Cardiac remodeling is defined as genome expression resulting in molecular, cellular and interstitial changes and manifested clinically as changes in size, shape and function of the heart.1 The progression of heart failure (HF) is associated with left ventricular (LV) remodeling, which manifests as gradual increases in LV end-diastolic and end-systolic volumes, wall thinning, and a change in chamber geometry to a more spherical, less elongated shape, with a progressive decrease in ejection fraction (EF).

When ventricular remodeling is advanced, it begins to be self-sustaining and capable of driving disease progression, regardless of the patient's neurohormonal status. This may explain why medical therapies lose their effectiveness in end-stage HF, and why some device-based therapies (cardiac resynchronization and mechanical ventricular assistance), which can affect LV remodeling, are beneficial.

The overall importance of ventricular remodeling as a pathogenic mechanism and prognostic determinant is not clear. Some drug therapies and cardiac devices that increase the survival of patients with HF can slow, and in some cases even reverse, certain parameters of remodeling. Controversially, as in the case of etanercept2 and in cardiac resynchronization,3,4 reverse remodeling has not translated into increased survival. Additionally, the molecular mechanisms of reverse remodeling have not been fully elucidated.

Left ventricular reverse remodeling (LVRR) is characterized by a decrease in LV dimensions, normalization of LV shape and improvement of systolic function.

A significant prevalence of recovery of LV function in patients with dilated cardiomyopathy (DCM) has been reported.5 However, such studies included patients with new-onset DCM like acute myocarditis, and other reversible causes of DCM, such as peripartum and alcohol-related DCM. The mechanisms underlying LVRR in such situations appear to be different from those involved in chronic idiopathic DCM.

The aim of this prospective study was to assess recovery of LV function and reversal of ventricular remodeling in patients with chronic idiopathic DCM, after optimized medical therapy. We set out to assess its prevalence, to identify its predictors and to determine whether it was associated with better prognosis.

The study included consecutive adult patients with idiopathic DCM (left ventricular diastolic diameter [LVDD] >33 mm/m2 in men, >32 mm/m2 in women) between 2000 and June 2012 followed in an HF clinic, diagnosed less than 24 months previously and with two initial values of left ventricular ejection fraction (LVEF) of <0.40 more than one year apart.

We excluded DCM patients with secondary etiologies, including a history of myocardial infarction or angina, those with ischemia or significant coronary disease on coronary angiography, a history of moderate or severe hypertension, at least moderate primary mitral or aortic valvular disease, heavy alcohol use (>100 g/day), chemotherapy-induced and peripartum cardiomyopathy, acute HF with biopsy positive for acute myocarditis or positive serology for acute bacterial or viral infection. We included patients with idiopathic DCM, diagnosed after respiratory infections but with LV dysfunction that persisted for over a year (in order to exclude myocarditis). We also excluded patients with uncontrolled atrial and ventricular arrhythmias.

At baseline, patients underwent clinical assessment, electrocardiogram (ECG), 24-hour ECG, transthoracic echocardiogram, blood laboratory measurements, cardiopulmonary exercise testing (CPET) and cardiac magnetic resonance (CMR).

Patients were managed according to current clinical practice guidelines and clinicians aimed to reach the recommended target doses for all therapies.

During follow-up, periodic clinical assessment, laboratory measurements and echocardiogram were performed.

This study was in accordance with the recommendations set by the Declaration of Helsinki and with local legal requirements.

In the present study, we describe the frequency of improvement in LV systolic function in patients with chronic idiopathic DCM in an unselected population.

LVRR has been described in secondary forms of DCM, such as peripartum cardiomyopathy, alcohol abuse, myocarditis and ischemic heart disease, but the mechanisms underlying such conditions are different from those in idiopathic DCM.8,9

A significant prevalence of recovery of LV function has also been described in recent-onset DCM. Those patients have a higher potential for LVRR, due to resolution of the underlying disease, as in myocarditis, or to favorable effects of therapy. Kubanek et al.10 reported a prevalence of 45% of LVRR at 12 months in 44 patients with recent-onset DCM, including some with active and resolving myocarditis. We only included patients with idiopathic DCM diagnosed less than 24 months previously, but with two initial values of EF of <0.40 more than one year apart, in order to exclude resolving myocarditis.

In our population, LVRR occurred in approximately one third of patients within 22 months of diagnosis. It was associated with improvement in NYHA functional class, with decrease in BNP compared with those who did not recover, and with excellent prognosis.

Recovery in EF and reverse remodeling was associated with maximal treatment with ACEI/ARB and beta-blockers. Patients with LVRR were less often medicated with aldosterone antagonists, probably because they achieved better NYHA functional class.

A favorable response to drug therapy with ACEI, beta-blockers and aldosterone antagonists was reported, with almost complete reversal of LV dysfunction. An increase in EF of more than 15 units has been described, associated with increases in functional capacity and cardiac index and a decrease in pulmonary capillary pressure, associated with a better prognosis.11–14 Treatment of HF can influence hemodynamics by decreasing LV afterload and preload. The experimental literature suggests that alterations in the biology and contractility of the failing cardiac myocyte may be reversible after beta blockade. Recent studies in patients treated with beta-blockers who had an increase in EF also showed favorable changes in myocardial gene expression: an increase in sarcoplasmic reticulum calcium ATPase mRNA and alpha-myosin heavy chain mRNA and a decrease in beta-myosin heavy chain mRNA.15

In our study, patients with LVRR more often had hypertension and appeared to be at an earlier stage of the disease, with lower LVDD, shorter QRS interval, less LBBB and more favorable ventilatory efficiency. Patients with hypertension and LV dysfunction respond to appropriate afterload-reducing therapy with improvements in LV function, and probably more frequently and more rapidly reach maximum drug titration with beta-blockers and ACEI.

Although only 14% of patients had AF at first consultation, the higher percentage of AF among patients who recovered EF was somewhat surprising. One possible explanation is that AF might have developed simultaneously with heart failure, causing functional changes (irregular and rapid rhythm, loss of atrioventricular synchrony, and loss of atrial transport), which would then show maximum benefit from medical therapy, with reversal of ventricular dysfunction.16

The predictors of RRVE in CPET were higher POUE and lower dVE/VCO2/VO2. Decreased oxygen efficiency slope and lower ventilatory efficiency, determined by the VE/CO2 slope, additionally normalized for peak VO2, are sensitive and early prognostic factors of heart failure, reflecting more advanced disease.17,18

Our results are consistent with other studies that set out to define the clinical variables associated with improvement in LVEF. Cicoira et al.19 evaluated 98 patients with idiopathic DCM, and found that those who recovered LV systolic function had shorter duration of symptoms, worse NYHA class and a history of hypertension. In a large study,20 LVRR was found in 89 of 242 idiopathic DCM patients (37%) and baseline predictors were higher systolic blood pressure and absence of LBBB. Binkley et al.21 showed that patients who recovered LV function were younger, had higher systolic blood pressure, lower serum creatinine, shorter QRS interval, lower prevalence of diabetes and higher prevalence of hypertension, were more frequently female and had a lower prevalence of ischemic cardiomyopathy.

It has been postulated that non-ischemic etiology has a higher probability of reverse remodeling. This appears to be related to a higher degree of adrenergic activation for an equivalent degree of myocardial dysfunction and to a greater extent of viable myocardium in patients with idiopathic DCM. A marked reduction in sympathetic activity appears to reduce mortality. The extent of heart rate reduction, rather than its baseline level, appears to be associated with a greater increase in LV function.22

Contractile reserve has been suggested as a key predictor of LVRR, according to studies with dobutamine echocardiography23 and positron emission tomography.24 CMR with gadolinium administration indirectly demonstrates contractile reserve in patients with idiopathic DCM, through the presence of myocardial fibrosis. Some studies have assessed the prognostic value of CMR in non-ischemic DCM. In the study by Assomull et al., 25 midwall fibrosis was present in 35% of 101 patients and was associated with a higher rate of the primary combined endpoint of all-cause death and cardiovascular events. In one study of recent-onset DCM, the lower extent of LGE and the higher edema ratio at CMR were the most important baseline predictors of LVRR.10 In our study, the presence or extent of LGE was not a predictor of LVRR, possibly due to the small study population.

QRS duration is one of the most sensitive independent predictors of survival in patients with DCM. In our population, mean QRS duration of patients who did not recover LV function was 130 ms. This finding is consistent with recommendations for biventricular pacing. Patients with LVRR also less often had non-sustained ventricular tachycardia on 24-hour ECG, probably also reflecting some positive electrical remodeling.

To summarize, these variables probably discriminate patients in whom EF can recover with medical therapy only from those who may require resynchronization devices or more aggressive strategies, including heart transplantation. Patients whose LV function recovers no longer have indication for ICD or CRT therapy, thus complicating the timing of implantation of these devices. Although current guidelines suggest that an ICD is indicated only in patients already receiving maximal medical therapy, it is not clear how safe it is to wait for optimization of therapy before ICD implantation. We can postulate that in patients with LBBB, low systolic blood pressure and larger LV diameters, it may not be safe to wait for ICD/CRT implantation.

In this study we did not perform the expected number of CMR and CPET exams.

Another study is ongoing in our HF clinic, in a cohort of idiopathic DCM patients, all in sinus rhythm, assessing emerging laboratory predictors of LVRR and obtaining detailed echocardiographic data, with volumetric measures and myocardial deformation changes.

LVRR occurred in approximately one third of patients with idiopathic DCM, and these patients appeared to be at an early stage of the disease, had higher blood pressure and had maximal therapy titration. In these cases there is no longer indication for ICD or CRT implantation, thus complicating the timing of implantation of these devices.

The authors have no conflicts of interest to declare.