MR Imaging in the Evaluation of Placental Abruption: Correlation with Sonographic Findings

Purpose: To evaluate the accuracy of magnetic resonance (MR) imaging and color Doppler ultrasonography (US) in the diagnosis of abruption, to assess the accuracy of the different MR imaging sequences in the visualization of clots, and to evaluate the correlation between MR imaging findings and clinical outcome. Materials and Methods: This study protocol was approved by the institutional review board, and written informed consent was obtained from all patients. Between March 2008 and June 2010, 60 consecutive patients (mean gestational age, 30.7 weeks [range, 27-38 weeks]; mean age, 29 years [range, 20-38 years]) who were referred for US and MR imaging owing to a putative diagnosis of abruption were assessed. Multiplanar half-Fourier rapid acquisition with relaxation enhancement, true fast imaging with steady-state precession, three-dimensional T1-weighted gradient-echo MR imaging, and sagittal diffusion-weighted MR imaging were performed. Two radiologists independently reviewed each case, resolving by consensus any diagnostic discrepancy. During a second imaging analysis, the same readers randomly and independently assessed the single sequences. The signal intensity of hematoma was correlated with clinical outcome. The reference standard for abruption was the presence of clots and/or fibrin at inspection of the placenta after delivery. The diagnostic efficacy of US and MR imaging was calculated with 95% confidence intervals. Interobserver agreement was assessed by using the Cohen kappa test. Results: The performance of US and MR imaging was calculated in 39 patients who gave birth less than 10 days after MR imaging; these women were considered to have an adequate reference standard. Abruption was found at delivery in 19 patients. Abruption was identified in 10 of the 19 patients (52%) with US and in all 19 (100%) with MR imaging (P = .002), with an interobserver agreement of 0.949. Diffusion-and T1-weighted sequences helped identify 19 (100%) and 18 (95%) of the 19 abruptions, respectively; interrater agreement was very good for all sequences (k = 0.892-1.0). Hematomas classified as hyperacute or acute worsened to abruption grade II, with the mother being symptomatic or the fetus distressed. Conclusion: MR imaging can accurately depict placental abruption, with excellent interobserver agreement, and should be considered after negative US findings in the presence of late pregnancy bleeding if the diagnosis of abruption would change management. (C) RSNA, 2011