Accurate detection of Copy Number Variants (CNV) in homologous recombination repair (HRR) related genes is essential for HRR deficiency diagnostics and treatment administration such as PARP inhibitors. CNV detection from high-throughput sequencing (HTS) is expected to complete simultaneous comprehensive genetic markers evaluation, reduce cost and turnaround time and to increase sensitivity. It is still challenging to detect somatic CNV from small targeted panels that are widespread in clinical settings. The current aim: To evaluate clinical utility of somatic CNV detection from amplicon-based HTS panels.
High-throughput sequencing of 220 tumor samples (breast, ovarian, pancreatic and prostate) was performed with an AmpliSeq panel covering ATM, BRCA1, BRCA2 and CHEK2 genes (409 amplicons, 36kbs). Data from 160 cases passed quality control were analyzed for CNV. Following metrics were utilized for CNV detection: SNP allelic imbalance (more than 10% from hetero- or homozygosity), by-pool coverage drop (more than 20% from baseline), score of coverage consistency. For accurate SNP allelic imbalance estimation we implemented amplicon-based variant allele fraction (VAF) calculation ensuring reduction of amplicon dropout errors and PCR-enrichment artifacts. This allowed to increase concordance of VAF of detected SNPs.
Over analyzed 160 samples, 7 samples displayed SNP allelic imbalance concordant with consistent coverage drop in BRCA1 (6 cases, 16% positive for BRCA1/2 mutations) and BRCA2 (1 case, 0% positive for BRCA1/2 mutations) genes. These samples were considered as putative for CNV. No cases with putative CNV in ATM or CHEK2, or with CNV in more than one gene were found. The observed number of positive samples corresponds to the level expected from literature data.
Integrative analysis of allelic imbalance and amplicon coverage from targeted high-throughput sequencing with small panels allows to detect samples putative for CNV for further validation with orthogonal methods. This allows to increase information yield from small gene panel sequencing and may lead for PARP inhibitors indication for additional 4% of patients with breast, ovariant, pancreatic and prostate cancer.
The authors.
Has not received any funding.
All authors have declared no conflicts of interest.