JOURNAL

OF SURGICAL

RESEARCH

60,111-118

Platelet-Activating KARL

Departments

(1991)

Factor in Porcine Pseudomonas

Acute Lung Injury’

BYRNE, M.D., CURTIS N. SESSLER, M.D., P. DECLAN CAREY, FRCSI, TIMOTHY D. SIELAFF, ALBERTO VASQUEZ, M.D., JAMES L. TATUM, M.D., JERRY I. HIRSCH, PHARM.D., AND HARVEY J. SUGERMAN, M.D. of Surgery,

Medicine

and Radiology,

Medical

Submitted

College

of Virginia,

for publication

July

Virginia

Commonwealth

University,

Richmond,

M.D.,

Virginia

28, 1989

INTRODUCTION We investigated the role of platelet-activating factor (PAF) in acute septic lung injury by examining the effects of the selective PAF antagonist SRI 63-675 and by measuring PAF in lung tissue in the porcine model. Four groups of pigs (15-25 kg) were studied: saline control (C, n = 5); Pseudomonas (Ps, n = 9), given 5 X 10s CFU/ml at 0.3 ml/20 kg/min intravenously over 1 hr; SRI (n = 3), given SRI 63-675 in a 40 mg/kg bolus; and SRI + Ps (n = 5). Ps infusion produced a fulminant lung injury characterized by a threefold increase in pulmonary arterial pressure at 30 min and persistent pulmonary hypertension (P < 0.05 vs C), a significant (P < 0.05 vs C) decrease in arterial oxygen tension (PaO,) from 60 min, a significant (P < 0.05 vs C) increase in extravascular lung water (EVLW) from 120 min, and a significant (P < 0.05 vs C) increase in albumin flux determined scintigraphically as slope index at 150-180 min. Systemic arterial pressure and cardiac index (CI) decreased significantly (P < 0.05) in the Ps group vs C at 60 and 180 min, respectively. Bolus injection of SRI 63-675 at the time of Ps infusion blocked the early pulmonary hypertension, attenuated the early and late fall in PaO,, ameliorated the increase in EVLW, and prevented the late (150-180 min) increase in albumin flux. SRI 63-675 had minimal effects on Ps-induced hypotension or alterations in CL The quantity of PAF bioactivity in lung tissue as determined by rabbit platelet serotonin release was significantly (P < 0.05) greater in Ps-infused animals than in saline controls. Collectively these results suggest that PAF plays a major role in septic lung injury. 0 1991 Academic PRESS, 1~.

The infusion of live Pseudomonas aeruginosa bacteria into experimental animals produces fulminant acute lung injury characterized by acute pulmonary hypertension, hypoxemia, increased pulmonary microvascular permeability with increased transvascular flux of solute and water, and systemic hypotension [6, 26, 271. The pathophysiology of sepsis-associated lung injury is complex and involves activation of circulating and resident intravascular cells as well as endothelial cells [4, 141. Various key inflammatory mediators are synthesized and released into the circulation producing changes in vascular and airway tone and microvascular permeability. Prominent among these mediators are the cyclooxygenase metabolites of the arachidonic acid cascade that produce bronchoconstriction and pulmonary arteriolar vasoconstriction [21, 341, serotonin and histamine [5, 24, 31,441, cytokines such as tumor necrosis factor and interleukin-1 [16, 22, 32, 43, 50],‘neutrophil-derived reactive oxygen species [12, 18, 231, and lysosomal enzymes [ 151. Recent work has implicated the biologically active lipid, platelet activating factor (1-0-alkyl-2-acetyl-snglyceryl-3-phosphocholine) (PAF) as a key mediator in the physiologic derangements in sepsis in various animal models [l, 7, 9, 10, 11, 13, 17, 19, 25, 40, 491. This substance is synthesized and released from the cell membrane phospholipid component of basophils, platelets, neutrophils, mast cells, endothelial cells, and macrophages following inflammatory and immunologic stimuli [36, 391. PAF activates many inflammatory cells, promotes phagocyte chemotaxis and aggregation, and causes enhanced synthesis and release of vasoactive substances, superoxide anion, and lysosomal enzymes. PAF alters endothelial cell cytoskeleton, leading to intercellular gap formation and increased permeability to albumin [8]. Infusion of synthetic PAF in animal models produces pulmonary and hemodynamic alterations similar to those of endotoxemia or bacterial sepsis [3, 71. PAF has been previously detected by bioassay in blood,

i These experiments were supported in part by Department of the Army (Contract DAMD 17-86-C-6168), The American Heart Association (Virginia affiliate), and the Virginia chapter of the American Lung Association and were performed in accordance with the NIH guidelines for the use of experimental animals. SRI 63-675 was kindly provided by Dean Handley, Ph.D., of the Sandoz Research Institute, East Hanover, NJ. The views, opinions, and/or findings contained herein are those of the authors and should not be construed as an official Department of the Army position, policy, or decision unless so designated by other documentation. 111

0022.4804/91$1.50 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

112

JOURNAL

OF

SURGICAL

RESEARCH:

various biological fluids, and lung homogenate following bacteremia [25] and endotoxin injection [13, 171 in rats and following endotoxin in sheep [40]. Recently the development of PAF receptor antagonists has allowed further direct investigation into the role of PAF in sepsisassociated and endotoxin-induced acute lung injury. Chang et al. demonstrated that pretreatment with two structurally different PAF receptor antagonists attenuated systemic hypotension, prevented hypoxemia, and increased the transcapillary protein flux following endotoxin administration to rats [ 131. Handley et al. reported similar effects with the use of an endotoxin model with PAF antagonist SRI 63-072 [20], and Adnot et al. demonstrated comparable results with a similar PAF antagonist in an endotoxin guinea pig model [l]. In the sheep lung lymph fistula model, infusion of the PAF antagonist SRI 63-441 blocked the early pulmonary hypertension and hypoxemia and attenuated the high flow of protein-rich lung lymph (interpreted as reflecting increased vascular permeability) whereas the PAF antagonist ONO-6240 was less effective [42, 491. To further investigate the role of PAF in the physiologic derangements seen after sepsis, SRI 63-675, a [tetrahydrofluranyl(methoxyphosphomyloxy) ethyl] quinolium selective PAF antagonist, was administered to anesthetized septic pigs, and hemodynamic, gas exchange, and permeability parameters were measured. Finally we quantified PAF bioactivity in lung homogenates from septic and control pigs. MATERIALS

AND

METHODS

Animal preparation. Young swine, weighing 15-30 kg, were anesthetized with intramuscular ketamine hydrochloride (25 mg/kg) and atropine (0.4 mg) and placed supine. Sodium pentobarbital (30 mg/kg) was administered intravenously to induce anesthesia, and endotracheal intubation with a cuffed endotracheal tube (National Catheter) was performed. Anesthesia was maintained throughout the study with bolus injections of pentobarbital as necessary. All animals were paralyzed with a continuous intravenous infusion of pancuronium bromide (0.2 mg/min) and mechanically ventilated with a Harvard large animal ventilator (Harvard Apparatus, Boston, MA), using 0.5 FiO, and 5 cm H,O positive-end expiratory pressure and 20 ml/kg tidal volume. Respiratory frequency was adjusted in all animals to give a PaCO, of 35-40 Torr at the beginning of each experiment. Indwelling catheters were placed in the left common carotid artery for systemic arterial pressure (SAP) monitoring and arterial blood gas determination, left and right external jugular veins for infusion of normal saline or Pseudomonas organisms, and technetium99m-labeled human serum albumin (Tc-HSA). An indwelling balloon tipped pulmonary artery catheter was positioned via pressure monitoring for measurement of

VOL.

50,

NO.

2, FEBRUARY

1991

pulmonary arterial pressure (PAP), pulmonary artery occlusion pressure (PAOP), and thermodilution cardiac output (Model K3500E, Mansfield Scientific). Systemic arterial blood samples were analyzed for pH, PaO,, and PaCO, using a pH/blood gas analyzer (Model 1103, Instrument Laboratories). Thermal-Cardiogreen extravascular lung water (EVLW). A No. 5 French lung water catheter (Model 96-020-5F, American Edwards Laboratories) was positioned in the lower abdominal aorta via the left femoral artery for measurement of EVLW. In this technique [ 28],10 ml of green dye solution (5 mg indocyanine green dye in 10 ml 5% dextrose) at 4°C is rapidly infused as a bolus through the proximal (central venous) port of the indwelling pulmonary artery catheter while blood is simultaneously withdrawn through the thermistor-tipped femoral artery catheter and a densitometer cuvette (Model 402A, Waters Instruments, Inc.) linked to a lung water computer (Model 9310, American Edwards Laboratories). The lung water computer measures mean transit times of the intravascular dye (MTD) and the freely diffusible thermal component (MTT) as well as the cardiac output (CO). EVLW is calculated according to: EVLW = CO(MTD-MTT)/body weight in kilograms. Gamma scintigraphy. Scintigraphically determined transpulmonary flux of Tc-HSA was used as an indicator of pulmonary capillary permeability [46-481. In this technique, animals are placed supine beneath a Searle Pho-Gamma V scintillation camera fitted with a low energy parallel hole collimator linked with a Digital Equipment Corp. (DEC) mobile acquisition system. Tc-HSA is infused via the right external jugular vein at 60 min (10 mCi) and 105 min (7 mCi). Data are collected at 1-set intervals for 60 set by the gamma camera for scintigraphic definition of the heart and lungs and at I-min intervals thereafter. The counts are stored on floppy disk, transferred to magnetic tape, and regions of interest (i.e., right lung and heart) are selected using a DEC gamma II medical computer system. Lung:heart radioactivity ratios are constructed using a VAX 8600 computer. The slope index (SI) is calculated by least-squares linear regression analysis over a 30-min period after a 15-min postinjection delay to permit isotope equilibration. SIs are reported for the 75- to 105- and 150- to 180-min periods after baseline. Measurement of platelet activating factor in lung homogenates: Lung tissue collection and processing. At the conclusion of the experiment and immediately following sacrifice of the animal, a left thoracotomy was performed, and a 2- to 3-g sample of the left lung was excised, placed in a tared container, weighed, and quickly placed into 25 ml of ice-chilled methanol. The mixture was homogenized for 60 set using a tissumizer (Tekmar Co., Cincinnati, OH) and frozen at -20°C until

BYRNE

ET

AL.:

PAF

IN

assay. Lung samples underwent Bligh-Dyer lipid extraction [2] followed by isolation and purification using thin-layer chromatography (TLC) on glass plates precoated with silica gel G. Authentic PAF prepared from bovine heart lecithin (Sigma) was spotted on a separate lane. Zones from sample lanes that comigrated with authentic PAF (Sigma) were scraped, and the samples dried under nitrogen and resuspended in buffer. The final sample was resuspended in Krebs-Ringer phosphate dextrose buffer with 0.25% bovine serum albumin. Platelet-activating factor assay. Bioassay for PAF was performed by measuring the release of [3H]serotonin from rabbit platelets [ 131. Blood was collected from the central ear artery of New Zealand white rabbits and added to acid citrate dextrose. The blood was centrifuged at 300g for 20 min, and the platelet-rich plasma was removed and incubated for 30 min with 1 @i of [3H]serotonin (New England Nuclear, Boston, MA) at 37°C. After being washed twice in Tyrode’s gelatin, the platelets were resuspended in Tyrode’s gelatin without calcium to a concentration of 1 X 10’ (determined by absorbance at 530 nm). Each assay was performed in a reaction tube that contained various dilutions of the sample in a 450~~1 volume of Tyrode’s gelatin with calcium to which 50 /*l of the platelet preparation was added. After 90 set of incubation at room temperature, the reaction was stopped by the addition of 20 ~1 of 9% formaldehyde. After centrifugation at 2500g for 15 min, 100 ~1 of the supernatant was removed and added to 10 ml of scintillant. The samples were counted for radioactivity in a liquid scintillation counter (Beckman Delta 300, Beckman Instruments, Berkley, CA). The reported counts per minute were converted to disintegrations per minute using external standards ratio and a quench curve for that counter. The platelets in one reaction tube containing only buffer and platelets were lysed by Triton X-100 and counted to determine the maximal radioactivity in the batch of platelets. Another tube contained only buffer, to determine spontaneous release. The bioactivity in each sample was expressed as percentage of release and calculated according to percentage release = (sample-spontaneous)/(maximum-spontaneous). Percentage release was converted to picograms of PAF after comparison with a standard curve of serotonin release elicited by known amounts of synthetic PAF (Sigma). All samples were run in triplicate with one of these tubes also containing the selective PAF receptor antagonist WEB 2086 (gift from Boehringer Ingelheim, Ridgefield, CT), at a final concentration of lop5 M, to confirm specificity of the assay for PAF. Experimental groups. The control group (C, n = 5) received sterile NaCl infusion in a volume equivalent to the volume of Pseudomonas and SRI 63-675. Careful attention to fluid balance was maintained in all groups throughout the study, and normal saline was infused where necessary to maintain the PAOP between 5 and

PORCINE

Pseudomonas

!b

113

AL1

o-oContro l -*Pseudomonas

*p ( 0.05 Ps “S c %I ( 0.05 SRI + Ps vs Ps

6b

A-ASRI 120

Time

IA0

+ Ps 2ko

(min)

FIG. 1. The early rise in pulmonary artery pressure domonas group was attenuated by the SRI 63-675. The pulmonary hypertension was not improved.

in the Pseulate phase of

10 mm Hg. The Pseudomonas group (Ps, n = 9) was infused continuously for 1 h with live P. aeruginosa (5 X 10’ organisms/ml at 0.3 ml/20 kg/min). SRI 63-675 (SRI, n = 3) animals received SRI 63-675 in a single 40 mg/kg iv bolus at the start of the saline infusion. Five animals received a coinfusion of Pseudomonus and SRI 63-675 (Ps + SRI) at the doses and times described above. SAP, PAP, PAOP, cardiac index (CI), and arterial blood gases were measured before infusion and at 30-min intervals thereafter. EVLW measurements were taken every 60 min until the end of the study. Scintigraphic data were collected and analyzed as described above. Lung homogenates were obtained from four control and five Pseudomonas-infused animals. Statistics. The results are expressed as means f SEM. Differences were tested for significance using ANOVA and repeated measures ANOVA for betweenand within-group differences. Differences between means were analyzed using Tukey’s Studentized range test or Student’s t test (PAF in lung homogenate). The level of statistical significance was set at P < 0.05. RESULTS

Hemodynamic variables. No significant changes from baseline were observed in any of the measured physiologic variables in saline control animals or those given SRI 63-675 alone. It was noted, however, on blood withdrawal for sampling, that SRI 63-675 had caused slight hemolysis. Infusion of Pseudomonas organisms produced profound physiologic alterations. Animals in the septic group showed a significant (P < 0.05) increase in PAP (Fig. 1) immediately following initiation of the Pseudomonas infusion. PAP fell slightly from its initial peak but remained significantly elevated compared to that of control animals throughout the entire study period. Treatment with SRI 63-675 prevented Pseudomonas-induced pulmonary hypertension during the first 90 min but had no effect thereafter. Systemic arterial pressure (Fig. 2) increased initially from baseline and

114

JOURNAL

150

OF

SURGICAL

RESEARCH:

o--oControl 0-e Pseudomonas a-~% +Ps

1

VOL.

50,

‘z‘

NO.

2, FEBRUARY

1991

250

I” E 125 L 4 m

100

6 2

75

lp

0

FIG. improved

2.

*o ( 0.05 Ps “B c “p ( 0.05 SRI + Ps “S Ps I I 0 60

( 0.05 Ps YS c ‘p ( 0.05 SRI + Ps vs Ps 60

Systemic hypotension by SRI 63-675.

120 Time (mln)

180

240

Time

in the Pseudomonas

group

was not

then declined progressively from 60 min in the Ps group becoming significantly different from controls at 60 min and thereafter. Cardiac index (Fig. 3) began to deteriorate significantly in the septic group compared to that in controls at 180 min and continued to decline until the end of the study. Treatment with SRI 63-675 had no beneficial effect on the deterioration in either SAP or CI. Oxygenation and lung injury. SRI 63-675 alone had no deleterious effect on systemic PaO,. Systemic PaO, (Fig. 4) decreased progressively in the Ps group and became significantly lower than controls at 60 min and thereafter. In the SRI + Ps group, PaO, was maintained at control levels until 240 min and was significantly higher than that in Ps alone at the 150- and 210-min time points. EVLW (Fig. 5) in control animals remained at or near baseline levels throughout the entire study. EVLW in septic animals was significantly higher than that in controls at 60 min and increased progressively thereafter. EVLW was significantly lower in the SRI + Ps group than in the Ps alone group at 60 min, but thereafter values remained at levels between control and Ps alone groups. Gamma scintigraphy. The slope index in control animals did not change significantly between the 75- to 105-