Population-based prevalence of 22q11 deletion syndrome and other chromosome abnormalities in a combined prenatal and postnatal cohort.

Abstract

Objectives: To create and analyse a population-based dataset of genomic tests performed on miscarriage, fetal, and infant samples in a state with >73,000 annual births. Method(s): Analysis of state-wide chromosome testing from Jan 2015 to Dec 2016, including all prenatal and postnatal samples including infants up to 12 months of age, was performed for: (i) the numbers and types of chromosome abnormalities present in miscarriage, prenatal diagnosis, and postnatal specimens; (ii) trends in chromosome abnormalities with advancing development; and (iii) the state-wide prevalence of the 22q11.2 deletion syndrome. Result(s): 8827 chromosomal analyses on 2573 miscarriage, 3661 prenatal, and 2593 postnatal samples were obtained. The majority (91.1%) were performed with whole genome SNP microarrays. The overall frequency of major chromosome abnormalities in the entire cohort was 28.4%. There was a significant decreasing trend in the percentage of chromosome abnormalities with later developmental stage from 51.0% to 21.3% to 16.2% in the miscarriage, prenatal and postnatal cohorts respectively (chi2 trend = 768.1, p < 0.0001). The total frequency of abnormalities in the live infant subgroup was 14.1%. The frequencies of pathogenic CNVs detected via CMA for the miscarriage and prenatal and postnatal cohorts were 2.8%, 2.3%, and 6.2%, respectively. There was a significant increasing trend in the frequency of pathogenic CNVs with later developmental stage (chi2 trend = 41.75, p < 0.0001). For the subgroup of live infants, the pathogenic CNV frequency on CMA analysis was 7.1%. There were 39 cases of 22q11.2 DS detected, including 1 miscarriage and 15 prenatal and 23 postnatal cases. After accounting for two cases of duplicate testing, and excluding the miscarriage case, the estimated prevalence of 22q11 DS was 1 in 4301 Victorian births. More than one third of diagnoses were confirmed during pregnancy. Conclusion(s): Our study marks a complete transition in genomic testing from the G-banded karyotype era, with SNP microarray now established as the first line investigation for pregnancy losses, fetal diagnosis, and newborn/infant assessment in a high-income setting. With advancing development, overall diagnostic yield of chomosome analysis declines, but pathogenic CNVs comprise a greater proportion of results. Integration of prenatal and postnatal datasets provides important opportunities for accurate prevalence estimates of congenital syndromes such as 22q11 deletion syndrome.