López Mármol Rocío, Ávila Cabreja José Alejandro, de Haro Romero Teresa, de Paco Matallana Catalina, Ocón Hernández Olga, González-Vanegas Otilia, Carretero Lucena Pilar, Carrillo María Paz, Delgado Juan Luis, Rolle Valeria, Gormus Uzay, Oraha Liza, Gil María M, Molina Francisca S
Department of Obstetrics and Gynecology, Hospital Universitario Virgen de las Nieves, Granada, Spain.
Instituto de Investigación Biosanitaria ibs.GRANADA, Granada, Spain.
Acta Obstet Gynecol Scand. 2025 Oct;104(10):1897-1906. doi: 10.1111/aogs.70026. Epub 2025 Jul 23.
Our objective was to compare the predictive performance of the Fetal Medicine Foundation (FMF) competing-risk model for preterm preeclampsia (PE) screening using placental growth factor (PlGF) measurements obtained at 11-13 weeks versus before 11 weeks of gestation.
This multicenter prospective cohort study included women with singleton pregnancies attending their routine first-trimester assessment (11 to 13 weeks) in four hospitals across Spain from 2021 to 2023. Maternal characteristics, biophysical parameters (mean arterial pressure and uterine artery pulsatility index), and biochemical markers (PlGF measured twice in each woman, before 11 weeks and between 11 and 13 weeks) were assessed. Risk assessment for preterm PE was estimated by the FMF algorithm. Predictive performance was evaluated by comparing detection rates (DR) at different fixed screen-positive rates (SPR), area under the receiver-operating characteristic curve (AUROC), and calibration plots. Statistical adjustments were made to account for prophylactic aspirin use.
The study population comprised 3448 women, including 19 (0.55%) who developed preterm preeclampsia and 47 (1.36%) who developed term preeclampsia. At 10% SPR, the detection rates (adjusted for aspirin use) were highest for the model incorporating PlGF between 11 and 13 weeks (72.9%; 95% CI, 42.2%-90.9%), compared to models with PlGF before 11 weeks (66.4%; 95% CI, 39.9%-85.4%) and without PlGF (66.0%; 95% CI, 39.3%-85.3%). Similar trends were observed at higher SPR thresholds. The best discrimination (AUROC: 0.863; 95% CI, 0.754-0.971) and calibration were also achieved by the model using PlGF between 11 and 13 weeks.
PlGF measured before 11 weeks did not improve preterm PE screening performance. Due to the small number of cases, further validation is needed. Maternal and biophysical markers remain a viable alternative when PlGF is unavailable.
我们的目标是比较胎儿医学基金会（FMF）竞争风险模型在使用妊娠11至13周时获得的胎盘生长因子（PlGF）测量值与妊娠11周前获得的测量值进行早发型子痫前期（PE）筛查时的预测性能。
这项多中心前瞻性队列研究纳入了2021年至2023年在西班牙四家医院进行常规孕早期评估（11至13周）的单胎妊娠妇女。评估了产妇特征、生物物理参数（平均动脉压和子宫动脉搏动指数）以及生化标志物（每位妇女在11周前和11至13周期间各测量两次PlGF）。通过FMF算法估计早发型PE的风险评估。通过比较不同固定筛查阳性率（SPR）下的检测率（DR）、受试者操作特征曲线下面积（AUROC）和校准图来评估预测性能。进行了统计调整以考虑预防性使用阿司匹林的情况。
研究人群包括3448名妇女，其中19名（0.55%）发生早发型子痫前期，47名（1.36%）发生足月子痫前期。在10%的SPR时，纳入11至13周PlGF的模型的检测率（经阿司匹林使用调整后）最高（72.9%；95%CI，42.2%-90.9%），相比之下，11周前有PlGF的模型（66.4%；95%CI，39.9%-85.4%）和没有PlGF的模型（66.0%；95%CI，39.3%-85.3%）。在更高的SPR阈值下观察到类似趋势。使用11至13周PlGF的模型也实现了最佳辨别力（AUROC：0.863；95%CI，0.754-0.971）和校准。
11周前测量的PlGF并未改善早发型PE的筛查性能。由于病例数量较少，需要进一步验证。当无法获得PlGF时，产妇和生物物理标志物仍然是可行的替代方法。
