Quantitative Dynamic-Enhanced MRI and Intravoxel Incoherent Motion Diffusion−Weighted Imaging for Prediction of the Pathological Response to Neoadjuvant Chemotherapy and the Prognosis in Locally Advanced Gastric Cancer

BackgroundThis study aimed to explore the predictive value of quantitative dynamic contrast-enhanced MRI (DCE-MRI) and intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) quantitative parameters for the response to neoadjuvant chemotherapy (NCT) in locally advanced gastric cancer (LAGC) patients, and the relationship between the prediction results and patients’ prognosis, so as to provide a basis for clinical individualized precision treatment.MethodsOne hundred twenty-nine newly diagnosed LAGC patients who underwent IVIM-DWI and DCE-MRI pretreatment were enrolled in this study. Pathological tumor regression grade (TRG) served as the reference standard of NCT response evaluation. The differences in DCE-MRI and IVIM-DWI parameters between pathological responders (pR) and pathological non-responders (pNR) groups were analyzed. Univariate and multivariate logistic regressions were used to identify independent predictive parameters for NCT response. Prediction models were built with statistically significant quantitative parameters and their combinations. The performance of these quantitative parameters and models was evaluated using receiver operating characteristic (ROC) analysis. Clinicopathological variables, DCE-MRI and IVIM-DWI derived parameters, as well as the prediction model were analyzed in relation to 2-year recurrence-free survival (RFS) by using Cox proportional hazards model. RFS was compared using the Kaplan–Meier method and the log-rank test.ResultsSixty-nine patients were classified as pR and 60 were pNR. Ktrans, kep, and ve values in the pR group were significantly higher, while ADCstandard and D values were significantly lower than those in the pNR group. Multivariate logistic regression analysis demonstrated that Ktrans, kep, ve, and D values were independent predictors for NCT response. The combined predictive model, which consisted of DCE-MRI and IVIM-DWI, showed the best prediction performance with an area under the curve (AUC) of 0.922. Multivariate Cox regression analysis showed that ypStage III and NCT response predicted by the IVIM-DWI model were independent predictors of poor RFS. The IVIM-DWI model could significantly stratify median RFS (52 vs. 15 months) and 2-year RFS rate (72.3% vs. 21.8%) of LAGC.ConclusionPretreatment DCE-MRI quantitative parameters Ktrans, kep, ve, and IVIM-DWI parameter D value were independent predictors of NCT response for LAGC patients. The regression model based on baseline DCE-MRI, IVIM-DWI, and their combination could help RFS stratification of LAGC patients.