Analyzing Clinical Characteristics Of Leiomyomas Coexisting Adenomyosis And Establishing A Predictive Model

ObjectiveThe design of the study aims to compare the differences in clinical characteristics of both leiomyomas and adenomyosis to women with leiomyomas alone,and to construct a logistic regression model for predicting the risk of adenomyosis coexisting with leiomyomas.In order to freeback its clinical characteristics,improve the accuracy of diagnosis.Materials and MethodsRetrospective analysis from June 2016 to January 2020 in the Third Affiliated Hospital of Zhengzhou University on 409 patients who underwent hysterectomy and all the patients were collected for observation according to inclusion and exclusion conditions,which a total of 150 patients with adenomyosis combined with leiomyoma were histologically diagnosed as the observation group,and 259 cases with leiomyoma alone were selected as the control group during the same period.And recording clinical data respectively to screen the independent risk factors by using the simple factor analysis and the multivariate logistic regression analysis to established the predicive model of indicating adenomyosis coexisting with leiomyomas.The efficiency of the model was tested by the sensitivity,specificity,positive predictive value,negative predictive value,and receiver operating characteristic curve.Results1.Pathological results showed that the group of leiomyomas associated with adenomyosis had enlarged uteri(P<0.05);however,women with only leiomyomas had a higher uterine weight,more leiomyomas and larger leiomyomas,and there were statistically significant differences between the groups(P<0.05),but not including the classification and location of leiomyomas.2.There was no significant difference in the same signs or symptoms and the categories of clinical manifestations between the two groups(P>0.05),which was relatively difficult to distinguish the disease diagnosis before operation.3.Single factor regression analysis of abortions,mean breastfeeding cycle per fetus,benign breast disease,CA125,HOMA-IR,HB,fasting insulin,NLR,PLR were statistically significant(P<0.05).4.Multivariate regression analysis showed that Multiple miscarriages,mod erate to severe anemia,NLR,PLR,high levels of CA125 and HOMA-IR were related with the disease of leiomyomas associated with adenomyosis,and prol onged breastfeeding cycle of per fetus was a protective factor;the above facto rs were selected to establish the Logistic regression equation.In order of OR value,X1=(CA125>35 U/ml),X2=(abortion times>2),X3=(HB<60g/L),X4=(H B:60～90g/L),X5=(HOMA-IR≥2.69),X6=NLR,X7=PLR,X8=mean breastfeedin g cycle per fetus.This model is P=1/1+E xp∑(-6.639+3.003X1+2.834X2+1.967X3+1.768X4+1.713X5+1.042X6+0.015X7+-1.580X8),and detection of accuracy=90.00%,sensitivity=86.70%,specificity=91.90%,positive predictive v alue=86.67%,negative predictive value=91.89%,ROC-AUC=0.876(95%CI:0.840～0.906,P<0.001).Conclusion1.It is relatively difficult to distinguish whether leiomyomas are combined with adenomyosis according to clinical manifestations and there are similarities and differences in the pathological characteristics between leiomyomas combined with adenomyosis group and simple leiomyomas group.2.NLR,PLR,severe anemia,frequent miscarriage,high levels of CA125,and HOMA-IR may be recommended as risk factors for leiomyomas with adenomyosis,while prolonged breastfeeding cycle of per fetus may be recommended as protective factors to prevent their occurrence.The established predictive model has good diagnostic value,and can effectively find the patients associated with adenomyosis in the early stage,and choose reasonable treatment.