Objective: To determine the value of serum screening for Down's syndrome at 8-14 weeks of pregnancy using seven potential serum markers (alpha-fetoprotein, unconjugated oestriol, total human chorionic gonadotrophin (hCG), free alpha-hCG, free beta-hCG, pregnancy associated plasma protein A (PAPP-A), and dimeric inhibin A).
Design: Stored blood samples collected from women at about 10 weeks of pregnancy, prior to having chorionic villus sampling procedure on account of advanced maternal age, were retrieved from pregnancies associated with Down's syndrome and from matched unaffected pregnancies.
Setting: Twenty-one obstetric centres in nine countries.
Subjects: Seventy-seven pregnancies associated with Down's syndrome each matched with five controls (except in two cases that were matched with four controls) for maternal age (same five year age groups), duration of storage of the serum sample (same calendar year), and gestational age (usually same week of pregnancy).
Results: The levels of two potential markers differed between affected and unaffected pregnancies sufficiently to be of value in screening--free beta-hCG and PAPP-A. The median free beta-hCG level in affected pregnancies was 1.79 times the median level for unaffected pregnancies, and the median PAPP-A level was 0.43 times the normal median. These two markers were combined with maternal age to estimate a woman's risk of having a fetus with Down's syndrome. A screening programme that used a risk cutoff level of 1:300 would detect 63% of affected pregnancies and also classify 5.5% of unaffected pregnancies as screen positive. None of the other five markers added more than 2% detection for the same false-positive rate.
Conclusion: The performance of screening using maternal age and serum-free beta-hCG and PAPP-A at 10 weeks of pregnancy was better than the double test (alpha-fetoprotein and hCG with maternal age) and similar to the triple test (alpha-fetoprotein, unconjugated oestriol and hCG with maternal age) at 15-22 weeks.