Comparison and combination of amide proton transfer magnetic resonance imaging and the apparent diffusion coefficient in differentiating the grades of prostate cancer

BACKGROUND: More effective risk stratification of prostate cancer (PCa) than that possible with current methods can reduce undertreatment and guard against overtreatment. The aim of this study is to validate the differences and combined effects of amide proton transfer (APT) imaging and apparent diffusion coefficient (ADC) in discriminating the PCa grade group (GG) ≤2 from GG ≥3 PCa. METHODS: This is an ongoing prospective study conducted in the radiology department of Nanxishan Hospital of Guangxi Zhuang Autonomous Region. Patients pathologically diagnosed with PCa were enrolled consecutively according to the eligibility criteria. A total of 180 patients (age range, 42–92 years) were included in this study. Using histopathology as the reference standard, we placed 71 cases in GG ≤2 (mean age 67.03±8.696 years) and 109 cases in GG ≥3 (age 69.60±9.638 years). Magnetic resonance imaging (MRI) parameters, including APT and ADC values, were analyzed using an independent samples t-test and binary logistic regression analysis stratified with GG. Receiver operating characteristic curve was used to analyze the diagnostic performance for different parameters distinguishing GG ≤2 and GG ≥3. RESULTS: APT [odds ratio (OR) for the transitional zone (TZ) PCa: 3.20, 95% CI: 1.14–8.98, P=0.02; OR for the peripheral zone (PZ) PCa: 86.32, 95% CI: 13.24–562.88, P=0.003] and ADC values (OR for TZ PCa: 89.79; 95% CI: 2.85–2,827.99, P=0.01; OR for PZ PCa: 39.92; 95% CI: 3.22–494.18, P=0.004) were independent predictors that differentiated the GG of patients. The sensitivity and specificity of the APT values were 61.1% and 81.0%, respectively, while the sensitivity and specificity of the ADC values were 83.3% and 61.9%, respectively. The optimal cutoff value of APT was 3.35% and which of ADC was 1.25×10(−3) mm(2)/s in TZ origin PCa. At the optimal cutoff values of 3.31% (APT) and 0.79×10(−3) mm(2)/s (ADC) in PZ PCa, the sensitivity and specificity of the APT values were 74.0% and 83.6%, respectively, while the sensitivity and specificity of the ADC values were 94.0% and 53.4%, respectively. The area under the curve of the combination of APT and ADC was significantly higher than either of APT or ADC alone in Delong test (TZ: P=0.002 and P=0.020; PZ: P=0.033 and P<0.001). CONCLUSIONS: APT and ADC have complementary effects on the sensitivity and specificity for identifying different PCa GGs. A combination model of APT and ADC could improve the diagnostic efficacy of PCa differentiation.