Optimal PSA density threshold and predictive factors for the detection of clinically significant prostate cancer in patient with a PI-RADS 3 lesion on MRI.

• In patients with a prostate imaging reporting and data system 3 lesion, the detection rate of clinically significant prostate cancer (csPCa) is about 17%. Therefore, the management of this patients can be discussed. • Prostate-specific antigen density (PSAD) ≥ 0.15 ng/ml/cm3 and age were independent predictive factors of csPCa and previous negative prostate biopsy (PBx) was negatively associated with csPCa. • Using PSAD alone to select patient to undergo PBx, would omit 71.5% of PBx at the cost of missing 15.0% of csPCa. • Other predictive factors as age and PBx history should also be considered in the discussion with the patient, to avoid several PBx while missing few csPCa. • Combining predictive factors could be an option to avoid unnecessary PBx. Indeed, in patient with PSAD < 0.15 ng/ml/cm3, younger than 65 years old with previous negative PBx, there was no csPCa. While Prostate Imaging Reporting and Data System (PI-RADS) 4 and 5 lesions usually justify prostate biopsy (PBx), the management of a PI-RADS 3 lesion can be discussed. The aim of our study was to determine the optimal prostate-specific antigen density (PSAD) threshold and predictive factors of clinically significant prostate cancer (csPCa) in patients with a PI-RADS 3 lesion on MRI. Using our prospectively maintained database, we conducted a monocentric retrospective study, including all patients with a clinical suspicious of prostate cancer (PCa), all of them had a PI-RADS 3 lesion on the mpMRI prior to PBx. Patients under active surveillance or displaying suspicious digital rectal examination were excluded. Clinically significant (csPCa) was defined as PCa with any ISUP grade group ≥ 2 (Gleason ≥ 3 + 4). We included 158 patients. The detection rate of csPCa was 22.2%. In case of PSAD ≤ 0.15 ng/ml/cm3, PBx would be omitted in 71.5% (113/158) of men at the cost of missing 15.0% (17/113) of csPCa. With a threshold of 0.15 ng/ml/cm3, the sensitivity and the specificity were 0.51 and 0.78 respectively. The positive predictive value was 0.40 and the negative predictive value was 0.85. According to multivariate analysis, age (OR = 1.10, CI95% 1.03–1.19, P = 0.007), and PSAD ≥ 0.15 ng/ml/cm3 (OR = 3.59, CI95% 1.41–9.47, P = 0.008) were independent predictive factors of csPCa. Previous negative PBx was negatively associated with csPCa (OR = 0.24, CI 95% 0.07–0.66, P = 0.01). Our result suggests that the optimal PSAD threshold was 0.15 ng/ml/cm3. However, in this case omitting PBx in 71.5% of cases would be at the cost of missing 15.0% of csPCa. PSAD should not be used alone, other predictive factors as age and PBx history should also be considered in the discussion with the patient, to avoid PBx while missing few csPCa. [Display omitted] [ABSTRACT FROM AUTHOR]