Baseline virtual MR elastography and extracellular volume fraction in the prediction of response to neoadjuvant chemotherapy in gastric cancer

Motivation: Although neoadjuvant chemotherapy (NAC) was recommended for gastric cancer (GC), only about 40% patients could achieve pathological response. Accurate prediction of the NAC response is crucial for patients’ benefits. Goal(s): To evaluate the predictive performance of virtual MR elastography (vMRE) and extracellular volume (ECV) for predicting the response to of NAC in GC patients. Approach: Patients underwent DWI-based elastography, pre- and post-contrast T1 mapping before treatment. DWI-based virtual shear modulus (μDiff) and ECV were calculated from DWI and T1 mapping. Results: Both μDiff (AUC: 0.833) and ECV (AUC: 0.794) were independent predictors for NAC response, their combination could improve the AUC to 0.968. Impact: Our result revealed that DWI-based virtual shear modulus (μDiff) and ECV exhibited a promising predictive ability for predicting response to NAC. This would aid in identifying responders before treatment, reducing unnecessary toxicity and side effects, and guiding individualized treatment.