Download Effects of Nitroprusside on Venous Return and Central

Effects of Nitroprusside on Venous Return
and Central Blood Volume in the Absence
and Presence of Acute Heart Failure
HUBERT POULEUR, M.D., JAMES W. COVELL, M.D., AND JOHN Ross, JR., M.D.
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
SUMMARY The effects of nitroprusside (380-760 ug/min) on the systemic venous return, the central blood
volume and the equilibrium point between the venous return and cardiac output curves were studied in eight
dogs using a right-heart bypass preparation at constant total blood volume. Experiments were performed
before and after the production of acute left ventricular failure. During control, nitroprusside infusion shifted
the venous return curve, with a reduction in its plateau (3.1 to 2.4 l/min;p < 0.005), the central blood volume
fell (-2.4 ± 0.4 ml/kg; p < 0.05) and the cardiac output at equilibrium was reduced (3.1 to 2.4 1/min;
p < 0.005). Cardiac failure (produced by acute coronary artery ligation) increased the central blood volume
(+6.3 ± 0.6 ml/kg; p < 0.01), leading to a decrease in both the systemic blood volume and the plateau of the
venous return curve (3.2 to 2.6 1/min; p < 0.003). The cardiac output at equilibrium was also reduced (3.2 to
1.9 1/min; p < 0.001). In this setting, nitroprusside infusion decreased the central blood volume (-7.4 ml/kg;
p < 0.01), but the venous return curve was not altered. The cardiac output curve was markedly shifted upward
due to reduced afterload on the failing left ventricle; however, in the absence of a downward shift of the venous
return curve, the cardiac output at equilibrium increased from 1.9 to 2.6 1/min (p < 0.05). We conclude that
the differing effects of nitroprusside in acute cardiac failure are due primarily to an increased shift of blood
from the central circulation in the failure state compared with the normal circulation. This shift counteracts
the systemic venodilatation produced by nitroprusside and allows the markedly improved left ventricular function consequent to reduced afterload to be expressed as increased venous return and cardiac output.
tion may occur;`4 however, this has not been critically
or quantitatively examined, and in one study no
change in the central blood volume was found during
nitroprusside use in a patient with hypertension.15
Thus, although the effects of vasodilators on cardiac
output and preload (left-heart filling pressures) have
been amply documented,'-6 some investigators
suggested that nitroprusside can cause peripheral
venous pooling and thereby a reduction in the venous
return to the right heart.'6 While a mechanism involving decreased venous return during nitroprusside infusion might be operative in the relatively normal circulation, in which a reduction in cardiac output
usually accompanies nitroprusside use," 17 it is clearly
incompatible with the increased cardiac output
generally observed during such therapy in patients
with congestive heart failure,1 2 because a steady-state
augmentation in the cardiac output must be associated
with a commensurate increase in venous return.'8
Therefore, we examined the responses to
nitroprusside infusion in the presence and absence of
left ventricular failure and tried to determine the
mechanism by which this drug improves the output of
the failing heart, despite an increase in the venous
capacitance and a decrease in cardiac filling
pressures.' We gave particular attention to the
character of the venous return curves," the volume of
blood in the central circulation,'9-21 and the shifts in
this blood volume between the central and systemic
circulations both in the normal and failure states.
VASODILATOR THERAPY, or afterload reduction, has an important role in the treatment of congestive heart failure.1-6 Several clinical and experimental studies have shown that in congestive heart failure
nitroprusside increases the cardiac output while
decreasing the left ventricular end-diastolic pressure
and relieving pulmonary congestion.' 6 There is
evidence that these improvements are not the same
when oral nitrates are used, because the cardiac output may not be significantly increased,7-9 and when
hydralazine alone is used, increased cardiac output
may occur without relief of pulmonary congestion."
Attempts have been made to explain the superiority of
nitroprusside by a balanced action of this agent on the
resistance and capacitance vessels,' but how such an
action might explain opposite responses of the cardiac
output in the normal and heart failure states has not
been investigated.
Increased capacitance of the venous bed due to
vasodilators has been shown in experimental animals,"," and vasodilators have been shown to
decrease the forearm venous tone in patients with congestive heart failure." It has been suggested that a
shift of blood from the central to peripheral circulaFrom the Division of Cardiology, Department of Medicine,
University of California, San Diego, School of Medicine, La Jolla,
California.
Supported by NIH research grants HL-17682 and HL-12373,
NHLBI, PHS/DHEW.
Dr. Pouleur is a Fogarty International Fellow, No. 5 F05 TWO
2549-02.
Address for correspondence: John Ross, Jr., M.D., Director,
Division of Cardiology, Department of Medicine, M-013, University of California, San Diego, La Jolla, California 92093.
Received October 13, 1978; revision accepted August 20, 1979.
Circulation 61, No. 2, 1980.
Methods
Experimental data were obtained from eight
mongrel dogs (mean weight 24 kg, range 19-33 kg)
anesthetized with sodium pentobarbital (25 mg/kg
328
EFFECTS OF NITROPRUSSIDE ON VENOUS RETURN/Pouleur et al.
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
i.v.). Ventilation was maintained by a Harvard
respirator with oxygen-enriched room air (20 ml/kg).
After a median sternotomy, the heart was suspended
in a pericardial cradle. Figure 1 shows a schematic
diagram of the preparation. A 3-mm i.d. steel cannula
was inserted into the left ventricle through a stab incision at the apex and connected to a Statham P23Db
strain gauge manometer. The frequency response of
this cannula coupled to the manometer was flat to 100
Hz and gain was linear to 200 mm Hg. A micromanometer (Millar 5F Model PC340A) was introduced through the right femoral artery and its tip was
located just above the aortic valve for measurement of
central aortic pressure. A silastic rubber catheter (1.1
mm i.d.) was also introduced in the right atrium via
the left mammary vein and connected to a Statham
P23Db transducer calibrated at high gain (20 mm Hg
full scale). Zero reference levels were taken at the level
of the mid-right atrium. Hydrostatic errors in right
atrial pressure measurements were determined at the
end of each experiment by exposing the catheter tip in
situ to atmospheric pressure. Necessary corrections to
the recorded right atrial pressure, if any, were then
made. An electromagnetic flow probe (Biotronex) was
implanted around the aortic root and connected to a
Biotronex BL 610 flowmeter; probe diameter was
selected to give a satisfactory signal during diastole at
low aortic pressures and to avoid significant baseline
shift during transient changes in aortic pressure; the
mean pressure gradient across the probe was less than
5 mm Hg. The frequency response of this system was
flat (-3 dB) from 0-100 Hz, with a transit time of 6
msec and a linear phase shift of 1.9°/Hz.
The dogs were then placed on right-heart bypass. In
brief, a wide-bore Bardic cannula (40F) with a multihole tip was introduced into the right atrium via the
right atrial appendage and its tip advanced into the
right ventricle. A 26F Bardic cannula was introduced
329
into the main pulmonary artery through the right ventricular outflow tract and secured with a tie around the
pulmonary artery. The circuit (volume 350 ml) was
primed with two-thirds dextran (MW 70,000 in 0.9%
saline) and one-third Ringer's lactate. The entire
venous return to the left heart was then assumed by a
roller pump and measured by a Biotronex cannulaflow probe inserted in the external circuit. Dextran
(50-70 ml) was infused to replace blood lost during
surgery, and before collecting data the preparation
was allowed to stabilize for 15 minutes at a right atrial
pressure of 0-2 mm Hg.
All signals together with a standard ECG lead II
were simultaneously recorded on paper and on FM
magnetic tape for later replay at a paper speed of 100
mm/sec.
Control Measurements
Venous Return Curves
In eight dogs, the relation between the venous
return to the right heart and the right atrial pressure
was determined using a procedure modified after a
method described by Guyton et al."1 In brief, starting
from the control point (hemodynamic data obtained
after stabilization at a right atrial pressure of 0-2 mm
Hg and a mean left ventricular end-diastolic pressure
of 9.4 ± 1.5 mm Hg), the speed of the roller pump was
abruptly reduced or increased to a desired level and
maintained there for 8-10 seconds. After the initial
change in right atrial pressure, venous return to the
right atrium responded reaching a new level and
equilibrium between right atrial pressure and the
venous return was rapidly achieved. At this transient
equilibrium point, the output of the venous pump
(measured with a flowmeter) equaled the venous
return. During this brief period, significant vasomotor
reflexes had not yet occurred," as evidenced by a con-
E.M.F
SG
Heat
Exchanger
,'
SVC
,P
FIGURE 1. Schematic diagram of the
modified right-heart bypass preparation.
IVC = inferior vena cava; SVC = superior
vena cava; SG = strain gauge; EMF = electromagnetic flowmeter; PA = pulmonary
artery; L V = left ventricle; R V = right
ventricle.
Millar A
Micromanometer
ClIRCULATION
330
VOL 61, No 2, FEBRUARY 1980
could displace the junction between the linear portion
and the plateau of the venous return curve. This artifact, caused by the transmission of pulsatile waves
from the pump backward to the right atrium, was
minimized by interposing the heat exchanger between
the roller pump and the right atrium.
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
stant heart rate. In addition, pump output changes
during this period had little effect on release or trapping of blood in the lungs; the increases or decreases in
central blood volume during the transient period were
less than 1 ml/kg and of the same magnitude in all the
maneuvers studied. Changes due to alterations in
right-sided pump output should have been similar in
our preparation to those in Guyton's transient
method, which was shown to produce venous return
curves similar to those in the isolated circulation."'
Between each such measurement the pump speed was
returned to the control level, and additional points
were determined only when the right atrial and the left
ventricular end-diastolic pressures had restabilized at
their previous control levels (generally after 1.5-3
minutes). At least five points on the descending portion of the venous return curve and two points high on
the venous return curve at right atrial pressures below
0 mm Hg were determined in each dog. It was found
in obtaining the latter two points on the plateau of the
venous return curve (fig. 2) that it was not possible to
augment the pump output further even at more
negative right atrial pressures, because the venous
system cbllapses; i.e., the cardiac output is now
limited by the amount of venous return." No attempt
was made to directly determine the mean systemic
pressure, which is a measure of the degree of filling of
the systemic circulation at any given time.'2 22
Instead, the linear portion of the venous return-right
atrial pressure relation was extrapolated to zero
venous return to provide a reasonable approximation
of this variable"' 23 (broken lines, fig. 2). No Starling
resistor was inserted in the circuit; Guyton et al.1" have
shown that the absence of a collapsible tube in the circuit does not quantitatively modify the results but
Cardiac Output Curves of the Left Ventricle
After recording the venous return curve, the cardiac
output curve was determined by obtaining measuremdnts of flow and left ventricular end-diastolic
pressure at three to five different levels of venous
return, each point being obtained after 2 minutes in
the steady state.
Nitroprusside Infusion Before Heart Failure
With the preparation stable at the control point,
nitroprusside was infused in six dogs in a dose
(380-760 gg/min) sufficient to produce a sustained
decrease in mean aortic pressure, averaging -30 mm
Hg. When a steady-state level of mean aortic and left
ventricular end-diastolic pressures was obtained, the
venous return curve and the cardiac output curve were
repeated as described above.
During the transient periods of decreases in mean
aortic pressure, the venous return and the aortic blood
flow were continuously recorded, and by subtracting
the aortic flow from the venous return flow, the shift in
the volume of blood contained between the pulmonary
artery and aorta (central blood volume) was computed
as previously described.24 25 Briefly, this approach is
based on the assumption that, during a steady state,
the amount of blood entering the lungs is equal to the
300T0
* CONVTROL
o AIITROPRUSSIDE
.E
co
(n=6)
2000k
FIGURE 2. Curves relating venous return
to the right heart to the mean right atrial
pressure; such curves reflect primarily the
characteristics of the peripheral circulation. 7 Average venous return curves (
SD) obtained under control conditions before
and during nitroprusside infusion in six dogs
are shown. The venous return data are normalized to an average body weight of 20 kg.
CBV= central blood volume.
a:
F-
Lii
co_
a:
C/)
1000
A CBV =
-2.4 ± 0.4 m/Kg-'
,
-2
,
,
0
2
4
jm*
6
RIGHT ATRIAL PRESSURE
8
mmHg
EFFECTS OF NITROPRUSSIDE ON VENOUS RETURN/Pouleur et al.
amount of blood leaving the left ventricle. Small
amounts of flow (bronchial and coronary flows) are
not detected by the flow probes, but the changes in
these flows were probably small, and the effects of
these two flows tend to cancel one amother.26 Because
the validity of the computation is dependent upon the
stability of the gain and of the zero flow levels of the
two flow transducers, the reproducibility of the equilibrium between the two signals was checked at the
different steady states in each dog and found to be
within ± 0.2%; i.e., there was a maximum cumulative
error of 4 ml/min at a cardiac output of 2000 ml/min.
Heart Failure Produced by Coronary Occlusion
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
After recording control data and the responses to
nitroprusside in five dogs, and in two additional dogs
directly after recording the control data, the left
anterior descending artery was abruptly occluded after
premedication with lidocaine (2 mg/kg i.v.). During
the ensuing period it was generally necessary to
decrease the speed of the roller pump to avoid acute
cardiac dilatation and to achieve a new steady state
with left ventricular end-diastolic pressure averaging
approximately 10 mm Hg higher than control. At this
new steady state, the venous return and cardiac output
curves were then determined. Similarly, changes in
central blood volume during the transient period were
computed by continuous subtraction of the two flow
signals.
In these seven dogs with coronary occlusion,
nitroprusside (380-760 ,ug/min) was then infused and
the venous return and cardiac output curves and the
central blood volume were measured in the same way
as during the control nitroprusside infusion. In all
dogs, total peripheral resistance was calculated by the
following formula: (mean aortic pressure - mean
right atrial pressure)/mean aortic flow, and expressed
in dyn-sec-cm-5.
Statistical analysis of the data was performed by
paired t test, and results having ap value less than 0.05
were considered significant.
Results
Relations Between Right Atrial Pressure and Venous Return
Before and During Nitroprusside
Control Conditions (Without Cardiac Failure)
The averaged venous return curves obtained under
control conditions before and during nitroprusside infusion for six dogs, normalized to an average weight of
20 kg, are illustrated in figure 2. Nitroprusside infusion caused a significant downward displacement of
this relation (absolute average values [+ SEM] for the
plateaus from 3100 + 300 to 2400 + 280 ml/min;
p < 0.005), and the extrapolated mean systemic
pressures fell (from 8.3 ± 0.9 to 7.2 ± 0.8 mm Hg;
p < 0.05). There were no significant changes in the
slopes of the linear portions of the curves (489 + 52 to
516 ± 59 ml/mm Hg; NS).
331
Cardiac Failure
After the increase in left ventricular end-diastolic
pressure due to coronary occlusion, and reduction of
pump output, the position of the venous return curve
was also below control, as shown in figure 3 (absolute
values from 3200 + 365 to 2550 ± 321 ml/min;
p < 0.003, n = 7), and the estimated mean systemic
pressure also fell from 9.3 ± 0.8 to 7.6 ± 0.8 mm Hg
(p < 0.005) without change in the slope of these
relations (464 ± 50 and 521 ± 82 ml/mm Hg; NS).
As shown in figure 3, when nitroprusside was infused in these seven dogs after coronary occlusion,
there was no change in either the plateau of the curve
(2550 ± 320 to 2580 ± 270 ml/min; NS) or the mean
systemic pressure (7.6 ± 0.8 to 7.3 + 0.7 mm Hg;
NS); the slopes also were not different (521 ± 82 to
535 ± 76 mm Hg-1; NS).
Also, in the five dogs in which nitroprusside infusion
was performed both before and after coronary occlusion, significant differences were not observed between
the two venous return curves.
Changes in Left Ventricular End-diastolic Pressure
and the Central Blood Volume
Control Conditions
Nitroprusside infusion during control conditions
decreased left ventricular end-diastolic pressure from
9.5 + 1.6 to 4.5 ± 1.5 mm Hg (p < 0.002), and the
central blood volume fell; when the new steady state
was reached, 2.4 ± 0.5 ml/kg of blood had been
shifted from the central to systemic circulations
(p < 0.05, compared with the theoretical maximal
cumulated error during the same period, i.e.,
-1.0 ± 0.03 ml/kg).
Cardiac Failure
After coronary occlusion, left ventricular enddiastolic pressure increased from 9.5 ± 0.9 to
19.7 ± 2.0 mm Hg (p < 0.003), and 6.3 ml/kg
(p < 0.01) of blood shifted from the systemic circulation into the central circulation between the control
and failure states.
Nitroprusside infusion after coronary occlusion
decreased left ventricular end-diastolic pressure to
9.5 i 0.9 mm Hg (p < 0.0003), and the central blood
volume fell by 7.4 i 1.3 ml/kg (p < 0.01).
Left Ventricular Function Curves and Equilibrium Points
Between the Cardiac Output and Venous Return Curves
Figure 4 illustrates a representative experiment
showing the curves relating cardiac output and left
ventricular end-diastolic pressure before and after coronary occlusion, and the effects of nitroprusside.
Coronary occlusion consistently produced a shift of
the function curve downward and to the right (curve A
to C); nitroprusside both before and after coronary
occlusion shifted the curves upward and toward the
left (A to B, C to D).
Figure 5 provides a schematic analysis of the rela-
VOL 61, No 2, FEBRUARY 1980
CI RCULATION
332
B
A
3000 *
*
CONJTROL
A
* COROINARY OCCLUSIOIN
(n=7)
CORONARY OCCLUSION
INITROPRUSSIDE DUR/IVG
CORONlARY OCCLUSION
(n=7)
1=c
.E
2000-
:m
C/)
nr
U)
L.U
2E
1000
CI =
7A+
-7.4 ± 1.3S mt.Kg-'
AACBV =
+6.3+0.6 mlKg-'
\
\
\
l
l
l
l
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
8
6
4
2
0
-2
l
2
10
6
RIGHT ATRIAL PRESSURE mmHg
RIGHT ATRIAL PRESSURE mmHg
FIGURE 3. A) Average venous return curves (fi SD) obtained under control conditions and after coronary
occlusion in seven dogs. B) A verage venous return curves obtained after coronary occlusion before and during nitroprusside infusion in seven dogs. The venous return data are normalized to an average body weight of
20 kg. CBV= central blood volume.
tioni between the venous return and cardiac output
curves. In this figure, the values for the systemic
venous return and for the left ventricular function
curves were obtained from the same dog as illustrated
in figure 4, and left ventricular end-diastolic pressure
was used to represent the mean left atrial pressure.
The right ventricular function curves were adapted
from the data of Guyton et al.'8-20 The pulmonary
venous return was not represented, because in the
steady state it is equal to the systemic venous return.
Under control conditions (fig. 5A) the systemic
venous return was the limiting factor for cardiac out-
A
3000 F
B
D
i=E
.
c
_
FIGURE 5. Schematic analysis of the relation between venous return and right or left
ventricular output curves before and during
nitroprusside infusion under control conditions (A) and after coronary occlusion (B).
Points A, B, C and D indicate the
equilibrium points between cardiac output
and venous return curves.
2000K
CD
W
c0
C-,
C-,
* CONITROL
1000o
0
NVITROPRUSSIDE CONTInuL
* CORONARY OCCLUSION
A
1
l
l
A
A
NITROPRUSSIDE DURINAG
CORONARY OCCUlSIO/I
l
l
16
12
24
20
LEFT VENT. END DIASTOLIC PRESSURE mmHg
4
8
28
EFFECTS OF NITROPRUSSIDE ON VENOUS RETURN/Pouleur et al.
333
RIGHT VENTRICULAR
OUTPUT
3000
cn_
C,,
2000
U)
cr_
C),
I-.C
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
--- NVitroprusside
c=
During Control
.
1000
W
-4
8
4
0
12
FIGURE 4. Relation between cardiac output and left ventricular end-diastolic
pressure obtained in one representative dog
under differing experimental conditions.
Points A, B, C and D indicate the positions
of the plateau of the venous return curve in
control, control + nitroprusside, coronary
occlusion, and coronary occlusion +
nitroprusside, respectively.
16
RIGHT OR LEFT ATRIAL PRESSURE (mmHg)
A
RIGHT VENTRICULAR
OUTPUT
:IC=
c)
2000
Cl
C)
ILJ-
CL
-Coronary Occlusion
1000
C)
cz_
C
-4
B
0
4
8
12
16
20
RIGHT OR LEFT ATRIAL PRESSURE (mmHg)
24
334
CIRCULATION
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
put, the equilibrium point being on the plateau of the
systemic venous return curve (point A). This was also
the case when nitroprusside was infused under control
conditions (point B). Despite a moderate shift upward
of the left ventricular function curve during
nitroprusside, cardiac output fell as left ventricular
filling pressure decreased (fig. 5A, points A, to B1).
The trend illustrated in figure 5 was observed in all six
dogs, and average hemodynamic data at the
equilibrium points in all dogs is summarized in table 1.
After coronary occlusion (fig. 5B), the left ventricle
was unable to maintain an output equal to the plateau
of systemic venous return, and the right atrial pressure
increased to reach an equilibrium between the systemic venous return and the left ventricular output
(point C); in our controlled experimental setting, this
was accomplished by reducing the speed of the roller
pump. Nitroprusside infusion under these conditions
markedly shifted the cardiac output curve upward,
and because the venous return curve was unchanged at
equilibrium, a substantial rise in the cardiac output
occurred (points C to D). A marked shift upward of
the left ventricular function curve occurred during
nitroprusside, and the cardiac output was higher at a
considerably lower filling pressure (fig. SB, C1 to D1).
Average equilibrium points in all seven dogs are
shown in table 1.
The relations between cardiac output and venous
return schematized in figure 5 were observed in all
dogs during control and in five of the seven dogs after
coronary occlusion. In the remaining two dogs (not illustrated), the left ventricular depression was less
pronounced, and equilibrium points before and after
nitroprusside were at the plateau of the venous return
curve. Thus, both during control and in the failing
heart the venous return was the limiting factor of the
cardiac output in all dogs during nitroprusside infusion; during failure, cardiac output at this new
equilibrium point was increased in five dogs and unchanged in two. The average increase in cardiac output in the failing heart for the whole group was +600
ml/min (p < 0.05), and significant decreases in mean
aortic pressure and in left ventricular systolic and enddiastolic pressure also occurred (table 1).
Discussion
The concept that peripheral factors are as important as cardiac factors in the regulation of cardiac output was discussed by Starling in 189527 and recently
has received extensive and quantitative confirmation
from the studies of Guyton and co-workers',18-21
Vasodilators now have an important role in the
therapy of acute and chronic congestive heart failure,
and the effects of these agents on arterial impedance,
cardiac filling pressures, and ventricular performance
have been investigated extensively.' 6, 28 However,
their effects on the systemic venous circulation and the
consequent role of the venous bed in regulating the
cardiac output in the absence and presence of heart
failure have not been quantitatively explored, and the
VOL 61, No 2, FEBRUARY 1980
aim of the present investigation was to apply the
general approach of Guyton to this problem.
The method used herein to measure venous return is
a modification of one previously used by Guyton and
co-workers to determine venous return curves in openchest dogs." In the present study, instead of suddenly
changing the right atrial pressure and measuring the
resulting venous return, venous outflow was abruptly
altered and held constant and the right atrial pressure
was measured as it reached the new equilibrium point.
The striking similarity between the venous return
curves thus obtained and those reported in dogs11 23, 29
make it likely that the two experimental approaches
provide comparable data. In all methods that use transient responses to determine the venous return curve,
some blood is unavoidably trapped within or released
from the pulmonary circulation during the 8-10second study period, which could modify the mean
systemic pressure and the venous return. However,
this volume was shown to be small (< 1 ml/kg) and of
similar magnitude during the various interventions.
Moreover, it has been previously shown that venous
return curves determined either in the steady state or
by the transient method are identical in areflexic
dogs," further supporting the use of the approach. The
fact that relatively small changes in blood volume can
produce marked reductions in venous return raises the
question of artifactual changes due to spontaneous
bleeding and/or to fluid shifts into the extravascular
spaces during the course of the study.'0 Significant
bleeding was prevented by careful homeostasis during
surgery; moreover, changes in total blood volume
sufficient to bias our conclusions did not appear to occur during the time of data collection (120 minutes)
for the following reasons: 1) In the five dogs in which
nitroprusside was infused both in the control period
and after coronary occlusion, the two venous return
curves during nitroprusside infusion were identical; if
significant changes in total blood volume had occurred, the increase in venous capacitance produced
by the same dose of nitroprusside should have resulted
in a lower venous return at the end of the experiment
during the second infusion. 2) In two dogs, the time
between control studies and coronary occlusion was
deliberately reduced by eliminating the control
nitroprusside infusion; the control venous curve and
that during acute heart failure were performed within
less than 30 minutes of each other, and the results
were identical to those observed in the other dogs.
When nitroprusside was infused in the absence of
heart failure, mean systemic pressure fell and the
systemic blood volume increased only slightly.
Nitroprusside markedly increases the capacitance of
the systemic bed, particularly on the venous side,"3
suggesting that the amount of blood transferred from
the central circulation to the systemic bed was smaller
than the increase in systemic venous capacitance. In
this connection, Tarazi et al." reported only a small,
nonsignificant fall in cardiopulmonary blood volume
(measured by indicator-dilution technique) during
nitroprusside infusion in hypertensive subjects without
EFFECTS OF NITROPRUSSIDE ON VENOUS RETURN/Pouleur et al.
5
,4 c;
0
00
N
0
CD
Cl
0)
0
(o
c1-
c
0
0
0
0
V
H
"?
A-
0
0)
CO
0V
cl
0
0
~o
cl
0
C)
CS7
v
ci
a'
1'-
CO
m
z
0)
IC c
m
0
m
CO
0
0
cl
V
V-
0
cO
-H
LC=
z
c'3
P-H
cC
CO
CO
0
r-H
(o
cl
z
m
m
1-0
CI
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
-H
+a
P0
x.
4^
0
0)'
U
0
0
C=
CO
m-H
0
10
o
0
0
Co
0
0
0
c-
0
0
0
C)
CD
co v
V
m
cq
V
kc
10
2)
2)
2)
1.
-H
---0
5.4
10
ko
0 0)e
0
o
U
c;0
;o cz
¢. v
V
CO
't
t-
0
CO
6
co
V
a)
S
0)
Q
0)
0)
0
X
0)a)
C
2)
2)
cr-
cH
Cv
C
Q)
v.
600
C;.
0 C=
01)
o)
0 0)
c;
0)
0)
0
C V
0
-H
0
-
0)
0
0
03
-
CO
-
CO
u:
In
0
C
1o
CS
-H~;
oi
v
tIl-H
0
0
cO
0
LO
6
0)
V
-H
c)
co
0
10
fo
C
6
0
0)
0
0
V
2)
0)1
X
blJ
0)
I.'
CO
CI
0
Cd
"5
2)
V
---
0)
-H
p
0
m m
kl0
cM
l
c
CD
v
V
-
q
2)
C
'C
0
2)
0 0
m
H
335
heart failure. If circulatory reflexes do not compensate
under such circumstances, the cardiac output (after a
transient increase reflecting the small transfer of blood
from the lungs) should decrease below control because
venous return is the factor limiting cardiac output under normal conditions.18 This was found to be the case;
the observed shift downward of the venous return
curve and the fall in cardiac output when vasomotor
tone was reduced by nitroprusside resembled the
response previously reported when venomotor tone
was altered by ganglionic blockade together with a
decreased rate of epinephrine infusion.'1' 12, 30, 33 The
responses we observed in cardiac output and filling
pressures are similar to those observed in patients with
heart disease who do not have elevated left-heart filling pressures.' The response is also similar to that in
some normal patients with nitroprusside administration, although when the reflexes are intact the cardiac
output sometimes increases. Such circulatory effects
of nitroprusside in the normal circulation are
characterized by strong secondary neurogenic and
humoral responses, with persistent tachycardia in
dogs34 and an increase in heart rate and plasma
catecholamine concentrations in patients.35 Thus, in
the normal circulation, marked reflex responses can
come into play which tend to restore the venous
return, cardiac output and the aortic pressure toward
normal.34 Other effects of nitroprusside include direct
vasodilation of the pulmonary vasculature,'2 which
may also play a role in observed hemodynamic
responses.
With the production of acute heart failure, the
systemic blood volume fell because blood was
translocated from the systemic vascular bed into the
pulmonary circulation and left heart, as pointed out in
other analyses of acute left-heart failure.'9' 20 We
reduced pump flow and cardiac output in order to prevent cardiac overdistension, but a fall in the cardiac
output also occurs during acute heart failure in the uncontrolled circulation.20 That a significant amount of
blood is trapped in the pulmonary circulation during
acute heart failure has been previously shown by
Lindsey et al.,2' who demonstrated an increase in central blood volume from 10.3 ml/kg to 17.5 ml/kg
after 3 minutes of major left-heart failure. Evidence of
s.tress relaxation in the pulmonary circulation36
suggests that this volume could increase with time. In
our study, a mean of 6.3 ml/kg of blood was displaced
from the systemic circulation into the central circulation during heart failure. Previous studies by Guyton
et al.12 Harlan et al.22, and Drees and Roth30 have
shown that such a loss of blood from the systemic circulation significantly decreases the mean systemic
pressure and the venous return. In the normal circulation, carotid sinus baroreceptor reflex activity can
significantly reduce the venous capacitance and compensate for a hemorrhage of 7.5-10 ml/kg;3'0 37 ventricular afferent fibers also may be stimulated during
coronary occlusion and cause increased sympathetic
tone.9 38 However, in the open-chest animal under
sodium pentobarbital anesthesia, sympathetic tone is
336
CIRCULATION
already near maximal,39 and it seems likely that the
changes in the position of the venous return curve during acute heart failure were due largely to the shift in
blood volume to the central circulation.
When nitroprusside was infused in the presence of
left ventricular failure during the transient period as
the aortic pressure fell, the left ventricular stroke
volume increased while the ventricular end-diastolic
pressure decreased; when the new steady state was
reached, approximately 7.4 ml/kg of blood had been
transferred from the central circulation to the
systemic circulation. No significant changes in the
venous return curves or the mean systemic pressure
were observed, indicating that the amount of blood
shifted from the central circulation was roughly equal
to the increase in systemic venous capacitance
produced by nitroprusside. This lack of effect on the
venous return curve allowed a steady-state increase in
the cardiac output.
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
Clinical Implications
This experimental preparation represents an artificial, controlled setting for studying the circulatory
effects of acute left ventricular failure and the actions
of nitroprusside, and extrapolations to the clinical setting of heart failure, therefore, should be made with
caution. This applies in particular to the setting of
chronic left ventricular failure, in which the venous
return curve may be displaced upward and to the
right,18' 19 although even under such circumstances the
responses to nitroprusside could be similar to those we
observed. The experimental findings reported might
have particular relevance, however, to the condition of
acute left ventricular failure in man, as may occur
after acute myocardial infarction.
The fact that administration of nitroprusside to
patients with relatively low left-heart filling pressures
after acute myocardial infarction can cause a fall in
the cardiac output has been well documented.1 Based
on our observations, this response can be largely explained by the lack of a sufficient shift in blood volume
from the undistended central circulation and left heart
to the peripheral circulation to compensate for the
venodilator action of nitroprusside. Under such circumstances, other adverse effects, such as reflex
tachycardia, can also occur. In acute heart failure in
man, if the shift in blood volume from the central to
the peripheral circulation during nitroprusside infusion is similar in magnitude to the increase in systemic
capacitance produced by this agent, an explanation is
provided for the opposite effect, a rise in the cardiac
output, which has been observed in such patients.1' 3 17
This lack of shift of the venous return curve was
associated with a marked upward shift in the function
curve of the left ventricle, which led to the increase in
cardiac output.
There may be other causes for the differences
between the normal and failing circulations. The unloading of the failing heart appears to have a more
dramatic effect on its pump function than on the normal ventricle.1' 40 As illustrated in figure 4, the failing
VOL 61, No 2, FEBRUARY 1980
ventricle was working at a high end-diastolic pressure
on a flat ventricular function curve, and the steep
diastolic pressure-volume curve allowed relatively
small reductions in ventricular volume (with correspondingly small effects on the Frank-Starling
mechanism) to be associated with large decreases in
the filling pressure.41 In addition, there is evidence that
an apparent "descending limb" of the function curve
in heart failure can be due to excessive afterload.40 42
In recent studies we have shown that unloading of the
ventricle by nitroprusside can be explained by a
decrease in the afterload, defined as the left ventricular
wall stress during systole, which is determined both by
the aortic input impedance and by geometrical factors.28 In acute heart failure, by markedly reducing the
wall stress, nitroprusside allows the ventricle to improve its stroke volume and to function at a lower enddiastolic volume. Because in acute cardiac failure the
pumping ability of the heart was usually the limiting
factor for cardiac output, in contrast to the normal
situation in which the venous return was limiting, the
lack of a downward shift in the venous return curve
during nitroprusside administration in heart failure
allowed the markedly improved pumping ability of the
left ventricle produced by systolic unloading to be expressed as an increase in the cardiac output (fig. 4).
The fact that the ventricle can achieve a higher cardiac output when nitroprusside is used in heart failure
leads to a smaller decrease in aortic pressure than occurs under normal conditions, despite a substantially
greater decrease in the calculated total peripheral
resistance (table 1). These responses should produce a
reduced change in baroreceptor activity and less
marked reflex responses in patients with heart failure
than under normal conditions. Moreover, the
documented reductions in the gain of high- and lowpressure baroreceptors in heart failure43' 4 may be
contributory. Such conclusions are consistent with the
minimal change in heart rate1' 3 and the decrease in
plasma catecholamine concentrations that have been
observed with nitroprusside therapy in patients with
heart failure.35
References
l.Chatterjee K, Parmley WW, Ganz W, Forrester J, Walkinsky
P, Crexells C, Swan HJC: Hemodynamic and metabolic
responses to vasodilator therapy in acute myocardial infarction.
Circulation 48: 1183, 1973
2. Cohn JN, Mathew JK, Franciosa JA, Snow JA: Chronic
vasodilator therapy in the management of cardiogenic shock
and intractable left ventricular failure. Ann Intern Med 81: 777,
1974
3. Franciosa JA, Guiha NH, Limas CJ, Rodiguera E, Cohn JN:
Improved left ventricular function during nitroprusside infusion
in acute myocardial infarction. Lancet 1: 650, 1972
4. Goodman DJ, Rossen RM, Holloway EL, Alderman EL,
Harrison DC: Effect of nitroprusside on left ventricular
dynamics in mitral regurgitation. Circulation 50: 1025, 1974
5. Miller RR, Vismara LA, Williams DO, Amsterdam EA,
Mason DT: Pharmacological mechanisms for left ventricular
unloading in clinical congestive heart failure: differential effects
of nitroprusside, phentolamine and nitroglycerin on cardiac
function and peripheral circulation. Circ Res 39: 127, 1976
6. Miller RR, Vismara LA, Zelis R, Amsterdam EA, Mason DT:
EFFECTS OF NITROPRUSSIDE ON VENOUS RETURN/Pouleur et al.
7.
8.
9.
10.
11.
12.
13.
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
14.
15.
16.
17.
18.
19.
Clinical use of sodium nitroprusside in chronic ischemic heart
disease. Circulation 51: 328, 1975
Gold HK, Lienbach RC, Sanders CA: Use of sublingual
nitroglycerin in congestive failure following acute myocardial
infarction. Circulation 46: 839, 1972
Gray R, Chatterjee K, Vyden JK, Ganz W, Forrester JS, Swan
HJC: Hemodynamic and metabolic effects of isosorbide
dinitrate in chronic congestive heart failure. Am Heart J 90:
346, 1975
Williams DO, Amsterdam EA, Mason DT: Hemodynamic
effects of nitroglycerin in acute myocardial infarction. Circulation 51: 421, 1975
Chatterjee K, Parmley WW, Massie B, Greenberg B, Werner J,
Klausner S, Norman A: Oral hydralazine therapy for chronic
refractory heart failure. Circulation 54: 879, 1976
Guyton AC, Lindsey AW, Abernathy B, Richardson T: Venous
return at various right atrial pressures and the normal venous
return curve. Am J Physiol 189: 609, 1957
Guyton AC, Polizo D, Armstrong GG: Mean circulatory filling
pressure measured immediately after cessation of heart pumping. Am J Physiol 179: 261, 1954
Mason DT, Braunwald E: The effects of nitroglycerin and amyl
nitrite on arteriolar and venous tone in the human forearm. Circulation 32: 755, 1965
Cohn JN, Franciosa JA: Vasodilator therapy of cardiac failure.
N Engl J Med 297: 27, 1977
Schlant RC, Tsagaris TS, Robertson RJ: Studies on the acute
effects of intravenous sodium nitroprusside. Am J Cardiol 9:
51, 1962
Guiha NH, Cohn JN, Mikulic E, Franciosa JA, Limas CJ:
Treatment of refractory heart failure with infusion of
nitroprusside. N Engl J Med 291: 589, 1974
da Luz PL, Forrester JS, Wyatt HL, Tyberg JV, Chagrasulis
R, Parmley WW, Swan HJC: Hemodynamic and metabolic
effects of sodium nitroprusside on the performance and
metabolism of regional ischemic myocardium. Circulation 52:
400, 1975
Guyton AC, Jones CE, Coleman TG: Graphical analysis of cardiac output regulation: In Circulatory Physiology: Cardiac
Output and its Regulation, edited by Guyton AC, Jones CF,
Coleman TG. Philadelphia, WB Saunders Co, 1973, pp
237-262
Guyton AC, Jones CE, Coleman TG: Cardiac output in heart
failure: I. bilateral failure. In Circulatory Physiology: Cardiac
Output and its Regulation, edited by Guyton AC, Jones CF,
Coleman TG. Philadelphia, WB Saunders Co, 1973, pp
45 1-468
20. Guyton AC, Jones CE, Coleman TG: The pulmonary circulation in left sided heart failure. In Circulatory Physiology: Cardiac Output and its Regulation, edited by Guyton AC, Jones
CF, Coleman TG. Philadelphia, WB Saunders Co, 1973, pp
473-475
21. Lindsey AW, Banahan BF, Cannon RH, Guyton AC:
Pulmonary blood volume of the dog and its changes in acute
heart failure. Am J Physiol 190: 45, 1957
22. Harlan JC, Smith EE, Richard TQ: Pressure-volume curves of
systemic and pulmonary circuit. Am J Physiol 213: 1499, 1967
23. Caldini P, Permutt S, Waddell JA, Riley RL: Effect of
epinephrine on pressure, flow, and volume relationships in the
systemic circulation of dogs. Circ Res 34: 606, 1974
24. Charlier AA, Jaumin P, Pouleur H: Circulatory effects of deep
inspirations, blocked expirations and constant positive pressure
inflations at equal transpulmonary pressure in conscious dogs. J
337
Physiol (Lond) 241: 589, 1974
25. Pouleur H, Jaumin PM, Charlier AA: Pulmonary blood
volume and haemodynamic changes during steady lung inflations in dogs. Acta Anaesth Scand 17: 253, 1973
26. Gregg DE, Fisher LC: Blood supply to the heart. In Handbook
of Physiology. Circulation. Section 2, vol II, edited by
Hamilton WF, Don P. Washington DC, Am Physiological
Society, 1963, pp 1517-1584
27. Starling EG: Some points in the pathology of herat disease.
Lancet 1: 652, 1895
28. Pouleur H, Covell JW, Ross J Jr: Effects of alteration of aortic
input impedance on the force-velocity-length relationship in the
intact canine heart. Circ Res 45: 126, 1979
29. Green JF: Pressure-flow and volume flow relationships of the
systemic circulation of the dog. Am J Physiol 229: 761, 1975
30. Drees JA, Rothe CF: Reflex venoconstriction and capacity
vessel pressure-volume relationships in dogs. Circ Res 34: 360,
1974
31. Rubin S, Misbach G, Lekven J, Parmley WW, Tyberg J:
Volume and resistance changes of the peripheral circulation
during nitroprusside infusion in the areflexic dog: role of the
spleen. (abstr) Clin Res 26: 266, 1978
32. Tarazi RC, Justan HP, Bravo EL, Niarchos AP: Vasodilating
drugs: contrasting haemodynamic effects. Clin Sci Mol Med
51: 575s, 1976
33. Trapold JH: Role of venous return in the cardiovascular
response following injection of ganglion-blocking agents. Circ
Res 5: 444, 1957
34. Pagani M, Vatner SF, Braunwald E: Hemodynamic effects of
intravenous sodium nitroprusside in the conscious dog. Circulation 57: 144, 1978
35. Cohn JN, Taylor N, Vrobel T, Moskowitz R: Contrasting
effects of vasodilators on heart rate and plasma catecholamines
in patients with hypertension and heart failure. (abstr) Clin Res
26: 547, 1978
36. Sarnoff SJ, Berglund E: Pressure-volume characteristics and
stress relaxation in the pulmonary vascular bed of the dog. Am
J Physiol 171: 238, 1952
37. Shoukas AA, Sagawa K: Control of total systemic vascular
capacity by the carotid sinus baroreceptor reflex. Circ Res 33:
22, 1973
38. Zucher IH, Gilmore JP: Atrial receptor discharge during acute
coronary occlusion in the dog. Am J Physiol 227: 360, 1974
39. Guyton AC, Jones CE, Coleman TG: Effect on cardiac output
of opening the chest. In Circulatory Physiology: Cardiac Output and its Regulation. Philadelphia, WB Saunders Co, 1973,
pp 432-433
40. MacGregor DC, Covell JW, Mahler F, Dilley RB, Ross J Jr:
Relations between afterload, stroke volume, and descending
limb of Starling's curve. Am J Physiol 227: 884, 1974
41. Ross J Jr, Braunwald E: Studies on Starling's law of the heart.
IX. The effects of impeding venous return on performance of
the normal and failing left ventricle. Circulation 30: 719, 1964
42. Ross J Jr: Afterload mismatch and preload reserve: a conceptual framework for the analysis of ventricular function. Prog
Cardiovasc Dis 18: 255, 1976
43. Greenberg TT, Richmond WH, Stocking RA, Gupta PD,
Meehan JP, Henry JP: Impaired atrial receptor responses in
dogs with heart failure due to tricuspid insufficiency and
pulmonary artery stenosis. Circ Res 32: 424, 1973
44. Higgins CB, Vatner SF, Eckberg DL, Braunwald E:
Alterations in the baroreceptor reflex in conscious dogs with
heart failure. J Clin Invest 51: 715, 1972
Effects of nitroprusside on venous return and central blood volume in the absence and
presence of acute heart failure.
H Pouleur, J W Covell and J Ross, Jr
Downloaded from http://circ.ahajournals.org/ by guest on April 29, 2017
Circulation. 1980;61:328-337
doi: 10.1161/01.CIR.61.2.328
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1980 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/61/2/328.citation
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/