Organ level features include increased left ventricular (LV) mass, distortion of chamber geometry with LV dilatation and relative wall thickening, impaired systolic and diastolic function, demonstrable defects in adrenergic responsiveness, and increased arrhythmogenecity (Fig.1). At the cellular level, a number of analogues to the organ level abnormalities exist. Specifically, increased myocyte size, distortion of cell shape and size, myocyte lengthening with increased sarcomeres, impaired cellular shortening and relaxation, a number of electrophysiologic (EP) abnormalities, and altered protein production. One common grouping of abnormal protein expression is referred to as a recapitulation of the fetal phenotype.

The LVAD is a blood pump connected to the LV apex via a 1-inch diameter cannula that effectively drains all returning blood from the left ventricle, decompressing the LV quite profoundly (Fig.2). When the pump is 90% full, it ejects into an outflow conduit anastomosed to the ascending aorta, porcine valves prevent any backflow.

Immediate improvement in effective cardiac output is seen after LVAD placement, even though the native aortic valve is often no longer opening. An immediate decrease in LV diameter and a relative wall thickening, reflecting profound decompression of the distended LV, is seen in the operating room upon LVAD placement.

A number of changes and improvements in the failing heart are produced by LVAD placement. Chronic remodeling of the chamber as a result of chronic unloading. Reduction in LV mass in the weeks and months following LVAD placement. Biochemical composition changes in the heart. One signature abnormality of the failing heart is the expression of atrial natriuretic peptides (ANP) and brain natriuretic peptides in the ventricular myocardium, even in the adult. After weeks of LVAD support a decrease in ANP immunostaining has been shown. Mechanical unloading reduces overall cell volumes by about 40%, cell length about 25%, and the cross-sectional profile of the typical cell is less flat.

Improvement in contractile phenotypes of the failing myocyte has been shown, despite refractoriness to medical therapy. In the typical contractile phenotype of the failing myocyte, the cell shortens as it contracts and lengthens as it relaxes at relatively slow rates compared to non-failing controls. In contrast, in LVAD-supported myocytes, a much greater overall shortening and much faster rates of shortening and re-lengthening are seen.

Adrenergic responses and increased frequency stimulation responses were examined, and showed that LVAD support causes a re-sensitization to adrenergic stimulation in failing myocytes. A small increase in the magnitude of shortening with little change in rates of shortening or re-lengthening in response to isoproterenol in non-supported failing myocytes was shown, while in LVAD-supported hearts there was a greater magnitude of shortening and much greater improvement in the rate of relaxation. The immediate change in shortening with isoproterenol was about 2-fold that in LVAD-supported cells.

Excitation contraction

Excitation contraction (EC) improvement may explain the improved contractile function of failing myocytes, as shown by recent, unpublished studies by Margulies' laboratory. There appears to be some plasticity at the level of the L-type calcium channel, explaining why contraction improves in LVAD-supported myocytes. Associated with the nearly normal recovery of myocytes with LVAD support is a lesser wholesale calcium transient in the failing cells. LVAD support tends to improve the absolute amplitude, peak calcium transient and the rate of decline.

Simultaneous measurement of calcium current with whole cell voltage clamp techniques, cytosolic calcium measurements using the calcium indicator fluo-3 and confocal microscopy, and contraction measurements with video edge detection were performed. The normal bell-shaped voltage current relationship for the L-type calcium channel was exhibited in the non-failing cells. In the failing cells, a bell-shaped curve was exhibited, but for any given membrane potential a lesser L-type channel current was seen. LVAD-supported myocytes exhibited improvement, with a curve similar to the non-failing cells.

EP processes and contractile performance

To look at the natural EP processes governing contractile performance, the current clamp mode action potential from single cells was used (Fig. 3). In the non-LVAD supported cell a marked prolongation of the action potential duration was seen, with some shape distortion. After LVAD support, the myocytes tended to have shorter action potential durations. The action potential is a composite of many different ion channels working together to govern the duration and shape of the action potential. Whether these results reflect a perturbed or unperturbed phenomenon is questioned, as the pipette also tends to dialyze the intercellular contents.

The first finding of global EP remodeling with LVAD support comes from very recent data from Margulies' laboratory. The surface 12-lead ECG from patients supported with LVAD was examined for the composite cardiac repolarization, as indicated by the corrected QT (QTc) interval. QTc immediate is the 12-lead ECG data obtained when the patient arrived at the surgical intensive care unit following LVAD placement. QTc delayed is the data from the 12-lead ECG obtained weeks later.

Overall, global EP remodeling was seen as indicated by a QTc decrease from an average of 509 ms to 453 ms (p<0.01). In the ischemic and non-ischemic cardiomyopathy groups, a similar 50 ms decrease from a QTc immediate of 499 to a QTc delayed of 450 was seen. The normal QTc value is 427 ms. The 50-60 ms decrease represents about a 60-70% excursion towards normal. In the presence of bundle branch block the same striking decrease in QTc from 530 to 460 ms (p<0.01) was seen. The presence of anti-arrhythmics enhanced the decrease, from 577 to 472, due to both EP remodeling and the discontinuation of anti-arrhythmics.

Various researchers have tried to elucidate the gene expression changes contributing to these striking functional changes. Bartling reported post-LVAD reduction in ANP at the mRNA level. Apoptotic-associated proteins are elevated in the failing heart, and several of these proteins migrate towards normal in LVAD-supported hearts.

Interestingly, calcium-handling proteins have not exhibited consistent changes pre- and post-LVAD placement in studies by Margulies and other investigators. No consistent trends with mRNA for calcium regulatory hormones have been seen to date. In some patients an increase in SERCA at the mRNA level is seen after LVAD support, while in others a decrease is seen occasionally. The same is true for calcium-sodium exchange. Margulies' interpretation is that the observed changes in calcium homeostasis likely reflect functional changes. These could be related to the dynamics between ion channels and the calcium regulatory hormones, or regulatory changes that govern calcium homeostasis vis a vis proteins, but do not necessarily govern the overall abundance of these proteins. More work is needed in this area.

Myocardial cytokines, particularly TNF-alpha, have shown marked decreases after LVAD placement. Importantly, cytokines can govern myocardial function, so consistent decreases in myocardial TNF-alpha, and presumably other changes in cytokines, clearly account for some of the functional improvement after LVAD placement. This finding coupled with Margulies' data suggests that some of the functional improvements are mediated via calcium homeostasis.

A novel approach to therapy is the potential use of LVADs as a bridge-to-recovery, that is, a tool to induce recovery in patients with end-stage failing hearts. Although this concept has been discussed, there is much to learn before it can be widely applied and a number of factors must be addressed. Smaller, less expensive and more easily removable devices are required. Candidate recognition must be defined. Optimization of recovery strategies, such as medications, must be defined. Reliable recovery markers must be identified, perhaps a decrease in natriuretic peptides or increases in myocardial or circulating cytokines. Weaning strategies and prevention of recurrent dysfunction must also be studied. Case reports show about a 50% recurrent dysfunction rate after device removal.