Combined analysis of acetylcholinesterase and alpha-fetoprotein improves the accuracy of antenatal diagnosis of neural-tube defects.

Acetylcholinesterase (AChE), butyrylcholinesterase (BChE) and alpha-fetoprotein (AFP) were measured in 293 amniotic fluids from the second and third trimesters of pregnancy in a prospective study of their diagnostic value in the detection of neural-tube defects (NTD). In normal samples, the mean AChE and BChE concentrations were 3.0 u/L (SEM = 9.1 u/L) and 15.2 u/L (SEM = 1.2 u/L) in the second trimester, and 1.6 u/L (SEM = 0.1 u/L) and 7.8 u/L (SEM = 0.6 u/L) in the third trimester. AFP levels fell throughout the second trimester to unmeasurable levels in the third. AChE levels were markedly elevated in samples from NTD pregnancies at all gestational ages, and a cut-off level of the mean + 3 SD optimally separated normal from abnormal samples. This cut-off correctly predicted 10 of 11 fetuses with neural lesions; there were three false positive results. A similar cut-off for AFP predicted nine of 11 cases of neural lesion with two false positive results. For BChE, a cut off of 50 u/L predicted eight of 11 neural lesions with two false positive results. The specificity, sensitivity, and predictive value positive (PVP) were calculated for each test. The assays of AFP and AChE were useful when used individually (PVP congruent to 80%), but a Bayesian combination of these two tests produced a superior PVP (98%). Because the cost of AChE assay is low, and the test is so simple, it is suggested that an AChE analysis should be performed whenever an amniocentesis is requested for the diagnosis of NTD.