Poster Session 4
Abraham Tsur, MBA, MD (he/him/his)
Director of The Women's Health Innovation Center
Sheba Medical Center, Ramat Gan, Tel Hashomer, Israel
Ramat Gan, HaMerkaz, Israel
David Nadav Sabag, BSc
Medical intern
Sheba Medical Center, Ramat Gan, Tel Hashomer, Israel
Givatayim, HaMerkaz, Israel
Ronen Fluss, PhD
Gertner Institute
Ramat Gan, HaMerkaz, Israel
Amit Huppert, PhD
Gertner Institute
Ramat Gan, HaMerkaz, Israel
Michal Fishel Bartal, MD (she/her/hers)
Maternal Fetal Medicine Faculty
UTH Houston & Sheba Medical Center Israel
Houston, TX, United States
Dan Dominissini, PhD
The sheba Medical Center
Ramat Gan, HaMerkaz, Israel
Yoav Yinon, MD
Head of Fetal Medicine Unit
Chair of the Israeli Maternal Fetal Medicine Society
The Sheba Medical Center
Ramat Gan, HaMerkaz, Israel
Shali Mazaki Tovi, MD
The Sheba Medical Center
Ramat Gan, HaMerkaz, Israel
Rakefet Yoeli-Ullman, MD
Sheba Medical Center
Ramat Gan, HaMerkaz, Israel
To determine real-world accuracy of circulating angiogenic factors in predicting preeclampsia with severe features (sPE) in singleton and twin gestations.
Study Design:
This retrospective cohort study consisted of individuals admitted to the antepartum unit of a tertiary academic center from 2023 to 2024. We included individuals with hypertensive disorders of pregnancy (HDP) at 23+0 to 34+6 weeks of gestation with available sFlt-1/PlGF measurements obtained as part of the clinical flow. Exclusion criteria included developing sPE on the day of measurement. The area under the curve (AUC) was used to assess accuracy of the prediction of sPE within two weeks. The discriminatory ratio of ≥40 was evaluated for clinical performance. Similar to the PRAECIS study, measurements were re-included if delivery did not occur within two weeks. In addition, we developed a logistic regression model integrating sFlt-1/PlGF with blood pressure (BP). Prediction of sPE by the integrated model was compared to prediction based on sFlt-1/PlGF alone using AUC and likelihood ratio test (LRT).
Results:
The study group included 59 individuals (71 sflt-1/PlGF measurements). In 35% of measurements, the individuals developed sPE within two weeks. Table 1 reports the discriminatory clinical performance of the ≥40 ratio stratified by singleton and twin pregnancies. The ≥40 ratio showed better sensitivity and NPV in twins than in singletons. Moreover, although specificity in twins was lower than in singletons, the PPV was higher in twins due to the increased prevalence of sPE. Integrating sFlt-1/PlGF with BP measurement increased the model accuracy (AUC 0.83 VS 0.88, figure 1) and enhanced the model fit (LRT p=0.005).
Conclusion:
Real-world evidence supports extending the clinical use of the sFlt-1/PlGF ratio to assess the risk of developing sPE in twin pregnancies as well. Integrating sFlt-1/PlGF with clinical risk factors like BP can enhance predictive performance and clinical decision-making.
