OBJECTIVETo investigate the value of MRI with CUBE sequence in early evaluation of the efficacy of neoadjuvant therapy (NAT) for locally advanced rectal cancer.

METHODSInclusion criteria: (1) rectal cancer proven by biopsy; (2) locally advanced rectal cancer (T3-4 or positive lymph nodes) with distance from lower edge of tumor to anal verge within 12 cm diagnosed by MRI before NAT; (3) acceptance of NAT treatment regulations and completion of NAT; (4) completion of routine MRI scan and CUBE scan before and after 2-course NAT chemotherapy (including new neoadjuvant chemotherapy and chemoradiotherapy); (5) completion of surgery 6-8 weeks after NAT; (6) exclusion of any previous NAT due to rectal malignant tumor or other tumors; (7) exclusion of poor image quality of preoperative routine MRI insufficient for rectal cancer staging or of CUBE image insufficient for tumor volume measurement. Fifty patients with advanced local rectal cancer were recruited in our hospital from February 2014 to January 2016. All the patients underwent MRI before and after 2-course neoadjuvent therapy. Tumor volume on CUBE were measured and the volume changes were calculated: volume difference= pre-treatment volume-post-treatment volume, volume change rate= (pre-treatment volume-post-treatment volume)/ pre-treatment volume. All the patients were categorized into sensitive and non-sensitive group according to postsurgical pathology. Comparisons were made between both groups before and after therapy. ROC curve was used to evaluate the value of CUBE-associated parameters in predicting the efficacy of rectal cancer.

RESULTSAmong enrolled 50 patients with rectal cancer, 31 were male and 19 were female, with mean age of 49.1 years (range 21 to 70 years). T-staging by MRI before NAT was T2N1-2 in 1 case, T3 in 43 cases, T4 in 6 cases. The number of patients after NAT from tumor regression grading (TRG) 0 to TRG3 was 14, 13, 18, 5, respectively. The sensitive group and insensitive group were 45 cases and 5 cases. Mean tumor volume before and after preoperative 2-course NAT was 18.70 (4.14 to 91.77) cmand 9.26 (1.02 to 52.58) cm, respectively, whose difference was significant (U=-5.826, P<0.001). Both measure values of overall tumor volume before and after preoperative NAT between sensitive group and insensitive group did not show significant differences(all P>0.05). While sensitivity group had significantly higher volume difference and change rate compared to insensitive group [ (11.90±10.01) cmvs. (0.65±3.93) cm, P=0.005; 0.45±0.28 vs. 0.09±0.36, P=0.010]. ROC curve revealed that the optimal cutoff value of accurate identification of patients with NAT sensitive was 1.96 cmfor volume difference with sensitivity 86.7% and specificity 80%, and 0.06 for volume change rate with sensitivity 93.3% and specificity 60%.

CONCLUSIONMRI CUBE can predict the efficacy of NAT for early rectal cancer patients accurately and sensitively through the detection of tumor volume change before and after NAT.