Acute Hemodynamic Effects of Red Cell Volume

Reduction in Polycythemia of Cyanotic

Congenital Heart Disease
By AMNON ROSENTHAL, M.D., DAVID G. NATHAN, M.D., ALAN T. MARTY, M.D.,
LAWRENCE N. BUr-LON, B.S., OLLI S. M1hTFINEN, M.D., PH.D.,
AND ALEXANDER S. NADAS, M.D.

SUMMARY
Acute reduction in red cell volume (RCV) without significant alterations of
blood volume in 22 patients with severe polycythemia secondary to cyanotic con-
genital heart disease resulted in a decrease in peripheral vascular resistance and an
increase in stroke volume, systemic blood flow (SBF), and systemic oxygen transport.
These changes are probably related to the decreased blood viscosity and yield shear
stress associated with lower red cell concentrations. Hypervolemia in hypoxic poly-
cythemia should be maintained in order to sustain an adequate SBF. In contrast to
acute phlebotomy which may be expected to decrease blood oxygen content and
SBF, the replacement of whole blood with plasma or 5% albumin is shown to result
in an increased systemic blood flow and oxygen delivery.

Additional Indexing Words:

Blood viscosity Phlebotomy Cardiac output Exchange transfusion
Downloaded from http://ahajournals.org by on March 31, 2023

Hematocrit Hypoxemia Plasma

I N POLYCYTHEMIA accompanying cya-
notic congenital heart disease, both the
hematocrit (Hct) and the circulating whole
blood volume (WBV) are increased."2 The
hypervolemia results from an increase in red
cell volume (RCV). The increase in RCV
provoked by hypoxia provides an increased
oxygen-carrying capacity which maintains an
adequate oxygen supply to the tissues. Since
From the Department of Cardiology and the
Division of Hematology of the Department of
Medicine of the Children's Hospital Medical Center
and the Department of Pediatrics, Harvard Medical
School, Boston, Massachusetts.
Supported in part by Grants HE 10436-04, HD
02777 and AM 04481 from the National Institutes of
Health, U. S. Public Health Service and by grants
from the John A. Hartford Foundation, Inc., and
Baxter Laboratories, Morton Grove, Illinois. Dr.
Nathan is the recipient of U. S. Public Health Service
Career Development Award K3AM 35361.
Received January 6, 1970; revision accepted for
publication April 20, 1970.
Circulation, Volume XLII, August 1970
studies in dogs and man suggest that at any
given Hct the systemic blood flow (SBF) is
greater with hypervolemia than normovole-
mia,3-5 the induced hypervolemia probably
helps sustain a sufficient SBF despite the
increased viscosity that occurs at high hema-
tocrits.6 7 Polycythemia in cyanotic congenital
heart disease, however, is often of such
severity that it may become a liability and
produce adverse physiologic effects. The
clinical manifestations of headaches, irritabili-
ty, anorexia, and dyspnea have been attrib-
uted to polycythemia. Thrombotic lesions in
the lungs,8' 9 kidneys, and central nervous
system'0 have been found in polycythemic
patients. Postoperative hemorrhagic diathesis,
possibly a result of intravascular coagulation,
is also frequently observed in these individu-
als."' 12 These complications have been attrib-
uted to the elevated blood viscosity and
intravascular red cell aggregation that accom-
297
 298 ROSENTHAL ET AL.
Table 1
Acute Hemodynamic Effects of Erythropheresis in 22 Patients with Severe Polycythemia Secondary to Cyanotic
Congenital Heart Disease
Age/sex Height Weight vHct (%) Art. sat. HR (beats/min)
Subject (yr:mo) Diagnosis (cm) (kg) Before After Before After Before After
1 3:5/F D-TGA 90.0 11.8 66.0 61.0 74 73 115 108
2 17:6/F D-TGA, PAH 115.0 22.7 77.0 62.0 59 61 109 118
3 17:3/M L-TGA, PS, VSD 177.0 63.4 71.0 61.0 87 83 63 67
4 14:10/M D-TGA, PAH 138.0 21.8 77.0 61.0 71 72 116 107
5 20:6/M PA, VSD 158.5 48.2 74.5 56.5 89 82 90 84
6 7:8/M T/F 122.0 19.9 72.5 53.0 73 67 109 126
7 16:6/F L-TGA, PS, VSD 156.0 38.8 80.0 63.0 81 67 110 105
8 13:3/M D-TGA, PAH 145.0 31.4 72.0 64.0 80 79 87 82
9 13:6/F T/F 130.0 20.8 85.0 65.0 72 63 97 98
10 10:8/M D-TGA, VSD, PAH 125.0 19.7 67.5 53.5 76 65 90 95
11 7:3/M D-TGA, PAH 103.0 15.1 63.5 58.0 55 57 116 125
12 24:10/M PA, VSD 170.0 47.8 74.0 69.0 66 56 99 92
13 5:0/F TA, PS, VSD 95.5 11.4 73.0 62.0 82 78 82 83
14 4:3/M TA, PS, VSD 103.0 14.4 66.0 60.5 81 79 122 107
15 5:6/F T/F 113.5 16.3 74.5 61.0 84 78 120 118
16 25:0/M T/F 163.0 63.8 76.5 68.0 86 84 67 68
17 9:11/M D-TGA, VSD, PAH 118.0 19.8 73.0 61.0 78 74 90 100
18 17:0/M T/F 174.0 61.4 71.5 61.0 87 84 80 78
19 15:8/F PA, VSD 141.0 25.3 77.5 66.0 81 76 98 108
20 4:0/M D-TGA, PS, VSD 102.0 16.1 74.5 65.0 68 64 100 99
21 3:6/F A-V Canal, PS 97.5 13.9 76.0 63.0 70 72 137 132
22 16:11/M PA, VSD 176.0 62.3 75.0 67.0 73 59 92 96
Mean 73.5 62.0 76 72 100 100
Downloaded from http://ahajournals.org by on March 31, 2023

Standard deviation 5 4 9 9 18 18
P value by paired t test <0.001 0.5
Abbreviations: vHct = venous hematocrit; Art. sat. = arterial oxygen saturation; Vo2 = oxygen consumption; Par
- mean arterial pressure; HR = heart rate; Pcv = mean central venous pressure; SVR = systemic vascular resistance;
SBF = systemic blood flow; SI = stroke index; SOT = systemic oxygen transport; D-TGA = dextrotransposition of
great arteries; PAH = pulmonary artery hypertension; PS = pulmonary stenosis; VSD = ventricular septal defect; L-TGA
= levotransposition of great arteries; PA = pulmonary atresia; T/F = tetralogy of Fallot; TA = tricuspid atresia; A-V
= atrioventricular.

panies high red cell concentrations.6 Attempts Methods

at therapy by acute phlebotomy without
volume replacement may be followed by
vascular collapse and cerebrovascular acci-
dents probably due to the sudden reduction in
the blood volume and SBF. The circulatory
response to reduction in RCV and fluid
replacement in cyanotic polycythemic patients
has not been evaluated. This investigation
was, therefore, designed to determine the
acute hemodynamic effects of RCV reduction
and its replacement with fresh frozen plasma
or 5% solution of human albumin in severe
hypoxic polycythemia.
WBV to be removed = weight (L/kg) x 0.11 x (vHct, - vHctD)/vHctl,
Twenty-two patients with polycythemia secon-
dary to cyanotic congenital heart disease were
studied in the catheterization laboratory of the
Children's Hospital Medical Center, Boston.
Table 1 lists the pertinent clinical data on these
patients. Hemodynamic measurements were made
in the supine position after a 12-hour fast, and
prior to, and 1 to 2 hours after the replacement of
whole blood with fresh frozen plasma or 5%
human albumin in physiologic saline. The RCV
was reduced in an amount sufficient to give a
desired venous hematocrit (vHct) value. Thirty-
milliliter aliquots of whole blood were removed,
and fresh frozen plasma was infused in 30 to 60
minutes. The amount exchanged was calculated
from the following equation:
Circulation, Volume XLII, August 1970
 RED- CELL VOLUME REDUCTION IN POLYCYTHEMIA 299

V02 Par Pcv SVR SBF SI SOT

(ml/min/m2) (mm Hg) (mm Hg) (mm Hg/L/min/m2) (L/min/m2) (ml/beat/M2) (ml 02/min/kg)
Before After Before After Before After Before After Before After Before After Before After
151 151 85 80 4.0 5.5 28 19 2.9 3.9 25 36 32.2 36.7
121 138 94 82 2.0 5.0 44 19 2.1 4.0 19 34 12.7 28.2
134 148 86 81 5.5 4.0 24 16 3.4 4.9 54 73 26.2 32.5
127 145 69 71 1.0 1.0 21 17 3.2 4.1 28 38 25.1 27.9
131 129 85 80 6.0 10.0 42 23 1.9 3.0 21 36 15.2 19.4
157 137 84 68 3.0 2.0 21 10 3.8 6.7 35 53 32.5 41.6
178 170 86 92 5.0 6.0 18 13 4.6 6.5 42 62 39.2 38.9
172 158 79 83 5.0 5.0 23 17 3.2 4.6 37 56 23.5 25.7
108 110 86 68 8.0 8.0 31 9 2.5 6.5 36 66 27.1 48.0
139 158 75 82 2.0 3.0 16 7 4.6 10.9 51 115 30.9 49.6
161 170 90 88 5.0 7.0 12 8 7.0 10.3 60 82 39.9 56.0
134 134 94 101 9.0 14.0 39 20 2.9 4.3 29 47 18.1 19.4
131 150 86 101 8.0 9.0 17 15 4.6 6.3 56 76 52.6 61.0
156 176 86 88 7.0 9.0 18 13 4.3 6.0 35 56 44.4 53.4
133 142 92 87 4.0 6.0 33 19 2.7 4.2 23 36 31.0 36.2
125 125 85 78 6.0 6.0 26 21 3.1 3.5 46 51 20.0 21.4
120 126 84 68 6.0 8.0 33 21 2.4 2.9 27 29 22.8 24.2
124 113 81 82 3.5 5.0 24 19 3.2 4.1 40 53 22.6 25.9
116 121 103 100 5.5 7.0 46 34 2.1 2.7 21 25 22.0 22.5
160 156 70 80 6.0 7.0 23 9 3.1 7.8 31 79 29.2 56.0
114 112 90 85 4.5 5.5 50 23 1.7 3.5 12 27 18.1 34.8
93 83 84 84 15.0 21.0 53 32 1.3 2.0 14 21 8.1 10.9
136 139 85 83 5.5 7.0 29 17 3.2 5.1 33 52 27.0 35.0
Downloaded from http://ahajournals.org by on March 31, 2023

22 23 7 10 3.0 4.0 12 7 1.3 2.3 14 23 10.6 14.0

0.2 0.2 <0.001 <0.001 <0.001 <0.001 <0.001

where weight (L/kg) =body volume in liters,
0.11 = assumed blood volume as fraction of body
volume,2 vHct, = initial vHct, and vHctD = de-
sired vHct. Demerol compound (meperidine, 25
mg, promethazine, 6.25 mg, and chlorpromazine,
6.25 mg/ 100 ml of mixture) in a dose of 1 ml/30
pounds was administered prior to the study
(maximal dose, 2 ml).
The hematocrit was determined on central
venous blood using the Adams microhematocrit
centrifuge (M-90). Oxygen-carrying capacity was
measured by the cyanmethemoglobin method,
and blood oxygen saturation was determined by
an oximeter (American Optical Corporation
model no. 10840). Central venous pressure was
measured via catheters placed in the superior
vena cava or right atrium. Arterial and venous
pressures were obtained using strain-gauge
transducers, and heart rate was monitored.
Oxygen consumption was measured and SBF
determined in duplicate by the direct Fick
principle using mixed venous (superior vena cava
or right atrium) and arterial blood oxygen
saturations. Arterial partial pressure of carbon
Circulation, Volume XLII, August 1970
dioxide (Paco9), oxygen (Pao0), and pH were
determined by the use of the Instrumentation
Laboratory gas electrode and pH meter, respec-
tively. The CO2 content was calculated from
PaO2' Paco2 and pH using radiometer blood gas
calculator.* Viscosity measurements (at 38C)
were obtained in four patients with the GDM
rotational viscometer.13 This technic permits the
measurement of shear stress as a function of shear
rate at a range of 0.1 to 100 inverse seconds
(sec-1). The yield shear stress was extrapolated
from a plot of the square roots of shear stress
against the square roots of shear rate. Yield shear
stress (at 38C) was also measured in four
additional subjects undergoing red cell pheresis
and in whom complete hemodynamic data were
not obtained. Fibrinogen determinations'4 were
done in duplicate on each blood sample obtained
for viscometric studies.
Simultaneous measurements of RCV with 51Cr
autologous-labeled red cells and plasma volume
*The London Co., West Lake, Ohio.
 300
wvith 1'2;--labeled albumin were performed on
eight patients by a previously described meth-
od.-' Measurements were obtained prior to and
following the completion of erythropheresis. The
combined dosage of the two isotopes given was
proportional to the weight of the subject and was
always below the recommended dose.16 In four of
the 22 patients (nos. 18, 19, 20, and 22) the
blood removed was replaced with a 5% solution of
human albumin instead of fresh frozen plasma.
All significance tests were based on the paired t
statistic.
Results
The data reflecting acute hemodynamic
effects of reduction of RCV and its replace-
ment with plasma or 5% human albumin are
shown in table 1. The quantity of blood
removed from the patients varied depending
on the initial vHct, the subject's weight, and
the initial arterial oxygen saturation. The
mean initial vHct of the entire group was
reduced from 73.5% (range, 66 to 85%) to 62%
(range, 53 to 69%). A concomitant reduction
occurred in the oxygen-carrying capacity from
28.8 vol % (range, 21.7 to 35.5) to 24.5 vol %
(range, 19.5 to 28.4). Simultaneous determi-
nations of red cell and plasma volumes,
Downloaded from http://ahajournals.org by on March 31, 2023
ROSENTHAL ET AL.
a mean of 126 ml/kg (range, 102 to 147) prior
to the pheresis to 118 mllkg (range, 97 to
133). This change in WBV was probably due
to leakage of some fluid out of the vascular
space during the redistribution of the infused
plasma protein. Mean value for circulating
RCV was reduced by 23% from 87.2 ml/kg
(range, 62.3 to 105.9) to 67.3 ml/kg (range,
49.4 to 78.9). Plasma volume increased and
vHct decreased. The persistence of a steady
state during the study was evidenced by the
lack of significant change in the patient's state
of consciousness, respiratory rate, heart rate,
temperature, and oxygen consumption. The
changes induced in the vHct and RCV
resulted in significant alterations in blood
viscosity, central venous pressure, systemic
vascular resistance (SVR), SBF, stroke vol-
ume, and the amount of oxygen delivered to
0.12-

0
0
obtained on eight of the patients (cases 1 to O.10-
.
8), revealed that following replacement of
whole blood with an equal volume of plasma 0 *
total blood volume had diminished slightly. 0.08- 0
Whole blood, volume (WBV) decreased from
cm
Shear Rate Shear Rate 0
0.1 sec.-1 10 sec.-' 0.061
C')
0

300[ 30

0
0.041
200V 20 *0 0
Viscosity 0
(centipoise)

looF 10 0.02-

Before After Before After '

20 40 60
Figure 1

A decrease in blood viscosity following reduction in

venous hematocrit in four patients with polycythemia
secondary to cyanotic congenital heart disease. Vis-
cosity was measured at shear rates of 0.1 sec-' and
10 sec-1. X = case 1, o case 2, * = case 4 and *=
case 9.
Hematocrit %
Figure 2
Yield shear stress values (at 38C) of 23 arterial blood
samples from eight subjects in whom erythropheresis
was performed. The yield shear stress is considerably
increased at higher hematocrit values.
Circulation, Volume XLII, August 1970
 RED CELL VOLUME REDUCTION IN POLYCYTHEMIA 301
A 120w
B
15f

10 80
SBF SI
(L/min/M2) (ml/beatM2)

5 40
-6-

0 I O'
Before After Before After

D
60r

50 40 z
SVR SOT
(mm Hg/L /min) -0- (ml/min/kg)
-41-- .

60 20
Downloaded from http://ahajournals.org by on March 31, 2023

Before After
o l
Before After

Figure 3
Changes in systemic blood flow (SBF), stroke index (SI), systemic vascular resistance (SVR),
and systemic oxygen transport (SOT) associated with erythropheresis in 22 patients. The
symbol (-*-) denoted the mean value for each parameter before and after the procedure.

the body by the systemic circulation (systemic
oxygen transport).
Blood Viscosity and Yield Shear Stress
Following vHct reduction a decrease in
blood viscosity occurred in all patients stud-
ied. Results from four patients (cases 2, 9, 10,
and 17) who had simultaneous hemodynamic
and viscometric studies before and after red
cell pheresis are shown in figure 1. Blood
viscosity varies inversely and nonlinearly, with
shear rate.6' 17 Viscosity values at shear rates
of 0.1 sec-1 and 10 sec-' were selected because
these shear rates define the approximate range
relevant to the blood flow in the microcircu-
lation.'8
Yield shear stress is the force necessary to
make the structure of static blood "yield" and
Circulation, Volume XLII, August 1970
thereby allow flow to begin.'7 Since at any
given temperature only one yield shear stress
value exists for each blood sample,'9 it is more
convenient to direct attention to yield shear
stress rather than viscosity when comparing
different samples of blood. Yield shear stress,
normally 0.04 + 0.01 dyne/ cm2 (mean + stan-
dard deviation), was considerably increased
at high hematocrits (fig. 2). The yield shear
stress was lowered following red cell pheresis
in all patients studied. Fibrinogen content,
which is also known to affect yield shear
stress,19 remained in the normal range for all
samples (306 + 40 mg%).
Systemic Blood Flow
The initial SBF was low (2.5 L/min/m2 or
less) in seven of the 22 patients. Two of these
 302 ROSENTHAL ET AL.
A B C

150

.
100
0/ Change 0/ Change 0. %/ Change
SVR SBF 0 SOT
y y 0 y
50
* :

-30 -10
% ChangexHematocrit 0/0Change Hematocrit
x

Figure 4
Relative changes in systemic vascular resistance (SVR), systemic blood flow (SBF), and sys-
temic oxygen transport (SOT) in relation to relative change in venous hemnatocrit (vHct) in
22 patients. A reduction in the vHct was followed by a decrease in SVR and an increase in
both SBF and SOT. The standard deviations of the differences between observed values and
regression estimates are 15.5, 41.2, and 33.0f for SVR, SBF, and SOT, respectively.

seven patients had clinically severe congestive
heart failure. Reduction of the vHct and RCV
was followed by an increase in the SBF in all
patients (P < 0.001) from a mean of 3.2 + 1.3
Downloaded from http://ahajournals.org by on March 31, 2023
individual patients are shown in table 1. SVR
decreased in all patients (P < 0.001) from an
initial mean of 29 + 12 mm Hg/L/min/m2 to
17 + 7 following the procedure (fig. 3C). The

L/min/m2 to 5.1 + 2.3 L/min/m2 (fig. 3A).
For a 10% relative reduction in the vHct there
was a 38% average increase in the SBF (fig.
4B). The increase in SBF was the result of an
increase in stroke volume. Stroke index (SI)
rose from a mean of 33+ 14 ml/beat/m2 to
52 + 23 ml/beat/m2 (P < 0.001; fig. 3B).
Following the procedure there was no consis-
tent or significant change in the heart rate or
the rate of oxygen consumption (Vo2)* In
only two subjects did the heart rate vary more
than 10 beats/min (table 1). There was no
evidence of a relationship between the initial
vHct and heart rate (r = 0.06), but V02
tended to be lower in those with a high vHct
(r = -0.42). The systemic arteriovenous (A-
V) oxygen difference was reduced in all
patients reflecting the increased SBF. The
initial mean A-V oxygen difference of 4.8 ±
1.7 vol % of blood decreased to a mean of
3.2+ 1.7 vol % (P<0.001).
Systemic Vascular Resistance
The changes in systemic vascular resistance
(SVR) following acute RCV reduction in the
fall in SVR correlated well with the decrease
in red cell concentration. For a 10% relative
decrease in the vHct there was a 23% average
change in SVR (fig. 4A). Great variability was
encountered in the arterial blood pressure
response to erythropheresis. The changes in
blood pressure recorded were stable over a
half-hour period of observation. A decrease of
10 mm Hg or more in mean systemic arterial
pressure was noted in four patients and was
associated with a reduction in both systolic
and diastolic pressures (table 1). Three
subjects (cases 2, 9, and 16) developed mild
postural hypotension lasting from 12 to 24
hours after completion of the study. Following
the exchange, central venous pressure rose in
16 subjects, remained unchanged in four, and
decreased in two (P < 0.001). In the two
patients with clinically severe congestive heart
failure, the initial high mean central venous
pressure showed a further increase. In none of
the patients, however, was congestive heart
failure precipitated by the procedure. The
reason for the rise in central venous pressure is
Circulation, Volume XLII, August 1970
 RED CELL VOLUME REDUCTION IN POLYCYTHEMIA 303

A B
100r Group 1I
901

Art. 02 801-
Art. 02
Sat.(%) Sat. (%k)
70 -+

60

Before
50 After

Before After
Figure 5
Arterial oxygen saturation (Art. 02 Sat) values before and after erythropheresis in polycythemic
patients with severe pulmonary stenosis or pulmonary vascular obstructive disease (A, group 1)
and those with D-transposition of the great arteries and no significant pulmonary vascular
obstructive disease (B, group II). The symbol (-.-) denotes the mean value before and after
the procedure.

not clear, but it may be due to increased
venous return caused by a decrease in the
venous resistance resulting from a reduction in
blood viscosity. The situation may be some-
what analogous to a sudden opening of a
systemic arteriovenous fistula.20
Systemic Oxygen Transport
Downloaded from http://ahajournals.org by on March 31, 2023
between initial vHct and SOT at rest in
patients with severe secondary polycythemia
(vHct > 65%).
Systemic Arterial Oxygen Saturation
To illustrate the effect of red cell pheresis
on the systemic arterial blood oxygen satura-
tion in cyanotic patients, it is useful to divide

Systemic oxygen transport (SOT) can be
defined as the amount of oxygen delivered to
the body by the systemic circulation per unit
of time.21 To facilitate comparisons between
individuals of different sizes, the amount of
oxygen delivered per unit of time per unit of
body weight is used. This expression for SOT
is derived by multiplying SBF by arterial
oxygen content and dividing the result by the
patient's weight. Thus, SOT (ml 02/min/kg)
= SBF (L/min) x arterial oxygen content
(ml 02/L)/weight (kg). The effects of a
change in the vHct on the SOT is shown in
table 1 and figure 3D. An increase in oxygen
delivery to the tissues occurred in 21 of the 22
patients (P < 0.001). SOT increased from a
mean of 27 + 10.6 mI/min/kg to a mean of
35.0 + 14.0 ml/min/kg. The quantitative re-
sponse in SOT to a given relative change in
the vHct was rather variable among the
patients. For example, for a 10% reduction in
the vHct, the increase in SOT ranged from
10% to 40%, while the mean was about 20% (fig.
4C). There was no significant relationship
Circulation, Volume XLII, August 1970
the patients into two groups (fig. 5A and B).
Group I consists of 17 patients with severe
pulmonary stenosis (tetralogy of Fallot, tri-
cuspid atresia with pulmonary stenosis, or
transposition of the great arteries with pulmo-
nary stenosis) or marked pulmonary vascular
obstructive disease (pulmonary vascular resis-
tance greater than one-third SVR). Group II
consists of five patients with dextrotransposi-
tion of the great arteries (D-TGA) without
significant pulmonary stenosis or pulmonary
vascular obstruction (table 1, cases 1, 2, 4, 8,
and 11). Patients in group I showed a
significant drop in systemic blood arterial
oxygen saturation (P <0.001) following RCV
reduction and its replacement with plasma or
albumin (fig. 5A). The fall in systemic
saturation was most pronounced in subjects
with pulmonary atresia and least in those with
tricuspid atresia. In group II (D-TGA) the
systemic arterial oxygen saturation remained
essentially unchanged by the procedure (fig.
5B).
 304
Arterial Blood Gases
The changes in the Pao, induced by
erythropheresis are similar to those described
for systemic arterial oxygen saturation. In
group I patients, the mean Pao2 before and
after pheresis was 51.6 10.2 mm Hg and
45.2 8.3 mm Hg, respectively (P < 0.01). In
group II the Pao2 remained essentially un-
changed (45.0 + 9.7 vs. 45.8 8.0 mm Hg).
There was an increase (1.5 mm Hg or more)
in Pac02 following the procedure in 10 of the
22 patients; it remained unchanged in 10
others (in four of whom the blood was ex-
changed with albumin), and it was decreased
in two (P = 0.05 from two-sided sign test) . An
increase in arterial pH was demonstrated in
all the cases in which blood was replaced with
plasma. The mean initial arterial pH of
7.33 + 0.06 increased to 7.38 + 0.05 (P < 0.001).
When 5% solution of human albumin was
utilized for replacement, the initial blood pH
was not significantly altered (7.35 + 0.05
and 7.34 + 0.04, respectively). The rise in
blood pH following plasma infusion would
appear to be secondary to the rise in
Downloaded from http://ahajournals.org by on March 31, 2023
blood CO2 content which rose in all
case of plasma exchange (P < 0.001).
The mean CO2 content before and after
exchange transfusion with plasma was 21.0 +
3.1 mM/L and 25.8 + 4.0 mM/L, respective-
ly. The infusion of 5% albumin solution was
not associated with a rise in CO2 content
(23.9 3.2 mM/L vs. 24.3 2.3 mM/L).
Clinical Findings and Side Effects
Following red cell volume reduction, sub-
jective symptomatic improvement occurred in
15 patients. The beneficial effects most fre-
quently cited were relief from headaches,
decreased dyspnea and fatigue, increase in
stamina, and improvement in appetite. No
significant relationship was observed between
symptomatic improvement and the magnitude
of increase in SOT. No serious complications
or side effects were encountered in any of the
patients during or following the procedure of
erythropheresis. Transient postural hypoten-
sion was noted in three patients and urticaria
in two. The pH of nonbuffered fresh frozen
plasma is low (6.8 to 7.1), and hyperpnea
ROSENTHAL ET AL.
sometimes occurs on rapid infusion of plasma.
A similar response follows the rapid adminis-
tration of 5% human albumin solution (pH 6.9
to 7.2).
Discussion
The main purpose of this study was to
delineate the acute effects of isovolumic red
cell volume reduction on circulatory dynamics
in patients with polycythemia secondary to
cyanotic congenital heart disease under stable
and controlled conditions. Red cell concentra-
tion and volume were reduced while whole
blood volume (WBV) was maintained as
constant as possible. Although the amount of
solution infused was carefully adjusted to
replace all blood withdrawn, determinations
of blood volume revealed a small but definite
decrease in WBV following the procedure.
The mean resting SBF and stroke volume
of our patients with severe hypoxic poly-
cythemia were comparable to those of normal
individuals except when there was marked
congestive heart failure. The increase in blood
viscosity at high Hct levels tends to increase
resistance to blood flow.3' 6 However, the
concomitant hypervolemia with its associated
increased vascular dilatation may have the
opposite effect and thus resting SBF may
remain essentially normal.20 A significant
increase in stroke volume and SBF followed
RCV reduction and its replacement with an
equal volume of plasma or 5% human albumin.
This differs from the effect of phlebotomy
without fluid replacement in normal patients
and those with polycythemia vera when stroke
volume and SBF are reduced.22 23 Our clinical
experience has shown that acute phlebotomy
in patients with hypoxic polycythemia may
result in vascular collapse, cyanotic spells,
cerebral vascular accidents, or seizures. These
findings may be the sequelae of an acute
reduction in SBF resulting from a decrease in
blood volume. A controlled study using
phlebotomy alone was, therefore, considered
too hazardous.
An explanation for the observed increase in
SBF and stroke volume is provided by the
decrease in peripheral vascular resistance
Circulation, Volume XLII, August 1970
 RED CELL VOLUME REDUCTION IN POLYCYTHEMIA
resulting from the procedure (figs. 4A and 3C).
Marked changes in peripheral vascular resis-
tance and SBF were induced experimentally
by Murray and co-workers2' by producing
normovolemic Hct changes in dogs. The
decrease in peripheral vascular resistance and
the consequent increase in SBF induced by
normovolemic reduction of the Hct may be
primarily attributed to the reduction in the
viscosity and yield shear stress. Elevated
blood viscosity in association with polycythe-
mia of cyanotic heart disease bas been
adequately documented.19' 24, 25 The effect of a
reduction in vHct on the viscosity at shear
rates of 0.1 and 10 sec7l in four of our patients
is illustrated in figure 1. Physiologically the
most significant viscosity measurements are
those obtained at shear rates of 0 to 20 sec-'
since these represent the changes relevant to
flow in the microcirculation.6' 18 Furthermore,
the changes in viscosity with greater red cell
concentrations are paralleled by progressive
increase in the yield shear stress (fig. 2).
Replogle and associates" have suggested that
in polycythemia the yield stress may thus
Downloaded from http://ahajournals.org by on March 31, 2023
represent a significant portion of the peripher-
al resistance to the flow of blood. We observed
in vitro a sharp decrease in yield shear stress
following a reduction in the vHct in four
patients in whom measurements were made
prior to, and following, red cell pheresis. This
suggests that a reduction in viscosity and yield
shear stress may be important factors in
lowering the peripheral vascular resistance.
Further evidence of this effect is provided by
the great similarity between the hematocrit-
peripheral resistance relationship obtained
experimentally2l and the viscosity-hematocrit
curves obtained by direct measurements.26
In addition to the decrease in viscosity, two
other factors must be considered in explaining
the fall in peripheral resistance and increased
SBF. These are the decreased arterial oxygen
content and increased blood pH. Both are
known to produce local vasodilation and
increased SBF.27 The rise in blood pH
produced by intravenous infusion of plasma
probably had no effect on SBF regulation
since alkalosis was not produced, and replace-
Ciculation, Volume XLII, August 1970
305
ment with albumin caused a similar rise in
SBF without any significant alterations in
blood pH. Previous experiments have shown
that in anemia the drop in arterial oxygen
content, rather than the decrease in viscosity
primarily results in an increase of SBF. A
decrease in oxygen content without any
changes in blood viscosity can produce
progressive rise in SBF28; therefore, it seems
likely that the mechanisms for decrease in
peripheral resistance and increased SBF in-
duced by erythropheresis are a reduction in
the viscosity and yield shear stress and
possibly a decreased arterial oxygen content.
The present study demonstrates for the first
time that the changes in vascular resistance
and SBF induced in severe hypoxic polycythe-
mia are similar to those previously observed in
animal experiments.
Substantial acute improvement in SOT was
produced by the reduction of vHct in most of
our patients. As previously defined, the SOT is
a measure of the amount of oxygen supplied,
but not extracted by, the tissues per minute
per unit of body weight. It represents the
maximum amount of oxygen available to the
tissues for utilization. The increase in SOT
produced by erythropheresis is primarily due
to the increase in SBF since oxygen-carrying
capacity is always reduced and arterial oxygen
saturation often declines as a result of
pheresis. The maximum SOT usually occurs at
the normal vHct.2' In hypoxic polycythemia
arterial blood is desaturated, and oxygen
transport is less effective relative to the vHct.
The increased blood volume, however, by
lowering the resistance to blood flow enough
to maintain an adequate SBF,29 provides for
more normal oxygen delivery at rest. The
advantage of an increase in SOT following
isovolemic RCV reduction may be in providing
for a greater oxygen reserve available for
tissue utilization on exercise. It should be
emphasized that the improvement in oxygen
delive ry was shown here as an acute result of
erythr pheresis. How long it may persist
follow ing RCV reduction has not been deter-
mined.
 306
In individuals with increased right ventricu-
lar outflow obstruction or pulmonary vascular
resistance (group I), associated with marked
secondary polycythemia, RCV reduction may
produce a considerable fall in systemic arterial
oxygen saturation. This is a result of the
reduction in SVR, in the probable absence of a
significant change in pulmonary resistance,
leading to an increased right-to-left shunt via
the ventricular septal defect. In patients with
pulmonary atresia the fall in systemic satura-
tion may be dramatic because the entire
pulmonary blood supply is derived from the
aorta. Among those with D-TGA without
significant pulmonary stenosis or pulmonary
vascular obstruction (group II), systemic
oxygen saturation remained essentially un-
changed following the procedure. The ability
to maintain the blood arterial oxygen satura-
tion in those with D-TGA may be related to
increase in pulmonary blood flow and the
improved mixing between the pulmonary and
systemic circulations.
The replacement of whole blood with
plasma was accompanied by a rise in blood
Downloaded from http://ahajournals.org by on March 31, 2023
CO2 content and pH. These changes were not
observed when blood was replaced with a 5%
human albumin solution. The rise in pH after
replacement with plasma is most likely
secondary to the increased CO2 content which
in turn was attributed to the metabolism of
the sodium citrate and anhydrous citric acid
present in the administered plasma30' 31 (100
ml of anticoagulant contains 0.8 g of citric
acid and 2.2 g of sodium citrate) .
Our experience has suggested that acute
RCV reduction and replacement with plasma
or albumin (erythropheresis) reduced coagu-
lation abnormalities and decreased operative
mortality in severely polycythemic patients.32
The present study indicates that erythrophere-
sis lowers blood viscosity and increases SBF
and SOT. These effects together with the
absence of any serious complications make
erythropheresis preferable to acute phleboto-
my (without fluid replacement) in manage-
ment of patients with severely hypoxic polycy-
themia.
ROSENTHAL ET AL.
References
1. VE~REL D: Blood volume changes in cyanotic
congenital heart disease and polycythemia
rubra vera. Circulation 23: 749, 1961
2. ROSENTHAL A, BUTTON LN, NATHAN DG, ET AL:
Blood volume changes in cyanotic congenital
heart disease. Amer J Cardiol. In press
3. MURRAY JF, GOLD P, JOHNSON BL JR: The
circulatory effects of hematocrit variations in
normovolemic and hypervolemic dogs. J Clin
Invest 42: 1150, 1963
4. RICHARDSON TQ, GUYTON AC: Effects of
polycythemia and anemia on cardiac output
and other circulatory factors. Amer J Physiol
197: 1167, 1959
5. SCHNABEL TG JR, ELIASCH H, THOMASSON B, ET
AL: The effect of experimentally induced
hypervolemia on cardiac function in normal
subjects and patients with mitral stenosis. J
Clin Invest 38: 117, 1959
6. REPLOGLE RL, MEISELMAN HJ, MERRILL EW:
Clinical implications of blood rheology studies.
Circulation 36: 148, 1967
7. PUTNAM TC, KEvY SV, REPLOGLE RL: Factors
influencing the viscosity of blood. Surg Gynec
Obstet 124: 547, 1967
8. RICH AR: A hitherto unrecognized tendency to
the development of widespread pulmonary
vascular obstruction in patients with congenital
pulmonary stenosis (tetralogy of Fallot). Bull
Hopkins Hosp 82: 389, 1948
9. FERENCZ C: The pulmonary vascular bed in
tetralogy of Fallot. Bull Hopkins Hosp 106:
81, 1960
10. BERTHRONG M, SABISTON DC JR: Cerebral
lesions in congenital heart disease. Bull
Hopkins Hosp 89: 384, 1951
11. HARTMAN RC: A hemorrhagic disorder occurring
in patients with cyanotic congenital heart
disease. Bull Hopkins Hosp 91: 49, 1952
12. DENNIS LH, STE WART JL, CONRAD ME: Heparin
treatment of haemorrhagic diathesis in cyanotic
congenital heart disease. Lancet 1: 1088,
1967
13. GILINSON PJ JR, DAUWALTER CR, MERRILL
EW: Rotational viscometer. Trans Soc Rheol
7: 319, 1963
14. RATNOFF OD, MENZIE C: A new method for
measurement of fibrinogen in small samples of
plasma. J Lab Clin Med 37: 316, 1951
15. BUTTON LN, GIBSON JG, WALTER CW: Simul-
taneous determination of the volume of red
cells and plasma for survival studies of stored
blood. Transfusion 5: 2, 1965
16. RADIATION SAFETY MANUAL: The Children's
Hospital Medical Center, Boston, Mass. Dec 1,
1968
17. MERRLL EW, GILLILAND ER, COKELET CR, ET
AL: Rheology of human blood near and at zero
Circulation, Volume XLII, August 1970
 RED CELL VOLUME REDUCTION IN POLYCYTHEMIA
flow: Effects of temperature and hematocrit
level. Biophys J 3: 199, 1963
18. WELLS R: Blood flow it the microcirculation of
man and the flow properties of blood: A
correlative study. Fourth European Conference
on Microcirculation, Cambridge, 1966. Basel,
S. Karger, 1966, p 520
19. MARTY AT, ERAKLIs AJ, PELLETiER GA, ET AL:
The effect of temperature and hematocrit on
the yield shear stress of blood. Submitted for
publication.
20. GUYTON AC: Cardiac Output and Its Regulation.
Philadelphia, W. B. Saunders Co. 1963, pp
352; 366
21. MURRAY JF, GOLD P, JOHNSON BL JR: Systemic
oxygen transport in induced normovolemic
anemia and polycythemia. Amer J Physiol
203: 720, 1962
22. COBB LA, KRAMER RJ, FINCH CA: Circulatory
effects of chronic hypervolemia in polycythe-
mia vera. J Clin Invest 39: 1722, 1960
23. RALSTON LA, COBB LA, BRUCE RA: The role of
blood volume in regulation of cardiac output:
The effects of arterial bleeding in normal
subjects. Clin Res 8: 124, 1960
24. RUDOLPH AM, NADAS AS, BORGES WH: Hemato-
logic adjustments to cyanotic congenital heart
disease. Pediatrics 11: 454, 1953
25. KONTRAS SB, CRAENEN J, BODENBEHDER JG, ET
AL: Hyperviscosity in congenital heart disease.
Downloaded from http://ahajournals.org by on March 31, 2023
Circulation, Volume XLII, August 1970
307
(Abstr) Circulation 38 (suppl VI): VI-118,
1968
26. LEVY MN, SHARE L: The influence of erythrocyte
concentration upon the pressure-flow relation-
ships in the dog's hind limb. Circulation
Research 1: 247, 1953
27. RICHARDSON DW, WASSERMAN AJ, PATTERSON
JL JR: General and regional circulatory
responses to change in blood pH and carbon
dioxide tension. J Clin Invest 140: 31, 1961
28. SUNAHARA FA, BECK L: Cardiovascular effects of
acutely produced anemia in the normal dog.
Amer J Physiol 176: 139, 1954
29. CASTLE WB, JANDL JH: Blood viscosity and
blood volume: Opposing influences upon
oxygen transport in polycythemia. Seminars
Hemat 3: 193, 1966
30. CALLADINE M, GAIRNDER D, MAIDOO BT, ET
AL: Acid base changes following exchange
transfusion with citrated blood. Arch Dis Child
40: 626, 1965
31. WEXLER IB, PINCUS JB, NATELSON S, ET AL: The
fate of citrate in erythroblastotic infants treated
with exchange transfusion. J Clin Invest 28:
474, 1949
32. KEvY SV, BUTTON LN, ROSENTHAL A: The value
of erythropheresis in correcting coagulation
abnormalities in cyanotic heart disease.
(Abstr) The Society for Pediatric Research,
1969, p 159