- Email: [email protected]
Antibody against neutrophil adhesion improves reperfusion and limits alveolar infiltrate following unilateral pulmonary artery occlusion
JOURNAL
OF SURGICAL
RESEARCH
52,1%+204
(19%)
Antibody against Neutrophil Adhesion Improves Reperfusion and Limits Alveolar Infiltrate following Unilateral Pulmonary Artery Occlusion MICHAEL
J. BISHOP, M.D., TIM F. KOWALSKI, M.D., SANDRA M. GUIDOTTI, B.S., AND JOHN M. HARLAN, M.D. of Anesthesiology and Medicine (Division of Pulmonary and Critical Care and Division of Hematology),
Departments
University
of Washington,
Submitted
Seattle, Washington
for publication
98195
October 31, 1990
muscle, the flow is initially much less than prior to the ischemic period [l-3]. This phenomenon, often called “no reflow,” is accompanied by tissue injury with neutrophi1 plugging of vessels and emigration of neutrophils into the tissue. Although the lung has generally been felt to be protected from ischemia/reperfusion injury because of its dual circulation and the presence of alveolar oxygen, lung reperfusion injury has been seen both clinically and experimentally. In humans, reperfusion injury occurs following pulmonary artery thrombectomy [4] and has been reported following streptokinase treatment of pulmonary embolism [5]. In dog and rabbit models, reperfusion following 24 hr of unilateral pulmonary artery occlusion produces edema, leukopenia, and microscopic lung injury [6-B]. As with the systemic beds, relief of pulmonary vascular occlusion does not always result in restoration of normal flow to the lung distal to the obstruction. We previously found in dogs and rabbits that following 2448 hr of unilateral pulmonary artery occlusion, relief of the occlusion results in a less than expected flow through the reperfused vascular bed [6, 71. Humans who have undergone thromboendarterectomy for chronic thrombotic obstruction of major pulmonary arteries initially demonstrate hypoperfusion of the previously occluded lung zones [4]. Following lung transplantation, an elevation in pulmonary vascular resistance, and hence a decrease in perfusion, to the transplanted lung is commonly seen [g-11]. All of these situations show that following varying lengths of lung ischemia, reperfusion to the previously occluded pulmonary vascular bed is diminished from expected levels. The mechanisms of the no reflow phenomenon in the pulmonary vascular bed has not been studied previously. Because microscopic sections of lung following reperfusion demonstrated constricted vessels, we hypothesized that vasospasm might play a role in no reflow. We therefore studied the effect of a nonspecific smooth muscle relaxant, papaverine, on the extent of reperfusion. Be-
Relief of unilateral pulmonary arterial occlusion results in bilateral lung injury and results in only partial restoration of pulmonary blood flow distal to the site of occlusion. We hypothesized that the “no reflow” phenomenon was in part due to neutrophil adherence and aggregation in the pulmonary vasculature. The study was carried out in two phases. First, we studied the effect of neutrophil depletion on left lung blood flow following 24 hr of left pulmonary artery occlusion. Hydroxyurea was used to deplete circulating neutrophils to 77 + 18/mms (means f sem) (n = 6) as compared to 708 + 165/mm3 in control rabbits (n = 8). In both groups left lung blood flow immediately following reperfusion was markedly reduced at 6.4 + 2.2% of cardiac output in control rabbits and 7.3 + 2.3 in treated rabbits. However, at 4 hr, neutrophil-depleted animals had significantly greater flow (18.7 f 3.6 vs 8.4 + 2.3% for control rabbits, P < 0.05). In both groups, flow remained substantially below the normal rabbit left lung blood flow of 39.8 + 2.2%. To test whether the improved reflow was due to decreased numbers of neutrophils limiting aggregation, or whether active neutrophi1 adherence played a role, we tested the effect of a monoclonal antibody that interferes with neutrophil adhesiveness (MoAb 60.3) on reflow and on neutrophil emigration into the alveoli. We found that MoAb 60.3 did not affect initial retlow. However, at 24 hr, flow in treated rabbits was virtually normal (38.8 I~I 1.5+%) but remained markedly diminished in untreated animals at 25.5 + 3.9%. MoAb 60.3 also markedly diminished neutrophil emigration into the alveoli following reperfusion. We conclude that neutrophils play a significant role in preventing reflow following lung ischemia and that inhibition of neutrophil adhesiveness reverses this phenomenon and decreases the inflammatory response in the alveolus. 0 1992 Academic press, h.
INTRODUCTION
When reperfusion follows brief periods of ischemia in circulatory beds of the brain, heart, kidney, and skeletal 199
0022.4804/92 $1.50 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
200
JOURNAL
OF SURGICAL
RESEARCH:
cause histologic examination also demonstrated large numbers of intravascular neutrophils, we hypothesized that neutrophil obstruction of vessels was an alternative cause of low reflow. The neutrophil studies were carried out in two phases. We first studied the effect of hydroxyurea-induced neutrophil depletion on flow following 24 hr of left pulmonary artery occlusion. When these studies demonstrated a significant improvement in flow following neutrophil depletion, we then questioned whether the obstruction was due to passive aggregation of the neutrophils in the pulmonaryvasculature or to active adhesion. The availability of a monoclonal antibody (MoAb 60.3) to the CD18 neutrophi1 adhesion protein enabled us to study this. The CD18 glycoprotein mediates adhesion of neutrophils to each other (homotypic adhesion) and to endothelium (heterotypic adhesion) [12]. Using MoAb 60.3, we studied reperfusion in rabbits with normal numbers of neutrophils that had been rendered nonadhesive. In addition to studying the effect of MoAb 60.3 on no reflow, we also studied whether such treatment would ameliorate the alveolar inflammation that accompanies reperfusion. METHODS New Zealand white rabbits weighing from 2.5 to 3.5 kg were studied. All animals were managed in accordance with NIH guidelines for the care of laboratory animals. We used the pulmonary artery occlusion/reperfusion model as described previously [7]. Briefly, the rabbits were anesthetized with ketamine (30 mg/kg) and xylazine (9 mg/kg) intramuscularly, then intubated and ventilated while anesthesia was maintained with halothane. Using sterile technique, a thoracotomy was performed in the fourth intercostal space and the left main pulmonary artery was isolated and occluded with an atraumatic microvascular clamp. The lung was reexpanded, the pleural space evacuated, and the chest closed, and the rabbit allowed to recover. Twenty-four hours following pulmonary artery occlusion, the animal was reanesthetized, reintubated, and had the clamp removed. Depending on the various protocols, the animal was either allowed to recover fully or remained under anesthesia. For studies that did not involve reawakening the rabbits following occluder removal, pentobarbital anesthesia was used. For rabbits that were to be awakened following occluder removal, halothane was used to permit more rapid awakening. The distribution of pulmonary blood flow during reperfusion was determined by injecting radiolabeled microspheres into an ear vein catheter. Microspheres (16.5 & 0.1 pm diameter DuPont-New England Nuclear) were radiolabeled with scandium-46, niobium-95, ruthenium103, or cerium-141. Microspheres of this diameter were virtually uniformly trapped in the pulmonary vascula-
VOL.
52, NO. 3, MARCH
1992
ture as they significantly exceed alveolar capillary diameter. Microspheres were injected into the inferior vena cava to ensure mixing in two heart chambers prior to reaching the lung. Injections of spheres was performed at 10,120, and 240 min following reperfusion in the neutrophil-depleted rabbits and in the MoAb 60.3-treated rabbits at 10 min and 24 hr following reperfusion. A different isotope was used for each measurement so that serial determinations of flow could be made. Lungs were removed at the end of the experiment and the appropriate energy levels counted with a gamma counter. Correction for energy spill down was made following counting of reference samples of the isotope. The proportion of cardiac output that was distributed to the previously occluded lung was then calculated by dividing the radioactivity in the previously occluded lung by the total radioactivity in both lungs. Papaverine
Studies
We used papaverine to assess the role of smooth muscle constriction in the low reflow phenomenon. Six rabbits were studied, using the occlusion and reperfusion protocol described above. Cardiac output was measured via thermodilution and systemic blood pressure via an indwelling femoral catheter. Following reperfusion, we made a measurement of blood flow distribution, then administered 15 mg/kg of papaverine, then repeated the measurement. Finally, an hour later when all hemodynamic effects of the papaverine subsided, we again repeated the measurement. Neutrophil
Depletion
Neutrophil depleted rabbits (n = 6) were treated with 400 mg/kg/day of hydroxyurea intravenously for 4 days prior to the occlusion. Treatment continued through the day of reperfusion. A sample of arterial blood was drawn just prior to removal of the occluder and the number of circulating neutrophils were counted. Control rabbits (n = 8) followed the same protocol but did not receive hydroxyurea. MoAb 60.3 Studies Rabbits were studied for 24 hr following reperfusion. We followed the same protocol as described above for the operative placement of the occluder in all rabbits. Rabbits were reanesthetized 24 hr after occluder placement and given either 2 mg/kg of MoAb 60.3 (treated rabbits) or the equivalent volume of saline, and the oceluder was removed 5 min later. Following removal of the occluder, the chest was evacuated of air and closed. A measurement of blood flow distribution was performed and the animals were allowed to recover. Twelve hours following removal of the occluder, a second dose of 2 mg/kg of MoAb 60.3 was given. Twenty-four hours after occluder removal, blood flow distribution was again
BISHOP
01
0 Time Relative
2 to Reperfusion
ET AL.:
4 (hours)
NEUTROPHILS
I
FIG. 1. Percentage of the cardiac output perfusing the left (previously occluded) lung following reperfusion. The two curves are significantly different (P < 0.01) by analysis of variance for repeated measures. Values are means * sem.
measured and the rabbits were then euthanized. Lungs were removed and each lung was lavaged with 20 ml of saline and the neutrophil content of the recovered fluid was noted. The technician performing the measurements was blinded as to whether the animals were treated or control. Statistics Analysis of variance was used to assess the effect of interventions and of right versus left lung data. Paired t tests were used to assess changes in hemodynamic parameters in papaverine-treated animals. Data are presented as means f sem. RESULTS Prior studies in our laboratory of rabbit pulmonary blood flow established a normal left lung blood flow of 39.8 f 2.2% of cardiac output (n = 6). Papaverine
Treatment
Addition of the nonspecific smooth muscle relaxant papaverine resulted in marked systemic vasodilation with a decrease in systemic blood pressure from 86 + 5 to 66 -t 6 mm Hg (P < 0.01) and an increase in cardiac output from 395 f 21 to 513 + 50 ml/min (P < O.Ol), a calculated decrease of 41% in systemic resistance. Left lung blood flow rose from 5.1 f 2.2 to 7.8 + 3.5% (P < 0.05) of cardiac output. Sixty minutes later, when the effects of the papaverine were no longer evident and output and blood pressure had returned to baseline, repeat measurement of flow showed a return to 5.1 f 1.9% of flow to the left lung. Neutrophil
Depletion
Control rabbits 3590 + 343/mm3 708 3 165/mm3. mean circulating
had a circulating leukocyte count of and a circulating neutrophil count of Neutrophil-depleted animals had a leukocyte count of 1667 f 235/mm3
AND
LUNG
REPERFUSION
201
and a mean neutrophil count of 77 f 18/mm3 immediately prior to reperfusion. One animal, excluded from statistical analysis because of technical problems with the initial and 2-hr measurements of blood flows, had no detectable circulating neutrophils. Hydroxyurea decreased platelet counts from a mean of 538 f 53 thousand/mm3 prior to treatment to 422 + 40 thousand/mm3 just prior to reperfusion (P < 0.05). Left lung blood flows were not significantly different between the two groups immediately following reperfusion (Fig. l), but by 2 hr the neutrophil-depleted rabbits’ left lung blood flow had risen to 17.0 f 4.8% versus 7.4 + 2.4 in the control group (P = 0.07). At 4 hr, control rabbits’ flows were unchanged at 8.4 f 2.3% whereas neutrophil-depleted animals had increased to 18.7 -t 3.6% (P = 0.03). MoAb 60.3-Treated
Rabbits
Inhibition of neutrophil adherence with MoAb 60.3 did not affect flow immediately following release of the occlusion (Fig. 2). However, 24 hr later the flow to the left lung had returned to normal in MoAb 60.3-treated rabbits (38.8 + 1.5% of cardiac output versus 39.8 f 2.2 in normal rabbits) but was still markedly diminished in untreated rabbits (25.5 f 3.9%). Alveolar lavage demonstrated a marked decrease in cellular content in both right and left lungs (Fig. 3), with the most dramatic decrease in the neutrophil fraction. Neutrophils constituted 23 -t 2% of the cells recovered from the reperfused left lung in the absence of treatment and only 8 f 3% in treated rabbits (P c 0.05). In the continuously perfused right lung, the percentage of neutrophils was 4 f 1% versus 1 + 1% in the two groups (P < 0.05). DISCUSSION Unilateral pulmonary artery occlusion and reperfusion for periods of a day or more result in lung injury
5 a
50
2
40
0 yj 2 ; E g 2 0.
Mean t SEM 30 20 10 0
10 min 24 hours Post-Repsrfusion
FIG. 2. Flow to the reperfused lung is presented as a percentage of total cardiac output. MoAb 60.3 did not affect flow to the reperfused lung immediately following release of occlusion but MoAb 60.3treated rabbits had normal flow by 24 hr. In contrast, untreated rabbits still had significantly diminished left lung blood flow at 24 hr. *P < 0.05 between groups. Values are means f sem.
202
JOURNAL
OF SURGICAL
RESEARCH:
VOL.
52, NO. 3, MARCH
1992
1,000 6 8 ;
600
x5
600
2x
400
iiii
200
A 0
Left
Riiht
FIG. 3. Total leukocyte counts (A) and neutrophils (B) recovered following 20 ml lavage of each lung. The left lung was occluded for 24 hr and then reperfused for 24 hr. MoAb 60.3 markedly diminished neutrophil emigration into the alveoli and also decreased total leukocytes present. *P < 0.95 for untreated vs treated. Values are means + sem.
distal to the occlusion with a minimal injury also occurring in the contralateral lung [8,13]. We and others have noted that reperfusion is accompanied by high pulmonary vascular resistance in the previously occluded area. This correlates with clinical evidence of low pulmonary blood flow initially following lung transplantation. Low reflow has also been demonstrated following pulmonary thromboendarterectomy for chronic thrombotic obstruction of the pulmonary artery [4]. In these patients, radionuclide scans demonstrate a progressive improvement in perfusion for a period of months following surgery. No reflow due to neutrophil aggregation is seen in organ systems other than the lung after much briefer periods of ischemia [l-3] and has, in some cases, been preventable by interfering with neutrophil adhesiveness [ 141. However, the lung, because of its dual blood supply, is resistant to brief periods of ischemia and thus lung reperfusion is not entirely analogous to brain, heart, or kidney reperfusion. We previously examined rabbit pulmonary vasculature following reperfusion and found extensive neutrophi1 aggregation and vascular plugging. However, counting of occluded vessels in micrographs revealed plugs in approximately one-third of the vessels, leading us to question whether this could be the sole reason for low reflow. Intravascular erythrocyte thrombi were rarely seen, and examination of the left pulmonary artery was routinely conducted to ensure that no large vessel occlusion or thrombosis is present. We have previously established that the reduced flow to the reperfused lung is not due to the effects of the thoracotomy [ 151. In that study, rabbits that underwent a thoracotomy and a brief left pulmonary artery obstruction (under 1 min) still had 42% of cardiac output delivered to the left lung. Our studies demonstrated that neither marked neutrophi1 depletion nor inhibition of neutrophil adherence could completely prevent the low reflow immediately following relief of the occlusion, suggesting that other factors may play a role. However, by 4 hr, rabbits that had been neutrophil depleted demonstrated significantly increased flow, suggesting that neutrophils do play a role
in the increased pulmonary vascular resistance of the reperfused lung. Our inability to deplete circulating neutrophils completely makes it impossible to assess whether in the total absence of neutrophils, flow might have been yet greater. Even with our low mean circulating count of 77/mm3, there may still have been significant numbers of marginated neutrophils in the lungs prior to reperfusion [16] which were then available to aggregate. In a study in rabbits using an identical dose of hydroxyurea, alveolar capillary PMNs were significantly depleted although at a somewhat lower efficiency than in the peripheral circulation [17]. The one rabbit in which we could detect no circulating neutrophils was noteworthy in that it had a return to normal flow at 4 hr. The high lethality of routinely depleting rabbits to this extent prevented attempting to reproduce this result. The improvement in reflow following neutrophil depletion occurred in the presence of a virtually normal number of platelets. Although the hydroxyurea caused a statistically significant decrease in the platelet count, the actual number of platelets was only diminished by 20% and was still well within a normal range. Because the initially low flow did not appear neutrophi1 dependent, we considered vasoconstriction and endothelial swelling as other possible causes. Endothelial cell damage can be seen on electron microscopy following reperfusion [ 181 but quantitating its effect on flow is not possible. We attempted to assess the effect of vasoconstriction using papaverine. Although the papaverine resulted in a statistically significant increase in flow to the left lung, the absolute magnitude of the effect was small suggesting that vasoconstriction plays only a minor role. In fact, the increased flow following papaverine could also have been mediated by the increased cardiac output, since increased cardiac output reduces passive neutrophil sequestration in the lung [19]. Having found that the low reflow was in part neutrophi1 dependent, we questioned whether this was due to passive sequestration or involved active adherence to endothelium and/or neutrophil-neutrophil adhesion. Neutrophils marginate in the lungs in inverse proportion to blood flow [ 191. However, actual adherence to the endo-
BISHOP
ET AL.:
NEUTROPHILS
thelium occurs following neutrophil activation and may be followed by migration into the tissues and subsequent tissue injury and vascular occlusion. The importance of neutrophils in producing vascular and tissue injury following ischemia/reperfusion has been demonstrated by inhibiting homotypic and heterotypic neutrophil adhesion. A membrane glycoprotein complex designated the CDll/CD18 complex appears to mediate adhesiveness in this setting [12, 20, 211. The monoclonal antibody designated MoAb 60.3 binds to a function-related epitope on the CD18 glycoprotein and inhibits stimulated aggregation [22], adherence to endothelial monolayers [23], and neutrophil-mediated endothelial injury in vitro [24]. In uiuo, administration of MoAb 60.3 has been shown to ameliorate vascular and tissue damage following ischemia/reperfusion in the gut [2] and ear [21] and to reduce gastric and liver injury following hemorrhagic shock and resuscitation [ 20, 251. Interestingly, MoAb 60.3 did not prevent neutrophil emigration into the lung in this latter model of generalized ischemia/reperfusion [26]. Nevertheless, we felt that given the apparent role of neutrophil adhesion in the low reflow phenomenon, we should study the possibility that the stimulus of reperfusion might increase neutrophil adhesiveness and promote homotypic or heterotypic adhesion in the lung via activation of the CDll/CD18 complex. The CD18 MoAb 60.3 markedly improved lung perfusion at 24 hr following release of the occluder. Potential explanations include either a decrease in capillary plugging by neutrophil aggregates, a reduction in neutrophilmediated vascular injury by adherent neutrophils, or a combination of the two. The decrease in neutrophil emigration to approximately one-fifth that is seen in untreated animals suggests that the CDll/CD18 complex does play a major role in adhesiveness and emigration in the pulmonary vasculature following reperfusion. Inhibiting neutrophil adherence did not improve flow immediately following reperfusion, just as neutrophil depletion did not affect the immediate post-reperfusion flow. We speculate that more than one mechanism contributes to the decreased reperfusion following relief of occlusion. One potential explanation is that flow is limited initially by vasoconstriction and/or endothelial swelling and later by neutrophil aggregation and/or adherence. Thrombosis did not seem to play a role: intravascular erythrocyte thrombi were rarely seen, and routine examination of the left pulmonary artery was routinely conducted to ensure that no large vessel occlusion or thrombosis is present. The results of our study suggest that the pulmonary vascular dysfunction and alveolar inflammatory infiltrate following pulmonary artery ischemia and reperfusion may be partially preventable. Potential clinical importance of these findings include following thrombectomy, following pneumothorax reexpansion, and following lung transplantation.
AND
LUNG
REPERFUSION
203
ACKNOWLEDGMENTS This work was supported by NIH Grant 30542 and American Heart Association Grant 8%WA-16. J.M.H. is an Established Investigator of the American Heart Association.
REFERENCES 1.
Barroso-Aranda, J., Schmid-Schonbein, G. W., Zweifach, B. W., and Engler, R. L. Granulocytes and no-reflow phenomenon in irreversible hemorrhagic shock. Circ. Res. 63: 437, 1988. 2. Hernandez, L. A., Grisham, M. B., Twohig, B., Arfors, K. E., Harlan, J. M., and Granger, D. N. Role of neutrophils in ischemia-reperfusion induced microvascular injury. Am. J. Physiol. 253(Heart Circ Physiol22): H699, 1987.
3. Schmid-Schonbein, G. W. Granulocyte: friend and foe. NIPS 3: 144,1988. 4. Moser, K. M., Spragg, G., Utley, J., and Daily, P. 0. Chronic thrombotic obstruction of major pulmonary arteries. Ann. Zntern. Med. 99: 299,1983. 5. Martin, T. R., Sandblom, R. L., and Johnson, R. J. Adult respiratory distress syndrome following thrombrolytic therapy for pulmonary embolism. Chest 83: 151,1983. 6. Bishop, M. J., Boatman, E. S., Ivey, T. O., Jordan, J. P., and Cheney, F. W. Reperfusion of ischemic dog lung results in fever leukopenia and lung edema. Am. Rev. Respir. Dis. 134: 752,
1986. 7. Kowalski,
T. F., Guidotti, S., Deffebach, M., Kubilis, P., and Bishop, M. J. The bronchial circulation in pulmonary artery occlusion and reperfusion. J. Appl. Physiol. 68: 125, 1990.
8. Bishop, M. J., Chi, E. Y., and Cheney, F. W. Lung reperfusion
in dogs causes bilateral lung injury. J. Appl. Physiol. 63: 942, 1987. 9. Castagna, J. T., Shors, E., and Benfield, J. R. The role of perfusion in lung preservation. J. Thorac. Cardiovasc. Surg. 63: 521, 1972. 10. Haverich, A., Scott, W. C., and Jamieson, J. W. Twenty years of lung preservation: A review. Heart Transplant. 4: 234, 1985. 11. Jurmann, M. J., Dammenhayn, L., Schafers, H. J., Wahlers, T., Fieguth, H. G., and Haverich, A. Prostacyclin as an additive to single crystalloid flush: Improved pulmonary preservation in heart-lung transplantation. Heart Transplant. 385: 1986. [Abstract]
12. Beatty, P. G., Ledbetter,
13.
14.
15.
16.
17.
J. A., Martin, P. J., Price, T. H., and Hansen, J. A. Definition of a common leukocyte cell-surface antigen (Lp95-150) associated with diverse cell-mediated immune functions. J. Zmmunol. 131: 2913, 1983. Horgan, M. J., Lum, J., and Malik, A. B. Pulmonary edema after pulmonary artery occlusion and reperfusion. Am. Rev. Respir. Dis. 140:1421,1989. Simpson, P. J., Todd, R. F., III, Fantore, J. C., Mickelson, J. K., Griffin, J. D., and Lucchesi, B. R. Reduction of experimental canine myocardial reperfusion injury by a monoclonal antibody (anti mo-1, anti -cdllb) that inhibits leukocyte adhesion. J. Clin. Invest. 81: 624, 1988. Bishop, M. J., Giclas, P. D., Guidotti, S. M., Su, M. L., and Chi, E. Y. Complement activation is a secondary rather than causative factor in rabbit pulmonary artery ischemia/reperfusion injury. Am. Rev. Respir. Dis. 143: 386, 1990. Albertine, K. H. Lung injury and neutrophil density during air embolization in sheep after leukocyte depletion with nitrogen mustard. Am. Rev. Respir. Dis. 138: 1444, 1988. Gie, R. P., Doershuck, C. M., Coxson, H. O., English, D., and Hogg, J. C. The role of neutrophils in acute lung injury caused by
204
JOURNAL an infusion
of zymosan-activated
OF SURGICAL plasma in rabbits.
RESEARCH:
VOL.
Am. Rev.
Respir. Dis. 139: A301, 1989. 18.
Haworth, S. G., Hall, S. M., Odom, N., and McGregor, C. G. Ischemia/reperfusion injury after unilateral lung transplantation in the rat: Ultrastructural changes in pulmonary capillaries. Am. Rev. Respir. Dis. 139(4): A415, 1989.
19.
Thommasen, H. V., Martin, B. A., Wiggs, B. R., Quiroga, M., Baile, E. M., and Hogg, J. C. Effect of pulmonary blood flow on leukocyte uptake and release by dog lung. J. Appl. Physiol. 56: 966,1984.
20.
Vedder, N. B., Fouty, B. W., Winn, R. K., Harlan, J. M., and Rice, C. L. Role of neutrophils in generalized reperfusion injury associated with resuscitation from shock. Surgery 106: 509, 1989.
21.
22.
Vedder, N. B., Winn, R. K., Chi, E. Y., Rice, C. L., Arfors, K. E., and Harlan, J. M. Inhibition of leukocyte adherence by antCD18 monoclonal antibody attenuates reperfusion injury in the rabbit ear. Proc. Natl. Acad. Sci. USA 87: 2643, 1990. Schwartz,
B. R., Ochs, H. D., Beatty,
P. G., and Harlan,
J. M. A
23.
24.
52, NO. 3, MARCH
1992
monoclonal antibody-defined membrane antigen complex is required for neutrophil-neutrophil aggregation. Blood 65: 1553, 1985. Harlan, J. M., Killen, P. D., Senecal, F. M., Schwartz, B. R., Yee, E. K., Taylor, R. F., Beatty, P. G., Price, T. H., and Ochs, H. D. The role of neutrophil membrane glycoprotein GP-150 in neutrophi1 adherence to endothelium in vitro. Blood 66: 167, 1985. Diener, A. M., Beatty, P. G., Ochs, H. D., and Harlan, J. M. The role of neutrophil membrane glycoprotein 150 (GP-150) in neutrophil-mediated endothelial cell injury in vitro. J. Zmmunol.
135:537,1985.
25 ’
Vedder, N. B., Winn, R. K., Rice, C. L., Chi, E. Y., Arfors, K. E., and Harlan, J. M. A monoclonal antibody to the adherence-promoting leukocyte glycoprotein, CD18, reduces organ injury and improves survival from hemorrhagic shock in rabbits. J. Clin.
Invest. 81: 939,1988.
26 .
Doerschuk, C. M., Winn, R. K., Coxson, H. O., and Harlan, J. M. CD 18-dependent and independent mechanisms of neutrophil adherence in the pulmonary and systemic microcirculation of rabbits. J. Zmmunol. 144: 2327, 1990.
OF SURGICAL
RESEARCH
52,1%+204
(19%)
Antibody against Neutrophil Adhesion Improves Reperfusion and Limits Alveolar Infiltrate following Unilateral Pulmonary Artery Occlusion MICHAEL
J. BISHOP, M.D., TIM F. KOWALSKI, M.D., SANDRA M. GUIDOTTI, B.S., AND JOHN M. HARLAN, M.D. of Anesthesiology and Medicine (Division of Pulmonary and Critical Care and Division of Hematology),
Departments
University
of Washington,
Submitted
Seattle, Washington
for publication
98195
October 31, 1990
muscle, the flow is initially much less than prior to the ischemic period [l-3]. This phenomenon, often called “no reflow,” is accompanied by tissue injury with neutrophi1 plugging of vessels and emigration of neutrophils into the tissue. Although the lung has generally been felt to be protected from ischemia/reperfusion injury because of its dual circulation and the presence of alveolar oxygen, lung reperfusion injury has been seen both clinically and experimentally. In humans, reperfusion injury occurs following pulmonary artery thrombectomy [4] and has been reported following streptokinase treatment of pulmonary embolism [5]. In dog and rabbit models, reperfusion following 24 hr of unilateral pulmonary artery occlusion produces edema, leukopenia, and microscopic lung injury [6-B]. As with the systemic beds, relief of pulmonary vascular occlusion does not always result in restoration of normal flow to the lung distal to the obstruction. We previously found in dogs and rabbits that following 2448 hr of unilateral pulmonary artery occlusion, relief of the occlusion results in a less than expected flow through the reperfused vascular bed [6, 71. Humans who have undergone thromboendarterectomy for chronic thrombotic obstruction of major pulmonary arteries initially demonstrate hypoperfusion of the previously occluded lung zones [4]. Following lung transplantation, an elevation in pulmonary vascular resistance, and hence a decrease in perfusion, to the transplanted lung is commonly seen [g-11]. All of these situations show that following varying lengths of lung ischemia, reperfusion to the previously occluded pulmonary vascular bed is diminished from expected levels. The mechanisms of the no reflow phenomenon in the pulmonary vascular bed has not been studied previously. Because microscopic sections of lung following reperfusion demonstrated constricted vessels, we hypothesized that vasospasm might play a role in no reflow. We therefore studied the effect of a nonspecific smooth muscle relaxant, papaverine, on the extent of reperfusion. Be-
Relief of unilateral pulmonary arterial occlusion results in bilateral lung injury and results in only partial restoration of pulmonary blood flow distal to the site of occlusion. We hypothesized that the “no reflow” phenomenon was in part due to neutrophil adherence and aggregation in the pulmonary vasculature. The study was carried out in two phases. First, we studied the effect of neutrophil depletion on left lung blood flow following 24 hr of left pulmonary artery occlusion. Hydroxyurea was used to deplete circulating neutrophils to 77 + 18/mms (means f sem) (n = 6) as compared to 708 + 165/mm3 in control rabbits (n = 8). In both groups left lung blood flow immediately following reperfusion was markedly reduced at 6.4 + 2.2% of cardiac output in control rabbits and 7.3 + 2.3 in treated rabbits. However, at 4 hr, neutrophil-depleted animals had significantly greater flow (18.7 f 3.6 vs 8.4 + 2.3% for control rabbits, P < 0.05). In both groups, flow remained substantially below the normal rabbit left lung blood flow of 39.8 + 2.2%. To test whether the improved reflow was due to decreased numbers of neutrophils limiting aggregation, or whether active neutrophi1 adherence played a role, we tested the effect of a monoclonal antibody that interferes with neutrophil adhesiveness (MoAb 60.3) on reflow and on neutrophil emigration into the alveoli. We found that MoAb 60.3 did not affect initial retlow. However, at 24 hr, flow in treated rabbits was virtually normal (38.8 I~I 1.5+%) but remained markedly diminished in untreated animals at 25.5 + 3.9%. MoAb 60.3 also markedly diminished neutrophil emigration into the alveoli following reperfusion. We conclude that neutrophils play a significant role in preventing reflow following lung ischemia and that inhibition of neutrophil adhesiveness reverses this phenomenon and decreases the inflammatory response in the alveolus. 0 1992 Academic press, h.
INTRODUCTION
When reperfusion follows brief periods of ischemia in circulatory beds of the brain, heart, kidney, and skeletal 199
0022.4804/92 $1.50 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
200
JOURNAL
OF SURGICAL
RESEARCH:
cause histologic examination also demonstrated large numbers of intravascular neutrophils, we hypothesized that neutrophil obstruction of vessels was an alternative cause of low reflow. The neutrophil studies were carried out in two phases. We first studied the effect of hydroxyurea-induced neutrophil depletion on flow following 24 hr of left pulmonary artery occlusion. When these studies demonstrated a significant improvement in flow following neutrophil depletion, we then questioned whether the obstruction was due to passive aggregation of the neutrophils in the pulmonaryvasculature or to active adhesion. The availability of a monoclonal antibody (MoAb 60.3) to the CD18 neutrophi1 adhesion protein enabled us to study this. The CD18 glycoprotein mediates adhesion of neutrophils to each other (homotypic adhesion) and to endothelium (heterotypic adhesion) [12]. Using MoAb 60.3, we studied reperfusion in rabbits with normal numbers of neutrophils that had been rendered nonadhesive. In addition to studying the effect of MoAb 60.3 on no reflow, we also studied whether such treatment would ameliorate the alveolar inflammation that accompanies reperfusion. METHODS New Zealand white rabbits weighing from 2.5 to 3.5 kg were studied. All animals were managed in accordance with NIH guidelines for the care of laboratory animals. We used the pulmonary artery occlusion/reperfusion model as described previously [7]. Briefly, the rabbits were anesthetized with ketamine (30 mg/kg) and xylazine (9 mg/kg) intramuscularly, then intubated and ventilated while anesthesia was maintained with halothane. Using sterile technique, a thoracotomy was performed in the fourth intercostal space and the left main pulmonary artery was isolated and occluded with an atraumatic microvascular clamp. The lung was reexpanded, the pleural space evacuated, and the chest closed, and the rabbit allowed to recover. Twenty-four hours following pulmonary artery occlusion, the animal was reanesthetized, reintubated, and had the clamp removed. Depending on the various protocols, the animal was either allowed to recover fully or remained under anesthesia. For studies that did not involve reawakening the rabbits following occluder removal, pentobarbital anesthesia was used. For rabbits that were to be awakened following occluder removal, halothane was used to permit more rapid awakening. The distribution of pulmonary blood flow during reperfusion was determined by injecting radiolabeled microspheres into an ear vein catheter. Microspheres (16.5 & 0.1 pm diameter DuPont-New England Nuclear) were radiolabeled with scandium-46, niobium-95, ruthenium103, or cerium-141. Microspheres of this diameter were virtually uniformly trapped in the pulmonary vascula-
VOL.
52, NO. 3, MARCH
1992
ture as they significantly exceed alveolar capillary diameter. Microspheres were injected into the inferior vena cava to ensure mixing in two heart chambers prior to reaching the lung. Injections of spheres was performed at 10,120, and 240 min following reperfusion in the neutrophil-depleted rabbits and in the MoAb 60.3-treated rabbits at 10 min and 24 hr following reperfusion. A different isotope was used for each measurement so that serial determinations of flow could be made. Lungs were removed at the end of the experiment and the appropriate energy levels counted with a gamma counter. Correction for energy spill down was made following counting of reference samples of the isotope. The proportion of cardiac output that was distributed to the previously occluded lung was then calculated by dividing the radioactivity in the previously occluded lung by the total radioactivity in both lungs. Papaverine
Studies
We used papaverine to assess the role of smooth muscle constriction in the low reflow phenomenon. Six rabbits were studied, using the occlusion and reperfusion protocol described above. Cardiac output was measured via thermodilution and systemic blood pressure via an indwelling femoral catheter. Following reperfusion, we made a measurement of blood flow distribution, then administered 15 mg/kg of papaverine, then repeated the measurement. Finally, an hour later when all hemodynamic effects of the papaverine subsided, we again repeated the measurement. Neutrophil
Depletion
Neutrophil depleted rabbits (n = 6) were treated with 400 mg/kg/day of hydroxyurea intravenously for 4 days prior to the occlusion. Treatment continued through the day of reperfusion. A sample of arterial blood was drawn just prior to removal of the occluder and the number of circulating neutrophils were counted. Control rabbits (n = 8) followed the same protocol but did not receive hydroxyurea. MoAb 60.3 Studies Rabbits were studied for 24 hr following reperfusion. We followed the same protocol as described above for the operative placement of the occluder in all rabbits. Rabbits were reanesthetized 24 hr after occluder placement and given either 2 mg/kg of MoAb 60.3 (treated rabbits) or the equivalent volume of saline, and the oceluder was removed 5 min later. Following removal of the occluder, the chest was evacuated of air and closed. A measurement of blood flow distribution was performed and the animals were allowed to recover. Twelve hours following removal of the occluder, a second dose of 2 mg/kg of MoAb 60.3 was given. Twenty-four hours after occluder removal, blood flow distribution was again
BISHOP
01
0 Time Relative
2 to Reperfusion
ET AL.:
4 (hours)
NEUTROPHILS
I
FIG. 1. Percentage of the cardiac output perfusing the left (previously occluded) lung following reperfusion. The two curves are significantly different (P < 0.01) by analysis of variance for repeated measures. Values are means * sem.
measured and the rabbits were then euthanized. Lungs were removed and each lung was lavaged with 20 ml of saline and the neutrophil content of the recovered fluid was noted. The technician performing the measurements was blinded as to whether the animals were treated or control. Statistics Analysis of variance was used to assess the effect of interventions and of right versus left lung data. Paired t tests were used to assess changes in hemodynamic parameters in papaverine-treated animals. Data are presented as means f sem. RESULTS Prior studies in our laboratory of rabbit pulmonary blood flow established a normal left lung blood flow of 39.8 f 2.2% of cardiac output (n = 6). Papaverine
Treatment
Addition of the nonspecific smooth muscle relaxant papaverine resulted in marked systemic vasodilation with a decrease in systemic blood pressure from 86 + 5 to 66 -t 6 mm Hg (P < 0.01) and an increase in cardiac output from 395 f 21 to 513 + 50 ml/min (P < O.Ol), a calculated decrease of 41% in systemic resistance. Left lung blood flow rose from 5.1 f 2.2 to 7.8 + 3.5% (P < 0.05) of cardiac output. Sixty minutes later, when the effects of the papaverine were no longer evident and output and blood pressure had returned to baseline, repeat measurement of flow showed a return to 5.1 f 1.9% of flow to the left lung. Neutrophil
Depletion
Control rabbits 3590 + 343/mm3 708 3 165/mm3. mean circulating
had a circulating leukocyte count of and a circulating neutrophil count of Neutrophil-depleted animals had a leukocyte count of 1667 f 235/mm3
AND
LUNG
REPERFUSION
201
and a mean neutrophil count of 77 f 18/mm3 immediately prior to reperfusion. One animal, excluded from statistical analysis because of technical problems with the initial and 2-hr measurements of blood flows, had no detectable circulating neutrophils. Hydroxyurea decreased platelet counts from a mean of 538 f 53 thousand/mm3 prior to treatment to 422 + 40 thousand/mm3 just prior to reperfusion (P < 0.05). Left lung blood flows were not significantly different between the two groups immediately following reperfusion (Fig. l), but by 2 hr the neutrophil-depleted rabbits’ left lung blood flow had risen to 17.0 f 4.8% versus 7.4 + 2.4 in the control group (P = 0.07). At 4 hr, control rabbits’ flows were unchanged at 8.4 f 2.3% whereas neutrophil-depleted animals had increased to 18.7 -t 3.6% (P = 0.03). MoAb 60.3-Treated
Rabbits
Inhibition of neutrophil adherence with MoAb 60.3 did not affect flow immediately following release of the occlusion (Fig. 2). However, 24 hr later the flow to the left lung had returned to normal in MoAb 60.3-treated rabbits (38.8 + 1.5% of cardiac output versus 39.8 f 2.2 in normal rabbits) but was still markedly diminished in untreated rabbits (25.5 f 3.9%). Alveolar lavage demonstrated a marked decrease in cellular content in both right and left lungs (Fig. 3), with the most dramatic decrease in the neutrophil fraction. Neutrophils constituted 23 -t 2% of the cells recovered from the reperfused left lung in the absence of treatment and only 8 f 3% in treated rabbits (P c 0.05). In the continuously perfused right lung, the percentage of neutrophils was 4 f 1% versus 1 + 1% in the two groups (P < 0.05). DISCUSSION Unilateral pulmonary artery occlusion and reperfusion for periods of a day or more result in lung injury
5 a
50
2
40
0 yj 2 ; E g 2 0.
Mean t SEM 30 20 10 0
10 min 24 hours Post-Repsrfusion
FIG. 2. Flow to the reperfused lung is presented as a percentage of total cardiac output. MoAb 60.3 did not affect flow to the reperfused lung immediately following release of occlusion but MoAb 60.3treated rabbits had normal flow by 24 hr. In contrast, untreated rabbits still had significantly diminished left lung blood flow at 24 hr. *P < 0.05 between groups. Values are means f sem.
202
JOURNAL
OF SURGICAL
RESEARCH:
VOL.
52, NO. 3, MARCH
1992
1,000 6 8 ;
600
x5
600
2x
400
iiii
200
A 0
Left
Riiht
FIG. 3. Total leukocyte counts (A) and neutrophils (B) recovered following 20 ml lavage of each lung. The left lung was occluded for 24 hr and then reperfused for 24 hr. MoAb 60.3 markedly diminished neutrophil emigration into the alveoli and also decreased total leukocytes present. *P < 0.95 for untreated vs treated. Values are means + sem.
distal to the occlusion with a minimal injury also occurring in the contralateral lung [8,13]. We and others have noted that reperfusion is accompanied by high pulmonary vascular resistance in the previously occluded area. This correlates with clinical evidence of low pulmonary blood flow initially following lung transplantation. Low reflow has also been demonstrated following pulmonary thromboendarterectomy for chronic thrombotic obstruction of the pulmonary artery [4]. In these patients, radionuclide scans demonstrate a progressive improvement in perfusion for a period of months following surgery. No reflow due to neutrophil aggregation is seen in organ systems other than the lung after much briefer periods of ischemia [l-3] and has, in some cases, been preventable by interfering with neutrophil adhesiveness [ 141. However, the lung, because of its dual blood supply, is resistant to brief periods of ischemia and thus lung reperfusion is not entirely analogous to brain, heart, or kidney reperfusion. We previously examined rabbit pulmonary vasculature following reperfusion and found extensive neutrophi1 aggregation and vascular plugging. However, counting of occluded vessels in micrographs revealed plugs in approximately one-third of the vessels, leading us to question whether this could be the sole reason for low reflow. Intravascular erythrocyte thrombi were rarely seen, and examination of the left pulmonary artery was routinely conducted to ensure that no large vessel occlusion or thrombosis is present. We have previously established that the reduced flow to the reperfused lung is not due to the effects of the thoracotomy [ 151. In that study, rabbits that underwent a thoracotomy and a brief left pulmonary artery obstruction (under 1 min) still had 42% of cardiac output delivered to the left lung. Our studies demonstrated that neither marked neutrophi1 depletion nor inhibition of neutrophil adherence could completely prevent the low reflow immediately following relief of the occlusion, suggesting that other factors may play a role. However, by 4 hr, rabbits that had been neutrophil depleted demonstrated significantly increased flow, suggesting that neutrophils do play a role
in the increased pulmonary vascular resistance of the reperfused lung. Our inability to deplete circulating neutrophils completely makes it impossible to assess whether in the total absence of neutrophils, flow might have been yet greater. Even with our low mean circulating count of 77/mm3, there may still have been significant numbers of marginated neutrophils in the lungs prior to reperfusion [16] which were then available to aggregate. In a study in rabbits using an identical dose of hydroxyurea, alveolar capillary PMNs were significantly depleted although at a somewhat lower efficiency than in the peripheral circulation [17]. The one rabbit in which we could detect no circulating neutrophils was noteworthy in that it had a return to normal flow at 4 hr. The high lethality of routinely depleting rabbits to this extent prevented attempting to reproduce this result. The improvement in reflow following neutrophil depletion occurred in the presence of a virtually normal number of platelets. Although the hydroxyurea caused a statistically significant decrease in the platelet count, the actual number of platelets was only diminished by 20% and was still well within a normal range. Because the initially low flow did not appear neutrophi1 dependent, we considered vasoconstriction and endothelial swelling as other possible causes. Endothelial cell damage can be seen on electron microscopy following reperfusion [ 181 but quantitating its effect on flow is not possible. We attempted to assess the effect of vasoconstriction using papaverine. Although the papaverine resulted in a statistically significant increase in flow to the left lung, the absolute magnitude of the effect was small suggesting that vasoconstriction plays only a minor role. In fact, the increased flow following papaverine could also have been mediated by the increased cardiac output, since increased cardiac output reduces passive neutrophil sequestration in the lung [19]. Having found that the low reflow was in part neutrophi1 dependent, we questioned whether this was due to passive sequestration or involved active adherence to endothelium and/or neutrophil-neutrophil adhesion. Neutrophils marginate in the lungs in inverse proportion to blood flow [ 191. However, actual adherence to the endo-
BISHOP
ET AL.:
NEUTROPHILS
thelium occurs following neutrophil activation and may be followed by migration into the tissues and subsequent tissue injury and vascular occlusion. The importance of neutrophils in producing vascular and tissue injury following ischemia/reperfusion has been demonstrated by inhibiting homotypic and heterotypic neutrophil adhesion. A membrane glycoprotein complex designated the CDll/CD18 complex appears to mediate adhesiveness in this setting [12, 20, 211. The monoclonal antibody designated MoAb 60.3 binds to a function-related epitope on the CD18 glycoprotein and inhibits stimulated aggregation [22], adherence to endothelial monolayers [23], and neutrophil-mediated endothelial injury in vitro [24]. In uiuo, administration of MoAb 60.3 has been shown to ameliorate vascular and tissue damage following ischemia/reperfusion in the gut [2] and ear [21] and to reduce gastric and liver injury following hemorrhagic shock and resuscitation [ 20, 251. Interestingly, MoAb 60.3 did not prevent neutrophil emigration into the lung in this latter model of generalized ischemia/reperfusion [26]. Nevertheless, we felt that given the apparent role of neutrophil adhesion in the low reflow phenomenon, we should study the possibility that the stimulus of reperfusion might increase neutrophil adhesiveness and promote homotypic or heterotypic adhesion in the lung via activation of the CDll/CD18 complex. The CD18 MoAb 60.3 markedly improved lung perfusion at 24 hr following release of the occluder. Potential explanations include either a decrease in capillary plugging by neutrophil aggregates, a reduction in neutrophilmediated vascular injury by adherent neutrophils, or a combination of the two. The decrease in neutrophil emigration to approximately one-fifth that is seen in untreated animals suggests that the CDll/CD18 complex does play a major role in adhesiveness and emigration in the pulmonary vasculature following reperfusion. Inhibiting neutrophil adherence did not improve flow immediately following reperfusion, just as neutrophil depletion did not affect the immediate post-reperfusion flow. We speculate that more than one mechanism contributes to the decreased reperfusion following relief of occlusion. One potential explanation is that flow is limited initially by vasoconstriction and/or endothelial swelling and later by neutrophil aggregation and/or adherence. Thrombosis did not seem to play a role: intravascular erythrocyte thrombi were rarely seen, and routine examination of the left pulmonary artery was routinely conducted to ensure that no large vessel occlusion or thrombosis is present. The results of our study suggest that the pulmonary vascular dysfunction and alveolar inflammatory infiltrate following pulmonary artery ischemia and reperfusion may be partially preventable. Potential clinical importance of these findings include following thrombectomy, following pneumothorax reexpansion, and following lung transplantation.
AND
LUNG
REPERFUSION
203
ACKNOWLEDGMENTS This work was supported by NIH Grant 30542 and American Heart Association Grant 8%WA-16. J.M.H. is an Established Investigator of the American Heart Association.
REFERENCES 1.
Barroso-Aranda, J., Schmid-Schonbein, G. W., Zweifach, B. W., and Engler, R. L. Granulocytes and no-reflow phenomenon in irreversible hemorrhagic shock. Circ. Res. 63: 437, 1988. 2. Hernandez, L. A., Grisham, M. B., Twohig, B., Arfors, K. E., Harlan, J. M., and Granger, D. N. Role of neutrophils in ischemia-reperfusion induced microvascular injury. Am. J. Physiol. 253(Heart Circ Physiol22): H699, 1987.
3. Schmid-Schonbein, G. W. Granulocyte: friend and foe. NIPS 3: 144,1988. 4. Moser, K. M., Spragg, G., Utley, J., and Daily, P. 0. Chronic thrombotic obstruction of major pulmonary arteries. Ann. Zntern. Med. 99: 299,1983. 5. Martin, T. R., Sandblom, R. L., and Johnson, R. J. Adult respiratory distress syndrome following thrombrolytic therapy for pulmonary embolism. Chest 83: 151,1983. 6. Bishop, M. J., Boatman, E. S., Ivey, T. O., Jordan, J. P., and Cheney, F. W. Reperfusion of ischemic dog lung results in fever leukopenia and lung edema. Am. Rev. Respir. Dis. 134: 752,
1986. 7. Kowalski,
T. F., Guidotti, S., Deffebach, M., Kubilis, P., and Bishop, M. J. The bronchial circulation in pulmonary artery occlusion and reperfusion. J. Appl. Physiol. 68: 125, 1990.
8. Bishop, M. J., Chi, E. Y., and Cheney, F. W. Lung reperfusion
in dogs causes bilateral lung injury. J. Appl. Physiol. 63: 942, 1987. 9. Castagna, J. T., Shors, E., and Benfield, J. R. The role of perfusion in lung preservation. J. Thorac. Cardiovasc. Surg. 63: 521, 1972. 10. Haverich, A., Scott, W. C., and Jamieson, J. W. Twenty years of lung preservation: A review. Heart Transplant. 4: 234, 1985. 11. Jurmann, M. J., Dammenhayn, L., Schafers, H. J., Wahlers, T., Fieguth, H. G., and Haverich, A. Prostacyclin as an additive to single crystalloid flush: Improved pulmonary preservation in heart-lung transplantation. Heart Transplant. 385: 1986. [Abstract]
12. Beatty, P. G., Ledbetter,
13.
14.
15.
16.
17.
J. A., Martin, P. J., Price, T. H., and Hansen, J. A. Definition of a common leukocyte cell-surface antigen (Lp95-150) associated with diverse cell-mediated immune functions. J. Zmmunol. 131: 2913, 1983. Horgan, M. J., Lum, J., and Malik, A. B. Pulmonary edema after pulmonary artery occlusion and reperfusion. Am. Rev. Respir. Dis. 140:1421,1989. Simpson, P. J., Todd, R. F., III, Fantore, J. C., Mickelson, J. K., Griffin, J. D., and Lucchesi, B. R. Reduction of experimental canine myocardial reperfusion injury by a monoclonal antibody (anti mo-1, anti -cdllb) that inhibits leukocyte adhesion. J. Clin. Invest. 81: 624, 1988. Bishop, M. J., Giclas, P. D., Guidotti, S. M., Su, M. L., and Chi, E. Y. Complement activation is a secondary rather than causative factor in rabbit pulmonary artery ischemia/reperfusion injury. Am. Rev. Respir. Dis. 143: 386, 1990. Albertine, K. H. Lung injury and neutrophil density during air embolization in sheep after leukocyte depletion with nitrogen mustard. Am. Rev. Respir. Dis. 138: 1444, 1988. Gie, R. P., Doershuck, C. M., Coxson, H. O., English, D., and Hogg, J. C. The role of neutrophils in acute lung injury caused by
204
JOURNAL an infusion
of zymosan-activated
OF SURGICAL plasma in rabbits.
RESEARCH:
VOL.
Am. Rev.
Respir. Dis. 139: A301, 1989. 18.
Haworth, S. G., Hall, S. M., Odom, N., and McGregor, C. G. Ischemia/reperfusion injury after unilateral lung transplantation in the rat: Ultrastructural changes in pulmonary capillaries. Am. Rev. Respir. Dis. 139(4): A415, 1989.
19.
Thommasen, H. V., Martin, B. A., Wiggs, B. R., Quiroga, M., Baile, E. M., and Hogg, J. C. Effect of pulmonary blood flow on leukocyte uptake and release by dog lung. J. Appl. Physiol. 56: 966,1984.
20.
Vedder, N. B., Fouty, B. W., Winn, R. K., Harlan, J. M., and Rice, C. L. Role of neutrophils in generalized reperfusion injury associated with resuscitation from shock. Surgery 106: 509, 1989.
21.
22.
Vedder, N. B., Winn, R. K., Chi, E. Y., Rice, C. L., Arfors, K. E., and Harlan, J. M. Inhibition of leukocyte adherence by antCD18 monoclonal antibody attenuates reperfusion injury in the rabbit ear. Proc. Natl. Acad. Sci. USA 87: 2643, 1990. Schwartz,
B. R., Ochs, H. D., Beatty,
P. G., and Harlan,
J. M. A
23.
24.
52, NO. 3, MARCH
1992
monoclonal antibody-defined membrane antigen complex is required for neutrophil-neutrophil aggregation. Blood 65: 1553, 1985. Harlan, J. M., Killen, P. D., Senecal, F. M., Schwartz, B. R., Yee, E. K., Taylor, R. F., Beatty, P. G., Price, T. H., and Ochs, H. D. The role of neutrophil membrane glycoprotein GP-150 in neutrophi1 adherence to endothelium in vitro. Blood 66: 167, 1985. Diener, A. M., Beatty, P. G., Ochs, H. D., and Harlan, J. M. The role of neutrophil membrane glycoprotein 150 (GP-150) in neutrophil-mediated endothelial cell injury in vitro. J. Zmmunol.
135:537,1985.
25 ’
Vedder, N. B., Winn, R. K., Rice, C. L., Chi, E. Y., Arfors, K. E., and Harlan, J. M. A monoclonal antibody to the adherence-promoting leukocyte glycoprotein, CD18, reduces organ injury and improves survival from hemorrhagic shock in rabbits. J. Clin.
Invest. 81: 939,1988.
26 .
Doerschuk, C. M., Winn, R. K., Coxson, H. O., and Harlan, J. M. CD 18-dependent and independent mechanisms of neutrophil adherence in the pulmonary and systemic microcirculation of rabbits. J. Zmmunol. 144: 2327, 1990.