The effect of hypertonic saline resuscitation on responses to severe hemorrhagic shock by the skeletal muscle, intestinal, and renal microcirculation systems: seeing is believing.

Abstract

Decompensated hemorrhagic shock is often refractory to resuscitation, and we show here that it is associated with loss of vascular tone in skeletal muscle precapillary arterioles. We tested the hypothesis that microvascular derangements in the skeletal muscle, intestinal, and renal microcirculation systems would be reversed by initial hypertonic saline-dextran infusion.
Male Sprague-Dawley rats underwent precollicular brain stem transection without anesthesia for study. Parameters measured by in vivo videomicroscopy included cardiac output, mean arterial pressure, and microvascular responses in the skeletal muscle, ileum, and renal (i.e., the hydronephrotic kidney) microcirculation systems. Hemorrhaged was induced to a mean arterial pressure of 50 mmHg until decompensation occurred. The rats were then initially resuscitated with (1) 4 mL/kg 7.5% NaCl in 6% dextran 70, (2) 33 mL/kg .9% NaCl in 6% dextran 70, or (3) 33 mL/kg .9% NaCl. Twenty minutes later they received shed blood plus 33 mL/kg .9% NaCl to maintain mean arterial pressure at baseline levels.
Decompensated hemorrhagic shock decreased cardiac output to between 24% and 35% of baseline values and profoundly decreased microvascular blood flow to between 10% and 19% of baseline. At the completion of resuscitation cardiac output increased to greater than baseline in all groups. Microvascular blood flow increased toward baseline transiently but then progressively deteriorated to between 36% and 69% of baseline in the 3 tissues. There was no significant difference between the three resuscitative fluids.
Despite return of cardiac output to greater than baseline levels, muscle, intestinal, and renal microvascular blood flows remained significantly depressed. Hypertonic saline and/or dextran did not improve these deficits.

Full text

Page 1

The effect of hypertonic saline resuscitation on responses to severe
hemorrhagic shock by the skeletal muscle, intestinal, and renal
microcirculation systems: seeing is believing
Henry M. Cryer, M.D., Ph.D.*, John Gosche, M.D., Jeff Harbrecht, M.D.,
Greg Anigian, M.D., Neal Garrison, M.D.
Department of Surgery, University of California Los Angeles Medical Center, 10833 Le Conte Ave., 72-178 CHS, Los Angeles, CA 90095, USA
Manuscript received April 13, 2005; accepted manuscript April 15, 2005
Abstract
Background: Decompensated hemorrhagic shock is often refractory to resuscitation, and we show here that it is associated with loss of
vascular tone in skeletal muscle precapillary arterioles. We tested the hypothesis that microvascular derangements in the skeletal muscle,
intestinal, and renal microcirculation systems would be reversed by initial hypertonic saline–dextran infusion.
Methods: Male Sprague-Dawley rats underwent precollicular brain stem transection without anesthesia for study. Parameters measured by
in vivo videomicroscopy included cardiac output, mean arterial pressure, and microvascular responses in the skeletal muscle, ileum, and
renal (i.e., the hydronephrotic kidney) microcirculation systems. Hemorrhaged was induced to a mean arterial pressure of 50 mm Hg until
decompensation occurred. The rats were then initially resuscitated with (1) 4 mL/kg 7.5% NaCl in 6% dextran 70, (2) 33 mL/kg .9% NaCl
in 6% dextran 70, or (3) 33 mL/kg .9% NaCl. Twenty minutes later they received shed blood plus 33 mL/kg .9% NaCl to maintain mean
arterial pressure at baseline levels.
Results: Decompensated hemorrhagic shock decreased cardiac output to between 24% and 35% of baseline values and profoundly
decreased microvascular blood flow to between 10% and 19% of baseline. At the completion of resuscitation cardiac output increased to
greater than baseline in all groups. Microvascular blood flow increased toward baseline transiently but then progressively deteriorated to
between 36% and 69% of baseline in the 3 tissues. There was no significant difference between the three resuscitative fluids.
Conclusions: Despite return of cardiac output to greater than baseline levels, muscle, intestinal, and renal microvascular blood flows remained
significantly depressed. Hypertonic saline and/or dextran did not improve these deficits. © 2005 Excerpta Medica Inc. All rights reserved.
Keywords: Cardiac output; Decompensated hemorrhagic shock; Dextran; Hypertonic saline; Intestinal microcirculation; In vivo videomicroscopy; Renal
microcirculation; Skeletal muscle microcirculation
It is a great honor to present at a meeting celebrating Dr.
Polk’s retirement the studies in the late 1980s that began our
understanding of the microcirculatory responses to hemor-
rhagic shock. His influence and the opportunities he pro-
vided us are directly responsible for the ideas and studies I
will present here.
When I asked to take 6 months off in the third year of my
surgical residency in 1983, Dr. Polk encouraged me to work
with Lewis Flint, who was taking a sabbatical with Pat Harris
in the departments of Physiology and Biophysics at the Uni-
versity of Louisville. During that time we performed experi-
ments showing that the decompensatory phase of hemorrhagic
shock was associated with dilation and loss of tone of the
terminal precapillary arterioles in the skeletal muscle micro-
circulation system and that this loss of tone was not reversed
with conventional resuscitation using blood and isotonic crys-
talloid resuscitation [1]. After I finished my residency in 1985,
Dr. Polk made it possible for me to work with Pat Harris and
Neal Garrison for 2 years, during which time I attained my
doctorate and developed a state-of-the-art laboratory at the
Louisville Veterans Administration Hospital (VAH) for study-
ing microcirculation. After that Dr. Polk invited me to join the
faculty at the University of Louisville. During that time I
collaborated with a number of faculty and residents in our
microvascular laboratory at the VAH. The studies that I
present here were performed from 1987 through 1990 in col-
* Corresponding author. Tel.: ?1-310-825-5981; fax: ?1-310-206-2472.
E-mail address: hcryer@mednet.ucla.edu
The American Journal of Surgery 190 (2005) 305–313
0002-9610/05/$ – see front matter © 2005 Excerpta Medica Inc. All rights reserved.
doi:10.1016/j.amjsurg.2005.05.032

Page 2

laboration with John Gosche, Jeff Harbrecht, Greg Anigian,
Bill Flynn, Neal Garrison, Pat Harris, Jean Jenkins, and Re-
becca Galloway. They represent our initial attempts to restore
microcirculation blood flow to the skeletal muscle, intestinal,
and renal circulation systems after decompensated hemor-
rhagic shock with hypertonic saline. These studies led to the
eventual successful restoration of microvascular blood flow
with pentoxifylline [2]. They take on added relevance today
with the resurging interest in hypertonic saline resuscitation for
its anti-inflammatory properties and its potential in battlefield
situations [3].
Prolonged hemorrhage and massive blood transfusion lead
to decompensated hemorrhagic shock when normal compen-
satory peripheral vasoconstrictor responses fail and peripheral
vasodilation markedly decreases systemic vascular resistance.
This peripheral vasodilation is typically unresponsive to vaso-
pressor agents and massive quantities of isotonic crystalloid
solution and usually results in death even if bleeding has been
controlled. In 1980, Velasco et al [4] demonstrated that small
volumes of very hypertonic saline solutions were as effective
as large volumes of isotonic crystalloid solutions for intravas-
cular volume expansion after hemorrhagic shock. De Felippe
et al [5] reported that small volumes of hypertonic saline could
reverse clinically decompensated hemorrhagic shock in pa-
tients and prevent mortality. In 1986, Kramer et al [6] demon-
strated that addition of dextran 70 plus hypertonic saline was
more effective than hypertonic saline alone in resuscitation
from compensated hemorrhagic shock, but no data existed
regarding decompensated hemorrhagic shock. Rocha-e-Silva
et al [7] reported that at the macroscopic level, hypertonic
saline causes a differential interorgan vascular response after
hemorrhage,thusallowingvasoconstrictiontooccurinskeletal
muscle, although vasodilation occurs in the splanchnic and
renal circulation. Finally, around that time we [8] found that
restoration of cardiac output (CO) to greater-than-normal lev-
els conferred a significant survival advantage in patients resus-
citated from hemorrhagic shock. These observations led us to
the working hypothesis that decompensated hemorrhagic
shock leads to differential responses in different microvascular
beds and that resuscitation with hypertonic saline may take
advantage of this to allow successful restoration of microvas-
cular flow and resuscitation. We developed a rat model that
allowed the measurement of systemic CO and microvascular
flow measurements in skeletal muscle, small intestine, kidney,
and liver that was ideally suited to test this hypothesis.
Materials and Methods
General animal preparation
Sprague-Dawley male rats weighing between 160 and
170 g were used for these studies. Rats were acclimated for
1 to 2 weeks in an Association for Assessment of Labora-
tory Animal Care–approved animal care center where they
were fed standard Purina Rat Chow and water ad libitum
until 12 hours before each experiment when food but not
water was withdrawn. All animals were handled according
to Association for Assessment of Laboratory Animal Care
guidelines for humane care and treatment. To avoid the
effects of anesthesia on the microcirculatory response to
hemorrhage, rats underwent precollicular brain stem tran-
section according to the technique described in a previous
publication [1]. This procedure renders animals cortically
unconscious, thus allowing study without the potential con-
founding effects of drug anesthesia on microcirculation sys-
tems. The animals were then allowed 5 to 6 hours to com-
pletely recover from drug anesthesia before beginning the
experiment. Rats then underwent tracheostomy for airway
control. The left femoral artery was cannulated for blood
pressure and heart rate measurements as well as withdrawal
of blood for the hemorrhage protocol. The left femoral vein
was cannulated to allow injection of resuscitative solutions.
For measurement of CO by transpulmonary thermodilution,
the right common carotid artery was cannulated with a
thermistor, the tip of which was positioned in the aortic root
just above the aortic valve. The right jugular vein was
cannulated with PE-50 tubing, which was positioned in the
superior vena cava just proximal to the right atrium. For
generation of CO curves, a 40-?L saline bolus was injected
by way of this jugular cannula, and changes in arterial blood
temperature were recorded as temperature time curves on a
polygraph.
Skeletal muscle preparation
For skeletal muscle experiments, the rat cremaster mus-
cle was prepared for in vivo videomicroscopy by a tech-
nique described in detail a previously publication [1].
Briefly, the cremaster muscle was dissected with its vascular
and neural supply intact and spread over an optical port in
a tissue bath filled with a modified Krebs solution. The
animal and bath were then positioned on the stage of a
trinocular microscope for transillumination. Nitrogen and
carbon dioxide were continuously bubbled through the bath
to maintain bath PO2at 15 to 25 mm Hg and bath carbon
dioxide at 35 to 45 mm Hg. Bath temperature was main-
tained at normal body temperature by a heating coil in the
bath. Rectal temperature was maintained in the normal
range with a back heating pad.
The microvascular anatomy of the cremaster muscle was
directly observed with a microscope that had 20? water
immersion objectives and 10? eye pieces. Microvascular
images were transmitted by way of a video camera to a
television monitor and videocassette recorder using a lens
system that produced a permanent videotape record of the
vascular anatomy at a magnification of 1500?. Vascular
anatomy of the cremaster muscle was defined by vessel
branch order.
To calculate microvascular blood flow, red blood cell
velocity was measured in first-order arterioles (A1) using an
optical Doppler velocimeter. Blood flow through the A1
306
H.M. Cryer et al. / The American Journal of Surgery 190 (2005) 305–313

Page 3

arteriole was calculated from the simultaneous measure-
ments of red blood cell velocity and A1 diameter at each
time interval during the experiment as described in a pre-
vious publication [1]. In each experiment, an A1 through A4
sequence of arterioles were selected so that the studied
arteriolar branches arose from the A1 arteriole in which
velocity was measured.
Small-intestine preparation
For small-intestine experiments, the small intestine was
prepared for in vivo videomicroscopic observation as pre-
viously described [9]. Briefly, the abdomen was opened,
and a short segment of small intestine was exteriorized and
opened along its antimesenteric border using electrocautery.
The opened intestinal loop was then suspended, with neu-
rovascular connections intact and the serosal surface up,
over an optical window in a specially designed tissue bath.
The intestine was suffused with a modified Krebs solution
as described previously and covered with washed cello-
phane. Intestinal microvascular anatomy was defined ac-
cording to the vessel branch order similar to skeletal muscle.
Renal microvascular preparation
For experiments looking at the unilateral renal microvas-
culature, chronic hydronephrosis was induced in 100-g male
Sprague-Dawley rats by ligation of the left ureter under
sterile conditions as previously described [10]. Hydrone-
phrosis was allowed to progress for 6 weeks when rats
weighed between 350 and 450 g. At the time of acute
experiments, the hydronephrotic kidney was exposed
through a left-flank incision and dissected with its neuro-
vascular pedicle intact. The kidney was bisected along the
greater curvature in an avascular plane, and the optimal half
was suspended over the optical port of a specially designed
stage with the outside of the kidney facing upward. The
kidney was suffused with warmed Krebs solution and sealed
in cellophane wrap. The microcirculation of the kidney was
positioned on the stage of a trinocular microscope for direct
in vivo video microscopy as described earlier. The micro-
vascular anatomy of the kidney flows from renal artery
branches into interlobar arteries, which terminate in arcuate
arteries and give rise to the afferent arterioles, which in turn
enter individual glomeruli. Efferent arterioles exit each glo-
merulus.
Experimental protocol
After the baseline (BL) readings were obtained, the an-
imals were heparinized (50 U), and hemorrhaged was in-
duced by rapid withdrawal of blood for 3 to 5 minutes to a
mean arterial pressure (MAP) of 50 mm Hg. Additional
blood was withdrawn as required to maintain blood pressure
at 50 mm Hg throughout the hypotensive period until de-
compensation occurred. Decompensation was defined as the
point at which blood pressure could no longer be maintained
without reinfusion of shed blood, and it was frequently
associated with decreased CO and changed respiratory pat-
tern. Measured parameters were recorded at the time of
decompensation, and the animals were then initially resus-
citated with either (1) 4 mL/kg 7.5% NaCl, (2) 4 mL/kg
7.5% NaCl in 6% dextran 70 (HTSD), or equivalent solute
loads as either (3) 33.3 mL/kg .9% NaCl in 6% dextran 70
(NSD) or (4) 33.3 mL/kg .9% NaCl (NS). Initial experi-
ments with skeletal muscle were performed without addi-
tional blood or saline resuscitation, and observations were
continued at 10-minute intervals until the animals died. In
subsequent experiments, rats were further resuscitated by
return of shed blood and additional crystalloid as .9% NaCl
as required to maintain MAP at or near BL values. Obser-
vations were then continued for an additional hour or until
death occurred. In the experiment with rats receiving 4
mL/kg 7.5% NaCl without dextran, the first 3 animals died
?10 minutes before measurements could be made, and we
aborted this protocol.
Statistical analysis
For statistical analysis, all data were expressed as percent
change from BL values. Group means, SDs, and SEs were
calculated for BL values and percent changes. Time points
for analysis included BL values, decompensation, 10 min-
utes, and 20 minutes after initial resuscitation as well as
after infusion of blood and NS. In addition, maximal com-
pensation was defined as the point during the hypotensive
period when MAP was maximally maintained despite de-
creasing cardiac output levels (generally this was closely
associated with that point at which there was a decrease in
requirement for further blood withdrawal to maintain MAP
at 50 mm Hg).
Changes between and within treatment groups were an-
alyzed initially using mixed-measures analysis of variance.
Within-group changes from BL values were subsequently
analyzed by analysis of variance for repeated measures with
Dunnett’s test. Differences in BL values, responses to re-
suscitation, and percentage changes at defined time points
between groups were analyzed by one-way analysis of vari-
ance with Tukey Honestly Significant Difference follow-up
test as indicated. Statistical correlations between variables
were calculated by way of multiple regression analysis.
Differences were accepted as significant P ? .05. Unless
otherwise noted, values are expressed as group means plus/
minus SEM.
Skeletal muscle
Baseline data for general, hemodynamic, and microvas-
cular variables are listed in Table 1. No important differ-
ences were noted between BL values in groups receiving
NS versus HTSD. Hemorrhage caused similar systemic and
307
H.M. Cryer et al. / The American Journal of Surgery 190 (2005) 305–313

Page 4

microvascular responses in both groups. Hemorrhage de-
creased MAP to 54% ? 1% in HTSD and 52% ? 1% in NS
rats during the compensatory phase of hemorrhagic hypo-
tension and 44% ? 4% of BL in HTSD and 42% ? 3% of
BL in NS rats during the decompensatory phase of hemor-
rhagic shock. Hemorrhage decreased CO to 56% ? 6% of
BL in HTSD and 60% ? 7% of BL in NS rats during the
compensatory phase and decreased CO significantly further
during the decompensatory phase to 37% ? 6% BL in
HTSD and 27% ? 3% of BL in NS rats.
Resuscitation with 7.5% NaCl in 6% dextran 70 in
HTSD and .9% NaCl in NS rats initially (10 minutes)
increased CO and MAP to a similar degree. However, this
effect was short lived in the NS group, and all animals died
within 25 minutes of initiating resuscitation. In contrast,
resuscitation with hypertonic saline–dextran HTSD allowed
all animals to survive 40 minutes, 3 animals to survive for
50 minutes, and 2 animals to survive ?1 hour.
Microvascular responses to hemorrhage were also simi-
lar in the 2 groups. Hemorrhage decreased skeletal muscle
microvascular blood flow during the compensatory phase to
30% ? 7% of BL values in HTSD and 33% ? 6% in NS
rats. During the decompensatory phase of hemorrhage, mi-
crovascular blood flow decreased significantly further to
17% ? 4% of BL in HTSD and 15% ? 4% of BL in NS
rats. Normal saline resuscitation did not increase skeletal
muscle microvascular blood flow in NS rats. However,
hypertonic saline–dextran resuscitation significantly in-
creased skeletal muscle microvascular blood flow in HTSD
rats but not to BL levels, and the effect was transient.
Hemorrhage constricted large A1 arterioles to 72% ?
5% of BL in HTSD and 67% ? 5% of BL in NS rats during
the compensatory phase and constricted these vessels fur-
ther to 66% ? 4% of BL in HTSD and 62% ? 5% of BL
in NS rats during the decompensatory phase. In contrast,
hemorrhage dilated third-order (A3) arterioles during the
compensatory phase to 129% ? 16% of BL in HTSD and
134% ? 11% of BL in NS rats and dilated A3 arterioles
further during the decompensatory phase to 139% ? 10% of
BL in HTSD and 147 ? 14% of BL in NS rats. In a similar
fashion, hemorrhage dilated fourth-order (A4) arterioles to
200% ? 19% of BL in HTSD and 173% ? 29% of BL in
NS rats during the compensatory phase and dilated those A4
arterioles further during the decompensatory phase to 223%
? 32% of BL in HTSD and 212% ? 17% of BL in NS rats.
All A3 and A4 vessels studied exhibited a normal rhythmic,
spontaneous vasomotion (constriction alternating with dila-
tion at a rate of approximately 12/m) during BL. Vasomo-
tion persisted in A3 and A4 vessels during the compensa-
tory phase of hemorrhage, but it ceased during the
decompensatory phase such that A3 and A4 vessels were
essentially maximally dilated because of a complete loss of
vascular smooth muscle tone.
Resuscitation with both normal saline and hypertonic
saline–dextran initially dilated A1 arterioles toward BL
values (77% ? 4% in HTSD and 75% ? 7% in NS rats) for
10 minutes, but these Al vessels then constricted back to
decompensated shock levels (62% ? 3%) within 20 minutes
in both groups. A3 arterioles, which had dilated maximally
and lost vasomotion during the decompensatory phase of
hemorrhage, responded to normal saline resuscitation with a
short-lived return of vasomotion and constriction back to
120% ? 23% of BL for 10 minutes, but they quickly dilated
back to 142% ? 12% of BL and lost vasomotion again by
20 minutes. Hypertonic saline–dextran resuscitation re-
stored vasomotion and constricted A3 arterioles to 86% ?
5% of BL for 40 to 60 minutes. Similarly, A4 arterioles,
which had maximally dilated and lost vasomotion during
decompensated hemorrhagic shock, regained vasomotion
and constricted toward BL (127% ? 13% in HTSD and
154% ? 20% in NS) 10 minutes after resuscitation. How-
ever, in the normal saline–resuscitated group, vasomotion
was lost, and A4 vessels dilated maximally again (219% ?
14% of BL) by 20 minutes, whereas A4 vasomotion and
constriction (96% ? 10% of BL) was restored to near BL
levels for 40 to 60 minutes with hypertonic saline–dextran
resuscitation. All HTS rats (n ? 3) died 3 to 8 minutes after
7.5% NaCl infusion before data could be obtained. These
data demonstrated that although hypertonic saline–dextran
was better than NS, neither fluid was sufficient to return CO
to normal. However, the fact that HTSD restored tone to A3
and A4 arterioles was encouraging. Therefore, in subse-
quent experiments we included blood and NS in the resus-
citation protocol.
Small intestine
Baseline values for CO, intestinal microvascular blood
flow, and vascular diameters were similar in the 3 groups
(Table 2). During hemorrhage, vasoconstriction of A1 ves-
sels occurred to 71% ? 3%, 83% ? 4%, and 76% ? 8% in
the NS, NSD, and HTSD rats, respectively, at decompen-
sation. A2 arterioles constricted to 69% &plusmn, 5%, 85%
? 3%, and 90% ? 9% in the NS, NSD, and HTSD rats,
respectively, at decompensation. A3 responses were vari-
able, without a significant diameter change, in all 3 groups.
Table 1
Baseline values (mean ? SEM): skeletal muscle
Variable HTSD (n ? 7) NS (n ? 5)
MAP (mm Hg)
CO (mL/min)
Flow (nL/s)
A1 diameter (?m)
A2 diameter (?m)
A3 diameter (?m)
A4 diameter (?m)
112 ? 6
110 ? 11
134 ? 7
133 ? 8
68 ? 9
22 ? 3
8 ? 2
111 ? 4
99 ? 7
114 ? 5
122 ? 12
76 ? 8
20 ? 4
6 ? 1
A1 ? first-order arterioles; A2 ? second-order arterioles; A3 ? third-
order arterioles; A4 ? fourth-order arterioles; HTSD ? 7.5% NaCl in 6%
dextran 70; CO ? cardiac output; MAP ? mean arterial pressure; NS ?
33.3 mL/kg .9% NaCl.
308
H.M. Cryer et al. / The American Journal of Surgery 190 (2005) 305–313

Page 5

Cardiac output decreased to 30% ? 2%, 24% ? 3%, and
30% ? 1% and intestinal microvascular blood flow to 10%
? 2%, 11% ? 2%, and 19% ? 2% in NS, NSD, and HTSD
rats, respectively, at decompensation (Figs. 1 and 2). At the
time of decompensation, the intestinal microcirculation
could be characterized as intense generalized vasoconstric-
tion with virtual cessation of flow. There was evidence of
capillary occlusion in some capillaries and margination of
polymorphonuclear (PMN) cells in both small arterioles and
venules.
Resuscitation resulted in an initial increase in CO,
MAP, and intestinal microvascular blood flow along with
dilation of arterioles and venules back toward BL with all
three fluids. The initial increase in CO was to 105% ?7%
NSD, 92% ? 6% , and 77% ? 5% (Fig. 1). Intestinal
microvascular blood flow initially increased to 48% ?
14%, 42% ? 3%, and 51% ? 11% in the NS, NSD, and
HTSD groups, respectively (Fig. 2). Subsequent addition
of blood and NS to the resuscitation resulted in an in-
crease in CO to greater than BL in all 3 groups (NS 140%
? 10%, NSD 138% ? 12%, and HTSD 124% ? 5% of
BL) (Fig. 1). Intestinal microvascular blood flow in-
creased toward BL after return of shed blood in the NS
(93% ? 33% of BL) and the NSD groups, but it de-
creased to 61% ?9% and 46% ? 7% of BL in the NSD
and HTSD groups, respectively, although CO was greater
than BL (Fig. 2). It is of note that in the NS group,
microvascular flow returned to above normal in 2 but
remained low in 3 animals. This variability precluded a
statistical significance between groups.
Renal microcirculation
Baseline values are listed in Table 3. MAP showed a
progressive decrease that did not vary significantly
among groups during hemorrhage at maximum compen-
sation and at decompensation. MAP returned to BL after
normal saline resuscitation (109% ? 6%) but not after
hypertonic saline dextran (71% ? 14%) or normal saline
dextran (81% ? 9%), and this difference among groups
was statistically significant. After return of shed blood
and NS, MAP returned to nearly BL levels. CO was
consistently depressed across all groups during hemor-
rhage at maximum compensation and at decompensation
(Fig. 3). At decompensation, CO was 30% ? 2%, 31% ?
5%, and 34% ? 3% in the NS, NSD, and HTSD groups,
respectively (Fig. 3). CO transiently increased to 136% ?
Fig. 1. Intestinal experiment cardiac output responses to compensated and
decompensated hemorrhagic shock followed by resuscitation with NS,
NSD, or HTSD. Data are expressed as percent of BL values. There were no
significant differences between groups. BL ? baseline; comp ? compen-
sated; decomp ? decompensated; hem ? hemorrhagic; HTSD ? 7.5%
NaCl/6% dextran 70; NS ? .09% NaCl; NSD ? 0.9% NaCl/6% dextran
70; res 1 ? 10 minutes; res 2 ? 20 minutes.
Fig. 2. A1 intestinal blood flow responses to compensated and decompen-
sated hemorrhagic shock followed by resuscitation with NS, NSD, or
HTSD. Data are expressed as percent of BL values. There were no signif-
icant differences between groups. BL ? baseline; comp ? compensated;
decomp ? decompensated; hem ? hemorrhagic; HTSD ? 7.5% NaCl/6%
dextran 70; NS ? .09% NaCl; NSD ? 0.9% NaCl/6% dextran 70; res 1 ?
10 minutes; res 2 ? 20 minutes.
Table 2
Baseline values: intestine
VariableNS
(n ? 5)
NSD
(n ? 5)
HTSD
(n ? 5)
MAP (mm Hg)
CO (mL/min)
A1 flow (nL/s)
A1 diameter (?m)
A2 diameter (?m)
A3 diameter (?m)
104 ? 6
65 ? 4
10 ? 2
96 ? 5
74 ? 4
27 ? 3
98 ? 7
69 ? 6
11 ? 2
90 ? 3
73 ? 4
26 ? 2
101 ? 2
75 ? 7
15 ? 5
106 ? 11
70 ? 7
30 ? 4
A1 ? first-order arterioles; A2 ? second-order arterioles; A3 ? third-
order arterioles; A4 ? fourth-order arterioles; HTSD ? 7.5% NaCl in 6%
dextran 70; CO ? cardiac output; MAP ? mean arterial pressure; NS ?
33.3 mL/kg .9% NaCl; NSD ? 33.3 mL/kg .9% NaCl in 6% dextran 70.
Table 3
Baseline values: kidney
Variable NS
(n ? 6)
NSD
(n ? 5)
HTSD
(n ? 7)
MAP (mm Hg)
CO (mL/min)
IL flow (nL/s)
IL diameter (?m)
AFF diameter (?m)
EFF diameter (?m)
101 ? 7
128 ? 4
17 ? 11
39 ? 1
11 ? 1
12 ? 1
99 ? 6
96 ? 9
19 ? 8
32 ? 6
12 ? 1
14 ? 2
97 ? 2
97 ? 2
16 ? 5
29 ? 6
10 ? 1
12 ? 1
AFF ? afferent arterial; CO ? cardiac output; EFF ? efferent arterial;
HTSD ? 7.5% NaCl in 6% dextran 70; IL ? intralobular; MAP ? mean
arterial pressure; NS ? 33.3 mL/kg .9% NaCl.
309
H.M. Cryer et al. / The American Journal of Surgery 190 (2005) 305–313