Abstract
BACKGROUND: B-Raf V600E mutations account for about half of all skin cutaneous melanoma cases, and patients with this mutation are sensitive to BRAF inhibitors. However, aberrations in other genes in the MAPK/ERK pathway may cascade a similar effect as B-Raf V600E mutations, rendering those patients sensitive to BRAF inhibitors. We rationalized that defining a signature based on B-Raf pathway activity may be more informative for prognosis and drug sensitivity prediction than a binary indicator such as mutation status. METHODS: In this study, we defined a B-Raf signature score using RNA-seq data from TCGA. A higher score is shown to not only predict B-Raf mutation status, but also predict other aberrations that could similarly activate the MAPK/ERK pathway, such as B-Raf amplification, RAS mutation, and EGFR amplification. RESULTS: We showed that patients dichotomized by the median B-Raf score is more significantly stratified than by other metrics of measuring B-Raf aberration, such as mutation status, gene expression, and protein expression. We also demonstrated that high B-Raf score predicts higher sensitivity to B-Raf inhibitors SB590885 and PLX4720, as expected, but also correlated with sensitivity to drugs targeting other relevant oncogenic pathways. CONCLUSION: The BRAF signature may better help guide targeted therapy for melanoma, and such a framework can be applied to other cancers and mutations to provide more information than mutation status alone.