Amniotic-fluid alpha-fetoprotein measurement in antenatal diagnosis of anencephaly and open spina bifida in early pregnancy. Second report of the U.K. Collaborative Study on Alpha-fetoprotein in Relation to Neural-tube Defects
Lancet 2(8144): 651-662
1979
ISSN/ISBN: 0140-6736 PMID: 90757 Document Number: 149909
Results of a collaborative study performed in the United Kingdom of the correlation between amniotic fluid alpha fetoprotein (AFP) levels and occurrence of open neural tube defects (NTDs) are presented. AFP measurement between 13 and 24 weeks of pregnancy was investigated as a tool for diagnosing NTDs early in pregnancy. Data on 13,105 singleton pregnancies without fetal NTDs and on 385 with fetal NTDs (222 with anencephaly and 152 with spina bifida) were collected. The percentage of unaffected pregnancies with amniotic fluid AFP values equal to or greater than a given cut-off level (expressed as a multiple of the normal median) increased with gestational age, but similar percentages could be obtained by using different cut-off levels at different gestational ages such as 2.5 times median at 13-15 weeks, 3 times at 16-18, 3.5 at 19-21, and 4 at 22-24 weeks. Using these cut-off levels, 98% (120/123) of spnia bifida cases and 98% (218/222) of anencephalics gave positive results; .48% (61/12,804) of unaffected singleton pregnancies not associated with miscarriage also gave positive results (excluding 7 positive results relating to cases of serious fetal malformations). Where amniotic fluid samples were blood-free, detection rates were the same (98% and 98%), but the false positive rate was much lower (.27%, i.e., 31/11,625). In the study area as a whole, the estimated approximate odds of having a fetus with open spina bifida, given 1 amniocentesis sample for AFP measurement, are 18:1 among women with AFP equal to or greater than 2.5 times the median at 16-18 weeks, 5:1 among women who have previously had an infant with an NTD, and 1:2 among other women. The corresponding odds for all open NTDs are approximately 35:1, 9:1, and 1:1, respectively. Raising the cut-off levels by .5 times the median at every gestational period reduced the detection en spina bifida to 93% and reduced the false positive rate to .24%.