BRAF Inhibition: Bridge or Boost to T-cell Therapy?

In this issue of Clinical Cancer Research, Atay and colleagues explore the relationship between mannose-6-phosphate receptor (M6PR) and inhibitors of the BRAF/MEK pathway in human melanoma (1). With a series of in vitro experiments, they demonstrate that M6PR is upregulated by PLX4720, a BRAF inhibitor, in a dose-dependent fashion in both BRAF-sensitive and BRAF-resistant melanoma. Treatment was associated with an increased susceptibility to lysis by HLA-matched tumor-infiltrating lymphocytes (TIL). Treated cell lines also demonstrated increased intracellular GranzymeB, associated with M6PR overexpression and abrogated in M6PR knockout lines. In models utilizing patient-derived xenografts, a combination of PLX4720 and TILs significantly slowed the growth of tumors when compared with either agent alone or no treatment.

While patients with tumors bearing the BRAF^V600E/K mutation may experience initial clinical benefit, most will develop resistance to these targeted therapies. Previous observations in murine models of melanoma led the group to hypothesize that a combination of BRAF-targeted therapy with adoptive transfer of TILs may be a feasible human clinical alternative. Using this strategy, this group demonstrated a 63% 12-week and 35% 12-month objective response rate (6 of 16 treated patients).

These interesting experiments by Atay and colleagues reflect upon an earlier era, before the clinical translation of immune checkpoint modulation radically changed the course of patients with metastatic melanoma. The role of BRAF/MEK inhibition in an overall treatment strategy has been altered, and it is necessary to place these clinical findings within that context. The initial trials leading to the approval of vemurafenib, a BRAF inhibitor, and ipilimumab, a mAb targeting CTL antigen 4 (CTLA-4), were being conducted concurrently. The first new drug for metastatic melanoma since IL2 in 1998, ipilimumab was capable of mediating objective responses in 7% of patients (2). Vemurafenib was approved later the same year after demonstrating significant improvements in overall and progression free-survival (HR 0.37 and 0.26, respectively) when compared with dacarbazine (3).

While continuing to monitor progression and survival in those clinical trials, mAbs targeting programmed death receptor 1 (PD-1) were developed. Nivolumab and pembrolizumab were both initially approved for patients with melanoma refractory to vemurafenib and/or ipilimumab. However, the difference between the strategies became clear over time as the survival curves matured. The checkpoint modulators were capable of “raising the tail” of the survival curves, signifying long-term durable survival benefit for patients, whereas BRAF/MEK–targeted strategies demonstrated transient responses. Pembrolizumab, nivolumab, and nivolumab in combination with ipilimumab are now approved as first-line treatment regimens for patients with metastatic melanoma (4, 5).

The translational challenge now is to develop meaningful therapeutic options for patients refractory to checkpoint blockade, and investigators have focused on the development of novel immunotherapeutics and combinations of reagents with known reactivity in melanoma. Unfortunately, attempts to combine vemurafenib with other approved treatments have met with little success. Two independent trials of vemurafenib with high-dose IL2 were stopped early for poor accrual in the era of checkpoint blockade and adding ipilimumab to vemurafenib induced dose-limiting hepatotoxicity.

An experimental strategy also capable of “raising the tail” is the adoptive transfer of TILs derived from freshly resected melanoma metastases. In trials with long-term follow-up, objective clinical responses could be seen in approximately 50% of patients, including durable, likely curative, complete responses in approximately 25% of patients. Only 2 patients (of 46) with complete responses have recurred, and 5- and 10-year overall survival approached approximately 30% (6, 7). The current study by Atay and colleagues combining BRAF inhibition with TILs appears to be comparable with these results. Similarly, our published efforts combining vemurafenib with adoptive transfer of TILs reported a 64% objective response rate (7 of 11 patients) with two complete responses (8). Even in a precheckpoint blockade era, a randomized trial would have been necessary to identify superiority of the combination compared with TIL alone. However, the majority of patients in each of these trials were naïve to checkpoint therapy hindering interpretation in today's landscape (Table 1; refs. 1, 6–11).

Checkpoint-refractory tumors and the lymphocytes that they harbor may be qualitatively or quantitatively different. In much the same way that BRAF inhibition may exert a selection pressure to create treatment-resistant clones, the application of checkpoint inhibitors results in immunoedited tumors comprised of clones without known antigen recognition or lacking important antigen-processing molecules. In a small sample of patients undergoing serial biopsy while receiving pembrolizumab, the tumors of nonresponding patients had lower CD8⁺ T-cell density at all time points (12). It is also possible that intratumoral T-cell diversity is also affected by anti–PD-1 treatment (13).

Exploration of adoptive transfer as a salvage therapy after checkpoint blockade has demonstrated that objective responses are possible and provide meaningful clinical outcomes for patients with limited options. While the overall response rate is likely to be lower than historical efforts, the ability to mediate regression suggests a component of adoptive cell transfer acts by mechanisms separate from checkpoint inhibition. The potential value of Atay and colleagues' findings may lie in the concept of increasing responsiveness to T-cell–mediated cytolysis via upregulation of M6PR regardless of the treatment mechanism employed.

Our initial efforts in broadening the applicability of TILs for cancer therapy were based in transitioning the concept from melanoma, where conventionally grown TILs could work, to more common cancers, where it did not. Careful retrospective analysis of responding patients identified TIL clonotypes that recognized nonsynonymous mutations in random intracellular proteins (14, 15). Identifying and reinfusing TILs of the latter category has been a successful strategy when applied to selected epithelial cancers, including a complete regression of metastatic breast cancer (16, 17). The intratumoral milieu of checkpoint-refractory melanoma may be similar to that of epithelial cancers; selecting TILs on the basis of neoantigen reactivity may improve response rates over current efforts. Further refinement of this neoantigen selection concept may involve the introduction of mutation-specific T-cell receptors into a patient's peripheral blood to provide a more enriched, perhaps less differentiated, infusion product.

The development of an autologous cell product, however, requires an investment of time, which may be limited for this patient population. The current treatment options for patients diagnosed with metastatic melanoma should be mapped carefully with strategic timing to harness the benefit of approved therapeutics and optimize access to experimental clinical trials. The high response rates of BRAF/MEK inhibitors make them an attractive bridge therapy for patients navigating entry and registration into early-phase clinical trials.

No potential conflicts of interest were disclosed.

This work was funded by the Center for Cancer Research, the intramural program of the National Cancer Institute.