Cardiac arrhythmia diagnosis is as old as the Bible and ancient Chinese literature.1 Treatment has progressed from the natural products used by the ancients, to formalized attempts at pharmacological therapy over the last 300 years (with digitalis and cinchona bark being the lead compounds), to the modern use of devices and ablation. However, with the exception of electronic pacing for atrioventricular (AV) block and related arrhythmias, all therapies used have had significant drawbacks. These range from the pro-arrhythmic consequences of many antiarrhythmic drugs to the inappropriate shocks delivered by cardioverter–defibrillators, and have provided the impetus for exploring the possibilities afforded by emerging cell and gene therapies.
Autologous biological therapies achieved widespread use in the 20th century in the form of bone grafting, skin grafting, and tendon transplants. The use of allogeneic biological sources has become commonplace in replacing kidneys, hearts, lungs, and livers, although in all of these instances a key component to success has been the advances in immunosuppression that prevent rejection. Two breakthroughs have been particularly responsible for advances in modern cell therapy: bone marrow transplant (either autologous or heterologous2) and the isolation and study of mammalian and, later, human3 embryonic stem cells. We can now look with near certainty to being able to coax cells into developing along specific lineages to repair and/or replace diseased organs. Currently, experimentation proceeds with the embryonic stem cell, whose pluripotency makes it a progenitor of all cell and organ types in the body, with induced pluripotent stem cells (the product of adult fibroblasts engineered to return to an embryonic status from which they can then be grown along specific lineages), and with adult human mesenchymal stem cells (hMSCs), which are multipotent and can differentiate into mature lineages or serve as vehicles for carrying genes and small molecules to target organs.
Limits to progress are seen only in the fits and starts of the scientific discovery process but, for human embryonic stem cell research, also in sociopolitical and religious reservations. How long will it take before we see consistently successful clinical outcomes? A different high-impact discovery may provide some perspective. Tyndall recognized the antibacterial properties of penicillium in 1876,4 but not until 1928 did Fleming discover that the medium in which the mold grew carried the bactericidal property.5 Another decade was required for Chain, Florey, and colleagues6 to standardize this soup into a medicine that could be reliably and reproducibly produced and administered. So, antibiotics—conceived within a few years of Pasteur announcing his germ theory of disease— arrived just in time for clinical use in World War II.
Modern gene therapy stems from the understanding of DNA and RNA as well as the comprehension that modifications in either may alter an individual’s gene expression with deleterious or favorable phenotypic consequences. The most practical applications have used a virus engineered to (hopefully) delete its pathogenic properties and to carry a specific gene or genes into a recipient’s cells with expression of that gene altering the phenotype. However, virally delivered gene therapy has various potential and demonstrated toxicities, including immune and inflammatory responses and neoplasia. A new generation of engineered viruses is now available to provide vectors for gene therapy in humans. Approved for experimental human administration are: latter-generation adenoviruses, whose gene delivery is episomal and only transiently expressed; adeno-associated viruses, whose gene delivery is longer-lasting but whose size limits them to carrying only small constructs; and lentiviruses, which result in genomic incorporation of genetic material but whose efficacy in the heart may be limited.
Current technology enables the coronary vasculature to be accessed using catheters that can record pressures, electrical activity, and pace, and inject either via the catheter lumen or via a needle expressed through the catheter tip. However, we still need delivery systems that do not produce shear injury to stem cells and can reproducibly deliver constructs (cells or viruses) to sites in the heart where expression will modify a site of abnormal impulse initiation, interrupt a re-entrant circuit, or, in the case of a biological pacemaker, achieve optimal activation patterns and cardiac output. We also need to understand the extent to which a construct administered to any site will remain localized rather than traveling to other sites in the heart or body with deleterious consequences. If we are delivering cells, will they form adequate gap-junctional connexions with adjacent myocardium so that the desired function is achieved? If we are delivering a virus, will there be adequate expression in the cells to which delivery is targeted? Certainly, there is far more information to date indicating that expression is often non-uniform and, when attempted globally, is disappointingly limited. So, there is much to be done, but, as will be described below, much that is promising.
Antiarrhythmic Gene and Cell Therapies
Bradyarrhythmia therapy has focused on the treatment of complete heart block. The state-of-the-art treatment is electronic demand pacemakers in settings of high-degree AV block and sinus node and/or atrial disease, or AV sequential pacing when normal sinus rhythm persists.7 The drive to make a biological pacemaker and/or a biological AV bypass is rooted in the desire to supplant electronic implants with a biological material that will function physiologically with regard to autonomic responsiveness and growth and development of the patient, while obviating the need for a power-pack or lead replacement and frequent monitoring. The template for building a biological pacemaker is the sinus node action potential.8 A variety of channels/exchangers contribute to the pacemaker function and any intervention that increases inward current and/or decreases outward current in a cardiac cell can confer pacemaker function. Many variations of biological pacing have been attempted. Those that have used viral vectors or injection of naked plasmids include:
- overexpressing β2-adrenergic receptors in murine and porcine hearts that increase heart rate but also enhance sympathetic responsiveness and have pro-arrhythmic potential;9
- overexpressing a dominant negative mutant of Kir2.1 (the major component of the outward potassium current IK1) in guinea pig hearts, which reduces outward current and increases pacemaker rate at the expense of excess prolongation of repolarization and resultant pro-arrhythmic potential;10,11 and
- overexpression of the HCN family of pacemaker genes, which increases inward current during diastole and does not appear to be pro-arrhythmic.12–14 Variations on the latter approach are the major form of experimentation today.
We and others have found that pacemaker function can be induced in the atria and ventricles of large animals in complete heart block. Genes delivered include HCN2, which not only drives the heart but also is vagally and sympathetically responsive.12–15 Chimeric and mutant HCN channels are engineered to optimize function,15–17 and potassium channels (a variant on the pacemaker channel) carry inward sodium current.18
Cell therapy has been used as an alternative to viruses, with two types of cell being emphasized. One is the human embryonic stem cell coaxed into a pacemaker lineage.19 Experiments in porcine hearts have shown good pacemaker function for three months. An important limiting factor is that in moving to a cardiac lineage the cells express surface antigens, which makes them a target of rejection. The resultant need for immunosuppressive therapy renders this approach clinically unacceptable.
We have used adult hMSCs as platforms loaded with pacemaker genes via electroporation.14,20 The hMSCs do not differentiate, couple with myocytes via formation of gap junctions, and, in most animals, remain immunoprivileged over six-week study periods. There is the intent to perform longer-term studies using this approach. Another cell therapy uses fibroblasts to carry pacemaker genes to myocytes with which they are fused via polyethylene glycol administration.21 Finally, in experimental AV block, cell-engineered bypass tracts are being used to carry sinus impulses to the ventricles.22 This avenue of investigation is not as well-developed as the pacemaker therapies.
Atrial fibrillation has been treated in animal models by inducing AV block as the gene/cell therapy equivalent of rate control. Methods have included:
- G-αi2 overexpression via AV nodal artery injection in pigs to suppress basal adenylyl cyclase activity and to indirectly reduce AV nodal Ca current via amplified vagal tone.23 With this approach in sinus rhythm, AV conduction slowed and ventricular refractoriness (ERP) was prolonged. During AF, ventricular rate was reduced by 20%.23
- Creating an AV nodal site of Ca channel blockade.24
- Implanting fibroblasts to induce AV nodal scarring and block.25
All of these approaches deliver the gene locally to produce rate control. The clinical equivalent is radiofrequency ablation and, as yet, nothing suggests that the gene/cell therapy approach will offer any advantages.
Another AF therapy focuses on rhythm control using an ion channel mutation, Q9E-hMiRP1, administered into the atrial epicardium of pigs. This construct confers the equivalent of an acquired long-QT syndrome. Infusion of the IKr blocker clarithromycin is proposed as a means for achieving regional atrial IKr blockade without prolonging the QT interval26 and achieving an antiarrhythmic outcome. More detailed animal testing is required to fully appreciate the therapeutic potential.
The paucity of effective therapies and continued problems with cardioverters or defibrillators make the re-entrant arrhythmias complicating ischemic heart disease an attractive target. A breakthrough was provided by Sasano et al., who delivered a dominant negative HERG mutant (HERG-G628S) via vascular infusion to a peri-infarct zone in pigs.27 Prior to gene transfer, monomorphic ventricular tachycardia (VT) had been consistently inducible in the infarcted animals, but one week after transfer all of the HERG-G628S-treated animals showed no such arrhythmia. This outcome was based on the apparent prolongation of repolarization and refractoriness at the local sites of gene therapy injection.
Our own group has focused on normalizing conduction in the epicardial border zones of canine hearts in the post-infarction period as a means by which to prevent re-entrant tachycardias. We use novel sodium channel constructs that are not inactivated by the low membrane potentials of the depolarized tissues.28,29 In infarcted tissue, when myocytes are depolarized to -65mV, virtually all of the normal cardiac sodium channels (designated SCN5A) are inactivated. In contrast, skeletal muscle (SkM1) Na channels have a more positive inactivation midpoint (-68mV), and almost half of them are available to open during an action potential in a depolarized cell. We have administered an adenoviral SkM1 construct to infarcted dogs in which the incidence of inducible polymorphic VT is 75%, and have reduced this incidence to 17% five days post-infarction.28 SkM1 reduced electrogram fragmentation and increased the upstroke velocity of phase 0 of depolarization, all consistent with an increased inward Na current and normalization of conduction.
Other studies suggest a role for connexins and the gap junctions they form in antiarrhythmia therapy. Evidence includes the following:
- overexpression of Cx45 results in VT in mice;30
- Cx40 mutations are associated with atrial fibrillation in humans;31
- epicardial border zones of healing canine myocardial infarcts manifest altered connexin distribution and density in regions important to generation of VT;32 and
- antiarrhythmic peptides have been used to increase junctional conductance, and one, rotigaptide, appears to target Cx43 specifically and antiarrhythmically.33
Gene and cell therapies provide new possibilities for applying principles that were previously targeted at rational drug design. We can now use mathematical modeling to simulate a desired change in an ion current and then select or design constructs that will mimic the proposed optimal change. We can regionally deliver these constructs using viral vectors, cell carriers, or modified cells to provide novel antiarrhythmic therapies. Conceptually, these approaches work; the challenge is executing the practical application and rigorous testing of these aproaches to see whether we can truly improve the therapeutic universe for cardiac arrhythmias.