Combining Polygenic Risk Score (PRS) with MRI to Improve Prostate Cancer Detection
Multi-parametric MRI (mpMRI) is a key imaging tool for identifying clinically significant prostate cancer (defined as Gleason score≥7). However, its sensitivity and specificity are not perfect. Prostate cancer is highly heritable, and polygenic risk scores (PRS) based on germline genetic variants can stratify risk, though historically PRS hasn’t distinguished aggressive from nonaggressive cancers well.
This study evaluated whether combining a 400-variant multi-ancestry PRS with mpMRI improves the identification of clinically significant prostate cancer.
Data from over 24,000 men were analyzed. The results of over 1200 men who underwent mpMRI (2015–2021) were included.
Men in the top PRS quartile were significantly more likely to undergo mpMRI and have clinically significant cancer on biopsy. Among those with positive mpMRI (PI-RADS 3–5), those in the top PRS quartile had a higher cancer detection rate (65.8% vs. 49.1%). Even among those with negative mpMRI, the top PRS group had more clinically significant cancers.
A proposed biopsy strategy using PRS could reduce missed cancers, but could also slightly increase unnecessary biopsies.
A 451-variant PRS (includes PSA-linked variants) showed stronger association with undergoing mpMRI but was less specific for clinically significant cancer, while the 400-variant PRS that eliminates variants that govern baseline PSA production independent of prostate cancer was better at balancing cancer detection and avoiding overdiagnosis. PRS may enhance prostate cancer diagnosis by identifying men at higher genetic risk who may benefit from closer monitoring, targeted mpMRI, or biopsy. This could be particularly helpful in managing PI-RADS 3 lesions, for which clinical consensus is lacking.
This study was not a randomized screening study; participants may not represent the general population. The main findings of this study are based on White men, therefore, applicability to other groups needs further validation.
This study provides initial support for using the 400-variant multi-ancestry PRS alongside mpMRI to better identify clinically significant prostate cancer. Prospective trials are needed to confirm its utility and assess its impact on outcomes.
JNCI Cancer Spectr. 2024 Feb PMID: 38429995
