II Jornades SOHIB
¿Se puede mejorar la estimación del
riesgo cardiovascular con información
Genètica?
Jaume Marrugat
Research Group on Cardiovascular Epidemiology & Genetics,
Program of Research on Imflamatory & Cardiovascular Disorders,
IMIM, Barcelona, Spain.
Spanish Cardiovascular Research Network HERACLES (RETICS-ISCIII)
Justification of primary prevention of
coronary heart disease
• Greatest cause of death in developed countries. In
~35% of cases its onset symptom is sudden death.
• Most cases are related to lifestyle & other modifiable
factors, whose improvement results in reduced CHD
incidence.
Cumulative incidence of CVD adjusted for the competing risk of death for
men and women according to aggregate risk factor burden at 50 years of
age, in the Framingham Heart Study
≥ 2 risk factors
1 major risk factor
≥ 1 increased risk factor
≥ 1 suboptimal risk factor
All risk factors optimal
Adjusted Cumulated Incidence
0,7
0,6
69%
Men
0,5
Women
50%
50%
46%
0,4
36%
39%
0,3
0,2
0,1
<3% a Framingham i 10% a Girona
5%
0
50
60
70
80
90
50
Attained age
Lloyd Jones.Circulation. 2006;791-798
60
70
80
90
8%
Changes in population LDL & HDL cholesterol 1995-2005
FEMALE
0.012
0.012
MALE
1995 (mode=133.3)
2000 (mode=142.4)
2005 (mode=131.1)
density
density
0.008
0.008
1995 (mode=151.6)
2000 (mode=148.5)
2005 (mode=126.8)
p-value<0.001
0.000
0.000
0.004
0.004
p-value<0.001
0
50
100
150
200
250
300
350
0
50
100
LDL cholesterol (mg/dL)
MALE
0.040
200
250
300
350
LDL cholesterol (mg/dL)
FEMALE
1995 (mode=51.6)
2000 (mode=50.1)
2005 (mode=55.9)
0.020
p-value=0.351
0.010
0.010
p-value=0.644
density
0.020
0.030
0.030
1995 (mode=42.3)
2000 (mode=44.1)
2005 (mode=44.7)
0.000
0.000
density
150
10
30
50
70
HDL cholesterol (mg/dL)
90
110
10
30
50
70
HDL cholesterol (mg/dL)
90
110
Grau M & REGICOR Inv. Eur J Cardiovasc Prev Rehabil. 2007. 14: 653-9.
Girona population aged 35-74 years
Población de Girona de 35-74 años
1995
2005
Hombres
n=1748
n=4458 p-trend
HTA Conocida
HTA Total
HTA Tratada
HTA Bien controlada
20%
45%
37%
15%
30%
44%
45%
38%
0,01
NS
0,018
0,011
Mujeres
HTA Conocida
HTA Total
HTA Tratada
HTA Bien controlada
22%
39%
38%
22%
26%
35%
47%
45%
0,002
NS
0,018
0,011
Grau M & REGICOR Inv. Eur J Cardiovasc Prev Rehabil. 2007. 14: 653-9.
Cambios en la prevalencia de
hipertensión arterial 1995-2005
30
Trends in MI incidence rates in the population aged
35 to 74 years i Gerona 1990-2003
Atack rate men
Atack rate women
300
250
200
150
100
50
0
Gil M REGICOR. Rev Esp Cardiol 2007-60-349.
Disease prevention
Screening
POPULATION Population
Genetic
Characteristics
individuals Patients
Risk
Factors
Primordial
prevention
Disease
Primary
Prevention
Inappropriate
Life Styles
Patients
Reccurrences
Secondary
Prevention
Population distribution of
total cholesterol
Japan
Finland
 CHD
 CHD
190
230
mg/dL
Population distribution of
total cholesterol & CHD incidence
CHD
incidence
 CHD
incidence
200
220
mg/dL
Population & MI patient distribution of total
cholesterol in high incidence countries
Population
MI patients
200
mg/dL
220
Homes
Dones
REGICOR adaptada versió 2010
The ideal risk function…
Braunwald, JACC 2006
Many events
Highest
High Risk
Intermediate
Lesser #
of events
Low
CHD risk, and 10-year CHD event incidence in
the REGICOR population cohort (n=3200)
Incidencia
Población
% Casos
60
50
Porcentaje
40
30
20
10
0
<5%
5-9,9%
10-14,9%
Riesgo REGICOR
15-20%
20% -
CHD risk, and 5-year CHD event incidence in
the primary care VERIFICA cohort (n=4500)
Incidencia
Población
% Casos
60
50
Porcentaje
40
30
20
10
0
<2,5%
2,5-4,9%
5-7,6%
Riesgo REGICOR
7,7-9,9%
10% -
Low
CVD risk
Stimulate healthy lifestyles,
appropriate weight and follow-up
Moderate
High
Intensive intervention on risk factors
What else can be done in XXIst century?
No risk factor reaches individually
the ideal AUC
…so we combine factors…
…that are usually correlated among them
Libby P. EHJ 2010; 31: 777
Risk modifiers influence atherogenesis through
effects on inflammation as reflected by biomarkers
of the acute phase response
Potential biomarkers in CV diseases
• The 22 more abundant proteins, including albumin &
immunoglobulins, constitute 99% of the plasmatic
proteome mass.
• Many biologically interesting molecules circulate in very
small concentrations:
– Troponin:
– Insulin:
– TNF:
nanomolar (10-6) concentration
picomolar (10-9) concentration
femtomolar (10-12) concentration
Anderson, N. L. et al. The human plasma proteome: a nonredundant list developed by
combination of four separate sources. Mol. Cell. Proteomics 3, 311–326 (2004).
Biomarkers
Plasma:
>2x104
>2x104
Anderson, N. L. et al. Mol. Cell. Proteomics 3, 311–326 (2004).
>3x105
Área bajo la curva de muerte o eventos CV mayorescon
factores emergentes.
Proteína C reactive; Péptido natriurético B, Péptido natriurético N-terminal pro–auricular,
aldosteronemia sérica,renina plasmática; fibrinógeno; inhibidor del activador del plasminogeno tipo
1; Dimero D; homocisteína; y el cociente albumina en orina / creatinina.
Reclassification
• Recently some researchers proposed level of
reclassification as a measure of added utility:
Net Reclassification Improvement*
• They categorized individuals into meaningful CHDrisk categories (say <5%, 5-10%, 10-15%,>15%)
based on risks predicted using models with and
without the new marker and measured the amount
of “reshuffling”.
•*Cook, Circulation 2007. Steyerberg et al Epidemiology 2010; 21:128-38
Adjusted AUC for different models with
traditional risk factors and CIMT
in ARIC Study
AUC area under the curve; CI confidence interval; CIMT carotid intima-media thickness; TRF traditional risk factors.
Nambi et al. JACC . 2010; 55: 1600-7
Number and % reclassification with CIMT and plaque
information are added to traditional risk prediction models
in ARIC Study
Reclassification
Overall sample
Men
Women
Nambi et al. JACC 2010; 55: 1600-7
Overall NRI
9.9%
8.9%
9.8%
Clinical NRI (5-10  10-20 or 10-20 >20%)
21.7%
16.4%
25.4%
Some informative non-invasive tests
• Ultrasound carotid intima media thickness (IMT).
• Ankle brachial index is inexpensive, easy-tomeasure with some training.
C-IMT & ABI do not indicate increased
CVD risk … but existing disease!
IMT
Estenosis
Possible factors for the reclassification of
candidates to CVD primary prevention
•
•
•
•
•
•
•
HS C reactive Protein > 1g/l,
Family history of early ECV,
Obesity (BMI > 30), or waist > recommended values,
Microalbuminuria or renal failure,
Inappropriate diet (self-administered short questionnaire)
Insufficient exercise (self-administered short questionnaire)
Adverse genetic profile (predisposition/genetic burden),
• Carotid IMT
• Feeble/undetectable paedial pulse or ABI<0.9
Preclinical
disease
Heritability (h2)
• The proportion of total phenotypic variance that is
explained by genetic variance at the population level.
MYOCARDIAL INFARCTION: h2 = 0.56.
(Nora JJ, et al. Circulation 1980;61:503-8.)
CORONARY HEART DISEASE MORTALITY: h2 =0.53-0.57
(Zdravkovic S, et al. J Intern Med 2002;252:247-54. Wienke A, et al. Twin Res 2001;4:26674.)
CAROTID ARTERIOSCLEROSIS: h2 = 0.09-0.67
(Xiang AH, et al. Arterioscler Thromb Vasc Biol 2002;22:843-8; Fox CS, et al. Stroke.
2003;34:397-401; Juo SH, et al. Stroke 2004;35:2243-7; Swan L, et al.
Atherosclerosis 2003;166:137-41;
North KE, et al. Arterioscler Thromb Vasc Biol 2002;22:1698-703; Hunt KJ, et al.
Stroke 2002;33:2775-80.)
INTROCORONARY CALCIUM: h2 = 0.42
(Peyser PA, et al. Circulation 2002;106:304-8.)
Coronary heart disease: Complex disease
•
Polygenic: different loci/genes are defining the
individual susceptibility.
•
Allelic heterogeneity: different variants in each
locus/gene are defining the individual
susceptibility.
•
Gen-gen interactions.
•
Environment-environment interactions
•
Gen-environment interactions
New variants for MI or CAD via GWAS
Myocardial Infarction Genetics Consortium
32
Allelic dosage and MI risk
3
p=2x10-18
2,23
2
OR
1,69
1
1,43
1,22
1.00
0
1
2
3
4
Allelic dosage quintiles (9 SNPs)
MIGen Consortium. Nat Gen 2009;41:334-41.
5
Graphic CV risk representation (35 to 74
y) by REGICOR risk function with CRF
and with CAD genetics alone
Average
10-year
CV risk
Riesgo de IAM
a 10 años (%)
50%
REGICOR/Genetica
Genetic risc+smoking+hypertension+dyslipaemia
45%
-Tabaquismo
REGICOR/Genetica
Genetic risc+hypertension+dyslipaemia
-Tabaquismo - HTA
REGICOR/Genetica
Genetic risc+dyslipaemia
40%
-Tabaquismo - HTA - Hipercolesterolemia
REGICOR/Genetica
Genetic risc alone
35%
30%
25%
20%
15%
10%
5%
0%
35
40
50
45
55
60
65
70
edad (años)
Age
(years)
J
P
J
P

exp   β CRFp *CRFp,i   βSNPj *SNPj,i   β CRFp *CR FP   βSNPj *SNPj 
j1
p1
j1
 p1


prob(event i | CRFp,i , SNPj,i )  1  S
• Proof of concept—Do novel genetic markers, non
associated with classical risk factors, differ
between subjects with and without CHD?
• In silico case-control study: on the Wellcome Trust
Case Control Consortium public data
• Prospective validation—Incremental value:
• Cohort studies:
• REGICOR
• Framingham
•…
Distribution of the number of risk alleles in cases and controls
Distribution of the number of alleles in CHD cases and controls
Number of risk alleles in the 9 variants of interest (0 to 18 alleles)
OR per allele = 1.18
Lluis-Ganella C…Elosua R. REC 2010 (in press)
Odds ratio for the cumulated number of MI risk alleles
7 alleles is the median (and reference group)
Alleles
# Controls
# Cases
OR (95% IC)
Lluis-Ganella C…Elosua R. REC 2010 (in press)
P value
Odds ratio for CHD across genetic score quintiles
Genetic score quintiles
Alleles
# Controls
# Cases
Lluis-Ganella C…Elosua R. REC 2010 (in press)
OR (95% IC)
P value
Comparison of the effect (OR) of
quintiles of the cumulated # of adverse
alleles with that of total cholesterol
Total Cholesterol
< 160
(7%)
160-199
(17%)
200-239
(38%)
240-279
(26%)
>=280
(12%)
Q5 vs. Q1
OR
0,52
1
1,20
1,65
1,93
3,74
1
1,21
1,40
1,88
2,20
2,20
Quintiles
OR
Lluis-Ganella C…Elosua R. REC 2010 (in press)
Comparison of the effect (OR) of quintiles
of the cumulated # of adverse alleles with
that of hypertension
PA
OR
Optimal
Normal
HT grade I
(28%)
HT grade
II-III
(13%)
Q5 vs. Q1
(20%)
Normal
High
(18%)
(21%)
1
1
1,32
1,68
1,86
1,86
1
1,21
1,40
1,88
2,20
2,20
Quintiles
OR
Lluis-Ganella C…Elosua R. REC 2010 (in press)
• Proof of concept—Do novel genetic markers, non
associated with classical risk factors, differ
between subjects with and without CHD?
• In silico case-control study: on the Wellcome Trust
Case Control Consortium public data
No improvement in c-statistics (AUC).
• Prospective validation—Incremental value:
Significant
• Cohort
studies:improvement in reclassification
• REGICOR
• Framingham
•…
NRI: 15.2 %
Some advantatges of determining the
cardiovascular risk genotype over
other biomarkers
Genetic characteristics :
• Does not change with age or sex
• Does not change with food
• Does not change with drugs
• Does not have intra-individual variability
• Need to determined only once in life (or until new
markers are found)
CHD/CVD screening
• Population
• Individuals
• Individuals
• Patients
• Patients
Risk charts: screening basic system
 First level of Reclassification
Biomarkers
Genetic predisposition
Second level of reclassification
Exercise test, ABI, C-IMT
Non-invasive angiography
Diagnostic Confirmation
Coronary angiography
PCI / other revascularization
RI
Imaging
Biomarkers,
Genetic risk
Preclinical disease: IMT-c & ABI
CVD
SK
Simplified figure for potential use of
emerging risk factors and imaging
techniques in CV primary prevention
Classical Risk
Factors
Low
<5%
10-year
CVD Risk
REGICOR
Promoting healthy life-styles
and appropriate BMI. Follow-up / 5 years
Interme
diate
5-10%
>1 additional
tests†
High
>10%
>1 additional
tests †
† HS C reactive Protein > 1g/l, Family history of
ECV prematura, Carotid IMT, Feeble/undetectable
paedial pulse or ABI<0.9, Obesity (BMI > 30), or
waist > recommended values, Microalbuminuria or
renal failure, Inappropriate diet, Insufficient
exercise
Adverse genetic profile (once in a life span)
Intensive intervention on risk
factors. Annual follow-up
Non-invasive
coronary angiography
for vulnerable plaque*
Treatment
of lesions
Very high
risk
Coronary
angiography
Exercise test
Assymptomatic population
with no CVD history
Extremely high risk
WWW.REGICOR.ORG
Number/1,000
<= 2.4
2.4 to 4.8
4.8 to 7.2
7.2 to 9.6
9.6 to 12.0
No data
STANDARDISED CARDIOVASCULAR
MORTALITY ~ 2004
MEN
ESC Report Cardiovascular Diseases in Europe 2006. WHO data; June 2006.
Positive remodeling (+), Soft plaque (+),
Fibrous plaque (+),
Calcification (-)
Coronary
angiography
LAD
Tomografía computerizada
multicorte con multidetectores
Motoyama et al. ACC 2006
Clarifying Uncertainity
• Non-invasive tests
– Coronary Calcium score (CT)
 increases with age, low predictive capacity,
– Magnetic Ressonance
 Limited resolution expensive
– CT MD
 expensive, high irradiation, but promising…
Standardized CAD mortality rates in Europe
(1995-2002)
Fed. Rusa
Lithuania
Bulgaria
Rep. Checa
Hungría
Rumanía
Polonia
Finlandia
Irlanda
Cuba
Estados Unidos
Reino Unido
Alemania
Suecia
Austria
Men 2002
Noruega
Men 1995
Canadá
Grecia
Holanda
Italia
Suiza
Luxemburgo
Portugal
España
Argentina
Chile
Francia
Mortality / 100.000
Japón
0
100
200
300
400
500
600
World Health Statistics. World Health Organization. Geneve, 2000. (versión on-line en http://www.who.int/ncd_surveillance/infobase/web/InfoBasePolicyMaker/Reports/).
Incidence & mortality of CHD in men (x100.000) and levels
oxidized LDL (U/L) adjusted for triglyceride, in chronic CHD
men, in Europe.
South
Helsinki
N=125
Estocolmo
N=133
800
700
600
500
Ausburgo
N=159
Barcelona
N=135
Central
North
70
Oxidized LDL
Incidence (x100.000)
Mortality (x100.000)
68
66
64
62
400
60
Roma
N=12
4
300
200
Atenas
N=120
Grau M, Marrugat J; AIRGENE Study Group
European Heart Journal 2007; doi: 10.1093/eurheartj/ehm446
58
56
100
54
0
52
Ecological analyses of IMTmean vs.
coronary heart disease mortality in Europe
D. Baldassarre et al. IMPROVE Study JACC 2010. doi:10.1016/j.jacc.2009.11.075
25-year CHD mortality in the 7-Countries Study
by cholesterol levels in different world areas
30
Northern Europe
CHD mortality rate, %
25
USA
20
15
10
Southern continental Europe
Serbia
Southern Mediterranean Europe
5
Japan
0
2,6
(100mg/dl)
5,15
(200mg/dl)
Kromhout D. Eur Heart Journal 1999; 20: 796-802.
7,75
(300mg/dl)
Colesterolemia
total, mmol/ l
25-year mortalirty risk
Diagram of the effect of cholesterol on
CHD mortality risk by regions.
20
18
16
14
12
10
8
6
4
2
0
RR=1,7
Anglo-saxon
Mediterranean
RR=1,7
<200
200-239
240-300
Total Cholesterol
>300 mg/dL
25-year mortalirty risk
Diagram of the effect of cholesterol on
CHD mortality risk by regions.
20
18
16
14
12
10
8
6
4
2
0
Anglo-saxon
Mediterranean
Absolute difference of risk among regions
<200
Kromhout D. Eur Heart Journal 1999; 20: 796-802.
200-239
240-300
Total Cholesterol
>300 mg/dL
Framingham function over-estimation
of the 10-year CHD-event risk in Girona
WOMEN
MEN
Estimateds/observed ratio
5
After calibration
4
3
Women
2
Men
5
0
35-44 y
45-54 y
55-64 y
65-74 y
Age Group
Estimated/observed rate
1
4
3
2
1
0
35-44 y
45-54 y
55-64 y
Age group
Marrugat et al. J Epidemiol Comm Health 2003;57:1-6.
65-74 y
<3% of
population
Many “false
Positive”
CCS Sensitivity of 98% & Specifity of 27% for
CCS>0 to predict CAD events at 37 months in men
aged ~52 years
CIMT Sensitivity of 96% & Specificity of 49% for
CIMT>0.55mm to predict CAD in men/women
aged ~62 years.
Kondos GT. Circulation 2003;107:2571-2576.
Belhassen L. J Am Coll Cardiol, 2002; 39:1139-1144
1: No history of angina, heart attack, stroke, or peripheral arterial disease.
2: Population over age 75y is considered high risk and must receive therapy without testing for atherosclerosis.
3: Must not have any of the following: Chol>200 mg/dl, blood pressure >120/80 mmHg, diabetes, smoking, family history,
metabolic syndrome.
4: Pending the development of standard practice guidelines.
5: High cholesterol, high blood pressure, diabetes, smoking, family history, metabolic syndrome.
6: For stroke prevention, follow existing guidelines.
HTA
Newton-Cheh C et al. Nat Genet. 2009 May 10.
HTA
Newton-Cheh C et al. Nat Genet. 2009 May 10.
Heredabilidad de diferentes fenotipos
relacionados con arteriosclerosis
INFARTO AGUDO DE MIOCARDIO: h2 = 0.56.
(Nora JJ, et al. Circulation 1980;61:503-8.)
MORTALIDAD POR CARDIOPATÍA ISQUÉMICA: h2 =0.53-0.57
(Zdravkovic S, et al. J Intern Med 2002;252:247-54. Wienke A, et al. Twin Res
2001;4:266-74.)
ARTERIOSCLEROSIS CAROTIDEA: h2 = 0.09-0.67
(Xiang AH, et al. Arterioscler Thromb Vasc Biol 2002;22:843-8; Fox
CS, et al. Stroke. 2003;34:397-401; Juo SH, et al. Stroke
2004;35:2243-7; Swan L, et al. Atherosclerosis 2003;166:137-41;
North KE, et al. Arterioscler Thromb Vasc Biol 2002;22:1698-703;
Hunt KJ, et al. Stroke 2002;33:2775-80.)
CALCIO INTRACORONARIO: h2 = 0.42
(Peyser PA, et al. Circulation 2002;106:304-8.)
Validación de los 9 marcadores genéticos de
riesgo coronario independientes de los
factores de riesgo cardiovascular
• Primera Fase: Validación de un chip diagnóstico
para analizar el riesgo coronario:
• Estudio caso-control: in silico en la base de datos pública
del Welcome Trust Consortium (WTCCC)
• Segunda fase: Validación de la capacidad predictiva
de los marcadores incluidos en el chip:
• Estudios de cohorte:
• REGICOR
• PROCAM
• Framingham (in silico)
Los 9 genes de riesgo coronario INDEPENDIENTES
de los factores de riesgo cardiovascular
SNP
Cromoso
ma
Alelo
de riesgo
Alelo
raro
Frecuencia
alelo raro
OR
(por copia alelo de
riesgo)
Beta
de alelo raro
rs1333049
9
C
C
0,47
1,35
0,30010
rs602633*
1
C
A
0,23
1,18
-0,16551
rs17465637
1
C
A
0,28
1,13
-0,12222
rs501120
10
T
C
0,13
1,19
-0,17395
rs1515098*
2
T
C
0,34
1,21
-0,19062
rs1474787*
6
A
A
0,25
1,23
0,20701
rs9982601
21
T
T
0,13
1,20
0.1823
rs12526453
6
C
G
0.45
1,12
-0,1133
rs6725887
2
C
C
0,14
1,17
0,1570
rs9818870
3
T
T
0.17
1,15
0,1398
rs3184504
12
T
T
0.38
1,13
0,1222
* rs602633, rs1515098 y rs1474787 se encuentran en desequilibrio de ligamiento con rs599839, rs2943634 y rs6922269, respectivamente.
Comparación de la OR del # de alelos
adversos con la de los niveles de
colesterol total
Colesterol
< 160
(7%)
160199
(17%)
200239
(38%)
240279
(26%)
>=280
(12%)
--
--
--
OR
0,52
1
1,20
1,65
1,93
Nº Alelos
4
(5%)
5
(11%)
6
(16%)
7
(20%)
8
(21%)
9
(14%)
10
(8%)
11
(5%)
OR
0,70
0,74
0,92
1
1,16
1,54
1,65
2,09
Extremos
3,74
3,00
Comparación de la OR del # de alelos
adversos con la de los niveles de HTA
PA
(20%)
NAlta
(18%)
(28%)
HTA
II-III
(13%)
1
1
1,32
1,68
1,86
Nº Alelos
4
(5%)
5
(11%)
6
(16%)
7
(20%)
8
(21%)
9
(14%)
10
(8%)
11
(5%)
OR
0,70
0,74
0,92
1
1,16
1,54
1,65
2,09
OR
Optima
N
(21%)
HTA I
--
--
--
Extremo
1,86
3,00