Figure 3: Significantly mutated genes (SMGs) exclusive to BRAF melanomas have implications for immunotherapy, and secondary drivers further segregate with V600E and V600K hotspot mutations.

a) In BRAF melanomas, but not non-BRAF melanomas, mutations in MECOM and/or BMP5 were associated with clinical benefit to immunotherapy, as assessed by RECIST criteria. Additionally, when restricting to MECOM/BMP5-mutated melanomas, BRAF melanomas are associated with significantly better clinical benefit compared to non-BRAF melanomas (Fisher’s exact test, p = 0.02, two-sided). b) Survival curves between MECOM/BMP5-mutant and wild-type tumors in (top) all immunotherapy treated tumors (n = 297), (middle) BRAF immunotherapy treated tumors (n = 109), and (bottom) non-BRAF immunotherapy treated tumors (n = 188). c) Overlap of BRAF-mutant, BRAF V600E and BRAF V600K SMGs showed that roughly 2/3 of both the V600E and V600K SMGs were also identified through the BRAF-mutant mutational significance analysis. However, only 16% (7/44) and 32% (7/22) of the V600E and V600K SMGs overlapped with each other, respectively. d) Despite V600K tumors experiencing a two-fold enrichment of nonsynonymous mutational load, some BRAF V600E cancer gene (CGC, OncoKB) SMGs are altered in a similar or greater proportion of samples (left; p > 0.05 adjusted for mutational load between subtypes, χ², two-sided). Conversely, some BRAFV600K cancer gene SMGs are altered in more than double the proportion of samples (right; p < 1.85 x 10⁻³ adjusted for mutational load between subtypes χ², two-sided). In a) and d) an asterisk denotes a Mann-Whitney U p-value < 0.05. OS, overall survival; PFS, progression free survival.