bs_bs_banner
© 2013, Copyright the Authors
Artificial Organs © 2013, International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.
Pediatric Single-Lumen Cannula Venovenous
Extracorporeal Membrane Oxygenation:
A French Center Experience
*†Pierre-Louis Léger, *Julia Guilbert, *†Ségolène Isambert, *†Nolwenn Le Saché,
*Fazia Hallalel, *Alain Amblard, *Jean-Yves Chevalier, and *†Sylvain Renolleau
*Service de Réanimation Néonatale et Pédiatrique, Hôpital Armand-Trousseau; and †Université Pierre et Marie
Curie-UPMC, Paris, France
Abstract: Single-lumen cannula venovenous (VV) extra-
corporeal membrane oxygenation (ECMO) is a special
extracorporeal life support (ECLS) technique used for
neonatal and pediatric refractory hypoxemia. This is an
alternative flow rate ECLS that consists of successive
clamping on the drainage and the injection lines. Currently,
the Armand-Trousseau’s pediatric intensive care unit
remains the only pediatric ECMO center proposing this
partial assistance. This article details a technical note and a
retrospective analysis of our experience in refractory
hypoxemia. The retrospective study, from 2007 to 2011,
included all pediatric and neonatal patients treated by
single-lumen cannula VV ECMO. The study was focused
on pre-ECMO patient characteristics and complications
during ECMO course. During the last 5 years, 67 pediatric
patients were assisted by this single-lumen cannula VV
ECMO. Sixty-one patients (91%) were newborns. Thirty-
nine patients presented with meconium aspiration syn-
drome (58%), which was the most frequent etiology.
Before cannulation, mean oxygenation index (OI) was
32 ⫾ 11, alveolar-arterial oxygen difference was 604 ⫾
47 mm Hg, and partial pressure arterial oxygen/fraction
inspired oxygen ratio was 59.2 ⫾ 35.8. Forty-eight patients
The extracorporeal membrane oxygenation
(ECMO) support is a lung assistance used in refrac-
tory hypoxemia and acute respiratory distress
doi:10.1111/aor.12024
Received July 2012; revised September 2012.
Address correspondence and reprint requests to Dr. Pierre-
Louis Léger, Service de Réanimation Néonatale et Pédiatrique,
Hôpital Armand-Trousseau, 28 Avenue du Docteur Arnold Netter,
75012 Paris, France. E-mail: pierre-louis.leger@trs.aphp.fr
Presented in part at the 8th International Conference on Pedi-
atric Mechanical Circulatory Support Systems and Pediatric
Cardiopulmonary Perfusion held June 13–16, 2012 in Istanbul,
Turkey.
(72%) presented pulmonary hypertension, and 66 patients
were treated by nitric oxide (98%). Fifty patients (75%)
were treated by vasopressors or inotropic drugs. Average
duration of ECMO was 13.2 ⫾ 7.8 days. There were forty-
six survivors (69%). The worst prognosis was for respira-
tory syncytial virus pneumonia. Complications like acute
renal injury and hematologic and transfusion acts were not
so different than those observed in classical ECMO tech-
niques. Nevertheless, 19 patients presented a stroke (28%
of the overall population), but this high rate did not seem to
be due to the ECLS technique used. Single-lumen cannula
VV ECMO is a partial and efficient ECMO support. Our
experience shows that this technique is as efficient and less
invasive than two cannulas ECMO. The single-lumen
cannula VV ECMO is a simple and safe ECLS support
used for neonatal or pediatric refractory hypoxemia.
Because this is a partial assistance, it is a promising
ECLS support. Key Words: Single-lumen cannula veno-
venous extracorporeal membrane oxygenation—Neonatal
refractory hypoxemia—Pediatric acute respiratory
distress syndrome—Extracorporeal life support—Partial
assistance.
syndrome (ARDS). The first description was made in
1974 (1). ECMO improved survival of neonatal
refractory hypoxemia (2–4). More than 20 years ago,
the French pediatric ECMO center, Armand-
Trousseau, has been using all ECMO techniques,
conventional continuous venoarterial (VA) or veno-
venous (VV) ECMO, to treat pediatric and neonatal
refractory hypoxemia. An alternative single-lumen
cannula VV ECMO was developed in France and was
named Assistance Respiratoire Extra-Corporelle
(AREC). Actually, our pediatric intensive care unit
(PICU) currently uses this very special partial and
long-term assistance.
Artificial Organs 2013, 37(1):57–65
58 P.-L. LÉGER ET AL.
In this review, we will describe the specificities and
the main physiological aspects of this technique. We
will expose our selection and exclusion criteria for
pediatric or neonatal refractory hypoxemia. After-
ward, we will also present the results of a retrospec-
tive study about patients treated by single-lumen
cannula VV ECMO in the last 5 years.
PARTIAL LUNG SUPPORT CONCEPT
The idea of partial lung support followed studies
about ARDS and the concept of baby lung syndrome
(5). In order to protect the lungs from barotrauma
and volotrauma (ventilator-induced lung injury)
(6,7), mechanical ventilation is achieved with low
tidal volume, plateau pressure below 30 cm H2O, and
low respiratory rate (8–11). Nevertheless, these
parameters induce hypercapnia that can be tolerated
when pH is above 7.25 (11). Thus, Gattinoni et al.
imagined to separate the oxygenation and CO2
removal functions of the lungs (12,13). Oxygenation
was achieved via natural lungs with the ventilator and
Artif Organs, Vol. 37, No. 1, 2013
an additional flow of oxygen via a tracheal catheter
and CO2 removal with a low-flow extracorporeal
support (14).
In addition, in neonates and infants, pulmonary
hypertension is frequently associated with neonatal
and pediatric respiratory failure. VV ECMO that
increases mixed PvO2 participate to pulmonary
vasodilation.
CIRCUIT DESIGN
Cannula
The cannula is introduced by surgical procedure
into the right jugular vein and is located into the right
atrium, controlled by chest X-ray. The jugular vein is
ligated upstream. The cannula size depends on the
patient’s weight. Usually, we use a 12Fr cannula for
newborns (Venous ECMO catheter, Maquet, Hirrlin-
gen, Germany).The cannula is connected via a Y-type
connector to silicone tubing: one tubing corresponds
to the drainage line and one tubing corresponds to
the injection line (Fig. 1).
FIG. 1. Single-lumen cannula VV ECMO
circuit. The photos represent the different
elements on the ECMO circuit: the alter-
native clamp (in the upper left), the single-
lumen cannula (in the upper right), the
roller pump with tubing in the pump head
and the oxygenator (at the bottom). The
single-lumen cannula is inserted into the
right internal jugular vein and is connected
to alternative clamps (for drainage and
injection) and to a roller pumps that deliv-
ers venous blood to a membrane lung and
returns it through the cannula to the right
atrium.
 SINGLE-LUMEN CANNULA VV ECMO 59

Tubing
Circuit size available: newborn circuit or infants
(AREC Nouveau-né TK 0034 or AREC Enfant TK
0035, Sofra Medical, Vienne, France). Circuits are sili-
cone tubing.

Nonocclusive roller pump

The nonocclusive roller pump is the original
element of the single-lumen cannula VV ECMO
(A100 pump, Sofra Medical). Pump and tubing in the
pump head specificities are the main elements allow-
ing alternative flow rate in the circuit (see technical
section of this article). The tubing on the pump is
changed every 2 days to avoid tubing rupture.

Clamp
Time-controlled alternative clamps are located on
the drainage and on the injection lines (Alternating
clamp, EFS, Montagny, France).

Other elements of the circuit

The other elements are the membrane oxygenator
(Hilite 800 LT or Hilite 2400LT, Medos, Stolberg,
Germany), the heat generator (Biocal 370,
Medtronic, Minneapolis, MN, USA), the oxygen
blender (Air Oxygen blender, Sechrist, Anaheim,
CA, USA), the capnograph connected to gas effluent
port of the membrane (Normocap Oxy, Datex-
Ohmeda, Helsinki, Finland), and the flow meter on
the drainage line (HT 110, Transonic System, New-
York, NY, USA). For example, median flow rate for
newborns is 200–300 mL/min, and around 600–
800 mL/min for infants.
TECHNIQUE
This special extracorporeal life support technique
(ECLS) was described in the early 1990s (15–17). The
single-lumen venous cannula characterizes this alter-
native and partial ECMO support. In fact, because of
the single-lumen cannula, the drainage and the injec-
tion are not concomitant but successive (Fig. 2). The
time-controlled clamps generate an alternative flow
rate in the cannula due to the successive clamping on
the drainage and the injection line. So the direction of
the blood flow into the cannula differs between these
two phases. First, the clamp is open on the drainage
line and closed on the injection line. The cannula
drains the blood from the right atrium to the pump.
The blood accumulates in the tubing of the pump
head. During the injection phase, the clamp on the
drainage line is closed when the clamp on the injec-
tion line is open. The blood in the tubing of the pump
head fills the oxygenator membrane and returns to
FIG. 2. The alternative clamp cycle. (a and b) The drainage line
is open, the infusion line is closed; pressure rises downstream
from the pump and blood pools in tubing until clamp is activated.
(c and d) The drainage line is closed, the infusion line is open;
pressure decreases downstream from the pump and blood is
injected until clamp is activated. I, injection line; D, drainage line;
C, closed; O, open.
From: Chevalier et al. (16). Used with permission.
the patient. The blood direction in the cannula is
reversed. This is a partial extracorporeal support
because the maximum flow rate is limited by tubing
distensibility and clamping time (Fig. 3). The clamp
spends 2 s on the drainage line and 1 s on the injec-
tion line.
The recirculation problems are less frequent
during the single-lumen cannula VV ECMO than in
continuous VV ECMO support using occlusive or
centrifugal pumps. The alternative clamping limits
this problem. For example, the theoretical recircula-
tion for a 12Fr cannula was measured at 18 mL/min
corresponding to 4–7 mL/kg/min flow rate for new-
borns. The theoretical neonatal cardiac output would
be estimated to 200 mL/kg/min. The extracorporeal
flow rate support is usually 70 mL/kg/min. So the

Artif Organs, Vol. 37, No. 1, 2013

60 P.-L. LÉGER ET AL.
FIG. 3. Diagram of the pump head tubing during the cycle. (a and
b) Drainage phase with filing of the tubing into the pump head,
drainage-line clamp is open and return-line clamp is closed, pres-
sure rises after the membrane. (c) Injection phase with emptying
of the tubing into the pump head, return-line clamp is open and
drainage-line clamp is closed, pressure falls after the membrane,
and returns to position 1.
From: Chevalier et al. (16). Used with permission.
maximum recirculation flow rate is only about 10%
(Y. Durandy, unpublished data).
We have shown that the flow rate is limited in
comparison with continuous VV ECMO. The sickest
patients who need total ECMO support (either for
oxygenation or hemodynamic failure) need to be put
on continuous VV or VA ECMO support. The non-
occlusive roller pump allows an alternative flow rate
in the circuit because the blood volume can change
into the tubing of the pump head. This tubing is a
silicone distensible tube. When there is no efficient
drainage (e.g., when right preload is too low), the flow
rate in the circuit is zero and the tube is flat, and it
does not change with the speed of the rotor. When
the drainage is efficient, the flow rate depends on
several factors: right preload, speed of the rotor
pump, and height between the baby and the pump
because preload of the pump also depends on gravity.
So during the drainage phase, the tubing in the pump
head is full and becomes ellipsoidal. If the pressure
after the pump is too high, for example, when there is
a membrane oxygenator clotting, the tubing in the
pump head is too full and becomes round. This is due
to the distension of the silicone tubing in the pump
head and also because the pump is nonocclusive. The
flow rate in the circuit is zero, but the rotor of the
pump does not stop.
So, the flow rate is self-regulated for a constant
speed of the rotor. The outflow mainly depends on
preload and afterload of the pump. This specific non-
occlusive pump avoids the suction phenomena
during the drainage when the preload is low. The
pump also avoids the disruption on the return line
when the afterload is high.
The systemic hemodynamic is not significantly
affected by the alternating clamp. Pressure measures
before and after the membrane oxygenator show
Artif Organs, Vol. 37, No. 1, 2013
variations between drainage and injection phases.
Mean pressure in the circuit oscillates between 50
and 250 mm Hg. Central venous pressure varies from
3 to 6 mm Hg, but mean arterial pressure remains
constant (18). Moreover, invasive pulse pressure and
Doppler evaluation of aortic blood flow velocity do
not vary during the drainage or injection phase (data
not shown).
The single-lumen cannula VV ECMO is a very
special lung support adapted for ARDS, refractory
hypoxemia, and persistent pulmonary hypertension.
The benefits of a single cannula are evident. First, one
vascular access is less invasive than both jugular and
femoral cannulas for continuous VV ECMO. The
femoral cannula implies upstream ligation, increasing
the risk of thrombosis or inferior member growth
inequality. Second, switch to continuous ECMO is
always possible. In our experience, it occurred in 30%
patients (data not shown).
MANAGEMENT OF ECMO PATIENTS
“Apneic” ventilation
Most of the time, oxygenation can be achieved via
the natural lungs despite severe injury. Apneic venti-
lation is a nonconventional mode of mechanical
ventilation (19,20) that associates a lung protection
strategy with low frequency and low pressure
support and oxygenation improvement. Oxygenation
is achieved partly by the ventilator and by a continu-
ous flow of oxygen delivered through a tracheal cath-
eter. The catheter is introduced in the tracheal tube
and placed 1 cm above the distal extremity. This posi-
tion reduces potential tracheal injury. A high positive
end-expiratory pressure (PEEP) allows alveolar
recruitment and permits an easy passive diffusion of
the oxygen. The low frequency ventilation prevents
alveolar collapse. For example, in newborns,
the respiratory frequency varies between 10 and 15
cycles/min, inspiratory pressure must be under 30
cm H2O, tidal volume is below 7 mL/kg, and PEEP
varies between 6 and 12 cm H2O. The continuous
humidified oxygen flow rate is delivered at 2 L/min/
m2. As in continuous VV ECMO, the improvement of
arterial oxygenation depends on the natural lungs, as
opposed to VA ECMO that bypasses the natural
lungs.
Anticoagulation and transfusions
First, a bolus of 50 IU/kg heparin is injected before
the surgical cannulation if the activated clotting time
(ACT) is above 180 s. The priming is composed for
100 mL of priming volume by 100 IU heparin, 1.3 mL
calcium chloride 10%, and 4 mL sodium bicarbonate
 SINGLE-LUMEN CANNULA VV ECMO 61

4.2%. Afterward, continuous heparin is infused at a was 39.3 ⫾ 1.8, and the median weight was 3.4 ⫾
dose between 20 and 40 IU/kg/h. ACT is maintained 0.5 kg. For infants, the median age was 6.1 ⫾ 4.7
between 180 and 200 s. The platelet count is main- months and weight was 5.6 ⫾ 2.4 kg (Table 1a).
tained above 60 000/mm3 and hemoglobin above The mean number of pre-ECMO ventilation days
10 g/dL. was 3.5 ⫾ 3.9 days. The FIO2 was 100% for all
patients. The mean PEEP was 5.5 ⫾ 1.7 cm H2O, OI
SELECTION AND EXCLUSION was 32 ⫾ 11, and AaDO2 was 604 ⫾ 47. The mean
CIRCULATORY SUPPORT CRITERIA PaO2/FIO2 ratio was 59.2 ⫾ 35.8. Nevertheless, we
Our selection criteria are the same that are used in observed some differences between newborns and
other centers (21). We consider a rapid or progressive pediatric patients. For infants, the mean PEEP and
respiratory deterioration while undergoing optimal PaO2/FIO2 ratio were worse than in newborns with
conventional treatments: maximal mechanical
ventilation with pure oxygen, severe hypoxemia TABLE 1a. Patient characteristics before and during
criteria like partial pressure arterial oxygen (PaO2) single-lumen cannula VV ECMO
<40 mm Hg more than 6 h, alveolar-arterial oxygen
difference (AaDO2) >620 more than 8 h, or oxygen- Mean ⫾ SD (min–max)
or percentage of patients [n patients]
ation index (OI) >40 more than 6 h, correction of
hemodynamic disturbances with vascular filings, Newborns
Term (GA) 39.3 ⫾ 1.8 (35–42) [61]
vasopressive or inotropic drugs, optimal sedation, Birth weight (kg) 3.4 ⫾ 0.5 (2.5–4.8) [61]
and paralysis drugs if necessary, and nitric oxide Pediatric patients
(NO) if pulmonary hypertension. Age (months) 6.2 ⫾ 4.7 (2–12.9) [6]
Weight (kg) 5.6 ⫾ 2.4 (3.3–8.6) [6]
The contraindications are mainly weight less than Sex
2 kg, gestational age (GA) <35 weeks, prolonged Male 56 [39]
mechanical ventilation, congenital cardiopathy, Female 44 [28]
Etiology
severe chromosomic or neurologic disabilities, intra- MAS 58 [39]
cranial hemorrhage, coagulopathy, or uncontrolled PPHT 18 [12]
bleeding. Nevertheless, assistance support is dis- CDH 11 [7]
RSV 9 [6]
cussed between intensive doctors in each case. RDS 4 [3]
Ventilation parameters
WEANING CRITERIA FIO2 (%) 100 ⫾ 0 [65]
PEEP (cm H2O) 5.5 ⫾ 1.7 (2–12) [43]
Weaning off begins when gas exchanges show nor- OI 32 ⫾ 11 (16–63) [15]
AaDO2 (mm Hg) 604 ⫾ 47 (438–640) [13]
mocapnia and normoxia with minimal extracorporeal PaO2/FIO2 59.2 ⫾ 35.8 (34–220) [33]
support. In summary, our weaning objective targets Gasometric parameters
fraction-inspired oxygen (FIO2) less than 40% on PaO2 (mm Hg) 54.8 ⫾ 23.9 (18–128) [34]
PaCO2 (mm Hg) 52.6 ⫾ 22.2 (14–111) [41]
apneic ventilation and tracheal catheter, and FIO2 pH 7.2 ⫾ 0.2 (6.9–7.5) [40]
less than 60% on membrane oxygenator. Extracorpo- Pulmonary hypertension
real blood flow on the circuit is reduced until 20% of Systemic PHT 61 [41]
Nonsystemic PHT 10 [7]
theoretical global cardiac output. No PHT 1.5 [1]
Unknown 26 [18]
STUDY DESIGN Pulmonary vasodilatators
NO 98 [66]
In this retrospective study, we have analyzed 67 Epoprostenol 18 [12]
patients treated by single-lumen cannula VV ECMO Sildenafil 1.5 [1]
Hemodynamic drugs
technique from January 2007 and December 2011. Norepinephrine 47 [32]
We have collected patients’ characteristics before Dopamine 30 [20]
cannulation, extracorporeal support durations, sur- Dobutamine 13 [9]
Milrinone 3 [2]
vival, and complications that occurred during their No support 25 [17]
intensive care unit stay.
Sixty-seven patients were included in this study. The number of
patients (n) was used to calculate values or percentages for each
RESULTS category. The denominator of percentages for sex, etiology groups,
pulmonary hypertension, pulmonary vasodilatators, and hemo-
Patient characteristics before ECLS dynamic drugs was 67.
PPHT, persistent pulmonary hypertension; CDH, congenital
Among the 67 patients, 61 were newborns (91%) diaphragmatic hernia; RSV, respiratory syncytial virus; RDS,
and 6 were infants. For newborns, the median GA respiratory distress syndrome; PHT, pulmonary hypertension.

Artif Organs, Vol. 37, No. 1, 2013

62 P.-L. LÉGER ET AL.

7.2 ⫾ 3.4 and 48.8 ⫾ 5.8 cm H2O, respectively. The OI TABLE 2. Survival and causes of mortality during
and AaDO2 were not different (data not shown). single-lumen cannula VV ECMO
Concerning the gasometric parameters, PaO2 was Percentage of patients
54.8 ⫾ 23.9 mm Hg, PaCO2 was 52.6 ⫾ 22.2, and pH [n patients/total patients]
was 7.2 ⫾ 0.2. Survival
Pulmonary hypertension is a frequent consequence Global 69 [46/67]
of severe neonatal pulmonary diseases. Forty-eight Newborns 73 [45/61]
Infants 17 [1/6]
patients (71%) presented pulmonary hypertension Survival by subgroups
before cannulation. Sixty-six patients were treated by MAS 79 [31/39]
NO (98%) even if pulmonary hypertension was not PPHT 58 [7/12]
CDH 57 [4/7]
proved. Other pulmonary hypertension treatments at RSV 17 [1/6]
the early phase of the respiratory disease were less RDS 100 [3/3]
common (epoprostenol or sildenafil). Mortality causes
Refractory hypoxemia 33 [7/21]
The systemic hemodynamics were altered before Stroke or brain death 23 [5/21]
ECLS because of pulmonary hypertension or associ- Cardiac arrest 15 [3/21]
ated sepsis; norepinephrine was the first-line Surfactant protein deficit 15 [3/21]
Hemorrhagic shock 9 [2/21]
treatment. Thirty two patients (47%) received nore- Capillar alveolar dysplasia 5 [1/21]
pinephrine before cannulation. Twenty patients
(30%) were treated with dopamine. Dobutamine was The number of patients (n) was used to calculate values or
percentages for each category. The number of total patients was
the first choice of inotropic drug in nine patients the denominator for each category.
(13%) and after milrinone in two patients (3%). Only PPHT, persistent pulmonary hypertension; CDH, congenital
17 patients (25%) had no hemodynamic support. diaphragmatic hernia; RSV, respiratory syncytial virus pneumonia;
RDS, respiratory distress syndrome.

Patient characteristics during ECLS

Concerning the 46 survivors, average duration of
extracorporeal life support was 13.2 ⫾ 7.8 days, with
a large range between 4 and 41 days. Number of
ventilation days was 26.2 ⫾ 13.3 (range between 8
and 69 days), and number of ICU days was 29.6 ⫾ 16
(range between 8 and 74 days) (Table 1b). We
observed a great standard deviation value because of
heterogeneous evolution in respiratory distress syn-
drome, both in neonatal and pediatric patients.
Survival and causes of mortality
The overall survival was 69% (46 survivors out of
67 patients). The survival was very different between
newborns (73%) and infants (17%). Only six infants
were treated with this technique (Table 2). Most
infants were assisted by continuous VV or VA
ECMO (data not shown).
The main etiology was meconium aspiration syn-
drome (MAS) with a survival rate of 79% (31 of 39
TABLE 1b. Main durations in survivors for single-lumen
cannula VV ECMO
Mean ⫾ SD (min–max)
[n patients]
Duration of mechanical ventilation 26.2 ⫾ 13.3 (8–69) [46]
Duration of ECLS 13.2 ⫾ 7.8 (4–41) [46]
Duration of ICU stay 29.6 ⫾ 16 (8–74) [46]
The number of patients (n) was used to calculate values for each
category. Durations are expressed in days.
patients). Congenital diaphragmatic hernia (57%)
and persistent pulmonary hypertension (58%) had
lower survival rate than MAS. The highest survival
rate concerned respiratory distress syndrome (100%
survived). The lowest survival concerned respiratory
syncytial virus pneumonia with only one surviving
patient (17%).
Complications
Here, we report the hematologic, renal, neurologic,
and infectious complications, and transfusion acts in
the studied population (Table 3).
Concerning hematologic injury, severe bleeding
and severe thrombosis occurred in respectively 8 and
11 patients (12 and 16% of the overall population).
Platelet transfusions were the most frequent transfu-
sion act. The mean platelets transfusion for each
patient was 9.5 ⫾ 10.2 units, with large range between
0 and 60 transfusions. The mean red blood cell trans-
fusions were 2.9 ⫾ 2.6. Fresh frozen plasma was very
uncommon.
In single-lumen cannula VV ECMO, the tubing of
the pump head changed every 2 days to avoid
rupture. With this protocol, rupture was rare. So, the
mean tubing change was 6.8 ⫾ 4.7 times during the
course of ECMO. The mean membrane oxygenator
change was 0.5 ⫾ 0.7 times because we used long-
time membrane oxygenator duration.
Infections were frequent, especially septicemias
for 30 patients (44% of overall population) and

Artif Organs, Vol. 37, No. 1, 2013

 SINGLE-LUMEN CANNULA VV ECMO 63

TABLE 3. Complications during single-lumen cannula VA ECMO. First, there is only one site of cannulation
VV ECMO unlike double cannula VV ECMO and VA ECMO. In
Mean ⫾ SD (min–max) double cannula VV ECMO, the cannulas are placed
or percentage of patients into the jugular vein and the femoral vein or in both
[n patients] femoral veins. VA ECMO also requires the insertion
Hematologic injury of two cannulas and one of them brings to carotid
Severe bleeding 12 [8] artery ligation. Second, this technique has less recir-
Severe thrombosis 16 [11]
Transfusion act culation problems (18) compared with double
Platelets 9.5 ⫾ 10.2 (0–60) [67] cannula VV ECMO or double lumen single cannula
Red blood cell 2.9 ⫾ 2.6 (0–11) [67] VV ECMO. Third, there is less risks of circuit com-
Fresh frozen plasma 0.4 ⫾ 0.8 (0–3) [67]
Membrane oxygenator change 0.5 ⫾ 0.7 (0–7) [67] plications with a low-flow assistance than with high-
Tubing pump head change 6.8 ⫾ 4.7 (0–22) [67] flow VV or VA ECMO. Finally, a switch to continuous
Infections ECMO is possible at any moment if lung assistance
Septicemia 44 [30]
VAP 27 [18] seems to be insufficient. Currently, the main limita-
Renal injury tion is addressed to infants and children because we
Acute renal injury 60 [40] no longer have large circuits. This is why we use con-
Hemofiltration 22 [9]
Neurologic injury tinuous ECMO for these patients. In the future, we
Left hemispheric stroke 21 [14] hope to dispose again larger circuits for this pediatric
Right hemispheric stroke 7 [5] population. VV ECMO with a double lumen cannula
Sixty-seven patients were included in this study. The number of such as Avalon is a possible alternative to our tech-
patients (n) was used to calculate values or percentages for each nique. This cannula is depicted to have less recircula-
category. For hematologic injury, infections, renal injury, and neu- tion than other double lumen cannulas. Nowadays,
rologic injury, the mean value represents the number of suffering
patients. The denominator for these groups is 67. Percentage of we do not have enough experience with its use.
hemofiltration was calculated from the 40 patients presenting Nevertheless, apart from its cost, the switch from VV
acute renal injury. Concerning transfusion acts, oxygenator change, ECMO to VA ECMO with this cannula is difficult.
and tubing pump head change, the mean value represents the
mean transfusion or change for each patient. The lumen dedicated to reinjection in its normal use
VAP, ventilation-associated pneumonia. is not available for drainage.
In our 5-year retrospective study, 67 patients have
been assisted by single-lumen cannula VV ECMO
ventilation-associated pneumonia for 18 patients
technique. Newborns represented the large majority
(27%). The most frequent bacterial agent was
of the patients (91%). Some differences can be
coagulase-negative staphylococci (septicemias) and
observed between the inclusion criteria values and
Pseudomonas aeruginosa (ventilation-associated
the patient’s values (OI, AaDO2, PaO2). This could be
pneumonias).
explained by lacking data, particularly for the most
Forty patients (60% of the population) had at
unstable patients who were immediately put on
least biological or clinical signs of acute renal failure
ECLS without arterial blood gas data.
during single-lumen cannula VV ECMO. Continu-
Survival in newborns was similar to ELSO registry
ous furosemide infusion in association with fluid
data. The survival of MAS was 79% and congenital
restriction was sufficient most of the time. Neverthe-
diaphragmatic hernia was 57% (24). During the last 5
less, nine patients (22%) needed hemofiltration
years, only six young infants (mean age 6.2 months)
during ECLS.
were treated by single-lumen cannula VV ECMO
Neurologic injuries were mainly represented by
with a survival rate of only 17%. The ELSO registry
strokes. Ischemic strokes occurred in 19 patients
points out a pediatric survival of 57% in the age
which represented 28% of the population. Seventy-
group between 30 days and 1 year old and 70% for
three percent of strokes occurred in the left hemi-
respiratory syncytial virus pneumonia (25).
sphere. The territory of the middle cerebral artery
Some elements could explain this result. First, it is
represented the main localization (data not shown).
difficult to conclude about single-lumen cannula VV
ECMO technique in pediatric patients because of our
DISCUSSION
limited cohort. Unfortunately, larger circuits for chil-
Single-lumen cannula VV ECMO is a specific alter- dren were not available. During the same period, 12
native flow rate ECLS technique used for 20 years of the 26 pediatric patients (46%) with continuous
with success in French pediatric reference ECMO ECMO survived (mean age 31 ⫾ 45 months, data not
centers (22,23). This technique presents several shown). These results are not that different than
advantages compared with either continuous VV or those of the ELSO registry and Flamant et al.’s study

Artif Organs, Vol. 37, No. 1, 2013

64 P.-L. LÉGER ET AL.
on bronchiolitis (25,26). Second, a lot of these
patients were transferred from other PICUs in poor
shape with refractory hypoxemia, shock, and multi-
visceral organ failure. In older children, our tech-
nique of partial pulmonary support should probably
be only used in case of isolated pulmonary failure.
That also points out the importance of the timing of
the transfer in an ECMO center.
Concerning the complications during single-lumen
cannula VV ECMO, strokes reveal a special impor-
tance.We have shown that 19 strokes occurred during
the ECLS, which represented 28% of the study popu-
lation. The ELSO registry related 10–15% cases
of strokes in VV ECMO (24,27). Actually, 11 of the
19 stroke patients presented with MAS, which
was associated with perinatal asphyxia and anoxo-
ischemic encephalopathy. We can suppose that the
association between anoxo-ischemic encephalopathy
and cerebral circulation disturbance caused by the
jugular vein cannulation could increase the risk of
stroke. Moreover, hemodynamic status plays a funda-
mental role in this situation. In a recent retrospective
study about neurologic complications in 80 patients
treated by ECLS, a statistical correlation was shown
between the duration of hemodynamic support
before ECLS and neurologic complications (data not
published yet). In our study, systemic hemodynamics
support before or during the first days after cannula-
tion was present in all patients with stroke, but only
85% in patients without stroke (data not shown).This
element points out the importance of hemodynamic
status before and within the first days of ECLS and
should encourage an earlier transfer to an ECMO
center when shock occurs during refractory hypox-
emia. Finally, we observed that 74% of strokes were
located in the left hemisphere. The venous cannula
and the technique do not entirely explain this fact.
CONCLUSION
The single-lumen cannula venovenous extracorpo-
real membrane oxygenation is an efficient partial
extracorporeal life support system. For 20 years, this
technique has been used for neonatal refractory
hypoxemia and for pediatric ARDS at the Armand-
Trousseau PICU. This specific extracorporeal support
with a single-lumen cannula, successive clamping, and
alternative flow rate represents another solution for
pediatric ECMO. It is less invasive than a double
cannula ECLS and with lower recirculation. Hemo-
dynamic tolerance is not different than in continuous
ECMO technique. Our retrospective study shows
similar survival to other ECLS techniques. Neuro-
logic complications like strokes were not due to
Artif Organs, Vol. 37, No. 1, 2013
specificities of the technique but rather linked to eti-
ology, neonatal birth conditions, and hemodynamic
support needs. While adult ECMO centers develop
partial ECLS such as pumpless extracorporeal lung
assist (28), we hope that our special extracorporeal
partial assistance will be developed in other ECMO
centers throughout the world.
REFERENCES
1. Bartlett RH, Fong SW, Burns NE, Gazzaniga AB. Prolonged
partial venoarterial bypass: physiologic, biochemical, and
hematologic responses. Ann Surg 1974;180:850–6.
2. UK collaborative randomised trial of neonatal extracorporeal
membrane oxygenation. UK Collaborative ECMO Trial
Group. Lancet 1996;348:75–82.
3. Bartlett RH, Gazzaniga AB, Wetmore NE, Rucker R,
Huxtable RF. Extracorporeal membrane oxygenation
(ECMO) in the treatment of cardiac and respiratory failure
in children. Trans Am Soc Artif Intern Organs 1980;26:578–
81.
4. Bartlett RH, Roloff DW, Custer JR, Younger JG, Hirschl RB.
Extracorporeal life support: the University of Michigan expe-
rience. JAMA 2000;283:904–8.
5. Gattinoni L, Pesenti A. The concept of “baby lung.” Intensive
Care Med 2005;31:776–84.
6. Brochard L, Roudot-Thoraval F, Roupie E, et al. Tidal volume
reduction for prevention of ventilator-induced lung injury in
acute respiratory distress syndrome. The Multicenter Trial
Group on Tidal Volume Reduction in ARDS. Am J Respir Crit
Care Med 1998;158:1831–8.
7. Dreyfuss D, Saumon G. Ventilator-induced lung injury:
lessons from experimental studies. Am J Respir Crit Care Med
1998;157:294–323.
8. Burns KE, Adhikari NK, Slutsky AS, et al. Pressure and
volume limited ventilation for the ventilatory management of
patients with acute lung injury: a systematic review and meta-
analysis. PLoS One 2011;6:e14623.
9. Ney L, Kuebler WM. Ventilation with lower tidal volumes as
compared with traditional tidal volumes for acute lung injury.
N Engl J Med 2000;343:812–3; author reply 813–4.
10. Putensen C, Theuerkauf N, Zinserling J, Wrigge H, Pelosi P.
Meta-analysis: ventilation strategies and outcomes of the acute
respiratory distress syndrome and acute lung injury. Ann
Intern Med 2009;151:566–76.
11. Randolph AG. Management of acute lung injury and acute
respiratory distress syndrome in children. Crit Care Med
2009;37:2448–54.
12. Gattinoni L, Kolobow T, Damia G, Agostoni A, Pesenti A.
Extracorporeal carbon dioxide removal (ECCO2R): a new
form of respiratory assistance. Int J Artif Organs 1979;2:183–
5.
13. Gattinoni L, Kolobow T, Tomlinson T, et al. Low-frequency
positive pressure ventilation with extracorporeal carbon
dioxide removal (LFPPV-ECCO2R): an experimental study.
Anesth Analg 1978;57:470–7.
14. Gattinoni L, Agostoni A, Pesenti A, et al. Treatment of acute
respiratory failure with low-frequency positive-pressure venti-
lation and extracorporeal removal of CO2. Lancet 1980;2:
292–4.
15. Chevalier JY. Extracorporeal respiratory assistance for pedi-
atric acute respiratory failure. Crit Care Med 1993;21:S382–3.
16. Chevalier JY, Couprie C, Larroquet M, Renolleau S, Durandy
Y, Costil J. Venovenous single lumen cannula extracorporeal
lung support in neonates. A five year experience. ASAIO J
1993;39:M654–8.
17. Chevalier JY, Durandy Y, Batisse A, Mathe JC, Costil J. Pre-
liminary report: extracorporeal lung support for neonatal
acute respiratory failure. Lancet 1990;335:1364–6.
 SINGLE-LUMEN CANNULA VV ECMO
18. Durandy Y, Chevalier JY, Lecompte Y. Single-cannula veno-
venous bypass for respiratory membrane lung support.
J Thorac Cardiovasc Surg 1990;99:404–9.
19. Draper WB, Whitehead RW, Spencer JN. Studies on diffusion
respiration: alveolar gases and venous blood pH of dogs during
diffusion respiration. Anesthesiology 1947;8:524–33.
20. Gattinoni L, Pesenti A, Kolobow T, Damia G. A new look at
therapy of the adult respiratory distress syndrome: motionless
lungs. Int Anesthesiol Clin 1983;21:97–117.
21. Mugford M, Elbourne D, Field D. Extracorporeal membrane
oxygenation for severe respiratory failure in newborn infants.
Cochrane Database Syst Rev 2008;(3):CD001340.
22. Flamant C, Lorino E, Nolent P, et al. Newborn infants sup-
ported by extracorporeal membrane oxygenation: survival and
clinical outcome. Arch Pediatr 2007;14:354–61.
23. Flamant C, Nolent P, Hallalel F, Lardeux C, Chevalier JY,
Renolleau S. Evolution of extracorporeal membrane oxygen-
ation (ECMO) in neonatal acute respiratory failure, fifteen
years of experience. Arch Pediatr 2004;11:308–14.
65
24. Bahrami KR, Van Meurs KP. ECMO for neonatal respiratory
failure. Semin Perinatol 2005;29:15–23.
25. Zabrocki LA, Brogan TV, Statler KD, Poss WB, Rollins MD,
Bratton SL. Extracorporeal membrane oxygenation for pedi-
atric respiratory failure: survival and predictors of mortality.
Crit Care Med 2011;39:364–70.
26. Flamant C, Hallalel F, Nolent P, Chevalier JY, Renolleau S.
Severe respiratory syncytial virus bronchiolitis in children:
from short mechanical ventilation to extracorporeal mem-
brane oxygenation. Eur J Pediatr 2005;164:93–8.
27. Zahraa JN, Moler FW, Annich GM, Maxvold NJ, Bartlett RH,
Custer JR. Venovenous versus venoarterial extracorporeal life
support for pediatric respiratory failure: are there differences
in survival and acute complications? Crit Care Med 2000;28:
521–5.
28. MacLaren G, Combes A, Bartlett RH. Contemporary extra-
corporeal membrane oxygenation for adult respiratory failure:
life support in the new era. Intensive Care Med 2012;38:210–
20.
Artif Organs, Vol. 37, No. 1, 2013