Representative example of reverse remodeling induced by LVAD therapy. Serial chest radiographs and echocardiograms (M mode, parasternal long-axis view) from a male patient with a multiyear history of idiopathic dilated cardiomyopathy and HF refractory to standard therapy, including inotropes (a–c). The patient underwent a combination of LVAD unloading and pharmacologic therapy and was weaned from the LVAD successfully 6 months after implantation. He continued on standard HF pharmacologic therapy and achieved a sustained functional recovery with no signs or symptoms of HF over a 7-year follow-up period. EF ejection fraction (left ventricular), LVIDd LV end-diastolic diameter (From Drakos et al. [14]; with permission).

(a) Quantitative relationship between drug/device effects on ejection fraction (EF) and mortality: each data point represents a placebo-corrected change in EF from an individual remodeling trial plotted against the mortality odds ratio (OR) for the specific therapy. Color-coded mortality effect based on meta-analysis of published data. Interventions were classified as favorable (blue circles) if the upper limit of the 95 % confidence interval (CI) of the OR for death from the mortality trials was less than 1, neutral (black circles) if the 95 % CI crossed 1, and adverse (red circles) if the lower limit of the 95 % CI was greater than 1. There was a significant correlation between short-term therapeutic effect on LVEF and longer-term therapeutic effect on mortality (ρ = −0.51, P < 0.001). Remodeling data were derived from analysis of 86 randomized controlled trials (RCTs) of 25 interventions involving 19,092 patients. (b) Predicted probability of a categorical mortality outcome based on drug/device effect on EF: the lines represent the likelihood of a categorical mortality outcome based on an intervention’s trial-level effect on EF compared with placebo (unadjusted, weighted, ordered logistic regression). Color-coded mortality effect based on meta-analysis of published data. Definition of mortality effect for a given intervention is as described in A. The graph suggests that if the mean change in EF is 10 % in the short-term studies, the probabilities that the long-term mortality studies will be significantly favorable (blue line), neutral (black line), or significantly unfavorable (red line) are approximately 85 %, 15 %, and 1 %, respectively. As such, the degree of drug/device-induced reverse remodeling as reflected by EF change is a quantitative marker of favorable prognosis (Adapted from Kramer et al. [4]).

(a) Quantitative relationship between drug/device effects on end-diastolic volume (EDV) and mortality: each data point represents a placebo-corrected change in EDV from an individual remodeling trial plotted against the mortality OR for the specific therapy. Color-coded mortality effect based on meta-analysis of published data: favorable (blue circles), neutral (black circles), or adverse (red circles). Definition of mortality effect for a given intervention is as described in Fig. 6.2. There was a significant correlation between effect sizes in the mortality trials and mean therapy-induced changes in EDV (ρ = 0.44, P = 0.002). Remodeling data derived from analysis of 50 randomized controlled trials (RCTs) of 19 interventions, including 10,855 total patients. (b) Predicted probability of a categorical mortality outcome based on drug/device effect on EDV: the lines represent the likelihood of a categorical mortality outcome based on an intervention’s trial-level effect on EDV compared with placebo (unadjusted, weighted, ordered logistic regression). Color-coded mortality effect based on meta-analysis of published data. Definition of mortality effect for a given intervention is as described in Fig. 6.2. The graph suggests that if the mean change in EDV is −10 mL in the short-term studies, the probability that the long-term mortality studies will be significantly favorable (blue line), neutral (black line), or significantly unfavorable (red line) is approximately 56 %, 43 %, and 1 %, respectively. As such, the degree of drug/device-induced reverse remodeling as reflected by EDV change is a quantitative marker of favorable prognosis (Adapted from Kramer et al. [4]).

(a) Quantitative relationship between drug/device effects on end-systolic volume (ESV) and mortality: each data point represents a placebo-corrected change in ESV from an individual remodeling trial plotted against the mortality OR for the specific therapy. Color-coded mortality effect based on meta-analysis of published data: favorable (blue circles), neutral (black circles), or adverse (red circles). Definition of mortality effect for a given intervention is as described in Fig. 6.2. There was a significant correlation between effect sizes in the mortality trials and the effect sizes on ESV in the remodeling studies (ρ = 0.48, P = 0.002). Remodeling data are derived from analysis of 40 randomized controlled trials (RCTs) of 16 interventions, including 5,435 total patients. (b) Predicted probability of a favorable mortality outcome based on drug/device effect on ESV. The blue line represents the likelihood of a favorable mortality outcome based on an intervention’s trial-level effect on ESV compared with placebo (unadjusted weighted logistic regression). There was insufficient remodeling data for ESV to model the probability for neutral or adverse outcomes individually. Color-coded mortality effect based on meta-analysis of published data. Definition of mortality effect for a given intervention is as described in Fig. 6.2. The graph suggests that if the mean change in ESV is +25 mL in the short-term studies, the probability that the long-term mortality studies will be significantly favorable is approximately 20 %. This is equivalent to stating that the probability that the long-term mortality studies are neutral or significantly unfavorable is approximately 80 % (100−20 %). As such, the degree of drug/device-induced reverse remodeling as reflected by ESV change is a quantitative marker of favorable prognosis (Adapted from Kramer et al. [4]).

Schematic representation of the β-adrenergic system in the setting of remodeling and reverse remodeling. Remodeling is associated with excessive activation of sympathetic nerve activity [15], reduction in functional neuronal density [16, 17], and downregulation of β-adrenergic receptors [18–20]. It also is noteworthy that the degree of increase in sympathetic nerve activity is related to that of downregulation of β-adrenergic receptors [19]. Reverse remodeling, on the other hand, likely is associated with normalization of such adaptive or maladaptive changes.