Procedure efficiency and diagnostic outcomes during the learning curve of transperineal MRI-guided prostate biopsy

This work aimed to investigate the role of surgeon experience on operating time and detection rate of clinically significant prostate cancer (csPCa) in a large contemporary cohort undergoing transperineal (TP) MRI-targeted prostate biopsy.

We retrospectively reviewed patients who underwent TP MRI-guided prostate biopsies between 10/2018 and 04/2024. Patients with suspicious MRI lesions (PI-RADS 3, 4 and 5) formed the study cohort. All patients underwent MRI-targeted biopsies of suspicious lesions combined with a standard 12-core systematic biopsy. Caseloads-both total and in increments of 50 cases-were calculated for each surgeon. We employed univariable and multivariable Poisson regression models to analyse biopsy duration, and logistic regression models to assess csPCa detection.

Out of 3336 MRI-guided TP biopsies, 1148 (34%) involved PI-RADS 3, 1398 (42%) PI-RADS4 and 790 (24%) PI-RADS 5 as index lesions. Most patients (55%) had one suspicious lesion. The median caseload per surgeon was 63 (IQR [interquartile range]: 30-110). Biopsy duration significantly decreased after the first 50 procedures (13 min [95% CI: 10-18] to 10 min [95% CI: 8-15]) and plateaued thereafter. Multivariable Poisson regression demonstrated increased caseload independently associated with shorter biopsy duration (IRR 0.891 [95% CI: 0.833-0.899], p < 0.001). The overall csPCa detection rate was 43%, with no significant correlation to caseload. These findings were consistent across subgroup analyses based on PI-RADS score.

This large, contemporary study demonstrated that biopsy duration decreased significantly after approximately 50 cases, indicating a learning curve without compromising the detection rate of csPCa.