Improving multivariable prostate cancer risk assessment using the Prostate Health Index

Area of Science:

• Urology
• Oncology
• Biomarker Research

Background:

• Accurate risk stratification is crucial for men undergoing prostate cancer assessment.
• Current prediction models rely on clinical factors and prostate-specific antigen (PSA) levels.

Purpose of the Study:

• To evaluate the clinical utility of a prediction model incorporating clinical data and the novel biomarker p2PSA.
• To assess the model's effectiveness in guiding prostate biopsy decisions in an Irish cohort.

Main Methods:

• Serum samples from 250 men were analyzed for PSA, free PSA (fPSA), and p2PSA.
• The Prostate Health Index (PHI) score was calculated.
• Clinical prediction models were developed using logistic regression, including age, family history, digital rectal examination, prior biopsy, and either PSA or PHI.

Main Results:

• The PHI model demonstrated superior accuracy in predicting prostate cancer (PCa) and high-grade PCa compared to PSA and PCPT models (AUCs 0.77 vs 0.71 vs 0.69 for PCa).
• The PHI model showed better calibration and a superior net benefit in decision-curve analysis.
• The PHI model outperformed traditional methods in risk stratification.

Conclusions:

• Integrating p2PSA into a clinical prediction model enhances risk stratification accuracy for men investigated for PCa.
• The PHI score improves decision-making regarding prostate biopsies in referral populations.
• This approach offers a more precise method for identifying men who would benefit from biopsy.