Abstract: Objective To explore the predictive value of intravoxel incoherent motion (IVIM) imaging for the pathologic response to neoadjuvant chemotherapy in locally advanced esophageal squamous cell carcinoma (ESCC). Methods The prospective study was conducted. The clinicopathological data of 33 patients with locally advanced ESCC who were admitted to Affiliated Hospital of Zhengzhou University from September 2015 to October 2017 were collected. Patients received magnetic resonance imaging (MRI) and IVIM imaging examination before and after neoadjuvant chemotherapy. Two radiologists read the imaging together, manually delineated the region of interest in the diffusion-weighted imaging, and the apparent diffusion coefficient (ADC), diffusion coefficient (D), perfusion coefficient (D*), and perfusion score of the tumor (f) were automatically measured. Patients underwent neoadjuvant chemotherapy with paclitaxel plus cisplatin, and underwent radical surgery for esophageal cancer after 2 cycles of chemotherapy. Observation indicators: (1) comparison of IVIM imaging parameters before and after neoadjuvant chemotherapy in patients with ESCC; (2) comparison of change value and change rate of IVIM imaging parameters before and after neoadjuvant chemotherapy in patients with different tumor regression grade (TRG); (3) predictive efficacy of change value and change rate of IVIM imaging parameters before and after neoadjuvant chemotherapy for TRG. Measurement data with normal distribution were presented as Mean±SD, and comparison before and after neoadjuvant chemotherapy was done using the paired t test, and comparison between different TRG patients was done using the t test. Measurement data with skewed distribution were presented as M(P25, P75), and comparison before and after neoadjuvant chemotherapy and between different TRG patients were done using the Wilcoxon rank sum test. The receiver operating characteristic (ROC) curve was used to evaluate predictive value of IVIM imaging parameters. Results Thirty-three patients were screened for eligibility, including 26 males and 7 females, aged from 44 to 74 years, with an average age of 60 years. All the 33 patients were diagnosed as ESCC by pathological examination. (1) Comparison of IVIM parameters before and after neoadjuvant chemotherapy in patients with ESCC: 33 patients with ESCC showed a significant difference in the ADC, D, and f value after neoadjuvant chemotherapy ［ADC: (1.95±0.56)×10-3 mm2/s vs. (2.54±0.50)×10-3 mm2/s, t=-6.98; D: (1.26×10-3 mm2/s (0.81×10-3 mm2/s, 2.44×10-3 mm2/s) vs. 1.68×10-3 mm2/s (0.83×10-3 mm2/s, 2.27×10-3 mm2/s), Z=-3.96; f: 0.33%±0.14% vs. 0.42%±0.15%, t=-3.13, P<0.05］. (2) Comparison of change value and change rate of IVIM imaging parameters before and after neoadjuvant chemotherapy in different TRG patients: of 33 patients, 15 were in TRG 2 and 18 were in TRG 3. The ADC change value, ADC change rate, D change value, D change rate were (0.85±0.52)×10-3 mm2/s, 52.91%±32.51%, 0.64×10-3 mm2/s (0.05×10-3 mm2/s, 1.41×10-3 mm2/s) , 48.20%(3.03%,16.95%) of TRG 2 patients, and (0.38±0.35)×10-3 mm2/s, 21.94%±19.08%, 0.26×10-3 mm2/s (-1.43×10-3 mm2/s, 0.81×10-3 mm2/s), 20.18%(-58.61%,77.14%) of TRG 3 patients, respectively, with significant differences between two groups (t=3.09, 3.41, Z=-3.04,-2.93, P<0.05). (3) Predictive efficacy of change value and change rate of IVIM imaging parameters before and after neoadjuvant chemotherapy for TRG: ROC curve analysis showed that ADC change value exhibited an area under curve (AUC) of 0.798, a sensitivity of 66.7% and a specificity of 94.4% in predicting TRG, when 0.86×10-3 mm2/s was used as the cut-off value. With 43.3% as the cut-off value, ADC change rate had an AUC of 0.793, a sensitivity of 66.7% and a specificity of 88.9% in predicting TRG. With 0.35×10-3 mm2/s as the cut-off value, D change value had an AUC of 0.809, a sensitivity of 73.3% and a specificity of 77.8% in predicting TRG. With 25.9% as the cut-off value, D change rate had an AUC of 0.800, a sensitivity of 80.0% and a specificity of 72.2% in predicting TRG. ConclusionsThe change value and change rate of ADC and D values before and after neoadjuvant chemotherapy are potential predictors of pathologic response in ESCC. The significantly increased ADC and D values after neoadjuvant chemotherapy are prone to good pathologic response. The change value and change rate of D values show a better predictive value for pathologic response to neoadjuvant chemotherapy in ESCC compared with those of ADC values.