Role of Ultrasound and MRI in Diagnosis of Severe Placenta Accreta Spectrum Disorder: An Intraindividual Assessment With Emphasis on Placental Bulge

BACKGROUND. The "placental bulge" sign (focal area of myometrial-placental bulging beyond the normal uterine contour) on ultrasound (US) or MRI is postulated to represent deeper venous invasion in placenta accreta spectrum (PAS) disorder and may represent severe PAS. OBJECTIVE. The purpose of this study was to evaluate the diagnostic performance and interobserver agreement of US and MRI features for diagnosis of severe PAS, with an emphasis on the placental bulge sign. METHODS. This retrospective study included 62 pregnant women (mean age, 33.2 +/- 5.5 [SD] years) with clinically suspected PAS who underwent both US and MRI. Five readers (two maternal-fetal medicine specialists for US, three abdominal radiologists for MRI) independently reviewed images for the given modality, blinded to the final diagnosis, and recorded the presence of a range of findings (nine on US, eight on MRI), including placental bulge. Intraoperative and pathologic findings were used to separate patients into those with and without severe PAS according to International Federation of Gynecology and Obstetrics classification. Diagnostic performance of US and MRI findings for severe PAS was evaluated, multivariable logistic regression was performed, and interobserver agreement was assessed. RESULTS. A total of 58.1% (36/62) of patients had severe PAS. On US, the finding with the highest accuracy for severe PAS was placental bulge (85.5%), which had a sensitivity of 91.7% and specificity of 76.9%. On MRI, the finding with highest accuracy was also placental bulge (90.3%), which had a sensitivity of 94.4% and specificity of 84,6%. In the multivariable regression analysis, placental bulge was an independent predictor of severe PAS on US (odds ratio [OR], 8.94; p = .02) and MRI (OR, 45.67; p = .003). Interobserver agreement analysis showed a kappa value for placental bulge of 0.48 for MRI and 0.40 for US. Given wide 95% Cls, differences among features for a given modality and differences between modalities were not statistically significant. CONCLUSION. The findings suggest a strong performance of placental bulge in diagnosing severe PAS on both US and MRI, with a potentially stronger performance on MRI. Nonetheless, interobserver agreement remains suboptimal for both modalities. CLINICAL IMPACT. Accurate prenatal diagnosis of severe PAS by imaging could help guide maternal counseling and selection of either hysterectomy or uterine-preserving surgery.