Decoding mutational signatures in breast cancer: Insights from a multi-cohort study

Abstract

Purpose

Diagnosis and treatment decisions of hormonal breast cancers (BC) are now guided by genomic mutations determination, combined into mutational signatures, and provide insight into the patients’ genomic landscape. This work aims to compare genomic data and signatures extracted from tissue samples collected in the CICLADES study to existing cohorts. Ultimately, the goal is to prove the accuracy of smaller cohorts and provide new relevant data.

Materials and methods

DNA from patients of the CICLADES cohort was extracted, sequenced, and custom filtering was applied to the resulting files. Genomic data was pulled from 6 BC cohorts available on cBioPortal.com. In total, 2303 samples were analyzed. Mutational signatures were extracted and matched to known signatures of the Catalogue of Somatic Mutations in Cancer (COSMIC). Tumor Mutation Burden (TMB) and hypermutation were estimated and compared between samples.

Results

PIK3CA and TP53 represented the two genes highly mutated across all cohorts. TMB was similar between the CICLADES and CBSM groups, however the MSKCC population showed a significantly higher TMB than both. Nine signatures were extracted, with recurring Single Base Substitutions (SBS) signatures like SBS1, SBS2 and SBS5. The presence of APOBEC-specific signatures was concordant with cohorts presenting APOBEC enrichment. The mean number of mutations was significantly higher in enriched samples for each analyzed cohort.

Conclusion

The use of comprehensive genomic profiling provided accurate evaluation of the TMB and extraction of signatures consistent with published literature. The genomic analysis of the tissue samples of the CICLADES cohort brings new and relevant data, comparable to results found in bigger cohorts.