- Email: [email protected]
Inhibition by dipyridamole of neutrophil adhesion to vascular endothelium during coronary bypass surgery
Inhibition by Dipyridamole of Neutrophil Adhesion to Vascular Endothelium During Coronary Bypass Surgery Massimo Chello, MD, Pasquale Mastroroberto, MD, Emanuele Malta, MD, Francesco Cirillo, MD, and Vittoria Celi, MD Units of Cardiovascular Surgery and Clinical Chemistry, Department of Experimental and Clinical Medicine, Medical School of Catanzaro, Catanzaro, Italy
Background. Release of reactive oxygen radicals by activated neutrophils and neutrophil adhesion to endothelial cells have been observed after cardiopulmonary bypass. The aim of the present study was to evaluate the effects of preoperative dipyridamole treatment on neutrophil superoxide anion generation and endothelial cell– neutrophil interactions. Methods. Two groups of patients scheduled for elective coronary artery bypass grafting were randomized to receive oral dipyridamole or a placebo. Nitro blue tetrazolium scores of circulating neutrophils, neutrophil CD11b/CD18 expression, and their adhesion to human umbilical vein endothelial cells were assayed before anesthesia, 30 minutes after the beginning of cardiopulmonary bypass, at the end of bypass, and 60 minutes postoperatively. Results. In both groups, cardiopulmonary bypass resulted in a significant increase in nitro blue tetrazolium scores in circulating neutrophils as well as a significant
increase in both neutrophil CD11b/CD18 expression and neutrophil adhesion to endothelial cells. The extent of neutrophil superoxide anion generation was higher in the control group; a significant (p < 0.01) reduction in neutrophil adhesion to endothelial cells was observed 1 hour postoperatively in the dipyridamole group. In 5 patients treated with dipyridamole, the incubation of activated polymorphonuclear leukocytes with adenosine deaminase significantly increased their adhesion to endothelial cells (p < 0.05). Conclusions. Our study demonstrated that preoperative treatment with oral dipyridamole significantly reduces both neutrophil superoxide anion generation and extent of neutrophil adhesion to endothelial cells after coronary bypass grafting procedures with cardiopulmonary bypass. The mechanism is probably mediated by endogenous adenosine. (Ann Thorac Surg 1999;67:1277– 82) © 1999 by The Society of Thoracic Surgeons
C
thelial cell (HUVEC) monolayers in vitro. Because the extent of neutrophil activation during CPB can be responsible for postoperative morbidity [1, 2], efforts to reduce this activation are of potential clinical importance. Dipyridamole is a pyridino-pyrimidine compound originally introduced as an antianginal agent owing to its coronary artery vasodilative properties. Dipyridamole has been shown to decrease platelet aggregation and adhesiveness and thus has recently come to be widely used as an antithrombotic agent [10, 11]. In addition, it has been reported to directly inhibit human PMNs from generating active oxygen metabolites and to affect the scavenging action on superoxide anions and hydroxyl radicals [12]. The purpose of the present study was to investigate the ability of dipyridamole to modulate superoxide anion generation by human neutrophils and endothelial cell– neutrophil interactions in patients undergoing CPB.
ardiopulmonary bypass (CPB) has been demonstrated to promote major activation of polymorphonuclear leukocytes (PMNs) [1, 2]. This in turn plays an important role in coronary and pulmonary endothelial injury associated with ischemia and reperfusion [3, 4]. Further, activated PMNs have been shown to injure endothelial cells in vitro [5]. Neutrophil adherence likely contributes to such injury by both capillary plugging of the myocardial microvasculature [3] and by releasing reactive oxygen species and proteolytic enzymes [4, 6]. Activation products from this reaction may be responsible for the postoperative organ dysfunction frequently observed after procedures with CPB [2, 7]. The adhesion of neutrophils to the endothelium is directly mediated by specific “adhesion” molecules on the neutrophil and endothelial cell surfaces. Several studies [8, 9] have indicated that the CD11b/CD18 complex of leukocyte adhesion receptors are involved in the adhesion of chemotactically stimulated normal human neutrophils to unstimulated human umbilical vein endoAccepted for publication Oct 26, 1998. Address reprint requests to Dr Chello, Via S. Giacomo dei Capri 29, 80128 Naples, Italy; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Forty patients scheduled for nonurgent aortocoronary bypass grafting were randomized to receive oral dipyrid0003-4975/99/$20.00 PII S0003-4975(99)00173-3
1278
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
Table 1. Summary of Patient Dataa,b Variable Age (y) NYHA class Accompanying drugs b-blockers Ca21 blockers Nitrates Total CPB time (min) Cross-clamp time (min) No. of grafts No. of IMA grafts a b
Control Group (n 5 20)
Dipyridamole Group (n 5 20)
51.7 6 2.3 2.3 6 0.6
53.2 6 2.4 2.4 6 0.4
5 13 16 98 6 7 54.4 6 5.5
6 15 15 101 6 8.1 58.7 6 6.3
2.5 6 0.5 16
2.6 6 0.4 17
Where applicable, data are shown as the mean 6 the standard error. There were no significant differences between groups.
CPB 5 cardiopulmonary bypass; IMA 5 internal mammary artery; NYHA 5 New York Heart Association.
amole or a placebo. All patients had angina on effort and were receiving some combination of b-adrenergic blocking agents, nitrate vasodilators, and calcium-channel blocking agents. The two groups were similar in mean age, male preponderance, preoperative hemodynamic data, and number of coronary arteries revascularized (Table 1). Also similar were the mean total pump run and the mean cross-clamp time. Patients had good left ventricular function as judged by preoperative cardiac angiogram (left ventricular ejection fraction . 0.55) and had no evidence of pulmonary disease as judged by chest roentgenogram and lung volumes. Treatment with dipyridamole (Persantine; Boehringer Ingelheim, Ingelheim, Germany) was started 36 hours before operation (100 mg orally four times per day). On the morning of the operation, the last preoperative dose was given with the antianginal medications. No dipyridamole was administered intraoperatively. This protocol has been demonstrated to maintain adequate plasma levels of dipyridamole during CPB and up to 48 hours postoperatively [13].
Operative Procedure The same standard anesthesia protocol was used in all patients. After premedication, a Swan-Ganz catheter was positioned in the central pulmonary artery, and a radial artery cannula was inserted. Anesthesia was induced with sodium thiopental, and muscle relaxation was achieved with pancuronium bromide; analgesia was provided with fentanyl. The pump (Sarns roller pumps and Dideco hollow-fiber oxygenators) was primed with 2,000 mL of Ringer’s lactate. The heart was exposed through a median sternotomy, and heparin sodium (300 U/kg) was administered intravenously before CPB to produce an activated clotting time greater than 400 seconds. The hematocrit level was maintained between 20% and 25%, and pump flows were kept between 2.0 and 2.5 L z min21 z m22 during moderate hypothermia to maintain a mean arterial pressure between 50 and 70 mm Hg.
Ann Thorac Surg 1999;67:1277– 82
Cardioplegia was achieved with ice-cold St. Thomas’ Hospital solution infused into the ascending aorta. The left ventricle was vented through the aortic root. After completion of the distal anastomosis, the aortic crossclamp was removed, and the aortovenous anastomoses were performed while the patient was being rewarmed to 37°C. After decannulation, protamine sulfate (10 mg/mL) (Eli Lilly and Company, Indianapolis, IN) was administered intravenously to neutralize the heparin (1 mg for each 300 units of heparin).
Study Protocol Blood samples were taken from the radial artery catheter before the induction of anesthesia, 30 minutes after the start of CPB, at the end of CPB, and 60 minutes postoperatively. The samples were collected in cooled heparinized syringes that were immediately capped and stored in ice until separation and analysis. Blood samples were assayed for oxygen and carbon dioxide tensions, pH, and oxygen saturation. The study protocol was approved by the Ethics Committee of the Medical School of Catanzaro. Informed consent was obtained from each patient.
Hemodynamic Measurements From the moment patients arrived in the intensive care unit, systolic, diastolic, and mean systemic and pulmonary artery pressures, mean pulmonary capillary wedge pressure, systemic vascular resistance, and cardiac output were monitored every 4 hours for 24 hours. Left ventricular stroke work index and cardiac index were calculated using standard formulas. Hemodynamic instability was defined as mean arterial blood pressure of less than 70 mm Hg necessitating dopamine hydrochloride support for stabilization.
Neutrophil Isolation Neutrophils were isolated by Ficoll-Hypaque densitygradient centrifugation, dextran sedimentation, and hypotonic lysis of erythrocytes. Neutrophils were suspended in Hanks’ balanced salt solution, free from phenol red, Ca21, and Mg21 and containing 0.25% bovine serum albumin. The final cell preparation had 98% 6 2% neutrophils. The neutrophils were maintained on ice in Hanks’ balanced salt solution at 1 to 5 3 106 cells/mL until usage. Isolated PMNs were more than 99% pure as assessed by Wrights-stained cytocentrifuge preparation and more than 99% viable as assessed by exclusion of trypan blue.
Neutrophil CD11b Expression Neutrophil CD11b expression was detected by indirect immunofluorescence and flow cytometry. Briefly, 1 mL of blood was drawn from the oxygenator into a heparinized syringe that was kept at room temperature to minimize the effects of cooling and rewarming on PMN integrin expression. Samples of blood in aliquots of 100 mL were put in polypropylene tubes and stained with anti-CD11b monoclonal antibody CBL145 (Cymbus Biosciences, Hants, UK) as the primary reagent and fluorescein isothiocyanate– conjugated goat anti-mouse immunoglobu-
Ann Thorac Surg 1999;67:1277– 82
lin (Cymbus Biosciences) as the secondary reagent. Controls included cells stained with the second reagent alone, and cells stained with an irrelevant isotypematched control monoclonal antibody. No other primary monoclonal antibodies were employed. Red blood cells were removed by hypotonic lysis. Flow cytometry was performed on a FACScan (Becton Dickinson, Mountain View, CA). The granulocyte population was identified by its forward and orthogonal light scatter characteristics. Green and red amplifier gains were calibrated with beads before each experiment (Flow Cytometry Standards Corp, Research Triangle Park, NC) to ensure that relative fluorescence values were comparable between experiments. The mean fluorescence for each specimen was calculated from the fluorescence distribution (5,000 events) using the FACScan Research Software, version B (Becton Dickinson). In the flow cytometry studies, the logarithmic mean fluorescence values obtained from the histograms were converted mathematically into a relative fluorescence value and expressed as percent increase over the observed baseline values.
Cell Culture Human umbilical vein endothelial cells were isolated from fresh umbilical cords by the method of Jaffe [14]. The HUVECs were placed in 75-mm plastic culture flasks and grown in Medium 199 supplemented with 20% fetal bovine serum, 25 mg/mL of endothelial cell growth supplement, and 90 mg/mL of heparin (growth supplement and heparin from Sigma Chemical Co, St. Louis, MO) at 37°C in a humidified atmosphere of 95% air and 5% carbon dioxide atmosphere. All media were prepared with endotoxin-free water (Baxter, Columbia, MD) and filtered with Zetapore Filters (Cuno Life Science Division). Endotoxin-free plasticware and glassware were used in all experiments. For experimental studies, confluent HUVEC monolayers (passage 3 to 7) were trypsinized and replated on 48-well plates (Costar, Cambridge, MA) precoated with a sterile solution of 1.5% gelatin and grown to confluence for neutrophil adhesion studies. The HUVECs were grown to confluence and used within 72 hours. The identity of some endothelial cultures was checked by indirect staining with fluorescein isothiocyanate–labeled factor VIII antibody (polyclonal IgG) (Sigma-Aldrich, Milan, Italy).
PMN Adherence Assay The PMNs were labeled with a hydrophobic fluorescent compound (3-39-dioctadecyloxacarbocyanine perchlorate (DiI) (Sigma-Aldrich, Milan, Italy) as described by Lo and co-workers [15]. Cells, 4 to 8 3 106/mL, were incubated with 50 mg/mL of DiI in HAP buffer (137 nM NACl, 2.7 mM KCI, 0.9 mM CaCl2, 0.5 mM MgCl2, 8 mM phosphate, 0.05% glucose, 0.5 mg/mL albumin, 0.3 U/mL aprotinin, pH 7.1) for 10 minutes at 0°C; unbound dye was removed by three washes with the buffer; and labeled PMNs were resuspended in Medium 199 for the adhesion assay. The DiI–labeled PMNs (10 mL of 106 cells/mL) were added to twice-washed HUVEC
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
1279
monolayers. Adhesion was allowed to proceed for 15 minutes at 37°C, and unbound PMNs were removed by three washes with Medium 199. Residual adherent PMNs on HUVEC surfaces were counted manually on an inverted microscope equipped for fluorescence using an IF355-550 filter. Values of 5 replicates were averaged, and variations between replicates were small (,10%). In addition, to evaluate whether increased endogenous adenosine is responsible for the activity of dipyridamole, other experiments were performed (n 5 5) in which PMN– endothelial cell adhesion was assayed in the presence of adenosine deaminase.
Nitro Blue Tetrazolium One specific form of neutrophil activation is the capacity to reduce nitro blue tetrazolium (NBT), which is associated with an enhanced generation of superoxide anion radicals. This is the pivotal phenomenon of the respiratory burst after neutrophil activation. The NBT test was performed on fresh blood as previously described [16].
Statistical Analysis All values are expressed as the mean 6 the standard error. Comparisons between and within groups were made using two-way analysis of variance for repeated measures followed by the Scheffe´ test for multiple comparisons. Relationships between independent variables were assessed by linear regression analysis. Significance was set at a p value of less than 0.05.
Results Adverse reactions to dipyridamole were infrequent. Four patients complained of nausea, and 1 patient had a urticarial skin reaction. All patients survived the operation and are well. There were no significant differences between the two groups in terms of cardiac index, left ventricular stroke work index, mean pulmonary capillary wedge pressure, heart rate, and pulmonary systolic pressure in the postoperative period. Four patients in the control group and 5 in the dipyridamole group required dopamine for a mild low cardiac output state. No significant difference was found between the two groups in terms of postoperative blood loss through the thoracic drains during stay in the intensive care unit.
NBT The NBT scores are shown in Figure 1. The maximum NBT score was recorded in both groups at the end of CPB (control group, 13.9% 6 1.2%; p , 0.01 versus baseline; dipyridamole group, 10.4% 6 1%; p , 0.01 versus baseline). In both groups, the NBT scores 60 minutes postoperatively were still significantly higher compared with baseline, though they were going down. Interestingly, the NBT scores were significantly higher in the control group than in the dipyridamole group at the end of CPB and 60 minutes postoperatively.
1280
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
Fig 1. Nitro blue tetrazolium (NBT) scores of circulating neutrophils in arterial samples from control and dipyridamole-treated (dipy) groups. Data are shown as the mean 6 the standard error of the mean. (base 5 baseline; end CPB 5 end of cardiopulmonary bypass; 301cpb 5 30 minutes after start of cardiopulmonary bypass; 601po 5 60 minutes postoperatively; ** 5 p , 0.01 versus control; * 5 p , 0.05 versus control.)
Neutrophil CD11b Expression Figure 2 shows neutrophil CD11b expression in control and treated patients. A significant increase in CD11b expression was observed in both groups throughout the whole experimental period, with the highest observed values at 60 minutes postoperatively. However, neutrophil CD11b expression was significantly greater in control patients than in dipyridamole-treated patients at all sampling points during and after CPB.
Ann Thorac Surg 1999;67:1277– 82
Fig 3. Adherence of circulating neutrophils to human umbilical vein endothelial cell monolayers during and after cardiopulmonary bypass (CPB) in control and dipyridamole-treated (dipy) groups. (601p.o. 5 60 minutes postoperatively; * 5 p , 0.01 versus control.)
utes postoperatively. In the linear regression analysis, no significant correlation was found between CD11b expression and neutrophil adhesion to HUVEC monolayers. To evaluate whether inhibition of dipyridamole was mediated by increased endogenous adenosine, PMNs collected 60 minutes postoperatively were incubated with HUVECs in the presence of adenosine deaminase (Sigma Chemicals, Milan, Italy). As shown in Figure 4, the percentage of neutrophil– endothelial cell adhesion was significantly affected by pretreatment with adenosine deaminase ( p , 0.05).
PMN Adhesion Figure 3 shows the percentage of neutrophil adhesion to HUVECs in the control and dipyridamole-treated groups. In both groups, a significant increase ( p , 0.05 versus baseline) in neutrophil adhesion was observed at the end of CPB, with the highest values seen 60 minutes postoperatively ( p , 0.001 versus baseline). In the between-group comparisons, a greater but not significant difference in neutrophil adhesion was observed in control patients during the operative period, whereas a significant difference ( p , 0.01) was observed 60 min-
Comment
Fig 2. Expression of CD11b/CD18 on the surface of circulating neutrophils in arterial samples from control and dipyridamole-treated (dipy) groups. Abbreviations are the same as in Figure 1.
Fig 4. Effect of adenosine deaminase (ada) on neutrophil– endothelial cell adhesion 60 minutes postoperatively in patient pretreated with dipyridamole (dipy). (* 5 p , 0.05.)
The systemic inflammatory response that occurs to some extent in all patients after CPB remains a major cause of morbidity and mortality, and although technical advances have improved the safety of cardiac operations, postoperative dysfunction of lung and other organs still occurs [2, 4]. A large body of evidence suggests that the extent of neutrophil activation can be responsible for the
Ann Thorac Surg 1999;67:1277– 82
postoperative development of cardiac, renal, and pulmonary dysfunction as well as abnormal bleeding and pulmonary insufficiency [1, 2, 4]. In the present study, we demonstrate that dipyridamole pretreatment preceding cardiac surgical procedures affects several aspects involved in neutrophil activation. This study led to three important findings: we defined the time course of neutrophil oxidative burst and neutrophil–endothelial cell adhesion; we demonstrated that dipyridamole is a potent inhibitor of PMN respiratory burst and adhesion to HUVECs after CPB; and we showed that the inhibitory activity of dipyridamole is mediated by endogenous adenosine. Our results are consistent with those of previous studies in which dipyridamole was found to inhibit neutrophil activation, including the process of cell adhesion and release of oxygen metabolites such as superoxide anions. Teoh and associates [17] evaluated the effect of dipyridamole on myocardial platelet and leukocyte deposition in a canine model of acute regional myocardial ischemia with reperfusion during cardioplegia on CPB. They found a significant reduction in the deposition of both platelets and leukocytes in the reperfused myocardial regions of animals pretreated with dipyridamole. Colli and Tremoli [18] also showed that dipyridamole significantly inhibited the generation of superoxide anions by neutrophils in response to stimuli that increase intracellular calcium levels. A similar inhibitory profile of the drug was observed when superoxide anion generation was evaluated in both whole blood and isolated cells, results indicating that plasma components and red cells did not interfere with the inhibitory activity of the drug on the respiratory burst. In a similar fashion, Suzuki and colleagues [11] demonstrated that pretreating human PMNs with dipyridamole at therapeutic concentrations significantly inhibited the generation of active oxygen metabolites after PMN stimulation. On the basis of the data from this study, we conclude that dipyridamole at plasma concentrations inhibits active oxygen metabolic generation by neutrophils and also modulates neutrophil adhesion. Although the exact mechanism of action is not fully clarified, our data support the hypothesis that endogenous adenosine modulates expression of different adhesion molecules on the neutrophil surface, as the administration of adenosine deaminase significantly antagonizes the effects of dipyridamole on neutrophil adhesion. Dipyridamole potently inhibits the uptake of adenosine by both erythrocytes and endothelial cells [10] and thus may increase the local concentrations of this substance. Increased plasma concentrations of adenosine have been reported after intravenous infusion of dipyridamole (0.4 mg/kg of body weight for 8 minutes) in healthy volunteers [19], and therapeutic concentrations of dipyridamole (3.2 mmol/L) have been reported to inhibit adenosine diphosphate and collagen-induced aggregation in whole blood [20]. Although the underlying molecular mechanisms have not been fully understood, a comparison between the effects of dipyridamole described here and those reported for adenosine by others indicates that dipyridam-
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
1281
ole mimics several effects of adenosine on neutrophils. Adenosine is an antiinflammatory agent that is released predominantly from the vascular endothelium and that might modulate neutrophil activation mainly by attenuating the neutrophil superoxide production [21], inhibiting the neutrophil adherence and cytotoxicity to endothelium [22, 23]. It is present in plasma at concentrations of at least 0.3 mmol/L [24] and is released locally by ischemic tissues and by activated or injured endothelium [25]. Recent data suggest that, similar to its effect on platelets, dipyridamole modulates neutrophil function indirectly, that is, through an increase of extracellular adenosine, which, in turn, inhibits both superoxide anion generation by leukocytes and expression of CD11b/ CD18. To test this hypothesis, studies with adenosine deaminase, which degrades extracellular adenosine to inactive inosine, have been performed. Thiel and coworkers [26] evaluated the effect of adenosine on the expression of adhesion molecules on human PMNs. When PMNs were activated by formyl-metyonil-leucylphenylalanine (FMLP), the number of cell surface b2 integrins increased fivefold, whereas L-selectin molecules were completely shed. When extracellular concentrations on endogenously formed adenosine were enhanced by the nucleoside uptake inhibitor dipyridamole, up-regulation of b2 integrins and shedding of L-selectin were inhibited. Both effects were reversed by the enzyme adenosine deaminase, a finding suggesting that endogenously formed adenosine may play an important role in the regulation of b2 integrins and L-selectin on human PMNs. Wollner and associates [27] showed that adenosine significantly inhibited the upregulation of Mac-1 expression on neutrophils stimulated with FMLP in a dosedependent fashion by a mechanism involving adenosine receptors of the A2 type. Colli and Tremoli [18] demonstrated that dipyridamole inhibited the generation of superoxide anions by neutrophils stimulated with FMLP in a concentration-dependent fashion. However, the effects of dipyridamole on superoxide anion generation by whole blood were completely prevented by the addition of adenosine deaminase, indicating that the effects of dipyridamole are mediated by the presence of adenosine in the suspension medium. Although in the present study, pretreatment with dipyridamole greatly reduced both CD11b/CD18 expression and PMN adherence to HUVEC monolayers, it failed to demonstrate a significant relationship between the adhesion glycoprotein CD11b/CD18 and neutrophil adhesion. An explanation for this finding may be that CD11b/CD18 upregulation is not an absolute requirement for adhesive interactions between neutrophils and vascular endothelium, whereas qualitative alteration in the CD11b/CD18 molecule may be more important to neutrophil– endothelial cell adhesion than quantitative modifications [5]. In this study, we demonstrated that dipyridamole inhibits the increase in expression of activated PMNs. Our observations, however, do not eliminate the possibility that dipyridamole might modulate
1282
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
the avidity of neutrophil CD11b/CD18 Mac-1 for its ligands. In conclusion, although further studies are advisable to clarify the exact mechanism of action, we can affirm on the basis of our data that pretreatment with dipyridamole significantly inhibits both neutrophil activation and adhesion to the venous endothelium in patients undergoing coronary artery bypass grafting with CPB. In this study, changes in neutrophil adherence were studied in an experimental model of cultured HUVECs. Although the adherence assay that we used is very reliable in regard to that specific vascular bed, we emphasize that it is not possible to affirm whether similar changes also occur with systemic endothelium. Therefore, examination of the endothelium from other vascular districts are necessary.
References 1. Herskowitz A, Mangano DT. Inflammatory cascade. A final common pathway for perioperative injury? Anesthesiology 1996;85:957– 60. 2. Hall RI, Smith MS, Rocker G. The systemic inflammatory response to cardiopulmonary bypass: pathophysiological, therapeutic, and pharmacological considerations. Anesth Analg 1997;85:766– 82. 3. Thomas DD, Sharar SR, Winn RK, et al. CD18-independent mechanism of neutrophil emigration in the rabbit lung after ischemia-reperfusion. Ann Thorac Surg 1995;60:1360– 6. 4. Tonz M, von Segesser LK, Fehr J, Schmid ER, Turina MI. Acute lung injury during cardiopulmonary bypass. Are the neutrophils responsible? Chest 1995;108:1551– 6. 5. Carlos TM, Harlan JM. Leukocyte-endothelial adhesion molecules. Blood 1994;84:2068 –101. 6. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci USA 1991;88:4651–5. 7. Albelda SM, Smith CW, Ward PA. Adhesion molecules and inflammatory injury. FASEB J 1994;8:504–12. 8. Zimmerman GA, McIntyre TM. Neutrophil adherence to human endothelium in vitro occurs by CDw18 (Mol 1, MAC-1/LFA-1/GP 150,95) glycoprotein-dependent and -independent mechanisms. J Clin Invest 1988;81:531–7. 9. McElroy FA, Philp RB. Relative potencies of dipyridamole and related agents as inhibitors of cyclic nucleotide phosphodiesterases: possible explanation of mechanism of inhibition of platelet function. Life Sci 1975;17:1479– 83. 10. FitzGerald GA. Dipyridamole. N Engl J Med 1987;316: 1247–57. 11. Suzuki S, Sugai K, Sato H, Sakatume M, Arakawa M. Inhibition of active oxygen generation by dipyridamole in human polymorphonuclear leukocytes. Eur J Pharmacol 1992;227:395– 401.
Ann Thorac Surg 1999;67:1277– 82
12. Tremoli E, Colli S. Dipyridamole inhibits superoxide anion generation by human neutrophils [Letter]. Thromb Haemost 1988;59:34. 13. Teoh KH, Christakis GT, Weisel RD, et al. Dipyridamole preserved platelets and reduced blood loss after cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;96:332– 41. 14. Jaffe EA. Culture and identification of large vessel endothelial cells. In: Jaffe EA, ed. Biology of endothelial cells. The Hague: Martinus Nijhoff, 1984, 1. 15. Lo SK, Detmers PA, Levin SM, Wright SD. Transient adhesion of neutrophils to endothelium. J Exp Med 1989;169:1779–93. 16. Chello M, Mastroroberto P, Celi V, Romano F, Marchese AR, Colonna A. Reduction by indobufen of neutrophil activation in peripheral arterial occlusive disease. J Cardiovasc Pharmacol 1996;27:417–23. 17. Teoh KH, Christakis GT, Weisel RD, et al. Dipyridamole reduced myocardial platelet and leukocyte deposition following ischemia and cardioplegia. J Surg Res 1987;42:642–52. 18. Colli S, Tremoli E. Multiple effects of dipyridamole on neutrophils and mononuclear leukocytes: adenosine-dependent and adenosine-independent mechanisms. J Lab Clin Med 1991;118:135– 45. 19. Sollevi A, Ostergren J, Fagrell B, Hjemdahl P. Theophylline antagonizes cardiovascular responses to dipyridamole in man without affecting increases in plasma adenosine. Acta Physiol Scand 1984;121:165–71. 20. Saniabadi AR, Lowe GDO, Barbenel JC, Forbes CD. A comparison of spontaneous platelet aggregation in whole blood with platelet rich plasma: additional evidence for the role of ADP. Thromb Haemost 1984;51:115– 8. 21. Cronstein BN, Kramer SB, Weissmann G, Hirschhorn R. Adenosine: a physiologic modulator of superoxide anion generation by human neutrophils. J Exp Med 1983;158: 1160–77. 22. Cronstein BN, Levin RI, Philips M, Hirschhorn R, Abramson SB, Weissmann G. Neutrophil adherence to endothelium is enhanced via adenosine A1 receptors and inhibited via adenosine A2 receptors. J Immunol 1992;148:2201– 6. 23. Bullough DA, Magill MJ, Firestein GS, Mullane KM. Adenosine activates A2 receptors to inhibit neutrophil adhesion to isolated cardiac myocytes. J Immunol 1995;155:2579– 86. 24. Hirschhorn R, Roegner-Maniscalco B, Kuritsky L, Rosen FS. Bone marrow transplantation only partially restores purine metabolites to normal in adenosine deaminase deficient patients. J Clin Invest 1981;68:1387–93. 25. Ely SW, Berne RM. Protective effects of adenosine in myocardial ischemia. Circulation 1992;85:893–904. 26. Thiel M, Chambers JD, Chouker A, et al. Effect of adenosine on the expression of beta (2) integrins and L-selectin of human polymorphonuclear leukocytes in vitro. J Leukocyte Biol 1996;59:671– 82. 27. Wollner A, Wollner S, Smith JB. Acting via A2 receptors, adenosine inhibits the upregulation of mac-1 (CD11b/CD18) expression on FMLP-stimulated neutrophils. Am J Respir Cell Mol Biol 1993;9:179– 85.
Background. Release of reactive oxygen radicals by activated neutrophils and neutrophil adhesion to endothelial cells have been observed after cardiopulmonary bypass. The aim of the present study was to evaluate the effects of preoperative dipyridamole treatment on neutrophil superoxide anion generation and endothelial cell– neutrophil interactions. Methods. Two groups of patients scheduled for elective coronary artery bypass grafting were randomized to receive oral dipyridamole or a placebo. Nitro blue tetrazolium scores of circulating neutrophils, neutrophil CD11b/CD18 expression, and their adhesion to human umbilical vein endothelial cells were assayed before anesthesia, 30 minutes after the beginning of cardiopulmonary bypass, at the end of bypass, and 60 minutes postoperatively. Results. In both groups, cardiopulmonary bypass resulted in a significant increase in nitro blue tetrazolium scores in circulating neutrophils as well as a significant
increase in both neutrophil CD11b/CD18 expression and neutrophil adhesion to endothelial cells. The extent of neutrophil superoxide anion generation was higher in the control group; a significant (p < 0.01) reduction in neutrophil adhesion to endothelial cells was observed 1 hour postoperatively in the dipyridamole group. In 5 patients treated with dipyridamole, the incubation of activated polymorphonuclear leukocytes with adenosine deaminase significantly increased their adhesion to endothelial cells (p < 0.05). Conclusions. Our study demonstrated that preoperative treatment with oral dipyridamole significantly reduces both neutrophil superoxide anion generation and extent of neutrophil adhesion to endothelial cells after coronary bypass grafting procedures with cardiopulmonary bypass. The mechanism is probably mediated by endogenous adenosine. (Ann Thorac Surg 1999;67:1277– 82) © 1999 by The Society of Thoracic Surgeons
C
thelial cell (HUVEC) monolayers in vitro. Because the extent of neutrophil activation during CPB can be responsible for postoperative morbidity [1, 2], efforts to reduce this activation are of potential clinical importance. Dipyridamole is a pyridino-pyrimidine compound originally introduced as an antianginal agent owing to its coronary artery vasodilative properties. Dipyridamole has been shown to decrease platelet aggregation and adhesiveness and thus has recently come to be widely used as an antithrombotic agent [10, 11]. In addition, it has been reported to directly inhibit human PMNs from generating active oxygen metabolites and to affect the scavenging action on superoxide anions and hydroxyl radicals [12]. The purpose of the present study was to investigate the ability of dipyridamole to modulate superoxide anion generation by human neutrophils and endothelial cell– neutrophil interactions in patients undergoing CPB.
ardiopulmonary bypass (CPB) has been demonstrated to promote major activation of polymorphonuclear leukocytes (PMNs) [1, 2]. This in turn plays an important role in coronary and pulmonary endothelial injury associated with ischemia and reperfusion [3, 4]. Further, activated PMNs have been shown to injure endothelial cells in vitro [5]. Neutrophil adherence likely contributes to such injury by both capillary plugging of the myocardial microvasculature [3] and by releasing reactive oxygen species and proteolytic enzymes [4, 6]. Activation products from this reaction may be responsible for the postoperative organ dysfunction frequently observed after procedures with CPB [2, 7]. The adhesion of neutrophils to the endothelium is directly mediated by specific “adhesion” molecules on the neutrophil and endothelial cell surfaces. Several studies [8, 9] have indicated that the CD11b/CD18 complex of leukocyte adhesion receptors are involved in the adhesion of chemotactically stimulated normal human neutrophils to unstimulated human umbilical vein endoAccepted for publication Oct 26, 1998. Address reprint requests to Dr Chello, Via S. Giacomo dei Capri 29, 80128 Naples, Italy; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Forty patients scheduled for nonurgent aortocoronary bypass grafting were randomized to receive oral dipyrid0003-4975/99/$20.00 PII S0003-4975(99)00173-3
1278
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
Table 1. Summary of Patient Dataa,b Variable Age (y) NYHA class Accompanying drugs b-blockers Ca21 blockers Nitrates Total CPB time (min) Cross-clamp time (min) No. of grafts No. of IMA grafts a b
Control Group (n 5 20)
Dipyridamole Group (n 5 20)
51.7 6 2.3 2.3 6 0.6
53.2 6 2.4 2.4 6 0.4
5 13 16 98 6 7 54.4 6 5.5
6 15 15 101 6 8.1 58.7 6 6.3
2.5 6 0.5 16
2.6 6 0.4 17
Where applicable, data are shown as the mean 6 the standard error. There were no significant differences between groups.
CPB 5 cardiopulmonary bypass; IMA 5 internal mammary artery; NYHA 5 New York Heart Association.
amole or a placebo. All patients had angina on effort and were receiving some combination of b-adrenergic blocking agents, nitrate vasodilators, and calcium-channel blocking agents. The two groups were similar in mean age, male preponderance, preoperative hemodynamic data, and number of coronary arteries revascularized (Table 1). Also similar were the mean total pump run and the mean cross-clamp time. Patients had good left ventricular function as judged by preoperative cardiac angiogram (left ventricular ejection fraction . 0.55) and had no evidence of pulmonary disease as judged by chest roentgenogram and lung volumes. Treatment with dipyridamole (Persantine; Boehringer Ingelheim, Ingelheim, Germany) was started 36 hours before operation (100 mg orally four times per day). On the morning of the operation, the last preoperative dose was given with the antianginal medications. No dipyridamole was administered intraoperatively. This protocol has been demonstrated to maintain adequate plasma levels of dipyridamole during CPB and up to 48 hours postoperatively [13].
Operative Procedure The same standard anesthesia protocol was used in all patients. After premedication, a Swan-Ganz catheter was positioned in the central pulmonary artery, and a radial artery cannula was inserted. Anesthesia was induced with sodium thiopental, and muscle relaxation was achieved with pancuronium bromide; analgesia was provided with fentanyl. The pump (Sarns roller pumps and Dideco hollow-fiber oxygenators) was primed with 2,000 mL of Ringer’s lactate. The heart was exposed through a median sternotomy, and heparin sodium (300 U/kg) was administered intravenously before CPB to produce an activated clotting time greater than 400 seconds. The hematocrit level was maintained between 20% and 25%, and pump flows were kept between 2.0 and 2.5 L z min21 z m22 during moderate hypothermia to maintain a mean arterial pressure between 50 and 70 mm Hg.
Ann Thorac Surg 1999;67:1277– 82
Cardioplegia was achieved with ice-cold St. Thomas’ Hospital solution infused into the ascending aorta. The left ventricle was vented through the aortic root. After completion of the distal anastomosis, the aortic crossclamp was removed, and the aortovenous anastomoses were performed while the patient was being rewarmed to 37°C. After decannulation, protamine sulfate (10 mg/mL) (Eli Lilly and Company, Indianapolis, IN) was administered intravenously to neutralize the heparin (1 mg for each 300 units of heparin).
Study Protocol Blood samples were taken from the radial artery catheter before the induction of anesthesia, 30 minutes after the start of CPB, at the end of CPB, and 60 minutes postoperatively. The samples were collected in cooled heparinized syringes that were immediately capped and stored in ice until separation and analysis. Blood samples were assayed for oxygen and carbon dioxide tensions, pH, and oxygen saturation. The study protocol was approved by the Ethics Committee of the Medical School of Catanzaro. Informed consent was obtained from each patient.
Hemodynamic Measurements From the moment patients arrived in the intensive care unit, systolic, diastolic, and mean systemic and pulmonary artery pressures, mean pulmonary capillary wedge pressure, systemic vascular resistance, and cardiac output were monitored every 4 hours for 24 hours. Left ventricular stroke work index and cardiac index were calculated using standard formulas. Hemodynamic instability was defined as mean arterial blood pressure of less than 70 mm Hg necessitating dopamine hydrochloride support for stabilization.
Neutrophil Isolation Neutrophils were isolated by Ficoll-Hypaque densitygradient centrifugation, dextran sedimentation, and hypotonic lysis of erythrocytes. Neutrophils were suspended in Hanks’ balanced salt solution, free from phenol red, Ca21, and Mg21 and containing 0.25% bovine serum albumin. The final cell preparation had 98% 6 2% neutrophils. The neutrophils were maintained on ice in Hanks’ balanced salt solution at 1 to 5 3 106 cells/mL until usage. Isolated PMNs were more than 99% pure as assessed by Wrights-stained cytocentrifuge preparation and more than 99% viable as assessed by exclusion of trypan blue.
Neutrophil CD11b Expression Neutrophil CD11b expression was detected by indirect immunofluorescence and flow cytometry. Briefly, 1 mL of blood was drawn from the oxygenator into a heparinized syringe that was kept at room temperature to minimize the effects of cooling and rewarming on PMN integrin expression. Samples of blood in aliquots of 100 mL were put in polypropylene tubes and stained with anti-CD11b monoclonal antibody CBL145 (Cymbus Biosciences, Hants, UK) as the primary reagent and fluorescein isothiocyanate– conjugated goat anti-mouse immunoglobu-
Ann Thorac Surg 1999;67:1277– 82
lin (Cymbus Biosciences) as the secondary reagent. Controls included cells stained with the second reagent alone, and cells stained with an irrelevant isotypematched control monoclonal antibody. No other primary monoclonal antibodies were employed. Red blood cells were removed by hypotonic lysis. Flow cytometry was performed on a FACScan (Becton Dickinson, Mountain View, CA). The granulocyte population was identified by its forward and orthogonal light scatter characteristics. Green and red amplifier gains were calibrated with beads before each experiment (Flow Cytometry Standards Corp, Research Triangle Park, NC) to ensure that relative fluorescence values were comparable between experiments. The mean fluorescence for each specimen was calculated from the fluorescence distribution (5,000 events) using the FACScan Research Software, version B (Becton Dickinson). In the flow cytometry studies, the logarithmic mean fluorescence values obtained from the histograms were converted mathematically into a relative fluorescence value and expressed as percent increase over the observed baseline values.
Cell Culture Human umbilical vein endothelial cells were isolated from fresh umbilical cords by the method of Jaffe [14]. The HUVECs were placed in 75-mm plastic culture flasks and grown in Medium 199 supplemented with 20% fetal bovine serum, 25 mg/mL of endothelial cell growth supplement, and 90 mg/mL of heparin (growth supplement and heparin from Sigma Chemical Co, St. Louis, MO) at 37°C in a humidified atmosphere of 95% air and 5% carbon dioxide atmosphere. All media were prepared with endotoxin-free water (Baxter, Columbia, MD) and filtered with Zetapore Filters (Cuno Life Science Division). Endotoxin-free plasticware and glassware were used in all experiments. For experimental studies, confluent HUVEC monolayers (passage 3 to 7) were trypsinized and replated on 48-well plates (Costar, Cambridge, MA) precoated with a sterile solution of 1.5% gelatin and grown to confluence for neutrophil adhesion studies. The HUVECs were grown to confluence and used within 72 hours. The identity of some endothelial cultures was checked by indirect staining with fluorescein isothiocyanate–labeled factor VIII antibody (polyclonal IgG) (Sigma-Aldrich, Milan, Italy).
PMN Adherence Assay The PMNs were labeled with a hydrophobic fluorescent compound (3-39-dioctadecyloxacarbocyanine perchlorate (DiI) (Sigma-Aldrich, Milan, Italy) as described by Lo and co-workers [15]. Cells, 4 to 8 3 106/mL, were incubated with 50 mg/mL of DiI in HAP buffer (137 nM NACl, 2.7 mM KCI, 0.9 mM CaCl2, 0.5 mM MgCl2, 8 mM phosphate, 0.05% glucose, 0.5 mg/mL albumin, 0.3 U/mL aprotinin, pH 7.1) for 10 minutes at 0°C; unbound dye was removed by three washes with the buffer; and labeled PMNs were resuspended in Medium 199 for the adhesion assay. The DiI–labeled PMNs (10 mL of 106 cells/mL) were added to twice-washed HUVEC
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
1279
monolayers. Adhesion was allowed to proceed for 15 minutes at 37°C, and unbound PMNs were removed by three washes with Medium 199. Residual adherent PMNs on HUVEC surfaces were counted manually on an inverted microscope equipped for fluorescence using an IF355-550 filter. Values of 5 replicates were averaged, and variations between replicates were small (,10%). In addition, to evaluate whether increased endogenous adenosine is responsible for the activity of dipyridamole, other experiments were performed (n 5 5) in which PMN– endothelial cell adhesion was assayed in the presence of adenosine deaminase.
Nitro Blue Tetrazolium One specific form of neutrophil activation is the capacity to reduce nitro blue tetrazolium (NBT), which is associated with an enhanced generation of superoxide anion radicals. This is the pivotal phenomenon of the respiratory burst after neutrophil activation. The NBT test was performed on fresh blood as previously described [16].
Statistical Analysis All values are expressed as the mean 6 the standard error. Comparisons between and within groups were made using two-way analysis of variance for repeated measures followed by the Scheffe´ test for multiple comparisons. Relationships between independent variables were assessed by linear regression analysis. Significance was set at a p value of less than 0.05.
Results Adverse reactions to dipyridamole were infrequent. Four patients complained of nausea, and 1 patient had a urticarial skin reaction. All patients survived the operation and are well. There were no significant differences between the two groups in terms of cardiac index, left ventricular stroke work index, mean pulmonary capillary wedge pressure, heart rate, and pulmonary systolic pressure in the postoperative period. Four patients in the control group and 5 in the dipyridamole group required dopamine for a mild low cardiac output state. No significant difference was found between the two groups in terms of postoperative blood loss through the thoracic drains during stay in the intensive care unit.
NBT The NBT scores are shown in Figure 1. The maximum NBT score was recorded in both groups at the end of CPB (control group, 13.9% 6 1.2%; p , 0.01 versus baseline; dipyridamole group, 10.4% 6 1%; p , 0.01 versus baseline). In both groups, the NBT scores 60 minutes postoperatively were still significantly higher compared with baseline, though they were going down. Interestingly, the NBT scores were significantly higher in the control group than in the dipyridamole group at the end of CPB and 60 minutes postoperatively.
1280
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
Fig 1. Nitro blue tetrazolium (NBT) scores of circulating neutrophils in arterial samples from control and dipyridamole-treated (dipy) groups. Data are shown as the mean 6 the standard error of the mean. (base 5 baseline; end CPB 5 end of cardiopulmonary bypass; 301cpb 5 30 minutes after start of cardiopulmonary bypass; 601po 5 60 minutes postoperatively; ** 5 p , 0.01 versus control; * 5 p , 0.05 versus control.)
Neutrophil CD11b Expression Figure 2 shows neutrophil CD11b expression in control and treated patients. A significant increase in CD11b expression was observed in both groups throughout the whole experimental period, with the highest observed values at 60 minutes postoperatively. However, neutrophil CD11b expression was significantly greater in control patients than in dipyridamole-treated patients at all sampling points during and after CPB.
Ann Thorac Surg 1999;67:1277– 82
Fig 3. Adherence of circulating neutrophils to human umbilical vein endothelial cell monolayers during and after cardiopulmonary bypass (CPB) in control and dipyridamole-treated (dipy) groups. (601p.o. 5 60 minutes postoperatively; * 5 p , 0.01 versus control.)
utes postoperatively. In the linear regression analysis, no significant correlation was found between CD11b expression and neutrophil adhesion to HUVEC monolayers. To evaluate whether inhibition of dipyridamole was mediated by increased endogenous adenosine, PMNs collected 60 minutes postoperatively were incubated with HUVECs in the presence of adenosine deaminase (Sigma Chemicals, Milan, Italy). As shown in Figure 4, the percentage of neutrophil– endothelial cell adhesion was significantly affected by pretreatment with adenosine deaminase ( p , 0.05).
PMN Adhesion Figure 3 shows the percentage of neutrophil adhesion to HUVECs in the control and dipyridamole-treated groups. In both groups, a significant increase ( p , 0.05 versus baseline) in neutrophil adhesion was observed at the end of CPB, with the highest values seen 60 minutes postoperatively ( p , 0.001 versus baseline). In the between-group comparisons, a greater but not significant difference in neutrophil adhesion was observed in control patients during the operative period, whereas a significant difference ( p , 0.01) was observed 60 min-
Comment
Fig 2. Expression of CD11b/CD18 on the surface of circulating neutrophils in arterial samples from control and dipyridamole-treated (dipy) groups. Abbreviations are the same as in Figure 1.
Fig 4. Effect of adenosine deaminase (ada) on neutrophil– endothelial cell adhesion 60 minutes postoperatively in patient pretreated with dipyridamole (dipy). (* 5 p , 0.05.)
The systemic inflammatory response that occurs to some extent in all patients after CPB remains a major cause of morbidity and mortality, and although technical advances have improved the safety of cardiac operations, postoperative dysfunction of lung and other organs still occurs [2, 4]. A large body of evidence suggests that the extent of neutrophil activation can be responsible for the
Ann Thorac Surg 1999;67:1277– 82
postoperative development of cardiac, renal, and pulmonary dysfunction as well as abnormal bleeding and pulmonary insufficiency [1, 2, 4]. In the present study, we demonstrate that dipyridamole pretreatment preceding cardiac surgical procedures affects several aspects involved in neutrophil activation. This study led to three important findings: we defined the time course of neutrophil oxidative burst and neutrophil–endothelial cell adhesion; we demonstrated that dipyridamole is a potent inhibitor of PMN respiratory burst and adhesion to HUVECs after CPB; and we showed that the inhibitory activity of dipyridamole is mediated by endogenous adenosine. Our results are consistent with those of previous studies in which dipyridamole was found to inhibit neutrophil activation, including the process of cell adhesion and release of oxygen metabolites such as superoxide anions. Teoh and associates [17] evaluated the effect of dipyridamole on myocardial platelet and leukocyte deposition in a canine model of acute regional myocardial ischemia with reperfusion during cardioplegia on CPB. They found a significant reduction in the deposition of both platelets and leukocytes in the reperfused myocardial regions of animals pretreated with dipyridamole. Colli and Tremoli [18] also showed that dipyridamole significantly inhibited the generation of superoxide anions by neutrophils in response to stimuli that increase intracellular calcium levels. A similar inhibitory profile of the drug was observed when superoxide anion generation was evaluated in both whole blood and isolated cells, results indicating that plasma components and red cells did not interfere with the inhibitory activity of the drug on the respiratory burst. In a similar fashion, Suzuki and colleagues [11] demonstrated that pretreating human PMNs with dipyridamole at therapeutic concentrations significantly inhibited the generation of active oxygen metabolites after PMN stimulation. On the basis of the data from this study, we conclude that dipyridamole at plasma concentrations inhibits active oxygen metabolic generation by neutrophils and also modulates neutrophil adhesion. Although the exact mechanism of action is not fully clarified, our data support the hypothesis that endogenous adenosine modulates expression of different adhesion molecules on the neutrophil surface, as the administration of adenosine deaminase significantly antagonizes the effects of dipyridamole on neutrophil adhesion. Dipyridamole potently inhibits the uptake of adenosine by both erythrocytes and endothelial cells [10] and thus may increase the local concentrations of this substance. Increased plasma concentrations of adenosine have been reported after intravenous infusion of dipyridamole (0.4 mg/kg of body weight for 8 minutes) in healthy volunteers [19], and therapeutic concentrations of dipyridamole (3.2 mmol/L) have been reported to inhibit adenosine diphosphate and collagen-induced aggregation in whole blood [20]. Although the underlying molecular mechanisms have not been fully understood, a comparison between the effects of dipyridamole described here and those reported for adenosine by others indicates that dipyridam-
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
1281
ole mimics several effects of adenosine on neutrophils. Adenosine is an antiinflammatory agent that is released predominantly from the vascular endothelium and that might modulate neutrophil activation mainly by attenuating the neutrophil superoxide production [21], inhibiting the neutrophil adherence and cytotoxicity to endothelium [22, 23]. It is present in plasma at concentrations of at least 0.3 mmol/L [24] and is released locally by ischemic tissues and by activated or injured endothelium [25]. Recent data suggest that, similar to its effect on platelets, dipyridamole modulates neutrophil function indirectly, that is, through an increase of extracellular adenosine, which, in turn, inhibits both superoxide anion generation by leukocytes and expression of CD11b/ CD18. To test this hypothesis, studies with adenosine deaminase, which degrades extracellular adenosine to inactive inosine, have been performed. Thiel and coworkers [26] evaluated the effect of adenosine on the expression of adhesion molecules on human PMNs. When PMNs were activated by formyl-metyonil-leucylphenylalanine (FMLP), the number of cell surface b2 integrins increased fivefold, whereas L-selectin molecules were completely shed. When extracellular concentrations on endogenously formed adenosine were enhanced by the nucleoside uptake inhibitor dipyridamole, up-regulation of b2 integrins and shedding of L-selectin were inhibited. Both effects were reversed by the enzyme adenosine deaminase, a finding suggesting that endogenously formed adenosine may play an important role in the regulation of b2 integrins and L-selectin on human PMNs. Wollner and associates [27] showed that adenosine significantly inhibited the upregulation of Mac-1 expression on neutrophils stimulated with FMLP in a dosedependent fashion by a mechanism involving adenosine receptors of the A2 type. Colli and Tremoli [18] demonstrated that dipyridamole inhibited the generation of superoxide anions by neutrophils stimulated with FMLP in a concentration-dependent fashion. However, the effects of dipyridamole on superoxide anion generation by whole blood were completely prevented by the addition of adenosine deaminase, indicating that the effects of dipyridamole are mediated by the presence of adenosine in the suspension medium. Although in the present study, pretreatment with dipyridamole greatly reduced both CD11b/CD18 expression and PMN adherence to HUVEC monolayers, it failed to demonstrate a significant relationship between the adhesion glycoprotein CD11b/CD18 and neutrophil adhesion. An explanation for this finding may be that CD11b/CD18 upregulation is not an absolute requirement for adhesive interactions between neutrophils and vascular endothelium, whereas qualitative alteration in the CD11b/CD18 molecule may be more important to neutrophil– endothelial cell adhesion than quantitative modifications [5]. In this study, we demonstrated that dipyridamole inhibits the increase in expression of activated PMNs. Our observations, however, do not eliminate the possibility that dipyridamole might modulate
1282
CHELLO ET AL DIPYRIDAMOLE AND NEUTROPHIL ADHESION
the avidity of neutrophil CD11b/CD18 Mac-1 for its ligands. In conclusion, although further studies are advisable to clarify the exact mechanism of action, we can affirm on the basis of our data that pretreatment with dipyridamole significantly inhibits both neutrophil activation and adhesion to the venous endothelium in patients undergoing coronary artery bypass grafting with CPB. In this study, changes in neutrophil adherence were studied in an experimental model of cultured HUVECs. Although the adherence assay that we used is very reliable in regard to that specific vascular bed, we emphasize that it is not possible to affirm whether similar changes also occur with systemic endothelium. Therefore, examination of the endothelium from other vascular districts are necessary.
References 1. Herskowitz A, Mangano DT. Inflammatory cascade. A final common pathway for perioperative injury? Anesthesiology 1996;85:957– 60. 2. Hall RI, Smith MS, Rocker G. The systemic inflammatory response to cardiopulmonary bypass: pathophysiological, therapeutic, and pharmacological considerations. Anesth Analg 1997;85:766– 82. 3. Thomas DD, Sharar SR, Winn RK, et al. CD18-independent mechanism of neutrophil emigration in the rabbit lung after ischemia-reperfusion. Ann Thorac Surg 1995;60:1360– 6. 4. Tonz M, von Segesser LK, Fehr J, Schmid ER, Turina MI. Acute lung injury during cardiopulmonary bypass. Are the neutrophils responsible? Chest 1995;108:1551– 6. 5. Carlos TM, Harlan JM. Leukocyte-endothelial adhesion molecules. Blood 1994;84:2068 –101. 6. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci USA 1991;88:4651–5. 7. Albelda SM, Smith CW, Ward PA. Adhesion molecules and inflammatory injury. FASEB J 1994;8:504–12. 8. Zimmerman GA, McIntyre TM. Neutrophil adherence to human endothelium in vitro occurs by CDw18 (Mol 1, MAC-1/LFA-1/GP 150,95) glycoprotein-dependent and -independent mechanisms. J Clin Invest 1988;81:531–7. 9. McElroy FA, Philp RB. Relative potencies of dipyridamole and related agents as inhibitors of cyclic nucleotide phosphodiesterases: possible explanation of mechanism of inhibition of platelet function. Life Sci 1975;17:1479– 83. 10. FitzGerald GA. Dipyridamole. N Engl J Med 1987;316: 1247–57. 11. Suzuki S, Sugai K, Sato H, Sakatume M, Arakawa M. Inhibition of active oxygen generation by dipyridamole in human polymorphonuclear leukocytes. Eur J Pharmacol 1992;227:395– 401.
Ann Thorac Surg 1999;67:1277– 82
12. Tremoli E, Colli S. Dipyridamole inhibits superoxide anion generation by human neutrophils [Letter]. Thromb Haemost 1988;59:34. 13. Teoh KH, Christakis GT, Weisel RD, et al. Dipyridamole preserved platelets and reduced blood loss after cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;96:332– 41. 14. Jaffe EA. Culture and identification of large vessel endothelial cells. In: Jaffe EA, ed. Biology of endothelial cells. The Hague: Martinus Nijhoff, 1984, 1. 15. Lo SK, Detmers PA, Levin SM, Wright SD. Transient adhesion of neutrophils to endothelium. J Exp Med 1989;169:1779–93. 16. Chello M, Mastroroberto P, Celi V, Romano F, Marchese AR, Colonna A. Reduction by indobufen of neutrophil activation in peripheral arterial occlusive disease. J Cardiovasc Pharmacol 1996;27:417–23. 17. Teoh KH, Christakis GT, Weisel RD, et al. Dipyridamole reduced myocardial platelet and leukocyte deposition following ischemia and cardioplegia. J Surg Res 1987;42:642–52. 18. Colli S, Tremoli E. Multiple effects of dipyridamole on neutrophils and mononuclear leukocytes: adenosine-dependent and adenosine-independent mechanisms. J Lab Clin Med 1991;118:135– 45. 19. Sollevi A, Ostergren J, Fagrell B, Hjemdahl P. Theophylline antagonizes cardiovascular responses to dipyridamole in man without affecting increases in plasma adenosine. Acta Physiol Scand 1984;121:165–71. 20. Saniabadi AR, Lowe GDO, Barbenel JC, Forbes CD. A comparison of spontaneous platelet aggregation in whole blood with platelet rich plasma: additional evidence for the role of ADP. Thromb Haemost 1984;51:115– 8. 21. Cronstein BN, Kramer SB, Weissmann G, Hirschhorn R. Adenosine: a physiologic modulator of superoxide anion generation by human neutrophils. J Exp Med 1983;158: 1160–77. 22. Cronstein BN, Levin RI, Philips M, Hirschhorn R, Abramson SB, Weissmann G. Neutrophil adherence to endothelium is enhanced via adenosine A1 receptors and inhibited via adenosine A2 receptors. J Immunol 1992;148:2201– 6. 23. Bullough DA, Magill MJ, Firestein GS, Mullane KM. Adenosine activates A2 receptors to inhibit neutrophil adhesion to isolated cardiac myocytes. J Immunol 1995;155:2579– 86. 24. Hirschhorn R, Roegner-Maniscalco B, Kuritsky L, Rosen FS. Bone marrow transplantation only partially restores purine metabolites to normal in adenosine deaminase deficient patients. J Clin Invest 1981;68:1387–93. 25. Ely SW, Berne RM. Protective effects of adenosine in myocardial ischemia. Circulation 1992;85:893–904. 26. Thiel M, Chambers JD, Chouker A, et al. Effect of adenosine on the expression of beta (2) integrins and L-selectin of human polymorphonuclear leukocytes in vitro. J Leukocyte Biol 1996;59:671– 82. 27. Wollner A, Wollner S, Smith JB. Acting via A2 receptors, adenosine inhibits the upregulation of mac-1 (CD11b/CD18) expression on FMLP-stimulated neutrophils. Am J Respir Cell Mol Biol 1993;9:179– 85.