Use of Common Genetic Variants to Identify Risk of Autism in Siblings of Children Diagnosed with Autism Spectrum Disorders

Friday, May 18, 2012: 10:30 AM
Grand Ballroom East (Sheraton Centre Toronto)
10:15 AM
F. Liebaert1, B. A. Dombroski2, G. D. Schellenberg2, T. Rio Frio1, J. Carayol1, C. Amiet1,3, B. Génin1, C. Vazart1, K. Fontaine1, C. Marcaillou1, F. Rousseau1, E. Couchon4 and G. Dawson5, (1)IntegraGen SA, Evry, France, (2)Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA, (3)Department of Child and Adolescent Psychiatry, Groupe Hospitalier Pitié-Salpêtrière, APHP, Paris, France, (4)IntegraGen, Inc, Cambridge, MA, (5)University of North Carolina, Autism Speaks, UNC Chapel Hill, Chapel Hill, NC, United States
Background: Autism spectrum disorders (ASD) are among the most common forms of severe developmental disability and are characterized by a 4:1 male:female ratio and a sibling recurrence risk estimated to 18.7%. Since multiple studies have shown that early intervention leads to a significantly improved long-term outcome, early identification of children at higher risk of ASD is a key goal. The inheritance pattern of ASD in most families is complex and not compatible with simple mendelian inheritance. While genetic testing for autism is primarily limited to the identification of copy number variants (CNVs) which may be causal for autism, autism associated CNVs are only found in a limited percentage of affected individuals. Recently, a number of common genetic variants or SNPs (single nucleotide polymorphisms) conferring autism risk have been identified. While individual SNPs are not on their own sufficient to be causal, recent studies have shown that the combination of autism associated SNPs allows for the identification of increased ASD risk in siblings of affected children.

Objectives: To focus on the emerging role of common genetic variants and how the identification and the combination of risk-associated common variants in ASDs can lead to identification of siblings of children with ASD who are at higher risk of autism.  

Methods: Two sets of multiplex families were used: an “exploratory population” consisted of 544 families from the Autism Genetic Resource Exchange repository (AGRE, www.agre.org) and a “replication population” consisted of 668 families from the University of Pennsylvania combined with a different subset from the AGRE repository. SNPs associated with an increased risk of autism were identified by performing gender-based genome-wide association (GWA) studies on the “exploratory population”. SNPs associated with autism were prioritized using relevant biological and functional data for genes where SNPs were located. The ability of highly prioritized SNPs to maintain their association with autism was determined in both “exploratory” and “replication populations” through a reproducibility index estimated using a resampling approach. A gender-specific genetic score, the sum of individual risk-associated alleles, was then constructed. The ability of these gender-specific genetic scores to discriminate siblings with or without ASD was evaluated in the “exploratory population” and in the “replication population”.

Results: Thirty eight SNPs were found to maintain their association with autism following reproducibility studies. Genetic scores (GS) were constructed for 1,974 children with autism and 584 unaffected siblings. In males GS of 23 was associated with an 90% specificity (95%CI:86-94), a 30% sensitivity (95%CI:27-41), and a 51% (95%CI:45-55) positive predictive value (PPV). In females, a GS of 28 was associated with a 81% specificity (95%CI:75-85), a 50% sensitivity (95%CI:45-56), and a 22% PPV (95%CI:18-26).

Conclusions: Our findings demonstrate that a combination of multiple risk-associated common variants in a gender-specific genetic score allows for the identification of siblings of children with ASD who have a significantly higher risk of developing autism.

See more of: Genetics I
See more of: Genetics
See more of: Biological Mechanisms
| More