521101

research-article2014
PRF0010.1177/0267659114521101PerfusionIssitt et al.

Original Paper

Perfusion
2014, Vol. 29(3) 194­–198
Clinical experience with Affinity Pixie™ © The Author(s) 2014
Reprints and permissions:
oxygenation system in paediatric and sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0267659114521101
infant patients prf.sagepub.com

R Issitt,1 A Robertson,1 N Cross,1 R Crook,1 V Molyneux,1

M Shaw,1 N Walton1 and V Tsang2

Abstract
Introduction: Great Ormond Street Children’s Hospital undertakes over 500 open heart cardiothoracic procedures
requiring cardiopulmonary bypass per year. Data from our centre show that many of our neonatal/paediatric patients
require higher cardiac indexes than previously thought. We evaluated the new Pixie™ oxygenation system, rated from
0.1 L/min to 2 L/min, to determine if it could be used for these patients.
Methods: Between 2010 and 2012, 250 Pixie™ oxygenators were used on consecutive patients requiring correction of
congenital cardiac defects. Data were collected on FiO2 requirements, oxygenator pressure drop and gaseous micro-
emboli handling. Retrospective analysis was also undertaken on the procedures and demographics of all patients during
2011-2012 to determine the percentage of patients on whom the Pixie™ could be used.
Results: Analysis of the procedures undertaken at Great Ormond Street Hospital (GOSH) showed that 89% were in pa-
tients under 20 kg, requiring a flow rate of <2 L/min (at a base cardiac index of 2.8 L/min/m2).The maximum FiO2 required
at 2.5 L/min was 85%. Gaseous microemboli were reduced by 82.5±9.9% and bubble volume was decreased by 94.3±8.4%
from the ‘venous’ pre-oxygenator to the ‘arterial’ post-oxygenator.
Discussion: The Pixie™ oxygenator proved effective at flows up to 2.5 L/min, with air-handling capabilities comparable
with other oxygenators. This represents a single oxygenator that could potentially be used to cover 89% of our surgical
procedures. However, we believe that, for the smallest patients (i.e. < 2kg), a smaller priming oxygenator should be used
in order to limit unnecessary haemodilution in this vulnerable group.

Keywords
cardiopulmonary bypass; CPB; gaseous microemboli; GME; congenital heart disease; CHD

Introduction
At least 8 out of every 1000 babies born each year have a
congenital cardiac defect. Approximately half of these
babies have a minor defect and will not need any treat-
ment, but the rest require medical therapy or surgery via
direct intervention, supported by extracorporeal circu-
lation. The techniques for perfusion and management
of paediatric open heart surgery are quite different when
Previous publications have suggested that paediatric
patients are more sensitive to extracorporeal circulation
1Department of Clinical Perfusion Great Ormond Street Children’s
Hospital, London, UK
2Department of Paediatric Cardiothoracic Surgery, Great Ormond

compared to adults, arising from differences in complex Street Children’s Hospital, London, UK
anatomical and congenital abnormalities in the heart Corresponding author:
and great vessels as well as due to the broad range in Richard Issitt
patient size and weight, associated pathological sequelae Department of Clinical Perfusion
and increased metabolic requirements.1 At Great Cardiac Theatres
Great Ormond Street Children’s Hospital
Ormond Street Children’s Hospital, over 500 congenital Level 3
cardiothoracic procedures requiring cardiopulmonary Morgan Stanley Clinical Building
bypass (CPB) are undertaken each year, ranging from London, WC1N 3JH, UK.
neonatal to adolescent ages. Email: richard.issitt@gosh.nhs.uk

Downloaded from prf.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016

Issitt et al. 195
Figure 1.  Retrospective data of the maximum flow rates
used during CPB compared with the body surface area (BSA).
Data taken from all cases performed in the first quarter of
2010.
because of the comparatively large foreign surface area
of the circuit as compared to adults.2 As a consequence,
many of the devices used for paediatric CPB have been
developed with a specific goal of reducing the size of the
device and, consequently, reducing the impact on the
paediatric patient.3 Equally, the application of biocom-
patible coating for extracorporeal circuits has become
routine to minimise the effects of the adhesion of plate-
lets and the subsequent systemic inflammatory
response.4 Retrospective data from this unit have shown
that the traditional cardiac index of 2.8 L/min/m2 may
not be sufficient for neonatal and small paediatric
patients, with some requiring a cardiac index in excess
of 3.5 L/min/m2 (Figure 1). For a number of the patients,
this has taken the required flow rates above 1.5 L/min,
which has been the cut-off point for our current neona-
tal/paediatric oxygenator. Therefore, it has become nec-
essary to use an oxygenator that is capable of providing
a broad range of blood flows without sacrificing the
aforementioned benefits.
The clinical results and practical experience obtained
after the first 2 years of clinical experience with a new
commercially available oxygenator (Affinity Pixie
Oxygenation Systemtm, Medtronic Inc., Kerkrade, the
Netherlands), designed to cover a comparatively broad
flow range, are presented in this paper. The adoption
into routine clinical use at this institution and the deci-
sion process used to add the system to the institutional
protocol is discussed, as well as preliminary evaluation
of gaseous microemboli (GME) activity and oxygenator
efficiency.
Methods
Ethical Guidelines
The Institution’s Research and Development Department
reviewed this project prospectively and ethical approval
was obtained from the local ethics committee.
Downloaded from prf.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Patients
Between January 2010 and December 2012, 250 con-
secutive patients undergoing cardiothoracic surgical
procedures with cardiopulmonary bypass for the cor-
rection of congenital cardiac abnormalities, requiring a
calculated flow <2 L/min at a cardiac index of 2.8 L/
min/m2, were included in the evaluation. Demographic
and operative data are shown in Table 1. There were no
inclusion/exclusion criteria to ensure that as broad a
population as possible was tested.
Anaesthesia
Anaesthesia was induced by inhalation of sevoflurane in
oxygen and, after induction, fentanyl 5 µg/kg and pan-
curonium 100 µg/kg were given and anaesthesia was
maintained with isoflurane 1.0% in oxygen and air.
After tracheal intubation, arterial and central venous
lines were inserted. Further incremental doses of fen-
tanyl, up to 25 µg/kg, were given during the case.
Cardiopulmonary bypass
The CPB circuit consisted of the Affinity Pixietm oxy-
genator and a hard-shell venous reservoir (Medtronic
Inc.) with either 3/16x1/4, 1/4x1/4, or 1/4x3/8 tubing
sets, dependent upon the patient size. The Stöcket S5tm
(Stöckert, Sorin Group GmbH, Munich, Germany)
heart-lung machine with pole-mounted roller pumps
was used with a 3Ttm heater/chiller (Stöckert, Sorin
Group GmbH). The total base prime volume was 330,
350 or 500 ml, depending on circuit size. This consisted
of packed red blood cells (PRBC) of the patient’s blood
group, a synthetic colloid (Gelofusine, B. Braun
Melsungen AG, Melsungen, Germany) with mannitol
20% w/v (Baxter, Thetford, United Kingdom) 2.5 ml/kg
and heparin 1000 unit/ml, 1.5ml (Wockhart, Wrexham,
United Kingdom). The prime was then washed with
1000 ml of balanced crystalloid solution (Plasmalyte
7.4, Baxter) by performing pre-bypass ultrafiltration
(PBUF), carried out as previously described, to ensure
an initial on-CPB haematocrit of 30%.5 Biochemical
compatibility was then attained using sodium bicar-
bonate as a buffering agent. All patients were system-
ically cooled to nasopharyngeal temperatures
between 28-35°C and myocardial protection was
achieved with 30 ml/kg of cold blood cardioplegia
(4:1 blood:cardioplegia to a final concentration of 20
mmol) of St Thomas’s Solution (IVEX Pharmaceuticals,
Larne, Northern Ireland). During the rewarming phase
of CPB, a gradient no greater than 10°C between the
patient’s nasopharyngeal temperature probe and the
heater/chiller (maximum arterial blood temperature
37.5°C) was maintained until a maximum nasopharyn-
geal temperature of 36°C was achieved.
196 Perfusion 29(3)

Table 1.  Operative and patient demographic data. Oxygenator pressure drop
Case Type Percentage (%) Pre- and post-oxygenator membrane pressures were mea-
ALCAPA 0.5 sured throughout CPB on the first 100 patients. Maximum
ASD* 22.3 pressure drops were obtained from data collected using
AVSD 5.7 the Data Management System within the Stöckert S5tm
Aortic Valve Regurgitation 0.5 heart-lung machine (DMS, Stöckert, Sorin Group GmbH).
Aortic Valve Stenosis 3.1
Sub AV Stenosis 1.3
Berlin Heart BiVAD Implantation 0.8 Microemboli detection
Berlin Heart LVAD Implantation 1.0
Bilateral Lung 0.3 Gaseous microemboli detection was measured using the
BT Shunt 0.8 GAMPTtm clinical bubble detector (BCC200, GAMPT
Damus Kaye Stansel 0.5 GmbH, Merseberg, Germany). The venous detector was
Fallots Tetralogy* 8.8
placed immediately proximal to the oxygenator inlet
Glenn 3.1
Heart Transplant* 2.9 and the arterial immediately distal to the oxygenator
Hypoplastic Ao Arch 0.3 outlet. The bubble detectors were calibrated by the man-
IAA 1.3 ufacturers before clinical use and measured GME under
L-Atrio-ventricular Valve Repair 0.3 500 μM for number and volume (μL), as well as giving
LVOTO Repair 0.8 over-range data (GME over 500 μM).
MV Repair 2.9
Norwood 1 (BT) 0.3
PA Reconstruction 1.6 Results
PDA 1.3
Rastelli 1.0 During the 2011-2012 period, 535 cardiothoracic pro-
Ross 1.0 cedures were carried out at GOSH. Of these, 479 (89%)
RVOTO 1.6 were in patients requiring a calculated cardiac output
RV-PA Conduit Replacement* 1.0 of <2 L/min (at a cardiac index of 2.8 ml/min/m2).
RV-PA Conduit + Closure of Shunt 0.3 Patients ranging from 5 days to 9 years, 10 months
Switch 2.1 and 22 days, with corresponding weights ranging from
TAPVD 1.6
2.9 kg to 18 kg, were perfused using the Pixietm oxygen-
TCPC 8.8
Tracheal Reconstruction 1.6
ation system (see Table 1).
Truncus Arterious 0.5
VSD* 18.2 Gas exchange
VSD + Debanding 1.8
Wardens Procedure 0.3 Maintenance of PaO2 at 25 kPa was attained using in-
  100.0 line blood-gas monitoring (CDI500, Terumo, Tokyo,
Height (cm) 87.9±17 Japan). We found that, upon re-warming, the FiO2 per-
Weight (kg) 12.4±4.7 centages required for maintaining a PaO2 of 25 kPa at a
BSA (m2) 0.5±0.2 flow around 2500 ml/min (above the official rated flow
CPB (mins) 74.0±50.4 of 2 L/min) were relatively high at 80-85%. The average
X-Clamp (mins) 39.8±25.9 values for PaO2, FiO2 and SvO2 were 30.7 kPa, 49.8% and
ALCAPA; anomalous left coronary artery arising from the pulmonary
91.8%, respectively. Venous saturations (SvO2) did not
artery, ASD; atrial septal defect, AVSD; atrio-ventricular septal fall below 75%.
defect, BiVAD; biventricular assist device, LVAD; left ventricular
assist device, BT; Blalock-Taussig, Ao; aortic, IAA; interrupted aortic
arch, LVOTO; left ventricular outflow tract obstruction, MV; mitral Oxygenator pressure drop
valve, PA; pulmonary artery, PDA; patent ductus arteriosis, RVOTO;
right ventricular outflow tract obstruction, RV-PA; right ventricle – Pressure gradients across the membrane showed a linear
pulmonary artery, TAPVD; total anomolous pulmonary venous drainage, relationship along the entire flow rate range (Figure 2),
TCPC; total cavopulmonary connection, VSD; ventricular septal defect.
which is comparable with other oxygenators within the
Data given as mean±SD. * indicates operations in which maximum flow
was >2 L/min. neonatal/paediatric range.6

Oxygenator efficiency Microemboli detection

The patient’s O2 tension (PaO2) was maintained at 25 Bubble count data showed the number of GME detected
kPa via in-line monitoring (CDI100, Terumo, Leuven, post oxygenator reduced by a mean of 82.5±9.9% and
Belgium), using alpha-stat blood gas management and the bubble volume decrease by 94.3±8.4%, even in cir-
recording the necessary fraction of inspired oxygen (FiO2). cumstances of gross air entrainment (Table 2).

Downloaded from prf.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016

Issitt et al. 197
Figure 2.  Pressure drop across the Pixie™ oxygenator. Data
taken from the first 100 patients.
Discussion
Cardiopulmonary bypass is required for the vast major-
ity of surgically corrective procedures undertaken within
an open heart and on the great vessels. Recent work by
Fujii et al.7 has shown that paediatric patients require far
higher cardiac indexes than traditionally given to adult
patients, especially those with cyanotic heart disease.
This is reflected in our retrospective data (Figure 1). A
number of the patients treated had cyanotic cardiac
lesions, which are commonly associated with aortopul-
monary collateral vessels. Sakamoto et al. showed the
deleterious effects of collateral shunt flow on lactate lev-
els in patients with major aorto-pulmonary collaterals,
further suggesting the need for increased flow rates in
this population.8 To this end, our departmental policy
has been modified to reflect this evidence and so, calcu-
lated flows based upon a cardiac index of 2.8 L/min/m2
have been adopted as standard, although far higher flow
rates can be required (Figure 1). This was demonstrated
in our evaluation by a number of patients (6%) requiring
>2 L/min, even though their 2.8 L/min/m2 index was <2
L/min. The need for this increased flow was based upon
rising serum lactate levels and the requirement of main-
taining SvO2 >75%. We could not define a particular char-
acteristic of the patients requiring higher than 2.8 L/min/
m2 flow rates; the lesions in which they were observed
were both cyanotic and acyanotic (Table 1). This move
towards higher flows challenges the maximum flow
capacity of established low-prime oxygenators. Therefore,
low-prime, high powered oxygenation systems are
required. The Affinity Pixietm oxygenator has a low
prime volume (48 mL) with a membrane surface area of
(0.67 m2), providing a recommended blood flow range of
0.1 – 2 L/min. It also incorporates a new biocompatible
coating, but we did not seek to test its efficacy in this
evaluation. Our institutional practice is to medically
Downloaded from prf.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
manage patients until an adequate weight has been
achieved before surgical intervention is undertaken. It is
worth stressing though that, in the smallest and rarest
patients, i.e. those under 2 kg, who are particularly at risk
from haemodilution and disruption of haemostatic
integrity, we would still advocate the use of a smaller
oxygenator which has a smaller membrane surface area
and lower prime volume.
Our analysis of procedures carried out over the
2011-2012 period showed that 89% of our patient pop-
ulation required a calculated flow less than 2 L/min,
based upon a base cardiac index of 2.8 L/min/m2.
During our evaluation of the Pixietm oxygenator, we
attempted to determine whether the oxygenator was
capable of safely providing adequate oxygenation up to
2 L/min. We found that, even above the recommended
2 L/min limit, the oxygenator was able to provide ade-
quate oxygenation (PaO2 >25 kPa), with the highest
FiO2 recorded at 85% during the re-warming phase of
CPB. However, it should be noted that the parameters
appeared to be linked to the patient rather than the flow
required; we observed cases where a FiO2 of 50% was
sufficient to maintain an adequate PaO2 at 2.5 L/min
and a FiO2 of 70% in a patient requiring 1.4 L/min. We
would, therefore, suggest that the specific pathology is
equally as important as the size of the patient in deter-
mining the oxygen requirement. However, at 85%,
there is little margin to improve if problems are encoun-
tered; it remains to be seen, therefore, if 85% is too
high. Further data should be sought against other oxy-
genators to compare.
The issue of GME is of paramount importance in
perfusion9 and so the GAMPT bubble detector BCC200tm
is available for use on every procedure. We found that
the Pixietm removed 82.5±9.9% of GME and decreased
the bubble volume by 94.3±8.4%, suggesting that the
fibre bundle acted as an excellent bubble trap. It is inter-
esting to note that the number of GME recorded was
higher than other tests of oxygenators, especially in
‘venous’ measurement.2 There are a number of reasons
for this; firstly, we use the GAMPT BCC200tm for GME
data acquisition, which has been previously shown to be
incomparable with data from centres using the EDACtm
GME detection system.10 Secondly, many other trials of
oxygenators are in vitro and examine the effects of GME
from the venous line in isolation and, therefore, are not
able to include the major source of GME encountered
within a circuit, which is not from the venous line, but
facilitated through the cardiotomy suction and vent
lines (especially in cyanotic patients with extensive col-
lateral circulation which requires increased suction).8
Thirdly, previous experiments have used a prime solu-
tion with little or no colloid, which will have an effect on
the viscosity of the prime and, therefore, the GME-
carrying potential.11
198 Perfusion 29(3)

Table 2.  Microemboli data from the GAMPT™ bubble detector system.

Microemboli Data

  Venous Arterial  

  Number Volume (μL) Number Volume (μL) Number Change (%) Volume Change (%)
Mean 159293 76.074 29817 1.181 82.50 94.32
SD 159510 151.040 41346 1.477 9.89 8.40
Microemboli Data
  Venous Arterial  
  Number Volume (μL) Number Volume (μL) Number Change (%) Volume Change (%)
Mean 159293.1 76.1 29816.7 1.2 82.5 94.3
SD 159509.9 151.0 41345.5 1.5 9.9 8.4
Median 102013.0 21.2 14643.0 0.7 83.2 97.6
Minimum 16458.0 1.4 1295.0 0.0 57.7 68.5
Maximum 746586.0 714.0 184650.0 6.2 98.0 100.0
SD; standard deviation.

In conclusion, our findings show that the Affinity References

Pixietm oxygenation system is capable of delivering flows
above the official rated maximum flow of 2 L/min with-
out compromising oxygen delivery and, therefore,
patient safety. We have demonstrated that this oxygen-
ator can supply 89% of our open heart procedures, espe-
cially those patients with increased flow requirements,
without upsizing to a small adult system. We would,
however, continue to recommend that smaller oxygen-
ators should be used in those patients whose weight is
<2 kg to avoid unnecessary haemodilution and surface
area exposure.
Limitations
We endeavoured to determine how the Pixietm oxygen-
ator functioned clinically and did not seek to compare
directly with other oxygenators. We feel, therefore, that
further testing in a randomised controlled fashion
between the Pixietm and competitor oxygenators should
be undertaken in a clinical, in vivo setting to establish
relative functionality.
Funding
This research received no specific grant from any funding
agency in the public, commercial or not-for-profit sectors.
Disclosures
The authors had full control of the methods used, analysis of
data and production of the written report.
1. Elliott M, Hamilton J, Clark I. Review article : Perfusion
for paediatric open-heart surgery. Perfusion 1990; 5: 1–8.
2. Wendel HP, Scheule AM, Eckstein FS, Ziemer G.
Haemocompatibility of paediatric membrane oxygenators
with heparin-coated surfaces. Perfusion 1999; 14: 21–28.
3. Boettcher W, Merkle F, Koster A, et al., Safe minimization of
cardiopulmonary bypass circuit volume for complex cardiac
surgery in a 3.7 kg neonate. Perfusion 2003; 18: 377–379.
4. Paparella D, Yau TM, Young E. Cardiopulmonary bypass
induced inflammation: pathophysiology and treatment.
An update. Eur J Cardio-Thorac 2002; 21: 232–244.
5. Naik S and Elliott M. Ultrafiltration and paediatric car-
diopulmonary bypass. Perfusion 1993; 8: 101–112.
6. Dogal N, Mathis RK, Lin J, et al. Evaluation of three hol-
low-fiber membrane oxygenators without integrated arte-
rial filters for neonatal cardiopulmonary bypass. Perfusion
2012; 27: 132–140.
7. Fujii Y, Kotani Y, Kawabata T, et al. The benefits of high-
flow management in children with pulmonary atresia.
Artif Organs 2009; 33: 888–895.
8. Sakamoto T, Kurosawa H, Shin’oka T, Aoki M, Isomatsu
Y. The influence of pH strategy on cerebral and collat-
eral circulation during hypothermic cardiopulmonary
bypass in cyanotic patients with heart disease: results of
a randomized trial and real-time monitoring. J Thorac
Cardiovasc Surg 2004; 127: 12–19.
9. De Somer, F. Impact of oxygenator characteristics on
its capability to remove gaseous microemboli. J Extra
Corpor Technol 2007; 39: 271–273.
10. De Somer, FM, Vetrano MR, Van Beeck JP, Van Nooten
GJ. Extracorporeal bubbles: a word of caution. Interact
Cardiovasc Thorac Surg 2010; 10: 995–1001.
11. Bull JL. Cardiovascular bubble dynamics. Crit Rev
Biomed Eng 2005; 33: 299–346.

Downloaded from prf.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016