The use of free β-hCG in antenatal screening for Down's syndrome

Abstract

Objective: To investigate the value of the measurement of free beta human chorionic gonadotrophin (hCG) as a serum marker of Down's syndrome in the second trimester of pregnancy. Design: A prospective observational study using stored antenatal serum samples. Setting: Serum samples collected from women receiving routine antenatal care in Oxford. Subjects: Seventy-five singleton pregnancies with fetal Down's syndrome and 367 unaffected singleton pregnancies. Each affected pregnancy was matched with five control pregnancies for maternal age, gestational age, and duration of storage of the serum sample. None of the pregnancies were associated with neural tube defects. Main study measures: Maternal serum free beta-hCG levels. These were compared with total hCG levels in the same pregnancies. The performance of screening using free beta-hCG was compared with that using the principal markers, namely alpha-fetoprotein (AFP), unconjugated oestriol (uE3), and total hCG together with maternal age. Results: The median free beta-hCG level in the affected pregnancies was 2.22 multiples of the normal median (MoM), significantly higher than in the unaffected pregnancies (95% confidence interval, 1.84-2.68 MoM). The discriminatory performances of free beta-hCG and total hCG, each considered separately, were similar; with a cut-off level of 2.5 MoM the detection rate was 43% and 5.7 of unaffected pregnancies had raised free beta-hCG levels (likelihood ratio of 7.5 (43/5.7)), somewhat better discrimination than the 32% and 4.6% respectively using total hCG (likelihood ratio of 7.0 (32/4.6)). With a higher cut-off level of 3.5 MoM, the rates were 19% and 2.7% respectively (likelihood ratio of 7.0), using free beta-hCG, worse than the 19% and 1.4% using total hCG (likelihood ratio of 13.6). Screening using maternal age, AFP, uE3 and free beta-hCG (instead of total hCG) yielded a detection rate of 62% (instead of 58%) at a screening risk cut-off level corresponding to a 5% false-positive rated). Conclusion: The main advantage in using free beta-hCG instead of total hCG is that there is a small increase in the detection rate (4%) for a given false-positive rate when used with maternal age, AFP and uE3. The main disadvantage is that there is less practical experience with free beta-hCG measurement and insufficient data to screen in certain categories of pregnancy (e.g. twins). The best practical advice is to use total hCG for the present but consider changing to free beta-hCG either (i) after further data are available that will permit the interpretation of screening results in the same way as is currently available with total hCG, or (ii) if its use with another marker confers a worthwhile increase in the detection rate for a given false-positive rate.