Predictors of false-positive results in men with Prostate Imaging–Reporting and Data System 4 lesions

Abstract

Background

To identify predictive factors of false-positive results in men with Prostate Imaging–Reporting and Data System (PI-RADS) 4 lesions on multiparametric MRI and to evaluate potential factors for improving risk stratification in this high-risk category.

Methods

We retrospectively analyzed 811 men with PI-RADS 4 lesions who underwent prostate biopsy (November 2020–February 2025). Clinical data [age, PSA, PSA density (PSAD), prostate volume] and radiographic features {anatomical location [peripheral zone (PZ) vs. transitional zone (TZ)], multiplicity} were assessed. Multivariate logistic regression analysis identified independent predictors of false-positive results. Interaction between location and multiplicity was evaluated for zone-specific risk profiles.

Results

Of the 811 patients, 523 (64.5%) had benign pathology (false-positive). Independent predictors of false-positive results included younger age, lower PSAD, prior negative biopsy, and TZ location. TZ lesions had significantly higher odds of a false-positive result compared to PZ lesions (OR: 6.919, 95% CI: 4.440-10.781, P < 0.001). A divergent pattern was observed between location and multiplicity: in the PZ, multiplicity did not affect cancer risk, whereas in the TZ, multiple lesions showed a trend toward benign pathology (P = 0.087), with the lowest predicted cancer probability (10.0%) in patients with multiple TZ lesions.

Conclusion

False-positive PI-RADS 4 assessments were associated with features characteristic of BPH, particularly TZ localization and lower PSAD. Current biopsy recommendations may be suboptimal for patients with these benign-favoring features. Incorporating lesion localization and volume-adjusted metrics into risk stratification could potentially reduce unnecessary biopsies.