Feasibility, accuracy and safety of chorionic villus sampling: a report of 10741 cases

To evaluate the feasibility, accuracy and safety of chorionic villus sampling (CVS).

Ten thousand seven hundred and forty one singleton pregnancies at risk of chromosome abnormalities (96.3%) and gene disorders (2.8%) were referred from 1990 to 1999 to the fetal medicine unit of a teaching hospital. CVS was performed transabdominally after 11 weeks, using a modified freehand ultrasonographically guided technique by 5 operators. Fetal karyotyping was obtained using a direct method before 1995 and was completed by cell culture after 1996. Failed results, feto-placental discrepancy and fetal loss were assessed.

Villi were sampled using extra-amniotic puncture (89.4%) and one sampling-device insertion (92.3%). The mean weight of the specimen was 15.2 +/- 6.0 mg. All attempts at sampling were successful, except eight (0.07%). The number of failed results following direct preparation, cell culture and both methods was 20 (0.19%), 23 (0.21%) and 2 (0.02%), respectively. Light maternal cell contamination occurred in less than 1% of the samplings after microscopic selection of the villi, and never interfered with the assessment of karyotyping. All 3 false-negative results (0.03%) were recorded after direct preparation and 2 were corrected by culture. The rate of chromosomal abnormalities confined to the placenta decreased from 1.08% before 1995 to 0.73% after 1996. True fetal mosaicisms were recorded in 7 cases (0.06%). The rate of fetal loss at <28 weeks was 1.64% in all pregnancies and 1.92% when CVS was performed before 13 weeks. Advanced maternal age was the single factor significantly associated with fetal loss.

CVS was feasible, accurate and safe in our institution, as a result of the increasing experience of the operators and the cytogeneticists.