Systematic evaluation of genome sequencing for the assessment of fetal structural anomalies

Abstract

Current prenatal and pediatric genetic evaluation requires three tests to capture balanced chromosomal abnormalities (karyotype), copy number variants (microarray), and coding variants (whole exome sequencing [WES] or targeted gene panels). Here, we explored the sensitivity, specificity, and added value of whole genome sequencing (WGS) to displace all three conventional approaches. We analyzed single nucleotide variants, small insertions and deletions, and structural variants from WGS in 1,612 autism spectrum disorder (ASD) quartet families (n=6,448 individuals) to benchmark the diagnostic performance of WGS against microarray and WES. We then applied these WGS variant discovery and interpretation pipelines to 175 trios (n=525 individuals) with a fetal structural anomaly (FSA) detected on ultrasound and pre-screened by karyotype and microarray. Analyses of WGS in ASD quartets identified a diagnostic variant in 7.5% of ASD probands compared to 1.1% of unaffected siblings (odds ratio=7.5; 95% confidence interval=4.5-13.6; P=2.8x10-21). We found that WGS captured all diagnostic variants detected by microarray and WES as well as five additional diagnoses, reflecting a 0.3% added yield over WES and microarray when combined. The WGS diagnostic yield was also inversely correlated with ASD proband IQ. Implementation in FSA trios identified a diagnostic variant not captured by karyotype or microarray in 12.0% of fetuses. Based on these data and prior studies, we estimate that WGS could provide an overall diagnostic yield of 47.6% in unscreened FSA referrals. We observed that WGS was sensitive to the detection of all classes of pathogenic variation captured by three conventional tests. Moreover, diagnostic yields from WGS were superior to any individual genetic test, warranting further evaluation as a first-tier diagnostic approach.

Rare coding variation illuminates the allelic architecture, risk genes, cellular expression patterns, and phenotypic context of autism

“…Beyond ASD, exome sequencing has enabled the discovery of genes associated with overlapping and distinct genetic architectures across a spectrum of developmental and neuropsychiatric disorders 6,[10][11][12][13][14] . These exome studies have largely focused on analyses of single nucleotide variants (SNVs) and insertion/deletion variants (indels), in particular de novo protein-truncating variants (PTVs) and missense variants, with several studies noting modest enrichment of rare inherited variants as well 10,15 .…”

“…These exome studies have largely focused on analyses of single nucleotide variants (SNVs) and insertion/deletion variants (indels), in particular de novo protein-truncating variants (PTVs) and missense variants, with several studies noting modest enrichment of rare inherited variants as well 10,15 . The relative enrichment of de novo PTVs in cases varies by ascertainment strategy: burden is greatest in individuals with developmental delay (DD), intellectual disability, or multi-system congenital anomalies; moderate in individuals with ASD or isolated developmental anomalies; and lowest in schizophrenia and other neuropsychiatric disorders 10,11,13,14,16,17 . Indeed, hundreds of genes have now been discovered across this spectrum of developmental disorders, with associations driven by phenotype severity and cohort size [18][19][20] .…”

Targeted long-read sequencing resolves complex structural variants and identifies missing disease-causing variants

“…Standard clinical testing, including WGS, achieves a precise genetic diagnosis in approximately 50% of families. 1–4 While large-scale, prospective study of varied patient populations will be required to fully assess the advantages of T-LRS over conventional testing strategies, we anticipate use of T-LRS will increase the diagnostic rate for Mendelian conditions. Indeed, given only the small increase in diagnostic rate achieved by using short-read WGS to screen candidate genes or regions identified via exome sequencing or aCGH, T-LRS could be a more sensitive and cost-effective approach.…”

“…Routine use of genetic testing in clinical and research settings has improved diagnostic rates and uncovered the genetic basis for many rare genetic conditions, yet approximately half of individuals with a suspected Mendelian genetic condition remain undiagnosed. 1–4 Broadly, undiagnosed individuals fall into two main categories: (i) those with a DNA sequence variant or structural difference that does not fully fit their phenotype (i.e., variant of unknown significance), and (ii) those in whom routine clinical evaluation—including exome sequencing—failed to reveal any candidate variants or identified only a single variant for a recessive condition that fits the phenotype. Thus, new tools and technologies that provide a comprehensive and accurate survey of genetic variation have the potential to improve diagnostic rates.…”

Targeted long-read sequencing identifies missing disease-causing variation

“…Routine use of genetic testing in clinical and research settings has improved diagnostic rates and uncovered the genetic basis for many rare genetic conditions, yet approximately half of individuals with a suspected Mendelian condition remain undiagnosed. [1][2][3][4] Broadly, undiagnosed individuals who have undergone testing by DNA sequencing fall into two main categories: (1 those with a DNA sequence variant or structural difference that does not fully fit their phenotype (i.e., variant of unknown significance) and (2) those in whom routine clinical evaluation-including exome sequencing-failed to reveal any candidate variants or identified only a single variant for a recessive condition that fits the phenotype. Thus, new tools and technologies that provide a comprehensive and accurate survey of genetic variation have the potential to improve diagnostic rates.…”