|Home | About | Journals | Submit | Contact Us | Français|
Left ventricular assist devices (LVAD) are increasingly used in the everyday clinical practice either to ‘bridge’ end-stage heart failure (HF) patients to heart transplantation or as a permanent (‘destination’) therapy 1,2. Yet, there is still significant uncertainty regarding the consequences of this intervention both at the level of the detailed myocardial biology (i.e. biological outcomes) and at the functional cardiovascular response of the patient at the organ level (i.e. clinical outcomes).
The LVAD patient population presents a series of significant advantages as far as research is concerned. First, LVAD therapy offers the ability to acquire paired human myocardial tissue at LVAD implantation and again upon LVAD removal. The ability to obtain human tissue and the possibility for its serial examination before and after any therapeutic investigational therapy combined with LVADs provides an important opportunity for in depth study of the changes in the structure and function of the diseased human heart caused by the specific investigational therapy. Second, this population represents a relatively ‘safe’ investigational platform as the hemodynamic support provided by VADs makes these patients significantly less vulnerable to any arrhythmic3 or hemodynamic adverse events potentially associated with aggressive new investigational therapies. Third, the volumes of potential study subjects for these investigations (i.e. patients who receive LVADs) are rapidly increasing; due to lack of donor organs and incremental progress in device design and durability the number of advanced HF patients with LVADs has been continuously increasing1,2. The three research advantages outlined above create an ideal setting for various new HF therapies, in order to test their potential efficacy in LVAD patients. Fourth, this population offers an opportunity to investigate the effects of the LVAD induced removal of excess mechanical load which drives the vicious cycle of myocardial remodeling and eventually leads to the clinical HF syndrome4. Increasing evidence suggests that a significant degree of improvement in myocardial structure and function can be observed after LVAD induced mechanical unloading5, to the point that some of these advanced HF patients can be eventually weaned from the mechanical support and achieve sustained myocardial recovery6,7.
These important research advantages may transform this LVAD patient population in a precious translational research ‘vehicle’ for investigating new anti-remodeling and regenerative therapies for HF. Yet, in order for these promises to be fulfilled, we must first establish the ‘baseline’ and get to better understand the fundamental impact of LVAD induced unloading on the failing human heart.
Witnessing a chronically sick, almost moribund, end-stage HF patient achieve sustained myocardial recovery post LVAD weaning is one of the most fascinating and rewarding experiences in contemporary treatment of heart disease (Figure 1). The main results of key clinical outcome studies investigating LVAD bridge-to-recovery are summarized in Table 18-20. With the exception of three recent studies from Berlin21, Harefield12 and Vancouver14, the majority of the devices utilized in the bridge-to-recovery studies have so far included 1st generation, pulsatile LVADs. As shown in Table 1, the most effective approach aiming at recovery of myocardial function reported so far is the Harefield protocol which tested mechanical unloading combined with aggressive anti-remodeling drug therapy and the beta-2 agonist clenbuterol, in non-ischemic cardiomyopathy patients 11,12,13,22. The Harefield protocol was also tested in the HARPS multicenter study23. Out of thirteen patients only one met explantation criteria with the authors attributing their inability to reproduce the recovery rates of prior Harefield protocol reports potentially to differences in the patient characteristics of the population studied or modifications of the Harefield protocol done in the HARPS study23. Reproducibility of the Harefield protocol results in larger patient cohorts and in a randomized fashion is of great importance.
Similarly, as evident from Table 1, the success of LVAD weaning and of achieving sustained myocardial recovery varied significantly across the reported studies. This may have been caused by a variety of factors such as: non- standardized heart function monitoring during LVAD support, differences in medical therapy added to LVAD therapy, variable duration of LVAD unloading, divergences in LVAD weaning criteria and diversity of the populations studied in their propensity for recovery (HF etiology, extent of pre-LVAD cardiac remodeling etc). These limitations were especially prominent in the multicenter LVAD trials focused on bridge-to-transplantation or destination therapy, which, for this reason are not included in Table 124. As we will discuss in following parts of the manuscript, the wide range of results described in Table 1 might have contributed to the observed disconnect between clinical and biological outcomes of LVAD studies.
Several studies described significant beneficial effects of LVAD unloading on specific parameters of cardiovascular function: left ventricular (LV) and left atrial geometry and function, volume and pressure unloading, systemic hemodynamics, cardiopulmonary function and exercise capacity25-33. As reviewed in detail elsewhere34 the impact of LVAD therapy on the arrhythmogenicity of the heart is controversial with data from a recent small prospective study showing significant decrease in premature ventricular contractions and ventricular couplets, but no change in the incidence of non-sustained or sustained ventricular tachycardia35. Regarding the cardiovascular functional effects of pulsatile versus continuous flow LVADs, several clinical studies that directly addressed this issue are summarized in Figure 2. It seems that pulsatile LVADs might have some advantages over continuous flow devices and this may be further translated to more favorable outcomes in terms of bridge-to-recovery21. However, this issue warrants further investigation and remains to be proven in a properly designed prospective study. Moreover, with pulsatile LVADs the device ejection is not generally coordinated with ventricular contraction and this device–heart dyssynchrony may paradoxically increase afterload. Continuous flow LVADs, are not subject to such dyssynchrony and whether this theoretical advantage translates to clinical benefits warrants further investigation. Other potential advantages of continuous flow devices include the increased pump durability1 which allows for longer recovery time if needed and the greater ability to modify the degree of unloading over time.
Parameters of cardiac remodeling that have been shown to be favorably altered – improved or normalized – during LVAD unloading are summarized in Table 2. These effects will be briefly described in this section.
Pulsatile LVAD unloading has been repeatedly shown to induce regression of cardiac myocyte hypertrophy - cell length, width and thickness 22,39. Regarding the exact mechanisms governing hypertrophy regression during pulsatile LVAD support, reviewed in detail elsewhere26, ongoing investigations have been examining the roles of several complex pathways including cyclooxygenase-2 induced Akt phosphorylation, MAPK/ Erks and Akt kinase/ GSK3β. Whether the primary stimulus for the regression of hypertrophy is directly related to mechanical unloading/ stretch or to circulating systemic factors needs to be further investigated.
Animal models of prolonged unloading of nonfailing, nonhypertrophic myocardium by means of heterotopic transplantation 40 or LVAD41 or severing the chordae tendinae of the mitral papillary muscle42 suggested that mechanical unloading could lead to cardiac myocyte atrophy. Whether this phenomenon applies exclusively to unloaded nonfailing and nonhypertrophic, or also to hypertrophic and failing myocardium is controversial 43, 44. In two human HF studies, unloading by means of pulsatile LVAD support decreased cardiac myocyte size but not to levels below the respective of normal donor cardiac myocytes45,46. In the latter study light microscopy findings complemented by ultrastructural and metabolic data did not identify any evidence suggesting cardiac myocyte atrophy or degeneration during pulsatile LVAD support 46. These data are in agreement with echocardiographic data in pulsatile LVAD patients8. However, whether prolonged mechanical unloading with the currently utilized continuous flow LVADs affects basic protein degradation pathways and/or fetal gene program overexpression that have been implicated in cardiac hypertrophy and atrophic remodeling 43,47,48 remains to be investigated.
The myocyte contractile defects observed in failing hearts were shown to be reversed after pulsatile LVAD unloading, showing improved shortening and relaxation in isolated myocytes and isolated strips of ventricular tissue 6,49. These interesting effects on contractile dysfunction can be partially explained by pulsatile LVAD studies demonstrating significant improvements in Ca2+ handling, such as faster sarcolemmal Ca2+ entry and shorter action potential durations, higher sarcoplasmic reticulum Ca2+ content, improved abundance of SERCA, decreased abundance of Na+/ Ca2+ exchanger, and beneficial changes in L-type calcium channel and ryanodine receptor function 6,7, 50, 51. The aforementioned LVAD-induced benefits in myocardial contractility have been associated with favorable changes in cytoskeletal proteins: sarcomeric and non-sarcomeric proteins and the membrane- associated integrin pathway known to play an important role in mechanotransduction by mediating stretch signals from the extracellular matrix 52-56.
Pulsatile LVAD unloading has been shown to induce improvements in beta-adrenergic receptor density, location and distribution pattern, improved contractile response to beta-adrenergic stimuli and higher adenyl cyclase activity6,7,49. In a recent investigation using iodine 123-meta iodobenzylguanidine (123I-MIBG) scintigraphy it was shown that pulsatile LVAD unloading resulted in improvements in sympathetic innervation in the failing heart accompanied by clinical, functional, and hemodynamic improvements57.
Pulsatile LVAD support has been shown to be associated with improved respiratory capacity and augmented nitric oxide dependent control of mitochondria respiration58,59. Furthermore, cardiolipin, a lipid component of the mitochondrial membrane important for ATP formation and substrate transport, has been shown to normalize after pulsatile LVAD unloading60. These changes along with post LVAD alterations in the expression of several metabolism-related genes and proteins49,51,61 require further investigation in order to elucidate their role within the broader metabolic changes occurring during cardiac remodeling62.
Markers of autophagy have been shown to be downregulated after LVAD unloading of failing hearts63. Several studies demonstrating changes compatible with reduced apoptosis during LVAD unloading were recently reviewed in detail by Soppa et al49. These favorable changes in myocyte attrition are complemented by data suggesting that pulsatile LVAD unloading reduces myocardial stress, as indicated by the reductions of the stress proteins metallothionein and heme oxygenase-164,65.
Pulsatile LVAD support was associated with changes in expression of genes involved in the regulation of vascular organization and migration66. In addition, animal data showed that mechanical unloading by means of heterotopic transplantation increased microvascular density67. In agreement with these experimental findings, a recent human study showed that microvascular density was decreased in the failing human hearts compared to normal donors and pulsatile LVAD unloading induced a significant increase in the microvascular density towards normalization46. The same study provided immunohistochemical and ultrastructural evidence of endothelial cell activation which is consistent with the observed increase in microvascular density46.
Pulsatile LVAD unloading has been associated with decreased levels of atrial and brain natriuretic peptides and tumor necrosis factor-α both in serum and in myocardial tissue 8,68,69. The changes in the levels of other key neurohormones implicated in the progression of HF syndrome appear to be more complex. Specifically, the circulating levels of epinephrine, norepinephrine, renin, angiotensin II, and arginin vasopressin have been shown to decrease during LVAD unloading70. However, as discussed below, the effects on the myocardial tissue levels of these neurohormones are not uniform.
Investigations of the effect of LVAD unloading on extracellular matrix have shown conflicting results: a few studies reported decreased fibrosis, while most other investigations found a significant increase in fibrosis 26,34,49. The explanation for the contradictory observations is not clear, with some attributing the inconsistent results to differences in the background medications or the applied methodology 26,49. This controversial issue was recently addressed using advanced image analysis techniques in whole-field digital microscopy, an approach that reduces observer bias, markedly increases the amount of myocardial tissue analyzed and permits comprehensive endocardium-to-epicardium evaluation46. It was found that myocardial tissue from HF patients undergoing LVAD implantation, compared to normal myocardium, had increased interstitial and total fibrosis46. The interstitial and total collagen content further increased after pulsatile LVAD unloading in these patients (46). Recent findings regarding the effects of pulsatile LVAD unloading on the myocardial tissue levels of neurohormones of the renin-angiotensin-aldosterone axis and matrix metalloproteinases support the above results71,72. However, whether the observed increase in fibrosis is a manifestation of further progression of this aspect of cardiac remodeling that pulsatile LVAD unloading failed to reverse or a direct result of pulsatile LVAD unloading actively inducing an increase in fibrosis warrants further investigation.
Studies in LVAD patients investigated mRNA, microRNA and protein expression profiling73-77. Hopefully, future investigations using these technologies will consistently include in their study design the collection of functional myocardial recovery data77 and thus increase their potential to provide mechanistic insights.
Any attempt to associate in a systematic and logical way the key LVAD-induced biological effects with their expected corresponding clinical outcomes would be challenging. It could be argued that the reported beneficial LVAD-induced biological outcomes (Table 2) should have more consistently led to better clinical outcomes in terms of functional myocardial recovery (Table 1). The anticipated ‘sequential pattern’ of biology findings defining clinical functional response does not appear to always be clear or consistent (Tables 1 and and2).2). Therefore, one wonders: why do we observe these discrepancies?
One possible reason for the observed disconnect between the clinical and biological outcomes is the attempt to correlate findings across separate clinical and biological studies, rather than focusing this effort on investigations using the same structured and well controlled approach. Specifically, as we have previously reviewed in detail13, in most of the reported LVAD tissue/ biological outcomes studies no functional myocardial recovery data were collected. And vice versa, most of the clinical outcomes / bridge to recovery studies (Table 1) were lacking a comprehensive structural or molecular investigational arm13. As a consequence, we cannot distinguish between structural, cellular and molecular changes that occur in all LVAD patients regardless of possible induced myocardial recovery versus changes that occur exclusively in patients that LVAD unloading induced myocardial functional recovery (figure 3 – target #1). These latter biological changes unique to LVAD patients that achieved functional recovery might help us identify mechanisms of reverse remodeling that lead to myocardial recovery. Examination of tissue from both patients with evidence of various degrees of LVAD-induced myocardial functional recovery (i.e. Responders) and from LVAD patients without functional myocardial improvement (i.e. Non-Responders) is critical. This type of studies8, 50, 51, 53, 54, 56, 57, 61, 77, 79-81 become the springboard for further in depth investigational steps through future animal and human studies that will determine causality and provide mechanistic insights.
In fact, the ability offered by LVAD studies to correlate human tissue to functional data is something rare in clinical medicine. This type of tight association between structure and function is achievable in animal models; however it is very unusual to achieve this level of understanding in human investigational models. From that perspective, the LVAD patient population offers an important opportunity for performing in depth structure-function investigations, which will hopefully lead to clarification of some of the observed discrepancies between LVAD clinical and biological outcomes. The current absence of such in depth structure-function investigations makes any attempt to connect the biological and clinical outcomes very difficult. In essence, at least a degree of the observed disconnect between clinical and biological outcomes is a result of this missing data.
The aforementioned disconnect between clinical and biological outcomes may also be a consequence of a series of major limitations that are confounding many of the reported studies. As analyzed in the following paragraphs, these specific limitations may have led to an ‘inaccurate description’ and thus poor understanding of both the clinical and biological effects of LVAD unloading. Thus, trying to understand the potential associations or connections between the reported clinical and biological outcomes might be flawed by several problems in the design of these studies. In essence, we may be trying to connect two locations on the map (i.e. biological and clinical outcomes), however the coordinates of these two locations have not been well defined.
Various medications which are known to affect the function and structure of the failing human heart (beta-blockers, renin-angiotensin axis inhibitors, aldosterone antagonists) have been routinely used in previous studies in patients with LVADs but no standardization or randomization of their use was attempted. In most LVAD tissue studies no information regarding concurrent anti-remodeling drug therapy was reported34. In fact, in many LVAD patients systemic blood pressure increases and as a consequence these patients are often being treated with high doses of these medications. This is an important confounder because in the reported so far studies the drug-induced effects on cardiac remodeling cannot be separated from the effects of mechanical unloading alone.
The patient populations studied so far differ in their potential for reversal of cardiac remodeling; factors such as specific HF etiology and duration of HF symptoms have been reported to play significant role in this propensity for recovery10,12,24. Both these factors varied significantly among the reported clinical and biological studies making comparisons or associations between their findings problematic.
Clinical and experimental studies demonstrated that duration of mechanical unloading significantly affects the changes on the remodeling of the failing heart8,25,44,53,82,83. Therefore, it might be misleading to either claim or negate associations between ‘biological’ and ‘clinical’ outcomes studies that had different durations of LVAD unloading. Even within a single study, more often than not, LVAD support duration varied considerably between patients.
The great majority of biological outcomes reported in the literature were derived from studies involving pulsatile LVADs. However, due to mainly engineering reasons newer 2nd generation, non-pulsatile, continuous flow LVADs are now being used almost exclusively for long term support. Compared to pulsatile LVADs these newer devices produce a qualitatively different type of unloading (Figure 2). Given that clinical outcomes related to myocardial recovery originating from continuous flow LVADs have now started to be reported (12,14,21) it is necessary for the corresponding biological outcomes to be ‘updated’ as well. Pulsatile LVAD era biological outcomes and their presumed associations or dissociations with clinical outcomes cannot be taken for granted in the continuous flow era.
Relative values and ratios should be used with caution when assessing biological effects of LVADs. For example, hypertrophy regression is universally seen after LVAD unloading, and the decrease in cardiomyocyte size is affecting other morphometric measurements that are based on the relative quantification of other myocardial components (e.g. fibrosis). When using ratios, no significant changes of the structures of interest can be detected when both the nominator and the reference parameter (denominator) show changes of the same size and direction. This problem, called in stereology the ‘reference trap’ has been previously reviewed in detail by Baba26. One proposed approach to circumvent this problem would be to estimate the heart volume using magnetic resonance or computed tomography imaging and extrapolate from the volume fraction of the measured parameter the absolute volume26.
Most of the above described limitations of the biological outcomes also apply to LVAD clinical outcomes studies as well: i) concurrent anti-remodeling medical therapy, ii) variable propensity for reversal of cardiac remodeling, iii) variable duration of LVAD unloading, iv) differences between pulsatile and non pulsatile unloading.
In addition to those factors the reported clinical outcomes are also limited by:
There was no standardization across reported studies regarding protocols to serially and reliably monitor the functional status of the heart during mechanical unloading. This is an important limitation of many studies (Table 1). Preferably these protocols should include studies done (a) with full LVAD support and (b) with prolonged minimal LVAD support (the so called ‘turndown’ or ‘off pump’ studies) to allow for assessment of the native cardiac function under renewed pressure and volume load (85-89). The complexity of this issue is reviewed in a separate manuscript in this Mechanical Circulatory Support series.
There is neither robust clinical evidence nor expert consensus that would delineate criteria of likely sustained recovery post LVAD explantation. In the published bridge-to-recovery studies (Table 1) the utilized LVAD explantation criteria (echocardiographic, hemodynamic, cardiopulmonary/ exercise capacity) varied significantly. As a result, no consistent conclusions can be made regarding outcomes such as frequency or sustainability of myocardial recovery after LVAD implantation.
While grappling to understand the potential associations or disconnects between the reported LVAD-induced clinical and biological outcomes, we should take into account that, despite the enormous progress in the understanding of HF pathophysiology during the last two decades, important pieces of information are still missing. As reviewed in detail by Mann and Bristow 78, our current hemodynamic, cardiorenal and neurohormonal model systems are necessary but not sufficient to explain all aspects of the HF syndrome. Most importantly, they fail to adequately explain forward disease progression78. Similarly, the reverse process of myocardial improvement or recovery resulting from the use of currently approved medical or device HF therapies is also incompletely understood. The issue is getting maybe more complicated by the fact that reversal of key features of cardiac remodeling, such as hypertrophy regression, is governed by distinct pathways, different from those implicated in forward remodeling and HF syndrome progression 90).
Furthermore, ‘reverse cardiac remodeling’ and sustained ‘clinical myocardial recovery’ are not necessarily synonyms 91, 92; as shown by several published LVAD studies the partial or sometimes near complete reversal of the HF phenotype at the structural, cellular or molecular level (i.e. ‘reverse cardiac remodeling’) is not always followed by a similar degree of sustained clinical ‘myocardial recovery’ at the organ level 34,91,92. Future studies need to specifically focus on advancing our understanding of these phenomena 34, 93. The realization of this acute need highlights the unique opportunity of current investigations of LVAD-induced unloading in elucidating the incompletely understood relationship between reverse remodeling and myocardial recovery. Importantly, this process can identify potential new therapeutic targets in HF. Given that a large part of prior research, both in heart failure and in cardiovascular disease in general, has focused on predicting adverse outcomes, we maybe now also need to focus on determining methods to better understand, predict and enhance myocardial recovery. In conclusion, it should probably have been anticipated that attempts at systematically correlating specific LVAD-induced structural or molecular alterations with specific favorable post LVAD functional myocardial responses would lead to an inevitable degree of disconnect between these ‘biological’ and ‘clinical’ outcomes insofar as the specific biological signature of myocardial recovery of the failing heart is still not very well defined34,91,92.
It is obvious from the analysis above that many important issues remain to be elucidated.
The impact of the etiology of HF on the potential for myocardial recovery is not well understood (Figure 3 - target #2). Direct comparisons between ischemic and non-ischemic patients were performed in only a few LVAD studies 8, 24, 46, 60, 66, 73. The likely candidates for reverse remodeling induced by LVAD unloading are usually patients with non-ischemic cardiomyopathy of different etiologies: idiopathic, hypertensive, peripartum, familial, alcoholic etc. However, ischemic cardiomyopathy patients who have suffered myocardial infarction and have large areas of non-infarcted myocardium that ‘remodeled’ over the years could also be considered candidates13. This latter concept deserves further investigation and could combine the excision of scarred myocardium, using LV reconstruction techniques (e.g. Dor operation), with LVAD unloading13. It can be argued that with this approach the initial insult that triggered the cascade of cardiac remodeling progression – i.e. the post-myocardial infarction scar, has been eliminated13. In contrast, in most non-ischemic cardiomyopathy cases, the initial insult that caused progressive ventricular remodeling and HF often remains undetermined, most likely persists despite an initially successful reversal of the process by mechanical unloading, and might recur and cause further progression of HF after the termination of LVAD support 13. This might explain why long-term freedom from recurrent HF in the largest bridge to recovery series in non-ischemic cardiomyopathy patients was 74% and 66% at 3 and 5 years, respectively9.
The importance of the duration of HF, on the prospect of cardiac reverse remodeling also deserves further study. Data from two series of LVAD patients that were successfully bridged to sustained recovery have identified ‘duration of HF history’ (i.e. time from HF symptoms onset) as an important predictor of favorable response10,12. In terms of cardiac remodeling course, the time from the initial insult that triggered the HF syndrome, rather than the time from symptom onset, would be an even more meaningful target (Figure 3 – target #3). However, we need to acknowledge that this may be a target too hard to identify. The initial insult can often be determined in ischemic cardiomyopathy patients, but this may be hard in non-ischemic patients. Even in ischemic cardiomyopathy, other factors such as ischemia induced by non-culprit lesions, repetitive stunning, etc. add to the complexity. Insofar as the ‘HF history duration’ can be viewed as a surrogate of potential irreversibility of chronic remodeling, it may be argued that a more direct research target would be the identification of a degree of pre-LVAD structural or molecular remodeling beyond which there is ‘no return’. Indeed, Bruckner et al have reported that patients with worse hypertrophy and higher degree of fibrosis at baseline (i.e. the time of LVAD implantation) were much less likely to show recovery of LV systolic function during LVAD unloading 94. More research needs to be done to determine what extent of pre-LVAD myocardial remodeling changes preclude unloading-induced reversibility and thus provide useful guidance for LVAD bridge-to-recovery patient selection (Figure 3- target #4).
Another important issue is recognition of the specific type of mechanical unloading that best promotes reverse remodeling (Figure 3 – target #5). Various LVADs have been used in the experimental and clinical arenas during the last half century: pulsatile, non pulsatile / continuous flow, and counterpulsatile. Due to favorable engineering characteristics that were translated to better morbidity and mortality outcomes, the clinical field has recently shifted from pulsatile to continuous flow LVADs. The key known effects of pulsatile and continuous flow LVADs specifically on cardiovascular functional parameters were summarized in Figure 2. Whether these devices also have different effects on the biological outcomes we definitely need to learn more29, 36, 95-97. Consequently, whether the prospects of LVAD induced reverse remodeling are better served by pulsatile9,11, non-pulsatile12,14,21, or counterpulsation devices98,99, and by full or partial unloading100, is unknown. Future studies should target specific ventricular assist device properties that best promote reverse remodeling.
The potential impact of the following important issues also needs to be clarified: the concept of targeted adjuvant therapies/ disease-modifying medications (introduced by Sir Magdi H. Yacoub)22, the optimal duration of mechanical unloading8,25,44,49,82,83 and the development of advanced protocols to monitor the unloaded myocardium during LVAD support (Figure 3 - targets #6, #7, #8). These latter protocols could include hemodynamic evaluations, exercise testing protocols, conventional imaging techniques (echocardiography, nuclear imaging, computed tomography) or molecular imaging. As pointed out in the prior section they will need to address the challenging issue of testing the cardiac performance under both decreased and increased loading conditions. These monitoring protocols should also carefully evaluate the short-term and long-term impact of LVAD support on the right ventricle (RV). Some investigations have shown evidence of improved right ventricular structure and function after LVAD support8,96,101. This can be attributed both to the normalization of the neurohormonal milieu and the reduction in left ventricular filling pressures resulting in decreased RV afterload. However, Klotz et al found that biventricular VAD support resulted in significant reverse structural and functional remodeling of both the RV and the LV, while RV reverse remodelling was not found during LVAD support alone102. The authors concluded that the lesser degree of volume unloading provided to the RV during LVAD support may not be sufficient to result in significant reverse structural and functional remodeling of the RV. Along the same lines a recently published study showed that pre LVAD RV dysfunction seen on intensive medical therapy that included inotropes and diuretics persisted after 3 months of LVAD unloading103. More research is warranted to elucidate the impact of chronic LVAD support on the right ventricle.
Both the bridge-to-transplant and bridge-to-recovery investigational settings offer important research advantages and by addressing most of the research targets outlined in Figure 3 can help advance the field. It is absolutely necessary that both types of investigations would include in their study design a comprehensive heart function monitoring protocol (such as serial echocardiographic studies) in order to allow for structure – function correlation.
The bridge-to-transplant investigational setting offers two important advantages. First, it offers access to paired myocardial tissue specimens of large quantity, and this from both recovery responders and non-responders (given that both functional responders and non-responders will be transplanted per the clinical bridge-to-transplant protocol). On the contrary, bridge-to-recovery studies offer scarce, often minute amounts of human myocardial tissue at the post LVAD time point, and this only from recovery responders upon LVAD explantation. Non-responders either remain on LVAD as destination therapy and they do not offer paired myocardial tissue, or enter the transplant list after the end of the bridge-to-recovery study. Access to adequate quantities of paired human tissue from both myocardial recovery responders and non-responders is of great importance, as it allows investigational approaches that can examine differences between LVAD-induced biological changes that are associated with recovery and changes that occur in LVAD patients regardless of the functional response (Figure 3 – target #1). The second advantage of adequately powered, large-scale, bridge-to-transplant studies is the opportunity to study the impact of the duration of LVAD unloading on cardiac remodeling. The patients are transplanted at different times since LVAD implant and the tissue specimens can thus be grouped by duration of unloading (Figure 3 – target #8).
Bridge-to-recovery studies also present unique advantages. They can lead to the identification of ‘clinical’ and ‘biological’ markers of sustained (post LVAD explantation) myocardial functional recovery. The identification of markers of myocardial recovery will help to establish reliable LVAD explantation criteria (Figure 3- target #9). Furthermore, myocardial recovery post LVAD unloading is not an “all or nothing” phenomenon. Prior LVAD studies have shown that while only a relatively small proportion of end-stage HF patients had complete normalization of heart function, a much larger proportion showed significant improvement (‘partial recovery’) to a degree similar to that of stable HF outpatients8. LVAD explantation for partial recovery warrants further study in future bridge-to-recovery investigations. Bridge-to-recovery studies should also test surgical techniques of LVAD explantation that would minimize iatrogenic myocardial damage and enhance the sustainability of the achieved myocardial recovery.
Increasing clinical use of LVADs presents a key opportunity for in-depth investigations of the biology of the failing human heart. Through an effort to better define and connect the biological and clinical outcomes in this unique patient population we may eventually identify new therapeutic strategies that augment myocardial recovery and regeneration.
The authors are indebted to Sir Magdi H. Yacoub, Dale G. Renlund and Edward M. Gilbert for their enormous clinical and academic support over the years. We apologize to all of our colleagues whose original work was not cited due to imposed space constrains.
Funding Sources: This work was funded by grants from: National Heart, Lung, and Blood Institute (NHLBI), National Institute of Allergy and Infectious Diseases (NIAID), Juvenile Diabetes Research Foundation, HA and Edna Benning Foundation, National Center for Research Resources Public Health Services research grant UL1-RR025764, and the Department of Defense (to Dr Li); European Union – Research Executive Agency (REA)/Seventh Framework Programme (FP7) – Marie Curie - #276776 (to Dr Drakos); NIH National Center for Research Resources (NCRR) grant that supports the Center Clinical Translational Sciences (CCTS) UL1-RR025764 and C06-RR11234 (to Drs Drakos & Kfoury); Deseret Foundation #00571 (to Drs Drakos & Kfoury); American Heart Association #09CRP2050127 (to Dr Stehlik); and NIH 5R01HL089592-02 (to Dr Selzman).
Conflict of Interest Disclosures: None