The safety and efficacy of chorionic villus sampling for early prenatal diagnosis of cytogenetic abnormalities.

Chorionic villus sampling is a method of prenatal diagnosis in the first trimester of pregnancy in which tissue for genetic study is aspirated from the developing placenta by means of a catheter inserted transcervically under the guidance of ultrasonography. In this seven-center study, we compared the safety and efficacy of chorionic villus sampling in 2278 women with those of amniocentesis at 16 weeks' gestation in 671 women. Both groups were made up primarily of well-educated private patients; they were recruited in the first trimester of pregnancy and had viable pregnancies verified by ultrasound examination. Cytogenetic diagnoses resulted from 97.8 percent of the chorionic villus sampling procedures and 99.4 percent of the amniocenteses (P less than 0.05); aneuploidy was found in 1.8 and 1.4 percent, respectively, of the cases in which diagnoses were made. Of the women who underwent chorionic villus sampling, 17 (0.8 percent) subsequently had an amniocentesis because the diagnosis was ambiguous. Two of the diagnoses of aneuploidy (one tetraploidy, one trisomy 22) were later proved to be incorrect. On the basis of pediatric examination of the infants subsequently born to the women in the sample, there were no errors in the determination of sex or the identification of the major trisomies (21, 18, and 13). The rate of combined losses due to spontaneous and missed abortions, termination of abnormal pregnancies, stillbirths, and neonatal deaths was 7.2 percent in the group that underwent chorionic villus sampling and 5.7 percent in the group that had amniocentesis. After adjustment for slight differences in gestational and maternal age, the total loss rate for the women in the chorionic villus sampling group exceeded that for the amniocentesis group by only 0.8 percentage points (80 percent confidence interval, -0.6 to 2.2). The rate of loss of chromosomally normal fetuses after chorionic villus sampling was 10.8 percent among women in whom three or four attempts were made to place the transcervical catheter, as compared with 2.9 percent in those in whom only one attempt was necessary (P less than 0.01). There were no serious maternal infections among the women in this study or among an additional 1990 women who underwent chorionic villus sampling (upper 95 percent confidence limit, 0.08 percent). We conclude that chorionic villus sampling is a safe and effective technique for the early prenatal diagnosis of cytogenetic abnormalities, but that it probably entails a slightly higher risk of procedure failure and of fetal loss than does amniocentesis.