The value of diffusion kurtosis magnetic resonance imaging for assessing treatment response of neoadjuvant chemoradiotherapy in locally advanced rectal cancer.

Abstract

To evaluate the feasibility and value of diffusion kurtosis (DK) imaging in assessing treatment response to neoadjuvant chemoradiotherapy (CRT) in patients with locally advanced rectal cancer (LARC). Forty-one patients were included. All patients underwent pre- and post-CRT DCE-MRI on a 3.0-Tesla MRI scanner. Imaging indices (D _app , K _app and ADC values) were measured. Change value (∆X) and change ratio (r∆X) were calculated. Pathological tumour regression grade scores (Mandard) were the standard reference (good responders: pTRG 1-2; poor responders: pTRG 3-5). Diagnostic performance was compared using ROC analysis. For the pre-CRT measurements, pre-D _app-10th was significantly lower in the good responder group than that of the poor responder group (p = 0.036). For assessing treatment response to neoadjuvant CRT, pre-D _app-10th resulted in AUCs of 0.753 (p = 0.036) with a sensitivity of 66.67 % and a specificity of 77.78 %. The r∆D _app had a relatively high AUC (0.859) and high sensitivity (100 %) compared with other image indices. DKI is feasible for selecting good responders for neoadjuvant CRT for LARC. • LARC responded well after neoadjuvant chemoradiotherapy with lower pre-D _app-10th . • LARC responded well with greater increases in mean ADC and D _app . • The change ratio of D _app (r∆D _app ) had a relatively better diagnostic performance.