Pharmacogenomics of Arrhythmia Therapy (PAT)
Abnormalities of cardiac rhythm are a major public health problem. Drugs suppress cardiac arrhythmias in some patients, yet exacerbate arrhythmias or even generate new ones – the phenomenon of proarrhythmia – in others. This variability is not confined to antiarrhythmic drugs alone: in fact, the risk of proarrhythmia related to excessive prolongation of the QT interval has been the single leading cause of drug withdrawal over the past decade.
While rare monogenic arrhythmia syndromes have been recognized for decades, more recent work strongly supports the overall hypothesis tested here, that common DNA variants underlie not only variability in cardiac rhythm but also the response of normal and abnormal cardiac rhythms to drug exposure. Therefore, the major goals of studies in the Pharmacogenomics of Arrhythmia Therapy (PAT) Center are:
- polymorphism discovery and characterization of genes that are candidate modulators of cardiac rhythm;
- patient accrual for a series of studies to define genomic contributors to variable QT responses to drug therapy; and
- evaluation of variability in drug responses in atrial fibrillation (AF).
Our goals are to continue polymorphism discovery efforts, focusing increasingly on non-coding regions. New candidate genes will be screened as they are identified in experiments such as linkage in family studies, basic biology, genome-wide scans, or our zebrafish studies. These polymorphism discovery efforts will be coupled with growth of databases with well-characterized drug response phenotypes to enable association studies.
A major effort is devoted to generation of clinical databases that include well-defined patient phenotypes, along with DNA samples. Each database contains common information, including patient history, family history, medication history, and arrhythmia history, along with database-specific phenotypes. All databases are accumulated de novo, i.e. from patients starting a drug or newly identified with an arrhythmia rather than those followed for these conditions; thus, patient recruitment has been underway since late 2001/early 2002. One important initiative of the program has been construction of web-based data entry systems, and an Oracle database to house the accumulating phenotypic and genotypic information. Databases currently being accumulated, and their recruitment to March 31, 2008, include:
- Patients undergoing cardiac surgery: Phenotypes include the complications of hemorrhage/thrombosis and AF. This collection was ongoing at the initiation of this award, and currently includes 1664 patients, 29.1% of whom had post-operative AF.
- All patients in whom therapy with a QT prolonging antiarrhythmic drug is initiated: Current guidelines suggest that these drugs should be initiated on an inpatient basis with few exceptions. We have implemented an interface with the electronic inpatient ordering system in the Vanderbilt University pharmacy that allows identification of each patient in whom therapy with these drugs is initiated. Phenotypic information includes baseline QT interval, maximum QT interval after drug challenge, and whether drug was continued at three days and at three months. This database includes 592 subjects.
- All patients starting the anticoagulant coumadin: This is a prospective evaluation of the effect of CYP2C9 and VKORC1 genotypes on the rates of bleeding and of thrombosis, and on steady state dose during coumadin. Data on the first 297 patients were reported in the New England Journal of Medicine in March 2008. This database currently includes 335 subjects; a major focus of activity is recruitment of African-American patients, in collaboration with investigators at Meharry Medical College, a Historically Black Institution in Nashville.
- All patients with AF referred to the arrhythmia clinics at Vanderbilt University Hospital for management: This registry continues to be an important source for not only phenotyping large AF pedigrees but also analyzing genomic factors modulating drug response in patients with AF. Data on a large cohort of patients treated with antiarrhythmic drugs suggests that response to therapy is modulated by a common angiotensin-converting-enzyme (ACE) I/D polymorphism. Recently, data on 556 patients with AF validated an association with two non-coding SNPs on chromosome 4q25. The effects of the variants on drug response in patients with AF are currently being analyzed. This database currently includes 1091 AF subjects and 238 lone AF patients.
- Normal volunteers challenged with a QT-prolonging agent, ibutilide, and a beta-blocker, atenolol: Normal volunteers are studied on the general clinical research center, after being brought into salt balance. Ibutilide is used because the maximum extent of QT prolongation is immediately evident after intravenous challenge, thereby maximizing patient safety, streamlining the conduct of the study, and eliminating the factors related to variable drug elimination as a contributor to QT response. The outcome variables include measures of autonomic function (catecholamines, upright and supine blood pressure and heart rate), baseline electrocardiographic data, electrocardiographic data obtained during and following supine exercise in the presence and absence of ibutilide and, on a second study day, in the presence and absence of atenolol challenge. 279 subjects have completed this study, and data are now being analyzed.
In the 2005-2010 cycle, polymorphism discovery was initially performed by screening exons and flanking introns in ethnically defined panels obtained from the Coriell resource: Caucasians, African-Americans, Han Chinese in the Los Angeles area, and Mexican-Americans in the Los Angeles area. Non-coding regions identified as conserved in cross-species comparisons and functionally important in DNA-reporter assays are also rescreened. Within the last 1-2 years, rescreening has turned to defined patient populations with defined arrhythmias (long QT and AF) and drug response phenotypes. Screening was and is accomplished by a combination of resequencing and temperature gradient capillary electrophoresis; the NHLBI resequencing program also supported generation of ~40 MB of data in ion channel genes in patients with AF and controls.
The priority in this center are genes encoding cardiac ion channels or their functional subunits expressed in heart, although we recognize that polymorphisms in other genes, such as those in the adrenergic and the renin-angiotensin-aldosterone signaling systems, are excellent candidate modulators of arrhythmia phenotypes and their response to drugs. When experimental platforms are available (ion channel coding regions; regulatory regions), functional studies with variant constructs are conducted. To date (2001-2008), PAT has deposited to PharmGKB deep rescreening data (380-1,236 alleles each) on ~34 ion channel and other candidate arrhythmia modulator genes. The long QT dataset has been subject to a Golden Gate Assay with interrogation of 1536 tag SNPs from common haplotypes in 19 high priority candidate genes. In addition, we are using the power of model organism genetics (zebrafish) to identify entirely new genes and pathways involved in modulating variable responses to antiarrhythmic drugs.
PAT activities are supported by a generous grant (HL065962) from the National Heart Lung Blood Institute, NIH. Day to day activity in our center includes the following major components:
- Polymorphism screening in genes that encode ion channels or other proteins that may modulate normal and abnormal cardiac electrophysiology;
- Assessment of function of variants, thereby identified;
- Collection of clinical information (demographics; outcomes) in clinical populations;
- Analysis of the association between variable physiologic and pharmacologic phenotypes in these populations as a function of clinical variables in DNA polymorphisms.