Omitting biopsy in PI-RADS 3? The role of Stockholm3 and PSA density-findings from a prospective Swiss non-referral cohort

Abstract Introduction PI-RADS 3 (‘equivocal’) prostate MRI lesions pose a diagnostic challenge, particularly in non-referral settings with variable MRI quality. We evaluated whether prostate-specific antigen density (PSAD) and the Stockholm3 blood test can aid biopsy decisions for men with PI-RADS 3 lesions in a prospective, non-referral Swiss cohort. Methods Retrospective analysis of a prospective registry at a non-referral urology practice in Switzerland. From 2023–2025, men with PSA >1.5 ng/mL or suspicious digital rectal examination underwent Stockholm3 testing, MRI, and MRI-fusion plus systematic prostate biopsy. Clinically significant prostate cancer (csPCa) was defined as ISUP grade ≥2. Diagnostic accuracy and decision curve analysis compared Stockholm3 (≥13%, ≥15%, ≥18%) and PSAD thresholds (0.10, 0.15, 0.20ng/mL/cc) within the PI-RADS 3 subgroup. Results Overall, 50/174 (29%) men had PI-RADS 3 lesions, 59/174 (34%) were diagnosed with csPCa. Among PI-RADS 3 lesions, csPCa prevalence was 14%. Outcome distributions shifted towards higher ISUP grades with increasing PSAD thresholds when compared to baseline (≥0.15ng/mL/cc: 44% ISUP≥2; p=0.009), whereas Stockholm3 cut-offs showed no significant change for csPCa. As continuous measures, PSAD demonstrated superior discrimination (AUC 0.751; 95% CI 0.46-0.95) than Stockholm3 (AUC 0.595; 95% CI 0.30–0.88), confirming higher overall diagnostic accuracy within the PI-RADS 3 subgroup. Decision curve analysis demonstrated higher net benefit for PSAD-based approaches. Conclusion In non-referral practice, Stockholm3 did not reliably exclude csPCa among men with PI-RADS 3 lesions. PSAD-based triage, particularly at ≥0.15ng/mL/cc, showed superior diagnostic utility and may support safe biopsy deferral in selected cases. Prospective validation in broader non-referral settings is warranted.