Prediction of Preeclampsia-Related Adverse Outcomes With the sFlt-1 (Soluble fms-Like Tyrosine Kinase 1)/PlGF (Placental Growth Factor)-Ratio in the Clinical Routine: A Real-World Study

• Published in: Hypertension (Dallas, Tex. 1979) Vol. 77; no. 2; pp. 461 - 471
• Main Authors: Dröge, Lisa Antonia, Perschel, Frank Holger, Stütz, Natalia, Gafron, Anna, Frank, Lisa, Busjahn, Andreas, Henrich, Wolfgang, Verlohren, Stefan
• Format: Journal Article
• Language: English
• Published: United States American Heart Association, Inc 01.02.2021
• Subjects: maternal death; blood pressure; pregnancy outcome; preeclampsia; HELLP syndrome
• ISSN: 0194-911X; 1524-4563; 1524-4563
• DOI: 10.1161/HYPERTENSIONAHA.120.15146

This retrospective real-world study investigated the clinical use of the sFlt-1 (soluble fms-like tyrosine kinase 1)/PlGF (placental growth factor) ratio alone or in combination with other clinical tests to predict an adverse maternal (maternal death, kidney failure, hemolysis elevated liver enzymes low platelets-syndrome, pulmonary edema, disseminated intravascular coagulation, cerebral hemorrhage, or eclampsia) or fetal (delivery before 34 weeks because of preeclampsia and/or intrauterine growth restriction, respiratory distress syndrome, necrotizing enterocolitis, intraventricular hemorrhage, placental abruption or intrauterine fetal death or neonatal death within 7 days post natum) pregnancy outcome in patients with signs and symptoms of preeclampsia. We evaluated the sFlt-1/PlGF-ratio cutoff values of 38 and 85 and evaluated its integration into a multimarker model. Of 1117 subjects, 322 (28.8%) developed an adverse fetal or maternal outcome. Patients with an adverse versus no adverse outcome had a median sFlt-1/PlGF-ratio of 177 (interquartile range, 54–362) versus 14 (4–64). Risk-stratification with the sFlt-1/PlGF cutoff values into high- (>85), intermediate- (38–85), and low-risk (<38) showed a significantly shorter time to delivery in high- and intermediate- versus low-risk patients (4 versus 8 versus 29 days). When integrating all available clinical information into a multimarker model, an area under the curve of 88.7% corresponding to a sensitivity, specificity, positive and negative predictive value of 80.0%, 87.3%, 75.0%, and 90.2% was reached. The sFlt-1/PlGF-ratio alone was inferior to the full model with an area under the curve of 85.7%. As expected, blood pressure and proteinuria were significantly less accurate with an area under the curve of 69.0%. Combining biomarker measurements with all available information in a multimarker modeling approach increased detection of adverse outcomes in women with suspected disease.