Prediction of preeclampsia and its prevention with aspirin

Background: Preeclampsia (PE) affects 2-3% of all pregnancies and is a major cause of maternal and perinatal morbidity and mortality. The current approach to screening for PE is based on the identification of risk factors from maternal characteristics and medical history. This approach, however, fails to identify a high proportion of cases of PE and does not provide individualised, patient-specific results. An alternative approach is to combine maternal factors with biophysical and biochemical markers to estimate the individual probability of developing PE with higher detection rates. To date, no intervention is proven to reduce the risk of the disease, and several studies evaluating the use of aspirin for prevention of PE led to inconclusive results. Objectives: The aims of the studies included in this thesis are, first, to prospectively validate in a large European population a first-trimester algorithm for prediction of PE that combines maternal demographic characteristics and medical history with biophysical and biochemical markers; second, to compare this method of screening to the performance of currently used guidelines; third, to evaluate a possible beneficial effect of aspirin initiated at 11 to 14 weeks of gestation and at a dose of 150 mg in the prevention of PE in a multicentre, double-blind, placebo-controlled randomised trial; and fourth, to analyse a potential role of cell-free DNA testing in the prediction of PE. Methods: Combined screening for PE was applied in the first or second trimester, and women found to be at high-risk in the first trimester were offered participation in a double-blind trial of aspirin against placebo in six European countries. We have recorded maternal characteristics and history, measured the uterine artery pulsatility index (UtPI) on ultrasound, the mean arterial pressure (MAP), serum concentration of pregnancy-associated plasma protein A (PAPP-A) and placental growth factor (PLGF). Pregnancy outcomes were obtained from the hospital maternity records. Bayes theorem was used to combine the a priori risk from maternal factors with the results of biomarker measurements and estimate individual probabilities. In the randomised trial, the analysis was performed in an intention-to-treat basis and the treatment effect on the primary outcome (the development of PE with delivery before 37 weeks of gestation) was reported with 95% confidence interval (CI), and on secondary outcomes with 99% CI. Cell-free DNA fetal fraction was compared with other first trimester markers for PE and in a case-control study. Results: Detection rates of combined screening, for a false-positive rate (FPR) of 10%, were 89% (95% CI 79-96%), 75% (95% CI 70-80%) and 47% (95% CI 44- 51%) for PE