PATHOPHYSIOLOGY AND NATURAL HISTORY
CONGESTIVE HEART FAILURE
Relationship between central hemodynamics and
regional blood flow in normal subjects and in
patients with congestive heart failure
MARK E. LEITHE, M.D., RAYMOND D. MARGORIEN, M.D., JAMES B. HERMILLER, M.D.,
DONALD V. UNVERFERTH, M.D., AND CARL V. LEIER, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
ABSTRACT Central and regional (hepatic, renal, and limb) hemodynamic data are presented for a
normal population (n = 16) and for a group of patients with congestive heart failure (n = 64). The
patient population represented a wide spectrum of severity of congestive heart failure. Various relationships between central and regional hemodynamics were analyzed. The results indicate that in congestive heart failure blood flow to hepatic, renal, and limb regions is significantly decreased, and that this
decrease is proportional and linearly related to the reduction in cardiac output. The vascular resistances
of these regions correlated directly with systemic vascular resistance. Changes in renal vascular
resistance and renal blood flow became attenuated as the severity of the heart failure advanced from
moderate to severe and at higher levels of systemic vascular resistance. There was little to no correlation between systemic blood pressure and liver, kidney, and limb blood flow for the range of systemic
pressures studied.
Circulation 69, No. 1, 57-64, 1984.
HEART FAILURE is accompanied by a variety of
compensatory mechanisms (e.g., sympathetic nervous
system activation and mechanisms of the renin-angiotensin-aldosterone and vasopressin systems) affecting
regional, as well as central, hemodynamics. In general, studies in animals and humans suggest that, in
congestive heart failure, there is a redistribution of
regional blood flow away from vascular beds sensitive
to sympathetic nervous system stimulation (e.g.,
splanchnic, renal), with a relative increase in the proportion of cardiac output to the more autoregulated
vascular beds such as the coronary and cerebral circulations. 1-4
Regional blood flow measurements in previous human studies have generally been restricted to one organ
system. Current methods of bedside cardiac catheterization and techniques that determine blood flow to
multiple vascular beds have provided the opportunity
to study, in a large population of patients with congestive heart failure, the central and regional hemodynamics in each subject. This investigation was performed
to examine the relationships between central hemoFrom the Division of Cardiology, Ohio State University College of
Medicine, Columbus.
Address for correspondence: Carl V. Leier, M.D., 621 Means Hall,
1655 Upham Dr., Columbus, OH 43210.
Received Feb. 24, 1983; revision accepted Sept. 22, 1983
Vol. 69, No. 1, January 1984
dynamics and regional blood flow in patients with congestive heart failure and in an age- and sex-matched
normal control population.
Methods
Patient population. The congestive heart failure population
consisted of 64 patients (54 men and 10 women) with a mean
age of 54 + 12 years (range 19 to 69). A wide spectrum of mildto-severe congestive heart failure as defined by clinical and
hemodynamic data was represented by this population. Twentyfour patients were classified as functional class IV, 28 as functional class III, and 12 as functional class II (New York Heart
Association). Fifty-six patients underwent diagnostic cardiac
catheterization within 1 year of being studied. After clinical and
laboratory evaluation, 40 patients had congestive heart failure
secondary to idiopathic dilated cardiomyopathy, 13 had ischemic cardiomyopathy, five had congestive heart failure after
valve replacement for longstanding rheumatic valvular disease,
three were judged to have alcoholic cardiomyopathy, and the
remaining three patients had hypertensive, peripartal, or doxorubicin-induced cardiomyopathy.
Forty-nine patients were receiving an oral digitalis preparation daily (0.25 mg digoxin in 38 patients and 0.125 mg in eight
patients; 0.05 to 0.1 mg digitoxin in three patients). Forty-six
patients were receiving 20 to 300 mg oral furosemide daily,
eight patients 600 to 1600 mg quinidine sulfate daily, five patients 1500 to 4500 mg oral procainamide daily, two patients
300 to 340 mg oral disopyramide daily, nine patients 20 to 80
mg oral isosorbide dinitrate daily, and four patients were receiving 15 to 30 mg topical nitroglycerin ointment daily. Digitalis
and antiarrhythmics were continued throughout the study while
furosemide and nitrates were discontinued at least 24 hr before
the study.
57
LEITIIE et al.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Normal subjects. The normal group consisted of 16 subjects
(12 men and four women) with a mean age of 51 + 16 years
(range 24 to 79). These individuals had an unremarkable history
and physical examination, a normal chest roentgenogram, electrocardiogram, and echocardiogram, and normal systolic time
intervals.
Written informed consent was obtained from each subject
before the study.
Procedures and measurements. All studies were performed
while patients were in a 20 to 30 degree recumbant position and
postabsorptive state and between 8:00 A.M. and 12 noon.
Central hemodynamic data for the patients with congestive
heart failure were obtained with a flow-directed triple-lumencatheter; this catheter was placed in the pulmonary artery 1 day
before the study to achieve adequate equilibration. Cardiac output, obtained in triplicate and averaged for each patient, was
measured by the thermodilution technique5 with an Instrumentation Laboratory 601 computer and 602 recorder.
In the normal subjects, cardiac output was measured in duplicate by the precordial radioisotope-dilution curve technique.68
An indwelling catheter was placed in the antecubital vein of
both arms. After equilibration, 1 ml of 311-labeled human serum albumin (12,uCi), followed immediately by a 10 ml bolus
of normal saline, was injected into each subject's right arm. The
resulting dilution curve was detected and recorded by a Nuclear
Chicago Model 8731 precordial scintillation detector, count rate
meter, and a Model 8416 recorder. The probe was positioned
perpendicular to the chest wall, over the fourth intercostal space
at the left sternal border. This position was marked on each
subject to ensure that a constant volume of blood would be
viewed by the detector. The concentration of isotope was demonstrated to be at equilibrium 10 min after injection, and at this
time a blood sample was withdrawn for use in the blood volume
calculation. A Picker Nuclear Model 625105 well counter was
used to analyze radioactivity of the injectate and to compute
blood volume. The cardiac output was calculated as:
Cardiac output =
_Ce_x_B x
60
A
where Ce = radioactivity in counts per minute at equilibrium;
BV = blood volume in liters; A = the area under the primary
curve, the ordinate of which represents counts per minute and
abscissa of which represents time in seconds. In our laboratory,
the correlation coefficient between the radioisotope-dilution
curve method and the thermodilution technique for measuring
1
cardiac output is 0.86 (r2 = .74, p < .05 when n l0).
Cardiac index (1/min/m2) was calculated as cardiac output/
body surface area and total systemic resistance (dynes-seccm-5) as mean systemic arterial pressure x 80/cardiac output.
Systemic blood pressure was measured by cuff and mercury
column sphygmomanometer. The disappearance of the Korotkoff sounds was taken as the diastolic blood pressure value. An
indwelling upper limb arterial cannula, interphased with Electronics for Medicine M2101 amplification systems was used in
six patients in whom Korotkoff sounds were not clearly audible.
Mean systemic blood pressure was determined electronically in
these six patients; in the remaining patients it was calculated as:
(systolic blood pressure -diastolic blood pressure)/3 plus diastolic blood pressure. Systemic blood pressure measurements
were made twice (and averaged) for each set of central hemodynamic values.
Hepatic plasma flow was determined by indocyanine green
(ICG) clearance.9 After blood for a plasma blank and the ICG
standard was obtained, ICG was injected intravenously at 0.5
mg/kg body weight. Venous blood samples were drawn from
the pulmonary artery at 5, 10, 15, and 20 min after injection in
58
the congestive heart failure patients and, because of a higher
clearance rate, at 3, 6, 9, and 12 min in the control subjects. The
plasma samples were read spectrophotometrically (Beckmanof
DB Model 1401) against the plasma blank at a wave length
805 mg to determine percent absorbance, which was then converted to plasma concentration by the use of the ICG standard.
p-Aminohippurate (PAH) was used to measure effective renal
plasma flow.'2 After control blood samples were obtained 40
mg/kg PAH was injected intravenously over 5 min and venous
blood was sampled at 15, 30, 45, and 60 min after injection in
the patients with heart failure. In the normal subjects 25 mg/kg
PAH was injected and venous blood was sampled at 7, 14, 21,
and 28 min after injection. The dose and sampling were modified in the normal group because of a more rapid excretory rate,
a lesser volume of distribution, and a more marked vasodilatory
effect of PAH in normal subjects. The plasma concentrations
were determined spectrophotometrically by the paradimethyl
aminobenzaldehyde method at a wave length of 465 mg. Calculation of effective renal and hepatic plasma flow was essentially
the same. For each determination, the four concentrations were
plotted on semilog paper and a line was drawn through theof
points by the best-fit method. Plasma flow (plasma clearance
dye) was then obtained by multiplying volume of distribution by
the elimination constant; the volume of distribution equals the
amount of dye injected divided by the concentration at time zero
and the elimination constant equals the natural log of 2 divided
by the plasma half-life. Blood flow was then calculated by
dividing plasma flow by one minus the hematocrit. Hepatic and
renal flows were corrected for body surface area. Limb blood
flow (right upper extremity) was determined by venous occlusive plethysmography and expressed in milliliters of blood!
deciliter limb tissue/minute. Room temperature was held constant at between 210 and 23.50 C. Plethysmography was performed with a Meda Sonics SPG-16 Amplification System interphased between a SG-21 gallium-indium strain gauge and a
RI 2A strip-chart recorder. A minimum of five determinations
were averaged for each limb blood flow value. Regional vascular resistance was calculated by dividing the mean arterial pressure (mm Hg) by the regional blood flow value (in 1/min for
renal and hepatic. ml/100 ml/min for limb).
The sequence of testing was as follows: central hemodynamic
measurements (10 to 15 min), hepatic and limb blood flow (15
to 25 min), renal blood flow (30 to 60 min), and repeat central
hemodynamic measurements (10 to 15 min). The two sets of
central hemodynamic data were averaged.
Statistical analysis. A Hewlett-Packard 85 computer and
curve-fitting program were used to determine the relationship of
cardiac index, systemic vascular resistance, and blood pressure
to regional blood flow and regional vascular resistance. Interand subgroup comparisons were made with one- and twogroup
way analysis of variance.
Results
The results of this study are presented in figures 1
through 4. Figure 1 depicts mean ( ± SD) central and
regional hemodynamic values for the normal subjects
and for the patients with congestive heart failure. The
congestive heart failure group was divided into subgroups of those with mild, moderate, and severe disease on the basis of respective cardiac indexes of >
2.5, 2.5 to 2.0, and < 2.0 1/min/m2.
The mean cardiac index in the normal poplation was
3.50 + 0.48 1/min/M2, while that of the entire population with congestive heart failure was significantly reCIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CONGESTIVE HEART FAILURE
A
B
SYSTEMIC VASCULAR
T
RESISTANCE
4.00 1
c
3.00
-i
2.00
c1
E
.00
C
600
N
at
400
e
2
200
E:
700
N
r.
.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
500
11
c:
&t
G
300
0
14
H
._
E
LIMB VASCULAR RESISTANCE
g
-
to
30
u
6
20
a:
0
NL
n = 16
Mild Moderate Severe
26
18
20
CHF Subgroups
X+± SD, *p<0.05vs
nl
+p<.OO5
10
NL
n =
16
Mild Moderate
18
20
Severe
26
CHF Subgroups
vsAdjacent Less
Severe Subgroups
FIGURE 1. Mean cardiac index (A), systemic vascular resistance (B),
regional blood flow (C, E, and G), and regional vascular resistance (D,
F, and H) measurements in normal subjects (NL) and subgroups of
patients with congestive heart failure (CHF).
duced to 2.19 0.47 1/min/m2. The systemic vascular
resistance for each congestive heart failure subgroup
was significantly elevated above control and above
that of each preceding subgroup (figure 1, B).
Figures 2 through 4 illustrate the regional blood flow
and vascular resistance data points plotted against central hemodynamic parameters for each individual in
this study. Best-fit regression formulas are included in
each figure. Table 1 lists the best- and second best-fit
formulas for each plot.
Hepatic hemodynamics. The mean hepatic blood flow
in the normal subjects was 595 + 116 ml/min/m2. The
mean hepatic blood flow for the entire population with
congestive heart failure was reduced to 340 + 153 ml/
min/m2 (p < .05). Hepatic blood flow was significantly diminished from control for each of the three subgroups (figure 1, C). Mean hepatic vascular resistance
for each congestive heart failure subgroup was significantly elevated above control (figure 1, D). Although
mean hepatic vascular resistance tended to increase
Vol. 69, No. 1, January 1984
with increasing severity of congestive heart disease, no
significant differences were noted between subgroups.
A good linear correlation was found between individual cardiac index and hepatic blood flow measurements (figure 2, top left). The relationship of hepatic
blood flow to systemic vascular resistance was best
described by a logarithmic curvilinear function, with a
fairly good negative correlation (figure 2, bottom left).
Hepatic blood flow correlated poorly with systemic
blood pressure (figure 2, top right), but a good positive
correlation was noted between hepatic vascular resistance and systemic vascular resistance (figure 2, bottom right).
Renal hemodynamics. Mean renal blood flow in the
normal subjects was 607 ± 172 ml/min/m2, while in
the population with congestive heart failure it was reduced to 395 + 131 ml/min/m2 (p < .05). Renal blood
flow was significantly reduced from control for patients in each congestive heart failure subgroup (figure
1, E). Renal flow was significantly less in the moderate compared with the mild heart failure group, but
little difference was noted between the moderate and
severe disease subgroups. Renal vascular resistance
was increased significantly from control in the moderate and severe heart failure subgroups (figure 1, F).
The resistance was increased significantly in the moderate compared with the mild failure subgroup and
decreased modestly in the severe compared with the
moderate congestive heart failure subgroup. For the
combined normal and congestive heart failure populations, a positive linear relationship with fair correlation
existed between cardiac index and renal blood flow
(figure 3, top left). A fair correlation was noted for
renal blood flow values and systemic vascular resistance values for the combined normal and congestive
heart failure groups; the relationship between these
parameters was best described by a logarithmic function (figure 3, bottom left). Little to no correlation
existed between renal blood flow and mean systemic
blood pressure (figure 3, top right), but a positive
correlation was found between renal and systemic vascular resistance (figure 3, bottom right).
Limb hemodynamics. The mean limb blood flow in
congestive heart failure of 4.52 ± 1.77 ml/100 ml/min
was reduced below the normal mean value of 8.22 ±
3.16 ml/100 ml/min (p < .05). For each congestive
heart failure subgroup, limb blood flow was significantly reduced below control and below that in each
preceding subgroup of patients with less severe disease
(figure 1, G). Limb vascular resistance increased significantly from control for each subgroup, but there
was no significant difference in resistance between the
59
LEITHE et al.
* NL
+ CHF
800
a
800
+
+
a
+600
1N
an
+
a
a]
ale400
+
200
+
0.22X
+
+
+
+
+.++
NL'Y 266.72 exp
CHF Yz-23.04+166.10 X
NL+ CHF:~Y--45.11 +179.20 X
-
+
400
< 0.05
I
200
*+.
+
+
+++
+$
'i293,9d2X-
0077 0.006
NL~ Y
C HF Y =- 236.12+129.45 LfX 0li 10013
+
NL+CHFY=-861.38+28039 0.20
+
.0
r
.1
0.041
p
NS
NS
NS
LnX
+
0
Is5
2.5
20
3.5
3.0
CARDIAC INDEX
40
60
45
80
MEAN B
(Liters/MinJM21
Equation
r
X-0.59
NL Y= 116831 exp °000
CHFY. 1913.57-.21117 LnX -0.35
a
+
800
NL+CHF-Y-3015.74-356.37LnX -0.58
r2
0.35
0.12
0.34
<
W
+
ax +++ +
,_
a=
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
pi
CHFY.0041X
NL+CHFY- 0.031 X
0.05
++
++
120
100
D PRESSURE
(mmHq
r
Equation
NL Y= 43.57+0.040X 0.52
p
< 0.05
< 0,05
++
\+
an
600-
m
*+
.7
NL Y 23U
CF Equotion
2G1+2.5n01 r
+
+
< 0,05
< 0.05
0
0, >
+
+
+++
++++
++++
+
Uj
p
0.31
0.26
0.50
0.556
0.'51
0. 71
=
+++
r2
r r
Equation
+
+
*
600
8E
82
=
+
Ei
0
i.09e
0.61
0.66
r2
0.27
0.37
0.43
400
<
<
0.05
0.05
+
t1+
U1
CO
p
< 0.05
-
+
+
rD
D3
+
200
e
> a:
200
4
1
500
1~~~
2500
1500
SYSTEMIC VASCULAR RESISTANCE
(Dynes-sec-cm51)
+
0
1
500
1500
2500
SYSTEMIC VASCULAR RESISTANCE
(Dymes-aec-cm5 )
FIGURE 2. Correlations between hepatic blood flow and cardiac index, systemic vascular resistance, and mean systemic blood
pressure and between hepatic and systemic vascular resistances. Best-curve equation, correlation coefficient, coefficient determination, and statistical significance are given for the normal (NL), congestive heart failure (CHF), and combined populations. The
regression lines superimposed on the data points represent the regression formulas for the combined normal and congestive heart
failure (NL + CHF) populations.
heart failure subgroups (figure 1, H). A linear function
with a fair-to-good direct correlation described the relationship between cardiac index and limb blood flow
for all subjects (figure 4, top left). An exponential
function curve, with a fair inverse correlation for all
subjects studied, described the relationship between
limb blood flow and systemic vascular resistance (figure 4, bottom left). A poor correlation was found between mean systemic blood pressure and limb blood
flow (figure 4, top right). Limb vascular resistance
showed a fair positive correlation with systemic vascular resistance (figure 4, bottom right).
For most of the regression plots, the second best-fit
formula comes very close to the best-fit formula in
describing the relationship between two parameters
(table 1).
Discussion
This study has demonstrated that in a large population of patients with congestive heart failure, blood
flow to hepatic, renal, and limb regions is significantly
decreased and that this decrease in flow is proportional
and, in general, linearly related to the reduction in
cardiac output. Little to no correlation was noted between systemic blood pressure and hepatic, renal, or
60
limb blood flow. It is noteworthy that considerable
individual variation in regional blood flow and in regional vascular resistance was observed in normal subjects and in all of the congestive heart failure subgroups studied.
The mean hepatic blood flow in normal subjects
(595 ml/min/m2) comprised 17% of their mean cardiac
index of 3.5 1/min/m2. Mean hepatic blood flow in the
congestive heart failure patients was found to average
340 ml/min/m2, which is 16% of their mean cardiac
index of 2.19 1/min/m2. A previous report'3 supports
the finding that hepatic blood flow drops in proportion
to the drop in cardiac output. However, in the earlier
report the larger mean hepatic blood flow in normal
subjects (841 ml/min/m2) and in congestive heart failure patients (535 ml/min/m2) accounted for approximately 20% of the cardiac index in each group. The
most likely explanation for the increased hepatic flow
noted in the prior study is that sulfobromophthalein
clearance was used to estimate hepatic flow. Extrahepatic extraction of sulfobromophthalein ranges from
10% to 30% of the total extraction,'4 while ICG is not
cleared from the extrahepatic circulation in any appreciable amount.1' Microsphere distribution studies in
rats have also shown that hepatic blood flow generally
CIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CONGESTIVE HEART FAILURE
* NL
+ CHF
a
Equation
1000
36,,
0
1000
+
a
800
800
N
+
r
r2
NL: Y.70787-l.091X
-0.045
0.002
CHF: Y = 551.90- 1.80 X
NLW+CHF: Y476.99- 0.47X
-0.17
-0.032
0.028
0.001
M-<,
6._
600
=
400
a
+
+
200
Equation
r2
r
NL: r 404.24 exp
CHF' Y= 241 89+6E77 X
NL+CHF Y. 154.12 + 114.71 X
0.19
0-24
0.50
0.038
0.058
0.25
p
4 =
NS
NS
<0.05
W
+
1
600
g
L,
a
U
++-.
400
0
mm
1
1
200
0
2.0
1.5
3.5
3.0
CARDIAC INDEX
Liters /Min/M2)
2.5
*M11083)
U
800
a
+
a*a
1
r
r2
0.22
0.09
0.28
p
300
NS
Equation
NL:Y " 0.031 X 12
CHFIY. 1.15X0-63
NL+CHFY* 0.41X076
< 0.05
+
42 400
ai0
++
+
<1
200-
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
X
1
p
<0.05
<0.05
+0.05
+
+
+
+
+.+
_)
g e100
tr
200
r2
0.33
0. 19
0.33
+
DaE
+
r
0.S7
0.44
0.57
+
< 0.05
4-'
+
120
100
80
MEAN BLOOD PRESSURE
60
NL Y=3976.24- 479.73LnX - 0.47
-0.29
CHFY= 543. 28-0.084 X
NL + CHF:Y .2597 82 - 293.37 LnX - 0. 53
*
600
4.5
(mm Hq)
Equation
+
\
4.0
+~~~~~~~~~
0i1
+
i1 +
l
~~
~
+
++
+
+
+++
4
0
1
1
500
1500
500
2500
1500
2500
SYSTEMIC VASCULAR RESISTANCE
SYSTEMIC VASCULAR RESISTANCE
(Dynes-sec-cm- s)
Dyns-sec-cm-5)
FIGURE 3. Correlations between renal blood flow and cardiac index, systemic vascular resistance, and mean systemic blood
pressure and between renal and systemic vascular resistances. Best-curve equation, correlation coefficient, coefficient determination, and statistical significance are given for the normal (NL), congestive heart failure (CHF), and combined populations. The
regression lines superimposed on the data points represent the regression formulas for the combined normal and congestive heart
failure (NL + CHF) populations.
decreases in proportion to the drop in cardiac output.3
Renal blood flow in our normal population averaged
607 ± 172 ml/min/m2, comprising 17% of mean cardiac index. These values compare favorably with those
averaged from several other studies (660 ml/min/m2
and 18%).' The mean renal blood flow in our congestive heart failure population dropped to 395 131 ml/
min/m2; this value represents 18% of the cardiac index,
indicating that in the congestive heart failure population as a whole, the drop in renal blood flow was
proportional to the reduction in cardiac output. These
results are supported by data from studies by Blegen
and Aas'5 and Werko et al.,16 who reported mean renal
blood flow values of 370 and 404 ml/min/m2, respectively, in patients with cardiac and valvular disease and
congestive heart failure. In contrast, the mean renal
blood flow obtained in several smaller studies compiled by Wade and Bishop1 was 231 ml/min/m2, accounting for 1 1% of the mean cardiac index; this suggests that in congestive heart failure renal flow is
selectively decreased out of proportion to the drop in
cardiac output. The explanation for these discrepancies are not clear, but differences in patient population
(e.g., severity of disease, prestudy equilibration) and
Vol. 69, No. 1, January 1984
investigative techniques (e.g., continuous infusion vs
single bolus PAH) are probably contributory.
Limb blood flow in this study dropped from 8.22
4.16 ml/100 ml/min in normal subjects to 4.52 ± 1.77
ml/100 ml/min in congestive heart failure patients, a
45% reducton. Zelis et al.17 found limb blood flow in
normal and in congestive heart failure populations to
be 6.30
0.58(SEM) and 2.94 0.53 ml/100 ml/
min, respectively; this represents a comparable drop
(53%) in congestive heart failure. The lower absolute
blood flow values obtained by Zelis et al. can be explained by differences in methods. In our laboratory,
the double-occlusion method (proximal and distalwrist occlusion) generates limb flow values that range
between 9% and 30% lower than the single-occlusion
method (proximal) used in this study.
It is important to appreciate the limitations of the
techniques used in the investigation of regional blood
flow in humans. Although the techniques used produce
results that correlate well with more direct measurements, 1' 11-13 they provide only indirect determinations
of blood flow. Hepatic and renal blood flow were
estimated by clearance of dye (ICG and PAH, respectively) and limb blood flow by changes in limb vol±
±
61
LEITHE et al.
* NL
+ CHF
CHF Y. 2.06 + 3.23 LnX
NL+CHF Y0.017+ 2.22 X
20.0
r2
r
Equalon
NL: Y .260+ 1.60X
018
0.034
0.38
0.55
0.14
0. 30
p
NS
20.0 -
16.0
16.0
-
_o
E 0E12.0
0
8°
-
-
a
a
-
8<
++
+
+
+.. +
-
40
4
E
a
Ua
8
a
0.026
NS
CHF Y
0.10
0.011
NL+CHF:Y -6.81 + 2.70 LnX
0.12
0.015
NS
NS
r
a
a
+SW+ +4.
+++~~U
8.0
+
_8
v+UMe U
+
p
0.16
3-g2E
a
iL E 12. 0
2
exp0009x
5.54 exp003
a
a
O \
Equation
NL Y = 3.094
< 0.05
< 0005
-
a
+
+**
4.0
1~~
(127)
li
0
_
1~~~~~~~~~~~~~~~~~~~~
1
1.5
2.0
2.5
3.5
3.0
CARDIAC INDEX
(Lites Min/ M2)
4.0
0
4.5
60
(mmHq)
Equation
NL:Y- 12.53 +O.OOIX
Equation
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
20.0
.NL: Y
1.69 X
r
021
-0077
a-0.000
CHF:Y- 6.46 exp -0o28
NL +CHF Y. 9.48 exp 0o000x - 0.44
a
16.0
X
3DO'
X*
r2
p
0.006
NS
0.079
0.20
<0.05
c0.05
CHF Y. 10.73exp °°°
NL+CHF Y7
x
8.36exp
Fo
0.001
NS
0O41
0.17
0005
0.51
0.26
0.05
r
0.032
+- 4+
40
+
+
p
+
+
Uz-
a
E 12.0
8
120
80
100
MEAN BLOOD PRESSURE
-
in0
+
B.O
*+.+
40
t
0
++
+++
++
-
QC-
-+>oRt~~~~~
20
J
a
i
+
+
a
2
_i
a
0
0
l1
500
000
1500
2000
2500
3000
SYSTEMIC VASCULAR RESISTANCE
(Dym*s-wec-ctsl)
500
1500
2500
SYSTEMIC VASCULAR RESISTANCE
( Dynes - sec - cm- 5 )
FIGURE 4. Correlations between limb blood flow and cardiac index, systemic vascular resistance, and mean systemic blood
and between limb and systemic vascular resistances. Best-curve equation, correlation coefficient, coefficient of
determination, and statistical significance are given for the normal (NL), congestive heart failure (CHF), and combined
populations. The regression lines superimposed on the data points represent the regression formulas for the combined normal and
congestive heart failure (NL + CHF) populations.
pressure
ume/time after proximal venous occlusion. Clearance
of dye is not only dependent on blood flow to the organ
metabolizing or removing the dye, but also on that
organ's cellular function, extraction of dye by other
organs, dosage and rate of dye delivery, and other less
prominent variables. Patients with detectable hepatic
or renal disease or peripheral vascular disease were
excluded from this study. In our laboratory, nonhepatic extraction of ICG was found to range from 2% to
8% and that of PAH from 5% to 18%. The determination of percent dye extraction by a given organ requires
the placement of an indwelling catheter in the venous
system of the organ and a systemic arterial cannula.
These catheters were not placed in the patients in this
study. We elected to use 100% for liver and kidney
extraction of ICG and PAH, respectively (i.e., 0%
other-organ extraction), recognizing that our flow values may be underestimating the actual flow values by
2% to 20%. However, the same techniques and considerations were applied to each participant in both groups
in the study, suggesting that intergroup and intragroup
62
comparisons of central and regional hemodynamic values and their various ratios (e.g., regional flow/cardiac
output) are quite valid. Various noninvasive methods
(e. g., limb conductance, strain gauge, volume displacement) are available to measure limb blood flow
and their differences account for some variation in the
absolute values that have been reported. In our laboratory, the results with the gallium-indium strain-gauge
technique of determining limb volume changes per
time correlate well (r = .88) with those with the standard method of volume-displacement water-chamber
plethysmography. 18
While changes in regional blood flow are proportional to changes in cardiac output for a heart failure
population as a whole, differences within the population are apparent. A curvilinear relationship exists between systemic vacular resistance and regional blood
flow (bottom left panels of figures 2 through 4); however, it is important to remember that mathematically
an inverse curvilinear relationship exists between flow
and resistance. By examining regional vascular resisCIRCULATION
PATHOPHYSIOLOGY AND NATURAL HISTORY-CONGESTIVE HEART FAILURE
TABLE 1
Best- and second best-fit regression formulas for regional vs central hemodynamic parameters
Second-best fit
Best fit
Relationship
Group
Hepatic blood flow
NL
CHF
NL + CHF
NL
Hepatic blood flow
and systemic vascu- CHF
lar resistance
NL + CHF
NL
Hepatic blood flow
and mean blood
CHF
pressure
NL + CHF
NL
Hepatic vascular resistance and systemic CHF
vascular resistance
NL+CHF
Renal blood flow
NL
and cardiac index
CHF
NL + CHF
Renal blood flow
NL
and systemic vasCHF
cular resistance
NL + CHF
Renal blood flow
NL
and mean blood
CHF
pressure
NL + CHF
Renal vascular reNL
sistance and systemic CHF
vascular resistance
NL + CHF
Limb blood flow
NL
and cardiac index
CHF
NL + CHF
Limb blood flow
NL
and systemic vascu- CHF
NL+CHF
lar resistance
Limb blood flow and
NL
mean blood pressure CHF
NL+CHF
Limb vascular resisNL
tance and systemic
CHF
vascular resistance
NL+CHF
and cardiac index
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
CHF
=
Equation
congestive heart failure; NL
Y =266.72exp0 -22X
Y -23.04 + 166.I1OX
Y =-45.11l±179.20X
Y= 1168.31expo fwx
Y =1913.57 -211.17LnX
Y =3015.74 -356.37LnX
Y =1293.92X -0.18"
Y =-236.12 + 129.45LnX
Y =-861.38 +280.39LnX
Y=43.57+O.040X
Y =0.04 1X'-098
Y =0.03 1X'. '4
Y =404.24exp0. !07X
Y =241.89 +68.77X
Y =154.12 +114.71X
Y 3976.24 -479.73LnX
Y =543.28 -0.08~4X
Y =2597.82 -293.37LnX
Y =707.87 -1.091 X
Y =a551.90 1.80X
Y =476.99 -0.47X
Y' 0.03 1X"1
Y =l. 15X06
Y =0.41 X076
Y =2.60 +1.60X
Y =2.06 +3.23LnX
Y =0. 17±+2.22X
Y =1. 69Xl'"2
Y =6.46exp OOGI x
Y =9.48exp 1000 1 x
Y =3.094expO(.1Ox
Y =p5.54exp reY
6.81 2.7OLnX
Y =12.53 +0.Ol X
Y= 10.73expllOI10x
Y =8.36expOl-00x
=
=
r2
.56
.51
.71
.59
.35
.58
.077
.11
.20
.52
.61
.66
.19
.24
.50
.47
.29
.53
.045
.17
.032
.57
.44
.57
.18
.38
.55
.077
.28
.44
.16
.10
.12
.032
.41
.51
.31
.26
.50
.35
.12
.34
.006
.013
.041
.27
.37
.43
.038
.058
.25
.22
.09
.28
.002
.028
.001
.33
.19
.33
.034
.14
.30
.006
.079
.20
.026
.011
.015
.001
.17
.26
Equation
127.89+ 133.43X
62.73 + 364.66LnX
Y= = 6.019 + 452.24LnX
Y= = 1007.023 - 0.362X
y=
Y= = 545.17 - 0.12X
Y= =735.90 -0.21X
Y = 686.13exp - 0().x)x
Y =42.75X044
Y
=
Y= =
y =
8.8 17xn
52.3 1 exp R (SOI IX
Y
=
Y
= 44.44expx
°I x
Y
=
42.48exp(1 (X)IX
367.10+68.54X
= 283.86 + 142.54LnX
= 195.56 + 279.57LnX
= 1079.78 -0.42X
= 3121.OlX O284
= 717.12 - 0.17X
= 667.40exp - `IO) x
= 1071.81 - 151.77LnX
=522.48 -19.39LnX
= 26.97exp(0 (OlIIX
= 64.29exp(' (IX)OIX
y =
y
y
y
y
Y
y
Y
y
Y
y
52.84exp(' ()(W(IX
1.67 + 5.26LnX
= 1.43 + 1.41 X
= 2.036exp(0 44x
y =
y =
Y
Y
Y
Y
Y
y
Y
Y
Y
y
y
0.892 + 1.30LnX
73.24X- (83
= 642.78X 067
= 0. 17X 827
= 12. 19X -O)24
= 2.94 + 0.026X
= 11.69exp" MIX
= 6.84 + 0.009X
= 4.23 + 0.0 lOX
=
r
r
.55
.50
.70
.58
.32
.55
.071
.10
.19
.50
.60
.65
.18
.22
.48
.45
.23
.49
.044
.16
-.01
.56
.43
.55
.17
.37
.55
.055
.27
.43
.15
.09
.11
.031
.40
.50
.30
.25
.49
.34
.103
.30
.005
.010
.036
.25
.36
.42
.032
.050
.23
.20
.052
.24
.002
.025
.0001
.31
.18
.31
.030
.14
.30
.003
.071
.18
.022
.008
.012
.001
.16
.25
normal subjects.
tances for normal subjects and patients in three heart
failure subgroups (figure 1, D, F, and H), and by
plotting regional resistances against systemic vascular
resistance (bottom right panels of figures 2 through 4),
it appears that regional hemodynamic changes are not
quite homogeneous in a degree relative to the severity
of heart failure or to the level of systemic vascular
resistance. The increase in renal vascular resistance
appears to be less for a given increase in systemic
vascular resistance at higher levels of resistance (figure
3, bottom right) and as the disease of patients proceeds
from moderate to severe (figure 1, F). The attenuation
of change in renal vascular resistance probably accounts for the observed lack of change in renal blood
Vol. 69, No. 1, January 1984
r
flow as the disease progresses from moderate to severe
heart failure (figure 1, E), despite a significant drop in
cardiac output. This suggests that the sympathetic nervous system,and perhaps other compensatory mechanisms (e.g., renin-angiotensin, vasopressin), play a
less dominant role in the regulation of renal blood flow
as heart failure becomes more severe. Vascular reactivity to norepinephrine and to nerve stimulation decreases in the animal preparation of congestive heart
failure. 19 It is likely that "autoregulation," the mechanisms of which is not yet precisely defined,20 also plays
an important role in the attenuation of changes in renal
vascular resistance and flow as heart failure worsens.
In contrast to the renal circulation, attenuation in
63
LEITHE et al.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
flow and resistance changes with increasing severity of
heart failure was not apparent for the hepatic or limb
circulations. The explanation for this disparity is not
provided by this study, but is probably related to differing vascular regulation (e.g., responsivity, regulatory mechanisms, autoregulation) for the various organ systems as the cardiac function changes from
normal to severely diseased.
In conclusion, this study has provided mean values
for hepatic, renal, and limb blood flow in normal subjects and patients with congestive heart failure. For the
heart failure population as a whole, the decline in
blood flow to the renal, hepatic, and limb vascular
beds was proportional to the decrease in cardiac output. No disproportionate shunting of blood away from
these "sympathetically innervated" regions was observed in patients in the resting state. In fact, the magnitude of increase in renal vascular resistance and decrease in renal blood flow appeared to become
attenuated as the severity of heart failure evolved from
moderate to severe. The findings of this study do not
pertain to exercising congestive heart failure patients.
Exercise appears to elicit more dramatic alterations in
renal, hepatic, and limb blood flow through accentuated changes in systemic and regional vascular resistance and blood pressure.24 17 More pronounced alterations in regional blood flow are also likely to occur in
the setting of acute and/or extreme decompensation of
cardiac failure. Finally, although central hemodynamics generally correlated with blood flow to the regions
studied, there was great variation in the individual
blood flow values at each level of cardiac output and
systemic resistance, reflecting the complex nature of
regional blood flow regulation. This variability does
not allow for a very precise prediction, via regression
formulas, of regional blood flow from central hemodynamic data alone and vice versa.
We acknowledge the technical assistance of Pat Huss, R.N.,
Max Bacher, Mitzi Prosser, and Elena Popescu, and also the
professional care our study patients received from the nursing
and house staff of the Coronary Care Unit of The Ohio State
University Hospitals.
64
References
1. Wade OL, Bishop JM: Cardiac output and regional blood flow.
Oxford, 1962, Blackwell Scientific, pp 134-148
2. Zelis R, Longhurst J: The circulation in congestive heart failure. In
Zelis R, editor: The peripheral circulations. New York, 1975,
Grune and Stratton, pp 283-314
3. Zelis R, Flaim SF, Nelis S, Longhurst J, Moskowitz R: Autonomic
adjustments to congestive heart failure and their consequences. In
Fishman AP, editor: Heart failure. Washington, 1978, Hemisphere
Publishing Corporation, pp 237-247
4. Higgins CB, Vatner SF, Franklin D, Braunwald E: Effects of
experimentally produced heart failure on the peripheral vascular
response to severe exercise in conscious dogs. Circ Res 31: 186,
1972
5. Ganz W, Swan HJC: Measurements of blood flow by thermodilution. Am J Cardiol 29: 241, 1972
6. Maclntyre WJ, Pritchard WH, Moir TW: The determination of
cardiac output by the dilution method without arterial sampling.
Circulation 18: 1139, 1958
7. Schreiner BF Jr, Lovejoy FW Jr, Yu PN: Estimation of cardiac
output in patients with cardiopulmonary disease. Circ Res 7: 595,
1959
8. Kloster FE, Bristow JD, Griswold HE: Cardiac output determination from precordial isotope-dilution curves during exercise. J
Appl Physiol 26: 465, 1969
9. Weigand BD, Ketterer SG, Rapaport E: The use of indocyanine
green for the evaluation of hepatic function and blood flow in man.
Am J Digestive Dis 5: 427, 1960
10. Caesar J, Sheldon S, Chiandussi L, Guevara L, Sherlock S: The
use of indocyanine green in the measurement of hepatic function.
Clin Sci 21: 43, 1961
11. Wheeler HO, Cranston WI, Meltzer JD: Hepatic uptake and biliary
excretion of indocyanine green in the dog. Proc Soc Exp Biol 99:
11, 1958
12. Chasis H, Redish J, Goldring W, Ranges HA, Smith HW: Use of
sodium p-aminohippurate for functional evaluation of human kidney. J Clin Invest 24: 583, 1945
13. Myers JD, Hickam JB: Estimation of hepatic blood flow and
splanchnic oxygen consumption in heart failure. J Clin Invest 27:
620, 1948
14. Hanson KM: Liver. In Johnson PC, editor: Peripheral circulation.
New York, 1978, Wiley Medical, pp 285-314
15. Blegen E, Aas K: Renal blood flow and glomerular filtration rate in
patients with valvular heart lesions. Acta Med Scand 138: 391,
1950
16. Werko L, Vamauskas E, Eliasch H, Ek J, Bucht H, Thomasson B,
Bergstrom J: Studies on the renal circulation and renal fraction in
mitral valvular disease. Circulation 9: 687. 1954
17. Zelis R, Mason DT, Braunwald E: Partition of blood flow to the
cutaneous and muscular beds of the forearm at rest and during leg
exercises in normal subjects and in patients with heart failure. Circ
Res 24: 799, 1969
18. Leier CV, Heban P, Huss P, Bush CA, Lewis RP: Comparative
systemic and regional hemodynamic effects of dopamine and dobutamine in patients with cardiac failure. Circulation 56: 466, 1978
19. Mark AL, Mayer HE, Schmid PG, Heistad DD, Abboud FM:
Adrenergic control of the peripheral circulation in cardiomyopathic
hamsters with heart failure. Circ Res 33: 74, 1973
20. Johnson PC: Principles of peripheral circulatory control. In Johnson PC, editor: Peripheral circulation. New York, 1978, Wiley
Medical, pp 1 1 1- 139
CIRCULATION
Relationship between central hemodynamics and regional blood flow in normal subjects
and in patients with congestive heart failure.
M E Leithe, R D Margorien, J B Hermiller, D V Unverferth and C V Leier
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Circulation. 1984;69:57-64
doi: 10.1161/01.CIR.69.1.57
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1984 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/69/1/57
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/