Full text of DOI:10.1177/026765910101600408

Perfusion 2001; 16: 313–318
Haemolysis due to active venous drainage during
cardiopulmonary bypass: comparison of two
different techniques
Silvia Cirri, Luca Negri, Marco Babbini, Giorgio Latis, Boutros Khlat Anaesthesiology and Intensive Care
Department, Giuseppe Tarelli, Paolo Panisi, Enzo Mazzaro, Alessandro Bellisario Cardiac Surgery
Department, Bruna Borghetti, Fernanda Bordignon, Mario Ferrara, Helga Pavan Perfusion Department and
Massimo Meco Cardiac Surgery Department, Instituto Clinico Sant’Ambrogio, Milan
To facilitate mini-access for cardiac surgery, two different methods of active
venous drainage are used: vacuum assisted drainage and centrifugal pump
aspiration on the venous line. The aim of this study was to compare the
haemolysis produced using these two techniques.
From June to December 1999, 50 consecutive patients were operated on
using a ministernotomy. All of these patients had valvular surgery for either
valve repair or valve replacement (9 MVRepair, 11 MVR, 29 AVR, 1 AVR ꢀ
MVR). They were randomized into two groups: Group A, 25 patients who
underwent surgery where vacuum assisted drainage was used, and Group B,
25 patients where kinetic asssisted venous drainage with centrifugal pump
venous aspiration was used. Patient characteristics of both groups were similar
for age, gender, body weight, body surface area, height, cardiopulmonary
bypass (CPB) time, aortic crossclamp time, priming volume, cardioplegia
volume, haemoglobin concentration, haematocrit, serum creatinine, bilirubin,
lactate dehydrogenase (LDH), serum glutamic oxaloacetic transaminase
(sGOT), serum glutamic pyruvic transaminase (sGPT), aptoglobin, reticulocytes,
and platelet count. We checked all these laboratory parameters preoperatively,
at the end of CPB, and 2 and 24h after operation. We also checked
haemoglobinuria at these same time points. We assessed blood loss at 6, 12,
and 24h after the operation and calculated total postoperative bleeding. There
was a tendency towards a greater increase in LDH, sGOT and sGPT in Group A
more than in Group B, but these data did not reach statistical significance.
Platelet count was always lower in Group A and aptoglobin increased in Group A
more than in Group B. More patients in Group A had haemoglobinuria.
These findings indicate that haemolysis is increased more in patients treated
with vacuum assisted drainage, when compared to the rise in haemolysis in
those treated with centrifugal pump venous drainage. Total bleeding is also
greater in Group A.
Address for correspondence:
M Meco, Cardiac Surgery
Department, Istituto Clinico Sant’Ambrogio, Via Faravelli,
16-20149 Milan, Italy. E-mail: mecoma@inwind.it
© Arnold 2001
0267–6591(01)PF490OA
Downloaded from prf.sagepub.com at UNIV WASHINGTON LIBRARIES on March 28, 2015

314 S Cirri et al.
creatinine, total bilirubin, lactate dehydrogenase
(LDH), serum glutamic oxaloacetic transaminase
(sGOT), serum glutamic pyruvic transaminase
(sGPT), macroscopic haemoglobinuria, and reticu-
locyte count. We also noted the amount of blood loss
at 6, 12, and 24h after the end of the operation and
calculated the total postoperative loss. Anaesthesia
induction was obtained with fentanyl and propofol
while pancuronium was used for muscle relaxation.
With the intention of procuring early extubation
after the operation, anaesthesia was maintained
using a continuous propofol infusion. Trans-
esophageal echocardiography was used as a matter
of routine.
Ministernotomy was used in all cases: the incision
was carried out from the jugular notch to the 4–5th
right intercostal space for mitral pathology, and the
4–5th left intercostal space for aortic pathology. Sys-
temic moderate hypothermia (32°C) was used for all
patients. Myocardial protection was achieved with
cold crystalloid antegrade and retrograde cardiople-
gia and a vent was placed in the pulmonary artery.
The operations performed are listed in Table 2.
Aortic cannulation was performed with a
Medtronic DLP 20-22 french cannula (Medtronic,
Eden Prairie, MN, USA). Where aortic valve
surgery was performed, a single venous cannula
(Medtronic DLP 29/29 Fr.) was used and for mitral
valve surgery, we performed a double venous
cannulation (Medtronic DLP 22 Fr.). CPB was rou-
tinely carried out with a roller pump.
Introduction
Minimally invasive surgical approaches have been
applied recently in the management of valvular dis-
ease.1–3 It is claimed that these techniques, in fact,
reduce surgical trauma as well as postoperative pain,
blood loss, and length of stay in the intensive care unit.
In order to facilitate mini-access to a very narrow
surgical field and aid in the management of car-
diopulmonary bypass (CPB) with small diameter
aortic and venous cannulae, two different methods
of active venous drainage are used: vacuum assisted
venous drainage (VAVD) and the kinetic assisted
venous drainage (KAVD). VAVD is carried out by
the direct application of the vacuum to the venous
drainage tubing, as described by Munster and
colleagues.4 KAVD is carried out by inserting a
centrifugal pump, e.g. the Biomedicus BP 80
(Medtronic Biomedicus, Eden Prairie, MN, USA)
in the venous return line.5,6 The aim of this study
was to compare the haemolysis and bleeding which
were produced during ministernotomy operations
for valve surgery, using these two techniques.
Patients and methods
Fifty patients (26 males and 24 females) undergo-
ing elective cardiac surgery for valvular disease,
using ministernotomy, were randomized into two
groups who had similar age, weight, height, and
body surface area (BSA) (Table 1). Group A con-
sisted of 25 patients who were operated upon while
making use of VAVD during CPB. Group B com-
prised of 25 patients operated upon while using
KAVD during CPB.
The following parameters were monitored preop-
eratively, after the end of bypass, 2h postoperatively
and 24 h after the operation: haematocrit (Hct),
haemoglobin (Hb), platelet count, aptoglobin,
For VAVD and KAVD, we used the ‘Hamlet box’
technique described by Munster and colleagues4
and that described by Toomasian and McCarthy.6
The ‘Hamlet box’ technique involves
a vacuum source (wall suction), and a vacuum
line connecting the vacuum source and the
vacuum regulator;
•
Table 2 Surgical procedures
Table 1 Demographic data
Surgical procedure
Number of
patients
%
Demographic data
VAVD
KAVD
Mitral valve repair
9
11
29
1
18
22
58
2
Age (years)
Weight (kg)
Height (cm)
BSA (m2)
60.04 ꢁ 11.02
70.58 ꢁ 11.79
166.32 ꢁ 8.21
1.76 ꢁ 0.16
57.8 ꢁ 16.52
72.08 ꢁ 13.36
167.4 ꢁ 9.86
1.79 ꢁ 0.21
Mitral valve replacement
Aortic valve replacement
Aortic ꢀ mitral valve replacement
As it can be noticed in ministernotomy it is possible to
perform complex surgical procedures such as plastic mitral
valve repair and double valvular replacement.
All values are expressed as mean ꢁ standard deviation.
There are no significant differences between the two groups.
Downloaded from prf.sagepub.com at UNIV WASHINGTON LIBRARIES on March 28, 2015

Two different techniques of active venous drainage during CPB: a comparison 315
a vacuum tube connecting the vacuum regulator levels recorded for the parameters we used to indi-
•
•
•
and the reservoir;
cate haemolysis production.
a closed hard-shell reservoir with a vacuum relief
valve set on the top of the reservoir;
a vacuum regulator pre-set at ꢂ20mmHg.
At the end of CPB, Group A showed higher
levels of LDH (627.12 ꢁ 302.76 vs 494.04 ꢁ
175.79, p ꢅ 0.06), sGOT (64.64 ꢁ 54.93 vs 36.54 ꢁ
13.95, p ꢅ 0.01) and sGPT (22.84 ꢁ 13.99 vs
16.82 ꢁ 8.5, p ꢅ 0.07) when compared with Group
B. The number of patients in Group A with macro-
scopic haemoglobinuria at the end of CPB was
higher than in Group B (7 patients – 29% vs 1
patient – 4.8%, p ꢅ 0.044).
For KAVD, the technique involves
a Biomedicus BP 80 pumphead (Medtronic
Biomedicus, Eden Prairie, MN, USA) inserted
into the venous line at the level of the console
roller-pump tubing;
•
the pumphead placed inside a Biomedicus
remote drive unit attached to a pump mast;
a luer-lock connector positioned approximately
10 cm before the inlet to the kinetic pumphead;
a pressure separation dome connected to 1/8ꢃ
tubing and attached to a pressure monitoring site
on the Biomedicus pump console to monitor the
negative pressure generated by the kinetic pump;
maximum negative pressure ꢂ20mmHg.
•
•
•
At 2 h after CPB, the results were the same with
regard to LDH (706.29 ꢁ 279.19 vs 511.39 ꢁ 189.64,
p ꢅ 0.006), sGOT (75.56 ꢁ 53.09 vs 46.08 ꢁ 19.16,
p ꢅ 0.01), and haemoglobinuria (8 patients – 33.3%
vs 2 patients – 10.5%) Platelet count at this sam-
pling time was higher in Group B (117.85 ꢁ
45.70 ꢆ 103/ml vs 153.13 ꢁ 57.62 ꢆ 103/ml, p ꢅ
0.023). In the first postoperative day, i.e. 24 h post-
operation, only LDH showed a statistically signifi-
cant difference in levels between Group A and
Group B (731.04 ꢁ 253.22 vs 584.17 ꢁ 124.53,
p ꢅ 0.012) although sGOT values came close to
significance (81.04 ꢁ 51.53 vs 58.83 ꢁ 29.77,
p ꢅ 0.06).
•
The priming volume in both groups was 1250cm3
(800 cm3 Ringer’s Lactate ꢀ 200 cm3 Trometamol
(Tham) ꢀ 200 cm3 mannitol 18% ꢀ 50 cm3 albumin
20%); crossclamp time and total bypass time were
similar in both groups (Table 3).
It is interesting to note that all values of the
haemolysis indicators are constantly higher in
Group A patients, even if these values do not often
reach statistical significance. The number of
patients with macroscopic haemoglobinuria was
higher in the VAVD group. With platelets, however,
the numbers are generally higher in Group B. Post-
operative blood loss was greater in Group A than
in Group B at all time points (Table 5) with the total
loss being 698 ꢁ 957.46 ml vs 390.40 ꢁ 241.47ml,
p ꢅ 0.03.
Statistical analysis
All data are presented as mean ꢁ standard devi-
ation. All discrete variables were analysed using
Student’s t-test, verified by Bonferroni’s correction.
All continuous variables were analysed using the
chi-square test and Fisher’s exact test where applic-
able. p values ꢄ0.05 were taken to show statistical
significance.
Results
Preoperative data, crossclamp times and pump
times were similar in both groups. Table 4 shows the
Discussion
During extracorporeal circulation in CPB, blood is
exposed to anomalous mechanical and environ-
mental factors, such as high shear stress, turbulence,
decreased oncotic pressure cause by dilution of
plasma, and hypothermia. Research has identified
the forces that precipitate haemolysis as pressure
and shear stress, with the latter entity being the
most detrimental to the red blood cells.7 The dam-
age caused is manifest in immediate and delayed
Table 3 CPB data
CPB data
VAVD
KAVD
Priming volume (ml)
Cardioplegia volume (ml)
Crossclamp time (min)
Bypass time (min)
1302 ꢁ 236
1570 ꢁ 437
77.35 ꢁ 15.24
103.0 ꢁ 19.47
1370 ꢁ 286
1360 ꢁ 391
73.44 ꢁ 20.23
91.44 ꢁ 24.24
All values are expressed as mean ꢁ standard deviation.
There are no statistically significant differences between
the two groups.
Downloaded from prf.sagepub.com at UNIV WASHINGTON LIBRARIES on March 28, 2015

Downloaded from prf.sagepub.com at UNIV WASHINGTON LIBRARIES on March 28, 2015

Two different techniques of active venous drainage during CPB: a comparison 317
Table 5 Bleeding values (ml)
VAVD
0–6h
KAVD
0–6h
7–12h
13–24h
Total
7–12h
13–24h
Total
460.4 ꢁ 903.9 111.4 ꢁ 129.2 121.0 ꢁ 151.7 698.0 ꢁ 957.5 233.0 ꢁ 164.1 804 ꢁ 72.3 81.8 ꢁ 68.4
*p ꢅ 0.009
390.4 ꢁ 241.5
*p ꢅ 0.003
All data are expressed as mean ꢁ standard deviation. *p values ꢄ 0.05.
haemolysis and by changes in the mechanical prop-
erties of red blood cells.
Conclusions
In minimally invasive surgery, for work to be car-
ried out in restricted spaces, it is necessary to reduce
the dimensions of the cannulae. This situation calls
for active venous drainage, of which there are two
techniques which are most widely used: VAVD and
KAVD.4–6
In a study of suction-induced haemolysis,
Gregoretti demonstrated that an increase in vac-
uum pressure resulted in an increase in haemolysis.8
Shin and colleagues found the same quantity of haem-
olysis in patients treated with VAVD and in patients
treated with a siphon system;9 results are confirmed
by Lau et al.10 From other studies, we also know that
a centrifugal pump causes less haemolysis, less
reduction in lytic complement inhibitors, and also
less platelet damage and platelet depletion than a
roller pump.11,12 Following the results of these stud-
ies, we sought to compare VAVD and KAVD to
ascertain which produced less haemolysis.
We found that haemolysis indicators are con-
stantly higher when VAVD is used and that KAVD
causes less damage to the erythrocytes. It would
appear from this study that haemolysis is not only
directly attributable to the suction applied, but is
also increased with an increase in the level of nega-
tive pressure applied beyond 20mmHg. It is prob-
able that the direct application of vacuum to the
venous reservoir causes an augmentation of turbu-
lence and shear stress to the red blood cells.
With KAVD, the turbulence and shear stress are
evidently smaller because the centrifugal pump is
included in the venous line, thus, there is no inter-
ruption to the continuity of flow. These consider-
ations may also be applied to the platelets. Since
the platelet count is always higher in the KAVD
group, the lower quantity of bleeding in this group
is probably due to the lower platelet structural
damage.
Both these methods are sure and effective for min-
imally invasive surgery that demands the use of
cannulae of small diameter. Flows can be main-
tained without difficulty and without technical dif-
ficulty. However, our study shows that the red cell
and platelet damage is reduced when the kinetic
assisted mode of venous drainage is used.
References
1 Olin CL, Peterffy A. Minimal access aortic valve
surgery. Eur J Cardiothorac Surg 1999; 15(Suppl. 1):
S32–38.
2 Bouchard D, Perrault LP, Carrier M, Menache P,
Bel A, Pelletier LC. Ministernotomy for aortic valve
replacement: a study of the preliminary experience.
Can J Surg 2000; 43: 39–42.
3 Loulmet DF, Carpentier A, Cho PW et al. Less
invasive technique for mitral surgery. J Thorac
Cardiovasc Surg 1998; 115: 772–79.
4 Munster K, Andersen U, Mikkelsen J, Petterson G.
Vacuum assisted venous drainage (VAVD).
Perfusion 1999; 14: 419–23.
5 McCuster K, Hoffman D, Maldarelli W, Topliz S,
Sisto D. High-flow femoro-femoral bypass utilizing
small cannulae and a centrifugal pump on the
venous side. Perfusion 1992; 7: 295–300.
6 Toomasian JM, McCarthy JP. Total extrathoracic
cardiopulmonary support with kinetic assisted
venous drainage: experience in 50 patients. Perfusion
1998; 13: 137–43.
7 Kameneva MV, Undar A, Antaki JF, Watach MJ,
Calhoon JH, Borevetz HS. Decrease in red blood
cell deformability caused by hypothermia,
hemodilution, and mechanical stress: factors related
to cardiopulmonary bypass. ASAIO J 1999; 45:
307–10.
8 Gregoretti S. Suction-induced hemolysis at
various vacuum pressures: implications for
intraoperative blood salvage. Transfusion 1996; 36:
57–60.
Downloaded from prf.sagepub.com at UNIV WASHINGTON LIBRARIES on March 28, 2015

318 S Cirri et al.
9 Shin H, Yozu R, Maehara T et al. Vacuum
assisted cardiopulmonary bypass in minimally
invasive cardiac surgery: its feasibility and
effects on haemolysis. Artif Organs 2000; 24:
450–53.
10 Lau CL, Posther KE, Stephenson GR et al.
Mini-circuit cardiopulmonary bypass with vacuum
assisted venous drainage. Feasibility of an
asanguineous prime in the neonate. Perfusion 1999;
14: 389–96.
11 Nishinaka T, Nishida H, Endo M, Miyagishima M,
Ohtsuka G, Koyanagi H. Less blood damage in the
impeller centrifugal pump: a comparative study with
the roller pump in open heart surgery. Artif Organs
1996; 20: 707–10.
12 Hansbro SD, Sharpe DA, Catchpole R et al.
Hemolysis during cardiopulmonary bypass: an
in vivo comparison of standard roller pumps,
nonocclusive roller pumps and centrifugal pumps.
Perfusion 1999; 14: 3–10.
Downloaded from prf.sagepub.com at UNIV WASHINGTON LIBRARIES on March 28, 2015