Magnetic resonance imaging for abnormally invasive placenta: the added value of intravenous gadolinium injection. (27th June 2016)

• Record Type: Journal Article
• Authors: Millischer, A‐E; Salomon, LJ; Porcher, R; Brasseur‐Daudruy, M; Gourdier, A‐L; Hornoy, P; Silvera, S; Loisel, D; Tsatsaris, V; Delorme, B; Boddaert, N; Ville, Y; Sentilhes, L
• Is Part Of: BJOG. Volume 124:Number 1(2017:Jan.)
• Page Start: 88
• Page End: 95
• Publication Date: 2016-06-27
• Subjects: Abnormally invasive placenta -- diagnosis -- gadolinium injection -- magnetic resonance imaging -- placenta accreta -- placenta percreta; Obstetrics -- Periodicals; Gynecology -- Periodicals; 618
• DOI: 10.1111/1471-0528.14164
• Languages: English
• ISSNs: 1470-0328

Abstract

Abstract : Objectives: To assess the added value of intravenous gadolinium injection to magnetic resonance imaging (MRI) ‐based diagnosis of abnormally invasive placenta (AIP) and to examine this in relation to the radiologist's experience. Design: Retrospective study. Setting: Between March 2009 and October 2012, 31 pregnant women who had previous caesarean delivery together with a placenta praevia and suspected placenta accreta on ultrasound in the third trimester of pregnancy. Population: All were offered MRI examination, and made aware of the limited (but so far reassuring) data regarding fetal safety of gadolinium. Twenty pregnant women agreed to undergo prenatal MRI (1.5 T), with and without gadolinium injection. Methods: Two sets of MRI examinations without and with gadolinium were reviewed independently 2 months apart by two senior and two junior radiologists; all were blinded to the outcome (known in all cases). Histopathological findings and clinical signs of AIP were considered as the defining criteria of diagnosis. Main outcome measure: accuracy of MRI with and without gadolinium was assessed. Results: Eight of the 20 women had confirmed abnormal placental invasion. The overall performance of both sets of readers in detecting AIP increased with gadolinium—sensitivity and specificity of 75.0% (42.0–100%) and 47.9% (19.9–75.9%) increasing to 87.5% (57.1–100%) and 60.4% (33.9–86.9%), respectively ( P = 0.04). The added value of gadolinium remained irrespective of radiologist's experience, although senior radiologists performed better overall (sensitivity and specificity of 87.5% and 62.5% versus 62.5% and 33.3%, respectively, increasing with injection to 93.8% and 70.8% versus 81.3% and 50%, respectively; P < 10 −4 ). Conclusion: There was an association between gadolinium use and improvement in MRI‐based diagnostic accuracy for the diagnosis of AIP, for both junior and senior radiologists. Tweetable abstract: Gadolinium injection improves MRI performance of radiologists for the diagnosis of placenta accreta. Tweetable abstract: Gadolinium injection improves MRI performance of radiologists for the diagnosis of placenta accreta.