Unraveling BRAF alterations: molecular insights to circumvent therapeutic resistance across cancer types

Abstract

Aim: As intrinsic resistance - often driven by concurrent genomic alterations in tumor suppressor genes or oncogenes - remains a major challenge in oncology, this work aimed to comprehensively analyze BRAF somatic alterations across cancer types and identify new potential therapeutic strategies to overcome drug resistance. Methods: We conducted an extensive analysis of genomics, transcriptomics, and clinical data retrieved from public repositories, including cBioPortal. Our comprehensive analysis examined BRAF alterations [point mutations, structural variants (SVs) and copy number alteration] in more than 217,000 tumor samples across 120 distinct tumor types from primary and metastatic sites in both adult and pediatric cohorts, focusing on mutual exclusivity and co-occurrence of mutations in other oncogenes or tumor suppressors. The work also explores the association of BRAF somatic alterations with survival, clinical and molecular features. Results: Analysis of mutation frequencies across cancer types revealed that BRAFV600E represents approximately 90% of all BRAF alterations. While melanoma and thyroid carcinoma show the highest prevalence of BRAF mutations, followed by colorectal and non-small cell lung cancer in terms of absolute number of patients harboring BRAF mutations worldwide, notably high mutation frequencies were identified in rare malignancies, including hairy-cell leukemia, ganglioglioma, and serous borderline ovarian tumors. The comprehensive analysis of genomic profiling data across these tumors uncovered distinct patterns of co-occurring and mutually exclusive alterations in oncogenes and tumor suppressor genes, illuminating resistance mechanisms and suggesting novel therapeutic combinations. Conclusion: Comprehensive genomic profiling is critical for optimizing targeted therapy and overcoming drug resistance in BRAF-mutated cancers. The identification of co-occurring alterations provides opportunities for rational combination therapies, emphasizing the importance of detailed mutation profiling in developing effective treatment strategies across diverse cancer types.

Keywords: BRAF-mutated tumors; drug resistance; genomic profiling; mutation co-occurrence; mutually exclusive mutations; targeted therapy.

Figure 1

Workflow of pan-cancer analysis of BRAF alterations across tumor types. Workflow of the analysis on the genomic data sets to evaluate - across multiple cancer types - the frequency of BRAF alterations (point mutations, deletions/amplifications and structural variants), associations with molecular/clinical features and mutational co-occurrence. Tumors were selected on a mutation frequency cutoff of 5% from the TCGA and non-TCGA datasets in cBioPortal. Cohort descriptions are provided in the main text and in the Methods section.

Figure 2

Association analysis of BRAF mutations with molecular and clinical outcomes in CRC and NSCLC. (A) OS in CRC patients carrying (red line) or not (blue line) BRAF mutations on Kaplan-Meier graphs; (B) Patterns of MSI and (C) CIMP in BRAF-mutated and -unaltered CRCs; (D) OS in NSCLC patients carrying (blue line) or not (red line) BRAF mutations on Kaplan-Meier graphs; (E) Incidence of specific tumor forms and (F) TMB in BRAF-mutated and -unaltered NSCLCs. CRC: Colorectal cancer; NSCLC: non-small cell lung carcinoma; OS: overall survival; MSI: microsatellite instability; CIMP: CpG island methylator phenotype; TMB: tumor mutational burden; MSS: microsatellite stable.

Figure 3

Comparative analysis of gene co-mutation patterns in CRC patients carrying or not BRAF mutations. (A and B) Oncoprint of gene mutations (cutoff 5%) indicating co-mutations or mutual exclusivity patterns in BRAF-mutant CRC patients of cohort #1 (A) and cohort #2 (B); (C and D) Bar plot of the number of BRAF-mutant CRC patients of cohort #1 (C) and cohort #2 (D) with mutational overlap. The spots under the bar plot show the presence of mutually exclusive or co-occurrence mutations (highlighted in blue and red, respectively). On the right, for each mutated gene, the table indicates the association parameters and statistical significance; (E) OS in CRC patients carrying BRAF mutations and co-occurrent mutations in SMAD4 (magenta), PIK3CA (blue), and FBXW7 (green) on Kaplan-Meier graphs; (F and G) Box plots showing mutation count (F) and TMB (G) in CRC patients carrying co-occurrent mutations in BRAF oncogene and ATM, SMAD4, PIK3CA or FBXW7 genes. CRC: Colorectal cancer; OS: overall survival; TMB: tumor mutational burden; ATM: ataxia-telangiectasia mutated.

Figure 4

Comparative analysis of gene co-mutation patterns in NSCLC patients carrying or not BRAF mutations. (A and B) Matrix-like heatmap (oncoprint) of gene mutations (cutoff 5%) indicating co-occurring or mutual exclusivity mutational patterns in BRAF-mutant NSCLC patients of cohort #1 (A) and cohort #2 (B); (C and D) Bar plot of the number of BRAF-mutant NSCLC patients of cohort #1 (C) and cohort #2 (D) with mutational overlap. The spots under the bar plot show the presence of mutually exclusive or co-occurrence mutations (highlighted in blue and red, respectively). On the right, for each mutated gene, the table indicates the association parameters and statistical significance; (E and F) OS in NSCLC patients carrying BRAF mutation alone (green) or co-occurrence of (E) SMARCA4 (orange) or (F) STK11 (blue) mutations, on Kaplan-Meier graphs; (G-I) Box plots showing Winter (G), Ragnum (H), and Buffa (I) hypoxia scores in NSCLC patients carrying BRAF mutation alone or co-occurrent with the p53 gene. NSCLC: Non-small cell lung carcinoma; OS: overall survival.

Figure 5

Subtype-specific analysis in BRAF-mutated rare tumors. (A and B) Oncoprint displaying gene mutations (cutoff 5%; A) and mutation frequency diagrams in MAPK and RTKs for hairy-cell leukemia (B); (C and D) Oncoprint displaying gene mutations (cutoff 5%; C) and mutation frequency diagram in MAPK and RTKs for ganglioglioma (D); (E and F) Oncoprint displaying gene mutations (cutoff 5%; E) and mutation frequency diagram in MAPK and RTKs for serous borderline ovarian tumor (F). In each box, the red color intensity is proportional to the mutation frequency of the gene. RTKs: Receptor tyrosine kinases; ATM: ataxia-telangiectasia mutated.