Factors associated with pathological complete response and construction of scoring system in patients with rectal cancer after neoadjuvant chemoradiotherapy / 中华普通外科杂志

ABSTRACT

Objective:To identify the clinical factors associated with pathological complete response (pCR) after neoadjuvant chemoradiotherapy (nCRT) in locally advanced middle and low rectal cancer and establish a scoring system.Methods:In this retrospective analysis the clinical data of patients with locally advanced middle and low rectal cancer treated with nCRT combined with surgery at Union Hospital of Tongji Medical College, Huazhong University of Science and Technology from Jan 2016 to Jan 2020 were studied. Patients were divided into pCR group and non-pCR group. Single factor analysis and Logistic multivariate regression analysis were performed to explore pCR related factors after nCRT, and a pCR prediction scoring system was established.Results:The pCR was achieved in 33 patients (20.8%). Univariate analysis showed that the maximum thickness of the tumor≤25mm before nCRT ( P=0.046), concurrent oxaliplatin-combined intensive chemotherapy ( P=0.013), the NLR≤1.65 before nCRT ( P=0.004) and the serum CEA≤5 ng/ml before nCRT ( P=0.016) were significantly associated with pCR. In multivariate analysis, concurrent oxaliplatin-combined intensive chemotherapy, the NLR before nCRT and serum CEA before nCRT were independent related factors of pCR. The probability of pCR for patients with score of 0, 1, 2, and 3 was 42% (10/24), 30% (19/63), 5% (3/57) and 7% (1/15), respectively. The probability of pCR in patients with score≤1 point was 33% (29/87), and 6% (4/72) for score?1 point ( P?0.001). The area under the curve of the scoring system is 0.729 (95% CI 0.638-0.820, P?0.001). Conclusions:Concurrent oxaliplatin-combined intensive chemotherapy, NLR≤1.65 before nCRT and serum CEA≤5 ng/ml before nCRT are independent predictors of pCR in locally advanced middle and low rectal cancer and the scoring system constructed in combination with above indicators can effectively predict pCR.