A Fibrinogen-based Nomogram Model For Predicting Prognosis Of Epithelial Ovarian Cancer

Objective:In this study,we collected plasma fibrinogen values before initial treatment of epithelial ovarian cancer（Epithelial Ovarian Cancer,EOC）,discussed the correlation with the clinicopathological characteristics of EOC patients,analyzed the influence on postoperative survival of EOC patients,constructed the EOC survival prediction model,demonstrated nomogram visualization,to accurately predict the prognosis of EOC patients,and guide the individualized treatment of patients.Methods:The clinical data of patients undergoing the first stage or subtraction of ovarian cancer in the Shanxi Bethune Hospital from July 2014 to February 2022.The clinicopathological data,laboratory examination results and follow-up data of the above patients were systematically collected.After collecting the data,the subject working characteristic curve（receiver operator characteristic curve,ROC）of fibrinogen was drawn,and the optimal cut-off value of fibrinogen was taken to compare the correlation between different levels of fibrinogen and the clinicopathological characteristics of EOC patients.Univariate analysis was performed byχ²test to screen risk factors affecting postoperative survival of EOC patients,and included them in the multivariate COX proportional risk model to obtain independent risk factors affecting postoperative survival of EOC patients.The overall survival（Overall survival,OS）curve of the above factors was drawn using Kaplan-Meier method,differences between groups were compared using the Log-rank methodand,the nomogram was drawn in R language prediction model.The prediction model was also internally validated and calibrated.All statistical analyses of the data were performed using SPSS 26.0 and RStudio,and P<0.05 was considered as statistically significant.Results:According to the results of the ROC curve,the optimal cut-off value of fibrinogen was 4.37g/L,according to which the patients were divided into low fibrinogen values and high fibrinogen values.Different levels of fibrinogen were associated with postoperative residual lesion size,presence of ascites,cancer antigen（Cancer Antigen125,CA125）level,human epididymal protein 4（Human Epididymal Protein 4,HE4）level,the International Federation of Obstetrics and Gynecology（Federation International of Gynecology and Obstetr,FIGO）stage,and whether maintenance therapy（P<0.05）,there was no correlation with the patient’s age（P=0.862）and the histological type（P=0.186）;Univariate analysis:the presence of ascites,FIGO stage,CA125 level,HE4 level,postoperative residual lesion size,and fibrinogen level were the risk factors for postoperative survival of patients with EOC（P<0.05）;Multivariate analysis:FIGO stage（HR:3.152,95%CI:1.528～8.502,P=0.002）,residual lesion size after surgery（HR:0.463,95%CI:0.250～0.859,P=0.015）,CA125 level（HR:3.366,95%CI:1.666～6.802,P=0.001）,fibrinogen level（HR:3.114,95%CI:1.704～5.692,P<0.01）,presence of ascites（HR:0.510,95%CI:0.276～0.942,P<0.031）were independent risk factors for the postoperative survival of patients with EOC（P<0.05）,and were successively assigned X₁,X₂,X₃,X₄and X₅,the survival prognostic model was constructed based to the included independent risk factors affecting the postoperative survival of EOC patients:h（t,X）=h0（t）exp（3.152X₁+0.463X₂+3.366X₃+3.114X₄+1.780X₅）,reflected by using the nomogram,agreement coefficient corrected for this prediction model（Concordance index,C-index）was 0.753[95%CI:0.726-0.781],after an internal correction,show a good fit of the predicted calibration curve to the standard curve,and good conformity of the model,suggested that the model has a good predictive accuracy,can be used to predict postoperative survival in patients with EOC.Conclusion:1.According to the ROC curve results,the optimal cut-off value of fibrinogen was 4.37g/L.2.Different preoperative levels of fibrinogen and the size of postoperative residual lesions,presence of ascites,CA125 level,HE4 level,FIGO stage,and maintenance therapy were correlated（P<0.05）,there was no correlation with the patient’s age（P=0.862）and the histological type（P=0.186）.3.Univariate analysis:presence of ascites,FIGO stage,CA125 level,HE4 level,postoperative residual lesion size,and fibrinogen level were risk factors for postoperative survival of patients with EOC（P<0.05）.4.Multivariate analysis:FIGO stage（HR:3.152,95%CI:1.528～8.502,P=0.002）,residual lesion size（HR:0.463,95%CI:0.250～0.859,P=0.015）,CA125 level（HR:3.366,95%CI:1.666～6.802,P=0.001）,fibrinogen（HR:3.114,95%CI:1.704～5.692,P<0.01）and ascites（HR:0.510,95%CI:0.276～0.942,P<0.031）were independent risk factors for postoperative survival of EOC patients（P<0.05）.The values were successively assigned X₁,X₂,X₃,X₄,X₅,and finally the survival prognosis model was established:h（t,X）=h0（t）exp（3.152X₁+0.463X₂+3.366X₃+3.114X₄+1.780X₅）,which was shown by nomogram.5.The survival prediction model based on FIGO stage,fibrinogen≥4.37g/L,CA12535U/ml,whether postoperative residual lesion reaches RO,and ascites has high accuracy and differentiation,and can be used to predict the postoperative survival of EOC patients.