Brett S' Abrahams, PhD

1
Brett S. Abrahams, PhD
2
The autisms encompass a diverse range of
behavioral disorders
Rett Syndrome
Core Impairments- Language Use- Social
Behavior - Behavioral Flexibility
Childhood Disintegrative Disorder
3
The number of implicated regionsis large and
growing
25-100x increased risk for the sibling of an
affected child
Abrahams and Geschwind, Nature Reviews Genetics,
2008
4
Important insights and opportunities from new
technologies
Abrahams and Geschwind, from Vogeland Motulskys
Human Genetics, 2009
Linkage Peaks
Common Variants
Rare Syndromes
Abrahams and Geschwind, Nature Reviews Genetics,
2008
5
Autism in most cases likely the result of
multiple independent risk factors
Abrahams and Geschwind, from Vogel and Motulskys
Human Genetics, 2009
6
Observations made in one cohort often fail to be
observed in others
Genetic heterogeneity makes replication difficult
Veronica Vieland, Nationwide Childrens Hospital
7
Risk Prediction for Siblings of Children With
Autism Spectrum DisorderA Combination of Four
Genes
G Dawson, GD Schellenberg, A Ziegler, IR König, F
RousseauJ Carayol, O Korvatska, J Munson, A
Estes, C-E Yu, J Hager
Sponsored by IntegraGen, SA
8
Staged design employed to test reproducibility

• Discovery Cohort (AGRE)
• 276 multiplex families genotyped at four
  positions
• risk scores compared between cases and
  pseudo-controls
• Replication Cohort (Univ. Washington)
• 295 multiplex families genotyped at as above
• Genetic scores compared between cases and controls

9
Sensitivity and specificity

• Risk score (RS)Sum number of risk alleles over 4
  SNPs
• SpecificityProportion of individuals without the
  disease correctly identified by the test
• SensitivityProportion of individuals with the
  disease correctly identified by the test

10
The risk score (RS) is a measure of genetic
similarity to other cases
Genotype
11
Individual variants inconsistently associated
with disease status
Odds ratios for individual variants each less
than two
12
Higher risk score reproducibly elevated in cases
versus controls
13
Highest risk scores associated with increased
likelihood of disease
14
Conclusions

• Autism genetics are complex but much has and is
  being learned
• High risk scores are associated with a 2-3x
  increased risk for autism, an effect much larger
  than that for individual common variants
• Estimates for increased risk (2-3x), specificity
  (93), and sensitivity (21) are consistent
  between populations
• Current and future tests may prove helpful in
  risk assessment in children with a sibling with
  autism

Looking forward

• Ongoing work will improve risk assessment
  further
• Increased sensitivity critical (RS lt 7 not
  currently informative)
• Back and forth between laboratory and clinic is
  essential

15
Acknowledgements

• Investigators
• Gerard Schellenberg, PhD
• University of Washington and VA Seattle, WA
• Olena Korvatska, PhD and Chang-En Yu
• VA Seatlle, WA
• Jeff Munson, PhD and Annette Estes, PhD
• University of Washington Autism Center, Seattle,
  WA
• Andreas Ziegler, PhD and Inke R. König, PhD
• Institut für Medizinische Biometrie und
  Statistik, Universität zu Lübeck, Germany
• StatSol, Sereetz, Germany
• IntegraGen Employees
• Jérome Carayol, PhD
• Jörg Hager, PhD
• Mélanie Letexier
• Frédéric Tores, B.Sc.
• Francis Rousseau, PhD