Introduction:

Prostate magnetic resonance imaging (MRI) has allowed improved prostate cancer risk-stratification utilizing the international Prostate Imaging Report and Data System version 2 (PI-RADS v2) classification system. Specifically, higher PI-RADS scores have been correlated with increased rates of positive targeted biopsies and clinically significant prostate cancer (CSPC). Nonetheless, with 10-30% rates of CSPC even for high PI-RADS, further opportunities to risk stratify patients for the presence of higher risk disease persist. The objective of this study was to stratify PI-RADS lesions based upon PSA density (PSAD).

Methods:

In an IRB approved study, 327 patients with lesions assigned PI-RADS scores from MRI and biopsied using magnetic resonance/ultrasound (MR/US) fusion to target all lesions were reviewed. Each biopsy sample from the lesions was given a Gleason grade and pathologic outcomes were stratified by various parameters, including age at time of MRI, race, and PI-RADS scores.  CSPC was defined as Gleason score ≥ 7. Logistic regression was used to determine odds ratios (OR) with 95% confidence intervals.  Age at time of MRI by quartiles, race, and PI-RADS score were included in multivariate analysis regardless of significance on univariate analysis. All p-values were reported as two-sided, and p<0.05 was considered statistically significant.

Results:

709 lesions were analyzed. When using a PSAD cut-off of ≥0.15 ng/ml/cc, 24% of all PI-RADS 4 and 47% of all PI-RADs 5 lesions were found to have CSPC, compared to 11% of PI-RADS 4 and 35% of PI-RADS 5 lesions with PSAD <0.15 (see figure). When controlling for PI-RADS, age, and race, multivariate analysis showed that PSAD was independently associated with CSPC using the cutoff of ≥0.15 ng/ml/cc (Odds Ratio (OR) 2.24, 95% confidence interval (CI) 1.41-3.54, p<0.001). This finding was also supported when performing multivariate analysis controlling for PI-RADS, age, and race using PSAD as a continuous variable (OR 1.03 per 0.01 PSAD increase, 95% CI 1.02-105, p<0.001). A plurality of all PI-RADS ≤3 lesions had a PSAD between 0-0.1, whereas PI-RADS 4 lesions had a PSAD central tendency between 0.1-0.2. PI-RADS 5 lesions appear to have no clear distinct distribution of PSAD compared to lower PI-RADS scores.

Conclusion:

PSAD appears to be a useful marker that can stratify for the risk of CSPC in a complementary manner to prostate MRI given the wide distribution of PSA densities for higher PI-RADs lesions.  PSA density ≥0.15 shows a 2.2x increased risk for CSPC in PI-RADs 4, and 1.4x increased risk for PI-RADS 5 lesions. Further studies are warranted to help determine optimal PSAD cut-offs by PI-RADS scores to best balance CSPC predictions without over-diagnosis.

Funding: N/A