Abstract
BACKGROUND: The aim of this study was to evaluate the best diagnostic approach for the genetic analysis of samples from first, second and third trimester intrauterine fetal deaths (IUFDs). We examined a total of 417 IUFD samples from fetuses with and without congenital anomalies. On 414 samples, karyotyping (N = 46) and/or rapid aneuploidy testing by QF-PCR (N = 371) was performed). One hundred sixty eight samples with a normal test result were subsequently tested by genome wide Single Nucleotide Polymorphism (SNP) array analysis. Three samples were only analyzed by array. RESULTS: In 50 (12.0%) samples an aneuploidy was detected by QF-PCR and/or karyotyping, representing 47.1% of first, 13.2% of second and 3.4% of third trimester pregnancies. Karyotyping and QF-PCR failed in 4 (8.7%) and 7 (1.9%) samples, respectively, concerning mostly contaminated amniotic fluid samples from third trimester pregnancies.Clinically relevant aberrations were identified in 4.2% (all fetuses with malformations) of the 168 samples tested by SNP array. Inherited copy number variants (CNVs) were detected in 5.4% and 8.9% showed CNVs of unknown clinical relevance as parental inheritance could not be studied yet. In a sample from a fetus suspect for Meckel-Gruber syndrome, the genotype information from the SNP array revealed various stretches of homozygosity, including one stretch encompassing the CEP290 gene. Subsequent CEP290 mutation analysis revealed a homozygous, pathogenic mutation in this gene. CONCLUSIONS: Based on our experience we recommend QF-PCR as the first-line test in IUFD samples of first and second trimester pregnancies to exclude aneuploidy before performing array analysis. The chance to detect aneuploidy in third trimester pregnancies is relatively low and therefore array analysis can be performed as a first-tier test. A tissue sample, instead of amniotic fluid, is preferred because of a higher success rate in testing.We emphasize the need for analysis of parental samples whenever a rare, unique CNV is detected to allow for better interpretation of such findings and to improve future pregnancy management. Furthermore, we illustrate the strength of SNP arrays for genotype analysis, even though we realize it is crucial to have detailed phenotypic information to make optimal use of the genotype data in finding candidate recessive genes that may be related to the fetal phenotype.
