REVIEW ARTICLE

John Youngberg, MD
Section Editor

Anesthesia for Left Ventricular Assist Device Placement

Berend Mets, MB ChB, PhD, FRCA, FFA(SA)

HISTORY
T HE FIRST successful use of a left ventricular assist device
(LVAD) as a bridge to transplantation occurred in 1985.
Today, there is little information describing the anesthetic
Experimental bridging devices were used at the Texas Heart
Institute in 1968, 1978, and 1981; however, all the patients died
management for LVAD placement. Patients presenting for this of sequelae secondary to sepsis before a donor heart became
operation have long-standing cardiac failure and often have available. The first successful bridge to transplantation by Oyer
associated hepatic and renal impairment, which may signifi- was performed (using a Novacor LVAD, Baxter Healthcare,
cantly alter the pharmacokinetics of administered drugs and Oakland, CA) in 1985,5 whereas the first pneumatically driven
render patients coagulopathic. These critically ill patients may TCI (Thermo Cardiosystems, Woburn, MA) LVAD was im-
be treated with cardiac drugs (vasodilators and antiarrhythmics) planted in a patient in 1986 and the subsequent electric version
that may have potentially significant interactions with the in 1991 by Frazier at the Texas Heart Institute.5,6 The cost of an
administration of anesthetics and the stress of surgery. The LVAD is device-dependent, ranging from $17,000 to $50,000.7
anesthesiologist should be prepared to manage cardiac decom- The cost of LVAD implantation and maintenance, although
pensation and acute desaturation before institution of cardiopul- considerable, is less than the cost of maintaining a status 1
monary bypass (CPB) as well as right ventricular failure and candidate for heart transplantation for 75 days, the median
severe coagulopathic bleeding after CPB. This article discusses waiting time in 1992.5 Currently, 91% of patients fitted with an
the anesthetic management of these challenging patients in the assist device for severe cardiac failure can be discharged from
light of the aforementioned concerns and describes the utility of the hospital, and 74% survive to undergo transplantation.8 There
transesophageal echocardiography (TEE) in patient manage- are indications that LVAD patients subsequently transplanted do
ment. better than those not subjected to the LVAD bridge.9 This
contention is supported by the authors through the demonstra-
tion of physiologic improvement in transplant candidates after
LEFT VENTRICULAR ASSIST DEVICES LVAD placement (as a bridge to transplantation) and better
survival after transplantation (16 of 20) when this group was
With the advent of the first heart transplant performed by
compared with historical transplant controls (3 of 6) with
Barnard in 1967,1 there has been a need for mechanical cardiac
preoperative criteria similar to the LVAD group ( p , 0.05).9
assist devices either as a bridge or as an alternative to
The vented-electric LVAD, powered by 2 batteries (providing 6
transplantation.2 Since the time of the first heart transplant,
to 8 hours of charge time), has given patients the potential for
anesthesiologists have refined anesthetic techniques to deal with
normal levels of activity outside of a hospital setting. One-year
patients in end-stage cardiac failure.3,4 To date, however, there
survival rate is 70%.9 Currently a multicenter trial is in progress
is a paucity of data on anesthetic techniques and considerations
(REMATCH trial [Randomized Evaluation of Mechanical Assis-
specifically for LVAD placement. The author incorporates into tance for the Treatment of Congestive Heart Failure]) to
this review what published information there is and draws compare the use of LVAD placement as an alternative to
heavily on the experience gained at the New York Presbyterian medical therapy in end-stage cardiac failure.10
Hospital.

PROCEDURE FOR LEFT VENTRICULAR ASSIST

DEVICE PLACEMENT
From the Department of Anesthesiology, College of Physicians and
Surgeons of Columbia University, New York Presbyterian Hospital, Indications
New York, NY. In transplant candidates, hemodynamic criteria for device
Address reprint requests to Berend Mets, MB ChB, PhD, FRCA, insertion usually include inotropic or intraaortic balloon pump
Department of Anesthesiology, PH5 Stem, 630 West, 168 St, New York
support (or both) and a systolic blood pressure less than 80
Presbyterian Hospital, New York, NY, 10032.
Copyright r 2000 by W.B. Saunders Company
mmHg. In one series, all patients had a cardiac index less than 2
1053-0770/00/1403-0019$10.00/0 L/min/m2 and a pulmonary capillary wedge pressure greater
doi:10.1053/cr.2000.5842 than 20 mmHg.11 Possible contraindications to LVAD place-
Key Words: heart failure, pharmacokinetics, heart assist devices, left ment include infection, renal failure (serum creatinine . 5.0),
ventricular assist device severe liver impairment, stroke, severe pulmonary dysfunction,

316 Journal of Cardiothoracic and Vascular Anesthesia, Vol 14, No 3 (June), 2000: pp 316-326
ANESTHESIA FOR LVAD PLACEMENT 317

and insulin-dependent diabetes mellitus.7,9 Because most of the measures 11.2 cm 3 4.0 cm and is made of titanium and lined
initial morbidity after LVAD placement involves infection7 and by a diaphragm with a textured polyurethane surface, which
most chronic cardiac failure patients are infected,12 patients are divides the pump into two halves and delivers a stroke volume
not accepted for LVAD placement unless blood cultures are of 85 mL (Fig 1). Porcine xenografts are present in the inflow
negative (especially for fungus) at least 1 week before device and outflow to ensure unidirectional flow. The textured surface
placement.7 Although after LVAD placement mild renal insuffi- obviates the need for long-term anticoagulation because there is
ciency secondary to long-standing cardiac failure is often a low risk (2.7%) of thromboembolism.14
reversed, because of improved renal blood flow, severe renal During surgery, before anticoagulation with heparin for CPB,
impairment remains a contraindication. This is because fluid an infradiaphragmatic preperitoneal pocket is created for the
management postoperatively in LVAD patients is often compli- LVAD (it can also be placed intraperitoneally),13 then the aorta
cated. The need for hemodialysis with associated decreased and right atrium are cannulated, and CPB is instituted. Next, a
renal perfusion (more likely if there is severe preoperative renal left ventricular vent is placed through the right superior
impairment) and further renal damage can however be avoided
pulmonary vein, and without cardioplegia or aortic cross-
if there is early institution of venovenous or arteriovenous
clamping, device placement is initiated.13 The LVAD inflow
hemofiltration.11 Liver impairment (and the resultant coagulop-
cannula is brought through the diaphragm and plugged into a
athy) is a contraindication because the already significant
polytetrafluoroethylene (Teflon) cuff in the left ventricular apex
morbidity from postoperative bleeding is enhanced in these
(Fig 2). The LVAD outflow graft is secured to the right lateral
patients.2 Severe respiratory impairment is also a contraindica-
tion because these patients may develop oxygen desaturation aspect of the ascending aorta. The device is operated in a fixed
more easily perioperatively (with accompanying hypoxic pulmo- or automatic mode. In the latter mode, which is more physi-
nary vasoconstriction), leading to pulmonary hypertension, ologic, the pump fills passively to 90% capacity and is then
which can further compromise right ventricular function. activated by electric motor or pneumatically. Blood is ejected
A screening scale has been developed in an attempt to define by a pusher plate mechanism. Before weaning off CPB, the
better patient suitability for LVAD implantation.11 Preoperative LVAD is hand cranked using a manual device with the patient in
parameters were assigned weights according to their relative a head-down position to ensure de-airing of the LVAD, minimiz-
risk for early postoperative mortality as an outcome. The ing the potential for cerebral air embolism.13 When the LVAD is
preoperative risk factors (with relative weights) were urine operational, the left atrium and left ventricle act as conduits for
output less than 30 mL/hr (3), central venous pressure greater blood, which drains through the ventricular apex into the LVAD
than 16 mmHg (2), mechanical ventilation (2), prothrombin pump. Blood is then pumped into the ascending aorta to the
time greater than 16 seconds (2), and reoperation (1). It was systemic circulation. Because of the pressure gradient across the
found that LVAD recipients with a score of greater than 5 aortic valve, a competent aortic valve remains closed. After
usually die.11 The authors believed that this screening scale CPB, the LVAD stroke volume, heart rate, and blood flow are
might be used as a starting point for discussing the suitability of prominently displayed on a digital monitor, allowing continu-
a patient for LVAD placement. ous assessment of LVAD function. Although initial LVAD
design was based on a pneumatic system that required recover-
Surgery ing ambulating patients to push a 73-lb console, advances in
The LVAD, the Novacor (Baxter Healthcare, Oakland, CA) design and the incorporation of an electric motor allow patients
or the Heartmate (Thermo Cardiosystems, Woburn, MA) is to leave the hospital and resume work using a battery pack small
implanted through a median sternotomy.2,13 The TCI-LVAD enough to be worn by the patient (Fig 2).5,15

Fig 1. TCI-Heartmate left ven-

tricular assist device. A cross sec-
tion and the location of unidirec-
tional porcine xenografts in the
device inflow and outflow are
shown. (Reprinted with permis-
sion from Thermo Cardiosys-
tems Inc.)
318 BEREND METS

improve cardiac performance because their hearts are often on

the flat or even descending limb of the Starling curve, and
preload reserve is exhausted.17 In contrast, increases in afterload
can decrease stroke volume dramatically (Fig 3) and thus
cardiac output. These patients with fixed low stroke volumes are
dependent on an increased heart rate to maintain cardiac output.
This is because slowing of the heart rate does not result in the
usual increase in stroke volume because preload reserve is
exhausted. Furthermore, increases in heart rate do not increase
cardiac output because the accompanying decreased diastolic
filling time tends to further decrease the stroke volume.
Because decreasing afterload can augment cardiac output,
these patients may be maintained on vasodilatory angiotensin-
converting enzyme (ACE) inhibitors to decrease afterload and
diuretics to minimize volume overload.18,19 The patients may
also be taking amiodarone in combination with an ACE
inhibitor. This combination of therapy may be associated with a
decrease in 1-year mortality as well as a decrease in the
incidence of sudden death in end-stage cardiac failure. This
Fig 2. Schematic of a TCI-Heartmate left ventricular assist device decrease was shown in a review of 737 consecutive patients
plugged into the left ventricular apex, located in a preperitoneal referred for heart transplantation and discharged home on
pocket and connected to the ascending aorta. The diagram shows the medical therapy from 1986 to 1993. An improvement in
use of a battery pack containing rechargeable batteries (worn as a
mortality was shown when captopril (86% of patients compared
shoulder holster) providing 4 to 6 hours of charge. (Reprinted with
permission from Thermo Cardiosystems Inc.) with 46% of patients before 1989) replaced the hydralazine and
isosorbide dinitrate combination or another ACE inhibitor, and
amiodarone was used instead of class 1 antiarrhythmic drugs
PATIENT, PHYSIOLOGY, AND PHARMACOLOGY (53% of patients after 1990 v 10% before 1989 on amiodarone).
The anesthesiologist is faced with a critically ill patient in The total 1-year mortality rate decreased from 33% before 1989
heart failure, usually receiving inotropes or an intra-aortic to 16% after 1990 ( p 5 0.0001), and sudden death decreased
balloon pump,16 often presenting for a repeat cardiac opera- from 20% to 8% ( p 5 0.0006).20 The anesthesiologist should be
tion.7,16 The patient may be suffering from associated end-organ aware of the potential for adverse interactions when patients
dysfunction secondary to cardiogenic shock, such as renal and taking ACE inhibitors21 or amiodarone22 (or both) are subjected
hepatic impairment or pulmonary insufficiency, and may be to surgical stress and anesthesia. These therapeutic agents
infected.7,12 Patients presenting for LVAD placement usually inhibit neurohumoral pathways crucial to maintaining blood
have dilated cardiomyopathy, ischemic cardiomyopathy, or pressure and adrenergic innervation. Vasoconstriction (Fig 4) to
congenital heart disease with an ejection fraction of approxi- support blood pressure is normally achieved through the
mately 15%. As such, these patients have a relatively fixed low interplay of three neurohormonal systems: the sympathetic
stroke volume and depend on appropriate preload and heart rate system, the arginine vasopressin system, and the renin angioten-
to maintain marginal performance. Increasing preload does not sin system (RAS).23,24 Usually, blockade of one or two of these

Fig 3. Schematic of left ven-

tricular (LV) pressure volume
loops comparing the normal
heart with that of a patient with
end-stage cardiac failure. The
heart with end-stage cardiac fail-
ure requires a greater preload
volume for ejection, and its
stroke volume at this preload is
impaired. Small increases in af-
terload resulting in an intraven-
tricular pressure maximum
change from 90 to 100 mmHg in
the end-stage cardiac failure
heart cause a dramatic decrease
in stroke volume (arrow in
dashed pressure volume loop)
when compared with an intra-
ventricular pressure maximum
change from 120 to 145 mmHg in
the normal heart. (Adapted and
reprinted with permission.17)
ANESTHESIA FOR LVAD PLACEMENT 319

Fig 4. Neurohumoral systems

responsible for maintaining
blood pressure (BP) through va-
soconstriction by receptor stimu-
lation: AT1, angiotensin receptor;
V1, vasopressin receptor; a1, ad-
renergic receptor. Dashed arrow
indicates important factors re-
sulting in endocrine organ stimu-
lation of vasoconstrictor produc-
tion. ANG II, angiotensin II; ACE-I,
angiotensin-converting enzyme
inhibitor; AVP, arginine vasopres-
sin. (Reprinted with permis-
sion.33)

systems is compensated for by the others.24,25 During cardiac through the myocardium and pericardium, which is independent
surgery, however, the chronic use of amiodarone, which blocks of neural, humoral, or circulatory effects.37 Right ventricular
the sympathetic system by noncompetitive a-blockade and function in this setting may be compromised further by
b-blockade,26,27 has been associated with fatal vasodilatory elevations in pulmonary vascular resistance.38
shock28 and the need for massive doses of vasopressors.29 Cardiovascular decompensation during induction or mainte-
Similarly, ACE inhibitors, which block the RAS and decrease nance of anesthesia can occur through the decreased sympa-
cardiovascular responsiveness to catecholamines,30 have been thetic outflow associated with induction of anesthesia4,39 with
associated with increased vasoconstrictor requirements during resultant decreased vasoconstriction (especially when the RAS
CPB.31 In one study, the use of ACE inhibitors has been shown system is blocked by ACE inhibition),40 resulting in decreases
to be an independent predictor of vasodilatory shock postby- in blood pressure,39 decreasing diastolic coronary perfusion.
pass.32 These data suggest that the combined use of amiodarone Alterations in preload or afterload or depression of myocardial
and an ACE inhibitor might result in severe hemodynamic contractility and dramatic increases in systemic vascular or
compromise during cardiac surgery.33 Another study suggests
pulmonary vascular resistance may also be deleterious.17 De-
that amiodarone therapy for more than 4 weeks preoperatively
creases in heart rate are especially deleterious because patients
is associated with higher mortality in heart transplant recipients,
cannot compensate by increasing stroke volume.17
although the cause of this was not established.34
The cardiac failure patient with associated hepatic and renal
Patients with cardiac failure have high circulating catechol-
impairment may have significantly altered pharmacokinetics for
amines, which is a response by the body to stimulate the failing
myocardium and maintain cardiac output.35 Increases in plasma administered drugs.41 Cardiac failure results chiefly in a de-
norepinephrine levels correlate with the degree of left ventricu- creased volume of distribution and clearance for most drugs.
lar dysfunction.36 Thomas and Marks36 showed in 16 patients This usually results in drug concentrations being higher in these
that plasma norepinephrine levels were significantly higher in patients when conventional doses are administered.42 The
patients with class III and IV New York Heart Association decreased volume of distribution may be due to altered distribu-
functional cardiac status than patients in functional class II. tion of perfusion or altered protein binding. Decreased clear-
These investigators36 showed a highly significant correlation of ance of drugs is likely due to altered metabolism by the liver43-45
plasma norepinephrine levels with indices of impaired ventricu- and elimination by the kidney.44,46
lar function (preejection period and left ventricular ejection In cardiac failure, hepatic clearance of drugs may be impaired
time) (r 5 0.67, p , 0.01). by the following44:
In patients with long-standing left-sided cardiac failure, right 1. Decreased blood flow to the liver secondary to a
ventricular impairment (systolic and diastolic function) devel- decreased cardiac output and blood shunting to more
ops through ventricular interdependence, defined as the forces favored organs as well as intrahepatic shunting second-
that are transmitted from one ventricle to the other ventricle ary to circulating catecholamines47
320 BEREND METS

2. Decreased hepatic metabolic capacity from hepatocel- artery pressure resulting from enhanced pulmonary vascular
lular damage resulting from liver congestion45 and resistance versus that resulting from left ventricular failure.4
secondary to intrahepatic hypoxia, which impairs the This gradient gives some indication of the need for administra-
metabolism of microsomal drugs, such as lidocaine.48 tion of pulmonary vasodilatory agents when weaning from
The effect of cardiac failure on the pharmacokinetics of bypass is attempted. Continuous cardiac output catheters52 may
lidocaine was shown by the classic work of Stenson et al,49 who be useful in these patients in that after LVAD placement these
showed that lidocaine elimination from a bolus dose and catheters allow continuous assessment of right-sided cardiac
continuous infusion is roughly inversely proportional to cardiac output, while the LVAD monitor continuously displays left-
index. sided flow. Discrepant values may indicate aortic regurgitation
In cardiac failure, renal drug elimination may be impaired as or intracardiac shunting. Continuous mixed venous oxygen
a result of the following: saturation monitoring, which tracks oxygen transfer in the
1. Decreased glomerular filtration rate secondary to hy- lungs, cardiac output, and peripheral oxygen uptake,53 may be
poperfusion of particular use in determining the adequacy of cardiac output
2. Increased tubular reabsorption owing to redistribution in LVAD patients. Mixed venous oxygen saturation measure-
of renal blood flow.44 ment showed better reproducibility as well as correlation with
Increased tubular reabsorption may be especially important increasing doses of milrinone in cardiac failure patients than
when the parent compound (procainamide) is metabolized to cardiac output determinations by thermodilution in the setting
active metabolites (N-acetyl-procainamide), which are more of tricuspid regurgitation, which is a common problem in LVAD
dependent on renal elimination than the parent compound, patients.54
resulting in an increase in metabolite-to-parent drug ratio if Anesthetic induction. Rapid hemodynamic decompensa-
renal failure occurs.46 tion may occur in these patients if there is a sudden decrease in
Hypovolemic shock has similar effects on drug pharmacoki- left ventricular preload or increase in left ventricular afterload
netics to cardiac failure.43 Blood volume is decreased, a1-acid (Fig 3) or if pulmonary vascular resistance or heart rate
glycoprotein is increased (trauma or surgery), and hepatic increases from undue surgical or intubation stress, hypoxia, or
clearance of drugs is significantly impaired, secondary to hypercarbia.17 Because end-stage cardiac failure patients rely
decreased hepatic perfusion, intrahepatic shunting, and hy- on high circulating catecholamines to maintain vasoconstric-
poxia. In experiments involving a 30% blood loss in primates, tion,35,36 decompensation during induction or maintenance of
lidocaine volume of distribution, clearance, and hepatic extrac- anesthesia can occur through reduction in sympathetic outflow
tion were decreased, whereas elimination half-life was pro- associated with anesthesia.4,39 An important concern is to ensure
longed. Steady-state infusion of lidocaine resulted in higher unconsciousness in these critically ill patients, while maintain-
blood concentrations.50 ing cardiac stability, because cardiac surgical patients may
The aforementioned issues are important because they may experience recall after anesthesia.55 Incorporating an end-tidal
lead to potential complications of therapy. Common examples isoflurane concentration of 0.4% as part of the anesthetic
of inappropriate pharmacotherapy are the occurrence of resis- technique may avoid this problem because it is associated with
tant vasodilation in a patient with renal insufficiency adminis- loss of implicit and explicit memory.55 This low concentration
tered a standard infusion of milrinone because this drug is of isoflurane is unlikely to have significant hemodynamic
predominantly renally eliminated51; or cardiovascular depres- effects because these are dose-dependent, whereas isoflurane
sion from lidocaine administration in a patient in shock.50 maintains cardiac output better than halothane and enflurane56
and as well as sevoflurane and desflurane.57 There is new
ANESTHETIC MANAGEMENT
evidence that volatile anesthetics may be beneficial in this
Preoperative assessment of the patient should be performed context in that they confer some measure of protection against
to exclude neurologic deficits7 and to evaluate cardiac status, ischemic reperfusion injury in vitro.58 Finally, the anesthesiolo-
renal function, hepatic function as well as coagulation status gist should be mindful of the effect that cardiac failure may have
and current drug therapy. The anesthesiologist should be on the pharmacokinetics of administered drugs.41
prepared to manage cardiac decompensation and acute desatura- To achieve the aforementioned goals, the anesthesiologist
tion before institution of CPB as well as right ventricular should maintain previous rates of inotrope and vasoactive drug
failure38 and severe coagulopathic bleeding16 after CPB. LVAD infusion and optimize intra-aortic balloon pump function before
patients have a high incidence of infection12 so that appropriate induction of anesthesia.59 The patient may be in severe respira-
antibiotic prophylaxis is imperative. tory distress and must be allowed to sit in a position that allows
him or her to be comfortable; oxygen should be administered;
Conduct of Anesthesia and oxygen saturation should be monitored, while two large-
Monitoring. Standard cardiac hemodynamic monitoring is bore intravenous catheters and the arterial catheter are placed.
used. A pulmonary artery catheter is usually placed before After adequate preoxygenation and reassurance of the often
induction of anesthesia in the severely compromised patient to extremely anxious patient, the anesthetic induction can be
allow rapid monitoring of hemodynamic deterioration should performed. It is the author’s practice to infuse low doses of
that occur. At the time of pulmonary artery catheter placement, norepinephrine (Sanofi Winthrop Pharmaceuticals, New York,
it is useful to establish the transpulmonary pressure gradient NY) at the time of induction to compensate for the drop of
(mean pulmonary artery pressure 2 pulmonary capillary wedge elevated catecholamine concentrations35,36 from a decreased
pressure) to determine the component of mean pulmonary sympathetic output associated with anesthetic administra-
ANESTHESIA FOR LVAD PLACEMENT 321

tion.4,39 In similar circumstances in which anesthetic manage- assist device (RVAD) placement. Of the 22 patients who
ment for heart transplantation is provided, others use phenyleph- suffered right-sided circulatory failure, 21 died. Right-sided
rine for this purpose.39 The administration of these vasoactive circulatory failure occurred in 6 of 28 patients (21%) who
agents can be adjusted to maintain preload and afterload at received LVADs as a bridge to transplantation in another study.9
suitable levels, while allowing the administration of an adequate In this study, patients having this complication had a signifi-
depth of anesthesia. The intention here is to maintain blood cantly lower heart transplant and survival rate.9 Perioperative
pressure to ensure adequate systemic perfusion as assessed by vitamin K administration may decrease bleeding in LVAD
urine output and the absence of metabolic acidosis on blood gas recipients.67 In a study of 66 LVAD recipients, 39 of whom
analysis rather than to achieve normal hemodynamic param- received vitamin K, 25.9% not treated with vitamin K versus
eters.39 A mean arterial pressure of 70 mmHg is acceptable.39 A 5.1% of patients treated with vitamin K required reexploration
cardiostable anesthetic induction can be performed in end-stage for bleeding 48 hours after the operation ( p 5 0.026).67
cardiac failure using etomidate, ketamine, fentanyl, midazolam, In addition to the above-mentioned pharmacotherapeutic
or sufentanil.60-62 Thiopental (Pentothal) was used for the first approaches to minimize bleeding, preparation for substantial
heart transplant.3 The anesthesiologist should be mindful of the bleeding postbypass needs to be made. Preoperatively, two
slow circulation time and the reduced volume of distribution for wide-bore intravenous catheters should have been established.
drugs found with cardiac failure.59 Incremental doses should be An autologous blood recovery unit (Cell Saver, Haemonetics
administered, allowing adequate time to assess the hemody- Co, Braintree, MA) decreases the need for exogenous blood
namic effect of the administered agent.39 A hemodynamically administration, whereas a rapid blood infusion device (System
stable induction can be performed using a small dose of 500, Level 1 Technologies Inc, Rockland, MA) may be useful.
midazolam, fentanyl, and etomidate. Muscle relaxation can be Blood product administration should be guided by regular
achieved using cisatracurium. Overventilation of the patient’s coagulation profile assessment to ensure the appropriate admin-
lungs, once muscle relaxation has been achieved, should be istration of fresh frozen plasma, platelets, blood, and, if
avoided because this decreases venous return, whereas the necessary, cryoprecipitate. To avoid hypothermia, all adminis-
induced hypocarbia decreases circulating catecholamines, caus- tered fluids should be warmed, airway humidification should be
ing further hypotension.63 A rapid-sequence induction, per- applied, and a warming blanket or a convection heating system
formed because of concerns for a possible full stomach, may be (or both) needs to be used. Once bypass has been established,
hazardous if hypoventilation or apnea results in hypercarbia or the anesthesiologist should be aware that the heart will not be
hypoxia because this may cause pulmonary hypertension and arrested, and preparation should be made for managing right-
compromised left ventricular filling and output. Intermittent sided heart failure, which occurs in 30% of patients.9,68 Further
ventilation while maintaining cricoid pressure at the time of problems postbypass are vasodilatory shock69 and cardiac
induction might be a feasible alternative to minimize the chance arrhythmias.2
of pulmonary aspiration.4 The complex effects of CPB on the perioperative pharmaco-
Anesthetic Maintenance. Anesthesia can be maintained kinetics of administered drugs70-72 are reflected in changing
with agents of the anesthesiologist’s choice. The issues of blood concentrations, which also depend on the particular
awareness, altered drug pharmacokinetics, cardiostability, and characteristics of the drug in question. From the above, it is
appropriate ventilation mentioned earlier should be borne in clear that steady-state concentrations of drugs given prebypass
mind. Perioperative bleeding is a major concern in these may have changed when attempting to come off bypass. A good
patients. This bleeding is multifactorial in origin (eg, hepatic example is the effect of CPB on digoxin levels still often
impairment, extensive surgery, effects of CPB) but may be administered in cardiac failure patients. Therapeutic digoxin
added to by excessive fibrinolysis, thrombogenesis, and platelet levels before bypass have been shown to decrease on bypass
activation that occur with mechanical support devices.64,65 The (approximately 30%) and return to within the range of preby-
benefit of aprotinin (Bayer Corp, West Haven, CT), a serine pass levels after bypass, when there is reduced renal clearance
protease inhibitor that has platelet-preserving abilities, has been and possibly enhanced myocardial sensitivity.70,73,74 The study
addressed in a retrospective review of 142 patients (42 receiv- of Coltart et al74 is illustrative. Digoxin levels in 11 patients
ing aprotinin, 100 untreated) who underwent LVAD place- treated long-term (.5 years) with digoxin were 1.7 ng/mL
ment.16 In these patients, after a 1.4-mg test dose (just before prebypass, 1.1 ng/mL after 2 hours of bypass and at 4 hours
bypass, to exclude an anaphylactoid reaction66 ), 280 mg of after bypass, and 1.7 ng/mL 16 hours after bypass. The patients
aprotinin was infused (over 20 minutes) as well as added to the had received no digoxin for 24 hours preoperatively. Although
pump prime. After this, until the conclusion of surgery, aproti- there are no data on the pharmacokinetics of dobutamine or
nin was infused at a rate of 70 mg/hr.16 Aprotinin administration dopamine on CPB, it has been found that amrinone has a
was associated with decreased chest tube drainage (first 24 smaller volume of distribution and clearance than in volunteers;
hours), 1.5 L versus 4 L, as well as packed red cell administra- however, these differences were considered unlikely to be
tion (4.8 v 8.2) and total blood product units needed (13 v 22) to clinically significant.75 CPB did not affect the pharmacokinetic
achieve similar postoperative hemoglobin levels. This associa- profile significantly probably because amrinone has limited
tion was despite a higher percentage of patients with previous protein binding (21%) and a low hepatic intrinsic clearance.75 A
sternotomy in the aprotinin group.16 An important finding was pharmacokinetic study of increasing bolus doses—25, 50, and
the fact that LVAD patients not receiving aprotinin had an 75 µg/kg of milrinone given on bypass and 50 µg/kg postby-
increased incidence of right-sided circulatory failure (18% v pass—concluded that the pharmacokinetic profile was little
9.5%), defined by the authors as the need for right ventricular affected by CPB.76 Das et al77 concluded that the pharmacoki-
322 BEREND METS

netic profile of milrinone during cardiac surgery was not

different from patients in chronic cardiac failure.

INTRAOPERATIVE TRANSESOPHAGEAL
ECHOCARDIOGRAPHY
TEE is a key monitoring and diagnostic modality in patients
having LVAD placement (Table 1).78-82

Prebypass
After intubation, TEE can be used to optimize left ventricular
filling in an attempt at maintaining cardiovascular stability.
Then a thorough cardiac evaluation should be performed to
ascertain whether a patent foramen ovale83,84 is present, using
an agitated saline injection. If a patent foramen ovale (or atrial
septal defect) is found, this needs to be repaired because
potentially severe right-to-left shunting, not formerly present, Fig 5. Intraoperative echocardiogram in a patient with a left
ventricular assist device (LVAD) shows significant aortic regurgita-
could occur on activation of the LVAD.2 The absence of aortic
tion into the left ventricle when the LVAD was functioning. There was
insufficiency needs to be established. Aortic insufficiency (Fig a discrepancy between the thermodilution cardiac output (2.3 L/min)
5) results in blood ejected by the LVAD into the aorta and the measured LVAD flow of 4.0 L/min after separation from
regurgitating back through the incompetent valve into the left cardiopulmonary bypass. RV, right ventricle; LA, left atrium.
ventricle. A circuit of blood flow is set up with a resultant
decrease in systemic output.2 There is a discrepancy between On Bypass
thermodilution-measured cardiac output and the flow measured
TEE (Table 1) can be used to ensure that the inflow cannula to
by the LVAD pump displayed on its digital monitor. A
the LVAD is appropriately directed away from the left ventricu-
thermodilution cardiac output of 2 L less than the registered
lar septum (to avoid inlet occlusion) and toward the mitral valve
LVAD flow suggests severe aortic regurgitation and the need for
and to ensure that the position allows the left ventricle to drain
aortic valve replacement placement. TEE can be used to assess
completely (Fig 6).79 Once the LVAD is activated and the
this problem and diagnose mitral stenosis (which would limit
de-airing protocol is instituted, TEE can be used to monitor
LVAD inflow) so that surgical repair of these lesions can be
completeness of de-airing by monitoring for echo-dense fireflies
instituted intraoperatively.7 Before CPB, right ventricular func-
in the left atria, ventricle, and aorta.78 Air coming through the
tion and the extent of tricuspid regurgitation should be estab-
device is seen best in the descending aorta because the proximal
lished as a baseline for postbypass evaluation.
aorta and arch may be visualized poorly and the outflow graft is
secured to the right lateral aspect of the ascending aorta.
Table 1. Transesophageal Echocardiography During Left Ventricular
Assist Device Placement

Prebypass
Optimize left ventricular filling
Exclude
PFO
Aortic insufficiency
Mitral stenosis
Monitor
Right ventricular function
Determine baseline tricuspid regurgitation (if present)
On Bypass
Appropriate inlet cannula orientation (oriented to the mitral valve)
Device activation
Deairing, visualize aorta (ascending and descending)
Exclude right-to-left atrial shunting
Monitor decompression of LV and LA
Exclude aortic regurgitation
Postbypass Fig 6. Intraoperative echocardiogram in a patient with a left
Monitor: right ventricular function and tricuspid regurgitation ventricular assist device (LVAD) shows the inflow cannulae (INFLOW)
Monitor: decompression of LV and LA at the left ventricular apex sited toward the mitral valve (MV) and
Monitor: possible air entrainment if LV collapses and subatmo- away from the septum. A dilated right ventricle (RV) and opacifica-
spheric intradevice pressures occur tion of a dilated right atrium from agitated solution, administered to
Doppler-determined LVAD flows exclude a patent foramen ovale postbypass, are shown. The left
atrium (LA) and left ventricle (LV) are decompressed and the intra-
Abbreviations: PFO, patent foramen ovale; LV, left ventricle; LA, left atrial septum shifted to the left, suggesting appropriate functioning
atrium; LVAD, left ventricular assist device. of the LVAD. An echogenic pulmonary artery catheter (PAC) is visible.
ANESTHESIA FOR LVAD PLACEMENT 323

Removal of air is crucial before weaning from bypass because additional inotropic support or nitric oxide89 is required postby-
one of the potential problems on coming off bypass is further pass. Adjustments to the doses administered are then made from
impairment of right ventricular function secondary to right close monitoring using the pulmonary artery catheter and TEE
coronary air emboli from retained intracardiac air. The possibil- postbypass. In this institution, dobutamine (Eli Lilly & Co,
ity of right-to-left shunting, not apparent earlier, should again be Indianapolis, IN) and amrinone90 or milrinone (Winthrop Pharm,
looked for, and left atrial and left ventricular decompression New York, NY) is used to support right ventricular function and
should be noted as increasing flow is generated by the LVAD to decrease pulmonary vascular resistance. Norepinephrine is
(Fig 6).79 Aortic regurgitation should be excluded as left used to maintain systemic vascular resistance, and vasopressin
ventricular end-diastolic pressure is now decreased and a larger (American Regent Lab, Shirley, NY) is added if required.69
pressure drop than before across the aortic valve is created Vasopressin (0.1 U/min) administration in LVAD patients
(Fig 5). (n 5 5) receiving greater than 8 µg/min of norepinephrine who
had mean arterial pressures less than 70 mmHg, despite LVAD
Postbypass flows of 2.5 L/min/m2 postbypass, resulted in rapid (,15 min)
On weaning from CPB, TEE becomes a key monitor of right increases in mean arterial pressure (57 6 4 to 84 6 2 mmHg)
ventricular function and tricuspid regurgitation, which may and systemic vascular resistance (813 6 113 to 1188 6 87
render thermodilution cardiac output inaccurate.85 The contin- dyne · sec · cm25 ) without a significant change in cardiac index.
ued decompression of the left atrium and ventricle needs to be This improvement was associated with a decreased requirement
monitored (Fig 6). A possible late complication that may occur for norepinephrine administration (26.6 to 10.7 µg/min).69
if the left ventricle collapses and the LVAD inlet becomes If pulmonary vascular resistance remains elevated postby-
occluded is entry of air through the generation of subatmo- pass, nitric oxide can be added.89-91 In six LVAD patients
spheric pressure in the device. This pressure may allow entry of randomized to receive nitric oxide administration (20 ppm),
air through sites such as suture lines and the interstices of the mean pulmonary artery pressure decreased from 35 6 6 to 24 6
graft that are otherwise airtight, with the potential for cata- 4 mmHg, and LVAD flow increased from 1.9 6 0.2 to 2.7 6 0.4
strophic air embolism.78 A further use for TEE is to derive an L/min/m2 in less than 10 minutes. Controls (n 5 5) meeting
independent measure of LVAD flow from knowledge of the similar inclusion criteria had no effect from N2 administration.
outflow cannula dimensions and a determination of the Doppler On crossover to nitric oxide administration in some of these
flow envelope using continuous-wave Doppler.78 patients (n 5 3), however, similar salutary hemodynamic ef-
Postoperatively, it has been suggested that TEE be used to fects were shown again without decreases in systemic arterial
assess recovery of left ventricular function after LVAD place- pressure.90 Wagner et al89 showed similar beneficial effects on
ment.86,87 TEE-derived left ventricular measurements using hemodynamic parameters in eight male LVAD patients with
automated border detection and femoral artery–derived arterial right ventricular dysfunction treated with nitric oxide (25 to 40
pressures can be used to determine myocardial elastance loops ppm). They also showed, using TEE, that right ventricular
when the device is temporarily switched off.88 ejection fraction increased from 24% 6 7% to 44% 6 7%
( p 5 0.01) after 48 hours of nitric oxide administration and that
WEANING FROM CARDIOPULMONARY BYPASS right ventricular end-diastolic volume decreased ( p 5 0.05).
The major concern on weaning from CPB is that right Prostaglandin E1 infusion has also been used to treat pulmonary
ventricular failure usually occurs to a varying degree and that hypertension in LVAD patients.5 Severe refractory vasodilation
these patients are invariably in a state of vasodilatory shock.2,69 may require the addition of a metaraminol or an epinephrine
There appears to be an association between the extent of infusion. Because LVAD flow is volume-dependent, appropriate
bleeding and the development of right ventricular failure, preload must be maintained to ensure flows of at least 4 L/min
necessitating RVAD placement, because Goldstein et al16 showed and never less than 3 L/min for fear of thrombosis and occlusion
that chest tube output (first 24 hours) in these patients (10.5 L) of the inflow cannulae. Lower LVAD flows than this may result
was much greater than in those who did not require RVAD in the generation of negative pressure by the pump and air
placement (2.4 L). It has been postulated that the development entrainment through the suture lines with the possibility of air
of right ventricular failure in the setting of excessive bleeding embolism.78
and resuscitation is secondary to cytokine-mediated pulmonary
POSTOPERATIVE MANAGEMENT
vasoconstriction.2 A further contributor to impaired right ven-
tricular function postoperatively is that the LVAD decreases Patients are managed postoperatively in an intensive care
right ventricular contractility, whereas compliance is increased unit, and ventilation is continued until the patient has recovered
and impairment of septal function occurs.38 Preexisting right from anesthesia and conventional criteria for extubation are
ventricular impairment may be exacerbated further by LVAD met. Inodilators and vasopressor agents as well as nitric oxide
placement. In an attempt to treat right ventricular impairment are weaned according to usual clinical criteria with the added
preemptively, inodilators and vasopressors are continued or advantage in these patients that a continuous readout of LVAD
started on CPB. The likely severity of right ventricular impair- flow allows easy assessment of the effects of changing adminis-
ment may be predicted from a determination of the transpulmo- tered drug dosing.
nary gradient as an indicator of increased pulmonary vascular
resistance preoperatively.4 The size and contractility of the right SUMMARY
ventricle and extent of prebypass tricuspid regurgitation as- Appropriate anesthesia for LVAD placement needs to incor-
sessed by TEE may be used as a guide to determine whether porate an understanding of the surgical procedure and LVAD
324 BEREND METS

physiology, the altered physiology and pharmacology of the function, but must ensure adequate anesthesia to avoid intraop-
cardiac failure patient, and a knowledge of the interaction of erative awareness. Intelligent use of TEE can assist the surgeon
anesthesia with the potent cardiac drugs used in end-stage heart to pinpoint potential pitfalls of LVAD placement and can have a
failure therapy. The anesthesiologist is faced with a critically ill significant effect on improving outcome in these challenging
decompensated patient, often with altered renal and hepatic patients.

REFERENCES
1. Barnard C: A human cardiac transplant: An interim report of a anesthesia in patients chronically treated with angiotensin-converting
successful operation performed at Groote Schuur Hospital, Cape Town. enzyme inhibitors. Anesth Analg 74:805-808, 1992
S Afr Med J 41:1271-1274, 1967 22. Perkins M, Dasta J, Reilly T, et al: Intraoperative complications
2. Goldstein D, Oz M, Rose E: Implantable left ventricular assist in patients receiving amiodarone: characteristics and risk factors. DICP
devices. N Engl J Med 339:1522-1533, 1998 23:757-763, 1989
3. Ozinsky J: Cardiac transplantation—the anaesthetist’s view: A 23. Paller M, Linas S: Role of angiotensin II, alpha-adrenergic
case report. S Afr Med J 41:1268-1270, 1967 system, and arginine vasopressin on arterial pressure in rat. Am J
4. Dickstein M: Anesthesia for heart transplant. Semin Cardiotho- Physiol 84:H25-H30, 1984
rac Vasc Anesth 2:131-139, 1998 24. Carp H, Vadhera R, Jayaram A: Endogenous vasopressin and
5. Frazier H: The development of an implantable, portable, electri- renin-angiotensin systems support blood pressure after epidural block in
cally powered left ventricular assist device. Semin Thorac Cardiovasc humans. Anesthesiology 80:1000-1007, 1994
Surg 6:181-187, 1994 25. Brand P, Metting P, Britton S: Support of arterial pressure by
6. Goldsmith M: First implant of portable heart-assist device. major pressure systems in conscious dogs. Am J Physiol 255:H483-
JAMA 265:2930-2931, 1991 H491, 1988
7. Oz M, Rose E, Levin H: Selection criteria for placement of left 26. Liberman B, Teasedale S: Anaesthesia and amiodarone. Can J
ventricular assist devices. Am Heart J 129:173-177, 1995 Anaesth 32:629-638, 1985
8. Mehta S, Aufiero T, Pae W, et al: Combined registry for the 27. Charlier R: Cardiac actions in the dog of a new antagonist of
clinical use of mechanical ventricular assist pumps and the total adrenergic excitation which does not produce competitive blockade of
artificial heart in conjunction with heart transplantation: Sixth official adrenoreceptors. Br J Pharmacol 39:668-674, 1970
report—1994. J Heart Lung Transplant 14:585-593, 1995 28. Dyck MV, Baele P, Rennotte M, et al: Should amiodarone be
discontinued before cardiac surgery? Acta Anaesthesiol Belg 39:3-10,
9. Frazier O, Rose E, Macmanus Q, et al: Multicenter clinical
1988
evaluation of the HeartMate 1000 IP Left Ventricular Assist Device.
29. Gallagher J, Lieberman R, Mearanze J, et al: Amiodarone-
Ann Thorac Surg 53:1080-1090, 1992
induced complications during coronary artery surgery. Anesthesiology
10. Rose EA, Moskowitz AJ, Packer M, et al: The REMATCH trial:
55:186-188, 1981
Rationale, design, and end-points. Randomized Evaluation of Mechani-
30. Licker M, Neidhart P, Lustemberger S: Long-term angiotensin-
cal Assistance for the Treatment of Congestive Heart Failure. Ann
converting enzyme inhibitor treatment attenuates adrenergic responsive-
Thorac Surg 67:723-730, 1999
ness without altering hemodynamic control in patients undergoing
11. Oz M, Goldstein D, Pepino P, et al: Screening scale predicts
cardiac surgery. Anesthesiology 84:789-800, 1996
patients successfully receiving long-term implanted left ventricular
31. Tuman K, McCarthy R, O’Connor C: Angiotensin-converting
assist devices. Circulation 92:169-173, 1995
enzyme inhibitors increase vasoconstrictor requirements after cardiopul-
12. McCarthy P, Schmidt S, Vargo R, et al: Implantable LVAD
monary bypass. Anesth Analg 80:473-479, 1995
infections: Implications for permanent use of the device. Ann Thorac 32. Argenziano M, Chen JM, Choudhri AF, et al: Management of
Surg 61:359-365, 1996 vasodilatory shock after cardiac surgery: Identification of predisposing
13. Oz MC, Goldstein DJ, Rose EA: Preperitoneal placement of factors and use of a novel pressor agent. J Thorac Cardiovasc Surg
ventricular assist devices: An illustrated stepwise approach. J Card Surg 116:973-980, 1998
10:288-294, 1995 33. Mets B, Michler R, Delphin E, et al: Refractory vasodilation
14. Slater JP, Rose EA, Levin HR, et al: Low thromboembolic risk after cardiopulmonary bypass for heart transplantation in recipients on
without anticoagulation using advanced-design left ventricular assist combined amiodarone and angiotensin-converting enzyme inhibitor
devices. Ann Thorac Surg 62:1321-1327, 1996 therapy: A role for vasopressin administration. J Cardiothorac Vasc
15. Frazier O: Outpatient LVAD: Its time has arrived. Ann Thorac Anesth 12:326-329, 1998
Surg 58:1309-1310, 1994 34. Chin C, Feindel C, Cheng D: Duration of preoperative amioda-
16. Goldstein D, Seldomridge J, Chen J, et al: Use of aprotonin in rone treatment may be associated with postoperative hospital mortality
LVAD recipients reduces blood loss, blood use, and perioperative in patients undergoing heart transplantation. J Cardiothorac Vasc Anesth
mortality. Ann Thorac Surg 59:1063-1067, 1995 13:563-566, 1999
17. Dinardo J: Anesthesia for heart, heart-lung and lung transplanta- 35. Bristow M: The adrenergic nervous system in heart failure. N
tion, in Anesthesia for Cardiac Surgery (ed 2). New York, NY, Appleton Engl J Med 311:850-851, 1984
& Lange, 1998, pp 201-239 36. Thomas J, Marks B: Plasma norepinephrine in congestive
18. Group CTS: Effects of enalapril on mortality in severe conges- cardiac failure. Am J Cardiol 41:233-243, 1978
tive heart failure. N Engl J Med 316:1429-1435, 1987 37. Bove A, Santamore W: Ventricular interdependence. Prog Cardio-
19. SOLVD Investigators: Effect of enalapril on survival in patients vasc Dis 23:365-388, 1981
with reduced left ventricular ejection fractions and congestive cardiac 38. Santamore W, Gray L: Left ventricular contributions to right
failure. N Engl J Med 325:293-302, 1991 ventricular systolic function during LVAD support. Ann Thorac Surg
20. Stevenson W, Stevenson L, Middlekauff H, et al: Improving 61:350-356, 1996
survival for patients with advanced heart failure: A study of 737 39. Blank T, Nyhan D, Kaplan J: Heart and heart lung-transplanta-
consecutive patients. J Am Coll Cardiol 26:1417-1423, 1995 tion, in Kaplan J (ed): Cardiac Anesthesia, (ed 3). Philadelphia, PA, WB
21. Colson P, Saussine M, Seguin JR, et al: Hemodynamic effects of Saunders, 1993, pp 905-916
ANESTHESIA FOR LVAD PLACEMENT 325

40. Mets B: The renin-angiotensin system and ACE inhibitors in the rhagic complications during left ventricular assist implantation. Circula-
perioperative period, in Skarvan K (ed): Arterial Hypertension. London, tion 94:II-227-234, 1996
Bailliere Tindall, 1997, pp 581-604 65. Tanaka K, Wada K, Morimoto T: Hemostatic alterations caused
41. Mets B: Cardiac pharmacology, in Thys D (ed): Text Book of by ventricular assist devices for postcardiotomy heart failure. Artif
Cardiothoracic Anesthesiology. New York, NY, McGraw-Hill, 2000 Organs 15:59-65, 1991
42. Shammas F, Dickstein K: Clinical pharmacokinetics in heart 66. Goldstein D, Oz M, Smith C, et al: Safety of repeat aprotinin
failure: An updated review. Clin Pharmacokinet 15:94-113, 1988 administration for LVAD recipients undergoing cardiac transplantation.
43. Wilkinson G: Pharmacokinetics of drug disposition: Hemody- Ann Thorac Surg 61:692-695, 1996
namic considerations. Ann Rev Pharmacol 15:11-27, 1975 67. Kaplon RJ, Gillinov AM, Smedira NG, et al: Vitamin K reduces
44. Benowitz N: Pharmacokinetics in cardiac failure. Clin Pharmaco- bleeding in left ventricular assist device recipients. J Heart Lung
kinet 1:389-405, 1976 Transplant 18:346-350, 1999
45. Knodell R, Farleigh R, Steele N, et al: Effect of liver congestion 68. Pavie A, Leger P: Physiology of univentricular versus biventricu-
on hepatic drug metabolism in the rat. J Pharmacol Exp Ther 221:52-57, lar support. Ann Thorac Surg 61:347-349, 1996
1981 69. Argenziano M, Choudhri AF, Oz MC, et al: A prospective
46. Perucca E, Grimaldi R, Crema A: Interpretation of drug levels in randomized trial of arginine vasopressin in the treatment of vasodilatory
acute and chronic disease states. Clin Pharmacokinet 10:498-513, 1985 shock after left ventricular assist device placement. Circulation 96:II-
47. Boobis A, Powis G: The effects of catecholamines upon the 286-290, 1997
metabolism of foreign compounds and upon the distribution of perfus- 70. Holley F, Ponganis K, Stanski D: Effect of cardiopulmonary
ate in the isolated perfused pig liver. Biochem Pharmacol 23:3377- bypass on the pharmacokinetics of drugs. Clin Pharmacokinet 7:234-
3384, 1974 251, 1982
48. Mets B, Hickman R, Allin R, et al: Effect of hypoxia on the 71. Buylaert W, Herregods L, Mortier E, et al: Cardiopulmonary
hepatic metabolism of lidocaine in the isolated perfused pig liver. bypass and the pharmacokinetics of drugs. Clin Pharmacokinet 17:10-
Hepatology 19:668-676, 1993 26, 1989
49. Stenson R, Constantino R, Harrison D: Interrelationships of 72. Mets B: The pharmacokinetics of anesthetic drugs and adjuvants
hepatic blood flow, cardiac output, and blood levels of lidocaine in man. during cardiopulmonary bypass. Acta Anaesth Scand (in press)
Circulation 63:205-211, 1971 73. Morrison J, Killip T: Serum digitalis and arrhythmia in patients
50. Benowitz N, Forsyth R, Melmon K, et al: Lidocaine disposition undergoing cardiopulmonary bypass. Circulation 47:341-352, 1973
kinetics in monkey and man. Clin Pharmacol Ther 16:99-109, 1974 74. Coltart D, Chamberlain D, Howard M, et al: Effect of cardiopul-
51. Young R, Ward A: Milrinone: A preliminary review of its monary bypass on plasma digoxin concentrations. Br Heart J 33:334-
pharmacological properties and therapeutic use. Drugs 36:158-192, 338, 1971
1988 75. Bailey JM, Levy JH, Rogers HG, et al: Pharmacokinetics of
52. Boldt J, Menges T, Wollbruck M, et al: Is continuous cardiac amrinone during cardiac surgery. Anesthesiology 75:961-968, 1991
output measurement using thermodilution reliable in critically ill 76. Bailey JM, Levy JH, Kikura M, et al: Pharmacokinetics of
patients? Crit Care Med 22:1913-1918, 1994 intravenous milrinone in patients undergoing cardiac surgery. Anesthe-
53. Baele PL, McMichan JC, Marsh HM, et al: Continuous monitor- siology 81:616-622, 1994
ing of mixed venous oxygen saturation in critically ill patients. Anesth 77. Das PA, Skoyles JR, Sherry KM, et al: Disposition of milrinone
Analg 61:513-517, 1982 in patients after cardiac surgery. Br J Anaesth 72:426-429, 1994
54. Nunez S, Maisel A: Comparison between mixed venous oxygen 78. George S, Black J, Boscoe M: Intraoperative transoesophageal
saturation and thermodilution cardiac output in monitoring patients with echocardiography for implantation of a pulsatile left ventricular assist
severe heart failure treated with milrinone and dobutamine. Am Heart J device. Br J Anaesth 75:794-797, 1995
135:383-388, 1998 79. Hauptman PJ, Body S, Fox J, et al: Implantation of a pulsatile
55. Ghoneim MM, Block RI: Learning and memory during general external left ventricular assist device: Role of intraoperative transesopha-
anesthesia: An update. Anesthesiology 87:387-410, 1997 geal echocardiography [see comments]. J Cardiothorac Vasc Anesth
56. Eger EI: The pharmacology of isoflurane. Br J Anaesth 56:71S- 8:340-341, 1994
99S, 1984 80. Pollock SG, Dent JM, Kaul S, et al: Diagnosis of ventricular
57. Weiskopf RB: Cardiovascular effects of desflurane in experimen- assist device malfunction by transesophageal echocardiography. Am
tal animals and volunteers. Anaesthesia 50:14-17, 1995 (suppl) Heart J 124:793-794, 1992
58. Schlack W, Preckel B, Stunneck D, et al: Effects of halothane, 81. Simon P, Owen AN, Moritz A, et al: Transesophageal echocardio-
enflurane, isoflurane, sevoflurane and desflurane on myocardial reperfu- graphic evaluation in mechanically assisted circulation. Eur J Cardiotho-
sion injury in the isolated rat heart. Br J Anaesth 81:913-919, 1998 rac Surg 5:492-497, 1991
59. Jhaveri R, Tardiff B, Stanley T: Anesthesia for heart and lung 82. Nakatani S, Thomas JD, Vandervoort PM, et al: Left ventricular
transplantation. Anesthesiol Clin North Am 12:729-747, 1994 diastolic filling with an implantable ventricular assist device: Beat-to-
60. Baum V: Anesthesia for heart transplantation recipients. Semin beat variability with overall improvement. J Am Coll Cardiol 30:1288-
Anesth 4:298, 1990 1294, 1997
61. Gutzke GE, Shah KB, Glisson SN, et al: Cardiac transplantation: 83. Shapiro GC, Leibowitz DW, Oz MC, et al: Diagnosis of patent
A prospective comparison of ketamine and sufentanil for anesthetic foramen ovale with transesophageal echocardiography in a patient
induction. J Cardiothorac Anesth 3:389-395, 1989 supported with a left ventricular assist device. J Heart Lung Transplant
62. Clark N, Martin R: Anesthetic considerations for patients 14:594-597, 1995
undergoing cardiac transplantation. J Cardiothorac Anesth 2:519-564, 84. Baldwin RT, Duncan JM, Frazier OH, et al: Patent foramen
1988 ovale: A cause of hypoxemia in patients on left ventricular support. Ann
63. Prough D, Mathru M: Acid-base, fluids, and electrolytes, in Thorac Surg 52:865-867, 1991
Barasch P, Cullen B, Stoelting R (eds): Clinical Anesthesia. Philadel- 85. Heerdt P, Pond C, Blessios G, et al: Inaccuracy of cardiac output
phia, PA, Lippincott-Raven, 1996, pp 157-187 determination by thermodilution during acute tricuspid regurgitation.
64. Livingston ER, Fisher CA, Bibidakis EJ, et al: Increased Ann Thorac Surg 53:706-708, 1992
activation of the coagulation and fibrinolytic systems leads to hemor- 86. Brack M, Olson JD, Pedersen WR, et al: Transesophageal
326 BEREND METS

echocardiography in patients with mechanical circulatory assistance. the treatment of right ventricular dysfunction following left ventricular
Ann Thorac Surg 52:1306-1309, 1991 assist device implantation. Circulation 96:II-291-296, 1997
87. Kyo S, Matsumura M, Takamoto S, et al: Transesophageal color 90. Argenziano M, Choudhri AF, Moazami N, et al: Randomized,
Doppler echocardiography during mechanical assist circulation. ASAIO double-blind trial of inhaled nitric oxide in LVAD recipients with
Trans 35:722-725, 1989 pulmonary hypertension. Ann Thorac Surg 65:340-345, 1998
88. Mandarino WA, Gorcsan J 3rd, Gasior TA, et al: Estimation of 91. Beck JR, Mongero LB, Kroslowitz RM, et al: Inhaled nitric
left ventricular function in patients with a left ventricular assist device. oxide improves hemodynamics in patients with acute pul-
ASAIO J 41:M544-547, 1995 monary hypertension after high-risk cardiac surgery. Perfusion 14:37-
89. Wagner F, Dandel M, Gunther G, et al: Nitric oxide inhalation in 42, 1999