The new wave of left-ventricular assist devices (LVADs), none yet
approved for sale in the US but some available in Europe, go for
small size and greater durability to make them more patient-friendly
and, potentially, cut the need for replacement due to wear and tear,
compared with the pulsatile-pump first-generation devices.

Some of these advanced, magnet-driven LVADs were able to sustain
>80% of transplant-listed patients for at least six months or until
transplantation if it occurred earlier, in several small
observational studies presented here last week at the International
Society for Heart and Lung Transplantation [ISHLT] 2008 Annual
Meeting [1,2,3]. And, although the reported studies involved
handfuls of patients, the risk of complications--especially the most
daunting ones in LVAD therapy (stroke, infections, and bleeding)--
appeared low with all the devices.

What makes these so-called "third-generation" LVADs different from
the long-available pulsatile HeartMate XVE (Thoratec) and Novacor
LVAS (WorldHeart) pumps is a continuous-flow propulsion system
driven by electromagnets that spin an impeller--the rotor that
actually moves the blood--that is levitated in position so that it
doesn't actually touch the pump housing. With their impellers
levitated by more magnets or the dynamic forces of the blood itself,
these pumps have only the one moving part and little of the wear and
tear that ages the mechanical pulsatile pumps.

In various stages of development, these new devices include:

Centrifugal pumps with magnetically powered, hydrodynamically
levitated impellers, such as the VentrAssist (Ventracor) and the
HeartWare HVAD (HeartWare).

"Maglev" pumps, with magnetically powered, magnetically levitated
impellers, such as the DuraHeart LVAS (Terumo Heart) centrifugal
pump, the Levacor VAD (WorldHeart) centrifugal pump, and the Incor
(Berlin Heart) axial-flow pump.
Dr Andrew J Boyle (University of Minnesota, Minneapolis), who
presented the VentrAssist study at the ISHLT meeting, would also
include in the mix mechanically driven axial-flow LVADs like the
HeartMate 2 (Thoratec), which have been dubbed "second generation."
That device, available in Europe for about three years, was given a
unanimous vote of confidence by an FDA premarket advisory panel in
December, 2007, heartwire reported at the time, and appears headed
for approval. The HeartMate 2, about the same size as the
VentrAssist, is probably as much of a clinical advance over
pulsatile pumps as are the new magnetically driven devices, Boyle,
who has participated in trials of both pumps, told heartwire.

Boyle says he avoids terms like "second generation" for the
HeartMate 2 and the "third-generation" label commonly given to the
magnetically driven devices. "That's just a marketing ploy. The
reality is that all of them, other than the [HeartMate] XVE and the
Novacor [LVAS], are the new generation," he said. "Whether they are
called second or third generation, I think that's all baloney." They
use different engineering approaches to address the same problems,
he said, "but I think where the rubber meets the road, which is in
outcomes, there will be similar results."

And, Boyle said, the new-generation devices will probably all have
the same longevity. "I don't think wear and tear within the pump is
going to be an issue for any of them." But whether greater
durability manifests clinically has yet to be shown--that's hard to
do in bridge-therapy studies, in which the time to transplantation
is often less than the lifetime of even the first-generation LVADs.

The HeartMate XVE, currently the gold standard and approved for both
bridge and destination therapy in the US, has a lifetime of about 15
to 18 months, Boyle said. "Then you have to replace the whole pump,
usually because of bearing wear."

But, "with more and more people getting LVADs, there are more people
at the top of transplant lists, so waiting times are lengthening,"
Boyle observed. It's increasingly possible that in some areas of the
country "the waiting time for a donor to be found for a particular
patient will exceed the durability of the XVE." The new-generation
LVADs, he said, are expected to last five to 10 years. That's based
on bench testing, since destination-therapy studies have yet to be
done; Boyle has had patients live with one for up to 2.5 years, "and
there are longer times at other centers."

In the prospective experience he reported at the ISHLT meeting, 28
patients awaiting a donor heart were implanted with the VentrAssist
at four US centers. Of those, 23, or 82%, met the primary end point
of survival to transplant or on circulatory support for at least 180
days. All were in NYHA functional class 4 at the start. At 14 weeks,
>90% were in either class 1 or 2, with none in class 4. At 180 days,
>78% were still in class 1 or 2, while those in class 3 had grown to
22%; but again, none were in class 4.

The 38-point mean improvement in quality-of-life score (based on the
Minnesota Living with Heart Failure Questionnaire)
was "statistically and clinically significant," Boyle said.

The overwhelming majority of serious adverse events occurred within
the first 30 days after device implantation; adverse-event rates
were comparable to what has been observed with other LVADs,
according to Boyle.

Similar complication rates were reported at the meeting for the
magnetically driven, hydrodynamically levitated HeartWare HVAD. In a
23-patient series presented by Dr Georg Wieselthaler (Vienna General
Hospital, Austria), 9% had strokes (all ischemic), 4% had transient
ischemic attacks, 13% developed bleeding requiring reoperation, and
13% had infections at the percutaneous entry site over six months.

Survival at 180 days with the HeartWare device was 91%; 9% had been
transplanted by that point. Both of the deaths were from sepsis. The
mean LVAD-support time was 272 days.

Reporting on the maglev DuraHeart pump at the ISHLT sessions, Terumo
Heart CEO Dr Chisato Nojiri said that the adverse-event rates in his
33-patient series were about 0.22 per patient year for bleeding
requiring surgery, 0.40 per patient year for infections at the
percutaneous site or device pocket, and 0.28 per patient year for
any stroke. There were no instances of device-related thrombosis.

With the addition of 22 patients implanted with the device after
European market approval based on the first 33 patients, the 180-day
rate of survival for the DuraHeart was 82% (mean support time, 214
days). About one-fourth of the patients had been on the device for
at least a year; among those, the mean duration of support was 568
days, according to Nojiri.

An FDA-approved bridge-to-transplant trial with a target enrollment
of 140 patients is in the works for the DuraHeart pump, the company
recently announced. At least 32 patients have been enrolled in the
HeartWare device's corresponding trial, according to Wieselthaler.
And, observed Boyle, a similar pivotal trial for the VentrAssist
pump has entered 48 of a projected 140 patients for the bridge
cohort and 20 out of an as-yet undetermined total number of
destination-therapy patients.

Boyle reports being on the VentrAssist scientific advisory board;
his coauthor Dr John Woodard is chief science officer for Ventracor,
and coauthor Dr Yoshifumi Naka (Columbia University Medical Center,
New York, NY) reports being a consultant to Ventracor, Terumo Heart,
Thoratec, and Cardiomems.

Boyle AJ. The VentrAssist LVAD as a bridge to cardiac
transplantation: Results of the US Feasibility Trial. International
Society for Heart and Lung Transplantation 2008 Annual Meeting;
April 12, 2008; Boston, MA. Abstract 518.
Wieselthaler GM. Experience with the novel Heartware HVAD with
hydromagnetically levitated rotor in a multi-institutional trial.
International Society for Heart and Lung Transplantation 2008 Annual
Meeting; April 12, 2008; Boston, MA. Abstract 515.
Nojiri C. Long-term circulatory support with the DuraHeart maglev
centrifugal left ventricular assist system for advanced heart
failure patients eligible to transplantation: European experiences.
International Society for Heart and Lung Transplantation 2008 Annual
Meeting; April 12, 2008; Boston, MA. Abstract 514.