Trisomy 18 in chorionic villus sampling: problems and consequences

Noninvasive prenatal testing for fetal trisomies in a routinely screened first-trimester population

OBJECTIVE

We sought to assess performance of noninvasive prenatal testing for fetal trisomy in a routinely screened first-trimester pregnancy population.

STUDY DESIGN

This was a cohort study of 2049 pregnant women undergoing routine screening for aneuploidies at 11-13 weeks' gestation. Plasma cell-free DNA analysis using chromosome-selective sequencing was used. Laboratory testing on a single plasma sample of 2 mL was carried out blindly and results were provided as risk score (%) for trisomies 21 and 18.

RESULTS

Trisomy risk scores were given for 95.1% (1949 of 2049) of cases including all 8 with trisomy 21 and 2 of the 3 with trisomy 18. The trisomy risk score was >99% in the 8 cases of trisomy 21 and 2 of trisomy 18 and <1% in 99.9% (1937 of 1939) of euploid cases.

CONCLUSION

Noninvasive prenatal testing using chromosome-selective sequencing in a routinely screened population identified trisomies 21 and 18 with a false-positive rate of 0.1%.

Chorionic villus sampling

Chorionic villus sampling (CVS) has been used a successful and safe first-trimester prenatal diagnostic technique for over 12 years. Developed to avoid the medical and psychological complications of later prenatal diagnosis by amniocentesis, CVS rapidly has become a primary tool for the diagnosis of fetal cytogenetic, molecular, and biochemical disorders. In addition, its development has led to an improved understanding of several biological processes, including confined placental mosaicism and uniparental disomy.

Aneuploid correction and confined placental mosaicism

Uniparental disomy (UPD) can be caused by various genetic mechanisms such as gamete complementation, chromosome duplication in a monosomic zygote or postzygotic aneuploid correction. This latter mechanism has been recently well documented in human reproduction and seems to be strictly related to placental mosaicism. We have therefore studied some aspects of confined placental mosaicism (CPM) which are useful to clarify one of the most common sources of UPD in humans.

Abnormal distribution of chromosomes in postzygotic mitotic cell divisions may result in a mosaic condition with two or more cell lines showing different chromosome constitutions. The effects on fetal phenotype and pregnancy development depend on the chromosomes involved, the distribution of the abnormal cells among tissues and on the precise stage at which chromosome mutation occurs.

As shown in Fig. 1, when the mutational event occurs in the blastocyst, prior to the differentiation of embryonic and chorionic compartments, the mosaicism is found in both the placental and fetal tissues. In contrast, when the chromosome mutation occurs at a later stage, after embryonic and chorionic compartment separation, the abnormal cells may be confined to the placenta or to the embryo, and are not necessarily found in both.

Prenatal detection of two different monosomic cell lines by chorionic villus sampling

We present a prenatal case of mosaicism with at least two monosomy cell lines: one with monosomy 21 (45,XY,-21) and one missing the Y (45,X) and a possible third 46,XY in chorionic villus cell culture. Cytogenetic studies were initiated following the ultrasound detection at 11 weeks of a large cystic hygroma and in utero growth retardation. Spontaneous fetal demise occurred at 12 weeks and the pregnancy was terminated. To our knowledge, this is the first report of two different monosomic cell lines found in chorionic villus cells.

Retrospective study of trisomy 18 in chorionic villi with fluorescent in situ hybridization on archival direct preparations

Trisomy 18 in direct chorionic villus preparations needs further investigation since the chromosome abnormality may be confined to the placenta and may not represent the actual fetal karyotype. We performed, retrospectively, fluorescent in situ hybridization (FISH) with the chromosome 18 centromere probe (L1.84) on interphase nuclei of destained slides of all cases of full trisomy 18 (n = 22) and mosaic trisomy 18 (n = 8) detected among 7600 first-trimester chorionic villus samples during an 8-year period (1985-1992). More nuclei displaying three signals were encountered in cases of full and mosaic trisomy 18 confirmed in fetal tissue than in non-confirmed cases. FISH can be useful for the verification of trisomy 18 in direct chorionic villus preparations.

Discrepant karyotypes after second- and third-trimester combined placentacentesis/amniocentesis

Cytogenetic data are presented from a total of 1306 consecutive pregnancies with successful diagnosis obtained from both chorionic villi after short-time culture (CVS-SC) and amniotic fluid cell cultures (AC); samples had been taken simultaneously at combined placentacentesis (placental biopsy) and amniocentesis during the second (92.8 per cent) and third (7.2 per cent) trimesters. Concordant results were obtained in 1218 pregnancies with a normal karyotype and in 62 pregnancies with an aberrant fetal karyotype. Discrepant, i.e. false-positive and false-negative, results were found in 26 cases (2 per cent). From these data the accuracy of CVS-SC, defined as the proportion of all correct diagnoses, is calculated to be 98 per cent. Three non-mosaic and 14 mosaic false-positive results obtained after CVS-SC could not be confirmed by AC. Related to 1235 true normal fetal karyotypes, the specificity of CVS-SC, i.e. the proportion of normal karyotypes correctly diagnosed, amounts to 98.6 per cent. In nine pregnancies, an aberrant fetal karyotype detected after AC was missed by CVS-SC. The sensitivity of CVS-SC, i.e. the proportion of abnormal fetuses correctly diagnosed (62 out of 71), amounts to 87.3 per cent in our study group.

Cytogenetic diagnoses after chorionic villus sampling are less reliable in very-high-or very-low-risk pregnancies

An increasing number of cytogenetic prenatal diagnoses are performed on chorionic villus samplings. The accuracy of this method is influenced by chromosomal mosaicism, mostly confined to direct preparation methods. Especially those investigators who have experienced false-negative and false-positive findings propagate the combined use of direct and culture methods. Yet large collaborative studies have shown that in approximately two-thirds of diagnostic cases only one procedure is applied. Moreover, the accuracy of a cytogenetic investigation depends not only on the ontogenetic origin of the tissues investigated, but also on interacting factors such as the 'a priori risk' and the 'predictive value of a cytogenetic finding'. On this basis a differentiated prenatal diagnostic procedure is discussed, including either sole short-term culture (STC), combined STC and long-term culture (LTC), primary amniocentesis (AC), or primary percutaneous umbilical blood sampling (PUBS). The predictive value of the cytogenetic diagnosis from CVS varies significantly dependent on the a priori risk of a chromosome aberration and, in the case of an abnormal karyotype, on the specific chromosome involved. A non-mosaic and 'non-lethal' trisomy detected in STC is highly representative of the embryo/fetus, but there are exceptions of limited predictive value, e.g., trisomy 18. Guided by the strategy of an optional follow-up by LTC, AC, or PUBS in 1317 successive CV samplings, we are not aware of a false-negative diagnosis, but probably had one false-positive diagnosis: 47,XXY after STC; 46,XY after LTC. When referring to the rate of fetuses with an unbalanced karyotype expected in the different indication groups, a relative increase of false-positive findings in the very-low-risk group (maternal age < or = 35 years of age) and of false-negative findings in the very-high-risk group (abnormal ultrasonographic findings) of pregnant women when only performing CVS becomes obvious. Because of this dilemma, AC or--especially in the latter group--PUBS might be primarily offered to these indication groups instead of CVS.