A Predictive Model For Pelvic Lymph Node Metastasis In Endometrioid Adenocarcinoma

Objective: To explore the correlation between clinicopathological parameters and immunohistochemical markers,to search for independent predictors of pelvic lymph node metastasis in endometrioid adenocarcinoma,and to construct a risk prediction model with nomogram,so as to realize individualized evaluation of pelvic lymph node metastasis probability.Methods: The clinicopathological data of 548 patients with endometrioid adenocarcinoma admitted to jiangxi provincial maternal and child health hospital on January 1,2013 and December 31,2015 were retrospectively analyzed.The age,postmenopausal status,preoperative serum CA125 level,preoperative cureage pathological results,surgical methods,postoperative pathological reports and the expression of immunohistochemical markers ER,PR,p53 and ki-67 were recorded in detail.The correlation among postoperative clinicopathological parameters,immunohistochemical markers and pelvic lymph node metastasis was analyzed by single factor analysis.Independent predictors with statistical significance for pelvic lymph node metastasis were screened by Logistic regression,ROC curve was drawn,and the area under the ROC curve of the combined predictors was calculated.A nomogram model was established to predict pelvic lymph node metastasis in endometrioid adenocarcinoma,and the accuracy and consistency of the model were verified by internal sampling.Results :1.The area under the ROC curve of preoperative serum CA125 in predicting lymph node metastasis was 0.601(95%Cl: 0.471-0.730).When the critical value was 19.42U/m L,the sensitivity and specificity were 75% and 48.7%,respectively.The results of univariate analysis showed that the preoperative serum CA125 level,histological grading,muscle infiltration depth,lymphoid vascular space infiltration(LSVI),cervical interstitial infiltration,ovarian metastasis and pelvic lymph node metastasis were correlated,and the differences were statistically significant(all P < 0.05).The protein expressions of immunohistochemical parametersER,PR,P53 and ki-67 were correlated with clinical pathological parameters and pelvic lymph node metastasis,and the differences were statistically significant(all P< 0.05).Logistic regression analysis showed that preoperative serum CA125 level,histological grading,and LSVI were independent risk factors for lymph node metastasis,and the regression coefficients were 1.261,1.372,and 1.964,respectively.The ROC curve corresponding to the above markers was drawn,and the area under the curve was 0.867(95%Cl: 0.808-0.927),all greater than the independent predictors.A nomogram model was constructed to predict the risk of pelvic lymph node metastasis in endometrioid adenocarcinoma.Bootstrap internal sampling was used to verify the model internally,and the c-index of the model was 0.843.2.The further analysis of LSVI related clinical pathological parameters and immunohistochemical markers,single factor analysis showed that the histological grading,muscular invasion depth,cervical interstitial infiltrates,ER,PR,ki-67 LSVI related,the differences were statistically significant(P < 0.05),a Logistic regression analysis showed histological grading,muscular invasion,cervical interstitial infiltrates,ER positive is an independent predictor of LSVI.Conclusions :1.Preoperative serum CA125,histological grading,and LSVI can be used to predict lymph node metastasis of endometrial cancer,and the nomogram model jointly constructed by the three can provide reference value for clinical lymph node dissection.Immunohistochemical markers combined with clinical pathological parameters can preliminarily predict LSVI.