The N stage of esophagus cancer (EC) should be related to the number of positive lymph node (PLN) and the number of negative lymph node (NLN), and both of them had been demonstrated as independent prognostic factors recently. The information of 13,491 patients with EC was reviewed based on the database of SEER. The parameters related to prognosis were investigated by using univariate and multivariate analysis in a Cox proportional hazard regression model. A modified N stage method was proposed based on the re-adjust ratio of the number of PLN to the number of NLN according to the result of a Cox proportional hazard model, and it was evaluated by applying the Cross Validation method to the training and validation cohort. The superiority of the modified N stage method was explored by comparing it to the N stage in 7th AJCC based on the ROC analysis. Age, race, tumor length, tumor site, AJCC tumor stage, tumor differentiation, the combination of treatment, and organ or distant metastasis were all significant prognostic factors. The hazard rate (HR) of the number of PLN on prognosis was 1.042, while the HR of the number of NLN was 0.968. The re-adjust ratio of the number of PLN to the number of NLN was determined by the relative impact. The cut-off number of the re-adjust ratio was investigated from a series of survival analysis. The modified patient's N staging was defined as below: N1stage: the ratio range was from 0 to 0.21, N2: 0.22 ∼ 0.48, N3: more than 0.49. The performance of this modified N stage method was evaluated and its superiority compared to the N stage coming from AJCC was identified. The modified N stage method based on the re-adjust ratio of the number of PLN to the number of NLN could improve the evaluation of the tumor N stage.
Read full abstract