e

e
m

m
m

om

m
co

co

co

co
c
e.

e.

e.

e.
fre

fre

fre

re
sf
ks

ks

ks
62

k
oo

oo

oo

oo
Long-term Mechanical
eb

eb

eb

eb
Circulatory Support and the
m

m
Total Artificial Heart
m

om
co

co

co

co

c
Andrew C. W. Baldwin • Courtney J. Gemmato
e.

e.

e.

e.
• William E. Cohn • O. H. Frazier
re

fre

re

re
sf

f
ks

ks

ks
k
oo

oo

oo

oo
eb

eb

eb

eb
Heart failure (HF) is a leading cause of death in the United States left ventricular assist device (LVAD) (Thoratec, Inc., Pleas-
m

and a significant public health concern throughout the world.
Results from the Framingham Heart Study suggest that the
lifetime risk of developing HF for Americans over the age
m
m
anton, CA), a total of 145 centers are now approved for
device therapy in the United States. This number will likely
continue to grow as this technology and other new MCS

m
of 40 years is approximately 20%, and the incidence of this systems become more widely accessible.8,9
disease steadily increases with age.1 An estimated 5.1 million MCS is an inclusive term that refers to any device used
co

co

co

co

co
Americans are living with HF. This figure is expected to grow to assist or replace the failing heart. The options are wide-
e.

e.

e.

e.
significantly in the coming decades because the average age ranging and include differences in pump location (implant-
fre

fre

fre

fre
of the population is increasing, and advancements in medi- able vs extracorporeal), support capabilities (left, right, or
cal therapy are allowing patients to live longer with their biventricular), and flow characteristics (pulsatile vs continu-
ks

ks

ks

ks
disease.2,3 The economic burden of disease is also high, and ous). This chapter will focus primarily on the available long-
oo

oo

oo

oo
total healthcare costs attributed to HF exceed $30 billion in term options for MCS, particularly current LVADs and the
the United States each year.2,3 In light of the increased mor- total artificial heart (TAH). Although MCS is not without
eb

eb

eb

eb
tality associated with HF, strategies for managing this disease inherent risks, rapid advancements in technology—paired
have grown increasingly sophisticated. with the practical limitations of transplantation—suggest
m

m
Treatment of advanced HF is based on three principal that device therapy has an important role in both the present
methodologies: medical therapy, corrective surgical interven- and the future of HF surgery.
tion, and cardiac support or replacement. Appropriate medi-
m

m
cal management with antihypertensive, diuretic, vasoactive,
co

co

co

co

co
and inotropic therapy has been shown to improve symptoms, HISTORY OF MECHANICAL
slow the progression of disease, and (in some cases) provide
CIRCULATORY SUPPORT
e.

e.

e.

e.
survival benefits.4 Patients with more severe disease may
re

fre

fre

benefit from specific surgical interventions such as coronary Open heart surgery was made possible by the pioneering
fre
sf

revascularization, valve replacement or repair, and ventricular efforts of individuals seeking a method to provide temporary
ks

ks

ks
restoration procedures.5,6 These procedures often pose a high cardiopulmonary support by means of a mechanical pump
k
oo

oo

oo

oo

risk, and surgical candidacy is typically assessed on a case-by- oxygenator. The first successful use of cardiopulmonary
case basis. bypass (CPB) was reported in 1953 by John Gibbon,10 who
eb

eb

eb

eb

Selected patients with severe HF refractory to guideline- used a CPB system of his own design to repair an atrial septal
directed medical therapy may ultimately be considered can- defect. Excessive mortality rates led some surgeons, includ-
m

didates for cardiac transplantation or mechanical circulatory ing Gibbon himself, to abandon open heart surgery. In 1954,
support (MCS). Because of a relative scarcity of available C. Walton Lillehei11 reported the use of cross circulation (human-
donor organs for transplantation, device therapy has become to-human perfusion), using a parent as a temporary source
m

an increasingly popular option for managing end-stage HF of cardiopulmonary support during the repair of congenital
co

co

co

co

co

and offers documented survival benefits over medical therapy heart defects. Controversy stemming from this approach (par-
alone.7 Whereas only 33 centers participated in the original ticularly the risk of a 200% mortality rate) encouraged ongo-
e.

e.

e.

e.

clinical investigation (completed in 2009) of the HeartMate II ing research in the field of mechanical support, and efforts
fre

fre

fre

fre
s

ks

ks

ks

1361
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1361 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1362
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
led by Lillehei, Richard DeWall, and John Kirklin quickly 4 days after its implantation. On August 8, 1966, DeBakey
fre

fre

fre

re
led to refinements in CPB technology, resulting in improved implanted an LVAD in a 37-year-old woman with severe HF

sf
ks

ks

ks
clinical outcomes.12 In April 1956, Denton Cooley reported who could not be weaned from CPB after aortic and mitral

k
the use of a modified bubble oxygenator in the repair of a valve replacement. An extracorporeal pneumatic device, the
oo

oo

oo

oo
postinfarction ventricular septal defect; he would perform 94 pump routed blood from the left atrium to the right axillary
open heart procedures (primarily in children) by the end of artery. After 10 days of support, the patient demonstrated suf-
eb

eb

eb

eb
the year.13 ficient cardiac reserve to allow elective removal of the pump
m

The resulting wave of open heart procedures quickly led
to the observation that some patients were unable to regain
sufficient cardiac function to permit withdrawal from the
heart-lung machine. Anecdotal experience suggested that
m
m
at the bedside.14 She later returned to work, resumed normal
activity, and survived for 6 years with good cardiac function
before dying in an auto accident. In 1969, in an event that
will be discussed later in this chapter, Cooley implanted the

om
slow weaning (over the course of hours) allowed some hearts world’s first TAH to great acclaim.16 Despite this landmark
co

co

co

co
to successfully recover function14 (Fig. 62-1). As a result, event, effective long-term total heart replacement proved to be

c
e.

e.

e.

e.
interest in prolonged circulatory support led to the concept an elusive goal over the ensuing decades.
of a ventricular assist device (VAD). Decades earlier, the By the early 1970s, the complexities of immunosuppres-
re

fre

re

re
eccentric Russian scientist Vladimir Demikhov had success- sive therapy led to waning enthusiasm for cardiac transplanta-
sf

f
ks

ks

ks
fully designed and tested a device capable of maintaining the tion and associated MCS technology. However, research and
k

perfusion of a dog for several hours.15 Although this break- development efforts focused on VAD therapy were continued
oo

oo

oo

oo
through occurred in 1937, his experiments were not widely steadily and quietly within a selected group of institutions.
eb

eb

eb

eb
known outside the Soviet Union until years later. Throughout With support from the National Heart, Lung, and Blood
the 1960s, research efforts at a number of institutions were Institute (NHLBI) and in collaboration with Boston-based
m

focused on the development of MCS technology. A feder-
ally funded program with this goal was established at Baylor
College of Medicine, in Houston, with Michael DeBakey as
m
m
Thermo Cardiosystems, physicians at the Texas Heart Institute
began development of an abdominally positioned, pneumati-
cally driven, pulsatile-flow LVAD. Twenty-two patients were

m
its director. supported by the Model 7 abdominal LVAD, including the
The first clinical application of an LVAD was performed first documented mechanical “bridge” to cardiac transplanta-
co

co

co

co

co
by DeBakey in 1963, who used the device as a temporiz- tion, performed in 1978.17 The introduction of cyclosporine
e.

e.

e.

e.
ing measure in a patient with cardiogenic shock after aortic in the 1980s eventually led to renewed widespread interest in
fre

fre

fre

fre
valve replacement. Although the pump functioned satisfac- the field of transplantation and reestablished a meaningful
torily, the patient succumbed to pulmonary complications clinical role for MCS technology. Ongoing research efforts to
ks

ks

ks

ks
oo

oo

oo

oo
Flow
eb

eb

eb

eb
FLOW 3000 cc/min. gradually-
3000
decreased
m

2000
m
m

m
1000
co

co

co

co

co
e.

e.

e.

e.
0
cc/min. Total bypass Partial bypass Partial bypass
re

fre

fre

80
70 70 70
80
fre
sf

ks

ks

ks
Mean arterial
k

60 55 55
45 45 pressure
oo

oo

oo

oo

40 40
eb

eb

eb

eb

40 35 Mean left atrial

pressure 18
20 20
m

20
10 12 10
0
mm/Hg 0 min. 10 20 30 40 50 60 70 80 90 100 180
m

m
Pre On Off On On Off
op. pump pump pump pump pump
co

co

co

co

co

FIGURE 62-1  Correlation between cardiopulmonary bypass flows, mean arterial pressure, and left atrial pressure in a patient with left ventricular
e.

e.

e.

e.

failure after aortic valve replacement. The observed efficacy of gradual weaning from cardiopulmonary bypass helped stimulate interest in mechani-
cal circulatory support. (Reproduced with permission from DeBakey ME. Left ventricular bypass pump for cardiac assistance: clinical experience,
fre

fre

fre

fre

Am J Cardiol. 1971 Jan;27(1):3-11.)

s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1362 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1363
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
develop pumps capable of long-term support ultimately led
fre

fre

fre

re
to the HeartMate Implantable Pneumatic pump (Thoratec

sf
ks

ks

ks
Corp; Pleasanton, CA), which became the first implantable

k
device to be approved by the United States Food and Drug
oo

oo

oo

oo
Administration (FDA) for use as a bridge to transplantation
(BTT).18,19 An electrically powered, percutaneously vented
eb

eb

eb

eb
adaptation of this device, the HeartMate Extended Vented
m

m
Electric (XVE) LVAD, was spawned directly by these research
efforts. Upon its inaugural implantation in 1991, it became
the first pump to be managed in the outpatient setting.20
Simultaneously, research efforts were engaged in the
m

om
design of a completely different mechanism. The concept
co

co

co

co
of extended nonpulsatile support was inspired by anecdotal

c
e.

e.

e.

e.
reports concerning the use of the extracorporeal BioMedi-
cus centrifugal-force pump (Medtronic; Minneapolis, MN)
re

fre

re

re
as both a temporary LVAD and an extracorporeal mem-
sf

f
ks

ks

ks
brane oxygenator (ECMO). This led to a series of experi-
k

mental studies that were performed by the senior author

oo

oo

oo

oo
of this chapter (OHF) in the Texas Heart Institute labo-
eb

eb

eb

eb
ratories throughout the 1980s. As a result of these efforts FIGURE 62-3 The Jarvik 2000 was the first left ventricular assist
(and in collaboration with engineer Richard Wampler) device to incorporate blood-washed bearings in its axial flow design.
m

m
(Used with permission from Jarvik Heart, Inc.)
the Hemopump, a small catheter-based, axial-flow device,
was developed and successfully implanted in 1988, dem-
onstrating the efficacy and safety of high-speed continuous-
m

m
flow (CF) support21 (Fig. 62-2). Also during this period, an the HeartMate II, another implantable axial-flow device,
association between Texas Heart Institute investigators and
co

co

co

co

co
which would eventually become the most commonly used
researcher Robert Jarvik led to the development of a long- LVAD worldwide. A final step in the development of con-
e.

e.

e.

e.
term implantable CF pump that could be placed directly temporary MCS systems occurred in 1994, when collabora-
fre

fre

fre

fre
into the left ventricle.22 The Jarvik 2000 VAD (Jarvik Heart, tion between Frazier, Wampler, and financier Robert Fine
Inc.; New York, NY) was unique in its utilization of blood- led to the development of a novel CF design. Originally
ks

ks

ks

ks
washed bearings (previously thought to be biologically produced by Kriton Medical, Inc. (and later purchased by
incompatible with the vascular system) and validated the
oo

oo

oo

oo
HeartWare, Inc., Framingham, MA), the HeartWare HVAD
clinical feasibility of axial-based CF pumps23 (Fig. 62-3). utilizes a magnetically levitated, centrifugal-flow design that
eb

eb

eb

eb
Meanwhile, the clinical success of the Hemopump led the permits intrapericardial placement of the discoid-shaped
Nimbus Company (later absorbed by Thoratec) to develop pump housing and can function as both a right- and left-
m

m
sided VAD.24 Within the first decade of the new millen-
nium, all three pumps—the Jarvik 2000, the HeartMate II,
and the HeartWare HVAD—saw clinical use.
m

m
Increasingly sophisticated technology and improving
co

co

co

co

co
clinical reliability gradually led MCS to shift from an experi-
mental endeavor to a viable treatment option. The Random-
e.

e.

e.

e.
ized Evaluation of Mechanical Assistance for the Treatment
re

fre

fre

of Congestive Heart Failure (REMATCH) trial was designed

fre
sf

to compare outcomes between HF patients treated with

ks

ks

ks
k

the HeartMate XVE and those treated with medical ther-

oo

oo

oo

oo

apy alone.7 Completed in 2001, REMATCH investigators

demonstrated a nearly 50% reduction in mortality among
eb

eb

eb

eb

patients with LVAD support (Fig. 62-4). By providing

the first statistical evidence of a survival benefit, this study
m

remains a landmark in the advancement of MCS therapy.

Following the success of REMATCH, the HeartMate II soon
became the first CF device to be approved by the FDA for
m

both BTT indications (in 2007) and for use in patients ineli-
co

co

co

co

co

gible for transplantation (2009).25,26 The HeartWare HVAD

was granted BTT status after the release of the results of
e.

e.

e.

e.

FIGURE 62-2 The catheter-based hemopump first revealed the

safety and feasibility of high-speed continuous-flow circulatory the ADVANCE trial in 2010, and it is expected to receive
fre

fre

fre

fre

support. (Used with permission from Dr. Richard Wampler.) broader approval in the near future.27 With more pumps and
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1363 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1364
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
100 100%

fre

fre

fre

re
90%

sf
Freedom from maior malfunction
ks

ks

ks
80
80%

k
oo

oo

oo

oo
Survival (%)

70%
60
LV assist device
eb

eb

eb

eb
60% p=0.0026
40 50%
m

m
40%
20 XVE (n=101)
30%
Medical therapy VE (n=167)
m

om
0 20%
0 6 12 18 24 30
co

co

co

co
10%

c
Months 0%
e.

e.

e.

e.
No. at risk 0 6 12 18 24
re

fre

re

re
LV assist device 68 38 22 11 5 1 Months post implant
Medical therapy 61 27 11 4 3 0
sf

f
ks

ks

ks
Time 0 1 month 6 months 12 months 18 months
FIGURE 62-4  Kaplan-Meier survival analysis comparing end-stage
k
oo

oo

oo

oo
heart failure patients randomized to receive either optimal medical VE patients 167 133 77 37 14
therapy or left ventricular assist device support. The REMATCH at risk
eb

eb

eb

eb
trial researchers demonstrated a significant survival benefit for device
recipients at both 1 and 2 years. (Reproduced with permission from XVE patients 101 84 59 29 4
m

m
Rose EA, Gelijns AC, Moskowitz AJ, et al: Long-term use of a left at risk
ventricular assist device for end-stage heart failure, N Engl J Med.
FIGURE 62-5  Freedom from all major device malfunctions for the
2001 Nov 15;345(20):1435-1443.)
HeartMate VE and HeartMate XVE left ventricular assist devices.
(Reproduced with permission from Pagani et al. Improved mechani-
m

m
cal reliability of the HeartMate XVE Left Ventricular Assist System,
co

co

co

co

co
newer designs in the pipeline, options for MCS continue to Ann Thorac Surg. 2006 Oct;82(4):1413-1418.)
grow with ever-improving outcomes.
e.

e.

e.

e.
fre

fre

fre

fre
PULSATILE VERSUS Temporary CF technology was first adopted by Gibbon,
who employed roller pumps in the design of his machine that
NONPULSATILE FLOW
ks

ks

ks

ks
first enabled CPB support during surgery.10 However, early
oo

oo

oo

oo
A critical issue in the field of MCS is the physical mechanism efforts to develop long-term MCS pumps focused on pulsa-
by which flow is created. Quite logically, early efforts in device tile blood flow to replicate natural physiology. The eventual
eb

eb

eb

eb
therapy sought to mimic nature’s model by designing pulsa- development of CF LVADs was largely due to their associated
tile systems. Generations of devices have utilized pneumatic engineering advantages, and both experimental and clinical
m

compression systems, hydraulic actuators, and pusher plate
technology to allow repetitive filling and forceful ejection of
blood from a ventricular collection chamber. However, the
m
m
studies quickly documented the physiologic tolerance of
nonpulsatility. The technical superiority of CF devices was
definitively shown in the HeartMate II Destination Therapy

m
engineering challenges associated with matching the simplic- (DT) trial,26 which documented improved adverse event pro-
co

co

co

co

co
ity and reliability of the human heart—which beats more files, hospital readmission rates, functional capacity, and over-
than 40 million times each year—have been considerable. all survival when compared with the HeartMate XVE over
e.

e.

e.

e.
A chief cause of late pulsatile-device failure is mechanical a 2-year period (Table 62-1; Fig. 62-6). The widespread use
re

fre

fre

wear, which results from the presence of complex moving com- and long-term efficacy of CF devices has further confirmed
fre
sf

ponents and multiple points of contact. For example, although the important contribution of continuous blood-flow tech-
ks

ks

ks
the HeartMate XVE represented a significant step forward, it
k

nology to MCS.
oo

oo

oo

oo

had a relatively predictable lifespan, with a high failure rate Whereas the success of CF devices has proven the body’s
between the first and second years of use, and few of these tolerance of nonpulsatile blood flow, this technology has
eb

eb

eb

eb

devices survived for more than 2 years28 (Fig. 62-5). Pulsatility created a new physiology, the full effects of which are not
also requires a number of disadvantageous design elements to
m

yet known. Multiple studies have shown an association

accommodate the mechanics of operation. Pneumatic systems between gastrointestinal (GI) bleeding and the loss of pulsa-
use tunneled drivelines to communicate with an external com- tility.29 Histologic examination of aortic tissue samples have
pressor responsible for delivering (and withdrawing) pressur- revealed unanticipated degeneration of smooth muscle and
m

ized air. Pusher-plate systems (such as that used for the XVE) elastic fibers in patients supported by CF devices.30 Ultraso-
co

co

co

co

co

require a mechanism to allow external venting and aspiration nographic studies have further revealed stasis and thrombus
of air displaced by the cyclic actuation of the pump. As a result formation representing potential sources of thromboembo-
e.

e.

e.

e.

of these inescapable challenges, pulsatile pumps are generally lism within the carotid bulb and aortic valve cusps of selected
fre

fre

fre

fre

large, complex, and vulnerable to wear. patients with nonpulsatile devices.31 These issues present a
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1364 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1365
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
mobility and offer the potential for completely implantable
fre

fre

fre

re
TABLE 62-1:  Comparison of Continuous Flow designs. Although much more remains to be studied, CF

sf
and Pulsatile LVAD Outcomes and Adverse
ks

ks

ks
technology represents the future of MCS.
Event Profiles

k
oo

oo

oo

oo
Continuous Pulsatile  
CURRENT OPTIONS FOR
eb

eb

eb

eb
Flow LVAD Flow LVAD
(n = 133) (n = 59) MECHANICAL CIRCULATORY
m

m
Adverse Event No. (%) No. (%) p Value SUPPORT
Pump replacement 12 (9) 20 (34) <.001 Currently, device therapy involves a wide variety of options,
m

om
LVAD-related 47 (35) 21 (36) .01 obtainable through federally approved indications or investi-
infection gational studies. The pumps differ in a number of important
co

co

co

co
ways, including the duration of support permitted, approach

c
Sepsis 48 (36) 26 (44) <.001
e.

e.

e.

e.
Respiratory failure 50 (38) 24 (41) <.001 to access, mechanisms of flow, and output capabilities. The
re

fre

re

re
Renal failure 21 (16) 14 (24) <.001 most frequently used pumps and selected other devices of
Cardiac arrhythmia 75 (56) 35 (59) .006 interest—notably all CF designs—are discussed below. A
sf

f
ks

ks

ks
Rehospitalization 107 (94) 42 (96) .02 more thorough listing of currently available devices is provided
k

in Table 62-2.
oo

oo

oo

oo
Quality of Life Score Score
Metric (12 months) (12 months) p Value
eb

eb

eb

eb
Minnesota 34.1 ± 22.4 44.4 ± 23.2 .03
ECMO
m

m
Kansas City 65.9 ± 20 59.1 ± 20.3 .06 Oxygenation of the blood by a membrane oxygenator can be
achieved through the use of either veno-venous (VV) or veno-
LVAD, left ventricular assist device; Minnesota, Minnesota Living with Heart
Failure questionnaire; Kansas City, Kansas City Cardiomyopathy Questionnaire. arterial (VA) ECMO. The VV option is typically used for the
m

m
Adapted with permission from Slaughter et al. Advanced heart failure treated management of isolated respiratory failure, and it relies on
with continuous-flow left ventricular assist device, N Engl J Med. 2009
co

co

co

co

co
Dec 3;361(23):2241-2251. the function of the native heart for systemic circulation of
oxygenated blood. In contrast, through an arterial return can-
e.

e.

e.

e.
nula, VA ECMO allows complete cardiopulmonary support,
significant challenge, but the engineering advantages afforded
fre

fre

fre

fre
similar to that provided by CPB. Cannulation may be central
by CF pumps are considerable. These devices are smaller, or peripheral, rendering ECMO particularly useful in both
ks

ks

ks

ks
more efficient, and much more durable than their pulsatile emergent and intraoperative situations. It is generally limited
counterparts. Without the need for compliance chambers
oo

oo

oo

oo
to periods of short-term support (measured in days).
and pneumatic vents, CF pumps permit increased patient
eb

eb

eb

eb
Paracorporeal and Percutaneous Devices
m

m
1.0
0.9 A subset of devices can be inserted percutaneously or involve
0.8 extracorporeal pumping components. Recent studies have
Continuous-flow
Probability of survival

0.7 shown a dramatic increase in the utilization of these short-

m

m
LVAD
0.6 term devices,32 which are typically used for temporary MCS
co

co

co

co

co
0.5
in the setting of acute cardiogenic or postcardiotomy shock.
Pulsatile-flow
The CentriMag (Thoratec Corp.) is an extracorporeal VAD
e.

e.

e.

e.
0.4 LVAD
utilizing a magnetically levitated centrifugal flow impeller that
re

fre

fre

fre
0.3
can provide both right- and left-sided support (Fig. 62-7a).
0.2
sf

Access is typically gained via centrally placed CPB cannu-

ks

ks

ks
P=0.008 by the log-rank test
0.1
k

las that can be tunneled to allow temporary chest closure.

oo

oo

oo

oo

0.0
0 6 12 18 24
Capable of flows reaching 10 L/min, the CentriMag can also
accommodate an oxygenator within the pumping circuit,
eb

eb

eb

eb

Months since randomization

permitting full cardiopulmonary support. The device is gen-
No. at risk
m

erally used for days to weeks, but support periods exceeding

Continuous-flow
LVAD
133 95 82 69 62 30 days have been reported.33
Pulsatile-flow The TandemHeart Percutaneous Ventricular Assist Device
59 32 19 5 2 (pVAD) (CardiacAssist Inc.; Pittsburgh, PA) is a percu-
m

m
LVAD
taneous device that functions as a left atriofemoral bypass.
co

co

co

co

co

FIGURE 62-6  Kaplan-Meier survival analysis comparison of patients

supported by continuous flow and pulsatile left ventricular assist
Inserted under fluoroscopic guidance, the TandemHeart has
a transseptal cannula that transfers oxygenated blood from
e.

e.

e.

e.

devices. (Reproduced with permission from Slaughter et al. Advanced

heart failure treated with continuous-flow left ventricular assist device, the heart to an extracorporeal, magnetically driven, centrifu-
fre

fre

fre

fre

N Engl J Med. 2009 Dec 3;361(23):2241-2251.) gal flow pump mounted on the thigh. Using a hydrodynamic
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1365 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1366
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
fre

fre

fre

re
TABLE 62-2:  Currently Available Options for Mechanical Circulatory Support

sf
ks

ks

ks
First human

k
Device Manufacturer Flow type Design implant Approval status
oo

oo

oo

oo
Paracorporeal         Worldwide United States
eb

eb

eb

eb
CentriMag® Thoratec Corp. Continuous Centrifugal 2003 CE Mark—2002 FDA 510(k)—2002
m

m
(Pleasanton, CA) FDA IDE—2008
FDA HDE
  RVAD—2008
Percutaneous            
m

om
Impella 5.0®/CP®/ ABIOMED Continuous Microaxial 1999 CE Mark FDA 510(k)
co

co

co

co
RP®   5.0—2003   5.0—2009

c
e.

e.

e.

e.
  CP—2012   CP—2012
  RP—2014 FDA HDE
re

fre

re

re
  RP—2015
sf

f
ks

ks

ks
TandemHeartTM CardiacAssist, Inc. Continuous Centrifugal 2005 CE Mark—2000 FDA 510(k)—2003
k

  (Pittsburgh, PA) FDA IDE—2012

oo

oo

oo

oo
Implantable LVADs            
eb

eb

eb

eb
HeartMate® II Thoratec Continuous Axial 2000 CE Mark—2005 FDA BTT—2008
m

m
FDA DT—2010
Jarvik 2000 Jarvik Heart, Inc. Continuous Axial 2000 CE Mark—2005 FDA IDE—2000
Flowmaker®   (New York, NY) BTT Trial Pending
m

m
INCOR® Berlin Heart GmbH Continuous Axial 2002 CE Mark—2003 —
co

co

co

co

co
  (Berlin, Germany)
EVAHEART® Evaheart, Inc. Continuous Centrifugal 2005 CE Mark—2015 FDA IDE—2009
e.

e.

e.

e.
  (Houston, TX) PMDA—2010 BTT Trial Pending
fre

fre

fre

fre
HVAD® HeartWare, Inc. Continuous Centrifugal 2006 CE Mark—2009 FDA BTT—2012
ks

ks

ks

ks
  (Framingham, MA) DT Trial Pending
HeartAssist 5® ReliantHeart, Inc. Continuous Axial 2009 CE Mark—2013 FDA IDE—2014
oo

oo

oo

oo
  (Houston, TX) BTT Trial Pending
eb

eb

eb

eb
HeartMate® III Thoratec Continuous Centrifugal 2014 CE Mark Pending FDA IDE—2014
BTT/DT trial pending
m

m
MVAD® HeartWare Continuous Axial 2015 CE Mark Pending Awaiting FDA IDE
Total Artificial Heart   
SynCardia TAH SynCardia Systems, Inc. Pulsatile Pneumatic 1982 CE Mark—1999 FDA HDE—2004
m

m
  (Tucson, AZ)
co

co

co

co

co
CE Mark, European Union Conformité Européenne Approval; FDA, United States Food and Drug Administration; 510(k), Premarket Notification Clearance; PMA,
e.

e.

e.

e.
premarket approval; IDE, investigational device exemption; HDE, humanitarian device exemption; BTT, bridge to transplant; DT, destination therapy; PMDA, Japanese
Pharmaceuticals and Medical Devices Agency Approval; TAH, total artificial heart.
re

fre

fre

fre
sf

ks

ks

ks
bearing system, the TandemHeart can provide the femoral Implantable Continuous-Flow LVADs
k

artery with return flows reaching 4 L/min.34 (FDA-Approved)

oo

oo

oo

oo

The Impella (Abiomed; Danvers, MA) is a platform of

The HeartMate II axial-flow device is the most widely used
eb

eb

eb

eb

devices based upon a catheter-mounted microaxial design

conceptually similar to that of the original Hemopump LVAD worldwide, with more than 15,000 implants to date
m

(Fig. 62-7b). A percutaneous arterial catheter serves as a
power source and flow conduit for the pump, and left-sided
support is provided in a retrograde fashion across the aortic
m
(Fig. 62-8a). The pump’s electromagnetic motor powers a sin-
gle hydrodynamically suspended impeller operating at speeds
between 6000 and 15,000 rpm (although clinical operating
ranges are traditionally much smaller), and capable of flows

valve. With a maximal diameter of just 12 F (4 mm), various

adaptations of the device allow flows ranging from 2.5 to 5 L/ greater than 10 L/min. Most often implanted via median ster-
co

co

co

co

co

min. A right-sided version, the Impella RP, delivers antegrade notomy, the sintered titanium pump inlet is placed within the
e.

e.

e.

e.

flow from the inferior vena cava to the pulmonary artery. It left ventricle with an outflow graft anastomosed to the aorta. A
percutaneous driveline communicates with the remainder of the
fre

fre

fre

fre

recently became the first percutaneous device approved by the

FDA for right ventricular (RV) support.35 system: an external device controller and two portable batteries.
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1366 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1367
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
fre

fre

fre

re
sf
ks

ks

ks

k
oo

oo

oo

oo
eb

eb

eb

eb
m

m
m

om
co

co

co

co

c
e.

e.

e.

e.
re

fre

re

re
sf

f
ks

ks

ks
k
oo

oo

oo

oo
A B

FIGURE 62-7  Percutaneous ventricular assist devices. (A) The Thoratec Centrimag utilizes an extracorporeal magnetically levitated centrifugal-
eb

eb

eb

eb
flow pump to provide temporary circulatory support. (Reproduced with permission from Thoratec Corporation.) (B) The Impella CP is a catheter-
based pump designed to be positioned across the aortic valve and capable of producing flows as high as 4 L/min. (Reproduced with permission
m

m
from ABIOMED, Inc.)
m

m
The HeartWare HVAD has a centrifugal-flow design, with pump body allows intrapericardial positioning and serves as
a smaller and more compact profile than the HeartMate II an advantage for right-sided use. Like the HeartMate II, the
co

co

co

co

co
(145 vs 375 g, respectively), integrating the ventricular inlet HVAD relies on external batteries and a controller module
e.

e.

e.

e.
with the pump housing (Fig. 62-8b). The impeller is sus- connected to the patient by a percutaneous driveline.
fre

fre

fre

fre
pended via a combination of magnetic and hydrodynamic A number of additional devices remain in various stages
forces, and it operates at speeds ranging from 1800 to 4000 rpm, of investigational study in the United States and are likely
ks

ks

ks

ks
again with flows as high as 10 L/min. The flat surface of the to be more widely available in the coming years. On the
oo

oo

oo

oo
eb

eb

eb

eb
m

m
m

m
co

co

co

co

co
e.

e.

e.

e.
re

fre

fre

fre
sf

ks

ks

ks
k
oo

oo

oo

oo
eb

eb

eb

eb
m

m
m

A B C
co

co

co

co

co

FIGURE 62-8  The HeartMate II (A) and the HeartWare HVAD (B) are the two most commonly implanted devices worldwide. While the
e.

e.

e.

e.

HeartMate II utilizes an axial-flow design with an extracardiac pump housing, the centrifugal HVAD has a discoid shape designed for intraperi-
cardial placement. Newer-generation devices, such as the HeartWare MVAD (C), reflect increasing miniaturization of mechanical circulatory
fre

fre

fre

fre

support technology. (Used with permission from Thoratec Corporation [a] and HeartWare International, Inc. [b and c].)
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1367 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1368
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
whole, newer-generation pumps are being designed to pro- which transplant-ineligible patients supported by MCS had
fre

fre

fre

re
duce excellent performance while offering a smaller pro- better outcomes than patients supported by medical therapy

sf
ks

ks

ks
file, thus simplifying implantation and expanding potential alone.7 Through this designation, patients with systemic

k
applications to include smaller-framed and pediatric popula- life-threatening illnesses, malignancies, fixed pulmonary
oo

oo

oo

oo
tions (Fig. 62-8c). TAH technology is discussed later in this hypertension, advanced age, and morbid obesity may ben-
chapter. efit from device therapy regardless of their transplant listing
eb

eb

eb

eb
status. Designation as DT does not imply a limitation on
m

m
the timeframe for treatment, as uninterrupted CF support
surpassing 9 years has been reported. Over the past 7 years,
GOALS OF THERAPY the proportion of patients categorized as DT candidates at
the time of implantation has steadily increased from 14.7
m

om
Before device implantation, treatment teams must designate
an intended goal of therapy. Traditional categories of strategic to 41.6%—likely as a result of increased access to FDA-
co

co

co

co
intent represent the evolution of MCS. Originally, devices approved devices for DT, as well as recognition of the pos-

c
e.

e.

e.

e.
were conceived as temporizing measures while the patient sibility of reassessing the transplant status throughout the
period of device support.
re

fre

re

re
awaited a transplant. The eventual inclusion of transplant-
ineligible patients for device consideration required a new
sf

f
ks

ks

ks
and separate indication, resulting in the somewhat dichoto-
Bridge to Decision
k

mous framework in use today. Ideally, the decision to seek

oo

oo

oo

oo
device therapy should be based on the need for mechanical The use of long-term MCS in the setting of acute hemo-
eb

eb

eb

eb
support rather than being indirectly affected by the patient’s dynamic collapse or active decompensation results in com-
transplant listing status. However, the decision currently paratively poor outcomes.37 Short-term MCS offers the
m

serves a critical purpose in regard to patient and device selec-
tion, timing of intervention, regulatory oversight, and data
acquisition. Clinical trials aimed at seeking device approval
m
m
advantage of rapid and relatively less invasive access to
device therapy while allowing stabilization of the patient
and consideration of long-term device candidacy. Thus,

m
have traditionally been designed on this model. Likewise, the patients with a myocardial infarction, myocarditis, acute
United States Centers for Medicare and Medicaid Services
co

co

co

co

co
cardiogenic shock, or unsuccessful percutaneous interven-
(CMS) uses the indicated strategy for device therapy to estab- tion can be quickly transitioned to support in anticipation
e.

e.

e.

e.
lish a given patient’s eligibility for treatment. The categories of recovery, decline, or optimization before the initiation of
fre

fre

fre

fre
are a part of the MCS lexicon and warrant inclusion here. long-term MCS.
ks

ks

ks

ks
Bridge to Transplantation Myocardial Reconditioning
oo

oo

oo

oo
The original intention for MCS was to provide a “bridge” In a selected subset of patients, a distinct improvement in
eb

eb

eb

eb
to eventual cardiac transplantation for patients awaiting a ventricular function has allowed device weaning and (occa-
suitable donor organ. Unpredictable waitlist times and a fre- sionally) deactivation. Beginning in the 1990s, studies
m

quently tenuous clinical condition make these patients partic-
ularly vulnerable to deterioration while awaiting a transplant.
The decision to seek a BTT strategy requires consideration of
m
m
have revealed histologic and clinical evidence of myocardial
rehabilitation after periods of both pulsatile and CF device
support.38-40 Often labeled a “bridge to recovery,” this phe-

m
estimated waitlist times, the potential for a decline in clinical nomenon is perhaps better described as a return to medical
co

co

co

co

co
status before transplantation, the patient’s quality of life, and HF therapy. A number of institutions have proposed device-
the risk of death while on the waiting list. Progressively lon- weaning strategies, ranging from aggressive medical therapy
e.

e.

e.

e.
ger waitlist times throughout the United States have resulted using high-dose beta blockade to mechanical offloading and
re

fre

fre

in the relatively standard use of BTT for the treatment of gradual reconditioning of the heart through systematic titra-
fre
sf

transplant-eligible patients with end-stage HF.36 Whereas tion of pump speeds.41-43 Despite promising reports, only a
ks

ks

ks
BTT remains the most common implant strategy for MCS
k

small percentage of patients have successfully experienced this

oo

oo

oo

oo

(nearly 65% of these patients are designated as transplant outcome. Although little is known with certainty regarding
candidates), the percentage of patients officially listed for the precise variables that govern the pace and degree of myo-
eb

eb

eb

eb

a transplant at the time of device implantation has steadily cardial response, the potential for de-escalation of therapy
declined, now totaling just 21.7%.37
m

supports an aggressive pursuit of weaning strategies in all

patients who require device support.
Destination Therapy
Fluidity of Implantation Strategy
m

Patients with end-stage HF who are not eligible for trans-

co

co

co

co

co

plantation may benefit from MCS as a means to prolong or It is important to recognize that these designations apply
improve the quality of their life. The designation of destina- only to the stated goal at the time of device implantation.
e.

e.

e.

e.

tion therapy (DT) originates from the REMATCH trial, in Whereas regulatory guidelines mandate a proposed goal
fre

fre

fre

fre
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1368 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1369
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
of therapy, the available choices do not reflect the reality and underscore the multiorgan and multidisciplinary
fre

fre

fre

re
that a patient’s clinical status and transplant candidacy may approach to device candidacy.

sf
ks

ks

ks
change during periods of mechanical support. Although In 2013, the American College of Cardiology Foun-

k
some patients may have a deteriorating clinical status or dation (ACCF) and American Heart Association (AHA)
oo

oo

oo

oo
develop unexpected contraindications to transplantation, released a broad slate of guidelines for the management of
other patients—such as those with preoperative “fixed” HF.4 The report endorsed the use of an alternative to the
eb

eb

eb

eb
pulmonary hypertension—may experience an improve- traditional New York Heart Association (NYHA) staging
m

ment in conditions that originally precluded transplant
listing.44,45 The dynamic nature of device strategy is repre-
sented in reported device usage statistics, in that nearly 26%
of patients initially listed for BTT were no longer listed at
m
m
system, emphasizing disease-specific markers rather than
patient-specific symptoms (Table 62-3). The guidelines
support the use of durable MCS in patients with so-called
Stage D HF (refractory disease requiring specialized inter-

om
1 year, and that number rose to 43.5% by 24 months. Simi- ventions) with reduced left ventricular ejection fraction
co

co

co

co
larly, nearly 15% of patients designated for DT at the time (LVEF) as a means to both prolong survival and BTT.

c
e.

e.

e.

e.
of device implantation were deemed eligible for transplanta- Although the authors refrain from further characterizing
tion by 12 months.46 Frequent reappraisal and modification patient selection criteria, the guidelines do emphasize the
re

fre

re

re
of treatment strategies corresponding to clinical status are perceived importance of refractory disease, reduced EF,
sf

f
ks

ks

ks
essential for successful MCS outcomes. and severe symptomatology.
k

In 2013, CMS released updated eligibility criteria for the

oo

oo

oo

oo
use of MCS as DT in Medicare beneficiaries.48 More spe-
INDICATIONS FOR DEVICE SUPPORT
eb

eb

eb

eb
cific than the ACCF/AHA guidelines, the federal standards
Optimal strategies and indications for the initiation of allow the use of DT only for patients with end-stage HF
m

MCS remain subjects of active investigation. Traditional
hemodynamic parameters for device support continue
to surface in the literature; although not absolute, these
m
m
(NYHA class IV) who are deemed ineligible for transplanta-
tion at the time of device implantation and who meet each
of the following conditions: (a) failure to respond to opti-

m
variables include blood pressure (BP) < 80 mm Hg, mean mal medical management for 45 of the last 60 days, balloon
arterial pressure (MAP) < 65 mm Hg, cardiac index pump dependence for 7 days, or IV inotrope dependence
co

co

co

co

co
< 2 L/min/m2, and pulmonary capillary wedge pressure for 14 days; (b) LVEF < 25%; and (c) measured peak oxygen
e.

e.

e.

e.
(PCWP) > 20 mm Hg.47 Strict following of hemody- consumption of <14 mL/kg/min or an inability to perform
fre

fre

fre

fre
namic markers can be misleading, however, as patients the test. The federal requirements for BTT designation are
may present not just in acute cardiogenic shock but also somewhat broader, mandating only that patients be listed
ks

ks

ks

ks
with a more insidious manifestation of congestive HF. for transplantation by an approved transplant center. How-
In 2013, the International Society for Heart and Lung ever, use of a device for any purpose is permitted only if
oo

oo

oo

oo
Transplantation (ISHLT) proposed a thorough and that specific device has been granted FDA approval for the
eb

eb

eb

eb
broad-ranging compilation of recommendations for the chosen therapeutic goal.
use of MCS.36 Although many of the proposed guide- The Interagency Registry for Mechanically Assisted Cir-
m

lines are limited by the relative weakness of available
supporting evidence, they represent the most compre-
hensive and ambitious summary of consensus opinion,
m
m
culatory Support (INTERMACS) database was established
in 2008 to facilitate the growth of MCS and improve patient
outcomes. Paired with the registry’s first report was a proposal

m
co

co

co

co

co
e.

e.

e.

e.
TABLE 62-3:  Comparison of ACCF/AHA Stages of HF and NYHA Functional Classifications
re

fre

fre

fre
sf

ks

ks

ks
ACCF/AHA HF stages NYHA functional classification
k
oo

oo

oo

oo

A At high risk for HF but without structural heart I No limitation of physical activity. Ordinary physical activity does not
disease or symptoms of HF cause symptoms of HF
eb

eb

eb

eb

B Structural heart disease but without signs or II Slight limitation of physical activity. Comfortable at rest, but ordinary
m

symptoms of HF physical activity results in symptoms of HF

C Structural heart disease with prior or current III Marked limitation of physical activity. Comfortable at rest, but less than
symptoms of HF ordinary activity causes symptoms of HF
D Refractory HF requiring specialized interventions IV Unable to carry on any physical activity without symptoms of HF, or
m

symptoms of HF at rest
co

co

co

co

co

ACCF, American College of Cardiology Foundation; AHA, American Heart Association; NYHA, New York Heart Association; HF, heart failure.
e.

e.

e.

e.

Reproduced with permission from Yancy CW, Jessup M, Bozkurt B, et al: 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College
of Cardiology Foundation/American Heart Association Task Force on practice guidelines, Circulation. 2013 Oct 15;128(16):e240-e327.
fre

fre

fre

fre
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1369 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1370
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
fre

fre

fre

re
TABLE 62-4:  INTERMACS Clinical Profiles of Advanced Heart Failure

sf
ks

ks

ks
Profile Description Timeframe for definitive intervention

k
oo

oo

oo

oo
1 Patient with life-threatening hypotension despite rapidly escalating inotropic Needed within hours
eb

eb

eb

eb
support, critical organ hypoperfusion with increasing lactate levels and/or
systemic acidosis. “Crash and burn”
m

m
2 Patient with declining function despite intravenous inotropic support, may be Needed within few days
manifest by worsening renal function, nutritional depletion, inability to
restore volume balance. “Sliding on inotropes”
m

om
3 Patient with stable blood pressure, organ function, nutrition, and symptoms on Elective over a few weeks
co

co

co

co
continuous intravenous inotropic support, but demonstrating repeated failure

c
to wean owing to recurrent symptomatic hypotension or renal dysfunction.
e.

e.

e.

e.
“Dependent stability”
re

fre

re

re
4 Patient can be stabilized close to normal volume status but experiences frequent Elective over weeks to months as long as
relapses into fluid retention, generally with high diuretic doses. Symptoms are treatment of episodes restores stable baseline,
sf

f
ks

ks

ks
recurrent rather than refractory. More intensive management strategies should including nutrition
k

be considered, which in some cases reveal poor compliance. “Frequent flyer”

oo

oo

oo

oo
5 Patient is living predominantly within the house, performing activities of Variable, depends upon nutrition, organ
eb

eb

eb

eb
daily living and walking from room to room with some difficulty. Patient is function, and activity
comfortable at rest without congestive symptoms, but may have underlying
m

m
refractory elevated volume status, often with renal dysfunction. “Housebound”
6 Patient without evidence of fluid overload is comfortable at rest and with Variable, depends upon nutrition, organ
activities of daily living and minor activities outside the home, but fatigues function, and activity
m

m
after the first minutes of any meaningful activity. “Walking wounded”
7 A placeholder for future specification, patients without recent unstable fluid Transplantation or circulatory support not
co

co

co

co

co
balance, living comfortably with meaningful activity limited to mild exertion currently indicated
e.

e.

e.

e.
Reproduced, with permission, from Stevenson LW, Couper G: On the fledgling field of mechanical circulatory support, J Am Coll Cardiol. 2007 Aug 21;50(8):748-751.
fre

fre

fre

fre
ks

ks

ks

ks
for a new system of classifying patients with advanced HF.49 anomalies, and restrictive heart disease with decreased ven-
oo

oo

oo

oo
Seven clinical profiles were designed to establish an enhanced tricular dimensions. Because successful outcomes have never-
eb

eb

eb

eb
method for characterizing patients requiring device therapy theless been reported in these populations, decisions should
(Table 62-4). These profiles represent a stratification of patients be made on a case-by-case basis in order to weigh the poten-
m

diagnosed with advanced HF—defined as NYHA class III and
IV symptoms despite optimal therapy. By documenting the
profiles of patients undergoing device implantation, INTER-
m
m
tial risks and benefits of intervention.51,52 Surgery should be
deferred in the presence of active infection or bacteremia
to minimize the risk of bacterial device seeding; current

m
MACS provides an opportunity to evaluate data and trends recommendations suggest a minimum of 5 days of culture-
concerning surgical risks and indications. Clinical practice documented clearance from infection.36
co

co

co

co

co
has indeed shifted since the introduction of this system, most
e.

e.

e.

e.
notably through de-escalation of the preoperative clinical pro-
PREOPERATIVE EVALUATION
re

fre

fre

file. Whereas more than 60% of patients who received device

implants before 2011 were listed as having INTERMACS fre
sf

Once potential candidacy for device support has been deter-

ks

ks

ks
1 or 2 status, the proportion of patients classified as having
k

mined, a thorough appraisal of perioperative risk is war-

level 3 status has steadily increased, now representing nearly a
oo

oo

oo

oo

ranted. If at all possible, implantation of long-term MCS

third of all who receive implants.8,50 Indications and guidelines device should be performed electively, not under emergent
eb

eb

eb

eb

continue to evolve, but these results illustrate a clear trend in circumstances. Identification of preoperative risk factors and
the management of HF toward intervening during periods of
m

treatment of reversible comorbidities can greatly improve

reduced clinical acuity. patient outcomes.
Due to the importance of proper patient selection, efforts
CONTRAINDICATIONS TO to predict early mortality after LVAD implantation have
m

been a consistent theme in MCS research. During the era

DEVICE SUPPORT
co

co

co

co

co

of pulsatile LVADs, several risk-evaluation scoring systems

Relative contraindications to device therapy include irre- were developed with variable predictive utility. One of the
e.

e.

e.

e.

versible major end-organ dysfunction, severe hemodynamic most widely used of these scales, the Leitz-Miller destination
fre

fre

fre

fre

instability, profound coagulopathy, complex congenital therapy risk score (DTRS), was derived from a cohort of over
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1370 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1371
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
this patient population and significantly escalate operative
fre

fre

fre

re
TABLE 62-5:  Multivariable Predictors of risk.

sf
90-Day Mortality (HeartMate II Risk Score)
ks

ks

ks
The utility of concomitant cardiac and valvular procedures

k
is poorly understood and remains controversial. Whereas
oo

oo

oo

oo
Parameter SE OR (95% CI) p Value
some studies have suggested that concurrent procedures are
associated with diminished overall survival, consensus opin-
eb

eb

eb

eb
Age (years) 0.12 1.32 (1.05-1.65) .018
Albumin (g/dL) 0.23 0.49 (0.31-0.76) .002 ion does support intervention in specific situations.36,59,60
m

m
Creatinine (mg/dL) 0.20 2.10 (1.37-3.21) <.001 Native aortic insufficiency has been shown not only to
INR 0.32 3.11 (1.66-5.84) <.001 adversely affect the performance of CF pumps but also to
Center volume 0.34 2.24 (1.15-4.37) .018 speed the progression of valvular incompetence.61 Although
clinical practice varies, ISHLT guidelines suggest concurrent
m

om
 
  Low risk Medium risk High risk intervention for regurgitation deemed moderate or greater.36
co

co

co

co
The presence of a mechanical prosthesis entails a prohibitive

c
HMRS score <1.58 1.58-2.48 >2.48
e.

e.

e.

e.
90-Day survival 96% (± 1%) 84% (± 2%) 71% (± 4%) risk of thromboembolic complications and warrants surgical
exclusion through ligation of the LV outflow tract, supraval-
re

fre

re

re
HMRS = (0.0274 × Age) – (0.723 × Alb) + (0.74 × Creat) + (1.136 × INR) + vular patch closure, or a placement of a felt sandwich plug.62-65
sf

f
(0.807 × Vol).
ks

ks

ks
*
Center Volume: Yearly LVAD implant volume <15, value = “0”; Tricuspid regurgitation has been implicated in the develop-
k

Yearly volume >15, value = “1”. ment of postoperative right-sided HF (RHF), and guidelines
oo

oo

oo

oo
SE,standard error; OR, odds ratio; CI, confidence interval; INR, international
normalized ratio; HMRS, HeartMate II Risk Score; Alb, albumin; Creat, creati- generally support intervention—either replacement or ring
eb

eb

eb

nine; Vol, yearly device implant volume; LVAD, left ventricular assist device.
Adapted with permission from Cowger J1, Sundareswaran K, Rogers JG,
m
eb
repair—for disease classified as moderate or greater.36,66-68
The presence of mitral regurgitation (MR) does not typically

et al: Predicting Survival in Patients Receiving Continuous Flow Left Ventricular
Assist Devices: The HeartMate II Risk Score, J Am Coll Cardiol. 2013 Jan 22;
61(3):313-321.
m
m
necessitate surgical intervention at the time of device implan-
tation. Whereas MR is common in patients with end-stage
HF, this finding is often the result of annular distortion sec-

m
ondary to dilation of the left ventricle and typically improves
with mechanical decompression and offloading.69 Due to an
co

co

co

co

co
300 HeartMate XVE patients and incorporated 9 weighted
variables to estimate 90-day in-hospital mortality.53 When increased risk for right-to-left shunting, atrial septal defects
e.

e.

e.

e.
systematically tested in a CF LVAD population; however, and patent foramen ovale (PFO) should be closed at the time
fre

fre

fre

fre
DTRS was found to provide only modest discriminatory of implant.70 Unmasking of a previously undetected PFO has
value.54 In contrast, the Model of End-Stage Liver Disease been reported after mechanical unloading of the left side of
ks

ks

ks

ks
(MELD) has been shown to be a predictor of adverse events the heart, and this defect must be surgically or endovascularly
repaired if systemic postoperative hypoxemia occurs.69,71
oo

oo

oo

oo
after implantation of both pulsatile and CF LVADs.55 Utiliz-
ing only the creatinine level, total bilirubin level, and inter-
eb

eb

eb

eb
national normalized ratio (INR), the MELD score (and later
the anticoagulation-compatible MELD-XI score) was found End-Organ Function
m

to correlate with overall and postoperative survival.56 More
recently, a global risk-assessment scale was developed using
only CF data culled from the HeartMate II clinical trials
m
m
The presence of multiple medical comorbidities is a common
finding in patients with end-stage HF and an ever-present
challenge in the surgical management of the disease. Perma-

m
database. The HeartMate II multivariable risk score (HMRS) nent, life-threatening comorbidities (such as preoperative
co

co

co

co

co
utilizes five variables—patient age, albumin level, creati- dialysis dependence) are often associated with excessive peri-
nine level, INR, and implanting center volume—to predict operative risks and frequently preclude eligibility for MCS.
e.

e.

e.

e.
90-day postoperative mortality57 (Table 62-5). While simi- However, improvement in end-organ function has been
re

fre

fre

lar to MELD in its direct comparison of predictive results, documented after periods of device support.72-75 Preoperative
fre
sf

HMRS is more specifically designed for HF patients and may determination of the severity of end-organ disease is a critical
ks

ks

ks
offer a modest benefit in outcome discrimination.58 To date,
k

process, and it relies on the collective clinical judgment of the

oo

oo

oo

oo

no risk-assessment scale has shown absolute superiority in multidisciplinary treatment team.

preoperative risk stratification, but the available models do Renal dysfunction, specifically, is strongly associated
eb

eb

eb

eb

emphasize the importance of preoperative optimization of with poor outcomes after device implantation.76,77 As a
overall end-organ function.
m

result, patients who require chronic preoperative dialysis are

generally excluded from implantable MCS consideration.
Additionally, multiple studies have shown a decline in post-
Operative Risk operative survival associated with deranged markers of liver
m

As with any proposed cardiothoracic procedure, the surgical function.78,79 While acute cardiogenic liver injury can be sta-
co

co

co

co

co

complexity of each case must be carefully addressed early in bilized, elevations in bilirubin levels, aminotransferase levels,
the evaluation process. Unfavorable anatomical scenarios— and INR warrant further workup and evaluation. Preop-
e.

e.

e.

e.

such as a history of previous sternotomies, congenital anoma- erative mechanical ventilation has been shown to be associ-
fre

fre

fre

fre

lies, or heavily calcified anastomotic sites—are common in ated with postoperative respiratory failure, RV failure, and
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1371 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1372
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
increased mortality.80 An assessment of pulmonary reserve is Multidisciplinary Assessment/Social
fre

fre

fre

re
often limited by severe cardiac dysfunction, and no specific Evaluation

sf
ks

ks

ks
pulmonary function criteria have been formally established

k
for device exclusion. Impaired cognitive function or a recent Due to the comprehensive nature of device therapy and the
oo

oo

oo

oo
history of stroke obligates a thorough neurologic and cog- commitment required for successful outcomes, candidates
nitive examination. Patients with significant residual deficits for MCS must undergo a comprehensive multidisciplinary
eb

eb

eb

eb
should undergo an occupational therapy evaluation to deter- assessment. Preoperative nutritional status should be evalu-
m

mine whether they possess sufficient physical dexterity to
manipulate an MCS device independently.
m
m
ated and optimized, as studies have shown an association
between cachexia and perioperative mortality rates.95,96 A
history of psychiatric illness should cause significant concern
and warrants a thorough evaluation of competency and abil-

om
Right Ventricular Function ity to adhere consistently to therapy. Routine psychosocial
co

co

co

co
Prevention and treatment of RV dysfunction is a signifi- evaluation should be performed by qualified healthcare per-

c
e.

e.

e.

e.
cant challenge in the field of MCS, and postoperative RHF sonnel to explore available personal, social, vocational, finan-
has long been associated with poor clinical outcomes.37,81,82 cial, and environmental support systems. Advanced patient
re

fre

re

re
Identification of patients at risk for RHF provides an oppor- age alone should not preclude device therapy.97 Cases should
sf

f
ks

ks

ks
tunity for implementation of preoperative optimization be reviewed on an individual basis by a multidisciplinary
k

strategies and enhanced surgical preparedness. However, committee to ensure that all concerns are addressed before
oo

oo

oo

oo
methods developed thus far to predict RHF after LVAD surgery.
eb

eb

eb

eb
implantation have been largely based on experience with
pulsatile devices and have proven inconsistent, likely due to
DEVICE SELECTION
m

m
the variety of clinical definitions, variables, and treatment
strategies.83-86 Appropriate device selection is determined by a number of
Despite the complex pathophysiology underlying this factors, including the patient’s overall clinical condition, the
m

m
disease process, several recent studies have significantly indication for MCS, the proposed goal of device therapy, and
improved our understanding of variables that predispose to relevant patient-specific considerations.
co

co

co

co

co
RV dysfunction. In a retrospective analysis of the HeartMate Patients with an unstable condition or acute decompensa-
e.

e.

e.

e.
II BTT trial, Kormos and colleagues87 sought to identify risk tion are often best served by short-term devices that are less
fre

fre

fre

fre
factors for the development of postoperative RHF in a CF invasive, faster to place, and less expensive than fully implant-
LVAD population. That study specifically identified preop- able pumps. As a reliable and historically familiar option, VA
ks

ks

ks

ks
erative ventilator support, an increased central venous pres- ECMO is the procedure of choice for emergent situations
sure (CVP) and CVP/PCWP ratio, a decreased RV stroke in many institutions worldwide. Peripheral cannulation may
oo

oo

oo

oo
work index, and elevated white blood cell and blood urea be performed at the bedside, in the catheterization lab, or
eb

eb

eb

eb
nitrogen values as significant predictors of RHF. Echocar- in the operating theater, and the presence of an oxygenator
diographic parameters have also been suggested as tools for allows for full cardiopulmonary support for a period of
m

assessing the risk of RHF after LVAD implantation. Ultra-
sonographic studies have implicated reduced free-wall strain,
RV-to-LV diameter ratios, severity of tricuspid regurgitation,
m
m
days to weeks. Catheter-based LVADs such as the Impella
and TandemHeart are used with increasing frequency for
short-term LV support, particularly after high-risk coronary

m
and RV EF as predictors of right-sided heart dysfunction.88-91 interventions.32 These percutaneous devices can provide tem-
co

co

co

co

co
Researchers at the University of Pennsylvania recently devel- porary support without the need for surgical access, but they
oped a risk-stratification tool to assess the preoperative like- are susceptible to hemolysis and positional disruption, and
e.

e.

e.

e.
lihood for tolerance of univentricular support.92 Although their use is typically limited to 1 to 2 weeks. For patients for
re

fre

fre

limited by its size, patient selection, and single-center design, whom more robust flows are required or percutaneous mea-
fre
sf

the study identified preoperative tachycardia, elevated CVP, sures have failed, the CentriMag is the most frequently used
ks

ks

ks
ventilator support, RV dysfunction, and tricuspid regurgita- temporary option. Cannulation is similar to that performed
k
oo

oo

oo

oo

tion as predictors of the need for biventricular support. for routine CPB, and the device can be used for temporary
Thus, whereas the specific variables that may predispose uni- or biventricular support for up to 1 month.
eb

eb

eb

eb

patients to RHF after LVAD implantation remain unclear, Fully implantable CF LVADs are the devices of choice
certain principles have emerged to help guide patient selec- for patients who require long-term MCS and whose condi-
m

tion and management. A consistent theme throughout the tion is stable enough to permit major surgery. Of the vari-
majority of these studies is an association between predic- ous options, the HeartMate II LVAD and the HeartWare
tors of RHF and predictors of overall mortality.93,94 Sicker HVAD are the most commonly used devices. Although
m

patients—with preoperative evidence of end-organ dysfunc- they function in a similar manner, the choice between these
co

co

co

co

co

tion, hemodynamic instability, and impaired RV mechanics— pumps is based on factors ranging from anatomical con-
are especially vulnerable to RHF after LVAD implantation. cerns to institutional availability. In the United States, the
e.

e.

e.

e.

Thoughtful evaluation and preoperative consideration of CMS strictly mandates preoperative designation of patients
fre

fre

fre

fre

these risks are paramount in surgical planning. as either BTT or DT candidates. Despite the fluid nature
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1372 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1373
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
of these preimplant strategies, the HeartMate II remains the ensure an optimal pump position and avoidance of kinking;
fre

fre

fre

re
only device approved by the FDA for use in both BTT and (4) thorough de-airing of the heart and pump circuit; and

sf
ks

ks

ks
DT patients. The HVAD, by contrast, is currently approved (5) titration of pump speeds under echocardiographic guid-

k
only for BTT in the United States, and clinical trials seeking ance to optimize heart function.
oo

oo

oo

oo
DT approval are currently underway. No study has yet shown A vertical midline incision is made, followed by a 6-cm
either pump to be superior to the other. Although the advent subxiphoid extension. A standard sternotomy is then per-
eb

eb

eb

eb
of smaller CF pumps has lessened the anatomical challenge formed with the ratchet of the sternal retractor oriented
m

of implantation, patient body habitus remains an important
consideration. For example, the intrapericardial position of
the HVAD makes the HeartWare device particularly appeal-
ing for smaller-framed patients.
m
m
toward the patient’s head to provide adequate inferior expo-
sure. The pericardium is opened in the midline and divided
along the length of the diaphragm. Both pleural cavities are
entered to assist with positioning of both the pump and out-

om
For patients with biventricular HF, the decision to pursue flow conduit. The anterior border of the diaphragm is then
co

co

co

co
uni- or biventricular MCS greatly affects the available treat- incised from the midline to the apex of the heart, permit-

c
e.

e.

e.

e.
ment options. For patients with mild-to-moderate RHF, a CF ting entrance into the peritoneal cavity and adequate space
LVAD with temporary right-sided support may be sufficient for introducing the pump (Fig. 62-9a). A point of entry for
re

fre

re

re
to allow RV reconditioning. In the event of failed temporary the inlet cannula to traverse the diaphragm is selected to cor-
sf

f
ks

ks

ks
support or severe biventricular disease, dual (right and left) respond with the diaphragmatic surface of the left ventricle
k

CF devices may be used with relative success despite the logis- and is marked with an electrocautery.
oo

oo

oo

oo
tical complexities associated with multiple controllers and After systemic heparinization is achieved, the ascending
eb

eb

eb

eb
batteries. The pulsatile SynCardia TAH (SynCardia Systems; aorta is cannulated distally, with consideration for potential
Tucson, AZ) currently is the only FDA-approved option for future reoperation and transplantation. Venous cannulation
m

implantable biventricular support in patients awaiting trans-
plantation. Although the Syncardia requires considerable
intrathoracic volume displacement, it allows mobility and
m
m
is typically achieved by using a dual-stage cannula in the
right atrial appendage, except in cases for which a concomi-
tant procedure is planned—such as a tricuspid valve repair

m
outpatient management by means of a portable pneumatic or PFO—when bicaval cannulation is preferred. In a reop-
driver. The continued development of a fully implantable CF erative sternotomy, the femoral vessels should be exposed to
co

co

co

co

co
TAH is underway and will be discussed later in this chapter. allow immediate cannulation if mediastinal dissection results
e.

e.

e.

e.
in hemodynamic instability.
fre

fre

fre

fre
CPB is then initiated, and the LV apex is brought out of
SURGICAL TECHNIQUE the chest and stabilized with an off-pump suction-cup
ks

ks

ks

ks
As experience with device implantation has grown, surgical stabilizer. For inlet placement along the diaphragmatic sur-
techniques have evolved to reflect lessons learned. Early in face, identification of the appropriate ventriculotomy site
oo

oo

oo

oo
the development of LVADs, apical positioning of the inlet is crucial to avoid disruption of the papillary muscles and
eb

eb

eb

eb
cannula was advocated, so as to take advantage of the longest allow the inlet to lie parallel to the interventricular septum
measured span within the left ventricle. This positioning, how- along the short axis of the left ventricle. This is accomplished
m

ever, often creates unfavorable angulation of the inlet toward
the interventricular septum, resulting in arrhythmogenesis
and mechanical obstruction of the inlet after a reduction
m
m
by selecting a point approximately one-third of the distance
from the apex to the base of the heart (thus, anterior to the
origin of the papillary muscles), with the medial edge of

m
in the LV end-diastolic dimension. Over time, a number of the sewing ring placed 0.5 to 1 cm lateral to the posterior
co

co

co

co

co
alternative implant strategies have been devised, ranging from descending artery. A coring knife—typically provided by the
a thoracotomy to subcostal incisions and from anterior to lat- device manufacturer—is used to open the ventricle, with care
e.

e.

e.

e.
eral pump positions.98-100 Surgeons at our institution favor a to stay parallel to the septum and follow posteriorly in the
re

fre

fre

method designed to produce a parallel orientation between

fre
direction of the mitral valve. The ventricular cavity is then
sf

the inlet cannula and the interventricular septum by placing thoroughly inspected for evidence of thrombus or potentially
ks

ks

ks
the inlet along the diaphragmatic surface of the heart.101,102 obstructive trabeculae.
k
oo

oo

oo

oo

Although it deviates from traditional technique, this method To secure the Silastic inflow cuff of the HeartMate II, 12
eliminates the need for a preperitoneal pump pocket and cre- braided polyester full-thickness sutures are placed circumfer-
eb

eb

eb

eb

ates a more geometrically advantageous alignment. entially around the ventriculotomy site in a horizontal mat-
Whereas the technique described is specifically intended for tress fashion with broad interlocking felt pledgets. The sutures
m

the HeartMate II, most of the procedure is indistinguishable are passed through the felt sewing ring of the inflow cuff,
from implantation of other CF LVADs. Differences involving which is then lowered into position and secured. Hemosta-
the HeartWare HVAD—the second most commonly used sis around the inflow is bolstered by placing a full-thickness
m

device worldwide—will be addressed as necessary. Essential purse-string stitch through the pledgeted ring, using a large-
co

co

co

co

co

components of the procedure include (1) positioning of the caliber monofilament suture (Fig. 62-9b). The HeartWare
inlet cannula along the diaphragmatic aspect of the left ven- HVAD utilizes a specialized sewing ring equipped with a
e.

e.

e.

e.

tricle; (2) transdiaphragmatic positioning of the HeartMate II locking mechanism to secure the pump in place. The HVAD
fre

fre

fre

fre

pump housing; (3) proper sizing of the outflow graft to sewing ring is typically placed before the ventricle is cored,
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1373 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1374
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
fre

fre

fre

re
sf
ks

ks

ks

k
oo

oo

oo

oo
eb

eb

eb

eb
m

m
m

om
co

co

co

co

c
e.

e.

e.

e.
re

fre

re

re
sf

f
ks

ks

ks
k
oo

oo

oo

oo
A
eb

eb

eb

eb
m

m
D
m

m
co

co

co

co

co
e.

e.

e.

e.
fre

fre

fre

fre
ks

ks

ks

ks
oo

oo

oo

oo
eb

eb

eb

eb
m

m
m

m
co

co

co

co

co
e.

e.

e.

e.
B
re

fre

fre

fre
sf

ks

ks

ks
k

E
oo

oo

oo

oo
eb

eb

eb

eb

FIGURE 62-9 HeartMate II implantation. (A) Retraction of the

m

diaphragm in anticipation of subdiaphragmatic pump positioning.

(B) The ventricle is cored along the diaphragmatic surface of the
heart—lateral to the posterior descending artery—and a plegeted ring
secures the Silastic inflow cuff. (C) Transdiaphragmatic positioning
m

of the HeartMate II inflow cannula. (D) Anastomosis of the outflow

co

co

co

co

co

graft to the ascending aorta, aided by a partial occluding vascular

clamp. (E) Final alignment of the pump and outflow conduit. The
e.

e.

e.

e.

pump housing lies beneath the diaphragm, and the outflow graft
courses toward the right side of the chest. Introduction of a 19-gauge
fre

fre

fre

fre

C needle permits de-airing of the system.

s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1374 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1375
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
and the ring is seated with a similar series of horizontal POSTOPERATIVE MANAGEMENT
fre

fre

fre

re
mattress sutures.

sf
Diligent and comprehensive patient care is essential for suc-
ks

ks

ks
Next, the diaphragmatic myotomy—ample enough to

k
admit two fingers—is completed at the previously marked cessful MCS outcomes, and proper management begins
oo

oo

oo

oo
location, and the pump is brought into the field after being before surgery. As described earlier, careful patient selection is
prepared and primed on a separate table. The inlet cannula is paramount, as is prompt identification of the most appropri-
eb

eb

eb

eb
guided through the diaphragm, inserted within the Silastic ate opportunity for device implantation. To minimize surgical
m

ring, and finally secured in place with two ratcheting cable
ties (Fig. 62-9c). The pump housing is pulled into the abdo-
men until the heart lies flush with the diaphragm, and the
pump is positioned above the left lobe of the liver. To pro-
m
m
risks, particular emphasis should be placed on the preoperative
optimization of end-organ function, reversal of coagulation
imbalances, correction of malnutrition, and eradication of
preexisting infection. Device recipients with decompensated

om
tect the bowel, the body of the pump is wrapped in available clinical profiles (INTERMACS levels 1 and 2) are known to
co

co

co

co
omentum. The discoid shape of the HeartWare HVAD allows experience worse outcomes than patients whose condition is

c
comparatively stable.37 As a result, aggressive medical man-
e.

e.

e.

e.
intrapericardial pump placement. Although the inlet is placed
along the diaphragmatic aspect of the ventricle, entrance into agement, intraaortic balloon therapy, or temporary device
re

fre

re

re
the abdomen is not required. support is often warranted to improve the patient’s hemody-
sf

f
namic condition before surgery. Careful planning should also
ks

ks

ks
Next, attention is turned to the outflow graft, which is
k

allowed to fill, then is measured and cut in beveled fashion, include the recognition of patient-specific risks—such as the
oo

oo

oo

oo
with enough length to produce a gentle curve toward the potential need for RV support—to avoid lengthy delays in the
recognition and treatment of postoperative complications.
eb

eb

eb

eb
right side of the chest but without excessive redundancy that
might lead to kinking. After placement of a partial occlud- Intraoperative measures are also crucial for prevent-
m

ing clamp along the ascending aorta, an aortotomy is cre-
ated along the anterolateral surface by sequentially firing a
4.4-mm punch. The outflow-graft anastomosis is performed
m
m
ing postoperative complications. In addition to meticulous
surgical technique and adequate hemostasis, judicious
blood-product administration minimizes the risk of volume
overload and RV failure. In the setting of profound coagulop-

m
in an end-to-side fashion with continuous 5-0 polypropyl-
ene suture (Fig. 62-9d). The partial occluding clamp is then athy, low-dose recombinant factor VIIa (NovoSeven; Novo
co

co

co

co

co
removed, and the graft is de-aired while backflow is prevented Nordisk, Plainsboro, NJ) and four-factor prothrombin com-
e.

e.

e.

e.
by an occlusive vascular clamp. plex concentrate (Kcentra; CSL Behring GmbH, Marburg,
Germany) have shown value as pharmacologic resources
fre

fre

fre

fre
The driveline exit site is identified two fingerbreadths
below the right subcostal margin within the mid-clavicular without causing an undue risk of device complications or
ks

ks

ks

ks
line. A skin punch allows the introduction of a tunneling thromboembolic events.105
device, which is then passed behind the right rectus mus- Initial device setup is a critical process that occurs within
oo

oo

oo

oo
cle, anteriorly to the posterior rectus sheath. The driveline the operating theater. After the pump is activated, power and
flow recordings should be carefully monitored, and any sus-
eb

eb

eb

eb
is withdrawn with the tunneling device, cleaned, and con-
nected to the device controller, with care to loop the velour- pected physical impediments to device performance, such as
m

coated portion of the lead within the patient’s body, as studies
have shown that preservation of the velour-silicone interface
beneath the level of the skin is associated with a decreased
m
m
outflow kinking or cannula malpositioning, should be imme-
diately identified and corrected. Intraoperative identification
of optimal pump speed is a dynamic process that requires

m
incidence of driveline infections.103 both direct visualization and TEE to assess ventricular cham-
ber size, RV function, and interventricular septal position.
co

co

co

co

co
After thorough de-airing, the pump is started at its low-
est setting (6000 rpm for the HeartMate II; 1800 rpm for The combination of CPB weaning, intravascular fluid shifts,
e.

e.

e.

e.
the HeartWare HVAD) with a venting needle in the highest vasoactive medications, and sternal closure can result in sud-
re

fre

fre

fre
point of the outflow graft (Fig. 62-9e). CPB flows are gradu- den and dramatic changes in cardiac function, and continual
reassessment is required to maintain an appropriate balance
sf

ally decreased while the pump speed is adjusted under trans-

ks

ks

ks
esophageal echocardiographic (TEE) guidance to optimize between the patient and device.
k

Immediate postoperative management in the intensive

oo

oo

oo

oo

the chamber size, position of the interventricular septum,

degree of MR, and RV function. care unit (ICU) requires an understanding of the delicate bal-
eb

eb

eb

eb

Protamine is administered, the cannulas are removed, and ance between pump flow, RV function, and intravascular vol-
drains are placed in the mediastinum and pleural spaces. The ume status—often interpreted within the context of a volatile
m

bare portion of the outflow graft is covered with a 20-mm- and complex hemodynamic milieu. Baseline comprehensive
diameter ringed Gore-Tex graft (Gore Medical, Newark, DE) laboratory panels should be assessed immediately upon the
to avoid kinking and damage during future sternal reentry. The patient’s arrival in the ICU to assess acid-base status, coagula-
m

defect in the diaphragm is partially reapproximated, and the tion markers, and hematologic reserves. Acidosis should be
aggressively corrected and coagulopathy reversed with appro-
co

co

co

co

co

sternum and soft tissues are closed in the standard fashion. If

the patient has excessive bleeding related to severe coagulopathy, priate blood component resuscitation.
e.

e.

e.

e.

temporary chest closure may be advisable to allow correction and Arterial BP is principally controlled by vasoactive medi-
fre

fre

fre

fre

stabilization before delayed sternal closure is attempted.104 cations and intravascular fluid management, not pump speed.
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1375 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1376
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
The flows produced by CF LVADs operating at constant be weaned from vasoactive and inotropic medications as tol-
fre

fre

fre

re
speed are highly dependent on the pressure differential erated. Lines and drains should be removed as early as pos-

sf
ks

ks

ks
across the pump. The resulting afterload sensitivity, particu- sible to reduce the risk of postoperative infection, and strict

k
larly in centrifugal pump designs, requires careful BP titra- aseptic technique is required for driveline dressing changes.
oo

oo

oo

oo
tion to ensure sufficient output without causing excessive An aggressive physical therapy regimen is initiated as early as
systemic pressures. Although optimal MAPs in this patient possible, as is patient device training.
eb

eb

eb

eb
population remain unknown, our practice is to target a
m

m
range of 60 to 80 mm Hg. Due to the reduction in pulse Management of Right
pressure created by CF LVADs, systemic arterial pressures
are expressed as a single mean value and should be continu-
Ventricular Dysfunction
ously monitored in the postoperative period. Noninvasive After LVAD implantation, outcomes are heavily dependent
m

om
BP cuff measurements have been found to be unreliable in on the preservation of adequate RV function. Whereas RHF
co

co

co

co
the setting of CF MCS and frequently underestimate sys- is most often the result of inadequate left-sided output, the

c
e.

e.

e.

e.
tolic pressures. As a result, Doppler measurements have initiation of LVAD support can adversely affect RV func-
emerged as the gold-standard noninvasive modality.106 The tion in several ways. The sudden increase in left-sided car-
re

fre

re

re
control of CVP is also of critical importance in the postop- diac output and venous return requires the unsupported right
sf

f
ks

ks

ks
erative period, as excessive volume administration—a fre- ventricle to respond immediately to an increased demand.
k

quent obstacle in this patient population—can quickly lead Although LVAD therapy has been shown to reduce the sever-
oo

oo

oo

oo
to development of RV dysfunction. ity of pulmonary hypertension, this resolution may not be
eb

eb

eb

eb
Assessment of appropriate device function is based on immediate, forcing the right ventricle to pump against a rela-
system-specific parameters in conjunction with echocardio- tively high-pressure system. Furthermore, interdependence
m

graphic findings. Pump speed (rpm) and power expenditure
(watts) are the only directly measured variables provided by
the HeartMate II and HeartWare HVAD; estimated flow
m
m
between the right and left ventricles can result in the distor-
tion of RV and septal architecture upon unloading of the LV,
leading to impaired RV contractility. In this way, initiation of

m
(L/min) is indirectly calculated and typically corresponds to device support can lead to sudden demands and strains on the
changes in pump power. The pulsatility index (PI) is a param- right ventricle to which it cannot adequately respond.
co

co

co

co

co
eter unique to the HeartMate II, and it is designed to indicate Management personnel must be vigilant in monitor-
e.

e.

e.

e.
ventricular filling and function by interpreting changes in ing for signs of acute RHF to allow prompt and aggressive
fre

fre

fre

fre
pump flow due to contraction of the left ventricle. Although treatment. Signs of impending or evolving RV dysfunction
optimal target levels are unknown, the PI can be expected include increasing vasopressor requirements or hemody-
ks

ks

ks

ks
to decline with increasing pump support.69 The HeartWare namic instability, elevated or rising CVP, decreased urine
HVAD provides continuous readings of pump speed, power, output, and low mixed venous oxygen saturation (SvO2).
oo

oo

oo

oo
and flow, and it records values every 15 minutes, thus main- In 2011, Potapov and colleagues108 suggested standardized
eb

eb

eb

eb
taining a retrievable data log for review of the preceding criteria for the diagnosis of RV dysfunction that included
30 days. Interpretation of these data recordings has shown the inability to wean from CPB, or any two of the following
m

a potential for predicting, diagnosing, and managing clini-
cal events, including volume changes, RV dysfunction, and
pump thrombosis.107
m
m
conditions: MAP < 55 mm Hg, CVP > 16 mm Hg, SvO2 <
55%, LV flow rate index < 2 L/min/m2, or significant ino-
tropic dependence. In addition to hemodynamic markers,

m
Any deviation in pump data warrants immediate interro- echocardiography should remain an essential component of
co

co

co

co

co
gation, including assessment of hemodynamics, volume sta- the evaluation, allowing direct assessment of ventricular size,
tus, heart rhythm, sources of intrathoracic compression, and contractility, and septal positioning. Along with traditional
e.

e.

e.

e.
ventricular dimensions. Whether emergent or elective, assess- signs of depressed right-sided heart function, the failure to
re

fre

fre

ment of device function is best achieved by echocardiogra- maintain appropriate LVAD power consumption and flows
in the absence of hypovolemia may suggest inability of the fre
sf

phy focused on end-diastolic dimensions, septal positioning,

ks

ks

ks
valvular competence, and aortic valve opening. Systematic right ventricle to deliver sufficient volume to the pump inlet.
k
oo

oo

oo

oo

assessment across a range of pump speeds—“speed-change” The majority of patients with RV dysfunction will require
or “ramped-speed” echo—allows optimization of mainte- inotropic support, typically in the form of milrinone, dobu-
eb

eb

eb

eb

nance pump speeds by comparing functional dynamics. The tamine, isoproterenol, or epinephrine. Early and aggressive
preferred frequency of these studies varies among institutions; pharmacologic diuresis is essential to reduce RV preload, par-
m

at a minimum, it should be performed in the operating room, ticularly in the setting of excessive volume loading and mas-
soon after arrival in the ICU, before hospital discharge, and sive transfusion. Maintenance of sinus rhythm preserves the
any time concern is raised about the adequacy of support. RV stroke volume and lessens myocardial strain. Although
m

Our practice is to seek early extubation, mobilization, the results of studies have been contradictory, inhaled nitric
co

co

co

co

co

and resumption of nutritional support. Aggressive diure- oxide therapy appears to have utility in alleviating pulmo-
sis is typically warranted for postoperative fluid offloading, nary vascular resistance and RV dysfunction through direct
e.

e.

e.

e.

and volume status is monitored using daily weights, physical pulmonary arterial vasodilatory effects.108,109 When all other
fre

fre

fre

fre

findings, and hemodynamic parameters. The patient should management options have been exhausted, consideration of
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1376 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1377
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
mechanical right-sided support, either temporary or perma- offloading without development of pulmonary edema—a
fre

fre

fre

re
nent, is warranted.110 common complication of RV assistance as a result of reduced

sf
ks

ks

ks
afterload established by the low-pressure pulmonary circula-

k
Anticoagulation tion. Left-sided speeds may then be actively titrated to com-
oo

oo

oo

oo
pensate for physiologic variations in the volume delivered to
The approach to anticoagulation in patients who require the left ventricle.
eb

eb

eb

eb
long-term MCS has not yet been standardized and varies Biventricular assist device (BiVAD) support also creates
m

considerably between institutions. Whereas thromboem-
bolic complications were a significant challenge for early
devices, the unique textured interior surface of the Heart-
Mate XVE resulted in the development of a neointimal blood
m
m
a number of logistical challenges. The surgical complexity
of accommodating two pump housings within the thoracic
cage can be significant. Whereas the use of dual HeartMate II
devices has been reported, the relatively small profile and

om
interface that effectively eliminated the need for systemic intrapericardial pump position of the HeartWare HVAD
co

co

co

co
anticoagulation.111 The introduction of the Jarvik 2000, offers a considerable advantage, making this device preferable

c
e.

e.

e.

e.
which utilized blood-washed bearings and a CF design, led to for high-risk cases or planned biventricular support.116-118 In
the widespread practice of systemic anticoagulation to reduce our experience, the outflow conduits are best configured in
re

fre

re

re
the perceived risk of pump thrombosis.112 As a result, initial opposing alignments, with the right-sided graft directed pos-
sf

f
ks

ks

ks
management strategies for CF devices involved a target INR terolaterally to allow anastomosis with the pulmonary artery
k

of 2.5 to 3.5 before experience gained from the HeartMate II trunk without interfering with the course of the LVAD graft.
oo

oo

oo

oo
pivotal trial led to a suggested target range of 2 to 3 along A further disadvantage of dual device support is the redun-
eb

eb

eb

eb
with early postoperative heparin infusion and antiplatelet dancy of system controllers (two) and batteries (four) required
therapy in the form of aspirin and dipyridamole.25,113 for operation. The resultant physical burden, complexity in
m

Despite these initial recommendations, the appropriate
level of anticoagulation for long-term CF support remains an
area of active investigation. Studies have shown that the inci-
m
m
daily use, and responsibilities for maintenance present signifi-
cant challenges for long-term management of patients who
require biventricular therapy.

m
dence of hemorrhagic complications is far greater than that of
thrombotic events in this patient population, leading to the
co

co

co

co

co
suggestion that a lower INR target may be appropriate.37,114 POSTOPERATIVE OUTCOMES
e.

e.

e.

e.
Moreover, patients transitioned directly to warfarin with- Over the past 20 years, multicenter studies have documented
fre

fre

fre

fre
out an early heparin bridge have been shown to experience a progressive improvement in postoperative mortality after
reduced postoperative transfusion requirements without an LVAD implantation (Fig. 62-10). The landmark REMATCH
ks

ks

ks

ks
increased risk of thrombotic complications.115 Guidelines trial showed a 1-year survival rate of 52% for patients ran-
published by ISHLT recommend the initiation of heparin domized to receive device therapy, compared with 25% sur-
oo

oo

oo

oo
therapy on postoperative day 1, with warfarin and aspirin vival at 12 months for those treated with medical therapy
eb

eb

eb

eb
started by postoperative day 3 (target INR 2.0 to 3.0).36 Our alone.7 Six years later, the HeartMate II BTT study showed
institution follows an alternative protocol, preferring the 68% 1-year survival with CF support, a total that rose to
m

early introduction of warfarin and aspirin therapy (by post-
operative day 1), with heparin infusion reserved for patients
in whom anticoagulation remains subtherapeutic by postop-
m
m
73% within the larger Continued Access Protocol report.9,25
In keeping with this trend, the HeartWare ADVANCE
trial results were released in 2012 and showed an 86% sur-

m
erative day 4. Target INRs for this protocol are specific to the vival rate at 1 year, which was comparable to contemporary
co

co

co

co

co
type of device, application, and presence of concomitant car- INTERMACS data and led to FDA approval of the HVAD
diac pathology. Standard intraoperative heparin reversal with for BTT indications.27 Likely due to improved patient selec-
e.

e.

e.

e.
protamine sulfate is accepted practice in most centers. tion, 30-day perioperative survival has also improved during
re

fre

fre

that timespan, rising from roughly 80% with the HeartMate

fre
sf

XVE to more than 90% with contemporary devices.53,119,120

Biventricular Device Management
ks

ks

ks
Analysis of the INTERMACS registry reveals several sta-
k
oo

oo

oo

oo

The concomitant use of a right-sided pump adds several com- tistically significant risk factors for mortality in patients sup-
plexities to device management. From a hemodynamic stand- ported by CF LVADs, including severity of the preoperative
eb

eb

eb

eb

point, determination of the appropriate balance between the clinical profile (levels 1 and 2), renal dysfunction, age over
two pumps can be difficult. Continuous-flow devices are 70 years, RV dysfunction, and increased surgical complexity.8,121
m

principally governed by two variables: the speed of the rotor Biventricular device support is consistently associated with
and the pressure difference across the device. As a result, the increased short-term mortality, although the data suggest
output produced by a CF pump operating at constant speed some utility in early planned (as opposed to delayed) imple-
m

will reflect the volume of blood delivered to the inlet. This mentation of biventricular device therapy.37,122 Importantly,
co

co

co

co

co

arrangement provides a degree of inherent automaticity that LVAD therapy has been shown to be equivalent to transplan-
is advantageous when pumps are operated in series. Man- tation in terms of 1- and 2-year mortality, and pretransplant
e.

e.

e.

e.

agement of a biventricular system, therefore, requires iden- MCS has not been shown to negatively affect outcomes for
fre

fre

fre

fre

tification of a right-sided pump speed that permits venous subsequent transplantation.123,124 Quality-of-life metrics also
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1377 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1378
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
100 0.10

fre

fre

fre

re
0.09

sf
ks

ks

ks
n=9372, deaths=2272 0.08
80

k
oo

oo

oo

oo
0.07

Deaths/months
Months % survival

% Survival
60 0.06
eb

eb

eb

eb
1 95%
12 80% 0.05
m

m
24 69% 0.04
40
36 59%
48 47% Hazard 0.03
20 0.02
m

om
0.01
co

co

co

co

c
0 0.00
e.

e.

e.

e.
0 6 12 18 24 30 36 42 48
n at risk: 9372 6203 4027 2739 1821 1194 711 408 174
re

fre

re

re
Months post implant
sf

f
ks

ks

ks
FIGURE 62-10  Actuarial survival curve for adult recipients of primary continuous-flow left ventricular support. Patients were censored at trans-
k
oo

oo

oo

oo
plantation or device explantation. (Reproduced with permission from Kirklin JK, Naftel DC, Pagani FD, et al: Sixth INTERMACS annual report:
a 10,000-patient database, J Heart Lung Transplant. 2014 Jun;33(6):555-564.)
eb

eb

eb

eb
m

m
reveal a significant improvement with LVAD therapy that is and management of MCS, thus, obligates a consideration of
sustained through 24 months after device implantation.37 the risks associated with device therapy and an appreciation
of the implications of certain events.
m

m
ADVERSE EVENTS AND
co

co

co

co

co
DEVICE COMPLICATIONS Perioperative Bleeding
e.

e.

e.

e.
The implementation of device therapy is associated with spe- Whereas detailed analysis is limited with regard to the cur-
fre

fre

fre

fre
cific perioperative and long-term risks for the patient. To stan- rently available data registries, bleeding has historically been
dardize the evaluation of patient outcomes, INTERMACS cited as the most common complication after LVAD implan-
ks

ks

ks

ks
has established definitions for adverse events associated with tation, and bleeding of any type (operative and late onset)
oo

oo

oo

oo
MCS. The cumulative incidence of these adverse events is the most common adverse event recorded by INTER-
within the first 60 days of MCS has been reported to be as MACS.25,37,126 Surgical bleeding and large-volume blood-
eb

eb

eb

eb
high as 89%, many such events being associated with a reduc- product resuscitation has been associated with increased
tion in long-term survival125,126 (Fig. 62-11). Implementation 30-day and 1-year mortality rates, RV dysfunction, acute
m

m
lung injury, infection, and pretransplant allosensitization.127
Reported bleeding rates were high in the early pulsatile LVAD
100 experience, with nearly 50% of patients suffering from major
m

m
90 *Major event: first occurrence of infection, bleeding complications.81 Although the introduction of CF
bleeding device malfunction, stroke or death
co

co

co

co

co
80 devices resulted in an overall improvement in outcomes,
% Freedom from event

Months % Freedom bleeding has remained a significant perioperative challenge.

e.

e.

e.

e.
70
Patients 5436, events = 3611 1 59% Early experience with the HeartMate II clinical trials showed
re

fre

fre

fre
60 3 48% that between 50 and 80% of patients required perioperative
6 40%
sf

50 blood product transfusion, with 30% of patients undergo-

ks

ks

ks
12 30%
k

40 24 19% ing reoperation for bleeding complications.25,26 Single-center

oo

oo

oo

oo

30 36 14% analyses have since revealed a modest reduction in transfusion

requirements.104,128
eb

eb

eb

eb

20
Recently, the comparative analysis of different devices
10
m

with regard to perioperative bleeding complications has been

0 an area of particular interest. The HeartWare BTT trial inves-
0 6 12 18 24 30 36 42 48
Months post implant
tigators reported a lower risk of reoperation for bleeding than
did the contemporary INTERMACS control group (largely
m

FIGURE 62-11 Actuarial freedom from major adverse events in comprising HeartMate II patients).27 A single-center analy-
co

co

co

co

co

patients supported by continuous-flow left ventricular assist devices.

(Reproduced with permission from Kirklin JK, Naftel DC, Kormos
sis subsequently determined that although red blood cell
administration was equivalent for the two pumps, HVAD
e.

e.

e.

e.

RL, et al: Fifth INTERMACS annual report: risk factor analysis from
more than 6,000 mechanical circulatory support patients, J Heart recipients required fewer blood products overall (8.3 ± 13 vs
fre

fre

fre

fre

Lung Transplant. 2013 Feb;32(2):141-156.) 12.6 ± 14 units) and had a lower chest-tube output than did
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1378 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1379
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
HeartMate II patients.129 As a result, these studies have been is typically greater than 6 months, implying that most such
fre

fre

fre

re
used to suggest that, due to a reduction in surgical dissection, infections occur in the outpatient setting, when patients

sf
ks

ks

ks
the HVAD is associated with fewer bleeding complications. are more active and responsible for dressing care.132-134 By

k
However, whereas differences between pumps may indeed 1 year, 19% of patients have developed a driveline infec-
oo

oo

oo

oo
affect bleeding rates, these studies fail to account for wide tion, while 1 in 4 are diagnosed with a driveline infection by
variations in surgical approach and perioperative manage- 2 years of CF LVAD support.37 Whereas most patients in the
eb

eb

eb

eb
ment that may contribute to these results. Regardless of the INTERMACS database were treated with antibiotics alone,
m

m
device used, bleeding is a significant challenge in the periop- surgical intervention was required in 12.5% of cases.37,132
erative period. Specific information regarding the extent of debridement is
generally not reported, but vacuum-assisted closure, antibi-
otic bead therapy, and driveline mobilization have all been
m

om
Right-Sided Heart Failure described.135,136 Infection has been shown to adversely impact
co

co

co

co
survival, increasing the 1-year mortality rate as much as

c
The identification and management of RV dysfunction has
e.

e.

e.

e.
been extensively discussed. Development of RHF presents 6-fold.37,132,137 The presence of a documented device infection
does not preclude transplantation.
re

fre

re

re
a major postoperative challenge and is strongly associated
with poor clinical outcomes and increased mortality.37,81,82
sf

f
ks

ks

ks
Reported criteria for the diagnosis of RV failure vary in the
Renal Failure
k

literature. The most commonly used definition of post-LVAD

oo

oo

oo

oo
RV failure originates from the HeartMate II BTT trial, which As described earlier, preoperative renal dysfunction has been
eb

eb

eb

eb
described RV failure as the need for RV support, inotropic shown to be strongly associated with poor outcomes after
support for at least 14 days after implantation, or the initia- LVAD implantation.76,77 Similarly, the development of post-
m

tion of inotropic support after 14 days of LVAD support.9
INTERMACS utilizes considerably broader criteria for data
acquisition purposes, relying primarily on signs and symp-
m
m
operative renal failure is a significant predictor of the 1-year
mortality rate after device implantation.125 However, risk fac-
tors for postoperative renal failure are poorly understood, and

m
toms of elevated CVP and more specific conditions to evalu- indices of preoperative renal status do not appear to be strong
co

co

co

co

co
ate the level of severity.130 The reported incidence of RV independent predictors of such failure.77 Although defini-
failure after LVAD implantation remains highly variable, tions vary, the incidence of postoperative renal failure ranges
e.

e.

e.

e.
ranging from 15 to 30%.82,87,123,126 A subset of patients with from 14 to 30%.125,138,139 Factors associated with the devel-
fre

fre

fre

fre
RV failure (between 8 and 10% of device recipients) will opment of this complication appear to include preoperative
require mechanical RV support.37,88,91,126 ventilatory support, elevated CVP, and advanced patient age.
ks

ks

ks

ks
The initiation of MCS has been shown to enhance end-organ
oo

oo

oo

oo
perfusion, and management options for postoperative renal
Infection failure range from supportive medical therapy to temporary
eb

eb

eb

eb
Device therapy carries a significant risk of infection, primar- or permanent renal replacement.73 As many as 10 to 20% of
ily due to the presence of foreign material within the body patients undergoing LVAD implantation require some form
m

and a percutaneous driveline. With early pulsatile systems,
infection rates ranged from 50 to 60% during the lifetime
of the pump, and more than 40% of deaths in the LVAD-
m
m
of postoperative renal replacement therapy, and approximately
half of that population can be expected to recover indepen-
dent renal function, typically within the first month.138,139

m
supported cohort of the REMATCH trial were due to ful- Development of postoperative renal failure has been shown
co

co

co

co

co
minant sepsis.111,131 Although the incidence has significantly to be associated with longer ICU stays, increased ventilator
improved with the introduction of newer-generation devices, dependence, prolonged inotropic support, and a significant
e.

e.

e.

e.
infection remains a major source of morbidity and mortal- decline in 30-day, 3-month, and 1-year survival.37,138
re

fre

fre

ity during MCS, and it is the second most reported adverse

fre
sf

event in the INTERMACS database.37 With regard to device

ks

ks

ks
Stroke and Neurologic Dysfunction
k

comparison, the rates of driveline infection and sepsis in the

oo

oo

oo

oo

original HeartMate II BTT trial were 14 and 17%, respec- Despite improvements in device reliability and management
tively, compared with 12.1 and 11.4% in the subsequent strategies, stroke remains a significant source of morbidity and
eb

eb

eb

eb

HeartWare BTT trial.25,27 Creation of a preperitoneal pump mortality in patients who require LVAD support. Clinically
m

pocket during implantation of the HeartMate II likely pre-
disposes those patients to the development of a pump-pocket
infection, which can often result in life-threatening sepsis and
the need for pump exchange. As a result, our institutional
m
apparent strokes can result in significant functional disabilities—
a particular challenge for routine device maintenance—along
with impaired quality of life, significant rehabilitation needs,
and decreased postevent survival. Collectively, the impact of

preference is to avoid the need for a pump pocket by placing these outcomes highlights neurologic dysfunction as a par-
co

co

co

co

co

the pump in a subdiaphragmatic intraperitoneal position. ticularly feared complication of device therapy.
A review of INTERMACS data revealed percutaneous While strokes are typically classified under a singular
e.

e.

e.

e.

driveline infections in 9.8% of all patients supported with adverse event heading, the disparate pathogenesis of hem-
fre

fre

fre

fre

CF devices.132 The mean time to development of the infection orrhagic and ischemic strokes suggests the likelihood for
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1379 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1380
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
multiple causative factors. A detailed analysis comparing difference in bleeding events.29 However, the frequency of
fre

fre

fre

re
stroke etiology in device patients is not yet available, but these events remains unexpectedly high, as patients supported

sf
ks

ks

ks
the incidence of hemorrhagic and ischemic strokes appears by CF LVADs have a substantially higher risk for GI bleeding

k
to be roughly equal.26,140,141 Overall, the reported incidence than do patients who require anticoagulation for cardiovascu-
oo

oo

oo

oo
of stroke among device patients in the INTERMACS data- lar disease.147 Furthermore, bleeding events have been known
base is 11% at 1 year and 17% at 2 years postimplantation.37 to occur even without the use of anticoagulation.150
eb

eb

eb

eb
Risk factors for stroke during MCS are poorly understood, Some researchers have suggested that the propensity for
m

and many presumed influences—including anticoagulation,
atrial fibrillation, diabetes mellitus, and device type—have
not been found to be statistically significant.141 Pulsatility
may play a role in the development of embolic strokes. Pro-
m
m
GI bleeding may be due to an altered thrombotic physiology
resulting from interaction between the device and the blood-
stream. Multiple studies have shown that patients supported
by CF LVADs can develop an acquired von Willebrand syn-

om
longed closure of the aortic valve has been shown to result in drome as a result of the proteolytic cleavage of high-molecular-
co

co

co

co
eddies and stasis within the aortic cusps, predisposing some weight von Willebrand factor multimers.151-153 However, this

c
e.

e.

e.

e.
patients to the formation of thrombus that can be released explanation alone cannot explain the variability of bleeding
with later opening of the valve.142 Similarly, some patients events within the LVAD population, as bleeding has been
re

fre

re

re
have developed thrombus in areas of stasis within the carotid shown to be nearly universal in the setting of axial-flow device
sf

f
ks

ks

ks
bulb; again, such thrombus is susceptible to dislodgement support.154
k

after the recovery of pulsatility.31 As a result, routine main- An alternative explanation for the frequency of GI bleed-
oo

oo

oo

oo
tenance of intermittent pulsatility may provide prophylactic ing events may lie in the narrow pulse pressure created by CF
eb

eb

eb

eb
washout of these areas and reduce the potential for thrombo- support. The association between aortic stenosis and AVM
embolic events. In recent studies, BP has been identified as formation was first described by Heyde in the 1950s, and a
m

an additional key determinant in the pathogenesis of stroke.
The importance of adequate BP control in this population is
suggested by the fact that patients with lower BP at hospital
m
m
syndrome bearing his name was characterized over the sub-
sequent decades.155,156 This process is believed to result from
hypoperfusion and ischemia of intestinal mucosa, leading to

m
discharge and those receiving aggressive pharmacologic anti- friable neovasculature susceptible to rupture. Application of
hypertensive therapy have significantly lower rates of neuro- this concept to CF MCS physiology was first suggested by
co

co

co

co

co
logic events.141,143 Stroke adversely affects survival, entailing a Letsou and colleagues,157 who described a series of patients
e.

e.

e.

e.
25% 30-day and nearly 50% 1-year mortality risk.140,141 supported by the Jarvik 2000. Recently, a retrospective review
fre

fre

fre

fre
also documented a strong association between GI bleeding
Nonsurgical and Gastrointestinal and both reduced PI and aortic valve closure, suggesting
ks

ks

ks

ks
that a reduction in pulsatility may lead to increased rates
Bleeding of nonsurgical bleeding within the first 3 months of LVAD
oo

oo

oo

oo
As noted previously, bleeding of any type is the most fre- therapy.149
eb

eb

eb

eb
quent adverse event reported to INTERMACS, and late- ISHLT guidelines suggest that anticoagulation therapy
onset spontaneous bleeding is a significant source of be withdrawn in the setting of clinically significant bleeding.
m

morbidity for patients supported by CF LVADs.37 Research-
ers have consistently documented the frequency of nonsur-
gical (primarily GI) bleeding, particularly in older patients,
m
m
Resumption of anticoagulation is permitted after resolution
of the index bleeding event, but a reduction (or perma-
nent cessation) of anticoagulation is justified in the case of

m
and the reported incidence of this complication ranges from recurrent events. When possible, pump speeds are generally
co

co

co

co

co
18 to 30%.144-149 GI sources account for the vast majority reduced to allow increased native pulsatility. Endoscopic
(70%) of nonsurgical bleeding episodes. Epistaxis, intra- evaluation is also warranted for direct control of the bleed-
e.

e.

e.

e.
cranial, and genitourinary episodes are progressively less ing source, although identification of the responsible AVM
re

fre

fre

common.149 The location of GI bleeding is divided fairly is often difficult.36

fre
sf

evenly above and below the ligament of Treitz. Although

ks

ks

ks
percentages vary among studies, erosive gastritis and arte- Pump Thrombosis and
k
oo

oo

oo

oo

riovenous malformations (AVMs) appear to be the most

frequent causative factors.144,145,149 Whereas GI bleeding is
Device Malfunction
eb

eb

eb

eb

a significant source of morbidity—principally because of Pump thrombosis is an uncommon but potentially cata-
more frequent hospitalizations and increased transfusion- strophic complication of LVAD therapy. The HeartMate II
m

related alloantigen exposure—no associated impact on over- BTT trial investigators reported an actuarial freedom from
all mortality has been reported. device malfunction (defined as resulting in device replace-
The relative frequency of GI bleeding in this patient popu- ment or death) of 95% at 6 months and 93% at 1-year post-
m

lation has led to multiple proposed explanations. Anticoagu- implantation, one-third of those events being attributed to
co

co

co

co

co

lation during long-term MCS undoubtedly exposes patients primary device thrombosis.9 This degree of reliability and
to increased bleeding risks. A comparison between nonpul- durability represented a significant advance over earlier pulsa-
e.

e.

e.

e.

satile pumps and the pulsatile HeartMate XVE (which did tile systems, including a nearly 10-fold reduction in the need
fre

fre

fre

fre

not necessitate anticoagulation therapy) revealed a significant for device exchange.158 Between 2011 and 2012, concern was
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1380 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1381
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
raised regarding an apparent increase in the rate of pump Treatment strategies for suspected pump thrombosis are
fre

fre

fre

re
thrombosis with the HeartMate II. A three-center collab- limited. Noninvasive measures, such as intravenous antico-

sf
ks

ks

ks
orative study documented an increase in the incidence of agulation (heparin) and direct thrombin inhibitor therapy

k
pump thrombosis at 3 months from 2.2 to 8.4%; over the (bivalirudin or argatroban), are typically initiated early in the
oo

oo

oo

oo
same timespan, a federally mandated INTERMACS review management process with variable independent results.166,167
similarly described a decline in freedom from pump exchange Success with thrombolytic therapy (both systemic and intra-
eb

eb

eb

eb
or death at 6 months from 99 to 94%.37,159 After extensive ventricular) has also been reported, but this approach entails
m

investigation, the cause of this discrepancy remained unclear
and was likely multifactorial. More recently, however, insti-
tutional reports have suggested a reversal in the thrombo-
sis-related trend and the HeartMate II remains the most
m
m
a risk of intracranial bleeding.130,168,169 Definitive manage-
ment of pump thrombosis is surgical, either through pump
exchange, pump explantation, or urgent heart transplanta-
tion. Pump exchange may be performed through either a redo

om
frequently implanted device worldwide. Nevertheless, under- sternotomy or a subcostal approach, and the pump housing
co

co

co

co
standing the potential factors involved in device thrombosis of the HeartMate II can be exchanged directly without dis-

c
e.

e.

e.

e.
is critical, as this complication has been shown to negatively rupting the inlet or outflow conduit.170,171 Despite the rela-
impact overall survival.159 tive short-term safety of the exchange procedure, survival has
re

fre

re

re
A major challenge in the prevention of thrombosis events been shown to decrease with each successive pump exchange,
sf

f
ks

ks

ks
is the complexity of the pathophysiology involved. Indeed, which also increases the likelihood of later infection and neu-
k

the term “thrombosis” has traditionally been used in a broad rologic sequelae.37,160 The mortality rate 6 months after trans-
oo

oo

oo

oo
sense to describe not just confirmed thrombus within the plantation or initial device replacement is similar to that for
eb

eb

eb

eb
pump housing but also suspected thrombus due to various patients without a history of thrombosis.159
nonstandardized clinical factors. In 2013, INTERMACS
m

m
updated its definitions, establishing formal categories of TOTAL ARTIFICIAL HEART
suspected thrombosis (the criteria for which include hemoly-
sis, unexplained HF symptoms, and abnormal pump param- Whereas LVAD support has emerged as the standard of care
m

m
eters) and confirmed thrombus found on direct inspection or for patients with end-stage HF, data registry analysis sug-
radiographic imaging.130 Factors that contribute to thrombo- gests that a sizeable percentage (typically cited as 8 to 10%)
co

co

co

co

co
sis are likely far-ranging, and proposed mechanisms include of patients will require biventricular support.37,87,88,91 Heart
e.

e.

e.

e.
inadequate heat dissipation at low speeds, ineffective antico- transplantation has traditionally served as the gold-standard
fre

fre

fre

fre
agulation strategies, surgical and anatomical considerations therapy for this subset of patients, but the scarcity of donor
(such as angulation of the inlet or kinking of the outflow organs has led to a stagnant rate of transplantation, totaling
ks

ks

ks

ks
graft), patient-specific predisposition, and pump manufac- just over 4000 procedures each year worldwide.172 Mechani-
turing and design elements.160-162 cal total heart replacement is a logical and relatively inex-
oo

oo

oo

oo
Despite the variable etiology of device malfunction, recent haustible solution to this challenge. The use of implantable
eb

eb

eb

eb
studies have led to improved methods for accurately diagnos- BiVADs has been reported in the treatment of patients with
ing thrombotic events. Serum markers for hemolysis have severe biventricular HF.116,173,174 However, available devices
m

been shown to be associated with pump thrombus formation,
presumably due to the increased shear stress placed on red
cells traversing the area of flow impedance. Serum-free hemo-
m
m
are designed to function independently, and logistical chal-
lenges in their implantation and management as BiVAD sys-
tems renders them cumbersome and difficult to operate. As

m
globin (sfHg) has traditionally been used by INTERMACS a result, the development of a self-contained, fully implant-
co

co

co

co

co
for diagnostic purposes, with levels greater than 40 mg/dL able total heart replacement device is an important, though
indicative of significant hemolysis.130 Recent data have shown elusive, goal.
e.

e.

e.

e.
that lactate dehydrogenase (LDH) may be a more sensitive
re

fre

fre

marker than sfHg for detecting hemolysis and pump throm-

Early Development and Experience fre
sf

bosis, and subsequent INTERMACS reviews have shown a

ks

ks

ks
strong correlation between rising LDH levels and thrombotic The concept of replacing the human heart with a permanent
k
oo

oo

oo

oo

events.163 Whereas a threshold value of 600 IU/L has been or temporary mechanical device has been postulated for cen-
proposed for diagnostic purposes (representing 2.5 times turies and was first reported in the early nineteenth century
eb

eb

eb

eb

the upper limit of normal), this cutoff does not appear to be by French physiologist Julien Jean Cesar Le Gallois.175 Since
absolute, as anecdotal evidence suggests that many patients that time, efforts to duplicate nature’s design—including the
m

may exhibit higher LDH levels without tangible evidence of famed collaboration between Charles Lindbergh and Alexis
thrombosis. Other diagnostic tools, including speed-change Carrell in the 1920s—have periodically captured the public
echocardiography and three-dimensional computed tomog- imagination but failed to produce a reliable self-contained
m

raphy, also likely contribute to diagnostic accuracy.164 Given device.176

co

co

co

co

co

the variability in presentation, a recent multicenter collabora- The first tangible step toward the development of a viable
tion led to the development of a useful diagnostic and clini- TAH came in 1957, when Tetsuzo Akutsu and Willem Kolff
e.

e.

e.

e.

cal management algorithm for patients with suspected pump replaced the heart of a dog with an experimental pulsatile
fre

fre

fre

fre

thrombosis.165 device.177 The dog survived for several hours. Although the
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1381 24/04/17 2:42 pm

 e

e
m

m
m

om

m
1382
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
as the SynCardia TAH. The device consists of two separate,
fre

fre

fre

re
pneumatically driven pumps bound only by a Velcro patch

sf
ks

ks

ks
along their adjoining walls (Fig. 62-13). Pump operation

k
depends on the pneumatic displacement of a polyurethane
oo

oo

oo

oo
diaphragm powered by an external pneumatic driver. Dacron
cuffs permit atrial anastomosis after ventriculectomy, and the
eb

eb

eb

eb
outflow conduits are similarly composed of woven Dacron
m

m
grafts. Both components are attached to the pump hous-
ing by specialized connecting brackets. Each pump contains
large-diameter SynHall valves (derived from the original
Medtronic-Hall design) that regulate both inflow and out-
m

om
flow (27 and 25 mm, respectively) and provide minimal resis-
co

co

co

co
tance to flow. The pneumatic drivelines are tunneled beneath

c
e.

e.

e.

e.
the costal margin to reach the external driver, where pump
rate, drive pressure, and systolic duration are controlled. The
re

fre

re

re
device produces a stroke volume of 70 mL and a cardiac out-
sf

f
ks

ks

ks
put as high as 15 L/min.178,179
k

The earliest clinical iteration of this device, labeled

oo

oo

oo

oo
the Jarvik-7, was first implanted by William DeVries in
eb

eb

eb

eb
1982, supporting a 61-year-old man with congestive HF for
112 days until he died of multiorgan failure.180 Despite ini-
m

m
tially mixed results, clinical trials began in 1985. Over the
years, the device was rebranded the Symbion, then the Car-
dioWest TAH, and finally the SynCardia TAH.181-183
m

m
co

co

co

co

co
FIGURE 62-12 The Liotta total artificial heart, the first to be
implanted in a human. (Courtesy Texas Heart Institute.)
e.

e.

e.

e.
fre

fre

fre

fre
ks

ks

ks

ks
pump was never tested in humans, its brief success high-
lighted the potential for total heart replacement. Capitalizing
oo

oo

oo

oo
on a growing interest in MCS, the National Heart Institute
eb

eb

eb

eb
(now termed the NHLBI) established the Artificial Heart
Program in 1964, helping to fund and support research and
m

m
development efforts throughout the field.
The first implantation of a TAH into a human was per-
formed by Denton Cooley on April 4, 1969.16 The device was
m

m
placed in a 47-year-old man who could not be weaned from
co

co

co

co

co
CPB after undergoing an LV aneurysmectomy. Designed by
Domingo Liotta, the pump was a pneumatically powered,
e.

e.

e.

e.
double-chambered device with Dacron-lined conduits and
re

fre

fre

Wada-Cutter hingeless valves (Fig. 62-12). It was intended

fre
sf

to support the patient only until a suitable donor heart was

ks

ks

ks
found. Indeed, the TAH performed adequately for 64 hours,
k
oo

oo

oo

oo

after which the patient underwent a transplant. Despite its

success, Liotta’s pump was never again used clinically, but the
eb

eb

eb

eb

experience showed that a TAH could be used safely and effec-

tively for BTT.
m

Contemporary Options for Total

Heart Replacement
m

m
co

co

co

co

co

SYNCARDIA TAH
FIGURE 62-13  The Syncardia Total Artificial Heart (formerly the
In the late 1970s, under the leadership of Willem Kolff,
e.

e.

e.

e.

Jarvik-7, Symbion, and CardioWest TAH) utilizes two pneumati-

researchers at the University of Utah began developing a cally driven pumps for biventricular support. (Courtesy: SynCardia
fre

fre

fre

fre

TAH design that eventually evolved into what is now known Systems, Inc.)
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1382 24/04/17 2:43 pm

 e

e
m

m
m

om

m
1383
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
Like many other TAHs, the SynCardia is somewhat
fre

fre

fre

re
restricted by its large size. Generally, candidates are screened

sf
ks

ks

ks
for a sufficient body surface area (>1.8 m2) and for a chest

k
anteroposterior diameter of >10 cm on computed tomog-
oo

oo

oo

oo
raphy.184 Studies have shown that the body surface area is
associated with an increased risk of bleeding, infection, and
eb

eb

eb

eb
overall mortality.185 A smaller (50 mL) version of the device
m

m
is currently under development. Another historic limitation
of the SynCardia TAH was the inconvenience of its exter-
nal components. The original console was large and relatively
immobile, effectively eliminating the possibility of prolonged
m

om
ambulation or outpatient management. The 2010 release of
co

co

co

co
the 13-pound Freedom Driver has done much to alleviate

c
e.

e.

e.

e.
these problems, greatly improving device portability.186
The SynCardia TAH is currently approved by the FDA
re

fre

re

re
and CMS for BTT in the treatment of biventricular failure
sf

f
ks

ks

ks
and has been successfully used in more than 1100 patients.187
k

In 2012, the device also received approval for DT use as a

oo

oo

oo

oo
humanitarian exemption. Although the majority of implants
eb

eb

eb

eb
have been performed for short-term support, nearly 50 patients
have been supported for more than 1 year with the SynCardia,
m

m
and the median duration of support is approximately 554 days.
The survival-to-transplantation rate has been 72% in the FIGURE 62-14 The AbioCor Implantable Replacement Heart.
(Reproduced with permission from ABIOMED, Inc.)
long-term population. Complications have included systemic
m

m
infections (35%), driveline infections (27%), thromboembo-
lism (19%), and device malfunction (10%).185 Survival rates
co

co

co

co

co
each heartbeat) drains directly into the left atrium, thereby
are most impacted during the early postoperative period, and avoiding the conventional venous system and resulting in
e.

e.

e.

e.
the vast majority of deaths occur within the first month.188,189 an LV output that exceeds right-sided output by 2 to 5%.
fre

fre

fre

fre
Despite these limitations, the SynCardia has the potential for Whereas this imbalance seems negligible in the short term,
long-term reliable total heart replacement. it accounts for as much as 250 L of blood flow each day
ks

ks

ks

ks
and must be carefully considered when designing a pulsatile
ABIOCOR TAH
oo

oo

oo

oo
device responsible for producing right- and left-sided flows.
The AbioCor Implantable Replacement Heart is a self- To accommodate this discrepancy, the design of the AbioCor
eb

eb

eb

eb
contained, single-body, electrohydraulic TAH developed involves a right-to-left flow-balancing mechanism is designed
by ABIOMED, Inc. (Danvers, MA) and the Texas Heart to account for interventricular physiologic flow differences by
m

Institute.190 The device’s principal thoracic component, which
is positioned orthotopically after biventriculectomy, consists
of two polyurethane hydraulic pumping chambers—each
m
m
shunting hydraulic fluid away from the right-sided pump.191
The design functioned appropriately in both animal studies
and clinical use.

m
flanked by trileaflet valves—that serve as artificial ventricles With support from NHLBI, preclinical trials and exten-
co

co

co

co

co
and function as opposing compliance chambers (Fig. 62-14). sive animal testing began at the Texas Heart Institute in 1993.
A continuous motor allows shuttling of hydraulic fluid to In 2001, the FDA approved a phase 1 feasibility study for the
e.

e.

e.

e.
alternately pressurize the right and left ventricles. Active dia- device, which was first implanted in a 59-year-old man in
re

fre

fre

stolic filling of the contralateral side is enabled by the negative

fre
Louisville, KY who survived for 151 days with the device.192
sf

pressure established by the hydraulic system. Unique to the Criteria for inclusion in the study were restrictive, and
ks

ks

ks
k

design of the AbioCor is the use of radiofrequency data trans- included the diagnosis of irreversible biventricular HF (>70%
oo

oo

oo

oo

mission and a transcutaneous energy transfer system designed predicted 30-day mortality), with documented ineligibility
to create a fully implantable system that can communicate for transplant. A total of 14 patients were eventually enrolled
eb

eb

eb

eb

with external hardware without penetrating the skin. An elec- in the trial, the longest period of survival reaching 512 days.
m

tronic and data component is implanted within the abdo-
men. The pump produces a stroke volume of 65 mL and a
cardiac output as high as 12 L/min.
The AbioCor is also the only device designed to address
m
Overall survival metrics were promising, with a 71% 30-day
survival rate (compared with 13% in the medical arm) and
40% survival at 2 months.193
A major disadvantage of this pump is its relatively large

the natural dissymmetry in native ventricular output. Man- size, which limited candidacy to patients with a thoracic cage
co

co

co

co

co

agement of a totally implantable pulsatile TAH system deep enough to accommodate the device. Preoperative siz-
requires an understanding of the difference in stroke volumes ing was generally performed by using three-dimensional
e.

e.

e.

e.

produced by the right and left sides of the heart. The arterial computed tomography reconstructions, focusing primarily
fre

fre

fre

fre

blood supply to the bronchial tree (totaling 2 to 3 cc with on anteroposterior thoracic diameter and the position of the
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1383 24/04/17 2:43 pm

 e

e
m

m
m

om

m
1384
co

co

co

co
Part X  Transplant and Mechanical Circulatory Support

c
e.

e.

e.

e.
pulmonary bifurcation.193 Sufficient distance between the clinical durability far surpassing that of pulsatile systems.194
fre

fre

fre

re
bifurcation and diaphragm was required to prevent compres- Hemodynamically, CF pumps also appear well suited for

sf
ks

ks

ks
sion of the left pulmonary veins and lower-lobe bronchus. TAH use, because (much as in the native heart) the inherent

k
Due to the level of anatomical complexity, a high degree of inflow sensitivity of these devices provides a degree of auto-
oo

oo

oo

oo
confidence with regard to sizing and fit was required before maticity in responding to imbalances in right- and left-sided
implantation was attempted. Although the AbioCor received flows. Thus, development of a CF TAH may lead to realiza-
eb

eb

eb

eb
FDA approval for use under a humanitarian device exemp- tion of a long-awaited goal.
m

tion in 2006, no implants have been performed since the
completion of its initial feasibility study, principally due
to the prohibitive economic barriers associated with its
implementation.
m
m
Over the past decade at the Texas Heart Institute, exten-
sive research and development efforts have been dedicated to
the concept of CF total heart replacement, involving both
simulated models and live animal studies. To date, more

om
than 70 calves have been implanted with paired CF VADs
co

co

co

co
to provide completely nonpulsatile circulatory support. The

c
Complexities of Mechanical Total
e.

e.

e.

e.
resulting CF TAH model has revealed considerable long-term
Heart Replacement effectiveness, with more than 30 animals surviving for longer
re

fre

re

re
Initial efforts to develop a viable TAH began in earnest as than 7 days, 8 surviving for 30 days, and 2 calves surviving
sf

f
ks

ks

ks
part of a series of bold scientific initiatives proposed by for more than 90 days with nonpulsatile blood flow after CF
k

the Kennedy administration during the idealistic 1960s. total heart replacement.195,196 Tolerance of this altered physi-
oo

oo

oo

oo
Although similarly daring goals, such as manned spaceflight ologic state is further evidenced by stable hemodynamic pro-
eb

eb

eb

eb
to the moon, were accomplished within the decade, con- files both at rest and during treadmill exercise studies, as well
struction of a reliable self-contained device for replacing the as preserved hematologic and biochemical markers for end-
m

human heart proved to be a prohibitive technical challenge.
A number of complexities, both physiologic and technologic,
are to blame for the ongoing delay in realizing this objective.
m
m
organ function. This extensive experience with large animal
models has stimulated a number of collaborative research
efforts in search of a viable CF TAH design.

m
From an engineering standpoint, designing a TAH presents To this end, a novel CF TAH is currently being developed
that offers tremendous potential to become a completely
co

co

co

co

co
a number of challenges, the chief of these being durability.
The human heart beats more than 100,000 times each day. As implantable long-term cardiac replacement device. The
e.

e.

e.

e.
a result, thin diaphragms and hydraulic actuators devised for BiVACOR TAH, developed by BiVACOR USA in Hous-
fre

fre

fre

fre
pulsatile systems are subject to considerable mechanical wear; ton, Texas, working closely with the Texas Heart Institute,
though capable of providing support beyond 1 year, they are is a small, compact device that utilizes a single magnetically
ks

ks

ks

ks
unlikely to withstand the physical demands of long-term levitated rotor to provide pulmonary and systemic blood
flow without the use of bearings or valves197 (Fig. 62-15).
oo

oo

oo

oo
operation. Size is also a hurdle in TAH design, and the hous-
ing of biventricular volume-displacement chambers capable Impellers are mounted on each side of the rotor (which
eb

eb

eb

eb
of producing sufficient cardiac output requires a substantial separates the right and left chambers), and its frictionless
device profile that can be difficult to accommodate, even in suspension permits accommodation for physiologic flow
m

m
larger patients. Moreover, portability and mobility are cru-
cial considerations for long-term support. Pneumatic systems
require drivelines for pressure venting. Even though recent
m

m
efforts have reduced the size of associated equipment, com-
co

co

co

co

co
pressors and power consoles can be cumbersome and limit
functionality. Minimization of extracorporeal drivers, drive-
e.

e.

e.

e.
lines, and tethering equipment would speed postoperative
re

fre

fre

recovery, minimize infectious risks, and improve overall qual-

fre
sf

ity of life. Remedying these concerns, however, has proved to

ks

ks

ks
k

be an immense scientific and technological undertaking.

oo

oo

oo

oo

Future Directions: the CF TAH

eb

eb

eb

eb
m

A quarter century of clinical experience with CF pumps has

revealed the human capacity for physiologic tolerance of
nonpulsatile flow and has confirmed the significant design
advantages of CF devices over pulsatile TAHs. Continuous-
m

flow devices are not only significantly smaller than pulsatile

co

co

co

co

co

pumps but also are more anatomically appropriate thus being FIGURE 62-15  The BiVACOR total artificial heart employs a sin-
simpler to implant and operate. Furthermore, with few (or
e.

e.

e.

e.

gle moving part—a central magnetically levitated rotor—to provide

no) points of mechanical contact, CF designs essentially elim- compete cardiopulmonary support. (Reproduced, with permission,
fre

fre

fre

fre

inate the potential for mechanical wear, resulting in a level of from BiVACOR, Inc.)
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1384 24/04/17 2:43 pm

 e

e
m

m
m

om

m
1385
co

co

co

co
Chapter 62  Long-term Mechanical Circulatory Support and the Total Artificial Heart

c
e.

e.

e.

e.
imbalances by axial displacement the rotor.198 Also unique 7. Rose EA, Gelijns AC, Moskowitz AJ, et al: Long-Term use of a left

fre

fre

fre

re
ventricular assist device for end-stage heart failure. N Engl J Med 2001;
to the BiVACOR is the use of dynamic speed profiles to

sf
345:1435-1443.
ks

ks

ks
allow establishment of pulsatility in device operation.199 8. Kirklin JK, Naftel DC, Kormos RL, et al: Fifth INTERMACS annual

k
Feasibility studies have been successfully performed in large report: risk factor analysis from more than 6,000 mechanical circulatory
oo

oo

oo

oo
animals, with a goal of human trials in the near future. A support patients. J Heart Lung Transplant 2013; 32:141-156.
9. Pagani FD, Miller LW, Russell SD, et al: HeartMate III: Extended
similar hydrodynamically suspended design is also in devel-
eb

eb

eb

eb
mechanical circulatory support with a continuous-flow rotary left ven-
opment at the Cleveland Clinic, which also has begun tricular assist device. J Am Coll Cardiol 2009; 54:312-321.
m

m
animal studies.200 Thus, innovative engineering efforts and 10. Gibbon JH, Jr: Application of a mechanical heart and lung apparatus to
successful utilization of CF technology may soon offer the cardiac surgery. Minn Med 1954; 37:171-185; passim.
11. Lillehei CW, Cohen M, Warden HE, Varco RL: The direct-vision
opportunity for permanent mechanical replacement of the intracardiac correction of congenital anomalies by controlled cross
failing human heart.
m

om
circulation; results in thirty-two patients with ventricular septal
defects, tetralogy of Fallot, and atrioventricularis communis defects.
co

co

co

co
Surgery 1955; 38:11-29.

c
KEY POINTS 12. Kirklin JW, Dushane JW, Patrick RT, et al: Intracardiac surgery with the
e.

e.

e.

e.
aid of a mechanical pump-oxygenator system (Gibbon type): report of
re

fre

re

re
1. Continuous-flow technology has resulted in smaller, more eight cases. Proc Staff Meet Mayo Clin 1955; 30:201-206.
13. Cooley DA, Belmonte BA, Zeis LB, Schnur S: Surgical repair of rup-
durable MCS devices. Although the full effects of the
sf

f
tured interventricular septum following acute myocardial infarction.
ks

ks

ks
resulting physiologic changes remain unknown, nonpul- Surgery 1957; 41:930-937.
k
oo

oo

oo

oo
satile blood flow is well tolerated. 14. DeBakey ME: Left ventricular bypass pump for cardiac assistance. Clinical
experience. Am J Cardiol 1971; 27:3-11.
2. Thoughtful patient selection and preoperative optimiza- 15. Shumacker HB, Jr: A surgeon to remember: notes about Vladimir
eb

eb

eb

eb
tion of the patient’s condition are crucial to ensure good Demikhov. Ann Thorac Surg 1994; 58:1196-1198.
outcomes after LVAD implantation. 16. Cooley DA, Liotta D, Hallman GL, Bloodwell RD, Leachman RD,
m

3. Surgical considerations at the time of LVAD implantation
should include careful alignment of the pump inlet and
outflow conduits to minimize the risk of physical impedi-
m
m
Milam JD: Orthotopic cardiac prosthesis for two-staged cardiac replace-
ment. Am J Cardiol 1969; 24:723-730.
17. Norman JC, Duncan JM, Frazier OH, et al: Intracorporeal (abdominal)
left ventricular assist devices or partial artificial hearts: a five-year clinical

m
ments to proper device function. experience. Arch Surg 1981; 116:1441-1445.
co

co

co

co

co
18. Frazier OH, Macris MP, Myers TJ, et al: Improved survival after extended
4. Perioperative preservation and optimization of RV func- bridge to cardiac transplantation. Ann Thorac Surg 1994; 57:1416-1422;
tion is an essential component of MCS management.
e.

e.

e.

e.
discussion 1412-1421.
5. Due to the scarcity of available donor organs, development 19. Frazier OH, Rose EA, McCarthy P, et al: Improved mortality and reha-
fre

fre

fre

of a CF TAH offers the best means of providing definitive
long-term treatment for patients who require total heart
ks
fre
bilitation of transplant candidates treated with a long-term implantable
left ventricular assist system. Ann Surg 1995; 222:327-336; discussion

ks

ks

ks
328-336.
replacement. 20. Frazier OH: First use of an untethered, vented electric left ventricular
oo

oo

oo

oo
assist device for long-term support. Circulation 1994; 89:2908-2914.
21. Frazier OH, Nakatani T, Duncan JM, Parnis SM, Fuqua JM: Clinical
eb

eb

eb

eb
REFERENCES experience with the Hemopump. ASAIO Trans 1989; 35:604-606.
22. Macris MP, Parnis SM, Frazier OH, Fuqua JM, Jr, Jarvik RK: Develop-
m

1. Lloyd-Jones DM, Larson MG, Leip EP, et al: Lifetime risk for developing
congestive heart failure: the Framingham Heart Study. Circulation 2002;
106:3068-3072.
2. Heidenreich PA, Trogdon JG, Khavjou OA, et al: American Heart
Association Advocacy Coordinating Committee, Stroke Council,
m
m
ment of an implantable ventricular assist system. Ann Thorac Surg 1997;
63:367-370.
23. Jarvik RK: System considerations favoring rotary artificial hearts with
blood-immersed bearings. Artif Organs 1995; 19:565-570.
24. Tuzun E, Roberts K, Cohn WE, et al: In vivo evaluation of the HeartWare

m
Council on Cardiovascular Radiology and Intervention, Council on centrifugal ventricular assist device. Tex Heart Inst J 2007; 34:406-411.
co

co

co

co

co
Clinical Cardiology, Council on Epidemiology and Prevention, Council 25. Miller LW, Pagani FD, Russell SD, et al: Use of a Continuous-Flow
on Arteriosclerosis, Thrombosis, Vascular Biology, Council on Cardio- Device in Patients Awaiting Heart Transplantation. N Engl J Med 2007;
e.

e.

e.

e.
pulmonary, Critical Care, Perioperative and Resuscitation, Council 357:885-896.
on Cardiovascular Nursing, Council on the Kidney in Cardiovascular 26. Slaughter MS, Rogers JG, Milano CA, et al: HeartMate III: Advanced
re

fre

fre

Disease, Council on Cardiovascular Surgery and Anesthesia, Interdisci- heart failure treated with continuous-flow left ventricular assist device.
fre
sf

plinary Council on Quality of Care and Outcomes Research: Forecast- N Engl J Med 2009; 361:2241-2251.
ks

ks

ks
ing the future of cardiovascular disease in the United States: a policy 27. Aaronson KD, Slaughter MS, Miller LW, et al: HeartWare Ventricular
k

statement from the American Heart Association. Circulation 2011; Assist Device Bridge to Transplant ATI: Use of an intrapericardial,
oo

oo

oo

oo

123:933-944. continuous-flow, centrifugal pump in patients awaiting heart transplan-

3. National Health and Nutrition Examination Survey (NHANES): tation. Circulation 2012; 125:3191-3200.
eb

eb

eb

eb

Prevalence of Common Cardiovascular and Lung Dieseases, U.S., 28. Pagani FD, Long JW, Dembitsky WP, Joyce LD, Miller LW: Improved
2007-2011. Available at: http://www.nhlbi.nih.gov/about/documents/ mechanical reliability of the HeartMate XVE left ventricular assist
m

factbook/2012/chapter4#4_5. Accessed March 17, 2015.
4. Yancy CW, Jessup M, Bozkurt B, et al: 2013 ACCF/AHA guideline for
the management of heart failure: executive summary: a report of the
American College of Cardiology Foundation/American Heart Associa-
m
system. Ann Thorac Surg 2006; 82:1413-1418.
29. Crow S, John R, Boyle A, et al: Gastrointestinal bleeding rates in recipi-
ents of nonpulsatile and pulsatile left ventricular assist devices. J Thorac
Cardiovasc Surg 2009; 137:208-215.

tion Task Force on practice guidelines. Circulation 2013; 128:1810-1852. 30. Segura AM, Gregoric I, Radovancevic R, Demirozu ZT, Buja LM,
5. Deja MA, Grayburn PA, Sun B, et al: Influence of mitral regurgitation Frazier OH: Morphologic changes in the aortic wall media after support
co

co

co

co

co

repair on survival in the surgical treatment for ischemic heart failure trial. with a continuous-flow left ventricular assist device. J Heart Lung Trans-
Circulation 2012; 125:2639-2648. plant 2013; 32:1096-1100.
e.

e.

e.

e.

6. Velazquez EJ, Lee KL, Deja MA, et al: Coronary-artery bypass sur- 31. Reul JT, Reul GJ, Frazier OH: Carotid-bulb thrombus and continuous-
fre

fre

fre

fre

gery in patients with left ventricular dysfunction. N Engl J Med 2011; flow left ventricular assist devices: a novel observation. J Heart Lung
364:1607-1616. Transplant 2014; 33:107-109.
s

ks

ks

ks
ok

oo

oo

oo
o
eb

eb

eb

eb
m

Cohn_ch62_p1361-1390.indd 1385 24/04/17 2:43 pm