Download Clinical Performance with the Levitronix Centrimag

Clinical Performance with the Levitronix Centrimag Short-term
Ventricular Assist Device
F. De Robertis, MD,a E. J. Birks, MRCP, PhD,b P. Rogers, RGN,b G. Dreyfus, FRCS,a J. R. Pepper, FRCS,a
and A. Khaghani, FRCSa
Background: The Levitronix ventricular assist device (VAD) is a centrifugal pump designed for extracorporeal
support and that operates without mechanical bearings or seals. The rotor is magnetically levitated
so that rotation is achieved without friction or wear, which seems to minimize blood trauma and
mechanical failure. The aim of this study is to report our early results with the Levitronix Centrimag
device.
Methods:
Between June 2003 and April 2005, 18 patients (pts) were supported using the Levitronix at our
institution. Fourteen were male. Mean age was 40.3 ⫾ 18.3 (range 8 to 64) years. Indications for
support at implantation were: post-cardiotomy cardiogenic shock in 12 cases (Group A), and bridge
to decision regarding long-term ventricular support in 6 cases (Group B).
Results:
Mean support time was 14.2 ⫾ 15.2 days for all patients (range 1 to 64 days). Operative (30-day)
mortality was 50% (9 pts). Six pts were in Group A and 3 pts were in Group B. Overall, 6 pts (33%)
were discharged home and are presently alive and well (mean follow-up 13 months, range 5 to 17
months). Bleeding requiring re-operation occurred in 8 cases (44%), cerebral thromboembolism in
1 and pulmonary embolism in 1. There were no device failures.
Conclusions: The Levitronix functioned well and proved to be useful in patients with extremely poor prognosis
previously considered non-suitable for a long-term assist device. The device was technically easy to
implant and manage. There was no device dysfunction and complications were acceptable or
consistent with other devices. Survival to explant or a definitive procedure (VAD or transplantation)
was encouraging. J Heart Lung Transplant 2006;25:181– 6. Copyright © 2006 by the International
Society for Heart and Lung Transplantation.
Short-term ventricular assist devices have been used
mainly for post-cardiotomy cardiogenic shock where
the operative mortality is around 75%.1–3 Miniaturization of existing components has resulted in the introduction of new devices with smaller cannulae. In
addition, magnetically suspended pumps with no bearings seem to minimize blood trauma4,5 and mechanical
failure. It is anticipated that these new devices will
reduce morbidity and mortality. We therefore investigated the Levitronix Centrimag short-term ventricular
assist device at our institution and report our initial
data.
From the Departments of aCardiothoracic Surgery and bCardiology,
Royal Brompton and Harefield NHS Trust, Harefield, Middlesex, UK.
Submitted April 25, 2005; revised August 8, 2005; accepted August
16, 2005.
Reprint requests: Asghar Khaghani, MD, Department of Cardiothoracic Surgery, Transplantation and Mechanical Circulatory Support,
Royal Brompton and Harefield NHS Trust, Harefield, Middlesex UB9
6JH, UK. Tel: ⫹44-1895-828799. Fax: ⫹44-1895-828932. E-mail:
[email protected]
Copyright © 2006 by the International Society for Heart and Lung
Transplantation. 1053-2498/06/$–see front matter. doi:10.1016/
j.healun.2005.08.019
METHODS
Clinical Summary
Between June 2003 and April 2005, 18 patients (pts)
underwent Levitronix Centrimag short-term ventricular
assist device implantation. Fourteen were male. The
mean age was 40.3 ⫾ 18.3 (range 8 to 64) years.
The indications for support at the time of implantation were: post-cardiotomy cardiogenic shock in 12
cases (Group A), and bridge to decision regarding
long-term ventricular support in 6 cases (Group B).
Patients in Group A were undergoing the following
operations: congenital reconstructive surgery in 3 (2 re-do
procedures), of which 1 involved a tissue aortic valve
replacement and 1 an autograft AVR; aortic valve replacement in 2, of which 1 was a mechanical AVR and
1 was an autograft AVR; isolated coronary artery bypass
grafting (CABG) in 1; CABG combined with mitral valve
repair in 1; long-term LVAD system implantation in 3 (1
affected by idiopathic dilated cardiomyopathy [DCM],
and 2 ischemic cardiomyopathy); and orthotopic heart
transplantation (primary graft failure) in 2.
The etiology of cardiogenic shock in Group B patients was: idiopathic DCM (2 pts); ischemic cardiomyopathy (2 pts, 1 re-do); post-partum DCM (1 pt); and
181
182
De Robertis et al.
The Journal of Heart and Lung Transplantation
February 2006
acute myocarditis (1 pt). They were all in end-stage
cardiac failure but were considered not suitable for a
long-term device for one or more of the following
reasons: 3 pts were septic; 3 pts were moribund and
needed salvage intervention; and in 2 pts central neurologic damage was suspected. In this group the preoperative blood results showed: white blood cell
(WBC) count 16 ⫾ 7.8 (range 6.5 to 28.3) ⫻ 109/liter;
creatinine 1.73 ⫾ 0.76 (range 0.88 to 2.76) mg/dl;
bilirubin 2.62 ⫾ 0.61 (range 2.1 to 3.62) mg/dl; serum
alanine aminotransferase (ALT) 809.2 ⫾ 956.6 (range 22
to 2,186) U/liter; and albumin 23.2 ⫾ 2.94 (range 19 to
26) U/liter.
In both groups all patients were on inotropic support
and 9 patients had an intra-aortic balloon pump (IABP)
inserted pre-operatively. The Levitronix implantation
was a salvage procedure for all patients in Group A and
for 3 patients in Group B.
The serum lactate levels were measured at baseline
and then every 6 hours for 48 hours in all patients.
mechanical gaps in the pump are ⬎0.6 mm to allow
shear forces to be low. This device can produce flows
of up to 10 liters/min under normal physiologic conditions, with a priming volume of 31 ml.
The cannulae are supplied by the manufacturer and
are not heparin-coated; however, any other cannula can
be connected to the circuit. The inflow is a 32-French
(F) angled cannula that is wire-reinforced and malleable, whereas the outflow is a 22F straight cannula.
Ventricular cannulae are not available at present but the
manufacturer is planning to supply them in the future.
The device is CE-mark-approved for short-term support for up to 14 days and is commercially available
throughout Europe. In the USA, it is for investigational
use only. A clinical trial to evaluate the safety and
effectiveness of the technology for use in the treatment
of post-cardiotomy cardiogenic shock is in progress
under an investigational device exemption.
Our study was approved by the ethics committee of
the Royal Brompton and Harefield NHS Trust.
The Device
The Levitronix Centrimag short-term VAD is an extracorporeal system6 composed of a single-use centrifugal
blood pump, a motor, a console, a flow probe and a
circuit. The device is based on the “bearingless motor”
technology, which combines the drive, the magnetic
bearing and the rotor function into a single unit. The
motor generates the magnetic bearing force that levitates the rotor into the pump housing while also
generating the torque necessary to produce the unidirectional flow (Figure 1). The characteristic of the
magnetic levitation is the absence of bearings and seals,
thus resulting in minimal friction or heat generation in
the blood path. The rotor surface is uniformly washed,
which minimizes the areas of blood stagnation and
turbulence in the pump. To reduce hemolysis, the
The Surgical Procedure
All procedures except one were performed through a
median sternotomy. The exception was a patient who
had undergone a previous CABG and subsequently
developed ischemic cardiomyopathy. His condition deteriorated while on the waiting list for transplantation
and he needed left ventricular support as a bridge. A left
thoracotomy approach was chosen to try to avoid
another median sternotomy before transplant.
In some of the Group B patients the operation was
performed without the use of cardiopulmonary bypass.
In the LVAD configuration the inflow cannula was
inserted in the left atrium at the level of the junction
between the right superior pulmonary vein and the left
atrium. The outflow cannula was placed directly into
the ascending aorta in all but 2 cases. In these 2
patients, a Thoratec cannula (Thoratec Corp., Pleasanton, CA), which has a woven polyester graft section,
was used instead due the surgeon’s preference and it
was sewn to the aorta with a continuous 4-0 polypropylene running suture. If a right-side support system
was needed, the inflow cannula was placed in the right
atrium and the outflow cannula was inserted in the
main pulmonary artery. Purse strings of 2-0 polyester
were placed at the cannulation sites for hemostasis and
the sutures were tied around the cannulae. The right
and left outflow cannulae were placed anteriorly at the
front of the heart, whereas the inflow cannulae were
lying in the pericardium on the right side of the right
atrium (Figure 2). All cannulae were brought outside
the patients’ bodies through separate stab wounds and
secured to the skin using 1-0 nylon sutures. After very
careful de-airing, the cannulae were connected to the
circuit.
pump housing
I n l et
impeller
o u tle t
rotor
w i nd i n g
stator
Figure 1. The bearingless motor combines drive, magnetic bearing
and rotor function into a single unit. The motor generates the magnetic
bearing force that is levitating the rotor into the pump housing and it
also generates the torque necessary to produce the uni-directional
flow. Copyright IHC 2004.
The Journal of Heart and Lung Transplantation
Volume 25, Number 2
De Robertis et al.
183
Bi-ventricular support was required in 6 cases despite
the use of nitric oxide in all patients: 3 of these patients
were in Group A and 3 in Group B.
Figure 2. Cannulation sites for the Levitronix in the bi-VAD configuration (see text). Copyright IHC 2004.
Anti-coagulation
We did not administer any anti-coagulation for the first
6 to 12 hours to normalize the clotting profile. Once the
drainage was ⬍50 ml/hour, a heparin infusion was
started without a bolus to maintain an activated partial
thromboplastin time (aPTT) ratio of 1.5:2.5, and it was
continued for the duration of support. No anti-platelet
agents were given.
Outcomes
Table 1 summarizes the outcome data. The overall
30-day mortality was 50% (9 pts) and the cause of death
was multi-organ failure in all cases. In Group A, 6
patients (50%) died: 4 died while on Levitronix device
support; 1 was re-transplanted and died 7 days later;
and 1 was weaned off of the device but died the same
day. In Group B, 3 patients (50%) died while on
Levitronix support.
Of the 6 surviving patients at 30 days in Group A, 1
died of multi-organ failure secondary to hemorrhagic
diathesis at 64 days after the initial operation. This
patient initially presented with acute mitral and tricuspid regurgitation secondary to acute myocardial infarction. He underwent a combined procedure (CABG,
mitral and tricuspid valve repair) that was complicated
by RV failure that required a Levitronix RVAD. He was
then assisted for 10 days, but because of deteriorating
bi-ventricular function he subsequently received a longterm pulsatile bi-VAD system. The other 5 patients
(42%) in Group A were all discharged home. In particular, 1 patient who received a Levitronix RVAD, along
with a long-term LVAD, had the right-side system explanted and has since undergone transplantation. One
patient, supported with a bi-VAD system after primary
graft failure, has been re-transplanted. Three patients
who were previously supported with a Levitronix
LVAD had the system explanted.
Table 1. Outcomes Grouped by Indication
Group A
Statistical Analysis
Categoric factors are expressed as the number and
percentage of patients. Continuous parameters are
given as mean ⫾ standard deviation. Paired data were
compared by a 1-tailed Student’s t-test. p ⬍ 0.01 was
considered statistically significant. Outcomes are presented as operative mortality within 30 days and survival to discharge from the hospital.
RESULTS
Application
Left-side support was used in 9 cases, including 6
patients in Group A and 3 in Group B. Right-side
support was used in 3 cases, all in Group A: 1 patient
had isolated right ventricular (RV) failure after a combined CABG, mitral and tricuspid valve repair, and 2
patients had bi-ventricular failure revealed once the left
side was mechanically assisted with a long-term device.
Group B
Application
LVAD
LVAD
LVAD
LVAD
LVAD
LVAD
RVAD
RVAD
RVAD
bi-VAD
bi-VAD
bi-VAD
LVAD
LVAD
LVAD
bi-VAD
bi-VAD
bi-VAD
DOS
27
3
16
7
2
5
11
10
12
2
1
4
9
20
12
19
64
32
Outcome
E
W⫹A
W⫹E
W⫹A
E
W⫹A
W⫹T⫹A
B⫹E
E
T⫹A
T⫹E
E
B⫹T⫹A
E
E
E
E
E
SAE
0
A
0
A
0
A
A
54
0
A
7
0
A
0
0
0
0
0
DOS, duration of support expressed in days; SAE, survival after explantation,
expressed in days; E, expired; W, weaned; A, alive; T, transplanted; B, bridged.
See text for further details.
184
De Robertis et al.
Of the 3 surviving patients at 30 days in Group B, 2
died of multi-organ failure while assisted with a Levitronix bi-VAD system, at 32 and 64 days after the initial
operation, respectively. Both suffered central neurologic damage before the operation, the extent of which
was unclear at the time of the implant. In 1 case, the
patient had undergone intra-aortic balloon pump (IABP)
malfunction during transfer from the referring hospital,
and in the other case the patient had suffered a
prolonged cardiac arrest immediately before surgery.
The only patient (17%) who survived to discharge from
the hospital is a woman who required left-side support
for post-partum dilated cardiomyopathy (DCM). On
admission, she was moribund and a Levitronix was
implanted as a salvage procedure. Once resuscitated
she was assessed for a long-term device and, 9 days after
the initial implant, she was successfully bridged to a
continuous-flow system and subsequently transplanted.
Overall, 6 patients (33%) were discharged home and
continue to be alive and well. The mean follow-up is 13
months with a range from 5 to 17 months.
Device Performance
Full circulatory support was provided for all patients
who achieved a good cardiac index (CI) with a low
pre-load. The RVAD system was run at a speed ranging
between 2,500 to 3,500 rpm, whereas the LVAD system
was run at a speed of between 3,000 and 4,500 rpm.
The flows ranged from 5 to 7 liters/min. In the LVAD
patients, a pulse-wave trace was seen (reflecting ejection), which varied between patients depending on
recovery of the underlying left ventricular function. In
the 4 patients who received an aortic prosthesis, the
valves were monitored for thrombus formation with a
transoesophageal echocardiogram (TOE) at different
stages. The prostheses were opening when the heart
was ejecting, but no thrombus formation was detected
either at TOE or at post-mortem, including the patient
with the mechanical valve. Only 1 patient in this
sub-group survived and his autograft valve was not
affected and he had no thromboembolic event. In 9
cases, an IABP was inserted before the Levitronix
implant, and it was kept in place until the clotting
profile was normalized. However, the presence of some
pulsatility did not seem to have any clinical impact.
Overall, the average duration of support was 14.2 ⫾
15.2 days for all patients with a range of between 1 and
64 days: in Group A it was 8.3 ⫾ 7.5 days, and in Group
B it was 26 ⫾ 20.2 days.
Complications
Bleeding requiring re-operation occurred in 8 patients
(44%) who had a gross coagulopathy— either because
of a prolonged cardiopulmonary bypass (CPB) time in
Group A, or because of the pre-operative state in Group B.
The Journal of Heart and Lung Transplantation
February 2006
In particular, 3 had had pre-operative liver dysfunction,
of whom 2 had associated sepsis; 2 had a long-term VAD
implant (1 re-do); 2 had had a prolonged cardiopulmonary bypass (CPB) time for congenital reconstructive
surgery (1 re-do); and 1 had a consumption coagulopathy. The bleeding source was not related to the cannulation site in any patient. One patient had a cerebral
embolic episode but recovered fully. This patient had
undergone a heart transplant and required Levitronix
implantation for primary graft failure. During the operation the patient spontaneously clotted the oxygenator
twice before being assisted with the device.
Pulmonary embolism (PE) was observed in 1 patient.
This patient was known to have had a PE before the
operation, which was demonstrated on computerized
tomography (CT) scan. After Levitronix implantation
his condition deteriorated and a transthoracic echocardiogram showed a large clot formation attached to the
right inflow cannula. It is not clear if the patient
underwent a new PE event related to the device or if his
deterioration was the evolution of his pre-operative
condition.
There were no device failures in any of the 18
patients assessed.
Heparin infusion was continued for the entire duration of support and no cases of heparin-induced thrombocytopenia were observed. Hemolysis was not clinically relevant. Plasma-free hemoglobin levels were not
measured because the majority of patients were hemofiltered, and this would have grossly affected the results
of the investigation.
Serum Lactate Levels
Serum lactate levels improved after Levitronix implantation. Before implantation, serum lactate was 8.3 ⫾ 5.2
mmol/liter and steadily decreased during the first 48
hours of assistance, reaching statistical significance only
at t48 (3.8 ⫾ 3.7, p ⬍ 0.01) (Figure 3). As expected,
Figure 3. Mean serum lactate measured every 6 hours.
The Journal of Heart and Lung Transplantation
Volume 25, Number 2
patients in multi-organ failure who did not respond to
treatment had higher serum lactate levels.
DISCUSSION
To our knowledge, this is the first report of the clinical
use of the Levitronix Centrimag short-term ventricular
assist device. Our experience suggests that this is a
versatile, effective and reliable system. It can be used
for left, right or bi-ventricular support. It can safely
provide adequate flows up to 10 liters/min that can be
adjusted according to different clinical conditions. This
allows full support at low revolutions per minute and
provides adequate decompression of the ventricles.
Moreover, we found quite useful the possibility of running low flows for extended periods of time during the
weaning process, or high flows in the septic patients.
Mortality at 30 days can be considered an acceptable
end-point to evaluate the effectiveness of a short-term
device. Our mortality rate at 30 days was 50%, which
compares favorably with that reported in previous
series. Deng et al7 reported that having either sepsis
or post-cardiotomy cardiogenic shock or concomitant
right ventricular failure reduces 1-year survival to 24%.
Our patients had many of these co-morbidities. In our
experience with pulsatile devices in patients with this
level of illness the results were no better than with the
Levitronix. Hence, although other devices could be
used, the Levitronix is a less expensive option that
might be more cost effective. Centrifugal pumps have
been used already, particularly in post-cardiotomy cardiogenic shock, but the mortality associated with their
use was very high. Moreover, the morbidity and rate of
device-related complications were substantial.8 –10 This
was due to a combination of factors, some related to the
period of time spent attempting the weaning from CPB,
and some intrinsic to the chosen device. Cell damage to
the blood components produces biochemical alterations
that are partially responsible for post-operative complications such as bleeding and thromboembolism.11
We found that, with the Levitronix, implementation
of support is prompt and easy. The operation per se
does not require the use of the cardiopulmonary bypass. In the post-cardiotomy group all patients were
already on CPB at the time of the device insertion, but
in some Group B patients we implanted the device-off
pump as we believe that some of them would have not
tolerated CPB due to their poor pre-operative state. The
implantation procedure is technically less challenging
than with other devices. The technique is very similar
to that used routinely to cannulate for CPB and does not
need to be performed by an experienced VAD surgeon.
Furthermore, the system is easy to use and can be
maintained by a perfusionist.
The distinctive characteristic of magnetically suspended pumps is the absence of bearings. The theoret-
De Robertis et al.
185
ical advantage with Levitronix is reduced cell damage
because there is minimal friction and heat generation
within the rotor housing. At the same time, there are
fewer areas of blood stagnation and fewer mechanical
parts that could be responsible for a malfunction.
The overall average duration of support was 14.2 ⫾
15.2 days and in the bi-VAD sub-group this was 20.3 ⫾
24.5 days. One patient in particular was assisted for 64
days, after which he died of multi-organ failure. No
device failures were noted. In our series, 44% of the
patients were re-operated for bleeding. However, they
were all particularly ill and the pre-operative conditions
increased their risk of bleeding. The thromboembolism
rate was low and it is not possible to relate the events
directly to the device.
The rationale for using the Levitronix Centrimag
short-term ventricular assist device in our “bridge-todecision” patients lies in the need for a reliable and
versatile system that can be quickly implemented in
situations of rapid deterioration. In particular, we believe it benefits those patients who have not completed
the full assessment for transplantation or for the implant
of a long-term device and those who have relative
contraindications to a VAD that may be temporary, such
as a recent neurologic insult. The absence of device
failures and the low complication rate makes the Levitronix safe to use for patients who need additional time
for evaluation or for sufficient recovery for purposes of
being considered suitable for a heart transplant or a
long-term device.
In conclusion, our study suggests that the Levitronix
Centrimag short-term ventricular assist device is an
effective, reliable and versatile system in the treatment
of end-stage cardiac failure. It provides the opportunity
to evaluate and treat critically ill patients that otherwise
would not be suitable for any treatment. It carries a low
incidence of VAD-related complications and is ideal for
bridge-to-decision situations. The results are encouraging in our small cohort of patients.
REFERENCES
1. Goldstein DJ, Oz MC. Mechanical support for postcardiotomy cardiogenic shock. Semin Thorac Cardiovasc
Surg 2000;12:220 – 8.
2. Helman DN, Morales DL, Edwards NM, et al. Left ventricular assist device bridge-to-transplant network improves
survival after failed cardiotomy. Ann Thorac Surg 1999;
68:1187–94.
3. Moazami N, Pasque MK, Moon MR, et al. Mechanical
support for isolated right ventricular failure in patients after
cardiotomy. J Heart Lung Transplant 2004;23:1371–5.
4. Nakazawa T, Makinouchi K, Takami Y, et al. The effect of
the impeller– driver magnetic coupling distance on hemolysis in a compact centrifugal pump. Artif Organs 1996;
20:252–7.
186
De Robertis et al.
5. Gobel C, Eilers R, Reul H, et al. A new blood pump for
cardiopulmonary bypass: the Hiflow centrifugal pump.
Artif Organs 1997;21:841–5.
6. Mueller JP, Kuenzli A, Reuthebuch O, et al. The
CentriMag: a new optimized centrifugal blood pump with
levitating impeller. Heart Surg Forum 2004;7:E477– 80.
7. Deng MC, Loebe M, El-Banayosy A, et al. Mechanical
circulatory support for advanced heart failure. Effect
of patient selection on outcome. Circulation 2001;103:
231–7.
8. Samuels L-E, Holmes EC, Thomas MP, et al. Management
of acute cardiac failure with mechanical assist: experi-
The Journal of Heart and Lung Transplantation
February 2006
ence with the Abiomed BVS 5000. Ann Thorac Surg
2001;71(suppl 3):S67–72.
9. Noon GP, Lafuente JA, Irwin S. Acute and temporary
ventricular support with BioMedicus centrifugal pump.
Ann Thorac Surg 1999;68:650 – 4.
10. Noon GP, Ball JW, Papaconstantinou HT. Clinical experience with BioMedicus centrifugal ventricular support in
172 patients. Artif Organs 1995;19:756 – 60.
11. Hampton CR, Verrier ED. Systemic consequences of
ventricular assist devices: alterations of coagulation, immune function, inflammation, and the neuroendocrine
system. Artif Organs 2002;26:902– 8.