OUP user menu

Reverse remodeling in heart failure – fact or fiction?

Burkert Pieske
DOI: http://dx.doi.org/10.1016/j.ehjsup.2004.05.019 D66-D78 First published online: 1 August 2004


Maladaptive remodeling is associated with impaired prognosis in heart failure, and prevention of remodeling is an established therapeutic target. However, it is much less clear whether remodeling may be reversed once it has developed.

In the last decade, anti-neuroendocrine therapy with ACE inhibitors, and even more potently, beta-blockers, was shown to improve surrogate markers for reverse remodeling, such as ejection fraction (EF), ventricular volumes, and mass. For beta-blockers, reverse molecular remodeling was also shown in biopsy specimens on the cellular and subcellular level. Both moderate endurance training and continuous positive airway pressure (CPAP) therapy in heart failure patients with sleep apnea induce reverse remodeling. Cardiac resynchronization therapy improves exercise capacity and quality of life in patients with ventricular dyssynchrony and is clearly associated with geometrical and functional reverse remodeling over time. Whether this translates into improved survival remains to be demonstrated. Surgical approaches for reverse remodeling, such as mitral valve replacement, aneurysmectomy, and volume reduction (Batista procedure) have been developed, but may also be associated with high perioperative mortality. Mechanical unloading of the failing ventricles by left ventricular assist systems (LVAD) induces well-characterized reverse remodeling on the cellular and subcellular level. However, persistent functional improvement in a significant subset of patients that would allow weaning form the device is still under debate. Novel, complementary approaches, such as gene transfer or stem cell therapy are under pre-clinical and clinical investigation.

Taken together, reverse remodeling can be induced by pharmacological and non-pharmacological therapy and may serve as a surrogate parameter for therapeutic success in the individual patient. Since maladaptive remodeling is associated with poor prognosis, identification of novel strategies to reverse this process remains a promising target.

  • Reverse remodeling
  • Heart failure


Heart failure remains a major clinical and economic health care problem, with increasing prevalence mainly related to better survival from myocardial infarction and increased longevity of the population. Despite improved therapeutical options, reduction in mortality from heart failure only marginally contributed to the average 3.9 years increase in life expectancy attributable to reduction in cardiovascular mortality from 1970 to 2000 in the United States.58 In the Framingham cohort, the 1-year age-adjusted heart failure mortality rates declined insufficiently from 30% (1950–1969) to 28% (1990–1999) in men, and from 28% to 24% in women.61 Therefore, a better understanding of the pathophysiology of development and progression of heart failure, and how therapy affects this process is of predominant clinical relevance.

Remodeling and reverse remodeling

The progression of heart failure is associated with left ventricular remodeling, which manifests as gradual increases in left ventricular end-diastolic and end-systolic volumes, wall thinning, and a change in chamber geometry to a more spherical, less elongated shape (Fig. 1). This process is usually associated with a continuous decline in ejection fraction. The concept of cardiac remodeling was initially developed to describe changes which occur in the days and months following myocardial infarction.76 It has been extended to cardiomyopathies of non-ischemic origin, such as idiopathic dilated cardiomyopathy or chronic myocarditis,27 suggesting common mechanisms for the progression of cardiac dysfunction.

Fig. 1
Fig. 1

Post-infarction chamber remodeling (see text). (Reproduced from: Pfeffer M. In: Colucci W, editor. Atlas of heart failure, 1999.)

The process of cardiac remodeling is influenced by hemodynamic load, neurohumoral activation, and other factors still under investigation. The myocyte is the major cardiac cell involved in the remodeling process. Other components include the interstitium, fibroblasts, collagen, and coronary vasculature (Table 1); relevant processes also include ischemia, cell necrosis and apoptosis.15 Due to continuous maladaptive remodeling, myocardial dysfunction is usually a progressive condition,49 where even mild initial dysfunction may develop to severe heart failure over a time course of months to years. Functional polymorphisms in modifier genes relevant for disease progression may impact on the remodeling process.85 The results of cardiac remodeling include progressive worsening of systolic and diastolic function, development of mitral regurgitation, and increased propensity for arrhythmias. A hallmark in remodeling is alteration in the phenotype of the myocytes with re-expression of a fetal gene program, defective excitation–contraction coupling, and disturbed intracellular Ca2+ handling,78 for review, see37. Despite myocyte hypertrophy, this leads to defective contractile function, which may contribute to further progression of myocardial remodeling.

View this table:
Table 1

Cellular changes during remodeling

MyocytesHypertrophy, cell lengthening
Re-expression of fetal gene program
Altered excitation–contraction coupling
Altered energy metabolism
Altered myofibrillar content and function
VasculatureEndothelial dysfunction
Intima thickening
Smooth muscle hyperplasia
Rarefication of capillaries
InterstitiumInduction of matrix metalloproteases
Myocyte slipping
Increased collagen synthesis, fibrosis
Collagen isoform shift

Since cardiac remodeling is an important aspect of disease progression, preventing or reversing maladaptive remodeling is an accepted therapeutical target. The most effective strategy to prevent pathological remodeling after myocardial infarction is immediate reperfusion to minimize myocardial damage. Albeit controversial,67 late reperfusion following myocardial infarction reduced infarct expansion and pathological remodeling both in animal models42,77 and in humans.44 Consequent early initiated pharmacological therapy with ACE inhibitors83 and beta-blockers22 may prevent or slow ventricular remodeling after myocardial infarction. However, it is much less clear whether remodeling may be reversed once it has developed. In clinical practice, changes in ejection fraction, LV end-diastolic and end-systolic volumes, mass, and spericity index are used as surrogate parameters for remodeling or reverse remodeling. In some circumstances (if myocardium is available at different time points), remodeling may also be assessed on the cellular levels (see Table 2).

View this table:
Table 2

Examples for reverse remodeling with therapy

Base6 months
LVEDVI (ml/m2)142±26135±26Embedded Image(Giannuzzi et al., 2003; Embedded Image)
LVESVI (ml/m2)107±2497±24Embedded Image(Giannuzzi et al., 2003; Embedded Image)
EF (%)25±429±4Embedded Image(Giannuzzi et al., 2003; Embedded Image)
LVEDD (mm)64.3±1.863.4±1.8(Kaneko, 2003; Embedded Image)
LVESD (mm)54.5±1.851.7±1.2Embedded Image(Kaneko, 2003; Embedded Image)
EF (%)25±334±3Embedded Image(Kaneko, 2003; Embedded Image)
37.6±2.542.6±0.3Embedded Image(Mansfield, 1997; Embedded Image)
20.6±11.328Embedded Image(Sin, 2000; Embedded Image; #=estimated)
LVEDVI (ml/m2)100.2±4.695.6±4.9(Doughty, 1997; Embedded Image; Embedded Image n.a.)
LVESVI (ml/m2)72.9±4.165.5±4.5(Doughty, 1997; Embedded Image; Embedded Image n.a.)
EF (%)28.6±0.934.1±1.5(Doughty, 1997; Embedded Image; Embedded Image n.a.)
Cardiac resynchronization therapy
LVEDD (mm)72.7±9.271.6±9.1(Gras et al., 2002; Embedded Image)
70±10−3.5Embedded Image(Abraham, 2002; Embedded Image)
74±1067±12(Linde, 2002; Embedded Image)
LVESD (mm)63±1058±12(Linde, 2002; Embedded Image)
LVESVI (ml/m2)100±3692±40Embedded Image(Saxon, 2002; Embedded Image)
116±4385±29Embedded Image(Pitzalis, 2002; Embedded Image)
LVEDVI (ml/m2)129±37121±45(Saxon, 2002; Embedded Image)
150±53119±37Embedded Image(Pitzalis, 2002; Embedded Image)
EF (%)21.7±6.426.1±9.0Embedded Image(Gras et al., 2002; Embedded Image)
21.8±6.3+4.6Embedded Image(Abraham, 2002; Embedded Image)
24.5±7.830.0±12.1(Linde, 2002; Embedded Image)
24±529±6Embedded Image(Pitzalis, 2002; Embedded Image)
  • LVEDVI, left ventricular end-diastolic volume index; LVESVI, left ventricular end-systolic volume index; EF, ejection fraction; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; CPAP, continuous positive airway pressure; OSA, obstructive sleep apnea; CSA, central sleep apnea; n.a., not available.

Non-pharmacological approaches for inducing reverse remodeling

Long-term moderate exercise training has been shown to induce reverse remodeling in patients with stable chronic heart failure, and this was associated with significant increases in work capacity and peak oxygen uptake.29 Central and obstructive sleep apnea is highly prevalent in patients with asymptomatic and symptomatic left ventricular dysfunction and is associated with impaired prognosis. Sin et al.84 demonstrated the effectiveness of continuous positive airway pressure (CPAP) therapy in heart failure patients and central sleep apnea: ventricular function significantly improved after 3 months, associated with a relative 81% risk reduction for mortality or cardiac transplantation. Recently, in a randomized trial in 55 patients, CPAP therapy for 3 months improved ventricular function and reduced sympathetic activity in patients with heart failure and obstructive sleep apnea (OSA).66 Likewise, CPAP therapy in OSA patients reduced daytime blood pressure and heart rate, and was associated with a significant reduction in left ventricular end-systolic dimensions and increases in ejection fraction.46

Reverse remodeling in heart failure with medical therapy

Both ACE inhibitors and beta-blockers have been shown to slow, or even temporarily reverse, remodeling in heart failure. Initial results with ACE inhibitors clearly indicated benefit, but were controversial with respect to true reverse remodeling: Captopril did not lead to a reduction in LV volume, but attenuated progressive LV dilatation, in 59 patients after anterior myocardial infarction and an ejection fraction Math45% (Pfeffer et al., 1988). In contrast, captopril, compared to placebo, produced a significant reduction in LV end-systolic volume and an increase in ejection fraction after 3 months of therapy in 100 patients after myocardial infarction.83 Substudies of the SOLVD trial evaluated the effect of enalapril versus placebo on serial changes in left ventricular volumes and ejection fraction. Despite small numbers included, these studies demonstrated early and sustained reduction in left ventricular volumes after initiation of ACE inhibition in both asymptomatic and symptomatic patients.53,54 In the larger echo substudy of the SOLVD trial, enalapril treatment prevented further LV enlargement, associated with a slight reduction in LV mass, over a follow-up period of 12 months32 (Fig. 2, left). However, in the SAVE trial, captopril attenuated LV dilatation within the first year after myocardial infarction, but not in the second year of follow-up.87 This indicates that patients may escape from the beneficial effects of ACE inhibition on maladaptive remodeling after prolonged periods of time. Taken together, ACE inhibitors attenuate or prevent further remodeling, and may induce modest reverse remodeling in subgroups of heart failure patients. Whether more complete inhibition of the RAAS system by combining ACE inhibitors with AT1 antagonists and aldosterone receptor blockers is more effective for induction of reverse remodeling awaits clarification. Recently, the large Math ValHeFT trial demonstrated that combination therapy of the AT1 receptor antagonist Valsartan with an ACE inhibitor was more effective than ACE inhibitor therapy alone to induce reverse remodeling in patients with symptomatic systolic heart failure: both ejection fraction increased, and left ventricular end-diastolic dimension decreased significantly more with combined RAAS blockade over prolonged (Math24 months) periods of time.98

Fig. 2
Fig. 2

Left: Changes in left ventricular end-diastolic (EDV) and end-systolic (ESV) volumes and mass. Attenuation of LV dilatation and hypertrophy during the 12-month period of observation is seen in the enalapril-treated group. The Embedded Image value is for the repeated-measures ANOVA comparing the response in the two groups over time. (Modified reproduction from: Greenberg et al. Circulation 1995.) Right: Changes in left ventricular end-diastolic (EDVI), end-systolic (ESVI) volume index, and ejection fraction (EF) from baseline during the 12 month period of observation. Reversal of LV dilatation and improved EF is observed with carvedilol therapy. The Embedded Image value is for the repeated-measures ANOVA comparing the response in the two groups over time. (Modified reproduction from: Doughty et al. J Am Coll Cardiol 1997.)

Compared to ACE inhibitors, even more pronounced effects on reverse remodeling may be observed with beta-blockers in heart failure patients.51 In an elegant study, Hall et al.34 reported a progressive increase in ejection fraction (Fig. 3), regression of dilatation and hypertrophy, and restoration of a more elliptical chamber shape in patients treated with metoprolol over a time course of 3–18 months. Reverse remodeling with β1-adrenoceptor blockade was confirmed in a substudy of the MERIT-HF trial, where patients treated with metopolol CR/XL were followed by magnetic resonance imaging. Significant decreases in LV volumes, and increases in ejection fraction were observed after 6 months in the verum, but not the placebo group.33 Similarly, using the non-selective adrenoceptor blocker carvedilol, Olsen et al.74 reported reverse remodeling associated with improved symptoms in chronic heart failure. In the echo substudy of the Australia–New Zealand trial in patients with ischemic cardiomyopathy and severely depressed LV function,21 carvedilol treatment over 12 months was associated with smaller left ventricular end-diastolic and left-ventricular end-systolic volumes, and increased ejection fraction (Fig. 2, right). A meta-analysis of all available beta-blocker trials showed an average 29% relative increase in ejection fraction, irrespective of the etiology of heart failure.56 In these studies, reverse remodeling with selective or non-selective beta-blockers was observed even in the presence of baseline therapy with an ACE inhibitor. These data support the hypothesis that the prognostic benefit of beta-blocker therapy in heart failure is related to the potential to prevent and reverse left ventricular remodeling.

Fig. 3
Fig. 3

Changes in ejection fraction from baseline to day 1, month 1, and month 3 in the metoprolol group and the placebo group.

Recently, reverse remodeling with beta-blocker therapy has also been documented on the subcellular level in isolated human myocardium. Some beta-blockers may increase the number of β-adrenergic receptors, which are downregulated in heart failure.30 Lowes et al.64 demonstrated an association between normalization of myocardial gene expression for SERCA2a and α- and β-myosin heavy chains and improvement in ejection fraction and clinical status in patients with idiopathic dilated cardiomyopathy under beta-blocker therapy. Reiken et al.82 showed that beta-blocker therapy partially restored diastolic filling, β-adrenergic responsiveness, and ryanodine release channel function in patients with dilated or ischemic cardiomyopathy. These data provide insight into subcellular mechanisms for reverse remodeling with beta-blockers.

Novel pharmacological approaches for reverse remodeling in heart failure have been developed. For example, pharmacological blockade of the sarcolemmal sodium/hydrogen exchanger may prevent or reverse maladaptive remodeling:47 Cariporide prevented fibrosis and heart failure in β1-adrenergic transgenic mice24 and reversed isoproterenol-induced hypertrophy in rats.25 Another NHE1 inhibitor, EMD-87580, induced reverse remodeling in a rat model of post-myocardial infarction heart failure.13

Taken together, ACE inhibitors and even more pronounced, beta-blockers, induce reverse remodeling in heart failure, and this may in part explain improved clinical outcome. In the future, more complete blockade of neuroendocrine activation, e.g. by addition of AT1-antagonists and aldosterone receptor blockers, possibly in combination with complementary pharmacological approaches, may be even more effective for reverse remodeling.

Reverse remodeling with cardiac resynchronization therapy

Patients with heart failure and asynchronous wall motion due to intraventricular conduction delay are at increased risk for exacerbated pump failure. Biventricular and left ventricular cardiac resynchronization therapy (CRT) can re-coordinate contraction, thereby acutely improving systolic ventricular function4,48 and energetic efficiency.95 CRT has been shown to improve symptom class, exercise capacity, and quality of life even in patients who are already receiving optimal pharmacological therapy. Albeit the mechanisms of benefit remain poorly understood, reverse remodeling of the failing ventricles may be a major factor. Yu et al.99 reported that both end-diastolic and end-systolic chamber volumes declined already after 1 month of CRT (Fig. 4). When pacing was transiently stopped after 3 months continuous CRT in this study, systolic contractile function (assessed by Mathmax) acutely declined, but without enlargement of ventricular volumes. However, when pacing was kept off for the next month, there was re-initiation of chamber dilatation and remodeling (Fig. 4). In another small trial using radionuclide scintigraphy, Toussaint et al.94 reported an immediate, albeit not significant effect of resynchronization on ejection fraction (from 17.8±6.3 to 19.9±8.3), but ejection fraction further increased to 24.2±10.8% Math after 1 year of CRT. These data indicate that CRT has both immediate and long-term effects on ventricular function, the latter possibly associated with reverse remodeling. Larger, chronic and blinded/controlled studies have confirmed the reverse remodeling effects of CRT on chamber geometry (Saxon et al., 2002).1,86 Recently, reverse remodeling with CRT was also observed in the prospective, double-blind randomized Multicenter Insync Randomized Clinical Evaluation (MIRACLE) study. In this trial, the effect of CRT on chamber geometry and remodeling was assessed by doppler echocardiography in 323 heart failure patients (mean EF, 24±7%) on optimized medical heart failure therapy including ACE inhibitors (Math90%) and beta-blockers (Math55%). 172 patients were randomized to CRT on, and 151 patients to CRT off. CRT induced significant and progressive reverse remodeling over 6 months: left ventricular end-diastolic and end-systolic volumes as well as mitral regurgitation decreased, associated with an increase in ejection fraction (Fig. 5). In addition, LV mass decreased, and myocardial performance and left ventricular sphericity index improved.89 Reverse remodeling was associated with improved NYHA functional class, exercise capacity, and quality of life. Reverse remodeling occurred regardless of the cause of heart failure, but was more extensive in patients with a non-ischemic origin. This observation was confirmed in a substudy of the MUSTIC trial.23

Fig. 4
Fig. 4

Changes in LV end-diastolic Embedded Image and end-systolic Embedded Image volumes, ejection fraction (EF), rate of pressure rise (Embedded Image), and mitral regurgitation (MR) before and after biventricular pacing as well as when pacing was suspended for 4 weeks. *Significant difference vs baseline. †Significant difference vs biventricular pacing for 3 months. The time axis scale is not in proportion. (Modified reproduction from: Yu et al. Circulation 2002.)

Fig. 5
Fig. 5

Changes in ejection fraction (EF), mitral regurgitation (MR), LV end-diastolic (LVEDV), and end-systolic (LVESV) volumes at 3 and 6 months after randomization to the control (closed symbols) or the CRT (black symbols) group. (Modified reproduction from: St. John Sutton et al. Circulation 2003.)

The mechanisms for reverse remodeling with CRT remain to be elucidated, but my comprise: (1) reduced wall stress due to reduced inhomogeneities in regional contractile activation, (2) decreased sympathetic nerve activity,35 (3) increased metabolic efficiency.95 In a recent publication, Penicka et al.75 identified pulse-wave tissue doppler derived intraventicular and interventricular asynchrony as the best predictive factors for reverse remodeling of the left ventricle during CRT. Nevertheless, reverse remodeling was less likely to occur in the presence of very dilated chambers: Stellbrink et al.86 identified CRT “non-responders” with respect to reverse remodeling to have significantly higher baseline LVEDV as compared to “responders” (351±52 vs 234±74 ml). This is consistent with the observation that patients with large ventricles may be reverse-remodeling non-responders to medical therapy.60

Reverse remodeling with surgical approaches

Besides revascularization to improve myocardial perfusion and reverse stunning and hibernation, a number of surgical approaches to reverse maladaptive remodeling have been developed (for review, see2). Batista et al.8 introduced the concept of volume reduction by partial left ventriculectomy in specific cases of dilated cardiomyopathy. The rationale of this operation is to restore a normal ratio between wall thickness and the radius of the left ventricle to normalize systolic wall stress. However, this procedure is associated with high perioperative mortality, and the outcome in terms of clinical improvement appears impredictable. Therefore, it is not regarded as an alternative to transplantation in western countries.70 In contrast to the Batista procedure, restoration of the morphology of the left ventricle by resection of aneurysmatic scar tissue and endoventricular patch plasty repair after myocardial infarction is widely accepted20 and was associated with good clinical and functional midterm results.3 Reverse remodeling with this procedure was documented by improved ejection fraction (from 30±8 to 48±8%), significant decreases in end-diastolic and end-systolic volumes, and a leftward shift of pressure–volume loops associated with increases in end-systolic elastance.12 Patients with congestive heart failure may present with severe mitral regurgitation secondary to altered ventricular geometry. Correction of mitral regurgitation, as first proposed by Bolling et al.,9 is expected to reverse remodeling over time by reducing chronic volume overload and resulted in improved ejection fraction and functional status.10 Recently, the effects of passive containment of the heart using a special net (Accorn Cardiac support device) were associated with reverse remodeling over the 6 months following operation, possibly associated with reduced diastolic wall stress.52 However, perioperative mortality was high (10%), and persistent clinical benefit remains to be established.

Reverse remodeling with cardiac assist devices and novel therapeutic approaches

Left ventricular assist devices (LVADs) are still primarily used to provide temporary support of the failing heart until cardiac transplantation (bridge to transplant). Since hemodynamic overload is regarded as one of the most prominent stimuli for maladaptive remodeling, ventricular unloading with LVADs could also be an effective means to induce reverse remodeling. This has stimulated the concept of LVADs as “bridge to recovery”. In fact, echocardiographic and hemodynamic studies have revealed that unloading the failing heart may improve ventricular function, and markers of neuroendocrine activation and inflammation decrease during LVAD support.45 In addition, there is much evidence for reverse remodeling during mechanical assist, both at the level of the myocyte and the structure and function of the heart: A number of recent reports have demonstrated the effect of LVAD therapy on cellular and subcellular reverse remodeling. In contrast to conventional therapy, paired human samples from the left ventricle, obtained during LVAD implant and explant (usually cardiac transplantation) can be used in these investigations.

One prominent feature of reverse remodeling in LVAD-supported hearts is regression of myocyte hypertrophy and cell length,91,100 and this was associated with echocardiographically documented reduction in left ventricular dilation and mass.100 In addition, albeit not unequivocal17 subcellular myocyte structure, as assessed by restoration of the integrity of the cytoskeletal protein dystrophin, is improved by LVAD.96,97 Another prominent and clinically relevant abnormality in heart failure is action potential prolongation, the clinical correlate being a pathological increase in the ECG QT interval. LVAD support results in normalization of action potential duration in isolated human myocytes (Fig. 6(a)), and this was associated with a marked decrease in heart-rate corrected QTc intervals from 504±11 before to 445±9 ms after prolonged LVAD support.36 Another classical feature in heart failure, β-adenoceptor downregulation, is reversed, associated with improved β-adrenergic functional response19,73 (Fig. 6(b), right). Furthermore, mechanical unloading normalizes the pathological phenotype of the myocytes. A central defect in the failing human heart is the load-dependent downregulation of the sarcoplasmic reticulum Ca ATPase (SERCA2a). This leads to reduced SR Ca uptake capacity, disturbed excitation–contraction coupling processes, and impaired contractile function with reduced developed force and impaired relaxation and diastolic function (for review, see37). This also results in a clinically relevant reversal of the positive force frequency relationship.78 All these features can be normalized by ventricular unloading. As reported by Heerdt et al.39 and Barbone et al.5 SERCA2a gene expression is significantly increased in the unloaded left ventricle, but to a lesser extent in the right ventricle (Fig. 7(a) and (b)). This was associated with a normalization (i.e., positivation) of the inverse force–frequency relation in isolated trabeculae from LVAD-supported human hearts5,39 (Fig. 7(c)). Normalization of SERCA2a with LVAD could only be observed in a subset of patients in another study.7 Normalization of gene expression of major Ca handling proteins are also associated with improved contractile function of isolated myocytes: Dipla et al.19 reported enhanced myocyte shortening, improved relaxation, as well as normalized Ca transients with LVAD support. This is supported by normalization of SR Ca content with LVAD therapy.91 In addition, improved mitochondrial function,57 and normalized mitogen-activated protein kinase signaling26 was reported with LVAD therapy.

Fig. 6
Fig. 6

Cellular reverse remodeling in isolated human myocardium after LVAD support. (a) Normalization of prolonged action potentials in isolated myocytes. (Reproduced from: Harding et al. Circulation 2001.) (b) LVAD restores β-adrenoceptor density (upper panel) and inotropic responsiveness to isoproterenol (lower panel). (Reproduced from: Ogletree-Hughes et al. Circulation 2001.)

Fig. 7
Fig. 7

(a) Typical western blots probed for sarcolasmic reticulum Ca ATPase (SERCA2a) protein in right and left ventricular samples (RV, LV) from non-failing, LVAD-supported, and failing non-supported hearts. (b) Average data for SERCA2a protein expression from 14 LVAD-supported and 14 non-supported failing hearts. SERCA2a significantly increased in the left, but not the right ventricles of LVAD supported hearts, demonstrating load-dependency of molecular reverse remodeling. (Reproduced from: Barbone et al. Circulation 2000.) (c) Force–frequency relation in isolated human trabeculae from a non-failing, two end-stage failing (DCM) hearts, and one end-stage failing heart after LVAD support. LVAD support reversed the negative force–frequency relation to a normal positive one. (Reproduced from: Heerdt et al. Circulation 2000.)

Conflicting data exist with respect to normalization of the extracellular matrix. Early reports demonstrated even an increase in extracellular fibrosis during LVAD,68 other reports showed improved ventricular structure and reduced fibrosis.55 Reduced collagen damage with LVAD support was associated with reduced expression and activity of matrix metalloproteinases and increased expression of tissue inhibitors of matrix metalloproteinases.62 Normalization of extracellular matrix was related to normalization of cytokine expression, such as TNFα with LVAD.55,93 Reverse remodeling on the cellular level is also associated with partial normalization of neuroendocrine activation.72 This might indicate reversal of the vicious cycle, which underlies progression of heart failure, with mechanical support.

LVAD support, also after brief period of support, improves diastolic properties of the failing ventricle, and this may primarily be associated with improved chamber geometry.5,6 Furthermore, in an ex vivo study, pressure–volume relationships were obtained in non-failing human hearts, end-stage failing human hearts, and failing human hearts after LVAD support (Fig. 8). The ex vivo pressure–volume ratio was significantly shifted to the right in non-supported failing hearts, but largely reversed to normal after successful mechanical support.39 Taken together, these data clearly indicate that mechanical support induces reverse remodeling on the subcellular, cellular, and organ level. However, most of the studies did not report on clinical follow-up data of the patients, and therefore, the clinical meaning of these observation remains to be determined.

Fig. 8
Fig. 8

Ex vivo pressure–volume relations in explanted non-failing, end-stage failing, and end-stage failing after LVAD human hearts. LVAD support normalized pressure–volume relations except in patient 7 and 18 with malfunctioning LVADs. (Reproduced from: Heerdt et al. Circulation 2000.)

Consistent with reverse remodeling, clinical trials have reported that unloading of the failing heart by LVAD reversed LV enlargement and improved contractility in vivo,59,65,71 supporting the possibility that patients might be successfully weaned from the device. Hetzer et al.41 reported successful explantation of 33 devices, which was performed if EF increased to Math45%, LVEDD decreased to Math55 mm, and maximal systolic contraction veocity was Math8 cm/s during intermittent stop of the device. In the report by Hetzer et al.,41 a persistent recovery of myocardial function for a time course of up to 7 years was reported in 18 of the 31 cases, and Helman et al.38 reported on the recurrence of symptomatic heart failure after device explantation in two patients. Consistently, until today, only the minority of patients could be weaned from the device for prolonged periods of time,40,65 and reliable indicators for persistent improvement are still lacking. A working group of the National Heart, Lung and Blood Institute of the Unites States recently published recommendations to define outcome markers for patients on LVAD support.81

One potential problem with the use of mechanical unloading is heart muscle atrophy despite reverse remodeling, possibly limiting clinical efficacy of LVADs. This led some investigators to combine mechanical unloading with anabolic therapy, such as the β2 agonist clenbuterol (Yacoub, 2001). In fact, clenbuterol administration was associated with a larger percentage of patients with reverse remodelling and successful weaning from the device.43

Novel therapeutic strategies for reverse remodeling, such as gene therapy or stem cell therapy, may prove effective in the future. In an animal model of heart failure, adenoviral-mediated restoration of SERCA2a gene expression not only induced reverse remodeling, but was also associated with significantly improved survival (Fig. 9). In addition, as indicated by first animal studies,92 combination of different strategies for reverse remodeling, such as stem cell or gene therapy in patients supported by LVAD may proof beneficial and allow stable restoration of cardiac function, and hence, weaning form the device in the future. In addition, the development of smaller, totally implantable LVADs16 might allow hemodynamic unloading in earlier, potentially more reversible stages of heart failure.

Fig. 9
Fig. 9

Upper panel: LV pressure (LVP) vs LV volume detected by piezoelectric crystals in a sham+Ad.βgal-GFP heart, a failing+Ad.βgal-GFP heart, and a failing+Ad.SERCA2a heart. LV volumes were significantly increased in failing rats and were decreased after SERCA2a gene transfer. The slope of the end-systolic pressure–volume relationship was lower in failing hearts infected with Ad.βgal-GFP Embedded Image than in sham Embedded Image, indicating a diminished state of intrinsic myocardial contractility. Gene transfer of SERCA2a restored the slope of the end-systolic pressure–volume relationship to control levels. Lower panel: Survival–function curve of six different groups studied: sham, Embedded Image; sham+Ad.βgal-GFP, Embedded Image; sham+Ad.SERCA2a, Embedded Image; failing, Embedded Image; failing+Ad.βgal-GFP, Embedded Image; failing+SERCA2a, Embedded Image. Sham-operated animals did not show premature mortality. In the failing group, the non-infected animals had a survival curve that decreased steadily, and at 4 weeks the survival rate was only 18%. In the failing+Ad.SERCA2a group, however, the survival curve was significantly improved. (Reproduced from: del Monte et al. Circulation 2001.)

Why is reverse remodeling important?

Drug efficacy for treatment of heart failure is commonly measured in large multicenter clinical trials by assessing endpoints such as all-cause mortality and hospitalization rate. However, given the relatively low death rates in recent heart failure trials under optimized baseline therapy (for example, 1-year mortality was 9% in the “placebo” arm of the Val-HeFT trial, and 4% in the CHARM preserved study) it becomes increasingly difficult to target mortality as primary end-point in heart failure trials. Therefore, regression of maladaptive remodeling might serve as a surrogate marker for morbidity and mortality and success of therapy. This hypothesis is supported by retrospective analysis from heart failure trials14 and recent observational studies.50,90 Nevertheless, most of the major heart failure trials to date have only correlated treatment and cardiac function and have not directly demonstrated a causal relationship between improvement in cardiac function and improvement in long-term health outcomes.28 In the SAVE study, attenuation of ventricular enlargement with enalapril was associated with a reduction in adverse events.88 Quinones et al.80 demonstrated the effect of hypertrophy and LV function on clinical events in the SOLVD trial. Reduced morbidity by combination of the AT1 antagonist valsartan with an ACE inhibitor in the ValHeFT study was associated with better reverse remodeling.98 In a recent observational study, Metra et al.69 showed that the marked improvement in left ventricular function with beta-blocker therapy in chronic heart failure patients correlated with improved 2-years prognosis.

Novel diagnostic modalities, such as gene profiling for assessing risk of disease progression as well as metabolic capacity for a particular drug will be helpful for risk stratification and tailoring of individualized therapy in the near future. In addition, markers of neuroendocrine activation, such as brain natriuretic peptide (BNP) or its inactive cleavage product, NT-proBNP are marketed for diagnosis and monitoring of response to therapy in patients with left ventricular dysfunction. Not all patients responded to beta-blocker therapy with reverse remodeling, and hence, improved prognosis, in the Metra study.69 Therefore, reverse remodeling, which is fundamentally related to the disease process, might also serve as a surrogate marker for response to therapy in the individual patient. In this scenario, serial assessment of parameters of reverse chamber remodeling, possibly in combination with BNP or NT-proBNP, may allow to tailor the armentarium of therapeutical modalities effective in reverse remodeling (Fig. 10) to the individual patient. Nevertheless, the clinical and economic efficacy of this approach must be demonstrated. In addition, as exemplified by the case of the TNFα blocker etanercept,11 reverse remodeling may not always translate into improved survival in heart failure patients.

Fig. 10
Fig. 10

Multimodal therapeutical possibilities for therapeutic interventions with a proven benefit in heart failure. Grey: Interventions with an established effect on reverse remodeling.

In conclusion, reverse remodeling is a fact and may be induced by different therapeutic approaches. With conservative therapy, beta-blockers appear more potent than ACE inhibitors to induce reverse remodeling. Resynchronization therapy, and even much better characterized, mechanical unloading by assist devices, induce reverse remodeling on the cellular and organ level. The role of new therapeutic strategies, such as miniaturized impeller pumps for chronic unloading, gene therapy, or stem cell therapy, in reverse remodeling awaits characterization. In addition, parameters for identification of responders for a specific anti-remodeling approach need to be defined.


This manuscript was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG Pi 414/2).


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
  65. [65]
  66. [66]
  67. [67]
  68. [68]
  69. [69]
  70. [70]
  71. [71]
  72. [72]
  73. [73]
  74. [74]
  75. [75]
  76. [76]
  77. [77]
  78. [78]
  79. [79]
  80. [80]
  81. [81]
  82. [82]
  83. [83]
  84. [84]
  85. [85]
  86. [86]
  87. [87]
  88. [88]
  89. [89]
  90. [90]
  91. [91]
  92. [92]
  93. [93]
  94. [94]
  95. [95]
  96. [96]
  97. [97]
  98. [98]
  99. [99]
  100. [100]
View Abstract