Introduction & Objectives: The ability of MRI to detect clinically significant prostate cancer (csPCa) is important to select patients for prostate biopsy. Here, we aim to compare clinical and pathological characteristics of patients with csPCa subjected to MRI-US/FB.
Materials & Methods: We retrospectively analyzed a confidential database of patients with prostate cancer subjected to MRI/US-FB between October 2020 and April 2022. We stratified patient in 2 groups (ISU 1 group or ISUP>=2 group). Baseline characteristics of patients and pathological outcomes were compares using Student-t-test (age, PSA, PSAD, prostatic volume (PV) and index lesion size, total of cylinders collected and percentage of positive cylinders). Chi-square test was used for categorical data (peripheric lesion; multifocality; biopsy-naïve; neural invasion; lymph vascular invasion; cribriform and intraductal pattern). A P-value < 0,05 was considered statistically significant. Those statistical analysis was performed using SPSS v.25 (IBM SPSS Statistics for macOS, Version 25.0. Armonk, NY: IBM Corp.) We also calculated PSAD, PI-RADS and index lesion size area under the curve (AUC) and predictive positive value (PPV) to predict csPCa.
Results: 68 patients with PCa were identified. Overall, 18 (26,5%) patients presented with non-csPCa against 50 (73,5%) who presented with csPCa. On univariate analysis there were no statistically significant differences between both groups of patients regarding age, pre-operative PSA, PSAD, PV, multifocality, presence of cribriform or intraductal pattern. PI-RADS score was very close to achieve statistical significance (p=0,053). In this cohort, as higher was PI-RADS score, higher was the percentage of patients with csPCa. Lesion size (p= 0,029), the percentage of positive cylinders (p=0,003) and neural invasion detected by MRI/US-FB were associated with the presence of csPCa. The AUC to predict any csPCa was 0,673 (CI 95% 0,549-0,782 P=0,0192) for PIRADS. The highest Youden’s Index was at PIRADS >4 with the sensibility of 42% and specificity of 83,3% for the diagnosis of csPCa. The PPV was 87,462% (CI 95% 70,28-95,37%). The AUC to predict any csPCa was 0,672 (CI 95% 0,548-0,781 P=0,023) for index lesion size. The highest Youden’s Index was achieved for lesions with >8mm with the sensibility of 90% and specificity of 38,89% for the diagnosis of csPCa. The PPV was 80,33% (CI 95% 73,64-85,66%). On the other hand, the AUC to predict any csPCA was 0,728 for PSAD (CI 95%: 0,607-0,829 P=0,0038). The highest Youden’s Index was at PSAD>0,1ng/mL2 (sensitivity of 84% and specificity of 61,1%). The PPV was 85,69% (CI 95%: 76,83-91,54%).
Conclusions: Patients with csPCa have higher index lesion size, % of positive cylinders, more prevalence of cribriform pattern and neural invasion than non-csPCa patients. As higher is the PI-RADS score, index lesion size and PSAD, higher is the probability of csPCA.

Exploring prostate cancer patients’ characteristics to identify parameters that can predict clinically significant prostate cancer