 I’m looking at our questions, and 1 of them is right on point of the next thought I had. If they’re 

+ [human epidermal growth factor receptor 2], should you give EGFR therapy? There are data that that’s not a good idea.

 Yes, and that’s 1 of the reasons why I put these 4 markers on the slides in the initial work-up. It’s 

 as a negative predictive marker; not very frequent. We talked about this 2% that’s more in the left-sided, 

 wild-type tumors that have not responded to EGF receptor antibody therapy. But the data are stronger and stronger every day that there is a negative predictive marker. And it’s a target for our treatment.

In the metastasectomy setting, I’m leaning on it, particularly with negative adjuvant data. But I’m doing it in patients, and following them it’s a useful marker. And I do feel better, although to me it’s important to have a patient in front of you who understands what you’re talk about: that a positive is real. It means you’re in trouble almost for sure. A negative is good but is not a guarantee. We still are watching, but we feel more comfortable not treating that patient. How are you handling it?

 I completely agree. You know the actual real data if you have a positive result. We talk about the highest specificities, a tumor-informed test in which you profile the tumor and develop a PCR [pathologic complete response]test using the genetic fingerprint of this cancer. The predictor, the likelihood for tumor recurrence if you have a positive ctDNA, is 99.8% even if you don’t see anything on scans.

I’ve been adopting this for about a year, and there are certain patients for whom it influences treatment decision, work-up, or finding extra recurrence way before I would have done it. I would have seen it before because I do a routine scan and routine colonoscopies.

The problem I have is that there are a lot of trials ongoing or in planning. Let’s say, they’re adjuvant trials. These trials will take many years until we eventually see that they are prospectively valid and relevant. But you’re already doing it. Clinical practice and what’s happening now could actually leapfrog the evidence that we generate in prospective trials. It’s going to be an issue, but it’s a valuable tool and has influenced my treatment decisions.

 Me too. We’re going to learn a lot over the next couple of years and clean up our act in many ways with more precise decision-making.

If you got to cycle 1, day 21 of regorafenib, and you’ve worked your way up to 160 mg, and you take your week off. Then you start back at 160 mg. And I will say that I do—that’s what the study would suggest we do. But I’ve also been unhappy with that after it accumulates a little more. I find myself not infrequently going back to 120 mg during cycle 2. You don’t let your guard down into cycle 2 if you’ve escalated up to 160 mg. Watch them through cycle 2. Also liver functions—I’ve seen those go up in size. It’s a judgment call based on how that patient looks at the third week. What do you think?

 I completely agree. Most of my patients who are on long-term treatment are on 120 mg. There are a few who tolerate 160 mg, and there are some for whom you have to go down to 80 mg. We don’t have any data below 80 mg. It may be colon cancer. There’s some data that GIST [gastrointestinal stromal tumors] went below 80 mg. But I consider 120 mg the optimum long-term dose for my patients. The 160-mg dose can be achieved, but it’s not very frequent.

 Should we drop bolus 5-FU [5-fluorouracil], FOLFOX [5-fluorouracil, leucovorin, oxaliplatin], FOLFIRI [5-fluorouracil, leucovorin, irinotecan]? You and I might differ on this. I tend to drop it quickly. I don’t always drop it. I drop it always in FOLFOXIRI [5-fluorouracil, leucovorin, oxaliplatin, irinotecan]. Even if it’s a 45-year-old in front of me, I still drop it. But in FOLFOX [5-fluorouracil, leucovorin, oxaliplatin] and FOLFIRI [5-fluorouracil, leucovorin, irinotecan], I still hold on to my leucovorin.

 I’m still half-convinced, even though I taught a fellow this morning that, in theory, if I’m not giving bolus 5-FU [5-fluorouracil], I don’t need the leucovorin.

 That’s interesting because there are 2 studies that are old. One looked at the Roswell Park regimen and how much leucovorin we need. Because for Roswell Park, weekly bolus had 500 mg/m2 squared, and there is no difference between 20 mg and 5 mg. That’s the dose.

Then there was an EORTC [European Platform of Cancer Research] study, a randomized phase 3 study with 3 arms. One was Mayo Clinic regimen; 1 was infusion of 5-FU [5-fluorouracil] with or without leucovorin. And the progression-free survival, the primary end point, was better with leucovorin plus infusion 5-FU [5-fluorouracil] without bolus 5-FU [5-fluorouracil]. Those are phase 3 data.

I keep leucovorin around, but I don’t use bolus strategy. My goals of regimen is modified FOLFOX7 [5-fluorouracil, leucovorin, oxaliplatin]. And even with FOLFIRI [5-fluorouracil, leucovorin, irinotecan]—which I just put a patient on today—you need the bolus. It’s not confirmed by the NCCN [National Comprehensive Cancer Network] Guidelines.

Do you think FOLFOXIRI-BEV [5-fluorouracil, leucovorin, oxaliplatin, irinotecan, bevacizumab] is going to become a gold standard? As you point out, you’re still going to escalation strategy if you use FOLFOXIRI [5-fluorouracil, leucovorin, oxaliplatin, irinotecan].

 As much as I don’t like to think FOLFOX-BEV [5-fluorouracil, leucovorin, oxaliplatin, bevacizumab] as the 1 standard, I think FOLFOXIRI-BEV [5-fluorouracil, leucovorin, oxaliplatin, irinotecan, bevacizumab] should not be the gold standard. We have a lot of standards out there. We need to know when and how to use the tools we have.

And a triplet is indicated in some patients. I like FOLFOXIRI-BEV [5-fluorouracil, leucovorin, oxaliplatin, irinotecan, bevacizumab]. They’re good, and there are a lot more data now than there were 5 years ago. But it’s not the 1 gold standard. If I had to design a clinical trial, I would probably not use FOLFOXIRI [5-fluorouracil, leucovorin, oxaliplatin, irinotecan] as a first-line comparator. But would give it.

 On the other side is that there are still patients for who I would start with fluoropyrimidine and bevacizumab by itself.

 Particularly little old ladies. They do incredibly well on that regimen. Get your same 8 to 9 months of initial progression-free survival with something simple. The spectrum is in play.

® has been helpful, and we hope that what we discussed will help you in managing your patients with metastatic colorectal cancer. Thanks for taking the time. I hope everybody enjoyed the evening. Thanks to the 

Videos

Experts in CRC discuss questions and management of mCRC.

Open Questions and Future Directions in mCRC 

 It’s funny. When I send off molecular profiling, and I receive the result back, I click on the website, and there’s almost an excitement when I get 1 of these positive markers. Now, a 

of course is bad news, but at least I have more to do about that for a V600E patient.

When I receive MSI [microsatellite instability], I almost close my eyes and look again and think, “Do we have the right result?” I never believe the IHC [immunohistochemistry]. I always repeat the genetics just to make sure. And tumor mutational burden [TMB] has become a last-line salvage. If I did a test 2 years ago and the TMB was 12, I’m in, because now got approval. 

 [human epidermal growth factor receptor 2], for me at least, has been rare. I look in almost everybody, and it’s rare that I have a 

+ colon. And I see a ton of patients with colon cancer.

 Whenever I get 1 of these, I get a little excited. Then I think, “Do I go ahead and use it now? Do I stop what I’m doing? Let me pick on some of the easy ones, like MSI high. Is it first line for everybody? What if I’ve got a patient with a big tumor burden who is symptomatic?

 It’s hard to make a case against I/O [immuno-oncology] therapy in first line. The data are convincing, and we should embrace that. We have randomized data now. It’s a good reason to do this and not wait too much longer. But there is 1 caveat. The KEYNOTE study has this higher rate of early progressors: 29.4% compared 12.6%. And the switch, the crossover of the PFS [progression-free survival] data—I would love to know who those patients are. The publication is pretty meager on exactly that.

 I would like to see TMB analysis. What kind of mismatch repair proteins were deactivated? There’s tons of analysis that we can do and hopefully will be done. We can potentially even look at patients who need combination with chemotherapy in lung cancer. If you have lung cancer expressing less PD-L1, non–small cell lung cancer, you combine with chemotherapy because you want a response. Because the benefit is if you have a response. In the KEYNOTE study, 83% of patients who have a response will continue to have a response after 2 years.

That’s the excitement I have. We can cure patients with stage IV disease with immunotherapy when they’re MSI high. Those are loaded words, you know?

 I’ve got some who obviously haven’t been. But there will be some.

I hope that’s right. I keep coming back to the point—maybe it’s the same 1 you’re making—that never before have we had a biomarker in MSI where it’s not very clearly written how that has to be tested. You can use immunohistochemistry, with 1 of 4 proteins missing, even though we know they’re not all the same. You can have DNA fragment analysis. You can do next-generation sequencing with computer analysis. At least from an FDA perspective, or equation, we’ll learn the marker may not be the same in all those people.

? To me it’s still a second-line therapy. These are usually bad tumors. If I thought ahead, use FOLFOXIRI [5-fluorouracil, leucovorin, oxaliplatin, irinotecan] in the front line. If I get a maintenance window, I feel lucky. Then it’s a 

-targeting therapy second line. Is that where you are too?

 Agreed. We had data from a little bit more than 40 patients in the first line, and the ANCHOR-CRC study, which looked at the BEACON CRC triplet [encorafenib, binimetinib, cetuximab]. And there was a high response rate, 50%. Almost every patient had some form of tumor shrinkage. But the progression-free survival was 12.9 months. That’s not great, and that’s something I see even in a second- or third-line setting.

Once tumors have figured out how to counteract the 

 receptor antibody inhibition of the MAP kinase pathway, they progress quite rapidly. I see this rapid emergence of resistance. That is something we need to work on: MET amplification, the EGF receptor amplification.

There are a lot of mechanisms that we pick up right now. We need to figure out what’s going to individualize even the treatment against the resistant clones over time.

 Right now, you’ve got tumor mutational burden of 13, OK? Let’s say you’ve had first- and second-line therapy, and I’m adding pembrolizumab into your choices. Now you have 4 drugs, right? Are you going to fight an I/O card with the TMB data, or are you going to use anEGFR inhibitor? What are you going to do?

 I’m a little skeptical about the cutoff of 10 for all tumors.

 I completely agree. I believe there’s a difference, which I see with a TMB of 10 or 20—with 20 I would have no problem using pembrolizumab earlier. Probably not in the first line. Potentially in the second-line setting, depending on patients’ preference and how they are. But in the third-line therapy study, those patients do benefit. They have a high chance of benefit.

John Marshall, MD, and Axel Grothey, MD, discuss how molecular testing results guide treatment decisions in patients with mCRC. 

Using Molecular Testing to Guide Treatment Selection in mCRC

: Let’s spend a few minutes on an equally important drug, TAS-102 [trifluridine, tipiracil]. I can remember we used to call it CPT-11 before we got over that. But TAS-102 is what we call it now, trifluridine tipiracil, Lonsurf, the trade name. This is a different drug, right? It’s fluoropyrimidine of sorts, myelosuppression, the main adverse effect. But also very similar, if not identical, curves in terms of PFS, [progression-free survival] in terms of OS [overall survival]. Again, similar populations, similar trial design. And I think of the 2 drugs as not interchangeable. You decide you’re going to use one versus the other based on the biology that’s already been declared for the patient, what adverse effects have they endured.

But I also see there are more and more data with TAS [trifluridine, tipiracil] being used in combination with BEV [bevacizumab], and in earlier lines of therapy and chemotherapy. I see it as one of those interchangeable chemotherapy chess pieces that might end up replacing capecitabine after you’ve progressed on CAPE [capecitabine]. But for the same things we talked about with REGO [regorafenib], we also have this other compound that needs managing, the toxicity, and different sets of toxicities, but one that is equally important to try not to leave on the table when you’re playing chess against chronic disease. What do you think?

: Yes, I completely agree. In the end patients benefit from having all active agents as part of their treatment. And TAS [trifluridine, tipiracil] is an active agent. We have phase 3 data, 2 sets of phase 3 data. A kind of Western world plus Japan study, the RECOURSE study, and the TERRA study, similar to CONCUR with regorafenib in the Asian patient population. This drug is active, no question.

You mentioned fluoropyrimidine. It’s even active in patients who have been pretreated with 5-FU [fluorouracil]. It has a little bit of a different mechanism of action. It doesn’t cause hand-foot syndrome, different than capecitabine; it’s an oral agent.

It has an interesting and strange dosing schedule if you think about it, twice a day for the first 5 days, and then from day 8 to 12. It’s repeated every 4 weeks, but a lot of patients won’t be able to continue after 4 weeks because they have severe neutropenia, in particular with all these pretreatments we have. It’s interesting. It is more a standard chemotherapy that we have. It’s not the fancy biologic agent, or an I/O [immunotherapy] agent. But as you mentioned, I think it deserves attention because it combines well with bevacizumab, like very fluoropyrimidine combines well with bevacizumab.

We have randomized data in a later-line setting from a Danish group that showed BEV [bevacizumab] plus TAS [trifluridine, tipiracil], it leads TAS [trifluridine, tipiracil] alone in progression-free and overall survival. That’s what I try to use, and normally can use, because we have published randomized data. I combine TAS [trifluridine, tipiracil] with bevacizumab unless there is a contraindication.

Even the first-line data, as you mentioned, TAS [trifluridine, tipiracil] and BEV [bevacizumab] versus CAPE and BEV [capecitabine, bevacizumab], with a trend in outcomes improvement numerically but not significant in progression-free survival. But again, I think it’s a drug that should be used, should be part of our portfolio of agents. And I believe that the pool of patients who are candidates for TAS, [trifluridine, tipiracil] because it has fewer subjective adverse effects than regorafenib, is probably larger. Then we would not treat patients with PS2 [performance status of 2] with regorafenib.

There are some patients when we talk about treatment options with PS2 who I would treat with TAS-102 [trifluridine, tipiracil]. This in some ways also influences my sequence of events because you can have a big discussion. Regorafenib first or TAS [trifluridine, tipiracil] first. They have similar activity. I normally in most patients use regorafenib first now that we know how to dose it. We give patients a break from cytotoxic therapy, bone marrow gets a break. And we have data for TAS [trifluridine, tipiracil] after [regorafenib]. We don’t have the data for regorafenib after TAS [trifluridine, tipiracil]. I believe even if the performance status of patients might deteriorate on regorafenib, they might still be candidates for TAS [trifluridine, tipiracil], but not vice-versa.

The likelihood is higher that patients get both agents if you do regorafenib first and then TAS [trifluridine, tipiracil]. That’s what I use. I don’t what you’re doing in your clinical practice.

: You never heard me say this and the 200 people who are online now, I’ll deny it. Just give TAS [trifluridine, tipiracil] every other week, it solves the problem. It really does. It lines up very nicely with BEV [bevacizumab] as well.

Experts in mCRC review data on TAS-102 and discuss how best to sequence therapies in the third or fourth line. 

TAS-102 in Third or Fourth Line for mCRC

 It wasn’t necessarily less toxicity. A little less toxic but not dramatically. That’s because frankly it was dosed to toxicity, it’s just how you got there, right? Did you inch your way up and find your dose? Or did you have such bad toxicity early on that you withdrew?

This doesn’t surprise me and as I’ve gotten smarter, certainly my patients’ hand-foot reaction is much lower. We’re watching it very closely. There aren’t very many people on 160 mg of regorafenib, but there are some. Most of them tend to end up around 120 mg in my experience, but some even at 80 mg in the chronic therapy.

But these are the ones when you can find this dose who will be that curve where 6 months later they’re still stable disease on a decent dose that they can live with day in and day out. What’s your reaction on this?

 I could not agree with you more. A couple of things. First of all I look at REGO [regorafenib] as a cytostatic agent. Where again, we mentioned this before, it’s not a drug that can induce a lot of response. Whenever you have a cytostatic agent, more than for a cytotoxic agent, I believe, duration of therapy matters. While we know VEGF inhibition beyond progression delaying; and regorafenib is also a VEGF inhibitor, to some degree. But as you said, we don’t even know exactly how it works. It’s probably a signature effect in each individual patient that could be different.

But cytostatic agents benefit from duration of therapy. One of the goals of ReDOS, the study you just mentioned, was to optimize longer duration of treatment. This is what we saw, why the higher likelihood of patients moving into cycle 3 eventually also translated into trend toward improved survival.

And 9.8 months, that’s 3.8 months better, or more than 3 months better than what we’ve seen in the initial CORRECT study, which also shows we are better in patient selection nowadays. We treat patients better. We keep patients on therapy better.

One other point. When you talk about hazard ratios and survival and what can regorafenib really do; even with the older data from CORRECT. The way I communicate with those patients to see what can I expect in terms of outcomes is, “OK, we’ll do the 2 cycles of therapy, 8 weeks of treatment, find a way to get you to the first scan, toxicity management, ReDOS dosing, etc. And you have an almost 50/50 chance to not have progression of the disease, because that’s where curves split.

If you don’t have progression of disease at the scan, based on the data we have, you have a 50% chance to be alive a year from now, a year later. That is I think meaningful because now a year is another birthday, is another holiday.

Axel Grothey, MD, and John Marshall, MD, discuss safety results from the ReDOS study of regorafenib in patients with mCRC.

Regorafenib in Third or Fourth Line for mCRC: Safety in the Re-DOS Trial

: Let’s discuss Tony Bekaii-Saab, MD’s, trial to try to solve this. Because if you’re rapidly declining and you start somebody at 160 mg of regorafenib, it’s not going to work. They’re going to have bad hand-foot syndrome, they’re going to be tired, they’re going to be unhappy, and they’re going to request being removed from the trial and the treatment because they’re unhappy.

: And early on. Within a week, patients have these hand-foot skin reactions; they feel fatigued. You turn a patient who you think has good performance status and is a candidate for it, into a sick patient, if you use the old dosing strategy.

: The blogs, the advocacy blogs and everything started talking about don’t take that drug sort of thing. The lesson we learn, and Tony made it real, if you will, is that you need to know how to dose the medicine. As everybody knows, it’s typically 160 mg daily, 3 weeks on, 1 week off. As Axel was saying, most people get fairly significant adverse effects even within a week or two, unlike what we’re used to seeing with other drugs. People coming in, torn up hands and feet, and poor liver function.

Tony cleverly studied what we were all playing with and did it in a more formal way of looking at an escalating dose. Instead of starting at 160 mg, start at 2 pills at 80 mg, but for an entire week. Then see the patient again, as part of this study, and if they were fine and their laboratory tests were fine, then go up to 120 mg. Then if they were again fine one week later for week 3 of cycle 1, go up to 160 mg. If they tolerated that, then that was their starting dose for the next cycle. And if they didn’t, you went to the 120 mg, or whichever dose was the right one. And it was a smaller trial. It was also looking at some hand creams to try to prevent some of the skin reactions.

But the primary answer here was that if you started low and you went up, the primary end point was how many people made it into cycle 3. What you saw was that a lot more did. Not quite double, but nearly double made stable disease beyond the first 2 cycles. That’s basically what your trial, Axel, showed, and a lot of what I tell patients is that about half of everybody who takes regorafenib has stable disease or better at the first scan, which is at 8 weeks or so. That was borne out because about half of these people made it into cycle 3, just under half.

By dosing it this way you were able to avoid that early stop, the patient who gave up on their treatment, and then therefore didn’t find out whether they benefitted.

Then of course we have the OS [overall survival]. Now remember the previous slide I showed you was statistically significant with a good hazard ratio. This one not because it wasn’t powered for overall survival, but you clearly get a trend. It wasn’t worse, and we can say that even with a nonstatistical trial of this size. Whether we believe it’s better or not, you can think that if you like, but it trended that way. Again, small numbers but certainly suggesting that dosing this way, which is more labor intensive, I don’t want anybody to forget that. It requires monitoring by either a nurse practitioner or a physician.

We’re working on an app where we load it on the patient’s phone, and each day they tell us how many pills they’re talking. We list 5 toxicities. If it gets to a certain level, it sends a note to one of us so we can whether they’ve gotten their labs, that sort of thing.

I do think technology is making this easier, but it’s not a, “Send them a prescription, I’ll see you in a month,” kind of prescription. You must watch them more closely.

John Marshall, MD, reviews efficacy data from the ReDOS study of regorafenib in patients with mCRC.

Regorafenib in Third or Fourth Line for mCRC: Efficacy in the Re-DOS Trial

: Let’s talk about REGO [regorafenib] specifically. I’m sure everybody listeningtonight is familiar with the compound. As with all things, it works better in GI [gastrointestinal cancers] if you put a fluorine on it. It’s a multitargeted tyrosine kinase inhibitor. It hits, as I like to say, 19 different targets, sort of a dirty bomb. We’re never sure which target it’s hitting or if it’s multiple, but I also like that intellectually because I think about the targets within colorectal cancer as not being particularly strong drivers, right? Unlike say breast cancer or lung cancer, where if you find one of those targets, you totally change that patient’s outcome for the better, our drivers are OK but not that strong. Hitting more than 1 pathway at once to me, intellectually, makes sense. And this drug certainly has been proven in a variety of clinical trials. One of lead authors of the CORRECT study is on the call with me tonight, so I won’t get too gritty on that one and the CONCUR study.

You ran these studies and they were demonstrations of, to me, some very important things. One is that there were a lot of these patients around. The trials accrued quickly. The patients in this setting from around the world were willing to randomize to a placebo arm 2:1, but still a placebo arm. And it demonstrated clinical benefit and survival. The keys to the studies, when I look back, is that the patients had to be eligible, they had to be in a good performance status. They had to have good organ function. Nobody was in the hospital considering hospice or next-line therapy. They were in our clinics, driving themselves, so they were good performance patients with colorectal cancer, and we know there are plenty of them out there.

One of the places where we got lost is we then saw it as a drug that might lift somebody back from a rapidly falling performance status, and it just doesn’t do that. These drugs don’t do that. To me it’s about replicating that scenario that you led in the study about performance status, good organ function, and being able to comply. Then we’ll talk in a minute about dosing. But what’s your take on; when you were designing the studies and running them, did you think that you could pull it off?

: I was quite comfortable and convinced that we could pull it off. It was as you highlighted, it was one of the first, I think the first, phase 3 study where you had a placebo control in a third-line setting with a 2:1 randomization. We were wondering, and people were afraid, would the study ever accrue, because you had a placebo onboard. But they have such an unmet need. Patients wanted a new treatment because there was no other treatment.

The study accrual was projected to take 26 months. The study accrued in 10 months—16 months ahead of schedule. It shows you the unmet need that we had. And for me every patient with cancer represents an unmet need. We need to do better. We need to do warp speed for our cancer drugs for our patients.

But in the end you’re right. The drug is active. We have phase 3 data, actually 2 phase 3 studies. I also believe that less heavily pretreated patients really are the ones who should be targeted. I strongly believe not every patient with colorectal cancer needs regorafenib. Because if you have a PS2 [performance status of 2], let’s say rapidly progressing disease, deteriorating performance status, this is not the drug that can turn things around. There’s no Lazarus effect with regorafenib.

We’ve learned this over time, sometimes in a painful way, that finding the right patient is important. That’s exactly the patients that you mentioned. And fortunately, I believe, in a third-line setting, there are still a lot of patients who have good performance status, who are able to tolerate, and have adequate liver function, etc. This is a benefit of colorectal cancer, not pancreas cancer, gastric cancer. Patients in a third-line setting with gastric cancer, pancreas cancer are much sicker than in colorectal cancer on average.

: Yes, there are plenty of windows for this and TAS-102 [trifluridine, tipiracil]. We don’t think about them. If you’re considering an early phase clinical trial and you need some washout, or a rebiopsy, or a break, or you’re waiting on a slot, this is a perfect window to try some of these. I’ve been sandwiching this around liver-directed therapy on occasion when they’ve already had a lot of fluoropyrimidine.

It takes work. This is not an easy drug to administer or monitor. Don’t think, “Oh, I’m sending you away, and I’ll see you back in a month.” But if you do it and you use it correctly, you can see more clinical benefit for our patients.

What really burned this medicine early is that we used it in the wrong patients as you describe.

Experts in metastatic colorectal cancer review therapeutic options after progression for patients with mCRC.

Therapy After Progression in  mCRC

: Let’s talk about inducing the patient with unresectable cancer, too many weeds in the yard, as I like to say. But we’ve had a nice response. We’ve had some window of maintenance. This can often be a year from diagnosis now, and we start to see some progression. And I know you are very descriptive. You and I believe there’s progression, and then there’s progression. And then do you rechallenge? Do you bring in the IRI [irinotecan] if you haven’t? Do you bring your OX [oxaliplatin]. I know you probably do different things with different people but walk us through your thinking when you’re seeing that patient progressing.

: First of all, there is a difference between reintroduction and rechallenge, in my eyes. The reintroduction is what we do when we have induction maintenance, reintroduction strategy. For instance, it’s common with oxaliplatin if a patient’s tumor had not progressed on a regimen, on the drug, then you reintroduce it and there are tons of data of that.

Rechallenge is more likely to be used if you had tumor progression at some point and then 2 or 3 years later, you’re running out of options. You’re thinking, let’s go back to the furthest treatment away in a timely manner, and you rechallenge the cancer.

Now, when I go from first- to second-line treatment and use FOLFOX BEV [fluorouracil, oxaliplatin, bevacizumab] first line then after 8 cycles I go to a maintenance therapy. And eventually patients show tumor progression. The question of what to do depends a little on the velocity of the progression, the timing of the progression, and the adverse effects that patients might have.

There are different parameters. The patients, for instance, who had progression within 2 months of maintenance therapy, maybe had the first scan and progression-free survival. I normally switch completely and go to an irinotecan-based regimen and do something very different.

Yes, if you have maintenance therapy that works beautifully and these patients survive for a year or 2, a year and a half, no more neuropathy, and you had a great response initially, why not go back to FOLFOX? That is something. You have options, and it’s, again, not 1 size fits all.

: That first case where there’s a fairly rapid progression, where it makes them left-sided RAS wild type, are you going right? Is that when you’re going to play your EGFR [Epidermal Growth Factor Receptor]?

: Yes. That’s where I’ll probably put the EGFR card.

: That’s your next chance at a decent response, right?

: Right. Because response rate has always been higher with EGF receptor antibodies than with BEV beyond progression. It’s a different, cytostatic effect with BEV compared with a cytotoxic effect with EGF receptor antibody.

: But even if you skip second-line FOLFIRI [folinic acid, fluorouracil, irinotecan]-BEV, what it is, a 10% response rate? Whereas if you just need EGFRs in the right patient, you get up to 40% and 50%.

: Right. That’s actually an interesting point. We are overestimating the response rate in second line. If you look at the criteria for partial response, it’s really in the range of 10%. It’s not a lot.

: That’s important; as we all get to our third-line options, everybody says, “Well, you don’t get responses with those.” Well, frankly you don’t get that many responses in what you are enthusiastically giving in the second line, right?

: That’s an important reset after that initial response, apart from a BRAF, or an EGFR, or an MSI, or HER2, if you find one, you’re not going to get a response no matter what you do. Those are the lines of therapy.

: Irinotecan may be included; other biologics maybe, depending on the progression. Do you ever, and this is a bit of a provocative because I have 1 patient on this right now where I’ve given regorafenib [REGO], in what is in essence second line, because the patient doesn’t want to go to the more intensive intravenous therapy. We kind of know how to handle the drug better and it’s sort of maintenance too for me with a low-volume, good performance status patient. Have you ever done something crazy such as that?

: It’s not crazy. Patients tell us and guide us. It’s shared decision-making. It’s not, OK, you can push your FOLFIRI back through into the patient. You know, some patients want to have oral treatments.

And now that we have better ways to handle regorafenib in terms of toxicity management, etc, dosing strategies, it is worthwhile at least considering that. It’s not that common because, again, it’s clearly not for patients where you need a response to salvage or to get patients out of a threatening situation. Because REGO doesn’t induce responses. Same as TAS-102 [trifluridine–tipiracil], it’s the partner molecule, more or less. But it is 1 other option, and the more options we have, the better it is.

: Yes. I mean we all, for fear of not getting paid, to be blunt, we want to stick to our guidelines and our lines of therapy. And so many of these companies require us to fill out what line of therapy you’re in in order to get reimbursed. And I get that, and we don’t want to be missing expensive medicines. But at the same time, having a little bit of leeway to me is appropriate. And I believe some of these, what we call third-line therapies, in that good responding, good biology patient might in fact have some interesting…settings when patients still have excellent performance status and want rates from the unit.

Experts in metastatic colorectal cancer discuss when and in which patients to use a chemotherapy-free treatment interval.

Maintenance or Chemo-Free Interval in mCRC

 What’s our first and best chance at a response induction therapy? It’s where we get our highest response rates. They are most durable for almost all the sequence that they’re about to head off on. And the old tradition, of course, was to give this for a magical 12 cycles for 6 months. And it was almost a bravado among physicians that they felt obligated to get to that 12 cycles, or that 6 months, as though it was some rule, as though it was the law.

And I know we see eye to eye on this. For me, I tell a patient right from the beginning when we’re setting out on induction, “We’re going to take a look at about 3 months. And if you’ve had a nice response, I’m not going to go much longer because I’m not going to get much more response. But I’m going to take those drugs that got us there back off the table and try to. Instead of how much chemotherapy can I give you, it now switches…to how little chemotherapy I’m going to give you, assuming that we have a noncurative approach. How do you handle that? Are you taking people to 6 cycles or 6 months, or are you stopping on your maintenance a bit earlier?

: First of all, there’s really no magic to 6 months or 12 cycles or whatever you might call it; there’s absolutely none and I actually educate my patients right up front about this induction maintenance therapy or when I plan a change in treatment intensity. Because I don’t want them to say, “Oh my God, what’s happening” when we’re changing treatment. They need to know about this. And when you educate them about this up front, when you provide the rational, you know the initial response is the strongest. You do a first scan after 4 cycles, 2 months, let’s say 6 months, 6 cycles, 3 months, whatever you like. I do it after 2 months.

You see the stronger response then when you did the next scan. Your initial response is stronger. And then eventually you maintain the response. I tell my patients I want to turn this into chronic disease; again, using the least amount of treatment necessary to control it. And I tell them, all right, especially when you start the FOLFOX [fluorouracil oxaliplatin]-based therapy, you’re obligated to stop oxaliplatin before patients run into issues.

With irinotecan-based therapy, if you’re a FOLFIRI fan, you know first line, there’s nothing wrong with that, I still do de-escalation. Probably not as rigidly after 8 cycles, because you can push past, you live longer, but you don’t have to.

I strongly believe that at some point the tumor cells that are…still around are resistant to the intense treatment. Then it’s really kind of using capecitabine [CAPE]-BEV [bevacizumab]. I know you and I see on that eye to eye, too, as maintenance therapy and to try to maintain quality of life.

: Yes. We can’t forget that we want that patient’s bone marrow and liver function to be good maybe 2 years from then, right?

 If they’re doing well and if we keep giving them chemotherapy cycle after cycle, we’ll wear them out physically and also wear out the nonrenewable, unlimited resources of bone marrow and liver. I do the same if it’s with irinotecan. And, it’s even become more interesting in the triplet front line because then I’m pulling these other drugs back and…I sort of describe it as a chest game. I haven’t lost a piece, they’ve just moved it; leave it on the board there for a while and we’re going to move it again later. But if we can get by with a few pawns for a while, what a great way to go. And I’m the same way with the CAPE-BEV.

: One thing I get asked, when you start with a triplet, which I do more and more, what do you do in second line? I say, this is not the question because second line happens on maintenance or in treatment-free interval. The median progression-free survival of these triplet combinations is 12 months plus. However, patients still get treated for 12 months with a triplet. We stop treatment after 4 or 6 months and then when they progress you can reinitiate. And you use oxaliplatin, irinotecan, whatever, again.

: I almost never totally stop chemotherapy when people have visible metastatic disease. There are a few exceptions. For instance, when patients with very small volume disease that’s already proven itself to be very quiet, I’ll take 3 months off, something similar to that.

I did that this past summer. I gave a young man with metastatic disease the summer off, partly because of COVID-19 [coronavirus disease 2019], and, yes, he progressed, but not very much. We didn’t lose much ground in 3 months.

: I do that, too, maybe try to see around special days, vacation, holidays, family events, whatever. But 1 issue is we need to keep a close eye on these patients, and we need to know tumor biology. And there’s a point when you go to maintenance, you already know tumor biology of this patient. And not every patient is the same. My default is to keep patients on some form of treatment.

: I’m the same way. And this is a little special scenario of the patient who is fortunate enough to have retractable disease. Just as a quick little aside before we move into beyond progression, if you will.

There are data now that adjuvant therapy really doesn’t do its adjuvant thing. It might delay relapse, but it didn’t cure more people. The 3 months of adjuvant therapy, now the norm, it sort of made me change my thinking if I was giving induction chemotherapy, neoadjuvant, for example, and now we resect metastasectomy. I’m probably doing a ctDNA [circulating tumor DNA] test afterward to see if it’s negative. But then I’m probably leaving him alone more and more. What do you think?

: You know, you brought up an interesting point. As you were talking, I was thinking this is exactly where we need to have ctDNA.

Because there are some patients who don’t benefit from adjuvant therapy who have recurrence anyway, and there are some patients who don’t need it, they will not have a recurrence. We’re overestimating the importance of postoperative adjuvant therapy once you’ve given neoadjuvant treatment, which we commonly do in the metastasectomy setting. Because we can see the response, see the chemosensitivity. Regarding ctDNA, the minimal residual disease, there are trials that are being run right now that will completely change the way we use adjuvant therapy. It’s adjuvant therapy when you need it, when you start turning positive and then you can spare adjuvant treatment for a lot of patients.

: I still get questions almost once or twice a week from our community partners asking, “Can’t I give 6 months of adjuvant therapy?” I say, “You’re done, you’re okay, it’s fine.” It’s been so hard for people to let go of those additional 3 months. I don’t know if you’re getting the same sort of feedback.

: It’s still interesting, and some of these trials are still using 6 months as a default. You know, as a control arm, and that’s hard to justify because…if you really talk about 6 months of adjuvant FOLFOX-based therapy, that’s a lot of toxicity, persistent toxicity.

Axel Grothey, MD, and John Marshall, MD, discuss strategies for optimizing first-line treatment in metastatic colorectal cancer and the role of adjuvant therapy.

Optimizing First-Line Therapy in mCRC

 There’s a breakdown that I like to see of the different types of treatments we could use right now. Fluoropyrimidine plus bevacizumab is an active treatment, and it’s definitely 1 of the treatment options I could consider in patients with low-volume disease or smoldering disease, or in elderly patients. And then anything goes in terms of doublet chemotherapy plus bevacizumab [BEV], triplet chemotherapy plus BEV. Then you see the EGF [epidermal growth factor] receptor antibodies, which have some sidedness restriction and some molecular restriction. We really need to have the molecular profile and information about sidedness available.

There’s the old standard, chemotherapy plus biologic agent. We need to pick our poison regarding the intensity of treatment according to what we want to achieve. I, for instance, see if there is a left-sided tumor, RAS, RAF, wild type, or if the patient is threatened by a tumor load and we need to force a response. I also need to see if the patient is still in decent performance status. But we know it’s going downhill very soon. I use FOLFOXIRI [fluorouracil, leucovorin, oxaliplatin, and irinotecan]. Cetuximab has an adverse effects. I’ve used FOLFOXIRI triplet plus panitumumab, and this is based on data from the German study that show when using this approach, you have a 90% response rate. And it is remarkable that we have tools in our hands where we can really, almost reliably, predict that this patient will have a response.

And then for more recently MIS [microsatellite instable]-high tumors, we have seen randomized data from KEYNOTE177 [NCT02563002] that showed pembrolizumab, a single agent, is superior to standard doublet chemotherapy plus a biologic in MSI-high mismatch repair deficient cancer. That is definitely 1 of the treatment options we can consider in first-line therapy.

Eventually, I hope we’ll have this BEACON regimen combination with BRAF inhibitor, plus an EGF receptor antibody, also moving to frontline treatment in patients with BRAF V600E-mutant tumors. It would be potentially not in the biologic, in a pure biologic doublet, but in combination with chemotherapy. That is currently being investigated. But you see the granularity in the person with aspect of first-line treatment.

: In adjuvant therapy, we are embracing the idea that capecitabine may be a better drug in the adjuvant setting; not by a lot, but by a little. I do like to use a lot of capecitabine. It’s FOLFOX [folinic acid, fluorouracil, and oxaliplatin] this, FOLFIRI [folinic acid, fluorouracil, irinotecan] that shouldn’t we be using more CAPE [capecitabine], or is it OK to use more CAPE in the metastatic setting and not just the maintenance window? What’s your thought on that?

: There are data and the perception that capecitabine is not tolerated well. US patients had poorer tolerability than the European and Asian patients, in particular.

And potentially, based on the folate enrichment and fortification that we have in our food, it’s because folate makes a leucovorin activated. Folate makes capecitabine more toxic.

On the other hand, now we’ve used capecitabine, CAPOX [capecitabine, oxaliplatin] in the adjuvant setting, more because you can get away with a shorter deletion of treatment compared with FOLFOX. Patients might actually lose their fear of the adverse effects. We use it commonly with radiation therapy. I have started using more CAPOX in my clinical practice, with a biologic. And there are not that great data on CAPOX plus EGF receptor antibody; that’s the 1 caveat that I have, the skin toxicity potential. But I believe, I agree. We use capecitabine a lot in the maintenance setting. We should potentially use it in combination with oxaliplatin, at least in the metastatic setting.

: I agree with that; many of us are fixed on the pump, but with FOLFOXIRI we’re still stuck with the pump, and I think when you’re using oxaliplatin it’s a perfectly legitimate alternative.

: Right. And you know that’s another interesting point. People think you can’t deliver oxaliplatin through a peripheral line, which is not true. You actually can do that, you just need to educate patients about it, and you need to educate your chemotherapy nurse that this can be done.

: Right. And the moment we walk into a cancer center, the first thing the patient should get is a port.

: I’ve been leaning lately toward the triplet because of the FOLFOXIRI studies that have been done with an overall survival or a response rate, a PFS, [progression free survival] even in places where lines of therapy are all available. And I come back to, is it depth of response? What’s the reason for that? But the more heavily I’m weighing maintenance therapy, the more I lean toward FOLFOXIRI being a standard of care for a lot of patients, not just these high-risk patients. Is that too far?

: No, it’s not too far. I have been using triplets more than before. I believe we lost the fear of these triplets by using them in pancreas cancer. But it means for patients more poor performance status commonly, and we use it quite commonly. In our institution we kind of merge FOLFOXIRI and FOLFIRINOX into 1 regimen.

You know the subtle differences: irinotecan and bolus here, or not. We use a nonbolus 5FU containing regimen, just pure infusional 5FU. With irinotecan 150 mg and oxaliplatin 85 mg, it’s very well tolerated.

: And then the third question, you know it’s this right-left, and it’s a bit provocative. But with BRAF suddenly, if you have a mutated V600E, right/left doesn’t matter anymore. Because you’re going to use an EGFR with a BRAF in a right-sided tumor, right? So right/left goes out the window if you have V600E.

: If you have a V600E, we’re not there yet that I would say it’s similar to HER2 in breast cancer, where initially when you heard HER2-positive tumor and we didn’t target HER2, it was a poor prognostic marker. And now it turns into a good prognostic marker, a patient with HER2 and amplification live longer, but BRAF is still bad.

 But we’ve been saying, “Don’t use EGFR.”

: No, that would be in the context of a BRAF inhibitor.

 I would say right-sided, but if it’s BRAF.

: Correct. It serves a different purpose. I mean, it still doesn’t work with chemotherapy, but it works to counteract the feedback when you inhibit BRAF with encorafenib for another BRAF medicine. It’s actually pretty cool translation biology that they used to figure that out.

: I know, I think so, too; it’s very clever.

: I try to make it clear, bevacizumab is not for everyone. We talked about single-agent capecitabine, plus or minus biologic triplets, with or without a biologic.

There’s lots that we can do and there is a lot we can personalize, even beyond going for molecular markers. And the option goal is making all drugs available to all patients, not just folks in a first-line treatment, but over time looking at this continuum of care to extend lives and maintain quality of lives as long as possible.

This is what I tell my patients. That is my catchphrase on this slide here: using the least amount of treatment necessary to control their disease. And that means sometimes you do induction maintenance therapy approaches. You give patients a break from treatment. You use all the armamentarium of treatment approaches that we have available. And that’s something that is the art of oncology, helping keep patients alive and maintain their quality of life.

Axel Grothey, MD, and John Marshall, MD, discuss first-line treatment options in patients with metastatic colorectal cancer.

First-Line Options in mCRC

: Good evening, everybody. My name is Dr John Marshall and joining me tonight is my good friend and colleague, and really the world’s expert on gastrointestinal cancers and colorectal cancer, Dr Axel Grothey.

® event. Our focus tonight is on the management of metastatic colorectal cancer [mCRC]. And as I mentioned, I am John Marshall from Georgetown University, director of the Ruesch Center for the Cure of Gastrointestinal Cancers. And Axel, of course, is known to everybody. He is the director of Gastrointestinal Cancer Research at the West Cancer Center & Research Institute in Germantown, Tennessee.

Axel and I are going to talk about the optimum treatment for metastatic colorectal cancer [mCRC]. And there are a lot of new data that you need to be aware of. We will remind you of some important older data as we talk about clinical scenarios through lines of therapy, including what is the right first-line therapy.

We’re going to talk about maintenance therapy and how best to optimize chemotherapy lite or chemotherapy free interval. And then we’re going to talk about later lines of therapy, second, third, and beyond.

Axel, it seems we went from a time of 5FU [fluorouracil] only, then FOLFOX [folinic acid, fluorouracil, and oxaliplatin] for everybody. And now with molecular profiling, it’s not a straightforward 1-size-fits all front line for everybody. Give us your sense of how you figure out that new patient in front of you with metastatic disease. What are you going to do with him? What are you going to give? What molecular testing do you need to know, etc?

: It’s not a 1-size-fits-all front line. I still believe that in the United States most patients will receive FOLFOX plus bevacizumab as their “go-to.” It’s the broad-spectrum antibiotics that we use when we don’t know what to do or which bug we are actually treating.

And this is for a good reason. FOLFOX plus bevacizumab has activity in left- and right-sided tumors independent of the molecular profile that we have, but we can do better. And this slide here that you highlighted just now looked at what I like to present as the factors influencing treatment choices in metastatic colorectal cancer. Of course, these are patient characteristics. We’re treating older patients differently than younger patients. This is not ageism; it’s just about how patients tolerate things. And we have some patients who have been pretreated with adjuvant therapy. Performance status is a big issue.

The goal of therapy is patient preference, but more and more it will be the molecular characteristics. And the 4 factors that you see on the right upper side, the RAS mutation analysis, BRAF mutation, and HER2, are the cornerstones of what we really need. It is kind of the basis of molecular testing that can really guide therapy.

I would even include the tumor location as 1 of the factors that has similar standing as the molecular factors to the idea that tumor location, right-sided tumors, meaning anything proximal to the hepatic and splenic flexure, based on some European data, those tumors do not respond to EGF receptor antibodies when added to chemotherapy. And data are consistent that those tumors, even when they’re RAS wild type, BRAF wild type, do not respond to cetuximab or panitumumab. It is tumor location as a predictive factor. And it’s also a prognostic factor; right-sided tumors do worse than left-sided tumors. That is really kind of setting the stage for our treatment decision.

Axel Grothey, MD, discusses factors that influence treatment selection in patients with mCRC.

Axel Grothey, MD, West Cancer Center & Research Institute

Articles