Magnetic resonance imaging in restaging rectal cancer after neoadjuvant chemoradiotherapy

To evaluate the accuracy of magnetic resonance imaging (MRI) for restaging locally advanced nonmucinous rectal cancer after neoadjuvant chemoradiotherapy (CRT).

A total of 94 consecutive patients with histologically proven locally advanced middle or low located nonmucinous rectal adenocarcinoma, who were treated with preoperative CRT, followed by radical surgery 6-8 weeks later, were analyzed in this retrospective study. Preoperative MR images were reinterpreted by one observer and the results were compared with the histologic findings. The overall MRI tumor (T) and nodal (N) restaging accuracy were calculated. The agreement between post-CRT MRI examination and histological assessment was evaluated by using kappa statistics.

The overall accuracy of MRI for T restaging was 49%, with overstaging and understaging occurring in 40.4% and 10.6% of the patients, respectively. Only 18% of the patients with pathological complete response (pCR) were staged correctly by MRI, nevertheless an excellent 100% specificity in predicting pCR was detected. For N restaging with MRI, the overall accuracy was 63.8%, whereas 26.6% of the patients were overstaged and 9.6% were understaged. Kappa statistics revealed poor concordance of MRI restaging after preoperative CRT and pathological results in both T (k=0.156) and N staging (k=0.289).

Restaging rectal cancer still remains a challenge and better methods are urgently required. The surgical plan before treatment should not be changed except in those cases who had pCR, intolerance or refusing radical operation, for whom an observation strategy could be taken into consideration after the excellent specificity in predicting pCR.