Résumé
Late bladder toxicity is a concern for patients receiving prostate cancer radiotherapy and negatively affects survivors. Few risk factors are known beyond the radiation dose and volume of bladder exposed. A polygenic risk score (PRS) could identify susceptible patients.
A PRS was built using genome-wide association results from the Radiogenomics Consortium (N = 3,988) and then tested in the prospective REQUITE and URWCI studies (N = 2,034). The primary outcome was time to patient-reported gross [grade ≥2, (≥G2)] hematuria, analyzed using Cox proportional hazards regression. Secondary outcomes were ≥G2 urinary retention and frequency. The PRS was externally validated for clinically diagnosed irradiation cystitis in the UK Biobank (N = 8,430). A gene-burden test evaluated rare coding variants.
A 115-variant PRS was associated with a significantly increased risk of ≥G2 hematuria [hazard ratio (HR) per SD = 1.22; P = 0.009] as well as urinary retention (HR per SD = 1.18; P = 0.016) and frequency (HR per SD = 1.14; P = 0.036). When binarized, men in the upper decile (PRShigh) had a >2-fold increased risk of hematuria after adjusting for clinical risk factors [HR = 2.12; P = 0.002; Harrel's concordance index = 0.71 (95% confidence interval, 0.65-0.76)]. A similar effect size was seen in the UK Biobank for clinically diagnosed irradiation cystitis [odds ratio (OR) = 2.15; P = 0.026]. The burden test identified BOD1L1 as a putative novel radiosensitivity gene.
This PRS identifies susceptible patients and could guide the selection of those needing reoptimized treatment plans that spare the bladder beyond currently recommended constraints.
PRS-guided treatment planning in radiation oncology could lower the incidence of clinically relevant bladder toxicity and reduce the impact of this outcome on prostate cancer survivors.