The application values of MRI in predicting intraoperative massive hemorrhage for cesarean scar pregnancy

曾飘娥,王丰,周延,张立华,何为,刘剑羽

DOI: https://doi.org/10.3760/cma.j.cn112149-20191016-00607

2020-01-01

Abstract:Objective:To investigate the application value of magnetic resonance imaging (MRI) in predicting intraoperative massive hemorrhage for cesarean scar pregnancy (CSP).Methods:From January to December 2018, 77 CSP patients confirmed by operation and pathology in Peking University Third Hospital were retrospectively enrolled in this study. All patients underwent MR scans in three days before operation. According to the amount of intraoperative blood loss, CSP patients were divided into intraoperative massive hemorrhage group ( n=11) and non-massive hemorrhage group ( n=66). Clinical data including patients age, the gestational age, number of cesarean delivery, interval between current CSP and last cesarean, number of abortions and preoperative human chorionic gonadotropin (β-hCG) were recorded. MRI features including CSP types, the maximum diameter of the gestational sac, the lower uterus scar thickness and gestational sac, as well as uterine hemorrhage were reviewed. The reliability of the maximum diameter of the gestational sac, and the lower uterus scar thickness measurements were evaluated by inter-class correlation coefficient (ICC). Univariate analysis was used to compare the differences of clinical data and MRI features between the two groups. Factors with significant statistical differences were further analyzed by multivariate logistic regression. The receiver operating characteristic (ROC) curve was used to evaluate the efficacy and optimal threshold of predicting CSP intraoperative massive hemorrhage. Results:The gestational age, the maximum diameter of the gestational sac, and the lower uterus scar thickness were significantly different between two groups ( P<0.05). There was no significant difference in age, number of cesarean delivery, interval between current CSP and last cesarean, number of abortions, preoperative β-hCG, CSP types, gestational sac, or uterine hemorrhage between two groups ( P>0.05). Multivariate logistic regression analysis showed that only the lower uterus scar thickness was significantly different ( P=0.002,odds ratio=4.222,95% confidence interval=1.679 to 10.614) between two groups. The ROC curve analysis showed that the area under curve of the gestational age, the maximum diameter of the gestational sac, and the lower uterus scar thickness were 0.750, 0.829 and 0.899, respectively. The best predictor was the lower uterus scar thickness, with the optimal cutoff value as 2.25 mm, and the diagnostic sensitivity, specificity, and the Youden index as 90.9%, 75.8%, 66.7%, respectively. Conclusion:The lower uterus scar thickness can be used as an independent risk factor for CSP intraoperative massive hemorrhage. Preoperative MRI, with accurately assessing morphology of the lower uterus scar and measuring the thickness, may predict the risk of intraoperative massive hemorrhage in CSP patients.