Microvillar enzymes in amniotic fluid. Considerations for the prenatal diagnosis of cystic fibrosis.

The activity of gamma-glutamyl transpeptidase (GGT) and the proportion of alkaline phosphatase (ALP) activity that remains in cell-free amniotic fluid (AF) after inhibition by amino acids ("residual ALP activity") have been evaluated as possible prenatal diagnostic tests for cystic fibrosis. A total of 1511 AFs were examined. In 1435 "reference" AFs (excluding those from pregnancies with a known fetal abnormality and those with a known one in four risk of cystic fibrosis) at 14-24 weeks' gestation, the mean residual ALP activity in the presence of 2.5 mmol/L L-phenylalanine was 32% (median, 28%) and in 10.0 mmol/L L-homoarginine it was 70% (median, 72%). AFs were arbitrarily considered to be "abnormal" if the residual activity was greater than 50% in L-phenylalanine and less than 50% in L-homoarginine. An abnormal residual ALP activity pattern was found in nine of 27 pregnancies which resulted in a neural tube (or other developmental) defect and in five of the 10 pregnancies with a chromosomal abnormality; a further 23 (1.6%) abnormal patterns occurred in the 1435 reference AFs from pregnancies that were believed to have had a normal outcome. However, of an additional 23 AFs that were sampled at the 25th week of pregnancy or later, 13 had an abnormal residual activity pattern; the outcome at term in each case was normal. The residual activity of ALP in the presence of either inhibitor did not change with increasing gestational age. However, the absolute amount of ALP that was inhibited by L-phenylalanine (the so-called "phe-inhibitable activity") was greatest at 18 weeks. GGT activity decreased with increasing length of gestation. Of the AF samples from 16 pregnancies at one in four risk for cystic fibrosis that were obtained at 18 weeks' gestation, 11 AFs had a normal residual ALP activity pattern, normal GGT and normal phe-inhibitable ALP activity. Of these 11, five have come to term and the infants are not affected by cystic fibrosis. Three of the 16 AFs had an abnormal residual ALP activity pattern, low GGT and low phe-inhibitable ALP activity; these pregnancies were assessed to be affected by cystic fibrosis and termination was chosen in each case. The two remaining AFs had borderline, paradoxical GGT and residual ALP activity, and initially could not be classified clearly into either of the two groups; however, phe-inhibitable ALP activity was low in each. These pregnancies are continuing.