Xin Wang¹, Jing Jin², Yong Yang², Wenyang Liu², Hua Ren², Yanru Feng², Qin Xiao³, Ning Li², Hui Fang², Hao Jing², Ningning Lu², Yu Tang², Lei Deng², Jian-yang Wang², Shulian Wang², Weihu Wang², Yong-wen Song², Yue-ping Liu², Xin-fan Liu², Zihao Yu², Yexiong Li²

¹Department of RadiationOncology, Cancer Hospital and Institute, National Cancer Center, ChineseAcademy of Medical Sciences (CAMS), ²National Cancer Center,Cancer Hospital and Institute, ³The Affiliated CancerHospital of Xiangya School of Medicine, Hunan Cancer Hospital

Objective: Whetheradjuvant chemoradiotherapy (ACRT) or chemotherapy (ACT) will benefit for stageII or III upper rectal cancer (URC) after radical surgery remains unclear. Theaim of this study was to develop t a clinical nomogram and a recursivepartitioning analysis (RPA) based risk stratification system for the predictionof 5-year cancer-specific survival (CSS), trying to reveal whether ACRT or ACTis necessary for these patients. Method: This retrospective analysisincluded 547 primary URC patients who were extracted from 3995 rectal cancerpatients admitted in Cancer Hospital, Chinese Academy of Medical Sciences fromJanuary 2000 to December 2010. Lower margin of the tumor located 10-400px fromthe anal verge determined by colonoscope, underwent radical resection (R0) andpathologically staged as II/III were included. A nomogram was developed basedon the Cox regression model predicting 5-year CSS and RPA stratified patientsinto risk groups based on their tumor characteristics. Kaplan-Meier curves wereused to estimate CSS rates in each risk group according to ACRT or ACT status. Result:With a median follow-up of 68 months (range 4.6~182.5), the actually 5-yearoverall survival, disease-free survival, CSS, local recurrence free survivaland distant metastasis free survival of the entire cohort were 79.7%, 76.1%,83.3% and 93.7% and 79.8%, respectively. Of the 547 patients, 379 (69.3%)received ACRT or adjuvant radiotherapy and 327 (59.9%) patients received ACT.Five independent prognostic factors including age, preoperative increasedlevels of CEA and CA19-9, number of positive lymph nodes (PLNs), tumor deposit(TD), pathologic T category were identified and entered into nomogram thatpredicts the probability of 5-year CSS. The bootstrap-corrected c-index was0.757. When patients were divided into three risk groups on the basis of RPA,Only the high-risk group (patients with PLN ≤ 6 and TD, or PLN>6) benefitedfrom ACRT plus ACT when compared with surgery followed by ACRT or ACT orsurgery alone (5-year CSS: 70.8% vs. 57.8% vs. 15.6%, P<0.001). While, ACRT and ACT following surgery achieved nosignificant difference in CSS outcome in the intermediate- (P=0.159) andlow-risk groups (P=0.576). Conclusion: The outcome of URC is goodfollowing radical surgery. Our study indicates that ACRT plus ACT post-surgerymay be an important treatment plan with potentially significant survival advantagesin high-risk URC.

Key Words: upper rectal cancer  adjuvant chemoradiotherapy