• Qingbin Wu, Pengju Chen, Chi Shu, Lin Chen, Zechuan Jin, Jun Huang, Xin Wang, Xue Li, Mingtian Wei, Tinghan Yang, Xiangbing Deng, Aiwen Wu, Yazhou He, and Ziqiang Wang.
• Colorectal Cancer Center, Department of General Surgery, West China Hospital, Sichuan University, Chengdu, China.
• Bmc Med. 2023 Jan 4; 21 (1): 33.

BackgroundApproximately 10% of stage I colorectal cancer (CRC) patients experience unfavorable clinical outcomes after surgery. However, little is known about the subset of stage I patients who are predisposed to high risk of recurrence or death. Previous evidence was limited by small sample sizes and lack of validation.MethodsWe aimed to identify early indicators and develop a risk stratification model to inform prognosis of stage I patients by employing two large prospective cohorts. Prognostic factors for stage II tumors, including T stage, number of nodes examined, preoperative carcinoma embryonic antigen (CEA), lymphovascular invasion, perineural invasion (PNI), and tumor grade were investigated in the discovery cohort, and significant findings were further validated in the other cohort. We adopted disease-free survival (DFS) as the primary outcome for maximum statistical power and recurrence rate and overall survival (OS) as secondary outcomes. Hazard ratios (HRs) were estimated from Cox proportional hazard models, which were subsequently utilized to develop a multivariable model to predict DFS. Predictive performance was assessed in relation to discrimination, calibration and net benefit.ResultsA total of 728 and 413 patients were included for discovery and validation. Overall, 6.7% and 4.1% of the patients developed recurrences during follow-up. We identified consistent significant effects of PNI and higher preoperative CEA on inferior DFS in both the discovery (PNI: HR = 4.26, 95% CI: 1.70-10.67, p = 0.002; CEA: HR = 1.46, 95% CI: 1.13-1.87, p = 0.003) and the validation analysis (PNI: HR = 3.31, 95% CI: 1.01-10.89, p = 0.049; CEA: HR = 1.58, 95% CI: 1.10-2.28, p = 0.014). They were also significantly associated with recurrence rate. Age at diagnosis was a prominent determinant of OS. A prediction model on DFS using Age at diagnosis, CEA, PNI, and number of LYmph nodes examined (ACEPLY) showed significant discriminative performance (C-index: 0.69, 95% CI:0.60-0.77) in the external validation cohort. Decision curve analysis demonstrated added clinical benefit of applying the model for risk stratification.ConclusionsPNI and preoperative CEA are useful indicators for inferior survival outcomes of stage I CRC. Identification of stage I patients at high risk of recurrence is feasible using the ACEPLY model, although the predictive performance is yet to be improved.© 2022. The Author(s).

13 keywords...

hide…