Development and Validation of a Novel Prognostic Model for Long-Term Prostate Cancer Mortality

Article Impact Level: HIGH
Data Quality: STRONG
Summary of  Annals of Internal Medicine https://doi.org/10.7326/ANNALS-25-02036 
Dr. Patrick Lewicki et al.

Points

Researchers developed a novel prognostic model using data from over thirty-three thousand men to predict the long-term risk of dying from prostate cancer following initial prostate-specific antigen screening tests.
The tool demonstrated a significantly higher area under the receiver operating characteristic curve of zero point six hundred sixty-six when compared to traditional biopsy risk models in long-term trials.
Validation in a separate cohort of nearly one hundred seventy-five thousand Veterans Affairs patients confirmed the model’s superior accuracy in predicting cancer-related mortality over a twenty-year period of observation.
By incorporating age, race, and family history, the model allows clinicians to interpret screening results more effectively while specifically adjusting for a patient’s individual life expectancy and overall health status.
This personalized approach to prostate cancer screening aims to reduce the prevalence of overtreatment and unnecessary medical procedures for the nearly ten million men who undergo annual diagnostic testing.

Summary

This study developed and validated a novel prognostic model designed to estimate the long-term risk of prostate cancer mortality following prostate-specific antigen (PSA) screening. Using longitudinal data from 33,000 men aged 55–74 in the PLCO trial, researchers integrated PSA levels with age, race, family history, and comorbidities. The model addresses a significant gap in current clinical practice by adjusting for patient life expectancy and predicting time-to-event outcomes, providing a framework to reduce overtreatment for the nearly 10 million patients screened annually in the United States.

In the development cohort, the model demonstrated superior predictive capability over a 29.5-year horizon. The area under the receiver operating characteristic curve (AUC) was 0.666, significantly outperforming the 0.643 achieved by the previously validated Prostate Biopsy Collaborative Group (PBCG) risk model ($P < .001$). External validation in a cohort of approximately 175,000 Veterans Affairs patients confirmed this efficacy. At 20 years post-screening, the novel PLCO model maintained an AUC of 0.776, compared to 0.749 for the PBCG model ($P = .031$).

The tool offers a precise interpretation of PSA results by quantifying the absolute risk of death from prostate cancer over several decades. By incorporating non-genetic factors and individual health states, the model provides clinicians with a statistically robust method to tailor screening intervals and treatment decisions. This transition toward personalized care helps identify high-risk individuals who require aggressive intervention while sparing low-risk patients from the morbidity associated with unnecessary biopsies and overdiagnosis.

Link to the article: https://www.acpjournals.org/doi/10.7326/ANNALS-25-02036

References

Lewicki, P., Jiang, R., Radhakrishnan, A., Bryant, A., Schipper, M., Morgan, T. M., & Stensland, K. (2026). Predicting long-term risk for prostate cancer mortality following a prostate-specific antigen screening test: Prognostic model development and external validation. Annals of Internal Medicine, ANNALS-25-02036. https://doi.org/10.7326/ANNALS-25-02036