Artificial Organs

36(2):130138, Wiley Periodicals, Inc.

2011, Copyright the Authors
Artificial Organs 2011, International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.

Effects of Exchange Transfusion With

Liposome-Encapsulated Hemoglobin on VO2/DO2
Hitoshi Ikegawa, Yasuyuki Kuwagata, Koichi Hayakawa, Kazuo Noguchi, Hiroshi Ogura,
and Hisashi Sugimoto
Department of Traumatology and Acute Critical Medicine (D-8), Osaka University Graduate School of Medicine, Suita,
Osaka, Japan

Abstract: We clarified the effect of exchange transfusion

with liposome-encapsulated hemoglobin (neo red cells,
NRCs) with low O2 affinity (P50O2 = 50 mm Hg) on O2
metabolism. Rabbits were randomly assigned to receive
serial exchange transfusions with NRC (NRC group, n = 5),
shed blood diluted 1:1 with saline (red blood cell (RBC)
group, n = 5), or saline alone (plasma group, n = 4) under
hemodynamic monitoring. Cardiac tamponade was then
induced and successively reversed to determine relationships between O2 consumption (VO2) and O2 delivery
(DO2) using the dual-line method. Mean values of Hb concentration after exchange transfusion were 5.7 (NRC), 6.0
(RBC), and 1.5 (plasma) g/dL. The plasma group could not
even survive the initial exchange hemodilution due to a
critical decrease in DO2. The NRC, but not the RBC group,

developed progressive metabolic acidosis and lactatemia,

as well as increases in PaCO2 and decreases in tissue PO2 in
skeletal muscle after exchange transfusion. Nonetheless,
systemic O2 uptake indices obtained from an analysis of the
VO2/DO2 relationship in the NRC and RBC groups were
comparable. These findings suggested that systemic O2
uptake was maintained in rabbits after exchange transfusion with NRC, although progressive tissue hypoxia with
systemic acidosis is indicative of inadequate peripheral circulation and insufficient aerobic metabolism during
extended hemodilution in which 86% of the circulating
blood is replaced. Key Words: Exchange transfusion
RabbitOxygen deliveryOxygen consumptionOxygen metabolism.

Several artificial oxygen carriers have been developed as an alternative to blood to overcome problems
associated with transfusion such as disease transmission,mismatched blood type, and limited supply (14).
Among potential materials, a hemoglobin-based
blood substitute that appeared superior with respect
to oxygen-carrying capacity was, in fact, associated
with several toxic effects related to free hemoglobin
that included a nonspecific vasopressor effect, induction of vasospasm mainly by nitric oxide scavenging,
and penetration of tissues due to its small size (5).
Liposome-encapsulated hemoglobin (LEH) was
developed as a novel oxygen carrier that minimizes

these unfavorable vasoactive effects. One study

showed that rabbits survived for 6 months (until sacrifice) after blood exchange with LEH that replaced
approximately 85% of the circulating blood (6). In
another study, animals that received LEH as a resuscitation fluid survived lethal hemorrhagic shock (6,7).
Furthermore, a large body of evidence indicates that
the oxygen-transporting ability of LEH is equivalent
to that of red blood cells (RBCs) (812), although the
evidence does not necessarily support the notion that
organs and/or tissues will effectively take up oxygen
from circulating LEH. We postulated that organs
and/or tissues can extract sufficient oxygen from circulating blood after most (90%) of the (RBCs) have
been replaced by LEH. The present study tests our
hypothesis by evaluating changes in oxygen metabolism in rabbits during and after isovolumic
hemodilution (exchange transfusion) with LEH. The
relationship between oxygen consumption (VO2) and
oxygen delivery (DO2) was determined by repeating
cardiac tamponade induction to vary cardiac output

doi:10.1111/j.1525-1594.2011.01405.x
Received August 2010; revised August 2011.
Address correspondence and reprint requests to Dr. Hitoshi
Ikegawa, Department of Traumatology and Acute Critical Medicine (D-8), Osaka University Graduate School of Medicine,
Yamadaoka 2-2, Suita, Osaka 565-0871, Japan. E-mail: ikegawa@
hp-emerg.med.osaka-u.ac.jp

130

aor_1405

130..138

EFFECTS OF EXCHANGE TRANSFUSION WITH LEH ON VO2 /DO2

and assess the actual efficacy of DO2 with neo red cells
(NRCs; Terumo, Inc., Tokyo, Japan).
MATERIALS AND METHODS
Neo red cells
Table 1 shows the physical properties of the NRC.
The relatively high P50O2 (partial pressure of O2) that
permits half of hemoglobin to bind to O2 also facilitates oxygen unloading from hemoglobin due to an
enclosed inositol hexaphosphate moiety (13). The
NRC used herein are currently unavailable because
the company has further altered the formulation and
manufacturing processes to enhance LEH stability in
circulating blood and provide a longer shelf life (14).
Surgical preparation
The Animal Care Committee of Osaka University
approved the study, which conformed to the US
National Institutes of Health standards for animal
experimentation. Fourteen female New Zealand
White rabbits weighing 2.63.2 kg were anesthetized
with pentobarbital sodium (initial intravenous bolus
of 25 mg/kg, followed by a constant infusion of 10 mg/
kg/h). Normal saline was intravenously infused at a
constant rate of 6 mL/kg/h throughout the study as a
maintenance fluid and anesthetic carrier. A custommade 4.0-mm endotracheal tube was inserted via a
tracheotomy and secured by ligation with 2-0 silk. A
20-G low-compliance catheter was inserted into the
right carotid artery to measure arterial pressure and
to sample arterial blood. A 3-Fr Swan-Ganz thermodilution catheter (Model TS094H-3F, Baxter
Healthcare Corp., Irvine, CA, USA) was advanced
into the pulmonary artery via the right jugular vein to
sample mixed venous blood and to determine cardiac
output. The diaphragmatic muscle was gently
detached from the xiphoid process with bipolar electrical coagulation via a small upper transverse
laparotomy. Through this window, a 3-mm pericardiotomy was prepared for the later induction of artificial cardiac tamponade. The abdomen was then
closed with 1-0 silk, leaving a small window with
TABLE 1. Physical properties of neo red cells
Property
Relative osmolality
pH
Hemoglobin concentration
Particle size
Mean methemoglobin rate
Viscosity
P50

1
7.5
6.0 g/dL
180 88 nm
5.5% (range, 3.67.2%)
2 cp
4555 mm Hg

131

which to approach the pericardiotomized heart. The

right adductor muscle was punctured with intravenous catheter needles for the later positioning of a
flexible oxygen electrode.
The rabbits were postsurgically paralyzed with
pancuronium bromide (0.2 mg) intravenous bolus
and periodic supplementary doses (0.1 mg) to
prevent changes in resting VO2, and mechanically
ventilated with a Harvard non-rebreathing ventilator
at a tidal volume of 25 mL (FIO2 = 50%) at frequencies adjusted to a baseline target PaCO2 of 35 to
40 mm Hg.The expiration outlet of the ventilator was
connected to a thermal mass flow meter (Model
3920E, Kofloc, Inc., Tokyo, Japan) to monitor ventilation volume and to collect mixed expired gas in which
O2 and CO2 fractions were determined.Animals were
maintained at a constant body temperature of 39.0
39.5C on a heated operating table (Model 596833,
Harvard Apparatus, Inc., Holliston, MA, USA).
Experimental protocol
The animals were prepared for surgery and stabilized for 45 min, and then baseline data were
obtained. The animals were then randomly assigned
to groups based on the type of blood substitute: the
NRC group (n = 5) received NRC ([Hbnrc] = 6 g/
100 mL) suspended in saline containing 3% human
albumin (Baxter International, Inc., Tokyo, Japan);
the RBC group (n = 5) received shed blood diluted
1:1 with saline containing 3% human albumin; and
the plasma group (n = 4) received saline containing
3% human albumin (Fig. 1). After baseline measurements, 9 mL/kg of circulating blood was withdrawn
over 150 s and an identical amount of blood substitute was subsequently infused over 150 s. This procedure was repeated eight times in 5-min cycles over
a period of 40 min to exchange about 90% of the
total circulating blood. Hemodynamics and oxygen
metabolism parameters were measured at 60, 90, and
120 min after starting the blood exchange. A custommade balloon for cardiac tamponade was then
inserted through the pericardial window. Cardiac
tamponade was induced by inflating the balloon and
adjusting the mean arterial pressure (MAP) value to
about 90% of that at 120 min. The animals were stabilized for 7 min under the new condition, and an
experimental dataset was collected during the next
3 min. Cardiac tamponade was then completely
released over 5 min to restore hemodynamic status.
The next dataset was collected when the MAP was
about 90% that of the previous experimental value.
The stepwise cardiac tamponade series was continued in 15-min cycles until stable blood pressure could
no longer be maintained. The animals were sacrificed
Artif Organs, Vol. 36, No. 2, 2012

132

H. IKEGAWA ET AL.

FIG. 1. Experimental protocol.

by an overdose of anesthetic after completion of the

experiment.We confirmed at necropsy that the SwanGanz catheter was correctly positioned in all animals.
Measurements and calculations
We measured arterial blood gas (Model ABL510,
Radiometer, Copenhagen, Denmark), blood lactate
concentration (Model ABL100, Radiometer), hematocrit, and the rate of solid constituents other than
plasma obtained by centrifugation of the mixture of
blood and NRC (volume% of NRC : NRC-crit) at
baseline, 60, 90, and 120 min after starting the blood
exchange and at every 15 min during graded cardiac
tamponade. Hematocrit and NRC-crit were determined in blood samples as follows. Dextran (10%)
was added to each blood sample to a final ratio of
10% dextran/blood = 1:1, and then, samples were
placed in two hematocrit tubes and centrifuged at
12 000 rpm for 30 min. Hematocrit and NRC-crit
were calculated as the sum of the readings of the
ratios (%) of each corresponding component in the
two hematocrit tubes. The hemoglobin concentration in blood was measured (Model CELL-DYN,
Abbott Laboratories, Chicago, IL, USA) at baseline
and after blood exchange as follows: RBC hemoglobin concentration in blood (Hbrbc) (g/100 mL) =
baseline hemoglobin observed hematocrit/baseline
hematocrit. The NRC hemoglobin concentration
in blood (Hbnrc) (g/100 mL) = 6/24 NRC-crit. The
baseline ratio of methemoglobin in RBCs (%metHbrbc) was measured (Model ABL510, Radiometer).
That ratio of methemoglobin in NRC (%metHbnrc)
was also measured 60 min after blood exchange as
follows. Blood samples were separated by centrifugation at 3000 rpm for 10 min. One volume of supernatant was then mixed 1:1 with 14% dextran (Sigma
Artif Organs, Vol. 36, No. 2, 2012

Co. Ltd, St. Louis, MO, USA) and separated by centrifugation at 4500 rpm for 10 min. The pellet was
resuspended in 2 mL of saline and then diluted with
a half volume of 12% dextran. The suspension was
frozen in liquid nitrogen and thawed in water to
lyse the liposome capsules and then separated by
centrifugation at 3000 rpm for 30 min. After adding
phosphate buffer (8 mM KH2PO4, 12 mM
Na2HPO412H2O) to the supernatant, the hemoglobin and methemoglobin concentrations in NRC
were measured using the cyanohemoglobin method,
and %metHbnrc was calculated. The ratio (%) of
oxygen saturation of Hbrbc (SO2rbc) in arterial
blood was assumed to be 100% because PaO2 was
held at >150 mm Hg by maintaining FIO2 at 0.5
throughout the study. The ratio (%) of oxygen saturation of Hbnrc (SO2nrc) was determined using a
nomogram of the NRC oxygen dissociation curve
provided by Terumo, Inc. with reference to corresponding PO2 values. The oxygen content in the
blood (ml/100 mL) was calculated as:

1.34 Hbrbc (100 %metHbrbc ) 100 SO2rbc 100 +

Hb nrc 64 500 4 22 400 (100 %metHbnrc )
100 SO2 nrc 100 .
Physiological monitoring and calculations
Heart rate (HR) and MAP were simultaneously
monitored and recorded (models VC640G,
RTA1200, AP641G, and AT601G, Nihon Kohden,
Tokyo, Japan). Cardiac output determined using the
thermodilution technique (Explorer cardiac output
computer, Baxter Healthcare Corp.) where three
1-mL injections of normal saline at room temperature were divided by body weight for conversion into

EFFECTS OF EXCHANGE TRANSFUSION WITH LEH ON VO2 /DO2

a cardiac index (mL/min/kg). Room temperature was
maintained at 24C 0.5C. Exhaled gases were collected in a gas-sampling chamber comprising a highcompliance plastic bag (Tedlar, Masuda, Osaka,
Japan) for 3 min and the expired fractions of O2
(FEO2) and CO2 (FECO2) were measured using a
medical gas analyzer (MG360, Minato Medical
Science Co. Ltd., Osaka, Japan) immediately after
sampling. The volume of sampled gases was read
on a digital display (Model ACM-10, Kofloc, Inc.)
attached to a thermal mass flow meter and an expiatory minute volume (VE) was calculated by dividing
the readings by 3. DO2 was calculated as the product
of arterial O2 content and cardiac index. VO2 was
calculated as follows:

VO2 ( mL min kg ) = ( VE weight )

{[(1 FEO2 FECO2 ) (1 FIO2 )] FIO2 FEO2 }.
Tissue oxygen partial pressure in the right adductor muscle (PtO2) was monitored using a flexible
Clarke-type polarographic electrode connected to a
computer-supported Licox system (GMS, Mielkendorf, Germany). Temperature within muscles was
simultaneously monitored, and PtO2 values were
adjusted using integral software. The electrode was
calibrated before and after each experiment in room
air. The Licox system was connected to a computer
for continuous data acquisition (PowerLab, ADInstruments Japan, Tokyo, Japan) and the PtO2 data
are described as %change compared with baseline
values.

Statistical analysis
All values are expressed as means SD. Statistical differences in mean hemodynamic and oxygen
metabolism variables between groups were tested
using analyses of variance of repeated measures.
Multiple comparisons between and within groups
were performed using the StudentNewmanKeuls
test. The critical point of DO2 (DO2crit) was determined for each animal from a plot of VO2 against
DO2 using the dual-line method of Samsel and
Schumacker (15). After defining individual critical
point values, each dataset was divided into two
subsets according to individual DO2crit values.
Values below the DO2crit were defined as supply
dependent. Simple linear regression analysis was
performed for each supply-dependent dataset and
differences in slopes between experimental groups
were compared by analysis of covariance (16).
A P value of <0.05 was considered statistically
significant.

133

RESULTS
Hemodilution
The exchange transfusion decreased hemoglobin
concentrations by about 50% compared with the
baseline values in both the NRC (5.7 0.5 vs.
11.4 0.7 g/dL) and RBC (6.0 1.2 vs. 12.1 0.7 g/
dL) groups and to an extremely low value in the
plasma group (1.5 0.3 g/dL vs. 11.6 1.3 g/dL).
Hemoglobin concentrations between the NRC and
RBC groups did not significantly differ before and
after blood exchange. Hematocrit before (35 1%)
and after (5 1%) exchange transfusion in the NRC
group indicated that the mean blood exchange rate
was about 86% of the baseline blood volume.
Hemodynamics
Table 2 shows sequential changes in hemodynamic
parameters at 60, 90, and 120 min after starting
exchange transfusion without cardiac tamponade.
MAP in the NRC group tended to decrease after
exchange transfusion, but the differences did not
reach statistical significance. MAP significantly
decreased in the RBC group at 60 and 90 min and
was significantly higher in the NRC than in the RBC
group at 90 min. Exchange transfusion did not significantly affect HR and cardiac index in either the NRC
or the RBC group. Exchange transfusion caused a
remarkably decreased MAP in all animals in the
plasma group, leading to circulatory collapse and the
inability to survive beyond 120 min.
Blood gases, systemic oxygen metabolism, and
tissue oxygenation
While PaO2 did not significantly differ among the
groups, metabolic acidosis with lactatemia occurred
and persisted for at least 120 min after exchange
transfusion in the NRC group. The PaCO2 values for
the NRC group tended to increase from the baseline
value after exchange transfusion, but the differences
did not reach statistical significance. In contrast,
blood gases and the acidbase balance did not change
in the RBC group. The PtO2 values were significantly
decreased from baseline values at 60 min in the
NRC and RBC groups (15 5 vs. 39 13 and 19
5 vs. 29 9 mm Hg, respectively). The significant
decrease in %PtO2 from the baseline value persisted
at 120 min after exchange transfusion in the NRC
group. The %PtO2 values for the RBC group sequentially recovered almost to baseline values by 120 min.
The %PtO2 values were significantly lower in the
NRC than in the RBC group at 60, 90, and 120 min.
DO2 decreased by about 50% in both the NRC and
RBC groups after exchange transfusion. VO2 was not
Artif Organs, Vol. 36, No. 2, 2012

134

H. IKEGAWA ET AL.
TABLE 2. Sequential changes in hemodynamics and oxygen metabolism before
and after isovolumic hemodilution

MAP (mm Hg)

NRC
RBC
Plasma
HR (bpm)
NRC
RBC
Plasma
CI (mL/min/kg)
NRC
RBC
Plasma
pH
NRC
RBC
Plasma
PaO2 (mm Hg)
NRC
RBC
Plasma
PaCO2 (mm Hg)
NRC
RBC
Plasma
BE (mmol/L)
NRC
RBC
Plasma
Lac (mg/dL)
NRC
RBC
Plasma
%PtO2 (%)
NRC
RBC
Plasma
DO2 (mL/min/kg)
NRC
RBC
Plasma
VO2 (mL/min/kg)
NRC
RBC
Plasma

Baseline

60 min

90 min

90 6
85 5
85 4

83 8
73 2*
30 8*

85 10
70 3*
21 16*

120 min
80 10
76 3
n/a

264 15
268 19
267 28

284 14
281 11
234 35

284 12
282 14
215 2*

288 10
277 13
n/a

168 10
150 12
157 16

160 15
162 16
132 60

173 18
164 15
54 76*

162 19
156 17
n/a

7.41 0.03
7.39 0.05
7.40 0.04

7.22 0.06*
7.38 0.03
7.21 0.07*

7.26 0.06*
7.41 0.04
7.18 0.16*

7.25 0.09*
7.41 0.04
n/a

203 3
210 25
225 4

209 14
221 19
256 7

190 30
217 17
240 5

189 11
215 21
n/a

39 5
42 9
38 3

46 3
40 5
24 7*

-0.1 1.8
0.3 1.6
-0.9 1.9

-8.5 3.0*
-1.4 2.2
-17.2 3.0*

26 3
26 3
22 4

50 4*
26 4
98 13*

47 8
37 6
15 12*

50 8
37 5
n/a

-6.0 2.7*
-0.9 1.8
-23.0 1.8*

-5.3 3.2*
-1.1 1.7
n/a

42 2*
27 5
155 6*

42 12*
27 5
n/a

37 18*
72 16*
7 3*

46 21*
80 14
2 2*

46 19*
93 17
n/a

26.6 2.1
24.4 3.6
25.1 2.7

13.4 0.7*
13.8 2.4*
3.4 1.0*

13.9 1.7*
13.9 2.4*
1.4 1.9*

12.3 1.6*
12.9 2.5*
n/a

10.7 0.8
9.9 0.9
10.3 0.8

10.3 1.0
9.4 1.3
3.8 1.1*

10.1 0.6
9.4 1.1
1.9 1.2*

9.2 1.1
9.2 0.8
n/a

100
100
100

Values are shown as mean SD.

* Statistically significant difference from the baseline value (P < 0.05).

Statistically significant difference from the time-matched corresponding value in the RBC
group (P < 0.05).
%PtO2, percent change in tissue PO2; BE, base excess; CI, cardiac index; DO2, oxygen
delivery; HR, heart rate; Lac, serum lactate concentration; MAP, mean arterial pressure; n/a,
not available; PaCO2, arterial PCO2; PaO2, arterial PO2; PCO2, partial pressure of carbon
dioxide; PO2, partial pressure of oxygen; VO2, oxygen consumption.

altered in response to this level of decrease in DO2 in

either the NRC or the RBC group. The plasma group
developed severe metabolic acidosis with lactatemia
and the substantial decreases in PtO2, DO2, and VO2
at such a high degree of hemodilution did not permit
survival.
Figure 2 shows representative sequential changes
in PtO2 during and after blood exchange in each
group. The muscle PtO2 value decreased immediately
Artif Organs, Vol. 36, No. 2, 2012

after starting blood withdrawal and remained low

throughout the blood exchange procedure in the
NRC group. The PtO2 value was very low
(~4 mm Hg) at the end of blood exchange, gradually
increased thereafter, but remained quite low at
120 min after starting the procedure in the NRC
group. In contrast, the muscle PtO2 value decreased
during blood withdrawal and increased during
replacement with autologous blood, and this

EFFECTS OF EXCHANGE TRANSFUSION WITH LEH ON VO2 /DO2

135

dependent lines did not significantly differ between

the two groups (NRC: y = 0.12x + 7.6 and 0.72x + 1.7,
respectively; RBC: y = 0.07x + 8.1 and 0.75x + 2.0,
respectively).
DISCUSSION

FIG. 2. Representative sequential changes in tissue partial pressure of oxygen during and after blood exchange in all three
groups. Tissue partial pressure of oxygen shown in NRC (A),
RBC (B), and plasma (C) groups.

sequence was repeated in the RBC group throughout

the blood exchange procedure. The PtO2 value at the
end of blood exchange in the RBC group was below
the baseline value, but recovered almost to the baseline value by 120 min. The muscle PtO2 progressively
decreased during and after blood exchange in the
plasma group. All animals in this group were dead by
120 min after starting the blood exchange procedure.
VO2/DO2 relation during cardiac tamponade
Figure 3 shows individual data plots of the VO2/
DO2 relationship at baseline and during cardiac
tamponade. The mean values of DO2crit and of the
oxygen extraction ratio at the critical point did not
significantly differ between the NRC and RBC
groups (NRC: 10.4 1.1 mL/min/kg and 88 6%,
respectively; RBC: 9.4 1.1 mL/min/kg and 92
3%, respectively). Figure 4 shows overlaid plots of
the VO2/DO2 relationship for these groups. The
slopes of the supply-independent and supply-

The principal finding of the present study is that

the oxygen transport and uptake is maintained even
when most of the circulating blood is replaced with
NRC solution. All animals in the plasma group died
after exchange transfusion without hemoglobin due
to a critical decrease in DO2 that caused a critical
decrease in VO2. These results indicate that the procedure of exchange transfusion applied herein was
lethal in the absence of an oxygen carrier as potent
as hemoglobin. The slopes of the regression lines
and the DO2crit values determined by the dual-line
method did not differ significantly between the
NRC and RBC groups. In particular, the consistency
of the slopes of the VO2/DO2 relationship during
the supply-dependent area between the two groups
reflected an oxygen extraction capability of LEH
equal to that of RBCs at a hemoglobin concentration of ~6 g/dL. Other studies of oxygen metabolism
in animals under blood exchange have emphasized
the potential of LEH as an artificial oxygen carrier
based on its ability to maintain VO2 (9,1719). VO2
in these studies was invariably calculated using the
reverse Fick method, which does not represent a
steady metabolic measure but rather reflects a
momentary value. The VO2/DO2 relationship determined in this manner might be erroneous because
the calculation includes shared measurement variables (20,21). Indirect calorimetry is presently the
recommended method of evaluating VO2/DO2 relationships by measuring VO2. Thus, the present study
initially determined the VO2/DO2 relationship in
animals during exchange transfusion with NRC
using indirect calorimetry and then quantified the
ability of NRC as an oxygen carrier within circulating blood.
Data obtained from intact rabbits during our previous studies showed that mean DO2crit values and
mean oxygen extraction ratios range from 11.5 to
13.3 mL/min/kg and from 72 to 83%, respectively
(2224). Oxygen uptake was higher in both the NRC
and RBC groups as indicated by a lower DO2crit
value and a higher oxygen extraction ratio at the
critical point compared with our historical data.
Whereas some studies (2527) have found that isovolumic hemodilution with an asanguinous solution
ameliorates systemic or regional O2 uptake, others
have indicated that hemodilution with LEH also
Artif Organs, Vol. 36, No. 2, 2012

136

H. IKEGAWA ET AL.

FIG. 3. Relationship between oxygen delivery (DO2) and oxygen consumption (VO2) in NRC and RBC groups. Plots show DO2crit (DO2
at critical point) determined using dual-line method. X, baseline values for () NRC and () RBC groups.

confers benefits such as improved rheological properties (10), decreased microheterogeneity of myocardial blood flow (28), and increased functional
capillary density (29) on tissue oxygenation.
However, the effect of hemodilution alone cannot be
distinguished from the favorable outcomes of these
previous studies because none of them included a
hemoglobin-matched control group. The present
study showed that isovolumic hemodilution with
NRC is likely to ameliorate systemic oxygen uptake

FIG. 4. Relationship between oxygen delivery (DO2) and oxygen

consumption (VO2) during stagnant hypoxia. Supply-dependent
and -independent lines were calculated from pooled data. and
solid lines, NRC group; and dashed lines, RBC group.
Artif Organs, Vol. 36, No. 2, 2012

to the same extent as hemoglobin-matched isovolumic hemodilution with an albumin solution, although
the underlying mechanism remains unclear.
MAP slightly decreased in both the NRC and RBC
groups in the absence of changes in HR and cardiac
index after exchange transfusion, but the value was
significantly greater at 90 min in the NRC than in
the RBC group, suggesting an increase in systemic
vascular resistance, or systemic vasoconstriction.
The NRC group, but not the RBC group, developed
metabolic acidosis with lactatemia during blood
exchange, as indicated by a significant decrease in pH
and base excess and by a significant increase in serum
lactate concentration at 60 min. These metabolic disturbances did not continue to increase or were somewhat mitigated by 120 min in the NRC group. These
observations suggest that some degree of vasoconstriction with systemic acidosis developed during the
exchange transfusion with NRC. However, others
have indicated that hemoglobin encapsulation in
liposomes minimizes the unfavorable vasoconstrictive and hypertensive effects of free hemoglobin (30
32). The PtO2 value of the adductor muscle abruptly
fell after starting the blood exchange procedure in
the NRC, but not in the RBC group (Fig. 3). The
decrease in %PtO2 of the adductor muscle was significantly greater in the NRC than in the RBC group
and persisted for at least 120 min, whereas the %PtO2

EFFECTS OF EXCHANGE TRANSFUSION WITH LEH ON VO2 /DO2

in the RBC group had returned to near the baseline
value by that time. These observations suggest an
unexpected decrease in oxygen supply to peripheral
tissues that cannot be explained either by the fluctuation in systemic DO2 caused by the blood exchange
procedure or by a decrease in the arterial O2 content
due to hemodilution. These conditions suggest that
vasoconstriction and tissue hypoxia in the body,
including in the skeletal muscles, develop and persist
in the presence of NRC during and after blood
exchange. In addition, the amount of tissue hypoxia
was not so extensive as to alter the systemic VO2/DO2
relationship.
Several reports have described the beneficial
effects of LEH on recovery from hemorrhagic shock
(7,9,19,32), but tissue hypoxia or metabolic acidosis
occurring after massive LEH administration has
never been described. Nogami et al. recently demonstrated the value of LEH as an artificial oxygen
carrier using a fatal progressive hemodilution rat
model (33). They found that all animals transfused
with LEH or washed RBC suspension were rescued
from lethal progressive hemodilution without generating a potent nitric oxide scavenging effect or
changes in levels of plasma nitric oxide metabolites.
In addition, they found that LEH transfusion suppressed the expression of hypoxia-inducible factor-1
alpha in the liver and kidney, which supports the
salutary effects of LEH transfusions on the
microcirculation. Rohlfs et al. tested arterial blood
pressure responses to seven cell-free hemoglobin
solutions in rats after 50% isovolumic exchange
transfusion (34). They also measured the reactivity of
these hemoglobin solutions with nitric oxide and concluded that the variations in blood pressure increases
that occurred during exchange transfusion with
various cell-free hemoglobin solutions could not be
the result of nitric oxide scavenging reactions at
heme, but were rather due to alternative physiological mechanisms such as excess O2 supply due to
facilitated O2 unloading from hemoglobin that
resulted in autoregulatory vasoconstriction in the
microcirculation. Cabrales et al. investigated the
differential effect of high (P50O2 = 8 mm Hg) and
normal (P50O2 = 29 mm Hg) oxygen affinity LEH in
extreme hemodilution using their unique hamster
window chamber model (35). They noted that slight
systemic acidemia developed during hemodilution
when the oxygen affinity of the LEH was normal, but
not when it was high. They also found a significantly
higher tissue PO2 value under the dorsal skin when
the oxygen affinity of the LEH was high. Fukumoto
et al. tested the differential effect of infusions
of LEH with high (P50O2 = 10 mm Hg) and low

137

(P50O2 = 40 mm Hg) oxygen affinity on the rodent

ischemic stroke model with photochemically induced
thrombosis of the middle cerebral artery (36). They
confirmed that the protective effect was better with
high than with low oxygen affinity LEH based on the
extent and distribution of cerebral infarction. The
NRC used herein was characterized by low oxygen
affinity (P50O2 = 50 mm Hg) to easily unload oxygen
from hemoglobin (6).The high arteriolar PO2 in some
tissues including skeletal muscle in the present study
might have disturbed the normal vasodilatory
response to ischemia during and after the blood
exchange procedure. This could have resulted in the
metabolic acidosis and lactatemia in the NRC group
that persisted for at least 120 min. The LEH with a
relatively high P50O2 might not be efficient enough to
deliver O2 as its oxygen saturation would be very low
by the time the LEH reached the tissues.
Thus, the strengths of the present study include
the accurate determination of the systemic VO2/
DO2 relationship using indirect calorimetry and a
study design that included a hemoglobin
concentration-matched control. The limitations of
the present study are as follows. First, this experiment addressed extreme conditions that would
hardly occur in the clinical setting. Second, metabolic acidosis that might affect the oxygen uptake
ability of peripheral tissues could not be eliminated
before determination of the VO2/DO2 relationship.
Third, the hemoglobin concentration was below the
physiologically normal range, and, thus, scientific or
clinical insights derived from this study are limited
to anemia. Nevertheless, the physiological and/or
pathological responses after exchange transfusion
need to be determined to understand the full spectrum of physiological and/or pathological responses
to exchange transfusion with blood substitutes. Our
findings provide valuable information for a potential life-saving role of LEH for patients experiencing massive blood loss.
CONCLUSION
Systemic oxygen metabolism was maintained in
rabbits during exchange transfusion with neo red
cells to about 86% of the circulating blood as indicated by the slopes of the dual regression lines and by
the critical inflection points of these lines in the
analyses of the VO2/DO2 relationship. However,
exchange transfusion with NRC to such an extent
was associated with the development of metabolic
acidosis, lactatemia, and tissue hypoxia in skeletal
muscle indicating inadequate peripheral circulation.
Although the underlying mechanism of NRCArtif Organs, Vol. 36, No. 2, 2012

138

H. IKEGAWA ET AL.

induced disruption of the peripheral circulation

remains unclear, precapillary O2 unloading because
of the low O2 affinity of NRC might at least partly
account for the deficit in aerobic metabolism. In this
regard, liposome-encapsulated hemoglobin with high
O2 affinity might be more acceptable as a red blood
cell substitute than NRC under conditions such as
uncontrolled hemorrhage.
REFERENCES
1. Yokoyama K, Yamanouchi K, Watanabe M, et al. Preparation
of perfluorodecalin emulsion, an approach to the red cell
substitute. Fed Proc 1975;34:146877.
2. Snyder SR, Welty EV, Walder RY, Williams LA, Walder TA.
HbxL 99 alpha: a hemoglobin derivative that is cross-linked
between the alpha subunits is useful as a blood substitute. Proc
Natl Acad Sci U S A 1987;84:72804.
3. Djordjevich L, Miller IF. Synthetic erythrocytes from lipidencapsulated hemoglobin. Exp Hematol 1980;8:58492.
4. Farmer MC, Gaber BP. Liposome-encapsulated hemoglobin
as an artificial oxygen-carrying system. Methods Enzymol
1987;149:184200.
5. Chang TM. Hemoglobin-based red blood cell substitutes. Artif
Organs 2004;28:78994.
6. Takahashi A. Characterization of neo red cells (NRCs), their
function, and safety in in vivo tests. Artif Cells Blood Substit
Immobil Biotechnol 1995;23:34754.
7. Sakai H, Masada Y, Horinouchi H, et al. Hemoglobin-vesicles
suspended in recombinant human serum albumin for resuscitation from hemorrhagic shock in anesthetized rats. Crit Care
Med 2004;32:53945.
8. Usuba A, Miyazawa M, Motoki R, et al. Oxygen transport
capacity and hemodynamic effects of newly developed artificial blood neo red cells (NRC). Int J Artif Organs
1993;16:5516.
9. Usuba A, Osuka F, Kimura T, et al. Effect of liposomeencapsulated hemoglobin, neo red cells, on hemorrhagic
shock. Surg Today 1998;28:102735.
10. Erni D, Wettstein R, Schramm S, et al. Normovolemic hemodilution with Hb vesicle solution attenuates hypoxia in ischemic
hamster flap tissue. Am J Physiol Heart Circ Physiol
2003;284:H1702H1709.
11. Izumi Y, Sakai H, Hamada K, et al. Physiologic responses to
exchange transfusion with hemoglobin vesicles as an artificial
oxygen carrier in anesthetized rats: changes in mean arterial
pressure and renal cortical tissue oxygen tension. Crit Care
Med 1996;24:186973.
12. Izumi Y, Sakai H, Kose T, et al. Evaluation of the capabilities
of a hemoglobin vesicle as an artificial oxygen carrier in a rat
exchange transfusion model. ASAIO J 1997;43:28997.
13. Ogata Y, Goto H, Kimura T, Fukui H. Development of neo
red cells (NRC) with the enzymatic reduction system of
methemoglobin. Artif Cells Blood Substit Immobil Biotechnol
1997;25:41727.
14. Kaneda S, Ishizuka T, Goto H, Kimura T, Inaba K, Kasukawa
H. Liposome-encapsulated hemoglobin, TRM-645: current
status of the development and important issues for clinical
application. Artif Organs 2009;33:14652.
15. Samsel RW, Schumacker PT. Determination of the critical O2
delivery from experimental data: sensitivity to error. J Appl
Physiol 1988;64:207482.
16. Zar JH. Biostatistical Analysis. Englewood Cliffs, NJ: PrenticeHall, Inc., 1984;292305.
17. Usuba A, Motoki R, Ogata Y, Suzuki K, Kimitani T. Effect and
safety of liposome-encapsulated hemoglobin neo red cells
(NRCs) as a perfusate for total cardiopulmonary bypass. Artif

Artif Organs, Vol. 36, No. 2, 2012

Cells Blood Substit Immobil Biotechnol 1995;23:33746.

18. Izumi Y, Yamahata T, Yozu R, Kobayashi K, Mukai M. The
oxygen transporting capability of neo red cells (NRC) evaluated under total cardiopulmonary bypass. Jpn J Thorac Cardiovasc Surg 1998;46:307.
19. Terajima K, Tsueshita T, Sakamoto A, Ogawa R. Fluid resuscitation with hemoglobin vesicles in a rabbit model of acute
hemorrhagic shock. Shock 2006;25:1849.
20. Archie JP Jr. Mathematic coupling of data: a common source
of error. Ann Surg 1981;193:296303.
21. Stratton HH, Feustel PJ, Newell JC. Regression of calculated
variables in the presence of shared measurement error. J Appl
Physiol 1987;62:208393.
22. Kuwagata Y, Oda J, Matsuyama S, et al. Interleukin-1b alters
the oxygen delivery-oxygen consumption relationship in
rabbits by increasing the slope of the supply-independent line.
Shock 2000;14:1939.
23. Oda J, Kuwagata Y, Nakamori Y, et al. Mild hypothermia
alters the oxygen consumption/delivery relationship by
decreasing the slope of the supply-dependent line. Crit Care
Med 2002;30:153540.
24. Kuwagata Y, Oda J, Irisawa T, et al. Effect of ibuprofen on
interleukin-1b-induced abnormalities in hemodynamics and
oxygen metabolism in rabbits. Shock 2003;20:55864.
25. Van der Linden P, Gilbart E, Paques P, Simon C, Vincent JL.
Influence of hematocrit on tissue O2 extraction capabilities
during acute hemorrhage. Am J Physiol 1993;264(Pt 2):
H1942H1947.
26. Lindbom L, Mirhashemi S, Intaglietta M, Arfors KE. Increase
in capillary blood flow and relative haematocrit in rabbit skeletal muscle following acute normovolaemic anaemia. Acta
Physiol Scand 1988;134:50312.
27. Hutter J, Habler O, Kleen M, et al. Effect of acute normovolemic hemodilution on distribution of blood flow and tissue
oxygenation in dog skeletal muscle. J Appl Physiol 1999;
86:8606.
28. Matsumoto T, Asano T, Mano K, et al. Regional myocardial
perfusion under exchange transfusion with liposomal hemoglobin: in vivo and in vitro studies using rat hearts. Am J
Physiol Heart Circ Physiol 2005;288:H1909H1914.
29. Plock JA, Contaldo C, Sakai H, et al. Is hemoglobin in
hemoglobin vesicles infused for isovolemic hemodilution
necessary to improve oxygenation in critically ischemic
hamster skin? Am J Physiol Heart Circ Physiol 2005;289:
H2624H2631.
30. Nakai K, Matsuda N, Amano M, et al. Acellular and cellular
hemoglobin solutions as vasoconstrictive factors. Artif Cells
Blood Substit Immobil Biotechnol 1994;22:55964.
31. Sakai H, Hara H, Yuasa M, et al. Molecular dimensions of
Hb-based O2 carriers determine constriction of resistance
arteries and hypertension. Am J Physiol 2000;279:H908H915.
32. Sakai H, Takeoka S, Wettstein R, Tsai AG, Intaglietta M,
Tsuchida E. Systemic and microvascular responses to hemorrhagic shock and resuscitation with Hb vesicles. Am J Physiol
Heart Circ Physiol 2002;283:H1191H1199.
33. Nogami Y, Kinoshita M, Takase B, et al. Liposomeencapsulated hemoglobin transfusion rescues rats undergoing
progressive hemodilution from lethal organ hypoxia without
scavenging nitric oxide. Ann Surg 2008;248:3109.
34. Rohlfs RJ, Bruner E, Chiu A, et al. Arterial blood pressure
response to cell-free hemoglobin solutions and the reaction
with nitric oxide. J Biol Chem 1998;273:1212834.
35. Cabrales P, Sakai H, Tsai AG, Takeoka S, Tsuchida E, Intaglietta M. Oxygen transport by low and normal oxygen affinity
hemoglobin vesicles in extreme hemodilution. Am J Physiol
Heart Circ Physiol 2004;288:H1885H1892.
36. Fukumoto D, Kawaguchi AT, Haida M, Yamano M, Ogata Y,
Tsukada H. Liposome-encapsulated hemoglobin reduces the
size of cerebral infarction in rat: effect of oxygen affinity. Artif
Organs 2009;33:15963.