Comparison of biparametric and multiparametric magnetic resonance imaging in the detection of clinically significant prostate cancer – a prospective study with prostate-specific antigen density analysis

Abstract

Introduction/Objective Multiparametric MRI (mpMRI) is the standard imaging method for the detection of clinically significant prostate cancer (csPCa). Biparametric MRI (bpMRI), which omits dynamic contrast-enhanced imaging (DCE), may provide a faster and more accessible alternative. The aim of this study was to compare bpMRI and mpMRI in the detection of csPCa and evaluate the role of PSA density (PSAD).

Methods This prospective study included 170 men with suspected prostate cancer. All patients underwent 3-T prostate MRI before biopsy. MRI findings were interpreted according to PI-RADS v2.1, and histopathology served as the reference standard. Diagnostic performance was assessed at PI-RADS thresholds of ≥ 4 and ≥ 3. ROC analysis and logistic regression were used to evaluate PSAD.

Results Agreement between bpMRI and mpMRI was 98.2%. At the PI-RADS ≥ 4 threshold, bpMRI showed sensitivity of 97%, specificity of 77.9%, and accuracy of 85.3%, comparable to mpMRI (97%, 76.9%, and 84.7%). PSAD showed good diagnostic performance (AUC 0.872, 95% CI, 0.813–0.931, p < 0.001). The optimal cutoff selected by the Youden index was 0.205 ng/mL/cm3. Non-inferiority analysis confirmed that bpMRI was non-inferior to mpMRI for overall diagnostic accuracy.

Conclusion Biparametric MRI demonstrated diagnostic performance comparable to that of and non-inferior to mpMRI for detection of csPCa. PI-RADS ≥ 4 was the optimal diagnostic threshold, while PSAD improved risk stratification.

Keywords: prostate cancer; biparametric MRI; multiparametric MRI; PSA density; PI-RADS