PI-RADS 3 MRI lesions: Are biopsies still necessary?

Introduction: A significant proportion of newly diagnosed prostate cancer (PCa) cases are slow growing with a low risk of metastatic progression. There is a lack of data concerning the optimal biopsy regimen for improving diagnosis yield in PI-RADS3 lesions. This study aimed to assess the diagnostic value of current biopsy regimens in PI-RADS 3 lesions and identify clinical predictors to improve clinically significant PCa (csPCa) detection.

Methods: This retrospective study included patients from two academic centers between 2017 and 2024 who benefitted from prostate biopsies for a PI-RADS 3 lesion. Prostate biopsies were performed via either a transrectal or a transperineal route according to local resources. All patients underwent systematic and targeted cores. Targeted biopsies were performed using MRI-ultrasound fusion software or cognitive fusion. Patients were categorized based on biopsy results: benign, clinically insignificant (insignPCa, i.e. ISUP 1), or csPCa (ISUP≥2). The primary endpoint was to assess the detection rate of csPCa in MRI-targeted biopsies alone and in MRI-targeted+ipsilateral systematic cores in comparison to the gold standard MRI-targeted+systematic biopsies. Then, the percentage of csPCa missed and insignPCa diagnoses avoided were calculated referring to the gold standard MRI-targeted+systematic template.

Results: A total of 163 men with at least one PI-RADS 3 index lesion were identified. Ninety-one (55.8%) lesions were benign, 52 (32%) were insignPCa and 20 (12.3%) were csPCa. Of the 20 patients with csPCa, the diagnosis was made with targeted cores in 12 patients and with systematic cores in 8 patients. If no biopsies were performed, 12.3% of csPCa would have gone undiagnosed. Targeted biopsies alone would have missed 4.9% of csPCa but avoided 24.5% of insignPCa. Targeted biopsies with systematic cores solely performed on the same side as the index lesion would have missed 3.8% of csPCa and avoided 12.9% of insignPCa. We reported no significant differences in detection rate of csPCa between MRI-targeted cores alone vs. the gold standard template (7.4% vs. 12.3%; P=0.9) and between MRI-targeted cores+ipsilateral systematic cores vs. the gold standard template (8.5% vs. 12.3%; P=1.2). At multivariable analysis age, clinical T stage, and mpMRI lesion size were predictors of csPCa.

Conclusion: As a small proportion of PI-RADS 3 lesions are associated with csPCa, the value of biopsies is questionable. Targeted biopsies with systematic cores on the same side as the index lesion may improve detection of csPCa and reduce overdiagnosis of indolent CaP. Clinical T stage and lesion size on MRI can potentially predict the presence of clinically significant CaP.