Differentiation of uterine carcinosarcoma and low-risk endometrial cancer using multimodal MRI combined with serum CA125 and ZEB1
[Objective]To analyze the differential value of multimodal magnetic resonance imaging(MRI)combined with serum carbohydrate antigen 125(CA125)and zinc finger E-box binding homeobox 1(ZEB1)in the diagnosis of uterine carcinosarcoma(UCS)and low-risk endometrial cancer(EC).[Method]Twenty-four patients with UCS and 55 patients with low-risk EC admitted to Jinshui District General Hospital from January 2018 to December 2022 were retrospectively selected as the UCS group and low-risk EC group,respectively.All patients underwent multimodal MRI examination and serum CA125 and ZEB1 detection.Using pathological results as the gold standard,the discriminant value of multimodal MRI,serum CA1 25,ZEB1,and their combination for UCS and low-risk EC was analyzed using the reciever operating characteristic curve.[Results]The maximum tumor diameter,proportion of bleeding,proportion of cystic change/necrosis,volume transport constant and blood reflux constant in the UCS group were higher than those in the low-risk EC group(P＜0.05),and the relative apparent diffusion coefficient in the UCS group was lower than that in the low-risk EC group(P＜0.05).The serum levels of CA1 25 and ZEB1 in the UCS group were higher than those in the low-risk EC group(P＜0.05).The sensitivity of multimodal MRI combined with serum CA125 and ZEB1 in differential diagnosis of UCS and low-risk EC was 95.83％,which was higher than that of single differential diagnosis(P＜0.05).The AUC of combined differential diagnosis was 0.904,which was higher than that of single differential diagnosis.The specificity of combined differential diagnosis was 74.55％,and there was no statistically significant difference compared with single differential diagnosis(P＞0.05).[Conclusion]Multimodal MRI and serum CA125 and ZEB1 have differential diagnostic value for UCS and low-risk EC,and the combination of the three can further improve the differential diagnostic value.
万方数据
维普
