Macrosomia-related adverse outcomes may be reduced by utilization of magnetic resonance imaging (MRI) and ultrasound (US)-based models among a primigravid population, according to a recent study in the American Journal of Obstetrics & Gynecology.1
Takeaways
- MRI and ultrasound models improve the accuracy of predicting when a fetus will reach a weight of 4000 grams or more, aiding in better delivery planning.
- MRI-based models were found to be more accurate than ultrasound models, with an area under the curve of 0.96 compared to 0.91.
- Using these predictive models can help reduce the risks of cesarean delivery, postpartum hemorrhage, and shoulder dystocia associated with macrosomia.
- Important factors for the models include prepregnancy BMI, gestational age at birth, geographic origin, gestational diabetes, and fetal body volume for MRI models.
- The study included a diverse group of primigravid women, predominantly from the Middle East, North Africa, and Europe, with significant improvements in outcome prediction using the new models.
Macrosomia, defined by neonatal birthweight (BW) of over 4000 to 4500 g, has been linked to maternal and neonatal risks including cesarean deliver (CD) and operative vaginal delivery (OVD). Traumatic deliveries can lead to additional adverse outcomes such as extensive genital tract and postpartum hemorrhage (PPH).
Large for gestational age (LGA), defined by an estimated fetal weight (EFW) or BW above the 90th or 95th percentile for gestational age (GA), is a condition similar to macrosomia. However, discrepancies between growth charts can lead to LGA infants on one chart being diagnosed as normal weight on another.
Gestational weight gain and gestational diabetes mellitus (GMD) have been linked to increased macrosomia risk.2 One study reported LGA risk increases from 1.33% to 24.71% from a body mass index (BMI) of under 19 kg/m2 at GDM diagnosis to 33 kg/m2 or above at GDM diagnosis.
Decision-making about optimal delivery timing can be made by accurately predicting when a fetus would reach a weight of 4000 g.1 However, current models are lacking in accuracy.
Investigators conducted a study to evaluate the performance of MRI- and US-based models at predicting the GA when a BW of 4000 g or greater would be attained. Data was obtained from the the PREdict neonatal MACROsomia study, which compared prediction of singleton pregnancies at 36 weeks’ gestation.
Participants included primigravida women eligible for normal vaginal delivery (NVD). Women with elective CD, preeclampsia, multiparity, or contraindications to NVD were excluded. Collected data included age, prepregnancy BMI, smoking, geographic origin, assisted reproductive technologies, GDM, and diabetes mellitus type 1 or 2 (DM1 or DM2).
Intrapartum CD, OVD, anal sphincter injury, PPH, brachial plexus injury, shoulder dystocia, 5-minute Apgar score under 7, neonatal intensive care unit admission, and intracranial hemorrhage were assessed as adverse outcomes. US examination was performed at 36 weeks’ gestation using transabdominal sonography.
Women received fetal MRI within 15 minutes following US. Additionally, total fetal body volume (FBV) was measured.
Participants had a median maternal age of 27.12 years and a median prepregnancy BMI of 23.92 kg/m2. Of participants, 42.8% were from the Middle East or North Africa, 41.4% from Europe, 13.5% from sub-Saharan Africa, and 2.3% from other geographic regions. GDM was reported in 21%, while DM1 or DM2 were reported in only 0.3%.
Induction of labor was reported in 37.8% of patients and a BW of 4000 g or more in 8.7%. The median GA at birth was 39.86 weeks and the median BW was 3340 g.
The optimal MRI-based model used prepregnancy BMI, GA at birth, geographic origin, and FBV. The optimal US-based model used GDM, DM1 or DM2, geographic origin, maternal age, GA at birth, fetal gender, and EFW.
Data indicated significantly improved performance from the MRI-based model vs the US-based model for predicting BW of 4000 g or more, with an area under the curve (AUC) of 0.96 vs 0.91, respectively. For the testing datasets, the AUCs were 0.98 vs 0.89, respectively.
When applying a prediction formula from the MRI-based model, delivery after the predicted GA was reported among 5.5% of patients. The risks of CD, PPH, and shoulder dystocia were increased in these patients, with adjusted odds ratios (aORs) of 3.15, 4.50, and 9.67, respectively.
When applying a prediction formula from the US-based model, delivery after the predicted GA was reported among 3.4% of patients. The risks of intrapartum CD and PPH were increased in these patients, with aORs of 5.27 and 6.74, respectively.
Rates of macrosomia diagnosis in the high-risk group were 87.5% for those delivering after predicted GA from the MRI-based model and 60% from the US-based model.
These results indicated a reduction in macrosomia-related adverse outcomes from MRI- and US-based models for predicting a GA when BW will exceed 4000 g. Investigators concluded the MRI-based model is superior for identifying the highest-risk patients.
References
- Badr DA, Cannie MM, Kadji C, et al. Reducing macrosomia-related birth complications in primigravid women: ultrasound- and magnetic resonance imaging–based models. Am J Obstet Gynecol. 2024;230:557.e1-8. doi:10.1016/j.ajog.2023.10.011
- Krewson C. Neonatal outcomes impacted by weight gain during gestational diabetes. Contemporary OB/GYN. August 28, 2023. Accessed May 23, 2024. https://www.contemporaryobgyn.net/view/neonatal-outcomes-impacted-by-weight-gain-during-gestational-diabetes
