Rare autosomal trisomies on non-invasive prenatal testing: not as adverse as expected.

Abstract

Genome-wide sequencing of maternal plasma cell-free DNA (cfDNA) enables non-invasive prenatal testing (NIPT) of all 24 chromosomes. An expanded NIPT approach can identify rare autosomal trisomies (RATs). However, this is not without interpretative challenges. More evidence is needed to determine pregnancy outcome in cases of RAT detected by NIPT. Between January 2016 and December 2017, a total of 42 924 maternal blood samples were received from Dongguan, Guangdong Province, China, for NIPT as a first-tier screening test as part of a government-funded pilot program. The suggested gestational age for NIPT was 12–22 weeks, and a nuchal translucency or second-trimester ultrasound examination before blood sampling was mandatory. We documented 72 (0.17%) NIPT results positive for RAT, 11 of which were excluded from this study because of loss to follow-up. Trisomy 7 was the most commonly detected RAT, followed equally by trisomies 3 and 16. Pregnancy characteristics, NIPT results and pregnancy outcome are summarized in Table 1. Placental karyotyping confirmed confined placental mosaicism in 54 cases, and the remaining seven had normal placental and maternal lymphocyte karyotype. Nineteen cases opted for amniocentesis, all of which showed normal karyotype; among these, only five (three at risk for trisomy 14 and two at risk for trisomy 15) opted for additional chromosomal microarray analysis, which did not show unbalanced rearrangements or uniparental disomy (UPD) in any case. At follow-up, isolated intrauterine growth restriction (IUGR) was identified in one case of trisomy 16 and one of compound trisomy 7/8. There was no case with a sonographic structural anomaly. All neonates were karyotyped, and none had evidence of true fetal mosaicism. In this study, we found no correlation between NIPT Z-score value and the percentage of abnormal cells observed in the placenta. However, these cell percentages might not represent the true condition of the affected placentas because of biopsy samplingbias. Several genome-wide cfDNA-based NIPT studies have reported RATs in association with a range of pregnancy complications, including miscarriage, true fetal mosaicism and UPD, as well as IUGR1–3. However, our results indicate favorable pregnancy outcome. Ninety-five percent of our cases had an uncomplicated pregnancy. Three reasons might contribute to this discrepancy. First, the a-priori risk in our population may explain the lower incidence of pregnancy complications. Our patients came from the general rather than a high-risk population, as in other studies. As a result, NIPT was positive for RAT in only 0.17% of cases, which is much lower than that (0.43%) in a previous study1. Second, the fact that all NIPT samples were taken only after a pretest ultrasound examination may also explain the relatively favorable outcome, as most of the affected fetuses would have been identified before performing NIPT. Third, the gestational age at NIPT in our study was relatively late, further reducing the incidence of finding an affected pregnancy. We observed a positive predictive value (PPV) of 0 for RATs. Expanding the range of aneuploidies that NIPT targets compounds this problem. Our results showed that RATs are most often mosaic and confined to the placenta. It seems that the risk of adverse pregnancy outcome associated with confined placental mosaicism for RATs is not higher than that in the general population, except for trisomy 164,5. Nevertheless, we found that the PPVs for RATs reflected closely those predicted by placental data. Eighty-eight percent of positive NIPT cases were confirmed by placental investigation. Although the clinical utility of expanding NIPT to include RATs has yet to be demonstrated and, as such, the clinical implementation of this testing is premature in the general population, screening for RATs using NIPT might benefit pregnancy management in some specific situations such as in cases of IUGR with a normal chromosomal microarray analysis result.