A welcome review of current evidence on prenatal diagnostic testing
The American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) have recently revisited and updated clinical information and recommendations on several related documents: Practice Bulletin 162, which will be reviewed in this communication; Screening for Fetal Aneuploidies (Practice Bulletin 163); Microarrays and Next Generation Sequencing Technology (Committee Opinion 682); and Carrier Screening for Genetic Conditions (Committee Opinion 691). In Practice Bulletin 162, Drs. Mary Norton and Marc Jackson are the acknowledged authors on behalf of the ACOG Committee on Genetics and SMFM. Practice Bulletin 162 should be applauded for recommendations and conclusions that previous ACOG bulletins could be accused of obviating in deference to tradition.
Invasive prenatal diagnostic techniques
In 2007, ACOG boldly stated in Practice Bulletin 77 that all pregnant women should have the option of an invasive procedure (chorionic villus sampling [CVS] or amniocentesis). This statement still holds and reflects sensitivity of detecting fetal abnormalities being greatest with diagnostic tests possible only using tissue obtained from an invasive procedure. New in Practice Bulletin 162 are updated risks for CVS and amniocenteses. The hackneyed and outdated allusions to a loss rate of up to 1% for CVS or 0.5% (“1 in 200”) for amniocenteses are no longer applicable. Pregnancy loss rate following CVS ≥ 10 weeks is now cited as 0.22% (1 in 455).1 The risk of limb reduction defects with CVS is stated to be 6 per 10,000, not significantly different from the general population and as concluded by the World Health Organization in 1994.2
The loss rate following traditional amniocenteses is now stated to be 0.13% (1 in 769) in experienced hands. Practice Bulletin 162 does cite a 1% to 2% rate of amniotic membrane rupture, which seems unduly high in my opinion and based on a 1998 study.3 On the other hand, perinatal survival following often transient membrane rupture is, in my opinion, plausibly stated to be greater than 90%. The long-accepted conclusion that 10- to 13-week amniocentesis is not recommended was confirmed. Loss rates in multiple gestations are said to be 2%, perhaps high in experienced hands.
Laboratory tests and diagnostic accuracy
The most transformative guideline in Practice Bulletin 162 is its recommendation for DNA-based microarrays to determine status of all 24 chromosomes. A karyotype is no longer recommended.
This conclusion is based first on the 2012 National Institutes of Child Health and Human Development (NICHD) trial of Wapner and colleagues including this author6 followed by replication.7 The NICHD trial report compared accuracy and additional yield of microarray versus karyotype. Given a normal fetal ultrasound and a normal fetal karyotype, chromosomal microarrays identified clinically significant (chromosomal) abnormalities in 1.7% additional cases over that detected by karyotype alone. The additional abnormalities involved genomic material smaller than the 5 to 7 million base pair resolution possible with a high-resolution karyotype. If ultrasound had revealed a fetal anomaly, the yield catapulted an additional 6%. The take-home message is that an invasive prenatal procedure performed for any reason warrants a chromosomal microarray, and not simply a karyotype.
Chromosomal mosaicism is stated to occur in 0.25% of amniocenteses and in 1% of CVS samples. In amniotic cells and in chorionic villi analysis, providers have long applied algorithms to clarify the clinical significance of abnormal, non-modal cells. If a non-modal cell line in chorionic villi is believed confined to trophoblasts (placenta), the embryo itself should theoretically be normal: confined placental mosaicism (CPM). Extant recommendations for determining clinical significance remain.
Practice Bulletin 162 was prepared, however, prior to generation of new information derived from next generation sequencing (NGS). With NGS, mosaicism is unavoidably encountered, given its greater sensitivity, more often than with chromosomal microarrays. If NGS has been recently introduced into a lab to which prenatal samples are being sent, the provider should inquire if altered criteria for prenatal diagnosis of CVS or amniotic fluid cell mosaicism is being applied. NGS is now widely used in preimplantation genetic diagnosis (PGD), for which Practice Bulletin 162 was presumably not intended.