Inaccurate estimation of risk in second trimester serum screening for Down syndrome among women who have already had first trimester screening

Problems can arise in prenatal screening for Down syndrome when tests are performed in the first and second trimester and some women who have a negative first trimester test have a second trimester serum test. The second test result does not usually take account of the previous one being negative. Even if it does, it is often inaccurate. Using published data the extent of the error was examined. The age-specific risk of an affected pregnancy in such women will be lower than if no first trimester test had been performed. The distributions of the screening markers in affected and unaffected pregnancies will be different from those in unscreened women. If the appropriate age-specific risk and marker distributions are not used, error will arise. For example, a 35-year-old woman with nuchal translucency (NT), pregnancy-associated plasma protein-A (PAPP-A) and free β-human chorionic gonadotrophin (hCG) levels at the normal median would have a risk of 1 in 6500. If she then had the Triple Test with alpha-fetoprotein (AFP), unconjugated oestriol, and hCG levels of 0.7, 0.7 and 1.5 multiples of the median (MoM), respectively, her risk, ignoring the previous result, would be overestimated (1 in 95 compared with the correct estimate of 1 in 705). If the previous result was included, but the age-specific risk and second trimester marker distributions were not revised, her risk would be underestimated (1 in 820). If the correct age-specific risk and screening marker distributions were used, risk estimates would be accurate, but two tests would be less efficient than integrating all the screening information into a single test. The practice of offering second trimester serum screening to women who have already been screened is best avoided. Copyright © 2001 John Wiley & Sons, Ltd.     

The first trimester ‘combined test’ for the detection of Down syndrome pregnancies in 4939 unselected pregnancies

The high detection rate (DR) for Down syndrome (DS) pregnancies which can be achieved by measuring fetal nuchal translucency (NT) early in pregnancy can be improved by combining it with placental hormones [pregnancy-associated plasma protein A (PAPP-A) and free β-human chorionic gonadotrophin (fβ-hCG)] and maternal age (‘combined test’). In this study we wanted to assess the DR using the ‘combined test’ in an unselected population of self-referred pregnant women at a false-positive rate (FPR) of about 5%. NT, PAPP-A, fβ-hCG and maternal age were measured in all women with singleton pregnancies who booked for delivery in our hospital from 1 December 1997 to 31 April 2000 and who were between 10 and 13 completed weeks of gestation [crown–rump length (CRL) 35–70 mm]. The specific DS risk was calculated using the computer program Alpha Version 5aa (Logical Medical Systems, London, UK). A total of 4939 women were tested. Out of 14 DS pregnancies that occurred during this period of time, 12 were detected with the test. A total of 246 women had a false-positive test result in a non-DS pregnancy (FPR 5.0%). This makes the ‘combined test’ by far the best test for the detection of DS pregnancies in a low-risk population. The constant increase in maternal age at the time of delivery can also lead to an improved DR if a simple age-dependant protocol for DS detection is used, but only at the price of a much higher number of amniocenteses and subsequent abortions. The DR for DS can be increased much more markedly using the ‘combined test’ with a FPR that still remains at the level as it was in the early 1970s. Copyright © 2002 John Wiley & Sons, Ltd.     

Gender impact on first trimester markers in Down syndrome screening

The influence of fetal gender on the level in the first trimester of the serological markers alpha-fetoprotein (AFP), pregnancy-associated plasma protein-A (PAPP-A) and free beta human chorionic gonadotropin (βhCG) and on nuchal translucency is described for 2637 singleton pregnancies with normal outcome. Mean log MoM values for pregnancies with female and male fetuses were calculated using regression of log marker values on gestational age expressed as crown rump length and on maternal weight. A pronounced gender impact was found for free βhCG, being 16% higher for female than for male fetuses. Copyright © 2002 John Wiley & Sons, Ltd.     

Crown—rump length in aneuploid fetuses: Implications for first-trimester biochemical screening for aneuploidies

This study examined the effect of estimation of gestational age from the menstrual history compared with that from crown—rump length (CRL) measurement on the detection rate of screening for aneuploidies in the first trimester. Pregnancy-associated plasma protein A (PAPP-A) was assayed in blood collected prior to chorionic villus sampling in 356 women with unaffected pregnancies and 28 women with an aneuploid pregnancy. There were 14 Down's syndrome (DS) pregnancies. All pregnancies were dated from menstrual history and CRL measurement. The average CRL gestation in the aneuploid population was 2.5 days less than that derived from the LMP (95 per cent confidence interval (CI) for LMP—CRL gestation: using the algorithm based on unaffected pregnancies 0–3.5 days; using the matched case—control approach 1–4.5 days). The average CRL gestation in the DS population was 2 days less but this did not reach statistical significance (95 per cent CI for LMP—CRL gestation: using the algorithm — 1 to 4.5 days; using the matched case—control approach 0 to 5.5 days). The detection rate of aneuploidies in the first trimester using maternal serum PAPP-A was reduced by 7 per cent (and by 3 per cent for DS) for a 5 per cent false-positive rate when using CRL rather than LMP to date the pregnancy. This phenomenon is a consequence of an apparent reduction of gestational age when estimated by CRL in the aneuploid population. Further studies are required to evaluate whether CRL is an unbiased estimate of gestation for Down's syndrome pregnancies.