Does variation in prostate volume calculated with MRI versus TRUS lead to reclassification of men based on current PSA density threshold for active surveillance criteria? Result from a prospective cohort.

Abstract

138 Background: PSA density (PSAD) is a strong predictor of aggressive prostate cancer (PCa) and is often used as a selection criterion for active surveillance. However, measurement of PSAD can vary depending on the modality used to estimate prostate volume (PV). We analyzed a prospective cohort of men undergoing MRI−US fusion biopsy to assess the variation in PV obtained with both imaging modalities, and investigate the impact of this variation on PSAD measurement in order to determine if it led to re-classification of patients above or below the current threshold of PSAD used in clinical practice (0.15 ng/mL/cc). Methods: All men were consecutively enrolled in this prospective study and had their PV measured on MRI prior to prostate biopsy (PB), and on Trans-Rectal Ultrasound (TRUS) at the time of PB. PSAD was calculated by dividing the last PSA prior to biopsy by the PV ascertained with each imaging modality. We used paired t-tests and Wilcoxon signed−rank tests to compare the difference in PV and PSAD obtained with TRUS and MRI. We also categorized PSAD measurements on each imaging modality above and below a cut-off of 0.15 ng/mL/cc. We used the McNemar’s test for paired proportions to estimate the significance of discordance in PSAD categorization based on each imaging modality. Results: Of the 124 men, the mean PV assessed with MRI (70 cc) was 7 cc more on average (SD: 11, median = 4 cc) than that obtained with TRUS (63 cc) (P < 0.0001). Furthermore, the mean PSAD obtained with MRI (0.17 ng/mL/cc), was 0.01 ng/mL/cc lower (SD: 0.07, median = 0.01) than that obtained with TRUS (0.18 ng/mL/cc) (P < 0.0001). 118 (95%) men had concordant PSAD values assessed with either imaging modality using the cut-off of 0.15 ng/mL/cc. All 6 (5%) men with discordant PSAD values who were above the cut-off with PV obtained via TRUS were reclassified as being below the cut-off with PV assessed via MRI (P < 0.01). Conclusions: MRI can overestimate PV compared to TRUS. This may translate to reclassification of men around the currently used PSAD threshold and have implications for treatment decision making and selection of patients for active surveillance.