Magnetic Resonance Imaging And Its Scoring System Application In The Prediction Of Placenta Accreta Spectrum

[Background]In 2018,organizations such as the International Federation of Gynecology and Obstetrics（FIGO）,the American College of Obstetricians and Gynecologists,and the American Society of Maternal-Fetal Sciences issued guidelines,Placental accreta spectrum（PAS）^[1,2],a large group of diseases that adhere to the placenta,was renamed.In recent years,with the opening of the birth policy,the incidence of placenta accreta has increased year by year,and the pathogenesis of PAS is relatively complex.If the accurate diagnosis of prenatal PAS is not realized,the risk of refractory massive hemorrhage may be greater during childbirth,which will endanger the life safety of the maternal.At present,the diagnostic methods for PAS include serology,ultrasound,Magnetic resonance imaging（MRI）,etc.However,it is still difficult for a single examination method to achieve accurate diagnosis evaluation.This study aims to add Likert scoring system on the basis of MRI to improve the diagnostic efficiency of PAS typing.It is also of great clinical value to identify the risk factors for PAS,and to choose the follow-up treatment for PAS and improve the outcome of mothers and infants.[Objective]This study discussed the application effect of magnetic resonance imaging in the diagnosis of prenatal PAS and the clinical value of scoring system in the prediction of implantation depth,and analyzed the risk factors for the occurrence of PAS,in order to prevent and reduce the occurrence of postpartum massive hemorrhage,and to provide the corresponding reference basis for the prevention and clinical treatment of PAS.[Materials and Methods]The clinical and MRI imaging data of 60 pregnant women with clinical or ultrasound suspicion of PAS admitted to the Shen zhen People’s Hospital from March 2018 to March2020 were retrospectively analyzed.The pregnant women ranged in age from 20 to 45 years old,the average age was（30.96±8.75）years old,and the gestational weeks ranged from 28to 40 weeks.Mean gestational age（35.49±4.11）weeks.Using clinical operation or postoperative pathological diagnosis results as the"gold standard",a total of 43 patients（71.67%）with PAS and 17 patients（28.33%）without PAS were confirmed,including 22patients with adhesion,14 patients with implantation and 7 patients with penetrating PAS.Likert scale scoring system was used to quantify the MRI signs of PAS,ROC curve was used to evaluate the total score of MRI signs,the diagnostic value of different types of PAS,and the risk factor of PAS was univariateχ² Test and multivariate logistic regression analysis,P<0.05 indicated statistically significant difference.[Results]1.With clinical operation or postoperative pathological diagnosis results as the"gold standard",a total of 43 patients（71.67%）with PAS and 17 patients（28.33%）without PAS were confirmed.41 cases（68.33%）were positive（PAS patients）and 19 cases（31.67%）were negative（non-PAS patients）by MRI.2.The sensitivity of PAS in MRI examination was 90.70%（39/43）,specificity 88.24%（15/17）,positive predictive value 95.12%（39/41）,negative predictive value 78.95%（15/19）and accuracy 90.00%（54/60）.3.MRI examination was performed on pregnant women with high suspicion of PAS.The detection rates of abnormal blood vessels in placenta in positive patients were significantly higher than those in negative patients,including 51.22%（21/41）,53.66%（22/41）,75.61%（31/41）,and 34.15%（14/41）of tortuously thickened blood vessels in myotomes,respectively.The difference was statistically significant（all P<0.05）.There was no significant difference between positive patients and negative patients in the detection rates of inhomogeneous signal in placenta 85.37%（35/41）,low signal zone in placenta31.71%（13/41）,and placental tissue toward cervical ostium carina 12.20%（5/41）（all P>0.05）.4.There were significant differences in the total score of MRI signs among patients with different PAS types（P<0.05）;The total MRI score of adhesive PAS patients（22.19±2.23）,implant PAS patients（27.68±2.35）and penetration PAS patients（31.55±2.44）were significantly higher than those of non-placental accreta patients（16.14±1.89）,respectively,with statistical significance（all P<0.05）.The total score of MRI signs in patients with implanted PAS（27.68±2.35）and penetrating PAS（31.55±2.44）were significantly higher than those in patients with adhesive PAS（22.19±2.23）,respectively,with statistical significance（all P<0.05）.The total score of MRI signs in patients with penetrating PAS（31.55±2.44）was significantly higher than that in patients with implanted PAS（27.68±2.35）,with statistical significance（P<0.05）.5.According to the Receiver operating characteristic（ROC）curve drawn,the total score of MRI signs for the diagnosis of adhesion and non-implantation was 0.949 in the Area under the curve（AUC）.When the score was≥20,the sensitivity was 95.45%（21/22）and the specificity was 88.24%（15/17）,and 20 points was taken as the critical value for the diagnosis of adhesion and non-implant type,that is,when the total score of MRI signs was≥20,the adhesion type could be diagnosed,and when the total score of MRI signs was<20,the non-implant type could be diagnosed.The AUC for the diagnosis of adhesion type and implant type by using the total score of MRI signs was 0.933.When the score was≥25,the sensitivity was 92.86%（13/14）and the specificity was 90.91%（20/22）,and 25 points was taken as the critical value for the diagnosis of adhesion type and implant type,that is,when the total score of MRI signs was≥25,the implant type could be diagnosed.The adhesion type can be diagnosed when the total score of MRI signs is less than 25.The AUC of the total score of MRI signs was 0.796.When the score was≥30,the sensitivity was 71.43%（5/7）and the specificity was 78.57%（11/14）,and 30 was taken as the critical value for the diagnosis of the implant type and penetration type,that is,when the total score of MRI signs was≥30,the penetration type could be diagnosed.When the total score of MRI signs is less than 30,it can be diagnosed as implant type.6.Univariate analysis showed that there were statistically significant differences in age,pre-pregnancy Body mass index（BMI）,number of pregnancies,number of births,history of abortion,history of cesarean section,combined placenta previa and history of uterine surgery between the implantation group and the non-implantation group（all P<0.05）.There was no significant difference in gestational age and early gestational bleeding history between the two groups（all P>0.05）.7.Multivariate logistic regression analysis with the occurrence of PAS in pregnant women as the dependent variable showed that age（OR=2.352,95%CI:1.373-4.249）,delivery time（OR=1.642,95%CI:1.274-3.057）,abortion history（OR=1.927,95%CI:1.487-3.398）,cesarean section history（OR=1.716,95%CI:1.224-2.973）,combined placenta previa history（OR=2.645,95%CI:1.791-3.711）,and uterine surgery history（OR=1.689,95%CI:1.252-2.916）were independent risk factors for PAS in pregnant women（all P<0.05）.[Conclusion]1.The sensitivity and specificity of PAS in MRI examination were 90.70%（39/43）,88.24%（15/17）,95.12%（39/41）,78.95%（15/19）and 90.00%（54/60）respectively.2.MRI showed that the detection rates of abnormal blood vessels in placenta,placental distension,tortuous and thickened blood vessels in myometrium and blurring of placenta-uterus interface in PAS positive patients were significantly higher than those in negative patients,respectively.3.The total score of MRI signs in non-placental accretion,adhesion type,implant type and penetrating type PAS patients increased significantly successively.By drawing the ROC curve,the total score of MRI signs was 20 points as the critical value for the diagnosis of adhesion and non-implantable type,25 points as the critical value for the diagnosis of adhesion and implantable type,and 30 points as the critical value for the diagnosis of implant and penetration type,all of which had high sensitivity and specificity.4.Univariate analysis of the occurrence of PAS showed that age≥35 years old,pre-pregnancy BMI>24kg/m²,number of pregnancies>2,number of births>1,history of abortion,cesarean section,combined placenta previa and history of uterine surgery were all risk factors for the occurrence of PAS.Gestational age and bleeding history in early pregnancy were not risk factors for PAS.5.Multivariate logistic regression analysis was conducted with the occurrence of PAS in pregnant women as dependent variables.The results showed that age,birth time,history of abortion,history of cesarean section,combined placenta previa and history of uterine surgery were all independent risk factors for the occurrence of PAS in pregnant women.