Clinical, Molecular, and Genomic Changes in Response … · Clinical, Molecular, and Genomic Changes ... PHD,¶# Mark Slaughter, ... the 1,092 primary LVAD patients

Journal of the American College of Cardiology Vol. 57, No. 6, 2011© 2011 by the American College of Cardiology Foundation ISSN 0735-1097/$36.00

STATE-OF-THE-ART PAPER

Clinical, Molecular, and Genomic Changesin Response to a Left Ventricular Assist Device

Jennifer L. Hall, PHD,*†‡ David R. Fermin, MD,§ Emma J. Birks, MD,�Paul J. R. Barton, PHD,¶# Mark Slaughter, MD,** Peter Eckman, MD,* Hideo A. Baba, MD,††Jeremias Wohlschlaeger, MD,†† Leslie W. Miller, MD‡‡

Minneapolis and Rochester, Minnesota; Louisville, Kentucky; London, United Kingdom;Essen, Germany; and Tampa, Florida

The use of left ventricular assist devices in treating patients with end-stage heart failure has increased signifi-cantly in recent years, both as a bridge to transplantation and as destination therapy in those who are ineligiblefor cardiac transplantation. This increase is based largely on the results of several recently completed clinicaltrials with the new second-generation continuous-flow devices that showed significant improvements in survival,functional capacity, and quality of life. Additional information on the use of the first- and second-generation leftventricular assist devices has come from a recently released report spanning the years 2006 to 2009, from theInteragency Registry for Mechanically Assisted Circulatory Support, a National Heart, Lung, and Blood Institute–sponsored collaboration between the U.S. Food and Drug Administration, the Centers for Medicare and MedicaidServices, and the scientific community. The authors review the latest clinical trials and data from the registry,with tight integration of the landmark molecular, cellular, and genomic research that accompanies the reverseremodeling of the human heart in response to a left ventricular assist device and functional recovery that hasbeen reported in a subset of these patients. (J Am Coll Cardiol 2011;57:641–52) © 2011 by the AmericanCollege of Cardiology Foundation

Published by Elsevier Inc. doi:10.1016/j.jacc.2010.11.010

Clinical Perspective

Currently, the average mortality from the time of diagnosisof heart failure, regardless of whether there is preserved orreduced systolic function, is �60% at 5 years, and this ismuch higher when patients are in the most advanced stages(1). Patients typically become progressively less responsiveto medical therapy and are left with limited options formeaningful improvement in survival or functional capacityand quality of life (2).

Cardiac transplantation is currently the most establishedtreatment for refractory heart failure (1). However, thisoption is available to fewer than 2,300 patients per year,

From the *Division of Cardiology, Department of Medicine, University of Minne-sota, Minneapolis, Minnesota; †Lillehei Heart Institute, University of Minnesota,Minneapolis, Minnesota; ‡Developmental Biology Center, University of Minnesota,Minneapolis, Minnesota; §Division of Cardiovascular Diseases, Mayo Clinic, Roch-ester, Minnesota; �Division of Cardiology, Thoracic and Cardiovascular Surgery,University of Louisville, Louisville, Kentucky; ¶Heart Science Centre, National Heartand Lung Institute, Imperial College London, National Institute for Health ResearchCardiovascular Biomedical Research Unit, Royal Brompton and Harefield NationalHealth Service Foundation Trust, London, United Kingdom; #National Heart andLung Institute, Imperial College London, London, United Kingdom; **Division ofThoracic and Cardiovascular Surgery, University of Louisville, Louisville, Kentucky;††Institute of Pathology, University of Duisburg-Essen, Essen, Germany; and the
‡‡Cardiovascular Clinical and Research Integrated Strategic Program, University ofSouthern Florida, Tampa, Florida. This work was funded in part by the National
with a relatively fixed organ supply (1) and no substantiveincrease anticipated in the near future. The prevalence ofheart failure increases with age and affects an estimated 10%of people over 70 years of age, which is the age that manytransplantation programs use as a cutoff for acceptingcandidates. Importantly, the population over age 65 is expectedto double in the next 20 years. As a result, the use of leftventricular assist devices (LVADs) in treating patients withend-stage heart failure will increase significantly in thecoming years, both as a bridge to transplantation and asdestination therapy in those who are ineligible for cardiactransplantation (3).

Institute for Health Research Cardiovascular Biomedical Research Unit at the RoyalBrompton and Harefield National Health Service Foundation Trust and Imperial College(Dr. Barton), grants 0515584Z and 0655582Z from the American Heart Association,grant NIH-RO1-HL065462-04 from the National Institutes of Health, and the LilleheiScholar Program (laboratory of Dr. Hall). Dr. Hall consulted for Catholic Health CareWest from January 2008 to September 2010. Dr. Birks has received research support andhonoraria from Thoratec. Dr. Slaughter has received research grant support fromHeartware Inc. and Thoractec. Dr. Eckman has received consulting, honoraria, andresearch support from Thoratec. Dr. Baba has received speaker fees from Medtronic. Dr.Miller has received research support and speaking honoraria from Heartware Inc. andThoratec. All other authors have reported that they have no relationships to disclose.W. H. Wilson Tang, MD, served as Guest Editor for this paper.

Manuscript received August 24, 2010; revised manuscript received October 25,2010, accepted November 8, 2010.

fT

pCt(ct(cptLssrpatNteoHp

tvi

csmisiraaaiagsortt

FiL

TaecWilc

ffiptmffbta(oi(

642 Hall et al. JACC Vol. 57, No. 6, 2011LVAD Review: Clinical, Molecular, and Genomic February 8, 2011:641–52

In 2009, the Centers for Medi-care and Medicaid Services set amandate that patients receivingventricular assist devices be en-tered into a national database,the Interagency Registry for Me-chanically Assisted CirculatorySupport (INTERMACS) (3).INTERMACS is a National Heart,Lung, and Blood Institute fundedcollaboration between the Na-tional Heart, Lung, and BloodInstitute, the U.S. Food and DrugAdministration, the Centers forMedicare and Medicaid Services,and the mechanical circulatorysupport scientific community (3).This registry has shown that pa-tients with LVADs alone haveoverall actuarial survival rates of74% at 1 year and 55% at 2 years(3). The primary cause of death inthe 1,092 primary LVAD patientsin INTERMACS from June 2006through March 2009 was cardiacfailure, including right ventricular

ailure and ventricular tachycardia or ventricular fibrillation (3).he secondary cause of death was infection (3).At the inception of INTERMACS, the HeartMate XVE

ulsatile-flow pump (Thoratec Corporation, Pleasanton,alifornia) was the only approved device for destination

herapy in the U.S. In April 2008, the HeartMate IIThoratec Corporation), a continuous-flow LVAD, re-eived Food and Drug Administration approval as a bridgeo transplantation based on the results of the clinical trial4). Thus, the registry began tracking the evolution ofontinuous-flow technology in addition to the originalulsatile flow pumps (3). A recent randomized trial showedhat patients receiving treatment with continuous-flowVADs (HeartMate II) as destination therapy exhibited

ignificant improvement in the probability of stroke-freeurvival and device failure at 2 years compared with thoseeceiving pulsatile LVADs (5). Both continuous-flow andulsatile devices significantly improved the quality of lifend functional capacity (Thoratec HeartMate II Left Ven-ricular Assist System [LVAS] for Destination Therapy;CT00121485) (5). However, the survival outcomes with

he pulsatile device were nearly identical to those in patientsnrolled nearly 10 years earlier in the Randomized Evaluationf Mechanical Assistance for the Treatment of Congestiveeart Failure for destination therapy patients, indicating little

rogress with that pump (6).Clinical heart failure may present with changes referred

o as remodeling that include eccentric dilation of theentricular chamber, a reduction in contractile function, and

Abbreviationsand Acronyms

ACE � angiotensin-converting enzyme

AGAT � arginine:glycineamidinotransferase

ANP � atrial natriureticpeptide

BNP � B-type natriureticpeptide

Epac2 � exchange proteindirectly activated by cyclicadenosine monophosphate

IGF � insulin growth factor

INTERMACS � InteragencyRegistry for MechanicallyAssisted CirculatorySupport

LVAD � left ventricularassist device

miRNA � micro–ribonucleicacid

mRNA � messengerribonucleic acid

RNA � ribonucleic acid

ncreases in cardiac filling pressures and wall stress. These

linical phenotypic changes of heart failure are paralleled byignificant histologic, cellular, and molecular changes inost structural and functional components of the myocyte,

ncluding alterations in myocyte geometry and size, progres-ive interstitial fibrosis, up-regulation of cytokines andnflammation, and changes in myosin isoforms, as well aseductions in myocardial energetics, beta-receptor density,nd calcium-handling proteins; this is otherwise describeds recapitulation of the fetal gene program. These cellularnd molecular changes may arise from a response to changesn volume, stretch, pressure, and inflammation that occur asresult of diet, additional environmental modifiers, and/or

enetic mutations. A progressive body of evidence hashown a trend toward normalization of these perturbations,r reverse remodeling, with LVAD support. However,esponse to an LVAD is unique to each patient, suggestinghat underlying factors are critical in the response toherapy.

unctional Improvements in Myocytesn Response to LVAD Unloading,inked to Gene Expression Changes

ranslational research from human heart tissue before andfter support with an LVAD has shown that myocytes innd-stage failing hearts have the capacity to remodel. Thesehanges are accompanied by alterations in gene expression.

e summarize findings in the field that have led tomportant steps forward in our understanding of the mo-ecular changes and contractile properties of cardiac myo-ytes before and after unloading with an LVAD.

There is evidence of early mechanical improvements inailing hearts treated with LVADs (7–15). These earlyndings were important in showing that the contractileerformance of myocytes was partially recovered in responseo therapy with an LVAD. Myocyte contractile perfor-ance was significantly greater in isolated myocytes from

ailing human hearts after LVAD support compared withailing human hearts without LVAD support in patientsridged to transplantation (Table 1) (8). Improvements inhe magnitude of shortening were seen in isolated myocytesfter LVAD support in response to beta-adrenergic agonistsTable 2) (8). Similar to work in isolated myocytes, devel-ped tension in response to beta-adrenergic stimulation insolated trabecular muscle preparations was also improved13) (Fig. 1). Basal relaxation was also improved in myo-

Contractile Characteristics Are Improvedin Myocytes After Support With an LVADTable 1 Contractile Characteristics Are Improvedin Myocytes After Support With an LVAD

Contractile CharacteristicHF Monocytes

(n � 57)HF LVAD Monocytes

(n � 35)

Magnitude of shortening (% RCL) 6.9 � 0.5 9.6 � 0.7*

Time to peak contraction (s) 0.75 � 0.04 0.37 � 0.01

Time to 50% relaxation (s) 1.45 � 0.11 0.55 � 0.02

*Significant difference between the HF and the HF LVAD monocytes at p � 0.05. Adapted from
Dipla et al. (8).
HF � heart failure; LVAD � left ventricular assist device; RCL � resting cell length.

http://www.clinicaltrials.gov/ct2/show/NCT00121485

soirhiaHaBtb

swnfibcpticrBcuhc

. (8).sist dev

643JACC Vol. 57, No. 6, 2011 Hall et al.February 8, 2011:641–52 LVAD Review: Clinical, Molecular, and Genomic

cytes after LVAD support (Table 2) (8). However, relax-ation in response to isoproterenol was not improved(Table 2) (8). This was confirmed in isolated trabecularpreparations in a separate study (Fig. 1) (13). Figure 1 alsohows the significant magnitude in maximal rates of devel-ped rise and fall in tension in response to the beta agonistsoproterenol. Significant improvements are seen in bothise and fall in tension with LVADs compared with failingearts. In sum, these findings suggest that LVAD supportmproves developed tension and relaxation rates at rest andlso improves developed tension in response to beta agonists.owever, LVAD support did not improve relaxation rates

fter stimulation with beta agonists.eta-adrenergic signaling. Improvements in developed

ension with LVADs have been shown to be accompaniedy increased beta-adrenergic receptor density (13,16). Early

Isoproterenol Response on Contractile ParametTable 2 Isoproterenol Response on Contrac

Contractile Parameter Disease Group

Magnitude of shortening (% RCL) HF

HF-VAD

Time to 50% relaxation (s) HF

HF-VAD

*Significant difference (p � 0.0125) between baseline and isoproteremyocytes on the delta scores (t test for independent samples). ‡Signindependent samples). Reproduced, with permission, from Dipla et al

HF � heart failure; RCL � resting cell length; VAD � ventricular as

Figure 1 Response of Muscles to 1 �mol/l Isoproterenol

Muscles from nonfailing hearts (NF), failing hearts without left ventricular assist d(TPT); (B) time to half relaxation (THR); (C) maximal rate of tension rise (�dT/dt);contraction (mean � SEM). This study shows no change in TPT (A) or THR (B) bein �dT/dt (C) and �dT/dt (D) were seen in the failing hearts that were improvedAdapted from Ogletree-Hughes et al. (13).

tudies showed that beta-adrenergic density was improvedith LVAD therapy in 17 patients compared with 8onfailing patients (13). A recent study confirmed thesendings (15) (Fig. 2). Whether this improvement ineta-adrenergic density translates to intermediate bio-hemical improvements in the beta-adrenergic signalingathway that are associated with improved developedension are not known. It is possible that other moleculesn the signaling pathway may still be “desensitized” thatontrol the “brake” on this important pathway thategulates myocyte contractility and remodeling.eta-adrenergic signaling in recovery patients. More re-

ent work assessed beta-adrenergic signaling pathways in annbiased analysis using a gene chip platform in 6 pairedeart samples from patients who achieved sufficient myo-ardial recovery to allow explantation of the devices (17,18).

arameters

Baseline Isoproterenol Delta Score

6.43 � 0.73 8.92 � 0.78* 2.49 � 0.58†

7.63 � 0.88 12.56 � 1.01*‡ 4.93 � 1.42

1.28 � 0.19 0.95 � 0.16 0.32 � 0.15

0.70 � 0.07‡ 0.53 � 0.03 0.17 � 0.08

red t test). †Significant difference (p � 0.05) between HF and HF-VADifference (p � 0.0125) between HF and HF-VAD myocytes (t test for

ice.

LVAD) support, and failing hearts with LVAD support. (A) Time to peak tensionaximal rate of tension fall (�dT/dt). Data are shown as change from baselineNF, failing hearts, and failing hearts with LVAD support. Significant decreasesVAD support. n � number of muscles/number of hearts. *p � 0.05 versus NF.

erstile P

nol (paiificant d

evice ((D) m

tweenwith L

mTANrpnN

ALCtstwptmcrdhws


A novel combination therapy consisting of an LVADcombined with pharmacologic therapy including the selec-tive beta-2 agonist clenbuterol has shown promise in restor-ing ventricular function in patients with heart failure (17).The aim of this study was to identify common genes andsignaling pathways whose expression was associated withreversal of heart failure and restoration of ventricular func-tion. Microarray analysis was performed on 6 paired humanheart samples harvested at the time of LVAD implantationand at the time of LVAD explantation for recovery ofventricular function. Follow-up data showed that the im-provements in ventricular function were maintained formore than 5 years after explantation. This dataset representsthe first description of signaling pathways associated withthe functional recovery of end-stage human heart failureand the identification of new targets in the human heart thatare modified by this combination therapy. Analyzing thegene expression profiles from these patients at the time ofLVAD explantation compared with LVAD implantation(heart failure) resulted in the identification of pathwayssignificantly enriched with genes whose expression wasstatistically different between heart failure and recovery.Instead of a gene-by-gene approach, this study used anetwork analysis, with the goal of identifying pathways thatare altered in the process of recovery from heart failure (18).Put simply, gene lists are imported into the Ingenuitydatabase (Ingenuity Systems, Inc., Redwood City, Califor-nia), and the program subsequently overlays these genes intoseveral described networks to identify the best overall fit. Asseen in Figure 3, significant changes in genes in the

Figure 2Density of Beta-Adrenergic Receptorsin LV and RV Myocardial Samples of NF,Medically Managed, and LVAD-Supported Hearts

Data are reported as mean � SD of maximal binding (Bmax) values determinedfrom Scatchard transformation of saturated binding data. HF � heart failure(medically managed); LV � left ventricular; LVAD � left ventricular assistdevice; NF � nonfailing; RV � right ventricular. *p � 0.01 versus nonfailingand LVAD-supported hearts. Adapted from Klotz et al. (15).

beta-adrenergic signaling pathway in recovered hearts in-

cluded Rap guanine nucleotide exchange factor 4 (exchangeprotein directly activated by cyclic adenosine monophos-phate 2 [Epac2], down-regulated at explantation 2.2-fold),protein kinase, regulatory subunit, type I alpha (down-regulated at explantation 1.5-fold), phosphodiesterase 1A(down-regulated at explantation 1.5-fold), phosphodiester-ase 3B (down-regulated 1.5-fold), and calcineurin A (up-regulated 1.4-fold) (18). The down-regulation in Epac2 wasunique to the recovered group and confirmed in a replica-tion cohort, suggesting that this target may be important forrecovery. (This target is discussed again in the followingsection on calcium handling.) It is important to keep inmind that expression changes in messenger ribonucleic acid(mRNA) need to translate to functional changes in bio-chemical signaling to be causally related to functionalrecovery. Studies of this nature will be critical for movingthe field forward.Calcium handling. Impaired myocardial relaxation is aphysiologic characteristic of heart failure. The reportedimprovements in basal relaxation rate with LVADs have notbeen definitively linked to changes in calcium handling.Elegant functional experiments have been carried out toassess calcium transients and expression of calcium handlingproteins (7,19). In a study by Chaudhary et al. (7), LVAD-supported failing myocytes exhibited faster decay in bothearly and late stages of the [Ca2�]i transient compared with

yocytes from failing hearts not supported with LVADs.hese same findings were seen in patients who recovered.n increase in ribonucleic acid (RNA) expression of thea�-Ca2� exchanger was identified in this study from

ecovered patients (19). In the study with nonrecoveredatients, changes were not measured at the RNA level, ando changes were detected at the protein level of thea�-Ca2� exchanger (7), thus not offering an explanation

for the faster decay times in the [Ca2�]i transients (20,21).summary of calcium handling changes in response to an

VAD is shown in Figure 4.alcium handling in recovery patients. Expanding on

hese findings, recent work by Terracciano et al. (26)howed the largest improvements in action potential dura-ion and sarcoplasmic reticulum calcium content in patientsho achieved clinical recovery in response to LVADs andharmacological therapy (Fig. 5) (Harefield Recovery Pro-ocol Study). These findings help distinguish contractile andolecular differences in myocytes from hearts that achieve

linical recovery compared with those that achieve partialecovery (26). Another recent study by Ogletree et al. (27)emonstrated that the benefits of unloading on calciumandling and contractility are time dependent. Patientsith shorter durations of support (�115 days) achieved

ignificant increases in sarco/endoplasmic reticulum Ca2�-adenosine triphosphatase expression levels and correspond-ing increases in developed tension (27). These improve-ments, however, reverted back to failing levels in thepatients with longer durations of support (Fig. 6) (27). The
reason these changes are not persistent is unknown, but it

gstcad

pEanpdmpCMp


could be an indication of adverse remodeling signals withprolonged unloading.

A recent phase II clinical trial, the CUPID (CalciumUpregulation by Percutaneous Administration of Gene Ther-apy in Cardiac Disease) study, is already testing sarco/endoplasmic reticulum Ca2�-adenosine triphosphatase 2aene therapy in patients with end-stage heart failure, withome initial indications of efficacy (28). A second potentialarget for drug development may be Epac2. Significant de-reases in the calcium-regulating gene Epac2 were identifiedfter LVAD support in 11 recovered hearts (Fig. 7) (18). This

Figure 3 Identification of Genes in the cAMP-Mediated SignalinWhose Expression Was Altered in Association With R

Shaded symbols represent genes whose expression was significantly altered in ex

ecrease in Epac2 was unique to recovered hearts, as the b

attern was not found in nonrecovered hearts (Fig. 7) (18).pac2 tethers cyclic adenosine monophosphate to mitogen-

ctivated protein kinase, calcium-mediated signaling throughuclear factor of activated T cells, and metabolic signalingathways. The roles of both Epac2 and exchange proteinirectly activated by cyclic adenosine monophosphate 1 inultiple cell types and tissues, including the heart, nerves, and

ancreas, are just beginning to be understood (20,21).ytoskeletal proteins and integrins in recovery patients.yocardial recovery has also been associated with a specific

attern of changes in sarcomeric, nonsarcomeric, and mem-

thwayery

d versus implanted samples (18). cAMP � cyclic adenosine monophosphate.

g Paecov

plante

rane-associated proteins (29). These changes are high-

n(rpebpawepcdein

r

(kismetsttcMpta


lighted in Figure 8 (30). Affymetrix (Santa Clara, Califor-ia) microarray analysis was performed on paired samplesbefore and after LVAD implantation) and analyzed witheference to sarcomeric and nonsarcomeric cytoskeletalroteins (29). In the recovery group, of the nonsarcom-ric proteins, lamin A/C and spectrin increased fromefore to after LVAD implantation (29). The sarcomericroteins beta-actin, alpha-tropomyosin, alpha-1-actinin,nd alpha-filamin A increased in the recovered group,hile troponin T3 and alpha-2-actinin decreased at

xplantation (29). Vinculin decreased after LVAD sup-ort in the recovered group but increased in the nonre-overed group (29). Further work is needed in this area toefine if these changes that occur in expression of cytoskel-tal proteins and are uniquely associated with recovery arentegral to promoting improvements in contractility that areecessary for recovery.Integrins are bidirectional signaling molecules and play a

Figure 4 Calcium Handling After LVAD Support

This figure summarizes work to date in the literature of ventricular assist devices.assist device; NCX � Na�-Ca2� exchanger; SERCA � sarco/endoplasmic reticulum(2) � Terracciano et al. (26); (3) � Chen et al. (22); (4) � Marx et al. (23); (5) � Terr(25); (9) � Heerdt et al. (10); (10) � Terracciano et al. (19).

ole in mechanotransduction by mediating mechanical F

stretch) signals from the extracellular matrix, via proteininase cascades that provoke changes in gene expression,ncluding those involved in the hypertrophic response. Datauggest that the integrin pathway may play a key role at theolecular and cellular levels in processes of reverse remod-

ling and subsequent functional recovery occurring inhese patients (29). Work in a pre-clinical rat model hashown similar findings (31). Online Figure 1 summarizeshe changes in the integrin signaling pathway along withhe cytoskeletal proteins that occurred in the recoveryohort (29).

etabolism and growth factor–related genes and signalingathways in recovery patients. In addition to the findinghat the process of recovery was associated with cyclicdenosine monophosphate, calcium handling genes, and

1

ted, with permission, from Terracciano et al. (19)1. LVAD � left ventricular-adenosine triphosphatase; SR � sarcoplasmic reticulum; (1) � Harding et al. (9);o et al. (24); (6) � Dipla et al. (8); (7) � Chaudhary et al. (7); (8) � Frazier et al.

ReprinCa2�

accian

The original online posting of this article did not correctly identify the source ofigure 4. This has now been corrected.


integrin signaling, several new targets were identified in anunbiased approach, including those that regulate metabo-lism. Arginine:glycine amidinotransferase (AGAT), a rate-limiting enzyme in the creatine synthesis pathway, wassignificantly down-regulated after unloading in the recov-ered hearts, returning to normal levels in direct contrast tothe up-regulation of AGAT in patients with heart failurecompared with donor hearts (18,32). These changes inAGAT mRNA levels suggest a response to heart failure thatinvolves elevated local creatine synthesis (18,32). The mech-anisms leading to induced AGAT expression are unknownbut may be a response to the depletion of the local creatinepool, which we and others have shown to be a feature ofheart failure.

Insulin growth factor (IGF)-I mRNA was elevated at thetime of LVAD explantation in the patients who receivedpharmacologic therapy, which included clenbuterol (33).There were 2 groups distinguishable, those with low IGF-ImRNA at implantation who showed significant increaseduring recovery and those with high IGF-I mRNA atimplantation who remained high (33). IGF-I levels corre-lated with stromal cell–derived factor mRNA measuredboth in LVAD patients and in a wider cohort of patientswith heart failure (33). Microarray analysis of implantationand explantation samples of recovery patients further re-

Figure 5SR Calcium Content From Myocytes Isolated FromLVAD Cores and Tissue From Explanted (Recovery)and Transplanted Hearts Without Recovery

LVAD � left ventricular assist device; SR � sarcoplasmic reticulum.Reprinted, with permission, from Terracciano et al. (26).

vealed elevated IGF-II and IGF binding protein-4 and IGF

binding protein-6. IGF-I may act to limit atrophy andapoptosis during reverse remodeling and to promote repairand regeneration in concert with stromal cell–derived fac-tor. It may have important effects on fibroblasts (33). Aprevious analysis in 19 paired patient samples before andafter LVAD support with the HeartMate XVE device in acohort of patients who were transplanted (did not achievefull recovery) revealed expression changes in the Forkheadtranscription factor family and genes governing the angio-tensin/insulin signaling axis (34). This analysis also identi-fied a significant correlation in expression between Fork-head box 03A and angiotensin II type 1 receptor (34). More

Figure 6 Developed Tension, Expressedas Percent Change From Baseline

(A) Stimulation delay–response curves at delay periods ranging from 2 to120 s (n � number of muscles/number of hearts). (B) Data from the 40-sdelay period. Right: separation of left ventricular assist device (LVAD)–sup-ported failing hearts (F�LVAD) by duration of LVAD support. Short duration � 7to 103 days; long duration � 127 to 334 days. (C) Relationship between dura-tion and post-rest potentiation recovery (p � 0.05) at 40-s delay in hearts withLVADs, by heart. *p � 0.05 versus NF. Adapted from Ogletree et al. (27).F � failing hearts; NF � nonfailing hearts.


work will need to be done to determine the role andintegration of IGF-I/Forkhead and angiotensin II in re-modeling and recovery and localize these signals to myo-cytes and/or fibroblasts. In a later section, we discussmaladaptive fibrotic remodeling in response to LVADs andthe potential role of angiotensin-converting enzyme (ACE)inhibition in alleviating ventricular stiffness and fibrosis. A

Figure 7 Epac2 mRNA Levels

(A) Exchange protein directly activated by cyclic adenosine monophosphate 2 (Eparecovered hearts after ventricular assist device support (n � 11, p � 0.01). (B) N

Figure 8 Changes in the Cytoskeletal Pathway That Occurred i

Reproduced, with permission, from Birks et al. (29), adapted from Towbin and Bow

recent study by Bhavsar et al. (35) showed that IGF-I isreleased preferentially from fibroblasts over cardiac myo-cytes. Thus, the interaction between angiotensin II andIGF-I both mechanistically and clinically is still not wellunderstood.Micro-ribonucleic acid (miRNA). miRNAs constituteone of the more abundant classes of molecules that

essenger ribonucleic acid (mRNA) levels were significantly reduced inificant differences were seen in nonrecovered hearts. (n � 5, p � NS) (18).

covered Patients Only

0).

c2) mo sign

n Re

les (3

of devi


regulate genes in animals. MiRNAs are small, endoge-nous, noncoding RNAs (36). Many of these miRNAshave been shown to inhibit post-transcriptional process-ing (36,37). Recent work by Matkovich et al. (38)suggests that adding miRNA profiling to mRNA profil-ing enhances the ability of mRNA profiles to categorizethe clinical status of heart failure before and afterbiomechanical unloading. The results of this study byMatkovich et al. (38) confirmed 3 earlier identifiedmiRNAs associated with heart failure (miR-24, miR-125b, and miR-195) (38,39). In addition, Matkovich etal. (38) extended earlier work in mouse models (39,40)showing that miR-21, miR-23a, and miR-199a-3p werealso regulated in human heart failure. One of the mostexciting aspects of the study by Matkovich et al. (38) wasthe reversibility of the miRNAs with LVAD support.Table 3 lists miRNAs that normalize after LVADsupport. It is not yet known which miRNA expressionchanges are associated with clinical end points thatcorrelate with recovery.Natriuretic peptides and chromogranin A are normalizedby LVAD support. Despite divergent etiologies, heartfailure is characterized by the activation of neurohormonalsystems: catecholamines, natriuretic peptides, and compo-nents of the renin-angiotensin axis are increased and havebeen found to have pathophysiological and prognosticimplications. Some of these molecules exert local paracrineactivation, but their plasma levels have been demonstratedto be markers of clinical outcomes. Increased cardiac atrialnatriuretic peptide (ANP) and B-type natriuretic peptide(BNP) expression in patients with heart failure is associatedwith increased expression of the natriuretic peptides metab-olizing natriuretic peptide receptor C receptors and bluntedresponsiveness of guanylate cyclase A to ANP by reducedcyclic guanosine monophosphate synthesis.

Unloading the failing heart with an LVAD is associatedwith a decrease in natriuretic peptides and re-establishmentof the local responsiveness of guanylate cyclase A to ANP(41). Cardiac expression of ANP, BNP, and natriureticpeptide receptor C mRNA also correlated significantly with

MiRNAs That Normalize After LVAD SupportTable 3 MiRNAs That Normalize After LVAD Support

Fold Change

NameFailing vs.Nonfailing

Post-LVAD vs.Nonfailing

Post-LVFai

miR-27b 3.15 1.14 �4

miR-30a-5p 3.06 1.07 �2

miR-30c 3.31 1.32 �2

miR-30d 2.96 1.23 �2

miR-103 3.54 1.20 �2

miR-130a 3.06 1.36 �2

miR-378 3.60 1.20 �3

let-7f 4.84 1.18 �4

n � 10 LVAD-supported hearts, 80% men, 40% ischemic, average support time 1.7 months, typeLVAD � left ventricular assist device; miRNA � micro–ribonucleic acid.

cardiomyocyte diameters. In contrast to the latter, the levels

of the natriuretic peptides are fully reversed to the level ofthe controls, indicating that their expression is partly inde-pendent of cardiac hypertrophy and regulated by heartfailure–associated factors such as cardiomyocyte stretch(41). Evidence that unloading may be sufficient for adecrease in natriuretic peptides (and that recovery is notnecessary) comes from pre-clinical data, in which natriureticpeptides are decreased in heterotopic transplants of failingrat hearts (42,43).

Chromogranin A is an acidic calcium-binding proteinand is the major soluble constituent in secretory vesiclesthroughout the neuroendocrine system. Chromogranin Awas found to be significantly up-regulated during heartfailure and is co-stored with catecholamines and natriureticpeptides. BNP and chromogranin A are co-stored in themyocardium of patients with dilated cardiomyopathy,whereas this colocalization was not found in healthy con-trols (44). Wohlschlaeger et al. (45) investigated the expres-sion of natriuretic peptides and chromogranin A by immu-nohistochemistry and morphometric quantification beforeand after LVAD implantation. In a different set of patients,chromogranin A was evaluated in the plasma (45). Wedemonstrated that in line with ANP and BNP, chromo-granin A is significantly increased in patients with conges-tive heart failure compared with healthy controls and decreasedby ventricular support (45). Moreover, sarcoplasmic colocaliza-tion of BNP and chromogranin A is diminished after unload-ing. However, because of its low expression, the negativeregulation of chromogranin A is not reflected by plasma levels,so chromogranin A does not appear to be an appropriatebiomarker for the monitoring of “reverse cardiac remodeling”after unloading (45) (Online Fig. 2).

Summary

In summary, left ventricular unloading in patients withend-stage heart failure on LVAD support has been shownto improve myocardial structure and function, includingimproving beta-adrenergic responsiveness and myocyte con-

p Value

. Failing vs.Nonfailing

Post-LVAD vs.Nonfailing

Post-LVAD vs.Failing

2.60 � 10�4 0.270 0.0620

4.76 � 10�4 0.566 0.0821

1.24 � 10�4 0.105 0.0853

6.47 � 10�4 0.211 0.116

1.24 � 10�4 0.156 0.0834

6.80 � 10�4 0.127 0.270

6.25 � 10�4 0.153 0.0544

1.50 � 10�4 0.211 0.00543

ce not disclosed. Reproduced, with permission, from Matkovich et al. (38).

AD vsling

.75

.85

.51

.40

.94

.26

.00

.11

tractility. The phenotypic changes in the heart associated


with LVAD support may arise from changes in mRNA,miRNA, or post-translational mechanisms such as ubiquiti-nation or phosphorylation events, as discussed by Margulies(11). Several working groups, including our own, havedefined changes in gene expression that occur with LVADsupport that may be important in improving beta-adrenergic responsiveness and calcium handling, includingpartial restoration of beta-receptor density and alteredregulation of the genes in G-protein–coupled signaling andcalcium handling. A gene expression profile analysis usinggene chips with 199 myocardial samples from failing,LVAD-supported and nonfailing human hearts demon-strated that the majority of transcriptional changes accom-panying morphological and functional alterations in theheart were modest (46). Changes in miRNA expression in arecent study were more robust (38). Protein expression hasbeen mainly restricted to calcium handling, beta-adrenergic,and extracellular matrix pathways, and much work is yet tobe done. Pairing pre-clinical work in model organisms inwhich these targets (genes or miRNA) are overexpressed ordeleted and their function is understood in the context ofheart failure will be helpful. Overexpressing genes in whichphosphorylation sites (or other site of post-translationalmodification) are mutated will aid in the identification ofhow post-translational mechanisms regulate myocardialfunction and remodeling. However, rodent models do notrecapitulate human heart failure. Secondly, heterotopictransplantation has limitations in modeling assist devices.The next generation of studies will take advantage ofstate-of-the-art technology to yield more information fromhuman myocytes from core biopsies and define how changes inRNA and protein may alter morphology and contractility.

A recent histological study by Drakos et al. (47) foundthat unloading with LVADs in 15 patients receiving theHeartMate I device (pulsatile) resulted in a 33% increase inmicrovascular density, although the total luminal area wasdecreased, so the functional patency of these vessels isunclear.

Some remodeling responses to mechanical unloading maybe maladaptive. The most prominent advancement of thishypothesis is the concept of myocardial atrophy and rever-sal, which is being investigated in the Harefield RecoveryProtocol Studies described earlier. Mechanisms regulatingcardiac atrophy are unknown, although it is thought that theubiquitin-proteasome degradation pathway that is active inskeletal muscle and the coronary vascular bed plays a roleand may represent an overcompensation because of the lackof opposing hypertrophic stimuli after chronic unloading(48,49). Another concept in maladaptive remodeling is thatof fibrosis and extracellular matrix turnover. One study byKlotz et al. (50) of collagen content and subtypes afterLVAD support found a significant increase in total andcross-linked collagen in the myocardium compared withnonfailing and medically managed patients with heart fail-ure, which correlated with increased left ventricular stiffness.
Interestingly, the majority of LVAD patients after implan-
tation were not on ACE inhibitors, which have beendemonstrated to improve fibrosis and remodeling (51,52). Asubsequent retrospective cohort study by the same group,comparing LVAD patients who did and did not receiveACE inhibitor therapy after implantation, demonstrated asignificant decrease in collagen content and myocardialstiffness in the cohort with LVADs plus ACE inhibitors(16). At the molecular level, quantitation of matrix metal-loproteinases and their inhibitors demonstrated near nor-malization of the ratio of matrix metalloproteinase 1 totissue inhibitor of metalloproteinase 1 in LVAD patientsreceiving ACE inhibitors, suggesting a mechanism for theimprovement collagen turnover. These findings support thehypothesis that maximizing optimal medical managementafter ventricular unloading with LVADs may promotemyocardial recovery. In sum, the modest gene expressionchanges are sufficient to change vascular and fibrotic prop-erties in the heart. However, a more thorough understand-ing of the unique therapies and underlying genetic andenvironmental properties will further aid in our understand-ing of the biological pathways through which unloadingwith LVADs leads to remodeling and recovery. A betterunderstanding of the biology may allow us to alter theengineering properties of the devices and/or patient selec-tion criteria (see the following discussion) to improveoutcomes.Study limitations. The LVAD population presents aunique and valuable opportunity to obtain myocardial tissueof patients with end-stage heart failure at the time ofimplantation, and often at the time of heart and/or LVADexplantation, after a period of unloading. There are, how-ever, some recognized limitations of carrying out molecularstudies on the LVAD patient population. Study sizes tendto be small, which is a particular challenge when performinglarge-scale transcriptome analysis, although this limitationis being overcome by multicenter collaborations and theoverall increase in the use of LVADs. Tissue procurementmust also be carefully planned, and the tissue received fromthe apical transmural tissue core may be fibrotic and notrepresentative of the whole myocardium. In addition, theclinical presentation, duration of support, and medicalmanagement are uncontrollable variables, creating a heter-ogenous patient population. Animal models have beencreated in an attempt to study ventricular unloading in amore controlled fashion. These include a thoracic inferiorvena cava constriction model described by Lisy et al.(53,54), which demonstrates myocardial atrophy withchronic unloading, with significant up-regulation of therenin-angiotensin-aldosterone system axis and decrease inANP and BNP expression. Although the myocardial atro-phy findings have important implications for research inhuman LVAD patients, this model creates a biventricularlow cardiac output state from a presumably normal baselineand thus is not analogous to the physiologic effect ofimplanting an LVAD. Another established animal model is
that of heterotopic cardiac transplantation, described by

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2


multiple groups, including Tevaearai et al. (55). In thismodel, heart failure is induced in a “donor” cohort ofrabbits, after which the hearts are transplanted heterotopi-cally by anastomosing the aorta and pulmonary trunk to therecipient rabbits’ carotid artery and internal jugular vein,respectively. The heart is perfused only through the coro-nary circulation, through the coronary sinus, and returned tothe recipient via the right heart, creating a completelyunloaded left ventricle (55). Although this model is alsointriguing, it also involves a transplantation, removing theorgan from the systemic physiology of the heart failure“donor,” and also subjecting the heart to the immunologicaland biochemical milieu associated with transplantation.Given the limitations of animal models, it is critical tocontinue advancing translational research in the LVADpatient population to apply the understanding of unloadingresponses in heart failure to the development of clinicalpredictors and medical therapy for these patients.Final clinical perspective and future directions. Recentclinical trials have shown that both pulsatile andcontinuous-flow LVADs significantly improve the qualityof life and functional capacity of patients with heart failure(4,5). Translational research in the area of LVAD supporthas provided new information on signaling pathways. Tis-sue procurement, deoxyribonucleic acid sequencing, RNAsequencing, and epigenetic tools will help in our furtherunderstanding of the underlying signals regulating thereverse remodeling and recovery potential of the failinghuman heart. Investigation in the LVAD patient popula-tion has been limited by the fact that there is no consensuson medical management toward, and clinical monitoring of,myocardial recovery. For instance, the findings correlatingACE inhibitor use and favorable decrease in myocardialstiffness and collagen subtypes suggest a greater role forcontinuing optimal medical management after LVADplacement. The elucidation of myocardial atrophy pathwaysmay also illuminate the concept of optimizing or weaningunloading after LVAD placement to minimize atrophy andpromote recovery. Our current understanding of the role ofextracellular matrix in heart failure and the ability of thefibroblast to remodel in response to LVADs represents anew focus area. Further elucidating these pathways willallow for the discovery of novel therapeutic strategies forpatients in earlier stages of heart failure. Identification of thebiomarkers and factors that will aid in risk stratification forpatients with LVAD therapy will help move this fieldforward and enhance patient care.

Reprint requests and correspondence: Dr. Jennifer L. Hall,University of Minnesota, 310 Church Street, Minneapolis, Min-nesota 55455. E-mail: [email protected].

REFERENCES

1. Hunt SA, Abraham WT, Chin MH, et al. 2009 focused updateincorporated into the ACC/AHA 2005 guidelines for the diagnosis
and management of heart failure in adults: a report of the American
College of Cardiology Foundation/American Heart Association TaskForce on Practice Guidelines. J Am Coll Cardiol 2009;53:e1–90.

2. Jessup M, Brozena S. Heart failure. N Engl J Med 2003;348:2007–18.3. Kirklin JK, Naftel DC, Kormos RL, et al. Second INTERMACS

annual report: more than 1,000 primary left ventricular assist deviceimplants. J Heart Lung Transplant 2010;29:1–10.

4. Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-flowdevice in patients awaiting heart transplantation. N Engl J Med2007;357:885–96.

5. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failuretreated with continuous-flow left ventricular assist device. N EnglJ Med 2009;361:2241–51.

6. Rose EA, Moskowitz AJ, Packer M, et al. The REMATCH trial:rationale, design, and end points. Ann Thorac Surg 1999;67:723–30.

7. Chaudhary KW, Rossman EI, Piacentino V III, et al. Alteredmyocardial Ca2� cycling after left ventricular assist device support inthe failing human heart. J Am Coll Cardiol 2004;44:837–45.

8. Dipla K, Mattiello JA, Jeevanandam V, Houser SR, Margulies KB.Myocyte recovery after mechanical circulatory support in humans withend-stage heart failure. Circulation 1998;97:2316–22.

9. Harding JD, Piacentino V III, Gaughan JP, Houser SR, MarguliesKB. Electrophysiological alterations after mechanical circulatory sup-port in patients with advanced cardiac failure. Circulation 2001;104:1241–7.

0. Heerdt PM, Holmes JW, Cai B, et al. Chronic unloading by leftventricular assist device reverses contractile dysfunction and alters geneexpression in end-stage heart failure. Circulation 2000;102:2713–9.

1. Margulies KB. Reversal mechanisms of left ventricular remodeling:lessons from left ventricular assist device experiments. J Card Fail2002;8:S500–5.

2. Maybaum S, Mancini D, Xydas S, et al. Cardiac improvement duringmechanical circulatory support: a prospective multicenter study of theLVAD Working Group. Circulation 2007;115:2497–505.

3. Ogletree-Hughes ML, Stull LB, Sweet WE, Smedira NG, McCarthyPM, Moravec CS. Mechanical unloading restores beta-adrenergicresponsiveness and reverses receptor downregulation in the failinghuman heart. Circulation 2001;104:881–6.

4. Zafeiridis A, Jeevanandam V, Houser SR, Margulies KB. Regressionof cellular hypertrophy after left ventricular assist device support.Circulation 1998;98:656–62.

5. Klotz S, Barbone A, Reiken S, et al. Left ventricular assist devicesupport normalizes left and right ventricular beta-adrenergic pathwayproperties. J Am Coll Cardiol 2005;45:668–76.

6. Klotz S, Danser AH, Foronjy RF, et al. The impact of angiotensin-converting enzyme inhibitor therapy on the extracellular collagenmatrix during left ventricular assist device support in patients withend-stage heart failure. J Am Coll Cardiol 2007;49:1166–74.

7. Birks EJ, Tansley PD, Hardy J, et al. Left ventricular assist device anddrug therapy for the reversal of heart failure. N Engl J Med 2006;355:1873–84.

8. Hall JL, Birks EJ, Grindle S, et al. Molecular signature of recoveryfollowing combination left ventricular assist device (LVAD) supportand pharmacologic therapy. Eur Heart J 2007;28:613–27.

9. Terracciano CM, Koban MU, Soppa GK, et al. The role of the cardiacNa�/Ca2� exchanger in reverse remodeling: relevance for LVAD-recovery. Ann N Y Acad Sci 2007;1099:349–60.

0. Holz GG. Epac: a new cAMP-binding protein in support ofglucagon-like peptide-1 receptor-mediated signal transduction in thepancreatic beta-cell. Diabetes 2004;53:5–13.

1. Seino S, Shibasaki T. PKA-dependent and PKA-independent path-ways for cAMP-regulated exocytosis. Physiol Rev 2005;85:1303–42.

2. Chen X, Piacentino V, 3rd, Furukawa S, Goldman B, Margulies KB,Houser SR. L-type Ca2� channel density and regulation are altered infailing human ventricular myocytes and recover after support withmechanical assist devices. Circ Res 2002;91:517–24.

3. Marx SO, Reiken S, Hisamatsu Y, et al. Pka phosphorylationdissociates fkbp12.6 from the calcium release channel (ryanodinereceptor): defective regulation in failing hearts. Cell 2000;101:365–76.

4. Terracciano CM, Harding SE, Adamson D, et al. Changes insarcolemmal ca entry and sarcoplasmic reticulum ca content in ven-tricular myocytes from patients with end-stage heart failure followingmyocardial recovery after combined pharmacological and ventricular
assist device therapy. Eur Heart J 2003;24:1329–39.
mailto:[email protected]


25. Frazier OH, Myers TJ. Left ventricular assist system as a bridge tomyocardial recovery. Ann Thorac Surg 1999;68:734–41.

26. Terracciano CM, Hardy J, Birks EJ, Khaghani A, Banner NR, YacoubMH. Clinical recovery from end-stage heart failure using left-ventricular assist device and pharmacological therapy correlates withincreased sarcoplasmic reticulum calcium content but not with regres-sion of cellular hypertrophy. Circulation 2004;109:2263–5.

27. Ogletree ML, Sweet WE, Talerico C, et al. Duration of left ventric-ular assist device support: Effects on abnormal calcium cycling andfunctional recovery in the failing human heart. J Heart Lung Trans-plant 2010;29:554–61.

28. Jaski BE, Jessup ML, Mancini DM, et al. Calcium upregulation bypercutaneous administration of gene therapy in cardiac disease(CUPID trial), a first-in-human phase 1/2 clinical trial. J Card Fail2009;15:171–81.

29. Birks EJ, Hall JL, Barton PJ, et al. Gene profiling changes incytoskeletal proteins during clinical recovery after left ventricular-assistdevice support. Circulation 2005;112:I57–64.

30. Towbin JA, Bowles NE. The failing heart. Nature 2002;415:227–33.31. Lara-Pezzi E, Terracciano CM, Soppa GK, et al. A gene expression

profile of the myocardial response to clenbuterol. J Cardiovasc TranslRes 2009;2:191–7.

32. Cullen ME, Yuen AH, Felkin LE, et al. Myocardial expression of thearginine:glycine amidinotransferase gene is elevated in heart failureand normalized after recovery: potential implications for local creatinesynthesis. Circulation 2006;114:I16–20.

33. Barton PJ, Felkin LE, Birks EJ, et al. Myocardial insulin-like growthfactor-I gene expression during recovery from heart failure aftercombined left ventricular assist device and clenbuterol therapy. Circu-lation 2005;112:I46–50.

34. Hall JL, Grindle S, Han X, et al. Genomic profiling of the humanheart before and after mechanical support with a ventricular assistdevice reveals alterations in vascular signaling networks. PhysiolGenomics 2004;17:283–91.

35. Bhavsar PK, Brand NJ, Felkin LE, et al. Clenbuterol induces cardiacmyocyte hypertrophy via paracrine signalling and fibroblast-derivedIGF-1. J Cardiovasc Transl Res 2010;3:688–95.

36. Saunders MA, Liang H, Li WH. Human polymorphism at microRNAsand microRNA target sites. Proc Natl Acad Sci U S A 2007;104:3300–5.

37. Lim LP, Glasner ME, Yekta S, Burge CB, Bartel DP. VertebratemicroRNA genes. Science 2003;299:1540.

38. Matkovich SJ, Van Booven DJ, Youker KA, et al. Reciprocal regula-tion of myocardial microRNAs and messenger RNA in humancardiomyopathy and reversal of the microRNA signature by biome-chanical support. Circulation 2009;119:1263–71.

39. van Rooij E, Sutherland LB, Liu N, et al. A signature pattern ofstress-responsive microRNAs that can evoke cardiac hypertrophy andheart failure. Proc Natl Acad Sci U S A 2006;103:18255–60.

40. Tatsuguchi M, Seok HY, Callis TE, et al. Expression of microRNAsis dynamically regulated during cardiomyocyte hypertrophy. J Mol CellCardiol 2007;42:1137–41.

41. Kuhn M, Voss M, Mitko D, et al. Left ventricular assist device supportreverses altered cardiac expression and function of natriuretic peptides
and receptors in end-stage heart failure. Cardiovasc Res 2004;64:308–14.
42. Tsuneyoshi H, Oriyanhan W, Kanemitsu H, et al. Does the beta2-agonist clenbuterol help to maintain myocardial potential to recoverduring mechanical unloading? Circulation 2005;112:I51–6.

43. Wang J, Tsukashita M, Nishina T, et al. Chronic partial unloadingrestores beta-adrenergic responsiveness and reverses receptor down-regulation in failing rat hearts. J Thorac Cardiovasc Surg 2009;137:465–70.

44. Pieroni M, Corti A, Tota B, et al. Myocardial production of chromo-granin A in human heart: a new regulatory peptide of cardiac function.Eur Heart J 2007;28:1117–27.

45. Wohlschlaeger J, von Winterfeld M, Milting H, et al. Decreasedmyocardial chromogranin a expression and colocalization with brainnatriuretic peptide during reverse cardiac remodeling after ventricularunloading. J Heart Lung Transplant 2008;27:442–9.

46. Margulies KB, Matiwala S, Cornejo C, Olsen H, Craven WA,Bednarik D. Mixed messages: transcription patterns in failing andrecovering human myocardium. Circ Res 2005;96:592–9.

47. Drakos SG, Kfoury AG, Hammond EH, et al. Impact of mechanicalunloading on microvasculature and associated central remodelingfeatures of the failing human heart. J Am Coll Cardiol 2010;56:382–91.

48. Hill JA, Olson EN. Cardiac plasticity. N Engl J Med 2008;358:1370–80.49. Herrmann J, Ciechanover A, Lerman LO, Lerman A. The ubiquitin-

proteasome system in cardiovascular diseases-a hypothesis extended.Cardiovasc Res 2004;61:11–21.

50. Klotz S, Foronjy RF, Dickstein ML, et al. Mechanical unloadingduring left ventricular assist device support increases left ventricularcollagen cross-linking and myocardial stiffness. Circulation 2005;112:364–74.

51. Greenberg B, Quinones MA, Koilpillai C, et al. Effects of long-termenalapril therapy on cardiac structure and function in patients with leftventricular dysfunction: results of the SOLVD Echocardiographysubstudy. Circulation 1995;91:2573–81.

52. van Krimpen C, Smits JF, Cleutjens JP, et al. DNA synthesis in thenon-infarcted cardiac interstitium after left coronary artery ligation inthe rat: effects of captopril. J Mol Cell Cardiol 1991;23:1245–53.

53. Lisy O, Redfield MM, Jovanovic S, et al. Mechanical unloading versusneurohumoral stimulation on myocardial structure and endocrinefunction in vivo. Circulation 2000;102:338–43.

54. Lisy O, Redfield MM, Schirger JA, Burnett JC Jr. Atrial BNPendocrine function during chronic unloading of the normal canineheart. Am J Physiol Regul Integr Comp Physiol 2005;288:R158–62.

55. Tevaearai HT, Walton GB, Eckhart AD, Keys JR, Koch WJ.Heterotopic transplantation as a model to study functional recovery ofunloaded failing hearts. J Thorac Cardiovasc Surg 2002;124:1149–56.

Key Words: beta-adrenergic y calcium y clinical trials y geneexpression y integrins y microarray y miRNA y ventricular assistdevice.

APPENDIX

For supplementary figures and their legends,
please see the online version of this article.

Documents

Clinical, Molecular, and Genomic Changes in Response … · Clinical, Molecular, and Genomic Changes ... PHD,¶# Mark Slaughter, ... the 1,092 primary LVAD patients