Study On The Relationship Between Lymph Node Staging Of Non-surgical Treatment Patients With Esophageal Squamous Cell Carcinoma And Prognosis

Objective To find the optimal criteria for determining the positive LN in each LN zone and the non-surgical treatment LN staging in patients with ESCC by analyzing the relationship between the length of lymph node under Multi-slice Spiral Computer Tomography(MSCT)and prognosis of nonoperative patients with ESCC.Methods A total of 393 patients who underwent Multi-slice Spiral Computer Tomography scan from December 2009 to December 2014 at the department of Thoracic Surgery,Zhang Zhou Hospital,in Fujian Province were included in the analysis.For each patient,the following variables were recorded: the diameter of LN(short-axis and long-axis),the number and location of LN basing on MSCT at the time of admission.Survival functions were estimated by the Kaplan-Meier method,Log-rank tests were performed to compare the differences of survival rates,and multivariate Cox regression test were used to performed prognostic analysis.Using short-axis diameter,long-axis diameter,short-axis diameter long-axis diameter preferred,short-plus-long axis diameter and ratio of long-to-short axis as the positive LN standard respectively.Subsequently,we picked out the better prognosis LN zones by using Cox Proportional Hazard Model,random survival forest(RSF)model and Lasso regression model based on SPSS and R software.And then comparing survival curves and receiver operating characteristic curve(ROC)curves to explore the optimal lymph node staging.Results 1.In our study,we first identified the optimal criteria of positive LN,according to the ROC curve.The result showed that the positive criteria of short-axis diameter long-axis diameter preferred LN were superior to those with only the short-axis diameter,only the long-axis diameter,short-plus-long axis diameter and ratio of short-to-long axis diameter.The area under curve(AUC)of the 1-,2-,3-,4-and 5-year positive criteria was 57.34%,59.25%,61.59%,61.94%,63.09%,respectively.The positive criteria were different in different regions of the lymph node.For instance,the positive criteria of short-axis diameter lymph node included 1 zone(9mm)of the lower cervical,2R zone(8mm),2L zone(7mm),3P zone(8mm),4R zone(8mm),5 zone(10mm),6 zone(9mm)of most the upper thoracic LN,7 zone(7mm)of the thorax,11 R zone(5mm),12 L region(3mm),13 R region(3mm),14 R region(3mm)of pulmonary segment,9 zone(9mm)of the pulmonary artery ligament,15 zone(8mm)and 16 zone(5mm)both the upper and lower of diaphragm and 18 zone(10mm)of the abdominal cavity.The positive criteria of long-axis diameter lymph node included 3A zone(7mm),4L zone(14mm)of the upper thorax,8M zone(15mm),8L zone(10mm),10 R zone(15mm)of the middle and lower thorax,11 L zone(10mm),12 R zone(13mm)of the pulmonary segment and 17 zone(9mm),19 zone(11mm),20 zone(7mm)of most abdominal zones.2.The results revealed that the Lasso regression model screened 1,7,16 and 17 zones as prognosis-related LN zones,the Cox Proportional Hazard Model screened 1,3A,6,8L,10 L,15,16 and 18 zones and the RSF model screened 1,2R,2L,3A,3P,4R,4L,6,7,8M,8L,9,10 L,11R,16,17 and 20 zones,respectively.Univariate and multivariate regression analysis showed that the adjusted LN staging according to Cox Proportional Hazard Model was not an independent prognostic factor for nonoperative patients with ESCC.The 1-,2-,3-,4-and 5-year AUC value of the Lasso regression model(63.06%,64.68%,67.93%,68.24%,66.41%,P < 0.001,respectively)and the RSF model(58.54%,60.62%,63.58%,63.05%,62.12%,P < 0.001,respectively),which showed the LN staging according to Lasso regression model is better than that of RSF.The positive rates of LN in the zones of 1,7,16 and 17 regions which considered to be prognosis-related regions for nonoperative patients with ESCC were 35.62%,38.68%,34.38% and 47.58%,respectively.3.We defined N staging basing on prognosis-related regions screened by Lasso,N0 was defined as no LN positive,N1 was defined as any of zone 1,7,16 and 17 LN positive or other zone LN positive,N2 was defined as any of zone 1,7,16 and 17 LN positive and other zone LN positive,respectively.Univariate and multivariate regression analysis showed that the adjusted LN staging according to main LN zone was an independent prognostic factor for nonoperative patients with ESCC,P<0.001.The mortality risk of N0 patients were lower than N2 patients(HR=0.320,95%CI: 0.198~0.518),and it revealed similar results in N1 patients when co-mpared with the N2 stage patients(HR=0.642,95%CI: 0.468~0.882,respectively).Conclusions The positive criteria of short-axis diameter and long-axis diameter preferred lymph node is more suitable for nonoperative patients with ESCC than draft in clinical practice.The positive criteria were varies by LN zones.The lower cervical,thoracic and pulmonary segment are mostly short-diameter criteria,nevertheless the lower part of the thorax and the majority of the abdominal cavity are mostly long-diameter criteria.The LN staging(N staging)based on the screening of 1,7,16 and 17 zones are also suitable for nonoperative patients with ESCC and can provide reference for clinical staging and treatment.