Vol. 279, Issue 4, H1796-H1803, October 2000
Methods for assessing hepatic distending pressure and changes
in hepatic capacitance in pigs
Harald
Kjekshus1,
Cecilie
Risoe2,
Tim
Scholz1, and
Otto A.
Smiseth1
1 Institute for Surgical Research, The National Hospital,
University of Oslo, N-0027, and 2 Department of Cardiology,
Ullevaal Hospital, 0407 Oslo, Norway
 |
ABSTRACT |
The equilibrium pressure obtained
during simultaneous occlusion of hepatic vascular inflow and outflow
was taken as the reference estimate of hepatic vascular distending
pressure (Phd). Phd at baseline was 1.1 ± 0.2 (mean ± SE) mmHg higher than hepatic vein pressure
(Phv) and 0.7 ± 0.3 mmHg lower than portal vein
pressure (Ppv). Norepinephrine (NE) infusion increased
Phd by 1.5 ± 0.5 mmHg and Ppv by 3.7 ± 0.6 mmHg but did not significantly increase Phv. Hepatic
lobar vein pressure (Phlv) measured by a micromanometer tipped 2-Fr catheter closely resembled Phd both at baseline
and during NE-infusion. Dynamic pressure-volume (PV) curves
were constructed from continuous measurements of Phv and
hepatic blood volume increases (estimated by sonomicrometry) during
brief occlusions of hepatic vascular outflow and compared with static
PV curves constructed from Phd determinations at five
different hepatic volumes. Estimates of hepatic vascular compliance and
changes in unstressed blood volume from the two methods were in close
agreement with hepatic compliance averaging 32 ± 2 ml · mmHg
1 · kg liver1. NE
infusion reduced unstressed blood volume by 110 ± 38 ml/kg liver
but did not alter compliance. In conclusion, Phlv reflects hepatic distending pressure, and the construction of dynamic PV curves
is a fast and valid method for assessing hepatic compliance and changes
in unstressed blood volume.
regional circulation; compliance; unstressed blood volume; resistance; norepinephrine
| |
INTRODUCTION |
THE SPLANCHNIC
REGION represents the largest blood volume reservoir in the body
(7, 23, 29) and converges in a common vascular outflow
path through the liver. The liver is highly compliant and contributes
to circulatory homeostasis and regulation of cardiac filling by both
active and passive mobilization and storage of substantial amounts of
blood in response to physiological and pharmacological stimuli
(7, 24, 25). Construction of hepatic pressure-volume (PV)
curves facilitates the distinction between blood volume responses due
to active changes in the hepatic capacitance vessel tone and passive
alterations due to changes in systemic pressures and flows that affect
the liver (10). An essential key in assessing hepatic
vascular compliance and unstressed volume is a representative estimate
of the distending pressure in the main blood volume
compartment. Reports indicate that 50-80% of the hepatic
blood volume is localized in the sinusoids (4, 17, 21,
22). Thus several catheterization techniques have been applied
to estimate sinusoidal pressure. However, the initial values reported
of sinusoidal pressure relative to the portal (Ppv) and
hepatic vein (Phv) pressures were divergent with a
prevailing notion that sinusoidal pressure was close to Ppv
and that hepatic resistance was primarily postsinusoidal (5, 12,
14, 18, 20). In recent years direct measurements of
microvascular pressures with servo-null micropipettes (3, 16,
27) have clearly refuted this notion, showing sinusoidal
pressure to lie somewhat in the middle between Ppv and
Phv under baseline conditions. However, the use of the
micropipette technique is limited by its highly invasive nature,
requiring extensive stabilization of the liver to maintain the position
of the pipettes. An alternative way of estimating sinusoidal pressure
is to measure the pressure equilibrium during complete circulatory
isolation of the liver, a mean hepatic distending pressure
(Phd). The technique of zero-flow pressure equilibrium was
first defined for the entire circulation as the mean circulatory
filling pressure (9, 26) but has also been used to estimate splanchnic distending pressure in the dog
(19). Although anatomic localization is lost with this
technique, Shibamoto et al. (28) showed that
Phd (termed "triple occlusion pressure" in their paper)
closely resembled sinusoidal or capillary pressure in isolated perfused
livers. However, this approach requires the determination of several
static zero-flow points at different blood volumes in order to
construct PV curves. This is time consuming and adds extra scatter to
the data due to the prolonged recording period. We have previously
shown that reproducible capacitance data can be calculated from dynamic
PV curves constructed by plotting the simultaneous pressure and volume
increases observed during brief occlusions of hepatic vascular outflow
(10). Occlusion of hepatic vascular outflow is a rapid
method of assessing hepatic PV relations and can be performed
repeatedly during the course of an experiment without apparent
long-term effects on hepatic PV relations or the hemodynamic status of
the animal. In situations where transhepatic resistance is low or
normal, outflow occlusion will rapidly (within a fraction of a second)
induce a near equilibrium between Ppv portal and
Phv, ensuring that both pressures are fair estimates of the
operating distending pressure. However, we have shown that during
norepinephrine (NE) infusion, a large pressure gradient exists between
the portal and the hepatic vein during occlusion of hepatic vascular
outflow (10), and there is uncertainty as to which
pressure best reflects the operating distending pressure.
Thus the goals of the present study were the following: 1)
to determine which of the hepatic inflow and outflow pressures best represents the distending pressure of the hepatic capacitance vessels; and 2) to investigate if changes in hepatic
vascular compliance and unstressed volume calculated from dynamic PV
curves differed from those derived from static PV points.
| |
METHODS |
Animal preparation and instrumentation.
Six Norwegian farm pigs of either sex weighing 20-24 kg were
premedicated with ketamine 400 mg im (Ketalar, Warner Lambert, Morris
Plains, NJ) and azaperone 60 mg im (Stresnil, Janssen-Cilag Pharma,
Vienna, Austria) 20 min before induction of anesthesia with a bolus
dose of pentobarbital sodium (250 mg iv) and petidinclorid (100 mg iv,
Petidin, Nycomed Pharma, Oslo, Norway), followed by laryngeal
intubation and ventilation (Siemens 900B, Solna, Sweden). Anesthesia
was maintained by a continuous intravenous infusion of pentobarbital in
Ringer acetate (1 mg/ml at a rate of ~8 ml kg body
wt1 · h1, Pharmacia & Upjohn,
Stockholm, Sweden) combined with an intravenous infusion of
petidinclorid in Ringer acetate (0.1 mg/ml at a rate of 12 ml · kg body wt1 · h1).
After surgery was completed, the pentobarbital infusion was reduced to
4-6 ml · kg body
wt1 · h1.
A midline laparatomy was performed. For measurements of
Ppv, an 8-Fr feeding tube (Pharma-Plast; Lynge, Denmark)
was advanced through a small pancreatic tributary into the portal vein
with the catheter tip palpable 1 cm from the liver. For
measurements of wedged portal vein pressure (Ppvw), a 5-Fr
balloon catheter (model 93-132-5F, Edwards Swan-Ganz
catheter, Baxter Healthcare, Irvine, CA) was inserted through a
mesenteric vein and advanced with the balloon deflated until it wedged
in a small branch within the liver. The balloon was then inflated with
0.2 ml of saline and a wedged position was confirmed by contrast
injection (Omnipaque; Nycomed) in the upstream Ppv catheter
with no visual leakage of contrast past the tip of the Ppvw
catheter. For measurements of Phv and hepatic lobar vein
pressure (Phlv), an introducer catheter (8-Fr, RS*C80N10NR;
Terumo, Tokyo, Japan) was inserted in the right external jugular vein
and secured in place. A 6.5-Fr catheter (model P6.5, outer diameter 2.3 mm; William Cook Europe, Bjaeverskov, Denmark) was inserted through the
introducer to a hepatic vein under fluoroscopic guidance. The catheter
was inserted gently until resistance was felt, and a wedged position
was verified by its pressure equaling that of the portal vein
(5). Before insertion of the 6.5-Fr catheter, a 2-Fr
micromanometer-tipped catheter (model SPR-407, outer diameter 0.7 mm;
Millar Instruments, Houston, TX) had been prepositioned within the
6.5-Fr catheter with its tip residing at the distal opening. Guided by
markings on the 2-Fr catheter and the 6.5-F catheter, the 6.5-Fr
catheter was slowly withdrawn from the wedged position while the 2-Fr
catheter was simultaneously fed into the 6.5-Fr catheter at the same
rate, leaving the 2-Fr catheter steadfast in a lobar hepatic vein.
Typically, the 6.5-Fr catheter was withdrawn 5-6 cm under
fluoroscopic guidance and recording of pressures. Final positions of
the catheters were verified by fluoroscopy and contrast injection in
the 6.5-Fr catheter ensuring that its tip resided freely floating in a
large hepatic vein 1-2 cm from its opening into the caval vein.
Cardiac oscillations could be identified in the Phlv and
Phv pressure tracings (Fig. 1) indicating nonwedged positions. The
final positions of the hepatic catheters are illustrated schematically
in Fig. 2. A separate catheter was
inserted into a mesenteric vein and advanced toward the beginning of
the portal vein for drug infusions. For measurements of systemic
pressures and cardiac output, catheters were placed in the pulmonary
artery (Edwards 7-Fr Swan-Ganz thermodilution catheter; Baxter) and
abdominal aorta (6-Fr angiographic catheter; Cordis, Miami, FL). All
catheters were connected to AE840 pressure transducers (SensoNor,
Horten, Norway), and reference zero level was set at the level of the
right atrium.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1.
Recording of hepatic pressures during the withdrawal of
the 6.5-Fr fluid-filled hepatic vein (Phv) catheter from
the wedged position. Note that all the pressures overlap in the wedged
position before withdrawal. When the Phv catheter was
withdrawn 2 cm, the pressure recorded by the upstream 2-Fr
micromanometer tip catheter (hepatic lobar vein pressure,
Phlv) was similar to the downstream Phv,
providing no evidence for the existence of a sphincter in the proximal
hepatic veins. Also, the physiological oscillations of the
Phlv trace indicates a nonwedged position of the
Phlv catheter. The Phv catheter was gradually
withdrawn a further 4 cm to a free-floating position ~1 cm from the
hepatic vein outlet. Note the gradual decrease in Phv with
withdrawal while the Phlv remains stable. The respirator
was then temporarily disconnected and hepatic vascular in- and outflow
occluded (open arrow) to achieve complete circulatory isolation of the
liver. When a stable pressure-equilibrium (solid arrow) between the
fluid-filled portal vein (Ppv) and Phv
catheters was obtained, any zero drift of the Phlv catheter
was checked for and if necessary corrected.
|
|
View larger version (66K):
[in this window]
[in a new window]
|
Fig. 2.
Schematic illustration of the individual positions of the
hepatic catheters. Ppv, portal vein catheter;
Ppvw, portal vein wedge catheter; Phlv, hepatic
lobar vein catheter; Phv, hepatic vein catheter.
|
|
Liver blood flow was measured with ultrasound transit time flow probes
on the hepatic artery and portal vein (4SB and 12SB, respectively,
connected to a T207 flowmeter, Transonics Systems, Ithaca, NY). Hepatic
inflow could be interrupted by inflation of vascular occluders mounted
on the hepatic artery and portal vein, and hepatic outflow could be
occluded by inflation of a balloon catheter positioned in the
intrahepatic portion of the inferior vena cava as previously described
(10). Liver thickness was measured by gluing ultrasonic
transducers (ED3-2 connected to Sonomicrometer 120; Triton
Technology, San Diego, CA) to the ventral and dorsal surfaces of the
left lateral and the medial liver lobe as previously described
(10). All animal handling, experiments, and proceedings
were approved by the local laboratory animal science specialist under
the surveillance and registration of the Norwegian Experimental Animal
Board, conforming to the National Institutes of Health Guide for
the Care and Use of Laboratory Animals (National Research Council,
Washington, DC, 1996).
Calculations.
Total hepatic resistance was calculated as the difference between
Ppv and Phv divided by the sum of portal vein
and hepatic arterial blood flow. Portal venous resistance
(Rpv) was calculated as the difference between
Ppv and Phd divided by portal venous blood
flow. Hepatic arterial resistance (Rha) was
calculated as the difference between mean arterial pressure and
Phd divided by hepatic arterial blood flow. Hepatic lobar
venous resistance (Rhlv) was calculated as the
difference between Phd and Phlv divided by the
sum of portal vein and hepatic arterial blood flow. Hepatic venous
resistance (Rhv) was calculated as the
difference between Phlv and Phv divided by the
sum of portal vein and hepatic arterial blood flow.
Alterations in hepatic vascular volume were assessed through
measurements of liver thickness by sonomicrometry. Changes in liver
thickness by sonomicrometry were converted to volume changes by
plotting each liver thickness as a function of the integral of the
hepatic inflow during complete outflow occlusion. The integral of
hepatic inflow will equal the change in volume when hepatic outflow is
occluded. The slope of the volume-thickness curves thus obtained can be
taken as the calibration factor to convert changes in liver thickness
to changes in volume. The calibration procedure and reproducibility of
sonomicrometric volume estimates has been presented in detail
previously (10). The procedure does not allow for direct
measurement of total blood volume, only measurements of volume changes
relative to baseline. Instead of reporting volume changes as a
percentage of baseline, we made the assumption that total hepatic blood
volume equaled 35 ml/100 g liver wt (7) at baseline, and
we reported the measured volume changes as different absolute volumes
based on this assumption. The sonometrically based estimates of total
hepatic blood volumes were then used to construct PV curves. The slope
and intercept of the PV curves were calculated by linear regression
analysis using the least-squares method and used as estimates for
hepatic vascular compliance and hepatic unstressed blood volume, respectively.
Experimental protocol.
After surgery was completed, the pigs were allowed 30 min for
stabilization. The baseline recordings consisted of initial cardiac
output (CO) measurement by thermodilution (mean of 3 injections of 10 ml of ice-cold 5% glucose). The respirator was then
disconnected, and a 10-s recording of stable pressures and volume was
performed during free flow, followed by occlusion of hepatic vascular
outflow and a 15-s recording of the resulting continuous pressure and blood volume increases (see Fig. 3). This
15-s recording was used to construct the dynamic PV curves as
previously described (10). The respirator was reconnected,
and a 5-min stabilization period was allowed.
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 3.
Assessment of mean hepatic distending pressure (Phd)
and static pressure-volume (PV) points in 1 pig. Assembly of
consecutive recordings performed with 5-min intervals at baseline
(A) and during norepinephrine (NE) infusion (B).
Each recording consists of an initial period of free flow followed by
occlusion of hepatic vascular outflow (gray vertical arrow) and
subsequent occlusion of hepatic vascular inflow (solid vertical arrow)
after 0- (simultaneous in- and outflow occlusion), 2-, 5-, 10-, and
15-s duration of outflow occlusion, respectively. The 15-s duration of
outflow occlusion recording was also used for construction of the
dynamic PV curve. Note the rapid equilibration between Phlv
and Phv after occlusion of outflow, while a significant
gradient persists toward Ppv, most marked during NE
infusion. With subsequent occlusion of inflow, Ppv rapidly
approximates the other pressures and this initial equilibration point
(open arrows) was taken as the reference estimate of Phd
used to assess pressure gradients. Thereafter the pressures gradually
decline in unison and the point where they leveled off (solid arrows)
was used for construction of static PV curves. The dimensional traces
(D-L1 and D-L2, referring to the thickness measurements) remained
stable during this latter phase indicating an unaltered hepatic blood
volume, which suggests a delayed compliance of the hepatic
vasculature.
|
|
To assess Phd, circulatory isolation of the liver was then
performed by simultaneously occluding hepatic inflow and outflow during
recording of hepatic pressures and volume (Fig. 3). The Phd estimate thus obtained was corrected for any
alterations in hepatic volume caused by the occlusion procedure and
used as reference when evaluating the continuous pressure measurements
during free flow. To construct static PV curves, additional PV data at
increasing liver volumes were obtained by occluding hepatic inflow
after 2-, 5-, 10-, and 15-s duration of hepatic outflow occlusion. The individual PV data were obtained at 5-min intervals to allow
hemodynamic restitution between measurements. A representative collage
of hepatic pressures and thickness recordings from one pig is shown in
Fig. 3. The collage clearly demonstrates that circulatory isolation of
the liver obtained by occlusion of hepatic inflow during ongoing outflow occlusion induces a rapid equilibration of hepatic pressures. The pressures then gradually declined in unison before leveling off.
The hepatic pressure and volume estimate used in the construction of
static PV curves were taken at the point where the hepatic pressures
leveled off (Fig. 3, solid arrows). To assess which pressure best
reflected Phd during the outflow occlusion procedure used
for construction of dynamic PV curves, the hepatic pressures recorded
immediately before inflow occlusion were compared with the initial
equilibration point where Ppv equaled Phv (Fig.
3, open arrows).
The animals were then permitted 15 min of rest before the start of an
intraportal infusion of 0.5 µg · kg body
wt1 · min1 of NE. After a 5-min
period of stabile infusion, we measured the same sequence of
measurements performed as described during baseline. At the end of the
experiments, the animals were euthanized with a lethal intravenous dose
of pentobarbital. All measurements except CO determinations were done
with the respirator temporarily disconnected and recorded on a Gould
ES2000 (Gould Instrument Systems, Cleveland, OH), digitized, and
subsequently analyzed off-line using CVSOFT (version 2.2, Odessa
Computer Systems, Calgary, Canada). Paper recordings were obtained
simultaneously for backup and verification of the digitized data.
Statistical analysis.
Data are reported as means ± SE unless otherwise specified. A
two-tailed paired t-test was used to compare baseline and NE values and to compare individual static and dynamic capacitance data.
Bland-Altman (2) analyses were used to further assess the
agreement between the two methods and to explore any systematic deviations between the methods. Departure from zero of the
pressure-gradients between Phd and the other hepatic
pressures was tested using a two-tailed one-sample t-test.
Multiple regression analyses with dummy variables to correct for
between-animal variations were used to compare the actual PV curves and
predict mean PV curves with confidence intervals from the two methods.
A P value < 0.05 was considered statistically
significant. All statistical analyses were performed using the SPSS
statistical software (version 8.0, SPSS, Chicago, IL).
| |
RESULTS |
Assessment of Phd and resistance during free flow.
Compared with baseline, NE induced a 46% increase in Ppv
(P < 0.01), a 25% increase in Phlv
(P = 0.03), a 20% increase in Phd
(P = 0.02), and a 15% increase in Ppvw
[not significant (NS)], whereas Phv values were unchanged
(absolute values in Table 1). The
Ppv to Phd gradient increased from 0.7 ± 0.3 mmHg at baseline to 2.9 ± 0.7 mmHg during NE
(P < 0.001). In comparison, the
Phd-to-Phv gradient increased from 1.1 ± 0.2 to 2.1 ± 0.5 mmHg (NS). As illustrated in Fig.
4 and Table 1, both Ppv and
Phv differed significantly from Phd.
Ppvw and Phlv, on the other hand, yielded
similar mean pressures as Phd both at baseline and during
NE infusion with standard errors of the estimates < 0.5 mmHg.
Furthermore, Bland-Altman analyses showed good agreement
and no systematic deviations in comparing Phd with
Phlv (Fig. 5) and comparing
Phd with Ppvw (data not shown). Portal venous
and hepatic arterial flows did not change significantly with NE. Thus
NE infusion induced a 2.7-fold (P < 0.01) increase in
total hepatic resistance, a 5.2-fold (P = 0.04) increase in
Rpv, and a 2.7-fold increase (P < 0.03) in Rhv. Rha
increased 1.4-fold (NS). Rhlv was very low and
did not change with NE infusion.
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 4.
Difference between the individual pressures and the mean
Phd during free flow. *P < 0.05 vs. mean
hepatic distending pressure, § P < 0.05 vs.
baseline.
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 5.
Bland-Altman plot showing the agreement between mean
Phd measured during circulatory isolation of the liver and
Phlv measured during free hepatic flow.  ,
Baseline;  , NE infusion. CI, confidence interval.
|
|
Assessment of Phd during outflow occlusion.
With outflow occlusion during baseline, there was a rapid equilibration
between Phv and Phlv, and both equaled
Phd. Furthermore, only a modest gradient existed between
Ppv and the other pressures (Fig.
6A). With outflow occlusion
during NE infusion (Fig. 6B), there was also a rapid
equilibration between Phlv and Phv, and both
compared well with Phd with standard errors of the
estimates < 0.3 and 0.5 mmHg, respectively. However,
Ppv differed significantly from the other pressures at all
time points. Ppvw measurements during outflow occlusion
were unreliable because outflow occlusion often caused the catheter to
dislodge from its wedged position.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6.
Hepatic pressures during vascular outflow occlusion and
isolation of the liver at baseline (A) and during NE
infusion (B). The curves depict the mean pressures from the
Ppv, Phlv, and the Phv during free
flow (two first points) and later during increasing duration of outflow
occlusion. Circulatory isolation of the liver was subsequently achieved
at each point by occlusion of hepatic vascular inflow. The resulting
equilibration pressure, i.e., the mean Phd is shown as
individual points with 95% CI. Note the rapid equilibration between
Phlv and Phv during outflow occlusion
both at baseline and during NE infusion and their close
resemblance with Phd. In contrast, Ppv differ
significantly from Phd at all time points.
|
|
The pressure equilibration showed a consistent pattern during occlusion
of hepatic vascular outflow and inflow (Fig. 3). With outflow
occlusion, the Phv and Phlv equilibrated almost
instantaneously both at baseline and during NE infusion, whereas a
significant gradient persisted toward the Ppv. With
subsequent occlusion of hepatic inflow, Ppv fell rapidly
toward the level of Phlv and Phv (initial
equilibration point, used for evaluating the pressure gradients; open
arrows in Fig. 3). Thereafter the pressures gradually declined in
unison 0-2 mmHg (median 0.7 mmHg) and started to level off at the
point taken as the Phd when calculating the capacitance data (solid arrows in Fig. 3).
Assessment of hepatic compliance and unstressed volume.
NE infusion caused a parallel shift of the PV curve versus baseline,
indicated by an unaltered hepatic compliance of 31.5 ± 1.8 versus
31.2 ± 1.3 ml · mmHg1 · kg
liver1 (NS), a reduced unstressed blood volume of
3 ± 45 versus 107 ± 10 ml/kg liver (P < 0.05), and a reduced capacity at 7 mmHg of 232 ± 26 versus
358 ± 10 ml/kg liver (P < 0.01).
There were no significant differences among hepatic compliance,
unstressed blood volume, and capacity calculated from dynamic and
static PV curves (Table 2). Bland-Altman
analyses showed good agreement and no systematic deviations between the
two methods in predicting hepatic vascular compliance (Fig.
7A) and unstressed blood
volume (Fig. 7B) with mean differences of 0.09
ml · mmHg1 · kg body wt1
and 0.08 ml/kg body wt, respectively. We further compared the PV curves
from the two methods using multiple regression analysis with dummy
variables to correct for between-animal variation. Again, there was a
good agreement between dynamic and static PV measurements with both
methods clearly demonstrating the NE-induced parallel shifting of the
PV curve toward the pressure axis (Fig. 8, A and B,
respectively).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 7.
Bland-Altman plots showing the agreement between dynamic
and static PV-measurements. A: hepatic vascular compliance,
B: hepatic unstressed blood volume. Baseline, , NE infusion.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 8.
PV
curves from dynamic (A) and static (B)
PV-measurements. Mean PV curves with 95% CI at baseline (solid line)
and NE infusion (gray line) based on multiple regression analysis with
dummy variables to correct for between-animal variations. Equations
with r2 values from the respective multiple
regression analyses are presented on the corresponding figures.
|
|
| |
DISCUSSION |
Hepatic distending pressure.
We have used the equlibration pressure during complete circulatory
isolation of the liver as a reference method for estimating the mean
Phd. The sinusoids appear to represent the main hepatic blood volume compartment (4, 21), and one can therefore
expect the operating distending pressure within the sinusoids to be a major determinant of the Phd. This is supported by the
Shibamoto et al. study (28), which concluded that
Phd (called the "triple occlusion pressure") closely
resembled sinusoidal pressure. However, the determination of
Phd require extensive instrumentation of the animal to
enable complete circulatory isolation of the liver. Furthermore, only
intermittent determinations can be performed. In the present study we
have therefore compared this reference estimate with continuous
pressure measurements during free flow through the liver.
Phlv was overall the best estimate of Phd
[95% confidence interval (CI) for the difference: 0.6 to
0.3]. Ppvw also showed a good agreement with
Phd during free flow, but its measurement during outflow
occlusion was unreliable. Ppv and Phv differed
significantly from Phd during free flow. Our findings that
Phlv was a good estimate of Phd both at
baseline and during NE infusion is in agreement with the micropipette
study of Maass-Moreno and Rothe (16) observing only minor
differences between sinusoidal and hepatic venular pressures. However,
it contrasts somewhat with the observations of Lautt and colleagues
(12, 14), who found only minor differences between
Ppv and Phlv in cats and dogs. This may relate
to the size of the catheter relative to the vessel diameter as
described by Maass-Moreno and Rothe (15). By the aid of a
hydraulic and a mathematical model, they found that the use of
catheters with an outer diameter exceeding 60% of the inner diameter
of the vessel result in significant overestimation of Phlv.
We measured Phlv with a micromanometer-tip catheter, which
has its pressure sensor located on the side of the catheter tip with a
diameter of only 30% of the 6.5-Fr catheter used to introduce it, thus
minimizing the chance of a wedged position in the hepatic lobar vein.
The resulting Phlv measurements relative to Ppv
and Phv in the present study compare well to the values of
hepatic venular pressures reported from micropipette studies in other
animals both at baseline and during NE infusion (3, 27).
Positioning of the Phlv catheter is easily performed under fluoroscopic guidance and does not require a laparotomy. This technique
therefore represents a feasible way of estimating sinusoidal pressure
in animals and human subjects.
Hepatic resistance.
We found an overall increase in hepatic resistance with NE
infusion. Again, taking Phd as an estimate of
sinusoidal pressure, we calculated the resistances of the various
hepatic vascular segments. We found a relative increase in
Rpv that was nearly twice the increase
in Rhv during NE infusion. This is
consistent with the study by Shibamoto et al. (28), and
with the findings from micropipette studies (3, 27),
lending credence to the presented methodology. The
Rhlv was low at baseline and did not increase
with NE infusion. Not having an exact anatomical localization of
Phd makes an interpretation of Rhlv
difficult. Most likely it reflects the resistance in only a distal part
of the sinusoidal circulation. However, our findings are not compatible
with a major resistance site at the sinusoidal outlet into the hepatic
lobar veins. Furthermore, contrary to studies in other species
(11-14), we could not detect any step-wise pressure
drops suggestive of localized sphincters in the hepatic veins during
withdrawal of the Phv catheter from its wedged position,
neither at baseline (Fig. 1) nor during NE infusion (data not shown).
Finally, the observed mechanism of action of NE with a generalized
vasoconstriction appearing most prominent in the portal vein and
upstream sinusoids is perfectly suited for an agent intended to
mobilize blood from the liver.
Hepatic compliance and unstressed volume.
As illustrated in Fig. 3, the estimation of a static PV point requires
an initial occlusion of hepatic outflow of varying duration before
occlusion of inflow. With inflow occlusion there is an initial rapid
equilibration of the hepatic pressures after which the pressures
decline in unison over a period of 4-8 s before stabilizing.
During this time the hepatic dimensions remain stable suggesting a
delayed compliance, i.e., the relaxation of a vessel requires a finite
time for completion in response to rapid increases in pressure. Thus
dynamic PV curves would tend to overestimate distending pressure with
the slope of the PV curves being less steep, i.e., have a lower
compliance than the static PV curves. On the other hand, the longer
occlusion time needed for pressure equilibration and stabilization
after circulatory isolation could activate cardiovascular reflexes to a
larger degree than occlusion of outflow alone. Increased reflex
activity would be expected to contract hepatic capacitance vessels and
thus increase Phd. The compliance estimates from the two
methods were essentially similar with a mean overall difference of 0.1 ml · mmHg1 · kg body wt1
(NS). This suggests that the effects of delayed compliance on dynamic
PV curves and the effect of augmented reflex activation on static PV
curves, if present to any degree, counteract each other. Furthermore,
there were no significant differences between the two methods in
estimating hepatic unstressed volume and hepatic capacity, nor were
there any systematic differences between the two methods related to NE infusion.
Baseline compliance and unstressed volume estimates in the present
study (31 and 106 ml · mmHg1 · kg
liver1, respectively) compare well with previous reports
(1, 7, 8) with compliance estimates ranging from 26 to 31 ml · mmHg1 · kg liver1 and
unstressed volume estimates ranging from 104 to 150 ml/kg liver.
Greenway et al. (8) also reported similar effects of NE
infusion with unstressed volume reduced to 15 ml/kg liver (3
ml · mmHg1 · kg liver1 in
the present study). We therefore conclude that dynamic and static PV
curves yield comparable and valid estimates of changes in hepatic
compliance and unstressed volume. However, we find the dynamic approach
preferable due to the speed and ease with which a PV curve can be
constructed. The recording of a sufficient number of PV points to
construct a reliable static PV curve is time consuming because
hemodynamic restitution must be allowed between the determination of
each point. In the present study, each dynamic PV curve was acquired
from a single 15-s period of hepatic outflow occlusion, whereas each
5-point static PV curve took about 25 min to obtain (allowing 5 min of
restitution between each individual measurement point).
Limitations of the study.
Sonomicrometry is a reproducible method to estimate changes in hepatic
vascular volume in the pig model. However, the method is limited by not
being able to measure total hepatic vascular volume. To ease data
analysis and presentation, we have chosen to assume a standardized
value for total hepatic vascular volume at baseline. Thus the absolute
values presented for total hepatic volume, unstressed volume, and
capacity may deviate somewhat from the true physiological value.
However, the relative changes in these parameters are accurate and
constitute the basis for our conclusions.
The PV curve intercepts were taken as estimates of hepatic unstressed
volume. The values for unstressed volume are thus based on the
assumption of a linear PV relationship all the way down to zero
pressure. A nonlinear PV relationship may well exist in the
low-pressure range and could explain the finding of negative unstressed
volumes during NE infusion by us and others (8), which is
a physiological impossibility. However, because the PV curve is near
linear in the operating pressure range (10), the PV-curve
intercept is a valid parameter characterizing the PV relationship in
the operating pressure range, even though it may not accurately reflect
the true unstressed blood volume, i.e., the volume contained in the
vasculature at zero distending pressure.
In summary, Phlv measured by a 2-Fr micromanometer-tipped
catheter closely resembled the equilibrium pressure during hepatic circulatory isolation. The observed pressures and the calculated distribution of hepatic resistance in our study are comparable to
previously published studies where hepatic microvascular pressures were
measured directly. We therefore conclude that the measurement of
Phlv is a simple and reliable way of estimating sinusoidal pressure in the porcine liver. Dynamic PV curves constructed from continuous measurements of Phv and hepatic volume during brief occlusions of hepatic vascular outflow were similar to static PV
curves. Estimates of hepatic compliance, unstressed blood volume, and
capacity from the two methods were in good agreement, with no
systematic differences. We therefore conclude that the construction of
dynamic PV curves is a fast, convenient, and valid way to quantify hepatic vascular compliance and estimate changes in unstressed volume.
| |
ACKNOWLEDGEMENTS |
The authors thank Roger Ødegård for technical assistance.
| |
FOOTNOTES |
H. Kjekshus was supported by a fellowship from the Norwegian Council
for Cardiovascular Diseases, and T. Scholz was supported by a
fellowship from the University of Oslo. This study was also funded by
the Blix Family Fund for the Promotion of Science, Oslo.
Address for reprint requests and other correspondence: H. Kjekshus, Institute for Surgical Research, The National Hospital, N-0027 Oslo, Norway (E-mail:
harald.kjekshus{at}klinmed.uio.no).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 14 April 1999; accepted in final form 17 April
2000.
| |
REFERENCES |
1.
Bennett, TD,
and
Rothe CF.
Hepatic capacitance responses to changes in flow and hepatic venous pressure in dogs.
Am J Physiol Heart Circ Physiol
240:
H18-H28,
1981.
2.
Bland, JM,
and
Altman DG.
Statistical methods for assessing agreement between two methods of clinical measurement.
Lancet
1:
307-310,
1986[ISI][Medline].
3.
Bohlen, HG,
Maass-Moreno R,
and
Rothe CF.
Hepatic venular pressures of rats, dogs, and rabbits.
Am J Physiol Gastrointest Liver Physiol
261:
G539-G547,
1991[Abstract/Free Full Text].
4.
Conway, JG,
Popp JA,
and
Thurman RG.
Microcirculation in periportal and pericentral regions of lobule in perfused rat liver.
Am J Physiol Gastrointest Liver Physiol
249:
G449-G456,
1985.
5.
Friedman, EW,
and
Weiner RS.
Estimation of hepatic sinusoid pressure by means of venous catheters and estimation of portal pressure by hepatic vein catheterization.
Am J Physiol
165:
527-531,
1951.
6.
Goresky, CA,
and
Silverman M.
Effect of correction of catheter distortion on calculated liver sinusoidal volumes.
Am J Physiol
207:
883-892,
1964.
7.
Greenway, CV,
and
Lautt WW.
Hepatic circulation.
In: Handbook of Physiology. The Gastrointestinal System Motility and Circulation. Bethesda, MD: Am. Physiol. Soc, 1989: sect. 6, vol. I, chapt. 41, p. 1519-1564.
8.
Greenway, CV,
Seaman KL,
and
Innes IR.
Norepinephrine on venous compliance and unstressed volume in cat liver.
Am J Physiol Heart Circ Physiol
248:
H468-H476,
1985.
9.
Guyton, AC,
Lindsey AW,
and
Kaufman BN.
Effect of mean circulatory filling pressure and other periferal circulatory factors on cardiac output.
Am J Physiol
180:
463-468,
1955.
10.
Kjekshus, H,
Risoe C,
Scholz T,
and
Smiseth OA.
Regulation of hepatic blood volume-contributions from active and passive mechanisms during catecholamine and sodium nitroprusside infusion.
Circulation
96:
4415-4423,
1997[Abstract/Free Full Text].
11.
Lautt, WW,
Greenway CV,
and
Legare DJ.
Effect of hepatic nerves, norepinephrine, angiotensin, and elevated central venous pressure on postsinusoidal resistance sites and intrahepatic pressures in cats.
Microvasc Res
33:
50-61,
1987[ISI][Medline].
12.
Lautt, WW,
Greenway CV,
Legare DJ,
and
Weisman H.
Localization of intrahepatic portal vascular resistance.
Am J Physiol Gastrointest Liver Physiol
251:
G375-G381,
1986.
13.
Lautt, WW,
Legare DJ,
and
Greenway CV.
Effect of hepatic venous sphincter contraction on transmission of central venous pressure to lobar and portal pressure.
Can J Physiol Pharmacol
65:
2235-2243,
1987[ISI][Medline].
14.
Legare, DJ,
and
Lautt WW.
Hepatic venous resistance site in the dog: localization and validation of intrahepatic pressure measurements.
Can J Physiol Pharmacol
65:
352-359,
1987[ISI][Medline].
15.
Maass-Moreno, R,
and
Rothe CF.
Contribution of the large hepatic veins to postsinusoidal vascular resistance.
Am J Physiol Gastrointest Liver Physiol
262:
G14-G22,
1992[Abstract/Free Full Text].
16.
Maass-Moreno, R,
and
Rothe CF.
Distribution of pressure gradients along hepatic vasculature.
Am J Physiol Heart Circ Physiol
272:
H2826-H2832,
1997[Abstract/Free Full Text].
17.
Miller, DL,
Zanolli CS,
and
Gumucio JJ.
Quantitative morphology of the sinusoids of the hepatic acinus. Quantimet analysis of rat liver.
Gastroenterology
76:
965-969,
1979[ISI][Medline].
18.
Mitzner, W.
Hepatic outflow resistance, sinusoid pressure, and the vascular waterfall.
Am J Physiol
227:
513-519,
1974.
19.
Mitzner, W,
and
Goldberg H.
Effects of epinephrine on resisitive and compliant properties of the canine vasculature.
J Appl Physiol
39:
272-280,
1975[Abstract/Free Full Text].
20.
Nakata, K,
Leong GF,
and
Brauer RW.
Direct measurement of blood pressures in minute vessels of the liver.
Am J Physiol
199:
1181-1188,
1960.
21.
Ohara, N,
Schaffner T,
and
Reichen J.
Structure-function relationship in secondary biliary cirrhosis in the rat. Stereologic and hemodynamic characterization of a model.
J Hepatol
17:
155-162,
1993[ISI][Medline].
22.
Pang, KS,
Lee WF,
Cherry WF,
Yuen V,
Accaputo J,
Fayz S,
Schwab AJ,
and
Goresky CA.
Effects of perfusate flow rate on measured blood volume, disse space, intracellular water space, and drug extraction in the perfused rat liver preparation: characterization by the multiple indicator dilution technique.
J Pharmacokinet Biopharm
16:
595-632,
1988[ISI][Medline].
23.
Price, HL,
Deutsch S,
Marshall BE,
Stephen GW,
Behar MG,
and
Neufeld GR.
Hemodynamic and metabolic effects of hemorrhage in man, with particular reference to the splanchnic circulation.
Circ Res
18:
469-474,
1966[Abstract/Free Full Text].
24.
Rothe, CF.
Reflex control of veins and vascular capacitance.
Physiol Rev
63:
1281-1342,
1983[Free Full Text].
25.
Rothe, CF.
Physiology of venous return. An unappreciated boost to the heart.
Arch Intern Med
146:
977-982,
1986[Abstract].
26.
Rothe, CF.
Mean circulatory filling pressure: its meaning and measurement.
J Appl Physiol
74:
499-509,
1993[Abstract/Free Full Text].
27.
Rothe, CF,
and
Maass-Moreno R.
Hepatic venular resistance responses to norepinephrine, isoproterenol, adenosine, histamine, and ACh in rabbits.
Am J Physiol Heart Circ Physiol
274:
H777-H785,
1998[Abstract/Free Full Text].
28.
Shibamoto, T,
Wang HG,
Tanaka S,
and
Koyama S.
Hepatic capillary pressure is estimated using triple vascular occlusion method in isolated canine liver.
Am J Physiol Regulatory Integrative Comp Physiol
271:
R1130-R1141,
1996[Abstract/Free Full Text].
29.
Wade, OL,
Combes B,
Childs AW,
Wheeler HO,
Cournard A,
and
Bradley SE.
The effect of exercise on the splanchnic blood flow and splanchnic blood volume in normal man.
Clin Sci (Colch)
15:
457-463,
1956[Medline].
Am J Physiol Heart Circ Physiol 279(4):H1796-H1803
0363-6135/00 $5.00
Copyright © 2000 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
