Case 1

Van Morris:

Welcome to the JNCCN Video Roundtable Series. Today’s forum is titled “Tailoring Targeted Therapies for Treatment of BRAF-Mutated Metastatic Colorectal Cancer.” My name is Van Morris, I’m Associate Professor in the Department of Gastrointestinal Medical Oncology at The University of Texas MD Anderson Cancer Center, and I’m really excited and honored to share this forum today with two of my esteemed colleagues and I’m going to let them introduce themselves. First, I’ll start with Dr. Parikh.

Aparna Parikh:

Hi, my name is Aparna Parikh. I’m Assistant Professor of Medicine at Harvard Medical School and an attending physician at the MGH Cancer Center in Boston.

Van Morris:

And Dr. Pedersen?

Katrina Pedersen:

I’m Katrina Pedersen. I’m Associate Professor of Medicine in the Division of Oncology, section of GI Medical Oncology at Washington University School of Medicine in St. Louis and the Siteman Cancer Center.

Van Morris:

Thanks. Today we’ll be talking about the treatment of BRAF-mutated colorectal cancers and using it in targeted therapies across four different patient scenarios. These are our disclosures.

CASE 1

Van Morris:

To start things off, the first case is the case of a patient with a treatment-refractory metastatic colorectal cancer. I’ll start off by reading this patient’s history. This is a 54-year-old woman with a history of a T4 N2 M0 stage III adenocarcinoma of a hepatic flexure. She underwent complete surgical resection and completed her adjuvant chemotherapy with FOLFOX. She comes back within 6 months of finishing chemotherapy and at the time of her first restaging, she does not have any complaints but is noted on imaging studies to have developed new bilobar liver metastases. We’ll talk about a little bit on the next slide as well, the patient is confirmed, unfortunately, to have metastatic colorectal cancer and in speaking with surgical colleagues, her cancer, unfortunately, is deemed to be unresectable.

Again, overall she appears well. She reports 1 month of some dull right upper quadrant pain, but she’s able to maintain all of her activities of daily living. Her medical history is otherwise significant for hypertension, hyperlipidemia, and a history of atopic dermatitis. As we mentioned, she did have a biopsy. This was notable for a poorly differentiated adenocarcinoma with immunohistochemistry showing intact expression of all four of the mismatch-repair proteins and she’s found to be HER2 neu 0+ by immunohistochemistry. Mutation profiling with next-generation sequencing confirms the presence of a BRAF V600E mutation, a KRAS/NRAS wild-type status, and a tumor mutation burden of 12 mutations per megabase. Again, her laboratory data all are within normal limits, and her ECOG performance status is 0. Her physical exam is otherwise unremarkable.

The first thing that I’ll ask to Dr. Parikh are certain clinical and pathologic characteristics associated with patients with BRAF V600E colorectal cancer? What should oncologists look for and think about in terms of some of the associations of patient characteristics and BRAF-mutated CRC?

Aparna Parikh:

Thanks Dr. Morris. Very unfortunate case here with this woman who presented with very locally advanced node-positive T4 initial colon cancer. A couple things to note, so she was found to have right-sided disease, so her tumor was found initially at the hepatic flexure. Next, she was also identified as a woman, and we know that those patients, women are more likely than men to be found to have BRAF-altered disease. A couple of the pathological features, so again, right-sided tumors, demographic, women. And then when we looked at the histology, so poorly differentiated adenocarcinoma. These tumors tend to be more poorly differentiated, so some of those features are what are notable in terms of this particular patient and her BRAF V600E status, but welcome additional thoughts on that.

Katrina Pedersen:

I agree. I think that you hit on all the high points. Whenever I do see an older woman right-sided tumor, I’m thinking BRAF or MSI-high usually, especially if it seems like it’s more locally advanced. Whenever I see nodes, I’m actually leaning a little bit more toward BRAF involvement and less expecting to find one that’s on the mutator pathway.

Van Morris:

I absolutely agree. Especially I love the comment about the nodes. When you look at patients who have BRAF mutations, they’re more likely to have lymph node and distant lymph node involvement. Also, even though it wasn’t the case with this particular presentation, higher rates of peritoneal involvement, as well, relative to the BRAF wild-type counterparts. But yeah, these are all great points that I think clinicians need to be thinking about and associating how patients with BRAF-mutated CRC may present to their clinic.

The patient progressed immediately after completion of FOLFOX adjuvant chemotherapy. The question here, what therapy would you recommend? If we were asking this as a multiple-choice question with our fellows, a BRAF/EGFR inhibitor with encorafenib + cetuximab. A BRAF/EGFR/MET combination, encorafenib + cetuximab + binimetinib chemotherapy. The patient had FOLFOX, not had FOLFIRI, has a RAS wild-type tumor. Could we do FOLFIRI + cetuximab or, and we'll go into the details more of this a little bit later. Pembrolizumab as a single agent, the patient's TMB was greater than 10. I'll start with Dr. Pedersen now, kind of which of what would you recommend for this patient if she were to come to your clinic?

Katrina Pedersen:

For me, there’s really one trial that brings a lot of the top three choices together and makes the decision somewhat easier from a clinician standpoint. And I would choose the encorafenib and cetuximab because of the BEACON data. To start with, as you mentioned, she had fairly rapid progression to metastatic disease at the end of adjuvant therapy. Whether we were trying to find a clinical trial for this patient or standard of care, most of us would convert our thinking of that FOLFOX really serving the role of being a first-line therapy. A trial like BEACON, which was for second-line patients with BRAF V600E mutations, really does apply nicely here. And that study did take a look at a comparison. I believe the primary endpoint was to be the triplet of the BRAF V600E inhibitor, encorafenib; with cetuximab, the EGFR inhibitor; and then the MEK inhibitor, binimetinib, compared actually to the control arm of FOLFIRI and cetuximab, or if I’m recalling correctly, it was a few years ago now that we were all participating in it. It could be iri-cetuximab as well.

But then they did have the doublet arm of encorafenib and cetuximab. And I remember initially the excitement of seeing that yes, the primary endpoint was met, that survival was improved with patients who had the triplet regimen. But then with the updated data a little bit later, it turns out that the doublet actually had pretty much identical overall survival without the additional toxicity from the MEK inhibitor. I really use that to guide my practice and go toward a doublet agent. As far as the TMB goes, I would not lean toward pembrolizumab for TMB of only 12 in a patient with BRAF-mutated MSS.

Van Morris:

Yeah, I completely agree with everything you say. I think as you mentioned, Dr. Pedersen, the BEACON study is the largest and only phase III trial to date reported for patients with BRAF V600D metastatic colorectal cancer. While it was a positive study, the median progression-free survival was only 4 months, the median overall survival was only 9 months. While it was definitely a step forward, I think it just highlights also the fact that these patients still didn’t live very long or not very much time elapsed before they were progressing. And it again, highlights the fact that the BRAF 600E subpopulation, which accounts for 8 to 10% of all of our CRC patients; it’s a sizable number. These patients still have very unfavorable clinical outcomes relative to their BRAF wild-type counterparts that we often see in the clinic.

While encorafenib-cetuximab is the data-driven solution at this point, it just highlights the need for further studies and doing better in this population. We’re talking a lot here about encorafenib. There have been other trials which have evaluated other BRAF inhibitors. Encorafenib is one BRAF inhibitor, vemurafenib is another, dabrafenib is another one. They are all used in oncology. And I guess my question for Dr. Parikh, is there interchangeability, or do we see similar efficacy with other BRAF/EGFR combinations or are you only choosing encorafenib in your practice?

Aparna Parikh:

Great question. And fundamentally I think as Dr. Morris alluded to, one of the challenges that we see in general with BRAF-targeted therapies is the challenge with rapid sort of upstream activation of the MAP kinase pathway at the surface level. And I think preclinically, maybe some of these BRAF agents act a little bit differently in terms of when you look at pathway inhibition. That being said, others such as vemurafenib do have clinical data for it. One of the earlier BRAF triplet trials was SWOG 1406, and that study was looking at vemurafenib in combination with cetuximab + irinotecan versus just cetuximab + irinotecan alone. That study—PFS was the primary endpoint—PFS was reached, and I believe the response rate was around 17%. I think prior to having the phase III data from BEACON was very much actually using the vemurafenib + cetuximab + irinotecan combination short of a clinical trial. But I think now that we have the BEACON data, have not necessarily swapped out them for enco or another dabrafenib, for example.

Van Morris:

You’re right that the SWOG 1406 study was the first randomized trial to show some survival benefit with PFS, as it’s relative to the irino-cetux combination. And prior to that there were other smaller, single arm, single-institution trials that looked at the various permutations of BRAF and EGFR inhibitors and again showed similar numbers as what we saw in BEACON.

So we talked a lot about the BRAF/EGFR combinations here. We know that BRAF inhibitors as monotherapy work very well in patients with other BRAF-mutated solid tumors. In melanoma, for example, response rates in the 40% to 50% range with BRAF inhibitors as monotherapy. In non–small cell lung cancer, reported overall response rates exceeding 40%.

I’ll turn back to Dr. Parikh again. If you had a patient who, for whatever reason you just felt that an EGFR therapy, an antibody, were completely contraindicated, would you recommend a BRAF inhibitor as monotherapy in the context of colorectal cancer?

Aparna Parikh:

No. I think unfortunately this has really been challenging for colorectal cancer, whereas tumor context actually does matter and is paramount. It’s not for lack of trying, we have looked at monotherapy for BRAF CRC for over 10 years and barely sort of hit single-digit response rates. Though it feels like you want to offer that patient something, I think in fact we would probably be causing harm knowing that we were giving something that just didn’t have any efficacy.

Van Morris:

Response rates were 5% with dabrafenib as a single agent. Yeah, ’''s exactly right. Single digits. I completely agree with you, Dr. Parikh. We talked some about the interchangeability of BRAF inhibitors, and I’ll direct this question to Dr. Pedersen next, which is we’ve been talking BEACON about encorafenib and cetuximab, but can you talk about your thoughts on the interchangeability of encorafenib-cetuximab versus encorafenib-panitumumab, especially since you practice in St. Louis and in Missouri and what that means in your day-to-day practice as an oncologist?

Katrina Pedersen:

So there have been studies not necessarily in the BRAF-mutated context of CRC that’s compared the noninferiority of the efficacy between cetuximab and panitumumab and did find them to be noninferior to each other. Biologically, it makes sense to me that a drug like panitumumab would have some favorable characteristics because it is a humanized monoclonal antibody. As opposed to cetuximab, which is a chimera based on a murine antibody. You’re going to get more hypersensitivity. Well, you’re going to expect to get more of it if you’re using a drug like cetuximab on this murine backbone versus a human backbone like panitumumab.

What’s ironic about this question is that I’d been favoring panitumumab for a long time, I do view them as interchangeable. However, this actually predates the time of my practice in Missouri. I was actually further north when all this data was initially coming out. In preparing for our chit-chat today, guys, I actually had to go and dig up with Dr. Parikh’s help, the original Missouri data that actually came out of the University of Missouri about 15 years ago, showing that the hypersensitivity rates of grade 3/4 reactions, not minor ones either, was on the order of about 25%. Like one in 4fourpeople were having some pretty rocking reactions we’ll say. Which is not something that you want to put your infusion nurses through too often, if you want to keep them on your side.

So in this part of the world and really throughout the southeast section of the country, there’s a lot of data that’s really been driven out of UNC, Virginia, Vanderbilt. I don’t think I saw any Texas-specific, but I did see some Oklahoma-specific data that was also showing similar patterns where five times, six times more hypersensitivity reactions with cetuximab. It makes sense if you’re in the lower right-hand quarter of the country that you might want to consider panitumumab.

Aparna Parikh:

Yeah. What’s interesting is I think thinking about clinical trials and things too, so many clinical trials sort of stick with your monoclonal antibody of choice. You’re either left with cetuximab or panitumumab based on the trials. What I’ve done for my practice, we’re in the northeast, so cetuximab is actually what is on our formulary. And if a patient is coming in from the south sort of belt somewhere in one of these states, there’s actually an Ig antibody test that you can do. It’s not perfect, but you can maybe get a sense of how prone they are to have that sensitivity. And it’s come up for trials that are using cetuximab for BRAF-targeted therapy versus Pmab. How do you think about whether it’s premedications or just monitoring these patients closely or desensitization and things, strategies to get them. But I think a plug for our trial folks to recognize that not everyone in the country has the same baseline exposure rates. Trying to be mindful around which anti-EGFR therapy you are giving or allow potentially use of both.

Van Morris:

That’s a great point for trial advocacy and certainly patient inclusion as well. We have a SWOG study that’s open now across the United States, and this was very much a question that our colleagues in the southeastern United States were asking us. We support the trial, which we’ll talk about a little bit later. But we have a lot of angst about being locked in to use cetuximab just given the exactly these concerns. I lived in North Carolina before I came to Houston 11 years ago, and I was at Duke and I’d say besides UNC basketball giving cetuximab was the thing that oncologists feared the most at Duke as a UNC basketball fan, I'll make that plug myself here.

But yeah, I mean these are real problems across the country and I certainly when talking with other oncologists, if patients come to Houston and they’re going back to Missouri or going back to South Carolina, somewhere like that, oncologists will say, “I know the data is for enco/cetux, but are you comfortable with encorafenib/panitumumab in this setting?” I think we’re all driving the same point here that that’s okay. And I believe that the NCCN even lists enco/panitumumab as an option as well to hopefully make insurance companies easier to approve and get that to our patients.

Katrina Pedersen:

I think in the context of this case specifically, no matter where you’re at, but especially where I’m at. Seeing someone who has this history of atopic dermatitis would make me definitely second guess even enrolling them potentially on a trial with cetuximab or perhaps doing the IgE testing. Because a history of atopic diseases has been associated with a higher rate of cetuximab reactions.

Van Morris:

Absolutely. And thank you. Thank you for bringing that point out too. We’ll kind of move on.

We’ve talked a lot about BRAF/EGFR combinations in a setting. We’ve talked about not recommending BRAF monotherapies. Recently, the FDA approved in a tumor-agnostic fashion, the use of BRAF/MET combination dabrafenib and trametinib for patients with BRAF V600E advanced solid tumors. My question to Dr. Parikh is, would you recommend dabrafenib and trametinib as an FDA-approved combination in a patient like this?

Aparna Parikh:

Yeah. I think again, just given the data we’ve seen, I think the anti-EGFR with the BRAF-directed therapy as we’ve seen from BEACON and others, really seems to be the most important strategies. Really at that surface receptor tyrosine kinase, I think we need that anti-EGFR therapy to shut down the pathway upstream as much as possible. Given that, again, rapid feedback activation you have would just be BRAF monotherapy. I think there are other downstream targets from RAF that are being explored such as ERK inhibitors and things, which are quite exciting. And then later on we’ll talk a little bit about the PD-1 addition to see if we can extend that durability. But for now, I don’t think we have the data to really support MEK/BRAF alone.

Van Morris:

It’s interesting in the fine print for the FDA indication for dabrafenib and trametinib, they have just kind of a little parenthesis there that says “with the lone exception of colorectal cancer.” It's just important for oncologists to remember that the FDA approval is a tumor-agnostic approval minus CRC for patients with BRAF V600E advanced solid tumors.

Katrina Pedersen:

I went back and dug up the CRC dabrafenib/trametinib data that was published by Corcoran back in 2015, and it was only very, very modestly better than the vemurafenib monotherapy. The response rate was 12%.

Aparna Parikh:

12%. Right, yeah.

Katrina Pedersen:

They did have a disease control rate of 68%, which was interesting. PFS benefit of 3.5 months. That’s still, I’d say significantly under performs when you’re doing a cross trial comparison, which we would never do with our more modern approaches.

Van Morris:

Yeah. Speaking of another FDA approval, several years ago based on the KEYNOTE-158 study of pembrolizumab, the FDA approved pembrolizumab for use in patients with advanced solid tumors who have a mutation burden that was 10 mutations per megabase or greater. This is a patient with a microsatellite stable, and according to the FDA definition, a TMB greater than 10. Technically this patient has pembrolizumab as an FDA-approved option for her as well. I know we talked a little bit about this, but I’ll ask Dr. Pedersen, what are your thoughts on the likelihood of pembrolizumab as a single agent having benefits here?

Katrina Pedersen:

I honestly would not do it for really any MSS, but especially a BRAF-mutated MSS with a TMB of 10 or greater, unless it’s significantly greater. I mean, I’m talking on order of 40 to 50+ mutations per megabase. Because when you take a look at the data around TMB 10 in these patients, they really don’t tend to respond all that well in colorectal cancer. But when you’re really extending it back to thinking about the biology of what is really driving the growth of these cancers, TMB in this case is really coming along more for the ride with this MLH-1 hypermethylation that you might get more often with a BRAF-driven carcinogenesis. You really want to go after that BRAF, that’s truly the driver in this scenario.

Aparna Parikh:

I think it's unfortunate, ’''s another scenario where perhaps they should have put in that CRC was not actually included in the initial TMB trial. And there was a really nice paper by the MSK group, I think it was Rousseau et al in The New England Journal, and I can’t remember, a couple years ago. But what they did was when they looked at the TMB patients and when they took out patients who were MMR-deficient or had POLE or POLD alterations, the difference in survival basically with TMB entirely sort of went away. As Dr. Pedersen was alluding to, those patients with POLE/POLD are going to be ultra-mutated and those patients that are MMR-deficient, you’re going to see a higher TMB. But I think a little bit unfortunate because we’ve seen a lot of patients, I think after another tumor-agnostic approval, when you see these 10, 11, 12, many patients are getting exposed to checkpoint inhibitors.

Van Morris:

Yeah. Dr. Parikh I think mentioned this a little bit earlier as well, but there have been some kind of single-institution trials. Every patient I see in my practice asks, “what is the role of immunotherapy in colorectal cancer?” and in the context of a patient like this who has an MSS, BRAF V600E metastatic colorectal cancer, your group, and then our group at MD Anderson has done studies looking at combining MAP kinase–directed therapies with PD-1 antibodies. Can you talk about your group’s data and what is the role for a MAP kinase immunotherapy combination in this context?

Aparna Parikh:

Yeah. Again, this was based on knowing that the response rates are low, we saw a lack of durability. And then there was fairly compelling preclinical data suggesting that MAP kinase pathway inhibition may actually help improve the tumor immune response. I think very similar to the study that you guys did and are now looking at nationally, we ran proof-of-concept, single-arm phase II study combining PD-1 inhibition, BRAF, and MEK. We used spartalizumab, which is PDR001, dabrafenib and trametinib, and overall response rate was the primary endpoint. And the study did meet its OR endpoint with a response rate of just shy of 25% or so.

But notably, I think I can’t quite remember exactly what our median ended up being, but the durability for these patients, particularly for patients without prior exposure to BRAF-directed therapy, was quite long. Some patients were having efficacy and responding to therapy for over a year. And as we talked about a little bit early, the median OS and the BEACON data was still around 9 months. Really trying to see if we can synergize on that immune response from MAP kinase inhibition and not just focus on the responses, but if we can push out that durability for these patients.

Van Morris:

And I’ll just say, I would just really recommend that viewers check out the paper that was published in Nature Medicine earlier this year by Dr. Parikh and her colleagues, where they really had a nice translational science component to this as well. Where they were showing the kind of deep inhibition of the MAP kinase pathway really triggers immune activation. We’re talking about this synergy between MAP kinase–directed therapies and immunotherapy. There really does seem to be exciting initial data.

And to highlight Dr. Parikh’s point as well, our experience at MD Anderson, we also did a single-institution, single-arm study looking at encorafenib, cetuximab, and nivolumab in this same setting. About a quarter of our patients continued on study for over a year. And as she was talking about durability for the first time, we started seeing a tail on the survival curve. Which when we think about the poor prognosis of patients with BRAF V600E metastatic colorectal cancer, the ability to talk about a tail on a survival curve past the 1-year mark is super exciting, and we want to make sure that viewers are aware there is a randomized phase II trial that’s currently accruing across the United States.

It's the SWOG S2107 study, and this is randomizing patients with treatment-refractory microsatellite-stable BRAF V600E colorectal cancer to either standard of care, which would be encorafenib and cetuximab as we’ve talked about, or encorafenib, cetuximab, and nivolumab. Really seeing does the addition of a PD-1 antibody result in a survival benefit for these patients, and participants on this study would have a two-thirds chance of being randomized to the immunotherapy arm. Please keep that in mind as you’re seeing these patients in your clinic.

Aparna Parikh:

Yeah, just pulling up as we were talking to confirm so we could let people know. The median OS from our study was like 13.6 months, and the median duration of time on treatment was 7.4 months. Again, this was a small single institution study, but look forward to Dr. Morris’ randomized trial and entirely echo those sentiments. I think that’s a great, great option for these patients.

Katrina Pedersen:

This is one of those times where you wish you could have a fast-forward button to get to the results.

Aparna Parikh:

I know.

Van Morris:

So true.