From the Journals

Combination model predicts imminent preeclampsia



A triple test was a significantly more effective predictor of preeclampsia than was either angiogenic placental growth factor (PlGF) alone or the antiangiogenic factor soluble fms-like tyrosine kinase-1(sFLT)/PlGF ratio, based on data from more than 15,000 pregnancies.

A pregnant woman is being tested for high blood pressure by her doctor. Jovanmandic/Getty Images

The use of either PlGF or sFLT/PlGF alone to predict preeclampsia fails to account for individual maternal risk factors or the measurement of blood pressure at presentation, wrote Anca Ciobanu, MD, of King’s College London, and her colleagues.

In a study published in the American Journal of Obstetrics & Gynecology, the researchers reviewed data from 15,247 singleton pregnancies with live births of healthy babies and compared the preeclampsia detection rates of PlGF, sFLT/PlGF and a competing risks model that included a combination of maternal factors and median values of PlGF, sFLT, and mean arterial pressure (triple test). Preeclampsia developed in 2.1% of pregnancies.

Overall, the triple-test detection rate for delivery with preeclampsia was 10% higher than the sFLT/PlGF ratio and 20% higher than PlGF alone based on assessment at 2 weeks or less or 4 weeks or less before delivery. The negative predictive value was similar for the three tests.

At 2 weeks or less before delivery, the area under the receiver operating characteristic curves (AUROC) for preeclampsia was significantly higher for the combination model (0.975), compared with PlGF (0.900) or the sFLT/PlGF ratio (0.932), with P less than .0001 in each case. Similarly, at 4 weeks or less before delivery, the AUROC for preeclampsia was 0.907 for the triple test, 0.827 for PlGF alone, and 0.857 for the sFLT/PlGF ratio, with P less than .0001 in each case.

The competing risks model allows clinicians more flexibility to identify patients at increased risk by considering factors including maternal characteristics and variations from normal blood pressure, Dr. Ciobanu and her associates noted. Also, the combination model, “can form the basis of future research that would quantify and incorporate into the model, symptoms such as headache and epigastric pain, as well as proteinuria, creatinine, liver enzymes and platelets.”

The study findings were limited by several factors including the potential predictive value of screening for women with hypertensive symptoms attending specialist clinics, and whether mean arterial pressure would be an effective measure in patients seen at these clinics, the researchers noted. However, the results support the value of the competing risks model, which “provides a personalized risk for delivery with preeclampsia that could lead to personalized stratification of the intensity of monitoring and timing of delivery.”

The study was supported by a grant from the Fetal Medicine Foundation; Thermo Fisher Scientific provided the reagents and equipment. The researchers had no financial conflicts of interest.

SOURCE: Ciobanu A et al. Am J Obstet Gynecol. 2019 Feb 7. doi. org/10.1016/j.ajog.2019.01.235.

Next Article: