LAUSR

PURPOSE Prostate biopsy should be discussed with the patient in case of negative MRI (nMRI) and low clinical suspicious of prostate cancer (PCa). OBJECTIVES Primary objective was to describe the risk of clinically significant PCa (csPCa) in a nMRI biopsy-naïve population at baseline and during long-term follow-up. Secondary objective was to evaluate clinical factors and PSA as predictors of csPCa at baseline. MATERIALS AND METHODS All 503 consecutive biopsy-naïve patients referred in 2007-2017 for biopsy with nMRI (PIRADS1-2) who had systematic 12-core-biopsies (SB) at baseline were included. Clinical factors were digital rectal examination (DRE), PCa-family-history and PSA. In case of suspicious DRE or PSA kinetics during follow-up, MRI and biopsy were performed. CsPCa was defined as either GG1 with cancer-core-length >5 mm or ≥3 positive SB in addition to GG ≥2 (csPCa-1) or any GG ≥2 (csPCa-2). Non-clinically-significant-PCa was defined as either GG1 with cancer-core-length ≤5 mm and <3 positive SB (non-csPCa-1) or any GG1 (non-csPCa-2). Definition of high-risk csPCa was GG ≥3. Univariate and multivariate models were fitted to identify predictors of csPCa risk. RESULTS At baseline, biopsy showed csPCa-1 in 9% (n=45) and csPCa-2 in 6% (n=29) and non-csPCa in 22% (n=111). At median follow-up of 4 yrs (IQR: 1.6-7.1), 31% (95% CI: 27-36) of 415 untreated patients had a second MRI and 24% (95% CI: 20-28) a second biopsy which showed csPCa-1 in 5% (21/415, 95% CI: 3-7), csPCa-2 in 2% (7/415, 95% CI: 1-3) and non-csPCa in 8%. Overall incidence was 13% (n=66/503, 95% CI: 7-21) for csPCa-1, 7% (n=36/503, 95% CI: 5-9%) for csPCa-2 and 2% (n=12/503, 95% CI: 1.1-3.7) for high risk PCa. Predictors of csPCa risk were PSAd ≥0.15 ng/mL/mL (OR=2.43 [1.19-4.21]), clinical stage ≥T2a (OR=3.32 [1.69-6.53]) and PCa-family-history (OR=2.38 [1.10-6.16]). Performing biopsy in patients with nMRI and PSAd ≥0.15 ng/ml/ml or abnormal DRE or PCa-family-history would have decreased from 9% to 2.4% the risk of missing csPCa-1 at baseline while avoiding biopsy in 56%. CONCLUSION Risk of csPCa in a negative MRI biopsy-naïve population was 6%-9% at baseline and 7%-13% at long-term follow-up depending on csPCa definitions.