Theneed for artificial hearts - Heart | A leading

Heart 1996;76:200-206

VIEWPOINT

The need for artificial hearts

S Westaby

SummaryChronic immunosuppression, allograftcoronary disease, and restricted avail-ability of donor organs continue to limitthe scope of cardiac transplantation.Meanwhile increasingly favourable expe-rience with implantable blood pumpsused as a bridge to transplant has reintro-duced the concept of permanent mechan-ical cardiac support. Existing models (forexample, the Thermo CardiosystemsHeartmate device) are now used for suchsupport in patients who are not candi-dates for transplantation. Miniaturisedaxial flow pumps such as the Jarvik 2000fit within the failed left ventricle and pro-vide an exciting prospect for the treat-ment of heart failure in the future.Preliminary experience suggests that the"offloaded" left ventricle may recover.Mechanical blood pumps can be usedbefore the onset of multisystem failureand removed ifthe myocardium recovers.This "bridge to recovery" concept shouldbe tested in patients with recoverable car-diomyopathy and those with coronarydisease and poor left ventricular functionwhere an implantable pump can be usedin conjunction with myocardial revascu-larisation.

(Heart 1996;76:200-206)

Keywords: heart failure; mechanical hearts; bridge torecovery

Oxford Heart Centre,Oxford RadcliffeHospital, The JohnRadcliffe, Headington,Oxford OX3 9DUS WestabyAccepted for publication25 April 1996

Though cardiac transplantation markedlyimproves quality of life and provides a costeffective alternative to medical therapy, thenumber of donor organs is restricted.'2 Overallsuccess is also limited by the complications ofchronic immunosuppression, opportunisticinfection, and the development of allograftcoronary artery disease requiring retransplan-tation in 40% of patients by six years.34 In theabsence of a realistic alternative, transplanta-tion discards some poorly functioning butpotentially recoverable hearts, particularly indilated cardiomyopathy, viral myocarditis, or

protozoal -infections. In contrast, modernblood pumps can offload the failed left ventricleand encourage an element of recovery. There

is now the prospect of improved mechanicaldevices that, without the risks of immunosup-pression or organ rejection, will relieve symp-toms and prolong life in many heart failurepatients. Existing devices for mechanical car-diac support used as a bridge to transplanta-tion function for prolonged periods withoutblood damage or risk of infection.5-7 The newintraventricular axial flow pumps now underinvestigation provide better prospects for per-manent mechanical support and will becomethe artificial hearts of the relatively near future.Such devices provide the opportunity to inter-vene early before multisystem failure developsand to combine coronary or valve surgery withventricular support. There remains the optionto explant the device should the myocardiumrecover.

Experience with devices used as amechanical bridge to transplantThe concept of a mechanical bridge to trans-plantation (for moribund patients awaiting adonor organ) began in the late 1960s with theclinical use of the Liotta pneumatic doubleventricle heart by Cooley et al.8 This first totalartificial heart supported the circulation for 64hours pending acquisition of a donor organ.When the patient died from infection andmultisystem failure, intense controversy inhib-ited further attempts for almost 10 years. In1978 both Stanford University and The TexasHeart Institute attempted without success touse a left ventricular assist device (LVAD) as amechanical bridge to transplantation. Thesame year Reemtsma et al used the intra-aorticballoon pump to sustain three patients totransplantation.9 All three survived andbecame the first cases of the successful use of amechanical bridge to transplant with any typeof mechanical device. In 1981 Cooley et alused the Akutsu pneumatic total artificialheart to support a patient for 55 hours beforetransplantation.'0 The patient died 10 dayslater from sepsis and multiple organ failure;none the less, problems were defined andavoided in future efforts. The first long termLVAD transplant survivor was a 51 year oldman sustained for nine days with the Novacor(Novacor Medical Corporation, Oakland CA)system at Stanford University in 1984."1 Soonafterwards Hill et al, at Pacific Medical Center,

200

on May 15, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.76.3.200 on 1 S

eptember 1996. D

ownloaded from

http://heart.bmj.com/

The needfor artificial hearts

Figure 1 The Novacorleft ventricular assist device.

San Francisco, used the externally situatedPearce/Donachy LVAD (Thoratec Labora-tories Corporation, Berkeley CA) to support a47 year old man with post-infarction cardio-genic shock. The patient received a donorheart 52 hours later and survived.'2 Theseoperations occurred within a brief time frameand established mechanical bridge to trans-plantation as an expanding (but expensive)part of cardiac transplantation programmes.An important finding was that chronic offload-ing of the myocardium often resulted in sub-stantial improvement of native left ventricularfunction similar to that experienced after pro-longed bed rest.'3Among the currently available blood pumps

are two implantable LVADs specificallydesigned for long-term circulatory support.The Thermocardio Systems (Woburn MA)and Novacor (Oakland CA) pumps emanatefrom the mechanical support initiatives of TheTexas Heart Institute and Stanford Universityrespectively. Both are pusher plate typeLVADs originally designed to function as per-manent artificial hearts. The Novacor LVAD(fig 1) is composed of a balanced solenoidenergy convertor, dual pusher plate, sac-typeblood pump with a microprocessor based con-trol and monitoring console.6 The integratedenergy convertor and pump are implanted intothe abdomen and connected to the patient byDacron conduits to the left ventricular apexand the ascending aorta. The conduits contain21 mm Carpentier Edwards pericardial valveswhich provide unidirectional blood flow.Electrical energy is converted into hydraulicwork which ejects blood from the pump. Theblood sac has a highly smooth surface which isdesigned not to require anticoagulation. Theelectrical power cable is brought out percuta-

neously from the right lower quadrant of theabdomen to connect to an extracorporeal con-trol console and battery power system.Thoughsuccessful as a bridge to transplantation, thenoise of the device, together with the length ofthe inflow graft and a significant incidence ofthromboembolism, limit its potential use as apermanent implant."IThe Thermo Cardiosystems Heartmate

vented electric LVAD (fig 2A and B) consistsof a positive displacement, pusher plate pumpactivated by a low speed torque motor that canproduce a stroke volume of 83 ml at ratesvarying from 50 to 120 beats per minute. 4

The electromechanical actuator fits beneaththe pumping diaphragm and converts therotary motion of the torque motor to the linearmotion of the pump diaphragm (fig 3). Afterthe motor switches on and makes one com-plete revolution pushing the diaphragm, bloodejects from the pump. The motor then turnsoff and the pump passively refills. There is ashort inlet conduit which fits directly into theapex of the left ventricle and a longer outlettube to the ascending aorta. Both the inlet andoutlet conduits are fitted with 25 mm porcinevalves (Medtronic Blood Systems). The flex-ing diaphragm is fabricated from biomerpolyurethane and has an integrally texturedfibrillar surface. All metal components have asintered titanium/blood interface. This is acounter intuitive approach to thromboem-bolism whereby the textured surface promotesthe formation of a fibre and cellular coagulumthat evolves into a blood compatible biologicallining.'5 The pseudo-intimal layer obviates theneed for anticoagulation. The implantedpump components are situated beneath thediaphragm, either in a retroperitoneal pocketor within the peritoneal cavity. A thick

201



j.com/


eptember 1996. D

ownloaded from


202 Weeb

Figure 2(A) The ThermoCardiosystems Heartmatevented ventricular LVAD.(B) x-ray of the Heartmatedevice in situ. M



j.com/


eptember 1996. D

ownloaded from



Blood chamber

/ Diaphragm

A Air

Pusher-flplate

Motor chamber

B

C

D

Figure 3 Mechanism of action of the Heartmate LVAD. The torque motor rotates amoves the diaphragm of the blood chamber. Blood is expelled during the systolic phaseand D) while air is vented outside the body during the diastolic phase (A).

fibrovascular capsule develops aroundpump, preventing migration and resiinfection. A percutaneous electricconnects the blood pump assembly toportable system controller, which easilyonto a belt (fig 4). The controller contamicroprocessor that regulates motor funand monitors pump performance. The pcan be operated in either a fixed ratautomatic mode, asynchronous with thetrocardiogram. In the automatic modedevice increases output in responseincreased venous filling, providing flow x1 0 I/min.

In the ambulant patient the controllconnected to two lead acid gel cells batiwhich are contained in a shoulder holstbelt bag worn around the patient's wai;portable power base unit is used to recdextra batteries or to power the device wheipatient sleeps or sits for prolonged perTwo batteries used in parallel last up tohours and can be exchanged on a rotbasis to provide untethered support in

nitely. The external noise from the motor isbarely audible and the relatively uncompli-cated design of both internal and externalcomponents make the device durable and easyto operate.The fully portable electric system was first

used in 1990 after extensive experience withJ the pneumatically powered pump.16 Up to

September 1995, 435 pneumatic and 49 elec-tric devices have been implanted with maxi-mum duration of support of 344 and 503 daysrespectively. All have been used as a bridge totransplantation, recently on an outpatientbasis.'7 Much useful information was providedby the Food and Drug Administrationapproved clinical study of 116 patients, whowere compared with 46 control patients.'8 Theend point was survival at 60 days after trans-plantation. Seventy one percent of theHeartmate patients survived to undergo trans-plantation compared with 36% in the controlgroup. After the transplant the group sup-ported by the LVAD had a survival rate of65% v 30% for controls. Improved survivalboth to transplantation and afterwards waspartly the result of dramatic physiologicalrehabilitation. The improved haemodynamicsreversed multisystem organ failure so thatpatients could resume physical activity andadopt an exercise programme within the hos-pital environment. Only one patient suffered adevice related complication (0-9%); this wasbecause of a loosened connector. Thrombo-embolism did not occur in the absence ofinfection.'9 Besides resolution of hepatic andrenal failure, prolonged circulatory supportreversed the neurohormonal effects of chronicheart failure. Serum aldosterone concentra-tions, plasma renin activity, and concentra-tions of atrial naturetic peptide revert tonormal after prolonged support.5

and In four patients with dilated cardiomyopa-e (C thy at The Berlin Heart Hospital antibody

against the beta I receptor was used as amarker (Personal communication, J Mueller,Berlin Heart Hospital.) Before insertion of the

the LVAD all had left ventricular ejection frac-isting tions less than 15% but with prolongedline support (between 160 and 340 days) auto-the antibody titres and heart rate fell and left

clips ventricular ejection fraction increased pro-Lins a gressively. The patients were weaned byLction switching their LVAD from automatic mode)ump to fixed rate and reducing this from 90 to 50te or beats/min. After removal of the device theelec- patients remained stable on medical treatmentthe (up to 10 months after operation). This

e to encouraging experience gives rise to the con-up to cept of a mechanical bridge to recovery.

ler isteries:er orvst. Ahargen theriods.eighttatingidefi-

THE PROSPECT FOR PERMANENT MECHANICALDEVICESThe drive towards a permanently implantedartificial heart began in 1982 when DeVriesimplanted the Jarvik 7 pneumatic total artifi-cial heart in a 61 year old dentist (fig 5).2°Three further Jarvik 7 total artificial heartswere implanted by DeVries betweenNovember 1984 and April 1985 and, thoughthe haemodynamic fimction was encouraging,

000, --",%

..l

op- ----qA

203



j.com/


eptember 1996. D

ownloaded from


Westaby

Figure 4 The Heartmatedevice in situ with externalsystem controller and bat-teries. Reproduced with thepermission of ThermoCardiosystems.

Pump

System Controller

thromboembolism and infection were seriousproblems. With technological improvementsthe incidence of neurological complicationsand haemolysis were substantially reduced andone patient was sustained for 620 days at theHumana Heart Institute. Subsequently, theJarvik 7 was used as a bridge to transplant inpatients with biventricular failure.2' Thelargest series was that of Cabrol in Paris where24 (60%) of 40 consecutive patients receivedsuccessful transplants.22About 80% of the haemodynamic work of

the heart is performed by the left ventricle.Right ventricular failure can usually be treatedmedically after implantation of an LVAD andmay be avoided by earlier intervention. Theuse of a biventricular artificial heart is thenunnecessary. It is now clear that moribundheart failure patients can improve to NYHA Iand return to the community with an LVAD.Pilot programmes in the United States showthat well-motivated LVAD patients can returnto work or higher education while they wait fora transplant. In some instances use of thedevice has reduced inpatient costs from $5000

Percutaneous ventand electric lead

per day in the intensive care unit to less than$50 per day with the patient at home.The success of the outpatient LVAD pro-

gramme provides a powerful argument for theuse of the electric LVAD as an alternative tocardiac transplantation. In the United Statesin 1991 3797 patients were registered on theheart transplant waiting list with a medianwaiting period of 198 days: 2107 patientsunderwent heart transplantation and 778patients died waiting.23 Data from the UnitedNetwork for Organ Sharing have shown that amale with type 0 blood group, weighing morethan 90 kg will wait a mean time of 595 daysfor a donor organ.23 Half of this patient groupdie before transplantation. Consequently, asmany as 66 000 end stage cardiac patients peryear may be candidates for mechanical circula-tory support in the USA alone.24 Unlike thesupply of donor hearts, the availability ofLVADs is limited only by the industrial capac-ity for production.

Ultimately, long-term mechanical cardiacassistance may prove less expensive than car-diac transplantation or the intensive medical

204



j.com/


eptember 1996. D

ownloaded from



Figure 5 The Jarvik 7pneumatic total artificialheart.

treatment of patients in NYHA classes III andIV. Medically treated patients are admitted tohospital a mean of 2-25 times during their lastsix months of life, at an average cost of£15 000 per patient. This amount does notcover home care, medical staff costs, or lostincome. There are many patients between theages of 60 and 75 who have an expected mor-tality of about 40% per year on conventionalmedical treatment. This is a large and rapidlygrowing population who are in need of analternative treatment.25 For these patients it isimportant to assess whether the use of a per-manent LVAD can improve symptoms andprolong life in the current environment of con-strained health care resources. A prospectiverandomised clinical trial of LVAD versus con-

Figure 6 The Jarvik2000 ventricular artificialheart. The Jarvik 2000Oxford model is implantedinto the apex of the leftventricle and conveys bloodto the descending aortathrough a vascular graft.There is a percutaneouselectrical system attachedto batteries.

ventional medical treatment will answer thisquestion. Meanwhile efforts are directedtowards the development of improvedmechanical devices.The Jarvik 2000 intraventricular artificial

heart provides an innovative approach to thedevelopment of a permanent, fully implantablesystem (fig 6). This miniature, self-containedaxial flow pump-a little larger than thepatient's thumb-is implanted into the apex ofthe failing left ventricle and retained in placewith a cuff sutured to the myocardium. Theelectromagnetic pump rotor lies within a tita-nium-lined motor bore. The rotor is supportedon blood-immersed miniature bearings byinflow and outflow stators. The mechanism issimple, silent, reliable, and efficient with aprojected power requirement in the range of6-8 W for most patients. The Jarvik 2000device weighs 85 g and is 25 cc in volumecompared with the Heartmate pump whichweighs over 800 g and is about 600 cc in vol-ume. It is capable of providing a cardiac out-put of 10 I/min at a mean aortic pressuregreater than 80 mm Hg. Blood entering thedilated left ventricle through the mitral valve iswithdrawn via a Dacron graft to the descend-ing thoracic aorta. The normal anatomicalrelations of flow into and out of the heart areretained but an additional second outlet isprovided across the apex to the descendingthoracic aorta. The pump unloads the naturalleft ventricle and allows the diseased heart toprovide part of the cardiac output and pulsatil-ity. In its apical position the device does notinterfere with the papillary muscles or chordaetendineae and because there are no valves andonly one moving part it has a potential lifespan of several decades. The principal issuesof system reliability involve the durability ofthe blood-immersed bearings, flex fatigue ofthe electric wires, and the reliability of the

205



j.com/


eptember 1996. D

ownloaded from


Westaby

electronic system including the batteries.Preliminary work in Oxford and The TexasHeart Institute suggests that the Jarvik 2000can function free of thrombus formation andwith insignificant heat generation or haemoly-sis. Patient mobility is expected to be entirelynormal and those implanted with this systemwould require only anticoagulation with war-farin (international normalised ratio between2-0 and 3 0). No immunosuppressive agentsare needed and the risks of infection are small.Since an organ donor is not required, the sys-tem should be implanted before the onset ofmultisystem organ failure and might preventfurther deterioration in left ventricular func-tion. This will be the first implantable devicesuitable for use in children.

These and other developments over the pastdecade indicate that mechanical support canprovide effective treatment for congestiveheart failure. There is every likelihood thatmechanical blood pumps will become asapplicable to the treatment of heart failure asthe pacemaker is to rhythm disturbances.

1 Evans RW, Manninen DL, Garrison LP, Maier AM.Donor availability as the primary determinant of thefuture of heart transplantation. JAAfA 1986;255:1982-6.

2 McManus RP, O'Hair DP, Beitziner JM, Schweiger J,Siegel R, Breen TJ, et al. Patients who die awaiting hearttransplantation. Y Heart Lung Transplant 1993;12:159-72.

3 Grattan MT, Moreno Cabrol CE, Starnes VA. Eight yearresults of cyclosporin treated patients with cardiac trans-plants. 7 Thorac Cardiovasc Surg 1990;99:500-9.

4 Gao SZ, Alderman EL, Schroeder JS, Silverman JF, HuntSA. Accelerated coronary vascular disease in the hearttransplant patient: coronary arteriographic findings. JAmColl Cardiol 1988;12:334-40.

5 Frazier OH, Macris MP, Myers TJ, Duncan JM,Radovancevic B, Pamis SM, et al. Improved survivalafter extended bridge to cardiac transplantation. AnnThorac Surg 1994;57:1416-22.

6 McCarthy PM, Portner PM, Tobler HG, Starnes VA,Ramasamy N, Oyer PE. Clinical experience with theNovacor ventricular assist system. J Thorac CardiovascSurg 1991;102:573-81.

7 Baldwin RL, Frazier OH, Radokancevic B, Ford S,Duncan JM. Extended pretransplantation left ventricular

support for the patient with advanced heart failure.Improved results using Thermo Cardiosystems left ven-tricular assist device. J Heart Lung Trans 1991;1:159.

8 Cooley DA, Liotta D, Hallman GL, Bloodwell RD,Leachman RD, Milam JD. Orthotopic cardiac prosthesisfor two staged cardiac replacement. Am _J Cardiol1969;24:723-30.

9 Reemtsma K, Krusin R, Edie R, Bregman D, Dobelle W,Hardy M. Cardiac transplantation for patients requiringmechanical circulatory support. N Engl J Med 1978;298:670-6.

10 Cooley DA, Akutsu T, Norman JC, Serrati MA, FrazierOH. Total artificial heart in two staged cardiac transplan-tation. Cardiovasc Dis Bull Tex Heart Inst 1981;8:305-19.

11 Stames VA, Oyer PE, Portner P, Ramasamy N, Miller PJ,Stinson EB, et al. Isolated left ventricular assist as bridge tocardiac transplantation. _J Thorac Cardiovasc Surg 1988;96:62-71.

12 Hill JD, Farrar DJ, Hershon JJ, Compton PG, Avery GJ II,Levin BS, et al. Use of a prosthetic ventricle as a bridge tocardiac transplantation for post infarction cardiogenicshock. N EnglJ Med 1986;314:626-30.

13 Burch GE, DePasquale NP. On resting the human heart.AmJ Med 1968;44: 165-7.

14 Frazier OH. Chronic left ventricular support with a ventedelectric assist device. Ann Thorac Surg 1993;55:273-5.

15 Graham TR, Dasse K, Coumbe A, Salih V, Marrinan MT,Frazier OH, et al. Neo intimal development on texturedbiomaterial surfaces during clinical use of an implantableleft ventricular assist device. Eur _J Cardiothorac Surg1990;40: 182-90.

16 McCarthy PM. Heartmate implantable left ventricularassist device: Bridge to transplantation and future appli-cations. Ann Thorac Surg 1995;59:S.46-51.

17 Myers TJ, Dasse KA, Macris MP, Poirier VL, Cloy MJ,Frazier OH. Use of left ventricular assist device in an out-patient setting. ASAIO Journal 1994;471-5.

18 Frazier OH, Rose EA, MacMannus Q, Burton NA, LefrakEA, Poirier VL, et al. Multicenter clinical evaluation ofthe Heartmate 1000 IP left ventricular assist device. AnnThorac Surg 1992;53:1080-90.

19 Dasse KA, Chapman SD, Sherman CN, Levine AH,Frazier OH. Clinical experience with textured blood con-tacting surfaces in ventricular assist devices. TransASAIO 1987;23:418-25.

20 DeVries WC. The permanent artificial heart: four casereports. JAA1A 1988;259:849-59.

21 Copeland JG, Levison MM, Smith R, Icenogle TB,Vaughn C, Cheng K, et al. The total artificial heart asbridge to transplantation: a report of two cases. JAMA1986;256:2991-5.

22 Jarvik RK, DeVries WC, Semb BK, Koul B, Copeland JG,Levinson MM, et al. Surgical position of the Jarvik 7 totalartificial heart. Jf Heart Transplant 1986;5: 184-95.

23 United Network for Organ Sharing (UNOS). UNOS update1991;7(May):2.

24 Smith WM. Epidemiology of congestive heart failure. Am JCardiol 1985;55:3.

25 Oz MC, Rose EA, Levin HR. Selection criteria for place-ment of left ventricular assist devices. Am HeartJ 1995;129:173-7.

206



j.com/


eptember 1996. D

ownloaded from


Documents

Theneed for artificial hearts - Heart | A leading