Imaging Characteristics And Clinical Applications Of MRI In Cesarean Scar During The Third Trimester

Objective:The purpose of this study is to use 3.0 T magnetic resonance imaging(MRI)to explore the image characteristics of late pregnancy scar after cesarean section,to predict the critical value of lower uterine scar dehiscence,and to provide objective imaging evidences to guide the further treatment.Materials and Methods:A total of 178 pregnant women in the third trimester of pregnancy(gestational week≥28W),at the Third Affiliated Hospital of Guangzhou Medical University from March 2016 to January 2021,were retrospectively analyzed.They underwent routine MRI examination.According to the previous history of at least one cesarean section(CS)and repeated cesarean section,they were divided into scar group(78 cases)with one CS or more than one CSs and control group(100 cases)without previous history of CS.The scar grades of the lower uterine segment(LUS)in scar group were divided into group A(23cases)with incomplete scar rupture and group B(55 cases)with good scar healing.Firstly,we evaluated the disparities of LUS thickness in different gestational weeks and gravida times between the scar group and the control group.Secondly,we compared the differences of LUS thickness,T2WI signal intensity ratio,LUS thickness measured by ultrasound,gravida times,the interpregnancy interval,prenatal uterine contraction between the group A and group B.At the same time,the maternal and infant outcomes between group A and group B were studied containing lower uterine repair,intraoperative blood loss,uterine contraction fatigue,abdominal and pelvic adhesions,Apgar score≤7,neonatal low weight rate,neonatal premature birth rate,and neonatal transfer rate.The statistical methods were independent sample t-test,nonparametric test and one-way ANOVA(P<0.05).Finally,binary logistics regression was used to analyze the independent influencing factors of incomplete scar rupture in group A and to draw the receiver operating characteristic(ROC)curve to predict the critical value of uterine scar dehiscence.SPSS 26.0(IBM Corp.Armonk,NY,USA)version was utilized to be the statistical software.Results:1、General conditions:there was no significant difference in age,gestational week and BMI between the scar group and the control group,group A and group B(P>0.05).2、(1)Comparison between the scar group and the control group:there was a statistically significant difference in LUS thickness between those with and without CS history,and the LUS thickness in the scar group(2.469±0.735mm)was thinner than in control group(3.305±0.532mm),t=-10.971,P<0.05.The LUS thickness was statistically significant in different gestational weeks.The LUS thickness in the scar group(2.406±0.657mm)was thinner than in control group(3.803±0.368mm)in28+0～31+6GW,t=8.153,P<0.05 and the LUS thickness in the scar group(2.619±0.837mm)was thinner than in control group(3.506±0.401mm)in 32～35+6GW,t=5.334,P<0.05.In≥36GW,the LUS thickness was not statistically significant(P>0.05).(2)Comparison in the scar group and the control group:In the scar group,the LUS thickness in the group with 1 CS and 2 CSs and more than 2 CSs was statistically significant,F=3.707,P<0.05.The LUS thickness in the group with 1 CS(2.996±0.703mm)was greater than that with more than 2 CSs(2.489±1.035mm),P<0.05.In the control group,the LUS thickness was statistically significant in the group with 0 delivery and 1 delivery and 2 deliveries,F=5.051,P<0.05.The LUS thickness was greater in the group with 0 delivery(3.638±0.472mm)and 1 delivery(3.523±0.555mm)than with 2deliveries(3.044±0.702mm),P<0.05.There was no significant difference in the thickness between 0delivery and 1 delivery(P>0.05).In the scar group,the LUS thickness was not statistically significant in different gestational weeks(P>0.05),but was statistically significant in the control group with Z=38.486,P<0.05.The LUS thickness in control group at 28+0～31+6 GW[M(IQR)=3.85(0.575)mm]and 32～35+6 GW[M(IQR)=3.50(0.375)mm]was greater than that at≥36 GW[M(IQR)=2.70(0.400)mm],P<0.05.The LUS thickness in control group at 28+0～31+6 GW[M(IQR)=3.85(0.575)mm]was greater than that at 32～35+6 GW[M(IQR)=3.50(0.375)mm],P<0.05.3、Comparison between group A and group B in the scar group showed that the differences in prenatal uterine contractions,LUS thickness,T2WI signal intensity ratio,gravida times were statistically significant(P<0.05).The LUS thickness in group A(1.817±0.355mm)was thinner than that in group B(2.742±0.680mm),t=7.844,P<0.05.The T2WI signal intensity ratio in group A(0.813±0.196)was less than that in group B(0.917±0.202),t=2.107,P<0.05.The number of CS in group A[M(IQR)=2(2)times]was greater than that in group B[M(IQR)=1(1)times],Z=-3.354,P<0.05.The constituent ratio of prenatal uterine contractions in group A(14 cases,60.9%)was greater than that in group B(6 cases,10.9%),X~2=21.231,P<0.05.and the constituent ratio of identified scar characteristics in group A(10 cases,43.5%)was less than that in group B(40 cases,72.7%),X~2=6.029,P<0.05.4、Comparison of maternal and infant outcomes between group A and group B:There were significant differences in lower uterine segment repair,intraoperative blood loss and abdominopelvic adhesion between group A and group B.The constituent ratio of lower uterine segment repair in group A was 100%,P<0.05 and the constituent ratio of abdominopelvic adhesion in group A(17 cases,73.9%)was greater than that in group B(21 cases,38.2%),X~2=8.288,P<0.05.The intraoperative blood loss in group A[M(IQR)=400(200)m L]was greater than that in group B[M(IQR)=300(100)m L],Z=-2.142,P<0.05.The uterine contraction fatigue rate in group A and B was not statistically significant(P>0.05).There were significant differences in premature delivery rate and underweight rate between group A and group B.The constituent ratio of premature delivery rate in group A(9 cases,39.1%)was greater than that in group B(10 cases,18.2%),X~2=3.863,P<0.05 and the constituent ratio of underweight rate in group A(6 cases,26.1%)was greater than that in group B(5 cases,9.1%),X~2=3.867,P<0.05.There was no significant difference in 1-minute Apgar score≤7 and neonatal transfer rate(P>0.05)between group A and group B(P>0.05).5、Binary Logistics regression analysis was performed on various related factors,and the results showed that the thickness of LUS,prenatal contraction,and the number of previous CS were three independent influencing factors of abnormal LUS scar dehiscence.The regression coefficient(β)of LUS thickness was-3.78,standard error was 1.183,OR was 0.023,P was 0.001,the regression coefficient(β)of prenatal contraction was 1.953,standard error was 0.845,OR was 7.050,P was 0.021and the regression coefficient(β)of CS times was 1.354,standard error was 0.713,OR was 3.872,P was 0.048.6、In group A,the ROC curve of LUS thickness,T2WI signal intensity ratio and the three combined independent influencing factors detections showed that the best critical value of LUS thickness for predicting uterine scar dehiscence was 2.10 mm,AUC was 0.898,sensitivity was 91.3%,specificity was 81.2%;the best critical value of T2WI signal intensity ratio was 0.936,AUC was 0.667,sensitivity was 87%,specificity was 45.5%;the three combined detections AUC is 0.936,and the sensitivity is 91.3%,the specificity is 85.5%.Conclusions:1、There were significant differences in LUS thickness and gestational weeks between pregnant women with and without history of CS.There was not difference in LUS thickness of pregnant women with CS in gestational weeks,while there were significant differences in LUS thickness of without CS both gravida times and gestational weeks,with better ductility and elasticity.2、In this study,the uterine LUS thickness,prenatal uterine contraction and times of cesarean section were the independent influencing factors for incomplete uterine rupture.The group with incomplete scar rupture was more likely to lead to adverse maternal and fetal outcomes.4、MRI can be used to predict incomplete uterine rupture.The optimal cut-off value of LUS thickness is 2.10 mm,the optimal cut-off value of T2WI signal intensity ratio is 0.936,the combined detection of LUS thickness,uterine contraction and times of CS has the highest reliability.