Poster Presentation Clinical Oncology Society of Australia Annual Scientific Meeting 2023

Scrutinizing the whole genome for potential genes to design an African exclusive prostate cancer germline panel (#203)

Kazzem Gheybi 1 , Vanessa Hayes 1 , Riana Bornman 2 , Weerachai Jaratlerdsiri 1 , Shingai Mutambirwa 3
  1. University of Sydney, Camperdown, NSW, Australia
  2. School of Health Systems and Public Health, University of Pretoria, Pretoria, South Africa
  3. Department of Urology, Sefako Mekgatho Health Sciences University, Dr. George Mukhari Academic Hospital, Medunsa, South Africa

Germline testing has recently become widespread for prostate cancer (PCa) to indicate precision treatment strategies and also provide further malignancy risk for patients and their relatives. The panels for germline testing however, are solely designed based on studies on European ancestral patients, while African ancestry is a known risk factor for the advanced disease and mortality. We have recently shown that these panels are not ideal for detection of pathogenic variants in Black South African patients and there is a significant loss of sensitivity when compared with non-African populations (5.6% vs 11-17%), which concurs with previous, yet limited, African American and west African studies. As such, a whole genome approach is required to identify African-relevant pathogenic variants that may be contributing to the associated health disparity. Here we interrogate whole genome data for 119 Black South African men diagnosed with a bias towards high-risk PCa. Of the 13.3 million single nucleotide variants (SNV) and 2.1 million Indels (insertion/deletions, <50bp) identified, we found 104 (82 SNVs and 22 Indels) known pathogenic variants in 98 genes of which only BRCA2, ATM, RAD50, CHEK2, TP53 are included in current PCa germline testing guidelines. Aware that current pathogenic variant databases have been derived from predominantly non-African patient data, after excluding for common and benign variants, we performed further functional (SIGT, PolyPhen2) and oncogenic (CGI) prediction identifying 399 potentially oncogenic variants with uncertain significance in 234 genes. Remarkably, while 94 of 119 patients (79.0%) presented with a known pathogenic variant, all patients presented with at least two potentially oncogenic variants with uncertain significance (mean=6.31). Based on the frequency of impact, we generate a 40 gene African-relevant candidate panel, recommending clinical-trial studies to determine applicability to predict PCa risk and therapeutic implication. Ultimately, we provide the first available data for novel gene candidates for inclusion in PCa germline testing panels to allow for African inclusion.