| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Nephrology Unit, Department of Medicine, University of Rochester Medical Center, Rochester, New York 14642
 |
ABSTRACT |
|---|
 Top
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Fluid loss from
the peritoneal cavity to surrounding tissue varies directly with
intraperitoneal hydrostatic pressure
(Pip). According to Darcy's law
[Q = 
KA(dPif/dx)],
fluid flux (Q) across a
cross-sectional area (A) of tissue
will increase with an increase in either hydraulic conductivity
(K) or the interstitial fluid hydrostatic pressure gradient
(dPif/dx,
where x is distance). Previously, we
demonstrated that in the anterior abdominal muscle (AAM) of rats,
dPif/dx
increases by only 40%, whereas K
rises fivefold between Pip of 1.5 and 8 mmHg. Because K is a function of
interstitial volume (
if), we
hypothesized that perturbations of
Pip would change
Pif and expand the interstitium,
increasing if. To test this
hypothesis, we used dual-label quantitative autoradiography (QAR) to
measure extracellular fluid volume
(ec) and intravascular volume
(iv) in the AAM of rats
within the Pip range from
2.8 to +8 mmHg. if was
obtained by subtraction (ec iv).
dPif/dx
was measured with a micropipette and a servo-null system. Local
iv did not vary with
Pip and averaged 0.010 ± 0.002 ml/g, and if averaged 0.19 ± 0.01 ml/g at Pif 
1.2 mmHg.
However, if doubled between
Pif of 1.2 and 4.2 mmHg (from 0.20 ± 0.00 to 0.39 ± 0.01 ml/g, respectively) but did not increase
with further increases in Pif.
This nonlinear pressure-volume relationship does not explain the
fivefold increase in K with
Pip. Because the interstitial
matrix contributes to the interstitial resistance to fluid flow, and
because hyaluronan (HA) is the only component of the matrix that is not
anchored to the tissue, we hypothesized that the loss of interstitial
HA was responsible for the continued decrease in interstitial
resistance to fluid flow. We determined HA concentration in the rat AAM
and adjacent subcutaneous tissue (SC) at
Pip = 0 mmHg and after 2 h of
dialysis at constant Pip = 6 mmHg.
The HA content (normalized to dry weight) in the AAM was reduced from
487 ± 16 to 360 ± 27 µg/g dry tissue
(n = 4, P < 0.05) and increased from 528 ± 72 to 1,050 ± 136 mg/g dry tissue
(n = 4, P > 0.001) in the SC. We conclude
that the mechanisms responsible for the increase in
K with
Pip include expansion of the
interstitium, dilution of interstitial macromolecules, and washout from
the AAM to SC of interstitial macromolecules responsible for resistance
to fluid flow.
convection; hydraulic conductivity; compliance; peritoneal dialysis
| |
INTRODUCTION |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
STUDIES ON FLUID LOSS from the peritoneal cavity to surrounding tissues have demonstrated that the rate of peritoneal fluid loss to tissues bordering the peritoneal cavity varies directly with the intraperitoneal hydrostatic pressure (Pip) (10). According to Darcy's law
|
(1) |
if; the portion of the total
tissue volume outside cells and blood vessels that is available to
water). Hence, we hypothesized that changes in the local interstitial
hydrostatic pressure (Pif) may
change if and expand the
interstitium. Expansion of the interstitium would provide an
explanation for the continued decrease in the interstitial resistance
to hydraulic flow observed at Pip
>1.5 mmHg. To investigate this, we determined
if versus
Pip and
Pif in the rat anterior abdominal
muscle (AAM). We found that if changes with increasing Pip in a
nonlinear fashion. This nonlinear response of the interstitium to
pressure did not fully explain the observed linear rise in
K. We therefore explored the
relationship between hyaluronan (HA) and tissue conductance to fluid.
HA is a structural component of the interstitial matrix that is not anchored to tissue. It is a linear, anionic disaccharide polymer (molecular mass between 106 and
107 daltons) and is believed to be
a major interstitial component that determines hydraulic resistance to
bulk flow. Studies in tissue treated with hyaluronidase [an
enzyme that degrades tissue HA and other glycosaminoglycans
(GAG)] demonstrated 10- to 20-fold increases in fluid
conductivity of skin fascia (5) and 24-fold increases in pulmonary
interstitial conductivity (17). Our hypothesis is that
elevation of Pif will cause a
mobilization and a washout of HA by the flow through the abdominal wall
interstitium. If this hypothesis is correct, then the fivefold increase
in K observed in our previous study
could be explained by the reduction in the interstitial resistance to
bulk flow due to the combined effects of expansion of the interstitium
(if), the dilution of
interstitial macromolecules, and washout of HA.
| |
METHODS |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Animals
All experiments were performed in 200- to 350-g female Sprague-Dawley rats (Charles River Laboratories). Animals had free access to water and standard rat chow until the morning of the experiment. At least three animals were used for each pressure level investigated. All procedures were approved by the University of Rochester Committee on Animal Resources.Materials
Immunoglobulin G (anti-goat IgG, no. G-6638; Sigma, St. Louis, MO) was labeled with 131I (Amersham microvial, type P15; Amersham Life Science, Arlington, IL). Iodination was performed using Iodo-Beads (Pierce, Rockford, IL). The isotope was purified by passing the solution over an ion-exchange column (1-XP, Bio-Rad, Hercules, CA). Before the experiment the isotope was checked for degradation and free 131I by trichloroacetic acid (TCA). If free 131I was >1%, the solution was purified further by mixing it with saline and concentrating the mixture with a Centricon 30 microconcentrator (Amicon, Beverly, MA) by centrifugation (IEC Centra CL2). Dilution and concentration were repeated until the free 131I was <1% by TCA precipitation. [14C]mannitol was purchased from Amersham Life Science. The volume of distribution and half-life of the product have previously been determined to be 0.174 ± 0.006 l/kg and 13 min, respectively (8). With these values, an infusion rate of labeled mannitol was chosen to maintain a constant plasma concentration during the course of the experiment.Surgery
Anesthesia was induced by an intramuscular injection of pentobarbital sodium (60 mg/kg) to the hindleg and maintained with subsequent intravenous injections. Surgery was initiated on loss of the blink reflex. A tracheotomy was performed to reduce airway resistance. Two arterial lines were established using PE-50 catheters. The left carotid artery was cannulated to allow for continuous blood pressure measurements on a pressure measurement system (PE-10z Statham pressure transducer; Window Graf, Gould Valley Instruments, OH), and a tail artery catheter was used for blood sampling. A venous catheter was secured into the left external jugular vein for continuous infusion of [14C]mannitol from an infusion pump (Harvard Apparatus 22; Harvard Apparatus, Holliston, MA). The rectal temperature of the animal was continuously monitored and maintained between 35.5 and 38.5°C with a servo-controlled warming blanket (Harvard Apparatus) and an overhead heating lamp. The peritoneal cavity was exposed through a midline abdominal incision (~1.5 cm), and the hollow viscera (duodenum to rectum) were removed using the technique described in our previous publication (38). The slitlike abdominal incision was closed using a continuous suture after careful inspection to ensure that there was no bleeding. This maneuver was necessary so that fluid in the cavity had access to the entire abdominal wall. With the aid of a trocar, a multihole catheter was placed through the abdominal wall into the peritoneal cavity and secured with a purse stitch. A three-way valve was connected to the multihole catheter to administer and sample the dialysate and to continually measure the Pip with a glass capillary manometer. A urethral catheter was inserted for collection of urine during the experiment.Dialysis Procedures
An experiment was initiated with infusion of the dialysis fluid (prewarmed to 37°C) in an amount sufficient to raise Pip to 0.7-1.5 mmHg below the desired pressure. A reservoir containing the rest of the dialysis fluid was then connected to the three-way valve attached to the intraperitoneal catheter. The reservoir was maintained at the exact level above the right heart to produce the desired Pip, which was recorded every 15 min. Pip typically matched the desired level 15 min after initiation of the experiment.The dialysis fluid used in all experiments was a 5% bovine serum albumin in Krebs-Ringer bicarbonate solution containing (in mol/l) 0.12 NaCl, 0.01 KCl, 0.0021 CaCl2 · 2H2O, 0.025 NaHCO3, 0.00028 KH2PO4, and 1.18 ml of 1 M MgSO4 · 7H2O. The osmolality of the solution was adjusted to 290 ± 5 mosmol/kg by addition of NaCl. The solution was filtered with a 0.45-µm pore size membrane (Nalgene) and stored at 4°C. At 60 min, [14C]mannitol was given as a bolus intravenous injection followed by a continuous infusion for 1 h. This procedure is important to keep the plasma tracer concentration constant during the sampling period. Ten minutes before termination of the experiment, a bolus injection of 131I-labeled IgG was given intravenously to mark the local intravascular space. Blood and peritoneal fluid were sampled every 15 min.
At the end of the experiment, the following steps were taken in rapid
succession. After an anesthetic overdose, the animal was euthanized by
decapitation to stop blood flow, the fluid was drained from the cavity,
and the animal was rapidly frozen using chlorodifluoromethane
(Dust-off, Falcon Safety Products, Branchburg, NJ) precooled to
75°C. The abdominal wall was carefully cut from the carcass
with an autopsy saw. Thin sections (20 µm) were taken horizontally
with a Bright-Hacker cryomicrotome (model OTF, Fairfield, NJ) and dried
on a slide warmer. Sections were used for autoradiography and for
histology after being stained with hematoxylin and eosin.
Measurements
Interstitial hydrostatic pressure. The hydrostatic pressure profile within the AAM was measured by the technique of Wiig and colleagues (36) as later modified by Flessner (7) to allow for in vivo measurements of the interstitial pressure profile in the rat anterior abdominal wall. Details of the procedure may be found in our previous publication (7).
Dual-label quantitative autoradiography.
Quantitative autoradiography (QAR) was used to determine the local
concentration of each tracer in the tissue at the time of animal death.
Briefly, tissue tracer concentrations were determined from the thin
tissue sections (see Dialysis
Procedures). The sections were placed with standards
(tissues with known isotopic concentration) against X-ray film (Kodak
Biomax MR; Eastman Kodak, Rochester, NY) to produce autoradiograms.
Autoradiograms for the tissue contents of the
131I-labeled IgG were produced
first with the use of proper shielding to prevent the emission of
-particles on the film. After 10-12 half-lives of
131I-labeled IgG, the tissue
slides and the 14C standards were
placed against X-ray film to produce autoradiograms of the tissue
containing the [14C]mannitol.
After the films were developed, the tissue slides were stained with
hematoxylin and eosin. Each slide was examined by light microscopy to
determine the mesothelial layer and the skin side. This procedure is
important in tissue samples in which tracer concentration profiles are
to be determined. The films were analyzed with a computerized
densitometer (MCID; Imaging Research, St. Catherines, ON, Canada) that
measures optical density (OD) versus position in the tissue. The
isotopic standards were used to construct a calibration curve
(concentration vs. OD) to convert the unknown OD values from the tissue
samples into concentrations. By superimposing the tissue histology over
the autoradiogram, we carefully determined the location of the reading
and obtained a curve showing concentration versus position
(concentration profile data) or mean concentration in a large area of
the tissue. Dividing these concentrations by the plasma
concentration provided an estimate of the extracellular volume
(ec). Despite our experience
with this technique, a problem exists at boundaries of a tissue that is
in contact with a bulk solution. The pixel size is 50 × 50 µm,
and the estimated error in placement of an imaging grid is approximately ±2 pixels. Thus it is possible to have misalignment at the edge of multiple profiles, which likely results in smoothing of
the profile but some inaccuracy within 100 µm of the peritoneum. Approximately 100-200 profiles were averaged to minimize this potential error. Slopes of the profile were determined over
400-500 µm of data to minimize this effect.
Control Concentration Profile Experiments
The design of the experiment called for a variable volume (dependent on the nominal Pip) in the peritoneal cavity. Steady infusion of [14C]mannitol after the intravenous bolus set up a constant plasma concentration that was presumed to equilibrate with the tissue interstitium over a 60-min period. However, with no tracer in the solution within the peritoneal cavity, a diffusion gradient was set up from the tissue into the cavity. This might affect the tissue concentration profile in the vicinity of the peritoneum, causing a decrease in the tissue concentration within 200 µm of the edge (9). We utilized a mathematical model (9) with parameters for mannitol derived from Ref. 8; we found that a steady state is approached within the tissue within 10-20 min but that the interstitial concentration profile does decrease below the 90% level of plasma within 200 µm of the peritoneum. By computing the diffusive fluxes at the surface, we found that these were only slightly greater than the hydrostatic pressure-driven convective flux in the opposite direction (37). Previous work (9), in which [14C]EDTA was injected intravenously and tissue concentration profiles were measured after 1 h, did not demonstrate any significant decreases in tissue concentration near the peritoneum; the opposing flow of water and solute likely played a role in this observation. However, the peritoneal volumes in these earlier experiments were not as large as those used in some of the present experiments, and therefore we could not rule out a significant effect at the edge of the peritoneum. To address this question, we designed experiments in which the concentration in the cavity was maintained equal to that in the plasma. This design has the problem of possible mass transfer from the cavity into the tissue, potentially increasing the concentration in the vicinity of the peritoneum if the plasma concentration falls significantly below the intraperitoneal concentration. An increase in extracellular concentration above the plasma concentration would result in an overestimation of the trueec.
To determine the equilibrium distribution volume of
[14C]mannitol in the
abdominal muscle, we designed the experiment to maintain the tracer
concentration in the plasma constant and to maintain the tracer
concentration in the dialysis solution equal to that in the plasma
throughout the experiment. Therefore, loading of the abdominal muscle
with the tracer occurs from the blood side as well as from the
peritoneal side, and the concentration in the tissue must come to an
equilibrium throughout the tissue. A total of four rats matched for
body weight (207 ± 1.2 g) were used in this series. The animals
were surgically prepared (see Surgery) and were rendered anephric
by bilateral ligation of the renal pedicle; this eliminated the need to
continuously infuse the tracer to make up for renal clearance.
[14C]mannitol (15 µCi) was
given as an intravenous bolus injection. For each animal the dialysis
solution was 130 ml of a 5% bovine serum albumin in Krebs-Ringer
solution containing 30 µCi of
[14C]mannitol. This
radioactivity dose ensured that the concentrations of
[14C]mannitol in the peritoneal
fluid and in the blood were identical throughout the experiment. The
dialysis solution was injected into the peritoneal cavity and allowed
to dwell for either 90 (n = 2) or 150 min (n = 2). These times were chosen
to be longer than the 60 min used in the bulk of the experiments to
test two assumptions: that an equilibrium between the plasma and the
extracellular space is attained by 60 min and that the space itself is
in a steady state under the constant intraperitoneal pressure. If the tissue-averaged values calculated for
ec are the same for each experiment duration, our assumptions are justified. The dialysis solution and blood were sampled at 10 min and then every 30 min. At the
end of the experiment, the animal was euthanized and the abdominal
muscle was harvested (see Dialysis
Procedures) and prepared for QAR.
Tissue preparation for HA assay. Tissue samples (~300 mg) were harvested from the AAM and adjacent subcutaneous plane (SC) of intact rats with no fluid in the cavity (Pip = 0 mmHg) and from the AAM and SC of a second set of rats after isotonic peritoneal dialysis at a nominal Pip of 6 mmHg for a single 2-h dwell. In the animals that underwent dialysis the SC was sampled before and after the dialysis procedure. SC samples from before dialysis were taken from the left abdominal wall before placement of the plastic plate necessary for the interstitial pressure (Pif) measurements. After dialysis, the right abdominal wall (previously untouched) was cut with the skin attached into 1-cm2 samples. From each of these samples, the skin was removed and the SC sample was collected; the muscle was then cut into smaller sections of 300 mg. The SC samples within the 1-cm2 section were collected from the tissue immediately adjacent to the muscle. In a subset of animals, the inner layer of abdominal muscle was separated from the outer layer along the muscle plane and each was processed individually. The tissue samples were placed in preweighed vials, and their wet weights were immediately determined. The vials were connected to a Vacu-Freeze (VirTis, Gardiner, NY) and freeze-dried to a constant weight. The dry weights of the tissue samples were then determined. A digestion buffer (pH 7.2) containing 0.05 M Tris · HCl, 0.01 M CaCl2, and 2.4 U pronase (Sigma) was added to the tissue, and the vials were incubated in a water bath at 55°C for 20 h. Tissue digestion was stopped by boiling at 100°C for 5 min. The vials containing the digest were centrifuged at 10,000 rpm for 1 h. The lipid layer was discarded. A 100-µl sample was taken from the digested sample for HA concentration determination (27). The total amount of HA in the tissue sample was then determined from this concentration and the total digestion volume. The mass of HA was then divided by either the dry weight or the wet weight of the original sample to calculate the HA concentration.
Radiometric tissue HA assay. The test kit was obtained from Pharmacia (Uppsala, Sweden) and used as suggested by the manufacturer. The test is based on a specific HA binding protein (HABP). The HA of the sample (unknown) reacts with 125I-labeled HABP in solution. A volume of 100 µl from either a solution containing the digested tissue or a standard solution was mixed with 200 µl of 125I-labeled HABP in polystyrene tubes and incubated for 60 min at room temperature. HA-Sepharose (100 µl) was then added, and the tubes were incubated for a further 45 min. Separation was performed by centrifugation after addition of 2 ml of washing solution followed by decanting. The radioactivity bound in standards and unknowns was expressed as a percentage of the radioactivity bound in the zero standard, and a standard curve was constructed. HA concentration in each sample was determined from the standard curve. The detection limit of the test kit is <10 µg/l. Our samples were diluted at 1:5 with the zero standard provided by the test kit. Samples containing >1,000 µg/l HA had to undergo further dilution.
Calculations
There is no specific marker for the interstitial space. However, the interstitial volume (if) can
be calculated from the difference between the extracellular space
(ec) and the intravascular space (iv)
|
(1) |
ec and iv were therefore calculated
from
![&thgr;<SUB>ec</SUB> = <FR><NU>V<SUB>ec</SUB></NU><DE><IT>M</IT><SUB>tot</SUB></DE></FR> (<IT>x</IT>) = <FR><NU>C<SUP>[<SUP>14</SUP>C]mannitol</SUP><SUB>tissue</SUB></NU><DE>C<SUP>[<SUP>14</SUP>C]mannitol</SUP><SUB>plasma</SUB></DE></FR>](/content/vol276/issue2/fulltext/H517/img003.gif)
|
(2) |
|
(3) |
ec as the extracellular
distribution of [14C]mannitol is
based on the assumption that the concentration of the tracer in the
interstitial fluid is equal to that in the plasma at the time of tissue
sampling. Because it is not possible to directly sample the
interstitial fluid, we corrected for renal loss of the tracer by
maintaining a constant plasma concentration and assuming a steady state
after 60 min of tracer infusion.
Estimation of Tissue Compliance
Ideally, tissue compliance should be calculated from data derived from a series of equilibrium states in which the interstitial volume and interstitial pressure are determined precisely. This type of data is most easily obtained during in vitro experiments, as illustrated by the scheme used by Maroudas (22). Our system is not at equilibrium because, by design, there is a pressure gradient across the abdominal wall (Pip Pskin), which causes flow from
the peritoneal cavity into the muscle. Our previous work (10, 37)
showed that flows at Pip ranging
from 2 to 8 mmHg are steady over periods ranging from 60 to 180 min.
Our control experiments (see Control Concentration
Profile Experiments) have shown that
if is stable over 90-150
min of dwell of an isotonic solution in the cavity at a
Pip of 6 mmHg. As long as the
Pip was maintained constant
throughout the experiment, our determinations of
dPif/dx
demonstrated stable pressure profiles. Because the flow rate into
tissue, the if, and
dPif/dx
were apparently constant for the period of observation, we feel that
the system was at a steady state for the given
Pip. Because the system was only at steady state and not at equilibrium, our calculation of tissue compliance should be considered as an estimate of the absolute value.
Average whole tissue compliance
(tissue) for the
subperitoneal tissue was calculated from the slope of the average
interstitial fluid volume
(
if)
plotted versus the average tissue pressure (
if)
as
|
(4) |
if and
if
were both calculated from the distance-averaged values over the 600 µm of tissue adjacent to the peritoneum of the respective profiles.
Because of the nonlinearity of the resulting curves, neither a
conventional least-squares regression analysis nor the method of
Snedecor and Cochran (31) could fit the data. Attempts to correlate the
data with a piecewise polynomial regression model [Number
Crunching Statistical Software (NCSS), version 6.0; NCSS Statistical
Software, Ogden, UT] did not fit the overall averages for each
pressure level, and therefore the data were broken up in specific
Pif ranges that apparently defined the curve: 2.4 mmHg to +1.2 mmHg (below threshold pressure), +1.2 mmHg to +4.2 mmHg, and +4.2 to + 7.4 mmHg. Each of these ranges
was analyzed with conventional least-squares analysis and the
"method of averages" suggested by Brace (2). This latter method
is essentially a graphic method that takes into account the fact that
both variables
(if
and
if)
are subjected to unknown errors. Because of these errors, the use of
the conventional least-squares regression analysis may underestimate
the slope (tissue).
Our long-term goal is the modeling of convection at the tissue level,
which requires parameters related to intratissue transport phenomena.
Because we have determined profiles of
if and
Pif versus distance
x from the peritoneal edge, we can
also estimate the interstitial compliance based on measurements in the
first 400-500 µm from the peritoneum
[
(Pif)]. We used
the nomenclature of Taylor and colleagues (32) to differentiate this
compliance from those based on whole tissue measurements
|
(5) |
if and
Pif in the vicinity of the
peritoneum and calculating their ratio, we obtained an estimate of
interstitial compliance at the tissue edge. By estimating the slopes of
the profiles over this distance, we have minimized the errors
introduced by errors in measurements at the edge of the peritoneum.
This issue is discussed in detail later (see
Interstitial Compliance).
Statistics
All data are presented as means ± SE unless stated otherwise. One-way ANOVA was used to analyze the effect of a single factor (i.e., Pif) on measuredif. Two-way ANOVA was used to
check for possible variation in
if resulting from animal number
and Pip. All calculations were
performed with NCSS. A statistic was considered to be significant if
the probability of a type 1 error was
P < 0.05.
| |
RESULTS |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Figure 1 displays the measured hydrostatic
pressure (mean ± SE) profiles across the abdominal wall. In
previous work (7) the variance in the pressure measurements (±1
mmHg) was estimated from a series of determinations in tissue in which
each side was maintained at atmospheric pressure, and presumably the
pressure across the tissue under this condition was relatively constant and equal to the atmospheric pressure. The estimated variance for the
position of the micropipette tip in tissue is ±200 µm. No
measurements of negative pressure near the peritoneal surface were
found in any of the animals that underwent dialysis. However, in rats
with a closed or open cavity and no fluid instilled into the cavity,
there seems to be no specific trend in the measurements, because the
mean pressure in the peritoneal cavity and in the tissue averaged
(mean ± SE) 2.8 ± 0.4 and 0.12 ± 0.19 mmHg,
respectively, across the entire abdominal wall muscle.
|
[14C]mannitol concentration
profiles normalized to the plasma concentrations at each of the
Pip levels investigated are
depicted in Fig. 2. All profiles were
obtained after a total 120-min dwell time. At 60 min, a bolus
intravenous injection was administered and a constant tracer infusion
was started for the subsequent 60 min, providing estimates of the local
extracellular volume (ec) in
the abdominal wall muscle. Although the thickness of the abdominal wall
of the rat is ~1.3-1.9 mm, only the first 1,000 µm from the
peritoneum were used due to an artifact in the tissue introduced with
removal of the skin from the subcutaneous side of the abdominal wall.
The negative slope at the peritoneum appeared to increase with
Pip >3 mmHg. The intercept of
the curve at the y-axis was almost the
same at the Pip range of
0-1.5 mmHg, but this value doubled at
Pip >3 mmHg. Tracer
concentration in the wall displayed a nearly flat profile for the low
Pip levels investigated. The
sampled 1,000 µm of tissue is not an amorphous collection of muscle
fiber bundles and capillaries; in most tissue sections there was a
fascial plane at ~400-800 µm from the peritoneum. The observed
changes in slope at 400-600 µm may be due to this fascial plane.
The distance-averaged if values
as calculated from the tissue profile data are compared with the means
of the whole tissue measurements in Table
1.
|
|
In Fig. 3 the concentration of
[14C]mannitol in the peritoneal
fluid for each of the Pip pressure
levels is plotted as a function of dwell time (in min). Peritoneal
tracer concentration is normalized by dividing by the plasma
concentration [ratio of dialysate to plasma concentration
(D/P)]. All 60-min D/P values are <0.55, indicating that the
dialysis fluid had a lower concentration than the plasma and the tissue
interstitium. From Eq. 2,
Ctissue equals ecCplasma
at equilibrium, and the concentration in the interstitium equals
Ctissue/ec.
The normalized peritoneal fluid concentration is significantly
(P < 0.04, n = 3) lower at high
Pip (4.4, 6, and 8 mmHg) versus
low Pip (0, 0.7, and 1.5 mmHg),
because the peritoneal volume required to produce a nominal
Pip varies directly with the
Pip. The larger intraperitoneal
fluid volume used in the high Pip
experiments provides a larger sink for the tracer transporting from the
tissue into the cavity. Whereas the surface area of the peritoneum
exposed to fluid increases with fill volume, the relationship is
nonlinear and likely becomes constant at higher volumes (15).
Therefore, the rate of volume increase in these experiments exceeds the
increase in the rate of mass transfer from the tissue, and the
resulting concentration in the cavity at 60 min of equilibration time
is lower. The effect of the diffusion from the tissue edge into the
cavity and the possible "edge effect" at the peritoneal
tissue-fluid interface will be discussed later (see
DISCUSSION). The D/P curve for 0.74 mmHg Pip seems to be lower than
that for 1.5 mmHg, but this difference is not statistically significant
and may be attributed to a set of larger animals in the 0.74 mmHg
group, which required a relatively larger fill volume to achieve the
nominal Pip.
|
Figure 4 displays the results of control
experiments designed to investigate the possible effects of diffusion
from the tissue into the cavity (see Control
Concentration Profile Experiments). The design of
these experiments differed from those shown in Figs. 2 and 3 in that
the tracer concentration in the cavity was made as close as possible to
that of the plasma to eliminate any effect of diffusion from the tissue
into the cavity. In addition, the animals were rendered anephric to
minimize changes in the plasma tracer concentration. Intraperitoneal
volumes were similar to the volumes listed in the experiments at 6 mmHg
Pip presented in Fig. 3. Figure
4A shows D/P of
[14C]mannitol for the animals
dialyzed for 90 (n = 2) or 150 min (n = 2). As shown, D/P
was not affected by the duration of the experiment and stayed close to
unity throughout the course of the experiment. Although we anticipated
changes in the tracer volume of distribution due to ligation of the
renal pedicles, we determined the distribution volume for
[14C]mannitol to be 0.181 ± 0.003 l/kg rat tissue, which is not different from the value of 0.174 ± 0.006 l/kg rat tissue based on our previous work (8). However,
the plasma half-life of the tracer has increased from 13 min (8) to
1,061 ± 190 min. The tracer concentration in the peritoneal fluid
only slowly decreased with time with an estimated half-life of the
tracer of 5,109 ± 1,939 min. In Fig. 4B,
[14C]mannitol concentration
profiles normalized to plasma concentration are shown for the two
groups of animals (dialyzed for 90 and 150 min). In addition, curves
are shown for the averaged profiles for all four animals as well as for
the 60-min equilibration in which no
[14C]mannitol was added to the
peritoneal fluid (see Fig. 4B). The overall mean tissue tracer concentration shows a flat profile, with a
tissue-averaged if value of
0.35 ± 0.004 ml/g, which is not different from that resulting from
Fig. 2, which is also shown in Fig.
4B. Although there are
differences in the profiles from the different techniques, they are not
statistically significant.
|
Figure 5 is a plot of the mean
if in the abdominal wall versus
distance-averaged Pif. Each was
calculated from the average value of the profile over the 600 µm of
tissue closest to the peritoneum. For correlation, our previous
K determinations are also shown. Under
normal physiological conditions when the peritoneal cavity is not
opened to atmospheric pressure,
if is 0.17 ± 0.00 ml/g wet weight tissue. At the Pif
levels of 0.3, 0.7, and 1.2 mmHg,
if is 0.18 ± 0.00, 0.20 ± 0.01, and 0.20 ± 0.00 ml/g, respectively. Increasing the
Pif to 2.1 mmHg resulted in an
increase in if to 0.26 ± 0.01 ml/g. A further increase in
Pif to 2.6, 4.2, 5.2, or 7.4 mmHg
resulted in if of 0.31 ± 0.01, 0.39 ± 0.01, 0.36 ± 0.01, or 0.33 ± 0.00 ml/g,
respectively. This response is nonlinear, with a threshold for change
at Pif of 1.2 mmHg. In the
Pif range between 4.2 and 7.4 mmHg, if seems to decrease with
increasing Pif. A one-way ANOVA
demonstrated significant (F = 22.91;
P < 0.0001) differences among
if of the three groups of
animals dialyzed at 4.2, 5.2, and 7.4 mmHg. A multiple-comparison
posttest (Bonferroni t-test) was used
to determine whether if of one
particular group differs significantly from another specified group.
Three comparisons were made. The results of this analysis are
summarized in Table 1. The data indicate that the response of the
interstitial space of unrestrained tissue to pressure may be biphasic
and characterized by two threshold pressures for change; at
Pif of 1.2 mmHg there is an abrupt
expansion of the interstitium in a linear fashion up to
Pif of 4.2 mmHg, but at
Pif >4.2 mmHg there is a
significant decrease in if. The
data for K do not follow the overall
pressure-volume curve; K is low and
almost constant at Pif <1.2
mmHg, whereas above this threshold there is linear increase in
K with increasing Pif. Furthermore, no decline in
K is observed when
Pif >4.2 mmHg. The local
intravascular volume (iv) in
the abdominal muscle (not shown) did not change with changing
Pif and remained at 0.010 ± 0.002 ml/g wet tissue.
|
The slopes from the curve of if
versus Pif were calculated by the
method of averages and were +0.67 ml · 100 g
1 · mmHg1
for Pif 2.4 to +1.2 mmHg,
+5.9 ml · 100 g1 · mmHg1
for Pif of +1.2 to +4.2 mmHg, and
1.8 ml · 100 g1 · mmHg1
for Pif of +4.2 to +7.4 mmHg.
Standard least-squares regression over each of these ranges provided
similar estimates (0.82, 6.3, and 1.8 ml · 100 g1 · mmHg1,
respectively). These results therefore demonstrate a low compliance below the apparent threshold of
Pif of 1.2 mmHg, a markedly
increased compliance between 1.2 and 4.2 mmHg, and an elimination of
further tissue expansion above a second apparent threshold of
Pif of 4.2 mmHg.
The local tissue compliance
[(Ptissue)] near
the peritoneum as obtained from the normalized concentration profiles
and the hydrostatic pressure profiles in the abdominal wall are listed in Table 2. The data clearly demonstrate
the dependency of the local
(Ptissue) on the local tissue
fluid pressure. For Pif <2.1, there is little expansion of the tissue, as evidenced by the low, flat
profiles of Fig. 2, and therefore calculations were considered inappropriate in this range. The general trend in the data show that
(Ptissue) increases from low
Pif to higher values. The highest (Ptissue) was obtained for
the animals dialyzed at Pif of 4.2 and 5.2 mmHg, but the value decreased in animals dialyzed at
Pif of 7.4 mmHg. This pattern
appears to correlate with the curve in Fig. 5.
|
Effect of Convection on Tissue HA
Because changes inif cannot
explain the fivefold change in K, we
examined the effect of Pip on the
contents of tissue HA. AAM and the associated SC tissue were sampled at
baseline (Pip = 0 mmHg) and after
isotonic peritoneal dialysis at constant
Pip of 6 mmHg for 2 h. In the AAM,
HA was reduced from 258 ± 15 (n = 6) to 204 ± 15 µg/g wet tissue
(n = 6, P < 0.05), and in the SC it
increased from 271 ± 21 (n =12) to
418 ± 110 µg/g wet tissue (n = 6, P > 0.05). Because of possible
potential effects of tissue edema on HA data, a second series of
experiments was designed to assess tissue HA contents on the basis of
tissue dry weight. In these animals, the inner layer of the anterior
abdominal wall muscle was dissected free from the outer layer of
muscle; the inner portion demonstrated a greater decrease in HA
contents, from 487 ± 16 to 265 ± 40 µg/g dry tissue, than did
the outer layer of abdominal wall muscle, which is decreased to only
349 ± 67 µg/g dry tissue after dialysis at
Pip of 6 mmHg. The data from this
series are shown in Fig. 6, which shows
that the HA content in the full-thickness AAM specimen was reduced from
487 ± 16 (n = 4) to 360 ± 27 µg/g dry tissue (n = 4, P < 0.05) and that in the SC was
increased from 528 ± 72 (n = 8) to
1,050 ± 136 µg/g dry tissue (n = 4, P < 0.001). In four
separate animals, HA content in the SC was determined before
(Pip = 0 mmHg) and after dialysis
at Pip of 6 mmHg. HA before
dialysis was 605 ± 138 µg/g dry tissue and significantly
increased to 1,214 ± 142 µg/g dry tissue
(P < 0.05).
|
Efforts were made to sample SC tissue immediately adjacent to the muscle sample, and the relative positions of each sample were always within the 1-cm2 tissue sample. However, there could have been microscopic nonhomogeneities in HA concentration within the square sample that this technique would not detect. Despite these limitations, the data strongly suggest the movement of the mobile muscle HA (washout) in the direction of fluid flow toward the subcutaneous tissue.
The dry weight-to-wet weight ratios of the AAM and SC tissue obtained
from control rats (Pip = 0 mmHg)
were 24.6 ± 0.3 and 55.6 ± 2%, respectively. Mean water
contents of the AAM and SC, calculated as 100 × [(wet
weight dry weight)/wet weight], were 75 and 44%,
respectively, and increased to 83.9 ± 0.8 and 85.8 ± 4.7%, respectively, after dialysis at nominal
Pip of 6 mmHg for 2 h. In Fig.
7, the distribution of the water content in
the AAM was calculated on the basis of milliliters per gram of dry tissue. The substantial increase in the influx of water from the cavity
into the tissue at Pip of 6 mmHg
seems to be confined mainly to the interstitial space. Therefore,
if increased substantially, whereas the local vascular volume stayed constant. However, the local
intracellular volume showed a slight increase in the animals dialyzed
at Pip of 6 mmHg, but this
increase was not statistically different from control.
|
| |
DISCUSSION |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
The present study addressed the effect of hydrostatic pressure on the
interstitial space of the abdominal muscle. The
Pip range (2.8 to +8 mmHg)
investigated varies from a normal physiological level to levels
typically observed during peritoneal dialysis. The results indicated
that if is not a constant
entity but one that changes with increasing
Pip >1.5 mmHg to bring the
adjacent tissue above the apparent threshold
Pif of ~1.2 mmHg. However, this
change is nonlinear and does not fully explain our previous finding of
a linear increase in the tissue hydraulic conductivity (K) at
Pip >1.5 mmHg. The continued
decrease in the interstitial resistance to hydraulic flow on elevation
of Pip is alternatively explained
by a combination of expansion of the interstitium
(if), dilution of the
interstitial macromolecules, and an apparent washout of mobile
interstitial HA in the direction of fluid flow from the muscle to the
associated subcutaneous space.
Relation of Pip to Pif
As in all serous cavities, the hydrostatic pressure in the peritoneal cavity under normal physiological condition is slightly negative. In this study, we determined that the hydrostatic pressure profile in the interstitium of the muscle of the anterior abdominal wall under normal physiological condition is rather flat and is approximately2.8 ± 0.4 mmHg (averaged over the entire muscle wall). The negative
Pif sets the capillary Starling
forces toward a slight filtration, which is balanced by an equal
lymphatic absorption. Under these conditions, the resistance to fluid
mobility in tissue is very high (14). It is not until
Pip is >1.5 mmHg (mean
Pif = 1.2) that a significant drop
in resistance to fluid mobility in tissue is observed (10). Because
Pip is exerted across the abdominal wall muscle where peritoneal fluid comes in contact with the
peritoneal surface, Pif is
expected to equal Pip at least in
the immediate submesothelial space. Our measurements of tissue pressure
demonstrated different pressure profiles from those that would be
calculated from the overall slope {overall pressure difference across the abdominal wall divided by tissue thickness
[(Pip Pskin)/tissue
thickness]}. The interstitial pressure gradient near the
peritoneum increased by only 49% over the range of
Pip from +2 to +8 mmHg, whereas
the overall pressure difference increased by a factor of four.
Relation of Pif to
if and Pressure-Volume Curve
if
essentially from the blood side, whereas the concomitant change in
Pif was simultaneously measured
with one of several techniques: an implanted capsule (13, 14), a
micropuncture technique (29, 33-35), or as calculated from
isogravimetric capillary pressure after venous outflow elevation (3, 6,
12, 24). In these studies, if
is reduced by perfusing the vascular space with dextran solution or by
performing very hypertonic (20% glucose) peritoneal dialysis to
dehydrate the interstitium of water by colloid or crystalloid osmosis,
respectively. if is increased
in these studies by either infusion of saline in large volumes or
venous outflow pressure elevations. In contrast, the present study
altered Pif and simultaneously
assessed the change in if while
the local vascular volume was constant. Although this new approach has
confirmed the principal shape of the pressure-volume curve, it provided
an additional quantitative description of the pressure-volume curve
under conditions of interstitial overhydration at
if >100% compared with
control. This portion of the curve is usually missing or largely
uncertain in the studies cited herein because of the experimental
difficulties involved in achieving a state of extreme overhydration
without affecting hemodynamics. The present study demonstrated an
apparent decrease in if in overhydrated tissue (
if
>100% of if at
Pip = 0 mmHg) at
Pif 
4.3 mmHg, reflecting a
reduced total tissue compliance
(tissue).
Determination of if
if can only be measured
indirectly using labeled tracers, which separately mark the
extracellular and intravascular spaces. For estimates of the
distribution volumes obtained by this technique to be valid, the
concentration of the tracer in the volume to be measured should equal
its concentration in the plasma at the time of sampling. To approximate
a steady state of equilibrium, several investigators have maintained a
constant tracer concentration in plasma by bolus dose injection and
continuous infusion (18) or by infusion after bilateral nephrectomy
(24, 29). Larsson et al. (18) found that the distribution volume of
51Cr-EDTA (molecular mass 341 Da)
in rat skeletal muscle does not change after infusion times of 60, 90, or 120 min. Wiig and Reed (34) reported that the extracellular spaces
measured with 51Cr-EDTA in
anephric cats were similar at 2 and 6-7 h after tracer injection.
Our mathematical model predicts that, for a substance with a small
molecular mass such as
[14C]mannitol (molecular mass
180 Da), the tracer is likely to have access to the total extracellular
water after 15-20 min; the 60-min equilibration time should
therefore be more than sufficient to approach an equilibrium.
Furthermore, in the control experiments reported in this study, we
varied the equilibration time between 90 and 150 min and allowed the
tissue to equilibrate with constant and identical concentrations of
[14C]mannitol in blood and
peritoneal fluid. After 90 and 150 min, the distribution volumes in the
tissue were similar to results at 60 min, with a coefficient of
variation of ~7%. Because there were no statistical differences
among the profiles of if at the 60-, 90-, or 150-min equilibration times, we feel confident that assessment of the distribution volume of
[14C]mannitol in the abdominal
muscle after 60 min of continuous infusion is at a steady state with
regard to its compliance properties and at near equilibrium with the
plasma concentration.
In a previous study (9),
[14C]EDTA was injected
intravenously and the tracer concentration profiles in different
tissues surrounding the peritoneal cavity were measured after 1 h. No significant decrease in the tissue tracer concentration was observed near the peritoneal edge. Theoretically, however, with no tracer in the
initial solution within the peritoneal cavity and a constant extracellular space, an edge effect or decrease in tissue concentration should have been observed adjacent to the cavity because there is a
diffusion gradient from the tissue into the cavity. The
present concentration profile measurements did not demonstrate a
consistent edge effect, especially when the pressure in the cavity was
>3 mmHg. Other factors that contribute to the uncertainty of the concentration at the peritoneal edge are
1) the 50- to 100-µm error in
matching the edge of the tissue with the image edge in macro-QAR,
2) the 20- to 30-s lag time in the
freezing of tissue after the cessation of an animal's blood flow,
which could cause a small decrease in the edge concentration, and
3) incomplete removal of peritoneal
fluid from the tissue and subsequent imaging of autoradiographic
imprints of the fluid overlying the tissue. These factors should
primarily affect the tissue space in the 50-100 µm immediately
adjacent to the peritoneum with little effect on deeper tissue. Because
the values for if and
(Ptissue) were obtained over
a minimum of 400-600 µm of tissue, the effects of these
uncertainties at the edge were minimized.
Interstitial Compliance
The shape of the pressure-volume curve (Fig. 5) assessed in the present study indicates that the interstitium of the abdominal wall during peritoneal dialysis is characterized by at least three apparent phases of expansion. Under normal physiological conditions in the intact rat, thetissue of the abdominal
muscle is low (0.7 ml · 100 g1 · mmHg1)
but increases to 6 ml · 100 g1 · mmHg1
during the initial phase of hydration (50-100% increase in
if). At
Pif = 4.2 mmHg, the expansion
appears to stop with a tendency to decrease in the final phase when
if >100% (overhydrated tissue).
The cause of this decrease in
tissue in overhydrated tissue
may be due to the rigid fascia (34). The fact that the abdominal wall
is unsupported in these experiments may in part explain differences in
our observations from those of others. On the other hand, the decrease
in interstitial volume may also relate directly to a group of
compensatory mechanisms that act by adjusting capillary pressure and
surface area, Pif and interstitial
oncotic pressure, as well as lymph flow to prevent progressive
accumulation of tissue edema and to counteract changes in
if (3, 24) by shifting the
transcapillary Starling equilibrium toward interstitial fluid reabsorption to decrease the hydration in the tissue.
The average whole tissue interstitial compliance is defined as the
ratio of a change in the tissue-averaged interstitial fluid volume
(if) to a corresponding
change in tissue-averaged interstitial hydrostatic pressure
(Pif). Most recent studies
have estimated in vivo compliance with measurements of changes in
if and micropipette measurements of changes in Pif in
the same tissue type and site (24, 27). It is known that estimates of
the tissue compliance strongly depend on the tissue pressure and the
method used to measure the pressure (35, 36). Previous estimates of
tissue in skeletal muscle in
rats (1.4 ml · 100 g1 · mmHg1;
Ref. 27), cats (1.4 ml · 100 g1 · mmHg1;
Ref. 6), or dogs (1.6 and 1.3 ml · 100 g1 · mmHg1;
Ref. 35), although similar to our
tissue value, were interpreted to reflect tissue values for
states of dehydration and early stages of hydration. In overhydrated
tissue, tissue was found to be
~2-7 ml · 100 g1 · mmHg1
to infinity. In rat skeletal muscle, Reed and Wiig (29) calculated tissue to be 5 ml · 100 g1 · mmHg1,
compared with 3.1 ml · 100 g1 · mmHg1 as obtained by Wiig
and Reed (34) for an increase in interstitial fluid volume from 0 to
70% in cat skeletal muscle. This value is higher than that of 2.5 ml · 100 g1 · mmHg1
as calculated from the data of Eliassen et al. (6) in the same species.
Although these variations may be due to species differences, they may
well be due to the interpretation of the pressure-volume curve in these
studies. To compare tissue
values from different studies, the control interstitial pressure is
commonly set at 0 mmHg. This arbitrary approach came from a generalized understanding of the pressure-volume curve. When
Pif is negative, small changes in
if cause a rapid increase in
Pif, but when
Pif is atmospheric, there is an
"inflection" of the pressure-volume curve such that large
increases in if are necessary
to increase Pif only slightly. In
our study we monitored the possible change in
if secondary to a change in
Pif. This allowed us to determine unique values of if for a given
Pif. Furthermore, we did not observe a significant change in the local
if when the
Pif was varied between 2.4
and +1.2 mmHg, but a significant change in if was observed at a threshold
Pif of 1.2 mmHg. In cat intestine Mortillaro and Taylor (24) observed an inflection of the
pressure-volume curve at a threshold pressure of 3.5-6
mmHg. Granger and Taylor (13) attributed this change to disruption of
the mucopolysaccharide-collagen cross-linkages caused by imbibation of
fluid beyond the tissue gel saturation point. This would imply an
increase in the gel compliance and, hence, account for the change in
the pressure-volume curve.
To our knowledge, the present estimates of the local tissue compliance
(Ptissue) are the first ones
based on local measurements of concentration and hydrostatic pressure
profiles at nominal Pip for the
same tissue in rats. This parameter is most useful in the
"distributed approach" toward the quantitative description of
interstitial transport (8, 32), in which local phenomena at the
microscopic level are modeled. Whole tissue or compartmental models do
not require such a parameter, and the term
tissue suffices. It is known
that estimates of the tissue compliance strongly depend on the tissue
pressure and the method used to measure the pressure (27, 36). The
calculated values of (Pif) in
Table 2 reflect this dependency of the imposed pressure and are of the
same magnitude as those found with other methods in whole tissue
preparations (tissue). The
values for (Pif) at pressure
>4.2 mmHg are likely more realistic than the negative slope obtained
from the curve of the averaged values of
if versus
Pif. Because of data variability and edge effect, the calculated
(Pif) qualify as best estimates.
Relation of Pip to Q and Pressure-Flow Curve
Cavity-to-tissue fluid flux across a definite area (Q/A) on the peritoneal side of the abdominal wall showed a direct dependency on Pip (7, 10). There is only a slight increase in fluid flux when Pip is changed from 0.7 to 1.5 mmHg, but at 1.5 mmHg a very significant change in the slope [sometimes called "yield phenomenon" (20)] occurs. Therefore, the pressure-flow curve for the abdominal wall is a nonlinear function of Pip, and 1.5 mmHg pressure in the peritoneal cavity (referenced to the right heart in the supine position) is a threshold for the hydraulically driven fluid flow from the cavity into the surrounding tissue. The nonlinearity of the pressure-flow curve does not contradict the consistent finding of a proportionate increase in peritoneal fluid loss rate with increasing Pip, because in most studies oriented toward dialysis, the threshold pressure is typically exceeded. In subcutaneous tissue, McMaster (23) demonstrated no fluid flow at inflow pressures <6.2 mmHg. The author further demonstrated in edematous subcutaneous tissue that there was no specific threshold for fluid flow into the tissue but that a linear relationship existed between the imposed pressure and the rate of fluid flow. The same phenomenon was later observed by Guyton et al. (14), who measured the flow induced by a pressure differential between two catheters inserted in the subcutaneous space. They observed a very slow flow at small (2-3 mmHg) pressure differences but markedly increased flow at pressure differentials large enough to create edema. They estimated that edema increases tissue fluid conductance in the subcutaneous space by >100,000 times. Levick and colleagues (16, 20, 21) have consistently found a linear increase in fluid flow across the synovial lining of rabbit knees when intra-articular pressure was increased above an apparent threshold pressure or a "yield point" of ~6.6 mmHg. The common theme in all of these in vivo studies is the yield point above which the interstitial space swells and the resistance to fluid flow decreases. Although the magnitude of this threshold pressure clearly depends on the type of tissue, the significant change in the slope of the pressure-flow curve when the pressure driving the flow exceeds the threshold level can only be attributed to an increased rate of conductance with pressure.It has been shown that HA, which is the mobile component of the interstitial matrix, is slowly removed by lymph to be subsequently degraded in lymph nodes and liver (11, 19). This mobility is reported to be enhanced by increased interstitial fluid flux (28), but in the skin, even with high interstitial fluid flux after elevation of the venous pressure and saline infusion, the daily output of HA by lymphatic drainage from the dog paw was found not to exceed 6% of the total tissue content (29). Given the paucity of lymphatics in the abdominal muscle, removal of macromolecules from the interstitium would not be expected to account for our data. Alternatively, the apparent movement of HA from the AAM to the SC, together with the dilution effect on interstitial macromolecules, could possibly explain the linear decrease in the interstitial resistance to flow in the abdominal wall. This apparent movement of HA is supported by the fact that in the dialyzed rats, the inner layer of the AAM demonstrated a greater decrease in HA content compared with its adjacent outer layer, and also by the apparent loss of HA from the anterior abdominal wall to the adjacent subcutaneous space. Furthermore, assessment of HA contents in the subcutaneous plane in the same animal demonstrated a threefold increase after dialysis compared with its baseline value before dialysis. It may be argued that the observed increase in subcutaneous HA is due to an increased local synthesis. Although the present data do not rule out this possibility, it seems unlikely because of the short duration of the experiment and because of the concomitant decrease in the HA content of the anterior abdominal wall. In a recent study, Bert and Reed (1) measured the in vitro transdermal flow rate of a phosphate-buffered saline as a function of applied pressure. They observed a significant drop in HA content in only one of the three pieces of dermis investigated. In the other two pieces, only 1% of HA and a tracer amount of collagen were apparently washed out of the tissue in two days. The authors also observed a decreasing average flow conductivity with increasing applied pressure, which they attributed to a compressible dermis. Such a finding is compatible with supported tissue in their in vitro study. Because tissue in our in vivo experiments is unsupported, and because of the difference in the tissue type investigated in the two studies, our observations will differ.
Interrelation Between Pif,
if, and
K
in anterior
abdominal muscle at different Pip
) or 150 min (
). Intraperitoneal volumes were similar to
that listed for 6 mmHg in Fig. 
) as determined from freeze-dried tissue samples. Data are
means ± SE.
