Heart failure secondary to systolic dysfunction is a disease of epidemic proportions in the U.S. Although medical therapy can prolong survival, decrease hospitalizations, and alleviate symptoms, the disease is progressive and many patients eventually present with fulminant and/or unremitting symptoms. In this group of patients, left ventricular assistant devices (LVADs) have been utilized to "bridge" patients between LVAD implantation and cardiac transplantation. Interestingly, in a subset of patients, the LVAD can be weaned over time, leading investigators to hypothesize that LVAD support might enable patients to be "bridged to recovery". Indeed, studies have demonstrated salutary changes in myocyte function after chronic LVAD support. However, LVAD supported hearts also demonstrate increased fibrosis, apoptosis, and extracellular matrix remodeling. These reports led us to speculate that therapeutic interventions directed at inhibiting matrix remodeling might improve the ability to wean patients from LVAD support.

During the past five years, we have demonstrated that the proinflammatory cytokine tumor necrosis factor (TNF) plays and important role in the development of the end-stage heart failure phenotype. Infusions of TNF or over-expression of TNF in animal models induces cardiac dilatation, diminished cardiac contractility, fibrosis, and extracellular matrix remodeling: changes that can be reversed by treatment with TNF soluble receptor (sTNFR). Indeed, systemic administration of sTNFR has demonstrated benefits in patients with symptomatic heart failure. As TNF is expressed at high levels in the LVAD supported heart, we hypothesized that treatment with sTNFR would substantially improve the ability to bridge patients to recovery and LVAD weaning.

Unfortunately, systemic administration of TNFR is problematic in LVAD supported hearts because of the inherent risk of infection in these patients and the important role of TNF in the immune response. Thus, we hypothesized that direct injection of the heart muscle with an adeno-associated virus (AAV) during sTNFR expression would have salutary effects. Studies in our own laboratory have demonstrated that AAV vectors: 1) are non-immunogenic; 2) provide persistent expression in muscle; and 3) can be constructed with cardiac specific and tetracycline-responsive promoters. Accordingly, this application will test the hypothesis that direct intramyocardial injection of AAV-sTNFR at the time of LVAD implantation could effect adaptive changes in the heart that could facilitate the ability to wean failing hearts from LVAD support.