LABORATORY INVESTIGATION
ATHEROSCLEROSIS
Protection against atherogenesis with the
drag-reducing agent Separan AP-30
FERRUKH I. FARUQUI, MAUREEN D. OTTEN, B.Sc.,
AND
PHILIP
I.
polymer
POLIMENI, PH.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
ABSTRACT The inhibitory effect of Separan AP-30, an anionic polyacrylamide, on atherosclerotic
plaque formation in aortas of rabbits on a high (2%) cholesterol diet was tested over a period extending
from 37 to 170 days. Atherogenesis was quantified morphometrically by application of a computerassisted image analysis of histologic cross sections of the aorta. The area of vessel wall-atheroma
interface, fraction of lumen occluded, and other indexes of atherogenesis were measured in each of 26
segments of aorta excised from the animals, half of which were administered injections (intravenous) of
Separan three times a week. Regression analysis of the morphometric data indicates that the polyelectrolyte exerts a powerful antiatherogenic effect in all regions of the aorta, inhibiting the formation of
plaque mass to less than half in the aortic arch and about one-fifth in the descending aorta as compared
with the aortic plaque masses in untreated rabbits. Results are compatible with the suggestion that a
novel hemodynamic principle in vivo, polymer drag reduction, might be effectively applied against
atherosclerosis.
Circulation 75, No. 3, 627-635, 1987.
THE ADDITION of certain linear macropolymers to
flow can under certain conditions greatly reduce frictional resistance by polymer drag reduction, an effect
also known among hydrodynamicists as the "Toms
phenomenon." Flow can thus be increased by threefold
or more without alteration of the driving pressure.'-3
The effect occurs only in the presence of disturbed or
turbulent flow, and it is associated with a laminarization of the flow. This phenomen has also been observed in pipe blood flow with at least four different
drag-reducing polymers, 7 including the anionic polyacrylamide Separan AP-30.
Application of the polymer drag-reduction principle
to blood flow in vivo was first attempted independently
From the Departments of Physiology and of Pharmacology and
Therapeutics, University of Manitoba Faculty of Medicine, Winnipeg,
Manitoba, Canada.
Supported primarily by a special donation to the Faculty of Medicine
for Research in Arteriosclerosis, with additional contributions from
grants of the Manitoba Health Research Council, Manitoba Heart Foundation, and Medical Research Council of Canada.
Address for correspondence: Dr. P. I. Polimeni, Department of Pharmacology and Therapeutics, University of Manitoba, Faculty of Medicine, 770 Bannatyne Avenue, Winnipeg, Manitoba, Canada R3E OW3.
Received May 6, 1986; revision accepted Nov. 13, 1986.
A preliminary report was presented at the 28th Annual Meeting of the
Canadian Federation of Biological Societies, Toronto, June 19, 1985.
This work was presented in partial fulfillment of the requirements for
the degree of B.Sc. (Med.) by F. I. Faruqui, who was the 1985 recipient
of the A. Allyn Roosen Award for Best Project in Cardiovascular
Research.
Vol. 75, No. 3, March 1987
at two laboratories in the mid- 1970s. "In one of these
laboratories, Mostardi et al.8 demonstrated with hotfilm anemometry that poststenotic flow disturbances in
the dog were diminished subsequent to intravenous
injection of Separan AP-30. They postulated'0 that it
might be feasible to inhibit atherogenesis with dragreducing polymers given the evidence that atheromas
tend to form preferentially at vascular bifurcations,
branches, and other regions of disturbed flow.'2 Accordingly, aortas of rabbits on a high-cholesterol diet,
some of which had been administered Separan AP-30,
were examined and compared. Based on a visual and
photographic inspection of the inner walls of aortas slit
lengthwise, it was concluded that the Separan-treated
animals were markedly less atherosclerotic than the
untreated ones. Similar results were obtained in White
Carneau pigeons fed an atherogenic diet.'3
Despite the profound theoretical and therapeutic implications of Mostardi's hypothesis and pictorial evidence, the experiments were never repeated and the
report was virtually ignored. The specific objective of
our research was to test the claimed antiatherogenic
effects of Separan, applying a computer-assisted image analysis of the contours of histologic cross sections
obtained from aortas of Separan-treated and untreated
atherosclerotic rabbits. This study is part of our more
general objective of testing the hypothesis that a hydrodynamic principle known to dampen flow disturbances
627
FARUQUI et al.
and turbulence is applicable against a pathologic process known to be associated with such unstable flows.
Materials and methods
Experimental protocol. Female New Zealand white rabbits
weight and organized into weightmatched cohorts, each consisting of a control (n = 8, 2.736 0.498 kg), a Separan-treated (n 8, 2.681 + 0.480 kg), and in
most cohorts a normal (n = 6, 2.519 + 0.812 kg) rabbit. The
normal group received 100% commercial rabbit chow (ICN
Biochemical Canada Ltd., Montreal, Quebec); both control and
Separan-treated groups received the same diet containing in
addition 2%c cholesterol. Separan treatment consisted of the
injection into the marginal ear vein of a 0.2% (weight/weight)
Separan AP-30 (Dow Chemical Co., Midland, MI) solution
were ranked for body
=
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
three times a week. Treatment was initiated at the same time that
the high-cholesterol diet was started. The volume injected was
adjusted so that the dose would result in a blood polymer concentration of about 60 ppm. The rabbits were housed in individual cages with access to chow and water ad libitum. The criteria
for killing a cohort were persistent weight loss or signs of illness
in any member. This protocol of termination was adopted, instead of fixed intervals, to maximize plaque mass in each cohort
while keeping the total number of "sick days" to a minimum. An
animal was considered to be ill when it became obviously listless. In all cases termination was necessitated by the illness, or
in two cases unexpected death, of the control cohort member:
one control rabbit was excluded (see below).
Preparation of Separan solution. To prepare the Separan
solution the polymer powder was dissolved in 0.9% NaCl solution by agitating for 2 or more days on a rotatory (60 rpm)
shaker. The polymer solution was dialyzed (Spectrapor 2, molecular weight cutoff at 12,000 to 14,000 daltons, Spectrum
Medical Industries Inc., Los Angeles) exhaustively over a 5 day
period against a solution containing 154 mM NaCI, 1 mM
CaCl2, and 10 mM HEPES buffer (solution adjusted to pH 7.8).
According to the manufacturer'4 the polymer molecular weight
ranges approximately between 1-105 and 5 107 daltons, with the
"average" weight estimated to be 4.106 daltons. The average
molar concentration of polymer injectate was thus calculated to
be about 0.5 ,M and it was further diluted in vivo to about 15
nm/liter of blood.
Excision and sampling of aorta. Each rabbit was anesthetized with 60 mg/kg sodium pentobarbitol via a marginal ear
vein. A combined midline thoracotomy and laparotomy exposed the heart; 8 ml blood was withdrawn from the right atrium
into a heparinized syringe. The hematocrit was determined and
the remaining blood was centrifuged at 600 g for 20 min, and
plasma was removed and frozen at -200 C for later cholesterol
and triglyceride analyses. The abdominal cavity was examined
for the presence of ascites and for gross abnormalities of the
liver and other organs. A medial lobe liver biopsy sample was
taken for histologic examination after the lungs and viscera had
been photographed in situ. The heart and aorta extending to the
iliac bifurcation, with stubs of its major branches left attached
for purposes of orientation, were dissected free and photographed. The aorta was then divided into 5 or 10 mm segments,
as shown in figure 1.
Morphometric assessment of aortic atherosclerosis. Each
aortic segment was cleansed of blood clots and immersion fixed
in ice-cold 2% gluteraldehyde dissolved in 25 mM Na-cacodylate, 4.7 mM KCI, 2.5 mM CaCl2 2H20, and 0. 8 mM MgC12
at pH 7.35. After 48 hr at 4° C the fixative was decanted, each
tissue specimen was rinsed three times with the cacodylate
vehicle adjusted with NaCI to maintain osmolality, dehydrated
in graded ethanol solutions, and infiltrated overnight in vacuo at
628
23 + 1° C in a glycol methacrylate solution catalyzed with
benzoyl peroxide (PolySciences Inc.). Each sample was embedded in a catalyzed glycol methacrylate solution polymerized
with polyethylene glycol 400 and N,N-dimethylaniline. Sections 2 juM thick from the anterior face of each segment were
stained with Lee's methylene blue-basic fuchsin, coded, and
examined at 25-fold magnification with a Leitz-Wetzler microscope fitted with a drawing tube. The morphometric analysis of
the aorta and its atherosclerotic lesions is described in some
detail elsewhere,15 but a brief description of the protocol follows
below.
The vessel wall and any plaques that were present were traced
first on paper and then on a digitizing tablet. The image of the
vessel, usually partially collapsed, in cross section, was transformed by a microcomputer program to a circular shape amenable to geometric analysis. The linear and areal variables, derived values, and both original and transformed images were
recorded (figure 2). Derived values included the percentages of
atherosclerotic plaque area covering the inner vessel wall and of
luminal occlusion. With the use of geometric calibration figures
with cross-sectional contours similar to those of atherosclerotic
aortas to assess the accuracy of the computer analysis, the mean
SEGMENTS:
ARCH
$;~5
nnm
DESC. THORACIC
COELIAC
RENOMESENTERIC
-10 mm
SUPRAILIAC
,25
RABBIT AORTA
FIGURE 1. Dissection of rabbit aorta into
gular markers)
and
segments
for
regional subdivisions (trianhistologic cross sections.
CIRCULATION
LABORATORY INVESTIGATION-ATHEROSCLEROSIS
A.
SECTION NAME: 2-1
FILE: 39
LA -363.4
LB -321.6
LPE-321.6
LE - 0.0
DA-115.6
8- 102. 0
OP- 69.5
TW- 6.8
AW - 2312
ABP- 8171
AP - 4375
AB - 3795
B. FILE: 70
LA -334.0
LB -306. 8
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
LPE-253.6
LE - 53.2
LPE/LB-100.0%
AP/AW -189. 1%
AP/ABP- 53.5%
ABP/AW-353. 3%
AB/ABP- 46.4%
SECTION NAME: 27-3
DA=106.0
B0 97.6
OP- 88.3
TW- 4.2
AW
ABPAP AB
1429
7481
1354
6126
LPE/LB= 62.6%
AP/AW - 94.7%
AP/ABP- 18.1%
ABP/'AW-523.2%
AB/ABP- 81.8%
FIGURE 2. Illustrative printouts of morphometric data and contours of
actual aortas in cross section (left) and their respective transformations
(right). These cross sections, which are representative of effects of 170
days of a high-cholesterol diet (see figure 7), were obtained from the
aortic arches of control (A) and Separan-treated (B) rabbits.
for the calculations of plaque-wall interface area and
luminal occlusion obtained by five observers were estimated to
be 5% and 13%, respectively, with 95% confidence limits.15
A primary consideration in this comparative study was to
avoid psychological and operational biases with respect to the
two major groups (Separan-treated and untreated rabbits on an
atherogenic diet). This was mainly accomplished by doubleblind coding of the more than 500 histologic slides and rearranging them in slide boxes randomly. The code was broken after all
contours and morphometric data were on paper and only then
were the data collated and assessed.
Statistical analysis. The specific statistical tests applied are
described in the Results section. All values are expressed as the
mean + SD unless otherwise indicated. Changes or differences
are considered to be statistically significant when p < .05.
errors
Results
To assess comparability of dietary intake between
the cholesterol-fed controls and Separan-treated animals, mean daily body weight gain and plasma cholesterol and triglyceride levels were determined. Plasma
cholesterol levels (mmol/liter plasma) in the three
groups just before death were as follows: normal, 3.3
+ 2.6; control, 30.9 + 13.3; and Separan-treated,
Vol. 75, No. 3, March 1987
24.0 ± 13.3. Applying a one-way analysis of variance
(ANOVA) with Duncan's new multiple-range (NMR)
test and p < .05, a statistical difference was established between the normal values and the values from
the two cholesterol groups, but there was no statistical
difference between the control and Separan-treated
groups. There also appeared to be no relationship between prolonged exposure to dietary cholesterol and
plasma cholesterol concentration. Plasma triglyceride
levels (mmol/liter plasma) were not significantly different among the three groups (normal, 0.68 ± 1.89;
control, 1.54 ± 0.98; and Separan-treated, 2.21 +
1.89). Hematocrits for the same groups were 41 + 5,
34 + 2, and 25 + 1, respectively. The microvolumetric method used to measure hematocrit is not considered to be technically reliable for tests on blood
obtained from atherosclerotic animals because of a
poorly defined lipid phase in the hematocrit microtubes. However, the low hematocrit values obtained
for rabbits on a high-cholesterol diet are comparable to
those previously reported.10' 16
The mean daily weight gains of the high-cholesterol
groups were not statistically different, but both groups
gained less weight than rabbits on a normal diet (figure
3). The fact that the Separan group gained at least as
much weight as the control group suggests that the
polymer-treated animals ingested at least as much
chow and cholesterol as their control partners. Although it was difficult to quantify, it was our impression that in the later stages of the experiment the Separan-treated animals appeared distinctly less lethargic
than the control rabbits. Fatty infiltration of the liver
did not seem to be diminished by Separan treatment.
Both high-cholesterol groups became jaundiced, but
the jaundice appeared to develop less rapidly in the
Separan-treated group. One rabbit was excluded from
the control (114 days) group because of strong evidence that it had become chronically ill and its food
intake apparently had been markedly reduced over a 3
week period before termination. This evidence included loss of body weight (observed in only one other
animal, which died), the lowest plasma cholesterol
level found in any cholesterol-fed animal, and all organs nearly free of fat, which was in contrast to the
case in all other animals in the control and Separantreated groups.
After quantification of the atheromas in 26 segmental samples of each aorta, the data were organized to
show the mean value obtained from n samples in each
aortic subdivision: arch (n = 4), descending thoracic
(n = 8), celiac (n = 4), renomesenteric (n = 4), and
suprailiac (n = 6). No plaques were found anywhere
629
FARUQUI et al.
MEAN DAILY WEIGHT GAIN:
25r
D NORMAL,
S
20-
S
15.8 ± 5.6 9
O CONTROL, 3.8 ± 4.0 9
* SEPARAN, 8.6 ± 5.7 9
z 15R
FIGURE 3. Average daily weight gains during con~
ventional ("normal") or high-cholesterol (2%) diets
for normal, control, and Separan-treated rabbits that
were weighed three times a week until the time of
termination. The mean values (+- SD) for each group
are given in the box on the upper right.
I-
X
0 10
0
3:
0
.
.0
0
-J
5
S
0
0
_0
0
0
0
0
-5'
0
1
1
1
50
100
150
200
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
DAYS ON DIET
in the aortas of the rabbits on a conventional diet.
Figure 4, A, shows a comparison of data on luminal
occlusion in aortic subdivisions (%) from rabbits,
matched for duration of diet, in the control and Separan-treated groups. In such a comparison the points
would be expected to cluster along the identity line if
the polymer had no effect on atherogenesis. In fact, 16
of 23 points fell below the line when the luminal occlusion was less than 20% and all 12 points fell below the
line when occlusion was greater than 20%. This skewing suggests that Separan tended to inhibit atherogenesis in all subdivisions of the aorta. A plot of all data
from the two groups, ranked for percent occlusion in
each group and rank-paired, is shown in figure 4, B.
Wilcoxon's signed-ranks test of the hypothesis that the
occlusions in the two groups had the same mean ranks
indicated that the ranks were significantly different.
The plot suggested generally that in the presence of
Separan the onset of plaque formation was delayed and
its rate inhibited.
A two-way ANOVA with p < .05 was applied to the
data from the control and Separan groups to assess the
plaque formations in the five aortic subdivisions. This
test indicated that plaque formation in the aortic arch
was greater than anywhere else in both groups, when
the atheromas were assessed either in terms of vessel
wall-plaque interface area or luminal occlusion. Application of Duncan's NMR test showed no differences
among the remaining four subdivisions in the control
group, but the protective effect of Separan was not the
same for all subdivisions of the descending aorta.
An analysis of covariance (ANCOVA) was performed for each of the five aortic subdivisions compar630
ing the two regression lines describing the relationship
between the area of vessel wall covered by plaque and
the duration of high-cholesterol diet. The slope for the
control group was significantly greater than the slope
for the Separan group in both the aortic arch and the
suprailiac subdivision. This indicates that Separan reduced the rate of atherogenesis in these two regions of
the aorta. When the four subdivisions of the descending aorta were analyzed as a single entity, the slope of
the pooled data was still significantly lower for the
Separan-treated group (table 1).
An ANCOVA comparing the percentage of lumen
occluded in aortas from control and separan-treated
rabbits during the high-cholesterol diet showed that the
slopes were different in each pair of the five aortic
subdivisions (table 2). Figure 5 shows the regression
lines for luminal occlusion as a function of days on a
high-cholesterol diet in the aortic arches from the control and Separan groups. The comparable regression
lines for the descending aorta are shown in figure 6. As
with the wall-plaque interface data, two-way ANOVA
indicated that the aortic arch was most susceptible to
plaque formation, with no differences among the remaining segments of the same aorta.
The protective effect of Separan treatment is indicated in figure 7, which summarizes and compares the
atherogenesis in the presence and absence of polymer
predicted by regression analysis after 170 days of a
high-cholesterol diet. The fractions (expressed as percentages) of vessel wall covered with plaque and lumen occluded by plaque in control and test groups are
illustrated in the solid-line histograms by the relative
heights and areas, respectively; the widths of the histoCIRCULATION
LABORATORY INVESTIGATION-ATHEROSCLEROSIS
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
j20
CONTROL (%)
30
40
50
60
70
90
80
100
FIGURE 4. A, A plot of the mean ( + SE) percentage of lumen occluded in the five subdivisions of each aorta from seven pairs of
rabbits matched for their length of time on a high-cholesterol diet. The subdivisions, with their respective symbols and segment
sample numbers in parenthesis, are the arch (solid circle, 4), descending thoracic (open circle, 8), celiac (open diamond, 4),
renomesenteric (open square, 4), and suprailiac (solid square, 6) aorta. The lack of variance bars on some symbols signifies that
the SE is less than the width or height of the symbol. The point coordinates for the values from control and Separan-treated pairs
are indicated on the abscissa and ordinate, respectively, and the identity line consists of points of equal paired values. The
coordinate scales are expanded for all values under 20%. All data from the two groups, irrespective of aortic subdivisions, are
plotted in B after pairing of ranked occlusion values. The open circles are single pair values and the closed circles represent the
mean of 10 such values, where the SE never exceeds the dimensions of the symbol.
TABLE 1
Efrect of Separan AP-30 on aortic atheromas in cholesterol-fed
rabbits: wall-plaque interface area
Regression slope of wall area,
100 (%/day)
n
Aortic subdivisions
C
S
C
Arch
27
147
56
27
24
40
27
165
63
30
27
45
26±+ 12
Descending
Thoracic
Celiac
Renomesenteric
Suprailiac
TABLE 2
Effect of Separan AP-30 on aortic atheromas in cholesterol-fed
rabbits: luminal occlusion
35 + 5
17 + 9
35 11
48 13
52 10
S
A
A
A
-12 + 12
13 + 5
8+7
17 + 8
12 14
22 9
n = total number of histologic sections obtained from seven control
(C) or eight Separan-treated (S) animals.
Ap < .05.
Vol. 75, No. 3, March 1987
Regression slope of
occlusion, 100 (%/day)
n
Aortic subdivisions
C
S
C
Arch
Descending
Thoracic
27
147
56
27
24
40
27
165
63
30
27
45
32±6
Celiac
Renomesenteric
Suprailiac
27 2
20±+-5
18 4
41 ± 7
35 5
S
A
A
A
A
A
A
8±5
3 1
1 1
4±+ 2
2+3
5+3
n = total number of histologic sections from control (C) or Separantreated (S) group.
Ap < .05.
631
FARUQUI et al.
70r
AORTIC ARCH:
0
U CONTROL
60[
U
0 SEPARAN
0
0
50o
z
0
0t)
-J 40k
0
0 30k
0
J
z
W_j
20k
U
--* t°~~
0
10~
o0
0
0
0
50
*
*
100
DAYS ON DIET
150
200
FIGURE 5. Development of atheromas during a high-cholesterol diet
in the aortic arches of control and Separan-treated rabbits. The slopes of
the regression lines for luminal occlusion in the control and Separantreated groups are 0.32 - 0.06 and 0.08 -+- 0.05%/day, respectively.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
grams are an index of plaque thickness. In relative
terms polymer treatment was most effective in the descending thoracic aorta, where the occlusion was only
approximately one-tenth that in the control group, and
least effective in the aortic arch, where the occlusion
was about 43% of that in untreated animals. In all
subdivisions of the aorta the extent as well as the thickness of the plaque was markedly less in the test group
than it was in the control group.
Discussion
Mostardi et al.'0 first proposed that drag-reducing
polymers, known to laminarize flow both in vitro4' 17. 1
and in vivo,8 might inhibit atherosclerosis by dampening flow disturbances. Photographs of plaque formations in exposed aortic intimae from Separan-treated
and untreated rabbits on a high-cholesterol diet showed
dramatic differences, in support of Mostardi's hypothesis. The objective of our study was to test this hypothesis, quantifying the effects of Separan treatment by
morphometric analysis. This analysis was specifically
designed to determine not only the effect of Separan on
the fraction of aortic intima covered by atheromas, but
also to assess the fraction of lumen occluded with and
without therapy. The latter assessment is infrequently
attempted in the aorta by quantitative techniques.
However, it is therapeutically the most relevant variable, because vascular blood flow is related exponentially to luminal cross-sectional area.
A problem with the experimental protocol concerned the difficulty of maintaining the control animals
on a prolonged high-cholesterol diet. It would be preferable to compare the atheromas in Separan-treated
and untreated rabbits that had all been on a 6 month
high-cholesterol diet. However, given the morbidity of
the control rabbits this would have required a much
larger colony of animals than we had anticipated. Nevertheless, statistical analyses establish with high probability that Separan AP-30 is a powerful antiatherogenic agent under the conditions applied in the present
experiments, predicting that Separan can inhibit
plaque formation to half that expected in the aortic arch
and to about one-fifth that expected in the descending
aorta.
The impression that Separan-treated animals were
distinctly healthier than untreated animals was a subjecPERCENT OCCLUSION
(HISTOGRAM AREA):
0 CO
~CONTROL, N=7
ae
100r
SEPARAN,TN=8
SEPARAN,
N ~~
e
_
r.T-O
TOTAL
_ _S__
1
OCCLUSION. lODi
60
DESCENDING AORTA:
50
z
0 40
-J
0 CONTROL, N=7
* SEPARAN, N=8
(19-22 SECTIONSIPOINT)
1._
~
SEMNS
~
14
TOAI
ENE IC
ILA
0)
0
30
0
z
W 20
-J
10
0
0
50
100
DAYS ON DIET
150
200
FIGURE 6. Development of atheromas during a high-cholesterol diet
in the descending aortas of control and Separan-treated rabbits. Each
point represents the mean value (+ SD) from 19 to 22 sections. The
slopes of the regression lines for the luminal occlusions in the control
and Separan-treated groups are 0.27 + 0.02 and 0.03 + 0.01 %/day,
respectively.
632
FIGURE 7. Atheroma-wall interface area and luminal occlusion in
Separan-treated and control (untreated) rabbits on a high-cholesterol
diet. Summary of regression values for each aortic segment after 170
days of the diet. The height of each histogram represents the percentage
of vessel wall interfacing with atherosclerotic plaque (As), where the
broken-line histogram represents total occlusion of a given vessel lumen. The histogram area is directly related to the cross-sectional area of
plaque relative to vessel lumen, with the percentage of lumen that is
occluded (A0) indicated numerically. The x axis is the ratio A0/As, an
index of plaque thickness. The variables are based on the mean values
obtained from three to eight cross sections per aortic subdivision in each
of eight polymer-treated and seven untreated animals.
CIRCULATION
LABORATORY INVESTIGATION-ATHEROSCLEROSIS
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
tive assessment. However, the fact that all eight cohorts had to be terminated because of the actual or the
apparent impending demise of the control member of
the cohort gave an objective dimension to the assessment. This was further supported by the greater gain of
body weight observed in all but one Separan-treated
animal relative to its cohort control partner. Parenthetically, although the amount of food ingested was not
quantified, these findings all but exclude the possibility that the observed atherosclerotic differences between the two groups on atherogenic diets were related
to differences in the amounts of cholesterol ingested.
The finding that plasma cholesterol levels were not
significantly different in the two groups is in accord
with this suggestion and implies that the effect of Separan is not dependent on differences in gastrointestinal absorption of cholesterol.
Evidence continues to mount that atherosclerosis
does not develop haphazardly, but tends to appear at
specific geometric sites19-22 associated with various
types of flow disturbance.'9' 23-25 The term "flow disturbance" is used here in a general sense, incorporating
the many types of flows reported to occur at atherosclerotic sites that are distinguished by marked departures from Poiseuillean character, including turbulence, eddying, swirling, oscillation, vibration,
separation, stagnation, and other nonlaminar or unstable flows. Studies of flow patterns in tubes simulating
the vasculature have already established the instability
of flows through nonlinear conduit systems, particularly when flows are pulsatile and the tubes are elastic.
The variations observed in correlations of atherogenic
site and vascular geometry are not surprising, given the
relatively small alterations in conduit geometry that are
capable of causing major changes in flow patterns.26
The controversy surrounding the relationship between atherogenesis and high27 versus low28 shear
stress seems to have subsided with growing evidence
that low shears stimulate and moderately high shears
protect against the formation of atheromas.24 2830
Fry's 27 demonstration of acute endothelial injury in the
presence of very high shear stress is likely to be of less
clinical pertinence than his observation of endothelial
changes in regions of low shear combined with strong
turbulence. Caro3' also has refined his hypothesis of
shear-dependent mass transport in recognition of the
susceptibility of blood-wall transport to nonsteady
shear stresses.19, 32, 3 These and similar findings are
compatible with a correlation between flow disturbance and atherogenesis. In this connection it
might be useful to consider the hypothesis that an
increased heart rate might be a factor in atherogeneVol. 75, No. 3, March 1987
sis, because oscillatory flow is a type of flow disturbance that has been repeatedly shown to worsen
atherosclerosis.19' 31-33
The relationship between heart rate and atherogenesis might be difficult to elucidate without quantitative
experimentation specifically designed to control the
variability of vascular geometry, cholesterol intake
and metabolism, blood pressure, and other complicating factors of atherogenesis. In monkeys on a highcholesterol diet, some subject to surgical ablation of
the sinoatrial node, the development of coronary arterial lesions and stenosis was significantly less in animals with heart rates below the preoperative mean than
in those with rates above that mean.34 The retardation
of atherosclerosis was unrelated to blood pressures or
plasma cholesterol levels, which were not significantly
different in the two groups. Of particular interest in
this regard is the recent report of Blumlein et al. ,' who
quantitated the effect of several drugs on the area of
aortic intima covered by plaque in cholesterol-fed rabbits. Although these investigators concluded that an
antiatherogenic action of verapamil could not be explained by its '"hemodynamic" actions, or more specifically its effect on blood pressure, in fact a linear
plot of mean plaque areas versus heart rates from seven
groups yields a regression line with a correlation coefficient of .84. An analysis of the original data would be
required to assess the statistical significance of this
correlation.
Since 1971, when Hoyt36 listed over 20 linear polymers of high molecular weight (> 1 10' daltons) demonstrating the Toms phenomenon, additional natural37
and synthetic38 macropolymers with the molecular attributes necessary for polymer drag reduction have
become available. However, only four such polymers
Separan,5 Polyox,7 an okra polysaccharide extract,6
and a calf thymus deoxyribonucleic acid4- have demonstrated drag-reduction in blood flow in vitro. Three
of these polymers - Separan,39 40 Polyox,* and an
okra extract identified as a rhamnogalactogalacturonan9, 11 -augment cardiac output in association with
a marked reduction in peripheral resistance in experimental animals. These macromolecules are characterized by extraordinary linear dimensions, with molecular lengths much longer than those of drag-reducing
polymers that have no effect on blood flow, and in all
cases their lengths exceed the diameters of several
erythrocytes. The hemodynamic effects of these substances are compatible with polymer drag reduction,39' 40 but not with inotropic or vasoactive mecha*Polimeni PI, Ottenbreit B: Unpublished observation.
633
FARUQUI et at.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
nisms.1 Given the diversity of the chemical
composition of these macropolymers and their physical similarities, it is difficult to avoid the conclusion
that their hemodynamic effects reflect a physical rather
than a conventional pharmacologic phenomenon. It
remains to be seen whether Mostardi's suggestion that
the antiatherogenic action of Separan is based on a
physical mechanism is reinforced or weakened after
one or more of the other hemodynamically active macropolymers are tested for therapeutic effect. However, given the ample evidence that intravascular flow
disturbance is a significant factor in atherogenesis19' 27 32, 42 43 and that Separan dampens flow disturbances,8 17. 44-47 such a mechanism is compatible with a
large body of work in medicine and fluid dynamics.
If drag-reducing polymers other than Separan also
demonstrate antiatherogenic effects in future studies, it
would strongly support the view48 that hemodynamics
play a primary role in causing atherosclerosis and that
a novel therapeutic approach is possible against this
disease. In principle this approach would have the advantage of ameliorating the pathologic condition in
two distinct ways. The immediate effect would be to
facilitate blood flow through partially occluded vessels
and enhance cardiac output without commensurate
stress on the heart. The long-term effect would be to
inhibit atherogenesis, or perhaps, given the dynamic
nature of atherosclerosis, cause a regression of the
atheromas. Recent findings in monkeys49 indicate that
although diet or drug therapies correcting hyperlipidemia cause much of the stainable lipid to disappear from
atherosclerotic lesions and reduce the size of necrotic
centers, the vascular lumens do not enlarge, presumably because of scarring and concentric contraction. If
this occurs in humans as well, then drag-reducing
polymers might be unique among antiatherogenic
agents in facilitating blood flow through such vascular
constrictions, as they do in constricted pipes,6 after
atherogenesis has been suppressed or atherosclerosis
reversed.
We thank Mr. S. Selvathurai for technical contributions and
Ms. C. Baraniuk for secretarial assistance. We are also grateful
to Drs. S. A. Carter, M. Cohen, R. J. Hoeschen, and G. P.
Sharma for their critical reading of an early draft of the manuscript. Statistical analyses were made upon the recommendations of Mrs. M. S. Cheang and Mr. R. Tate. The Separan AP30 was a gift from Dow Chemical Co., Midland, MI.
3. Sellin RHJ, Hoyt JW, Scrivener 0: The effect of drag-reducing
additives on fluid flows and their industrial applications: I. Basic
aspects. J Hydraulic Res 20: 29, 1982
4. Hoyt JW: Turbulent-flow properties of deoxyribonucleic acid solutions. Nature 211: 170, 1966
5. Greene HL, Nokes RF, Thomas LC: Biomedical implications of
drag reducing agents. Biorheology 7: 221, 1971
6. Castro WE: Reduction of flow friction with polymer additives.
Water Resources Research Institute, Clemson University, Report
No. 24, 1972
7. Stein PD, Parsons ED, Blick EF: Modifications of dynamic flow
properties of turbulently flowing human blood by long chain polymers. Med Res Eng 11: 6, 1972
8. Mostardi RA, Greene HL, Nokes RE, Thomas LC, Lue T: The
effect of drag-reducing agents on stenotic flow disturbances in
dogs. Biorheology 13: 137, 1976
9. Polimeni PI, Al-Sadir J, Cutilletta AF: Rhamnogalactogalacturonan (RGGu): a drug increasing cardiac output in the rat, perhaps
by a novel mechanism. Physiologist 20: 75, 1977 (abst)
10. Mostardi RA, Thomas LC, Greene HL, Van Esson F, Nokes RF:
Suppression of atherosclerosis in rabbits using drag reducing polymers. Biorheology 15: 1, 1978
11. Polimeni PI, Al-Sadir J. Cutilletta AF: Polysaccharide for enhancement of cardiac output. Washington, D.C., 15 May 1979, US
Patent 4, 154, 922
12. Schettler G, Nerem RM, Schmid-Schonbein H, Morl H, Diehm H,
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
References
1. Hoyt JW: The effect of additives on fluid friction. Trans Am Soc
Mech Eng D94: 258, 1972
2. Berman NS: Drag reduction by polymers. Annu Rev Fluid Mech
10: 47. 1978
634
29.
30.
editors: Fluid dynamics as a localizing factor for atherosclerosis.
Berlin, 1983, Springer-Verlag
Greene HL, Mostardi RF, Nokes RF: Effects of drag reducing
polymers on initiation of atherosclerosis. Polymer Eng Sci 20: 499,
1980
Separan polymers. Midland, MI, 1981, Dow Chemical Co.
Polimeni PI, Cunningham P, Otten MD, McCrea D: Morphometric
quantification of atherosclerotic plaques by computer-assisted image-analysis of histographs. Comput Biomed Res (in press)
Lee RJ, Zaidi IH, Baky SH: Pathophysiology of the atherosclerotic
rabbit. Environ Health Perspect 26: 225, 1978
Gadd GE: Effects of drag-reducing additives on vortex stretching.
Nature 217: 1040, 1968
Berner C, Scrivener 0: Drag reduction and structure of turbulence
in dilute polymer solutions. Prog Astronaut Aeronaut 72: 290,
1980
Ku DN, Giddens DP, Zarins CK, Glagov S: Pulsatile flow and
atherosclerosis in the human carotid bifurcation. Positive correlation between plaque location and low and oscillating shear stress.
Arteriosclerosis 5: 293, 1985
Velican D, Velican C: Accelerated atherosclerosis in subjects with
some minor deviations from the common type of distribution of
human coronary arteries. Atherosclerosis 40: 309, 1981
Halon DA. Sapoznikov D, Lewis BS, Gotsman MS: Localization
of lesions in the coronary circulation. Am J Cardiol 52: 921, 1983
Bargeron CB, Hutchins GM, Moore GW, Deters OJ, Mark FF,
Friedman MH: Distribution of the geometric parameters of human
aortic bifurcations. Arteriosclerosis 6: 109, 1986
Gutstein WH, Farrell GA, Armellini C: Blood flow disturbance and
endothelial cell injury in preatherosclerotic swine. Lab Invest 29:
134, 1973
Friedman MH, Deters OJ, Mark FF, Bargeron CB, Hutchins GM:
Arterial geometry affects hemodynamics. Atherosclerosis 46: 225,
1983
Svindland A, Walloe L: Distribution pattern of sudenophilic
plaques in the descending thoracic and proximal abdominal human
aorta. Atherosclerosis 57: 219, 1985
Malcolm AD, Roach MR: Flow disturbances at the apex and lateral
angles of a variety of bifurcation and their role in development and
manifestations of arterial disease. Stroke 10: 335, 1979
Fry DL: Acute vascular endothelial changes associated with increased blood velocity gradients. Circ Res 22: 165, 1968
Caro CG, Fitz-Gerald JM, Schroter RC: Arterial wall shear and
distribution of early atheroma in man. Nature 223: 1159, 1969
Zarins CK, Bomberger MD, Glagov S: Local effects of stenosis:
increased flow velocity inhibits atherogenesis. Circulation 64
(suppl II): II-221, 1981
Zarins CK, Giddens DP, Bharadvaj BK, Sottiurai VS, Mabon RE,
Glagov S: Carotid bifurcation atherosclerosis: quantitative correla-
CIRCULATION
LABORATORY INVESTIGATION-ATHEROSCLEROSIS
31.
32.
33.
34.
35.
36.
37.
38.
39.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
40.
tion of plaque localization with flow velocity profiles and wall
shear stress. Circ Res 53: 502, 1983
Caro CG: Mechanical factors in atherogenesis. In Hwang NHC,
Normann NA, editors: Cardiovascular flow dynamics and measurements. Baltimore, 1977, University Park Press, p 473
Chien S, Lee MML, Laufer LS, Handley DA, Weinbaum S, Caro
CG, Usami S: Effect of oscillatory mechanical disturbance on
macromolecular uptake by arterial wall. Arteriosclerosis 1: 326,
1981
Friedman MH, Hutchins GM, Bargeron CB, Deters OJ, Mark FF:
Correlation between intimal thickness and fluid shear in human
arteries. Atherosclerosis 39: 425, 1981
Beere PA, Glagov S, Zarins CK: Retarding effect of lowered heart
rate on coronary atherosclerosis. Science 226: 180, 1984
Blumlein SL, Sievers R, Kidd P, Parmley WW: Mechanism of
protection from atherosclerosis by verapamil in the cholesterol-fed
rabbit. Am J Cardiol 54: 884, 1984
Hoyt JW: Drag-reduction effectiveness of polymer solutions in the
turbulent-flow rheometer: a catalogue. Polymer Lett 9: 851, 1971
MacGregor EA, Greenwood CT: Polymers in nature. New York,
1980, John Wiley & Sons
Molyneux P: Water-soluble synthetic polymers: properties and behavior. Boca Raton, 1983, CRC Press, vol II
Polimeni PI, Ottenbreit B, Coleman P: Enhancement of aortic
blood flow with a linear anionic macropolymer of extraordinary
molecular length. J Mol Cell Cardiol 17: 721, 1985
Coleman PB: Enhancement of cardiac output by the drag-reducing
Vol. 75, No. 3, March 1987
41.
42.
43.
44.
45.
46.
47.
48.
49.
agent, Separan AP-273, a linear polyelectrolyte of extraordinary
length, thesis. University of Manitoba, 1985
Bose R, Bose D, Polimeni PI: Do polymer drag-reducing agents
enhance cardiac output by flow laminarization? Can Fed Biol Soc
28: 76, 1985 (abst)
Wesolowski SA, Fries CC, Sabini AM, Sawyer PN: The significance of turbulence in hemic systems and in the distribution of the
atherosclerotic lesion. Surgery 57: 155, 1965
Gutstein WH, Farrell GA, Schneck DJ: In vivo demonstration of
junctional blood flow disturbance by hot wire anemometry.
Atherosclerosis 11: 485, 1970
Barnes HA, Townsend P, Walters K: Flow of non-Newtonian
liquids under a varying pressure gradient. Nature 224: 585, 1969
Latto B, Shen CH: Effect of dilute polymer solution injection on
external boundary layer phenomena. Can J Chem Eng 48: 34, 1970
Walters K, Barnes HA, Dodson AG: Drag reduction in unconventional flow situations. Chem Eng Prog Symp Ser III 67: 1, 1971
Rudd MJ: Velocity measurements made with a laser dopplermeter
on the turbulent pipe flow of a dilute polymer solution. J Fluid
Mech 51: 673, 1972
Stehbens WE: The role of hemodynamics in the pathogenesis of
atherosclerosis. Prog Cardiovasc Dis 18: 89, 1975
Wissler RW, Vesselinowich D, Davis HR, Lambert PH, Bekermeier M: A new way to look at atherosclerotic involvement of the
artery wall and the functional effects. Ann NY Acad Sci 454: 9,
1985
635
Protection against atherogenesis with the polymer drag-reducing agent Separan AP-30.
F I Faruqui, M D Otten and P I Polimeni
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Circulation. 1987;75:627-635
doi: 10.1161/01.CIR.75.3.627
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1987 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/75/3/627
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/