Combined genetic score better distinguishes prostate cancer risk
Among 6,590 men, 20% without harmful gene variants moved to higher risk groups and 60% carrying them to lower groups, compared with carrier-only models.
Our read
Biomarkers- Studied:
- Cohort study of 218,484 UK Biobank men without prostate cancer, against carrier-only models (Dose-response, independent replication)
- What changed:
- The combined score better distinguished future prostate cancer cases than carrier-only models: concordance index 0.69 versus 0.52
- The catch:
- Health effects of score-guided screening were not measured
Editor of record: Boris Djordjevic · AgingResearch.News
In an observational study, researchers developed a prostate cancer risk score using 218,484 UK Biobank men without prostate cancer at recruitment. Their median age was 58.6 years, and median follow-up was 15.5 years. The score combined harmful variants in specific genes with a measure of inherited risk spread across many genetic variants. Researchers compared it with models that classified risk based on whether men carried harmful variants.
The combined score better distinguished men who developed prostate cancer from those who did not. Its concordance index, a measure of that distinction, was 0.69 versus 0.52 for carrier-only models. Researchers tested an adapted version in 6,590 men from a healthcare biobank. Their median age was 60.2 years, and median follow-up was 8.6 years. It also performed better there, scoring 0.64 versus 0.54.
Why it matters
Estimating prostate cancer risk by age 75 bears on how screening might be tailored as men age. The authors suggest the framework may enable more precise risk-based screening.
Caveats
This observational study tested prediction, not whether score-guided screening improves health. It excluded men already diagnosed with prostate cancer, and most participants were of European ancestry.
- Human genetics
- Resources: datasets, atlases and tools
- Prostate cancer
- GenProb-PCa
- Polygenic risk score
- Humans
The paper
Endeavor Health
Journal of Medical Genetics, Jul 30, 2026, Open access
Comments
Be the first to comment.