President & CEO University of Ottawa Heart Institute, Canada
In half a century, advances in cardiology have revolutionised the approach and treatment of the once-deadly diseases. In the last 30 years, we have cut the cardiac mortality rate in North America by half. Cardiovascular disease remains the No. 1 killer, however, and by 2010 it will have earned that dubious distinction worldwide.
DNA sequencing of human genes could provide the 21st century with the ultimate in evidence-based medicine allowing us to tackle not only cardiovascular disease, but the many other life threatening diseases. For the first time, we have the technology and the basic science to personalise therapy based on an individual’s genetic makeup and variants. “Personalised Medicine,” combined with prevention, offers us the chance to defeat coronary artery disease.
We can, in large part, prevent coronary artery disease by attacking its major risk factors: hypercholesterolemia (elevated cholesterol levels), obesity, hypertension and diabetes. The advent of statin therapy and the use of Angiotensin-Converting Enzyme (ACE) inhibitors to treat heart failure have reduced the mortality rate. However, not all patients respond to treatment with statins. All these risk factors have a genetic component. We need to identify genetic risk factors before we can implement a comprehensive genetic screening and prevention program.
The evidence about genetics’ role in cardiovascular disease is compelling. In a study of premature Coronary Artery Disease (CAD), only 38% of patients had abnormal lipid values.1 It is reasonable to suspect that the remainder of CAD patients were influenced by family history, as several studies of the Utah population suggest.2
In the Framingham study, a family history of CAD, cerebral vascular accidents or peripheral arterial disease was associated with 2.4-fold increased risk of CAD in men, and 2.2 in women. More than 50% of a person’s predisposition to coronary artery disease is genetic, although the quantification of the genetic versus environmental component awaits more precise definition.
Until recently, it was not possible to identify the genes involved in coronary artery disease and other common multi-gene disorders. The introduction, of 500K DNA markers on a microarray chip and the multi-slice Fast Computed Tomography (CT), which permits non-invasive coronary angiograms, has made it possible to search for the responsible genes.
In 2004, adverse drug reactions caused more than 100,000 deaths and 2 million hospitalisations in the United States.3 Our increasing knowledge of pharmacogenetics suggests that patients’ variable response to drug therapy is, in large part, genetically determined.4 Genetic screening for those responses can eliminate such death and morbidity. For example, 20% of people are resistant to aspirin.5 The dose of warfarin required to be effective to prevent thrombosis in patients also varies markedly. It depends upon the influence of the gene that encodes for Vitamin K Epoxide Reductase, responsible for 25% of the variation in the dose. In the future, we will likely screen patients for the 10 forms of this gene to determine the effective dosage. Already, genetic screening across ethnic groups has led to the following recommendations: African Americans require a high dose of warfarin, Asian Americans a low dose and European Americans a medium dose.6
Genotyping is already an established practice before administering chemotherapy for some forms of cancer. 20% of breast cancer patients, for example, exhibit the gene that encodes for HER2 protein. Herceptin is given to block HER2 protein. If the protein is not present, the therapy will not be effective. The Food & Drug Administration (FDA) is also convinced of the importance of pharmacogenetics; recently, the drug regulator approved BiDil only for use in heart failure in African Americans. The drug was shown to reduce mortality and hospitalisation in African Americans with heart failure by 43%, while having no effect in the Caucasian population.7
Below age 35, most Sudden Cardiac Death (SCD) is attributable to familial diseases. More than 40% of SCD is due to hypertrophic cardiomyopathy, followed by familial arrhythmias such as long QT syndrome or Brugada syndrome.8 Most individuals are otherwise asymptomatic, and SCD occurs without any warning.9,10 Approximately 5% of the 13 million people with cardiovascular disease in the United States11 are significantly predisposed to SCD, which accounts for more than 50% of deaths in patients with heart failure.12 In patients over age 35, SCD is predominantly due to coronary artery disease and usually occurs within the first 60 minutes of symptoms, often precluding the availability of medical help.
Genetic screening and prevention through therapy, such as the use of a defibrillator, is the only hope these individuals have, because drug therapy for arrhythmias is relatively ineffective. But cardiac defibrillators cost more than $75,000. If we can determine through early screening which patients have familial cardiomyopathies and arrhythmias and so are vulnerable to SCD, we could determine who would benefit from a defibrillator.
Hundreds of mutations have been identified in the sarcomeric proteins responsible for cardiomyopathies and SCD. The increased risk associated with family history provides a second indication of genetic involvement. In cohorts of SCD, Friedlander et al.13 and Jouvin et al.14 have shown there is a 1.6-1.8-fold increase in SCD susceptibility among offspring of parents who died from SCD. Although the sample size is small, the relative risk in offspring from families where both parents experienced SCD increased by 9-fold. Thirdly, variations in DNA sequences known as Single Nucleotide Polymorphisms, or SNPs, of the hepatic P450 clearance pathways increase the risk of ventricular (Torsades-de-pointes) arrhythmias.15 Fourthly, a single SNP variant in the SCN5A sodium channel gene found in African Americans, affecting 4 million people16, is associated with an increased incidence of arrhythmias, particularly in individuals receiving proarrhythmic drugs that prolong the QT interval. Over the next 10 years, several predisposing SNPs will likely be identified.
The application of molecular genetics to inherited cardiovascular disorders has been successful largely in the field of single-gene disorders17, 18, in which a single gene induces the phenotype. It is estimated that there are 6,000 single-gene disorders, of which we have identified more than 2,000. Hypertrophic cardiomyopathy19 was the first such disorder identified in cardiology. There are more than 1,200 mutations recognised as responsible for single-gene disorders that induce cardiovascular disease.
Multiple genes, however, confer susceptibility to CAD. We expect hundreds of SNPs to contribute only 5% to 10% of increased risk each. But in combination, they are responsible for the phenotype.
In genome-wide studies, searching for these SNPs would require a DNA marker every 6,000 base pairs, amounting to 500,000 markers that would need to be genotyped for each DNA sample.20 Such high-throughput genotyping has until recently been prohibitively expensive. A study would require samples from several thousand individuals to detect a single SNP.
Case control association studies are the most sensitive and appropriate mechanism to identify genes for coronary artery disease. These studies collect samples from thousands of unrelated individuals with CAD and thousands of controls without CAD. Then investigators compare the SNP frequency in controls versus cases (affected individuals).
Today, microarray chips containing 500,000 SNPs and 1 million SNPS are available. They provide, on average, a marker at intervals of 6,000 bps or 3,000 bps respectively. Studies by Hinds, et al.21 and that of the International HapMap project22, 23 indicate a minimum of 375,000 properly placed markers to genotype an American-European population.
Using the 500K microarray, in August 2005 we initiated a study at the University of Ottawa Heart Institute (the Ottawa Heart Genomics Study)24. The sample size was calculated assuming 90% power, a gene frequency of ≥ 5%, odds ratio of ≥ 1.3 and ≥ 0.2 size differences between controls and cases.
The initial population, estimated to be 2,000 (1,000 affected individuals and 1,000 controls) was genotyped to detect association having a p-value of 0.001 or more significant. Those markers showing an association in the initial population were genotyped in a second independent population to ascertain the degree of replication. A second sample size of 12,000 (8,000 affected and 4,000 controls) would, we estimated, detect SNPs showing a stronger association at p-values such as 10-8 or greater.25 As far as we know, this is the first study to utilise the 500K as a genome-wide scan for CAD, with the latter documented by coronary arteriography.
At the Canadian Cardiovascular Genetics Centre, established at the University of Ottawa Heart Institute in 2004, we perform more than 10,000 angiograms per year, providing the necessary high-throughput phenotyping. The Institute, which serves 1.8 million people, has more than 100,000 coronary angiograms available on patients.
To date, we have completed more than 900 million genotypes on 1,800 individuals. We expect to complete Phase I (n = 2,000) shortly. Unfiltered analysis of the first 500 controls and 500 affected cases indicates that we have found several thousand SNPs with p-values of 0.001 or greater, and more than 130 clusters with p-values ranging from 10-3 to 10-12. We will analyse the second population (n = 12,000) to determine replication, coupled with further customised SNP genotyping. We recognise that many of these associations are false positives. We don’t expect to confirm them all.
Completing the initial phase, however, provides us with a population with an all-inclusive set of SNPs exhibiting strong associations to the phenotype of CAD. We hope to collaborate with investigators in Canada and other countries to identify and analyse the genes contributing the most risk for CAD.
It is a unique opportunity to provide the armamentarium for comprehensive genetic screening to help prevent this deadly condition. Subsequent to completing the replication studies, researchers can compare these genes to those involved in specific risk cohorts, such as hypertension, obesity or hyperlipidemia.