Diagnostic Utility of MRI Features of Placental Adhesion Disorder for Abnormal Placentation and Massive Postpartum Hemorrhage.

Abstract

OBJECTIVE. The study's aim was to assess MRI, in cases in which MRI was deemed clinically necessary, to determine its diagnostic accuracy for placental adhesion disorder (PAD) and prognostic accuracy for massive postpartum hemorrhage (PPH). Additionally, we investigated the diagnostic utility of MRI in the antenatal workup of PAD as an adjunct to clinical assessment and ultrasound. MATERIALS AND METHODS. We retrospectively identified patients who underwent antenatal MRI for suspicion of PAD. Images were reviewed by two radiologists who were blinded to surgical and pathologic outcomes. Diagnostic utility of various clinical, ultrasound, and MRI features of PAD were estimated by ROC analysis. Logistic regression analysis was performed to assess various diagnostic models for PAD and prognostic models for massive PPH, with model selection based on Bayesian information criterion. RESULTS. Fifty-six patients met the inclusion criteria. Sensitivity and specificity of MRI in the diagnosis of PAD were 93% and 81%, respectively. The most accurate MRI features for PAD were myometrial thinning (AUC = 0.881), heterogeneous placenta (AUC = 0.864), and placental bulge (AUC = 0.845). The most accurate MRI features for massive PPH were heterogeneous placenta (AUC = 0.872) and dark intraplacental bands (AUC = 0.736). The addition of MRI to a model based on clinical and ultrasound features was preferred for both diagnosis of PAD and prognosis of massive PPH. CONCLUSION. This study shows the utility of certain MRI features for identification of PAD and massive PPH. Furthermore, our data show a substantial incremental benefit of the addition of MRI in the antenatal workup for PAD compared with clinical assessment and ultrasound alone.