Combination of diffusion-weighted imaging and arterial spin labeling at 3.0 T for the clinical staging of nasopharyngeal carcinoma.

PURPOSE To probe the utility of diffusion-weighted imaging (DWI) and 3D arterial spin labeling (ASL) in assessing the clinical stage of nasopharyngeal carcinoma (NPC). MATERIALS AND METHODS This prospective study included sixty-five newly diagnosed NPC patients who underwent DWI and 3D ASL scans on a 3.0-T magnetic resonance imaging (MRI) system. The apparent diffusion coefficient (ADC) and the tumor blood flow (TBF) of NPC were measured. Tumors were classified as low or high T, N and American Joint Committee on Cancer (AJCC) stages. Student's t-test was used to evaluate the differences between tumors with low and high clinical stages. Pearson correlation analyses were performed to determine the correlation between MRI parameters and clinical stages. Receiver operating characteristic (ROC) curves were then used to evaluate diagnostic capability. RESULTS High T stage (T3/4) NPC showed significantly lower ADCmin (P = 0.000) and higher TBFmax (P = 0.003) and TBFmean (P = 0.008) values than low T stage (T1/2) NPC. High N stage (N2/3) NPC showed significantly lower ADCmin values (P = 0.023) than low N stage (N0/1) NPC. High AJCC stage (III/IV) NPC showed significantly lower ADCmin (P = 0.000) and higher TBFmax (P = 0.005) and TBFmean (P = 0.011) values than low AJCC stage (I/II) NPC. ADCmin values showed moderate negative correlations with T stage (r = -0.512, P = 0.000), N stage (r = -0.281, P = 0.023), and AJCC stage (r = -0.494, P = 0.000). TBFmax values showed moderate positive correlations with T stage (r = 0.369, P = 0.003) and AJCC stage (r = 0.346, P = 0.005). Compared with ADCmin and TBFmax alone, the combination of ADCmin and TBFmax improved the accuracy from 72.3% and 75.4% to 78.5%, respectively, for T staging, as well as from 72.3% and 69.2% to 83.1% for AJCC staging. CONCLUSIONS ADCmin and TBFmax values in patients with NPC could help evaluate clinical stages. ADCmin and TBFmax values combined could clearly improve the accuracy in the assessment of AJCC stage.