MP09-04 PROSTATE HEALTH INDEX DENSITY (PHID) IS MORE PREDICTIVE OF DISEASE RE-CATEGORIZATION THAN MULTIPARAMETRIC MRI IN MEN UNDERGOING ACTIVE SURVEILLANCE

Abstract

INTRODUCTION AND OBJECTIVE: Prostate Health Index (phi) is an FDA-approved serum biomarker that is calculated using total PSA, free PSA, and pro2PSA in an algorithm to estimate the likelihood of having prostate cancer (PCa) and high-grade disease on biopsy. The phi value can be divided by the prostate volume to calculate a phi density (phiD) that may be even more accurate for detecting high grade tumors. Although phiD has been described, its utility for identifying men at risk of up-grading during active surveillance (AS) has not been evaluated or compared against other commonly used metrics. METHODS: Men who had a phi value and multiparametric prostate MRI (mpMRI), with clinical stage ≤T2a and Gleason grade group 1 (GG1) cancer on a diagnostic biopsy, and were enrolled in an IRB-approved AS program were included. To avoid sampling error bias, any subject who was re-categorized to ≥GG2 on biopsy within 6 months of the diagnostic biopsy was excluded. Clinical variables including age, race, PSA and family history of PCa were documented. A positive mpMRI was defined as PIRAD≥3. The predictive value of phiD and mpMRI on disease re-categorization was assessed using both univariate and multivariable logistic regression in SAS 9.4 (Cary, NC). RESULTS: 418 men were included of which 57 (13.6%) were re-categorized to ≥GG2 on surveillance biopsies. There were no differences in the distribution of race, family history, and age between those who were re-categorized and those who remained GG1. However, the disease re-categorization group exhibited significantly more positive mpMRI (p=0.0001) and had higher PSA values (p=0.006), phi score values (p<.0001), and phiD values(p<.0001) (Table). In univariate analyses, phiD yielded the highest predictive value of disease re-categorization (AUC=0.79), followed by phi score (AUC=0.78), mpMRI (AUC=0.66) and PSA (AUC=0.65). Multivariable analysis demonstrated that phiD (OR=2.75, 95%CI 1.82-4.14) and mpMRI (OR=2.60, 95%CI 1.21-5.58) remained significant predictors (combined AUC=0.82). CONCLUSIONS: phiD is more predictive of disease re-categorization than mpMRI in men being followed by AS. As such, phiD should be strongly considered as part of algorithms to identify men who are at risk of disease re-categorization in AS.Source of Funding: None.