Amniotic fluid alpha-fetoprotein determination at the time of genetic amniocentesis: has it outlived its usefulness?

The purpose of this retrospective study was to evaluate the utility of routine measurement of amniotic fluid alpha-fetoprotein levels at the time of second trimester genetic amniocentesis (mean gestational age, 17.3 weeks +/- 2.5 weeks standard deviation; median, 16.8 weeks; range, 15 to 22 weeks). During the study period 7174 patients underwent second trimester genetic amniocentesis. Outcome data were available in all cases. In 79 (1.1%) cases the amniotic fluid alpha-fetoprotein level was > or = 2.0 multiples of the median. Thirty-three of the 79 (42%) patients had normal ultrasonograms, and in 31 of 33 (94%) the amniotic fluid alpha-fetoprotein level was between 2.0 and 3.0 multiples of the median. Forty-six of the 79 (58%) patients had abnormal ultrasonographic findings, and of these, 82% were neural tube defects, abdominal wall defects, or cystic hygromas. Acetylcholinesterase was positive in 37 cases, all of which had abnormal ultrasonographic findings. None of the fetuses with negative findings on sonographic screening had detectable abnormalities at birth. In this study, with over 7000 patients, amniotic fluid alpha-fetoprotein and acetylcholinesterase levels did not increase the detection of fetal abnormalities. On the basis of these results, routine measurement of amniotic fluid alpha-fetoprotein level at the time of routine genetic amniocentesis (15 to 22 weeks) does not appear justified.