Implementation of multigene panel testing for breast and ovarian cancer in South Africa: A step towards excellence in oncology for the public sector

Translation of genomic knowledge into public health benefits requires the implementation of evidence-based recommendations in clinical practice. In this study, we moved beyond BRCA1/2 susceptibility testing in breast and ovarian cancer patients to explore the application of pharmacogenetics across multiple genes participating in homologous recombination DNA damage repair. This involved the utilisation of next-generation sequencing (NGS) at the intersection of research and service delivery for development of a comprehensive genetic testing platform in South Africa. Lack of international consensus regarding risk categorization of established cancer susceptibility genes and the level of evidence required for prediction of drug response supported the development of a central database to facilitate clinical interpretation. Here we demonstrate the value of this approach using NGS to 1) determine the variant spectrum applicable to targeted therapy and implementation of prevention strategies using the 15-gene Oncomine™ BRCA Expanded Panel, and 2) searched for novel and known pathogenic variants in uninformative cases using whole exome sequencing (WES). Targeted NGS performed as a routine clinical service in 414 South African breast and/or ovarian cancer patients resulted in the detection of 48 actionable variants among 319 (15%) cases. BRCA1/2-associated cancers were identified in 70.8% of patients (34/48, including two double-heterozygotes), with the majority (35.3%, 12/34) representing known South African founder variants. Detection of actionable variants in established non-BRCA1/2 risk genes contributed 29% to the total percentage (14/48), distributed amongst ATM, CHEK2, BARD1, BRIP1, PALB2 and TP53. Experimental WES using a virtually constructed multi-cancer NGS panel in 16 genetically unresolved cases (and four controls) revealed novel protein truncating variants in the basal cell carcinoma gene PTCH1 (c.4187delG) and the signal transmission and transduction gene KIT (c.930delA) involved in crucial cellular processes. Based on these findings, the most cost-effective approach would be to perform BRCA1/2 founder variant testing at referral, followed by targeted multigene panel testing if clinically indicated and addition of WES in unresolved cases. This inventive step provides a constant flow of new knowledge into the diagnostic platform via a uniquely South African pathology-supported genetic approach implemented for the first time in this context to integrate research with service delivery.