Investigators evaluated whether germline genomic risk adds predictive value to measured prostate-specific antigen (PSA) for prostate cancer detection. They built P-CARE, a composite germline risk tool that combines a 601-variant polygenic hazard score with family history and genetic ancestry. The primary analysis used data from the Million Veteran Program (MVP).
The study team reported that men at high P-CARE risk experienced greater hazards of receiving treatment and of developing metastasis compared with men at low risk; both hazards were reported with a hazard ratio of 3.32. The authors interpreted these hazard estimates as evidence that germline genomic stratification aligns with clinically meaningful downstream events recorded in the VA cohort. Short, direct statement: high P-CARE tracked with higher clinical event rates.
The authors examined a clinically relevant subgroup: men whose baseline PSA was measured between ages 40 and 49 and whose PSA was 1–3 ng/mL. This is a near-term screening window where management decisions often rely heavily on PSA alone. Data indicated that 15-year probabilities of treatment or metastasis differed across P-CARE strata in this subgroup: 2.6% for the low-risk group, 4.5% for the intermediate group, and 10.2% for the high-risk group.
These absolute risk differences, presented by the investigators, suggest that genetically informed risk categories can meaningfully reclassify men who share the same PSA value. In practice, that could translate into different thresholds for surveillance or biopsy referral when genotype-derived risk is taken into account, though the authors did not provide prospective management algorithms in the preprint.
Beyond long-term treatment and metastasis outcomes, the cohort analysis also addressed biopsy-detected disease. The study team reported that high P-CARE risk predicted clinically significant prostate cancer on biopsy, with an odds ratio of 2.96. In contrast, the score labeled PRSPSA was reported as inversely associated with the clinical outcomes the authors examined.
The investigators framed these biopsy results as complementary evidence that germline genomic information correlates with phenotype at the time of diagnostic workup as well as with longer-term events.
The authors sought replication and applied their approach to the Prostate Cancer Prevention Trial (PCPT) as an external validation cohort. Data indicated that the patterns observed in MVP—greater event rates among higher P-CARE strata and prediction of clinically significant disease—were consistent in PCPT. The preprint presents the PCPT results as supportive replication rather than definitive proof; the manuscript has not undergone peer review.
This work is issued as a preprint and the authors explicitly note it has not been certified by peer review. The UCSD institutional review board waived ethics oversight for the reported analyses. Full summary statistics related to the MVP studies are available at dbGAP accession phs001672, as stated by the authors.
The authors conclude that germline genomic risk measured by P-CARE complements measured prostate-specific antigen and they advocate for prospective evaluation of genetically informed screening strategies. The preprint does not set out specific clinical-action thresholds, nor does it report prospective trial protocols or regulatory milestones; those details were not provided in the source.
Limitations highlighted by the publication status and available information include the absence of peer review at time of posting and the need for prospective studies to test whether using P-CARE to alter screening or biopsy decisions changes clinically important outcomes. The manuscript does not report additional operational parameters such as decision thresholds, cost-effectiveness analyses, or implementation pathways—those remain unanswered in the source and will need addressing in subsequent work.