Third-Trimester Screening for Preeclampsia in Twins: Comparative Performance of Fetal Medicine Foundation Algorithm, Roche, and Quidel Triages

Abstract

Introduction: We aimed to compare three preeclampsia screening methods in the third trimester: (a) Fetal Medicine Foundation (FMF) multi-marker algorithm (maternal factors, biophysical markers, serum placental growth factor [PlGF], and soluble fms-like tyrosine kinase-1 [sFLT-1]); (b) Roche triage (sFLT-1/PlGF), on fresh and stored samples; and (c) Quidel triage (PlGF) on stored samples. Methods: Women with two live fetuses were enrolled at a twin clinic at 11–13 weeks’ gestation. They visited the clinic every 2–4 weeks until delivery for examination. Blood testing was conducted at enrollment in the first trimester at 11–13 weeks’ gestation, and at 20–22 and 32–34 weeks’ gestations. Here we evaluated the screening characteristics in the third trimester and the detection rate (DR), false-positive rate (FPR), and positive and negative predictive values (PPV and NPV) and compared the screening efficacy based on recommended cutoffs and area under the receiver operation characteristic curve (AUROC). Results: Third-trimester analysis included 146 women with twin pregnancies, of which 24 (16.4%) had preeclampsia. Preeclampsia prediction was conducted by three methods. (1) A prospective prediction was conducted on fresh samples using the FMF algorithm risk assessment in the third trimester for twin pregnancies, which had a DR of 79.2% and a FPR of 5.3%, with PPV and NPV of 76.0% and 95.5%, respectively. (2) Roche triage was tested on fresh samples prospectively using soluble fms-like tyrosine kinase-1 (sFLT-1) versus placental growth factor (PlGF), with a cutoff ratio >38, had a DR of 62.5% and a FPR of 22.1%, with PPV and NPV of 37.5% and 90.7%, respectively. In addition, we repeated the testing with Roche’s Triage using stored samples and the same sFLT-1/PlGF ratio>38. (3) Quidel triage was performed retrospectively using stored samples, acting according to the definition of high risk according to PlGF <12 pg/mL and medium risk according to PlGF 12–100 pg/mL, and subsequently combined the high and medium risk groups into one. The AUROC for the FMF algorithm was 0.91 (95% CI: 0.82–0.99), which was statistically significantly higher (p < 0.0001) compared to the Roche triage of fresh samples at 0.74 (0.61–0.84) and stored samples at 0.76 (0.65–0.86). The AUROC for Quidel triage was 0.75 (0.64–0.86). Conclusions: In twin pregnancy, the third-trimester FMF multi-marker algorithm is more effective for preeclampsia screening than Roche or Quidel triages. It is up to medical centers or professional societies to decide which preeclampsia screening method to adopt, according to local resources. This study describes the strength and limitations of each screening method.