Introduction:

Multiparametric magnetic resonance imaging (mpMRI) may identify anatomic lesions amenable to targeted biopsies to augment the standard template for men with prior negative prostate biopsies. However, it is unclear whether risk of prostate cancer for a given Prostate Imaging-Reporting and Data System (PI-RADS) version 2.0 score lesion differs between men with prior negative biopsies compared to the biopsy naïve population.

Methods:

The Prospective Loyola University mpMRI (PLUM) Prostate Biopsy Cohort includes men referred for mpMRI prior to prostate biopsy. Men with clinical suspicion for prostate cancer received mpMRI followed by mpMRI-transrectal ultrasound (TRUS) fusion-guided biopsies if PI-RADS 3-5 lesions were identified (systematic and targeted cores) or systematic biopsies only for the remainder. Patients from January 2015 to June 2020 receiving their first mpMRI without a previous diagnosis of prostate cancer were included in the present study and separated into biopsy naïve or prior negative biopsy cohorts. Baseline clinical and mpMRI parameters were compared between groups, and risk of prostate cancer diagnosis was assessed stratified by PI-RADS score. Predictors of detecting prostate cancer were assessed in multivariable logistic regression models for each cohort. Adjusted odds ratios were calculated by PI-RADS score between cohorts for the outcomes of any prostate cancer and Gleason ≥3+4 prostate cancer.

Results:

A total of 480 biopsy naïve and 420 prior negative biopsy patients were identified. Prior negative biopsy patients had greater age (median 67.1 vs. 64.3), PSA (median 7.5 vs. 5.5), and PI-RADS 3 lesions (36.7% vs. 24.4%). Men with prior negative biopsy had lower detection rates for any prostate cancer (27.9% vs. 54.4%) and Gleason ≥3+4 prostate cancer (20.0% vs. 38.3%). When stratified by PI-RADS, men with prior negative biopsies remained less likely to be diagnosed with any cancer and Gleason ≥3+4 cancer. Multivariable logistic regression models confirmed these findings (ORs 0.11, 0.20, and 0.31 among PI-RADS 5, 4, and 3, respectively; all p<0.01). Notably, younger age and Asian race were associated with less probability of cancer among biopsy naïve patients but not for prior negative biopsy patients. Deferring repeat biopsy for PI-RADS ≤3 lesions could have avoided 54.6% (225/420) of biopsies while missing 14.3% (12/84) of Gleason ≥3+4 cancers.

Conclusion:

For any given PI-RADS score, men with prior negative prostate biopsy are less likely than biopsy naïve men to be diagnosed with any prostate cancer or Gleason ≥3+4 prostate cancer. Notably, among men with prior negative biopsy, rates of Gleason ≥3+4 prostate cancer were 57.8% for PI-RADS 5, 26.7% for PI-RADS 4, and <6% for PI-RADS ≤3 suggesting biopsy could be deferred in the latter group.

Funding: Siemens