Portal de Periódicos da CAPES

Reversal of ventricular remodeling: Important to establish and difficult to define

2006; Elsevier BV; Volume: 9; Issue: 4 Linguagem: Inglês

10.1016/j.ejheart.2006.09.008

1879-0844

Paul J. Hauptman, Hani N. Sabbah,

Cardiovascular Function and Risk Factors

"I do applaud your noble goals. Now let's see if you achieve them" Guenevere to the Knights in "Camelot", Since the term "remodeling" was used more than 20 years ago 1, a multitude of studies in both animal models of experimentally-induced heart failure and humans with heart failure have described extensive pathophysiologic changes that occur following myocardial injury at both the "global" and "cellular" levels. These changes occur regardless of the type of injury such as ischemia, pressure overload, volume overload or other factors that lead to intrinsic abnormalities in the contractile machinery 23456. The abnormalities that occur in left ventricular (LV) size and shape, cardiac interstitium, electrophysiological milieu, and myocyte biology indicate that phenotypically and genotypically, by autocrine, paracrine and endocrine mechanisms, the heart dramatically undergoes a major maladaptive transition as heart failure ensues and progresses relentlessly toward the intractable phenotype that ultimately culminates in death. A major challenge for the clinical sphere has been to demonstrate that the two levels of remodeling, "global", as gauged by the extent of increase in ventricular size, mass and sphericity of the left ventricle and "cellular or structural" as manifested by changes in the individual myocyte, interstitium, signal transduction pathways, sub-cellular apparatus and nucleus are essentially synonymous. The impetus comes from the recognition that although the extent of global remodeling, on non-invasive imaging, often correlates in multivariable models with clinical outcomes and prognosis 789, it is the structural component that underlies the fundamental change in cardiac function. Certainly, not all examples of LV dilation represent maladaptive remodeling in the true sense; for example, the athlete's heart likely represents a largely physiological increase in LV mass and LV end-diastolic dimensions. At the same time, there has been an apparent adoption of the belief that the converse is true. Specifically, it is presumed that a partial or complete reversal of ventricular dilation is advantageous and an appropriate surrogate for outcomes even in the absence of evidence for a return to a normal structural profile. We will argue that this concept is correct on many levels but also flawed and that more studies are required before assuming that reversal of global remodeling is synonymous with reversal of structural remodeling. Defining the Term: "Reverse remodeling" is a term that has been applied in clinical trials to a dilated LV that seemingly has reversed its abnormal trajectory by reverting to a more normal size and shape. However, this is a definition that lacks specificity. For example, does evidence for a reversal of global remodeling suffice in the absence of proof of a favorable and durable structural change or a measured survival benefit to establish efficacy? Does recovery of function exclusively involve a reversal of the abnormalities associated with remodeling or are other pathways related to the reparative or recovery process involved? These are important questions because 1) an improvement in patient survival has become increasingly difficult to achieve as the primary endpoint without mega-clinical trials, 2) proof of concept, early stage clinical trials are often designed to demonstrate that the LV is smaller and less spherical and 3) we lack the means to readily measure changes at the cellular and/or sub-cellular level in patients with heart failure. Animal Models for Examining the Reversal of Remodeling: The most controlled situation for the examination of the reversal of LV remodeling exists in studies of animals with experimentally-induced heart failure. Koch's postulates can often be satisfied in these experimental models. With pharmacological and/or device interventions, we can establish if an intervention reproducibly affects both the global and structural remodeling process; if removal or cessation of the intervention brings about a recurrence of the processes that accompanied the initial remodeling; and if recovery of function is re-established upon re-administration of the intervention. However, not all interventions are so easily removed and re-administered; further, the relationship between acute or sub-acute and chronic changes, on both the cellular and global levels, is not known. The Reversal of Remodeling in Human Heart Failure: A fundamental issue for any therapeutic intervention is whether a reversal of global remodeling correlates with improved outcomes in human heart failure. Several analyses have suggested that patients' outcomes with pharmacologic therapy correlate with the direction of change in LV volumes 78910. This concept has obvious appeal since it is difficult to measure many of the other processes that accompany remodeling in patients. Hence, imaging studies obtained prior to and after administration of selected pharmacological or device therapies can be used to demonstrate that LV size and function are favorably impacted. However, these changes do not always correlate with improved clinical outcomes 11. Part of the explanation may be that drugs and interventions can have apparently favorable acute and sub-acute effects on ventricular size but unfavorable and potentially unrelated effects long term. In general, the study in humans of reverse remodeling at the cellular/structural level, such as suppression of the expression of the fetal gene program and recovery of individual myocyte function has not been practical given the need to examine myocardial tissue in serial fashion. There are notable exceptions. The use of ventricular assist devices in patients who are bridged to transplant allows for the collection of tissue at the time of initial device implantation and at the time of device explant during a transplant procedure. These studies have confirmed that structural changes occur at the myocyte level in response to mechanical unloading 121314. Similarly, in studies of the effects of β-adrenergic antagonists, examination of serial endomyocardial biopsy specimens has demonstrated a favorable increase in gene expression for sarcoplasmic reticulum calcium ATPase and α-myosin heavy chain in concert with an increase in LV ejection fraction 15. However, for the most part, we have to interpret surrogate data that largely address structure and function at the global level. This is an important limitation. Udelson and Konstam have acknowledged, "…it is unlikely that one measurement such as a volume change will provide a complete description of natural history outcomes in such a biologically complex process" 7. Indeed, volume changes alone cannot be interpreted as reversal of remodeling; one need only point to LV volume reduction surgery that brought about an instant change in LV size but potentially adverse effects on cardiovascular mortality and morbidity 16. Nonetheless, a goal of other surgical interventions, such as mitral valve reparative surgery for functional mitral regurgitation and the Dor procedure is ventricular volume reduction and change in LV shape 17181920. By that mechanism, it is often suggested or assumed that there is an associated improvement in myocardial function and possibly LV mechanical efficiency. Perhaps no other pharmacological or device intervention has a larger repository of established data for both global and structural changes of reverse remodeling than the β-adrenergic antagonists. Selected agents in this class improve patient outcomes of survival and hospitalization 212223; effect increases in LV ejection fraction, decreases in LV mass and sphericity 24,25; and as noted earlier are associated with changes in structural parameters 15. Exercise tolerance may not be consistently and significantly impacted, but the data clearly demonstrate that an improvement in indices of ventricular size and function at rest and with exercise 26 and favorable changes in the natural history of heart failure are achieved. Early observations have also suggested that the clinical changes observed in patients with dilated cardiomyopathy regress upon removal of beta blocker and return on re-exposure 27, thereby satisfying Koch's postulates. However, at the same time, these data suggest that the improvements in ventricular size and outcomes do not necessarily represent a "cure" but rather a favorable albeit temporizing change in the trajectory of the underlying cardiomyopathic process. More recently, resynchronization therapy has been advocated as a useful treatment on the basis of data suggesting an improvement in symptoms and exercise performance, a decrease in ventricular size and improved outcomes in open label studies 2829303132333435. However, the extent of volume reduction needed to ensure improvement in outcomes and the degree to which these changes are sustained before prognosis is meaningfully impacted are not yet known. In the echocardiographic sub-study of the MIRACLE trial 28 both LV end-diastolic and end-systolic volumes decreased in the paced cohort and LV ejection fraction improved at 3 months. Nevertheless, correlations between the changes in echocardiographic parameters and clinical outcomes were weak, with absolute r-values ranging from 0.12 to 0.20. Despite this, the authors used the term "reverse remodeling" multiple times to describe these echocardiographic changes. A related issue is whether the echocardiographic changes are sustained. Preliminary data are not clear. Observations of Yu and colleagues 32 suggest an immediate reversion toward baseline of multiple hemodynamic and echocardiographic parameters (ejection fraction, dp/dt, isovolumic relaxation time and degree of mitral regurgitation) when biventricular pacing is discontinued among patients with NYHA Class III and IV heart failure. Within 1 to 4 weeks, the reversion to baseline is complete. Therefore, the degree to which structural changes have occurred with pacing must be brought into question. Challenges for future investigations of interventions in heart failure: In light of the uncertainty over the precise nature, extent and sustainability of the processes underlying a reversal of remodeling on both the global and structural levels in heart failure, it would be helpful to reach consensus about key markers so that data can be compared across different intervention studies of drug and device therapies. However, not only is withdrawal of some interventions technically unfeasible (as in cardiac containment devices), but also theoretically ethical objections could be raised in clinical investigation. Nevertheless, in human heart failure, more studies of reverse remodeling are needed. Several "layers" of evidence can be considered. First, in lieu of mortality data, any intervention that proves to be reproducible and demonstrates sustainable effects on LV function, size, shape and mass (even when the intervention is withdrawn for a predetermined, albeit brief period) would provide important supporting evidence for a true biological effect. Increasingly, magnetic resonance imaging has been recognized as a highly sensitive tool to document response to treatment in both animal models and humans 25,36. In some instances, using a double crossover design, re-introduction of the intervention can be tested. Second, data supporting cellular and structural improvement may require myocardial tissue, a difficult but not impossible task, in order to demonstrate reversal of cellular, structural, biochemical and molecular changes. Less appealing but potentially important is the use of plasma or serum-based biomarkers that might be coupled with clinical outcomes of survival or quality-of-life. At this time, the exact parameters to measure are not known but it can be argued that they should at least be consistent with the hypothesized effect of the intervention. The regulatory climate remains fixed: surrogate outcomes are not satisfactory. The true test of an intervention remains its performance in clinical testing. It is in this setting that the claim about reversed remodeling should be accepted cautiously. Ultimately, developing support for the concept of reverse remodeling as a surrogate endpoint is reasonable. Perhaps investigators will have to show in a focused "clinical trial", yet to be designed but incorporating repeat endomyocardial biopsies, that cellular and structural measures of reverse remodeling are consistent with measures of global remodeling which themselves are associated with improved outcomes. This effort, which would likely require sponsorship of granting agencies, research consortia and regulators, would allow us to assess whether we have in our possession the Holy Grail of heart failure clinical trials. Though difficult to define, a convergence of the evidence would at least move us away from counting proverbial "body bags" and "number of hospital admissions" and toward a scientifically sound and clinically responsible evaluation of therapeutic interventions.