Estimating an individual's risk of having a fetus with open spina bifida and the value of repeat alpha-fetoprotein testing: FOURTH REPORT OF THE UK COLLABORATIVE STUDY ON ALPHA-FETOPROTEIN IN RELATION TO NEURAL TUBE DEFECTS*

摘要

Data from the United Kingdom Collaborative Alpha-fetoprotein (AFP) Study were used to estimate an individual woman's risk of having a fetus with an open neural tube defect based on her particular maternal serum or amniotic fluid AFP concentration, and also to indicate the value of obtaining fresh specimens for repeat AFP testing. This was done by first showing that the AFP distribution adequately fitted a log Gaussian distribution and then estimating the mean and standard deviation of those distributions as well as the components of variance due to assay imprecision and within-person AFP fluctuations. In the United Kingdom as a whole the odds of having a fetus with open spina bifida among women with maternal serum AFP concentrations of 2·0, 3·0, and 4·0 multiples of the normal median (MoM) at 16-18 weeks' gestation are roughly 1:400, 1:59, and 1:13 respectively. Among women who have a single serum AFP concentration ≥2·5 MoM at 16-18 weeks' gestation which remained raised after the revision of gestational age by ultrasound and who proceed to amniocentesis, the odds are 1:84, 1:1, and 24:1 for women who have amniotic fluid AFP values of 2·0, 3·0, and 4·0 MoM respectively at 16-18 weeks' gestation. If the amniotic fluid sample is blood-stained these odds are lower, if it is not they are higher— for example, 1:4 and 5:1 respectively at 3·0 MoM. If an amniotic fluid AFP result is borderline there is practical value in obtaining a fresh amniotic fluid sample and performing a repeat AFP test, but if a maternal serum AFP result is borderline or high there is little value in testing a second maternal serum sample. Centres in which women with raised serum AFP concentrations routinely have a repeat test may therefore wish to reconsider this policy.