Predictive ability of magnetic resonance imaging (MRI) for detecting prostate cancer and its clinical significance in MRI-targeted biopsy for prostate imaging reporting and data system (PI-RADS) ≥3 lesions

Abstract

AIM

Identifying the index lesion in prostate cancer (PCa) is vital for its treatment. Therefore, various coefficients and parameters are used to improve the diagnostic accuracy of magnetic resonance imaging (MRI). This study aimed to analyze MRI data, utilized as a triage test before prostate biopsy, to identify independent risk factors affecting negative biopsy results in PCa and investigate the ability of these factors to predict clinically significant and insignificant PCa (csPCa and ciPCa, respectively).

MATERIALS AND METHODS

A retrospective analysis was conducted on data from 364 patients with a prostate imaging reporting and data system (PI-RADS) v2.1 score of 3 or higher, who underwent cognitive MRI-targeted biopsy (MRI-TB). Of the patients, 226 (62.1%) had benign lesions, 75 (20.6%) were diagnosed with ciPCa, and 63 (17.3%) with csPCa. The study assessed patients' demographic, biochemical, and radiologic characteristics, including apparent diffusion coefficient (ADC) and ADC coefficient of variation (ADC_CoV) values.

RESULTS

The multivariate analysis performed to differentiate PCa from benign pathologies revealed that only MRI parameters, specifically the presence of PI-RADS 4 and 5 lesions (odds ratio [OR]: 12, p < 0.001 and OR: 73, p = 0.008, respectively), a lower ADC value (OR: 0.996, p = 0.041) and a higher ADC_CoV value (OR: 1.07, p = 0.003) were independent risk factors. No MRI findings had significant predictive power for csPCa, with total prostate-specific antigen (PSA) (OR: 1.17, p = 0.019) found to be the only independent risk factor.

CONCLUSION

The results of this study suggest that data obtained from MRI can predict PCa but not csPCa.