Pathological Ventricular Remodeling: Therapies

Introduction

Current therapies targeting pathological ventricular remodeling^{1, 2} manifest significant effectiveness in reducing morbidity and mortality in patients with systolic heart failure³. However, in many instances, disease progression continues unabated. Whereas novel disease targets are continually being discovered, most innovative therapies do not demonstrate consistent efficacy in patients; indeed, many prove to be ineffective, even deleterious, before reaching phase III clinical trials. Here, we review therapeutic strategies targeting cellular pathways governing left ventricular remodeling in the two major types of heart failure, that with reduced (HFrEF) or preserved (HFpEF) systolic function. In an accompanying article, we highlight recent advances in our understanding of mechanisms underlying pathological ventricular remodeling⁴.

Advances in this field are conditioned by the highly heterogeneous nature of heart failure. Notably, within each of the two broad categories of HFrEF and HFpEF, a wide variety of disease etiologies dictate pathogenesis⁵. In other words, heart failure, a syndrome defined on clinical terms, derives from numerous different diseases, such as myocardial infarction, hypertension, cytokine or neuroendocrine dyscrasias, genetic disorders, and more⁵. It seems likely that the therapies which have emerged with efficacy are those targeting features which are shared among these disorders. As a corollary, it is conceivable that some of the therapies which have failed in clinical trials target relevant elements of pathogenesis but which are not common to all. As “personalized medicine” emerges in the discipline of heart failure, we envision therapies tailored to the specifics of molecular and cellular pathogenesis.

Anti-remodeling Therapies

Pharmaceutical agents

Over the last three decades, numerous randomized clinical trials have demonstrated substantial efficacy of ACE (angiotensin converting enzyme) inhibitors, ARBs (angiotensin receptor blockers), MRAs (mineralocorticoid receptor antagonists), and β-adrenergic blockers in reducing morbidity and mortality in patients with systolic heart failure. Presently, ACC/AHA guidelines for the diagnosis and treatment of heart failure in adults emphasize use of these agents in patients with HFrEF⁶ (Figure).

Figure.

Therapeutic interventions in pathological ventricular remodeling. Clinically proven pharmacological agents reduce morbidity and mortality, including ACE inhibitors, ARBs, β-blockers, and MRAs by reducing cell death, hypertrophy, and fibrosis. GLP-1 may prove effective in treating metabolic derangements. Mechanical support using ventricular assist device therapy unloads the failing myocardium and limits ventricular dilation. ICDs and CRTs target electrophysiological remodeling events. Finally, cell replacement therapy to replenish lost cardiomyocytes remains experimental and holds promise for the future.

Inhibitors of the renin-angiotensin-aldosterone axis

Originally, ACE inhibitors and ARBs were used to treat hypertension. However, it was subsequently found that these agents afforded substantial benefit in animal models of heart failure, including increased survival, by targeting adverse cardiac remodeling. Angiotensin receptor activation can induce cardiac remodeling independently of changes in blood pressure⁷, and ACE inhibitors and ARBs both act to antagonize the effects of Ang II, albeit at different points in the cascade. Numerous clinical trials have demonstrated that ACE inhibitors and ARBs reduce heart failure morbidity and mortality³. More recently, antihypertensive agents targeting renin enzymatic activity⁸, the rate limiting step in Ang II production, have become available and are being studied for effects on adverse cardiac remodeling. For example, the renin inhibitor, aliskiren, blunts remodeling in experimentally infarcted mouse hearts⁸ and has been tested for efficacy in ALOFT (Aliskiren Observation of heart Failure Treatment) and ASPITE (The Assessment of Services Promoting Independence and Recovery in Elders), however, with disparate results in efficacy (favorable and unfavorable, respectively)^{9, 10}. Additional planned trials, such as ASTRONAUT (AliSkiren Trial ON Acute heart failure OUtcomes)¹¹ [aliskiren vs placebo in addition to an ACEI or ARB] and ATMOSPHERE (Aliskiren Trial of Minimizing OutcomeS for Patients with HEart failuRE)¹² [aliskiren vs enalapril or combination] will evaluate endpoints of death and rehospitalization due to heart failure.

Low-dose MRAs are recommended for treatment in select patients with moderately severe or severe heart failure symptoms (NYHA class III-IV), recent decompensation, or with LV dysfunction early after myocardial infarction⁶. The Eplerenone in Mild Patients Hospitalization And SurvIval Study in Heart Failure (EMPHASIS-HF) revealed that eplerenone reduces mortality and hospitalization in patients with systolic dysfunction and mild symptoms, expanding the role of MRAs to include asymptomatic patients¹³.

Aldosterone is a mineralocorticoid secreted by the adrenal gland in response to Ang II or cytokines; it can also signal directly within the myocardium via resident mineralocorticoid receptors and the requisite 11 beta-hydroxysteroid dehydrogenase activity¹⁴. Increases in cardiac aldosterone have been reported in experimental models of myocardial infarction¹⁵, correlating with left ventricular remodeling¹⁶. The effects of aldosterone are similar to those observed with Ang II, including inhibition of nitric oxide synthase and promotion of inflammation, fibrosis, and cardiac myocyte apoptosis¹⁷. However, the use of spironolactone is limited due to metabolic and endocrine side effects and variations in patient response¹⁸, which are largely absent with eplerenone¹³. In addition, patients with chronic heart failure have increased aldosterone synthase activity leading to non-mineralocorticoid receptor-mediated responses¹⁹. Blockade of mineralocorticoid receptors increases aldosterone synthase activity. Consequently, MRA therapy results in high levels of aldosterone. Aldosterone may also have non-mineralocorticoid receptor-mediated deleterious effects on cardiomyocytes²⁰.

The aldosterone synthase inhibitor, FAD286²¹, reduced LV remodeling in a rat model of heart failure, and another, LCI699, demonstrated safety and tolerability in 14 patients with primary aldosteronism²². Future studies will determine whether aldosterone synthase inhibitors are more effective than MRAs in treating heart failure.

A new class of therapeutics that targets the renin-angiotensin-aldosterone axis comprises dual-acting agents for angiotensin receptor-neprilysin inhibition. Neprilysin activates the kinin and natriuretic peptide systems. One such inhibitor, LCZ696, combines a moiety of valsartan with an endopeptidase inhibitor²³. An initial safety study of this inhibitor has led to the ongoing PARADIGM-HF trial (Prospective comparison of ARNI with ACEI to Determine Impact on Global Mortality and morbidity in Heart Failure) (NCT01035255) to compare LCZ696 with enalapril²³.

Selective angiotensin receptor blockers are also currently in development. Angiotensin II can act through two different receptors: Ang II receptor type 1 (AT₁) and Ang II receptor type 2 (AT₂). Whereas the AT₁ receptor is ubiquitously expressed in the cardiovascular system, the expression of AT₂ is low in the normal adult, but elevated in heart failure. In many contexts, activation of AT₁ and AT₂ receptors produces opposite effects, with the AT₂ receptor promoting vasodilation and blunting sof cardiac hypertrophy²⁴. A first-in-class AT₂ receptor agonist, Compound 21, has shown promising early results by decreasing infarct size in a rodent model²⁵.

Beta-adrenergic receptor blockers

Inhibition of β-adrenergic signaling has long been used in the treatment of hypertension and cardiac arrhythmias and more recently in patients with mild to severe heart failure. β-blockers have been shown, both experimentally and clinically, to reduce adverse cardiac remodeling and improve heart failure mortality²⁶. Specifically, cardiac myocytes express the β1-adrenergic receptor subtype, and thus manifest responses to β1-selective inhibitors. However, mechanisms underlying the full benefit of β1-adrenergic receptor blockade in reducing left ventricular remodeling remain elusive, because fibroblasts predominately express the β2-adrenergic receptor subtype²⁷.

Agonist-promoted β-receptor desensitization occurs via G-protein Receptor Kinases (GRKs) which tag these receptors for cellular internalization and degradation. Specific to the heart, GRK2 (also known as β-ARK) has been engineered as a catalytically inactive fragment, β-ARK_ct, which interferes with GRK2 binding to β-adrenergic and subsequent receptor degradation, thus serving to block β-adrenergic uncoupling and desensitization²⁸. Experimentally, β-ARK_ct gene therapy in infarcted rabbit or rat hearts improved function and blunted heart failure²⁹. Although this gene therapy has not yet been tested in humans to date, it holds potential for diminishing β-receptor desensitization thereby augmenting efficacy of β-blockers in heart failure patients.

The role of β-blockers in the treatment of heart failure in patients with HPpEF remains unclear. For example, 6-month treatment with the β-blocker nebivolol did not improve 6-minute walk test performance in ELANDD (The Effect of Long-term Administration of Nebivolol on clinical symptoms, exercise capacity and left ventricular function in patients with Diastolic Dysfunction)³⁰. Some evidence, however, suggests that its negative chronotropic effects may have contributed to this result³⁰. There is a large ongoing trial evaluating β-blocker therapy in HFpEF (beta-PRESERVE) to test the efficacy of metoprolol succinate on mortality and hospitalization rates³¹.

Of note, pharmacogenetic studies of both ACEI and β-blocker therapies have uncovered specific gene mutations that impact therapeutic efficacy. For example, a deletion variant in the gene coding for angiotensin converting enzyme is associated with greater survival benefit from ACE inhibitors and β-blockers³². Ser49 (versus Gly49) in the β1 adrenergic receptor, a specific insertion in the α2C adrenergic receptor, and Gln41 (versus Leu41) in G-protein coupled receptor kinase 5 are each associated with relatively greater survival benefit from β-blocker therapy³².

Positive inotropes

Positive inotropic agents play a critical role in controlling symptoms in decompensated or end-stage HF patients⁶. However, chronic administration of these agents increases mortality, and long-term use of digoxin has no beneficial effect on mortality³³.

Two new non-glycoside inotropic agents have entered the therapeutic arena recently. One is therapy to deliver the cDNA of the sarcoplasmic reticulum Ca²⁺ (SERCA2a) pump via adeno-associated virus in an effort to replenish the down-regulated SERCA2a levels typical of failing myocardium³⁴. The second is a luso-inotropic compound, istaroxime, which inhibits the Na⁺/K⁺-ATPase, leading to accumulation of intracellular Na⁺, decreased activity of the Na⁺-Ca²⁺ exchanger to remove cytosolic Ca²⁺, and consequent activation of sarcomeric contraction (the same mechanism as digitalis glycosides). The CUPID study (Calcium Up-regulation by Percutaneous Administration of Gene Therapy in Cardiac Disease) demonstrated safety and suggested benefit of adeno-associated virus type 1-delivered sarcoplasmic reticulum Ca²⁺-ATPase in advanced heart failure, supporting larger confirmatory trials³⁴. Istaroxime manifests a dual mode of action, combining inotropic (stimulation of myocardial contractility during systole) and lusitropic (improvement of diastolic relaxation) effects. A Phase II trial to assess the hemodynamic effects of istaroxime in 120 heart failure patients (HORIZON-HF, Hemodynamic, Echocardiographic, and Neurohormonal Effects of Istaroxime, a Novel Intravenous Inotropic and Lusitropic Agent: a Randomized Controlled Trial in Patients Hospitalized with Heart Failure) revealed lowered capillary wedge pressure and decreased heart rate³⁵.

Another novel and potentially exciting class of agents, cardiac myosin activators, includes omecamtiv mecarbil (CK-1827452)³⁶. This compound binds the myosin catalytic domain, increasing the transition rate of myosin binding to actin into a more tightly bound state thereby increasing force while simultaneously inhibiting ATP turnover, together leading to more myosin heads generating force per beat. This novel mechanism increases systolic ejection time, resulting in improved systolic function without increasing myocardial oxygen demand, resulting in significant increases in cardiac efficiency³⁷. Preliminary evidence suggests that omecamtiv mecarbil is safe in patients with ischemic cardiomyopathy and angina, increasing stroke volume and fractional shortening and decreasing ventricular volumes³⁷.

Other promising therapeutic agents

Lipid lowering drugs

Inhibitors of HMG-CoA (3-hydroxy-3-methylglutaryl coenzyme A) are well-established at lowering cholesterol, providing robust protection to patients with a variety of forms of ischemic heart disease³⁸. Beneficial anti-remodeling effects of these “statins” may occur on top of those observed with traditional therapy (ACE inhibitors and β-blockers)³⁹. However, in CORONA (Controlled Rosuvastatin Multinational Trial in Heart Failure), rosuvastatin did not reduce the primary endpoint or the number of all-cause deaths in older patients with systolic heart failure, although the drug did reduce cardiovascular hospitalizations⁴⁰. Also in GISSI-HF (Gruppo Italiano per lo Studio della Sopravvivenza nell'Insufficienza Cardiaca - Heart Failure), rosuvastatin 10 mg daily did not alter clinical outcomes in patients with chronic heart failure of any cause⁴¹.

Vasopressin and ET1 receptor antagonists

Beyond the two major neurohormonal cascades, the sympathetic and renin-angiotensin-aldosterone axes, cascades elicited by endothelin and vasopressin are also activated during the pathogenesis of heart failure⁴². Stimulation of the vasopressin V1A receptor increases intracellular Ca²⁺ levels and promotes myocyte hypertrophy and remodeling. Stimulation of V2 receptors increases the expression and membrane incorporation of aquaporin-2 water channels into the collecting duct cells in the kidney, resulting in free water absorption. Vasopressin receptor antagonists, such as conivaptan, lixivaptan, mozavaptan, and tolvaptan, block V1A and/or V2 receptors. In patients hospitalized with heart failure in EVEREST (Efficacy of Vasopressin Antagonism in Heart Failure: Outcome Study with Tolvaptan) trial, addition of oral tolvaptan to standard therapy improved many, though not all, heart failure signs and symptoms without serious adverse events⁴³. However, tolvaptan initiated for acute treatment of patients hospitalized with heart failure had no effect on long-term mortality or heart failure-related morbidity⁴⁴.

Endothelins, including endothelin-1 (ET-1), increase contractility and stimulate growth and myofibrillogenesis in cardiomyocytes, thereby promoting cardiac hypertrophy⁴⁵. ET-1 binds to the ET-A receptor to elicit a wide range of responses, including increased expression of nuclear transcription factors, activation of protein kinases and ion channels responsible for cardiomyocyte contractility and growth⁴⁵. Disappointingly, multiple trials testing a variety of endothelin antagonists have yielded negative results, including ENCOR (ENrasentan Clinical Outcomes Randomised)⁴⁶, EARTH (Efficacy And safety of iRbesarTan and olmesartan in patients with Hypertension)⁴⁷, and RITZ-4 (Randomised Intravenous TeZosentan 4)⁴⁸.

Anti-cytokine agents

Pro-inflammatory cytokines promote hypertrophy, apoptosis, and extracellular matrix remodeling, thereby contributing to heart failure pathogenesis⁴⁹. Elevated levels of pro-inflammatory cytokines are also associated with poor clinical outcomes⁵⁰. However, clinical trials studying TNFα inhibitors, etanercept and infliximab, reported neutral or negative effects on all-cause mortality or hospitalization for heart failure⁵¹. Some evidence suggests that suppression of inflammatory responses globally, as opposed to inhibition of a specific pathway might hold promise⁵². In addition, some inflammatory responses arise secondary to other cellular events, and inhibition of the inciting signaling pathways might prove more fruitful⁵².

Glucagon-Like Peptide-1 (GLP1)

Metabolic derangements are significant elements of heart failure pathogenesis. Indeed, insulin resistance and lipotoxicity are recogonized as hallmarks of dilated cardiomyopathy⁵³ and diabetic cardiomyopathy^{54, 55}. GLP1 stimulates myocardial glucose uptake in dilated cardiomyopathy through p38 MAP kinase⁵⁶, and chronic exposure to GLP1 in a rat model of heart failure sustains left ventricular systolic function and prolongs survival⁵⁷. These exciting preclinical findings have led to an ongoing clinical trial, FIGHT (Functional Impact of GLP-1 for Heart Failure Treatment). (http://clinicaltrials.gov/ct2/show/NCT01800968)

Novel vasodilators

Disorders in cGMP-dependent mechanisms contribute to myocardial dysfunction and remodeling. Specific phosphodiesterases govern the amplitude, duration, and compartmentalization of cyclic nucleotide signaling. An inhibitor of phosphodiesterase-5A (PDE-5A), sildenafil, improved hemodynamics, LV diastolic function, and RV systolic function, in a small, single center trial in HFpEF⁵⁸. These encouraging findings prompted the multicenter clinical trial RELAX (phosphodiesteRasE-5 inhibition to improve quality of Life And eXercise capacity in diastolic heart failure). The primary end point was change in peak oxygen consumption after 24 weeks of therapy, and secondary end points included change in 6-minute walk distance and a hierarchical composite clinical status score⁵⁹. Unfortunately, sildenafil therapy did not result in significant improvement in exercise capacity or clinical status⁵⁹. Another clinical trial, PITCH-HF (Phosphodiesterase Type5 inhibition with tadalafil changes outcomes in Heart Failure), is underway in patients with HFrEF (http://neriresearch.net/overview.html).

Nitric oxide production by endothelial nitric oxide synthase (eNOS) can limit cardiac hypertrophy, apoptosis, and fibrosis, thereby impacting adverse cardiac remodeling⁶⁰. In light of this, a novel approach to target and increase eNOS activity has been developed. Preclinical studies of AVE9488, an oral agent targeting eNOS, suggested improvement in LV remodeling, including decreased hypertrophy, fibrosis, ventricular dilation, and enhanced contractility, yet without effects on left ventricular mass⁶¹.

Special considerations in HFpEF

HFpEF is a clinical manifestation of many different pathophysiologies⁵. These range from LV stiffness, impaired diastolic and/or systolic function, and metabolic dyscrasias⁴. In light of multiple failed clinical trials, efforts have focused on identifying biomarkers, risk factors, specific syndrome features, and appropriate study endpoints to parse these patients into defined subgroups and test therapeutic efficacy.

The most widely used HFpEF biomarkers are BNP (brain natriuretic peptide) and NT-proBNP (N-terminal pro-BNP). Outcomes in RELAX were worse relative to prior HFpEF trials when the more sensitive NT-proBNP assay was required for diagnosis⁵⁹, possibly because this may indicate that the patient's shortness of breath is more likely attributed to heart failure⁶². Other circulating biomarkers have been used to diagnose HFpEF, including procollagen, inflammatory factors (interleukins 6 and 8 and TNFα), matrix metalloproteinase, triiodothyronine, heart-type fatty acid binding proteins, troponin T, and carbohydrate antigen-125⁵. However, these biomarkers remain to be validated in large cohorts.

Compared with HFrEF, HFpEF patients are classically more likely to be older, hypertensive females⁶². A retrospective study of new-onset HF in Framingham participants between 1981 and 2008 showed that approximately 50% were classified as HFpEF. Older age, diabetes mellitus, and a history of valvular disease predicted both HFrEF and HFpEF⁶³. Higher body mass index, smoking, and atrial fibrillation predicted HFpEF only, whereas male sex, higher total cholesterol, higher heart rate, hypertension, cardiovascular disease, left ventricular hypertrophy, and left bundle-branch block were predictive of HFrEF⁶³.

To date, no therapeutic agents have been shown to reduce mortality and morbidity in HEpEF; negative trials include DIG-PEF (Digitalis Investigation Group-Preserved EF), CHARM-Preserve (Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity-Preserve), I-PRESERVE (Irbesartan in heart failure with PRESERVEd EF), and RELAX⁵. Likely, substantial heterogeneity within this patient population is a significant contributor to these negative results. Also, morbidity and mortality may not be optimal endpoints, given the age of these subjects and the presence of multiple comorbidities⁵.

Therapies targeting electrophysiological remodeling

Medical treatment

Agents that alter HF progression also impact disease-associated electrophysiological remodeling and may alter the risk of sudden cardiac death. For example, β-blockers reduce sudden death in both postinfarction patients and patients with HF regardless of etiology⁶. Aldosterone antagonists decrease sudden death and overall mortality in HF early after MI and in advanced HF⁶. ACE inhibitors and ARBs decrease the risk of developing atrial fibrillation in HF patients⁶⁴.

Device-based therapies

HF patients have elevated risk of life-threatening ventricular arrhythmia⁶. Several clinical trials, including DEFINITE (Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation)⁶⁵, MADIT (Multicenter Automatic Defibrillator Implantation Trial) II⁶⁶ and SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial)⁶⁷ demonstrated that ICDs (implantable cardioverter-defibrillators) reduced sudden cardiac death and overall mortality in HF patients. Accordingly, ICDs are class I recommendations in patients with LVEF < 35%⁶.

Intracardiac conduction delay in HF perturbs the coordinated mechanical contraction of the ventricle and worsens systolic performance. Therapy based on bi-ventricular pacing ameliorates the resulting contractile dyssynchrony. Several clinical trials, including MIRACLE-ICD (Multicenter InSync ICD Randomized Clinical Evaluation)⁶⁸, CARE-HF (Cardiac Resynchronization in Heart Failure)⁶⁹, and MADIT-CRT (The Multicenter Automatic Defibrillator Implantation Trial – Cardiac Resynchronization Therapy)⁷⁰ have demonstrated that CRT (cardiac resynchronization therapy) can improve symptoms and reduce mortality. The improved outcomes derive largely from improvements in LV remodeling as assessed by enhanced systolic and diastolic function⁷¹. In animal studies, CRT can restore intracellular Ca²⁺ homeostasis⁷², renormalize the ratio of β1/β2 adrenergic receptors⁷³, and alter the expression of genes involved in mitochondrial energetics, extracellular matrix remodeling, and myocardial stress responses⁷⁴.

Analysis of all studies assessing outcomes using CRT reveals poor agreement between echocardiographic and clinical outcomes⁷⁵. For example, whereas MIRACLE reported beneficial effects of CRT based on clinical indices such as NYHA symptom class and 6-minute walk test, reductions in LV volumes were found to be dependent on disease etiology, with greater reductions noted in non-ischemic patients versus ischemic patients⁷⁶. On the contrary, reductions in mortality and improvements in cardiac function in CARE-HF were similar in non-ischemic and ischemic patients⁶⁹. In an attempt to optimize patient selection, a multicenter trial (PROSPECT, Predictors of Response to CRT) enrolling 498 patients was conducted⁷⁷, but no specific parameter emerged that conclusively improved patient selection criteria.

Mechanical support devices

Substantial evidence supports the notion that antagonism of neurohormonal stimulation improves survival in heart failure⁶; nevertheless, heart failure progresses in the majority of patients. In response, a variety of mechanical devices, including physical restraints, have been developed to reshape the ventricle and reverse pathological cardiac remodeling. These devices vary in terms of their design, material composition, abilities to redistribute strain across the ventricle, and technique of implantation. Overall, a critical feature is the ability to decrease ventricular wall stress, ultimately counteracting the pathological remodeling process.

Ventricular assist devices (VADs)

In the setting of end-stage heart failure, cardiac transplantation is one of the options for patients resistant to medical therapy. However, due to an enduring shortage of donors, left ventricular assist device (LVAD) therapies have emerged as an important therapeutic option, both as a bridge to transplant and more recently as “destination” therapy. Initially, pulsatile volume-displacement pumps were engineered, providing critical circulatory support to improve survival until transplantation becomes an option.

Each of the first generation VADs proved capable of improving survival, end-organ dysfunction, and quality of life⁷⁸; that said, they are not without drawbacks. Adverse events such as bleeding, infection, thrombosis, and mechanical failure limit their utility to less than one year, and in the long term, they do not replace cardiac transplantation. The Randomized Evaluation of Mechanical Assistance in Treatment of Chronic Heart Failure (REMATCH) trial revealed 35% 2-year failure rate for the HeartMate XVE device with mortality in excess of 10%⁷⁹. These shortcomings, coupled with chronically inadequate availability of organ donations and excessive waiting times, have prompted attempts to develop new, more reliable and smaller devices, including those with nonpulsatile dynamics.

The HeartMate 2 (Thoratec Inc.) is the most successfully used nonpulsatile device to date with over 6000 implants worldwide⁷⁸. It is compact with fewer moving parts compared to first generation devices. In other second and third-generation VADs (the centrifugal pumps), devices lack bearings and are driven by a magnetically levitated impeller. Patients implanted with these second generation devices have a 65%-69% first-year survival rate with rare mechanical failure and few fatal events⁷⁸. However, infection remains an important problem, even though it rarely leads to increased mortality.

Myocardial regeneration

A major mechanism of pathological cardiac remodeling involves myocyte death. And given that cardiac myocytes have only limited capacity for regeneration, there is great interest in developing progenitor cells – resident to the myocardium or otherwise – to enhance the regenerative capacity of the injured heart. Embryonic stem (ES) cells hold great promise, as they are by definition capable of differentiating into any cell type in the body. However, there are several disadvantages of using this cell type, including the possibility that they would differentiate into unwanted cell phenotypes to form teratomas. Their manipulation typically involves growth on a layer of feeder cells containing animal products, necessitating immunosuppressive therapy and predisposing to rejection. In some circles, use of ES cells has ignited ethical concerns which have led to legal restrictions. Although there are no current clinical trials underway using ES cells to treat heart disease, a first phase I application to treat spinal cord injury using human ES cells was approved by the US Federal Drug Administration (FDA) in January 2009.

One of the earliest cell types to be considered for cardiac regenerative therapy was skeletal myoblasts. These so-called satellite cells were attractive candidates, because they could be harvested from the host, expanded in vitro, followed by autologous transplantation into the heart. Furthermore, these cells are relatively resistant to ischemia. However, the MAGIC (Myoblast Autologous Grafting in Ischemic Cardiomyopathy) trial demonstrated no significant benefit from use of these cells⁸⁰. Worse, patients receiving the skeletal myoblasts were at significantly increased risk of malignant ventricular arrhythmias⁸⁰.

It has been demonstrated that the cardiac stem cell pool diminishes with aging^{81, 82}. In addition, several populations of cardiac stem cells have been described; however, it is still not clear whether these cells are truly endogenous to the heart or derive from bone marrow or circulating cellular elements. Furthermore, cell surface markers used to define stem cell populations are subject to molecular regulation and may be increased or decreased depending on context. Therefore, rather than the existence of numerous different cardiac stem cell populations/infiltrating cells, it is possible that these cells derive from one population and represent a continuum of cellular phenotypes.

Stem cells being considered for myocardial regeneration derive from bone marrow, circulating pools of progenitor cells, and tissue-resident stem cells derived from adipose tissue, skeletal muscle, umbilical cord blood, myocardium, and epicardium. To date, the majority of the clinical trials using stem/progenitor cells to treat patients following an ischemic event have used adult stem cells derived from bone marrow. Recent meta-analyses of the completed clinical trials using bone marrow-derived stem cells to treat ischemic heart disease suggested that the transplantation of these cells is safe and affords benefits beyond those achieved using standard therapy^{83,84, 85}. These studies went on to document decreases in infarct size, improvements in ejection fraction, and decreased left ventricular end-systolic volumes, suggesting improvement in overall global function. However, not all clinical trials using autologous stem cells have demonstrated efficacy, because most patients receiving autologous adult stem cells are at an advanced age and more likely to suffer from co-morbidities such as hypertension, diabetes, and ischemic disease which lead to decreased stem cell viability and function^86-89. Also, whether the transplanted cells actually replace dead or dying cardiac myocytes is hotly debated. In fact, a number of other mechanisms have been proposed to underlie the benefit, including elicitation of paracrine factors which mediate endogenous repair, angiogenesis, or differentiation of native progenitors. Whether these stem cells actually differentiate into cardiac myocytes remains controversial. Further, low rates of engraftment decrease the likelihood that a sufficient number of dead myocytes can be replaced, which can approach one billion cells lost during a myocardial infarction. Therefore, strategies to reprogram cells native to the heart cells may hold significant promise.

Three recent phase I clinical trials using autologous cardiac stem cells yielded promising results in terms of myocardial repair. The Bolli and Anversa groups conducted a phase 1 trial named SCIPIO (Stem Cell Infusion in Patients with Ischemic cardiOmyopathy) of autologous cardiac stem cells (c-kit positive and lineage negative) for the treatment of heart failure resulting from ischemic heart disease. In 14 patients receiving cell therapy, LVEF increased from 30.3% to 38.5% at 4 months after infusion. Importantly, the beneficial effects of stem cell therapy were even more pronounced at 1 year in eight patients whose LVEF increased by 12.3%. MRI measurements of infarct size demonstrated 24% and 30% decreases at 4 and 12 months, respectively⁹⁰.

The recently reported CADUCEUS trial (CArdiosphere-Derived aUtologous stem CElls to reverse ventricUlar dySfunction) tested the effects of autologous stems cells cultured as cardiospheres (CDCs) as a mechanism of myocardial regeneration. These investigators reported significant reductions in scar, increases in viable heart mass, and improvements in regional contractility and regional systolic wall thickening based on MRI at six months. However, changes in end-diastolic volume, end-systolic volume, and LVEF did not differ between groups⁹¹.

The POSEIDON trial (Pilot Study of the Comparative Safety and Efficacy of Transendocardial Injection of Autologous Mesenchymal Stem Cells Versus Allogeneic Mesenchymal Stem Cells in Patients With Chronic Ischemic Left Ventricular Dysfunction Secondary to Myocardial Infarction) compared autologous and allogeneic mesenchymal stem cells (MSCs) in ischemic cardiomyopathy⁹². Injection of either autologous or allogeneic stem cells was associated with low occurrence of major adverse effects, and allogeneic MSCs did not stimulate significant donor-specific immune reactions. Relative to baseline, autologous, but not allogeneic, MSC therapy was associated with an improvement in the 6-minute walk test and the Minnesota Living with Heart Failure Questionnaire score; however, neither approach improved exercise VO_2max. Allogeneic and autologous MSCs reduced mean infarct size and sphericity index but did not increase EF. Overall, allogeneic MSCs reduced LV end-diastolic volumes. Interestingly, in a subgroup analysis, low-dose MSCs (20 million cells) produced the greatest reductions in LV volumes and increased EF⁹².

Cellular reprogramming

Due to technical and ethical issues surrounding use of human ES cells for regenerative medicine, scientists have explored alternative strategies to generate cells with the capability of differentiating into cardiomyocytes. Scientists have successfully reprogrammed adult fibroblasts to de-differentiate into a embryonic stem cell-like state^{93, 94}. These cells, termed induced pluripotent stem (iPS) cells, are viewed as a major scientific breakthrough in the field of regenerative medicine. These cells offer the advantage of being taken directly from the patient and reprogrammed into a de-differentiated state with subsequent manipulation into the desired cell type. These cells overcome obstacles associated with human ES cells in that they may escape immune rejection, and there are no ethical concerns associated with the use of human embryos. However, currently, the study of iPS cells is in its infancy, and there are several barriers to overcome before the cells can be used to regenerate the human heart. For example, iPS cells appear to be less efficient than ES cells in their capacity to differentiate into all cell types, and it is not known for each iPS cell clone whether reprogramming is complete. Even if a small number of cells are incompletely reprogrammed within the recipient tissue, the risk of teratoma formation remains present. Furthermore, to date, most iPS cells have been generated using viruses carrying transgenes, which are integrated into the host genome; reactivation could lead to tumorigenesis. However, this latter concern may be overcome by the engineering of tiny nonviral DNA vectors (rings of DNA about one-half the size of those usually used to reprogram cells)⁹⁵. Other challenges include disease-related mutations and polymorphisms harbored within iPS cells derived from diseased patients and the fact that many diseases are attributable to more than one cell type.

Adult stem cells typically maintain “epigenetic memory” of their tissue or origin. As such, full cardiac myocyte differentiation may be insufficient without nuclear reprogramming. This is one reason why manipulation of cells from within the heart offers significant advantages, potentially eliminating the need for exogenous stem cell transplantation. Recently, it was demonstrated that cardiac fibroblasts could be reprogrammed into functional cardiac myocytes^96-98, raising the possibility that fibroblasts within the heart might be redirected from contributing to scar formation to muscle regeneration.

Potentially important work is presently underway to discover means of reprogramming the heart using small molecule therapies^{99, 100}. By activating resident progenitor cells, coaxing myocytes to re-enter the cell cycle, and/or transforming other cell types into a myocyte lineage, these strategies hold promise as novel ways to rebuild injured myocardium.

Tailored HF therapies

Despite the widely recognized fact that the syndrome of heart failure derives from a broad range of disease etiologies with vastly different molecular mechanisms of pathogenesis, current treatment strategies are largely uniform. Several hurdles must be overcome before etiology-specific therapy can be applied to defined patient subgroups. First, it is difficult to parse etiologic subgroups based on clinical presentation. Biomarkers may prove useful, as is emerging with NT-proBNP, procollagen, and inflammatory factors in HFpEF patients^{5, 59}. In 2011, the UK-based National Institute for Health and Clinical Excellence issued clinical practice guidelines in heart failure based on serum NT-pro-BNP levels, echocardiography, and specialist assessment¹⁰¹. Second, innovative approaches may overcome deficiencies in conventional clinical trial design and/or deficient subgroup analysis. For example, Bayesian adaptive trial design uses information existing at the time of trial initiation, combined with data accumulated during the trial, to identify treatments most beneficial for specific patient subgroups¹⁰². Another approach employs multivariate prediction tools for risk stratification¹⁰³. In the end, these approaches may prove superior to conventional one-variable-at-a-time subgroup analysis to minimize false positives and false negatives¹⁰⁴.

The ultimate goal of targeted HF therapy will be individualized treatment based on each patient's clinical and genetic signatures. This, however, requires knowledge of each patient's genetic composition, meaningful interpretation of genetic variations in disease, and their interactions with therapies. For example, sexual dimorphism has been reported in the transcriptional responses to heart failure: in women, genes involved in cyclic nucleotide metabolism, glucose transport, and neurohumoral pathways are up-regulated, whereas in men, genes involved in arrhythmia, self-immunity, and cellular homeostasis are altered¹⁰⁵. Also, pharmacogenetic studies have uncovered genetic variants contributing to variable responses to ACE inhibitor and β-blocker therapies³². Ultimately, therapeutic modalities, duration of treatment, medication doses will need to be customized for each patient.

Conclusions and perspective

In recent years, significant strides have been achieved in our understanding, and therapeutic targeting, of pathological ventricular remodeling. ACE inhibitors, ARBs, aldosterone antagonists, and MRAs, β-blockers have become standard of care; in patients with advanced heart failure, device-based therapy is important. Yet, heart failure continues to expand rapidly in incidence and prevalence, exacting enormous individual and societal tolls. Further, efforts to translate preclinical discoveries into the clinical arena have disappointed in the majority of instances. Undoubtedly, much of this failure derives from incomplete understanding of the underlying, complex biology and the currently accepted practice of aggregating all forms of heart failure together irrespective of underlying etiology. The extent and proportionate contributions of multiple events, including myocyte loss, hypertrophy, hyperplasia, extracellular matrix changes, metabolic derangements and immunological events differ by etiology and impact the response to therapy in a meaningful way.

Due to redundancy in cellular and molecular pathways involved in LV remodeling, it is unlikely that future therapeutics will target just one cell type or signaling pathway. Also, it is likely that blockade of neurohormones (e.g. catecholamines, angiotensin, aldosterone), or loading (vasodilators or diuretics), may have overlapping cellular targets (e.g. cardiac myocytes, fibroblasts, etc). Furthermore, targeting one signaling pathway or an individual molecule may be inefficient due to the existence of intricate, interlacing cellular signaling networks and regulatory loops; a multidisciplinary experimental/systems biology may be key. In the end, elucidation of the complex and fascinating biology of LV remodeling is likely to yield significant benefit to the growing number of patients with heart failure.