Prenatal prediction of postnatal large-for-date neonates using a simplified method at MR imaging: comparison with conventional 2D ultrasound estimates

Purpose To evaluate the performance of a simple method of estimating fetal weight (EFW) using MR imaging as compared with 2D US in the prediction of large-for-date neonates. Materials & Methods Written informed consent was obtained for this EC-approved study. Between March 2011 and May 2016, 2 groups of women with singleton pregnancies were evaluated: women that underwent US-EFW and MR-EFW within 48 h before delivery and those undergoing these evaluations between 35 + 0 weeks and 37 + 6 weeks of gestation. US-EFW was based on Hadlock et al. and MR-EFW on the formula described by Backer et al. Planimetric measurement of the fetal body volume (FBV) needed for MR-EFW was performed using a semi-automated method and the time required for measurement was noted. Our outcome measure was performance in prediction of large-for-date neonates by MR imaging versus US-EFW using receiver-operating characteristic (ROC) curves. Results 270 women were included in the first part of the study with 48 newborns (17.8%) of birthweight ≥90th centile and 30 (11.1%) ≥95th centile. Eighty-three women were included in the second part with 9 newborns (10.8%) of birthweight ≥95th centile. The median time needed for FBV planimetric measurements in all 353 fetuses was 3.5 (range; 1.5-5.5) min. The area under the ROC curve for prediction of postnatal large-for-date neonates by prenatal MR imaging performed within 48 h before delivery was significantly better than by US (difference between the AUROC = 0.085, P < 0.001; standard error = 0.020 for birthweight ≥90th centile and 0.036, P = 0.01; standard error = 0.014 for birthweight ≥95th centile). Similarly, MR-EFW was better than US-EFW, with both performed remote from delivery, in predicting birthweight ≥ 95th centile (difference between the AUROC = 0.077, P = 0.045; standard error = 0.039). Conclusion MR planimetry using our purpose-designed semi-automated method is not time-consuming. MR-EFW performed immediately prior to delivery or remote from delivery predicts large-for-date neonates significantly better than US-EFW.