Multiparametric magnetic resonance imaging (MRI) integrating amide proton transfer-weighted (APTw) imaging, synthetic MRI, and multiplexed sensitivity encoding diffusion-weighted imaging (MUSE DWI) in the diagnosis and grading of prostate cancer.

BACKGROUND: Prostate cancer (PCa) is a malignant tumor that significantly threatens the quality of life of elderly men. This study evaluated the efficacy of multiparametric magnetic resonance imaging (MRI) combining synthetic magnetic resonance imaging (SyMRI), amide proton transfer-weighted (APTw) imaging, and multiplexed sensitivity encoding diffusion-weighted imaging (MUSE DWI) in the diagnosis and grading of PCa.

METHODS: A total of 195 patients with pathologically confirmed prostate lesions who underwent the SyMRI, APTw, and MUSE DWI were retrospectively analyzed. Quantitative parameters derived from SyMRI [T1, T2, proton density (PD)], APTw (APTmin, APTmax, APTmean), and apparent diffusion coefficient (ADC) from MUSE DWI were extracted by manually delineating regions of interest (ROIs). Group comparisons were performed using the independent t-test or Wilcoxon rank-sum test, as appropriate. Parametric variables were correlated with the International Society of Urological Pathology (ISUP) grades using the Spearman rank correlation coefficient. A one-way analysis of variance (ANOVA) test with Bonferroni correction or Kruskal-Wallis test was performed to determine the differences of each parameter among different ISUP grades. Receiver operating characteristic (ROC) curve analysis, DeLong test, and logistic regression modeling were conducted to assess the diagnostic performance of individual and combined parameters.

RESULTS: Totals of 96 malignant and 99 benign patients were included in this study. In the transitional zone (TZ) and peripheral zone (PZ) of the prostate, all quantitative parameters except PD value showed statistical differences for distinguishing benign and malignant prostate lesions (P<0.05). T1, T2, and ADC values were significantly lower, whereas APTmin, APTmax, and APTmean were significantly higher in PCa compared to benign lesions. For the diagnosis of PZ PCa, the area under the curve (AUC) of the multiparametric model [AUC =0.948; 95% confidence interval (CI): 0.878-0.979] showed no significant difference from ADC value (AUC =0.896; 95% CI: 0.784-0.957) (DeLong test: P=0.118). Compared with Prostate Imaging Reporting and Data System (PI-RADS), the multiparametric model can significantly enhance the diagnostic efficacy (DeLong test: P=0.002). For the diagnosis of TZ PCa, the AUC of the multiparametric model (AUC =0.846; 95% CI: 0.754-0.910) was significantly higher than that of the single parameters (DeLong test: P<0.05). Compared with PI-RADS, the multiparametric model can significantly enhance the diagnostic efficacy (DeLong test: P=0.049). APTmax and APTmean were significantly positively correlated with ISUP grade (r=0.419, P<0.001; r=0.311, P=0.002). T2 and ADC values were significantly negatively correlated with ISUP grade (r=-0.285, P=0.005; r=-0.495, P<0.001). Compared to PCa lesions with grade 1, ISUP grade 4 exhibited significantly higher APTmax and APTmean (P<0.05) but lower T2 value (P=0.048), whereas ADC value was significantly higher in ISUP grade 1 than in ISUP grade 2-5 (P<0.05).

CONCLUSIONS: Multiparametric MRI incorporating SyMRI, APTw imaging, and MUSE DWI may offer enhanced diagnostic performance in distinguishing benign from malignant prostate lesions and grading of PCa.