Risk Factors And Prediction Model Of Lymphatic Vascular Space Invasion In Endometrial Carcinoma

Objective: To investigate the risk factors leading to lymphovascular interstitial infiltration in endometrial cancer patients,and to establish and validate a prediction model for the risk of lymphovascular interstitial infiltration in endometrial cancer patients,in order to provide individualized assessment and treatment for clinical diagnosis and treatment.Methods: A total of 480 patients with endometrial cancer who were hospitalized at the Affiliated Hospital of Jining Medical College from January 2016 to June 2021 were selected,and relevant clinical data were collected from the patients,who were randomly divided into a training group（360 cases）and a validation group（120 cases）according to a ratio of 7.5:2.5.The differences in clinicopathological characteristics between lymphovascular interstitial infiltration and non-infiltrating patients were assessed.Statistically significant study variables were screened by univariate logistic analysis and included in multifactorial logistic regression analysis to identify independent predictors of lymphovascular interstitial invasion,and a column line graph（Nomogram）prediction model was constructed based on the results of the multifactorial analysis.The performance of the Nomogram prediction model was evaluated using calibration curves with subject operating characteristic（ROC）curves.Results:1.Univariate analysis revealed that age,family history of malignancy,depth of myometrial infiltration ≥1/2,type of pathology,cervical interstitial infiltration,and maximum tumor diameter were strongly associated with the occurrence of lymphovascular interstitial infiltration in patients with endometrial cancer（P<0.05）.Multi-factor logistic regression analysis showed that family history of malignancy,myometrial infiltration ≥1/2,type of pathology,cervical interstitial infiltration,and maximum tumor diameter were independent risk factors for lymphovascular interstitial infiltration.2.A logistic retrospective model was constructed using the above five independent factors to derive the prediction equation: Y=-4.27222+1.20150*family history of malignancy +1.38343*myometrial infiltration+1.05743*pathological type+0.96548*interstitial cervical infiltration +0.16790*maximum tumor diameter.ROC curves were plotted and yielded an area under the ROC curve（AUC）of 0.836（95% CI: 0.761-0.911）in the training group and 0.840（95% CI: 0.737-0.943）in the validation group for predicting the risk of lymphovascular interstitial infiltration.The calibration curves indicated a high agreement between the predicted risk values of lymphovascular interstitial infiltration by this line graph model and the actual observed values.Conclusion: The successful construction of this lymphovascular interstitial infiltration prediction model can help to determine the presence of lymphovascular interstitial infiltration before standard staging surgery for endometrial cancer patients.The risk line graph model based on family history of malignancy,myometrial infiltration depth ≥1/2,pathological type,cervical interstitial infiltration and maximum tumor diameter can more visually,graphically and conveniently predict the risk of lymphovascular interstitial infiltration in endometrial cancer patients initially,enabling clinicians to identify high-risk patients earlier and formulate more precise surgical approaches and individualized treatment plans.