| BMC Cancer | |
| A nomogram to predict lymph node metastasis risk for early esophageal squamous cell carcinoma | |
| Xiaofeng Duan1  Xiaobin Shang1  Peng Tang1  Hongjing Jiang1  Zhao Ma1  Chuangui Chen1  Jie Yue1  Zhentao Yu1  | |
| [1] Department of Esophageal Surgery, Tianjin Medical University Cancer Hospital and Institute, Key Laboratory of Cancer Prevention and Therapy, National Clinical Research Center for Cancer; | |
| 关键词: Esophagus; Cancer; Lymph nodes; Metastasis; Nomogram model; | |
| DOI : 10.1186/s12885-021-08077-z | |
| 来源: DOAJ | |
【 摘 要 】
Abstract Background A nomogram was developed to predict lymph node metastasis (LNM) for patients with early-stage esophageal squamous cell carcinoma (ESCC). Methods We used the clinical data of ESCC patients with pathological T1 stage disease who underwent surgery from January 2011 to June 2018 to develop a nomogram model. Multivariable logistic regression was used to confirm the risk factors for variable selection. The risk of LNM was stratified based on the nomogram model. The nomogram was validated by an independent cohort which included early ESCC patients underwent esophagectomy between July 2018 and December 2019. Results Of the 223 patients, 36 (16.1%) patients had LNM. The following three variables were confirmed as LNM risk factors and were included in the nomogram model: tumor differentiation (odds ratio [OR] = 3.776, 95% confidence interval [CI] 1.515–9.360, p = 0.004), depth of tumor invasion (OR = 3.124, 95% CI 1.146–8.511, p = 0.026), and tumor size (OR = 2.420, 95% CI 1.070–5.473, p = 0.034). The C-index was 0.810 (95% CI 0.742–0.895) in the derivation cohort (223 patients) and 0.830 (95% CI 0.763–0.902) in the validation cohort (80 patients). Conclusions A validated nomogram can predict the risk of LNM via risk stratification. It could be used to assist in the decision-making process to determine which patients should undergo esophagectomy and for which patients with a low risk of LNM, curative endoscopic resection would be sufficient.
【 授权许可】
Unknown