AJP - Heart AJP: Cell Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Heart Circ Physiol 274: H209-H216, 1998;
0363-6135/98 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kleen, M.
Right arrow Articles by Messmer, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kleen, M.
Right arrow Articles by Messmer, K.
Vol. 274, Issue 1, H209-H216, January 1998

PEEP only partly restores disturbed distribution of regional pulmonary blood flow in lung injury

M. Kleen1,2, B. Zwissler2, and K. Messmer1

1 Institute for Surgical Research and 2 Institute of Anesthesiology, Klinikum Grosshadern, Ludwig-Maximilians-University, 81366 Munich, Germany

    ABSTRACT
Top
Abstract
Introduction
Methods
Results
Discussion
References

The effects of lung injury, positive end-expiratory pressure (PEEP), and norepinephrine on heterogeneity of regional pulmonary blood flow (rPBF, radioactive microspheres) were investigated. We hypothesized that lung injury increases heterogeneity of rPBF and that PEEP ventilation reduces these effects. Heterogeneity of rPBF is scale dependent and was therefore assessed in detail. Local correlation (rho ), relative dispersion (RD), fractal dimension (D), perfusion gradients, and histograms of rPBF each measures a different aspect of heterogeneity. In eight anesthetized dogs, lung injury was induced with oleic acid and glass bead injection. Afterward, PEEP of 10-20 cmH2O was instituted. Norepinephrine was infused at 20 cmH2O PEEP. Heterogeneity increased upon lung injury (rho , 0.44 ± 0.09 vs. 0.24 ± 0.09; RD, 0.36 ± 0.06 vs. 0.64 ± 0.12; both P <=  0.05), but fractal dimension remained constant. PEEP did not change rho , RD, or D. Perfusion gradients were reversed after lung injury (right, -27 ± 18 vs. 196 ± 115%; left, -24 ± 18 vs. 282 ± 184%; P <=  0.05). PEEP (10 cmH2O) reduced gradients (116 ± 73 and 143 ± 62%, respectively; P <=  0.05). Norepinephrine, in part, further reduced gradients (right, 50 ± 58%; P <=  0.05; left, 102 ± 94%; P = NS). We conclude that oleic acid- and glass bead-induced lung injury produces abnormal distribution of rPBF. Of these changes, application of PEEP only reverses perfusion gradients.

positive end-expiratory pressure; acute respiratory distress syndrome; microspheres; regional blood flow; oleic acid; dogs

    INTRODUCTION
Top
Abstract
Introduction
Methods
Results
Discussion
References

ACUTE RESPIRATORY DISTRESS syndrome (ARDS) is a major problem in critical care medicine. The syndrome has been reported to be associated with a wide range of mortality rates of between 26 and 70% (21). Mechanical ventilation with positive end-expiratory pressure (PEEP) has long been a standard part of symptomatic therapy (18). In 1986, decreased gas content in computed tomography images of dependent lung areas of ARDS patients was first demonstrated (6, 23). It was shown recently that PEEP of 5-20 cmH2O significantly reduces this maldistribution of tidal volume (7). Furthermore, high levels of PEEP have been shown to reduce mortality from severe ARDS in a retrospective, uncontrolled study (5).

Infusion of oleic acid (OA) may mimic some vascular aspects of ARDS. The combination of OA-induced pulmonary edema and glass micro-beads embolism has been shown to more closely simulate human ARDS (31). In dogs with OA edema, the administration of PEEP (10 cmH2O) normalized the injury-induced redistribution of perfusion to edematous lung regions (25). Norepinephrine is used as a standard therapy for circulatory support in septic patients who often also suffer from ARDS, since norepinephrine increases peripheral vascular resistance, which is decreased during septicemia (28). The effects of norepinephrine on regional pulmonary blood flow (rPBF) distribution in an ARDS model have not yet been studied.

The purpose of the present study was to determine the effects of various levels of PEEP and the combination of PEEP with norepinephrine infusion on experimental ARDS-induced changes of heterogeneity of pulmonary perfusion using absolute blood flow values. Hitherto, rPBF during ARDS and PEEP therapy has been assessed with either indirect methods or by using only some representative samples from the lung rather than the whole lung.

Heterogeneity of rPBF as measured with relative dispersion (RD; SD/mean) is scale dependent, increasing with higher resolution of measurement. Comparison of data between groups or even between experiments may therefore be inappropriate. The increase of apparent heterogeneity with increasing resolution can be measured with fractal dimension (D). In contrast to RD, D is a resolution-independent measure of heterogeneity of regional perfusion.

Heterogeneity may also be viewed on a small-scale level. One may ask the question how perfusion to neighboring lung regions compares. Spatial correlation (rho ) of perfusion measures local heterogeneity (or, vice versa, local self-similarity) of neighboring organ regions.

Both D and rho  have recently been added to the methodological spectrum for assessment of heterogeneity of pulmonary blood flow (9, 10). The effects of experimental lung injury and PEEP ventilation on these parameters have not been studied so far.

The hypotheses were as follows: 1) lung injury increases all measures of heterogeneity of regional pulmonary blood flow; 2) mechanical ventilation with PEEP returns heterogeneity of rPBF to pre-injury values; and 3) norepinephrine further homogenizes blood flow distribution.

    METHODS
Top
Abstract
Introduction
Methods
Results
Discussion
References

Animal Preparation

Eight foxhounds of either sex weighing 17.7 ± 1.5 kg were studied. The study was approved by the governmental animal care and use committee. All animals received care in compliance with the "Guide for the Care and Use of Laboratory Animals" [Department of Health and Human Services Publication No. (NIH) 85-23, Revised 1985]. The dogs were premedicated with 20 mg/kg propiomazine (Combelen; Bayer, Leverkusen, Germany). Anesthesia was induced by intravenous injection of 20 mg/kg pentobarbital sodium (Nembutal; Ceva, Bad Segeberg, Germany), 0.75 mg/kg piritramide (Dipidolor; Janssen, Neuss, Germany), and 0.25 mg/kg alcuronium (Alloferin; Roche, Grenzach-Whylen, Germany) and maintained by continuous infusion of these drugs at a rate of 5, 0.15, and 0.075 mg · kg-1 · h-1, respectively. Fluid losses were replaced with Ringer lactate solution at 5 ml · kg-1 · h-1. The dogs were endotracheally intubated and mechanically ventilated with 12 breaths/min and a tidal volume of 15-18 ml/kg using 100% oxygen (Servo 900C; Siemens-Elema, Solna, Sweden). Core body temperature was maintained by means of a heating pad.

Surgical Preparation

A catheter for monitoring mean arterial pressure was inserted into the left brachial artery and advanced to the descending aorta. A second catheter was advanced to the superior vena cava via the left brachial vein. A Swan-Ganz catheter (7 Fr; Edwards, Anasco, Puerto Rico) was inserted into the pulmonary artery for measurement of core body temperature and pulmonary artery pressure as well as for injection of microspheres and withdrawal of the reference sample during injection of microspheres. Both injection of microspheres and sampling of blood were performed with the distal port of the catheter in the pulmonary artery. A tip manometer was inserted into the right ventricle via the right external jugular vein for measurement of right ventricular pressures. A side-winder catheter (6 Fr; Cordis, Miami, FL) was introduced under fluoroscopic control via the right common carotid artery into the left atrium for measurement of left atrial pressure. Intrathoracic pressure was estimated using a balloon catheter (National Catheter) inserted into the esophagus and positioned at the level of the atrium. After the catheters were inserted, the dogs were placed in the left lateral decubitus position. In the lateral position, esophageal pressure reliably reflects intrathoracic pressure during PEEP (4). Blood loss due to drawing reference samples was isovolemically replaced with blood. Autologous blood was collected by isovolemic hemodilution with 6% Dextran 60 (Macrodex; Schiwa, Glandorf, Germany) to a hematocrit of 28%. The dogs were allowed to stabilize for 30 min after hemodilution. An additional 400-500 ml of blood were obtained from a premedicated, awake donor dog on the day of the experiment. Autologous and homologous blood were mixed, the absence of hemolysis was verified, and the mixture (hematocrit 28 ± 2%) was used for volume substitution and augmentation of intravascular volume during ventilation with PEEP.

Experimental Protocol

Thirty minutes after baseline measurements, the lungs were injured with OA and glass beads. After 70 min and measurements, PEEP of 10 cmH2O was induced, and measurements were taken. Thereafter, the level of PEEP was increased by 5 cmH2O until 20 cmH2O was reached. After each increase, measurements were taken. Subsequently, norepinephrine was infused (with PEEP at 20 cmH2O), and a final set of measurements was made. The dose of OA was 0.01 ml/kg. Glass beads with 100-µm mean diameter were administered until mean pulmonary artery pressure reached 35-40 mmHg. Both OA and glass beads were injected into the right atrium. This technique has been shown to produce a stable ARDS-like syndrome and pulmonary hypertension (31). Norepinephrine was infused at a dose of 0.2-1.0 µg · kg body wt-1 · min-1. During PEEP, transmural right ventricular end-diastolic pressure was kept constant by transfusion of blood (3.6 ± 2.1, 4.5 ± 2.4, and 4.5 ± 1.1 ml/kg at 10, 15, and 20 cmH2O PEEP, respectively). Ten minutes were allowed to elapse after each intravascular volume increase. Subsequently, animals were killed by intravenous injection of saturated potassium chloride. The lungs were removed and dried in room air while being kept inflated at a pressure of 20 cmH2O for 4 days. A time control group was not studied because the course of this experimental ARDS has been studied in detail before (31).

Measurements

Heart rate, right and left ventricular pressures, as well as intrathoracic pressures were monitored continuously on an eight-channel recorder (481, Gould-Brush, Cleveland, OH) and were evaluated at end expiration. To correct for increased absolute intrathoracic pressure during PEEP ventilation, intrathoracic pressure was subtracted from central venous pressure, mean left atrial pressure, and right ventricular pressures to yield transmural pressures. All pressures reported are transmural pressures. Intermittently, thermodilution cardiac output was measured in triplicate (SP 1435, Gould-Statham, Oxnard, CA). Blood gases were measured with an ABL 300 blood-gas analyzer (Radiometer, Copenhagen, Denmark). Stroke volume (SV) was calculated as
SV = cardiac output (ml)/heart rate (ml/min) (1)
Pulmonary vascular resistance (PVR) was calculated as
PVR = (PAP − PCWP)/CO × 80 (dyn ⋅ s ⋅ cm<SUP>−5</SUP>) (2)
where PAP is pulmonary arterial pressure, PCWP is pulmonary capillary wedge pressure, and CO is cardiac output. The factor 80 in Eq. 2 transforms into the SI unit. Intrapulmonary shunt flow fraction (Qs/Qt) was calculated as
<A><AC>Q</AC><AC>˙</AC></A><SUB>s</SUB> /<A><AC>Q</AC><AC>˙</AC></A><SUB>t</SUB> = (Cc′<SUB>O<SUB>2</SUB></SUB> − Ca<SUB>O<SUB>2</SUB></SUB>)/(Cc′<SUB>O<SUB>2</SUB></SUB> − C<OVL>v</OVL><SUB>O<SUB>2</SUB></SUB>) × 100% (3)
where Cc'O2 is pulmonary end-capillary oxygen content calculated as arterial hemoglobin concentration multiplied by the Huefner number 1.39 plus physically disolved oxygen (0.0031 times alveolar PO2). Arterial (CaO2) and mixed venous (C<OVL>v</OVL>O2) oxygen contents were calculated analogously. Microsphere injections were performed as described below at the time of hemodynamic measurements.

Microsphere Methodology

All lungs were completely dissected into an average of 384 ± 77 specimens. Each tissue block was assigned x-, y-, and z-coordinates. Right and left lungs were each cut into four sagittal slices of equal thickness. The numbered slices (1-8) served as x-coordinates. Each slice was dissected into 15 horizontal blocks reaching from posterior to anterior. The blocks were numbered from bottom to top; these numbers were used as y-coordinates. The blocks were cut into 10 tissue samples (z-coordinates). Tissue specimens were prepared to have approximately equal size with the help of a grid drawn on transparent plastic. Smaller sagittal slices were cut into less blocks and smaller blocks into less final samples. Therefore, only an average of 384 samples was obtained rather than the theoretical number of 1,200 (8 slices × 15 blocks × 10 samples).

Regional perfusion was measured with the reference sample method reported by Heymann et al. (17). For injection, microspheres were transferred into glass vials and suspended in saline to give a total volume of 10 ml. Before injection, the vial was agitated for a minimum of 3 min. Microspheres were then injected over 50 s into the right atrium. During injection, the vial was continuously agitated to prevent settling of the spheres. For every measurement of rPBF, 528,000-817,000 (mean, 700,000) microspheres labeled with a randomly selected isotope (95Nb, 114mIn, 51Cr, 141Ce, 103Ru, 46Sc, 16.5 ± 0.1-µm diameter; 46Sc, 16.5 ± 0.2-µm diameter; New England Nuclear-TRAC, Du Pont, Wilmington, DE) were injected. No cardiorespiratory changes were noted after injection of microspheres. Beginning 10 s before microspheres injection and lasting for a total of 3 min, reference samples were drawn from the pulmonary artery at a rate of 3.24 ml/min (pump model 940A; Harvard Apparatus, South Natick, MA). Reference and tissue sample radioactivity were counted for 5 min using a 1,024 channel gamma counter (model 5260; Packard Instruments, Downers Grove, IL) with a 3-in. sodium iodine (thallium drifted) detector. Data were half-life corrected, and cumulated radioactive spectra were separated to obtain activities from single nuclides using the software package MIC III (14) (Department of Surgical Research, Heidelberg, Germany). Sample flow (Qp; ml/min) was calculated for each nuclide as
<A><AC>Q</AC><AC>˙</AC></A><SUB>p</SUB> = <A><AC>Q</AC><AC>˙</AC></A><SUB>pr</SUB> × I<SUB>p</SUB> × I<SUP>−1</SUP><SUB>pr</SUB> (ml/min) (4)
where Qpr is withdrawal rate of the pulmonary reference sample (ml/min), Ipr is counts per minute in the pulmonary reference sample, and Ip is counts per minute in the tissue sample. Qp was normalized to 100 g dry lung wt.

The number of microspheres in each sample was calculated using the specific activity per microspheres. This was determined before the experiments in vitro by assessing the activity of a specimen with a defined number of microspheres of each batch. The number of microspheres in each specimen was counted three times by microscopic examination by different persons, and the results were accepted if the difference between counts did not exceed 1% and microsphere counts were between 600 and 1,000. The gamma activity of specimens was counted for 20 min, and the result was divided by the number of microspheres.

Analysis of Heterogeneity of rPBF

Relative dispersion. To determine global heterogeneity of rPBF, RD (standard deviation/mean) was calculated for each time point in the protocol. By dividing a measure of dispersion (SD) by a measure of central tendency (mean), one expresses the dispersion as a fraction of the mean. Thus it is possible to compare heterogeneity values from different pulmonary blood flow states. Because for calculation of RD all regional blood flow values are averaged, RD represents a global measure of heterogeneity.

Fractal dimension. Fractal properties of pulmonary blood flow heterogeneity were demonstrated by Glenny and Robertson in 1990 (10). A commonly used approach to analyze the fractal nature of pulmonary perfusion is to dissect the lung in small parts and measure the perfusion on this level. Data from these samples are then recombined to obtain information about perfusion of larger samples. If the perfusion is fractal, the natural logarithm of the decreasing heterogeneity of perfusion to larger samples is linearly related to the natural logarithm of the increasing sample size.

Heterogeneity of blood flow to the lung is described with the RD of regional perfusion. The scale of the applied measurement is expressed as the relative mass of the recombined tissue samples (m) in relation to a reference tissue sample mass (m0). Heterogeneity and scale of measurement of rPBF have been shown to be related according to the following equation
(RD<SUB>m</SUB> /RD<SUB>m<SUB>0</SUB></SUB>) = (m/m<SUB>0</SUB>)<SUP>1−D</SUP> (5)
where RDm is the relative dispersion of regional perfusion at some resolution and RDm0 is relative dispersion of regional perfusion when analyzed with the reference resolution (13). The variable D in the exponent of Eq. 5 is the fractal dimension of the perfusion. The slope of the regression line through multiple data points on a plot of
ln (RD<SUB>m</SUB>) vs. ln (m/m<SUB>0</SUB>) (6)
is therefore 1 - D. Such plots are therefore used to determine fractal properties of perfusion of the lung (3, 11, 24).

We calculated D with a computer program developed in C/C++ (Watcom C/C++ compiler v. 10; Powersoft, Concord, MA). For calculation of D, individual samples as well as recombinations to larger tissue units comprising 4, 8, 16, 32, 64, etc., individual samples were used. Recombinations were achieved by adding the microsphere contents of the samples comprised in an aggregate sample and dividing the result by the sum of the weights of samples.

Both the program for calculating spatial correlation and the program for assessing fractal dimension were validated with computer-generated data sets with known D and rho , respectively.

Spatial correlation. In 1992, Glenny (9) suggested rho  of regional pulmonary perfusion as a new tool for characterizing distribution of rPBF. Spatial, or three-dimensional, correlation measures the average relationship of blood flow to lung tissue pieces separated by a certain distance. We chose the smallest distance for calculation, thus obtaining spatial correlation for neighboring pieces. Spatial correlation as a measure for averaged, local heterogeneity of pulmonary blood flow has been shown to change differently upon intervention than global heterogeneity measured with RD (12). For calculation of spatial correlation, all blood flow values from tissue specimens separated by a given distance (multiples of sample size) are treated as data pairs and are used as variables for a three-dimensional extension of the standard correlation formula as reported by Glenny (9). We implemented this formula with a computer program written in C++ (Watcom C/C++ compiler v. 10; Powersoft). For analysis, we used the correlation values of adjacent samples; thus rho  is a measure of local gravity-independent heterogeneity of regional pulmonary perfusion.

Relative gravity-dependent perfusion gradients. Relative gravity dependent perfusion gradients (RPG) were determined separately for each lung as
RPG = (PBF<SUB>bottom</SUB> − PBF<SUB>top</SUB>)/PBF<SUB>top</SUB> × 100% (7)
where PBFtop is mean blood flow in the uppermost quarter of the right and left lungs and PBFbottom is mean blood flow in the most dependent quarter of the right and left lungs. With dogs in the left lateral decubitus position, top and bottom were defined as right and left side of the thorax, respectively.

Histograms. Histograms of pooled rPBF values from all animals were analyzed using the median as well as the 5th and 95th percentiles. rPBF was calculated as shown in Microsphere Methodology.

Statistics

All statistical analyses were performed using the software package SAS v. 6.10 (SAS Institute, Cary, NC). After a significant F value was obtained from analysis of variance, the Student-Newman-Keuls test was applied a posteriori to determine differences between a time point in the protocol and the respective preceding time point. The type 1 error level was set to 5%.

    RESULTS
Top
Abstract
Introduction
Methods
Results
Discussion
References

In Table 1, central hemodynamic data and blood gases are summarized. Induction of experimental ARDS resulted in significant tachycardia, a decrease of stroke volume, and increases of mean pulmonary artery pressure (MPAP), pulmonary vascular resistance, mean right ventricular pressure, and right ventricular maximal rate of pressure increase (RV dP/dtmax). Cardiac output, arterial pressure, and left atrial pressure were unchanged. Ventilation with 20 cmH2O PEEP caused a significant additional rise of MPAP.

                              
View this table:
[in this window]
[in a new window]
  		Table 1.   Cardiovascular and gas exchange parameters

There was a 55% rise of intrapulmonary shunt fraction (Qs/Qt) upon lung injury induction (P <=  0.05), a significant acidosis (P <=  0.05), a 16% fall of arterial PO2 (PaO2) (P < 0.05), and a 28% rise of arterial PCO2 (PaCO2) (P <=  0.05). Mixed venous PO2 (P<A><AC>v</AC><AC>¯</AC></A><SUB>O<SUB>2</SUB></SUB>) was unchanged. Mechanical ventilation with PEEP elicited significant reduction of Qs/Qt (P < 0.05) and elevation of PaO2 (P < 0.05), whereas arterial pH, PaCO2, and P<A><AC>v</AC><AC>¯</AC></A><SUB>O<SUB>2</SUB></SUB> remained unchanged.

Figure 1 shows RD of rPBF (global heterogeneity), D of rPBF (scale-independent global heterogeneity), and rho  of the perfusion of adjacent pulmonary tissue samples (local heterogeneity of rPBF). Global heterogeneity of rPBF was altered by lung injury: RD was 0.36 ± 0.06 at baseline and 0.64 ± 0.12 upon induction of lung injury (P < 0.05). Neither administration of PEEP nor therapy with norepinephrine induced any rehomogenization (Fig. 1). D was stable after lung injury (1.25 ± 0.06 vs. 1.34 ± 0.08; P = NS) and was not influenced by PEEP therapy or norepinephrine. Induction of lung injury decreased rho  from 0.44 ± 0.09 at baseline to 0.24 ± 0.09 (P < 0.05). Therapy with 10, 15, and 20 cmH2O PEEP and additional administration of norepinephrine did not change rho  (Fig. 1).


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1.   Effects of lung injury, different levels of positive end-expiratory pressure (PEEP), and PEEP plus norepinephrine (NE) on local and global heterogeneity of regional pulmonary blood flow (rPBF), analyzed with fractal dimension, spatial correlation, and relative dispersion. Data are means ± SD. b, Baseline; P10, P15, and P20, PEEP of 10, 15, and 20 cmH2O, respectively. * P <=  0.05 vs. preceding time point in protocol.

Figure 2 shows the RPG separately for each lung. At baseline, there was a preferential perfusion of the uppermost quarter of both lungs (right, -27 ± 18%; left, -24 ± 18%). Upon embolization, there was a reversal of this distribution. The RPGs changed to 196 ± 115 and 282 ± 184% for the right and left lung, respectively (P < 0.05). PEEP of 10 cmH2O lowered the gradients of both lungs significantly to 116 ± 73 and 143 ± 62%, respectively, for right and left lungs (P < 0.05). PEEP of 15 and 20 cmH2O did not induce further significant decrease of RPG for both lungs (PEEP 15 cmH2O: right, 103 ± 78%; left, 90 ± 52%; PEEP 20 cmH2O: right, 109 ± 105%, left, 124 ± 58%; P = NS). Infusion of norepinephrine, however, lowered the RPG to 50 ± 58% (P < 0.05) in the right lung, which was not significantly different from baseline. In the dependent left lung, the RPG remained at 102 ± 94% with infusion of the vasoconstricting agent. In Fig. 4, the mean perfusion values of each slice of both lungs are depicted. The perfusion gradients from dependent to nondependent parts of the lungs, and vice versa, can be deduced from Fig. 4.


View larger version (18K):
[in this window]
[in a new window]
  		Fig. 2.   Effects of lung injury, different levels of PEEP, and PEEP plus NE on relative perfusion gradients (RPG) of rPBF. Data are means ± SD. Definitions are as in Fig. 1. * P <=  0.05 vs. preceding time point in protocol for left lung. # P <=  0.05 vs. preceding time point in protocol for right lung.

The cumulated histograms of rPBF from all animals values are shown in Figure 3. The median blood flow was 3.8 ± 0.6 l · min-1 · 100 g-1 at baseline and 3.7 ± 0.9 l · min-1 · 100 g-1 upon lung injury induction (P = NS). Ventilation with PEEP of 10 cmH2O elevated median blood flow to 4.2 ± 2.8 l · min-1 · 100 g-1 (P < 0.05). PEEP of 15 and 20 cmH2O each reduced perfusion (2.6 ± 0.4 and 2.3 ± 0.3 l · min-1 · 100 g-1, respectively; P <=  0.05). Infusion of norepinephrine did not further change blood flow (2.1 ± 0.8 l · min-1 · 100 g-1; P = NS). The configuration of the rPBF histograms was distorted with lung injury (Fig. 3). The 5th percentile (P5) of rPBF was lowered by lung injury. Application of PEEP did not reverse this, but PEEP of 20 cmH2O further reduced the lower border of the histograms. Additional application of norepinephrine increased P5 slightly but consistently. The median lung blood flow was not altered by lung injury. PEEP of 10 cmH2O increased the median blood flow by 0.46 l · min-1 · 100 g-1. Increasing PEEP then reduced median perfusion to a minimum of 2.09 l · min-1 · 100 g-1 as compared with 3.74 l · min-1 · 100 g-1 at baseline. The upper border of the histograms, the 95th percentile, was significantly reduced by 15 cmH2O after an insignificant increase upon lung injury.


View larger version (32K):
[in this window]
[in a new window]
  		Fig. 3.   Effects of lung injury, PEEP, and PEEP plus norepinephrine on histograms of rPBF. Data are cumulated regional blood flow values from 8 dogs. M, median of histogram; P5, 5th percentile of histogram; P95, 95th percentile of histogram. * P <=  0.05 vs. preceding time point in protocol.

    DISCUSSION
Top
Abstract
Introduction
Methods
Results
Discussion
References

The main results of the present study are as follows: 1) lung injury caused an increase of both global and local heterogeneity of rPBF, reversal of gravity-dependent blood flow gradients, and alterations of statistical measures of histograms of rPBF. 2) PEEP and, partly, norepinephrine counteracted changes of perfusion gradients and blood flow histograms. However, injury-induced heterogeneity remained high. Neither the global, resolution-dependent RD nor global, resolution-independent fractal D were influenced by PEEP therapy. Local correlation of blood flow was also unchanged.

Data on heterogeneity of regional pulmonary perfusion during respiratory failure are limited. The prone position has been shown to reduce gravitational gradients of regional pulmonary perfusion (30) and reduce venous admixture (1) in dogs with OA edema. High-level PEEP was demonstrated to improve gas exchange in lavage-induced lung injury (2) and to homogenize density distribution in computed tomography images in ARDS patients (7). In 1994, Schuster and Howard (25) showed with indirect perfusion measurements (positron emission tomography) that PEEP reverses OA-induced redistribution of pulmonary perfusion to edematous tissue. Using the multiple inert gas elimination technique, Matamis et al. (22) noted beneficial effects of 17 ± 2 cmH2O PEEP on ventilation-perfusion distribution in patients with acute respiratory failure. Hedenstierna and co-workers (15, 16) reported on the effects 20 cmH2O PEEP on gravitational gradients in healthy dog lungs by analyzing representative samples rather than whole lungs. In their studies, decreased right ventricular preload was not compensated for, but reduced cardiac output without PEEP did not produce similar patterns of perfusion redistribution (15, 16).

There is one study dealing with distribution of rPBF in OA injury in dogs (27). The authors have calculated peripheral to central perfusion gradients. Early after OA infusion they found a reduction of peripheral perfusion. This seems to be in contrast to our data. We observed an increase in peripheral perfusion in the dependent lung and a decrease of peripheral blood flow in the nondependent lung (Fig. 4). Tarver and colleagues (27) studied only representative samples of one lobe of one lung. We studied the complete lung and observed heterogeneous distribution changes including effects similar to those observed by Tarver et al. We therefore cannot definitely conclude whether our results are in contrast to or in agreement with those of the cited study. Another important difference between our study and that of Tarver et al. is that in the former, animals were placed in the left lateral decubitus position and in the latter the prone position was used. The time frame in both studies is very different. We induced OA-glass bead lung injury over a time that was certainly much longer than the time used in Tarver's study. The latter study was explicitly designed to investigate early effects of OA infusion, whereas we established a more chronic form of lung injury. In conclusion, we think that our results are not necessarily in contrast to those published by Tarver and colleagues.


View larger version (33K):
[in this window]
[in a new window]
  		Fig. 4.   rPBF gradients at baseline, lung injury, PEEP of 10-20 cmH2O, and PEEP plus norepinephrine. Each lung was separated in 4 sagittal slices (dogs were in left lateral decubitus position). Average blood flow to each slice is plotted against its vertical position in lying dog. Data are cumulated from 8 dogs.

Spatial Correlation, Relative Dispersion, and Fractal Dimension

Spatial correlation (rho ) of perfusion of adjacent lung tissue samples is a measure of local similarity of blood flow (9). Lung injury decreased rho  notably (0.44 ± 0.09 vs. 0.24 ± 0.09) (Fig. 1). This is in contrast to findings of Schuster and Howard (25) who reported a redistribution of perfusion to the most edematous lung regions induced by OA after application of 10 cmH2O PEEP. Global heterogeneity of rPBF greatly increased upon lung injury (RD = 0.36 ± 0.06 vs. 0.64 ± 0.12) (Fig. 1) and was influenced neither by PEEP nor by infusion of norepinephrine. The lack of effect of the treatment on these three parameters of heterogeneity is an important finding. From previous data on OA edema, a positive effect of PEEP on distribution of rPBF should have been expected (25). When assessed with more specific methods, however, heterogeneity proved uninfluenced. Therefore, RD and rho  are valuable parameters of rPBF distribution in this model of lung injury. For testing other therapeutic strategies, it may be useful to use them to show superiority of new concepts.

Relative Perfusion Gradients

At baseline, we found preferential perfusion of the nondependent parts of the lung, which is in agreement with recent findings by Hlastala et al. in horses (19). Relative perfusion gradients were calculated separately for each lung. The dependent part of the left lung was the part adjacent to the lateral thoracic wall, and the nondependent parts were mediastinal regions. For the right lung (superior to the left lung because of the left lateral decubitus position), the dependent part was the mediastinal part and the nondependent part corresponded to those regions next to right lateral thoracic wall. Induction of lung injury reversed RPGs and greatly enhanced the extent of gravity-dependent perfusion inhomogeneity in both lungs (Fig. 2). The difference of the bottom and top parts (left lateral decubitus position of the dogs) of the lungs amounted to two to three times the blood flow in the uppermost parts of the lungs. PEEP of 10 cmH2O counteracted these gradients significantly. This is in contrast to the reported effects of PEEP on the noninjured lung in dogs in which 20 cmH2O PEEP resulted in marked reduction of perfusion of the uppermost regions of the lung (15, 16). Similarily, PEEP (10 cmH2O) redirected perfusion to dependent regions of the lung in a model of OA-induced single-lung injury (20).

In the right (nondependent) lung, norepinephrine further redistributed perfusion to nondependent areas of the lung, thus rendering the resulting RPG not significantly different from baseline values despite the presence of lung injury and 20 cmH2O PEEP. The catecholamine was infused in a central vein. The lung is known to clear 30% of this drug during a single passage (8, 26); however, this may or may not be the case in the injured lung. As can be concluded from unchanged values for pulmonary vascular resistance (Table 1), norepinephrine did not vasoconstrict pulmonary vessels. There was, however, a trend toward increased right ventricular contractility (RV dP/dtmax, see Table 1) and increased cardiac output. The effects of norepinephrine infusion on RPGs may therefore probably be explained by a recruitment of pulmonary vessels. Norepinephrine might be of benefit when combined with PEEP in attempts to normalize disturbed gradients of pulmonary blood flow in ARDS. However, norepinephrine failed to further improve Qs/Qt or PaO2 in our study and may also cause increased cardiovascular stress, which may be undesirable in critically ill patients.

Histograms of rPBF Values

At baseline, the histogram of rPBF values (Fig. 3) was close to a Gaussian distribution as judged by inspection. Such a normal distribution was also found by Walther et al. (29) in awake sheep. Upon lung injury, the number of values close to zero and the number of extremely high values were increased, which resulted in a left-shifted and skewed histogram. The most prominent change upon ventilation with PEEP (15 cmH2O) was the reduced frequency of extremely high values. Thus the rPBF histogram approached prelung injury configuration upon therapy with PEEP. It appears that PEEP therapy removed the tail from the rPBF distribution and added to the frequency of values close to the mean. The median value was reduced by higher levels of PEEP, but the fifth percentile was not greatly reduced; thus the lower median seems due to less high values. Considering improved gas exchange and reduced shunt fraction, we interpret these changes upon PEEP as improvement of distribution of rPBF in this model of lung injury.

Conclusion

In the present study, we provide the first data on spatial correlation and fractal dimension of regional pulmonary blood flow in a model of acute lung injury and PEEP therapy. Our data show that most parameters of heterogeneity are greatly changed by lung injury. PEEP therapy normalizes only the perturbations of perfusion gradients and some of the statistical measures of rPBF histograms. However, neither of the new parameters of blood flow distribution (fractal dimension, spatial correlation) that were changed by lung injury were normalized by PEEP therapy. In addition to restoring ventilation to collapsed or fluid-filled alveoli that are still perfused, PEEP may contribute to improved oxygenation in patients with ARDS by improving disturbed perfusion gradients.

    ACKNOWLEDGEMENTS

The valuable technical assistance of R. Schwarz, C. Schwickert, and P. Spengler in performing these experiments is gratefully acknowledged. We are indebted to Dr. R. Schosser for the expert support in microsphere measurements.

    FOOTNOTES

Address for reprint requests: M. Kleen, Institute for Surgical Research, Klinikum Grosshadern, Ludwig-Maximilians-University, Marchioninistr. 15, 81366 Munich, Germany.

Received 16 April 1997; accepted in final form 9 September 1997.

    REFERENCES
Top
Abstract
Introduction
Methods
Results
Discussion
References

1.   Albert, R. K., D. Leasa, M. Sanderson, H. T. Robertson, and M. P. Hlastala. The prone position improves arterial oxygenation and reduces shunt in oleic-acid-induced acute lung injury. Am. Rev. Respir. Dis. 135: 628-633, 1987[Medline].

2.   Barnas, G. M., P. L. Donahue, C. S. Kong, I. G. Ryder, D. H. Choi, B. H. Hoff, and C. F. Mackenzie. Effects of PEEP on acinar gas transfer in healthy and lung-injured dogs. Am. J. Respir. Crit. Care Med. 152: 1241-1247, 1995[Abstract].

3.   Caruthers, S. D., and T. R. Harris. Effects of pulmonary blood flow on the fractal nature of flow heterogeneity in sheep lungs. J. Appl. Physiol. 77: 1474-1479, 1994[Abstract/Free Full Text].

4.   Craven, K. D., and L. D. H. Wood. Extrapericardial and esophageal pressures with positive end-expiratory pressure in dogs. J. Appl. Physiol. 51: 798-805, 1981[Abstract/Free Full Text].

5.   DiRusso, S. M., L. D. Nelson, K. Safcsak, and R. S. Miller. Survival in patients with severe adult respiratory distress syndrome treated with high-level positive end-expiratory pressure. Crit. Care Med. 23: 1485-1496, 1995[Medline].

6.   Gattinoni, L., D. Mascheroni, A. Torresin, R. Marcolin, R. Fumagalli, S. Vesconi, G. P. Rossi, F. Rossi, S. Baglioni, F. Bassi, G. Nastri, and A. Pesenti. Morphological response to positive end expiratory pressure in acute respiratory failure. Computerized tomography study. Intensive Care Med. 12: 137-142, 1986[Medline].

7.   Gattinoni, L., P. Pelosi, S. Crotti, and F. Valenza. Effects of positive end-expiratory pressure on regional distribution of tidal volume and recruitment in adult respiratory distress syndrome. Am. J. Respir. Crit. Care Med. 151: 1807-1814, 1995[Abstract].

8.   Ginn, R., and J. R. Vane. Disappearance of catecholamines from the circulation. Nature 219: 740-742, 1968[Medline].

9.   Glenny, R. W. Spatial correlation of regional pulmonary perfusion. J. Appl. Physiol. 72: 2378-2386, 1992[Abstract/Free Full Text].

10.   Glenny, R. W., and H. T. Robertson. Fractal properties of pulmonary blood flow: characterization of spatial heterogeneity. J. Appl. Physiol. 69: 532-545, 1990[Abstract/Free Full Text].

11.   Glenny, R. W., and H. T. Robertson. Fractal modeling of pulmonary blood flow heterogeneity. J. Appl. Physiol. 70: 1024-1030, 1991[Abstract/Free Full Text].

12.   Glenny, R. W., and H. T. Robertson. A computer simulation of pulmonary perfusion in three dimensions. J. Appl. Physiol. 79: 357-369, 1995[Abstract/Free Full Text].

13.   Glenny, R. W., H. T. Robertson, S. Yamashiro, and J. B. Bassingthwaighte. Applications of fractal analysis to physiology. J. Appl. Physiol. 70: 2351-2367, 1991[Abstract/Free Full Text].

14.   Gross, W., R. Schosser, and K. Messmer. MIC-III: an integrated software package to support experiments using the radioactive microsphere technique. Comput. Programs Biomed. 33: 65-85, 1990.

15.   Hedenstierna, G., and F. C. White. Spatial distribution of pulmonary blood flow in the dog with PEEP ventilation. J. Appl. Physiol. 47: 938-946, 1979[Abstract/Free Full Text].

16.   Hedenstierna, G., F. C. White, R. Mazzone, and P. D. Wagner. Redistribution of pulmonary blood flow in the dog with PEEP ventilation. J. Appl. Physiol. 46: 278-287, 1979[Abstract/Free Full Text].

17.   Heymann, M. A., B. D. Payne, J. I. E. Hofmann, and A. M. Rudolph. Blood flow measurements with radionuclide-labeled particles. Prog. Cardiovasc. Dis. 20: 55-79, 1977[Medline].

18.   Hickling, K. G. Ventilatory management of ARDS: can it affect the outcome? Intensive Care Med. 16: 219-226, 1990[Medline].

19.   Hlastala, M. P., S. L. Bernard, H. H. Erickson, M. R. Fedde, E. M. Gaughan, R. McMurphy, M. J. Emery, N. Polissar, and R. W. Glenny. Pulmonary blood flow distribution in standing horses is not dominated by gravity. J. Appl. Physiol. 81: 1051-1061, 1996[Abstract/Free Full Text].

20.   Johnston, W. E., J. Vinten-Johansen, and E. Tommasi. Sodium nitroprusside and positive end-expiratory pressure are not detrimental in canine asymmetric pulmonary edema. Crit. Care Med. 17: 241-246, 1989[Medline].

21.   Lemaire, F. The prognosis of ARDS: appropriate optimism? Intensive Care Med. 22: 371-373, 1996[Medline].

22.   Matamis, D., F. Lemaire, A. Harf, B. Teisseire, and C. Brun-Buisson. Redistribution of pulmonary blood flow induced by positive end-expiratory pressure and dopamine infusion in acute respiratory failure. Am. Rev. Respir. Dis. 129: 39-44, 1984[Medline].

23.   Maunder, R. J., W. P. Shuman, J. W. McHugh, S. I. Marglin, and J. Butler. Preservation of normal lung regions in the adult respiratory distress syndrome. Analysis by computed tomography. JAMA 255: 2463-2465, 1986[Abstract].

24.   Parker, J. C., J. L. Ardell, C. R. Hamm, S. A. Barman, and P. J. Coker. Regional pulmonary blood flow during rest, tilt, and exercise in unanesthetized dogs. J. Appl. Physiol. 78: 838-846, 1995[Abstract/Free Full Text].

25.   Schuster, D. P., and D. K. Howard. The effect of positive end-expiratory pressure on regional perfusion during acute lung injury. J. Crit. Care 9: 100-110, 1994[Medline].

26.   Sole, M. J., M. Dobrac, L. Schwartz, M. N. Hussain, and E. F. Vaughan-Neil. The extraction of circulating catecholamines by the lungs in normal man and in patients with pulmonary hypertension. Circulation 60: 160-163, 1979[Abstract/Free Full Text].

27.   Tarver, R. D., J. Tsai, L. W. Hedlund, D. C. Sullivan, M. M. Lischko, C. C. Harris, E. L. Effmann, and C. E. Putman. Regional pulmonary distribution of iodine-125-labeled oleic acid. Its relationship to the pattern of oleic acid edema and pulmonary blood flow. Invest. Radiol. 21: 102-107, 1986[Medline].

28.   Vincent, J. L., A. Roman, and R. J. Kahn. Dobutamine administration in septic shock: addition to a standard protocol. Crit. Care Med. 18: 689-693, 1990[Medline].

29.   Walther, S. M., K. B. Domino, R. W. Glenny, N. L. Polissar, and M. P. Hlastala. Pulmonary blood flow distribution has a hilar-to-peripheral gradient in awake, prone sheep. J. Appl. Physiol. 82: 678-685, 1997[Abstract/Free Full Text].

30.   Wiener, C. M., W. Kirk, and R. K. Albert. Prone position reverses gravitational distribution of perfusion in dog lungs with oleic acid-induced injury. J. Appl. Physiol. 68: 1386-1392, 1990[Abstract/Free Full Text].

31.   Zwissler, B., H. Forst, and K. Messmer. A new experimental model of ARDS and pulmonary hypertension in the dog. Res. Exp. Med. 189: 427-438, 1989[Medline].

AJP Heart Circ Physiol 274(1):H209-H216
0363-6135/98 $5.00 Copyright © 1998 the American Physiological Society



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kleen, M.
Right arrow Articles by Messmer, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kleen, M.
Right arrow Articles by Messmer, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online