Key determinants of pathogenicity Session 6: Determining pathogenicity and genotype-phenotype correlation J. Peter van Tintelen MD PhD Clinical geneticist Academic Medical Center Amsterdam, the Netherlands
Faculty Disclosure The presenter has advised that the following presentation will NOT include discussion on any commercial products or service and that there are NO financial interests or relationships with any of the Commercial Supporters of this years Conference.
ACMG criteria for variant interpretation pros & cons pathogenicity in monogenic disease conditional pathogenicity (interaction) susceptibility
Richards et al. Genet Med 2015
Richards et al. GIM 2015
Gnomad: 140,000 exomes national initiatives: GoNL, UK biobank etc Allele frequencies ORs
amino acid properties evolutionary conservation functionally relevant domains affects splicing etc CADD SIFT Polyphen2 Grantham MAss etc etc
3 categories experimental evidence: 1. Normal function gene consistent with know biology of the disease (expression relevant tissue, co-localization protein) 2. Gene product functionally affected in pts 3. Demonstrate disruption of gene/variant in model organism/system leads to relevant phenotype
Outcome is not a binary phenomenon!
Determining variant pathogenicity - clinical applicability - Class 1 Class 2 Class 3 Class 4 Class 5 Benign Likely Uncertain Likely Pathogenic benign significance pathogenic <5% 5 10% 10 90% 90 95% >95% No mutation no clinical applicability unless susceptibility variant Biggest clinical applicability Maron, Maron, Semsarian JACC 2012
ACMG framework variant interpretation Not intended for somatic variation/ pharmacogenetics/ riskalleles Generic and require disease and gene specific refinement Genet Med 2018;20:351-359
ACMG framework variant interpretation Know thy genes: Using protein domain knowledge to improve variant interpretation MYH7 Nonrandom mutation cluster analysis Walsh R et al Genet Med 2017;19:192 203
ACMG framework variant interpretation Know thy genes: Using protein domain knowledge to improve variant interpretation RyR2 Kapplinger JD et al Circ Genom Precis Med 2018; e001424
ACMG framework variant interpretation Not intended for somatic variation/ pharmacogenetics/ riskalleles Generic and require disease specific refinement Only for inherited Mendelian disease But is there such thing as monogenic Mendelian disease?
1:1000 Dx DCM/ARVC 10% PLN R14del: 1:10,000 Registry n=1000 Low penetrance (10%) dominant pathogenic variant? Susceptibility? PLN R14del Van der Zwaag PA et al Neth Heart J 2013;21:286-93. 1:1,300 1:500 Van der Zwaag PA et al Eur J Heart Fail. 2012;14:1199-207 Milano A te al Circ Cardiovasc Genet 2016;9:147-53.
ARVC 1:200-1:1000 truncating PKP2 variant J Am Coll Cardiol. 2013 Oct 1;62(14):1290-1297 Eur J Heart Fail. 2014 Dec;16(12):1337-44 Eur Heart J. 2015 Jul 14;36(27):1735-43
N=6/31 2 excessive alcohol 2 infectious disease 1 anthracyclin 1 additional SCN5A pathogenic variant Eur Heart J. 2014;35(32):2165-73 N=31 Hoorntje E. et al. Eur J Heart Fail 2018
Key determinants of pathogenicity: misclassification >20% literature Heart 2011; 97:844-9
Key determinants of pathogenicity: Mis phenotyping (brief) SV parox. VT: NSVT Dx LQTS SCD 13 yrs >24 family members identified! Yet: mutation not present in 13 yo boy! WES: de novo DES N342D mutation (class 5!) KCNQ1-V133I probable deleterious mutation. ICD Ackerman JP et al Mayo Clin Proc 2016;91:1606-16 Neth Heart J. 2012 May;20(5):219-28
THIS SHOULD BE DONE BETTER!
ClinGen: cardiovascular clinical domain WG
ClinGen: evaluation of the strength of evidence for genes implicated in hypertrophic cardiomyopathy Jodie Ingles on behalf of the ClinGen Cardiovascular Clinical Domain WG: in preparation
ClinGen cardiovascular WG: variant classification N=60 MYH7: 93% concordance 3 star ClinVar Genet Med 2018;20:351-359
Ruklisa D. et al Genome Med 2015; 7(1):5 https://www.cardiodb.org/ appraise/
summary & conclusions no such thing as monogenic disease phenotype vital in genetic test interpretation (pretest probablity) ACMG guidelines should be more gene/disease specific susceptibility vs conditional pathogenic vs pathogenic in monogenic disease not being LP/P does not preclude susceptibility re assess every variant yourself (databases) (advice: make one virtual DNA lab)
Thank you
SCD & fever GPD1-L: c.370a>g; Ile124Val Barry London et al. Circulation. 2007;116:2260-2268 - - - - - - 27 y I ECG: nl/bs/fever II Ajmaline testing 14 y
GPD1 L: A280V Co-expression A280V GPD1-L with SCN5A HEK cells: 50% Na inward current Cell surface Na channels expression 31±5% LOD>4 All affected variant + Not in >500 controls Barry London et al. Circulation. 2007;116:2260-2268
journal Absent 600 alleles
SCD fever I ECG: nl/bs/fever II Ajmaline testing - - 14 y + - - - - 27 y
SCD fever I ECG: nl/bs/fever II Ajmaline testing - - - - - + - - 27 y Dx/ encephalitis 14 y + + +
SCD fever Exac: c.370a>g; Ile124Val 250/120,000 alleles; (1/240 individuals!) 2 homozygotes Concluded: insuffient evidence that - this - variant sole contributer to phenotype - - 14 y + + + - - + + - - 27 y -
Could this be a relatively common susceptibility variant? KCNE2 Vs a Rare autosomal dominant pathogenic variant
DES N342D mutation (klasse 5!) Neth Heart J. 2012 May;20(5):219-28
Yield in proven familial cases High Penetrance/ effect Monogenic polygenic/multifactorial low 89% of familial cases (n=493 indexpts; 295 familial disease) Groeneweg et al. Circ Cardiovasc Genet 2015;8:437-446 Rare mutations; frequently present; single 0.001 0.01 more. 0.1
Threshold model: sex +/ gene(s) +/ environment disease pregnancy healthy TTN individuals 1 2 3 4 5 male gender mutation 1 mutation 2 exogeneous