Predictors Of heart failure after ST elevation Myocardial Infarction

Heart failure (HF) is a phenotypically heterogeneous disease with multiple
possible etiologic factors such as coronary artery disease (CAD) with or
without myocardial infarction (MI), hypertension, valve dysfunction or viral
infection. An ischemic etiology such as CAD and MI underlie the development of
HF in about 70% of all cases. In a community-based sample from the US, a 19.2%
incidence of HF was observed 30 days post-MI and in a cohort of elderly, 64%
developed HF in the first year after their first MI. Post-MI HF has a
detrimental effect on prognosis. In a study of >15,000 patients with an acute
MI, 23% developed HF during index hospitalization after thrombolytic therapy
and more importantly patients who developed HF had a higher mortality rate at
30-days (18.9% vs. 3.1%) and 1 year (25.2% vs. 5.3%).

Despite the identification of several risk factors for post-MI HF (e.g. infarct
size, age, diabetes, renal dysfunction), interindividual variability exists,
with some patients experiencing significant left ventricular dilation despite
the absence of risk factors and some patients classified as high risk do not. A
substantial part of this variability is thought to be genetically based and
state of the art genetic research has the potential to lead to a more
fundamental understanding of underlying pathophysiological mechanisms related
to post-MI HF and identify new treatment targets and diagnostic biomarkers.
However, genetic causes of HF have only been identified in rare causes of
non-ischemic HF with monogenic inheritance. So far, only one genome-wide
association study (GWAS) has been reported in unrelated individuals with HF[5].
This GWAS studied 73 subjects with HF from 1345 Framingham Heart Study
participants using a 100k gene chip and did not report any significant
associations. The lack of success in this study may be explained by 1) the
inclusion of a heterogeneous group of HF patients and 2) the focus on main
effects of single loci using strict statistical thresholds for significance
which considers only one single nucleotide polymorphism (SNP) at a time and
thereby ignoring the genomic and environmental context of each SNP. This single
locus-based analysis may be appropriate in diseases caused by a single gene,
but are not applicable to the analysis of complex diseases. A large number of
risk alleles presumably with an odds ratio less than 1.3 are not detected by
current single-locus GWAS analysis methods due to the use of a conservative
correction for multiple tests using Bonferroni or the false discovery rate.
Novel alternative analysis approaches to GWAS data that focus on the combined
effects of many loci, each making a small contribution to overall disease
susceptibility, may provide a solution for aforementioned statistical
limitations. Associated SNPs may be selected based on prior expert knowledge
and only those SNPs that are significantly enriched in particular biological
groups (e.g. biochemical pathways, gene function) will be selected for
replication. This bioinformatics approach will allow us to select SNPs with
more liberal p-values than the traditional statistical analyses. In addition,
it is likely that loci will contribute to complex diseases such as HF only
through their interaction with other genes and environmental factors, while
main effects of the individual loci may be small or absent.

In the present proposal, two different state of the art bioinformatics
approaches to analyze genome-wide data will be utilized that each addresses the
complexity of HF. First, the Exploratory Visual Analysis (EVA) database will be
used and software to organize loci by biochemical pathway and gene function to
capture those genetic effects that might be missed using traditional
statistical methods. Second, the multifactor-dimensionality reduction (MDR) to
detect those genetic effects that are dependent on other loci will be utilized.
This analysis will be guided by prior knowledge on protein-protein interaction
information to reduce the computational complexity.

To assist physicians in identifying those patients who are at high risk for
post-MI HF, a clinical prediction model will be developed. Next to observed
genetic factors, demographic data, cardiovascular risk factor status,
medication use, laboratory measurements and angiographic data will be included
in the prediction model.

Primary objective:
The main aim of the current study is to identify novel pathophysiological
pathways associated with the development of post-MI HF using genome-wide data
and innovative bioinformatics approaches.

Specific aims
1) Identifying and validating novel pathophysiological pathways by detecting
patterns of SNP associations in biochemical systems and gene-function-groups
using genome-wide data.
2) Prioritizing genes by extracting prior biological knowledge from
protein-protein databases to decrease the computational burden of the
interaction analyses.
3) Detecting gene-gene and gene-environment interactions using a model-free,
non-parametric data mining method.

Secondary objectives:
- A clinical prediction tool will be developed to predict patients at high risk
for developing postinfarct heart failure.
- Potential diagnostic biomarkers identified by the primary objective will be
measured and related to incidence of heart failure.
- The association between SNP*s and secondary endpoints such as reinfarction,
stent thrombosis or mortality will be investigated. Potential diagnostic
biomarkers for the secondary endpoints will be measured and related to the
incidence of the specific secondary endpoint.

The study is a multi-center, prospective, longitudinal, observational study. A
nested case-control study will be performed when 800 cases with HF are
included. For each incident case of HF, we will select a control free of
post-MI HF matched on age at time of the MI, gender, size of the MI using peak
CK (in 1000 U/dL increments), and duration of follow-up.

This is an observational study. Participants will not be exposed to any risks
associated with participation to the current study. Participants will not have
any advantages or disadvantages by participating in the current study. Extra
blood will be drawn to measure DNA and newly identified biomarkers. Next to
standard clinical care, no additional venapunction is necessary.