Rectal cancer patients receiving neoadjuvant chemoradiotherapy (NCRT) are currently classified using the same Tumor-Node-Metastasis staging system as those patients without NCRT. We determined whether the combination of tumor treatment response (TRG) and ypT stage more accurately assesses primary tumors in rectal cancer after NCRT. We analyzed data from 329 rectal cancer patients treated with NCRT followed by radical resection. Cox proportional hazards models were used to evaluate the effects of different staging parameters on disease-free survival (DFS). ypN stage and TRG were independently associated with 3-year DFS, but ypT stage was not. We developed a new modified T stage classification metric (M-TTRG) that categorized patients into 5 subgroups based on ypT stage and TRG, with weighting by β-coefficients from multivariate analyses. The incidence of patients developing local or distant recurrence increased with increasing M-TTRG level. All five M-TTRG classes correlated with 3-year DFS. Improvement was seen in the model with M-TTRG classification compared with ypT stage, based on area under the curve after computing receiver operating characteristic curves. Our modified ypTNM staging system significantly improved prediction of 3-year DFS. This suggests TRG could complement ypT stage, and we propose the new M-TTRG metric could be used to better classify NCRT-treated patients, thereby improving treatment and assessing prognosis. The M-TTRG metric might be applicable to other types of cancer.
Keywords: TNM staging system; neoadjuvant chemoradiotherapy; rectal cancer; tumor regression grade; ypT stage.