luis

Society of Critical Care Anesthesiologists
Section Editor: Michael J. Murray
The Systemic and Regional Hemodynamic Effects of
Phenylephrine in Sheep Under Normal Conditions and
During Early Hyperdynamic Sepsis
Hiroshi Morimatsu, MD,* Ken Ishikawa, MD, Clive N. May, PhD, Michael Bailey, PhD,
and Rinaldo Bellomo, MD*
BACKGROUND: Phenylephrine treatment of hypotension in sepsis raises concern because it
may decrease vital organ bloodflow. Accordingly, we investigated the effects of phenylephrine on
systemic and regional bloodflow in normal and septic sheep.
METHODS: Responses to phenylephrine or vehicle infusion for 6 hours were determined in
conscious normal sheep and sheep with early sepsis induced by administration of live
Escherichia coli. Cardiac output and coronary, mesenteric, and renal bloodflow were measured
with implanted flow probes.
RESULTS: In normal sheep, phenylephrine decreased cardiac output and heart rate (HR) but
increased stroke volume and mean arterial blood pressure (MAP) (84
6 to 108
6 mm Hg,
magnitude of mean difference [diff.] 19 [22.6%]; 95% confidence intervals [CI], 17-21). There were
significant decreases in regional conductance values with a transient decrease in mesenteric
bloodflow, no change in coronary bloodflow, and increased renal bloodflow (222
53 to 271
55
mL/min; diff. 31 [13.9%]; 95% CI, 26 -36). During hyperdynamic sepsis, vasodilatation and
increased bloodflow occurred in all vascular beds. Phenylephrine restored MAP and stroke volume to
baseline values, but HR, cardiac output, and total peripheral conductance progressively decreased.
Phenylephrine decreased mesenteric and coronary conductance, with no sustained reduction in
flows, but renal conductance was significantly decreased and overall renal bloodflow increased
(293
22 vs 347
100 mL/min; diff. 55 [18.8%]; 95% CI, 47- 65).
CONCLUSIONS: In sheep with early hyperdynamic sepsis, phenylephrine, at a dose that restored
MAP, increased stroke volume and renal bloodflow while decreasing HR and coronary bloodflow
but not mesenteric bloodflow. Similar responses were seen in normal animals. (Anesth Analg
2012;115:330 -42)
Inpatientswithsepsis,themostcommonhemodynamic -1 receptor stimulation may increase oxygen demand,
profile is a hyperdynamic state characterized by in-
promote arrhythmias, and be associated with a poor prog-
creased cardiac output and peripheral vasodilatation.1
nosis.6-8 Thus, phenylephrine might be a useful drug in
When vigorous fluid replacement fails to restore target
this setting because, in contrast to norepinephrine, phenyl-
arterial blood pressure, vasoactive drugs are frequently
ephrine can increase arterial blood pressure without increas-
used to achieve an adequate perfusion pressure.2,3
ing HR. Despite such considerations, unlike norepinephrine,
Phenylephrine is a selective
-adrenergic receptor ago-
phenylephrine is not recommended as a first-line drug in
1
nist, with negligible or no effect on -1 adrenergic recep-
vasopressor treatment of septic shock or other hypotensive
tors.4,5 In critically ill patients, including those with septic
states because of concerns about the risk of decreased stroke
shock, prolonged increases in heart rate (HR) induced by
volume3 and splanchnic ischemia.9 These concerns were
challenged by a single-center randomized study in which
phenylephrine did not reduce gastric pH in septic patients in
From the *Department of Intensive Care and Department of Medicine,
comparison with norepinephrine.4 Likewise, in an anesthe-
Austin Health, Melbourne, Victoria, Australia; Howard Florey Institute,
University of Melbourne, Parkville, Victoria, Australia; and Australian and
tized porcine model, phenylephrine did not alter bloodflow in
New Zealand Intensive Care Research Centre, School of Public Health and
splanchnic macro- or microcirculations.10
Preventive Medicine, Monash University, Melbourne, Victoria, Australia.
This uncertainty regarding the effect of phenylephrine
Accepted for publication March 7, 2012.
on vital organ bloodflow, in particular its effects in hyper-
Funding: This study was supported by a grant from NHMRC, Australia
dynamic sepsis, indicates the need for further studies.
(454615). Ken Ishikawa was supported by a research fellowship from Iwate
Medical University, and Clive N. May was supported by an NHMRC
Accordingly, we have investigated the effects of phenyl-
Research Fellowship (566819).
ephrine on systemic and regional hemodynamics in normal
The authors declare no conflict of interest.
conscious sheep and in sheep with hyperdynamic sepsis.
Reprints will not be available from the authors.
Address correspondence to Rinaldo Bellomo, MD, Department of Intensive
Care and Department of Medicine, Austin Health, Melbourne, Victoria,
METHODS
Heidelberg, Victoria 3084, Australia. Address e-mail to rinaldo.bellomo@
austin.org.au.
Animal Preparation
Copyright (c) 2012 International Anesthesia Research Society
Experiments in the normal and septic condition were
DOI: 10.1213/ANE.0b013e31825681ab
completed on 6 adult Merino ewes (27 to 35 kg) housed in
330
www.anesthesia-analgesia.org
August 2012 * Volume 115 * Number 2

---

individual metabolic cages, with free access to food and
water. Experimental procedures were approved by the
Preparation for the experiment
Protocol 1(randomized crossover)
control Phenylephrine or Vehicle
Animal Experimental Ethics Committee of the Howard
Two carotid arterial loops
Florey Institute in accordance with the Prevention of Cru-
2 weeks
1st: Phenylephrine or Vehicle
1 weeks
elty to Animals Act 1986, under the guidelines of the
Implantations of flow probes for
2nd: Phenylephrine or Vehicle
National Health and Medical Research Council of Austra-
renal and mesenteric artery
1 weeks
lia's Code of Practice for the Care and Use of Animals for
2 weeks
Experimental Purposes, which conforms with the Guide for
Implantation of flow probes for
Protocol 2(randomized crossover)
pulmonary and coronary artery
control
Phenylephrine or Vehicle
the Care and Use of Laboratory Animals published by the
sepsis
2 weeks
US National Institutes of Health (NIH Publication No.
1st: Phenylephrine or Vehicle
85-23, revised 1996).
2 weeks
Animals underwent 3 separate sterile surgical procedures
2nd: Phenylephrine or Vehicle
at intervals of 2 weeks. For all procedures, anesthesia was
Figure 1. Diagram illustrating the study protocol.
induced with IV sodium thiopental (15 mg/kg), and main-
tained with oxygen/air/isoflurane (1.5 to 2.0%). Fractional
inspired oxygen was altered to maintain Pao2 at approxi-
Septic Sheep
mately 100 mm Hg, and ventilation was controlled to main-
To determine the effects of phenylephrine on vital organ
tain Paco2 at approximately 40 mm Hg. First, 2 carotid arterial
bloodflow in hyperdynamic sepsis, we subjected sheep to
loops were created to facilitate subsequent arterial cannula-
sepsis 1 week after the above experiments in the normal
tion. In the second operation, transit time flow probes (Tran-
condition. After a 2-hour control period, sepsis was in-
sonics Systems, Ithaca, NY) were placed around the left renal
duced by IV bolus injection of Escherichia coli (3
109
artery (4 mm) and the superior mesenteric artery (6 mm).
colony-forming units from an organism obtained from a
Finally, transit time flow probes (Transonics Systems, Ithaca,
patient with E. coli bacteremia and shock). Sheep received
NY) were placed around the pulmonary artery (20 mm) and
IV 0.9% saline (2 to 3 mL/hg/h). The onset of hyperdy-
circumflex coronary artery (3 mm). The animals were allowed
namic sepsis was prospectively defined by the simultane-
at least 2 weeks to recover after the final operation before they
ous presence of the following criteria: (i)
50% increase in
were used for the planned experiments.
HR, (ii)
30% increase in cardiac output, and (iii)
10%
The day before the experiment, a Tygon catheter (Cole-
decrease in MAP.
Parmers, Melbourne, Australia) (internal diameter 1.0 mm,
After the development of hyperdynamic sepsis, sheep
external diameter 1.5 mm) was inserted into the carotid
randomly received either phenylephrine (1 to 2 g/kg/min)
arterial loop to measure arterial blood pressure. A poly-
or vehicle (0.9% saline 12 mL/h) for 6 hours. The dose of
thene catheter (Portex(R), Smiths Medical International Ltd,
phenylephrine was titrated to restore MAP to the control
Hythe, Kens, UK) (internal diameter 1.19 mm, external
level. No fluid boluses, mechanical ventilation, or antibiotics
diameter 1.70) was inserted into a jugular vein for measure-
were given during the experiment. All catheters were re-
ment of central venous pressure and infusion. Analog
moved, and gentamicin (150 mg IM) was administered at
signals (mean arterial blood pressure [MAP], HR, and
the end of the experiment. Sheep were allowed at least 2
cardiac output) and regional blood flows (coronary blood-
weeks for full recovery from sepsis. The animals were
flow, mesenteric bloodflow, and renal bloodflow) were
examined before crossover. If showing full appetite,
collected using a customized data-acquisition system (Lab-
normal behavior, normal vital signs, and normal vital
view, National Instruments, Austin, TX). The data were
organ blood flows and systemic hemodynamics, they
recorded at 100 Hz for 10 seconds every minute during the
were crossed over to the other arm of the study.
experiments. Total peripheral conductance ( cardiac
After all experiments had been completed, animals were
output/MAP) and vascular conductance in other organs
killed according to an ethics committee-approved protocol.
(organ bloodflow/MAP) and stroke volume ( cardiac
Statistical Analysis
output/HR) were calculated.
Responses to phenylephrine or vehicle infusion were com-
pared separately in conscious normal sheep and sheep with
Protocol and Measurements
early sepsis. To ascertain whether phenylephrine was sig-
The study protocol is summarized in Figure 1.
nificantly different from vehicle over the 6-hour period, or
whether phenylephrine behaved differently over time, we
used repeated-measures analysis of variance (RM-ANOVA)
Normal Sheep
models fitting main effects for group (phenylephrine or ve-
Cardiovascular variables were measured during a 2-hour
hicle), time (2 times using 0 hour and 6 hours) and an
control period and a 6-hour infusion of phenylephrine (1
interaction between group and time to determine whether the
g/kg/min) or vehicle (0.9% saline 12 mL/h). The dose of
2 groups behave differently over time. Modeling was per-
phenylephrine (1
g/kg/min) was chosen to match the
formed using the PROC mixed procedure in SAS version 9.2
minimum dose used in patients with septic shock.4 Sheep
(SAS Institute Inc., Cary, NC) with the 6 individual sheep
received IV 0.9% saline (2 to 3 mL/kg/h) to match the
treated as random effects. Twelve outcome variables relat-
protocol used in septic sheep. Treatments were random-
ing to cardiac output and coronary, mesenteric, and renal
ized, and after 1 week, the animals were crossed over to the
bloodflows were considered (MAP, stroke volume, car-
other arm of the study.
diac output, total peripheral conductance, HR,
central
August 2012 * Volume 115 * Number 2
www.anesthesia-analgesia.org
331

---

Effects of Phenylephrine in Sepsis
Table 1A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Mean arterial blood pressure
Control
Phenylephrine
0.0001
19.24
21.24
17.24
Sepsis
Mean arterial blood pressure
Control
Phenylephrine
0.0001
11.84
14.27
9.42
Percentage difference in normal group:
26% (CI,
31% to
21%); percentage difference in septic group:
5% (CI,
11% to 1%). Normal
conscious sheep
without sepsis; sepsis
sheep after infusion of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall
difference; Diff
magnitude of mean difference; lower and upper 95
95% confidence intervals for the difference (calculated as control
phenylephrine; a
positive sign therefore indicates that the mean value for control is greater than phenylephrine, and a negative sign indicates that the mean value for phenylephrine
is greater than control. CI
95% confidence interval.
Table 1B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Mean arterial blood pressure
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in normal sheep?
Mean arterial blood pressure
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
<0.0001
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Mean arterial blood pressure
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in septic sheep?
Mean arterial blood pressure
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.08
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
140
140
120
120
) 100
g
100
P
80
80
60
MA
60
mmH 40
(
40
*
20
+ *+ *+ *+ *+ *+
20
*
+ *
+ *
+ +
I
II
I
II
III
0
0
0
2
4
6
(h)
0
2 4
6
(h)
Figure 2. Left, Mean arterial blood pressure (MAP) effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle (open white circles)
in normal sheep. Data were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious sheep (n
6). Results
are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the control
period (I). For more detail on the comparison, see accompanying tables. Right, Mean arterial blood pressure (MAP) effects of phenylephrine
(1 to 2
g/kg/h, solid black circles) or vehicle (open white circles) in sepsis. Data were collected during a 2-hour control period (I), a 2-hour
sepsis period when the criteria for sepsis were reached at 8 to 10 hours after E. coli (II) and a 6-hour intervention period (III) in conscious sheep
(n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean
of the last 2 hours of the sepsis period (II). For more detail on the comparison, see accompanying tables.
100
100
e
80
m
80
l
u
t)
o
60
e
a
60
e
v
l
/
b
40
k
40
r
o
(m 20
20
St
+ + + +
I
II
I
II
III
0
0
0 2 4
6 (h)
0 2 4
6 (h)
Figure 3. Left, Stroke volume effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle (open white circles) in normal sheep. Data
were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious sheep (n
6). Results are means
SD. P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the control period (I). For more detail
on the comparison, see accompanying tables. Right, Stroke volume effects of phenylephrine (1 to 2
g/kg/h, solid black circles) or vehicle
(open white circles) in sepsis. Data were collected during a 2-hour control period (I), a 2-hour sepsis period when the criteria for sepsis were
reached at 8 to 10 hours after E. coli (II) and a 6-hour intervention period (III) in conscious sheep (n
6). Results are means
SD. *P
0.001
for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the last 2 hours of the sepsis period (II). For
more detail on the comparison, see accompanying tables.
332
www.anesthesia-analgesia.org
ANESTHESIA & ANALGESIA

---

Table 2A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Stroke volume
Control
Phenylephrine
0.0001
2.68
3.78
1.58
Sepsis
Stroke volume
Control
Phenylephrine
0.0001
2.96
4.35
1.56
Percentage difference in normal group:
1% (CI,
9% to 7%); percentage difference in septic group:
21% (CI,
25% to 17%). Normal
conscious sheep without
sepsis; sepsis
sheep after infusion of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall difference;
Diff
magnitude of mean difference; lower and upper 95
95% confidence intervals for the difference (calculated as control
phenylephrine; a positive sign
therefore indicates that the mean value for control is greater than phenylephrine, and a negative sign indicates that the mean value for phenylephrine is greater
than control). CI
95% confidence interval.
Table 2B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Stroke volume
Is there an overall difference between phenylephrine and vehicle over the
Group effect RM-ANOVA
<0.0001
6-hour intervention period in normal sheep?
Stroke volume
Does phenylephrine behave differently from vehicle over time in the 6-
Interaction between group
0.38
hour intervention period in normal sheep?
and time: RM-ANOVA
Stroke volume
Is there an overall difference between phenylephrine and vehicle over the
Group effect RM-ANOVA
<0.0001
6-hour intervention period in septic sheep?
Stroke volume
Does phenylephrine behave differently from vehicle over time in the 6-
Interaction between group
0.005
hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
120
l
120
g)
e
100
era
H
100
h
nc
m 80
p
80
t
a
/m 60
eri
60
in 40
l
p
nduc
40
ta
l/m 20
c
o
*
o
m
+ *+ *+ *+ *+ *+
20
+ + * * *
+ + +
T
(
I
II
I
II
III
0
0
0
2
4
6
(h)
0
2 4
6 (h)
Figure 4. Left, Total peripheral conductance effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle (open white circles) in normal
sheep. Data were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious sheep (n
6). Results are
means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the control period
(I). For more detail on the comparison, see accompanying tables. Right, Total peripheral conductance effects of phenylephrine (1 to 2
g/kg/h,
solid black circles) or vehicle (open white circles) in sepsis. Data were collected during a 2-hour control period (I), a 2-hour sepsis period when
the criteria for sepsis were reached at 8 to 10 hours after E. coli (II) and a 6-hour intervention period (III) in conscious sheep (n
6). Results
are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the last 2 hours
of the sepsis period (II). For more detail on the comparison, see accompanying tables.
venous pressure, coronary bloodflow, coronary vascular
and 4B), and HR (Fig. 6, Table 5A and 5B) with limited
conductance, mesenteric bloodflow, mesenteric vascular
effect on central venous pressure (Fig. 7, Table 6A and 6B).
conductance, renal bloodflow, and renal vascular con-
ductance) under both normal and septic conditions,
Regional Effects
meaning that 24 models were constructed (all presented
Infusion of phenylephrine caused vasoconstriction in the
from Table 1 to Table 9 and graphically from Fig. 2 to Fig.
coronary (Fig. 8, Table 7A and 7B), mesenteric (Fig. 9, Table
10). To lessen the change of a type I error caused by the
8A and 8B), and renal circulations (Fig. 10, Table 9A and
large number of tests, we have chosen a reduced 2-sided
9B). These changes affected regional flows differently with
P value of 0.001 to indicate statistical significance. Re-
no significant alteration in coronary bloodflow (Fig. 8;
sults have been reported as least-square means (95%
Table 7A and 7B), a decrease in mesenteric bloodflow (Fig.
confidence intervals, CIs).
9, Table 8A and 8B), and an increase in renal bloodflow
(Fig. 10, Table 9A and 9B).
RESULTS
Normal Sheep
Septic Sheep
Systemic Effects
Systemic Effects
Infusion of phenylephrine (1
g/kg/min) for 6 hours
After bolus injection of live E. coli, all sheep reached the
increased MAP by 19 mm Hg (22.6%) (95% CI, 17-21) (Fig.
predefined criteria for hyperdynamic sepsis within 8 hours
2, Table 1A and 1B) and stroke volume (Fig. 3, Table 2A and
with increases in cardiac output, HR, and central venous
2B). However, it decreased total peripheral conductance
pressure and decreases in stroke volume and MAP (Figs. 2
(Fig. 4, Table 3A and 3B), cardiac output (Fig. 5, Table 4A
to 7). Hypotension resulted from systemic vasodilatation,
August 2012 * Volume 115 * Number 2
www.anesthesia-analgesia.org
333

---

Effects of Phenylephrine in Sepsis
Table 3A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Total peripheral conductance
Control
Phenylephrine
0.0001
9.85
8.32
11.39
Sepsis
Total peripheral conductance
Control
Phenylephrine
0.0001
12.68
9.81
15.55
Percentage difference in normal group: 34% (CI, 28% to 40%); percentage difference in septic group: 12% (CI, 2% to 22%). Normal
conscious sheep without
sepsis; sepsis
sheep after infusion of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall difference;
Diff
magnitude of mean difference; lower and upper 95
95% confidence intervals for the difference (calculated as control
phenylephrine; a positive sign
therefore indicates that the mean value for control is greater than phenylephrine, and a negative sign indicates that the mean value for phenylephrine is greater
than control). CI
95% confidence interval.
Table 3B. Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance (RM-ANOVA)
Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an Interaction
Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Total peripheral conductance
Is there an overall difference between phenylephrine and vehicle
Group effect RM-ANOVA
<0.0001
over the 6-hour intervention period in normal sheep?
Total peripheral conductance
Does phenylephrine behave differently from vehicle over time in
Interaction between group
<0.0001
the 6-hour intervention period in normal sheep?
and time: RM-ANOVA
Total peripheral conductance
Is there an overall difference between phenylephrine and vehicle
Group effect RM-ANOVA
<0.0001
over the 6-hour intervention period in septic sheep?
Total peripheral conductance
Does phenylephrine behave differently from vehicle over time in
Interaction between group
<0.0001
the 6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
10
10
t
u
8
8
t
p
)
u
in 6
6
c o
/m 4
i
a
L
4
rd
( 2
a
+ + + + + +
2
* *
C
I
II
I
II
III
0
0
0 2
4
6 (h)
0
2 4
6
(h)
Figure 5. Left, Cardiac output effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle (open white circles) in normal sheep. Data
were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious sheep (n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the control period (I). For more detail
on the comparison, see accompanying tables. Right, Cardiac output effects of phenylephrine (1 to 2
g/kg/h, solid black circles) or vehicle
(open white circles) in sepsis. Data were collected during a 2-hour control period (I), a 2-hour sepsis period when the criteria for sepsis were
reached at 8 to 10 hours after E. coli (II) and a 6-hour intervention period (III) in conscious sheep (n
6). Results are means
SD. *P
0.001
for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the last 2 hours of the sepsis period (II). For
more detail on the comparison, see accompanying tables.
as shown by the increase in total peripheral conductance
phenylephrine had a limited effect on renal vascular con-
and in the increased conductance in all regional beds, with
ductance and significantly increased renal bloodflow by 55
increased bloodflow to heart, gut, and kidney.
(18.8%) mL/min (95% CI, 47- 65) (Fig. 10, Table 9A and 9B).
In the vehicle group, all of the above changes caused by
administration of E. coli continued throughout the 6-hour
infusion. In contrast, phenylephrine infused at a dose (1 to 2
DISCUSSION
g/kg/h), titrated to return MAP to control levels (Fig. 2,
Key Findings
Table 1A and 1B), decreased total peripheral conductance
These studies provide key information on the systemic and
(Fig. 4 and Table 3A and 3B) and HR (Fig. 6 and Table 5A and
regional hemodynamic effects of phenylephrine infusion in
5B) and increased stroke volume (Fig. 3 and Table 2A and 2B).
normal and septic conscious sheep. In early hyperdynamic
Phenylephrine also reduced cardiac output (Fig. 5 and Table
sepsis, phenylephrine induced increasing vasoconstriction
4A and 4B) despite increasing central venous pressure (Fig. 7
over time, which returned MAP to control levels without
and Table 6A and 6B).
additional increase in HR and cardiac output, and im-
proved stroke volume, which had been decreased by the
Regional Effects
induction of sepsis. Phenylephrine, however, had differen-
In comparison with vehicle, infusion of phenylephrine
tial effects on vital organ flow, causing mesenteric and
significantly reduced coronary bloodflow and conductance
coronary vasoconstriction, an effect that increased over
(Fig. 8, Table 7A and 7B) and mesenteric conductance (Fig.
time, with limited changes in bloodflow towards normal. In
9, Table 8A and 8B) toward baseline levels. Infusion of
the septic kidney, however, phenylephrine did not cause
334
www.anesthesia-analgesia.org
ANESTHESIA & ANALGESIA

---

Table 4A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Cardiac output
Control
Phenylephrine
0.0001
0.35
0.22
0.47
Sepsis
Cardiac output
Control
Phenylephrine
0.0001
0.27
0.14
0.40
Percentage difference in normal group: 17% (CI, 11% to 23%); percentage difference in septic group: 1% (CI,
11% to 6%). Normal
conscious sheep without
sepsis; sepsis
sheep after infusion of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall difference;
Diff
magnitude of mean difference; lower and upper 95
95% confidence intervals for the difference (calculated as control
phenylephrine; a positive sign
therefore indicates that the mean value for control is greater than phenylephrine, and a negative sign indicates that the mean value for phenylephrine is greater
than control). CI
95% confidence interval.
Table 4B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Cardiac output
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in normal sheep?
Cardiac output
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.07
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Cardiac output
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in septic sheep?
Cardiac output
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
<0.0001
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
200
200
) 150
i
n
150
/
m 100
100
HR
ts
ea 50
(b
50
+ *+ + + + +
+ +
* *
+ + +
I
II
I
II
III
0
0
0
2
4
6
(h)
0
2 4
6
(h)
Figure 6. Left, Heart rate effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle (open white circles) in normal sheep. Data were
collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious sheep (n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the control period (I). For more detail
on the comparison, see accompanying tables. Right, Heart rate effects of phenylephrine (1 to 2
g/kg/h, solid black circles) or vehicle (open
white circles) in sepsis. Data were collected during a 2-hour control period (I), a 2-hour sepsis period when the criteria for sepsis were reached
at 8 to 10 hours after E. coli (II) and a 6-hour intervention period (III) in conscious sheep (n
6). Results are means
SD. *P
0.001 for
comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the last 2 hours of the sepsis period (II). For more
detail on the comparison, see accompanying tables.
vasoconstriction, and bloodflow increased by approxi-
increased stroke volume and decreased HR, which resulted in
mately 20% because of increased perfusion pressure. Simi-
maintained cardiac output.12,13 Treatment with phenyleph-
lar responses were seen in normal animals.
rine had differential vasoconstrictor effects on individual
vascular beds, with the greatest effect on the mesenteric and
Relationship to Previous Studies
the coronary bed and a lesser effect on the renal vascular bed.
The findings in normal animals that phenylephrine caused
The causes of these differences are unclear, but could be
peripheral vasoconstriction, increased MAP, and induced a
related to different vascular densities of -adrenoceptors in
baroreceptor reflex-mediated decrease in HR and cardiac
individual organs, different levels of withdrawal of sympa-
output, with no change in stroke volume, are consistent
thetic vasoconstrictor tone, different levels of compensatory
with previous reports in conscious dogs.11 With higher
autoregulation in response to the changes in perfusion pres-
doses of phenylephrine [bolus, 50 g/kg IV,12 and 10 mg drip
sure,14-16 or all of these. In septic sheep, a higher dose of
infusion within 10 minutes13], the increased afterload led to a
phenylephrine was required to increase MAP. This is in
decrease in stroke volume, HR, and cardiac output.12,13 This is
accord with previous findings in humans showing de-
one of the main reasons why phenylephrine is not recom-
sensitization of the response to 1-adrenoceptor agonists
mended as a first-choice vasopressor drug in sepsis.3 The
in sepsis, postulated to result from decreased expression
relevance of higher-bolus doses to the clinical situation re-
of vascular
-adrenoceptors.17,18 Moreover, a higher
mains dubious. In contrast, increasing the dose of phenyleph-
dose of phenylephrine is sometimes needed to maintain
rine (from 0 to 8 g/kg/min) in patients with septic shock
MAP (3 to 5 g/kg/min) in human sepsis.19
August 2012 * Volume 115 * Number 2
www.anesthesia-analgesia.org
335

---

Effects of Phenylephrine in Sepsis
Table 5A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Heart rate
Control
Phenylephrine
0.0001
8.62
5.85
11.40
Sepsis
Heart rate
Control
Phenylephrine
0.0001
11.28
7.69
14.87
Percentage difference in normal group: 15% (CI, 8% to 22%); percentage difference in septic group: 17% (CI, 11% to 23%). Normal
conscious sheep without
sepsis; sepsis
sheep after infusion of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall difference;
Diff
magnitude of mean difference; lower and upper 95
95% confidence intervals for the difference (calculated as control
phenylephrine; a positive sign
therefore indicates that the mean value for control is greater than phenylephrine, and a negative sign indicates that the mean value for phenylephrine is greater
than control). CI
95% confidence interval.
Table 5B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Heart rate
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in normal sheep?
Heart rate
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.02
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Heart rate
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in septic sheep?
Heart rate
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
<0.0001
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
50
50
us
o
25
n
25
e
re
)
l

v
s
u
0
0
t
ra
(%
pres
-25
-25
cen

I
II
I
II
III
-50
-50
0 2 4
6 (h)
0 2 4
6 (h)
Figure 7. Left, Central venous pressure effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle (open white circles) in normal
sheep. Data were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious sheep (n
6). Results are
means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the control period
(I). For more detail on the comparison, see accompanying tables. Right, Central venous pressure effects of phenylephrine (1 to 2
g/kg/h,
solid black circles) or vehicle (open white circles) in sepsis. Data were collected during a 2-hour control period (I), a 2-hour sepsis period when
the criteria for sepsis were reached at 8 to 10 hours after E. coli (II) and a 6-hour intervention period (III) in conscious sheep (n
6). Results
are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the mean of the last 2 hours
of the sepsis period (II). For more detail on the comparison, see accompanying tables.
The lack of tachycardia with phenylephrine may be a
In a clinical trial, comparing phenylephrine with norepi-
benefit in comparison with catecholamine vasopressors
nephrine, as first-line drugs to treat hyperdynamic septic
(adrenaline, high-dose dopamine, or noradrenaline) used in
shock, treatment with phenylephrine at doses
3 g/kg/min
sepsis that increase HR and may induce arrhythmias.
did not cause deterioration of renal function or urine
As in our previous studies, septic shock was accompanied by
output.
increases in renal vascular conductance and renal blood-
The tendency for coronary vascular conductance and
flow.20,21 Interestingly, phenylephrine had little vasoconstric-
coronary bloodflow to decrease with phenylephrine was
tor effect on the renal vascular bed, and the increase in MAP
probably due to a direct vasoconstrictor effect of phenyl-
caused a further increase in renal bloodflow. This is in accord
ephrine, as well as a response to the decrease in cardiac
with findings in other experimental models of septic shock in
output, which would reduce coronary conductance by
which phenylephrine did not change renal vascular resis-
decreasing cardiac oxygen demand. In hyperdynamic sep-
tance10,22,23 and with a growing understanding of the patho-
tic patients, coronary bloodflow appears high in proportion
genesis of acute kidney injury in sepsis.24,25
to estimated myocardial workload,8,28 so it is unlikely that
Of relevance to our findings, norepinephrine did not
the decrease in coronary bloodflow associated with the
affect renal bloodflow in previous studies with a similar
decrease in cardiac output would compromise cardiac
protocol,26 and epinephrine actually decreased renal
oxygen supply. In sepsis, phenylephrine decreased mesen-
bloodflow despite similar effects on systemic hemody-
teric vascular conductance, returning its levels to values
namics.27 Therefore, there appears to be a different
slightly above baseline, which resulted in a corresponding,
response of the renal vascular bed to each vasoactive
small and short-lived, decrease in mesenteric bloodflow.
drug, which varies according to the degree of
and
This is consistent with previous reports in patients with
receptor stimulation.
hyperdynamic septic shock, which showed a decrease in
336
www.anesthesia-analgesia.org
ANESTHESIA & ANALGESIA

---

Table 6A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Central venous pressure
Control
Phenylephrine
0.0180
3.99
0.63
7.35
Sepsis
Central venous pressure
Control
Phenylephrine
0.0001
8.42
10.97
5.87
Percentage difference in normal group:
170% (CI,
252% to
88%); percentage difference in septic group:
177% (CI,
406% to 52%). Normal
conscious
sheep without sepsis; sepsis
sheep after infusion of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of
overall difference; Diff
magnitude of mean difference; lower and upper 95
95% confidence intervals for the difference (calculated as control
phenylephrine;
a positive sign therefore indicates that the mean value for control is greater than phenylephrine, and a negative sign indicates that the mean value for
phenylephrine is greater than control). CI
95% confidence interval.
Table 6B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Central venous pressure
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
0.018
the 6-hour intervention period in normal sheep?
Central venous pressure
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.67
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Central venous pressure
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in septic sheep?
Central venous pressure
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.002
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
100
l
ow
100
f
80
)
80
l
ood
in 60
60
b
l/m 40
r
y
40
a
m
n
( 20
o
20
I
II
I
II
III
0
0
Cor
r
1.0
1.0
l
a
)
e
g 0.8
0.8
s
c
u
nc
0.6
va
t
a
0.6
/
mmH
r
y
uc
n 0.4
a
0.4
n
nd
/
mi 0.2
r
o
c
o
0.2
*
* *
(
ml
+
I
II
I
II
III
0.0
Co
0.0
0
2
4
6 (h)
0
2 4
6 (h)
Figure 8. Left, Coronary bloodflow and coronary vascular conductance effects of phenylephrine (1
g/kg/min, solid black circles) or vehicle
(open white circles) in normal sheep. Data were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in conscious
sheep (n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for each time point.
P
0.001 for comparisons with the
mean of the control period (I). For more detail on the comparison, see accompanying tables. Right, Coronary bloodflow and coronary vascular
conductance effects of phenylephrine (1 to 2
g/kg/h, solid black circles) or vehicle (open white circles) in sepsis. Data were collected during
a 2-hour control period (I), a 2-hour sepsis period when the criteria for sepsis were reached at 8 to 10 hours after E. coli (II) and a 6-hour
intervention period (III) in conscious sheep (n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for each time.
P
0.001 for comparisons with the mean of the last 2 hours of the sepsis period (II). For more detail on the comparison, see accompanying tables.
hepatosplanchnic perfusion.19,29 The clinical consequences
levels, suggesting a degree of safety with regard to the
of returning markedly increased mesenteric bloodflows in
splanchnic circulation.
sepsis to slightly increased values remains unknown. More-
International guidelines for management of sepsis have
over, in sepsis, microcirculatory shunting is postulated to
not recommended phenylephrine as a first-line drug in
occur.30 In hyperdynamic septic shock, phenylephrine may
septic shock, because it may decrease stroke volume.3
theoretically worsen such shunting. However, in clinical stud-
However, in a number of clinical studies, phenylephrine
ies, phenylephrine has not been shown to increase lactate
did not reduce stroke volume in patients with sepsis or
August 2012 * Volume 115 * Number 2
www.anesthesia-analgesia.org
337

---

Effects of Phenylephrine in Sepsis
Table 7A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Coronary blood flow
Control
Phenylephrine
0.68
0.29
1.10
1.68
Normal
Coronary vascular conductance
Control
Phenylephrine
0.0001
0.05
0.03
0.06
Sepsis
Coronary blood flow
Control
Phenylephrine
0.0001
6.05
2.99
9.11
Sepsis
Coronary vascular conductance
Control
Phenylephrine
0.0001
0.14
0.11
0.18
Bloodflow: percentage difference in normal group: 7% (CI,
5% to 19%); percentage difference in septic group: 0% (CI,
11% to 11%). Conductance: percentage difference in
normal group: 24% (CI, 13% to 35%); percentage difference in septic group: 12% (CI,
2% to 26%). Normal
conscious sheep without sepsis; Sepsis
sheep after infusion
of E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall difference; Diff
magnitude of mean difference; lower and upper
95
95% confidence intervals for the difference (calculated as control
phenylephrine; a positive sign therefore indicates that the mean value for control is greater than
phenylephrine, and a negative sign indicates that the mean value for phenylephrine is greater than control). CI
95% confidence interval.
Table 7B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Coronary bloodflow
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
0.68
the 6-hour intervention period in normal sheep?
Coronary bloodflow
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.36
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Coronary bloodflow
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
0.0001
the 6-hour intervention period in septic sheep?
Coronary bloodflow
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.05
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Coronary vascular
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
conductance
the 6-hour intervention period in normal sheep?
Coronary vascular
Does phenylephrine behave differently from Vehicle over time in the
Interaction between group
0.01
conductance
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Coronary vascular
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
conductance
the 6-hour intervention period in septic sheep?
Coronary vascular
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.0001
conductance
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
w
o
1000
1000
fl
d
800
o
)
800
l
o
in 600
600
c b
l/m 400
400
m
t
eri
( 200
200
e
n
* *
I
II
I
II
III
0
0
Mes
r
12
l
a
)
12
e
g 10
10
s
cu
a
nc
8
8
t
a
c v
/
mmH 6
n
6
4
t
eri
nduc
4
/
mi
e
n
2
c
o
*
+ *
+ *+ *
+ *+
2
* *
es
(
ml
+
I
II
I
II
III
0
M
0
0
2
4
6
(h)
0
2 4
6 (h)
Figure 9. Left, Mesenteric bloodflow and mesenteric vascular conductance effects of phenylephrine (1
g/kg/min, solid black circles) or
vehicle (open white circles) in normal sheep. Data were collected during a 2-hour control period (I) and a 6-hour intervention period (II) in
conscious sheep (n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for each time point. P
0.001 for comparisons
with the mean of the control period (I). For more detail on the comparison, see accompanying tables. Right, Mesenteric bloodflow and
mesenteric vascular conductance effects of phenylephrine (1 to 2
g/kg/h, solid black circles) or vehicle (open white circles) in sepsis. Data
were collected during a 2-hour control period (I), a 2-hour sepsis period when the criteria for sepsis were reached at 8 to 10 hours after E. coli
(II) and a 6-hour intervention period (III) in conscious sheep (n
6). Results are means
SD. *P
0.001 for comparisons with vehicle for
each time.
P
0.001 for comparisons with the mean of the last 2 hours of the sepsis period (II). For more detail on the comparison, see
accompanying tables.
338
www.anesthesia-analgesia.org
ANESTHESIA & ANALGESIA

---

Table 8A. Differences in Key Hemodynamic Variables When Comparing Control to Phenylephrine in
Normal and Septic Sheep
Group
Variable
Type
_Type
Probt
Diff
Lower 95
Upper 95
Normal
Mesenteric blood flow
Control
Phenylephrine
0.0001
33.04
21.57
44.51
Normal
Mesenteric vascular
Control
Phenylephrine
0.0001
1.21
1.06
1.37
conductance
Sepsis
Mesenteric blood flow
Control
Phenylephrine
0.0082
29.89
7.39
52.40
Sepsis
Mesenteric vascular
Control
Phenylephrine
0.0001
1.41
1.00
1.83
conductance
Bloodflow: percentage difference in normal group: 1% (CI,
6% to 8%); percentage difference in septic group:
6% (CI,
15% to 3%). Conductance: percentage difference in
normal group: 20% (CI, 12% to 28%); percentage difference in septic group: 5% (CI,
7% to 17%). Normal
conscious sheep without sepsis; sepsis
sheep after infusion of
E. coli; phenylephrine
phenylephrine infusion; control
infusion of vehicle; Probt
significance of overall difference; Diff
magnitude of mean difference; lower and upper
95
95% confidence intervals for the difference (calculated as control
phenylephrine; a positive sign therefore indicates that the mean value for control is greater than
phenylephrine, and a negative sign indicates that the mean value for phenylephrine is greater than control). CI
95% confidence interval.
Table 8B. Differences in Key Hemodynamic Variables Using a Repeated-Measures Analysis of Variance
(RM-ANOVA) Model Fitting Main Effects for Group (Phenylephrine or Vehicle), Time (Hours 0 - 6), and an
Interaction Between Group and Time and Testing Key Hypotheses
Variable
Question
Test
P value
Mesenteric bloodflow
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
the 6-hour intervention period in normal sheep?
Mesenteric bloodflow
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
0.001
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Mesenteric bloodflow
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
0.0082
the 6-hour intervention period in septic sheep?
Mesenteric bloodflow
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
<0.0001
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Mesenteric vascular
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
conductance
the 6-hour intervention period in normal sheep?
Mesenteric vascular
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
<0.0001
conductance
6-hour intervention period in normal sheep?
and time: RM-ANOVA
Mesenteric vascular
Is there an overall difference between phenylephrine and vehicle over
Group effect RM-ANOVA
<0.0001
conductance
the 6-hour intervention period in septic sheep?
Mesenteric vascular
Does phenylephrine behave differently from vehicle over time in the
Interaction between group
<0.0001
conductance
6-hour intervention period in septic sheep?
and time: RM-ANOVA
Boldfaced values indicate significance. Because of multiple comparisons, only a P
0.001 was considered statistically significant.
septic shock,4,19,31-33 and we found that phenylephrine
of experimental animals, thus increasing the possibility of
increased stroke volume to baseline presepsis levels. This
type II error. We selected a crossover randomized con-
likely resulted from the venoconstrictor action of phenyl-
trolled trial to minimize such error. In addition, to our
ephrine, indicated by the tendency for central venous
knowledge, this is the first study of the effects of phenyl-
pressure to increase. This effect has been seen with
ephrine on regional bloodflow with evaluation by continu-
phenylephrine in human sepsis.31 This effect of phenyl-
ous direct measurements in the setting of hyperdynamic
ephrine may be particularly important in the splanchnic
sepsis. We did not measure the effect of phenylephrine on
capacitance vessels, where pooling of blood has been
tissue oxygen extraction because of the difficulty of main-
noted in sepsis.31,34,35 In septic sheep, treatment with
taining renal venous cannulae patent for the 5 to 6 weeks
phenylephrine decreased cardiac output, although by the
required. Similarly, we did not measure the impact of
end of the 6-hour infusion, cardiac output remained
phenylephrine on oxygen consumption. Previous work in
significantly above control levels. This decrease in car-
humans suggested that phenylephrine treatment of sepsis
diac output occurred despite the increased stroke vol-
may induce significant and unpredictable increases in
ume and reflected the greater decrease in HR. This is in
oxygen consumption
contrast to patients with septic shock in whom adminis-
. Thus, measurement of regional and
systemic oxygen consumption may be an important vari-
tration of phenylephrine for 8 and 12 hours did not
able to measure in future studies.32 We did not measure the
decrease cardiac output despite a decrease in HR and a
effects of phenylephrine on the renal microvasculature.
higher dose of phenylephrine ( 3 to 5 g/kg/min) being
Accordingly, we do not know whether medullary flow was
administered than in the present study.4,19
maintained or whether there was shunting of bloodflow or
Strengths and Limitations
the cortex or whether, within the cortex, there was further
The present study has both strengths and limitations. As
shunting. We also did not conduct metabolic studies and
with any experimental animal model, its relevance to
are unable to comment on how these changes in renal
human disease remains unclear, but the hyperdynamic
bloodflow affected glomerular filtration rate.
state that developed is similar to that in human sepsis,
We focused on the effects of phenylephrine on hemody-
suggesting clinical relevance. In addition, the present in-
namics rather than organ function, but the information on
vestigation had the limitation of studying a small number
stroke volume provides a clinically relevant measure of its
August 2012 * Volume 115 * Number 2
www.anesthesia-analgesia.org
339

---