Effects of the addition of quantitative apparent diffusion coefficient data on the diagnostic performance of the PI-RADS v2 scoring system to detect clinically significant prostate cancer

Abstract

Purpose: To evaluate the impact of the addition of quantitative apparent diffusion coefficient (ADC) data into the diagnostic performance of the Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) scoring system to predict clinically significant prostate cancer (CSPCa). Methods: We retrospectively included 91 consecutive patients who underwent prostate multiparametric magnetic resonance imaging (mp-MRI) and histopathological evaluation. Mp-MRI images were reported by the PI-RADSv2 scoring system and patients were divided into groups considering the likelihood of CSPCa. ADC value and ratio were obtained. Findings were correlated with histopathological data. Results: CSPCa was found in 41.8% of cases (n = 38). PI-RADSv2 score 3–5 yielded a sensitivity of 97.4% (95% confidence intervals 86.5–99.5), a specificity of 50.9% (37.9–63.9), and AUC of 0.74 (0.67–0.81) to predict CSPCa. ADC value < 750 µm2/s and an ADC ratio < 0.62 were the most accurate thresholds for differentiation of CSPCa, with AUC of 0.81 and 0.76, respectively. Combined PI-RADSv2 score 3–5 and ADC value < 750 µm2/s yielded a specificity of 84.9 (72.9–92.2), sensitivity of 70.3 (54.2–82.5), and AUC of 0.77 (0.68–0.86). Combined PI-RADSv2 score 3–5 and ADC ratio < 0.62 yielded a specificity of 86.5 (74.7–93.3), sensitivity of was 64.9 (48.8–78.2), and AUC of 0.75 (0.66–0.84). Conclusion: Quantitative ADC data might not be beneficial to be used routinely in mp-MR imaging as criteria to detect clinically significant lesions due to the reduced sensitivity. Instead, when prostate lesions present a PI-RADSv2 score ≥ 3, additional quantitative ADC criteria can be helpful to increase the PI-RADS score specificity.