Prospective evaluation of PI-RADS v2 and quantitative MRI for clinically significant prostate cancer detection in Indian men – East meets West

Abstract

ABSTRACT Purpose To validate the detection of clinically significant prostate cancer (Gleason’s score ≥7) by PI-RADS v2 and to assess the ability of quantitative MRI parameters to detect clinically significant prostate cancer (CSPCa) in Indian men. Methods Adult men (n = 95) with serum PSA >4 ng/ml were prospectively evaluated with multiparametric MRI (mpMRI) followed by histopathological evaluation using systematic 12-core prostate biopsy in 69 patients and prostatectomy specimens in 26 patients, performed within six weeks of mpMRI. The imaging and the pathology were divided into 12 sectors per prostate. For the validation of PI-RADS v2, a cut-off of PI-RADS v2 score ≥ 3 and PI-RADS v2 score ≥ 4 were compared to histopathology as a reference standard. Further, quantitative parameters, apparent diffusion coefficient (ADC), Ktrans, and Kep were correlated with the Gleason score and evaluated for their ability to distinguish between sectors with CSPCa and sectors without CSPCa. Results PI-RADS score ≥ 4 showed higher specificity (89%) than PI-RADS score ≥ 3 (72.2%) at the cost of mild but not significant reduction of sensitivity (sensitivity–87.6% vs 91.9), (n = 1,140 sectors, 95 patients). PI-RADS v2 and quantitative parameters demonstrated the ability to discriminate sectors positive vs negative for CSPCa: AUC (area under the curve) for ADC was 0.928, PI-RADS v2 was 0.903, Ktrans was 0.897 and Kep was 0.695. Gleason score correlated well with PI-RADS (r = 0.74), ADC (r = −0.73) and Ktrans (r = 0.69). Conclusion PI-RADS v2 is a reliable method for the detection and localization of clinically significant prostate cancer in Indian men, suggesting applicability beyond European or American demographics. Quantitative mpMRI parameters can detect clinically significant prostate cancer with similar test characteristics as PI-RADS v2.