Introduction

MR/US fusion biopsy (FB) is a promising modality for detection of clinically significant prostate cancer (csPCa), defined as Gleason >=7 in patients who have had a prior negative biopsy. The purpose of this study is to assess the learning curve with adoption of FB using PI-RADS Version 2 (v2) for detecting csPCa and to identify patient and technical factors that predict success.

Methods

A total of 113 consecutive patients with at least one prior negative biopsy and a multiparametric MRI (mpMRI) exam of the prostate with a PIRADS 3 or greater index lesion underwent FB at a single academic center previously naive to FB technology. Outcomes are detection rates for Gleason 6 cancer, csPCa, and any cancer. The following 22 covariates were analyzed: age, body mass index (BMI), PSA, prostate volume (MRI-estimated), prostate volume (US-estimated), PSAD (MRI-estimated), PSAD (US-estimated), time interval since the last negative SB, number of prior negative systematic biopsies, number of targeted biopsy cores of the index lesion, size of index lesion, PI-RADS v2 score, number of suspicious lesions on mpMRI, institution experience, surgeon, obesity, digital rectal exam (DRE), atypical small acinar proliferation (ASAP) on prior biopsy, high-grade prostatic intraepithelial neoplasia (HGPIN) on prior biopsy, and location of index lesion (zone, region, and sector). Multiple logistic regression with model selection was used to select covariates having significant effects on the outcome._x000D_

Results

Prostate cancers were identified in 52% of cases. Among patients diagnosed with prostate cancer, 80% were clinically significant. The detection rates of csPCa using FB when a PIRADS 3, 4, or 5 index lesion was present on mpMRI were 6%, 46%, and 66%, respectively. PI-RADS v2 score had a predictive accuracy (AUC) of 0.79 for csPCa detection. Institutional experience over time, MRI-estimated prostate volume, and PI-RADS v2 score were independent predictors of success at detecting csPCa.

Conclusions

Since FB is a highly technical and experience-driven process, development of internal quality measures to assess the institutional learning curve and the quality of PI-RADS v2 scoring is critical with adoption of this technology.

Funding

none