Purpose: This study evaluated the feasibility of magnetic resonance (MR) volumetry using a diffusion-weighted data set (V(DWI)) and compared it with conventional T2-weighted volumetry (V(C)) in patients affected by rectal cancer treated with chemoradiation therapy (CHRT).
Materials and methods: Fourteen patients with a biopsy diagnosis of rectal cancer underwent MR examination before and after CHRT. T2-weighted images were used to extrapolate V(C). A diffusion-weighted (DW) sequence was acquired [spin-echo diffusion-weighted echo-planar imaging (SE-DW-EPI)] with a b-value of 800 s/mm(2) and volume (V(DWI)) was calculated by semiautomatic segmentation of tumour hyperintensity. Two radiologists independently assessed volumes and analysed data in order to establish interobserver agreement and compare and correlate volumes to tumour regression grade (TRG), as evaluable at pathological examination of the surgical specimen.
Results: Interobserver agreement was 0.977 [(95% confidence interval (CI) 0.954-0.989) and 0.956 (95% CI 0.905-0.980) for V(C) and V(DWI) and 0.964 (95% CI 0.896-0.988) and 0.271 (95% CI-0.267 to 0.686) between V(C) and V(DWI) before and after CHRT. The correlation between TRG and V(C) and V(DWI) was, respectively, rho = 0.597 (p<0.05) and r(2)=0.156 (p=0.162) and rho=0.847 (p<0.001).
Conclusions: V(DWI) seems to be a promising tool for assessing response to CHRT in rectal cancer. Further studies on large series of patients are needed to refine the technique and evaluate its potential predictive value.