PentaLyte Decreases Lung Injury after Aortic Occlusion ... - ATS Journals

PentaLyte Decreases Lung Injury after Aortic Occlusion–Reperfusion ROBERT N. AXON, MANUEL S. BAIRD, JOHN D. LANG, AMY E. BRIX, and VANCE G. NIELSEN Departments of Anesthesiology and Comparative Medicine, and the Center for Free Radical Biology, University of Alabama at Birmingham, Birmingham, Alabama

Lung injury often occurs after hepatoenteric ischemia, with xanthine oxidase (XO, an oxidant-generating enzyme), released from reperfusing liver and intestines, mediating a significant component of this injury. Since pentastarch administration decreases intestinal reperfusion injury, we determined whether resuscitation with PentaLyte (a pentastarch-containing solution) would decrease hepatoenteric reperfusion injury, xanthine oxidase release, and concomitant lung injury after aortic occlusion– reperfusion. Aortic occlusion was established in rabbits for 40 min, and was followed by 3 h of reperfusion, during which either PentaLyte or lactated Ringer’s solution-based resuscitation was administered. Sham-operated animals served as controls. Hepatoenteric reperfusion injury, as manifested by release of the enzyme aspartate aminotransferase and decreased gastric intramucosal pH, was significantly (p , 0.0167) attenuated by PentaLyte administration after aortic occlusion–reperfusion, as compared with its occurrence in animals given lactated Ringer’s solution. The release of XO after aortic occlusion–reperfusion was 4-fold smaller after PentaLyte administration than after resuscitation with lactated Ringer’s solution (p , 0.05). Pulmonary injury, as defined by an increase in bronchoalveolar lavage fluid (BALF) protein content and lactate dehydrogenase (LDH) activity, was 4-fold less after PentaLyte administration following aortic occlusion–reperfusion than after administration of lactated Ringer’s solution (p , 0.05). We conclude that remote pulmonary injury is significantly decreased by concomitant PentaLyte-mediated reduction of hepatoenteric reperfusion injury and XO release. Axon RN, Baird MS, Lang JD, Brix AE, Nielsen VG. PentaLyte decreases lung injury after aorAM J RESPIR CRIT CARE MED 1998;157:1982–1990. tic occlusion–reperfusion.

Postoperative lung injury is frequent after trauma and major vascular surgery (1, 2). Hepatoenteric ischemia commonly occurs in these settings as a result of overt shock or as a consequence of surgical interventions (e.g., aortic cross-clamping). The release of xanthine oxidase (XO), an oxidant-generating enzyme, after hepatoenteric ischemia–reperfusion plays a major role in the evolution of remote lung and heart injury (3–8). The liver and intestine contain the highest tissue activity of XO and its precursor, xanthine dehydrogenase (9). Both forms of the enzyme will be referred to as xanthine oxidase for the remainder of this report. In addition to experiencing ischemia and subsequent oxidant stress, trauma and vascular-surgery patients share another common perioperative experience: resuscitation. Of interest in this regard is that hydroxyethyl starch macromolecules have been documented to reduce reperfusion injury in several animal models (10–15). Notably, pentastarch solutions have attenuated intestinal reperfusion injury (15). Mechanisms invoked to explain hydroxyethyl starch-medi(Received in original form August 21, 1997 and in revised form January 29, 1998) Supported in part by National Institutes of Health grant HL P01 48676 and by a grant from BioTime, Inc., Berkeley, CA. Correspondence and requests for reprints should be addressed to Vance G. Nielsen, M.D., Department of Anesthesiology, The University of Alabama at Birmingham, 619 South 19th Street, Birmingham, AL 35233-6810. E-mail: vance. [email protected] Am J Respir Crit Care Med Vol 157. pp 1982–1990, 1998 Internet address: www.atsjournals.org

ated tissue protection include a capillary “sealant” effect that prevents tissue edema (10, 11) and reduced neutrophil adherence to endothelial cells (13). Pentastarch scavenges hydroxyl radical (?OH) in vitro (16), raising the possibility that hydroxyethyl starches may be antioxidants. Taken as a whole, these findings provide a rational basis for the hypothesis that a pentastarch solution could attenuate hepatoenteric ischemia–reperfusion injury, decrease circulating XO activity, and concomitantly reduce lung injury after aortic occlusion–reperfusion. The purpose of the present study was to determine whether an infusion of PentaLyte (BioTime, Inc., Berkeley CA)—a 6% pentastarch solution containing balanced electrolytes (e.g., sodium, chloride, calcium, magnesium, and potassium), glucose, and lactate buffer—would decrease hepatoenteric reperfusion injury, resultant XO activity release, and concomitant lung injury after aortic occlusion–reperfusion when compared with a lactated Ringer’s solution-based resuscitation. For simplicity, PentaLyte will be referred to as a pentastarch solution for the remainder of this report. A clinically relevant rabbit model occlusion–reperfusion of the descending thoracic aorta was used to test these hypotheses.

METHODS Surgical Protocol The study was approved by the Animal Review Committee of the University of Alabama at Birmingham. All animals received humane care in compliance with the Principles of Laboratory Animal Care

1983

Axon, Baird, Lang, et al.: PentaLyte Decreases Lung Injury formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1985, U.S. Government Printing Office, Washington, DC 20402-9325). Male New Zealand White rabbits (Myrtle’s Rabbits, Thompson Station, TN) weighing 2 to 3 kg (n 5 39) were fasted for 14 to 16 h prior to experimentation, but were allowed free access to water. The animals were anesthetized with intravenous ketamine (Parke-Davis, Morris Plains, NJ), 10 mg/kg, given via a marginal ear vein. Animals were subsequently given inhaled 1% isoflurane (Abbott Laboratories, North Chicago, IL) carried in 99% oxygen. Isoflurane administration (inspired concentration) was monitored with an anesthetic-agent monitor (Model 8100; BCI International, Waukesha, WI) that was calibrated daily. After induction of anesthesia, incision sites were infiltrated subcutaneously with 1% lidocaine for additional analgesia. Rabbits were randomly assigned to a sham-operated group that was given lactated Ringer’s solution (n 5 7) or a group given pentastarch solution (n 5 6). Additional rabbits were randomly assigned to an aortic-occlusion group that was given lactated Ringer’s solution (n 5 13) or pentastarch solution (n 5 13). The dosage and duration of infusion of lactated Ringer’s solution or pentastarch solution are outlined subsequently. The arterial pressure was monitored via a 22-gauge central-earartery catheter. After tracheostomy, mechanical ventilation (FIO2 5 1.0) with a ventilator (Model 661; Harvard Apparatus, Millis, MA) was performed with the PaCO2 maintained at 32 to 45 mm Hg. Pancuronium bromide (Elkins-Sinn, Inc., Cherry Hill, NJ) was given intravenously at a rate of 0.1 mg/kg/h to maintain relaxed chest-wall muscle tone during ventilation. Central venous access was obtained through the right internal jugular vein with a size 5French double-lumen catheter (Cook Critical Care, Bloomington, IN) for pressure monitoring and fluid administration. A right femoral arterial catheter was also placed, to verify complete aortic occlusion in the aortic-occlusion groups. All pressures were recorded with continuous-flush transducers (Abbott Critical Care Systems, North Chicago, IL) on a Grass Model 7D polygraph (Grass Instruments, Quincy, MA). A size 7French TRIP sigmoid catheter (Tonometrics, Inc., Hopkinton, MA) was similarly placed into the stomach for tonometric determination of gastric intramucosal pH. The position of the gastric tonometer was confirmed by palpation. Gastric intramucosal pH served as a monitor of gastric ischemia. All rabbits received a maintenance infusion of lactated Ringer’s solution at a rate of 4 ml/kg/h prior to ischemia or sham ischemia, and esophageal temperatures were maintained at 38 to 398 C with a heating pad. A 30-min equilibration period followed the completion of the surgical preparation and device insertion.

Aortic Occlusion–Reperfusion Protocol Sham-operated animals had the left femoral artery exposed, with sham occlusion of the thoracic aorta beginning with ligation of the femoral artery. After 20 min of sham occlusion of the thoracic aorta, either lactated Ringer’s solution (4 ml/kg/h) or pentastarch solution (2 ml/kg/h) was infused throughout the remainder of the experiment. The aortic-occlusion groups also underwent a left femoral cutdown, with insertion of a size 4French Fogarty embolectomy catheter (American Edwards Laboratory, Irvine, CA) into the thoracic aorta. Balloonposition was predetermined by measuring the distance between the surface-anatomy landmarks of the xiphoid process and the inguinal ligament. Catheter insertion to this length (usually 18 to 19 cm) at the inguinal ligament placed the balloon 1 to 2 cm above the diaphragm, as confirmed by postmortem examination. Occlusion of the thoracic aorta was achieved by inflating the catheter balloon with up to 750 ml of saline. Subdiaphragmatic ischemia was confirmed by continuous monitoring of the femoral arterial pressure, which measured 0 to 10 mm Hg during balloon inflation. After 20 min of occlusion, a preload dose (20 ml/kg) of either lactated Ringer’s solution or pentastarch solution was infused during the last 20 min of ischemia. After 40 min of occlusion, the balloon was deflated and the catheter was removed from the aorta. The ensuing postocclusion shock was treated according to the algorithms described in the resuscitation protocol described subsequently. Arterial blood samples were removed at 180 min of reperfusion. The blood was centrifuged and the plasma assayed for the

hepatoenteric enzymes aspartate aminotransferase (AST) and XO, as described in the section on biochemical analysis. Plasma lactate dehydrogense (LDH) activity served as a measure of systemic injury. Wholeblood lactate samples were similarly collected and assayed as described in the section on biochemical analysis. Lactate concentration served as a measure of the adequacy of systemic perfusion. After 3 h of reperfusion, the rabbits were killed with an overdose of pentobarbital (65 mg/ kg; The Butler Company, Columbus, OH).

Resuscitation Protocol Fluids and medications were administered with an Omni-Flow 4000 infusion-pump system (Abbott Laboratories). Fluid administration. At the beginning of reperfusion, the infusion rate of either lactated Ringer’s solution or pentastarch solution was adjusted to maintain central venous pressure (CVP) at the value observed at 30 min of equilibration, 6 2 mm Hg. Phenylephrine administration. Phenylephrine (Elkins-Sinn) administration was continuously adjusted as follows: if the CVP was equal to the preocclusion value 6 2 mm Hg, and the mean arterial blood pressure (MAP) was , 80% of the value observed at 30 min of equilibration, phenylephrine was administered. Sodium bicarbonate administration. Sodium bicarbonate 8.4% (Abbott Laboratories) was infused intravenously at 5 mEq/kg/h during the last 20 min of aortic occlusion. An additional 5 mEq/kg bolus was infused at the beginning of reperfusion, and the infusion was continued at 5 mEq/kr/h. The infusion of sodium bicarbonate was altered according to changes in the arterial base excess (BE) as follows: if the BE > 6, the infusion was stopped; if the BE . 2, the infusion was decreased by 50%; if BE , 2 or . 22, no change was made; if BE , 22 but . 25, the infusion rate was increased by 50%; if BE , 25, the infusion rate was increased by 50% and a 1 mEq/kg bolus of sodium bicarbonate was administered.

Gastric Tonometry Tonometric measurements of gastric intramucosal pH were made after 30 min of equilibration, and at 30, 60, 90, 120, 150, and 180 min of reperfusion. The tonometer balloon was inflated with physiologic saline and the sample was anerobically withdrawn 30 min later. This time allowed for equilibration of the PCO2 between the gastric mucosa and the saline in the balloon. An arterial blood sample was obtained simultaneously with the tonometer sample for the determination of blood gases and bicarbonate (HCO23 ) concentration. The tonometer and blood samples were analyzed at 378 C, using a blood gas analyzer (Model 1306; Instrumentation Laboratory, Lexington, MA). Gastric intramucosal pH was determined according to the method of FiddianGreen and colleagues (17), using a modification of the Henderson– Hasselbalch equation: gastric intramucosal pH 5 6.1 1 log ([arterial HCO23 ]/[1.29 3 0.03 3 tonometered saline PCO2]). The constant 1.29 is a time-dependent equilibration factor determined by Tonometrics, Inc. from in vitro studies in which the tonometer was exposed to a physiologic saline solution with a known PCO2 for 30 min.

Bronchoalveolar Lavage Analyses At the end of each experiment, the lungs were removed and the left main-stem bronchus was clamped. The right lung was lavaged with 20ml aliquots of 0.9% saline until 50-ml of bronchoalveolar lavage fluid (BALF) were obtained. The lavage was centrifuged at 1,000 3 g for 10 min. The protein concentration of the cell-free supernatant was determined as a marker of alveolar–capillary membrane compromise. The LDH activity of the supernatant served as a marker of alveolarcell injury. The final values for the BALF protein content were determined with the following formula: BALF protein content 5 (50 ml 3 protein concentration) 4 rabbit weight in kg. BALF LDH activity was determined in a similar manner.

Tissue Analyses Tissue biopsies were obtained from the left lung and the middle third of the anterior wall of the stomach. After lavage, the right main-stem bronchus was clamped and the left lung was inflated gently (z 15 mm Hg) with 10% buffered formalin. The left lung and stomach samples were then immediately placed in 10% buffered formalin and sub-

1984

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

jected to routine processing, after which they were embedded in paraffin and tissue sections were cut and stained with hematoxylin and eosin (H&E). Morphometric injury was graded by one of the authors (A.E.B.), who was masked as to the identity of the resuscitative fluid or experimental condition (e.g., sham versus aortic occlusion). The grading scale for tissue injury was as follows: Lung. The lung of each animal was evaluated with the following three criteria of tissue injury: (1) heterophils in the interstitial space (heterophils are the rabbit equivalent of human neutrophils); (2) airway epithelial-cell changes (e.g., apical blebbing and vacuolation); and (3) capillary congestion. Each criterion had a possible score of 0 to 4, where 0 5 no injury; 1 5 minimal injury, 2 5 mild injury, 3 5 moderate injury, and 4 5 marked injury. Stomach. The stomach of each animal was evaluated with the following four criteria of tissue injury: (1) mucosal edema; (2) mucosal hemorrhage and ulceration; (3) mineralization; and (4) parietal/chiefcell degeneration and vacuolation in areas other than surface ulcerations. Each criterion had a possible score of 0 to 4, where 0 5 no injury; 1 5 minimal injury, 2 5 mild injury; 3 5 moderate injury, and 4 5 marked injury. Additional stomach biopsies were obtained for analysis of wet-todry weight ratios. After recording its wet weight, the tissue sample was placed in a drying oven at 708 C for 2 wk and reweighed. An increase in the ratio of stomach wet-to-dry weight served as a gross measure of tissue edema.

Biochemical Analysis Plasma aspartate aminotransferase (AST) and lactate dehydrogenase (LDH) activity. Fresh plasma and BALF samples were allowed to equilibrate to room temperature prior to spectrophotometric assay. Plasma AST activity, plasma LDH activity, and BALF LDH activity were measured according to a modification of the procedure of Henry and associates (18). Lactate. Whole blood was deproteinized by addition of ice-cold 8% perchloric acid (1:2, vol/vol), followed by vigorous mixing. The samples were centrifuged at 15,000 3 g for 15 min, and the supernatant was stored at 2858 C prior to assay. Lactate concentration was determined according to a modification of the method of Marbach and coworkers (19). Plasma XO activity. The plasma samples were stored at 2858 C prior to XO assay. Samples were prepared and processed as previously described (7, 8). XO activity was determined from the rate of production of uric acid in the presence of xanthine (75 mM) and b-nicotinamide adenine dinucleotide (NAD1, 0.5 mM). One unit of activity (U) is defined as 1 mmol/min urate formed at 378 C, pH 7.4. Allopurinol (150 mM), an inhibitor of XO was used in parallel samples to confirm that urate formation was specifically due to XO activity. The uric acid content of deproteinized plasma and samples was measured with a high-performance liquid chromatography (HPLC)-based electrochemical technique, as previously described (20). Protein assay. BALF and plasma samples were assayed for total protein concentration by a modification of the method of Smith and colleagues (21).

Statistical Analysis All variables are expressed as means 6 SEM except for the tissueinjury scores, which are expressed as the median with the 25th and 75th percentiles. All the following analyses were performed in accordance with common biostatistical principles (22): Analysis of the effect of administration of pentastarch solution versus lactated Ringer’s solution on all measured parameters in the sham-operated animals was done with Student’s t test, the Mann–Whitney rank-sum test, or repeated measures analysis of variance (ANOVA), as appropriate. If no significant differences were found, the values of the two sham-operated groups were combined for all analyses in which the aortic-occlusion groups were compared. Analysis of the effects of resuscitative fluid and aortic occlusion–reperfusion on BALF protein content, BALF LDH activity, plasma AST activity, plasma LDH activity, plasma XO activity, and gastric wet-to-dry weight ratio was done through oneway ANOVA. Tukey’s test was utilized for post hoc comparisons of groups for these parameters. Differences in resuscitation requirements (e.g., phenylephrine) between the two aortic-occlusion groups

VOL 157

1998

were analyzed with Student’s t test. Stratification of lung injury according to circulating XO activity was done post hoc, and statistical analysis of the difference in lung injury between low and high circulating XO activity groups was determined with Student’s t test. An a error of , 0.05 was considered significant for the aforementioned statistical analyses. Analysis of the effect of resuscitation and aortic occlusion– reperfusion on gastric intramucosal pH, MAP, central venous pressure (CVP), and BE was conducted by repeated measures ANOVA. Paired comparisons were conducted with an a correction to , 0.0167 (Bonferroni’s correction) for the analyses, utilizing repeated-measures ANOVA to account for three comparisons (pentastarch solution/aortic occlusion versus lactated Ringer’s solution/aortic occlusion versus sham-operated group). If the repeated-measures ANOVA demonstrated significance, a one-way ANOVA was performed at each time point, and a Tukey’s test was utilized for post hoc comparisons of groups. The effects of resuscitative fluid and aortic occlusion–reperfusion on lung- and gastric-injury scores were analyzed with the Kruskal–Wallis test. Dunnett’s test was utilized for post hoc comparisons of groups for differences in lung- and gastric-injury scores with an a error of , 0.05 considered significant.

RESULTS There were no significant differences in any of the measured parameters in the two sham-operated groups given pentastarch solution or lactated Ringer’s solution. These two groups were consequently combined, and are referred to as the sham group for the following analyses. Effect of Resuscitative Solutions on Plasma AST, LDH, XO, and Protein after Aortic Occlusion–Reperfusion

The release of AST, LDH, and XO activity was significantly different at 180 min of reperfusion among the three study groups (Figure 1A to C). The aortic-occlusion group given lactated Ringer’s solution had a significantly greater circulating AST activity (26-fold increase), LDH activity (8-fold increase), and XO activity (21-fold increase) than the sham-operated group. The aortic-occlusion group given pentastarch solution had an approximately 4-fold lower circulating AST activity, LDH activity, and XO activity than the aortic-occlusion group given lactated Ringer’s solution. Although the mean value of circulating AST, LDH, or XO activity was greater in the aorticocclusion group given pentastarch solution than in the shamoperated group, this difference was not statistically significant. Pentastarch solution did not inhibit XO activity in commercially available bovine milk in vitro (data not shown). There was a significant difference in plasma protein concentration at 180 min of reperfusion among the three groups. The sham-operated group had a significantly greater plasma protein concentration (34.3 6 1.93 mg/ml) than either the aortic-occlusion group given lactated Ringer’s solution (26.4 6 1.25 mg/ml) or the aortic-occlusion group given pentastarch solution (15.0 6 0.84 mg/ml). The plasma protein concentration was significantly different in the two aortic-occlusion groups. Effect of Resuscitative Solutions on Gastric Intramucosal pH, Gastric Wet-to-Dry Weight Ratio, and Circulating Lactate after Aortic Occlusion–Reperfusion

Gastric intramucosal pH decreased significantly after aortic occlusion–reperfusion in rabbits resuscitated with lactated Ringer’s solution as compared with the sham-operated group (Figure 2A). Animals resuscitated with pentastarch solution had a significantly greater gastric intramucosal pH after aortic occlusion–reperfusion than did rabbits given lactated Ringer’s solution. There was no difference in gastric intramucosal pH in the sham-operated group as opposed to the aortic-occlusion group given pentastarch solution.

1985

Axon, Baird, Lang, et al.: PentaLyte Decreases Lung Injury

Figure 1. Effect of resuscitative solutions on release of AST, LDH, and XO after aortic occlusion–reperfusion. (A to C) The aorticocclusion group given lactated Ringer’s solution (black bar) had significantly (*p , 0.05) greater circulating AST activity, LDH activity, and XO activity than did the sham-operated group (white bar). The aortic-occlusion group given lactated Ringer’s solution had significantly (†p , 0.05) greater circulating AST activity, LDH activity, and XO activity than did the aortic-occlusion group given pentastarch solution (gray bar).

There were significant differences between the three groups in gastric wet-to-dry weight ratio. The gastric wet-to-dry weight ratio was significantly smaller in the sham-operated group (5.94 6 0.34) than in animals given lactated Ringer’s solution (8.53 6 0.41) after aortic occlusion–reperfusion. The gastric wet-to-dry weight ratio was significantly smaller in the animals given pentastarch solution (5.60 6 0.11) than in those given lactated Ringer’s solution after aortic occlusion–reperfusion. The gastric wet-to-dry weight ratio was not different in the sham-operated group and animals given pentastarch solution after aortic occlusion–reperfusion. Circulating lactate concentration was significantly different among the three groups at 180 min of reperfusion (Figure 2B).

Figure 2. Effect of resuscitative solutions on gastric intramucosal pH and circulating lactate after aortic occlusion–reperfusion. (A) Gastric intramucosal pH significantly (p , 0.0167) decreased after aortic occlusion–reperfusion in rabbits resuscitated with lactated Ringer’s solution (closed squares) as compared with the shamoperated group (closed circles); *p , 0.05 at specific time points. Animals resuscitated with lactated Ringer’s solution had a significantly (p , 0.0167) lower gastric intramucosal pH after aortic occlusion–reperfusion than did those given pentastarch solution (closed triangles); †p , 0.05 at specific time points. (B) The aorticocclusion group given lactated Ringer’s solution (black bar) had a significantly greater circulating lactate concentration than did either the sham-operated group (white bar, *p , 0.05) or the aorticocclusion group given pentastarch solution (gray bar, †p , 0.05). The aortic-occlusion group given pentastarch solution also had a significantly greater circulating lactate concentration than did the sham-operated group (‡p , 0.05).

The aortic-occlusion group given lactated Ringer’s solution had a greater circulating lactate concentration than either the sham-operated group (6-fold increase) or the aortic-occlusion group given pentastarch solution (45% greater). The aorticocclusion group given pentastarch solution also had a significantly greater circulating lactate concentration than the shamoperated group. Effect of Resuscitative Solutions on BALF Protein Content and LDH Activity after Aortic Occlusion–Reperfusion

Lung injury was significantly different among the three groups as assessed by BALF protein content and LDH activity (Fig-

1986


Figure 3. Effect of resuscitative solutions on BALF protein content and LDH activity after aortic occlusion–reperfusion. (A, B) The aortic-occlusion group given lactated Ringer’s solution (black bar) had a significantly (*p , 0.05) greater BALF protein content and LDH activity than did the sham-operated group (white bar). The aorticocclusion group given lactated Ringer’s solution had a significantly (†p , 0.05) greater BALF protein content and LDH activity than did the aortic–occlusion group given pentastarch solution ( gray bar).

ure 3A and B). The aortic-occlusion group given lactated Ringer’s solution had an approximately 4-fold greater BALF protein content and LDH activity than did the sham-operated group. The aortic-occlusion group given pentastarch solution had an approximately 4-fold lower BALF protein content and LDH activity than did the aortic-occlusion group given lactated Ringer’s solution. There was no significant difference in BALF protein content or LDH activity in the sham-operated group and the aortic-occlusion group given pentastarch solution. Effect of Resuscitative Solutions on Histologic Measures of Lung and Gastric Injury after Aortic Occlusion–Reperfusion

Lung injury was significantly different among the three groups as assessed by morphometric score (Figure 4A). Both the sham-operated group and the aortic-occlusion group given pentastarch solution had a significantly lower lung morphometric injury score than did the aortic-occlusion group given lactated Ringer’s solution. Gastric injury was significantly different among the three groups as assessed by morphometric score (Figure 4B). Both the sham-operated group and the aortic-occlusion group given pentastarch solution had a significantly lower gastric morphometric injury score than did the aorticocclusion group given lactated Ringer’s solution. Representative morphology for both tissues is presented in Figure 5.

VOL 157

1998

Figure 4. Effect of resuscitative solutions on histologic measures of lung and gastric injury after aortic occlusion–reperfusion. (A) Lung injury (morphometric score) was significantly (*p , 0.05) greater in the animals given lactated Ringer’s solution after aortic occlusion than in the group given pentastarch solution or the shamoperated group. (B) Gastric injury (morphometric score) was significantly (*p , 0.05) greater in the animals given lactated Ringer’s solution after aortic occlusion than in the group given pentastarch solution or the sham-operated group. The data are depicted as the median (middle line in the box), with the top and bottom of the box encompassing the 25th and 75th percentiles and the whiskers encompassing the 5th and 95th percentiles.

Hemodynamic and Acid–Base Data

During reperfusion, there were no differences between the two aortic-occlusion groups in CVP, MAP, or BE (Table 1). Sham-operated animals had a significantly lower MAP in the reperfusion period than did either aortic-occlusion group. Sham-operated animals had significantly lower BE values during reperfusion than did the aortic-occlusion group given pentastarch solution. Resuscitation Requirements

Rabbits in the aortic-occlusion group given pentastarch solution required significantly less supplemental fluid (71 6 3 ml/ kg/3 h) than those given lactated Ringer’s solution (221 6 19 ml/kg/3 h) during reperfusion. Rabbits in the aortic-occlusion group given pentastarch solution required significantly less phenylephrine (1.8 6 0.4 mg/kg/3 h) during reperfusion than

1987


Figure 5. Representative lung and gastric histopathology. (A) Section of lung from a rabbit given pentastarch solution during aortic occlusion–reperfusion. There is a lack of congestion and inflammatory cell infiltrate. (B) Section of lung from a rabbit given lactated Ringer’s solution during aortic occlusion–reperfusion. There is marked capillary congestion and infiltration of heterophils (black arrow). (C) Stomach from a rabbit given pentastarch solution during aortic occlusion–reperfusion. There is minimal mucosal edema, with no evidence of ulceration or hemorrhage. (D) Stomach from a rabbit given lactated Ringer’s solution during aortic occlusion–reperfusion. There is ulceration of the entire mucosal surface, with associated hemorrhage and cellular degeneration. Bar z 200 mm in all panels.

those given lactated Ringer’s solution (3.2 6 0.3 mg/kg/3 h). Rabbits in the aortic-occlusion group given pentastarch solution also required significantly less sodium bicarbonate (11.8 6 0.6 mEq/kg/3 h) during reperfusion than those given lactated Ringer’s solution (16.0 6 1.4 mEq/kg/3 h). Association of BALF Protein Content and LDH Activity with Plasma Parameters

In order to ascertain whether BALF parameters might be related to circulating protein concentration or LDH activity, we performed the relevant linear regressions. BALF protein content was not significantly associated with plasma protein concentration at 180 min of reperfusion (r2 , 0.01, p 5 0.87). In contrast, LDH activity in BALF was significantly associated with plasma LDH activity at 180 min of reperfusion (r2 5 0.56, p , 0.05). Of interest was that BALF protein content and LDH activity were significantly associated with one another (r2 5 0.84, p , 0.05). Because circulating XO activity has been previously associated with increased lung injury after aortic occlusion–reperfusion (7), we determined whether BALF protein content and LDH activity were associated with increased circulating XO activity at 180 min of reperfusion. Because the typical plasma XO activity following surgical procedures is not well characterized, the following definitions of low and high plasma XO activity were derived, based on a modification of a previous definition (7). Animals were considered to have low circulating XO plasma activity if the observed value was < 20 mU/L at 180 min of reperfusion. High circulating XO activity was de-

fined as an increase in plasma activity of . 20 mU/L at 180 min of reperfusion. The value of 20 mU/L was obtained by summarizing the mean value plus two standard deviations of the plasma XO activity in the sham-operated group at 180 min of reperfusion. When all animals were stratified by high or low circulating XO activity, there was significantly (p , 0.05) greater lung injury in the high-XO-activity group (BALF protein content 5 10.05 6 2.60 mg/kg/lung, BALF LDH activity 5 1.63 6 0.52 U/kg/lung) than in the low-XO-activity group (BALF protein content 5 3.97 6 0.60 mg/kg/lung, BALF LDH activity 5 0.57 6 0.08 U/kg/lung). Association of Injury with Resuscitation Requirements

Linear regression analysis was done to determine whether there was an association between the severity of tissue injury and resuscitation requirements. In general, there were weak but significant associations between the lung and hepatoenteric injury parameters and the amount of resuscitation required by the two aortic-occlusion groups (Table 2). Animals with greater injury tended to require a greater quantity of resuscitative intervention than animals with less injury.

DISCUSSION The present study presents the first in vivo evidence that pentastarch solution attenuates remote lung injury after aortic occlusion–reperfusion. The administration of pentastarch solution decreased the release of XO activity after aortic occlusion–reperfusion when compared with resuscitation based on

1988


VOL 157

1998

TABLE 1 CENTRAL VENOUS PRESSURES, MEAN ARTERIAL PRESSURES, AND BASE EXCESS VALUES* Time† EQ CVP, mm Hg Sham AO 1 lactated Ringer’s solution AO 1 pentastarch MAP, mm Hg Sham‡ AO 1 lactated Ringer’s solution AO 1 pentastarch BE Sham§ AO 1 lactated Ringer’s solution AO 1 pentastarch

60 min

120 min

180 min

260 261 260

360 361 360

260 361 361

260 461 360

88 6 2 88 6 3 87 6 3

86 6 5 82 6 5 81 6 4

77 6 3 90 6 6 83 6 3

78 6 2 88 6 4 82 6 3

20.2 6 0.6 20.8 6 0.7 20.4 6 0.6

21.1 6 0.7¶ 1.8 6 0.9 3.5 6 0.6

20.2 6 0.5¶ 1.1 6 0.8 3.7 6 0.5

1.1 6 0.5¶ 0.5 6 0.9 3.9 6 0.8

Definition of abbreviations: CVP 5 central venous pressure; MAP 5 mean arterial pressure; BE 5 base excess; Sham 5 sham-operated group; AO 5 aortic occlusion group; pentastarch 5 pentastarch solution; EQ 5 30 min equilibration time point, remaining time points are during reperfusion. * All values are expressed as means 6 SEM. † Although parameters were measured more frequently, for brevity, only four time points are shown. ‡ p , 0.0167, sham versus AO 1 lactated Ringer’s solution and AO 1 pentastarch over time. § p , 0.0167, sham versus AO 1 pentastarch over time. ¶ p , 0.05, sham versus AO 1 pentastarch at specific time point.

lactated Ringer’s solution. XO has been previously shown to play a significant role in the etiology of remote lung and heart injury in the rabbit model of hepatoenteric ischemia–reperfusion (7, 8) and in other animal models (3–5). XO typically exists in an innocuous form (xanthine dehydrogenase) in nonischemic tissues, with purine catabolism occurring concomitantly with the reduction of nicotinamide adenine dinucleotide (9, 23–25). When tissues are exposed to metabolic stress, such as hypoxia or ischemia, xanthine dehydrogenase converts to the oxidase form. XO, in the presence of adequate purine substrate and molecular oxygen, generates the oxidants superoxide anion (?O22 ), hydrogen peroxide (H2O2), and ?OH (9, 23, 24). Conversion of xanthine dehydrogenase to XO occurs within seconds through rapid and reversible oxidation of sulfhydryls once the former enzyme is released into plasma (26). The release of XO alone is not sufficient to cause tissue injury under normal conditions, since the activity of XO is limited by low purine substrate concentrations (1 to 3 mM) in the systemic circulation (26). However, plasma concentrations of the

XO substrates hypoxanthine and xanthine have been shown to increase following reconstructive aortic surgery (27, 28). XO and high concentrations of purine substrates may consequently be released into the circulation following hepatoenteric ischemia–reperfusion, and may contribute to tissue injury. Another attractive hypothesis for the effect of pentastarch solution in preventing remote lung injury, in addition to its attenuating the release of XO activity, may be by directly detoxifying oxidants or interfering with neutrophil-mediated events. Pentastarch has been shown to be a hydroxyl radical scavenger in vitro (16). It has also been found to decrease neutrophil binding to endothelium in vitro, which could potentially decrease neutrophil-mediated oxidant injury in vivo (13). Neutrophils have been shown to sequester in the lung after episodes of ischemia to remote vascular beds (3, 4), as well as accumulating in reperfused tissue (29). Activated neutrophils produce oxidants (4), which can contribute to acute pulmonary injury in the form of edema, increased physiologic shunt, and increased capillary membrane permeability (30, 31). The

TABLE 2 LINEAR REGRESSION ANALYSIS OF INJURY PARAMETERS WITH PHENYLEPHRINE, SODIUM BICARBONATE, AND SUPPLEMENTAL FLUID REQUIREMENTS* Phenylephrine

BALF protein content BALF LDH activity Lung histologic injury score Plasma AST activity Plasma LDH activity Plasma XO activity Circulating lactate Gastric intramucosal pH† Gastric histologic injury score Gastric wet-to-dry weight ratio

2

r

0.37 0.16 0.38 0.30 0.33 0.20 0.26 0.05 0.08 0.20

Sodium Bicarbonate

Supplemental Fluid

p

2

r

p

r2

p

NS S S S S S S NS NS S

0.25 0.32 0.30 0.35 0.45 0.10 0.27 0.08 0.31 0.11

S S S S S NS S NS S NS

0.19 0.23 0.48 0.36 0.51 0.29 0.18 0.23 0.11 0.36

S S S S S S S S NS S

Definition of abbreviations: AST 5 aspartate aminotransferase; S 5 p , 0.05; NS 5 not significant; XO 5 xanthine oxidase. * All values are derived from comparison of the injury parameter observed at 180 min of reperfusion with the total dose of resuscitative medication over the 180 min of reperfusion (e.g., mg/kg/3 h for phenylephrine). † The association of gastric intramucosal pH with the amount of supplemental fluid administered is inverse (e.g., greater fluid requirements were associated with lower gastric intramucosal pH values).

1989


mechanism responsible for pulmonary leukosequestration after remote tissue injury seems to involve reactive oxygen species (31, 32). Circulating XO is a likely source of oxidants in these settings, since leukosequestration and acute lung injury are significantly attenuated by inhibition or inactivation of XO (3, 4). However, the precise linkage between hepatoenteric ischemia–reperfusion, circulating XO, and neutrophils in the evolution of remote organ injury remains unclear. Infusion of XO into rats increased lung XO activity and pulmonary leukosequestration, but paradoxically did not compromise the pulmonary capillary membrane (3). Our observations are in concordance with the concept that the remote organ injury observed after hepatoenteric ischemia–reperfusion involves XO-mediated events, given that circulating XO activity and remote lung injury were concomitantly reduced by pentastarch-based resuscitation. Determination of the role of pentastarch solution in attenuating neutrophil-mediated tissue injury after aortic occlusion–reperfusion remains a subject for future investigation. Aside from modulation of hepatoenteric reperfusion injury and release of XO activity, other possible confounding factors may have influenced both lung and hepatoenteric injury in the present study. In particular, there were weak but significant associations between the severity of tissue injury and the amount of resuscitative intervention required to maintain hemodynamic and acid–base parameters within the predetermined limits. Animals with greater tissue injury generally tended to require greater quantities of resuscitative interventions during reperfusion. In the case of supplemental fluids, the associations support the contention that there are important differences between the two fluids tested: despite the fluids’ producing similar hemodynamics, the pentastarch solution exhibited protective properties when compared with lactated Ringer’s solution. Circulating lactate concentrations may have been affected by the quantity of supplemental fluid administered (pentastarch solution has the same concentration of sodium lactate as lactated Ringer’s solution), but the association between fluid administered and circulating lactate concentration was weak (Table 2). In the case of phenylephrine administration, one could argue that hepatoenteric perfusion may be compromised in proportion to the amount of phenylephrine administered, as a result of vasoconstriction. This is unlikely, since the circulating lactate concentration was weakly associated with phenylephrine administration, and gastric intramucosal pH was not associated with phenylephrine administration (Table 2). Although resuscitative measures may in some minor way contribute to tissue injury, the converse may also be true—that attenuation of tissue injury with pentastarch solution reduces resuscitation requirements. In addition to the influences of resuscitative measures on tissue injury, one must also consider the effects of circulating LDH activity and protein concentration on BALF parameters. BALF protein content could potentially be influenced by increases in plasma protein concentration as a result of inadequate fluid resuscitation after aortic occlusion–reperfusion. This was not the case in the present study, since the shamoperated group had a greater plasma protein concentration than did either aortic occlusion–reperfusion group. Consequently, the increase in BALF protein content noted in the present study most likely reflect increased alveolar–capillary membrane permeability. In contrast, BALF LDH activity was associated with plasma LDH activity after aortic occlusion– reperfusion. Although it is possible that the release of LDH activity into BALF may reflect an increase in permeability, it more likely reflects alveolar cell lysis in the lung. BALF LDH activity was strongly associated with BALF protein content, a

parameter that did not change passively with changes in plasma protein concentration. Consequently, if the lung was nonselectively permeable to a protein the size of LDH (140,000 daltons), one would expect a significant decrease in BALF protein content in the aortic occlusion–reperfusion groups (low plasma protein concentration) as compared with sham-operated group (greater plasma protein concentration), given that the predominant plasma protein is albumin (60,000 daltons). Supporting this line of reasoning are previous studies in which BALF AST and XO activity remained at the level of detection and did not increase after aortic occlusion–reperfusion, in contrast to plasma AST and XO activity (unpublished data). Given that AST and XO are the same size as LDH, and colocalize with LDH in the liver and intestines, it would be expected that all three enzymes would increase in the BALF after aortic occlusion–reperfusion if the alveolar–capillary membrane was compromised enough to allow passage of LDH. This is not the case. Consequently, increased BALF protein content and LDH activity most likely reflect increased alveolar– capillary membrane permeability and alveolar cell lysis, respectively. In conclusion, the present study provides the first in vivo evidence that administration of pentastarch solution abrogates remote lung injury after aortic occlusion–reperfusion as compared with lactated Ringer’s solution-based resuscitation. One possible mechanism(s) by which pentastarch solution protects against remote lung injury after aortic occlusion–reperfusion includes decreasing the release of XO activity from the liver and intestines. The reduction in release of XO activity after aortic occlusion–reperfusion may result from the restoration by pentastarch solution of hepatoenteric perfusion, and from decreasing hepatoenteric cytolysis, as compared with lactated Ringer’s solution-based resuscitation. This therapeutic paradigm presents a novel approach to the reduction of XO-mediated tissue injury, wherein the focus is not the direct inhibition of enzymatic activity, but rather the attenuation of enzyme release. Clinically, it is not possible to either inactivate (7) or inhibit (9) XO activity within the time constraints (e.g., minutes) associated with the management of patients in the setting of major vascular surgery or trauma. The reduction of tissue XO activity requires hours to days of pretreatment (7, 9). These data serve as a rational basis for further investigation of the precise mechanisms responsible for pentastarch solution-mediated attenuation of both the remote lung injury and hepotoenteric injury associated with aortic occlusion–reperfusion. References 1. Gelman, S. 1995. The pathophysiology of aortic cross-clamping and unclamping. Anesthesiology 82:1026–1060. 2. Ryder, I. G., and D. L. Brown. 1993. Anesthetic risks for trauma patients. In C. M. Grande, editor. Textbook of Trauma Anesthesia and Critical Care. Mosby-Year Book, St. Louis. 445–452. 3. Terada, L. S., J. J. Dormish, P. F. Shanley, J. A. Leff, B. O. Anderson, and J. E. Repine. 1992. Circulating xanthine oxidase mediates lung neutrophil sequestration after intestinal ischemia-reperfusion. Am. J. Physiol. 263:L394–L401. 4. Koike, K., F. A. Moore, E. E. Moore, R. A. Read, V. S. Carl, and A. Banerjee. 1993. Gut ischemia mediates lung injury by a xanthine oxidase-dependent neutrophil mechanism. J. Surg. Res. 54:469–473. 5. Weinbroum, A., V. G. Nielsen, S. Tan, S. Gelman, S. Matalon, E. Bradley, and D. A. Parks. 1995. Liver ischemia-reperfusion increases pulmonary permeability in the rat: role of circulating xanthine oxidase. Am. J. Physiol. 268:G988–G996. 6. Nielsen, V. G., S. Tan, M. S. Baird, A. T. McCammon, and D. A. Parks. 1996. Gastric intramucosal pH and multiple organ injury: impact of ischemia-reperfusion and xanthine oxidase. Crit. Care Med. 24:1339– 1344.

1990


7. Nielsen, V. G., S. Tan, A. Weinbroum, A. T. McCammon, P. N. Samuelson, S. Gelman, and D. A. Parks. 1996. Lung injury after hepatoenteric ischemia-reperfusion: role of xanthine oxidase. Am. J. Respir. Crit. Care Med. 154:1364–1369. 8. Nielsen, V. G., S. Tan, M. S. Baird, P. N. Samuelson, and D. A. Parks. 1997. Xanthine oxidase mediates myocardial injury after hepatoenteric ischemia-perfusion. Crit. Care Med. 25:1044–1050. 9. Parks, D. A., and D. N. Granger. 1986. Xanthine oxidase: biochemistry, distribution and physiology. Acta Physiol. Scand. 548:87–99. 10. Zirka, B. A., C. Subbarao, M. C. Oz, S. J. Popilkis, R. Sachdev, P. Chauhan, H. P. Freeman, and T. C. King. 1990. Hydroxyethyl starch macromolecules reduce myocardial reperfusion injury. Arch. Surg. 125:930– 934. 11. Oz, M. C., B. A. Zirka, P. F. McLeod, and S. J. Popilkis. 1991. Hydroxyethyl starch macromolecules and superoxide dismutase effects on myocardial reperfusion injury. Am. J. Surg. 162:59–62. 12. Schell, R. M., D. J. Cole, R. L. Schultz, and T. N. Osborne. 1992. Temporary cerebral ischemia, effects of pentastarch or albumin on reperfusion injury. Anesthesiology 77:86–92. 13. Oz, M. C., M. F. FitzPatrick, B. A. Zirka, D. J. Pinsky, and W. N. Duran. 1995. Attenuation of microvascular permeability dysfunction in postischemic striated muscle by hydroxyethyl starch. Microvasc. Res. 50: 71–79. 14. Zirka, B. A., C. Subbarao, M. C. Oz, S. T. Shih, P. F. McLeod, R. Sachdev, H. P. Freeman, and M. A. Hardy. 1989. Macromolecules reduce abnormal microvascular permeability in rat limb ischemia-reperfusion injury. Crit. Care Med. 17:1306–1309. 15. Wilms, C. D., J. A. Davidson, J. M. Armstrong, M. Kwon, R. Risser, Z. F. Sandor, and J. T. Sentementes. 1991. Pentafraction-Du Pont versus albumin for resuscitation of a lethal intestinal ischemic shock in rats. Circ. Shock 33:216–221. 16. Pieper, G. M., G. J. Gross, and B. Kalyanaraman. 1990. An ESR study of the nitroxide radical of pentastarch-conjugated deferoxamine. Free Radic. Biol. Med. 9:211–218. 17. Fiddian-Green, R. G., G. Pittenger, and W. M. Whitehouse, Jr. 1982. Back diffusion of CO2 and its influence on the intramural pH in gastric mucosa. J. Surg. Res. 33:39–48. 18. Henry, R. J., N. Chiamori, O. J. Golub, and S. Berkman. 1960. Revised spectrophotometric methods for the determination of glutamic-oxalacetic transaminase, glutamic-pyruvic transaminase and lactic acid dehydrogenase. Am. J. Clin. Pathol. 34:381–391.

VOL 157

1998

19. Marbach, E. P., and M. H. Weil. 1967. Rapid enzymatic measurement of blood lactate and pyruvate. Clin. Chem. 13:314–325. 20. Tan, S., R. Radi, F. Gaudier, R. A. Evans, A. Rivera, K. A. Kirk, and D. A. Parks. 1993. Physiological levels of uric acid inhibit xanthine oxidase in human plasma. Pediatr. Res. 34:303–307. 21. Smith, P. K., R. I. Krohn, G. T. Hermanson, A. K. Mallia, F. H. Gartner, M. D. Provenzano, E. K. Fujimoto, N. M. Goeke, B. J. Olson, and D. C. Klenk. 1985. Measurement of protein using bicinchoninic acid. Anal. Biochem. 150:76–85. 22. Glantz, S. A. 1992. Primer of Biostatics, 3rd ed. McGraw-Hill, New York. 23. Parks, D. A., T. K. Williams, and J. S. Beckman. 1988. Conversion of xanthine dehydrogenase to oxidase in ischemic rat intestine: a reevaluation. Am. J. Physiol. 254:G768–G774. 24. Phan, S. H., D. E. Gannon, J. Varani, U. S. Ryan, and P. Ward. 1989. Xanthine oxidase activity in rat pulmonary artery endothelial cells and its alteration by activated neutrophils. Am. J. Pathol. 134:1201– 1211. 25. Adachi, T., T. Fukushima, Y. Usami, and K. Hirano. 1993. Binding of human xanthine oxidase to sulphated glycosaminoglycans on the endothelial-cell surface. Biochem. J. 289:523–527. 26. Yokoyama, Y., J. S. Beckman, T. K. Beckman, J. K. Wheat, T. G. Cash, B. A. Freeman, and D. A. Parks. 1990. Circulating xanthine oxidase: potential mediator of ischemic injury. Am. J. Physiol. 258:G564–G570. 27. Frank, R. S., M. M. Moursi, R. M. Podrazik, G. B. Zelenock, and L. G. D’Alecy. 1988. Renal vasoconstriction and transient declamp hypotension after infrenal aortic occlusion: role of plasma purine degradation products. J. Vasc. Surg. 7:515–523. 28. Schoenberg, M. H., B. B. Fredholm, and G. Hohlbach. 1985. Changes in acid-base status, lactate concentration and purine metabolites during reconstructive aortic surgery. Acta Chir. Scand. 151:227–233. 29. Grisham, M. B., L. A. Hernandez, and D. N. Granger. 1986. Xanthine oxidase and neutrophil infiltration in intestinal ischemia. Am. J. Physiol. 251:G567–G574. 30. Klausner, J. M., I. S. Paterson, L. Kobzik, C. R. Valeri, D. Shepro, and H. B. Hechtman. 1989. Oxygen free radicals mediate ischemia-induced lung injury. Surgery 105(2, Pt. 1):192–199. 31. Klausner, J. M., H. Anner, I. S. Paterson, L. Kobzik, C. R. Valeri, D. Shepro, and H. B. Hechtman. 1988. Lower torso ischemia-induced lung injury is leukocyte dependent. Ann. Surg. 208:761–767.