Establishment Of Diagnostic Model And Evaluation Of Diagnostic Efficacy Of Primary Fallopian Tube Carcinoma

Background :Primary fallopian tube carcinoma(PFTC)is a rare malignant tumor of gynecology,the incidence of which is only 0.14%-1.8%.PFTC lacks specific tumor markers,has atypical imaging findings,and has no special clinical manifestations in the early stage.Triad only occurs in about 15%of patients,and the rate of misdiagnosis and missed diagnosis is relatively high,with a preoperative diagnosis rate of only 2-6%.Therefore,to find the high risk factors of PFTC,establish a model to identify early tubal cancer,achieve early diagnosis,early treatment,and improve the prognosis of PFTC.ObjectiveThis study intends to analyze the patients with tubal mass in our hospital in recent ten years,collect related indicators,construct a scoring system diagnosis model,improve the diagnosis rate and reduce the misdiagnosis rate.Methods:1.176 patients with tubal masses admitted to our hospital from 2012 to 2021 were collected through the medical record system,including 138 cases of PFTC and 38 cases of benign tubal masses.They were divided into training set(123 cases)and verification set(53 cases)in a ratio of2.Univariate analysis was performed on the general clinical data,serological indicators and imaging indicators collected in the training set(P < 0.05),multivariate analysis is performed on the selected variables,and the high-risk factors of PFTC are screened out for binary logistic regression.The 10-point scoring system is constructed,and the receiver operating curve(ROC curve)is used to determine the cut off value of the model,and the specificity,and sensitivity at this value are calculated.3.Apply ROC curve to the verification set to determine the cut off value and calculate the specificity,sensitivity.Results1.Through univariate and multivariate logistic regression analysis,five risk factors for PFTC were screened out: CA125≥ 82U/ml,HE4≥56.9pmol/l,visible separation in the mass,menopause,irregular vaginal bleeding.2.Establish a scoring system for PFTC diagnosis based on the above risk factors,and assign a score to each risk factor.The area under ROC curve(AUC value)of the model in the training set was 2.8 5,and the sensitivity was 88.3%,specificity was 88.9%,Yoden index was 82.2%and AUC value was 0.946.The optimal cut-off value of the model in the verification set was 2.65,and the sensitivity and specificity were 89.5%and 88.90% respectively.The Yoden index was 77.2 percent.Conclusions:1.Through univariate and multivariate logistic regression analysis,five risk factors for PFTC were screened out: CA125≥ 82U/ml,HE4≥ 56.9pmol/l,visible separation in the mass in 3D color ultrasound images,menopause,irregular vaginal bleeding.2.The establishment of PFTC diagnostic model based on serological,imaging and clinical indicators has high sensitivity and specificity,has good predictive ability,and can be applied in clinical practice.