Associations between clinical characteristics and tumor response to neoadjuvant chemoradiotherapy in rectal cancer

Abstract

Following standard neoadjuvant chemoradiotherapy and total mesorectal excision, some patients with locally advanced rectal cancer achieve good response (pathological T0–2N0), while others show nonresponse (pathological T3–4N0 or node-positive). To date, the clinicopathological predictors of good response and the necessity of adjuvant chemotherapy treatment (ACT) in good responders remain unclear. In this retrospective study, clinicopathological characteristics were surveyed to investigate the correlation with good response; furthermore, a propensity score matching (PSM) model was designed to balance the confounding factors between good responders treated with ACT or observation. A total of 2255 patients were enrolled, including 1069 good responders and 1186 nonresponders. The results of the survival analysis showed a good response predicted a better 3-year prognosis (p < 0.001). The logistic regression analysis showed less advanced T and N stages (T3 vs. T4; N0 vs. N1–2), more neoadjuvant chemotherapy (nCT) cycles (≥4 vs. 1–3), and delayed surgery (≥8 weeks vs. <8 weeks) were independent predictors of a good response (p < 0.05). Especially, patients treated with both more nCT cycles and a delay in surgery included the greatest number of good responders (p < 0.001). For good responders, after PSM (1:3), 235 observation cases were matched to 705 ACT cases. As compared with observation, ACT had no greater impact on prognosis analysis (p > 0.05). In conclusion, more cycles of nCT and a delay in surgery predicted a better response, and the delivery of ACT might be omitted in good responders.