Outcome of first-trimester chorionic villus sampling for genetic investigation in multiple pregnancy.

Objectives: To evaluate the efficacy of and risk associated with chorionic villus sampling for genetic investigations in multiple pregnancies, and to evaluate the accuracy of the ultrasonographic detection of chorionicity during the first trimester.

Patients And Methods: A total of 198 sets of twins and nine sets of triplets from 10 087 consecutive first-trimester pregnancies undergoing chorionic villus sampling were considered. Gestational age ranged from 7 to 12.6 weeks. Assessment and confirmation of chorionicity was based on a multiplicity of features. Dichorionicity was established in 169 sets of twins (85.3%) and trichorionicity in all triplet cases, while 29 twins were considered monochorionic. Chorionic villus sampling was performed transabdominally in all but one case, and identification of the placental insertion of the umbilical cord was the main benchmark for sampling. Sampling risks were evaluated by comparing clinical outcome with that of a control population of 63 dichorionic twin pregnancies which underwent no invasive procedure.

Results: Determination of the presence or absence of the lambda sign led to a correct assignment of chorionicity in all cases, while the presence of a membrane thickness of 2 mm or more reflected a 100% specificity with a 22% false negative rate. Sampling was successfully performed in all cases and in only four cases (1.0%) were two needle insertions needed. At follow-up no evidence of incorrect sampling was reported. Karyotyping was provided to all patients, and in 94.1% of cases both short and long-term culture methods were carried out. No difference in fetal and perinatal losses between the study and control populations was found, but a higher rate of deliveries before 37 weeks and of low birth weight babies was noted amongst controls.

Conclusions: Chorionicity in twin pregnancy can be determined with certainty between 7 and 12 weeks of gestation; in cases of confluent placentas reliability is provided by determining the presence or absence of the lambda sign. This study indicates that first-trimester transabdominal chorionic villus sampling is a highly efficient, reliable, and relatively safe approach for genetic diagnosis in twin pregnancies. Although a precise evaluation of the relative risks of chorionic villus sampling and mid-trimester amniocentesis in twins must await randomized control studies, the advantages of a first-trimester diagnosis to enable early decision-making about selective fetal reduction are obvious.