Am J Physiol Heart Circ Physiol 285: H2663–H2669, 2003.
First published July 17, 2003; 10.1152/ajpheart.00128.2003.
A synergistic effect of albumin and fibrinogen on
immunoglobulin-induced red blood cell aggregation
Ronen Ben-Ami,1 Gershon Barshtein,4 Tamar Mardi,1 Varda Deutch,2
Ori Elkayam,3 Saul Yedgar,4 and Shlomo Berliner1
Departments of 1Internal Medicine D, 2Hematology, and 3Rheumatology,
Sourasky Medical Center, Tel Aviv 64239; and 4Department of Biochemistry,
Hadassah Medical School, Hebrew University, Jerusalem 91120, Israel
Submitted 23 February 2003; accepted in final form 3 July 2003
Ben-Ami, Ronen, Gershon Barshtein, Tamar Mardi,
Varda Deutch, Ori Elkayam, Saul Yedgar, and Shlomo
Berliner. A synergistic effect of albumin and fibrinogen on
immunoglobulin-induced red blood cell aggregation. Am J
Physiol Heart Circ Physiol 285: H2663–H2669, 2003. First
published July 17, 2003; 10.1152/ajpheart.00128.2003.—
Therapeutic administration of immunoglobulins (Ig) has the
potential to precipitate thrombotic events. This phenomenon
may be explained by red blood cell (RBC) aggregation, which
can be potentiated by Ig. The contribution of plasma albumin
and fibrinogen to Ig-induced RBC aggregation is unclear. We
examined RBC aggregation in three settings: 1) patients
receiving therapeutic infusions of Ig; 2) patients receiving
plasma supplemented in vitro with Ig; and 3) patients receiving RBC suspensions in standard buffer with varying concentrations of albumin, Ig, and fibrinogen. Ig infusion augmented aggregation of RBCs from patients with normal or
high plasma levels of albumin but decreased aggregation in
those with lower plasma albumin concentrations. In vitro,
RBC aggregation was significantly increased only when all
three components, fibrinogen, albumin, and Ig, were present
at or above normal concentrations in the suspension but was
unaffected when any one of the components was absent from
the suspension. Our results suggest a three-way interaction
among fibrinogen, Ig, and albumin that synergistically induces RBC aggregation in plasma. Understanding these interactions may help predict clinically important phenomena
related to RBC aggregation, such as thrombotic complications of Ig infusion.
hemorheology; thrombosis; adverse effects
IMMUNOGLOBULIN (Ig) preparations are routinely administered to patients suffering from a variety of autoimmune disorders, immune deficiencies, and GuillanBarre syndrome. Clinical data implicate therapeutically administered intravenous Ig in the precipitation
of life-threatening thrombotic events (7, 9, 10, 13, 21,
25, 28, 29, 31, 35). Ig preparations have been shown to
induce red blood cell (RBC) aggregation, which may
contribute to the increased blood viscosity associated
with their use (11, 23, 37). RBC aggregation affects
blood viscosity most markedly in areas of low shear
stress (⬍4 dyn/cm2), such as arterial bifurcations. The
Address for reprint requests and other correspondence: S. Yedgar,
Dept. of Biochemistry, Hadassah Medical School, Hebrew Univ.,
Jerusalem 91120, Israel (E-mail: [email protected]).
http://www.ajpheart.org
same sites are also prone to atherosclerosis and thrombosis (18). Increased RBC aggregation and the resulting elevated blood viscosity promote low-flow downstream from atherosclerotic lesions, a condition that
favors thrombogenesis (12). Regional low flow resulting from RBC aggregation can therefore set the stage
for vascular thrombosis, especially in persons with
underlying vascular atherosclerotic disease. Elucidating the mechanism for Ig-induced RBC aggregation
could help identify those patients at risk for thrombotic
complications.
Aggregation of RBCs is governed by opposing forces:
on the one hand, the repulsive electrical force between
negatively charged cells and the shear force exerted by
blood flow, which together disaggregate RBCs; and, on
the other hand, the cohesive force induced by the presence of various plasma proteins that promote the formation of rouleaux structures and larger aggregates (8,
20, 26, 30). The equilibrium between these forces determines the extent of RBC aggregation, which in turn
is the major determinant of blood viscosity at low shear
rate (2) and thus a key element in hemorheology.
The relative roles of plasma proteins in RBC aggregation are not clear, as disparate findings have been
reported. It is believed that fibrinogen, a 340-kDa fibrous hexamer, is the most potent aggregator of RBCs
in plasma (26). Some studies (11, 37) have shown that
Ig induces RBC aggregation, whereas others have
found no effect (16). Data on albumin are even more
conflicting. While it is clear that albumin does not
directly aggregate RBCs, inconsistent results have
been obtained regarding its effect when combined with
other plasma proteins. Two studies (14, 38) found an
inverse relationship between plasma albumin and
RBC aggregation in diabetics. Some investigations described enhancement of fibrinogen-induced RBC aggregation by albumin (17, 22), but one study (14) showed
inhibition. Ig-induced RBC aggregation has been inhibited by albumin in one study (17) but enhanced by
albumin in another (34). Reinhart and Nagy (24) found
that albumin increased the erythrocyte sedimentation
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
0363-6135/03 $5.00 Copyright © 2003 the American Physiological Society
H2663
H2664
IMMUNOGLOBULIN-INDUCED RED BLOOD CELL AGGREGATION
rate when added to fibrinogen and Ig but inhibited it
when added to either alone. However, the sedimentation rate is an imperfect correlate of RBC aggregation; because it is affected by the shape of RBC
aggregates, RBC aggregates are not visualized directly, and aggregation is measured under static, noflow conditions (36).
The present study was undertaken to examine the
combined effects of albumin and fibrinogen on Ig-induced RBC aggregation. First, we examined the effect
of Ig administered in vivo on RBC aggregation, as a
function of plasma albumin and fibrinogen concentrations. The combined effects of albumin and fibrinogen
on Ig-induced RBC aggregation were further investigated in vitro, using either human plasma or a suspension containing controlled concentrations of Ig, fibrinogen, and albumin as a medium for RBC aggregation.
MATERIALS AND METHODS
All patients and healthy (control) subjects gave their informed consent for participation, and the study protocol was
reviewed and authorized by the institutional review board at
the Sourasky Tel Aviv Medical Center.
Aggregation of RBCs from patients receiving intravenous
Ig therapy. To study the effect of intravenous Ig administered
in vivo, we recruited patients scheduled for Ig treatment
from the Tel Aviv Medical Center hematology and rheumatology clinics. All patients filled out a clinical questionnaire.
Venous blood was drawn for determination of RBC aggregation immediately before the infusion of Ig (Omr-IgG-am,
Omrix Biopharmaceuticals; Rehovot, Israel) was started and
again at its termination. Plasma fibrinogen concentration
was determined by the method of Clauss (5) using a thrombin
reagent (Dade Behring; Newark, DE) according to the manufacturer’s instructions. Plasma albumin was measured by
the Bayer Advia 1650 system (Bayer Diagnostics; Tarrytown,
NY). Plasma Ig was measured by nephelometry using the BN
II system (Dade Behring).
Preparation of RBC suspension for determination of RBC
aggregability. Samples of venous blood were drawn from the
antecubital vein and collected into EDTA-containing Vacutainers. The RBCs were isolated by centrifugation (2,000 rpm
for 10 min), washed with PBS (pH 7.4), and resuspended at
the desired hematocrit in either autologous plasma or PBS
supplemented with predetermined concentrations of human
fibrinogen (F4883, Sigma; St. Louis, MO), human serum
albumin (A3782, Sigma), and Ig, as described below.
Determination of RBC aggregation. All aggregation measurements were conducted immediately after venipuncture.
RBC aggregability was studied using a cell-flow properties
analyzer as previously described (3). Briefly, RBC suspension
was prepared at 5% hematocrit. The suspension was then
introduced into a cell-flow properties analyzer, consisting of a
narrow-gap (30 ␮m) flow chamber connected to a pump
exerting laminar flow and a pressure transducer that monitored shear stress during the experiment. The RBC dynamic
organization (aggregation/disaggregation) in the flow chamber was directly visualized and recorded through a microscope connected to a charge-coupled device videocamera,
which transmitted the RBC images to a computer. Images
were then analyzed by image analysis software [a modified
version of SigmaScan Pro (SPSS; Chicago, IL)], which computes the average aggregate size (number of RBCs/aggregate) by dividing the total aggregate volume by the volume of
a single RBC.
AJP-Heart Circ Physiol • VOL
RBC aggregation was monitored in the flow chamber under increasing shear stress and characterized by the area
under the curve (AUC) of average aggregate size plotted as a
function of the shear stress exerted (AUCAAS). The wall shear
stress taken for this calculation ranged from 0.15 to 4.00
dyn/cm2, at which normal RBCs are singly dispersed. The
AUC expresses both the extent of RBC aggregation and its
dependence on shear stress, thus reflecting the strength of
the intercellular interactions. We (1) previously found this
index to faithfully represent clinically relevant aggregation
in various disease states. The change in AUC after Ig infusion was expressed as the relative increment in RBC aggregation, calculated as follows
100 ⫻ [(AUC after Ig) ⫺ (AUC before Ig)]/(AUC before Ig)
RBC aggregation after the addition of Ig to plasma in vitro.
To study the effect of Ig on RBC aggregation over a wide
range of plasma fibrinogen concentrations, we drew blood
samples from healthy volunteers and hospitalized individuals with acute coronary syndromes. The latter group is
known to have elevated levels of plasma fibrinogen (1).
RBC aggregation was examined in autologous plasma at
baseline. Aggregation studies were then repeated after adding Ig in vitro to achieve a final concentration of 25 mg/ml,
approximately the concentration resulting in vivo from the
administration of Ig at a dose of 0.5 g/kg body wt.
RBC aggregation under controlled concentrations of fibrinogen, Ig, and albumin. To define more precisely the contribution of the major plasma proteins to RBC aggregation, we
studied the aggregation of RBCs from healthy volunteers
under a range of controlled concentrations of fibrinogen, Ig,
and albumin. To establish the dose response, RBC aggregation was examined at concentrations of fibrinogen ranging
from 0 to 1,000 mg/dl (increments of 200 mg/dl) in a suspension containing 0 or 25 mg/ml Ig and 0 or 4.5 g/dl albumin
and at concentrations of albumin ranging from 0 to 5.5 g/dl
(increments of 1 g/dl) in a suspension containing 0 or 25
mg/ml Ig and 250 or 500 mg/dl fibrinogen.
Statistical analysis. Differences in RBC aggregation
(AUCAAS) and plasma protein concentrations before and after Ig administration were calculated with Student’s t-test.
Correlations between continuous variables were analyzed
with Pearson’s bivariate correlation. All values are expressed
as means ⫾ SE. All statistical tests were two-sided. P values
were considered significant when ⬍0.05. Statistical calculations were performed with the SPSS software package.
RESULTS
Effect of Ig administered in vivo on RBC aggregation.
Thirteen patients (10 men and 3 women) received intravenous Ig treatment for a total of 18 sessions (5
patients were sampled twice). The indication for Ig
treatment was chronic lymphocytic leukemia in four
patients, non-Hodgkin’s lymphoma in one patient,
Hodgkin’s disease in one patient, multiple myeloma in
four patients, autoimmune vasculitis in two patients,
and common variable immune deficiency in one patient. Ig was given at a mean dose of 0.4 g/kg (range
0.3–0.6 g/kg). Results were analyzed per treatment
session rather than per patient.
Baseline values of plasma proteins are shown in
Table 1. After Ig treatment, plasma globulin increased
from 3.5 ⫾ 0.5 to 4.2 ⫾ 0.5 g/dl (P ⬍ 0.001), plasma
fibrinogen decreased from 325 ⫾ 26 to 290 ⫾ 19 mg/dl
285 • DECEMBER 2003 •
www.ajpheart.org
IMMUNOGLOBULIN-INDUCED RED BLOOD CELL AGGREGATION
Table 1. Baseline concentrations of plasma proteins
in the various patient groups studied
Patients receiving
intravenous Ig
Patients with ischemic
heart disease
Control subjects
n
Fibrinogen,
mg/dl
Ig,
g/dl
Albumin,
g/dl
13
325.8 ⫾ 26.5
3.5 ⫾ 0.5
3.8 ⫾ 0.1
15
4
380.6 ⫾ 31
268.9 ⫾ 28.4*
2.7 ⫾ 0.1
2.7 ⫾ 0.1
4.0 ⫾ 0.1
5.2 ⫾ 0.1†
Values are means ⫾ SE; n, no. of subjects. * Value smaller than in
patients with ischemic heart disease (P ⫽ 0.02); † value greater than
in patients with ischemic heart disease (P ⫽ 0.001) and in patients
receiving intravenous Ig (P ⬍ 0.001).
(P ⬍ 0.01), and plasma albumin decreased from 3.8 ⫾
0.1 to 3.4 ⫾ 0.1 g/dl (P ⬍ 0.001).
Overall, RBC aggregation was not significantly altered after the administration of Ig. The relative
change in AUC after Ig infusion ranged from an increase of 78% to a decrease of 43%. The results depicted
in Fig. 1 show that the response to Ig differed among
patients with low plasma albumin (ⱕ3.5 g/dl) and
those with normal to high plasma albumin (⬎3.5 g/dl).
Before Ig infusion, AUC was higher in the low albumin
group compared with the normal-high albumin group.
However, Ig infusion decreased AUCAAS by 12.29 ⫾
10.14% in the low albumin group and increased it by
23.18 ⫾ 9.95% in the normal-high albumin group (P ⫽
0.02; Fig. 1). In the group as a whole, RBC aggregation
did not correlate with plasma fibrinogen either before
or after Ig infusion. Baseline AUCAAS before Ig infusion did not correlate with plasma fibrinogen in either
of the albumin groups. However, after Ig infusion, in
patients with plasma albumin above 3.5 g/dl, AUCAAS
correlated positively with plasma fibrinogen (r ⫽ 0.74,
P ⫽ 0.006). In contrast, in the low albumin group,
Fig. 1. Relative increment in red blood cell (RBC) aggregation,
expressed as the area under the curve for average aggregate size
(AUCAAS), of RBCs from patients receiving intravenous Ig therapy.
The relative increment was positive among patients with plasma
albumin (Alb) concentrations above 3.5 g/dl and negative in patients
with plasma Alb concentrations below 3.5 g/dl (P ⫽ 0.02).
AJP-Heart Circ Physiol • VOL
H2665
there was no correlation between plasma fibrinogen
and AUCAAS after Ig infusion (r ⫽ ⫺0.55, P ⫽ 0.2).
The relative increment in AUCAAS after Ig infusion
correlated with plasma fibrinogen in the normalhigh albumin group (r ⫽ 0.59, P ⫽ 0.04) but not in
the low albumin group (r ⫽ 0.04, P ⫽ 0.9).
Effect of Ig on RBC aggregation in vitro. Plasma was
collected from 15 patients hospitalized with an acute
coronary syndrome and 4 healthy individuals. Concentrations of plasma proteins are shown in Table 1.
Baseline RBC aggregation in human plasma correlated
positively with plasma globulin (r ⫽ 0.58, P ⫽ 0.009)
and plasma fibrinogen (r ⫽ 0.47, P ⫽ 0.04) and negatively with plasma albumin concentrations (r ⫽ ⫺0.54,
P ⫽ 0.01). As with patients receiving Ig in vivo, addition of Ig to human plasma in vitro was not associated
with an overall change in mean AUCAAS. However, the
effect of Ig on RBC aggregation was dependent on the
concentrations of fibrinogen and albumin in plasma;
the relative increment in RBC aggregation correlated
with plasma fibrinogen levels (r ⫽ 0.64, P ⫽ 0.003). A
marked increase in RBC aggregation was noted in
plasma samples with fibrinogen concentrations of
above 450 mg/dl (Fig. 2). The plasma sample with the
highest fibrinogen concentration (600 mg/dl) exhibited
a 29-fold increase in AUCAAS after the addition of Ig.
Furthermore, the effect of Ig on RBC aggregation in
the presence of a high fibrinogen concentration was
strongly dependent on the albumin concentration.
When plasma fibrinogen was ⬍450 mg/dl, there was
an inverse correlation between albumin levels and
AUCAAS (r ⫽ ⫺0.69, P ⫽ 0.004; Fig. 3A). However, in
the four plasma samples in which the fibrinogen
concentration exceeded 450 mg/dl, AUCAAS correlated linearly with the plasma albumin concentration (r ⫽ 0.99, P ⫽ 0.006; Fig. 3B).
RBC aggregation under controlled concentrations of
fibrinogen, Ig, and albumin. The response of RBC aggregation to increasing concentrations of albumin (0 to
Fig. 2. RBC aggregation (AUCAAS) as a function of fibrinogen (Fib)
concentration in plasma supplemented with Ig. Markedly elevated
AUCAAS is noted at Fib concentrations above 450 mg/dl.
285 • DECEMBER 2003 •
www.ajpheart.org
H2666
IMMUNOGLOBULIN-INDUCED RED BLOOD CELL AGGREGATION
min concentrations of 0–3.5 g/dl, AUCAAS did not differ
significantly among the four suspensions. At albumin
concentrations of 4.5 and 5.5 g/dl, AUCAAS in the suspension containing a high fibrinogen concentration
supplemented with Ig was increased twofold over the
other three suspensions (P ⬍ 0.001; Fig. 4).
We also assessed the contribution of fibrinogen to
RBC aggregation in four sets of suspensions, produced
by the following combinations: 1) the presence or absence of a physiological concentration of albumin (4.5
g/dl); and 2) the presence or absence of Ig at a concentration of 25 mg/ml.
RBC aggregation (log AUCAAS) correlated with fibrinogen concentration in a suspension containing
25 mg/ml Ig and 4.5 g/dl albumin (r ⫽ 0.86, P ⫽ 0.02).
In contrast, fibrinogen concentration did not affect
AUCAAS in albumin-free suspensions (Fig. 5). At
fibrinogen concentrations above 400 mg/dl, the absolute value of AUCAAS was significantly greater in a
suspension containing both Ig and albumin compared
with the other three suspensions (P ⬍ 0.01; Fig. 5).
DISCUSSION
Our results suggest a three-way interaction among
the three main plasma proteins involved in RBC aggregation. The observations in vivo imply that infusion
of Ig increases RBC aggregation in the presence of an
elevated plasma fibrinogen concentration and a normal
to high plasma albumin concentration. Both these conditions are required for Ig to augment RBC aggregation, and their effects are synergistic. These findings
are supported by the results of the in vitro studies:
RBC aggregation in autologous plasma was enhanced
after the addition of Ig if the plasma fibrinogen concen-
Fig. 3. RBC aggregation in plasma (AUCAAS) as a function of plasma
Alb concentration. A: without Ig supplementation, no correlation is
evident between AUCAAS and plasma Alb. B: with Ig supplementation, AUCAAS correlates negatively with plasma Alb at plasma Fib
concentrations below 450 mg/dl (r ⫽ ⫺0.69, P ⫽ 0.004) but increases
linearly with increasing plasma Alb concentration at plasma Fib
concentrations above 450 mg/dl (r ⫽ 0.99, P ⫽ 0.006).
5.5 g/dl) was examined in four sets of suspensions,
produced by the following combinations: 1) low or high
fibrinogen concentrations (250 and 500 mg/dl, respectively); and 2) the presence or absence of Ig at a
concentration of 25 mg/ml. In a suspension containing
Ig (25 mg/ml) and a high fibrinogen concentration,
AUCAAS correlated positively with albumin concentration (r ⫽ 0.94, P ⫽ 0.01; Fig. 4). However, in the other
three suspensions, there was no effect of increasing
albumin concentration on RBC aggregation. At albuAJP-Heart Circ Physiol • VOL
Fig. 4. RBC aggregation (AUCAAS) as a function of Alb concentration
in four sets of suspensions with controlled concentrations of Fib and
Ig. AUCAAS correlates with Alb concentration only when both Ig (25
mg/ml) and high Fib concentration (500 mg/dl) are present in the
medium but not in the absence of either.
285 • DECEMBER 2003 •
www.ajpheart.org
IMMUNOGLOBULIN-INDUCED RED BLOOD CELL AGGREGATION
Fig. 5. RBC aggregation (AUCAAS) as a function of Fib concentration
in four sets of suspensions with controlled concentrations of Alb and
Ig. Log AUCAAS correlates with Fib concentration in a suspension
containing Ig (25 mg/ml) and Alb (4.5 g/dl) but not in an Alb-free
suspension. AUCAAS is significantly greater at Fib concentrations
above 400 mg/dl in a suspension containing Ig and Alb compared
with the other three suspensions.
tration exceeded 450 mg/dl; at this fibrinogen concentration, RBC aggregation correlated linearly with the
plasma albumin concentration. However, at a lower
plasma fibrinogen concentration, we found no enhancement of RBC aggregation by Ig, and AUCAAS correlated
negatively with the plasma albumin concentration.
These interactions were further delineated at controlled concentrations of fibrinogen, albumin, and Ig.
RBC aggregation was significantly greater in suspensions containing Ig, albumin, and fibrinogen compared
with suspensions deficient in any one of these macromolecules. The presence of albumin (4.5 g/dl) and Ig (25
mg/ml) in the RBC suspension facilitated enhancement
of RBC aggregation in response to increasing fibrinogen concentrations. Likewise, a fibrinogen concentration of 500 mg/dl and an Ig concentration of 25 mg/dl
facilitated the enhancement of RBC aggregation in
response to increasing albumin concentrations; in an
RBC suspension free of either fibrinogen or Ig, there
was no correlation between albumin and RBC aggregation. Taken together, these results show that enhanced RBC aggregation occurs only when all three
plasma proteins are present at or above threshold
concentrations, indicating that RBC aggregation is dependent on a synergistic interaction between Ig, albumin, and fibrinogen.
The effect of albumin on RBC aggregation merits
special consideration. Previous studies (14, 17, 22, 34,
38) varied widely in their findings concerning the role
of albumin in RBC aggregation. These discrepancies
may result from a failure to take into account the
AJP-Heart Circ Physiol • VOL
H2667
interactions among various plasma proteins. Our findings in vivo and in vitro confirm a negative correlation
between plasma albumin concentration and RBC aggregation, as reported by others (14, 38). However,
when studied at high concentrations of both fibrinogen
and Ig, albumin takes on an important role as an
inducer of RBC aggregation.
Aggregation of RBCs in a suspension increases with
increasing molecular size of macromolecules in the
suspension (4). Therefore, it is reasonable to speculate
that synergistic enhancement of RBC aggregation by
plasma proteins is a result of the formation of multiprotein complexes. Being an anionic protein, it is likely
that albumin by itself directly disaggregates negatively charged RBCs. Albumin may, however, interact
with other plasma proteins, with the effect of enhancing their ability to aggregate RBCs. Fibrinogen and Ig
are cationic proteins and thus can bind to albumin. It is
possible to construct a plausible model to explain the
synergistic effects of all three plasma proteins, based
on the following assumptions: 1) albumin can bind to
both fibrinogen and Ig simultaneously, using two separate binding sites; and 2) fibrinogen and Ig interact
with albumin using binding sites that are distinct from
the RBC membrane binding sites of these molecules. If
these assumptions hold true, then sufficient concentrations of these plasma proteins could result in the formation of large, positively charged complexes of fibrinogen, Ig, and albumin, which efficiently bridge RBCs
and promote the formation of large aggregates.
Several observations support this model. AlbuminIgG complexes are present in normal human plasma
(27). Such complexes probably increase in number after the administration of Ig. Albumin-fibrinogen complexes can be produced experimentally, and these complexes have been shown to interact with platelets (32).
Both Ig and fibrinogen have been extensively shown to
interact with RBCs. Ig infused into healthy subjects
coats RBCs without causing hemolysis (19). Fibrinogen
binding to RBC membranes has recently been shown to
be inhibited by the peptide Arg-Gly-Asp-Ser, suggesting both specific and nonspecific fibrinogen binding to
RBCs (15). Thus both fibrinogen and Ig can each bind
to albumin and RBC membranes and could theoretically form an intercellular matrix that stabilizes RBC
aggregates. The significant decrease in plasma fibrinogen and albumin after intravenous Ig administration,
reported also by Madl et al. (16), is consistent with the
formation of Ig-albumin-fibrinogen complexes, resulting in reduced levels of free albumin and fibrinogen.
The level of plasma albumin, a so-called “negative
acute phase protein,” decreases in acutely ill patients,
and albumin levels are inversely related to the risk of
death. A recent metaanalysis has shown that the timehonored practice of administering albumin to critically
ill patients does not reduce mortality and may in fact
increase it (6). In view of our findings, it is tempting to
hypothesize that reduction in plasma albumin levels in
acutely ill patients is an adaptive response that reduces the viscosity of blood containing high concentrations of RBC-aggregating acute phase proteins such as
285 • DECEMBER 2003 •
www.ajpheart.org
H2668
IMMUNOGLOBULIN-INDUCED RED BLOOD CELL AGGREGATION
fibrinogen and Ig. Increased RBC aggregation may
reduce oxygen delivery to tissues by increasing the
thickness of the marginal cell-free layer (33), thus
reducing the already compromised tissue oxygenation
in critical patients.
In clinical practice, fibrinogen and albumin concentrations tend to vary reciprocally, and it is uncommon
to find elevated levels of both in the same patient. Madl
et al. (16) studied patients with sepsis who received Ig
and found no change in RBC aggregation. These patients probably had high concentrations of fibrinogen
and low concentrations of albumin, as is typical of
acutely ill patients.
Induction of RBC aggregation by therapeutically administered Ig may have important clinical consequences. Thrombotic complications of Ig treatment are
increasingly reported and include myocardial infarction (9, 10, 25), ischemic stroke (7, 28, 31), and retinal
vein occlusion (21). Prophylactic administration of Ig to
bone marrow transplant recipients was associated
with increased mortality in two large registries. Death
was attributed to venoocclusive disease of the liver,
which could be explained by increased blood viscosity
resulting from RBC aggregation (13, 35). Ig is currently contraindicated in children with cyanotic heart
disease, who have an elevated hematocrit and blood
viscosity, as it was found to promote cyanotic episodes
and poor surgical outcomes (29). Knowing what constellation of plasma protein concentrations places a
patient at an increased risk of these complications may
facilitate the selection of candidates for Ig treatment.
In conclusion, albumin, fibrinogen, and Ig synergistically induce RBC aggregation. Therefore, patients
with high plasma concentrations of both albumin and
fibrinogen are susceptible to enhanced RBC aggregation after Ig infusion, which may place them at risk for
thrombotic complications. Our results should aid in
constructing a model of RBC aggregation by plasma
proteins that takes into account interactions among
the proteins themselves as well as between proteins
and RBCs.
We are indebted to O. Friedman and S. Levi for technical assistance.
DISCLOSURES
This work was supported by Israel Ministry of Science Grant
1460-1 (to S. Yedgar and G. Barshtein) and United States-Israel
Binational Science Foundation Grant 2001203 and by the Hebrew
University Friends Association and The Walter and Greta Stiel
Chair for Heart Studies (to S. Yedgar).
REFERENCES
1. Ben-Ami R, Barshtein G, Zeltser D, Goldberg Y, Shapira I,
Roth A, Keren G, Miller H, Prochorov V, Eldor A, Berliner
S, and Yedgar S. Parameters of red blood cell aggregation as
correlates of the inflammatory state. Am J Physiol Heart Circ
Physiol 280: H1982–H1988, 2001.
2. Chien S. Rheology in the microcirculation in normal and low
flow states. Adv Shock Res 8: 71–80, 1982.
3. Chen S, Barshtein G, Gavish B, Mahler Y, and Yedgar S.
Monitoring of red blood cell aggregability in a flow-chamber by
computerized image analysis. Clin Hemorheol 14: 497–508,
1994.
AJP-Heart Circ Physiol • VOL
4. Chen S, Gavish B, Zhang S, Mahler Y, and Yedgar S.
Monitoring of erythrocyte aggregate morphology under flow by
computerized image analysis. Biorheology 32: 487–496, 1995.
5. Clauss A. [Rapid physiological coagulation method in determination of fibrinogen.] Acta Haematol 17: 237–246, 1957.
6. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review
of randomised controlled trials. Br Med J 317: 235–240, 1998.
7. Dalakas MC. High-dose intravenous immunoglobulin and serum viscosity: risk of precipitating thromboembolic events. Neurology 44: 223–226, 1994.
8. Dintenfass L (Editor). Development of the blood viscosity factors. In: Blood Viscosity. Boston, MA: MTP, 1985, p. 45–112.
9. Elkayam O, Paran D, Milo R, Davidovitz Y, AlmozninoSarafian D, Zeltser D, Yaron M, and Caspi D. Acute myocardial infarction associated with high dose intravenous immunoglobulin infusion for autoimmune disorders. A study of four
cases. Ann Rheum Dis 59: 77–80, 2000.
10. Fisman DN and Smilovitch M. Intravenous immunoglobulin,
blood viscosity and myocardial infarction. Can J Cardiol 13:
775–777, 1997.
11. Fusman R, Zeltser D, Rotstein R, Chapman J, Avitzour D,
Shapira I, Eldor A, Elkayam O, Caspi D, Arber N, and
Berliner S. INFLAMET: an image analyzer to display erythrocyte adhesiveness/aggregation. Eur J Intern Med 11: 271–276,
2000.
12. Karino T and Goldsmith HL. Role of blood cell-wall interactions in thrombogenesis and atherogenesis: a microrheological
study. Biorheology 21: 587–601, 1984.
13. Klaesson S, Ringden O, Ljungman P, Aschan J, Hagglund
H, and Winiarski J. Does high-dose intravenous immune globulin treatment after bone marrow transplantation increase mortality in veno-occlusive disease of the liver? Transplantation 60:
1225–1230, 1995.
14. Lacombe C, Bucherer C, Ladjouzi J, and Lelievre JC.
Competitive role between fibrinogen and albumin on thixotropy
of red cell suspensions. Biorheology 25: 349–354, 1988.
15. Lominadze D and Dean WL. Involvement of fibrinogen specific binding in erythrocyte aggregation. FEBS Lett 517: 41–44,
2002.
16. Madl C, Koppensteiner R, Wendelin B, Lenz K, Kramer L,
Grimm G, Kranz A, Schneeweiss B, and Ehringer H. Effect
of immunoglobulin administration on blood rheology in patients
with septic shock. Circ Shock 40: 264–267, 1993.
17. Maeda N and Shiga T. Opposite effect of albumin on the
erythrocyte aggregation induced by immunoglobulin G and fibrinogen. Biochim Biophys Acta 855: 127–135, 1986.
18. Malek AM, Alper SL, and Izumo S. Hemodynamic shear
stress and its role in atherosclerosis. JAMA 282: 2035–2042,
1999.
19. Mueller-Eckhardt C, Salama A, Mahn I, Kiefel V, Neuzner
J, and Graubner M. Lack of efficacy of high-dose intravenous
immunoglobulin in autoimmune haemolytic anaemia: a clue to
its mechanism. Scand J Haematol 34: 394–400, 1985.
20. Nash GB, Wenby R, Sowemimo-Coker SO, and Meiselman
HJ. Influence of cellular properties on red cell aggregation. Clin
Hemorheol 7: 93–108, 1987.
21. Oh KT, Boldt HC, and Danis RP. Iatrogenic central retinal
vein occlusion and hyperviscosity associated with high-dose intravenous immunoglobulin administration. Am J Ophthalmol
124: 416–418, 1997.
22. Rampling MW and Martin G. Albumin and rouleaux formation. Clin Hemorheol 12: 761–765, 1992.
23. Reinhart WH and Berchtold PE. Effect of high-dose intravenous immunoglobulin therapy on blood rheology. Lancet 339:
662–664, 1992.
24. Reinhart WH and Nagy C. Albumin affects erythrocyte aggregation and sedimentation. Eur J Clin Invest 25: 523–528, 1995.
25. Rosenbaum JT. Myocardial infarction as a complication of
immunoglobulin therapy. Arthritis Rheum 40: 1732–1733, 1997.
26. Shalak R, Zarda P, Jan KM, and Chien S. Mechanism of
rouleaux formation. Biophys J 35: 771–781, 1981.
27. Sharma NC, Mohammad SF, Chuang HY, and Mason RG.
Albumin-IgG complexes in human serum and plasma that in-
285 • DECEMBER 2003 •
www.ajpheart.org
IMMUNOGLOBULIN-INDUCED RED BLOOD CELL AGGREGATION
28.
29.
30.
31.
32.
33.
hibit blood platelet adhesion. Proc Natl Acad Sci USA 78: 7750–
7753, 1981.
Silbert PL, Knezevic WV, and Bridge DT. Cerebral infarction complicating intravenous immunoglobulin therapy for polyneuritis cranialis. Neurology 42: 257–258, 1992.
Simoes EA, Sondheimer HM, Top FH Jr, Meissner HC,
Welliver RC, Kramer AA, and Groothuis JR. Respiratory
syncytial virus immune globulin for prophylaxis against respiratory syncytial virus disease in infants and children with congenital heart disease. The Cardiac Study Group. J Pediatr 133:
492–499, 1998.
Snabre P, Bitbol M, and Mills P. Cell disaggregation behavior
in shear flow. Biophys J 51: 795–807, 1987.
Steg RE and Lefkowitz DM. Cerebral infarction following
intravenous immunoglobulin therapy for myasthenia gravis.
Neurology 44: 1180–1181, 1994.
Takeoka S, Teramura Y, Okamura Y, Handa M, Ikeda Y,
and Tsuchida E. Fibrinogen-conjugated albumin polymers and
their interaction with platelets under flow conditions. Biomacromolecules 2: 1192–1197, 2001.
Tateishi N, Suzuki Y, Cicha I, and Maeda N. O2 release from
erythrocytes flowing in a narrow O2-permeable tube: effects of
AJP-Heart Circ Physiol • VOL
34.
35.
36.
37.
38.
H2669
erythrocyte aggregation. Am J Physiol Heart Circ Physiol 281:
H448–H456, 2001.
Weng X, Roederer GO, Beaulieu R, and Cloutier G. Contribution of acute-phase proteins and cardiovascular risk factors to
erythrocyte aggregation in normolipidemic and hyperlipidemic
individuals. Thromb Haemost 80: 903–908, 1998.
Wolff SN, Fay JW, Herzig RH, Greer JP, Dummer S,
Brown RA, Collins RH, Stevens DA, and Herzig GP. Highdose weekly intravenous immunoglobulin to prevent infections
in patients undergoing autologous bone marrow transplantation
or severe myelosuppressive therapy. A study of the American
Bone Marrow Transplant Group. Ann Intern Med 118: 937–942,
1993.
Woodland NB, Cordatos K, Hung WT, Reuben A, and Holley L. Erythrocyte sedimentation in columns and the significance of ESR. Biorheology 33: 477–488, 1996.
Yamamoto M. Effects of fibrinogen, globulin, albumin and hematocrit on the kinetics of erythrocyte aggregation in man.
Angiology 37: 663–671, 1986.
Zimmermann J, Schramm L, Wanner C, Mulzer E, Henrich HA, Langer R, and Heidbreder E. Hemorheology,
plasma protein composition and von Willebrand factor in type I
diabetic nephropathy. Clin Nephrol 46: 230–236, 1996.
285 • DECEMBER 2003 •
www.ajpheart.org