Lancet. 1991 Jun 22;337(8756):1491-9.
First-trimester chorion villus sampling has the advantage over second-trimester amniocentesis of allowing earlier prenatal diagnosis of various genetic and cytogenetic disorders in the fetus (and therefore earlier termination in affected pregnancies) but the relative safety and diagnostic accuracy remain unclear. Between 1985 and 1989, 3248 women seeking prenatal diagnosis, principally because of their age, were recruited to an international, multicentre, randomised comparison of the safety and diagnostic accuracy of the two techniques--5% of women allocated chorion villus sampling and 8% of those allocated amniocentesis were not tested, usually because of spontaneous miscarriage. 6% and 2% were retested, in most because of sampling failure. The endpoint of a liveborn infant who survived was achieved by 86% of women allocated chorion villus sampling and 91% of those allocated amniocentesis; statistical analysis, after appropriate weighting for a centre's contribution, showed that the typical difference between the groups was 4.6% (95% confidence interval 1.6-7.5%; p less than 0.01). This difference reflected more spontaneous fetal deaths before 28 weeks' gestation (2.9% [0.6-5.3%]); more terminations of pregnancy for chromosomal anomalies (1.0% [0.0-2.1%]); and more neonatal deaths (0.3% [-0.1 to 0.7%]). The difference in neonatal deaths was due to a preponderance of very immature liveborn infants in the chorion villus sampling group, and this factor also explained that group's longer mean stay in hospital. More abnormal diagnoses followed chorion villus than amniotic fluid analyses (5.6% vs 3.9%). This difference was largely due to diagnoses of trisomy 18 and of (usually mosaic) abnormalities known to be confined to the placenta. 3 terminated pregnancies were false positives, 1 tested by chorion villus sampling and 2 by amniocentesis, and 2 other mosaic cases diagnosed by chorion villus sampling may have been false positives. There was 1 false-negative result in the chorion villus sampling group. The possibility of earlier exclusion or diagnosis of some fetal disorders afforded by first-trimester chorion villus sampling must be set against its clinical risks.
孕早期绒毛取样相对于孕中期羊膜腔穿刺术的优势在于能够更早地对胎儿的各种遗传和细胞遗传学疾病进行产前诊断（从而在受影响的妊娠中更早地终止妊娠），但其相对安全性和诊断准确性仍不明确。1985年至1989年期间，主要因年龄原因寻求产前诊断的3248名妇女被纳入一项关于这两种技术安全性和诊断准确性的国际多中心随机对照研究——分配接受绒毛取样的妇女中有5%未接受检测，分配接受羊膜腔穿刺术的妇女中有8%未接受检测，通常是由于自然流产。6%和2%的妇女接受了重新检测，大多数是因为取样失败。接受绒毛取样的妇女中有86%和接受羊膜腔穿刺术的妇女中有91%实现了活产婴儿存活的终点；在对各中心的贡献进行适当加权后进行的统计分析表明，两组之间的典型差异为4.6%（95%置信区间1.6 - 7.5%；p < 0.01）。这种差异反映出孕28周前更多的自然胎儿死亡（2.9% [0.6 - 5.3%]）；因染色体异常导致的更多妊娠终止（1.0% [0.0 - 2.1%]）；以及更多的新生儿死亡（0.3% [-0.1至0.7%]）。新生儿死亡的差异是由于绒毛取样组中极不成熟的活产婴儿占多数，这一因素也解释了该组平均住院时间更长的原因。绒毛取样后异常诊断的比例高于羊水分析（5.6%对3.9%）。这种差异主要是由于18三体的诊断以及已知局限于胎盘的（通常为嵌合型）异常的诊断。有3例终止妊娠为假阳性，1例通过绒毛取样检测，2例通过羊膜腔穿刺术检测，另外2例通过绒毛取样诊断的嵌合型病例可能为假阳性。绒毛取样组有1例假阴性结果。孕早期绒毛取样能够更早地排除或诊断某些胎儿疾病，但必须权衡其临床风险。
