with 6 cycles or 3 months of oxaliplatin versus 6 months, you’ll have much less neuropathy and make a big difference in patient quality of life. And really, there are almost no examples where you need that much therapy to cure people with adjuvant treatment.

So the design was, basically, leave the specifics to the different groups, so the 6 different groups, and then pool the data in a preplanned analysis. Dan Sargent, one of the giants of biostatistics in cancer, actually engineered all of this. Unfortunately, he passed away suddenly before it came to fruition.

At the end of the day, clearly I think there was very little difference in survival between patients who got 3 or 6 months. And when you looked at the subsets of patients — and these are big subsets — almost 12,000 patients — what you see is that patients with a low-risk disease, N1 or fewer than 4 lymph nodes, for example, not T4, those patients clearly did not lose any survival benefit from the adjuvant therapy if they got just 3 months of oxali. The higher-risk patients it’s less clear, and probably most of us would say, “Give 6 months of adjuvant oxali, but you back down when the oxaliplatin neuropathy begins.”

A much tougher nut to crack is whether the chemotherapy mattered. And some people looked at the data — and, in fact, it does appear in Europe that CAPOX patients did better than FOLFOX patients. And that’s really hard to square. In the US, we don’t use much capecitabine, because the dosing is more poorly tolerated in the US than in Europe. A lot of speculation why that is. So some people would say, “CAPOX is a better adjuvant treatment than FOLFOX.” Again, in this case, though, it wasn’t a randomized study, CAPOX versus FOLFOX, in any of these countries. It was dealer’s choice, physician’s choice.

So for all of the 12,800 patients or whatever that were included in this big consortium pooled analysis, at the end of the day I think we can say probably that patients with low-risk Stage III colon cancer do fine with 3 months of oxaliplatin. And those with higher risk should get more, if not up to 6 months of oxaliplatin.

Identification of patients to receive longer versus shorter duration of adjuvant chemotherapy

DR LOVE: So in your own clinical practice outside a trial setting, how do you define high risk, where you start thinking about 6 months?

DR VENOOK: Yes. So to me, one of the most important things is the percentage of nodes. So let’s say if you have more than 4 nodes, by the parameters they used in this study, that would have qualified as a high risk.

Now, the high-risk patients were a little underrepresented compared to the standard-risk or low-risk patients as we usually define them. And that’s undoubtedly because the doctors who were putting patients on the study, if they had somebody with 12, 15, 20 positive lymph nodes, they probably didn’t include them on the study because they didn’t think they could get away with less chemotherapy. That’s my guess.

But my view is that patients who certainly have fewer than 4 nodes, patients with a lot of nodes — let’s say 20, 30, 40 nodes with 1 or 2 positive — those are patients I would lean more towards 3 months. I think, at the same time, you can easily justify, in patients who are at 6 cycles or 3 months of oxaliplatin who have no toxicity at all, perfectly reasonable to say, “Let’s keep going until we start to get some semblance of neurotoxicity, and then we’ll stop.” I mean, I think it’s a continuum.

The study didn’t do what we do in practice, which is, you stop the oxaliplatin when the neuropathy persists between cycles, and then continue the 5-FU alone. That would have been way more complex, way more patients, but we would have at least addressed it the way we actually do it clinically.

Use of CAPOX versus FOLFOX adjuvant regimens

DR LOVE: So as you say, that all these studies and all the data were really complex, the discussions that went on about this. One of the things that came up that — I don't know, I was kind of surprised at, I didn’t realize that, I guess, the common CAPOX regimens, you get more oxali in earlier than with FOLFOX?

DR VENOOK: That is technically true. The oxaliplatin with CAPOX, it’s every 3 weeks, oxaliplatin. And it’s, I think, 135 mg/m² compared to 85 mg/m² in the every other-weekly oxali of FOLFOX. So over the course of the 6 cycles of FOLFOX or 4 cycles of CAPOX, which should constitute 3 months, you get almost the identical amount of oxaliplatin.

I mean, it’s pretty hard to know if — it is true that you front load the oxaliplatin. So let’s say you’ve got 270 mg of oxali within the first 3 weeks and a day, right? So you treat. Then you take 3 weeks off. Then you treat again. And so there’s a little more dose up front in oxali compared to the FOLFOX regimen. Whether that’s meaningful or not, whether the early hit of chemotherapy matters, if it does, maybe that makes a difference. It’s really hard to square that. And I don't know. It’d be hard to figure that out.

DR LOVE: In your own practice, do you use FOLFOX or CAPOX?

DR VENOOK: We use very little capecitabine. Again, the flaw in the whole analysis as it applies to the US patients is it’s well documented that capecitabine is not as well tolerated in the US, right? So in the studies with CAPOX, the dose that’s used of capecitabine is just not a dose we can use in the US when you combine it with oxali. So if you cut it by 100 or 150 mg/m², are you losing efficacy? We don’t know, but that’s our concern. So most of us do not use cape.

Now, though, some people look at this data and say, “This is evidence we should be using capecitabine” and push it. Perfectly reasonable to do. I myself worry about getting too much toxicity in adjuvant therapy and throwing out the baby with the bath water. I mean, if patients get too sick and can’t complete the therapy, then I don’t know if you’ve done them a favor.

Prognostic and predictive value of primary tumor sidedness

DR LOVE: So another very practical issue that really only has come on the scene relatively recently in colon cancer is the issue of sidedness. You’ve been involved with the research. There’s been a lot of work that’s been published in the last couple of years. Can you kind of summarize it and particularly focus on where it fits in from a practical clinical perspective outside a trial?

DR VENOOK: Right. So this is something that shouldn’t have surprised us, if you think of the embryology. Of course, you remember your embryology.

DR LOVE: Right.

DR VENOOK: We all do. I think I missed that class in medical school.

DR LOVE: I missed a lot of classes in medical school.

DR VENOOK: Yes, you and me both. The right colon, right side of the colon, comes from the midgut. And the left side comes from the hindgut. And they merge during embryogenesis. And the transverse colon — somewhere is right and somewhere is left of origin. So it shouldn’t surprise us that there’d be a biology that’s different, right versus left.

For many years, people had observed that patients with right-sided primaries don’t do as well, but the assumption was always, “The disease presents later in its course.” The stool is liquid on the right side. The colon is large, and it’s much less likely that you’ll get obstructive symptoms. And so people said, “You just have more disease because there’s a time lag.” And that could have been.

I don’t think any of us, really — I had no clue that this was as meaningful a difference as it turns out to be. I was motivated to look, (1), because we started realizing there were genetic molecular features that were more so on the right than the left, let’s say, BRAF mutations, more commonly on right sided. So it led some of us to think, “Geez, maybe there’s a difference.”

And then there’s an old study, published in 2001, of the 5 different regimens of 5-FU that, in fact, showed a 5-month delta in survival, left primaries better than right. That’s what motivated me to go in and look at every chart and figure out sidedness in retrospect, because we hadn’t collected the data prospectively. And I was flabbergasted that there was a 13- or 14-month difference in all the patients, left versus right.

And then as we started drilling down, it’s very clear that patients with right-sided cancers, independent of their RAS status, get almost no benefit, if any at all, from the EGFR antibodies. So the practical impact is, while I think it’s appropriate to do molecular analyses on our patients to really identify where they might head initially or certainly beyond first line, I think it’s clear that patients with right-sided primaries, you don’t need a RAS-mutated or not to know that you shouldn’t use EGFR antibodies on first line.

Whether you can use EGFR antibodies subsequently in patients with right-sided tumors is not so clear. Remember, all the studies that have really looked at this, except one, were really restricted to first line. And now how somebody could get benefit subsequent line but not first line I’m not quite sure, biologically. But we don’t know that these patients stand to get no benefit from EGFR antibodies. Our study clearly showed that the group that did the worst, by far, were the group that had right-sided primaries who got cetuximab in our study.

DR LOVE: When you talk about a 13- or 14-month difference, specifically in what?

DR VENOOK: So in overall survival. So in our study, in the whole group, if you look at all the KRAS, just the fundamental first analysis, which was just published in JAMA a few months ago, the KRAS wild-type group was about a 32- versus 19-month survival, left versus right, which is really stunning. I mean, that’s a big difference. And you’d think, like, what were we thinking that we didn’t realize this?

And what we further did at ASCO this year, Federico Innocenti and Heinz-Josef Lenz presented our data on looking at all the molecular features of the cancers. And, in fact, we found that sidedness is prognostic even when you incorporate all the other molecular information, BRAF, which, of course, patients with BRAF mutations have a poorer prognosis. Those tend to be right sided. MSI-high patients tend to be right sided. Those patients do better without metastatic disease but don’t do so well with metastatic disease.

Even when you add all of these factors in, it turns out sidedness remained prognostic. I think that just means we haven’t figured out yet what the biology is. It’s not like right versus left. There’s got to be some biological explanation. We haven’t gotten it yet, though.

DR LOVE: Now, I assume that the explanation is not related to the issue of EGFR antibodies.

DR VENOOK: Correct. I think it’s not. My bet is it’s got something to do with the stem cells, which would be the tissues of origin or something like that. The other black hole, the black box that we have no idea really where it’s contributing, but I bet it is, is the biome.

And clearly, there’s some evidence that the bacteria from the left colon to the right colon is different. And, I mean, it’s astonishing, if you really dig into this literature, we see that you can find bacterial DNA in mets in the patients with liver, liver mets. I mean, how bizarre is that? So I can’t explain it, but that’s probably a factor.

DR LOVE: So from a practical point of view, other than just the prognostic implications of this and also EGFR antibodies, which, at least in the United States, I think, are not very commonly used first line anyhow, what is the practical implication? Do you think that in a patient with a right-sided lesion, maybe you think about triplet therapy, for example, more often, even without BRAF?

DR VENOOK: Yes. That’s a good question. I think yes, probably so. That’s how we treat most of our patients these days. When you get the new patient, it may be 2 to 3 weeks before you’ll get the molecular analysis. So even if you have a right-sided patient, you don’t need to know their RAS status unless, if they’re RAS mutant, then they don’t have a BRAF mutation. The two of those don’t go together. If they’re RAS wild type, you’d like to know BRAF and be poised to move them on to something else, because those patients may do reasonably well first line but really don’t do well down the road.

And, as you know from the TRIBE study, from Italy, clearly, patients with a triplet, FOLFOXIRI, do better with BRAF mutations than if you treat them with the doublets. So as a default, if you don’t know the BRAF status in patients, especially if they have a lot of disease, we tend to err on the side of triplet therapy up front. And then usually we drop the oxali in a month or two if they’re doing well.

Intestinal microbiota and colorectal cancer (CRC)

DR LOVE: You mentioned the biome. Can you talk a little bit more about what we’ve learned about that? You also talk about this fusobacterium nucleatum. I’m not sure I pronounced it correctly. Maybe just talk a little bit more about the biology of what we know about this.

DR VENOOK: Yes. I can take 2 seconds, and that’s about what I understand of it. You just summarized it all, right? So the Dana-Farber has done — the Harvard folks have done the best work on it, I think. We have got 3 different scientists working on the biome here at UCSF. It’s very clear that the microbiome has interactions with the immunologic system, for sure. And/or with drug metabolism, metabolism of toxins.

For instance, we think that capecitabine may be metabolized in some people by the microbiome partially, so that it may be activated in the gut before it’s even absorbed as a pro drug. So could that account for changes in toxicity, et cetera?

In terms of the way the disease has evolved, the data from the Farber suggests that patients with fusobacterium nucleatum, which tends to be dominant more on the right side than the left side of the colon, those patients tend to have higher risk of cancer and don’t do so well. Again proving that its cause and effect is difficult.

I mean, when you get into the nitty-gritty here, you realize it’s really even hard to sample the biome without thinking that you might change it, right? Because you put a swab in, does that change the environment? Or you have a patient who, of course, has had a bowel prep. Tell me about the biome in that patient. I mean, it’s ridiculously complicated, and it’s evolving.

The pathologists at Hopkins think there’s like a biofilm. So some people have bacteria embedded in the crypts in the bowel, and that could be related to the risk of cancer, because it may disturb the immune function, the immune cells’ function. It’s immensely complex. And again, we’re not even close to having a handle on it. But it’s some of the most fascinating research there is, I think.

DR LOVE: Speaking of Hopkins, I was going to ask you in terms of — I think you alluded to the possibility that somehow the biome is related to the immune system/checkpoint inhibitors, et cetera. What do we know about that?

DR VENOOK: Yes. So we know that there’s an interaction. We don’t know consistently what the makeup of the biome is to optimize the outcome, but it’s thought that that probably explains why not everybody with MSI-high colon cancer benefits from the checkpoint inhibitors but, more importantly, why even when patients benefit, many will develop resistance.

Treatment approach for BRAF V600E-mutated metastatic CRC (mCRC)

DR VENOOK: So the assumption is that the biome has something to do with resistance, somehow in the traffic of lymphocytes in the crypts, again, we’re starting to wave our hands when we start talking about these details, but immense opportunity if we could figure this out. As I said, a number of groups around the country have huge efforts in the microbiome. It’s not as if we haven’t known about this for a long time, but it’s just now dawning on everybody that it really is important.

DR LOVE: You mentioned the issue of BRAF. And maybe you can provide an update on where we are with patients with BRAF mutations and how, in your own practice, outside a trial setting, you factor this into your clinical decisions.

DR VENOOK: Yes. So Scott Kopetz presented the SWOG randomized Phase II study at ASCO this year looking at the use of a BRAF inhibitor in combination with chemotherapy or with an EGFR antibody and clearly showed — in this case with irinotecan and/or an EGFR antibody. Or in his case, it was a MEK inhibitor. I mean, clearly you see that combinations with BRAF inhibitors improve at least disease-free survival or progression-free survival, I should say, because these are metastatic patients.

It’s a different question what the survival impact is. We don’t know that there’s a survival impact yet. It’s not as simple as in melanoma, where patients can respond to monotherapy BRAF inhibitors even though it’s the same mutation, the V600E mutation. I should say, BRAF, there are many, many mutations. The one that we’re talking about here is the V600E mutation.

Now, the problem here is, if you treat these patients, let’s say, with FOLFOXIRI up front, even if you get a good response, what do you do next? And these folks can go down really rapidly. The disease can turn on a dime. So it’s a real struggle, because for the data that supports either dabrafenib or vemurafenib in combinations, various combinations, neither of those are before the FDA, nor are the studies robust enough to go to the FDA.

There are a few studies out there right now that are ongoing that may serve to go to the FDA, but they’re not even close to done. So it’s a predicament. You can look at the molecular features of a patient’s cancer and say, “Gee. You’ve got a BRAF mutation. That’s bad,” and then what do you do about it, unless you can get them on a study? And that study may very well randomize patients so they don’t receive the RAF inhibitor. And this is really an ethical, moral problem for many of us.

So part of the question is, can we somehow shake loose BRAF inhibitors so we can put them into patients? It’s no panacea, but at least some patients have a much better outcome.

DR LOVE: So then assuming you can access a BRAF inhibitor or maybe in combination with other targeted therapies, do you offer any type of targeted therapy to patients outside a trial who have BRAF mutations?

DR VENOOK: Guilty as charged. So yes. In general, I would never have said — that’s not something I would suggest, except that with BRAF-mutant patients, I think it’s we’re stuck. So we do.

We have petitioned the sponsors, pharmaceutical companies, to get the RAF inhibitors. Insurers, on occasion, have paid for it. And we have treated some patients off study. Much preferable would be to either get these approved or to get them, let’s say, insurers, to pay for it. It’s a need population. It’s not a huge percentage of patients, right, it’s only about 3% or 4% of patients with metastatic colon cancer.

DR LOVE: Yes, but, I mean, 3% to 4% of colon cancer is a lot of people.

DR VENOOK: Big number. Absolutely right, big number.

DR LOVE: In lung cancer, they treat subsets that are 1%.

DR VENOOK: Totally agree. Totally agree.

DR LOVE: And actually, it’s interesting, mentioning lung cancer. We’ve watched the BRAF story evolve there. Initially, it wasn’t clear which way it was going to go, more like colon or more like melanoma. Now, I mean, there’s approved combination, BRAF/MEK. And when you ask investigators, they’re using it first line.

DR VENOOK: Yes.

DR LOVE: So that’s kind of following the melanoma pattern.

DR VENOOK: Yes. It’s fascinating. I think that’s one place where I could see moving these agents up front, especially patients with such a bad disease. You wonder if the RAF inhibitors are only modestly effective because patients have already had enough mutational events and other problems with their cancer that you’ve already induced resistance.

I mean, what we’re doing and some groups are doing really aggressively is doing serial biopsies of patients with colon cancer, because we have to understand why and how patients develop resistance if we’re ultimately going to make a big difference, I think. So BRAF, it’s really great that we’ve identified it. Now, if we can really be aggressive in targeting it and get things on the market that can help us do that.

DR LOVE: And just to be clear, when you talk about trying to access these types of agents, you’re talking about single-agent monotherapy with a BRAF inhibitor?

DR VENOOK: No. So for colon cancer, the monotherapy — at least dabrafenib and vemurafenib have essentially no activity as monotherapies. So at our place, we work with combinations, with EGFR antibodies and irinotecan, other combinations with MEK inhibitor, which is commercially available. So we mix and match. I mean, it’s not scientific, but essentially we just feel compelled to at least give patients a chance at a good response or at least controlling their disease reasonably well if they have a BRAF mutation.

Prolonged response to T-DM1 in a patient with rapidly progressive HER2-amplified mCRC

DR LOVE: So speaking of targeted therapy of colon cancer, as long as we brought BRAF up, why don’t we touch a little bit on HER2? It’s kind of been an interesting evolution over the last few years. And I also want to ask you about your paper in Journal of the National Comprehensive Cancer — JNCCN — looking at a patient who got T-DM1.

DR VENOOK: Yes.

DR LOVE: Really interesting. Can you kind of provide an overview of where we are with HER2, amplified and mutant, for that matter, colorectal cancer?

DR VENOOK: Yes. So it’s also a small subset, 3% or 4%. And I think it probably took us a while to figure out that these folks could benefit. I remember there was a study maybe a decade ago when trastuzumab was first approved, maybe 20 years ago. And we saw really no activity in relatively unselected patients with colon cancer.

So it matters. Sidedness probably matters. Gender may matter. There are a lot of variables with HER2, ERBB2 and the way to affect the receptors. We see probably 3% or 4% of patients who are either HER2 amplified or HER2 mutant when we do our big assay. And this is one of the few advantages of centers that have very large comprehensive next-generation sequencing assays.

So we have what’s called the UC500. So we get genes that you don’t need to know about, you would assume. But really pretty surprisingly often, we were seeing HER2 amplification or HER2 mutations. And, in particular, we had a patient who was actually a physician scientist from Germany whose insurance in Germany was perfectly willing to give us whatever he wanted.

So when he came, he had rapidly progressive cancer, just really primarily refractory disease, which you almost never see. And we jumped in with T-DM1, and he had about a 7- or 8-month — a minor response and then stable disease, which at least steadied things a little bit so he could deal with what was going on, really pretty striking. And he was HER2 amplified.

Interestingly, we rebiopsied him when he progressed. And he had lost the HER2 amplification. Whether it was a sampling difference or actually the disease had developed resistance or morphed away from the HER2 amplification we don’t know. But it’s a nice opportunity, really, because we have a number of drugs available for it, not common, but something that there are studies now looking at bringing this into early use in the subset of patients with HER2-amplified or -mutated colon cancer, again, 3% or 4% of patients.

DR LOVE: Just out of curiosity, I guess T-DM1 usually is well tolerated, no alopecia. At least we know from breast cancer, sometimes, hepatic problems, thrombocytopenia. How did this man do?

DR VENOOK: He did incredibly well. We combined pertuzumab with T-DM1.

DR LOVE: Wow!

DR VENOOK: And he did really almost remarkably well. As you know, we wrote up the case in the JNCCN, which looks for the exceptional responders kind of stuff. This is not average, I’m sure. But he did really remarkably well, given the tempo and the course of his disease, although not for as long as we would have liked. He tolerated it very well.

DR LOVE: What do you think would be reasonable, if anything, to consider outside a trial setting with HER2-amplified colon cancer? There was a trial. I think it was looking at trastuzumab and lapatinib, if I’m remembering that correctly.

DR VENOOK: Right, the so-called, I think, HERACLES trial.

DR LOVE: Is that a strategy you would use outside a trial, or T-DM1?

DR VENOOK: I think it’s not crazy. I don’t think lapatinib’s a very good inhibitor. It’s hard to get the dose in that’s adequate that’s enough to lead to suppression. But there are a few studies looking at HER2 combinations going on. Again, this is another one of those things that it will be interesting to see if any company is going to go forward with a big enough study to get it approved. I mean, there’s very little reason for companies to go through the millions of dollars and the man hours necessary to get a drug approved for another indication when they’ve already got a bunch of indications already settled.

So this may be another example where we’ll have to somehow work around whether FDA approval exists or not. But I think in HER2 we need a little more data to make us confident. That’s another example where moving it up front may be of benefit. And I suppose in the adjuvant setting it might be of benefit in patients who are HER2-positive. I actually don’t know the likelihood. I don't know how many patients who present with less than Stage IV disease are HER2 amplified or mutated. Maybe somebody knows that, probably, but I don’t.

DR LOVE: Do you think that T-DM1 would be a reasonable nonprotocol option for a patient? I mean, if they got your paper here.

DR VENOOK: Yes. I think it is. I mean, that’s obviously why we did it. And it’s all things wrapped up in one. It’s an antibody conjugate, really pretty well tolerated, at least in our patient, so yes. I think these are interesting questions.

The problem for all of these questions is when you’ve got a couple percentage of patients and they’re getting many lines of therapy. And some of these patients, it’s one thing to do a study in breast cancer with HER2 targeting, because everybody knows to get HER2 analysis. Patients, the average bear with colon cancer in the community, may get — if we’re lucky these days, hopefully they’ll get at least all RAS and MSI status and BRAF. But they’re not going to get HER2. And so doing this, really moving it in force into patients earlier will require a lot more molecular testing than goes on right now.

EGFR tyrosine kinase inhibitors in patients with HER2-mutated mCRC

DR LOVE: You mentioned HER2 mutations. And we’ve heard a lot about that, again, in lung cancer, not so much in breast cancer. And there they talk about using TKIs. Now you have neratinib approved as postadjuvant therapy in breast cancer. But I’ve heard people talk about it with relation to HER2 mutations. What do you do if you find somebody has a HER2 mutation?

DR VENOOK: We’ve not had the same success. At least, we’ve had a modest success in the amplified patients. We do the same thing. We’ve tried a variety of the things you’ve talked about. In general, the TKIs, the EGFR TKIs are very poorly tolerated in patients with colon cancer, at least. Whether — they have terrible bowel problems for the most part. I don't know if it’s got to do with distribution of the receptors or whatever. But we’ve not had nearly the success with mutated patients as we have with amplified patients.

Effects of diet and lifestyle on outcome in colon cancer

DR LOVE: So I want to ask you also about work that you’ve been involved with. A lot of it came out of Dana-Farber, also looking at the issue of physical activity and nutrition in colon cancer in terms of progression/disease-free survival, for example. Can you talk about globally what you’ve learned from this research?

DR VENOOK: This is, I think, the brilliance of the group at Dana-Farber who, about 20 years ago, really pushed and pushed and forced the cooperative groups to include very detailed questionnaires in the study called 89803, which was the adjuvant colon study that looked at irinotecan. In the US, Len Saltz led that study.

And the study was absolutely negative. There was no difference whatsoever in outcome across the irinotecan versus 5-FU alone arms. In fact, there was much more toxicity with irinotecan. But in that study, they really had very detailed information on diet, exercise, lifestyle habits otherwise, how much TV you watched. I mean, you can’t believe the wealth of data. And analysis after analysis has shown that with that data set, it’s hard to explain exactly why this would be.

In fact, these patients with Stage III colon cancer, they were likelier to not recur if they had a variety of good habits and more likely to recur if they had bad habits. Now, you could say, “That’s because maybe these were patients who were couch potatoes before they got their colon cancer, and that’s why they did so badly.”

It doesn’t seem to be related to what their status was when they got the diagnosis. I think most of us really have always been troubled by the idea that diet or lifestyle could have that much impact on outcomes in cancer. And I think this data is pretty compelling.

In fact, one of the young women in my group here, Erin Van Blarigan, worked with the data set from 89803 and with folks at the Farber who are really quite expert in this kind of molecular epi, Jeff Meyerhardt, Charlie Fuchs. And she actually had one of the lead papers at ASCO this year, where they did an analysis in this same population and looked at what the American Cancer Society would call “healthy living,” which is, let’s say, a couple of drinks a day for a male, one drink a day for a female, no cigarettes, non-Western diet, so no In-and-Out burger kind of thing, maybe an aspirin a day, good exercise. Don’t sit and watch TV, all these factors that people say, “This is a good lifestyle, and this isn’t.”

And the patients who met the ACS criteria for good lifestyle had a far superior survival in this Stage III study than patients who didn’t, really pretty striking. Again, how do you implement that? I don't know. And there must be some molecular predisposition that we don’t understand, but I think very interesting data.

Recommendations for vitamin D supplementation

DR VENOOK: And just as I believe in vitamin D, I mean, we tried to do vitamin D as part of an adjuvant study in 80702. This was a dozen years ago. And the NCI and others rejected it.

And now, Kimmie Ng has data just with vitamin D in patients with Stage IV colon cancer getting chemotherapy with FOLFOX, with or without vitamin D. Patients who get vitamin D do better, at least in progression-free survival. So I think a lot of our assumptions that diet and all that can’t have that much of an effect, I think we’re just wrong. But it’s not so easy to separate out the factors, to understand what the main factor is.

DR LOVE: For practical purposes, what do you do in terms of vitamin D, both in the adjuvant and metastatic setting?

DR VENOOK: So I’m a little bit loony about that. We check everybody’s vitamin D and we replete vitamin D. I just fundamentally believe. And it’s easy enough to do, and especially in this era where everybody’s afraid of skin cancer and puts on lots of sun block, those are people who tend to be vitamin D insufficient. I think it’s said, I don't know. I mean, maybe what we consider a normal vitamin D may not be correct, the value at least. Eighty percent of people are said to be vitamin D insufficient. And so I routinely replace vitamin D and really try to get it up to normal.

And there’s talk of a randomized study that would do that. I mean, the challenge, of course, is that once people hear about vitamin D, like why wouldn’t they take vitamin D? It’s not like taking chemotherapy. But, practically speaking, we have a huge effort in exercise in our patients with both early-stage and advanced cancer as well as the nutritional things and supplements like vitamin D.

DR LOVE: And you tell your patients, “I’d like you to try to do this, because it maybe will help in terms of the disease”?

DR VENOOK: Right. I mean, I think patients really are empowered. I think it’s really important for patients to at least have some control over what they’re doing. And so being able to make decisions about vitamins or other supplements, I think it’s really valuable. And what do you know? It may actually be helpful.

Counseling patients about increased physical activity

DR LOVE: You mentioned some of the broad issues with a, quote, healthy diet. What level of exercise? I know the data suggests it’s not that necessary to be that intense or for that much. What do you actually tell your patients?

DR VENOOK: Right. So I can’t remember the exact criteria in the paper that the group that looked at the exercise in 89803, I think it was the equivalent of mowing a lawn, a big lawn, once a week or, let’s say, walking 2 miles at a leisurely pace during a week. So I just say what the contrary, the other side of the coin, is. If you sit and watch TV all the time, you don’t do so well. So our advice is to just take 3 walks a week. Let’s say you might want to walk at a mall or at a park where you can sit down periodically, that kind of thing. And I think it’s got to do with the number of calories you burn. And people think it may have to do with revving up insulin or using insulin. I mean, the mechanisms are curious.

But we do recommend exercise. In fact, the same woman who presented the data at ASCO, Erin Van Blarigan, we have a big number of studies that are looking at different ways of reminding patients to exercise. They have Fitbits^®. They get text messages, a variety of social media to — we encourage them. So “Big brother’s watching you. You’re sitting around. You’re sitting around. You should be exercising.” So we really encourage our patients to do that. And then when they come to our center, they have to park so far from the place, then they get some exercise in that regard, too. Just kidding.

DR LOVE: I think maybe not, but just out of curiosity, is it your observation or belief that exercise also helps with anxiety?

DR VENOOK: I really believe strongly that giving patients elements that they control, it is so important because much of, I believe, a lot of the anxiety of cancer patients, other than the horror, the uncertainty, is that so much is out of their control. So I really believe strongly in seeing what patients can have control over and letting them do that. I think it’s a very valuable thing.

Case: A 46-year-old man with right-sided primary MLH1-deficient cecal cancer and bulky nodal disease receives FOLFOXIRI/bevacizumab but discontinues because of toxicity and is switched to pembrolizumab

DR LOVE: So let’s spend the rest of the time talking about your cases, beginning with your 46-year-old man.

DR VENOOK: Okay. Yes. So the first case, this is a 46-year-old fellow who had a cecal primary and a lot of nodal disease in the retroperitoneum and supraclav, clavicular area. That’s kind of unusual to have node-dominant disease in colon cancer. But he had no apparent visceral metastases.

So the issue, of course, is, in these patients, do you take the primary out? In somebody who’s got this kind of bulky disease, we typically wouldn’t. He was started in the community on FOLFOX/bev, which is what 75% of the oncologists use for front-line colon cancer.

He came to see us. And this is an example — we talked earlier. This is a right-sided cancer, weird behavior. This is a patient we treated with — I recommended that they go to FOLFOXIRI/bev, which they did. And we got our molecular tests done. And it turns out that this was a patient who had MLH1 deficiency, which is to say he was MSI high. Otherwise, everything was wild type. So this was a difficult case, because here you’ve got a patient — you’re giving aggressive chemotherapy, and you know you might jump in with a checkpoint inhibitor at some point.

In fact, now there are checkpoint inhibitors that will be in front-line studies. So the challenge in this patient was, how long do you treat with the current chemotherapy and when do you jump in with the checkpoint inhibitor? And our approach in a patient like this is to get as much mileage out of standard chemotherapy as we can. Mind you, not everybody responds to the checkpoint inhibitors, even if they’re MSI high.

So I think this is one of the reasons you want to look at early information on the molecular features of the cancer. And again, a right-sided cancer is likelier to be MSI high and BRAF mutant, as well. And MSI high and BRAF mutations overlap about 25% or 30%, and those are patients who the BRAF may be an aberration. So this is an example of the patient we would continue aggressive therapy and then take the first opportunity to use a checkpoint inhibitor.

And, as you know, this is unprecedented. I believe this is the only circumstance — MSI-high patients are the only circumstance where drugs are approved based on molecular features and not agnostic to the site of origin of the cancer. So this is why it’s really important in right-sided patients, especially with bad-acting disease, to really figure out what the molecular features are.

DR LOVE: So just out of curiosity, again, outside a trial setting — or maybe you can tell me what kind of trial — if you had seen this patient up front and known he was MSI high from the beginning, what would you have done treatment wise, or would there be a trial you would have put him in?

DR VENOOK: Yes. That’s a really good question. At this moment, there’s not a trial available yet. They’re working their way through the regulatory situation. We don’t have one, at least. I think there may be one industry study.

It’s a tough call. We give the patients the option. Most patients, interestingly enough, opt for chemotherapy. We’ve had half a dozen patients like this. Most of them opt for chemotherapy. And then we may jump in later on, especially with somebody with a big bulky disease. You worry that if you don’t get on top of it, you may lose the patient. But I could see a time not too far away where we might start with a checkpoint inhibitor. And if it doesn’t work, then go to chemotherapy.

One of the dilemmas with immunotherapies is, (1), it’s hard to know if it’s working or not. And although it’s more so with the anti-CTLA-4 and other immune modulators. You worry about pseudoprogression or delayed response. And in a patient with bulky disease, you worry whether you have the luxury of waiting 3 to 4 months to see if the treatment works.

So we tend to still lean with chemotherapy. Again, the studies now are starting to combine chemotherapy and checkpoint inhibitors. And there’s no reason we shouldn’t do that. I think company sponsors were afraid, “Chemotherapy may interfere with the immune system.” If anything, I think chemotherapy could augment the effects of the checkpoint inhibitors, and so in combination may ultimately be where we get to. But that’s going to take a while.

DR LOVE: And, of course, much to a lot of people’s surprise, actually, the combination of chemotherapy and a checkpoint inhibitor, pembrolizumab, was actually approved in lung cancer, nonsquamous disease.

DR VENOOK: Correct.

DR LOVE: And a lot of people are using it. There was a lot of debate and controversy about it. This man actually, because he’d already gotten started on treatment, you switched him to FOLFOXIRI/bev, I guess because of the location of the primary being right sided. But even though he’s 46 years old, it looks like he had some problems with that.

DR VENOOK: Yes. So neutropenia, FOLFOXIRI/bev is not easy. And this patient got neutropenic fever and, I think, had a 4- or 5-day hospitalization. Again, he was cared for in the community, so this was as it was reported. Ultimately, he came back and we put him on pembrolizumab just at the time, because he also was losing weight. He didn’t look so good.

He’d responded, actually, but the chemotherapy had beaten him up. And he’s actually done remarkably well on the pembro. So, I mean, he had a good response, and it’s augmented the response so far. We’re not far enough into it to know how he’s going to do long term, but he’s done very, very well.

Rates of response and prolonged disease control with immune checkpoint inhibitors in mCRC

DR LOVE: What are the global figures that you have in your own head about response rate in these patients and particularly the issue of prolonged disease control more than, let’s say, a couple of years?

DR VENOOK: Yes. So there’s not much data on colon cancer yet. I mean, you hear about the melanoma population with — I don't know. I hear 10% long, durable remissions. I actually don’t know in colon cancer yet what the numbers are going to be. In our hands, we’ve seen it across the board, though, most patients in our studies, where — we have 32 studies, I think, in our GI group here at UCSF. And 20 of them have different immunotherapies.

And by and large, the responders are really robust responders and very durable. But it’s, I’d say, maybe 30 — and these are patients selected for features that you would expect them to benefit, let’s say MSI high in particular. Probably 1 in 3 are getting really good responses, but they are very often very robust and durable. And again, we have some exceptional responders.

A young fellow with hepatocellular carcinoma who’s almost 3 years out with a CR — I mean with bulky metastatic disease, with hepatocellular, on a checkpoint inhibitor who’s a complete remission, sustained, 3 years later off treatment. I mean, this is unprecedented, in my experience, to see people who literally almost rise from the ashes with these treatments. It’s phenomenal how some people get the benefit some people get.

DR LOVE: I was doing a few interviews a couple of weeks ago, 3 different cancers. I don’t even remember. I think it was lung, head and neck and maybe heme. And each one of those interviews had a patient who was about to go into hospice, who got a checkpoint inhibitor and whatever. Amazing story.

DR VENOOK: Yes. I mean, you’ve done this long — and you don’t see that. It’s almost to the point, like, how do you not offer patients this treatment? It’s astonishing that you can have such an amazing impact so quickly in patients who are otherwise just beyond hope.

Biomarkers for predicting the effectiveness of anti-PD-1/PD-L1 checkpoint inhibitors in mCRC

DR LOVE: Although the big difference, for example, in something like lung is, you have a lot higher frequency of response in MSI high. What’s your vision of the pathophysiology of why you see this in these tumors? What are the PD-L1 levels? What’s your concept?

DR VENOOK: I don’t think we understand — obviously, the slide that every one of us has suggests that the number of mutations is related to the likelihood of the checkpoint inhibitors working. The concept is that the more mutations, the more foreign proteins, then the better chance the immune system has of identifying one to attack.

It’s not that simple, obviously. In hepatocellular carcinoma, for example, the responses don’t matter — it doesn’t matter if they’re viral mediated or not. These are not highly mutated cancers. So it’s not as simple as that. And nor is it as simple, I don’t think, as PD-1 or PD-L1 levels, although that may also matter. I think at the end of the day, I think a lot of it has to do with the tissue, where the tissue of origin and the trafficking of lymphocytes into the tissues, whatever motivates the lymphocytes to be there or not.

And again, this is a black box. And I believe the biome may have a lot to do with this, actually.

So I, at least, am miles away from understanding it.

Spectrum of toxicities associated with anti-PD-1/PD-L1 checkpoint inhibitors

DR VENOOK: Now, at UCSF, in our cancer center, one of the things that we’re building is a program to study the toxicities of the checkpoint inhibitors and these immune therapies, because the other problem is they’re not a free lunch. While they can be dramatically effective, they can — like, an internist’s nightmare. We’ve seen everything from myasthenia gravis to thyroid storm to severe pulmonary toxicity. So these drugs are not without problems, but really, in some cases, dramatically effective.

DR LOVE: It’s interesting. What are you looking at in terms of toxicity? And do you have any experience? Are you gathering any data? I don't know how often it happens in patients with prior autoimmune conditions, Crohn’s disease, et cetera.

DR VENOOK: Yes. A great question. So for years we were afraid to look at those patients. In fact, it’s a mixed bag. Some patients have gone on these treatments with, let’s say, autoimmune Crohn’s, in particular, and had no problems. And others have had horrific exacerbations of their autoimmune diseases. So it’s not a one size fits all. Clearly, we don’t understand a lot of the moving parts.

I do think that we have certainly been loath to use these agents in these patients, but we’ve seen a few responses to the point that we’re now willing to do so. But it’s, again — I’ve had a patient go into renal failure. I mean, almost every organ has and will be affected by these treatments.

Phenotypes of microsatellite instability (MSI)-high colon cancer

DR LOVE: I was going to ask you one final question. What do you see in terms of the phenotype of MSI-high disease? I’ve heard people say that pathologists maybe suspect it under the microscope. This patient had this weird presentation with all these lymph nodes. What’s the typical phenotypes that you see?

DR VENOOK: Right. So the main phenotype that we see that should spring to mind is, right sided, but also mucinous, mucin-containing tumors. And you can look at — if you look at a scan, CT scan, it’s pretty obvious who has a mucinous colon cancer and who doesn’t. The mucinous tumors just look like gook. I mean, these are heterogeneous streaky tumors that — it looks like they have internal areas of necrosis. And that’s what brings to mind the MSI-high patient.

The other thing I should comment on is the MSI-high and BRAF-mutant patients tend not to have measurable disease or often don’t. In fact, our group had a paper — Chloe Atreya had a paper a few months ago with that characterization. And one of the things that we probably will realize is that most of the studies in advanced colon cancer have not included many of these patients, because most studies require measurable disease.

In our study, 80405, we did not require measurable disease. We had about 40 patients without measurable disease. Most of them were either BRAF mutant or MSI high. So these are classically, let’s say, ascites, malignant ascites or peritoneal implants or something like that, where you knew they had cancer but you couldn’t quantify it.

Case: A 65-year-old man with multiple recurrences of mCRC receives regorafenib followed by TAS-102

DR LOVE: So let’s move on and talk about your 65-year-old man.

DR VENOOK: Right. So this is a fellow who presented with Stage IV disease and had a resection of his primary, who initially responded to chemotherapy and then ultimately went to a liver metastasectomy and more chemotherapy. So he did not have FOLFOX initially. He went only about 6 months and then recurred and went back on FOLFOX/bev and ultimately went on maintenance capecitabine/bev, which we’re doing more and more.

One of the things that I’m always cognizant of is, when you start with bevacizumab, we tend to be hesitant to stop it because of the TML data, the idea that patients may progress more rapidly after you stop the bev. This patient ultimately went on a — he had had such a great response to irinotecan initially, we put him on a study which had one of these new irinotecan knockoffs, a pegylated irinotecan and an EGFR inhibitor, which — it was unclear if he got any benefit. He opted out after 2 months.

And at that point, we offered him other trials, but he went on the conventional route and went to regorafenib and then ultimately TAS-102. I mean, you can tell from our — as I’ve described, many of our patients here at UCSF, regorafenib and TAS-102 certainly impact — in some patients can have a positive impact. But we don’t have any hesitation about putting patients on a study rather than using either one of those 2 drugs, because we don’t think the upside is that significant.

This patient got tired and didn’t want to go on more studies. And he liked the idea of pills. And so we did that. And I would say that they’re not nontoxic. With regorafenib, we tend to start at a low dose and work our way up. And that may be a mistake. Maybe we’re underdosing patients. TAS-102 we’re looking at in a variety of combinations, especially with radioembolization, for example. But these are treatments we do really for patients who need to know they’re being treated. And it’s always a challenge when you don’t expect drugs to give you a response, so how do you know if they’re benefiting or not?

And so this is always a struggle for me and for many of our group and lots of people everywhere, in terms of are you making a patient sick or not? Are you impacting the disease? Tough drugs to know how to use. It’d be interesting to see if TAS, for example, could be moved up into earlier line, because it does seem to be different than 5-FU.

Dosing and tolerability of regorafenib

DR LOVE: You mentioned dosing of regorafenib. What dose did you use in him, and how did he do?

DR VENOOK: So this is a fellow we started, I think, routinely we start at 60. I mean at 80. And we worked him up to 160. So what I start is really half dose, pretty much. And most people start at around 120, I think, these days. I started at 80. He was doing well. And we ramped it up.

With the TKIs like this, as we learn — let’s say with imatinib, in GI stromal tumor, you can start low and work up. If you have a lot of toxicity at the start with a higher dose, sometimes it’s hard to go down from there and not induce more toxicity. Again, I worry that I’m not sure that any of us — certainly, I’m not sure that we or I am using regorafenib or TAS in the right way, because we tend to reserve it for patients at the end of the road. And we may be waiting too long or not dosing it aggressively enough. This is a problem.

But I do think regorafenib, in particular, is quite toxic at 160. And so they’ve explored lower doses. And probably it’s fine to use lower doses, but the label still calls for 160 a day.

DR LOVE: How long was he on treatment? And from your point of view, do you think he benefited?

DR VENOOK: I thought he did. I remember, I think he was on for about 4 or 5 months. And he got really tired. And you never know if he was just tired — if he was depressed and tired or if the drug induced it. So 4 or 5 months, I considered that a win, because his disease was progressing when he went on it. So I think probably we get more stasis of disease with these drugs than we get responses. But patients as they’re down the road, fatigue and other things start to interfere. And it’s hard to know if it’s disease progression or if it’s just the wear and tear.

But we know that drugs, let’s say, like regorafenib or sorafenib — certainly sorafenib, which has been used a lot in, let’s say, kidney cancer patients, we know that there’s a muscle wasting that happens over months and months of use. And so I don’t doubt that these drugs could add to the wear and tear and make patients a little more fatigued before their time.

DR LOVE: What do you see in terms of tolerability issues with regorafenib? Do you see hypertension, for example?

DR VENOOK: So it’s interesting. So the answer is, usually not, but that’s probably because these are patients, almost all of whom have had bevacizumab and didn’t have problems with bevacizumab hypertension. So we tend to shy — the few patients I’ve had who have, let’s say, Grade 3 or greater hypertension on bev, I tend not to use regorafenib, because it’s a class effect. So these patients have proved to us that they’re not going to get in trouble.

And not to mention that most patients at the end of the road, as they are getting more progressive cancer, very few patients have hypertension as a major problem, because I think their vascular tone, all these other factors come into play. Hypertension is not usually something we face at third or fourth line.

Regorafenib-associated tumor cavitation

DR LOVE: So one of the things that I’ve heard described — and I’m curious if you’ve ever observed it, and I think not just with TKIs like regorafenib, but maybe even with bevacizumab — is a phenomena where you see cavitation in the tumor. I’ve heard it’s kind of like part of a response. Have you ever observed that?

DR VENOOK: Yes. We see that. Not unusual. Certainly with bev we see that. And I’ve seen it with regorafenib. You see it even with a number of other therapies. I think it’s a bit exaggerated, the risks. Clearly, the original bevacizumab data in squamous cell cancer of the lung that delayed its approval because of the fact that patients were at risk for bronchial bleeds or lung bleeds with the necrosis from within.

In colon cancer in general, we see that. And I don’t usually worry about it. I take that as a general positive. I think maybe once I’ve seen a couple of these get infected, superinfected, but I’ve never really had a patient get a severe bleed with colon cancer who got bev or regorafenib or even for ramucirumab, which is the same mechanism.

Sequencing regorafenib and TAS-102 as later-line therapy

DR LOVE: So this patient got TAS-102 after getting regorafenib. First of all, how often are you able to get both agents into people? And how do you decide which one goes first?

DR VENOOK: Yes. So here at UC, where we take care of a lot of complex patients, usually if these patients get to third or fourth line, they’re usually back in their community because a lot of our patients come from a distance. And to give them regorafenib or TAS-102, we generally discourage them from schlepping into San Francisco. So we don’t get a lot of patients like this.

It’s different, which one we use. In a patient who I think has gotten a benefit from bev in combination, I tend to lean toward regorafenib, because I do think that TAS-102 is just another fluoropyrimidine. And at least we tend to see a lot of marrow toxicity. But again, I would emphasize that our patients are generally very heavily treated. We’ve been very aggressive in a lot of these patients, with liver resections and all.

And I’m not sure that we’re using TAS or regorafenib correctly and/or to the right degree. And I personally tend to be very averse to making patients sick. So if patients are third or fourth line and they’re fatigued or they’re having neutropenia or they’re admitted to the hospital with febrile neutropenia, I’m hard pressed to keep treating them, because I really don’t think I’m helping them that much.

TAS-102-associated neutropenia

DR LOVE: Any comments on the issue of neutropenia with TAS-102 and also the suggestion that maybe patients do better if they get neutropenia?

DR VENOOK: Yes. So that’s true. It’s an interesting observation. We see it pretty often, again, because I’m pretty aggressive with dosing TAS-102. In particular, we have a study, for example, with radioembolization and TAS-102, where neutropenia and thrombocytopenia are pretty common. And we see good responses. So I do think it’s a radiosensitizer, at least with radioembolization as we’re doing it. And in that setting, we’ll use growth factors and certainly will support it.

I just have trouble using drugs third or fourth line with clearly marginal advantage for most patients and putting them in the hospital with fever or using growth factors. But I think, again, some people can get months and months out of these treatments. So it’s not something you should just disdain. The patients who don’t get benefit, you tend to abandon it early.

California End-of-Life Option Act and aid-in-dying medication

DR LOVE: So one other thing about this case that I thought was fascinating — I couldn’t wait to ask you about — is his current situation.

DR VENOOK: Yes. He’s still alive. He’s probably 6 months down — after all this stuff, he’s alive. I think he’s taking Chinese herbs and mushrooms. And he’s living large. He’s playing golf. And his fatigue is gone, so in retrospect, certainly, it seemed to be chemotherapy induced. So that’s the other amazing thing. Sometimes we take credit. We think we’re making a big difference in patients with our interventions. And sometimes if you leave them be, they’ll do fine. So he checked in just a few weeks ago just to find out if we had anything new going on.

DR LOVE: Yes, I was really struck by that. And I totally agree with what you’re saying. I think that was really interesting. At what point would you send him to hospice?

DR VENOOK: So in California, we have an End-of-Life Act, which — and this is something he’s opted for. So the End-of-Life Act, basically, when we believe there’s less than 6 months to live, they essentially have — there’s a law that has us — the capacity to prescribe them a cocktail that they can self administer. I actually don’t write the prescriptions. That’s been my own philosophy.

He did opt for the End-of-Life Act. He has a prescription, and he’s still alive 6 months later. So it’s the best of both worlds. This person has the autonomy and the ability when it’s all said and done, but he’s living life well and enjoying it. And so it’s kind of — it’s a reminder that patients — giving them some time on their own is good. And some patients can get a whole lot of it. So he is a super-appreciative guy, because we gave him permission to leave and go back home, and he did. And he’s doing very well.

DR LOVE: That’s really fascinating. I was interviewing — I don't know if you know Ross Camidge from Colorado. And they have a similar law. He was telling me about a patient he had. He was very uncomfortable, told me the same thing about he didn’t want to write the prescription, but he was okay, I guess, doing the other part of it. And the interesting thing was, he checked back with the patient, I guess, a month or so later, the family, expecting to get the family. And the patient was there. And the patient never used the prescription. And the patient just said, “I just wanted to know it was there.”

DR VENOOK: Yes. I know. And I think that’s perfectly reasonable. But the problem here in California is, the cocktails are, I think, very difficult, cumbersome. There’s a heavy dose of secobarbital in the cocktails and —

DR LOVE: Right. And he was telling me the same thing. It was like 100 capsules and all this stuff.

DR VENOOK: Right. So here, we can put it all in a solution. The problem is, patients have to drink it within 10 minutes to have the full effect.

DR LOVE: Right.

DR VENOOK: And so they may fall asleep before they get — I mean, it’s not ideal.

DR LOVE: Crazy.

DR VENOOK: It’s one of those laws that, I think, perfectly well intentioned, but implementing it is not always so straightforward.

DR LOVE: Have you had any patients who actually did that?

DR VENOOK: I have not, no. We’ve had some patients in our group have opted to do that. I have not had any do that.

DR LOVE: When I heard that story originally from Ross and now from you, I was thinking that maybe it’s just like a safety thing for the patient to know they have it, but that if they’re well managed, maybe they never will have to use it.

DR VENOOK: We did some research a year ago when the law was just started in California, implemented in June of a year ago. And in Oregon, where patients have had the flexibility of doing this for a long time, there are very few people who have opted in. And I believe in California the number of patients who have opted for the cocktail is in the hundreds.

DR LOVE: Hmm.

DR VENOOK: I mean, California is a big state. I have no idea if those people also took their lives or used the prescription. I just know that I think that’s the number of people who got the prescription.

Case: A 66-year-old man undergoes simultaneous resection of primary sigmoid cancer and a solitary liver metastasis followed by pseudoadjuvant therapy with FOLFOX

DR LOVE: Let’s finish out with your 66-year-old man.

DR VENOOK: Okay. This is a patient who had a solitary right lobe of the liver lesion. We took out the primary, and he got the metastasis resected at the same surgery, which we do sometimes, if it’s an easy chip shot. This right lobe of the liver is not — usually a left lobe will be more easily taken out, but we took out the primary and the met at the same time.

He got 6 cycles of FOLFOX in pseudo adjuvant, I guess we’d call it, developed neuropathy and discontinued treatment. This is a few years ago. And he went almost 2 and a half years without evidence of cancer but developed a right lower-lobe pulmonary nodule, which, of course, with a negative workup otherwise. So we didn’t know if this was a met or a primary.

So a new pulmonary nodule, you’d always treat it as a new primary, as a solitary. We took him to surgery and it was a colon met, a small colon met. And he really didn’t want any adjuvant therapy. And so after two and a half years, it seemed reasonable to leave him alone.

Then, about, I think, 8 or 10 months later, he developed a few liver lesions and lung lesions. And so at that point, we started chemotherapy. He got FOLFIRI/bev for almost a year. And when I do that, let’s say we go every 3 or 4 weeks with FOLFIRI. He had stable disease but ultimately had an anastomotic recurrence. And he came back to us at that point, and we took him to surgery.

I think the message here is, this patient lived, I think, 5 or 6 years from his primary diagnosis. And the message is, always remember that these patients, if they recur systemically — the reason I chose to talk about this case is, when patients progress systemically, go back and look at the primary, because it’s not unusual that they’ll progress in the primary as well as systemically. And that’s what he did.

And again, that’s always a caveat. These patients who live a long time — he was back in the community. We’d forgotten — frankly, hadn’t recommended that they do colonoscopy, just — I hadn’t thought of it. And the idea that patients who live with colon cancer for a long time really can live an amazingly long time, but you should be on the lookout for local recurrence. And so that’s, I think, the message there.

Similarly, in patients where we leave the primaries in place up front, it doesn’t mean you just forget about the patients. Even in the study that McCahill published on showing the safety of leaving primaries in place in patients treated with FOLFOX/bev, I think 20% or 18% of patients needed emergent surgery. So you can leave the primary in place at the outset, but patients who are doing very well, think about the primary, because you’d like to take it out on your schedule rather than when they obstruct. Then they’re usually left with an ostomy of some sort, and it’s not a great situation.

DR LOVE: What’s the current situation with this patient?

DR VENOOK: This guy, he’s not doing so well. He had low-volume disease, right, except for his primary. So we took him and resected his primary. He did not respond. He’s had maybe marginal response or stable disease on the current chemotherapy.

We went back to FOLFOX, because it had been the better part of 3 years or 4 years since he’d had oxaliplatin. Interestingly, he didn’t develop any neuropathy, but I’d say he’s metastable disease. Again, this is an example of a patient who’s done remarkably well, almost without our interventions. I mean, essentially, surgery is about all we’ve done for him, with a little bit of chemotherapy. And this is the biology of the disease. You can cherry pick these patients. And he’s 6 years out.

DR LOVE: How was he doing on the cetuximab?

DR VENOOK: So the problem is, he didn’t get a rash. So he’s a very informed guy. And his doctor in the community was planning to do the EVEREST deal, was planning to dose escalate the cetuximab to try to get him a rash. And that’s where we stood. I recommended against that, because I don't know how much marginal benefit he’ll get.

And I should also say that he’s got low-volume disease. We took the primary out. He’s got a few mets and, in fact, is not threatened by his volume of metastatic disease, but — and once they get on the merry-go-round of chemotherapy, being aggressive, that’s where they are now.

To me, this is a patient who was 6 or 7 years out. I would have liked to hold off until there was something really cool to do, maybe figure out how to get checkpoint inhibitors to work in non-MSI-high patients, for example. But this patient is on treatment. Be interesting for me, I think, at some point, he’ll probably opt off the treatment. And how well he does just by — if we leave him alone after that.

Investigational combination strategies for microsatellite-stable (MSS) mCRC

DR LOVE: So final question: You mentioned something I was going to actually ask you about before, which is the issue of checkpoint inhibitors in non-MSI-high colorectal cancer. Anything new and exciting there?

DR VENOOK: So no. We’d expect it. I really thought by now we’d have the results of the atezolizumab/cobimetinib randomized study that has not matured yet. The data’s not available yet. I don't know what that means. I would suspect it may mean that there was no benefit to the experimental arm, but I don’t know that. That was the lead strategy that was in place.

Otherwise, I think we really haven’t figured out how to make patients who are MSS look — make the immune system think or make checkpoint inhibitors think they’re MSI high. So that’s a lot of work to do. That’s where we have to make progress, because — right? I mean, the checkpoint inhibitors are great, but that’s for 3% or 4% or 5% of patients with colon cancer. And we’ve got to do a lot better than that.

DR LOVE: But I was really surprised and interested to see that combination of a checkpoint inhibitor with a MEK inhibitor. I mean, it looked, I guess, in a single-arm study, maybe it was going to be useful.

DR VENOOK: Yes. So I think that was one of those studies that was in the eye of the beholder. I mean, I did not think it looked very promising. I didn’t take care of any patients on the study, though. People who had patients on the study really thought they worked. I confess that we’ve had a couple of people who elected to opt for that combination off study, and they’ve not done well. That’s my N of 2 that have not done well.

DR LOVE: What’s the thinking there? I was really surprised to see that combination.

DR VENOOK: Yes. So I think there’s a certain — part of the thinking is, of course, is the whack-a-mole idea that if you block different pathways, you can turn off the cycling, can turn off the feedback loop. I think that as we get into these combination cycle inhibitors, I think a little bit of — we’re kidding ourselves, to some extent, because it’s as if the cartoons we draw about the orderly pathway, the nodes from RAF to RAS in the downstream events as if that’s really etched in stone.

So I do think that probably different combination biologics will impact some patients, but I really am not all that positive about that or a lot of these other ideas. I think what we see even in the BRAF, where we know we have an inhibitor of the BRAF, it has some effect in some patients and it’s not durable. So it’s way more complicated than we would like, than it would need to be for these simple combinations to work.

CheckMate 142: Activity of nivolumab in patients with MSI-high mCRC

DR KOPETZ: So the nivolumab study, the CheckMate 142, has really looked at the population of a microsatellite-high tumor. So these were patients that by standard testing had either loss of a mismatch repair protein or other tests that demonstrated the microsatellite instability phenotype. These were patients that had multiple different lines of therapy and then enrolled on this single-agent nivolumab study.

The study itself, which has now been reported out in manuscript, had demonstrated about a 25% response rate with the nivolumab, but, importantly, a much higher proportion of patients that actually had disease control. So even though the tumor hadn’t responded, the tumor stopped growing. And the excitement, I think, in the field is really around the maturing progression-free survival and overall survival from these.

Dosing schedules and choosing between anti-PD-1/PD-L1 checkpoint inhibitors for mCRC

DR KOPETZ: So the, I think, really striking thing with this population is that about 50% of them are still alive at 2 and a half years at the last report. So it’s this amazing dichotomy that if patients respond or if their disease is stable beyond a few months, that they’re really on for a very long period of time. And that’s been really exciting for our patients.

DR LOVE: So maybe you can compare the data that we have available right now for nivolumab to pembrolizumab.

DR KOPETZ: Right. So there’s some differences in the population that we have for the nivolumab and pembrolizumab. But overall, I think we see kind of high response rates in both. Probably if we look at the disease control rates, we’d say that they’re fairly equivalent between the population. The pembrolizumab, which reported out first in that initial series in the New England Journal, that was beautifully reported, doesn’t have quite as many patients, and so we don’t have quite as much data on the long-term outcomes. But the sense is that they’re really behaving very equivalently.

DR LOVE: How do you decide, outside a clinical trial setting, which one to use?

DR KOPETZ: That’s a great question. There’s differences in the dosing regimens. The initial nivolumab was done every 2 weeks, but now there’s a kind of 4-week schedule. And pembrolizumab used a 3-week schedule. So there is kind of the very practical components of how one is dosed.

For colorectal cancer, we’re fortunate that both are approved for MSI-high colorectal cancer. There is a little broader indication for the pembrolizumab in terms of covering MSI-high and other tumor types outside of colorectal cancer. So for a GI oncologist, that becomes relevant when we’re dealing with a small bowel patient or some of the other GI tumors.

DR LOVE: Can you talk a little bit more — I’ve heard about this issue of prolonging the interval between the nivolumab in other tumors. And you mentioned 4 weeks. What exactly is done? What data do we have on it? Is that something you think is reasonable to do? Can you even do it outside a trial setting?

DR KOPETZ: Right. So it is something that there is some data on. And I think a lot of that is also based on some of the modeling of the duration and half-life. And it’s actually moved to a kind of a fixed dose even. So I think the idea is that you’re really saturating the receptors and the antibodies past the saturation point. And so I think there is comfort in the community with the alternate dosing regimens.

Incidence of high MSI in other cancers: Implications for testing

DR LOVE: And you mentioned, also, in terms of — I think this was the first approval that was really based on a biomarker as opposed to a cancer specifically, in terms of a cancer of origin. First of all, where else do you see MSI high, particularly MSI-high metastatic disease?

DR KOPETZ: Right. While colorectal is one of the key ones, we certainly also see it in endometrial cancer. That’s one of the ones that also is associated with, kind of, the Lynch syndrome, the hereditary form of MSI high. We see it also in small bowel, some gastric. The interesting thing is that in many different tumor types, there’s just a few percent of patients that are MSI high.

So one of the things we are struggling with as a community is, how broadly do you test for MSI high? Do you test a pancreas cancer, which may have a 1% prevalence, or prostate cancer, with a fraction of a percent? And I think that’s where we’re still trying to develop best practices on where should we test and where shouldn’t we.

DR LOVE: I heard an investigator — I can’t even remember, maybe it was a GI investigator — make the statement “Every single patient with metastatic solid tumor should have MSI testing.”

DR KOPETZ: Yes. It’s certainly a defendable position, because I think it’s one of those things that if you happen to be that 0.2% of patients, of prostate cancer, for example, that really could have a substantial impact for you. And so I think when we’re at that point where patients are exhausting other options and trying to explore what therapies are available, I think it’s certainly reasonable to do testing broadly.

Pros and cons of MSI testing methods

DR LOVE: Let’s talk a little bit about the types of testing that are done. I see that you were part of a poster at the ASCO meeting that looked at local and central testing within the CheckMate 142 study. What do we know right now about how people are testing and what the best way is to test?

DR KOPETZ: Yes, that’s a great question. And I think what we know is that there’s 2 different methodologies that are kind of standardly utilized. One of them is looking at the proteins themselves. So are the repair enzymes present? And to do that, the immunohistochemistry test can be done.

There’s 4 proteins that are part of the mismatch repair complex. And so the IHC is a fairly simple test of looking for expression of those 4 proteins with some subtleties there. Most of the time it’s a bit of a binary result. It’s either present or absent. And if any one of them are absent, then that defines an MSI-high phenotype.

The other test is a test that actually looks at the consequence of a deficient mismatch repair system. So it’s actually a kind of a phenotypic test. And that’s looking at these regions called microsatellites that I kind of view as the canary in the coal mine of DNA repair, right? So that miners were worried about the air quality in the mind. They would have a canary. If the canary keeled over from problems, then that meant the air was bad.

So what we’re looking at are regions that are really prone to DNA repair problems. And so if we see these regions, these microsatellites that are really hard to replicate because of the mononucleotide and dinucleotide repeats, if there is a problem with DNA repair, this is the place that they’ll be seen. And so the idea is, you look at the most sensitive regions of the DNA.

So the test itself is really just looking at how reliably replicated these regions are. And it’s a PCR test that I think is fairly standardized and available in several different labs.

Now, there’s pros and cons to this, right? The PCR test, you need normal tissue. That’s not always available. You either have to draw blood or have normal tissue from the colon. But that, the data suggests that if you’ve got a really good source of tissue, that either of those tests are probably appropriate. Now, we run into problems when we’re doing small biopsies, as we talked about in the poster, but that really, I think, in clinical practice what we’re dealing with is surgical specimens or endoscopic biopsies where there’s enough material to make these determinations.

DR LOVE: Now, these assays, they determine whether the alteration’s in the tumor or also germline, or both?

DR KOPETZ: Right. So that’s a great question. And because there is always a little bit of uncertainty about what does this mean for a patient’s family, both of these tests really are just looking at whether or not there’s a microsatellite instability phenotype.

Then the second step is really to determine is there a familial component? Is this a sporadic MSI high? Remember that most are. Or is it a hereditary MSI high, like a Lynch syndrome, that needs to be pursued? So to do that actually requires additional testing. That’s when we bring in genetic counselors and actually do sequencing of the patient’s germline to understand if there are any mutations detectable in those genes.

DR LOVE: So just getting back to this effort focused on the 142 study, what did you find in terms of correlation between the local assessment and central?

DR KOPETZ: Right. So I think one of the things that we learned in this setting is that you have to really have good tissue to start with for some of these tests. What was seen is that if you take these small core biopsies and try to do PCR tests that require a fair bit of tissue, these tests can fail. And this is a situation that’s unlike others where actually at the end of the day, said, “You know what? The local testing that was done on the primary tissue may be better than trying to do it on these small fragments of a biopsy.”

And so specifically what we saw was that even in those tumors that didn’t have central confirmation, patients still responded, meaning that these were still MSI-high patients. It’s just that we have to be careful of the type of tissue that we put into these tests.

DR LOVE: And I’m going to guess that the primary is pretty much the same as the met.

DR KOPETZ: Correct. Yes. This is a feature that is universal in the primary and the metastases. So it’s maintained. So as a clinician, if you have tissue available from either source, you’re fine to test either one. And it really focuses on where do you have the most tissue?

DR LOVE: You mentioned the 2 different methodologies. Which do you prefer, and which do you recommend to a general oncologist?

DR KOPETZ: So the immunohistochemistry test, I think, is the fastest in many settings. It’s probably the cheapest and easiest to interpret. It has very good concordance with the other methodology. So what I do in my practice is just do the immunohistochemistry, but on all patients who come in the door. And really, then, only if I have a patient that is really early onset or there’s some other high concern for a hereditary syndrome, it’s only in those situations where I’m just a little concerned that the test may be wrong that I do the additional PCR test. But very rarely are those discordant.

DR LOVE: Do you prefer, do you recommend to have the results of this before you make a decision? To start first-line therapy in metastatic disease.

DR KOPETZ: Yes. It’s a great question. And we try to get this testing done as fast as we can. The approval for the PD-1 inhibitors is kind of second line and beyond. And there’s still some benefit in chemotherapy in these patients. And so we don’t feel too — if the testing’s not available before we start treatment, that’s okay. And I think that we still see patients that respond to PD-1 inhibitor after some initial first-line chemotherapy.

Now, there’s studies ongoing asking that question, actually: Is PD-1 better than chemotherapy or, even, can you combine that with chemotherapy? So the idea that maybe the combination off the bat will be the best.

First-line treatment with anti-PD-1/PD-L1 checkpoint inhibitors in mCRC

DR LOVE: And, of course, that’s now being done, actually approved in non-small cell. We’ll see what happens in this situation.

But what about the issue, right now, of using it first line, for example, elderly, frail patient, really not a good candidate for chemotherapy? Putting aside the question of whether you can access it or not, do you think that might be a reasonable clinical decision based on what we know?

DR KOPETZ: Yes. I think it is. And I do have patients in my practice just as you described, where chemotherapy would be rather tough for them. Also in patients who have recently been on adjuvant therapy and had disease come back a lot of times. In those settings where we think, kind of, quote-unquote, first-line metastatic treatment may not be quite as effective, we end up then moving into the PD-1.

Clinical characterization of MSI-high CRC

DR LOVE: What are the clinical situations, if any, where you might suspect that the patient has an MSI tumor? I hear whiffs of things about the phenotype being differently. I mean, can you kind of predict who might have it?

DR KOPETZ: Right. And I think there are some hints. Actually, I think we probably overappreciated our ability to clinically sniff these out in years past. The phenotype, as you describe, is kind of right-sided tumors, perhaps a little more female predominance, perhaps a little older. That tends to be more of the sporadic MSI high. And, of course, the hereditary syndromes, the really young patients who have a colon or endometrial cancer history in their family or it’s their second primary, for example.

But I think what we’ve also found is that when we try to apply that kind of clinical sniff test to find an MSI high, we miss a lot of patients. And now that we’re doing universal testing, which is now recommended by ASCO and CAP and all the major guidelines, this idea that just everyone should be tested. And don’t rely too much on your ability clinically to detect and pick these patients out.

Reliability in assessment of MSI status after chemotherapy

DR LOVE: I want to ask you about a few specific studies that you’ve done in this area that look really interesting. One is a paper you had a few months ago, you were part of, in Clinical Cancer Research, “Can Microsatellite Status of Colorectal Cancer be Reliably Assessed After Neoadjuvant Therapy?”. I thought that was pretty interesting.

DR KOPETZ: Yes. No. That it is. And part of this gets to the idea of what’s the optimal specimen for testing? And there’s some subtleties in these, but I think that the main message for the clinicians a lot of times is, take advantage of the tissue that you have and without trying to worry too much about the subtleties of these different settings. It’s just do the testing, right? Just remember that everyone needs to get tested for this group.

DR LOVE: But I would assume — I mean, does neoadjuvant therapy change the MSI status?

DR KOPETZ: Right. So the MSI phenotype itself biologically is there, right, that at the cellular level, if you actually ask the question about are these undergoing mismatch repair? Yes, those are all maintained so that the biologic MSI characteristics are a feature that is independent of the prior treatments.

DR LOVE: I mean, I would assume that you can send a metastatic biopsy after the patient’s had chemo.

DR KOPETZ: Correct.

Investigation of predictive biomarkers for anti-PD-1/PD-L1 checkpoint inhibition in mCRC

DR LOVE: So you had a couple of interesting papers that you were part of from the ESMO meeting that just occurred. And one I wanted to ask you about was “Analysis of Tumor PD-L1 Expression and Biomarkers in Relationship to Clinical Activity of Patients with Deficient Mismatch Repair.” This is again data from the 142 study. But, I mean, not everybody responds. What did you all find out about the biomarker predictors?

DR KOPETZ: Right. So this is a huge area of interest, to say not everyone responds, why is that? And, of course, there is data that has been looking at PD-L1 expression. So we looked at that as part of this cohort and found that it was really — that we saw responders in both cohorts so that in this setting in colorectal cancer, like we’re now seeing in other tumor types, that the expression levels of PD-L1 are not strongly correlated with activity. So we don’t use that clinically to select patients or not.

Related work that we presented a few months before was asking about what about if it’s a hereditary syndrome versus sporadic? Right? Are there differences there? Turns out, not. What about BRAF mutations, which co-occur with MSI high more than one would predict by chance alone? It turns out that that also doesn’t impact who’s responding or isn’t. So I think what we can say is that the simple biomarkers really aren’t associated with differential response in colorectal cancer.

The hunt now is to really say, “What is it?” Right? If it’s not these initial biomarkers, what type of biomarkers are they? And here I think what we’re seeing is that — and I think there’ll be additional work that will be coming out in this cohort asking questions about the RNA profiling, about the microbiome. So we’ll be presenting some data at the immunotherapy meeting here in a few weeks looking at the microbiome and how that impacts response or not to these agents.

So I think there’s a lot we don’t understand, but MSI high is such a wonderful opportunity, given the high response rates and the activity, to really try to get a sense of where is immunotherapy working? Where isn’t it? And why is that?

Relationship between mutational load and response to anti-PD-1/PD-L1 checkpoint inhibitors

DR LOVE: Could you talk a little bit about your vision of the — I don't know — the biology of MSI-high colorectal cancer? I’ve heard people say that, quote, there’s a high mutational load. First of all, if you measure mutational load, does that correlate with response? And is that the essence of why it responds?

DR KOPETZ: Yes. So the mutational load is substantially higher. It’s about a hundredfold higher than what we see with traditional colorectal cancer, microsatellite stable. And this is even higher than the smoking-associated ones, like lung, bladder, for example, or melanoma, kind of UV induced. So the mutation rates are very high.

When I describe this to patients, what I tell them is this is like having a pull of the lever in Vegas, that every time you have a mutation, there’s a chance that that mutation could be antigenic, right, that that could be enough to stimulate the immune system. It turns out that probably for every 20 mutations that a patient has, only 1 of those has a meaningful immunologic response, right, that the immune system actually can identify it.

So that means if you have a tumor that has a very high level of mutations, you have more rolls of the dice or pull of the lever and more likely to be able to generate the immune response. But it’s not a clear-cut relationship because of that stochastic nature. It’s not a clear-cut relationship between mutation load and response. There’s still a lot of chance involved in that process.

Role of the gut microbiome in the pathogenesis of colon cancer and response to checkpoint inhibitors

DR LOVE: So you mentioned the microbiome. And another speaker on this same program, your colleague Alan Venook, was also talking about it. And it sounded really fascinating. Maybe you can elaborate a little bit more about what’s known about, quote, the microbiome, how it correlates with right and left sided and even — I mean, actually, the etiology and pathogenesis of this disease.

DR KOPETZ: Right. So the microbiome has probably been an underappreciated component of oncology, broadly. And what we are seeing is that the microbiome — and by that I should say we’re talking about the composition of the bacteria in the gut. And that’s both in the lumen of the gut but also these ideas of there’s biofilms. There’s actually kind of mucus layers that are present on the gut that have their own separate colonization and own separate composition of bacteria.

But this microbiome certainly plays a role in colon cancer and risk initiation of the potential to have an inflammatory milieu that’s present. The biofilms are really more present on the right side of the colon, so there are differences between right and left. And if you actually sample the microbiome in these different regions, you’ll see differences as well. So there’s some thought that that may be causative for some of the differences in the types of tumors we see on the right and left, although that remains to be seen.

Now, the intriguing thing — and this is where I think we are very excited, but we don’t have as much information as we’d like — is this idea that the microbiome in the bowel may impact the body’s immune response to anything, right? So it’s not just colon cancer we’re talking about PD-1 inhibitors. This has now been shown for other tumor types, the work done by Dr Wargo from our institution in melanoma response, right?

So that is fascinating. Why would bacteria in the gut influence your immune response to a melanoma? But indeed, that’s what’s seen. And there’s certainly thoughts about how the immune system gets modulated by the different microbiome in the gut. So this is an early field and we don’t have all the answers, but it certainly is compelling data.

DR LOVE: Any attempts, even empirically, to alter the microbiome, either by itself or in, let’s say, patients getting checkpoint inhibitors? I don’t even know how you would alter it, antibiotics or you tell me.

DR KOPETZ: Yes. So there are a few different approaches that are early. So one is trying to use diet to alter the microbiome. So there is a colleague at our institution that actually has a study where patients are going on a very bean-intensive diet, with this idea that there’s data saying that that helps alter the microbiome and it may have some beneficial impacts.

DR LOVE: But these are people also getting checkpoint inhibitors?

DR KOPETZ: So the first step is not to do it in the patients getting checkpoint inhibitors but then to learn about the patterns of microbiome change that you see. But that’s the hope. Can you then apply those interventions, if you can shift the microbiome in the right direction as it appears? So I think that’s the next step.

Then there’s others that are actually more of a fecal transplant model, right? So this is the idea that there’s ways that you can kind of package and deliver someone — a healthy person’s microbiome into a diseased patient’s intestines, so the idea if you clean it out, do antibiotics, other things and then you transplant in a healthy microbiome, fascinating areas. And it works beautifully in the mice, but we just have to see if it translates into patients.

DR LOVE: In mice, does it affect what happens when you give a mouse a checkpoint inhibitor?

DR KOPETZ: It does, right. It does indeed. And you can do these transplant studies in the mice and actually show that it’s not just an association, but the microbiome is causing differential response to the PD-1.

DR LOVE: That's interesting. Certainly your colleague Jim Allison has been a real leader in this field. And I’m kind of curious how he views all this stuff. How does he view the microbiome? Do you all sit around and talk about these things?

DR KOPETZ: That’s right. Yes, so I think it’s one of the clear signals that — and kind of a recognition that it used to be that we were just focusing on the immune cells, looking at the tumor-infiltrating lymphocytes in the tumor or in the periphery and the cytokines in the blood, but this recognition that if we really want to comprehensively understand the immune system, we’ve got to look beyond that.

And I’ll say it’s not just gut microbiome either, right? There’s also work looking at oral microbiome and how that may play a role. And so I think we’re going to be learning how much or how little we know before we get all this cleared up, because it’s a fascinating new area.

Rapidity of response to checkpoint inhibitors in patients with MSI-high tumors

DR LOVE: It really is interesting. A couple of clinical questions about checkpoint inhibitors with MSI high, first in terms of speed of response: I mean, in general it sounds like what we’re hearing throughout oncology is that the speed of response, the depth of response, kind of seems fairly quick, kind of like chemo. Is that the case with these patients?

DR KOPETZ: It is. When we initially started working in this, the thought was, “It’s going to take a little bit of time.” And there was this idea of, “Maybe initially you have to let it progress a little bit before it starts to respond.”

So many of these studies were actually written with the assumptions that it was going to take time. It turns out that you’re right, that clinically when these patients respond, we see this fairly quickly, very much like chemotherapy. Sometimes within a few weeks patients are having substantial improvements in their tumor-associated symptoms.

DR LOVE: So one of my favorite — we like to create these clinical scenarios to really get inside people’s heads. And one of my favorite ones recently has been, if you had a patient with a liver-only met where the surgeon said, “I need you to shrink this in order for me to take it out” and you knew the patient was MSI high, right now outside a clinical trial setting, how would you strategize? Would you include a checkpoint inhibitor as part of that strategy or just give the typical chemo/biologic?

DR KOPETZ: Yes, so I actually have a case just like that.

DR LOVE: Really?!

DR KOPETZ: So yes. So we’ve done that. And maybe I can share the case briefly.

DR LOVE: Yes.

DR KOPETZ: But this was a young gentleman, had a hereditary syndrome. But he presented with a transverse colon mass, very large, had kind of locally invasive, so it had invaded into other organs nearby. And the surgeon said, “Okay. We’ve got to get a response.”

He’d come already started on FOLFOX, got a few months of that, I think, like, 6 to 8 weeks maybe of it, really no change in CEA. A touch of shrinkage on the CT scan. He was MSI high. We put him on a PD-1 inhibitor and, within 2 months, the tumor had substantially regressed. They were able to go in and do a surgery that was much less involved. And there were only very rare isolated spots of tumor cells that remained, so less than 1% viable tumor at the end of the day with what, in essence, I would have thought as a very short duration of preoperative therapy.

So yes, I think we’ve gotten some confidence that in responding patients, you can see responses quickly. And these response rates may be very similar to what we see with chemotherapy.

DR LOVE: Incidentally, this case made me think of something. Is a transverse lesion right or left?

DR KOPETZ: That’s a great question. It depends on who you ask. And the differences in biology appear to be a bit of a continuum between the right and left side. Embryologically, the transverse is made of the 2 different embryologic segments. They are kind of fed by different vascular pedicles. So the surgeons will say, based on the anatomic definition, that it’s a mix of — part of the transverse is one side, part of the other.

Of course, there’s lots of clinical questions that come up about transverse and what do you do with first-line chemotherapy. And I’m not sure we have the clear answer to that yet. But I think there’s a sense that these may be more right sided in their behavior than left. So I tend to treat them more right sided than left, but I think you’ll get a divergent set of opinions on that.

DR LOVE: So what was I going to ask you? Oh. So this case you’re talking about, I don't know if maybe you came in in the middle, but I’m just kind of curious. If you knew up front that a patient had an MSI-high tumor and you needed a response, theoretically, what kind of strategy would you use? Would you use chemo, maybe checkpoint first, then chemo, combined? What would you do?

DR KOPETZ: Yes. I think it’s a great scenario. I think still, if we look at absolute response rates, that trying to get a course of chemotherapy, I mean, if it’s a healthy patient, kind of the triplet cytotoxic, FOLFOXIRI with bevacizumab, for example, will get response rates that are probably numerically higher than a PD-1 inhibitor.

But I think as you saw in this situation, although we did come in kind of after things had gotten going, what we did is had a very short window. So okay, after, kind of, 6 weeks, we started getting things ready and making the requests for the PD-1 inhibitor so that we weren’t waiting for 4 or 6 months of chemotherapy to make the switch.

Neoadjuvant studies of checkpoint inhibitors in MSI-high colon cancer

DR LOVE: So I just heard about a really cool study in non-small cell where they did neoadjuvant checkpoint inhibitors, saw a lot of response. And they were looking at the tissue, really cool. Has that been done with colorectal cancer?

DR KOPETZ: Yes. We have a study ongoing now in neoadjuvant for resectable liver metastases. We don’t know the answer yet. In this setting, the question is really kind of what can we learn that helps us understand why it doesn’t work in most patients, but, of course, a few percent, with the CTLA-4/PD-1 you can have some benefit, even in microsatellite stable.

So I think there’s a lot that we can learn when we have the whole liver metastases that come out that get back to the prior piece. The more tissue you have, the more you can learn, the more confident you are in the results.

DR LOVE: I was thinking, actually, about the MSI situation, using neoadjuvant therapy there. Has anybody looked at that? I mean, you would expect you would just see a lot of antitumor effect.

DR KOPETZ: Right. Yes. So that hasn’t been done yet in a colon cancer. We’ve been kind of talking a little bit about it. The colon cancers, the difficulty is that in standard practice, a lot of times we don’t know the MSI status until after the surgery. The medical oncologists don’t usually get involved until the postop period for these. So there’s some flow issues in terms of how you actually get the patients identified as MSI high and on that have really been the hurdle. I think there’s a lot of interest in doing it. It’s kind of the feasibility that’s been the limitation.

Combination strategies with checkpoint inhibitors for MSS mCRC

DR LOVE: You mentioned the issue of checkpoint inhibitors in MSI-stable tumors. And I know there have been a variety of attempts. I think there was one where there was a MEK inhibitor combined with it. Anything promising or exciting there?

DR KOPETZ: Right. So there was certainly some encouraging data with the MEK combined with a PD-L1 inhibitor, so that cobimetinib/atezolizumab data. There’s a randomized study that has been completed looking at that combination versus a standard of care regorafenib in kind of a late-line setting for colon cancer. So we’ll await those results to see if that is going to hold up. But I’d say that that enthusiasm has moved into many other studies.

So I think there’s probably at least a half dozen studies now ongoing looking at various mechanisms of combining checkpoint inhibitors with MAP kinase inhibitors as well. So that’s probably the one that we have the most initial data and enthusiasm on for those combinations.

Pseudoprogression with checkpoint inhibitors

DR LOVE: So the one final clinical thing I wanted to ask you about — you were talking about speed of response. But one question that started out with melanoma and always gets brought up no matter what cancer we’re talking about is the question of pseudoprogression. And I see, actually, at ESMO you actually had a paper looking at that. What did you find?

DR KOPETZ: Right. So this was getting back to the point before of do you need to wait for the immune system to kick in in order to call whether or not a patient’s benefiting or not? And what we’d seen across GI tumors in general is that pseudoprogression is just really not a phenomenon that we see, that when you see progression on an immunotherapy agent it’s real progression in GI, that this phenomenon of pseudoprogression, I think, is unfortunately not as common as we had initially hoped or thought based on early data.

DR LOVE: But one thing I have heard, though, as a clinical approach to this — it always made a lot of sense to me. And I don't know if you approach this the same way in colorectal cancer. I remember hearing it first in melanoma — is the idea that if you have a patient who’s progressing radiologically but is clinically stable that maybe you ought to not stop therapy and wait a while. And if the patient’s feeling worse, getting bad, that’s different. But if they feel fine, maybe don’t stop it. Is that what you actually do?

DR KOPETZ: Right. I think we do still give a little bit of credit — a little bit of leeway for some small progression, just like we would, actually, with many other therapies, whether it’s a cytotoxic of immunotherapy. If the patient’s doing okay and even if there’s a few millimeters of growth here or there in some of the sites of disease, we’ll kind of continue on with the idea that we’re still getting some benefit, that likely next time we do the scan, that 2-mm change may be 5 millimeters but that there’s a sense that we’re still deriving some clinical benefit that we shouldn’t toss the agent aside too soon.

Consensus molecular subtypes in colon cancer

DR LOVE: So I’m really curious about your work and your interest in the whole area of consensus molecular subtypes and also the correlation — I see you have a paper looking at consensus molecular subtypes in right versus left colon cancer, really getting into kind of the newer views, more contemporary view of the biology of this disease. Can you try to explain it to a simple person like myself?

DR KOPETZ: Oh, certainly. Yes. So the idea is that the DNA, the mutations that we see, give part of the story but that we know that that doesn’t give us the full picture of how the tumor’s wired and how it behaves clinically. And so the effort was to say, “Let’s look at the RNA,” right, how the machinery is actually being driven in the cancer and look and to see if there are common patterns that we see.

And indeed, there were. We looked at 4,000 colon cancers and asked the question, just how are they wired? How are they behaving? And what we saw were that there’s 4 major patterns or 4 subtypes that we saw. Now, these subtypes were defined without any regards to clinical characteristics, mutation profiles, outcomes. It was really just all about the wiring.

DR LOVE: When you say “wiring,” could I just clarify again at a macro level? The first place I think oncologists heard about this was in breast cancer.

DR KOPETZ: Right.

DR LOVE: Is this kind of comparable to the luminal, et cetera, in breast cancer?

DR KOPETZ: Right. Exactly. Or the germinal center in lymphoma — right. Exactly. So that is exactly it. Now, in colon cancer, there have been several papers, but we hadn’t really defined a consensus group, consensus classification. And so part of the reason that this has now kind of caught on and been engaged in many fronts for drug development and further translational research is that when we’re talking multiple different languages, it’s hard to advance.

So part of this effort, consensus molecular subtypes, just means taking existing data, coming to some conclusions on what these subtypes are. But indeed, that’s exactly it. How does the RNA signatures define these subgroups?

DR LOVE: And what about the correlation between this — and maybe you can talk a little bit more about how the 4 subtypes vary, particularly in terms of phenotype, if we know anything about it, but how it correlates with sidedness.

DR KOPETZ: Yes. Absolutely. So the 4 different subtypes, the CMS1 is one that is a very immune-driven subtype. So this includes MSI-high tumors but is also beyond that as well. So it also includes ones that have a very robust immune infiltrate for other reasons. The BRAF-mutant tumors, many of them fall into this category as well.

The second subgroup is what we call the canonical. So this is the one that’s kind of classically defined as an APC mutation, MYC overexpression, p53 mutations. This is the one that tends to have a little more of kind of the smokers in this. The CMS3 is a novel one that’s really this metabolic group. These are patients that have a lot of central obesity, have diabetes. There’s a very high rate of KRAS mutation in this setting, 80-some percent, and this kind of intriguing metabolic derangement. The CMS2 and 3 we call the immune desert. Really, there’s no signs that the immune system is even engaging at all.

CMS4 is an intriguing one. We call it kind of a mesenchymal subtype, where there is a very inhospitable and active microenvironment, so very active stroma. And the immune system actually, intriguingly, looks like it’s engaged to a certain amount but is excluded by the microenvironment. Right? So there is a degree of inflammation to these tumors, but the T cells just can’t get in there and engage. And so this may be TGF beta driven or others.

So the intriguing thing is that these have very different prognostic impacts in metastatic disease that can segregate between a year, year and a half to 3 and a half-year survivals. So there’s a lot of differences in their biology. There’s an association, then, of where they’re located. So the 1 and 3s tend to be more right sided. The 2s tend to be more left sided. And the 4s can be distributed a bit throughout.

But it’s a really interesting opportunity to now study these individual subgroups, because they’re now much more uniform in their biology. So we can then design drugs and inhibitors and strategies that may be more relevant for them.

Prediction of treatment benefit among molecular subtypes

DR LOVE: What do we know about predictors of treatment benefit? I see at ESMO there were a couple of papers trying to correlate these consensus molecular subtypes with the benefit from bev.

DR KOPETZ: Right. So there’s some data that says a CMS1, out of randomized data looking at, for example, the 80405 study, that showed that CMS1 is a subgroup that responds much better to bevacizumab than cetuximab. Conversely, in overall survival out of 80405, they thought that CMS2 responded better to cetuximab than bevacizumab. So it was the inverse of the CMS1. So that’s kind of an intriguing hint.

Now the question is, how do you put all this together? Right? And how do you put together sidedness and the CMS subtypes and the mutation profiles and really kind of build kind of a comprehensive viewpoint of the tumor biology? So we’re not quite there yet. CMS classifiers are now moving into clinical labs, but what to do with them in day-to-day practice is still something that we’re figuring out.

Case: A 67-year-old woman with BRAF V600E-mutated mCRC enrolls on the SWOG-S1406 clinical trial of irinotecan/cetuximab with or without vemurafenib

DR LOVE: So I want to ask you a little bit about the issue of patients with BRAF tumor mutations and particularly the study that you just reported looking at anti-BRAF therapy with chemotherapy and EGFR antibody. And maybe to do that, a good way to get into it would be to hear about this patient who was actually on this trial.

DR KOPETZ: Yes. No. Absolutely. So there’s a patient that I had the privilege of taking care of. She’s a 60-some-year-old woman, kind of presented back in 2011, initially, with some abdominal pain. A colonoscopy demonstrated disease in the ascending colon. She was initially a IIIB at staging and then underwent some adjuvant therapy.

She did okay with that and was without disease for a period of time until about 2014, when she then developed a kind of pelvic disease, peritoneal disease, ovarian metastases. So the cancer did indeed recur. And she had some liver metastases as well, kind of multifocal recurrence. So she did have — fairly symptomatic from her left ovary, so that was removed by the surgeon for palliative reasons and then was kind of initiated on some first-line FOLFIRI chemotherapy, so fairly standard up to this point.

Unfortunately, she progressed fairly rapidly on that FOLFIRI, not uncommon. Tumor testing was done when she came to MD Anderson and demonstrated that there was a BRAF mutation and, specifically, the most common BRAF mutation, V600E. As a little aside, there are other BRAF mutations besides V600 that have very different biologies. So when I talk about BRAF-mutant tumors, it’ll be the V600E mutation.

So that’s a fairly common approach, to see patients kind of progressing through chemotherapy. And so she went on the NCI cooperative group study, our SWOG-1406 study, where patients are randomized to getting irinotecan and cetuximab or irinotecan/cetuximab with vemurafenib, a kind of BRAF inhibitor, approved in melanoma.

She received the experimental arm, so the triplet, and actually had a very nice response. So she went on study in about 2015, had a response and was on therapy for about 18 months, so had a kind of good disease control for a while. Unfortunately, she then did progress on the disease. We did, kind of, mutational profiling, like, she did a liquid biopsy on her, looking at what had changed. And she had acquired some resistance mechanisms, MEK mutations that had reactivated that MAP kinase pathway.

So she received other standard-of-care therapies and, unfortunately, never responded to any further therapies and ended up dying a few months after coming off of the clinical trial.

Utility of liquid biopsies in mCRC

DR LOVE: You mentioned liquid biopsies. In general, in terms of metastatic colorectal cancer, where do you see liquid biopsies potentially fitting in?

DR KOPETZ: Yes, it’s — great opportunities there. So the liquid biopsies have some advantage of patients where you don’t have access to tissue, right? So we know sometimes patients will come and they’ll have a liver biopsy that was used for diagnosis, maybe some immunohistochemistry, and there’s nothing left. And so it’s a better option than rebiopsying some patients to get the molecular profiling.

We also see it used where you need that information quickly, right? You know it may take 4 to 6 weeks to get tissue testing done, and you’re trying to make decisions about up-front first-line therapy. So that gets a faster turnaround. And we also use it a lot in the research setting, as you heard here, right? What is it that’s changing? What are the resistance mechanisms that at play?

And what we find is that the liquid biopsy is actually more insightful than a tissue biopsy, because there can be multiple different resistance mechanisms that can develop. And when you do a tissue biopsy, you may capture one of those, but you may miss some of the others. And so the liquid biopsy is really nice in that it really assesses the entirety of the disease.

Efficacy and tolerability of irinotecan/cetuximab/vemurafenib

DR LOVE: So how did this lady do on this triplet regimen from the point of view of tolerability?

DR KOPETZ: So she did fairly well. She did have initially some diarrhea and fatigue. And after several months of therapy, we ended up decreasing her irinotecan dose in order to make sure that she could continue tolerance. And that improved her symptoms, and she was able to continue on.

We were concerned on these studies about potential for skin toxicity or secondary skin malignancies that can occur with BRAF inhibitors. But this combination, we found at the end of the day, was really well tolerated. We did not have an excess of skin toxicities or other side effects that preclude the standard dosing of the regimen.

DR LOVE: I mean, the one thing that we had heard about vemurafenib before the MEK inhibitors came along, actually, was the issue of sun sensitivity, extreme sunburn. Do you see that in these patients?

DR KOPETZ: Not that much. Certainly there is some, but it’s not that extreme level that we’d seen before. It may be that it’s that combination of the EGFR inhibitor that kind of blunts some of those responses.

DR LOVE: Can you talk about globally what was seen in the study in terms of the activity of this regimen and tolerability?

DR KOPETZ: Right. So we’d reported out the results at ASCO this year. So the study, overall, was positive. The primary endpoint was progression-free survival. And the hazard ratio was 0.48, which was statistically significant. So the difficulty is that these are patients that do very poorly on standard therapy. And so the progression-free survival went from 2 months to about 4.3 months. The disease control rate was more than tripled, so it went from 22% to 67% with the experimental arm. And then we had 4% versus 16% response rates with the combination.

Patients were allowed to then go — upon progression to cross over. So we didn’t have an overall survival readout as a result of that crossover, but certainly our primary endpoint of progression-free survival was strikingly positive.

DR LOVE: Do you believe that this is adequate evidence to support using this regimen outside a trial setting, and do you use it?

DR KOPETZ: I do. Right. I do indeed. I think this is a fairy large randomized study, although it’s a Phase II study. The study was powered with 80% power and a 2-sided alpha, I mean, of 0.05, all those characteristics that would be expected in kind of a Phase III study. And it was positive.

I think as a community, there’s not a whole lot of interest in going back and doing the same study again as a Phase III. Kind of the thought is, that is kind of compelling data in the setting of multiple other studies showing benefits of BRAF and EGFR inhibition.

DR LOVE: So yes. It’s amazing. This lady had, I guess, 18 months of disease control. That’s pretty impressive.

Influence of side-effect profiles in sequencing regorafenib and TAS-102

DR LOVE: I see also that prior to her death, she received both regorafenib and TAS-102 and didn’t respond. I’m curious in general, though, how you approach the use of these 2 agents in practice. What do you use first? And how do you manage people on these therapies?

DR KOPETZ: Right. It’s a setting where you have to — or set expectations with the patients. These are ones that clearly in these randomized studies we know extend survival on a population level. But we also have to kind of set the expectations for the individual. Like, what is the side-effect profile versus potential benefits?

So I tend to use a little bit more regorafenib first than TAS-102, partly because that the tolerability of the TAS-102 tends to be a little better, in my experience. Certainly the neutropenia and the anemia can be present, but it tends to be a little better tolerated, so I reserve that.

Theoretically, I’m also of the belief that there is some biologic crossover between the TAS-102 and 5-FU, similar but not identical mechanisms of action. So if you can put some other therapies in between the two, exposure to the fluoropyrimidine, then that may be beneficial, don’t know. This is just how I justify, kind of, the sequencings.

Case: A 52-year-old man with mCRC is found to have HER2 amplification and responds to trastuzumab/pertuzumab

DR LOVE: So I wanted to ask you about the issue of HER2, HER2-amplified colorectal cancer. Any patients that you have that you can talk about where you gave anti-HER therapy, either on or off a trial?

DR KOPETZ: Oh, certainly. So one that comes into mind is kind of a 52-year-old man that had metastatic disease that presented synchronously. It was a sigmoid primary, had unresectable bilobar liver metastases with some retroperitoneal nodes. The initial therapy of FOLFOX was administered. The patient got about 6 months to 7 months of benefit from the treatment. That was done with bevacizumab.

The patient then was switched over to receive FOLFIRI and cetuximab and unfortunately progressed at first restaging, then kind of came for opinion. We did some testing, which now in our practice is also routinely including the HER2 IHC, and this patient had a 3+ by IHC HER2. It turns out that FISH just confirmed that the amplification was there, but that’s not something needed with the 3+.

On the basis of that, we then put him on a study of trastuzumab/pertuzumab, so kind of a dual HER2. And the patient had a response.

DR LOVE: Objective response?

DR KOPETZ: Yes. Yep.

DR LOVE: Wow!

DR KOPETZ: Had objective response with that. And that is consistent with then both the overall data that’s been reported and also the response rates that we see with trastuzumab and lapatinib in Europe.

DR LOVE: So specifically — because, I mean, I was surprised, because I know that the new trial, I think it’s going to randomize your triplet versus — or no. I’m thinking — I’m sorry. I’m thinking of BRAF. But there’s a trial out there where they’re using pertuzumab and trastuzumab as one of the randomizations arms. And I was kind of surprised, because in breast cancer those 2 monoclonal antibodies are not very effective compared to when you add chemo. What do we know about that in colorectal cancer?

DR KOPETZ: Right. So the response rates with that are in around the 30% range with the doublets.

DR LOVE: Really?! The 2 monoclonal antibodies?

DR KOPETZ: Correct. Right.

DR LOVE: Huh.

DR KOPETZ: So clear activity. Now, HER2 is not common. It’s about 3% to 4% of patients. HER2 amplification is seen almost exclusively in KRAS wild-type patients. Right? Which is — this is the tumor biology talking to us, right, just like we see KRAS and BRAF mutually exclusive. They’re drivers. When you see HER2 mutually exclusive with KRAS, it’s a sign that the HER2 is really a driver in the colon biology.

DR LOVE: Although this is amplification, not mutation.

DR KOPETZ: Right. This is the — right. Exactly. So this is the amplification data.

The mutation data is very different. And mutation, there was some data present on neratinib, where HER2 mutation in breast cancer is looking very promising. But the data in HER2-mutant colon cancer is very underwhelming. So the HER2 mutations, not sure we’re going to be able to target effectively, but the HER2 amplifications absolutely. This is something that is moving forward.

So the study that you mentioned, the SWOG-1613, was just activated last week and so available through the cooperative group system. And that’s looking at irinotecan/cetuximab second or third line. So as long as a patient hadn’t received cetuximab previously, they’re eligible in second or third line, randomized to the trastuzumab/pertuzumab doublet or the control arm of irinotecan/cetuximab. So as you mentioned, we’re going up against chemotherapy, cytotoxic agent, but the doublet data really looks compelling.

DR LOVE: So what happened with this patient? First of all, how did he tolerate this combination?

DR KOPETZ: Did well, a little bit of mild rash, but was tolerated. Energy level was good, no diarrhea or nausea with it and ended up then progressing after just shy of 6 months of treatment but did have a good response going into the study.

DR LOVE: And what did you do next?

DR KOPETZ: So the patient was started then on regorafenib and then went on to another series of agents trying to target HER2, no response to regorafenib. The other agent is ongoing, so we’ll kind of see if he can derive some additional benefits.

Activity of T-DM1 in HER2-amplifed mCRC

DR LOVE: Dr Venook talked in my chat with him about — actually, it was a case he wrote up of a patient who got T-DM1 and seemed to benefit. Any thoughts about that?

DR KOPETZ: Right. Also an area of interest. And there’s a number of other agents that are delivering a payload along with HER2. So there’s a number of things that I think there’s some enthusiasm about. Actually, I think the idea is great, the antibody-drug conjugates for HER2. I think T-DM1, it has a microtubule inhibitor that’s part of the payload. And what we know is that taxanes are not particularly active in colon cancer, like every other tumor type it seems. And so there is some concern. Is that the right payload?

There’s a very active drug transporter that can spit out that payload, potentially. So again, the real evidence will come in the clinic. But I am enthusiastic about the strategy, but maybe not T-DM1 as the best antibody-drug conjugate.

DR LOVE: Although, I think these cytotoxics that are in these antibody-drug conjugates are like poisons. I guess — I don't know. You wonder whether or not you can really compare taxanes, systemically, to the way these things are given with these tubulins.

DR KOPETZ: No. Absolutely. And the cell line data suggests that it’s not as active in colon cancer than other tumor types. But we have to take that with a grain of salt. Absolutely, there’s enough rationale to get this into the clinic, kind of study it in patients. And we’ll have some of that data coming, I think, in the next years.

Multiplex testing for patients with mCRC

DR LOVE: Do you think that HER2 should be an assay routinely done in patients with metastatic colorectal cancer?

DR KOPETZ: So I think in patients where you’re kind of getting to the point where you’re running out of standard therapies, made it through first/second line, I think the fact that we have multiple studies with active regimens showing activity against HER2 amplified means that yes, we should do it. I mean, the cost for running an IHC on a tumor — I mean, this is an assay that’s readily available and fairly inexpensive to do. So I do it on all my patients, and I think especially now that we have a cooperative group study starting, which can do the testing as part of the study. But I think there’s a lot of interest now of trying to do some of this testing more routinely.

DR LOVE: One other common thing that’s done by docs in practice in a variety of tumors when a patient has metastatic disease and is starting to get through the approved therapies, they’re still in good condition, maybe younger, they want active treatment, a lot of people think of multiplex testing such as next-generation sequencing. For practical purposes in colorectal cancer, is that something that really ends up being helpful to patients if they’re in a community-based setting?

DR KOPETZ: Yes. So there’s 2 aspects. How helpful is it even if you have access to the clinical trials? Right? And I think we’d reported out, kind of, our experience in colon cancer with this type of testing. And it was only about a third of patients with an access to a huge kind of portfolio of experimental therapies at MD Anderson, really a minority of patients that you could make a rationale for utilizing experimental therapies.

It becomes even harder than when you start to say, “Your access to these trials becomes a little more problematic.” I think for patients who are willing to travel to find the right study that may match their tumor, it makes sense. I think the hard part, though, is kind of doing the testing without having the conversation about what do we do with the results afterwards, because I think you’re right. It’s very hard and putting a lot of burden on clinicians to try to always hunt down off-label therapies or try to convince insurance companies to try to provide these. And I think at the end of the day, we just have to make sure that we’re not setting expectations too high with the patients.

Case: A 68-year-old man with resected right-sided primary BRAF V600E-mutated, MSI-high mCRC and disease progression on FOLFOX/bevacizumab and FOLFIRI experiences a 2-year response to nivolumab

DR WAINBERG: So this was a typical patient who presented to the clinic with newly diagnosed disease having had, already, a surgery at an outside institution and a liver nodule that was biopsied and confirmed to be moderately differentiated adenocarcinoma. So we knew at that point that he had at least 1 to 2 good liver lesions that were not resected entirely because of their size but were present and at the time of diagnosis.

So at that point, we did a thorough analysis of his tumor and his lymph nodes. He had, actually, only 7 lymph nodes removed that were — none of that — which were positive. His tumor was RAS wild type and RAF wild type. And he did have, interestingly, the suggestion, at least on the immunohistochemistry, that he might be microsatellite unstable. But at that point, we didn’t know much about MSI testing, to be honest with you. But that’s where we were with that case.

DR LOVE: So what happened?

DR WAINBERG: So he had disease that was in the liver. And by the time we restaged him in early 2014, February 2014, he actually had multiple liver lesions at that point. So he went pretty fast from being a guy who had 1 or 2 liver lesions in the end of 2013 to by March of 2014 or so having about 10 liver lesions.

So we felt at that point there’s no way we can resect him at that point. So we started him on chemo, combination chemo, and he was on FOLFOX/bev for a while, had a good response clinically, by imaging and CEA. And he was on that regimen for about 6 months.

DR LOVE: Could I just clarify? Was the primary right or left sided?

DR WAINBERG: So he was a right-sided primary.

DR LOVE: And he had a BRAF V600E mutation?

DR WAINBERG: At that point in time, he had a BRAF V600E mutation and no family history of colorectal cancer. And as I indicated, we did immunohistochemistry mismatch repair genes on him from the primary diagnosis. And there was the loss of MLH1. So he didn’t have any family history, though, so at that point we were pretty suspicious that this was a microsatellite-unstable tumor.

Bear in mind that at that point, in 2014, microsatellite instability was being done more out of interest rather than for any practical reason.

DR LOVE: I guess for family and genetic reasons, really.

DR WAINBERG: Yes, family, genetics and —

DR LOVE: Wow! Things changed. That’s for sure.

DR WAINBERG: Yes.

DR LOVE: So I was just curious about the location, because I’ve been polling GI investigators lately about an interesting thing that Alan Venook stated to me in an interview a couple of weeks ago, which is, nowadays when he sees right-sided metastatic disease, he’s thinking about FOLFOXIRI even without a BRAF mutation. Is that something — I find not everybody’s quite on board with that. What do you think about that?

DR WAINBERG: In the absence of a BRAF mutation, I don’t personally do that. And I don't know if — I mean, I understand conceptually why you would, because the prognosis is worse and they’ve demonstrated that the prognostic significance of that right side is just worse, for whatever reason, for lots of reasons.

But the practical problem with doing that, the way I see it as someone who treats a lot of these patients, is, even though you’re going to give FOLFOXIRI to those folks, you need to consider the fact that if they don’t have a BRAF mutation, their prognosis isn’t quite as bad. So you’ve got to start thinking about second-line regimens and all of that. And you wonder if FOLFOXIRI is the right regimen for all patients with right-sided tumors.

DR LOVE: Yes. And I don’t think he’s saying that, I mean, exactly. But I think it was more that his trigger finger, so to speak — I mean, this patient sounds like he really didn’t do that well.

DR WAINBERG: Yes. I mean, it’s one of those patients where his liver lesions were at the time of surgery. So we already knew we were dealing with a metastatic disease at diagnosis. And things got worse in a matter of a few months.

He did have a nice response to FOLFOX/bev in terms of PR by imaging. Nothing went away. Definitively, we were still dealing with an unresectable situation. So at that point in time, we did consider a liver resection, but more or less there was no way to resect all gross disease in this patient. And even though he didn’t have extensive extrahepatic disease, he had a number of liver lesions that were just felt to be a big problem.

So we changed him to FOLFIRI at that point, because by 6 months later, after FOLFOX/bev, it was clear it had stopped working. He had new liver lesions. He had a rising CEA. So it was pretty clear that he was progressing.

DR LOVE: Just out of curiosity, why not bev/aflibercept or ramucirumab?

DR WAINBERG: So he ended up having some hypertension towards the end of his regimen. And I think after 6 months of bev, his hypertension actually became a little more challenging to deal with. And so VEGF inhibitors were problematic. At that point in time — and again, this is a few years ago — with the BRAF mutation and the MSI, we weren’t sure what to do with the EGFR inhibitors, either.

We were facing this question: Should we pull the trigger on cetuximab in combination with FOLFIRI, which, as you know, in the BRAF literature it’s gone back and forth a number of times? The NCCN Guidelines took it out for a number of years, and then they put it back as an option. So the question is, EGFR inhibitor role in this kind of BRAF mutation was difficult to know what to do.

So we had a problem with the VEGF inhibitors by that point that made further VEGF inhibitor use look like it would be a problem. He was already on 2 antihypertensives, and he had some vague cardiac history. So I didn’t feel comfortable pushing too much further.

Targeted therapy for BRAF-mutated mCRC

DR LOVE: So this man, obviously, now that we kind of got into the modern era, had the option of a checkpoint inhibitor. But I’m just kind of curious, since he did have a BRAF mutation, what your take is on targeted therapy of patients with a colon cancer and BRAF mutation. We saw some data — actually, on the same program, Scott Kopetz talked about a trial he presented at ASCO. You might want to talk about that. Globally, what’s your take on targeted therapy? And assuming the patient’s not MSI high, how do you integrate targeted therapy into nonprotocol management of these patients?

DR WAINBERG: In the BRAF mutation patients in particular? So, I mean, I think of the non-MSI BRAF-mutant patients, which are the more common and a very bad prognosis, as we know, it’s not the biggest group of patients. But if we’re going to say it’s 8% to 10% of patients with metastatic colon cancer, I think the data that Scott presented was very compelling, that the combination of irinotecan/cetuximab/vemurafenib is an active regimen in these people.

I personally looked at that regimen, that SWOG trial that was presented. And I thought — I saw the first data. I was less impressed. But then we saw the randomized data separating the curves, and no survival advantage yet, but that’s small numbers and short follow-up. But PFS advantage is pretty clear in the patients who get irinotecan/cetuximab and vemurafenib. I found that compelling enough that I think those patients should get a BRAF inhibitor in the mix.

I think we’ll see those approved pretty soon. As you know, there’s a bunch of BRAF inhibitors out there that are looking at the same paradigm. Whether it’s with EGFR inhibitor or with chemo is one school of thought that Scott put forward, which I think makes good sense. The other one is to forego the chemo part of it and give them the BRAF inhibitor and the MEK inhibitor and try to block downstream of that. So there’s different ways to get at that BRAF mutation. I do think that we’ll be in an era pretty soon where BRAF-mutant colon cancer patients will be treated with a BRAF inhibitor as part of their regimen.

DR LOVE: And I see that you’ve been involved with work looking at — first I see encorafenib and cetuximab, another trial looking at dabrafenib/trametinib. What do you see with these patients clinically, particularly with their skin?

DR WAINBERG: So these are not easy drugs. I mean, the simple problem is that the skin toxicity is a big one when you add in 2 drugs that cause skin toxicity. So all patients need to be really — we’re very aggressive with the antibiotics in these people from the get-go, perhaps even more aggressive than we would be if they were on single-agent EGFR inhibitors.

The dabrafenib/trametinib combination, which is — probably a lot folks know pretty well about that combination from multiple cancers — is there are some other issues beyond rash that are challenging. I would say fevers are a very common problem, as is fatigue and LFT abnormalities. So those drugs need to be dosed carefully.

DR LOVE: Could I just ask you — I mean, that combination’s actually approved in lung cancer now. And some people are using it first line. But would you use that combination outside a trial setting in BRAF-mutated colon cancer?

DR WAINBERG: Not yet. I think if I was going to use a BRAF inhibitor in colon cancer off trial, I would choose Scott’s regimen. I would choose the irinotecan/cetuximab/vemurafenib, because I think that is probably the furthest along. It’s got good randomized data. And I personally think the idea of giving some cytotoxic chemotherapy in those people is probably a good idea. And by that I mean if they’re irinotecan naïve in particular and they’re BRAF mutated, irinotecan/cetuximab/vemurafenib to me is a good way to go.

There’s a bunch of trials. There’s actually a randomized trial. Encorafenib is doing a randomized trial with their MEK inhibitor and with cetuximab. And they’re randomizing people to irinotecan/cetuximab versus their triplet or their doublet. So there is a lot of effort now going into this.

Emergence of checkpoint inhibitors as treatment options for MSI-high mCRC

DR LOVE: So, of course, this patient had another option, because of the MSI-high status. What happened?

DR WAINBERG: This patient got referred to me and I sent them back, actually, to the community oncologist while he was getting FOLFIRI. And then this business of the MSI started to emerge as a real thing, as a real actionable thing as opposed to just a hypothetical theoretical. That was the early days, I’d say, in 2014 or mid-2015, I guess, a couple of years ago.

And the data started to emerge that these people with MSI had some anecdotal responses to checkpoint inhibitors. So we tried to get this patient on a clinical trial, actually, because that would have been the ideal way. And then he was excluded from clinical trial participation, because he had some comorbidities that were considered ineligible for a clinical trial. So we ended up, at that point, applying for nivolumab. And by that I mean once it became clear that the nivolumab was active in those patients, it became relatively easy, as we all know, to get it.

So we start him on that. And this has been I’d say my best response to nivolumab. He’s been on that drug for 2 years. And he has disease that responded immediately. After 1 to 2 doses, his liver disease started to shrink. His CEA started to come down. He had evidence of clinical response, too, because, by the time we started, he was already having some clinical problems.

DR LOVE: Any tolerability issues, any autoimmune toxicities?

DR WAINBERG: Other than the laboratory abnormalities of some mildly elevated TSH, he hasn’t had any, so a very well-tolerated patient.

Tolerability of checkpoint inhibitors in gastrointestinal cancers

DR LOVE: What have you see in general with checkpoint inhibitors? I’m guessing most of them have been in MSI patients. Anything unusual in terms of toxicities or problems?

DR WAINBERG: I mean, I think now a year ago, people were even using this in MSS patients, because there was some hope and I dosed some myself. And I haven’t seen any benefit there. It’s clear that doesn’t work single agent.

I think in the MSI high, the responses are pretty striking, I find. Not every patient, but a majority of patients have some clinical benefit even if, like this patient, used in the third line. I mean, there are some toxicities, obviously. There’s some cumulative fatigue and some patients can get your pneumonitis and colitis. I haven’t seen it that much in these MSI patients. And to be fair, my experience with these drugs is usually in the GI settings, which is a little different.

Treatment duration and scheduling of checkpoint inhibitors

DR LOVE: Any thoughts about the possibility of discontinuing therapy in this man? And in general, how do you think through that issue?

DR WAINBERG: That’s a good question. And I think what we’re going to have to —that’s the next topic of discussion, particularly in MSI high, is that the data now that we have doesn’t really — we don’t have any data on should you discontinue patients who still have active disease. So he still has liver mets that are clearly visible.

He’s been on it, though, for 2 years. And there are randomized trials in other malignancies that are looking at that very question, 1-year versus 2-year discontinuation. And I think biologically, this is the kind of patient who you could probably stop it and perhaps sustain the same degree of benefit. But I was probably going to do that, pending that result in the other cancer that showed if there’s no difference between the patients who stop at 2 years or 1 year.

I was going to suggest that in this particular patient, because it’s been a long time and I wonder if we’re doing anything — we’re not seeing further response. We stopped seeing that long ago. We’re seeing continued stable disease. He still has liver issues.

DR LOVE: Any issue in terms of the duration between therapies when you’re using nivolumab? Right now it’s every 2 weeks. Here’s this man coming in now every 2 weeks at more than 2 years. I’ve heard people talk about using every 4 weeks. What do we know about prolonging the schedule?

DR WAINBERG: So I think in certain cancers that’s now becoming a backbone. And it’s clear that in pembrolizumab, they have a fixed dose every 3 weeks. So with nivo every 4 weeks, I think there is pretty good data for that. He doesn’t really come in every 2 weeks for 2 years in the sense that this is the kind of patient we give frequent breaks to, because he’s doing so well.

DR LOVE: Really?!

DR WAINBERG: So he travels. He — yes, I’d say he comes in 2 weeks and then takes a break for a week and goes on vacation and then comes — I mean, he’s living his life.

DR LOVE: Wow! That is really interesting.

DR WAINBERG: So he’s able to travel and do the things — he’s retired, so he gets around.

DR LOVE: I never thought about the idea. Obviously with chemo, in colon cancer, we had all those old trials on treatment holidays. I never thought about checkpoint treatment holidays.

DR WAINBERG: I think it has to be individualized, right? So if you have a patient like this, who’s clearly been going for so long that he needs a break on occasion — and this is the kind of patient where I’m very comfortable doing that. There’s just — once we got past 6 months — and that was the interesting thing. We saw some response early on and then even better response by 6 months.

After that, he hasn’t really — the lesions are there. They’re not shrinking. We scan him every 4 months or so. And so we get a sense that the best response is — I don’t think we’re going to make him a complete response no matter what we do.

DR LOVE: Obviously more and more, we’re seeing patients who in the past had poor prognoses. You occasionally have things like this happen, particularly with checkpoint inhibitors. We usually focus on all the data, the trials, et cetera. I’m just kind of curious at a human level how this played out in this man. What did you observe in terms of what it actually meant to him to be able to go through this response?

DR WAINBERG: Yes. I mean, he’s overjoyed. I mean, before the trial, he was progressing on FOLFOX and FOLFIRI — not the trial. Excuse me — the checkpoint inhibitor. And I think it’s very interesting how the field changes so quickly and in — just like you said, in a human level. He was told he had those tough discussions that, “We have continuing growth of your liver metastases and we have very few options.”

And he was offered some trials or some other drugs that are available. But these patients are very gratifying. And we all wish there were more of these MSI high now, because these are few and far between, unfortunately.

Novel research strategies with checkpoint inhibitors in CRC

DR LOVE: Maybe we can talk a little bit about some of the clinical research strategies trying to get more out of checkpoint inhibitors, both in MSI high as well as MSS tumors. There are some trials out there looking at MEK inhibitors, et cetera. But in both of these, what are some of the clinical trial strategies right now that you feel optimistic about?

DR WAINBERG: So in MSI high, I mean, the challenge, I think, in MSI high is, in many ways, the success of single-agent checkpoint inhibitors. It makes it very difficult to do much better. And by that, I mean every drug company wants to add their second drug or their better checkpoint inhibitor, looking at MSI high. And to me, it’s not that every patient responds as durably as this patient.

But there’s enough of a response in the majority of these patients that I find it difficult to imagine that a double checkpoint inhibitor or a double immunotherapy strategy, when you weigh in the toxicity and cost of these drugs, is going to be the best approach. And that’s a struggle, because everybody wants these MSI-high patients now on their studies, obviously.

The MSS patients are obviously the exact opposite in the sense that we know that single checkpoint inhibitors don’t have any activity. And there’s all this interest amongst every company out there, pretty much, and trying to make these tumors immunoresponsive. But there’s a reason they don’t respond to immunotherapy. And it’s very challenging to turn these kinds of patients to immunosensitive diseases, I think.

So there’s a lot of strategies out there and, obviously, some include chemotherapy backbones together with immunotherapy. Some include double immunotherapy strategies, and some include immunotherapy plus targeted therapy strategies. I think the key will be to think about the cancers that are not quite MSI but have some characteristics that would make them sensitive to checkpoint inhibitors or other immunotherapeutic strategies, so looking at now in the field of colon cancer, where there’s been a lot of progress in subclassifying tumors, whether it’s based on a CMS subtype, which is a popular thing now, or other subtypes.

There’s these groups. And there are groups that theoretically are more likely to respond to checkpoint inhibitors and maybe not by single agent but in combination. So that’s where I think the potential holds.

DR LOVE: So you mentioned combination immunotherapy. And, of course, in many different cancers we hear about anti-PD-1/anti-CTLA-4. Has that been looked at either with MSI-high or MSS tumors?

DR WAINBERG: I mean, I think it has been looked at in MSS. And there’s been no suggestion of any activity. In MSI high, it’s also been looked at. And again, there is some suggestion and I think that data’s been presented with ipi and nivo that there is a better response, perhaps, as indicated by response rate. Questions become challenging when you start to consider the additional toxicities, because adding a CTLA-4 inhibitor is much more, in my opinion, a much more toxic and challenging regimen, particularly for durability, than single-agent PD-1 inhibitors alone.

So the question is, is the extra small, perhaps, prolongation of PFS in very small numbers going to be worth the toxicity and the extra cost? That’s a challenging question.

Gastrointestinal microbiome and new drug development

DR LOVE: So I’m just kind of curious. I’ve been hearing mention of the so-called biome in colon cancer and in the bowel in general and the research that’s going on there. Just thought I’d give you an opportunity, if you have any interest or thoughts about that.

DR WAINBERG: I mean, it’s very interesting. It’s, I think, probably unique to the GI system. So in many ways, if we’re looking at GI malignancies — and perhaps not just colon cancer, but gastric cancer as well — the microbiome has always been this area of interest. And now the challenge is understanding it better and developing therapeutic strategies against it.

I think we’re still a ways away from that point, but there is good logic in trying to target the microbiome or using microbiome as a way to develop some new drugs. I think the research is pretty provocative but very preliminary still.

Treatment of MSI-high mCRC that progresses on checkpoint inhibitor therapy

DR LOVE: So just kind of curious how you would think through this man’s case, if he were to come in with clear-cut disease progression.

DR WAINBERG: If he progresses on a checkpoint inhibitor, because I have thought about this and I’ve said to myself, “If this man progresses on a checkpoint inhibitor, given the long” — see, when I have an MSI-high patient who’s been on therapy for a long time, on a checkpoint inhibitor, I think that’s the perfect candidate for some sort of enhanced immunotherapeutic strategy.

So because he was on for so long, adding perhaps a second checkpoint inhibitor or changing around the checkpoint inhibitors that are out there makes sense, because as opposed to going back to chemotherapy, at least for now, because to me, if you have an MSI-high phenotype, you’re clearly more likely to respond to these strategies.

I tend to think there are — I have had patients who progress. I’ve had a number of patients who have been on checkpoint inhibitors, and then they progress. And the progression is not usually as dramatic, in my experience, as it is when you progress with MSS cancer, in the sense that you have little time to make your decisions. There is some opportunity to look around and find some better strategies.

DR LOVE: So you were talking about further immunotherapy. On a trial, or you think you might think of something outside a trial?

DR WAINBERG: So at that point, I would probably try to find a trial, because I think that — so the only thing you could really do off trial that is easy to do, I suppose, is combine the CTLA-4 at that point. And that’s something to consider, if you can have that option and if the toxicity is worth it. So you could add ipi to that regimen based on that previous data that I alluded to.

There are other checkpoint inhibitors out there now that are distinct from PD-1 and CTLA-4 that actually look promising, whether it’s TIGIT or one of these other checkpoints adjacent to the PD-1 blockade. And once you identify a patient as being immunosensitive, as rare as it is in this disease, I think it makes sense to consider at that point another strategy.

TAS-102 after disease progression on standard fluoropyrimidine, oxaliplatin and irinotecan-based regimens

DR LOVE: What’s your usual approach in general to patients once they get beyond FOLFOX, FOLFIRI/bev, EGFR if indicated?

DR WAINBERG: In MSS patients, you mean?

DR LOVE: Yes.

DR WAINBERG: In all comers. So in patients that progress on all that chemotherapy, I tend to use outside of a trial if there’s no option, I tend to use TAS-102 in that patient. So in the patients who have been on standard chemotherapy, which is the large majority of our patients, of course, I prefer, personally, using TAS-102 as opposed to regorafenib if that’s the sequence of decisions that need to be made.

DR LOVE: Any clinical pearls that you have that maybe you pass on to your fellows about using TAS-102?

DR WAINBERG: I find TAS-102 to be a little bit easier than regorafenib, obviously, in toxicity profile, because I think we’re much more comfortable dealing with the cytopenias that might emerge. I think at that point, a large majority of the patients are obviously in different shape and not clearly — we’re not — I tell our fellows, “We’re not shooting for the moon here, unfortunately. We have to be very practical with this and consider this something that is hopefully going to stabilize the disease, but still, if you look at the data, a good chunk of the patients will progress at first scan. So don’t go overboard with the cytopenias and toxicity in a patient like that where you’re really not shooting for a response,” quite frankly. The majority of the patients will not respond.

Dose reductions for TAS-102-related cytopenia

DR LOVE: How do you approach the issue of cytopenias, specifically neutropenia, and what about the patient who’s been heavily pretreated? Do you see more problems there?

DR WAINBERG: Yes. I mean, we always see more problems in those patients. So I think that we tend to come down to the next dose level down pretty fast in that drug. Assuming they’re on the regular schedule, 5 days on, 2 days off, 5 days on, 2 days off, typically a day after their first 4 weeks of therapy, their second go-around, a lot of the patients will have 1 drop in a dose reduction if the cytopenias have become an issue, because the last thing the patients want to have to deal with is coming in and worrying about a CBC twice a week.

So if you can avoid that at that point and just — I often find CEA to be valuable in those patients that are in a situation in the absence of any clinical progression, using that as a way of looking for disease stabilization, too.

Initial dosing of regorafenib

DR LOVE: You mentioned regorafenib. First of all, how often do you see or do your patients end up getting both, sequentially, TAS-102 and then regorafenib?

DR WAINBERG: I’d say increasingly rarely. And by that I mean I think that even though both drugs are FDA approved, it’s difficult to have a patient that is in sufficiently good condition that’s going to get through both of those salvage drugs. So I’d say my experience, it’s only about 10% of patients that I find where the toxicity profile and the benefit is, in that situation, worth doing.

DR LOVE: What is your experience with regorafenib? And how do you dose the drug?

DR WAINBERG: So in regorafenib, we all know in many ways, not an easy drug to dose. Whether it’s the patient population that we tested it in or the drug itself has some toxicity, some of which are not Grade 3 toxicities but clearly affect quality of life. I start every patient, personally, at 120, so I do not start on the package label, FDA-approved dosage of 160. I maybe saw 1 patient who can handle 160. And if they did, I started them on 120 and then put them on 160.

So 120 mg of regorafenib, to me, is where I start the majority of my patients. And some patients need to go down quickly to 80 mg. And a lot of patients can hang in there at 120, at least until we see if the drug is doing any clinical benefit.

RAS testing in colon cancer

DR LOVE: I want to move on and ask you to comment on the issue of RAS testing in colon cancer.

DR WAINBERG: I think for about 10 years now, or more than 10 years, that RAS oncogenes are certainly the dominant oncogene in cancer and, in particular, in GI cancers, colorectal cancer. This is the most relevant, I would say.

RAS is a very complicated protein, as we all know, but generally speaking, it’s divided into HRAS, NRAS and KRAS. And KRAS mutations make up the bulk of them in exons 2/3. The NRAS is relatively newer — I’d say in the last 5 years — where it became clear that there’s a subset of patients that have NRAS mutations as well. And those are also all the same point mutations.

So if we look at colon cancer, and we all know that about 40% to 45% of patients will have a RAS mutation and perhaps a little bit higher when you add in NRAS now. So nearly half of patients will have some degree of mutations in one of the RAS oncogenes. They’re usually in certain codons, as we all know: 12, 13, 61 are the most common. And I think nowadays it’s a good idea, obviously, when we test for RAS, to ask the pathology to do — whether we want to call it all RAS testing or extended RAS testing. And that usually includes all the KRAS codons as well as NRAS and BRAF.

DR LOVE: Could I just ask, how much of this information would you get from a next-gen sequencing assay like FoundationOne?

DR WAINBERG: Oh, you get all of it. So that’s standard now, to look at — Foundation or even a lot of the internal hospital pathology departments will be able to give us — certainly everybody knows how to do KRAS. And I would say NRAS and BRAF are becoming more frequent even in community pathologists/hospitals, because they’re very easy to isolate. Certainly any NGS will give you — if they have an NRAS mutation, it’s going to be on there.

RAS oncogene in the pathogenesis of colon and other cancers

DR WAINBERG: So we know that RAS is, like I said, the dominant oncogene in colon cancer. And presumably, it plays an important role in the progression of the disease. And I think most of the data, if you look in aggregate at the literature, there’s pretty good concordance between studies. And there’s pretty good concordance between the sensitivity and the concordance between a metastasis, for example, and a primary tumor. It’s not perfect, of course, but it’s pretty good at this point.

DR LOVE: I was curious about this point you made here about KRAS being part of the evolution of polyps.

DR WAINBERG: Yes. So if you look at the old typical Bert Vogelstein hypothesis, progression from polyp all the way through adenoma, whether it’s dysplasia and ultimately neoplasia, RAS is on that progression and fairly early, actually, in the point between when the mucosa changes. So it’s right after APC, I think, in terms of the frequency with which that has been observed. And pretty elegant scientific studies show that it’s common to occur along that progression.

DR WAINBERG: And this is the bulk of the patients that have RAS mutations, colorectal, lung and pancreas. We know that, for example, in pancreatic cancer, virtually 95% of them have a KRAS mutation. And in colon cancer, it’s about — like we said, about 40% or so. And then it’s usually G12D and G12V and G13. And so those are the typical ones that when you look at the hot spot analysis or whatever mechanism we’re using to look at RAS, you’re going to see those present.

Lung also 25%. So I think everything we’ve learned from lung cancer in these situations should be also extrapolated on some level to colon cancer. We certainly know the pathway pretty well, the MAP kinase pathway, that blocking RAS and what would be a nice, attractive idea in theory — and it’s just upstream of BRAF, so patients that have RAS mutations do not have BRAF mutations as well. As you know, those in colon cancer are mutually exclusive.

So blocking BRAF, if you have a BRAF mutation, at least in — or I should say in the testing mechanism, a lot of centers do not test for BRAF if you already have a RAS mutation.

Oncogene addiction in BRAF-mutated colon cancer compared to other solid tumors

DR LOVE: Is there anything about the pathophysiology or the pathology of colorectal cancer that you can point out in terms of why it doesn’t seem to kind of respond the way, say, melanoma or lung cancer does to anti-BRAF treatment?

DR WAINBERG: Yes. I mean, we know that BRAF-mutant colon cancer is so different from BRAF-mutant melanoma, both probably biologically and all the animal models. And research has suggested all sorts of — even if you block BRAF in these patients, you have immediate increased pathway activation of some other pathways, whether it’s PI3 kinase or EGFR. And I guess that was the impetus behind the combination trials in the early going.

But I think to me, even in the BRAF-mutant colon cancer patients, the notion of true oncogene addiction is less relevant here, than in some of these other cancers that you alluded to. And so blocking one single pathway in colon cancer has not been fruitful, in our experience so far, as you know. Even in the BRAF-mutant combination studies, you see that the patients do progress and often much quicker than they did in other BRAF-mutant malignancies.

So if you’re talking about the BRAF, I think it’s a totally different paradigm than these other cancers. It’s really not as dependent on that pathway. It’s dependent on the pathway, but not to the same extent as some of these other cancers.

Efforts to develop KRAS-targeted agents

DR LOVE: Can you comment on the issue of drug development to treat KRAS-mutant colon cancer?

DR WAINBERG: Yes. And there’s been probably dozens of companies and lots of money spent on anti-KRAS strategies, which is a big challenge, obviously. It’s not the easiest and most druggable target, as we all know. We imagine that blocking this target in a successful, durable way that can be administered to patients would be a good approach to treating KRAS-mutant cancers, although we don’t know, to be fair, because very few of those drugs have made it to the point where they can be administered safely or we’re confident that they sufficiently block the pathway.

Some efforts have made the assumption if you block downstream with a MEK inhibitor, that certainly will cover your RAS-mutant cancers. But that hasn’t been the case, as we all know. MEK inhibitors as single agents have not shown any activity at all in RAS-mutant colon cancer and pancreas cancer. So efforts to block RAS are — that’s why great scientists are spending their careers dedicated to that problem. And there are a lot of them.

And I think there is a way to figure it out. And it’s just going to take some new pharmacology, new drug delivery mechanisms maybe. There’s some excitement in the immunotherapy community about thinking about RAS not as a druggable target with a typical drug but rather with some adoptive T-cell strategies. So there is some excitement in this field that is renewed. We all want to see some durable RAS inhibitor make it to the point where it can be given to patients and be sustained.

Case: A 55-year-old woman with moderately differentiated N1 colon cancer and 2 of 24 positive lymph nodes receives 3 months of adjuvant FOLFOX

DR LOVE: So I wanted to ask you, kind of get your viewpoint on a couple of other common questions that come up with colorectal cancer.

DR WAINBERG: Yes.

DR LOVE: Perfect. Okay. One is the issue of duration of adjuvant chemotherapy and the 3- versus 6-month question. And I was just wondering if you had any patient you’ve seen, let’s say, since ASCO where you actually utilized 3 months that you could maybe briefly present.

DR WAINBERG: I did. I’ve had a couple where this has already come up. And the data, to me, is a bit hard to figure out what’s going on there. I mean, I find that it’s an incredible data set. And an incredible endeavor that the international GI community put together here. I mean, we’ll never have that again, probably.

And the idea, collaboration, I think, is so unique that people don’t realize how many patients went into this. It’s unbelievable. But I took home from that what a lot of people took home from that, which is that 3 months of chemotherapy in some people is probably just as good. And those people are typically the lower-risk, so the N1 patients, the patients with 2 lymph nodes or below who you know that oxaliplatin is going to induce worse neuropathy if you keep it going. And up until last year, every patient, you told them, very straightforward, FOLFOX for 6 months and that was the end of the story.

And I think what I take out of that is these N1 patients — I had one just a few weeks ago, where I’m planning to give that patient 3 months of FOLFOX.

DR LOVE: Could you maybe briefly present, right off the top of your head, that patient, what happened?

DR WAINBERG: So that’s a patient, very straightforward patient that we all see every day. This is a patient who came into the office having had a left hemicolectomy for a sigmoid cancer, had 2 out of 24 lymph nodes positive.

A 55-year-old lady, intelligent, full-time working attorney and has a typical presentation of colon cancer in the sense that some obstruction type of GI symptoms with a little bit of rectal bleeding. Gets taken for a scope and shows a mass in the sigmoid colon, no metastasis.

It’s a great situation in many ways for that patient. She gets taken quickly to the operating room and a very moderate typical — I’d call it a bland colon cancer, the ones we’re familiar with seeing often, which is a moderately differentiated adenocarcinoma. It was removed with 2 out of 24 lymph nodes positive, no other warning signs there in terms of suggestion of any metastatic disease.

DR LOVE: Just out of curiosity, had she ever had a screening colonoscopy?

DR WAINBERG: No, so she was in her early fifties and “not yet” is the answer I got.

DR LOVE: Just out of curiosity — she’s an attorney and intelligent, I guess — any regrets that she expressed to you about not getting screened?

DR WAINBERG: I mean, I asked her if this was her primary colonoscopy. And it was the typical answer, which we face with — it was, “It’s been on my list of things to do.”

DR LOVE: That's interesting.

DR WAINBERG: Yes. I mean, a lot of patients, we all know that they are told over and over to get their screening colonoscopies. And they’re reluctant to do it. And you’re right. I mean, that’s a patient who maybe a few years ago had Stage I colon cancer or something like that. Who knows? There are some regrets, but I think that the positive side of it is that she’s still very — good chance, likely cured, is —

DR LOVE: So she’s N1. She’s got 2 positive nodes. What about the choice of therapy and specifically whether you used capecitabine or 5-FU?

DR WAINBERG: So I will usually use FOLFOX as — I mean, I’d say if I had to look at my patients, it’s probably three quarters to a quarter FOLFOX over CAPOX. I just find it easier for the majority of patients in terms of toxicity profile and especially in someone who’s undergone surgery. I’m always worried about getting adequate capecitabine doses in.

DR LOVE: I’ve heard people say that in this incredible analysis, that the 3 months of capecitabine looked better somehow, or at least the CAPOX, the way it was delivered looked better. Do you buy into that, or you think that’s overinterpreting?

DR WAINBERG: I mean, I think that’s — okay. So I don't know what to make of that yet other than I think we all know that capecitabine dosing and tolerability is a little bit different in Europe than it is in the United States. So those were European patients for the most part. And in the United States, for a bunch of reasons, as we all know, getting to those big doses of capecitabine isn’t easy.

So when you’re trying to treat for cure, sometimes — I often will go for FOLFOX. Now, in this patient, I did start with CAPOX because she was so opposed to having the port placed or a PICC line placed. So we did decide to do CAPOX. And after 1 cycle, I switched her to FOLFOX, because the problem became — this happens in a lot of patients, that getting that adequate dose of capecitabine in for 2 weeks, she started to develop a lot of GI toxicity with colitis type of picture, diarrhea and hand-foot syndrome. And she’s not DPD deficient or anything. It’s just in some people a tough combination.

And she had the expectation that she’d be fully working during this period, because she didn’t have a portacath or a PICC line. And then after seeing her response to this combination, I told her, “The easiest and safest way for you to get through these next 3 months and to do it properly will be to do FOLFOX.” So I talked her into that.

DR LOVE: What dose capecitabine did you use?

DR WAINBERG: So I used 1 gram BID. One g/m² BID. So it came to — I don’t remember the exact dosing in her, but it was at least 1,500 BID, which isn’t a huge dose, but — it might have been actually — I don’t remember the exact dosage in her, but it might have been 4 and 4, even. I think capecitabine is, in some people, the right choice and in some people isn’t. And the way I approach it is, if a patient is able to swallow pills and they want to try it, I try it, for the most part.

But I find that a lot of patients, if you really want to get them properly through their adjuvant therapy, you sometimes have to make some decisions there, which include either dose reducing your capecitabine to the point where you’re worried that you’re not getting adequate dosage or switching them off of it.

DR LOVE: It’s interesting, because your views are very similar to what I’ve heard from other GI cancer investigators, and yet I would guess that if you did a patient preference survey, you would find people — I mean, theoretically, assuming it was the same side effects, which, as you say, maybe is an issue, that there would be — think about this woman, a strong preference not to have a port, et cetera. And I just wonder. I realize it’s in the curative setting, but also the benefit, whether — how much additional benefit you get by pushing the dose. Of course, we don’t know, but I kind of wonder about it.

And thinking back to what you were saying before about regorafenib, obviously different situation, palliative, but yet the idea of starting at a dose you know will be tolerable and maybe working your way up, I guess that just doesn’t wash in the, quote, curative setting.

DR WAINBERG: Right. I mean, I think it’s harder to justify doing that in the curative setting, because — and perhaps the number isn’t as great as we would like, but you know there are some patients where you’re going to cure them with chemotherapy, just like in any adjuvant scenario. And particularly in a young patient, we try to push the envelope and try to give them really the proper dosage. And who knows what that dose is, to be fair?

The other problem with CAPOX in the adjuvant setting, as you know, is that that dose of oxaliplatin every 3 weeks at 130 mg/m² isn’t easy, either. So sometimes I find that that interest in not having the pump or the port and all that is an emotional reaction and not based on the practical reality, which is, let’s just get you through this safely and well and do it properly.

So it’s interesting. You have seen some switch away if you look nationally from CAPOX to FOLFOX, and whether it’s adjuvant or metastatic and as more people start thinking about it a little differently.

Degree of benefit needed for patients to undergo adjuvant chemotherapy

DR LOVE: So little-known fact: Our group does a lot of surveys. And we’ve presented many posters at ASCO and other meetings on just kind of asking people what they do and stuff. We had one data set that I actually presented orally at the ASCO GI meeting a few years ago. And I completely screwed up the whole presentation. I got lost in the slides. It was totally out of my element. But it was a kind of an interesting study. I don't know if you ever — you probably don’t —

DR WAINBERG: Yes. I remember some of the details.

DR LOVE: Oh, really? Anyhow, I actually ended up getting commented on. I think we had a poster discussion at ASCO this year. But we had a lot of experience in breast cancer and seeing the literature on how breast cancer patients make adjuvant decisions.

So we actually surveyed some patients with colon cancer who’d gotten adjuvant therapy and said, “Thinking about that experience, how much of a benefit would you have needed to see in order to justify going through this kind of therapy?” And in breast cancer, they found that there were a cohort of patients who would go through the chemo for just a very, very minimal benefit. So I don't know. Do you know what we saw with the study?

DR WAINBERG: I’m going to guess, but, I mean, we know that in breast cancer, patients will endure greater toxic regimens for a smaller benefit. And the adjuvant colon cancer data is stronger in terms of number needed to treat and all of those variables. So I’m guessing that you found that a larger percentage of patients would go through it again.

DR LOVE: Actually, what was interesting — there were a couple of things that we found, first of all, that about a third of the patients would go through it for a 1% reduction in risk of recurrence, but 10% of the patients would not go through it for, like, a 10% absolute benefit. So we saw both sides of the spectrum.

I think the most interesting thing that we saw — I was really curious — was after all this stuff in breast cancer, was whether there was going to be a difference in men or women. And we found no difference. The attitudes were the same. And I’m just kind of curious if that, in any way, reflects what you see in day-to-day practice in terms of the variation.

We didn’t see a lot of age difference, incidentally, although maybe it wasn’t perfect science. But I’m just kind of curious how you see people — this 55-year-old lady and other patients — kind of coming into these adjuvant decisions.

DR WAINBERG: I see no difference in gender. We see each gender equally in our disease. And I see no difference in how they approach the topic or how they think about it. Your data will be particularly interesting in light of the 3-versus-6 question, because presumably that 10% will go down dramatically.

Counseling patients about adjuvant chemotherapy

DR WAINBERG: I think — look. It’s always hard to have that adjuvant discussion. It’s a conceptual discussion. It’s hard for patients to understand who are outside of the field. We teach our fellows to talk to patients about the concept of an insurance policy and all those kinds of typical teaching points. But it’s a tough discussion.

And I find that, though, in this day and age, the very, very large majority of patients will choose some therapy, because they’ve been able to understand that — they’ve either seen people or know people who have had a cancer come back. And they want to avoid that fate. And so there’s a lot of buy-in to that point.

And I think the 3 months is going to help, because I think that when I tell patients, “Oh, you’re going to need adjuvant chemotherapy,” and they say, “For how long?” And I tell them 6 months, I see their face drop, because a lot of them think they’re out of the woods and they’re done. We all know that, that they’re referred from the surgeon, “It’s great. We got it all. It’s done.” And a lot of them think it’s over.

And some of the surgeons will say, “Go talk to your oncologist about the pros and cons of chemotherapy,” which is — then we come in and we always say, I would say with rare exception, the patients with a positive lymph node in colon cancer, we always push the chemo, whether it’s — you could have a debate on FOLFOX, XELOX or single agent or doublets, but we always say, “You’ve got to get something.”

And so they come in a little bit shocked at the duration, I find, because some of them have this concept that it’s going to be a few months shorter. So now I think you can have that discussion. And that’s a nuanced discussion. That, I think, should be individualized to the degree of risk, just like so many other things we do in Stage II colon cancer. It’s the rare patient now with Stage II colon cancer who gets treated. But I would say it’s the high risk.

So in the same way, if I have a patient with 8 positive lymph nodes, I’m not having, usually, a discussion yet about 3 months.

DR LOVE: And as you say, it’s really interesting. We should go back and do a study like this on 3 versus 6 months, because then it would be interesting to interview people who got neuropathy that hasn’t gone away with 6 months in terms of how much is that worth to them.

What about — I know there’s a big Asian population in your area. Any different ethnically?

DR WAINBERG: Yes. I mean, we see a big diversity here of patients. And in big urban centers, there’s sometimes a little more skepticism about chemotherapy, as you know. Whether it’s holistic approaches or alternative approaches, we get a lot of patients who are opposed to chemotherapy. So we do get some of that “No way, no how” type of answer. But I’d say ethnically, there’s no differences. We see no differences in terms of their willingness to accept it.

DR LOVE: It’s interesting, too, the whole dynamic of the adjuvant decision. You have it in lung cancer and breast cancer. I was just reflecting on the fact that yesterday I was interviewing Hal Burstein from Dana-Farber, a breast cancer investigator. And he presented a case to me of this woman who came in wanting adjuvant chemo, feeling she needed it. He does an Oncotype DX^® assay on her and it’s low. And he says, “You don’t need chemo,” and then she turns around and says, “I still want it.”

And he said probably the risk of secondary leukemia or whatever is going to be greater than the benefit. And yet that idea of the regret of if a patient develops a recurrence, I guess, is so dominant in this situation.

DR WAINBERG: Yes. I think it’s more dominant in breast cancer, to be honest, because, I mean, in colon cancer, it’s not as much decision-making goes into the decision fundamentally. It’s been, until now, a pretty clear consensus that patients with Stage III colon cancer get 6 months of FOLFOX or XELOX or CAPOX.

And so maybe now we’re going to have to start having those nuanced discussions, but it’s not going to be about chemo versus no chemo. It’s going to be about 3 versus 6. And what’s very interesting is, there are, I feel, this group of patients — and there’s no data on this yet, or not long-term data, but where maybe the 3 versus 6 is okay if you drop the oxali, but maybe some of them should get 6 of the 5-FU type of thing.

DR LOVE: Hmm. That's interesting.

DR WAINBERG: Because it is — the toxicity is pretty much from the oxali at that point. So I don't know if too many studies have looked at that or whether that’s a future study. I think it is. But this data set, which is the IDEA data set, is such a robust thing, they should be able to milk that for a lot of subset analyses to really nail down who are the people who don’t need that 6 months.

Case: A 58-year-old man with T3N1 sigmoid colon cancer receives 3 months of adjuvant CAPOX

DR HOCHSTER: So I personally am — when I do the 3-month treatment — I started somebody last week on this — I’m trying to give people CAPOX. And that was not my normal practice. I really don’t like capecitabine, because of the extra toxicity and the need for patients to take it twice a day for 14 days every 3 weeks and all those kinds of things. But I have really adopted that as a kind of standard for low-risk Stage III.

DR LOVE: I’m curious about this patient where you just started on 3 months. What was the patient’s situation?

DR HOCHSTER: He was a gentleman who had some rectal bleeding. And they found that he had a sigmoid colon cancer that was T3N1. And the N1 was deposit equivalent. There were actually 31 other lymph nodes that were negative. But he’s just into the Stage III category, so I feel pretty comfortable that his cure rate with 3 months of CAPOX will be equivalent. But for us it’s a little bit new, kind of trying to manage the capecitabine, because we’ve tended to always give people the 48-hour 5-FU infusion.

DR LOVE: I’m just kind of curious more a little bit about this man. How old did you say he was?

DR HOCHSTER: He’s 58.

DR LOVE: Fifty-eight. Okay. So he’s a younger man. And is he the kind of patient that if you said, “I’ve got 2 possibilities here. One might be a tiny, tiny bit better,” is he the kind of patient who says, “I don’t care what kind of toxicity it is, just give me everything,” or more like, “I want to maintain quality of life”?

DR HOCHSTER: It’s hard to say in this situation what the quality-of-life issues are. I mean, the cape could have more side effects than the 5-FU. But it’s 3 months and less long-term neuropathy. So I actually told him that we could do the regular 6 months, but I think that the 3 — the shorter course is really the equivalent in his case. And he was willing to accept the shorter course and take the pills. And I believed also that he would be very compliant with taking his capecitabine.

DR LOVE: Hmm. Let me ask you: If you see a benefit with capecitabine — and it’s always been interesting how US investigators don’t like CAPOX, and for the same reasons that you were just mentioning — but if you think there’s a slight benefit in the lower-risk patients, then why wouldn’t you give it to a higher-risk patient, too?

DR HOCHSTER: The data showed that with the 6 months, it doesn’t really pan out to be much better. It’s really in the 3 months that you see a much better outcome with the capecitabine compared to FOLFOX.

DR LOVE: I mean, globally, do you believe that — I mean, chemotherapy causes a relative risk reduction in general, that, for example — I don't know — 3 months with CAPOX, maybe you’re looking at, say, a 5% relative risk reduction, or 10% at the most difference? Do you think that that then, whatever it might be, if that’s what it is, you then could just apply it across the absolute risk so that if it’s a lower absolute risk you get less benefit, higher higher, or you don’t think it works that way?

DR HOCHSTER: I think that’s a question that we’ve debated a lot. I tend to believe that the relative risk reduction is the same, be it Stage II, Stage III. So today with Stage II, people appear to be at higher risk by either pathologic or Oncotype, whatever. I will also offer them adjuvant therapy. So yes, I think the relative risk reduction is the same no matter if it’s Stage II, low-risk Stage III or high-risk Stage III. But giving 3 months of treatment may not be the equivalent for the higher-risk patients.

DR LOVE: So any Stage II patient you treat, you’re going to treat for 3 months now?

DR HOCHSTER: I would. Yes.

DR LOVE: Okay. That’s at least a practical personal guideline anyhow. We’ll see how it actually plays out in practice.

SWOG-S1613: A Phase II study comparing pertuzumab/trastuzumab to cetuximab/irinotecan for HER2-amplified mCRC

DR LOVE: I want to ask you also where things stand right now with HER2-positive amplified, but maybe mutant, also colorectal cancer, where we are today in terms of how often this occurs and what we know about interventions. And I’m really interested to hear about the new SWOG study looking at that.

DR HOCHSTER: Thank you for asking that. So we’ve been working on developing a study in SWOG for the last year or so. And we finally managed to negotiate a lot of issues, including with the FDA. So this study should open up in the next 2 weeks. It’s SWOG-1613.

So what we know is that about 4% to 5% of colorectal cancer overexpresses HER2. And it’s probably a negative risk factor like in breast cancer. I mean, there’s a lot of preclinical data and some clinical data to suggest that. Additionally, we know that most of these will be RAS wild type. There’s very few HER2-overexpressed patients in the RAS-mutated group. So if you look at it that you take the half of the colon cancer patients that have RAS wild type, now we’re talking about maybe 10% that would be HER2 overexpressing. So that’s a reasonable group to study.

In Europe, they did a multicenter Phase II study of 27 patients, the HERACLES study, where they used trastuzumab and lapatinib. And it looked like the response rate was 30%, and another 30%, more or less, had stable disease, so some real benefit for heavily pretreated patients.

In the SWOG study, we’re looking for second- or third-line patients that have not had cetuximab previously. And those patients will be randomized to a control arm of irinotecan and cetuximab for the RAS wild-type group and — as all of them will be RAS wild type. And the experimental arm will be a combination of trastuzumab and pertuzumab, so that they’ll get a double antibody blockade of HER2. And we’re going to be enrolling 100 patients. And it’ll be a controlled trial, very similar to the BRAF study we recently completed. And we hope that will give a lot of really solid, controlled evidence as to the benefit of treating this patient group, specifically with therapy targeted to their HER2.

DR LOVE: I’m curious why you all chose to use pertuzumab and trastuzumab. In breast cancer that’s not too effective, unless you add chemo to it first. What about that?

DR HOCHSTER: Right. Of course, we don’t have clinical data to suggest one thing over the other. However, we haven’t seen a lot of benefit, either, with lapatinib. And I think that the kind of general feeling was that this was probably the best way to target the pathway.

DR LOVE: You mentioned the HERACLES study looking at lapatinib/trastuzumab. Have you yourself utilized it? I don’t even know if you can access these drugs. Have you treated any patients with this approach?

DR HOCHSTER: I haven’t. We haven’t been systematically screening for HER2, though we’ve started in the last couple of months.

DR LOVE: Now, what about HER2 mutations? You hear a lot about that in lung cancer. I’m starting to hear a little bit about that in breast cancer. What about in colorectal cancer?

DR HOCHSTER: Yes. Once in a while, you can find an actual point mutation in HER2 that’s an activating mutation. These drugs should be equally effective. They’re not specifically excluded if they have overexpressed protein. So it’s just, in this particular study, they have to qualify on the IHC test.

DR LOVE: And when you talk about overexpression, is it the same as breast cancer, IHC 3 or FISH greater than 2?

DR HOCHSTER: Yes. And we just submitted all the data to the FDA for an investigational device exemption.

DR LOVE: Just to clarify, though, you’re not impressed enough with the HERACLES result to actually try to use it? Or you think you couldn’t get it paid for, or what do you think?

DR HOCHSTER: I think that it’s an equally valid way to approach the blockade of this pathway, kind of with a vertical pathway versus dual antibody targeting. So I don't know that one’s better than the other, but this was just a study we were able to support in the US.

STEAM Phase II trial results: Sequential or concurrent FOLFOXIRI/bevacizumab versus FOLFOX/bevacizumab as first-line treatment of mCRC

DR LOVE: I’m curious to pick up on some of the research that you’ve been involved with, one being some work I know you presented earlier this year at the ASCO GI meeting, looking at the so-called STEAM study, S-T-E-A-M. What was that, and what have we learned from it?

DR HOCHSTER: That was looking at sequencing of first-line therapy with FOLFOX or FOLFIRI or FOLFOXIRI all with bevacizumab. So based on the European trials that showed — FOLFOXIRI/bev seemed to be better in terms of response rate and PFS and maybe even survival compared to FOLFIRI/bev, we wanted to look at if you really needed to give the triplet or you could use sequential doublets.

And so this was a randomized Phase II of 3 arms, FOLFIRI, FOLFOX/bev or FOLFOX 1 month, FOLFIRI the other month, alternating months. And it turned out that what they call concurrent FOLFOXIRI, which was giving all 3 drugs at once, or the sequential, the monthly alternating one, did have a better response rate and PFS than the control arm, which was FOLFOX. So this does seem to suggest that early introduction of irinotecan may be beneficial in patients who have very aggressive tumors or need rapid debulking, high response rate, issues, and that if somebody can’t really tolerate a triplet, you could probably get very close to the same results by alternating FOLFOX and FOLFIRI, one month of one regimen and one month of the other.

DR LOVE: That’s interesting about that alternating regimen. What’s your experience with it in terms of toxicity? And are you actually using it in your practice?

DR HOCHSTER: I haven’t really been using it in practice, because I think that FOLFOXIRI — I think there’s something actually pharmacologically that is an interaction between a TOPO1 inhibitor, like irinotecan, with oxaliplatin. So I like the idea of giving them concurrently. And if you’re used to giving FOLFIRINOX to pancreatic patients, giving FOLFOXIRI to a lot of well, fairly healthy colon cancer patients is really not more difficult, and it’s probably easier. So I haven’t really used the alternating that much. But I do have patients, a couple a month, that I will start on FOLFOXIRI.

FOLFOXIRI/bevacizumab for BRAF-mutated mCRC

DR LOVE: So you talked about earlier introduction of irinotecan, some kind of triplet, FOLFOXIRI/bev, for example in tumors with more adverse outcomes, prognoses. So I’ll ask you first what I have been hearing about for a couple of years now, BRAF-mutant disease. In general, are you using that approach there?

DR HOCHSTER: Yes, for a patient where we discover a BRAF mutation, I will treat originally with FOLFOXIRI, and then we would go on to a BRAF study, previously SWOG-1406, which closed. But now there’s another study that’s looking at a BRAF inhibitor with or without a MEK inhibitor.

DR LOVE: So here’s one that kind of surprised me. I just interviewed your colleague Dr Alan Venook yesterday. And I’m kind of still processing. I think he said to me that he’s doing it now for right-sided metastatic disease, not BRAF. I mean, BRAF also but right sided. He’s so convinced by the adverse impact that he wants to start with FOLFOXIRI/bev. Are you doing that? What do you think about it?

DR HOCHSTER: I haven’t really adopted a policy like that, like, to say all right-sided patients should get it. But I think it’s rational. And my threshold for using FOLFOXIRI in right-sided patients who have a RAS mutation would be pretty low, actually. So if a young, relatively healthy person comes in with disease that I’m trying to get to surgical resection and they have a right-sided tumor, I really wouldn’t hesitate.

DR LOVE: So I’m curious, though, that your trial calls for the use of cetuximab. I think I saw — maybe that was another paper — Johanna Bendell in. But in any event, what are you doing about investigators in the reaction belt?

DR HOCHSTER: Oh. Right. Yes. That could be an issue for people in the Southeast that have the reactions. Dr Bendell is from Nashville. And that’s kind of the epicenter of the reactions, so…

DR LOVE: I mean, do you allow panitumumab, or they just can’t go on the study?

DR HOCHSTER: If they aren’t prepared to try cetuximab, they couldn’t go on the study. We had to put on some parameters, so that was kind of the choice.

Onset and delayed recovery of neutropenia as an indicator of response to TAS-102

DR LOVE: So going through some of the other research that you’ve been involved with, I wanted to ask you: I see you’ve been involved with a number of papers looking at TAS-102. And maybe you could comment a little bit on what we’ve learned about this agent. In particular, one of the things I was really curious about was this issue of the neutropenia that’s seen with TAS-102 and the issue of does that predict a better outcome, and why? Can you talk about the work that you’ve done looking at that?

DR HOCHSTER: Yes. TAS-102 is really an interesting drug. It’s basically trifluorothymidine, so this was synthesized in the early sixties by Heidelberg, or the same guy who gave us 5-FU and FUDR. It’s another antimetabolite. But the half-life is so short that you’d have to give it by prolonged infusion. So it was kind of abandoned until the Japanese came up with a metabolic inhibitor to increase the half-life. So it’s a little bit those combinations of oral 5-FU with DPD inhibitors. But this drug has an inhibitor of thymidine phosphorylase, which breaks down the TFT. And so you can take it twice a day.

The drug’s schedule is 5 days a week for 2 weeks. So you take it 5 days, get 2 days off, then another 5 days. And that’s supposed to be every 28 days. Now, it’s not like the white count gets that low that people are getting neutropenic fevers and really get into a problem with neutropenia. The main thing that we’re seeing is kind of a delay in neutrophil recovery. So when it comes to week 4 when you’re supposed to restart the next cycle, the ANC is less than 1,500. And in all the studies, you had to wait then.

If it was 1,200 or 1,300, which we saw very frequently, you had to delay the patient another week. And so it turns out that for the patients that did have a little bit delayed recovery, their overall response rate was higher than the patients who didn’t. And that suggests that there’s some pharmacologic event going on here with those who have a higher AUC. You have lower ANC nadirs and longer ANC recovery and, therefore, those are the ones who benefit more. And there is some drug levels pharmacology from the population pharmacokinetic studies that were done by Taiho that suggest that’s true.

So we did another analysis not only in the pivotal trial but in our own institutional experiences, with Dr Grothey from Mayo and also from another institution. And it looks like we had the same experience. People who have delayed recoveries tend to have a little bit better response rate.

DR LOVE: From a clinical point of view, if you don’t see that, do you think about increasing the dose?

DR HOCHSTER: That’s a good question. I suppose that one could postulate that.

Ongoing studies combining TAS-102 with other agents in earlier disease settings

DR HOCHSTER: Personally, my interest in the drug is not so much continuing to use it in the heavily pretreated population but kind of trying to move it earlier in combination, because it is another antimetabolite that can pair with drugs like irinotecan and oxaliplatin.

And it’s different than 5-FU as a fluoropyrimidine. It hits a different target, so actually goes directly into DNA as a phosphorylated base. And it’s a lot more like gemcitabine in that way, in terms of the way it interacts with the DNA. So I think that these combinations may be used a little bit earlier when patients have had a lot of fluoropyrimidine already or a lot of 5-FU already. It’s noncross resistant in preclinical studies.

DR LOVE: Yes. I was going to ask you about that, because that makes a lot of sense. What are some of the combinations or the trials out there that conceivably could lead to a change in clinical practice? And what are some of the kind of decision points that you could see potentially changing with these trials?

DR HOCHSTER: We’ve done a study with a group at Memorial on irinotecan and bevacizumab with TAS, so TAS/iri/bev. And I was impressed that patients who had been heavily pretreated seemed to at least have stable disease or response. And so that’s a regimen that may go forward.

There’s a publication in Lancet in the last couple of months from the Japanese investigators who’ve had the drug longer. So they’ve already published the study now of TAS and bevacizumab, which looks active as kind of a combination similar to 5-FU/bev. And in Europe, they’re doing a study of TAS/bev versus TAS, which is kind of like the AVEX study that had been done previously. So that’s another study that’s going on.

We’re doing a little institutional study at Yale of TAS and oxaliplatin. And that seems to have been tolerated very well. And the Europeans are actually doing a multicenter study of that combination as well. So I think that we’ll be seeing some of these combinations being validated as options. And then it’ll be kind of figure out how to work those in as noncross-resistant regimens, if it’ll be in second line, alternating first line, kind of what, like, we were just discussing with the irinotecan. Now we have another fluoropyrimidine to work with.

Case: A 52-year-old man with refractory mCRC receives TAS-102, bevacizumab and irinotecan on a clinical trial

DR LOVE: So before we move on to some of the other papers you’ve been involved with, since we talked about TAS-102, why don’t we talk about one of your cases of a patient who actually received it, this 52-year-old man? So that’s your second case. I don't know if you know who that is.

DR HOCHSTER: Yes. Yes.

DR LOVE: Can you kind of briefly present what happened with this patient?

DR HOCHSTER: Sure. So he had rectal bleeding while he was on an NSAID in June 2011. And he had a clinical Stage T3N1, and he was treated with neoadjuvant chemoradiation followed by surgery and then adjuvant FOLFOX, pretty much the way we would treat a standard rectal cancer for T3 or node-positive. And he had a 5-cm moderately differentiated adenocarcinoma and no lymph nodes involved. So he was pathologic ypT3N0.

And then he did well for about another year, but he had a bone lesion then in his sacrum, which was treated with chemoradiation. And then he got 6 months of more, quote, secondary adjuvant therapy with FOLFIRI and bevacizumab. And that lasted about another year.

And then he had additional recurrence. So we went back to FOLFOX/bev, which he got, with a remission. Then he was on maintenance therapy with cape/bev for a while. In 2015 he got FOLFIRI again, with aflibercept. And then he got irinotecan and panitumumab, which he did well on but eventually progressed again. He was on the study of TAS and irinotecan for 8 months, and he did pretty well on that, but he eventually progressed. So he has the TAS-102 in combination and then eventually he got regorafenib after that but had a pelvic abscess, probably some kind of tumor perforation down in his pelvis.

Considerations in sequencing TAS-102 and regorafenib

DR LOVE: So he sounds like he’s a pretty complicated case, but maybe you could talk a little bit — I’m kind of curious how you approach the sequencing of TAS-102 and regorafenib in your own practice. What kind of patients might you start one on as opposed to the other? Do most people end up getting both?

DR HOCHSTER: I think that many people can get both. Some people will progress on the first one and won’t make it to the second one. So that’s obviously the question. First of all, we have some data from — because the TAS approval was later. So we have some data from the pivotal trial for TAS-102 versus placebo that the patients who had gotten regorafenib — and that was about 17% of the patients — did just as well as the people who hadn’t got regorafenib. So that’s the one hard data we have in this discussion.

I think after that, it’s kind of more what you believe about what the patient’s tolerance is going to be, because in both those studies of the drug versus placebo, it appears that the benefit in terms of improvement in PFS and response rate is really fairly comparable. And there are some patients with regorafenib who have gotten prolonged remissions or stable disease with it. We see patients who have lung cavitation. That shows that it’s really working as well. So I think that there is a real benefit to both drugs.

And for me, personally, I think that the patient who has gotten a lot of fluoropyrimidines and — up to now, sometimes, it’s good to try a little bit different strategies. And I would tend to use regorafenib, especially if I’m not worried about their compliance and their ability to — if they’re going to have some hand-foot syndrome that they can take care of themselves. If they seem like that will be a problem for them, then the TAS is pretty user friendly, where the main thing is that if they don’t get treated on week 4, they’ll get treated on week 5.

CALGB/SWOG-80405: Relationship among somatic DNA mutations, MSI status, mutational load and survival in mCRC

DR LOVE: So let’s talk a little bit about immunotherapy of colon cancer and particularly the issue of MSI-high tumors. And to get into that, maybe you could talk a little bit — I was curious about this data set that was just presented at ASCO that you were part of, the CALGB and SWOG-80405, the Alliance study. But here you all were looking at somatic DNA mutations, MSI status, mutational load. Do you want to talk about what you looked at there, what you found and what you think it means?

DR HOCHSTER: The 80405 study had a lot of data on the right versus the left side for the RAS wild type. So, I mean, I think we found in that patient population that the MSI-high patients are pretty much what we’ve come to expect. There are few of them, but they respond to chemotherapy pretty much as well, if not a little better, than other patients.

But the other studies have tended to show that these are patients where immunotherapy may have a real large effect. So there have been studies for refractory patients with both pembrolizumab and nivolumab that have shown a very high response rate and a very prolonged disease-free survival.

Potential rationale for the efficacy of atezolizumab/bevacizumab in MSI-high mCRC

DR HOCHSTER: One of the other studies that I presented was with bevacizumab and atezolizumab, so the anti-PD-L1 from Genentech, together with bevacizumab. And we had 10 patients. It was kind of 1 arm of a basket study that involved a number of GI tumors. But for that patient population, it seems to have worked very well with 90% response or stable disease.

DR LOVE: Yes. I was going to ask you about that study with the PD-1, atezolizumab and bevacizumab. What’s the thinking? I’ve heard people talk about anti-angiogenics like bev in some way having an immune-based effect or that some of the benefit from an anti-immune effect or from an immune effect. What’s the thinking behind all this?

DR HOCHSTER: There are no real randomized trials at the moment. I think that we could see one maybe in rectal cancer or melanoma. But, I mean, one theory is that because of the regulation of the kind of neovasculature that you see tumors by blocking VEGF, you might be able to get the immune cells in better. They might be able to penetrate the tumors better because the tumor vessels are more normal.

Another theory is it may upregulate some of the expression of the PD-1 antigen or other immune antigens that allow for better immunotherapy. But again, I think that there’s a lot of theory and some preclinical data but very little clinical data to support that. It’s just that in our study, it wasn’t randomized or controlled, but we did seem to see a lot of benefit.

Incidence of MSI-high tumors in mCRC

DR LOVE: Can you summarize what we know right now about, first of all, the frequency of MSI high, how much of that is somatic, how much of it’s Lynch and what we know about checkpoint inhibitors from these patients?

DR HOCHSTER: Right. So the MSI-high patients are about 4% or 5% of metastatic colon cancer. I mean, they start out more in Stage II and Stage III, but a lot of those are cured in the earlier stages, especially the Lynch patients, because with the DNA repair defect, apparently, the tumors are not as likely to develop metastases. But the ones who have metastatic disease, they have many DNA mutations. And they tend to be responsive to these anti-PD-1 drugs.

The most common way that people get this condition is by inheriting gene defect for mismatch repair enzymes. The 4 that we usually stain for, MSH and MLH, if one of those is inherited — a defective one is inherited from your parents, that’s the basis of a Lynch syndrome. And it’s probably about two thirds Lynch and one third spontaneous in most of the studies that are coming out.

Again, this is where kind of a situation where we have some clinical data but not a lot of large prospective trials, so we really have the numbers nailed down very carefully. But I would say it’s about two thirds Lynch and one third sporadic. And if you develop a sporadic mutation in these mismatch repair enzymes, you’re going to have the similar defect and a similar biology. And it seems to be equally responsive to anti-PD-1. Many of those aren’t due to point mutations. Some of those are due to gene silencing. But it doesn’t seem to matter for the PD-1 response.

Ongoing Phase III studies of anti-PD-1/PD-L1 checkpoint inhibitors as first-line therapy for mCRC

DR LOVE: Can you talk about what we know right now about specifically nivolumab and pembrolizumab and how the FDA indications are crafted and how that gets translated into nonprotocol practice?

DR HOCHSTER: Right. Both drugs are now approved for previously treated MSI-high or mismatch repair-deficient colorectal cancer. So I think that both can be used in practice. My particular question is how to use this. Do we need to use it in combination? Do we need to use it sequentially? And so I think those are the questions at the moment.

I mean, we know that people who’ve had a lot of chemo and are pretty sick can be literally resurrected with these drugs and can be treated for a long time, for really years on these therapies with minimal side effects. So it’s amazing how well patients can do with this.

But there’s a first-line study going on from pembrolizumab versus chemo, so I think that’s an important first-line study to determine. I have a patient on that study who responded really well to chemotherapy, as you might expect, because with platinum-based chemo and no ability to repair the DNA, they can do really well also. And he hasn’t made it to crossover yet, and it’s been a year and a half. So here’s a patient that’s done very well with chemo.

So I don’t know if the right thing to do is jump in with the anti-PD-1. We’re going to be specifically answering that in a new NCTN study that’s a collaboration between NRG and SWOG called the COMMIT study, which will be a first-line MSI-high study.

And in that study, we’ll be using atezolizumab, but it will be chemo alone, atezolizumab alone or the combination. And I think that’ll give us some real answers to how to use the drug in first line also. So I encourage people to check the MSI status before you start chemotherapy for first-line metastatic disease and think about enrolling people to these trials.

DR LOVE: That’s an interesting point, though, that you made about patients doing better with platinum-based chemotherapy than non-MSI patients. Do we have any quantification of how much better?

DR HOCHSTER: I don’t think this has been really studied systematically, because they’re pretty rare and they were kind of folded into all the other studies. So you got a few here and a few there. But these studies should really kind of give us a better handle on that.

Case: A 40-year-old man with metastatic rectal cancer and Lynch syndrome receives atezolizumab/bevacizumab on a clinical trial

DR LOVE: So I wanted to go through another one of your cases. And hold on a second. Yes. We were just talking about checkpoint inhibitors. How about this 40-year-old man? Can you talk a little bit about this patient?

DR HOCHSTER: Yes, the one who is getting the atezolizumab/bev combination.

DR LOVE: Right.

DR HOCHSTER: So he is a 40-year-old. And he presented with rectal bleeding in December 2012. At that time, he was found to have a pathologic T4N2 rectal cancer. And he had 12 positive lymph nodes, so really high-risk patient. I wasn’t seeing him at that time, but he was treated with adjuvant FOLFOX appropriately, and he completed the 6 months.

And they tested him, and he was found to have Lynch syndrome even though his family history was not that suggestive. So about a year later, he recurred in the abdomen with peritoneal disease. And he underwent debulking surgery with splenectomy, bowel resection, omentectomy and resection of peritoneal metastases. And then he got a course of HIPEC at that same time.

And he was again in remission for about another year, but he developed a new peritoneal metastasis and mesentery nodal mass, which was then growing. It was found in April 2015 and growing in July 2015, and at that time he was referred for our trial, where he started on bevacizumab with atezolizumab. So that’s more than 2 years ago. It’s like 2 and a third years already.

So he had really a nice response. Peritoneal disease all went away. His mesenteric mass over the first year was stable but then grew a little bit. So at that point, he was allowed to continue on study, so we gave him some radiation with SBRT to that nodal mass and continued the immunotherapy. And he’s continued to do well. He hasn’t had any further progression since that time.

He has really no side effects. The bevacizumab was stopped about 6 months ago because of proteinuria. And we had been giving it to him intermittently for the 6 months before that. And so that’s been discontinued, but he’s still on atezolizumab and doing well.

DR LOVE: And I noticed that about the proteinuria and the bev. How much proteinuria did he have and, from your point of view, how much proteinuria do you need to have in order to stop bev in a patient who’s doing well?

DR HOCHSTER: I think the proteinuria, to a large extent, is a cumulative side effect if people stay on it for a long time and/or related to not controlling the blood pressure too well. So those are the things that will predispose to proteinuria. Usually if it’s getting up to a gram a day on a urine collection or a UPC, urine protein creatinine ratio greater than 1, we’ll hold it. And if you hold it for a few months, they can get some more bev. But I think they tend to redevelop the proteinuria after that. It just causes kind of protein lead and, at some point, it just continues on at a certain level.

When I follow those patients after you stop the bev, the proteinuria does resolve. It’s just like it’s very hard to keep starting and stopping it. And you can’t treat them for that long before it comes back.

DR LOVE: So you say this man’s doing well on atezolizumab. No autoimmune problems whatsoever?

DR HOCHSTER: No, he hasn’t had any autoimmune side effects. He hasn’t had really any other side effects. He comes in for his infusion and keeps doing his normal stuff.

I would say he has a little bit of fatigue, he reports. But it’s pretty benign compared to being on other treatments for a couple of years.

Anti-PD-1/PD-L1 checkpoint inhibitor-associated immune-related adverse events

DR LOVE: What have you observed in your own clinical practice in terms of autoimmune problems with checkpoint inhibitors? And there was some speculation when the anti-PD-L1 agents came out, like atezolizumab, that maybe it might have a different or more favorable toxicity ratio. How did that play out?

DR HOCHSTER: I think that it’s a little hard to tell in GI cancers, because we haven’t used it that much. We don’t have that many patients where it’s approved. I mean, what we’ve seen — and I usually see this whenever I’m ward attending for my colleagues who treat lung cancer and melanoma. But, clearly, the patients who have combined immunotherapy with an anti-PD-1 or an anti-PD-L1 with an anti-CTLA-4, those patients tend to be at higher risk.

But I must say that we’ve had some, even, single-agent toxicity. We’ve seen some cardiomyopathy. We’ve seen interstitial lung disease occasionally. So all these things can happen once you set the immune system free. Sometimes it gets a little bit exuberant and begins to attack the normal tissue, too.

DR LOVE: It’s interesting you mentioned cardiomyopathy, because somebody just presented a case to me — I don’t think it was colon cancer. I can’t remember. As you stated, everybody’s trying these things — of what, I guess, sounded like an autoimmune myocarditis. You had a patient with that, or patients?

DR HOCHSTER: We had one who had had colon cancer. And then there have been a few more around our institution with lung cancer, melanoma, other diseases. But yes, you can — I mean, what we’ve seen is they tend to get admitted with respiratory difficulty. And then the differential is an interstitial lung disease or cardiomyopathy. And then we’ll end up seeing that their ejection fraction is kind of low, usually less than 30. With steroids, they usually recover.

DR LOVE: Interesting.

DR HOCHSTER: But it can definitely cause some kind of autoimmune myocarditis as well.

Perspective on hyperthermic intraperitoneal chemotherapy

DR LOVE: Hmm. Looking back at this man’s history, this man with Lynch syndrome, first of all, I don't know if you were taking care of him when he had his first recurrence. That was in 2013. He had this big debulking surgery and HIPEC. Were you taking care of him then?

DR HOCHSTER: No, no, I wasn’t.

DR LOVE: What do you think about that? It really caught my attention. He got the HIPEC and debulking surgery. Nowadays, from your point of view, is that kind of a legitimate or something you would ever consider doing?

DR HOCHSTER: I’m personally not a big fan of HIPEC or any regional therapy. I mean, I’ve done a lot of intraperitoneal chemotherapy since I was a fellow and many trials with intraperitoneal therapy in ovarian and gastric. HIPEC is just giving warm intraperitoneal therapy once at the time of surgery.

I think there’s a definite benefit from the debulking in terms of PFS and, possibly, survival, but there are no really good studies that have isolated the benefit of doing the HIPEC procedure. But it’s becoming kind of a popular approach now that there are more of these surface malignancy centers. And people are kind of looking for expertise in this area.

Ongoing Alliance A021502 Phase III study of adjuvant chemotherapy alone or with atezolizumab for Stage III colon cancer with deficient DNA mismatch repair

DR LOVE: So also I’m curious, looking back with the retrospectoscope at this case, when you look at what happened with this patient, so initially having 12 positive nodes, he gets adjuvant FOLFOX. Looks like he had a pretty quick and extensive recurrence. If you were to see him or a patient like this again in that situation, would you have incorporated a checkpoint inhibitor, and which one? And would you have done any local therapy?

DR HOCHSTER: Again, there’s no adjuvant data for use of anti-PD-1.

DR LOVE: I’m talking about when he had his first recurrence.

DR HOCHSTER: Yes. I think at that point that would fall within the indications that should be considered. But I do want to again call to people’s attention that the NCTN cooperative groups will be starting an adjuvant therapy trial very shortly. The Alliance is going to be running that study. And I think that’ll be very interesting, to see if a checkpoint inhibitor is really effective in the adjuvant setting for colon cancer.

DR LOVE: But this is MSI high?

DR HOCHSTER: All MSI-high patients.

DR LOVE: Wow! That’s going to be interesting to try to accrue, huh?

DR HOCHSTER: Yes, that will be interesting. But I think right now for new diagnoses, many, many institutions are routinely checking the MSI status. So I think we have a chance of getting that. The study’s going to be chaired by Frank Sinicrope from Mayo. And we need a lot of patients for that study, but I think it’s a worthwhile study.

Case: A 48-year-old woman with BRAF V600E-mutated mCRC initially receives FOLFIRINOX followed by cetuximab/encorafenib on a clinical trial

DR LOVE: So let’s talk a little bit about BRAF disease. And I know you’ve had an interest in that in terms of clinical research. And maybe to get into it you can begin by talking about your 48-year-old lady.

DR HOCHSTER: Yes. So this lady is actually a young woman who has 3 children in middle school and high school. And last summer she kind of developed bloating and various nonspecific symptoms, and eventually in August she underwent a colonoscopy and was found to have a sigmoid lesion that was obstructing and underwent a resection.

At that time, the CAT scan on her initial staging showed extremely extensive disease with extensive, extensive involvement of the liver, like, a packed liver, lung disease, peritoneal metastases. And she underwent a palliative resection because of her obstructive symptoms. And at that point, I saw her. And we did her sequencing and, not too unexpectedly, she was found to have this BRAF V600E mutation, the kind of standard BRAF mutation.

So I started her on FOLFIRINOX.

DR LOVE: Could I just ask you about the way she was diagnosed here? Now, she was diagnosed on a circulating DNA assay, or what?

DR HOCHSTER: No, no, no. She was diagnosed on colonoscopy. And then she had the palliative resection. And then we sequenced her tumor. But when I started her treatment, I wanted to check her circulating tumor DNA more as a DNA marker, because we know that the panels can detect the circulating BRAF-mutated DNA. And at that point when she had this very extensive amount of disease, her circulating tumor DNA assay showed that she had 68% BRAF V600E present.

So we treated her with the FOLFIRINOX, and she did really very well. Her liver lesions actually became more prominent, because they were kind of necrotic. That had changed the density. And our radiologist said, “We do see this once in a while, where even though they have a great antitumor response, she seems to have more lesions even.”

But her CEA had diminished from about 90 to 9 at the same time, and she was feeling much better. I mean, started to gain weight again. We knew she was responding. And when she got to the end of 12 cycles of the FOLFIRINOX, her circulating tumor DNA percent was down to 8% for the BRAF fraction.

DR LOVE: Yes. I don't know whether everybody knows this and I don’t, but I’ve got to say, when I saw this I was trying to remember from melanoma or even lung cancer, measurement of circulating DNA of BRAF. Is that commonly done?

DR HOCHSTER: The panels now that are being done by the various companies are looking for specific mutations that are common. So they’re all going to include that one amongst the other ones. So normally you’ll get a report back that has whatever DNA is in the tumor that’s being produced and circulating. So you’ll see KRAS DNA. If they have other mutations, it’ll all come out in this panel, where they give you the relative percentages.

DR LOVE: So, for example, Guardant, is that this type of assay?

DR HOCHSTER: Yes. Guardant’s one of them. And Foundation also has an assay like that.

DR LOVE: And they report the percent of what, tumor cells with this in it?

DR HOCHSTER: It’s just if they isolate the tumor DNA that’s circulating. It’s in the bloodstream. It’s not in any cells. It’s just, like, from the cells giving off the DNA. It’s not all digested. And some of it’s circulating around. And you can look for these hot-spot mutations that are point mutations that give the specific driver mutations that we’re looking for.

So when we know that people have, like, a KRAS codon 12 mutation, or 13 or whatever it is and the BRAF V600E, they have probes to look for those specifically, or they do it in the panel. And they report those specific DNA mutations. It’s like, I think, about a 50- or 60-gene panel.

DR LOVE: But what I’m trying to figure out is, when they put percent, like, it’s percent of what?

DR HOCHSTER: The amount of circulating tumor DNA.

DR LOVE: Huh.

DR HOCHSTER: It’s not quantitated how much DNA you have there, but it’s the relative frequency of that compared to the other DNA that’s found there.

DR LOVE: And does that percent, in general, correlate with response?

DR HOCHSTER: That’s a question. But there is some data for people who have Stage II where they have a resection. If the DNA doesn’t clear completely, they have a much higher risk of recurrence. And so we’re looking, again in the cooperative groups — and other people are looking at it also — if we can use that as a predictive marker.

In this case, I was trying to look at it as a response marker. And in her case it seems to, because she kind of nadired out at 8%. But then it began to go up again when I gave her maintenance with just 5-FU/bev, because she really couldn’t continue to tolerate the FOLFIRINOX, obviously.

And then we put her on a trial with a BRAF inhibitor. She was randomized to the arm of cetuximab with the BRAF inhibitor encorafenib. And she responded transiently for that, but after 4 months she was progressing again. And we saw that her DNA went up to 38%. So, I mean, she’s had 2 responses to the therapy, but they haven’t been very persistent.

High mutational burden in patients with BRAF mutations

DR LOVE: And I see she’s in nivolumab right now. How’s she doing?

DR HOCHSTER: That’s right. She just started. And she seems to be tolerating it okay. I mean, she literally started it 2 weeks ago. So I don't know how she’s going to respond.

DR LOVE: How does her getting nivolumab tie into FDA indication, clinical research? Where did the nivolumab come from?

DR HOCHSTER: We know that the patients who have BRAF mutations tend to have higher tumor mutation burden. It’s not exactly in an FDA indication. And it was not approved by her insurance. But we managed to get her drug replacement off label. So it would be considered an off-label use in the BRAF, but we know the BRAF patients have more mutations and a higher mutational burden than conventional colon cancer. So I think the studies in the future, we’re going to try to focus on the use of these drugs in BRAF-mutated colon cancer also.

DR LOVE: My understanding is, in general, colorectal cancer, metastatic colorectal cancer that’s not MSI high doesn’t respond hardly at all to checkpoint inhibitors. Is there any reason to think that BRAF does, other than theoretical? Any data?

DR HOCHSTER: There are no clinical data yet.

DR LOVE: It kind of reminds me of KRAS in lung cancer. I don't know why, but I guess because people think that maybe checkpoint inhibitors respond more. Really interesting.

SWOG-S1406: Results of a Phase II trial of irinotecan/cetuximab with or without vemurafenib for BRAF-mutated mCRC

DR LOVE: Where are we today, right now, with targeted therapy of BRAF-mutant colorectal cancer?

DR HOCHSTER: So we reported at ASCO GI this year and at the ASCO meeting, my colleague Scott Kopetz at MD Anderson presented the data from the SWOG trial, SWOG-1406. So in that trial, we randomized 100 patients to receiving a kind of standard regimen of irinotecan/cetuximab versus the same 2 drugs with the specific BRAF inhibitor vemurafenib. And people know that drug because it is an approved drug for melanoma. But it is considered off label for colon cancer.

In our study, the group that got the vemurafenib, the primary endpoint was PFS. So the PFS was really twice as long. The hazard ratio for the comparison was 0.48. So they went twice as long to progress. And the response rate was around 25% as opposed to 4% in the control arm. So we think that that’s going to be very beneficial for patients. That’s, again, second- or third-line patients who hadn’t received cetuximab. That was the basic eligibility for that study.

And I believe that that will make it into guidelines soon, so that we hope that patients will be able to access the vemurafenib for the treatment of BRAF-mutated colon cancer after they’ve received some first-line chemotherapy. But I do want to add that you need to give an anti-EGF drug with it. If you just give the BRAF inhibitor alone, it’s not very active. It’s not like melanoma. So you’ve got to give it with either cetuximab, as we did in our study, or with panitumumab, as has been done in some other studies.

DR LOVE: So I take it from what you’re saying that you would like to be able to use this combination of vemurafenib/irinotecan/cetuximab outside a trial setting right now for your patients.

DR HOCHSTER: Yes. Yes.

Preliminary results of the Phase III BEACON study evaluating cetuximab, encorafenib and binimetinib combination therapy for BRAF V600E-mutated mCRC

DR LOVE: And what about MEK inhibitors, of course, in not just melanoma but even lung cancer? We’re seeing those combinations be used. Vemurafenib’s been paired with cobimetinib in melanoma. And then you have dabrafenib and trametinib. What about that strategy or combined with an anti-EGFR antibody?

DR HOCHSTER: Right. So in this study that is currently ongoing, the so-called BEACON study, that is including a MEK inhibitor in one of the arms also. One of the arms is cetuximab and the BRAF inhibitor encorafenib. And the other arm is cetuximab/encorafenib and binimetinib, which is the MEK inhibitor. And then the third arm is the control with just chemo and cetuximab.

So that, I think, will be a direct comparison of how much benefit a MEK inhibitor would provide for BRAF. And hopefully it’ll be like melanoma and they’ll do better. The data from — there was a 30-patient safety lead-in to that randomized Phase III study that was just presented at ESMO.

And in that combination of cetuximab/encorafenib and binimetinib — so an anti-EGFR drug, a BRAF drug and a MEK drug — the response rate was over 40%. And the PFS wasn’t reached. The median time for patients to be on study was almost 6 months, over 5 and a half months, but three quarters of the patients were still on treatment. So that actually may be a much more effective regimen than just the triplet we had with chemo, but we’ll need to see how that turns out.

DR LOVE: I’m kind of trying to get a global feel for BRAF-mutant disease in colon cancer and targeted therapy. I’ve had the impression that it’s just completely a different ballgame than melanoma and even lung cancer. I mean, lung cancer right now, BRAF-positive disease investigators are treating with targeted therapy in first-line metastatic disease. But colon never seemed to kind of have that, I don't know, sensitivity. You’re bringing in the antibody, et cetera. Globally, do you really think that targeted therapy is going to be helpful once you kind of get it right?

DR HOCHSTER: Yes, I do. I think that it will be helpful. I mean, the median survival for BRAF-mutated metastatic disease is around 50% of nonmutated. So that’s a very poor-prognosis group. I think the reason that we haven’t made as much progress yet is twofold: first that the BRAF inhibitor by itself did not work, so we needed to discover that you needed to pair it with an anti-EGFR. And the second thing is, it’s still pretty rare, 4% or 5% of all colon cancer, so the studies take a lot longer to do.

And those are the things that have slowed down the development. But I think aggressive chemotherapy with these combinations and possibly adding some immunotherapy eventually should really help these patients do substantially better than where we started at.

COTEZO IMblaze 370: A Phase III study of atezolizumab with or without cobimetinib versus regorafenib for previously treated MSS mCRC

DR LOVE: Just to close out here, any other research strategies or agents in colon cancer that you’re excited about that we haven’t talked about today?

DR HOCHSTER: I think the other discovery or observation that’s going to be very interesting and that we’ll see clinical data on very soon is the use of the MEK inhibitor with immunotherapy for garden-variety or microsatellite-stable colon cancer. So as we heard from Johanna Bendell’s study, if you added cobimetinib to atezolizumab, 25% to 30% of patients seemed to respond to immunotherapy. And without the cobimetinib, it was 1% or 2%.

So it seems as though the MEK inhibitor can really have a major effect on upregulating the immunogenicity of colon cancer in activating the T cells so that they are more responsive to the atezolizumab. So the 3-arm study that was done internationally by Roche has been accrued. It accrued very rapidly. And we’re waiting to see how that pans out.

In our institution, I think we’ve managed to get 8 patients on before the study closed, which was really — it opened and closed pretty quickly. But we do have a couple of patients still on. And they were people who were on the combination arm. So I think that this might be a real lead.

DR LOVE: What is the actual randomization?

DR HOCHSTER: The control arm was regorafenib. One arm got the cobimetinib and atezolizumab. And then there was atezolizumab-alone arm.

DR LOVE: Hmm.

DR HOCHSTER: So it was 3 arms.

DR LOVE: And how many people did they accrue?

DR HOCHSTER: Over 500. It was a large Phase III trial.

DR LOVE: Hmm. Interesting. I wonder — I mean, it’s interesting that they did a randomized study. I was thinking to myself, if they did a big Phase II and they showed a 25% response rate and the responses lasted a couple of years, would that be enough? Or do you need to really do this kind of randomization?

DR HOCHSTER: Right. It’s always better to have more definitive controlled data. But it’s hard to predict, with the regulatory environment today, what you really need to have. A Phase III study that shows a real benefit over a standard approved regimen in third or fourth line will be very helpful. And, I think, especially with the cost of these drugs, it’s going to be important to show how beneficial it is. So we will see.

DR LOVE: When do you think we will see?

DR HOCHSTER: I think some of the data should be coming out by next year sometime, probably next year for ESMO. I don’t think it’ll make it for the ASCO meeting. But I imagine by the end of the year we’ll have some data.