Michael A. Davies, MD, PhD, discusses how he currently approaches treatment for patients advanced BRAF-mutant melanoma as well as exciting areas of current research to address ongoing questions in the community.
Therapeutic approaches to target the metabolic pathways, an enhanced understanding of how to utilize triplet regimens, and improving outcomes for patients with brain metastases are all areas being tackled in the setting of advanced BRAF-positive melanoma, explained Michael A. Davies, MD, PhD.
The invigorating research efforts build up on the 3 BRAF/MEK inhibitor combinations currently being used in this patient population: dabrafenib (Tafinlar) plus trametinib (Mekinist), vemurafenib (Zelboraf) and cobimetinib (Cotellic), and encorafenib (Braftovi) plus binimetinib (Mektovi).
“It is sort of challenging to really talk about objectively are the differences in efficacy because of the fact that they haven't been compared head-to-head,” said Davies of the 3 combinations. “However, there are some fundamental differences between them.”
In an interview with OncLive, Davies, professor and chairman of the Department of Melanoma Medical Oncology, at The University of Texas MD Anderson Cancer Center, discussed how he currently approaches treatment for patients advanced BRAF-mutant melanoma as well as exciting areas of current research to address ongoing questions in the community.
Davies: We have dabrafenib and trametinib, which was our first regimen was approved in 2014, then vemurafenib/cobimetinib in 2015, and then the newest regimen is encorafenib and binimetinib, approved in 2018. To this point, we still have no direct head-to-head trials comparing these regimens against each other. The fact of the matter is, we look at the data and use a variety of factors to judge between them.
First, they have very different side effect profiles. With dabrafenib and trametinib, the defining toxicity that we see is pyrexia that develops in about 50% of patients. It's usually self-limited and relatively easy to manage. However, there are a subset of patients in whom that becomes a recurring problem. There have particularly been issues in elderly patients and in those who are socially isolated; those fevers can cause a real problem, people can have the equivalent of cytokine storms, so that's something we consider.
For vemurafenib and cobimetinib, the most unique toxicity is significant photosensitivity. Finally, encorafenib and binimetinib has much lower rates both have fever and photosensitivity, but here is also fatigue and nausea [associated with them], which is similar to the other regimens. One of the other features that's different between these agents is that encorafenib and binimetinib has a much shorter half-life than the other 2 regimens. In patients who do experience toxicity, holding the drug will result in washout much more quickly. That gives you some advantages.
In terms of patients who have significant toxicities, there are also issues in which dabrafenib and trametinib needs to be taken on an empty stomach and trametinib has to be refrigerated. Despite some of those practical limitations, dabrafenib/trametinib is still the combination with the longest-term outcome data with robust 5-year data, and has actually the most data in patients with brain metastases, which is a significant issue in this disease. Each of [these combinations] have advantages and disadvantages to them.
In the absence of having head-to-head data, I look at all 3 of them as good options. You can pick which one you want to start with, knowing that if that isn't tolerated, you do have 2 other good options to work with.
One of the questions I often get asked is: If a patient is progressing on one regimen, can I switch them over to another regimen to get a response? In general, that doesn't work. The one exception I would say, in terms of directly switching and getting a response, is if you have a patient on a regimen where you have had to do significant dose reductions to get a tolerable dose. We have seen sometimes that switching to another regimen, if it's better tolerated and you get to a better dose, you can get to efficacy.
Now, what is very clear, though, is that resistance to BRAF and MEK inhibitors is not necessarily permanent. In trials, if patients were switched to something like an immunotherapy regimen for at least 3 months, if they then progressed on that regimen, then you can absolutely rechallenge with targeted therapy. In a phase 2 clinical trial in that setting, dabrafenib/trametinib had a disease control rate of over 70%. Switching directly from one regimen to another in the setting of resistance generally doesn't work unless you decrease the doses. However, this idea of being able to go back to targeted therapy after an interval is completely valid and often very clinically beneficial.
We started out with single-agent BRAF inhibitors. Then, we studied resistance to those agents and most of those studies really indicated that the majority of patients were developing resistance due to mechanisms that reactivate the same MAP kinase pathway that BRAF is on. It really provided a rationale for combining BRAF inhibitors with MEK inhibitors.
What's interesting is the available data about acquired resistance to BRAF/MEK combination, again suggests that most often we're still seeing reactivation of that same MAP kinase pathway that even though some of those mechanisms were overcome and led by MEK inhibitors, they are not necessarily overcome in patients.
Therefore, we are evaluating other ways to target the MAP kinase pathway. In particular, if you start at the level of RAF, RAF activates MEK, which activates AURKC. We now have ERK inhibitors that are in clinical trials started as single agents, but will be combined most likely with BRAF inhibitors. There are also paradox breakers for BRAF inhibitors, which are thought to do a better job of helping to block other RAF isoforms as a strategy.
In terms of other pathways, there are data that at least a subset of patients will develop resistance due to activation of the PI3K/Akt pathway. Agents targeting that pathway are a potential strategy, but in general, clinically, they have actually been quite difficult to manage the toxicities. To this point, there are no real "home runs" except for the isoform-specific inhibitors. Over the last few years, interestingly, there is sort of growing evidence that metabolic pathways may both mediate resistance and be therapeutic targets, particularly the pathway of oxidative phosphorylation, but is a key metabolic pathway. The question is: how do we target that with a good therapeutic index?
We have seen efficacy with BRAF inhibitors and with MEK inhibitors; it does make sense that if you go down further in the pathway in the cascade to ERK inhibitors, that there would be benefit there as well. There are some preclinical data that suggests that single-agent ERK inhibitors may be more effective than MEK inhibitors as a possibility. What we have seen with MEK inhibitors is the same challenge with ERK inhibitors, which is how to make them sort of clinically tolerable at doses that are effective. We do have some reason for hope, because it was interesting when we combine the BRAF inhibitors with the MEK inhibitors, it was actually better tolerated than either agent alone. Hopefully, we'll see the same with the ERK inhibitors, as well.
This is one of the true questions still to be answered in the field. We had multiple phase 1/2 clinical studies that looked very promising with combinations of BRAF, MEK, and PD-1/PD-L1 inhibitors, with [high] response rates in some of these studies. What we've seen over the last sort of 12 months are the first 2 phase 3 studies. The first one, the TRILOGY study of vemurafenib/cobimetinib with atezolizumab (Tecentriq) did meet its primary endpoint of improvement in progression-free survival [PFS]. It was approved by the FDA in July 2020.
We then sort of waited to see the next triplet study, which was dabrafenib/trametinib with or without a novel PD-1 inhibitor called spartalizumab. Interestingly, it had a hazard ratio for PFS, very similar to what was seen in TRILOGY, but it did not meet the prespecified criteria for statistical significance and, therefore, was negative.
In looking at that trial, in some ways, what was seen was unexpectedly durable responses with a dabrafenib/trametinib regimen. Overall, one of the clear things is that when we combine PD-1 with BRAF/MEK consistently, we have seen an improvement in the duration of responses. It's not actually improving response rates, but responses do appear to become more durable.
However, this does appear to be at the cost of additional toxicity. In both studies, to this point, we have not seen evidence of a benefit in overall survival. Both studies, again had a trend for improvement and I would take a "cup half full" approach. There were concerns that by using targeted therapy and immunotherapy together up front, would that actually close the door to second-line regimens and actually result in worse survival? We haven't seen any evidence of that.
However, some of the key limitations to this is that it's relatively short follow-up in comparison to most of the regimens we talked about. The most critical limitation is that neither of these studies included an immunotherapy control arm. Even for our BRAF-mutant patients, we often use immunotherapy in the first-line setting, because immunotherapy can be very effective in these patients. The challenge and trying to understand where triplets will fit into the therapeutic landscape is the fact that the trials compared BRAF/MEK alone to BRAF/MEK plus PD-1, but did not compare with immunotherapy.
Based on the shorter follow-up that we've had, and the prevalent use of immunotherapy with much longer outcome data, it is really hard to see how those trials, at this point, will have significant impact on practice in the frontline setting. There are lots of open questions about whether the triplets can have particular benefit in challenging situations. In particular, we still don't have an established standard of care for patients who've progressed on PD-1 [alone or with CTLA-4 inhibition]. We do know in that setting that BRAF/MEK can be effective, but it tends to be much less effective than in the frontline setting.
We know that BRAF/MEK can achieve responses in patients with brain metastases, but they're much shorter than what we see in other parts of the body. Will adding PD-1 improve outcomes there? We have a group of patients with a very high burden of disease that generally stand out in trials, and really none of our therapies are very effective in those patients. Is that the group where we're going to need to combine targeted therapy and immunotherapy together? That remains a good, but open question.
I don't know that I have definitive data in terms of the overlap of those two. In many cancers, PD-L1 positivity is really required to have clinical benefit from anti–PD-1. That's not the case in melanoma. The reality is, we don't use PD-L1 testing in general to decide whether a PD-1 inhibitor is going to be effective or not. There are some data specifically from the CheckMate-067 study that suggests that PD-L1 expression may help to differentiate between patients who will get the same benefit from single-agent PD-1 versus the nivolumab (Opdivo)/ipilimumab (Yervoy) combination versus those who won't. But, in terms of trying to use that to decide between immunotherapy and targeted therapy, there is not as much data.
There are some retrospective studies that have looked at some of these markers in patients treated with targeted therapy. There is evidence that if you see CD8+ T cells infiltrating your tumor, not only is that a predictor of better outcomes with anti–PD-1, it's actually also predictor of better outcomes with targeted therapy as well. It's unclear how those immune markers act as predictive markers.
Brain metastasis is a really big problem in this disease. We have very promising data in patients with asymptomatic brain metastases, even in setting up a reputation that combination immunotherapy can be very safe and effective in these patients. However, we know in patients who need steroids, that those agents actually don't work very well at all. We know that our BRAF/MEK combinations have high response rates, but the responses are generally of very short duration. Trying to understand how to improve outcomes with brain metastases remains a huge challenge for us.
Building upon successes that we've seen in breast cancer and lung cancer, what's now starting to be evaluated are BRAF inhibitors that cross the blood–brain barrier very efficiently. We have seen this really be a game changer for brain metastasis in HER2-positive breast cancer and in ALK- or EGFR-positive lung cancer. The question is: Will that have a similar impact in managing patients with melanoma with brain metastases in the setting of a BRAF mutation?
We focus a lot on the role of targeted therapy, certainly in the patients with stage IV or unresectable, stage III disease. We know that both targeted therapies and immunotherapies are very effective and safe in both the adjuvant and neoadjuvant setting. What we're starting to see now in our practice, though, is patients presenting with stage IV disease after previously receiving either targeted therapy or immunotherapy in the adjuvant setting. There are a lot of questions. Are those patients different than the patients we have been treating for stage IV disease in the past? If so, what did those previous treatments mean, in terms of their safety and efficacy for our standard therapies in frontline setting for stage IV disease?