RE: Differential influence of vemurafenib and dabrafenib on patients' lymphocytes despite similar clinical efficacy in melanoma

The recent article by Schilling et al. relates to an issue of increasing importance to the melanoma community, as it ponders rational strategies for combination or sequential application of the modalities of BRAF inhibitor (BRAFi) therapy and immunotherapy [1]. In their analysis of 277 vemurafenib- and 65 dabrafenib-treated patients, the authors report a reduction in peripheral blood (PB) lymphocyte counts in the vemurafenib, but not dabrafenib cohorts. Flow cytometric analysis suggest that CD4+ (and to a lesser extent, CD8+) T cells were preferentially affected compared with B cells or NK cells. Multiplex cytokine assays on the CD4+ subset of vemurafenib-treated patients revealed reduced levels of interferon-γ/interleukin (IL)-9 but unchanged levels of IL-2/4/10/17A. The authors concluded that vemurafenib, but not dabrafenib, adversely affects CD4+ T cells both numerically and functionally.

There are several methodological and statistical issues that merit discussion in this work.

differential effects

Inferences should be tempered by the retrospective nature of the analysis, wherein patient selection was not based on prespecified criteria which could have resulted in major imbalances that would have direct impacts on lymphocyte counts. First, prior therapies (not reported by the authors) could have included IL-2 or ipilimumab which would have had a long-lasting impact on the level and function of circulating lymphocytes. Secondly, study population likely represents patients with a high tumor burden given the choice to initiate BRAFi outside auspices of a clinical trial. Tumor burden is associated with increased host immune suppression and information relevant to tumor burden (initial lactate dehydrogenase level, absence/presence of brain metastases) was not reported.

lymphocyte numerical and functional assessment

Effects of BRAFi on T cells were assessed by comparing pretreatment values to post-treatment nadirs obtained between 0 and 12 weeks, and these values were not standardized between patients. The relatively large difference between the number of patients treated with each agent (vemurafenib: 277, dabrafenib: 65), compared with the relatively small difference in values (vemurafenib: −24.3%, dabrafenib: +1.2%) from pretreatment median baselines (vemurafenib: 1.27, dabrafenib: 1.4) raises the question of whether the reported results were numerical fluctuations that gained statistical significance but were of no clinical or biological significance given the unequal treatment arms.

clinical implications

Schilling et al. appropriately note that both patient cohorts had similar progression-free survival (PFS) and overall survival (OS). Hence, dividing PB lymphocyte (presumably total lymphocytes rather than CD4+ or CD8+ or NK-cell subsets although this is not explicitly stated) into quartiles and analyzing PFS/OS is unhelpful since statistically significant differences may not necessarily reflect clinically meaningful relationships, unless data reflecting the longitudinal lymphocyte response are provided.

BRAF-mutated melanoma is intrinsically immunogenic [2]. BRAFi therapy enhances melanoma antigen expression (increasing melanoma immunogenicity) while upregulating expression of T-cell PD1/TIM3 and tumor PD-L1 (augmenting immunosuppressive tumor microenvironment) [2]. Unlike MEK inhibition, BRAFi has no documented effects on T-cell function [3, 4]. These observations provide strong preclinical rationale for synergy between BRAFi and checkpoint inhibition and several trials of combinations are underway.

In the absence of a concurrent, randomized, unbiased enrollment of subjects to a clinical trial with prespecified criteria that allow an objective assessment, the conclusion that different BRAFi exert differential impact on the immune system is observational and the argument for causality is not supported by the evidence presented. Although useful in understanding the potential effects of BRAFi therapy on lymphocyte function, the analyses presented in this work should not preclude further evaluation of synergy between BRAFi and immunomodulatory agents.

disclosure

JMK Scientific advisory board, remunerated (GSK, BMS, Ziopharm, Vical, Merck, Novartis, Celgene) (each <5000 USD). DD, TC and HAT have no disclosures to report.