Background and objectives Genomic studies are essential for identifying mutations that may influence key aspects of breast tumours, such as susceptibility, aggressiveness, and response to treatment. There are deficient molecular and genomic data from Indian breast cancer patients. Methods mRNA from primary breast cancer samples were subjected to next-generation transcriptome (mRNA) sequencing on an Illumina platform, in duplicates and triplicates to generate 30–60 M reads/sample. PAM50, and absolute intrinsic molecular subtyping (AIMS) gene expression-based classifiers were used for intrinsic subtyping. Variants were called using, GATK, MuTect2, VarScan2, and VarDict, followed by filtering for somatic and non-synonymous changes. Germline variants were excluded using public databases. ClinVar annotations prioritised pathogenic variants, and the STRING algorithm was used for network analysis. Results A total of 207 RNA-Seq datasets from 97 breast cancer patients were analysed. There was good concordance between the immunohistochemical receptor and AIMS classification for all subtypes, but there was discordance between immunohistochemical and PAM50 subtypes within the ER-positive/HER2-positive subgroup, wherein only 38.5% (n= 5) were classified as HER2-like by gene expression classification. Variant analysis identified 145 high-confidence somatic mutations, with TP53 (n=46, 47%) and PIK3CA (n=33, 34%) being the most frequent. Additional actionable mutations in BRCA1 , BRCA2 , FGFR2 , PTEN , AKT1, and mTOR pathways were identified. At least one actionable mutation was found in 52% of patients. Fusion transcript analysis identified 91 recurrent fusions, including novel partners with ERBB2 , MED1 , and CDK12 , suggesting the possibility of unique molecular events. Interpretation and conclusions This study demonstrates that Indian breast cancer patients exhibit molecular subtypes and actionable mutations comparable to Caucasian cohorts.
Gardi et al. (Sat,) studied this question.