(1167) Preeclampsia associated morbidity requires a double hit from first-trimester maternal characteristics and placental dysfunction.

Friday, January 31, 2025

4:00 PM – 6:00 PM

Submitting Author and Presenting Author(s)

Rebecca Horgan, MD

Assistant Professor
Macon & Joan Brock Virginia Health Sciences Eastern Virginia Medical School at Old Dominion University
Norfolk, VA, United States

Coauthor(s)

EK

Erkan Kalafat, MD, MSc

Associate Professor
Koc University Hospital
Istanbul, Istanbul, Turkey
Elena Sinkovskaya, MD, PhD

Professor
Macon & Joan Brock Virginia Health Sciences Eastern Virginia Medical School at Old Dominion University
Norfolk, Virginia, United States
Alfred Z. Abuhamad, MD

President, Provost and Dean
Macon & Joan Brock Virginia Health Sciences Eastern Virginia Medical School at Old Dominion University
Norfolk, Virginia, United States
George R. Saade, MD (he/him/his)

Professor and Chair, Associate Dean for Women's Health Obstetrics and Gynecology
Macon & Joan Brock Virginia Health Sciences Eastern Virginia Medical School at Old Dominion University
Norfolk, Virginia, United States

Objective: To evaluate whether unsupervised machine learning clustering of first-trimester data can identify pregnancies at high risk for preeclampsia associated morbidity.

Study Design: This is a prospective cohort study which enrolled patients at ≤ 13+6 weeks’ gestation without fetal, umbilical cord or placental abnormalities. Baseline maternal characteristics, blood pressure, and first trimester ultrasound including uterine and placental artery (spiral artery and intervillous arteriole) Dopplers were obtained. The primary outcome was the prevalence of preeclampsia with associated morbidity defined as delivery at < 34 weeks, eclampsia, HELLP syndrome or placental abruption. Patients were followed throughout pregnancy and outcomes collected by trained research coordinators. Maternal characteristics (BMI and blood pressure) and uterine & placental Dopplers (uterine, spiral, and intervillous artery pulsatility index) were used to create their distinct risk groups using unsupervised machine learning models. Outcomes of patients who were labeled as high risk in both (maternal characteristics and Doppler) or just one (Maternal characteristics or Doppler) were compared using cox-proportional hazard models.

Results: Analysis included 617 patients. Machine learning identified 164 patients as high risk of preeclampsia by both maternal characteristics and Dopplers, 126 by maternal characteristics alone, 161 by Dopplers alone, and 166 as low risk. Preeclampsia rates were highest in the dual high-risk group (17.1%), compared to maternal characteristics alone (13.5%) or Dopplers alone (7.5%) (HR: 1.80, 95% CI: 1.07-3.03, P=0.026). The low-risk group had a 2.4% preeclampsia rate. 70% of preeclampsia cases with associated morbidity were in the dual high-risk group, significantly higher than those identified by only one criterion (HR: 4.15, 95% CI: 1.07-16.0, P=0.039).

Conclusion:

Preeclampsia with associated morbidity mostly occurs in patients with abnormalities in both first trimester blood pressure values and placental Doppler values.