9/8/2012 Disclosures Gregory M Marcus, MD, MAS Research: SentreHeart, Baylis, Gilead Genetics of Human Consultant: InCarda Atrial Fibrillation Advisory Board: Janssen UC Speaker’s Fees: St. Jude Medical SF Medical Center …a clinical research Heart & Vascular Center sf San Francisco perspective gh General Hospital San Francisco VA Medical Center Cardiovascular Research Institute Gladstone Institute of Cardiovascular Disease Genetics of AF: Why Should We Genetics of AF: Why Even Look? Care? • AF is the most common arrhythmia • AF is responsible for substantial morbidity and mortality • The etiology(ies) of AF remain poorly understood – Treatments are not aimed at the root cause – No clear prevention strategies • Can genetic studies help? Fox et al. JAMA 2004 1
9/8/2012 Genetics of AF: Why Even Look? Genetics of AF: Monogenic Studies Marcus et al. Heart Rhythm 2008 Chen et al. Science 2003 Genetics of AF: Monogenic Studies • Chinese family with an autosomal dominant AF • Mapped to a mutation in KCNQ1 • Functional studies revealed a gain in potassium channel function • Consistent with a shortening of the refractory period THEREFORE: • AF can be caused by a genetic defect • Confirmed a mechanism of AF Chen et al. Science 2003 Mahida et al. Cardiovasc Res 2011 2
9/8/2012 Genetics of AF: Monogeneic Studies Genetics of AF: Monogeneic Studies • So what? • AF can be genetic • Screening for genes identified from family studies • Mechanisms of AF can be better understood in other cohorts failed to identify mutations. 1-3 • AF has multiple etiologies • Maybe we really do need to tailor therapies Ellinor et al. Heart 2004 Ellinor et al. Heart 2004 1. 1. 2. Ellinor et al. BMC Med Genet 2006 2. Ellinor et al. BMC Med Genet 2006 3. Otway et al. J Am Coll Cardiol 2007 Otway et al. J Am Coll Cardiol 2007 3. Genetics of AF: Monogeneic Studies Genetics of AF: Candidate Genes • Michael Gollub’s point: • Nobel Laureates Brown and Goldstein’s study of a homozygous familial hypercholesterolemia • A rare genetic disease affecting 1 in 1 million individual’s led to understanding the role of HMG CoA reductase in cholesterol metabolism • Largely because of that work, 30 million patients WITHOUT THAT SPECIFIC DISEASE are now on statins Gollub Heart Rhythm 2011 Mahida et al. Cardiovasc Res 2011 3
9/8/2012 Genetics of AF: Candidate Genes SNPs and Haplotypes • Single Nucleotide Polymorphism (SNP) • CRITICISMS: • Low pretest probability • Present in >1% or >5% of the population • 30,000 genes • Distinct from a mutation • Non-coding regions generally not • Haplotype (SNPs that travel together) considered T A C • Corollary: Deciphering the meaning of 3,000 nucleotides positive findings C T T • Sequencing � may find many mutations 3,000 nucleotides • Linkage Dysequilibrium (LD) • SNPs • Great: Can use “tagging SNPs” • Not so great: Maybe the finding is in LD with something else Genetics of AF: Candidate Genes Multiple Hypothesis Testing • CRITICISMS: • P value: IF the null hypothesis is true, the probability that the findings observed are due to • Low pretest probability chance alone • 30,000 genes • 0.05 or 1 in 20 • Non-coding regions generally not • 20 INDEPENDENT tests � expect at least one will considered be positive 5% of the time due to chance alone • Corollary: Deciphering the meaning of • There is SOME biological plausibility/ pre-test positive findings probability • Sequencing � may find many mutations • SNPs are often NOT independent • SNPs • Multiple hypothesis testing 4
9/8/2012 Multiple Hypothesis Testing Genetics of AF: Candidate Genes • In candidate gene studies: • Multiple SNP testing • CRITICISMS: • What model was used? • Low pretest probability • “We found an association with this SNP!” • Corollary: Deciphering the meaning of positive findings • Theoretical exercise- assume: • Multiple hypothesis testing • Population with GG, GC, and CC genotypes • Population stratification • Evidence favors an association with the C allele • Publication bias • Dominant model = CC+ GC v. GG (any disease • Lack of replication 1 allele v none) • Recessiv e model = CC v. GC + CC (both disease alleles v the rest) • Additive model = CC � GC � GG (2 � 1 � 0) (incremental effect for each additional allele) 1. Sinner et al. Heart Rhythm 2011 Genetics of AF: Candidate Genes Genetics of AF: Candidate Genes 6 6 p=0.002 p=0.002 5 5 IL-6 (pg/ml) IL-6 (pg/ml) 4 4 3 3 2 2 1 1 0 0 GG GG GC GC CC CC -174G/C -174G/C IL-6 levels associated with prevalent AF Marcus et al. Am Heart J 2008 Guadino et al. Circulation 2003 Marcus et al. Am Heart J 2008 5
9/8/2012 Genetics of AF: Candidate Genes Genome Wide Association Studies • IL-6 promoter and AF: • Hypothesis free • Second open heart surgery study showed an • Multiple hypothesis testing must be addressed association between the C allele, higher IL-6 • Replication is critical levels and AF 1 • THOUSANDS OF PATIENTS NEEDED Gudbjartsson et al. Nature 2007 1. Bittar et al. Heart Surg Forum 2005 Genome Wide Association Studies Genome Wide Association Studies • Hypothesis free • Hypothesis free • Multiple hypothesis testing must be addressed • Multiple hypothesis testing must be addressed • Replication is critical • Replication is critical • CRAZY LOW (BONFERRONI-CORRECTED) P • REPLICATION WAS PERFORMED VALUES NEEDED Gudbjartsson et al. Nature 2007 Gudbjartsson et al. Nature 2007 6
9/8/2012 Genome Wide Association Studies Genome Wide Association Studies • Hypothesis free • Multiple hypothesis testing must be addressed • Replication is critical %cases (%controls) • SNPs are in a non-coding “gene desert” • Closest gene (50,000 base pairs away) is PITX2 • SMALL EFFECT SIZES …so not the whole explanation Gudbjartsson et al. Nature 2007 Gudbjartsson et al. Nature 2007 Genome Wide Association Studies Genome Wide Association Studies • Opened up a new area of study in AF • The causal pathways are not yet fully known • PITX2 encodes a transcription factor critical • But the association appears to be true in development • Involved in determining right-left asymmetry, 1-3 suppressing development of a left sided sinus node, 4 is important in development of pulmonary vein myocardium, 5 and PITX2c knock-out animal models are prone to atrial arrhythmias 6 Piedra et al. Cell 1998 1. Logan M et al. Cell 1998 2. Yoshioka et al. Cell 1998 3. Mommersteeg et al. Circ Res 2007 4. Mommersteeg et al. Circ R es 2007 5. Lubitz et al. ME Circ Arrhythm Electrophysiol 2010 Wang et al. Proc Natl Acad Sci USA 2010 1. 6. 7
9/8/2012 • Rigorous GWAS study in lone AF (although included HTN) • Rigorous GWAS studies of AF • SNP in intronic region of KCNN3, a calcium- • SNPs in intronic region of ZFHX3, which plays a activated potassium channel expressed in the role in regulation of cell growth in neurons and heart skeletal muscle • No clear link to AF yet 8
9/8/2012 A Different Approach • African Americans are known to have a significantly lower prevalence of atrial fibrillation compared to Whites. 1-4 Psaty et al. Circulation 1997 1. Go et al. JAMA 2001 2. Alonso et al. Am Heart J 2009 3. Marcus et al. Am J Med 2010 4. A Different Approach A Different Approach • African Americans generally have a higher burden of risk factors for atrial fibrillation – Hypertension 1 • African Americans represent a population – Diabetes 2 admixed with African and European ancestry 1,2 e – Larger body mass index 3 cause of atrial fibrillation – Heart Failure 4 • Understanding this paradox may provide new information regarding the cause of atrial fibrillation Kramer et al. Am J Hyperten 2004 1. Carter et al. Ann Intern Med .1996 2. Ford et al. Circulation 2009 3. Bibbins-Domingo et al. N Engl J Med 2009 4. Rosenberg et al. Am J Hum Genet 2003 1. Parra et al. Am J Hum Genet 1998 2. 9
9/8/2012 African Ancestry European Ancestry Ancestral Informative Marker Ancestral Informative Marker Cytosine 1 2 3 4 5 6 7 8 9 10 11 12 Guanine 13 14 15 16 17 18 19 20 21 22 X Y African Ancestry European Ancestry Predictor: 10% European Ancestry Outcome: Incident AF 1 2 3 4 5 6 7 8 9 10 11 12 Meta-Analysis Adjusted Hazard Ratio Adjusted for: age, sex, BMI, hypertension, diabetes, heart failure, left 13 14 15 16 17 18 19 20 21 22 X Y ventricular hypertrophy, history of MI, and study site 10
9/8/2012 • Concordance rate for AF was 22% among monozygotic (identical) twins versus 11% of dizygotic twins e cause of atrial fibrillation Marcus et al. Heart Rhythm 2008 • Three of 15 lone AF patients were found to have somatic mutation in GJA5 • Functional studies supported dysfunctional effects of the mutations observed fibrillation Sinner et al. Cardiovasc Res 2011 11
9/8/2012 Ablatogenetics Pharmacogenetics Ultimately…. • The goal will be to better guide therapy • To identify patients at risk • To identify effective targets for prevention and treatment ….potentially putting those of us that like to do AF ablation out of business Ablatogenetics Gene panel: Gene panel: Gene panel: KCNN3 PITX2 IL-6, Cx, RAS Thank You Ca Na K Drug Rx 12
Recommend
More recommend
