Early prediction value of ASL-MRI for induced chemotherapy response and short-term efficacy of locally advanced nasopharyngeal carcinoma

Objective To monitor tumor blood flow (TBF) before and after induction chemotherapy in locally advanced nasopharyngeal carcinoma (LA-NPC) by arterial spin labeling magnetic resonance imaging (ASL-MRI), and to investigate the value of ASL-MRI in early predicting response of induction chemotherapy and short-term efficacy in LA-NPC. Methods Thirty-eight newly diagnosed LA-NPC patients were enrolled. ASL-MRI were performed both before and after induction chemotherapy to obtain Pre-TBF and Post-TBF. The TBF change value (ΔTBF) and the change rate (ΔTBF%) before and after induction chemotherapy were calculated. After induction chemotherapy, complete response (CR) and partial response (PR) were classified as response group (RG), and stable disease (SD) as well as progressive disease (PD) were classified as non-response group (NRG). The short-term efficacy was evaluated at 3 months after radiotherapy and divided into CR group and non-CR group (PR, SD and PD). Univariate and multivariate binary logistic regression analysis was used to evaluate the effects of TBF parameters on the response of induction chemotherapy and short-term efficacy. The receiver operating characteristic (ROC) curve was used to evaluate the diagnostic efficacy. Results Of the 38 patients, 23 (60.5%) were in the RG and 15 (39.5%) in the NRG. There were 22 cases (57.9%) in the CR group and 16 cases (42.1%) in the non-CR group at 3 months after radiotherapy. The CR rate at 3 months after radiotherapy was significantly higher in the RG than that in the NRG (73.9% vs. 33.3%, P=0.02). Pre-TBF was significantly higher than Post-TBF in 38 patients (Z＝4.227, P < 0.001). Pre-TBF, ΔTBF and ΔTBF% in the RG were significantly higher than that in the NRG (all P < 0.05); Pre-TBF, ΔTBF, and ΔTBF% were significantly higher in the CR group than that in the non-CR group at 3 months after radiotherapy (all P < 0.05). Multivariate binary logistic regression showed that Pre-TBF was an independent risk factor of the efficacy of induction chemotherapy (P=0.027), with the area under ROC curve (AUC) value of 0.745 (P=0.012). T stage and ΔTBF% were independent risk factors of short-term efficacy (all P < 0.05), and the AUC value of ΔTBF% was 0.807 (P=0.001). Conclusion Pre-treatment TBF hyperperfusion in LA-NPC suggests better outcomes. Pre-TBF and ΔTBF% can predict the efficacy of induction chemotherapy and short-term efficacy of LA-NPC, respectively.