Download Changes in Pulmonary Venous Return during

Clinical Science (1997) 93,
13-20 (Printed in Great Britain)
13
Changes in pulmonary venous return during head-up tilting
in man
Mareo GUAZZI, Gloria TAMBORINI and Anna MALTAGLIATI
lstituto di Cardiologia dell'llniversita degli Stud;, Centro di Studio per le Ricerche Cardiovascolari del
Consiglio Nazionale delle Ricerche, Fondasione 'Monzino', I.R.C.C.S., Via C. Parea 4, 20 138 Milan, Italy
(Received I 2 August
I996/2I February 1997; accepted I0 March 1997)
~~
1. In a supine position, the heart fills to close to the
limits of pericardial constraint and the pericardium
may act to redistribute central blood volume from
the left side of the heart back to the more compliant
lung.
2. We probed whether, and through which mechanisms, a redistribution of blood from the lungs to the
left heart occurs during vertical displacement and
compensates for reduced venous return.
3. We investigated 16 normal volunteers with
Doppler-echocardiography during 20", 40" and 60"
head-up tilting. Tilting was stopped at 10 min in 10
subjects (group I) and at 45 min in 6 subjects
(group 2).
4. At 10 niin we observed a reduction in right ventricular diastolic dimension and left ventricular
end-diastolic pressure, as estimated by the difference between the duration of the pulmonary venous
flow during atrial contraction (7, wave) and that of
the mitral A wave. We also recorded a decrease
during systole (X wave) and an increase during diastole (Y wave) of the pulmonary venous forward
flow velocity. These variations were evident at 20"
and became progressively greater with increasing
degrees of tilting. In group 2, changes at 10 min and
at 45 min for any degree of displacement were similar.
5. A decrease in right ventricular dimensions (ventricular interdependence) and underfilling of the
lung compartment due to volume redistribution to
the periphery (diminished lung contribution to pericardial constraint) augment compliance within the
pericardial space, reduce downstream pressure for
pulmonary venous return and move the puimonary
venous flow predominantly to ventricular diastole,
allowing diastolic filling.
6. During head-up tilting a favourable interaction
between heart and lungs increases compliance
within the pericardial space and facilitates redistribution of blood from the lungs, resulting in a
sustained compensation for the reduced venous
return.
lNTWODUCTlON
Head-up tilting is a method widely utilized for
investigating circulatory adaptation to the body's
vertical position [l].The importance of the neural
reaction to changes in posture is well established
[2-41. Redistribution of blood from the lung is also
a basic homoeostatic compensation for a fall in right
ventricular stroke output iminediately after tilting
from supine to erect position [2]. Even if the ability
of the lung to perform the function of blood reservoir has long been known to physiologists and clinicians IS-81. the mechanisms underlying the
orthostatic redistribution of blood in man have not
receivcd adequate attention.
At lower degrees of displacement during graded
head-up tilting, venous return is diminished. but left
ventricular output is maintained in the absence of
variations in heart rate and peripheral vascular
resistance [9]. This suggests that haeinodynamic
adjustments, possibly of the downstream pressure
for pulmonary venous return and left ventricular filling, occur, which are independent of such responses.
Under normal conditions in a supine position,
pericardial constraint becomes the dominant influence on the degree of cardiac filling and ventricular
end-diastolic pressure, and the pericardium may act
to redistribute volunie from the left side of the heart
back to the more compliant lungs. Our hypothesis is
that the diminished venous return with tilting may
reduce constraint and increase compliance within
the pericardial space, resulting in a reverse redistribution of blood from the lung pool to the left side
of the heart and systemic circulation.
To test this hypothesis we investigated, non-invasively, normal volunteers with paired assessment
(echo-Doppler) of the pulmonary vcnous flow and
left ventricular filling patterns.
METHODS
Study subjects
Sixteen healthy men were the subjects of this
study. All were students, hospital or university staff
Key words: head-up tilting, pericardial constraint, pulmonary venous return, ventricular interdependence.
Correspondence: Dr M. Guani.
14
M. Guazzi et al.
or their relatives: age 34 k8 years, weight 85 -t 7 kg
and height 179k6 cm. They were non-smokers, had
no history of cardiovascular or pulmonary diseases
and had normal aerobic capacity. None was taking
cardiovascular drugs, had diabetes, varicose veins,
disturbances of cardiac rhythm or conduction. Physical examination, chest roengenogram, blood
chemistry and electrocardiogram were within normal
limits.
Echocardiogram (including pulsed Doppler and
colour Doppler examinations) was also normal and,
in particular, excluded the presence of pericardial
effusion and cardiac hypertrophy. None could be
considered a professional sportsman.
The protocol was approved by the institutional
Ethics Committee and informed written consent was
obtained from each subject who participated in the
study.
Head-up tilting protocol
Subjects underwent tilting between 9 :00 hours
and 11:00 hours in a quiet temperature-controlled
environment (21-23"C), after an overnight fast. For
head-up tilting we used a motorized cantilevered
table with foot support; heart rate was monitored
continuously and electrocardiographic recordings
were obtained at 1 min intervals throughout the
study; heart rate and blood pressure were determined at 1min intervals with the Dinamap system
(Critikon, Tampa, FL, U.S.A.) which computes
blood pressure over a period of 15-30s. Subjects
were not strapped to the table and were instructed
not to strain muscles during the tilt.
Ten subjects (group 1) were tilted to angles of
20", 40" and 60" to the horizontal for 10min each.
All subjects rested initially for at least 15 min on the
table in the horizontal position until the heart rate
in consecutive minutes varied by no more than
3 beatdmin. Tilting steps were separated by intervals
of 15min, during which subjects were returned to
the horizontal position and allowed to reach a
steady state before further estimates were made.
Records were taken within the first 2 min of the initiation of head-up tilting and after 10min. We did
not wish to tilt beyond 60" because the angle would
have required active muscle tension in the subjects.
Six other volunteers (group 2) were subjected to
20", 40" and 60" tests of 45min duration, each on
separate days, with the aim of probing whether
adaptive adjustments tend to exhaustion with more
prolonged duration of the test.
Among the 20 subjects who were originally enrolled for the study, four were excluded because satisfactory pulmonary venous flow velocities throughout
the cardiac cycle could not be obtained.
Doppler and echocardiographic recordings
Pulsed Doppler echocardiographic recordings
were obtained with a phased-array echocardio-
graphic-Doppler system (Sonos 1000 Hewlett-Packard, Palo Alto, CA, U.S.A.) and a transducer array
of 2.5 MHz. The subjects lay in a very slight left
lateral decubitus position [lo] during quiet respiration. A paper speed recording velocity of 50mm/s
was used for flow velocity measurements and
100 mm/s with simultaneous recording of the electrocardiogram and phonocardiogram (using a
100 Hz filter at 12 dB/octave with a contact microphone applied to the precordium where the aortic
component of the second heart sound was loudest)
for isovolumic relaxation time.
The velocities of the pulmonary venous flow were
examined with the transducer placed at the cardiac
apex and were obtained by placing the sample
volume 0.5-1 cm into the upper right pulmonary
vein. The vein was visualized by a slightly cephalad
elevation of the interrogation plane from a standard
four-chamber view. The position of the sample
volume was confirmed by obtaining a characteristic
pulmonary venous flow pattern. The high-pass filter
was minimized and the settings of velocity, baseline
and time resolution were adjusted during recording
in order to achieve the largest possible screen display of the velocity curves. After the pulmonary
venous flow velocity was examined, the mitral flow
velocity pattern was obtained using an apical fourchamber plane. The sample volume was placed just
distal to the tips of the open mitral valve, with
minor adjustments being made until maximal peak
flow velocity with a narrow spectrum was reproducibly obtained. Care was taken to insure that the
sample volume positions remained constant for
examinations at each heart rate.
Two-dimensionally directed M-mode echocardiograms were recorded of the septum and left ventricular posterior wall, immediately below the mitral
valve leaflets from a parasternal short-axis window.
A two-dimensional apical four-chamber view was
also recorded in all patients.
Data analysis
Off-line quantification of the Doppler recordings
was performed with a computer-integrated digitizing
pad and specifically designed software to measure
time intervals, velocities and velocity integrals. The
pulmonary venous flow velocity profiles were traced
along the instantaneous highest velocity spectra by
hand, to determine peak forward flow velocities
during systole (X) and diastole (Y), flow velocity
integrals of systolic and diastolic forward flow waves
and duration of the flow reversal during atrial systole (Z) (Fig.1). This interval, if the velocity signal
of atrial reversal was poor (2 cases) was obtained by
using the time interval between the start of atrial
contraction with cessation of early diastolic flow and
the start of forward systolic flow [ll]. The flow velocity integral of the systolic forward flow wave was
defined as the area under the traced velocity profile
from the onset of the forward flow to the onset of
Pulmonary venous flow during tilting
the diastolic forward flow wave, and the flow velocity integral of the diastolic forward flow wave was
defined as the area from the onset of the diastolic
forward flow wave to the end of the forward flow
(Fig. 1). The systolic filling fraction of pulmonary
venous forward flow was the ratio of the systolic to
the sum of the systolic and diastolic velocity integrals.
For the mitral valve, the peak early diastolic filling velocity (E), peak filling velocity at atrial contraction (A), flow velocity integrals of the early
diastolic filling wave, and duration and flow velocity
integral of the filling wave at atrial contraction were
determined. The velocity integral of the early diastolic filling wave was the area under the traced flow
velocity profile during the early diastolic filling, and
the velocity integral of the filling wave at atrial contraction was the area during the period of atrial contraction. The difference between duration of the
pulmonary Z wave and mitral A wave was utilized as
an index of the left ventricular end-diastolic pressure [ll].The left ventricular posterior and septa1
endocardia1 surfaces were digitized at a level
immediately below the mitral valve leaflets. Variables included left ventricular end-diastolic and endsystolic dimensions and wall thickness. The right
ventricular area at end-diastole was measured from
two-dimensional images by using a cine-loop display
SUPINE
40"
15
[12]. The left ventricular isovolumic relaxation time
was measured from aortic valve closure to the start
of the mitral flow. Left ventricular stroke output was
calculated as the velocity time integral of the systolic
velocity spectrum recorded in the outflow tract of
the left ventricle times the subvalvular area of the
outflow tract [13]. The averages from any six clear
consecutive cardiac cycles were used for quantitative
analysis. This method accounts for respiratory variations in the pulmonary venous and mitral flow tracings; however, variations are very negligible when
respiration is quiet and regular.
Reproducibility
All studies were reviewed by two independent
echocardiographers. The intraobserver (comparing
paired readings obtained by the same observer on
two separate occasions) and the interobserver (comparing results obtained and analysed by two observers for the same subject) coefficients of variation
were 8% and 11%for pulmonary venous flow and
6% and 9% for transmitral flow respectively.
Statistical analysis
Data are expressed as mean valuesf 1 SD. The
significance of differences between serial measurements was assessed by using analysis of variance for
20"
60"
Fig. 1. Raw data traces demonstrating the typical pulmonary venous flow changes from supine through 60" tilting across one patient in group I,with
schematic diagrams illustratinghow key measurementsare carried out. X = peak systolic forward flow velocity; Xvti = systolic velocity integral; Y = peak diastolic
forward flow velocity; Yvti = diastolic velocity integral; Z = peak velocity of flow reversal during atrial systole; Zd =duration of reverse flow during atrial systole.
M. Guazzi et al.
16
repeated measures and Newman-Keuls multiple
comparison procedure. Post hoc analysis was not
undertaken unless analysis of variance reached
statistical significance. Paired or unpaired Student's
t-test was also used, as appropriate. Differences at
the P <0.05 level were considered statistically significant.
RESULTS
Head-up tilting was well tolerated and no subject experienced symptoms or discomfort. In group
1, records were taken within the first 2min of the
initiation of displacement and after 10 min. There
was a slight potentiation at 10min of the changes
observed at 2min, however, in no case were differences statistically significant. For reasons of simplicity, only results at 10 min are reported here.
Group I
Pulmonary and mitral flow velocities and stroke
volume. As shown in Table 1, baseline left ventricular stroke volume averaged 74 _+39ml and was not
affected by head-up tilting at 20". This angle of tilting, however, was associated with a significant
decrease of the peak velocity and time velocity
integral during forward systolic flow (X) and an
increase during diastolic flow (Y), resulting in
reduction of the X/Y ratio, a decrease in systolic fill-
ing fraction, duration of flow reversal during atrial
contraction, peak E wave velocity and E/A ratio of
mitral flow.
Angles of tilting of 40" and 60" were associated
with reduction of stroke volume, no further decrease
of peak X and E wave velocities and progressively
greater peak Y wave velocity. At 60" head-up position, peak systolic flow velocity and time velocity
integral, systolic fraction during forward pulmonary
flow, XN ratio and duration of flow reversal during
atrial contraction were, respectively, 20%, 43%,
23%, 32% and 50% less than values in the supine
position. Peak velocity and time velocity integrals
during diastolic forward flow were 27% and 50%
greater than values in the horizontal position; mitral
peak E velocity and E/A ratio were reduced by 18%.
Changes in mitral E/A and pulmonary vein XN
peak velocity ratios were not related to those in
heart rate.
Peak velocity and duration of the A wave did not
vary significantly, whereas the duration of flow
reversal during atrial contraction ( Z ) became progressivcly shorter in parallel with increasing degrees
of tilting. Because of this, the difference in duration
of the two waves was reduced at 20" as compared
with the supine position, and became increasingly
negativc at 40" and 60" positioning (Fig. 2), suggesting a progressive reduction in left ventricular enddiastolic pressure [ll]. These changes were
associated with a decrease in the time velocity
integral of the pulmonary X wave and an increase in
Table I. Doppler variables in group I subjects in the supine position and at difierent degrees of head-up tilting. *P <0.05
compared with supine: #P<O.Ol compared with supine; t P 10.01 compared with immediately lower tilting degree. E = peak early
inflow velocity: A = peak late inflow velocity: E/A = ratio of early to late peak mitral flow velocity; X = pulmonary venous peak
forward flow velocity during systole; Y = pulmonary venous peak forward flow velocity during diastole: Z = pulmonary venous flow
reversal during atrial contraction; XM = ratio of pulmonary venous systolic to diastolic forward flow velocity. Results are presented as
mean values fSD.
Supine
Stroke volume (ml)
74 2 39
20" Tilt (I 0 min)
73f28
40" Tilt (10 min)
60"Tilt (10min)
+
68 k 26#t
65 24#
Pulmonary flow
Peak X wave velocity (mls)
Peak Y wave velocity (mls)
Peak XM wave velocity ratio
Y deceleration slope (mlsl)
Z wave duration (mls)
Time velocity integral of X wave (cm)
Time velocity integral of Y wave (cm)
Percentage of total flow velocity integral
of X wave
0.51 20.1
0.44 f0.7
1.1 f0.3
283 k37
188+34
I4k I
10f2
60
0.39 k0.08#
0.48 2 0.I*
0.85 0.2#
307 2 24
157+38#
10.9+2.5#
13+4#
46#
0.36 +0.07#
0.49 +O. I*
0.76 +O. IS#
305 38
121 rfi 13#t
8+3#
14+5#
36#
0.4 I k0.07*
0.56 0.12#
0.75 f0.18#
281 2 4 7
94 7#t
0.7 & 2.8#
Mitral flow
Peak E wave velocity (mls)
Peak A wave velocity (m/s)
Peak E/A wave velocity (mls)
Early deceleration slope (m/sz)
Time velocity integral of E wave (cm)
Time velocity integral of A wave (cm)
Percentage of total flow velocity integral
of A wave
0.87kO.l
0.53 k0.09
1.720.4
238rfi18
1624
7 1I
31
0.75 k0.14*
0.47 f0.07
I .6 k0.3*
257+ 19
14f2.7
6 f0.9
29
0.70+0. I*
0.48 0.05
I.5 k0.2*
252 23
1422
5 f0.8
27
0.72&0.1*
0.5 I k0.06
I .4 +0.3#
258 f 32
14f2.6
6* 1.4
+
+
+
+
+
1525#
37#
19
Pulmonary venous flow during tilting
Supine
20"Tilt
17
40" Tilt
60"Tilt
50
40
I
30 N
Q
20
5'
10
g
OF
-l
-10
2
6'
3
-20
2
$
-30 S.
-40
-50
#
A
Fig. 2. Mean (1SD)differences in duration of pulmonary venous flow reversal (Z wave) and mitral A wave during atrial contraction, and the mean time
velocity integral of the X and Y pulmonary venous waves, in the supine position and at different degrees of tilting in group Isubjects. *Indicates differences
from supine significant at P<O.O5; #indicates differences from supine significant at P<O.Ol; Aindicates differences from immediately lower tilting degree significant at
P<O.Ol.
the time velocity integral of the Y wave. The time
velocity integral of the E mitral wave was not significantly diminished with increasing orthostatic stimulation. Tilting did not augment the slope of
deceleration of the E mitral wave and did not affect
left ventricular isovolumic relaxation time.
Haemodynamics and ventricular dimensions. Heart
rate, systolic and diastolic blood pressure and
cardiac output were not altered by head-up 20" tilting (Table 2). Positions at 40" and 60" were associated with a significant increase in heart rate and
diastolic pressure and a decrease in systolic pres-
sure, cardiac output (Table 2) and stroke volume
(Table 1).
In the supine position, the left ventricular enddiastolic diameter averaged 48 + 2 mm and the diastolic area of the right ventricle averaged 18+4 cm2
(Table 2). During 20" head-up positioning, there
were no dimensional changes of the left ventricle,
whereas the right ventricular area reduced to
14 3 cm2(P < O.Ol), a 22% decrease from baseline.
With an increase in the orthostatic stimulus the
trend of the right ventricular cavity was towards
some further shrinking, which, however, did not
Table 2. Haemodynamics and ventricular dimensions in group I subjects while supine and at different degrees of head-up
tilting. *P <0.05 compared with supine; #P<O.Ol compared with supine. Results are presented as mean values+SD.
Supine
66f5
20" Tilt (10 min)
69+9
40" Tilt (10 min)
60" Tilt (10min)
73 7*
+
79 f 6#
I l0+6*
84+7
5400 980*
108+6*
89+5*
5600 +930*
Heart rate (beatslmin)
Arterial pressure (mmHg)
Systolic
Diastolic
Cardiac output (ml/min)
I l9+3
79+3
6100+830
I l6+5
77+6
5400 850*
Left ventricle
Systolic diameter (mm)
Diastolic diameter (mm)
28+3
48+2
28+3
4753
31 +3
46+2
26+4
45 2*
Right ventricle
Diastolic area (cm2)
18+4
14+3#
12+3#
13+4#
+
+
+
M. Guazzi et al.
18
reach statistical significance; the diastolic dimension
of the left ventricle showed some reduction at 60"
tilting, and the end-systolic diameter did not vary
significantly during the study.
flow, however, suggest that some decrease in venous
return had taken place and tachycardia and vasoconstriction did not participate in the maintenance of
left ventricular filling and output. Downstream
adjustments for pulmonary venous return and redistribution of blood from the lungs had probably
occurred, which compensated for the fall in right
ventricular stroke output [2, 5-81 immediately after
tilting and variations from the supine position in the
velocity profile of the pulmonary venous flow, i.e.
decreased velocity during ventricular systole and
increased velocity during diastole, probably reflected
a different pattern of flowing and not a diminished
amount of flow.
Because the cardiac chambers are situated in a
space limited by the pericardium, changes in the
volume of a chamber may affect the volume of other
chambers. The interaction between the two ventricles, which is called ventricular interdependence, is
significantly modulated by the pericardium [S].
There is also an atrioventricular interaction which is
characterized by how atrial and ventricular filling
and emptying are coupled. Coupling may be complete (the atrium is filled with a volume equal to
stroke volume during ventricular ejection), or less
than complete (the atrium receives some blood from
Group 2
Table 3 summarizes data in group 2 subjects. In
this group the patterns of heart rate, blood pressure,
stroke volume and flow velocities through the pulmonary vein and the mitral valve at 10 min after 20",
40" and 60" tilting were similar to those in group 1.
Within subjects in group 2 there were no statistically
significant differences between values at 10 min and
at 45 min head-up tilting.
DISCUSSION
Basic adjustments
By considering heart rate, blood pressure and
cardiac output, tilting at 20" appeared to be a stimulus of too low an intensity to reduce venous return.
Changes in right ventricular diastolic dimension and
in the pattern of pulmonary venous and transmitral
Table 3. Haemodynamic and Doppler variables in group 2 subjects in the supine position and at 10 min and 45 min of
different degrees of head-up tilting. *P <0.05compared with supine; #P <0.01 compared with supine; tP <0.01 compared with
immediately lower tilting degree. E = peak early inflow velocity; A = peak late inflow velocity; E/A = ratio of early to late peak mitral
flow velocity; X = pulmonary venous peak forward flow velocity during systole; Y = pulmonary venous peak forward flow velocity
during diastole; Z = pulmonary venous flow reversal during atrial contraction; XN = ratio of pulmonary venous systolic to diastolic
forward flow velocity. Results are presented as mean values fSD.
20" Tilt
Supine
40" Tilt
10 min
45 min
10 min
60" Tilt
45 min
10min
45 rnin
80+10#
Heart rate (beatshin)
Arterial pressure (mmHg)
Systolic
Diastolic
66k3
68+5
68f5
67+4
72+4#
80flO#
120k14
70f3
121k11
72+2
118+11
72+2
118+12
78f4*
118f9
77&7*
117+12
77+8#
118+12
83+7#
Stroke volume (ml)
85+8
79+8
76+10
74+9*
68+1I*t
65+II*
65+II*
Pulmonary flow
Peak X wave velocity (mls)
0.46+0.1
Peak Y wave velocity (m/s)
0.50 0.06
Peak XN wave velocity ratio
0.9k0.3
Z wave duration (m/s)
170k30
TimevelocityintegralofXwave(cm)
14+5
TimevelocityintegralofYwave(cm)
15f3
Percentageof total flow velocity integral 43
of X wave
0.40+0.I
0.5 I 0.07
0.8f0.3
165+25*
10.6+4*
18+4*
36*
Mitral flow
0.77k0.2
Peak E wave velocity (mls)
Peak A wave velocity (m/s)
0.49kO.l
Peak E/A wave velocity ratio
1.6k0.2
TimevelocityintegralofEwave(cm)
17k3
Time velocity integral of A wave (cm)
7 k2
Percentage of total flow velocity integral 28
of A wave
0.72f0.2* 0.66+0.2# 0.60+0.I#
0.45f0.08 0.47k0.08 0.47k0.07
1.6f0.2
1.4+0.3*
1.3f0.2*
16+4
15f3
13+3
7 f2
7k0.2
6 f I .5
26
32
29
+
+
0.40+0.1
0.34*0.05* 0.38+0.06# 0.38*0.06#
0.49 kO.08 0.50f0.06 0.49 0.09 0.57 f0. I #t
0.8k0.3
0.7fO.I*
0.8k0.2
0.7f0.06*
150+27*
120+27*t I I S + l3#t 95+9#t
11+4*
9+3*
11&3*
10+3*
17f5*
16+4
14+4
17+4*
40
36*
44
37#
+
0.60f0.2#
0.45f0.07
1.4+0.2*
14+3
6kI
29
0.60+0.l#
0.42k0.07
1.4+0.2*
12k3
5k1
28
0.37+0.I#
0.56 f0. I#t
0.7+0.1*
90+7#t
10+3*
16*3*
38+
0.56f0.1#
0.41 f0.05
1.4+0.3*
I I +4*
5k0.6
29
Pulmonary venous flow during tilting
upstream during ventricular diastole) [14]. The
degree of coupling is extreme during tamponade.
Under normal conditions the presence of substantial
diastolic atrial inflow, which increases minimally
with removal of the pericardium, implies that myocardial factors, more than the pericardium, determine the degree of atrioventricular interaction [14].
Under normal conditions in a supine position, the
heart fills to close to the limits where pericardial
constraint becomes the dominant influence on ventricular end-diastolic pressure. Under these conditions, the pericardium may act to redistribute central
blood volume from the left side of the heart back to
the more compliant lungs. With diminished venous
return with tilting, the decrease in right ventricular
volume, via ventricular interdependence, and the
movement of the pressure-volume relationship from
being mediated through the pericardium to being
mediated by myocardial properties, would reduce
the downstream pressure for pulmonary venous
return and increase compliance within the pericardial space. Redistribution of central blood volume
from the lungs to the periphery would also diminish
the contribution of the lungs to pericardial constraint [15]. The considerable changes in the difference in duration between pulmonary flow reversal
during atrial contraction and the A mitral wave
occurring during tilting (Fig. 2) are consonant with
these interpretations, as they reflect reduction in
left ventricular end-diastolic pressure. Variations
described in our findings in pulmonary venous and
mitral flow velocities are also compatible with a
decrease in intraventricular pressure and improved
compliance within the pericardial space.
In the supine position the heart is to a minor
degree exhibiting ‘cardiac tamponade’ mediated by
pericardial constraint; from this position the reduction in total pericardial volume and constraint
during tilting will move the venous return predominantly to ventricular diastole, during which the
atrium functions more as a conduit [16], allowing
diastolic filling. Reduction in volume loading with
tilting and changes in the surroundings in which the
heart contracts might also facilitate the left ventricular storage of potential energy during ejection that is
subsequently released during early diastole (elastic
recoil) to assist ventricular filling [17]. As to the
reduced pulmonary venous forward output during
systole, another explanation may be that the
hydraulic connection between right ventricular outflow and pulmonary venous flow represents a driving
force for pulmonary venous return [16,18] when the
pulmonary intravascular volume is replete, but not
when the compartment is underfilled, as during
orthostatic displacement.
Neural reaction
Tachycardia and peripheral vasoconstriction
would ensue when reduction in pericardial con-
19
straint and redistribution of central blood volume
become unable to compensate for venous pooling
and arterial underfilling. Unloading of the cardiopulmonary and/or systemic baroreceptors may
generate the signal [19] for the neural reaction. It is
unknown whether tachycardia is mainly due to
increased sympathetic or inhibited parasympathetic
activity.
Head-up tilting and fainting
Tilting may be associated with fainting, which
would result from inappropriate stimulation of
receptors located in the left ventricle [20], or from
sudden collapse of unfilled atria and great veins
which could signal the false information to the brain
that the heart is overfilled rather than underfilled
[21]. Our study suggests that a disordered heartlung interaction might be an additional reason for
orthostatic intolerance. Results in group 2 indicate
that in normal persons the mechanical adaptations
are sustained during prolonged head-up tilting.
Study limitations
The study suffers from the obvious limitations
related to the non-invasive technique regarding
detection of the immediate changes that precede
achievement of a new steady state, measurement of
ventricular size, Doppler assessment of pulmonary
venous flow [22] and left ventricular end-diastolic
pressure [111. Superiority of the transoesophageal
approach for records of the pulmonary venous
drainage is well known [23] and measurements with
the transthoracic method are generally obtained at
the level of the right superior vein. This may not be
representative of left-sided vessels or of vessels
draining the lower lobes, particularly in our subjects
who lay in a slight left lateral decubitus position [lo,
221. To obviate these difficulties as much as possible,
we always used a transducer position and manipulated the ultrasound beam so as to record the maximal right ventricular dimension [12] and to obtain
an interrogation as parallel as possible to the direction of flow. The differences between the duration
of pulmonary flow reversal during atrial contraction
and the duration of the mitral A wave might not be
strictly applicable to the dynamics presented herein,
because it has been validated in between-patient and
not in within-patient studies as an index of left ventricular end-diastolic pressure; the concept that a
negative time duration can be used to estimate left
ventricular end-diastolic pressure has also not been
validated.
Another appropriate consideration is that peripheral muscle tension, a major factor in compensation for changes in posture, was eliminated by
instructing the subjects not to strain muscles during
tilting; it is unknown how it will affect the results
and the in vivo integrated system.
20
M. Guazzi et al.
Conclusions
Despite these limitations, the study proves, we
believe, that reverse redistribution of blood from the
lung pool is essential for the maintenance of left
ventriklar filling during head-up tilting. This effect
is mediated through an improvement from supine
position of compliance within the pericardial space
and reduction of intraventricular pressure, as a consequence of a decrease in total pericardial volume
and constraint.
ACKNOWLEDGMENT
This study was supported in part by a grant from
the National Research Council, and the Monzino
Foundation, Milan, Italy.
REFERENCES
I. Kapoor WN, Smirh MA, Miller NL. Upright tilt testing in evaluating syncope: a
comprehensive literature review. Am J Med 1994; 97: 78-88.
2. Bora C, Wieling W, Brederode VJFM, Hond 4Rijk DLG, Dunnin J. Mechanisms
of initial heart rate response to postural change. Am J Physiol 1982; 243
H676-81.
3. Hainsworth R, Al-Shamma YMH. Cardiovascular responses to upright tilting in
healthy rubiects Clin Sci 1988 74: 17-22.
4. Guazri M. Wed-up tilt': tes: cvocativo ad impiego clinic0 e sperimentale.
Cardiologia 1995: 4U: 129-35.
5. Sjostrand T. The regulation of the blood distribution in man. Acta Phyrioi Scand
1952; 15: 3 12-27,
6. Lagerlof H, Eiiasch H, Werko L, Berglund E. Orthostatic changes of the
pu!monary and periphcral circulation in man: a preliminary report. Scand J Clin
Lab Invest 1951; 3: 85-91.
7. Weissler AM, Leonard JJ, Warner JV. Effect of posture and atropine on the
cardiac outpilt. J Clin Invest 1957; 3 6 1656-62.
8. Bove AA, Santamore WP. Ventricular interdependence. Prog Cardiovasc Dis
I98 I ; 23: 365-87.
9. Guazzi M, Pepi M, Maltagliati A, Celeste F, Muratori M, Tamborini G. How the
two sides of the heart adapt to graded impedance to venous return with head-up
tilting. J Am Coll Cardiol 1995; 2 6 1732-40.
10. Tanabe K, Yoshitomi H, Oyake N. hanuma T, Ohta T, lshibashi Y, Shimada T,
Morioka S, Moriyama K. Effects of supine and lateral recumbent positions on
pulmonary venous flow in healthy subjects evaluated by transesophageal Doppler
echocardiography.J Am Coll Cardiol 1994; 2 4 1552-7.
I I. Roussvoll 0, Hatle LV. Pulmonary venous flow velocities recsrded by
transthoracic Doppler ultrasound relation to left ventricular diastolic pressures.
JAm Coll Cardiol 1993; 21: 1687-96.
12. Bommer W, Weinert L, Neumann A, Nee( J, Mason DT, De Maria A.
Determination of right atrial and right ventricular size by hvo-dimensional
echocardiography. Circulation 1979; 6 0 9 I - 100.
13. lhlen H, Amlie JP. Dale J, Forfang K, Nitter-HaugeS, OtterstadJE, Simonsen S,
Myhre E. Determination of cardiac output by doppler echocardiography. Br
Heart J 1984; 5 I: 54-60.
14. Belocif S, Takata M, Shimada M, RobothamJL. Influence of pericardial constraint
on atrioventricular interactions.Am J Physiol 1992; 263: H125-34.
15. Gilbert JC, Glantz SA. Determinants of left ventricular filling and of the diastolic
pressure-volume relation. Circ Res 1989; 64: 827-52.
16. Keren G, Sherez J, Megidish R Levitt B, Lamiado S. Pulmonary venous flow
pattern: its relationshipto cardiac dynamics. Circulation 1985; 7 I : I 105- 12.
17. Moon MR, ingels NB, Daughters GT, Stinson EB, Hansen DE. Miller C.
Alterations in left ventricular twist mechanisms with inotropic stimulation and
volume loading in human subjects. Circulation 1994; 89 142-50.
18. Masuyama T, Lee JM. Nagano R, Nariyama K, Yamamoto K, Naito J.Mano T,
Kondo H, Hori M, Kamada T. Doppler echocardiographic pulmonary venous
flow-velocitypattern for assessment of the hemodynamic profile in acute
congestive heart failure. Am Heart J 1995; I 2 9 107- 13,
19. Sneddon IF, Counihan Pj, Bashir Y, Haywood GA. Ward DE, Camm J.
Assessment of autonomic function in patients with neurally mediated
cardiopulmonary baroreceptor responses to graded orthostatic stress. J Am Coll
Cardiol 1993; 21: 11934.
20. Abboud FM. Neurocardiogenic syncope. N Engl J Med 1993; 328: I 117-20.
2 I. D i c l h o n CJ.Fainting precipitated by collapse-firingof venous baroreceptors.
Lancet 1993; 3 4 2 970-2.
22. Pye M, Pringle SD, Cobbc SM. Reference values and reproducibilityof Doppler
echocardiography in the assessment of the tricuspid valve and right ventricular
diastolic function in normal subjects. Am J Cardiol 1991; 67: 269-73.
23. Castello R, Pearson PS,Lenzen P, Labovitz A], Evaluation of pulmcnsry venous
flow by transerophgsal echocardiography in subjects with a normal heart:
comparison with transthoracic echocardiography. J Am Coll Cardiol I99 I ; I 8
65-71.