Radiographic predictors of response to endoluminal brachytherapy for the treatment of rectal cancer

Objective Endoluminal brachytherapy (EBT) has been shown to be an effective neoadjuvant monotherapy for rectal adenocarcinoma. Radiographic predictors of response are used to guide treatment in rectal cancer; however, no predictors of response to EBT have been identified. Methods This is a single-institutional prospective study from 2010 to 2013. Analysis included 17 patients undergoing EBT and 13 patients undergoing conventional external beam chemoradiation (CRT). Clinical response to therapy was assessed with serial MRI and PET/CT variables. Pathological response to therapy was assessed using tumor regression grade (TRG) and compared with clinical response. Results EBT and CRT patients did not differ with respect to age, sex, race, carcinoembryonic antigen, or clinical stage of disease. There was a similar rate of pathologic complete response for both groups, with a trend towards more TRG 0 with EBT compared with CRT (35.3% vs. 7.7%, p = 0.08). Four days of EBT resulted in a significantly greater reduction in tumor volume on MRI than did CRT (92.7% vs. 63.1%, p = 0.004). Using receiver operating characteristic analysis, change in peak standardized uptake value was the best predictor for complete pathologic response in EBT patients (sensitivity 67%, specificity 82%). False-positive findings for nodal disease on MRI were seen in 59% of EBT patients and 23% of CRT patients. Conclusions This study demonstrates that EBT is an effective alternative with similar response rates to CRT. However, the inability of MRI and PET/CT to discern reactive from malignant tumor and nodes may pose limitations in their use for guiding further therapy. Larger randomized studies are needed.