Search tips
Search criteria 


Logo of thijTexas Heart Institute JournalSee also Cardiovascular Diseases Journal in PMCSubscribeSubmissionsTHI Journal Website
Tex Heart Inst J. 2004; 31(3): 289–302.
PMCID: PMC521775
Archival Article



The concept of circulatory assistance is not new.1 The need for such temporary support for hours to days has been recognized for more than a decade; and still exists. The most common form of acute, iatrogenic, potentially reversible circulatory failure in man is that following cardiopul-monary bypass. If means were available to support temporarily the circulation of such patients, the current low mortality associated with open-heart surgery would be further decreased and, perhaps, the indications for high-risk patients broadened. For example, following otherwise uncomplicated procedures it is common for patients to require reinstitution of cardiopulmonary bypass for brief periods to allow the biologic left ventricle to resume its pumping function to support the peripheral and central circulations. There is a correlation between the duration of the necessary period of re-support with the degree of acute left ventricular decompensation encountered. The limitations of prolonged bypass are hemolysis, protein denaturation and the appearance of hemorrhagic diatheses. Cardiopulmonary bypass is a controlled form of shock and applicable for only short periods of time.

The techniques proposed for circulatory assistance have included aortic bypass pumping (vide supra), diastolic intraaortic balloon pumping, counterpulsation with an ectopic ventricle, external regional counterpulsation, synchronous airway pulsation, veno-venous oxygenation, veno-arterial oxygenation and mechanical synchronous sternal compression. All have been utilized in man with varying degrees of success.

The purpose of this communication is to provide an anticipatory refrence responsive to questions regarding the proposed short-term use of an abdominal left ventricular assist device,2,3 and to submit these concepts and principles to peer review. This is appropriate in the current climate of an increasing involvement of medicine, technology, industry and government with society. Clearly, the possible use of this abdominal left ventricular assist device could involve not only the aforementioned sectors, but also impinge on moral, legal and ethical interests. It should be emphasized that, over the last two years, these implications have been given careful consideration. The new National Heart, Blood Vessel, Lung and Blood Act of 1972 reaffirms the responsibility for research and development in this area; there is to be both research into, and establishment of, programs dealing with devices and instrumentation. The details can be found in Sections 413 (A), items 3 and 4. Legislation is not specified.


The abdominal left ventricular assist device (ALVAD) is an implantable blood pump actuated by an external pneumatic power source.4 This pump is a modification of the Model VII left ventricular assist device developed under NHLI sponsorship over the past 6 years. Blood is received from the apex of the left ventricle during systole and ejected into the infrarenal portion of the abdominal aorta during diastole. The pump is of cylindrical design and operates axio-symmetrically. The pump ventricle consists of a polyurethane bladder which collapses when pneumatic pressure is applied to the space between the bladder and the stainless steel casing. Unidirectional flow is imparted by disc valves at the inflow and outflow orifices. In vitro tests of this pump have shown that its output is essentially a linear function of pumping rate up to maximum flows of approximately 13 liters per minute. Two pump sizes have been used for animal experiments, with 65 and 100 ml stroke volumes for dogs and calves, respectively.

All blood contacting surfaces, including the inflow tube, the valve struts and the pump bladder, excluding the valve discs and the outflow tube graft, are “flocked” with a coating of polyester fibrils. These filaments are 10 mils in length and 1 mil in diameter and are applied with a density of 100–150 fibrils per square millimeter. This flocked surface promotes a rapid deposition of fibrin which serves as base for the development of a stable, blood compatible surface. Microscopically, the neointima formed on flocked polyurethane appears to consist of a fibrin lattice containing pyknotic nuclei of fibrocytes. A thin neointima of endothelial-like cells is gradually deposited over this fibrin layer. The observation of lymphocyte aggregates at the blood-biomaterial interface suggests the possibility that the neoendothelial cells originate from circulating leukocytes by a process of redifferentiation.


The pump drive console provides both EKG triggered synchronous pumping and fixed rate asynchronous pumping over a range of 40 to 140 strokes per minute. Variable ejection delay and pulse duration are provided.5 The console has redundant pneumatic and electronic systems. The primary system has a vacuum assisted fill cycle and is capable of operating in either synchronous or asynchronous modes over the complete range of rates. The redundant system provides drive pressures at adjustable fixed rates and pulse durations. A four channel oscilloscope displays EKG, arterial pressure, pump drive pressure and left ventricular pressure. Left ventricular pressure is obtained from an inflow pressure transducer incorporated into the ALVAD. Fail-safe systems are included for: 1) EKG malfunction or arrhythmias; 2) primary system mechanical or electronic failure; 3) loss of pneumatic power; and 4) loss of AC line power.6


Implantation of the ALVAD is accomplished by utilizing a modification of the method developed by Bernhard and LaFarge.7 For acute experiments in dogs, a median sternotomy with midline abdominal extension is used. For chronic experiments in calves, a left thoracotomy and a left transverse laparotomy are preferred. The woven outflow tube graft is sutured end-to-side to the abdominal aorta below the renal arteries. The pump and inflow tube are primed with saline. A Teflon felt sewing ring is sutured to the left ventricular apex. A small incision is made in the apex and a Foley catheter is inserted into the left ventricular cavity utilizing a central stylet. The catheter balloon is inflated and with gentle withdrawal pressure is applied against the apical wall. A circular knife is passed down the catheter and a full thickness segment of apical myocardium is excised. The catheter balloon provides hemostasis. Under a carbon dioxide atmosphere, the balloon is deflated and quickly withdrawn and the inflow tube of the pump is rapidly inserted into the left ventricle. A purse string suture is then drawn tight around the collar of the sewing ring creating a seal around the inflow tube. The pumping chamber resides in the abdomen with its inflow tube traversing the diaphragm. The pneumatic drive line is covered with a sleeve of polyester velour. The drive line and the pressure transducer instrumentation cable are brought out of the abdomen through a separate small incision.


Two advantages result from the abdominal pump placement: 1) pulmonary function is minimally impaired; and 2) removal may be accomplished through the original abdominal incision without re-entering the thorax.


Removal of the ALVAD is performed by re-opening the previously made celiotomy incision. The pump inflow tube is attached to the pump by a rotary quick connect-disconnect fitting. The pump is removed and the inflow tube is left in situ in the ventricular apex, occluded with a special obturator. The outflow tube graft is divided and oversewn.

When viewed objectively, a series of questions can be raised regarding the use in man of such a device. Those which are immediately apparent were first considered two years ago and were included in proposals* to the Division of Technological Applications of the National Heart and Lung Institute.

  1. What is the magnitude of the problem:
    In evaluating the potential value of this abdominal left ventricular assist device (ALVAD), it is necessary to estimate the number of patients in this country who undergo open-heart surgery. The data of Roche and Stengle8 were extrapolated. From these it is reasonable to assume that this year, approximately 50,000 open-heart procedures will be undertaken. Two thousand of these procedures will be performed in our institutions. All of these procedures involve risk; for example, the recent review of aortocoronary saphenous vein bypass by Wilson9 cites operative deaths varying between 0–19% in 15 series of patients with preinfarction angina at different institutions. Mortality rates between 2 and 6% are currently reported for coronary artery and valvular heart disease patients under-going elective surgery.
  2. In what circumstances could this abdominal left ventricular assist device be of use?
    It is within the aforementioned group of patients who are expected to live, but die despite all currently available efforts that the use of this device could be of value.
  3. Do the risks outweigh the benefits regarding the use of this assist device?
    It should be made clear that in the patient population envisioned, the alternatives rapidly diminish until there are no viable options. The techniques of weaning a patient from cardiopulmonary bypass are well established and successful in most instances. Should these attempts fail, various short-term pharmacologic methods are instituted and intensified. During this interval, the degree to which the patient is at risk increases sharply. The time frame for definitive measures is compressed; concomitantly, prospects for survival decrease.
    More specifically, cardiac output and ejection fraction decrease, left and right atrial pressures increase, left and right ventricular end-diastolic pressures increase, left and right end systolic volumes increase. Decreased tissue perfusion, anaerobic glycolysis and metabolic acidosis rapidly ensue. And the heart, as a pump, fails. At this point, the patient is at 100% risk. The benefits of use of the left ventricular assist device prior to, and at this point, are those of substituting a competent, mechanical pump, independent of the progressively deranged physiology of the intrinsic circulation, for the failing (-ed) biologic left ventricle. In this setting, the abdominal left ventricular assist device may be utilized or the patient can expire.
  4. Are there other viable alternatives?
    In the setting described, we have recently utilized intraaortic balloon pumping in a group of patients in similar severe (pre-terminal) straits. During the past year 16 such patients with left ventricular failure following cardiopulmonary bypass were encountered. All underwent intraaortic balloon pumping instituted late in the evolution and pharmacologic treatment of this terminating set of circumstances. All were transiently improved hemodynamically; however, all expired. The reasons for this failure of response were: 1) the intraaortic balloon augments existing circulation, but cannot take over the entire left ventricular output, as can the ALVAD; and 2) intraaortic balloon pumping does not markedly lower left ventricular after-load to decrease significantly left ventricular work, as does the ALVAD. In short, to our knowledge, there are no viable alternatives of equal potential in these particularly difficult preterminal circumstances.
  5. Assuming that there are no viable alternatives, what can be gained or learned from the use of the abdominal left ventricular assist device for nearly irreversible left ventricular failure following cardiopulmonary bypass?
    It is generally conceded that laboratory models can approach but not actually simulate clinical settings. Nonetheless, experience gleaned in the laboratory has direct clinical relevance. Many, if not most, questions have been answered, i.e. when the abdominal ventricular assist device is actuated 1) mean arterial pressure is increased, 2) left ventricular systolic and diastolic pressures are decreased, and 3) left ventricular work is decreased. In addition, there is a decrease in left atrial pressure, left ventricular volume and tension. Myocardial oxygen extraction is decreased. Secondary effects are increased tissue perfusion, decreased metabolic acidosis and decreased pulmonary congestion. In experimentally induced myocardial ischemia, Vmax and Vce are increased. In this manner, the ALVAD functions effectively in vivo as a pump.10,11,12,13,14,15,16,17,18
  6. Are there technical difficulties associated with the implantation of this pump?
    Because of the usual and accepted difficulties associated with cardiopulmonary bypass in experimental animals, all of the experimental techniques which have been developed for implantation of this pump have avoided the use of heart-lung units. Workable, reproducible methods have been evolved to facilitate the implantation procedures in calves and dogs. Over the past two years, approximately 80 acute and 20 chronic experiments have been undertaken with these techniques which involve insertion of the ALVAD inlet tube into the apex of the beating left ventricle and connection of the outlet to the abdominal aorta with the avoidance of an atrial incision, ventricular automaticity, exsanguination and air embolism.
    In comparison, the implantation of this pump in man should be far more easily accomplished, i.e. the patient will have already been cannulated for cardiopulmonary bypass with cannulae in the superior and inferior vena cava, a perfusing cannula in the ascending aorta and decompression of the left atrium and ventricle with an aspirating cannula inserted through the right superior pulmonary vein. Implantation of the pump will require 1) continuation of cardiopulmonary bypass; 2) cross-clamping again of the ascending aorta proximal to the perfusing cannula; 3) insertion of the ALVAD inlet into the empty left ventricle under direct vision; 4) anastomosis of the outlet graft to the easily accessible infrarenal aorta; 5) quick-connection of the pumping chamber to the inlet and outlet beneath the diaphragm; 6) evacuation of air from the system, and 7) actuation of ALVAD pumping while slowly discontinuing cardiopulmonary bypass.
  7. What are the criteria for selection of patients for ALVAD implantation? These are matters of considerable concern;! and it is here that this emerging technology impinges on moral and ethical considerations. In our patient population, those at high risk can be identified preoperatively and consent forms for possible ALVAD utilization obtained. We cannot, however, predict with unerring accuracy which patient will need early or late ventricular assistance. Nor can we predict which patient will or will not respond to pharmacologic support of the circulation following cardiopulmonary bypass. We are, nonetheless, continually aware of patients who have expired who possibly could have benefitted and perhaps recovered, if ALVAD pumping had been available and instituted early following cardiopulmonary bypass. We are also aware of the MIRU experience with intraaortic balloon pumping in which some of the criteria for selection of patients in cardiogenic shock following myocardial infarction were so re-strictive that little or nothing could be offered to the vast majority studied. (Personal communication, Peter L. Frommer, M.D., Associate Director for Cardiology, Division of Heart and Vascular Diseases, NHLI, August 20, 1973.)
  8. What is the state of development of ALVAD pump … (should it be utilized in one circumstance while continuing to develop it, with modi-fications, for other circumstances)?
    The ALVAD modified Model VII pump is the result of one of the contract-funded, goal-oriented programs of NHLI. As such, it incorporates many of the advances of other implantable circulatory support systems under continued development since 1966. Its abdominal configuration is unique, although other systems under development for longer-term use may utilize this anatomic placement and basic design. It has been developed for implantation following cardiopulmonary bypass for this group of patients, with design modifications for rapid and early (hours to days) removal without re-entering the thorax.
    LaFarge and Bernhard have reported similar pumps, being developed for different purposes, operating continually, intrathoracically, for six months, or longer. We have observed calves during uninterrupted intra-abdominal pumping for two months. Inlet tubes have remained in the left ventricular apex for one year with no apparent untoward effects. The complete pump assembly has been implanted, pumped for one week and removed with no deleterious effects. As a result, it does not seem unreasonable to suggest that short-term clinical trials are justifiable in parallel with further development of longer-term (electrical19 and nuclear20) left ventricular assist systems. The major development problems of these latter systems are not germane to short-term utilization discussions. In summary, the currently available technology should allow clinical trials with reasonable reliability and durability for periods up to one week with temporal safety factors of four times that period.
  9. What are the costs of implantation and removal of this ALVAD?
    A survey of two institutions, one in the northeast and one in the south-west, indicates that hospital costs for open-heart surgery are remarkably similar regardless of the type of procedure performed. An average between $5,000–$6,000 is currently applicable. Estimates of the combined professional fees of the surgeons and the anesthesiologists ($1,200)* increase the cost to $6,000–$7,000. These rough calculations are in agreement with those made by the Ad Hoc Task Force on Cardiac Replacement of the National Heart and Lung Institute.21
    Inasmuch as: 1) the left ventricular assist device is designed for implantation during and/or after cardiopulmonary bypass; 2) the basic techniques of monitoring its effectiveness are those routinely utilized following clinical open-heart surgery; and 3) a cost-effectiveness analysis must consider the necessity of one additional person at the bedside during ALVAD pumping similar to intraaortic balloon pumping, a rough prediction of costs would result in an increment of 10–20%o in the above costs for pumping periods of hours to days.
  10. What are the problems of pump dependence?
    This is a difficult and complex problem. There are certain points which deserve emphasis. As presently conceived, the ALVAD would not be implanted unless there were no other viable options. Given this set of circumstances, the initial implantations (Phase I) would be undertaken in pre-terminal patients in the hope that any could be improved or perhaps even retrieved from almost certain death; the alternative to pump implantation being discontinuance of cardiopulmonary bypass and demise. Given this 100% mortality without ALVAD pumping, a 95% mortality with ALVAD pumping would seem justifiable. Phase II trials with earlier implantation in patients somewhat less irretrievable might then be considered with less trepidation.
    Nonetheless, the possibility of pump-dependence is real, somewhat similar to respirator dependence, and perhaps could be considered an acceptable risk commensurate with these intentionally selected terminal circumstances for initial clinical trials. Should dependence occur, one avenue of treatment would remain open, i.e., elective, unhurried transplantation. This latter possibility could, however, involve heroic measures … measures both impractical and unwarranted. The patient, his family, and his physicians would then face a reconcilable situation similar to that faced by others in analogous pre-terminal episodes, e.g. the patient with pre-terminal residua of Guillain-Barré syndrome (encephalomyeloradiculoneuritis) or other motor neuron diseases such as amyotrophic lateral sclerosis and bulbar syringomyelia in which respiration is entirely dependent on mechanical support. Phase III trials would possibly include ALVAD pumping in high-risk patients during and/or after elective open-heart surgery with weaning and removal following the stresses of surgery, i.e. ALVAD-standby-utilization.
  11. What is the reliability and durability of the ALVAD … are there deleterious or toxic effects?
    It must be clearly stated that our view of the readiness of this device for initial clinical trials is restricted to short-term use. Implicit in this position is a recognition that long-term (weeks to months) use requires further development with particular emphasis on valve function and neointima kinetics, areas not germane to short-term use.
    Unlike conventional heart-lung machines now in widespread use, there is no blood-air interface in the ALVAD and consequently no protein denaturation. There is minimal destruction of formed blood elements and consequently little hemolysis. There are no hemorrhagic diatheses with prolonged pumping. There are no toxic materials in contact with circulating blood. The risk of thrombo-embolism in the presence of ALVAD pumping is exceedingly low; following clinical cardiopulmonary bypass, far more problems are encountered due to hypocoagulability than to hypercoagulability.


An implantable left ventricular assist pump has been developed to the point where initial clinical trials can be seriously considered. These latter considerations involve a patient population faced with imminent demise. The pump functions well in vivo, according to design specifications, and can be easily removed. If used clinically in the projected circumstances, legal, moral and ethical questions arise. The potential of life sustaining benefits would seem to outweigh the inherent risks.

figure 22FFU1
Figure. Diagram showing anatomical placement of the ALVAD. The pump is in the abdomen with its inflow tube traversing the diaphragm to accept blood from the left ventricular apex. The outflow of the pump is anastomosed to the infrarenal aorta. The pneumatic ...
figure 22FFU2
Figure. Composite diagram of the experimental set up for hemodynamic studies on the ALVAD. Respiration is maintained with a volume respirator. Catheters in the aorta and coronary sinus allow blood sampling for blood gases and oxygen contents. Transducers ...
figure 22FFU3
Figure. Composite diagram shows placement of catheters and transducers around the heart during an ALVAD experiment. Note the left ventricular pressure catheter introduced by direct puncture of the myocardium. The segment length gauge (SLG) is a mercury-in-Silastic ...
figure 22FFU4
Figure. Results of a set experiments illustrating that ALVAD pumping elevates and phase-shifts aortic plessure with peak pressure occurring during prosthetic systole rather than biologic systole.
figure 22FFU5
Figure. Results of a series of determinations showing the effects of ALVAD pumping in lowering left ventricular systolic pressure.
figure 22FFU6
Figure. Determinations of left ventricular dP/dt as derived from left ventricular pressure curves illustrate a reduction with ALVAD pumping.
figure 22FFU7
Figure. Tension-Time-Index derived from left ventricular pressure wave forms is diminished with ALVAD pumping.
figure 22FFU8
Figure. Mean left atrial pressure is diminished with ALVAD assistance.


* RFP NHLI-MDAP-72-21, April 1972 and Proposal No. TE 2436-200 Continuation of Contract No NO1-HL-73-2946, May 1973

* Mr. Dayton W. Selby, Administrative Office, Washington Greater Metropolitan Blue Shield, 550 12th St., S.W., Washington D.C., Source: Dr. J. K. Roche, NHLI.

This article has been reprinted from Cardiovascular Diseases: Bulletin of the Texas Heart Institute 1974;1:251–64.


1. DeBakey ME: Left ventricular bypass pump for cardiac assistance: Clinical experience. Am J Cardiol 27:3, 1971 [PubMed]
2. Norman JC, Whalen WR, Daly BDT, Migliore JJ, et al: An implantable abdominal left ventricular assist device (LVAD). Clin Res XX (5):855, 1972
3. Migliore J, Robinson W, Fuqua J, et al: An abdominal left ventricular device (ALVAD): Experimental physiologic analysis I. European Soc Exp Surg VIlIth Congress, Oslo:97, May, 1973
4. Hughes DA, Edmonds CH, Igo SR, et al: An abdominal left ventricular assist device: Experimental physiologic analyses. J Sur Res 1974 (in press)
5. Coleman S, Whalen R, Robinson W, et al: A pre-clinical drive console dor pneumatically-powered left ventricular assist device. Clin Res XX(5):854, 1972
6. Robinson WJ, Migliore JJ, Arthur J, et al: Abdominal left ventricular assist device: experimental physiologic analysis II. Transactions, Am Soc Art Int Org XIX:229, 1973
7. Filler RM, Bernhard WF, Bankole M, et al: An implantable left ventricular aortic assist device. J Thorac Cardiov Surg 54:795, 1967 [PubMed]
8. Facilities for Open Heart Surgery in the United States: Distribution, Utilization, and Cost. Dr. JK Roche, NIH, NHLI, Bethesda, Maryland [PubMed]
9. Wilson WS: Aorto-coronary saphenous vein bypass, Heart-Lung II 1:90, 1973 [PubMed]
10. Arthur J, Dove GB, Migliore JJ, et al: A comparison of acute hemodynamic effects of abdominal left ventricular assist device (ALVAD) pumping in the non-failing and failing canine heart. Clin Res XXI (31):401, 1973
11. Dove G, Arthur J, Migliore JJ, et al: End-diastolic segment length (EDSL) and derived muscle length (ML) analyses during abdominal left ventricular assist device (ALVAD) pumping. Clin Res XXI (3):415, 1973
12. Dove G, Migliore JJ, Fuqua JM, et al: An abdominal left ventricular assist device (ALVAD): experimental physiologic analysis III: cardiac output. Proc 26th Annual Conf of Eng Med and Biol 15:148, 1973
13. Galioto FM, Dawson JT, Migliore JJ, et al: Serial echocardiographic assessment of left ventricular function in the calf before and after abdominal device implantation. Fed Proc Am Soc for Exp Biol 32(3):332, 1973
14. Migliore JJ, Arthur J, Fuqua JM, et al: Myocardial arteriovenous oxygen content and derived consumption (qO2) during abdominal left ventricular assist device (ALVAD) pumping. Clin Res XXI (3):438, 1973
15. Migliore JJ, Dove GB, Fuqua JM, et al: An abdominal left ventricular assist device (ALVAD): experimental physiologic analysis IV: myocardial blood flow. Proc 26th Ann conf of Eng Med and Biol 15:149,1973
16. Pirzada FA, Igo SR, Dove GB, et al: An abdominal left ventricular assist device (ALVAD), physiologic analyses XI: Hemodynamic effects. Clin Res XXI (5):953, 1973
17. Fuqua JM, Igo SR, Edmonds CH, et al: An abdominal LVAD: Experimental physiologic analyses V. Clin Res XXI (5):953, 1973
18. Igo SR, Edmonds CH, Migliore JJ, et al: An abdominal LVAD: Effects on contractile element velocity. Clin Res XXII (1):7A, 1974
19. Aldrich JG et al: A totally implantable electrically actuated left ventricular assist device (ALVAD): In vivo testing. Proc 26th Ann Conf Eng in Med and Biol, 1973
20. Hagen KG, et al: Nuclear-fueled circulatory support systems VIII: status of the tidal re-generator engine system. Proc IECEC, Philadelphia, 1973.
21. Cardiac Replacement: Medical, Ethical, Psychological and Economic Implications, October 1969. U.S. Government Printing Office.

Articles from Texas Heart Institute Journal are provided here courtesy of Texas Heart Institute