The report by McCowan and colleagues that low-dose aspirin did not increase birthweight when used to treat a small fetus with an abnormal umbilical Doppler is noted with interest. This finding would appear to contradict the first study of this type by myself1 and a recent meta-analysis2. I write to highlight some important limitations of this present study.
The authors chose to study fetuses with an abnormal umbilical Doppler and a small size on ultrasound. The authors cite a similar study3 to support their findings. These studies were made on a very different group to that reported in our study. We selected pregnancies in which the umbilical Doppler was abnormal without a requirement for an already established fetal effects (growth restriction) prior to treatment. We did not report 75% as having growth restriction at the commencement of therapy. It was the finding of our aspirin trial that the “early” use of aspirin benefited fetal growth. Indeed in our trial we reported that the most extremely affected fetuses did not benefit from treatment.
The question of umbilical Doppler and established fetal effect (growth restriction - small abdominal circumference) confounding the results of clinical trials can be examined in another context to highlight the importance of the difference in patient selection. The clinical utility of umbilical Doppler in high risk pregnancy is supported by meta-analysis which suggests an impressive 32% reduction in perinatal mortality4. Of the many randomised trials carried out there is one which reports contradictory findings. In that study by Newnham et al.5 women were enrolled if the fetal abdominal circumference was small (i.e. already established growth restriction). Umbilical Doppler was of no added value. In our first trial evaluating umbilical Doppler in clinical practice6 we suggested that the value of umbilical Doppler study was early identification of a fetus at risk so that clinical attention could be focused on a group of fetuses with risk recognised early. Using a fetal abdominal circumference measurement means that the fetal risk had already been identified and a fetal effect of the placental vascular disease has occurred.
The dosage of aspirin may be important. I have always been impressed by the fact that the largest benefit in terms of fetal weight increase was associated with the use of the largest dose of 150 mg of aspirin. The study by Beaufils et al.7 used that dosage and influenced the design of our trial. The CLASP8 study used only 60 mg of aspirin. Since the umbilical Doppler study signals a problem in the fetal umbilical placental circulation we have suggested that the larger dose is important to ensure activity in the fetal circulation. We have shown fetal platelet consumption is increased in these cases9 with an abnormal umbilical Doppler and speculated that the aspirin needs to exert an effect in the fetal circulation. Maternal thromboxane concentration after aspirin therapy does not correlate with maternal or fetal benefit10. There is not a similar study of fetal effect related to fetal thromboxane levels.
The assertion by McCowan and colleagues that the beneficial effect in our study “might be explained by chance” is not established by their study which is different in both patient selection and aspirin dosage.