Survival Analysis And Risk Prediction Model Of Thymoma

Objective:(1)To evaluate the survival rate of thymoma patients at 1year,5 years and 10 years and median survival time and explore the related factors affecting the survival of patients with thymoma;(2)Construct and evaluate individual prognostic index equations and nomogram models which predicts the risk of survival in patients with thymoma.Methods:(1)Refer to the relevant literatures and consult experts to make a form which was “Clinical Data and Follow-up Data Collection Form for Patients with Thymoma”.Using the form,we were collected the clinical characteristics of 271 patients with benign thymoma who were admitted to the hospital from January 2006 to December 2016 in the Hunan Provincial Cancer Hospital Case Management System.The patient’s survival was followed up by telephone interview,letter survey and outpatient review.(2)The Kaplan-Meier method was used to estimate the survival curve,survival rate and median survival time.Comparison of differences in survival rates between groups using the Log-rank method.The screening of factors affecting the survival of patients with thymoma was performed by multivariate Cox regression analysis.(3)According to the results of multi-factor analysis,the individual prognostic index equation is established,and the judgment of the application value is judged by the false positive probability.The analysis methods of above are all completed in the software SPSS20.0.(4)Using the rms package in R software 3.5.1 to construct nomogram,the prediction effect isevaluated by a concordance index(C-index).Results:(1)A total of 271 patients with thymoma were followed up,127 patients died of this disease,138 survived,and 6 patients were lost to follow-up.The rate of loss of follow-up was 2.21%.The median survival time of patients with thymoma was 7.17 years,and the 1-,5-,and 10-year survival rates were 94.1%,63.1%,and 31.3%,respectively.(2)The results of univariate analysis showed that there were significant differences in survival curves between age,length of hospital stay,tumor resection integrity,type of surgery,Masaoka staging,smoking status,WHO classification,maximum tumor diameter,muscle weakness,tumor metastasis,and lymphadenopathy.(3)Cox regression multivariate analysis showed that age,tumor resection integrity,adjuvant therapy,Masaoka staging,WHO typing,tumor maximum diameter and presence or absence of muscle weakness were factors influencing the survival of patients with thymoma.Low age,complete tumor resection,adjuvant therapy,small maximum tumor diameter,no companion muscle weakness,earlier Masaoka staging and low-level WHO classification are protective factors for survival of patients with thymoma.(4)The prognostic predictive ability of the individual prognostic index equation was evaluated.The results showed that the sensitivity was 83.3%,the specificity was 82.6%,and the probability of misjudgment was 0.17.(5)Built-in nomogram,the consistency index is 0.837,close to 1,and the forecasting ability is strong.Conclusions:(1)Younger,complete resection,adjuvant therapy,small tumor diameter,Masaoka staging and WHO typing which are younger,without early muscle weakness are beneficial to patients’ survival.(2)The prognosis index equation of thymoma constructed by this study can provide reference for the prediction of patient prognosis.(3)The nomogram model constructed in this study is more convenient for clinical use.