T1期和T2期食管鳞癌根治术后淋巴结转移和预后影响因素分析及列线图预测模型构建

Influencing factors for lymph node metastasis and prognosis in stage T1 and T2 esophageal squa-mous cell carcinoma after radical surgery and construction of nomogram prediction models

  • 摘要:
    目的 探讨T1期和T2期食管鳞癌根治术后淋巴结转移和预后影响因素及构建列线图预测模型。
    方法 采用回顾性队列研究方法。收集2014年1月至2019年12月川北医学院附属医院收治的672例T1期和T2期食管鳞癌患者的临床病理资料;男464例,女208例;年龄为(65±8)岁。患者均行根治性食管癌切除+二野或三野淋巴结清扫术。观察指标:(1)淋巴结清扫、转移及随访情况。(2)食管鳞癌根治术后淋巴结转移影响因素分析。(3)食管鳞癌根治术后预后影响因素分析。(4)食管鳞癌根治术后淋巴结转移及预后预测模型构建及评价。采用门诊、电话和网络问诊等方式进行随访,了解患者生存情况。随访时间截至2021年4月。正态分布的计量资料以x±s表示,组间比较采用t检验;偏态分布的计量资料以M(范围)表示。计数资料以绝对数或百分比表示,组间比较采用χ²检验。采用Kaplan‑Meier法计算生存率和绘制生存曲线。采用Log‑Rank检验进行生存分析。食管鳞癌根治术后淋巴结转移的单因素和多因素分析采用Logistic回归模型,食管鳞癌根治术后预后的单因素和多因素分析采用COX回归模型。根据多因素分析结果构建食管鳞癌根治术后淋巴结转移和预后列线图预测模型,以受试者工作特征曲线(ROC)的曲线下面积(AUC)评价列线图预测模型的区分度,以校准曲线评价列线图预测模型的一致性。
    结果 (1)淋巴结清扫、转移及随访情况。672例患者淋巴结清扫数目为(14±8)枚,淋巴结转移数目为2(1~19)枚。672例患者中,182例发生淋巴结转移,其中T1期和T2期分别为58例和124例。672例患者均获得术后随访,随访时间为38(1~85)个月。672例患者平均生存时间为65个月,1、3、5年总生存率分别为89.0%、74.3%、66.0%。325例T1期和347例T2期患者平均生存时间分别为70个月和61个月;1、3、5年总生存率分别为95.0%、83.5%、73.4%和87.4%、69.9%、59.2%;两者生存情况比较,差异有统计学意义(χ2=14.51,P<0.05)。(2)食管鳞癌根治术后淋巴结转移影响因素分析。单因素分析结果显示:肿瘤位置、肿瘤组织学分级、肿瘤T分期是影响食管鳞癌根治术后淋巴结转移的相关因素(优势比=1.40,1.54,2.56,95%可信区间为1.07~1.85,1.20~1.99,1.79~3.67,P<0.05)。多因素分析结果显示:肿瘤位置、肿瘤组织学分级、肿瘤T分期是食管鳞癌根治术后淋巴结转移的独立影响因素(优势比=1.42,1.61,2.63,95%可信区间为1.07~1.89,1.25~2.09,1.82~3.78,P<0.05)。(3)食管鳞癌根治术后预后影响因素分析。单因素分析结果显示:术前合并症,术后并发症,肿瘤组织学分级为G3,肿瘤T分期,肿瘤N分期为N1期、N2期、N3期,肿瘤TNM分期为Ⅲ期、Ⅳ期是影响食管鳞癌根治术后预后的相关因素(风险比=1.48,1.64,2.23,1.85,2.09,4.48,4.97,3.54,5.53,95%可信区间为1.08~2.03,1.20~2.23,1.47~3.39,1.34~2.54,1.44~3.04,2.89~6.95,1.57~15.73,2.48~5.05,1.73~17.68,P<0.05)。多因素分析结果显示:术前有合并症,肿瘤组织学分级为G3,肿瘤T分期为T2期,肿瘤N分期为N1期、N2期、N3期是影响食管鳞癌根治术后预后的独立危险因素(风险比=1.57,1.89,1.63,1.71,3.72,3.90,95%可信区间为1.14~2.16,1.23~2.91,1.17~2.26,1.16~2.51,2.37~5.83,1.22~12.45,P<0.05)。(4)食管鳞癌根治术后淋巴结转移及预后预测模型构建及评价。根据多因素分析结果,应用肿瘤位置、肿瘤组织学分级、肿瘤T分期构建食管鳞癌根治术后淋巴结转移列线图预测模型,其得分分别为82、100、100分,得分总和对应淋巴结转移率;应用术前合并症、肿瘤组织学分级、肿瘤T分期、肿瘤N分期构建食管鳞癌根治术后1、3、5年总生存率列线图预测模型,其得分分别为23、38、27、100分,得分总和对应1、3、5年总生存率。ROC结果显示:食管鳞癌根治术后淋巴结转移列线图预测模型AUC为0.66(95%可信区间为0.62~0.71,P<0.05)。1、3、5年总生存率列线图预测模型AUC分别为0.73、0.74、0.71(95%可信区间分别为0.66~0.80、0.68~0.79、0.65~0.78,P<0.05)。校准曲线结果显示:列线图模型预测的淋巴结转移率和1、3、5年总生存率与实际淋巴结转移率及1、3、5年总生存率具有较好一致性。
    结论 肿瘤位置、肿瘤组织学分级、肿瘤T分期是T1期和T2期食管鳞癌根治术后淋巴结转移的独立影响因素,其列线图预测模型可预测患者术后淋巴结转移率。术前有合并症,肿瘤组织学分级为G3,肿瘤T分期为T2期,肿瘤N分期为N1期、N2期、N3期是T1期和T2期食管鳞癌根治术后预后的独立危险因素,其列线图预测模型可预测患者术后总生存率。

     

    Abstract:
    Objective To investigate the influencing factors for lymph node metastasis and prognosis in stage T1 and T2 esophageal squamous cell carcinoma after radical surgery and construct nomogram prediction models.
    Methods The retrospective cohort study was conducted. The clinico-pathological data of 672 patients with T1 and T2 esophageal squamous cell carcinoma who were admitted to the Affiliated Hospital of North Sichuan Medical College from January 2014 to December 2019 were collected. There were 464 males and 208 females, aged (65±8)years. All patients under-went radical esophagectomy+2 or 3 field lymph node dissection. Observation indicators: (1) lymph node dissection, metastasis and follow‑up. (2) risk factors for lymph node metastasis of esophageal cancer after radical resection. (3) prognostic factors of esophageal cancer after radical resection. (4) construction and evaluation of the prediction models of lymph node metastasis and prognosis of esophageal cancer after radical resection. Follow‑up was conducted using outpatient examination, telephone and internet consultations to detect survival of patients up to April 2021. Measurement data with normal distribution were represented as Mean±SD, and comparison between groups was conducted using the t test. Measurement data with skewed distribution were represented as M(range). Count data were described as absolute numbers or percentages, and comparison between groups was conducted using the chi‑square test. Kaplan‑Meier method was used to calculate survival rate and draw survival curve. Log‑Rank test was used for survival analysis. Logistic regression model was used for univariate and multivariate analyses of risk for lymph node metastasis, and COX regression model was used for univariate and multivariate analyses of prognosis. Based on the results of multi-variate analysis, the nomogram prediction models for lymph node metastasis and prognosis predic-tion were constructed. The prediction discrimination of the nomogram models were evaluated using the area under curve (AUC) of the receiver operating characteristic curve (ROC). The calibration curve was used to evaluate the prediction consistency of the models.
    Results (1) Lymph node dissection, metastasis and follow‑up. The number of lymph node dissected was 14±8 and the number of lymph node metastasis was 2(range, 1‒19) in 672 patients. Of the 672 patients, there were 182 cases had lymph node metastasis, including 58 cases in T1 stage and 124 cases in T2 stage. All 672 patients were followed up for 38 (range, 1‒85)months. The average overall survival time of 672 patients was 65 months, with the 1‑, 3‑, 5‑year overall survival rate as 89.0%, 74.3%, 66.0%, respectively. The average overall survival time of 325 patients in T1 stage and 347 patients in T2 stage were 70 months and 61 months. The 1‑, 3‑, 5‑year overall survival rate of 325 patients in T1 stage and 347 patients in T2 stage were 95.0%, 83.5%, 73.4% and 87.4%, 69.9%, 59.2%, respectively, showing a significant difference in survival between them (χ²=14.51, P<0.05). (2) Risk factors for lymph node metastasis of esophageal cancer after radical resection. Results of univariate analysis showed that tumor location, tumor histological grade, tumor T staging were related factors affecting lymph node metastasis of esophageal cancer after radical resection (odds ratio=1.40, 1.54, 2.56, 95% confidence interval as 1.07‒1.85, 1.20‒1.99, 1.79-3.67, P<0.05). Results of multivariate analysis showed that tumor location, tumor histological grade, tumor T staging were independent factors affecting lymph node metastasis (odds ratio=1.42, 1.61, 2.63, 95% confidence interval as 1.07‒1.89, 1.25‒2.09, 1.82‒3.78, P<0.05). (3) Prognostic factors of esophageal cancer after radical resection. Results of univariate analysis showed that preoperative comorbidities, postoperative complications, tumor histological grade (G3), tumor T staging, tumor N staging (N1 stage, N2 stage, N3 stage), tumor TNM staging (Ⅲ stage, Ⅳ stage) were related factors affecting prognosis of esophageal cancer after radical resection (hazard ratio= 1.48, 1.64, 2.23, 1.85, 2.09, 4.48, 4.97, 3.54, 5.53, 95% confidence interval as 1.08‒2.03, 1.20‒2.23, 1.47‒3.39, 1.34‒2.54, 1.44‒3.04, 2.89‒6.95, 1.57‒15.73, 2.48‒5.05, 1.73‒17.68, P<0.05). Results of multivariate analysis showed that preoperative comorbidities, G3 of tumor histological grade, T2 stage of tumor T staging, N1 stage, N2 stage, N3 stage of tumor N staging were independent risk factors affecting prognosis of esophageal cancer after radical resection (hazard ratio=1.57, 1.89, 1.63, 1.71, 3.72, 3.90, 95% confidence interval as 1.14‒2.16, 1.23‒2.91, 1.17‒2.26, 1.16‒2.51, 2.37‒5.83, 1.22‒12.45, P<0.05). (4) Construction and evaluation of the prediction models of lymph node metastasis and prognosis of esophageal cancer after radical resection. Based on the results of multivariate analysis, tumor location, tumor histological grade, tumor T staging were applied to construct a nomo-gram model for lymph node metastasis prediction of esophageal cancer after radical resection, the score of tumor location, tumor histological grade, tumor T staging were 82, 100, 100, respectively, and the sum of the scores corresponding to the lymph node metastasis rate. Preoperative comor-bidity, tumor histological grade, tumor T staging, tumor N staging were applied to construct a nomo-gram model for 1‑, 3‑, 5‑year overall survival rate prediction of esophageal cancer after radical resection, the score of preoperative comorbidity, tumor histological grade, tumor T staging, tumor N staging were 23, 38, 27, 100, respectively, and the sum of the scores corres-ponding to the 1‑, 3‑, 5‑year overall survival rate. Results of ROC showed that the AUC of nomogram model for lymph node metastasis prediction after radical esophagectomy was 0.66 (95% confidence interval as 0.62‒0.71, P<0.05). The AUC of nomogram model for 1‑, 3‑, 5‑year overall survival rate prediction after radical esophagectomy were 0.73, 0.74, 0.71 (95% confidence intervals as 0.66‒0.80, 0.68‒0.79, 0.65‒0.78, P<0.05). Results of calibration curve showed that the predicted lymph node metastasis rate and the predicted 1‑, 3‑, 5‑year overall survival rate by nomogram models were consistent with the actual lymph node metastasis rate and 1‑, 3‑, 5‑year overall survival rate.
    Conclusions Tumor location, tumor histological grade, tumor T staging are independent factors affecting lymph node metastasis in T1 and T2 esophageal squamous cell carcinoma after radical surgery and nomogram model constructed by these indicators can predict the lymph node metas-tasis rate. Preoperative comor-bidities, G3 of tumor histological grade, T2 stage of tumor T staging, N1 stage, N2 stage, N3 stage of tumor N staging are independent risk factors affecting prognosis and nomogram model constructed by these indicators can predict the overall survival rate of patients after surgery.

     

/

返回文章
返回