Assessment of a Polygenic Risk Score in Screening for Prostate Cancer

Abstract

Background: The incidence of prostate cancer (PrCa) is increasing. Screening by assay of prostate-specific antigen (PSA) has a high false-positive rate. Genome wide association studies have identified common germline variants, which can be used to calculate a polygenic risk score (PRS) associated with PrCa risk. Methods: The BARCODE1 study recruited persons aged 55 to 69 yrs via primary care in the UK. PRS were derived from 130 PrCa risk variants in germline DNA extracted from saliva. Participants with a PRS >90th centile were invited for PrCa screening using multiparametric Magnetic Resonance Imaging (MRI) and transperineal biopsy, irrespective of PSA result. Results: Of 40,292 persons invited to participate, 8,953 (22.2%) expressed interest in participating and 6,393 had their PRS calculated, of whom 745 (11.7%) had a PRS >90th centile and were invited for screening. Of these 745 participants, 468 underwent MRI and prostate biopsy; PrCa was detected in 187 (40.0%) of them. Median age at diagnosis was 64 yrs (range 57 to 73 yrs). Using NCCN criteria (2023), 103 (55.1%) cancers were of intermediate or high risk and so required treatment; 73 (70.9%) of these cancers would not have been detected using the UK PrCa diagnostic pathway. Of the 187 cancers, 40 (21.4%) were ‘Intermediate Unfavorable’/’High’/’Very High-Risk’. Conclusions: Risk stratification by PRS of 6,393 persons led to the detection of PrCa requiring clinical management in 103 participants, of whom 73 (70.9%) would have been missed using the standard diagnostic pathway used in the UK.