Evaluation and response of mucin pools in primary rectal cancer tissue during chemoradiotherapy (CRT).

Abstract

798 Background: Mucinous carcinoma (MC) has been defined as carcinoma with > 50% of tumor volume showing mucin pool (MP). The incidence of MC had been reported to be 4 to 12% of colorectal cancer and to be less effective for chemotherapy and CRT. The incidence of MC was 3 to 8% in rectal cancer treated with surgery alone and 13 to 24% in patients treated with CRT. These findings suggest that MP may be induced by CRT in some patients. We therefore compared the degree of MP in MRI findings before CRT, after CRT and in resected specimens. Methods: The study group comprised 205 patients with rectal cancer who received CRT (40 to 45 Gy) from 2006 through 2014. MPs on MRI were evaluated using T2-weighted fast-spin echo images, and the MP area ratios on maximum cross-section of the primary tumor were calculated. An MP area ratio of 0% was designated as class 0, 1 to 9% as class I, 10 to 29% as class II, 30 to 50% as class III, and 50% or higher as class IV. Results: On MRI before CRT, the MP area ratio was class 0 in 181 patients (88%), class I in 11 (5%), and class II or higher in 13 (6%). On MRI after CRT, the MP area ratio was class 0 in 186 patients (91%), class I in 4 (2%), and class II or higher in 15 (7%). The prevalence of MP classes did not change appreciably. The MP class evaluated on MRI before and after CRT changed by 2 or more levels in only 2 patients, both of whom had marked tumor shrinkage. Among 175 patients with an MP area ratio of class 0 or I in the resected specimens, the MRI findings after CRT were consistent with the findings of the resected specimens. However, the MP area ratio was underestimated by 2 or more levels on MRI in 14 (47%) of 30 patients who had an MP area ratio of class II or higher. A pathological complete response (pCR) was obtained in 24 patients (12%), all of whom had an MP area ratio of class 0 or I on MRI before CRT. No patient with an MP area ratio of class II or higher before CRT had a pCR. Conclusions: The degree of MP evaluated on MRI did not change during CRT. In patients who had an MP area ratio of > 10% in their resected specimens, the evaluation of MP on MRI was often underestimated. Patients with an MP area ratio of > 10% before CRT had a poor response to CRT, with no pCR.