Mulivor R A, Cook D, Muller F, Boué A, Gilbert F, Mennuti M, Pergament E, Potier M, Nadler H, Punnett H
Am J Hum Genet. 1987 Feb;40(2):131-46.
Amniotic-fluid intestinal alkaline phosphatase activity, gamma-glutamyltranspeptidase activity, and leucine-aminopeptidase activity were quantitated to assess their reliability for the prenatal diagnosis of cystic fibrosis. The results indicate that for each of these enzymes an arbitrary cutoff point could be chosen that would enable one to correctly predict the outcome for the majority of at-risk pregnancies. However, some false positives and false negatives occurred with each enzyme. To obtain optimal diagnostic discrimination, the three enzyme values obtained for each sample were combined into a single linear discriminant function that proved to be a more accurate indicator of the outcome of the pregnancy. This was especially important for those cases in which the predicted outcome as based on the individual enzyme results was in disagreement. From the cases studied here, it appears that this method can be expected to give a correct prediction in approximately 96.5% of all 25%-at-risk pregnancies. False positives can be expected in approximately 1.4% of the pregnancies and false negatives in approximately 2.2%.
对羊水的肠碱性磷酸酶活性、γ-谷氨酰转肽酶活性和亮氨酸氨肽酶活性进行定量分析，以评估它们在产前诊断囊性纤维化方面的可靠性。结果表明，对于这些酶中的每一种，都可以选择一个任意的临界值，从而能够正确预测大多数高危妊娠的结果。然而，每种酶都出现了一些假阳性和假阴性情况。为了获得最佳的诊断辨别力，将每个样本获得的三种酶值合并为一个单一的线性判别函数，该函数被证明是妊娠结果更准确的指标。这对于那些基于个体酶结果预测的结果不一致的病例尤为重要。从这里研究的病例来看，似乎这种方法在所有25%高危妊娠中约96.5%的情况下有望给出正确的预测。预计在约1.4%的妊娠中会出现假阳性，在约2.2%的妊娠中会出现假阴性。
