Combination of PI-RADS version 2.1 and amide proton transfer values for the detection of clinically significant prostate cancer

Abstract Background The goal of this study was to assess whether combining amide proton transfer (APT)-weighted MRI with the Prostate Imaging Reporting and Data System scoring system version 2.1 (PI-RADS V2.1) could increase diagnostic accuracy compared to PI-RADS V2.1 alone in predicting clinically significant prostate cancer (csPCa). Methods The present study retrospectively analyzed data from patients who underwent prostate magnetic resonance imaging(MRI) examinations from July 2022 to August 2023. All patients underwent T2-weighted imaging (T2WI), amide proton transfer (APT), diffusion-weighted imaging (DWI), and dynamic contrast-enhanced (DCE) MRI. Two radiologists independently examined the images. The independent samples t test or the Wilcoxon rank sum test was employed to investigate the statistical variance in the demographic and APT parameters of the two groups. We utilized receiver operating characteristic (ROC) curve analysis to assess the diagnostic accuracy of PI-RADS V2.1 and the combination model (APT-weighted signal values and PI-RADS V2.1). The comparison of the area under the curve (AUC)s were conducted using the Delong method. Results A total of 289 patients were eventually included in this study; 102 had csPCa, and 187 had either benign lesions or clinically insignificant prostate cancer (cisPCa). The APTmean, APTmax, and APTmin values were significantly different between the two groups in both the peripheral zone (PZ) and transition zone (TZ). The combined models were significantly more effective than the use of PI-RADS V2.1 alone for the whole gland and PZ, with areas under the curve (AUC)s of 0.874–0.883 compared to 0.803 and 0.885 compared to 0.798, respectively (P < 0.05). However, there was no substantial improvement in diagnostic accuracy when APT-weighted signal values were incorporated into PI-RADS V2.1 for the TZ, as the AUC increased from 0.791 to 0.865, with a P value of 0.202. Conclusion By incorporating APT-weighted signal values with PI-RADS V2.1, there was a notable improvement in the diagnostic accuracy of csPCa detection in both the whole gland and the PZ compared to PI-RADS V2.1 alone. However, there was no significant enhancement in terms of csPCa in TZ.