@article{3123073,
    title = "MRI features predictive of invasive placenta with extrauterine spread in high-risk gravid patients: A prospective evaluation",
    author = "Bourgioti, C. and Zafeiropoulou, K. and Fotopoulos, S. and Nikolaidou, M.E. and Antoniou, A. and Tzavara, C. and Moulopoulos, L.A.",
    journal = "AJR.  AMERICAN JOURNAL OF ROENTGENOLOGY",
    year = "2018",
    volume = "211",
    number = "3",
    pages = "701-711",
    publisher = "American Roentgen Ray Society",
    issn = "0361-803X",
    doi = "10.2214/AJR.17.19303",
    keywords = "adult;  Article;  clinical article;  diagnostic accuracy;  echography;  ectopic pregnancy;  female;  gestational age;  high risk patient;  histology;  human;  intraoperative period;  nuclear magnetic resonance imaging;  observational study;  placenta disorder;  prediction;  predictive value;  priority journal;  prospective study;  radiologist;  sensitivity and specificity;  bladder;  diagnostic imaging;  pathology;  peritoneum;  placenta previa;  pregnancy;  prenatal diagnosis;  receiver operating characteristic;  reproducibility, Adult;  Female;  Humans;  Magnetic Resonance Imaging;  Peritoneum;  Placenta Previa;  Predictive Value of Tests;  Pregnancy;  Prenatal Diagnosis;  Prospective Studies;  Reproducibility of Results;  ROC Curve;  Urinary Bladder",
    abstract = "OBJECTIVE. The objective of our study was to evaluate MRI diagnostic ability in predicting invasive placenta with extrauterine spread in high-risk gravid patients. SUBJECTS AND METHODS. Between March 2016 and June 2017, 49 patients (mean age, 35.7 years; mean gestational age, 32.5 weeks) with sonographically confirmed placenta previa underwent dedicated MRI. All MRI examinations were reviewed by two experienced radiologists prospectively. Intraoperative and pathologic findings were the standard of reference. Kappa values were calculated to assess the agreement between MRI findings and histologic results as well as interrater reliability. ROC curve analysis was used to test the discriminative ability of MRI features for invasive placenta with extrauterine spread. Stepwise multiple logistic regression analysis was performed to identify any MRI findings predictive of invasive placenta and of bladder and parametrial involvement. RESULTS. MRI exhibited significant overall accuracy (AUC = 0.77, p = 0.006) in identifying invasive placenta with 100% sensitivity and negative predictive values; it was highly specific (100%) in identifying placental extension to both bladder and parametrial tissues. Lumpy tapering of the placental edges, intraplacental dark T2 bands, prominent intraplacental vascularity, and serosal hypervascularity were independently associated with an increased risk for invasive placenta. Serosal hypervascularity and vesicouterine space hypervascularity were independent predictors of bladder invasion; abnormal vascularization within the parametrial fat was significant for parametrial invasion. CONCLUSION. MRI is highly accurate in depicting placental extrauterine spread. The presence of abnormal vessels at the uterine serosa was the most important MRI feature for identifying invasive placenta. An abnormal vascular network within the vesicouterine space or parametrium was the most reliable MRI sign for detecting bladder or parametrial involvement. © American Roentgen Ray Society."
}