Role of multiparametric prostate magnetic resonance imaging in malignancy estimation in prostate biopsy

Objective: PI-RADS (Prostate Imaging-Reporting and Data System) v2.1 is a classification used in the prediction of prostate malignancy. It is expected that there will be a correlation between the risk group given as a result of this classification and prostate biopsy pathology. The aim of our study is to reveal the relationship between PI-RADS v2.1 classification and prostate biopsy pathologies in Mp-MRI examinations performed in our center. Materials And Methods: 138 patients who underwent prostate biopsy and also underwent Mp-MRI due to PSA value over 2.5 ng/ml and/or suspicious digital rectal examination (DRE) findings were included in our study. After ethics committee approval, patient data were retrospectively scanned from April 2018 to April 2021. Prostate biopsy results and PI-RADS v2.1 scores and lesion locations of Mp-MRI images taken and reported in our center's radiology clinic were compared. Those with PI-RADS v2.1 scores of 1,2 and 3 were benign; those 4 and 5 were considered malignant. At the end of our study, patients' ages, free PSA, total PSA, free/total PSA ratio, prostate volumes, PSA densities, biopsy pathologies and Mp-MRI results were compared. The compatibility of the lesion in the imaging performed in patients with malignancy and the biopsy quadrant and side was examined. When the distribution was normal, the independent sample t-test or if it was nonparametric Mann Whitney-U test was used to compare continuous variables in the two groups. The chi-square test was used to compare categorical variables. Binomial logistic regression analysis, which also provides the opportunity to calculate the relative risk, was used to examine the causal relationship and statistical significance of each of the mentioned independent variables with the presence of malignancy. ROC analysis was used to evaluate the success of the examined variables in diagnosing malignancy and to determine the cut-off points. Area under the curve (AUC), sensitivity, specifity, positive predictive value (PPV), negative predictive value (NPV) were calculated. Results: In our study conducted with 138 patients in total, the rate of malignancy detection in the pathology results of the patients increased as the PI-RADS v2.1 score in Mp-MRI results increased. While the number of patients with malignant pathology was 40 (29%), their mean age and PSA density were significantly higher than patients without malignancy. The rate of suspicious findings or hardness in digital rectal examinations was significantly higher in patients with malignant pathology. Parameters that will give an idea to the physician before biopsy are digital rectal examination, total PSA value, free/total PSA ratio, PSA density and, if Mp-MRI is applied, its PI-RADS score. In this direction, prostate biopsy may be required if necessary. As a result of the analysis, the contribution of PSA density and PI-RADS scoring to the other mentioned parameters in the prediction of malignancy comes to the fore. The negative predictive value (NPV) of Mp-MRI in detecting malignancy was 87.37% (95% Confidence Interval: 81.05% - 91.79%), and the positive predictive value (PPV) was 65.12% (95% CI: 52.9-75.62%) was calculated. In our study, the cut-off point in terms of distinguishing benign and malignant pathology results of PSA density was found to be 0.15 ng/ml, consistent with the literature. The joint use of PSA density with the PI-RADS v2.1 score was considered to be clinically important. It was determined that the diagnostic accuracy of Mp-MRI increased from 80.43% to 82.54% in patients with PSA density of 0.15 ng/ml2 and above. Conclusions: Parameters that will give the physician an idea before biopsy are digital rectal examination, total PSA value, free/total PSA ratio, PSA density and, if Mp-MRI is applied, its PI-RADS score. In this direction, prostate biopsy may be required if necessary. As a result of the analysis, the contribution of PSA density and PI-RADS scoring rather than the other mentioned parameters in the prediction of malignancy stands out.