Abstract
Background Overdiagnosis and unnecessary prostate biopsies remain significant challenges in prostate cancer screening. This study addresses the limited integration of prostate-specific antigen density (PSAD) with magnetic resonance imaging (MRI) Likert scoring in large retrospective UK cohorts. We hypothesized that combining PSAD thresholds with MRI Likert scores would improve diagnostic specificity and reduce unnecessary biopsies compared to MRI assessment alone. Methodology We conducted a retrospective cohort study of 958 men who underwent transperineal prostate biopsy between 2019 and 2023 at a single tertiary center. Serum PSA and MRI-derived prostate volumes were used to calculate PSAD, with volumes obtained on T2-weighted MRI using an ellipsoid formula. Measurements were performed on institutional reporting software by two independent radiologists to assess interobserver consistency. PSAD thresholds of 0.10, 0.15, and 0.25 ng/mL/cc were evaluated. Clinically significant prostate cancer (csPCa) was defined as a Gleason Grade Group (GG) ≥ 3. Diagnostic accuracy metrics with 95% confidence intervals were calculated, and receiver operating characteristic (ROC) analysis was performed. Results The median age of the cohort was 69 years. Median PSAD rose progressively with increasing Gleason Grade Group (GG0: 0.10; GG1: 0.10; GG2: 0.153; GG3: 0.257; GG4: 0.35; GG5: 0.37 ng/mL/cc). A PSAD < 0.10 ng/mL/cc yielded a negative predictive value of 0.846 for csPCa, while a PSAD > 0.25 ng/mL/cc produced a positive predictive value of 0.624. The area under the ROC curve (AUC) for PSAD in predicting csPCa was 0.712, indicating fair discriminative ability. Conclusions PSAD serves as a cost-effective adjunct to MRI, enhancing risk stratification for csPCa. Integrating PSAD thresholds with MRI Likert scoring supports more precise biopsy decision-making and may reduce unnecessary procedures. These findings are supported by retrospective data and warrant prospective, multicenter validation to inform standardized clinical pathways.