Class matters: Sensitivity of BRAF-mutant melanoma to MAPK inhibition

Summary

Mutations in BRAF outside of the 600 codon (BRAF non-V600) occur across cancer types, including in 3-5% of melanomas. The optimal treatment strategies are not clear, but could include MEK inhibitors based on pre-clinical studies. Combining BRAF and MEK inhibitors in this population may provide additional benefit.

In this issue of Clinical Cancer Research, Dankner and colleagues report in vitro, in vivo, and clinical confirmation that BRAF L597 mutations respond to dual BRAF and MEK inhibition (BRAFi and MEKi) (1).

The advent of effective molecular-targeted therapies has been a major success story for a subset of cancer patients. Precisely targeted small molecule inhibitors can inhibit constitutive signaling caused by various genomic alterations and lead to dramatic regressions in tumors harboring these driver mutations. Over the past few years, however, it has become abundantly clear that targeted therapy efficacy is far more complex than simply matching the right drug to the right mutated gene. BRAF, perhaps more than any other gene, typifies this complexity.

A useful nomenclature subdividing the various BRAF mutations has been proposed recently (2). Class I BRAF mutations, essentially limited to V600, are the most common activating BRAF mutation in melanoma (65.9%) enable RAS-independent, monomeric signaling. BRAFi produces clinical benefits in patients harboring these mutations in melanoma, non-small cell lung cancer, thyroid cancer, and hematologic conditions (Erdheim Chester Disease, hairy cell leukemia) (3). These responses can be further augmented by the addition of MEKi, which in melanoma improves response and survival. By contrast, BRAF V600 mutated colon cancer rarely benefits from these approaches, demonstrating the importance of upstream inputs and parallel signaling networks (EGFR signaling, in the case of colon cancer), with variable benefit seen in less common cancers. Class II BRAF mutations are characterized as non-V600, are less common in melanoma (11.4%), and are generally less activating than BRAF V600 alterations. These mutations also typically signal in a RAS-independent fashion, but do so as dimers. These class II mutations can be further subdivided into class IIa and IIb, those found in the activation segment (L597 and K601) and the glycine rich region (G466 and G469) of the kinase, respectively. Various pre-clinical studies and case reports have demonstrated sensitivity to MEKi, particularly for the class IIa mutations.(4, 5) Class III mutations (N581, D594) are RAS dependent, have low or absent kinase activity, and cooperate with either concurrent RAS or NF1 mutations (in melanoma), or upstream receptor tyrosine kinase (RTK) mediated signaling (in most epithelial tumors).(2) The most appropriate therapeutic strategy for the class III mutants is unknown but could include ERK inhibition or appropriate, context-specific RTK inhibition. Adding to this complexity are studies suggesting that many BRAF V600 wild type cells (particularly NRAS mutations, and potentially even BRAF non-V600 mutations) experience paradoxical MAPK activation when exposed to BRAFi monotherapy. In all, although melanomas harboring BRAF V600 mutations have a vetted kinase inhibition strategy, tumors harboring class II mutations have a less clear kinase inhibitor approach in the clinic.

Class II BRAF mutations, the topic of this manuscript, have been the subject of several tantalizing clinical case reports in melanoma patients. In these studies, in one case supported by in vitro confirmation, individual examples of remarkable efficacy were demonstrated by MEKi, inducing deep and durable responses (4, 5). Several studies are attempting to study MEKi in a prospective fashion in melanoma (NCT02296112) or across tumor types (NCI-MATCH study, Arm R, NCT02465060) in class II and class III BRAF mutations. However, given the relative infrequency of these mutations (<5% of melanomas), no systematic studies have yet been published assessing the overall benefit for either BRAFi or MEKi for non-V600 BRAF mutations.

This study by Dankner et al (1) suggests that, similar to class I BRAF mutations, combined BRAFi and MEKi may be a superior approach for class II mutations (Figure). Specifically, tumors harboring class IIa mutations may benefit more from the combination than tumors harboring class IIb mutations. BRAFi alone only provided short-term in vitro efficacy with rapid rebound in ERK signaling whereas MEKi as a single-agent had efficacy in cell lines harboring class I or class II mutations. However, in a variety of cell lines and in vivo models, dual therapeutic inhibition appears superior to either BRAFi or MEKi alone. Class IIb mutation models showed more intrinsic resistance to BRAFi (interestingly, with the exception of the novel BRAFi encorafenib) but also sensitivity to combination therapy. These results were similar to BRAF wild-type tumors, although most of the class IIb models in this study harbored concurrent RAS mutations, which may have influenced these data. In addition, the authors report two patients with BRAF L597 mutant melanoma who experienced dramatic although fairly transient responses to dabrafenib and trametinib. In summary, BRAF class I and class IIa mutations models had similar responses to single-agent BRAFi and MEKi, and the BRAFi+MEKi combination. In contrast, cells harboring WT BRAF or class IIb BRAF mutations behaved more similarly in their response to targeted MAPK inhibition. Importantly, BRAFi alone did not induce paradoxical activation of MAPK signaling in type II BRAF mutations, and combined BRAF and MEK inhibition did not trigger this phenomenon irrespective of BRAF status.

Figure. Melanoma cells harboring class IIa BRAF mutations are most sensitive to combination BRAF plus MEK inhibition.

The relative effectiveness of BRAF inhibitor (BRAFi), MEK inhibitor (MEKi), and combination BRAFi+MEKi on inhibition of cell growth and downstream signaling in melanoma harboring wild-type BRAF and class I, IIa, and IIb activating BRAF mutations. A “sensitivity” scale is shown to the right with dark red being the least sensitive to the kinase inhibitor and blue being the most sensitive.

What are the limitations of this study? First, on a clinical level, both patients who responded to BRAFi + MEKi could have also responded to single-agent MEKi (as has been previously reported). Second, both of these patients harbored BRAF L597S mutations, potentially limiting the generalizability to other class II mutations. MEKi monotherapy has primarily demonstrated efficacy with L597 and K601 mutations in prior studies as well (4, 5). Third, MEKi has impressive activity across a wide range of BRAF wild type melanoma cell lines with much less activity in patients (likely <10% response rate), raising questions about the translatability of the in vitro findings to inform results in patients.

How does this study impact clinical care for the 2-3% of melanoma patients with non-V600 BRAF mutations? First-line treatment for these patients remains unchanged, as immune checkpoint inhibitors have demonstrated excellent clinical activity in patients irrespective of BRAF mutation status. Further, in the absence of robust, prospective clinical data for other available agents, entry into clinical trials, potentially evaluating novel targeted therapy strategies, is a second preferred treatment option. However, after immunotherapy failure and in a setting where no appropriate trial exists, one could make a convincing argument that MAPKi therapy should be considered for melanoma patients harboring L597 and K601 (class IIa) mutations. Prior to this manuscript, MEKi would have been the clear therapeutic choice, but one could make the case now that combined BRAFi+MEKi should be the preferred strategy (perhaps with encorafenib and binimetinib). Combination therapy has superior pre-clinical efficacy in this setting and, in patients, a more favorably toxicity profile. Thus, the combination approach is likely at least equivalent from a safety and efficacy standpoint. Treatment for class IIb and especially class III mutations in melanoma, and various non-V600 BRAF mutations across other tumors remains less clear, as it remains difficult to extrapolate these data to such distinct mutation- and tissue-specific contexts.

In conclusion, this study demonstrated enhanced benefits of combined BRAFi+MAPKi in cells with class II BRAF mutations. Although this is not a common clinical situation and is not likely to be studied in a randomized trial, we would argue that BRAFi+ MEKi is an appropriate treatment option for this population. Further, no paradoxical activation of MAPK signaling was noted with the combination irrespective of BRAF status. Further study is needed to optimize treatments for other cancers and other BRAF mutations.