Vol. 85, Issue 3, 993-1003, September 1998
Vascular resistance and the efficacy of red cell substitutes
in a rat hemorrhage model
Robert M.
Winslow,
Armando
Gonzales,
Maria L.
Gonzales,
Michael
Magde,
Michael
McCarthy,
Ronald
J.
Rohlfs, and
Kim D.
Vandegriff
Department of Medicine, School of Medicine, University of
California, San Diego, and Department of Veterans Affairs Medical
Center, San Diego, California 92161
 |
ABSTRACT |
We have compared
polyethylene glycol-modified bovine hemoglobin (PEG-Hb; high
O2 affinity, high viscosity, high
oncotic pressure) and human hemoglobin cross-linked between the
-chains (-Hb; low O2
affinity, low viscosity, low oncotic pressure) with a
non-O2-carrying plasma expander
(pentastarch, high viscosity and oncotic pressure) after a 50% (by
volume) exchange transfusion followed by a severe (60% of blood
volume) hemorrhage. Mean arterial pressure and systemic vascular
resistance rose significantly in the -Hb but not in the PEG-Hb
animals. Two-hour survival was greater in the PEG-Hb animals (93%)
than in control (35%), pentastarch (8%), or -Hb (6%) animals.
In the PEG-Hb animals, there was no disturbance of acid-base balance,
significantly less accumulation of lactic acid, and higher cardiac
output than in the other groups. The data suggest that the rise in
vascular resistance that follows -Hb exchange transfusion offsets
the additional O2 transport provided by the cell-free hemoglobin. When resistance does not rise, as
with PEG-Hb, even relatively small amounts of cell-free hemoglobin
appear to be a very effective blood replacement.
hemodilution; hemorrhage; blood substitutes; hemoglobin; -hemoglobin
| |
INTRODUCTION |
HEMOGLOBIN-BASED
O2 carriers have been proposed as
temporary alternatives to red blood cells in certain clinical settings (reviewed in Ref. 29). However, elevation of systemic and pulmonary blood pressure has been observed with some candidate solutions (9, 14),
and because of the known reactivity of hemoglobin with NO (6) and the
identity of NO as the endothelium-derived relaxing factor (21) it is
often assumed that vasoconstriction is a general property of cell-free
hemoglobin solutions that might limit their clinical potential.
Therefore, an understanding of the hemoglobin-induced vasoactivity is
important not only for the development of
O2 carriers as therapeutics but
also for a deeper understanding of the mechanisms of
O2 delivery
(
O2)
to tissue.
On the basis of our observation that hemoglobin derivatives with
different vasopressor effects have nearly equal NO-binding properties
(23), we recently developed an alternative hypothesis to explain
hemoglobin-induced vasoactivity. Theoretical (5, 12) and in vitro data
(20) provide strong arguments that cell-free hemoglobin would be
expected to facilitate diffusive
O2 transport to tissues, but
attempts to confirm this prediction in vivo have failed (1, 10, 11).
Our hypothesis is that demonstration of diffusive
O2 transport by cell-free
hemoglobin is prevented by local autoregulatory mechanisms that lead to
vasoconstriction as a counterbalance to increased
O2 supply (27). Because
autoregulation is multifactorial, an important consequence of this
hypothesis is that alteration of one or more of the rheological (13)
and O2 transport properties (35)
of hemoglobin solutions might overcome vasoconstriction.
It is now possible to examine this proposition more closely because of
the availability of cell-free O2
carriers with properties that differ in critical ways. Among the
properties of hemoglobin solutions that can be varied are viscosity,
oncotic pressure, and O2 affinity.
In the experiments to be reported here, we explore some of these
properties using two well-characterized hemoglobins and a nonhemoglobin
plasma expander. The first is a human hemoglobin derivative modified by
cross-linking between -99-lysine residues (-Hb), which has low
viscosity, low oncotic pressure, and an O2 affinity approximately equal to
blood. The second is bovine hemoglobin surface modified with
polyethylene glycol (PEG-Hb), which has high viscosity, high oncotic
pressure, and high O2 affinity. These hemoglobins are compared with a conventional plasma expander, pentastarch (PS; 250,000 mol wt), a hydroxyethyl starch with ~45% hydroxyethyl substitution, resulting in high viscosity and high oncotic
pressure.
The experimental model we have chosen to evaluate these solutions is
50% isovolemic hemodilution with the test solution followed by an
exponential 60% blood volume hemorrhage (7). This hemorrhage is severe
enough so that ~50% of control animals die within 1 h after
conclusion of hemorrhage. Thus the effectiveness of perturbations such
as hemodilution is readily apparent not only in overall survival but
also in the physiological changes that occur during and after the
hemorrhage. An additional reason to choose this model is that perioperative hemodilution is a clinical application for which "blood substitutes" might be ideally suited (31).
Our results show that exchange transfusion with PEG-Hb does not raise
blood pressure, and after hemorrhage, animals show many features of
improved tissue oxygenation, i.e., better survival, normal acid-base
status, lower lactic acid production, and higher cardiac output
(
), than non-exchange-transfused animals. In contrast, exchange transfusion with -Hb raises blood pressure significantly, but after hemorrhage, animals have shortened survival, higher lactic acid production, and lower than
non-exchange-transfused animals. Control experiments with PS indicate
that the favorable results with PEG-Hb are not due merely to its
increased viscosity and/or oncotic pressure.
This study is an important first step toward the goal of a better
understanding of the role of viscosity, oncotic pressure, and
O2 affinity in
O2
by cell-free O2 carriers.
| |
MATERIALS AND METHODS |
Test solutions.
PEG-Hb is bovine hemoglobin conjugated to 5,000-mol-wt
methoxypolyoxyethylene (PEG) chains (36). The PEG chains are covalently attached to 
-amino groups of surface lysine residues such that 10-12 PEG chains are attached to each tetramer. It was supplied as
a gift from Enzon. -Hb was prepared as described previously (32)
and supplied as a gift from the US Army Blood Research Detachment,
Walter Reed Army Institute for Research. PS is the commercial product
Pentaspan obtained from DuPont Merck. The
O2 affinity of rat blood was
determined using an apparatus described previously (34) in which pH and
PCO2 are held constant at 7.4 and 40 Torr, respectively. O2 affinity of
-Hb and PEG-Hb were determined by simultaneous measurements of
absorbance and PO2 using a diode
array spectrophotometer (model 3000, Milton Roy) and a
micro-O2 electrode
(Microelectrodes) and amplifier. Saturation was determined as a
function of PO2 by analysis of the
optical spectrum collected at every
PO2 data point (26). Colloidal
osmotic (oncotic) pressure was measured using a colloid osmometer
(model 4420, Wescor, Logan, UT). Average molecular weight was
calculated according to general thermodynamic analysis, as reported
elsewhere (25). Viscosity was measured at 37°C using a capillary
viscometer (22). According to the specifications of the products,
PEG-Hb and -Hb contain 0.3 endotoxin units per milliliter of
endotoxin. Starting methemoglobin levels were <5.0% in each of the
solutions.
Autoxidation kinetics for -Hb and PEG-Hb were measured by
following the absorbance changes that accompany the conversion of
oxyhemoglobin to methemoglobin for air-equilibrated samples in 0.1 M
bis-Tris propane-0.1 M NaCl (pH 7.4, 37°C). Data were analyzed as
single-wavelength absorbance changes at 401 and 420 nm and by
multicomponent analysis of the visible spectrum between 480 and 650 nm
for the fraction of methemoglobin. Time courses were fitted to
single-exponential expressions, and results were similar at both
wavelengths or by multicomponent analysis.
Rates of heme loss were determined using fully oxidized hemoglobins.
Rates were measured using a doubly mutated sperm whale apomyoglobin
(H64Y/V68F, a kind gift from J. S. Olson, Rice University) as a heme
acceptor (8). The rate-limiting step in this reaction is heme
dissociation, and the time course of heme exchange between methemoglobin and the apomyoglobin reflects the rate of heme loss from
methemoglobin. Rates were measured in 0.15 M phosphate-0.45 M sucrose
(pH 7.0, 37°C). All proteins exhibited biphasic kinetic behavior,
with the kinetic phases displaying equal amplitudes due to chain
differences. According to Hargrove et al. (8), the rapid and slow
kinetic phases for unmodified HbA0
correspond to - and 
-subunits, respectively.
Experimental animals.
Groups of male Sprague-Dawley rats (210-350 g; Charles River) were
studied. Control animals were treated exactly as the experimental animals, except no exchange transfusion was performed before the hemorrhage period. In the test animals, exchange transfusion was carried out according to the protocol described below with PS, PEG-Hb,
or -Hb. Because of the small size of these animals, the amount of
blood taken for analysis was strictly controlled. Therefore, separate
animals were used for 1) blood
pressure and acid-base measurements,
2) ,
3) blood volume determinations, or
4) arterial and mixed venous
blood-gas content.
Surgical preparation of animals.
All animal protocols were approved by the Animal Care Committee of the
San Diego Veterans Affairs Medical Center. Animals were fed ad libitum
before all experiments. Rats were anesthetized with 250 µl of a
mixture of ketamine (71 mg/ml; Aveco, Fort Dodge, IA), acepromazine
(2.85 mg/ml; Fermenta, Kansas City, MO), and xylazine (2.85 mg/ml;
Lloyd Laboratories, Shenandoah, IA). The area of the left femoral
artery and vein was exposed by blunt dissection, and a polyethylene
catheter (PE-50) was placed into the abdominal aorta via the femoral
artery to allow rapid withdrawal of arterial blood. An identical
catheter was placed in the inferior vena cava via the femoral vein to
allow injection of the Evans blue dye in blood volume experiments.
Catheters were tunneled subcutaneously, exteriorized through the tail,
and flushed with ~100 µl of normal saline. Animals were allowed to
recover from the procedure and remained in their cages for an
additional 24 h before being used in experiments.
Blood volume.
The method used to measure plasma and blood volume has been described
previously (15). Briefly, a small amount of Evans blue dye is
administered to the test animal by injection into a femoral vein
catheter, and blood is sampled from the contralateral femoral artery at
5, 10, 15, and 20 min after the injection. Plasma is separated by
centrifugation, and the visible optical spectrum is recorded. After a
baseline correction of the absorbance at 620 nm for the presence of
plasma hemoglobin, linear regression of the absorbance values is done
to obtain the extrapolated zero-time value. This absorbance value is
used to calculate the dilution relative to the starting concentration
of Evans blue dye, and with the measured hematocrit and appropriate
corrections for plasma trapping and the
"Fcell" ratio,
the total body-to-venous hematocrit ratio, total blood volume is
calculated.
Hemodynamic measurements.
For the hemodynamic measurements, the femoral artery catheter was
connected, through a stopcock, to a pressure transducer (model 1050, UFI, Morro, CA), and arterial pressure was sampled continuously at 100 Hz using a data-collection system (model MP100WSW, BIOPAC Systems,
Goleta, CA). The data were stored in digital form for subsequent
off-line analysis. The arterial pressure signal was analyzed using a
program written in FORTRAN. Heart rate for each beat was calculated as
the reciprocal of the interval between successive pressure peaks.
Systolic and diastolic pressures were the maximum and minimum
pressures, respectively, and the mean arterial pressure (MAP) was
diastolic + 
(systolic 
diastolic) pressure. Myocardial
contractility (dP/dt) was calculated from the maximum positive slope for each pressure cycle. Mean values of
heart rate, systolic and diastolic pressures, MAP, pulse pressure, and
dP/dt were averaged for each minute of
data.
was measured by injection of cold saline via the
jugular vein. Mixing occurs in the pulmonary capillary beds, and
was measured by a thermodilution catheter and
computer (Columbia Instruments, Columbus, OH).
Systemic vascular resistance (SVR) was calculated as
MAP/.
O2 transport.
Because of the volume of blood needed to analyze
O2 and
CO2 content, a separate group of
animals was used for measurements of
O2 transport. In these
experiments, hemodilution and subsequent hemorrhage were carried out as
with the other animals, but at intervals blood samples were removed
simultaneously from the femoral artery and the inferior vena cava.
These were analyzed for blood gases (pH,
PO2,
PCO2), hematologic quantities (hematocrit, hemoglobin, plasma hemoglobin), and
O2 content
(Lex-O2-Con, Hospex).
Blood-gas, hematologic, and lactate measurements.
Arterial pH, PCO2, and
PO2 were measured in a blood-gas
analyzer (model ABL5, Radiometer) using 100-µl samples of heparinized
blood. Lactic acid was measured in femoral artery blood using a lactate
analyzer (model 1500-L, Yellow Springs Instruments). Total
CO2 and base excess were
calculated from PCO2, pH, and
hemoglobin concentration using algorithms described previously (30).
Total hemoglobin concentration was measured with a -hemoglobin photometer (HemoCue, Mission Viejo, CA) using 50-µl samples of blood
collected from the femoral artery. Hematocrit was measured using
~50-µl samples of arterial blood by microcentrifugation.
Exchange transfusion.
Fully conscious animals were placed in Plexiglas restrainers. The
arterial and venous cannulas were flushed with 200 and 100 µl,
respectively, of heparinized saline (100 U/ml). The arterial and venous
catheters were connected to an infusion pump (model 4262000, Labconco,
Kansas City, MO), and exchange transfusion was carried out at a rate of
~2 ml/min to a total volume of solution that equaled 50% of
estimated blood volume. The peristaltic pump was operated so that blood
was removed at exactly the same rate at which test material was
infused. Test solutions were filtered through a 0.22-µm filter
immediately before infusion, and the infusate tubing passed through a
37°C water bath.
Hemorrhage.
The hemorrhage protocol was begun ~10 min after completion of the
exchange transfusion by pumping out arterial blood from the femoral
artery at a rate of 0.5 ml/min. The pump was started at the beginning
of each 10-min period and run for a time calculated to complete the
removal of 60% of the blood volume by the end of 60 min. The total
blood volume (TBV) at the end of each 10-min period is
|
(1)
|
where
TBV0 is the initial blood volume,
assumed to be 60 ml/kg (15), and t is
time (minutes). Removal of the required quantities of blood was
accomplished by running the pump for a period of time corresponding to
the amount of blood to be removed divided by the pump rate, i.e., 0.5 ml/min. At the end of the 60-min hemorrhage period, the animals were
monitored for an additional 1 h before euthanasia. Blood samples were
taken every 10 min, and the volume of blood removed was added to that
removed by the pump to adhere strictly to the hemorrhage protocol.
Statistical and survival analysis.
Statistical and survival analyses were done using PROPHET (Bolt,
Baranek and Newman, Cambridge, MA). Unless otherwise noted, values are
means ± SE. Differences between means were considered significant
at P < 0.05. For the survival
analyses, animals were observed for a minimum of 120 min after the
start of the hemorrhage, and survival data were analyzed according to
the method of Kaplan and Meier (16). The elapsed times from the start
of hemorrhage to death or censoring for each animal were recorded. Data
were considered censored if the animal was alive at 120 min. The data were grouped into 10-min intervals, and for each interval the cumulative proportion alive and its standard error were calculated. Significance of survival differences between groups was analyzed using
the Mantel-Cox test of significance (17).
Because the amount of blood needed for some determinations would
significantly influence the hemorrhage experiments, some calculated
values [SVR, stroke volume (SV),
O2,
O2 consumption (
O2)] were calculated
from two measurements in different groups of animals. Thus, when a
quantity C (e.g., SVR) is calculated as the product of two sets of data A
(e.g., MAP) and B (e.g., 1/), standard errors of the quantity
C were calculated by analysis of the
variances in the quantities A and
B, and the
t-statistic for the differences
between sets of C (i.e.,
C1 and
C2) was
calculated from the pooled estimate of the standard deviations. These
procedures are documented in the PROPHET system.
| |
RESULTS |
Test solutions.
Table 1 gives some physical properties of
the three test solutions and compares them with rat blood.
Note that the O2 half-saturation pressure of hemoglobin (P50) of
PEG-Hb and its Hill coefficient are much lower than the values for
blood or for -Hb. The molecular weight given for PS is a mean
weight-average value (Pentaspan prescribing information, DuPont Merck).
The oncotic pressure of PEG-Hb is much higher than that of rat or human
blood and higher than that of -Hb. The viscosities of PEG-Hb and
PS are about equal to that of blood, whereas the viscosity of -Hb
is that of water.
Hematologic measurements.
The hematologic measurements are shown in Table
2. Hematocrit, total hemoglobin, and plasma
hemoglobin were measured at baseline, immediately after exchange
transfusion, and at 30, 60, and 120 min after the start of hemorrhage.
The drop in hematocrit after hemodilution is greatest in the animals
treated with PEG-Hb, presumably because of its greater oncotic pressure
and, therefore, hemodilution (Table 1). As expected, total hemoglobin
follows hematocrit closely in all animals. Plasma hemoglobin, however,
differs significantly between the two cell-free hemoglobin groups:
-Hb and PEG-Hb animals. In the former, plasma hemoglobin reaches
a maximum of 3.9 g/dl after hemodilution; in the latter it reaches only
1.9 g/dl. Also the disappearance of plasma hemoglobin is much faster in
the -Hb than in the PEG-Hb animals (3.6 and 8.9 h, respectively). The lower concentration in the PEG-Hb animals is partly due to the
lower hemoglobin concentration of the stock solution (Table 1) and
partly to the greater blood volume expansion (15).
Blood volume.
The effect of exchange transfusion on blood volume is presented in
Table 3. PS and PEG-Hb expanded the blood
volume to almost the same extent, whereas exchange transfusion with
-Hb contracted the blood volume significantly. The difference in
volume expansion caused by exchange transfusion with PS compared with
PEG-Hb is not statistically significant
(P = 0.625); however, the contraction after hemodilution with -Hb compared with both PS and PEG-Hb is
highly significant (P < 0.001 in
both cases). As has been reported previously (15), the volume expansion
consequent to exchange transfusion with PEG-Hb, or PS in this
experiment, is temporary, lasting only ~2 h after the exchange. The
plasma volume is a dynamic quantity (15); after an initial expansion it
gradually contracts after administration of PEG-Hb, as the hemoglobin
is cleared from the circulation.
Survival.
Overall survival in the various groups of animals is shown graphically
in Fig. 1, and the mean survival times are
given in Table 4. In untreated controls the
survival was 46% at 2 h after the hemorrhage was started. The overall
survival is worsened after 50% exchange transfusion with PS (8%) and
-Hb (6%) but markedly improved in the PEG-Hb animals (93%; Fig.
1). Differences in survival between groups are all statistically
significant, except for the -Hb-PS comparison, which is not
significant (P = 0.15). Note in Table
4 the lack of relationship between survival and postexchange hematocrit
or hemoglobin concentration. In particular, the PEG-Hb animals had
markedly improved survival even compared with the control group, with a
mean hemoglobin concentration only about one-half that of the controls.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 1.
Overall survival of animals during and after 60% hemorrhage.
n, No. of rats. Control animals or those exchange transfused
with pentastarch (PS) or human hemoglobin cross-linked between
-chains (-Hb) have significantly greater mortality than
animals treated with polyethylene glycol-modified bovine hemoglobin
(PEG-Hb). Data from 3 different experimental protocols (hemodynamic and
acid-base measurements, cardiac output, and
O2 transport) were pooled for this
analysis.
|
|
Hemodynamics.
MAP values are shown in Fig. 2. Exchange
transfusion with -Hb produces a mean rise from 110.8 ± 2.9 to
125.6 ± 4.2 mmHg (difference, P < 0.02). A mean drop from 114.1 ± 4.6 to 100.5 ± 5.0 mmHg was
observed during exchange transfusion in the PS animals, but the
difference was not statistically significant (P = 0.1). No significant change was
observed between baseline and postexchange periods in the PEG-Hb
animals, although Fig. 2 shows a small rise immediately after the
beginning of the exchange. After initiation of hemorrhage, MAP fell
immediately in the control and -Hb animals, possibly because of
their contracted blood volumes (Table 3). The fall in MAP is much
slower in the PS animals and insignificant in those receiving PEG-Hb.
The immediate fall in pressure in the control animals was
followed by a gradual increase, but values never returned to baseline
levels.
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 2.
Mean arterial blood pressure (MAP) during exchange transfusion (ET) and
60% hemorrhage. Vertical lines mark beginning of exchange transfusion
and hemorrhage periods.
|
|
Over the course of the hemorrhage and subsequent observation period,
the MAP was very sensitive to the withdrawal of blood (each 10 min) in
the control animals but was more stable in the PS and PEG-Hb animals.
Again, this could be due to the greater initial blood volume in those
two groups of animals. After an initial fall in the -Hb animals,
a prompt rise was observed, reaching a maximum of 129.1 ± 4.1 mmHg
at 50 min of hemorrhage. In fact, the MAP in the -Hb animals did
not fall much below the baseline, showing that maintenance of the MAP
is not a good correlate with overall survival. MAP values in the PEG-Hb
animals did not fall after initiation of the hemorrhage. Rather, a
gradual decrease was observed over the entire duration of the
experiment, reaching a minimum MAP of 89.1 ± 5.2 mmHg at 105 min
after the start of hemorrhage.
Heart rate.
The heart rate changes are shown in Fig. 3.
The baseline rate in the control animals was 425 ± 2.9 (SE)
beats/min. At the start of hemorrhage, the rate dropped slightly, but
by the end of the hemorrhage the rate had increased to 495 ± 2.0 beats/min, significantly faster than the baseline rate
(P < 0.001). At the end of the
hemorrhage, the PS animals had a mean heart rate of 495 ± 2.0 beats/min, a rate significantly faster than the PS baseline of 408 ± 5.3 beats/min (P < 0.001) but
not significantly different from the controls at the end of hemorrhage.
In contrast, exchange transfusion with -Hb led to a significant
drop in the heart rate (434 ± 3.9 to 345 ± 1.8 beats/min,
P < 0.001). By the end of the
exchange transfusion the heart rate in the -Hb animals had
returned to a value not significantly different from baseline and then
increased slightly during the hemorrhage and subsequent observation
period. Heart rate was fairly stable in the PEG-Hb animals, despite a
slight but significant fall after exchange transfusion (from 427 ± 1.6 to 383 ± 1.0 beats/min, P < 0.001) and a slightly lower rate (412 ± 1.1 beats/min,
P < 0.001) at the end of hemorrhage.
View larger version (47K):
[in this window]
[in a new window]
|
Fig. 3.
Mean heart rate (HR; beats/min) during exchange transfusion and 60%
hemorrhage. Vertical lines mark beginning of exchange transfusion and
hemorrhage periods.
|
|
dP/dt.
dP/dt, the maximum positive slope of
the pulse contour, is shown in Fig. 4. In
the control animals, this value increases within ~15 min after the
beginning of the hemorrhage, rising significantly (P < 0.001) from a baseline of 1,336 ± 8.2 to 1,513 ± 6.7 mmHg/s before returning to baseline
shortly before death of the animals. In the PS animals, exchange
transfusion produces a slight but significant
(P < 0.001) rise in
dP/dt from 1,264 ± 7.2 to 1,328 ± 8.9 mmHg/s. Hemorrhage leads to a further significant
(P < 0.005) rise in
dP/dt to 1,673 ± 7.1 mmHg/s, which
is prolonged compared with the control animals. Exchange transfusion
with -Hb produces a small but significant
(P < 0.001) rise in
dP/dt from 1,211 ± 9.0 to 1,257 ± 8.9 mmHg/s and then a significant
(P < 0.02) fall in
dP/dt to 825 ± 21.0 mmHg/s during
hemorrhage. Finally, exchange transfusion with PEG-Hb does not produce
a significant change in dP/dt from the
baseline value of 1,094 ± 9.0 mmHg/s, but there is a gradual but
significant (P < 0.001) rise during
and after hemorrhage to 1,405 ± 5.1 mmHg/s. In terms of the
absolute change from baseline, the control, PS, and PEG-Hb treatments
produce about the same rise, but the rise is delayed and prolonged in the PEG-Hb animals compared with the other two groups.
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 4.
Mean myocardial contractility
(dP/dt) during exchange transfusion
and 60% hemorrhage. Vertical lines mark beginning of exchange
transfusion and hemorrhage periods.
|
|
.
measurements by the direct thermodilution technique
are shown in Fig. 5. After exchange
transfusion with both PEG-Hb and PS, increases in
parallel with the blood volume expansion shown in Table 3. However,
because of the large variation in individual measurements, only the
rise in PEG-Hb animals was significant when subjected to a paired
t-test
(P = 0.031). The fall in
in the control and -Hb animals is prompt and
significant at all time points (P < 0.001). also falls in the PS animals relative to
baseline at 
20 min after hemorrhage
(P < 0.002). In the PEG-Hb animals,
is never significantly different from the baseline values (except immediately after the exchange, as noted) even at 120 min after the start of the hemorrhage. The principal component of the
changes in is the SV (Fig.
6). The changes follow closely the changes
in (Fig. 5). Of particular interest, the SV in the
PEG-Hb animals is essentially the same as baseline after the end of the
hemorrhage.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 6.
Cardiac stroke volume (SV). Values follow very closely
, since changes in heart rate are relatively small.
|
|
Figure 7 shows the SVR. Some variation in
the data is caused by the need to use data from different groups of
animals for the and MAP values. Even with this
limitation, the difference between the PEG-Hb, control, and -Hb
experiments is striking. During the course of the hemorrhage, the SVR
more than quadruples in the controls as the MAP drops significantly.
The SVR in the PEG-Hb animals remains at prehemorrhage levels and
doubles in the PS animals.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 7.
Systemic vascular resistance (SVR, MAP/) rises
sharply during hemorrhage in control and -Hb animals. Rise is
much less in PS animals and is not detectable in PEG-Hb animals. Error
bars for PEG-Hb animals are within limits depicted by symbols.
|
|
O2 transport.
Figure 8 presents
O2,
the product of and arterial
O2 content
(CaO2).
O2
falls for all three dilution groups after exchange transfusion.
O2
continues to fall markedly in the PS, -Hb, and control groups
after the initiation of the hemorrhage, whereas it remains stable at
~50% of baseline in the PEG-Hb animals. This is explained by the
increased in the latter group (Fig. 5). In the
control, -Hb, and PS animals, the
O2
values are virtually identical after the first 10 min of hemorrhage.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 8.
Total O2 transport
(O2,
× arterial
O2 content) plotted against time
after start of hemorrhage.
O2
falls in all 3 hemodiluted groups of animals but remains highest in
PEG-Hb group because of higher (see Fig. 5). Despite
a higher total and plasma hemoglobin concentration in -Hb animals
(Table 2),
O2
falls rapidly after start of hemorrhage because of relatively
low .
|
|
Although plasma methemoglobin was not measured directly in these
studies, the O2 transport data
also permit the calculation of the amount of
O2 transported by the plasma
hemoglobin fraction. Thus, from the measured
PO2,
PCO2, and pH, red cell saturation can
be calculated (30), and red cell
O2 content can then be obtained as
the product of red cell saturation and red cell
O2 content. The difference between
red cell O2 content and the
measured CaO2 is plasma hemoglobin
O2 content and, when divided by
plasma hemoglobin O2 capacity
(plasma Hb × 1.34), gives plasma hemoglobin saturation. This
calculation yields ~75% plasma hemoglobin saturation for the
-Hb and PEG-Hb animals and provides evidence that there can be no
more than 25% methemoglobin in either of the plasma hemoglobin
fractions.
The O2 extraction ratio
[(CaO2 CvO2)/CaO2,
where CvO2 is venous
O2 content] is shown in Fig.
9. This index is apparently sensitive to
changes in overall O2 transport,
since there is a significant rise in the PS and -Hb animals with
exchange transfusion alone. There is no significant change in the
PEG-Hb animals. In the -Hb animals the ratio rises to ~100%
almost immediately after the start of the hemorrhage, whereas the rise,
while still marked, is slightly less in the PS and control animals. The
rise in the ratio is more gradual in the PEG-Hb animals and does not exceed ~75% of
O2
at any time during the hemorrhage observation period.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 9.
O2 extraction ratio [OER,
(CaO2 CvO2)/CaO2,
where CaO2 and
CvO2 are arterial and venous
O2 contents, respectively.]
Note rapid rise to nearly complete extraction in -Hb animals.
|
|
O2 (Fig.
10) falls significantly and rapidly after
the start of the hemorrhage in the control, PS, and -Hb animals.
A small drop in O2 is seen
after the exchange transfusion in the PEG-Hb animals, but recovery is
complete by the end of the first 10-min hemorrhage period. Only the
PEG-Hb animals are able to maintain stable
O2 values during the entire
experiment.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 10.
O2 consumption
[O2, × (CaO2 CvO2)]. Note significant drops in
O2 in control, -Hb, and
PS, but not in PEG-Hb, animals.
|
|
O2 as a function of
O2
is shown in Fig. 11. It is striking that
none of the PEG-Hb O2 values
falls below the "critical" O2
of ~20
ml · min
1 · kg1,
whereas in each of the other groups of animals it declines below this
point during the hemorrhage.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 11.
O2 supply dependence. Data from
all experimental groups are plotted. In later stages of hemorrhage,
O2 becomes supply dependent
in control, -Hb, and PS animals. Even at lowest level of
hemorrhage, PEG-Hb animals were not supply dependent.
|
|
Respiratory and acid-base response to hemorrhage.
The respiratory and acid-base response to hemorrhage is in proportion
to overall survival. Thus the PEG-Hb animals demonstrate little rise in
PO2 or fall in
PCO2 or base excess and maintain
their arterial pH essentially constant throughout the experiments (Fig.
12). In contrast, the -Hb animals
hyperventilated significantly, as demonstrated by the rise in
PO2 and fall in
PCO2. This response did not
compensate for the restricted
O2,
however, and these animals showed a dramatic fall in base excess and
eventually unstable pH regulation. The control and the PS animals were
intermediate with regard to these blood-gas values, with the PS animals
showing less disturbance than controls.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 12.
Acid-base changes during exchange transfusion, 60% exponential
hemorrhage, and for 1 h thereafter. Almost no detectable departure from
baseline values is seen in PEG-Hb animals for any of these parameters,
whereas control, PS, and -Hb animals demonstrate a rise in
PO2, a decrease in
PCO2, greater base deficit, and
acidosis during experimental period. These results are consistent with
hyperventilation in latter 3 groups, but not PEG-Hb animals. BE, base
excess.
|
|
The greatest perturbation of respiratory and acid-base regulation is
seen in the animals exchange transfused with -Hb, despite the
fact that the -Hb animals have the same hematocrit as the PEG-Hb
animals (Table 2) and a higher total hemoglobin concentration than the
PS or PEG-Hb animals (Table 2).
A separate measure of the effectiveness of
O2 transport and tissue
ischemia is demonstrated by the arterial lactate measurements shown in Fig. 13. The PS animals
demonstrate higher lactate levels than the controls, presumably because
of their lower hemoglobin concentrations. The -Hb animals show
the highest accumulation, despite their higher total hemoglobin
concentration. The PEG-Hb animals produce less lactate than the
controls, despite a significantly lower total hemoglobin concentration.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 13.
Lactic acid accumulation during exchange transfusion, 60% exponential
hemorrhage, and for 1 h thereafter. Lactate accumulation is
significantly greater in -Hb, PS, and control (in that order)
than in PEG-Hb animals.
|
|
| |
DISCUSSION |
We have described a severe hemorrhage model, which has been chosen
because only such a model has the potential to demonstrate the efficacy
of cell-free hemoglobin. With no intervention, our 60% exponential
hemorrhage protocol is lethal in ~58% of rats 1 h after completion
of the hemorrhage. Thus improvement in survival to 93% after a 50%
(by volume) hemodilution with PEG-Hb is especially noteworthy,
considering that the PEG-Hb animals began the hemorrhage with a total
hemoglobin concentration of only 7.6 g/dl compared with 13.8 g/dl in
the controls (Table 2). This is the first demonstration of which we are
aware in which cell-free hemoglobin is more effective in preventing the
consequences of severe hemorrhage than red blood cells alone. This
model is also extremely relevant to the clinical development of red
cell substitutes, because to be efficacious the effects of any proposed
new agent will have to be clearly tied to some positive clinical
outcome, such as survival. Moreover, removal of autologous blood before
surgery and replacement with a substitute represent one principally
planned application for these solutions.
Solution properties.
It is not possible to obtain solutions for study as red cell
substitutes that differ from each other such that the influence of
isolated variables can be clearly understood. The solutions we have
chosen in the present study do allow some simple comparisons, however.
For example, -Hb and PEG-Hb have different solution and
functional properties. The comparison between PEG-Hb and PS is of
interest, because both of these solutions have equivalent oncotic
pressures and viscosities (Table 1). They differ primarily in that only
PEG-Hb carries bound O2. That the
effects of PEG-Hb are not due only to its high oncotic pressure and
viscosity is shown clearly by its effect on survival, lactic acid
concentration, , and
O2 compared with PS.
-Hb and PEG-Hb carry O2, yet they differ strikingly in their
O2-binding properties and
physiological effects. More favorable results were obtained with
PEG-Hb, even though its O2
affinity is higher than that of -Hb and its plasma concentration
was lower in our animals (Table 2).
The PEG-Hb and PS solutions have increased oncotic pressure and
viscosity. The extent to which increased viscosity of the solution
relative to -Hb plays a role is not completely clear. A
theoretical examination of this question by Intaglietta and co-workers
(13) suggests that viscosity may be a significant factor, but PEG-Hb
and PS are matched for viscosity and yet have markedly different
effects on survival, lactic acid production, and global
O2.
Physiological response.
Hyperventilation is a normal response to hemorrhage (7). This is
presumed to be mediated by protons produced in tissues as a consequence
of reduced
O2.
Although the initial effect of this
H+ production is not detectable as
a change in arterial pH, the respiratory center in the brain is highly
sensitive to it, and hyperventilation occurs almost immediately after
the onset of hemorrhage.
PEG-Hb is the most effective blood replacement in this model, as
demonstrated by the virtual lack of response in pH,
PCO2, PO2, and base excess. In other words,
animals maintain essentially perfect acid-base homeostasis, despite the
loss of an estimated 60% of their total blood volume by hemorrhage.
However, most striking is the fact that animals exchange transfused
with PEG-Hb maintain their acid-base balance and survive as well as untreated controls, even though the total hemoglobin concentration in
these animals is much less.
It appears that the best indicator of mortality in our animals is their
inability to compensate for the acidosis resulting from insufficient
O2
to tissues. Hannon and co-workers (7), working with the US Army, found
that disordered acid-base balance (as measured by base excess and
PCO2, for example) is the best
predictor of death in hemorrhagic pigs. Our data clearly show that
PEG-Hb improves the acid-base balance, as judged by arterial
PCO2, pH, base excess, and lactic
acid accumulation, relative to control, PS, and -Hb animals. A
significant part of this effect may be due to the plasma expansion
properties of PS and PEG-Hb, but it cannot be the total explanation,
because the two are matched for oncotic pressure, and yet their effects on blood volume are the same.
The significantly reduced effect on arterial
PCO2 in the PEG-Hb animals indicates
that the drive to hyperventilate in these animals is lower than in the
control or the PS animals. Although this could conceivably be due to a
central depressive effect, it is more likely due to improved
O2.
This conclusion is supported by the better survival and lactic acid
response in the PEG-Hb animals than in the control or the PS group.
Maintained O2 and reduced
lactic acid accumulation are striking features of the PEG-Hb animals.
Lactic acid accumulation is a measure of
O2 debt and has been shown to be
related to survival in humans with hypovolemic shock by a number of
studies. Broder and Weil (2) showed that only 11% of shock patients
survive when the lactic acid production is >4 meq/l, and Weil and
Afifi (28) showed that, as lactic acid concentration rises from 2 to 8 meq/l, the survival drops from 90 to 10%. In these cited studies,
lactic acid production is presumably due to hypoperfusion and
O2 debt (18) due to decreased
O2.
Vasoconstriction may severely limit
O2 supply. Indeed, the control,
-Hb, and PS animals demonstrated reduced
during the hemorrhage, whereas the PEG-Hb animals did not (Fig. 5).
Figure 11 shows that, in the later stages of the hemorrhage,
O2
is still not supply limited in the PEG-Hb animals, whereas it clearly
is in the other three groups. This is primarily because
remains high in the PEG-Hb animals (Fig. 5). Figure
2 shows that neither the PEG-Hb nor the PS solution raises the resting
MAP in rats during or after the 50% exchange transfusion. In the same
model, -Hb produces an immediate rise in MAP of ~20 mmHg.
Furthermore, there is only a slow, slight drop in MAP during the 60%
hemorrhage in the PEG-Hb animals but a significant drop in the PS
animals. Thus SVR remains low in the PEG-Hb animals relative to the
other three groups (Fig. 7). Our hypothesis is that this lack of
vasoactivity, along with the volume-expanding properties of the
hyperoncotic colloids (PEG-Hb and PS), permits efficient perfusion of
tissues and unloading of O2 in
capillary beds.
The improved survival with PEG-Hb exchange transfusion is all the more
striking, since the hematocrit is only 15.8% and plasma PEG-Hb
concentration is 1.9 g/dl. Moreover, the small amount of PEG-Hb that is
present has a P50 of only 10 Torr.
Our results are in contrast to a similar study using red blood cells
(4) in which significantly reduced survival was found in animals in which anemia and increased red cell
O2 affinity were observed at the
start of the hemorrhage. A key conclusion from our study, therefore, is
that critical properties of red cell substitutes, such as
O2 affinity,
O2 capacity, and oncotic pressure,
may be different from critical properties of red blood cells. The
hierarchy of critical properties of cell-free red cell substitutes in
still unclear. For example, P50
seems to be less important than oncotic pressure and, possibly,
viscosity. Only further experiments with clearly defined solutions will
define the order of importance.
The mechanism of vasoactivity of hemoglobin solutions remains to be
clarified. We show here that PEG-Hb has essentially no effect on MAP
and vascular resistance, even though its intrinsic reactivity with NO
is almost identical to that of -Hb (23). Thus the widely held
opinion that NO scavenging is responsible for the vasoactivity of
hemoglobin solutions (3) may be an oversimplification. We have proposed
separately that autoregulation of
O2 supply in the microcirculation
may play a role at least as important as NO scavenging (13, 27, 35).
The precise biochemical mechanism for such regulation remains a
mystery, but one candidate might be ATP-sensitive
K+ channels, which are known to be
sensitive to O2 supply (19). Alternatively, -Hb and PEG-Hb may extravasate at different rates and thus scavenge subluminal NO to different extents.
Implications for design of red cell substitutes.
An important observation from our experiments is the lack of
correlation between a pressor effect and survival. In fact, the group
with the best pressor effect, -Hb animals, had the poorest survival. This is in opposition to the reported benefits of rapid restoration of blood pressure with the commercial product DCLHb (24).
However, our results may not be strictly comparable, since our protocol
is different (hemodilution vs. resuscitation) and since, whether
-Hb is the same product as DCLHb, even though they contain the
same chemical modification, has never been resolved in the literature.
However, there is no question that, in our protocol, -Hb produces
detrimental results, even compared with the controls: lactic acid
production is higher, acid-base disturbances are more pronounced, and
survival is worse.
It would appear that a major goal of red cell substitute design should
be to maintain blood volume and support circulation with low vascular
resistance. Our data suggest that this can be accomplished by
optimization of the key properties of oncotic pressure, viscosity, and
O2 binding and capacity.
| |
ACKNOWLEDGEMENTS |
The authors acknowledge with gratitude the expert assistance of
Renée Schad in the preparation of the manuscript.
| |
FOOTNOTES |
This work was supported by National Heart, Lung, and Blood Institute
Grant HL-48018 and an unrestricted gift by Enzon to the Medicine,
Education, and Research Foundation at the University of California, San
Diego.
A preliminary report of these results has been presented in abstract
form (33).
Address for reprint requests: R. M. Winslow, VA Medical Center (111-E),
3350 La Jolla Village Dr., San Diego, CA 92161.
Received 10 November 1997; accepted in final form 18 May 1998.
| |
REFERENCES |
1.
Biro, G. P.,
P. J. Anderson,
S. E. Curtis,
and
S. M. Cain.
Stroma-free hemoglobin: its presence in plasma does not improve oxygen supply to the resting hindlimb vascular bed of hemodiluted dogs.
Can. J. Physiol. Pharmacol.
69:
1656-1662,
1991[Medline].
2.
Broder, G.,
and
M. H. Weil.
Excess lactate: an index of reversibility of shock in human patients.
Science
143:
1457-1459,
1964[Abstract/Free Full Text].
3.
Cases, A.,
J. M. Stulak,
Z. Katusic,
E. Villa,
and
J. C. Romero.
Hemodynamic and renal effects of cross-linked hemoglobin infusion.
Am. J. Physiol.
272 (Regulatory Integrative Comp. Physiol. 41):
R793-R799,
1997[Abstract/Free Full Text].
4.
Collins, J. A.,
and
L. Stechenberg.
The effects of the concentration and function of hemoglobin on the survival of rats after hemorrhage.
Surgery
85:
412-418,
1973.
5.
Federspiel, W. J.,
and
A. S. Popel.
A theoretical analysis of the effect of the particulate nature of blood on oxygen release in capillaries.
Microvasc. Res.
32:
164-189,
1986[Medline].
6.
Gibson, Q. H.
The kinetics of reactions between haemoglobin and gasses.
In: Progress in Biophysics and Biophysical Chemistry, edited by J. A. V. Butler,
and B. Katz. New York: Pergamon, 1959, p. 1-54.
7.
Hannon, J. P.,
C. E. Wade,
C. A. Bossone,
M. M. Hunt,
R. I. Coppes,
and
J. A. Loveday.
Blood gas and acid-base status of conscious pigs subjected to fixed-volume hemorrhage and resuscitated with hypertonic saline dextran.
Circ. Shock
32:
19-29,
1990[Medline].
8.
Hargrove, M. S.,
E. W. Singleton,
M. L. Quillin,
L. A. Ortiz,
G. H. Phillips, Jr.,
J. S. Olson,
and
A. J. Mathers.
His 64 E7 
Tyr apomyoglobin as a reagent for measuring rates of hemin dissociation.
J. Biol. Chem.
269:
4207-4214,
1994[Abstract/Free Full Text].
9.
Hess, J. R.,
V. W. Macdonald,
and
W. W. Brinkley.
Systemic and pulmonary hypertension after resuscitation with cell-free hemoglobin.
J. Appl. Physiol.
74:
1769-1778,
1993[Abstract/Free Full Text].
10.
Hogan, M. C.,
S. S. Kurdak,
R. S. Richardson,
and
P. D. Wagner.
Partial substitution of red blood cells with free hemoglobin does not improve maximal O2 uptake of working in situ dog muscle.
Adv. Exp. Med. Biol.
361:
375-378,
1994[Medline].
11.
Hogan, M. C.,
D. C. Wilford,
P. E. Keipert,
N. S. Faithfull,
and
P. D. Wagner.
Increased plasma O2 solubility improves O2 uptake of in situ dog muscle working maximally.
J. Appl. Physiol.
361:
2470-2475,
1992.
12.
Homer, L. D.,
P. K. Weathersby,
and
L. A. Kiesow.
Oxygen gradients between red blood cells in the microcirculation.
Microvasc. Res.
22:
308-323,
1981[Medline].
13.
Intaglietta, M.,
P. C. Johnson,
and
R. M. Winslow.
Microvascular and tissue oxygen distribution.
Cardiovasc. Res.
32:
632-643,
1996[Medline].
14.
Kasper, S. M.,
M. Walter,
F. Grune,
A. Bischoff,
H. Erasmi,
and
W. Buzello.
Effects of a hemoglobin-based oxygen carrier (HBOC-201) on hemodynamics and oxygen transport in patients undergoing preoperative hemodilution for elective abdominal aortic surgery.
Anesth. Analg.
83:
921-927,
1996[Abstract].
15.
Migita, R.,
A. Gonzales,
M. Gonzales,
K. D. Vandegriff,
and
R. M. Winslow.
Blood volume and cardiac index in rats after 50% exchange transfusion with hemoglobin-based oxygen carriers.
J. Appl. Physiol.
82:
1995-2002,
1997[Abstract/Free Full Text].
16.
Miller, R. G.
Survival Analysis. New York: Wiley, 1981, chapt. 3.
17.
Miller, R. G.
Survival Analysis. New York: Wiley, 1981, chapt. 5.
18.
Mizock, B. A.,
and
J. L. Falk.
Lactic acidosis in critical illness.
Crit. Care Med.
20:
80-93,
1992[Medline].
19.
Mouren, S.,
R. Souktani,
M. Beaussier,
L. Abdenour,
M. Arthaud,
M. Duvelleroy,
and
E. Vicaut.
Mechanisms of coronary vasoconstriction induced by high arterial tension.
Am. J. Physiol.
272 (Heart Circ. Physiol. 41):
H67-H75,
1997[Abstract/Free Full Text].
20.
Page, T. C.,
C. B. McKay,
W. R. Light III,
and
J. D. Hellums.
Experimental simulation of oxygen transport in microvessels.
In: Blood Substitutes: New Challenges, edited by R. M. Winslow,
K. D. Vandegriff,
and M. Intaglietta. Boston: Birkhäuser, 1996, p. 132-145.
21.
Palmer, R. M. J.,
A. Ferrige,
and
S. Moncada.
Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor.
Nature
327:
524-526,
1987[Medline].
22.
Reinhart, W. H.,
S. J. Danoff,
S. Usami,
and
S. Chien.
Rheologic measurements on small samples with a new capillary viscometer.
J. Lab. Clin. Med.
104:
921-931,
1984[Medline].
23.
Rohlfs, R. J.,
E. Bruner,
A. Gonzales,
M. L. Gonzales,
D. Magde,
M. D. Magde, Jr.,
K. D. Vandegriff,
and
R. M. Winslow.
Arterial blood pressure responses to cell-free hemoglobin solutions and the reaction with nitric oxide.
J. Biol. Chem.
273:
12128-12134,
1998[Abstract/Free Full Text].
24.
Schultz, S. C.,
I. N. Hamilton, Jr.,
and
D. S. Malcolm.
Use of base deficit to compare resuscitation with lactated Ringer's solution, Haemaccel, whole blood, and diaspirin cross-linked hemoglobin following hemorrhage in rats.
J. Trauma
35:
619-625,
1993[Medline].
25.
Vandegriff, K. D.,
M. McCarthy,
R. J. Rohlfs,
and
R. M. Winslow.
Colloid osmotic properties of modified hemoglobins: chemically cross-linked versus polyethylene glycol surface-conjugated.
Biophys. Chem.
69:
23-30,
1997[Medline].
26.
Vandegriff, K. D.,
R. J. Rohlfs,
M. D. Magde,
and
R. M. Winslow.
Hemoglobin-oxygen equilibrium curves measured during enzymatic oxygen consumption.
Anal. Biochem.
256:
107-116,
1998[Medline].
27.
Vandegriff, K. D.,
and
R. M. Winslow.
A theoretical analysis of oxygen transport: a new strategy for the design of hemoglobin-based red cell substitutes.
In: Blood Substitutes: Physiological Basis of Efficacy, edited by R. M. Winslow,
K. D. Vandegriff,
and M. Intaglietta. Boston: Birkhäuser, 1995, p. 143-154.
28.
Weil, M. H.,
and
A. A. Afifi.
Experimental and clinical studies on lactate and pyruvate as indicators of the severity of acute circulatory failure (shock).
Circulation
41:
989-1001,
1970[Abstract/Free Full Text].
29.
Winslow, R. M.
Hemoglobin-Based Red Cell Substitutes. Baltimore, MD: Johns Hopkins University Press, 1992.
30.
Winslow, R. M.
A model for red cell O2 uptake.
Int. J. Clin. Monit. Comput.
2:
81-93,
1985[Medline].
31.
Winslow, R. M.
Blood substitute oxygen carriers designed for clinical applications.
In: Blood Substitutes: New Challenges, edited by R. M. Winslow,
K. D. Vandegriff,
and M. Intaglietta. Boston: Birkhäuser, 1996, p. 60-73.
32.
Winslow, R. M.,
and
K. W. Chapman.
Pilot-scale preparation of hemoglobin solutions.
Methods Enzymol.
231:
3-16,
1994[Medline].
33.
Winslow, R. M., A. Gonzales, and M. Gonzales.
Physiologic effects of hemoglobin-based oxygen carriers compared
to red blood cells in acute hemorrhage (Abstract).
Blood 88, Suppl. 1: 626A, 1996.
34.
Winslow, R. M.,
M. L. Swenberg,
R. L. Berger,
R. I. Shrager,
M. Luzzana,
M. Samaja,
and
L. Rossi-Bernardi.
Oxygen equilibrium curve of normal human blood and its evaluation by Adair's equation.
J. Biol. Chem.
252:
2331-2337,
1977[Abstract/Free Full Text].
35.
Winslow, R. M.,
and
K. D. Vandegriff.
Hemoglobin oxygen affinity and the design of red cell substitutes.
In: Blood Substitutes: Industrial Opportunities and Medical Challenges, edited by R. M. Winslow,
K. D. Vandegriff,
and M. Intaglietta. Boston
Top
Abstract
Introduction
