Leukocyte Integrin Expression in PatientsUndergoing Cardiopulmonary BypassGeorge Asimakopoulos, FRCS, Alejandro Kohn, FRCS, Demetrios C. Stefanou, FRCS,Dorian O. Haskard, FRCP, R. Clive Landis, PhD, and Kenneth M. Taylor, MDCardiothoracic Unit and Cardiovascular Medicine Unit at Hammersmith Hospital, National Heart and Lung Institute,Imperial College School of Medicine, London, England

Background. The recruitment of leukocytes to vascularendothelium is controlled by adhesion events mediatedthrough the b2 integrins, whereas the response of extrav-asated leukocytes within the tissues is controlledthrough the b1 integrins. Although cardiopulmonarybypass (CPB) has been shown to be associated with asystemic inflammatory response and elevated levels of b2

integrins on leukocytes, its effect on the b1 integrins isnot known. This study investigated the effect of theprotease inhibitor aprotinin on the expression of the b1

and b2 integrins on circulating leukocytes in patientsundergoing CPB.

Methods. Patients undergoing primary elective coro-nary artery bypass grafting were randomized into full-dose aprotinin or placebo groups. Blood samples wereobtained at nine time points preoperatively, intraopera-tively, and up to 6 days postoperatively. The surfaceexpression of the b1 integrins VLA-1, -3, -4, -5, and -6 andof the b2 integrins CD11a/CD18, CD11b/CD18, and

CD11c/CD18 was measured by flow cytometry on gatedneutrophil and monocyte subpopulations in wholeblood.

Results. Expression of the b1 integrins was not signif-icantly altered during the study period and, therefore,aprotinin had no effect on the expression of these mole-cules. Of the b2 integrins, CD11b/CD18 expression wassignificantly increased on neutrophils at 15 minutes afteronset of CPB in the placebo group (p < 0.01) but not inthe aprotinin group.

Conclusions. This study showed that expression of theb1 integrins on neutrophils and monocytes did not alterduring the first 6 days after CPB. Expression of the b2

integrin CD11b/CD18 increased significantly on neutro-phils during CPB in control patients but not in patientstreated with full-dose aprotinin.

(Ann Thorac Surg 2000;69:1192–7)© 2000 by The Society of Thoracic Surgeons

Aprotinin is a nonspecific serine protease inhibitorthat blocks pathways of complement activation and

fibrinolysis, and inhibits the action of proteinases such astrypsin, plasmin, and kallikrein [1]. The efficacy of apro-tinin in reducing postoperative bleeding after cardiacoperations was discovered during trials investigating itsantiinflammatory properties [2]. Over the last decade,aprotinin has been used extensively as a hemostaticagent, although less attention has been paid to its poten-tial effects on the inflammatory response to cardiopulmo-nary bypass (CPB).

The systemic inflammatory response to CPB is a mod-ification of the physiologic response to tissue injury orinfection. Activation of leukocytes, platelets, comple-ment, and factor XII by contact with the bypass circuitand surgical trauma is followed by systemic secretion ofcytokines and other inflammatory mediators. Inducedexpression of adhesion molecules on activated leuko-cytes and endothelial cells can result in sequestration ofwhite cells within the tissues and a clinical syndrome, thesystemic inflammatory response syndrome (SIRS), which

differs quite widely among patients [3]. In its extremeform, it can lead to multiple organ failure that oftenincludes adult respiratory distress syndrome, a conditionassociated with massive leukocyte infiltration in the lungand high mortality [4].

Recruitment of white cells into tissues requires thestepwise interaction between adhesion molecules on thesurface of leukocytes and their corresponding receptorson the lumenal surface of inflamed endothelium. Thisprocess is mediated through three main groups of adhe-sion molecules: (1) the selectins, which mediate the initialattachment and rolling of leukocytes along the vesselwall under hydrodynamic shear flow, (2) the b2 integrins,which mediate firm adhesion of leukocytes to endothe-lium, and (3) the immunoglobulin superfamily of adhe-sion molecules expressed on the endothelial side which,in conjunction with the b2 integrins, regulate firm adhe-sion and transendothelial migration [5].

Although the effects of CPB on b2 integrins, and inparticular on CD11b/CD18, have been investigatedpreviously, there are no reported data on leukocytesurface expression of the b1 integrins after cardiacoperation [6]. The b1 integrins (also known as the verylate antigens [VLA]) are expressed at low amounts onresting leukocytes but are induced upon extravasation[7] or over a more extended time course (between 3 and

Accepted for publication Oct 20, 1999.

Address reprint requests to Dr Taylor, Cardiothoracic Unit, Hammer-smith Hospital, Imperial College School of Medicine, Du Cane Rd,London W12 0NN, England; e-mail: scarroll@rpms.ac.uk.

© 2000 by The Society of Thoracic Surgeons 0003-4975/00/$20.00Published by Elsevier Science Inc PII S0003-4975(99)01553-2

14 days) after activation by mitogens [8]. They bind tocomponents of the extracellular matrix and conveysignals from the subendothelial matrix to alter pro-tease and cytokine production by leukocytes, whichmay contribute toward the process of tissue injury[9, 10].

We conducted this double-blind randomized trial toinvestigate the hypothesis that CPB is associated withincreased leukocyte surface expression of b1 and b2

integrins and that aprotinin may modify this effect.

Material and Methods

Patient GroupsThe protocol of the study was approved by our EthicalCommittee and informed consent was obtained fromall patients. Eighteen patients, undergoing primaryelective coronary artery bypass grafting (CABG), wererandomized into two groups in a double-blind fashion.One group (n 5 8) received full-dose aprotinin [2] andthe other group (n 5 10) served as a control. Patientswere excluded from the study if they met any of thefollowing criteria: episodes of unstable angina or myo-cardial infarction within 6 weeks preceding operation,cerebrovascular accident within 3 months precedingoperation, combined valve operation, infective endo-carditis, known or suspected allergy to aprotinin, pre-vious exposure to aprotinin, coagulopathy, knownbleeding diathesis, use of glucocorticoids or nonsteroi-dal antiinflammatory drugs, use of anticoagulants oraspirin in the week before operation, serum creatininein excess of 177 mmol/L, pregnancy, presence of malig-nancy, or more than 75% carotid obstruction as shownby carotid Doppler scan. The two patient groups werewell matched with regard to age, sex, preoperative riskfactors, angina class, dyspnea class, and left ventricularejection fraction (Table 1).

Anesthetic and Operative TechniquesStandardized techniques were used for anesthesia andCPB. Anesthetic premedication included morphine(10 mg) and hyoscine (0.3 mg) administered intramuscu-larly in the morning of the operation. Anesthesia wasintroduced with midazolam (100 to 200 mg/kg), fentanyl(150 to 200 mg), and pancuronium (50 to 100 mg/kg), andsustained with propofol (5 to 10 mg z kg21 z h21). The CPBcircuit consisted of a roller pump (Stockert Instruments,Munich, Germany), a Bard William Harvey HF-570 mem-brane oxygenator (C. R. Bard, Crawley, England), andpolyvinylchloride tubing. Pulsatile extracorporeal circu-lation was used at 2.4 to 2.8 L z m22 z min21. Moderatehypothermia of 32°C was employed in all patients.

Myocardial protection was administered with a Bardcardioplegia delivery system using cold blood antegradecardioplegia, mixed with St. Thomas’ crystalloid solutionin a 4:1 ratio, with additional “hot shot” before theremoval of the cross-clamp.

Flow Cytometric Analysis of Leukocyte b1 and b2Integrin ExpressionCentral venous blood samples (5 mL) were obtained frompatients and placed immediately into heparin-containingtubes at the following time points: (1) before skin incision(pre-CPB), (2) 15 minutes after initiation of CPB, (3) 60minutes after initiation of CPB, (4) 2 hours, (5) 4 hours, (6)24 hours, (7) 48 hours, (8) 72 hours, and (9) 6 dayspostoperatively. Tubes containing blood were placed onice and flow cytometric analysis was carried out within 2hours from blood sampling. For analysis, blood samples

Table 1. Intraoperative and Postoperative Characteristics ofPatients

Characteristics of PatientsAprotinin

(n 5 8)Control(n 5 10)

pValue

Age (y) 59 6 3.9a 65 6 1.9 NSMen/women 7/1 10/0 NSPreoperative risk factors

Family history 3 6Hypercholesterolemia 5 6Hypertension 5 7 NS(Ex-)Smoker 7 8Previous MI 3 6

PreoperativecharacteristicsAngina class III or IV 2 0 NSDyspnea class III or IV 1 0 NSEjection fraction , 0.3 2 0 NS

Operative characteristicsNo. of grafts 3.3 6 0.3 3.0 6 0.2 NSLAD artery grafted 8 10 NSCPB time (min) 82 6 3.6 79 6 7 NSCross-clamp time (min) 47 6 3.3 45 6 4.5 NS

Postoperativecharacteristics andcomplicationsBlood loss at 12 h (mL) 371 6 62 848 6 126 0.003Blood transfused (units) 0.75 6 0.7 1.7 6 1.5 0.03Platelets transfused

(units)0.125 6 0.3 0.1 6 0.3 NS

Coagulation factors(units)

0 0.4 6 0.8 NS

Perioperative MI 0 0Reoperation for

bleeding0 0

Inotropic support 0 1 NSAtrial fibrillation 1 6 0.04Chest infection 3 1 NSCerebrovascular

accident0 1 NS

Sternal infection 0 1 NSRenal failure 0 0Time to extubation (h) 5.5 6 0.5 6.7 6 0.6 NSTime to discharge

(days)6.3 6 0.2 7.6 6 0.9 NS

a Mean 6 standard deviation.

CPB 5 cardiopulmonary bypass; LAD 5 left anterior descending;MI 5 myocardial infarction; NS 5 not significant.

1193Ann Thorac Surg ASIMAKOPOULOS ET AL2000;69:1192–7 INTEGRINS AND CARDIOPULMONARY BYPASS

were placed in 12 3 75 mm polystyrene tubes (Falcon,Becton Dickinson UK Ltd, Cowley, UK) using, per con-dition, 90 mL whole blood and 10 mL primary antibody(100 mg/mL) directed against the unique a subunits of theb1 and b2 family of integrins (Table 2). Primary incuba-tions were carried out on ice for 15 minutes. Eachantibody was matched at each time point to an irrelevantisotypic control. (An irrelevant isotypic control is anantibody that is produced by the same host as the mainantibody and displays nonspecific binding characteris-tics, typical for the host.) After two washes with phos-phate-buffered saline, fluorescein isothiocyanate-conjugated secondary antibodies were added at themanufacturer’s recommended concentration (SigmaChemical Co, Dorset, UK) and incubation was continuedfor a further 15 minutes. Erythrocytes were lysed for 60seconds by the addition of 1 mL Coulter whole bloodlysing reagent and fixed in 250 mL fixative solution(Coulter Electronics Ltd, Luton, UK). Lysed samples werepreserved in 5% formaldehyde and read on a flowcytometer (EPICS XL, Coulter Electronics Ltd) within 24hours. Neutrophil and monocyte cell populations wereidentified by their characteristic forward and side scatterprofiles and confirmed by staining with anti-CD14 mono-clonal antibody. Fluorescent intensity of experimentalversus isotypic control antibodies was presented as therelative fluorescence intensity (RFI; ratio of experimentalmean fluorescence intensity over irrelevant isotypic con-trol mean fluorescence intensity).

Statistical AnalysisData are expressed as mean 6 standard deviation. Fordemographic data, the Mann-Whitney U test was used tocompare continuous variables and the x2 test to comparenominal data between groups. A mixed model analysis ofvariance techniques was used to look for effects over timeand between treatment groups for the different integrins.Bonferroni corrections were applied to the p values toallow for the multiple comparisons that were made with

time point 1. Significance was assumed for p values lessthan 0.05.

Results

Intraoperative and Postoperative Characteristics ofPatientsThere was no significant difference between aprotininand placebo groups with regard to number of grafts, CPBtime, and cross-clamp time (Table 1). All patients re-ceived internal thoracic artery graft to the left anteriordescending artery. Patients in the aprotinin group lostsignificantly less blood through the chest drains andreceived significantly less blood transfusions at 12 hourspostoperatively (Table 1). Patients in the two groups hada similarly low rate of postoperative complications, withthe exception of atrial fibrillation, which was significantlymore common in the control group. On average, patientswho received aprotinin were extubated 1 hour earlier andwere discharged home 1 day earlier than control patients,although these differences were not statisticallysignificant.

Expression of Integrins on Circulating Neutrophils andMonocytesB1 INTEGRINS. Integrin expression was quantitated by nor-malizing the staining intensity obtained with test anti-body to an appropriate irrelevant isotype matched con-trol, thereby establishing an RFI for each integrin at eachof the nine time points studied on both neutrophil andmonocyte cell populations (where an RFI of 1.00 5 noexpression). The integrity of all test antibodies was veri-fied before use in the study by staining primary T-celllymphoblasts or other cell lines known to express VLAantigens (data not shown). Neutrophils expressed lowlevels of the b1 integrins in the preoperative samples(RFI , 2.00) and these levels were not significantlyaltered either intraoperatively or up to 6 days postoper-atively (Table 2). Monocytes expressed low levels of

Table 2. Integrins Tested on the Surface of Neutrophils and Monocytes in Patients Undergoing Cardiopulmonary Bypass

Integrin (clusterdesignation)

OtherNames Antibody Sourcea–e

Change From Preoperative Level

Neutrophils Monocytes

b1 integrinsCD49a/CD29 VLA-1 TS 2/7 a NS NSCD49c/CD29 VLA-3 11G5 b NS NSCD49d/CD29 VLA-4 PS2 b NS NSCD49e/CD29 VLA-5 MAb 16 c NS NSCD49f/CD29 VLA-6 GoH3 d NS NS

b2 integrinsCD11a/CD18 LFA-1 38 e NS NSCD11b/CD18 Mac-1 44 e p , 0.01 at 15 min NSCD11c/CD18 p150,95 3.9 e NS NS

a Martin Hemler, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA. b Martin Humphries, Manchester University, Manchester,UK. c Ken Yamada, Institute of Dental Research, NIH, Bethesda, MD. d Serotec, Oxford, UK. e Nancy Hogg, Imperial Cancer Research Fund,London, England.

NS 5 not significant.

1194 ASIMAKOPOULOS ET AL Ann Thorac SurgINTEGRINS AND CARDIOPULMONARY BYPASS 2000;69:1192–7

VLA-1, -4, and -6 (RFI , 2.00) and moderate levels ofVLA-3 and -5 (RFI 2.00 to 4.00) in the preoperativesamples, but again showed no significant differences inexpression in the intraoperative or postoperative periods.Administration of aprotinin, accordingly, had no effect onthe expression of these molecules at any of the timepoints studied up to 6 days after CPB (Table 2).

B2 INTEGRINS. The three b2 integrins tested, CD11a/CD18,CD11b/CD18, and CD11c/CD18, were present at moder-ate to high levels in preoperative samples on both neu-trophils and monocytes (RFIs between 3.2 and 27.4), andthese levels were significantly altered only at the 15-minute time point for CD11b/CD18 expression on neu-trophils in the placebo group ( p , 0.01, Table 2) but notin the corresponding aprotinin group. The complete timecourse for CD11b/CD18 expression on neutrophils withinthe two treatment groups is represented in Figure 1.CD11b/CD18 expression was measured on the gatedneutrophil subpopulation in whole blood by flow cyto-metric analysis, normalizing expression at each timepoint with test antibody (mAb 44) to an irrelevant isotypematched control antibody. Figure 1 confirms that CPBsignificantly induced CD11b/CD18 expression at 15 min-utes in the placebo group ( p , 0.01) but not in theaprotinin group ( p . 0.48). No significant differenceswere detected at any of the later time points with respectto expression of CD11b/CD18 or the other two b2 inte-grins tested on either neutrophils or monocytes (Table 2).We have noted that, although significant, the CD11b/CD18 response to CPB at 15 minutes was heterogeneousbetween individuals in the placebo group, with only 4“responders” showing elevated CD11b/CD18 levels (be-tween 281% and 642% of preoperative levels) and 6“nonresponders” showing levels largely unchanged (be-tween 69% and 124%).

Comment

Neutrophil activation, adhesion to endothelium, and ex-travasation are decisive steps in the systemic inflamma-

tory response. Activated neutrophils, along with theirbactericidal effects, also may contribute significantly toinjury of inflamed tissue [11]. Adhesion molecules playan important role during inflammation on both sides ofthe endothelial barrier, first by helping to recruit leuko-cytes from the vascular compartment to the vessel walland subsequently by modulating their responsivenesswithin the tissues by transducing signals from the sub-endothelial matrix. In investigating the etiologic processof the systemic inflammatory response to CPB, it istherefore important to understand whether the twoclasses of integrin receptors involved in adhesion toendothelium or matrix, the b2 and b1 integrins, respec-tively, are induced on leukocytes by cardiac operation.We therefore investigated the effects of CPB on theseintegrin families on neutrophils and monocytes. Thisstudy reports the expression of b1 integrins in patientsundergoing CPB and comprises one of the few random-ized trials investigating the effect of aprotinin on b2

integrins in this patient population [12, 13].Consistent with their original name (“very late anti-

gens”), most leukocyte b1 integrins increase their expres-sion gradually after several days of activation [8]. The lowlevels of basal integrin expression observed in the pre-operative samples in the present study have confirmedthe prevailing view that the VLA integrins are largelyabsent from circulating neutrophils [14, 15], althoughcertain conditions of activation can lead to low levels ofexpression and increased adhesiveness to collagen andlaminin matrix components [7, 16, 17]. The absence of b1

integrin expression in the immediate postoperative pe-riod in this study was not necessarily surprising, but overthe extended time course of the investigation suggeststhat uneventful cardiac operation constitutes a signifi-cant, but relatively short-lived, inflammatory insult thatdoes not appear to induce VLA production during thefirst 6 postoperative days.

It should be noted that a lack of observed effect onintegrin expression levels cannot be interpreted as aninevitable lack of effect of CPB on leukocyte adhesion, asthe b1 integrins have been described to alter their adhe-sive capacity by exposure to chemokines, metal ions, andother pharmacologic agents in the absence of alterationsin expression [18–20]. Further studies involving the exvivo testing of patient leukocytes in static adhesionassays to purified ligands would be required to fullyunderstand the effect of CPB on the adhesive propertiesof the b1 integrins on leukocytes.

VLA-4 is unique among b1 integrins in its ability tobind, in addition to matrix proteins, the endothelialligand vascular cell adhesion molecule-1 (VCAM-1). Al-though not expressed on neutrophils, except after endo-thelial transmigration or pharmacologic stimulation [7],VLA-4 is critically involved in monocyte interactions withendothelium, both in the initial rolling phase as well as inthe subsequent firm adhesion phase [21, 22]. VLA-4antagonists also have been shown to ameliorate asth-matic responses and inflammatory bowel disease in an-imal models of disease [23]. It was therefore of interest tonote that expression levels of VLA-4 remained un-

Fig 1. Time course of neutrophil Mac-1 expression in cardiopulmo-nary bypass (CPB) patients. Blood samples were collected as de-scribed in the text from patients randomized into placebo (emptycolumns) or aprotinin (filled columns) groups.

1195Ann Thorac Surg ASIMAKOPOULOS ET AL2000;69:1192–7 INTEGRINS AND CARDIOPULMONARY BYPASS

changed on monocytes throughout the study period,although it cannot be ruled out that CPB might havealtered the affinity of preexisting VLA-4 on monocytemembranes, a mechanism that has been described toregulate adhesion through this integrin [24].

The b2 integrins are well established as importantparticipants in leukocyte-endothelial cell adhesion andextravasation. All three integrins are expressed on neu-trophils and monocytes, of which CD11a/CD18 promi-nently binds the ligands intracellular adhesion mole-cule-1 (ICAM-1) and ICAM-2 expressed on endotheliumand ICAM-3 expressed on other leukocytes [25]. CD11b/CD18 also binds to ligands such as complement factoriC3b and fibrinogen [26]. Although b2 integrins exist onthe surface of leukocytes in a relatively inactive formunder normal conditions, inflammatory insults result inincreased expression and change in their activation state[5, 22]. The effect of CPB has been investigated mostextensively on CD11b/CD18 expression by neutrophils[6]. Upregulation of CD11b/CD18 indicates neutrophilactivation that is associated with increased ability of theneutrophil to adhere to endothelium, which is a prereq-uisite for neutrophil-induced tissue damage.

Nearly all literature reports have shown that CD11b/CD18 is upregulated at an early postoperative stage onthe surface of neutrophils. Some articles also have shownan increase in the expression of CD11c/CD18 whileCD11a/CD18 appears to be unaffected [6]. Our resultsdemonstrated that CPB significantly upregulated CD11b/CD18 expression on neutrophils at 15 minutes of CPB,consistent with the main body of literature. The reasonthis effect became insignificant at 60 minutes of CPB isprobably that activated neutrophil were marginalizedwithin tissues and the CPB circuit. We also noted that thisresponse, although significant ( p , 0.01) in the placebogroup, was heterogeneous between individuals, with 4 of10 “responders” exhibiting a large increase in CD11b/CD18 expression relative to preoperative levels (between281% and 642%) and 6 of 10 “nonresponders” exhibitingvirtually unchanged levels. As CD11b/CD18 is consid-ered a well-established marker of neutrophil activation, itis therefore likely that neutrophils from “responders” aremore sensitive to the effects of CPB, resulting in exacer-bated cellular activation. It is reasonable to assume thatthe variety in response may be attributable to geneticdifferences among patients or to a different “activationstate” of neutrophils preoperatively. Further investiga-tions will be necessary to determine whether such pa-tients can be identified before operation and whetherthey could be targeted with specific antiinflammatorystrategies, such as glucocorticosteroids or aprotinin.

Our results showing that CPB can significantly induceCD11b/CD18 on neutrophils at 15 minutes in the placebogroup ( p , 0.01) but not in the aprotinin group ( p .0.48) are consistent with previous findings from tworandomized trials that aprotinin can blunt the effect ofCABG on Mac-1 upregulation in neutrophils [12, 13]. Wehave taken care in our investigation to minimize anyinherent variability in the flow cytometric technique overthe extended time course of the study by normalizing all

fluorescent staining with specific antibodies to isotypematched controls at every time point. Our studies there-fore add conviction to previous investigations, whichroutinely reported only mean fluorescent intensities withtest antibodies without normalizing data between timepoints, yet reached the same basic conclusions as thepresent study.

The small number of the studied patient group pre-cluded clinical differences between groups from becom-ing significant. The lower rate of atrial fibrillation in thecontrol group may be attributable to a statistical errortype II, as aprotinin is not known to reduce dysrhythmiasafter CPB. Duration of hospital stay, however, has beenshown previously to be shorter in patients treated withaprotinin [27].

In summary, this study showed that the expression ofb1 integrins on neutrophils and monocytes did not in-crease during the first 6 days after CPB. Neutrophilexpression of the b2 integrin CD11b/CD18 increasedsignificantly during CPB in the control group but not inpatients treated with full-dose aprotinin. The heteroge-neity in the CD11b/CD18 response between individualssuggests that the degree of neutrophil activation inducedby CPB may vary significantly among patients. Patientswith high responses may be appropriate targets forspecifically directed antiinflammatory strategies.

References

1. Taylor KM. Effect of aprotinin on blood loss and blood useafter cardiopulmonary bypass. In: Pifarre R, ed. Anticoagu-lation, hemostasis, and blood preservation in cardiovascularsurgery. Philadelphia: Hanley & Belfus, 1993:129–45.

2. Royston D, Bidstrup BP, Taylor KM, Sapford RN. Effect ofaprotinin on need for blood transfusions after repeat openheart surgery. Lancet 1987;2:1289–91.

3. Boyle EM Jr, Pohlman TH, Johnson MC, Verrier ED. Thesystemic inflammatory response. Ann Thorac Surg 1997;64:S31–7.

4. Bernard GR, Artigas A, Brigham KL, et al. The American-European Consensus Conference on ARDS: definitions,mechanisms, relevant outcomes, and clinical trial coordina-tion. Am J Respir Crit Care Med 1994;149:818–24.

5. Springer TA. Traffic signals for lymphocyte recirculation andleukocyte emigration: the multistep paradigm. Cell 1994;76:301–14.

6. Asimakopoulos G, Taylor KM. The effects of cardiopulmo-nary bypass on leukocyte and endothelial adhesion mole-cules. Ann Thorac Surg 1998;66:2135–44.

7. Kubes P, Niu XF, Smith CW, Kehrli ME, Reinhardt PH,Woodman RC. A novel b1-adhesion pathway on neutro-phils: a mechanism evoked by dihydrocytochalasin B orendothelial transmigration. FASEB J 1995;9:1103–11.

8. Hemler ME. VLA proteins in the integrin family: structures,functions, and their role on leukocytes. Annu Rev Immunol1990;8:365–400.

9. Pacifici R, Basilico C, Roman J, Zutter MM, Santoro SA,McCracken R. Collagen-induced release of interleukin 1from human blood mononuclear cells. Potentiation by fi-bronectin binding to the alpha 5 beta 1 integrin. J Clin Invest1992;89:61–7.

10. Khan KMF, Falcone DJ. Role of laminin in matrix inductionof macrophage urokinase-type plasminogen activator and92-kDa metalloproteinase expression. J Biol Chem 1997;272:8270–5.

11. Asimakopoulos G, Smith PL, Ratnatunga CP, Taylor KM.

1196 ASIMAKOPOULOS ET AL Ann Thorac SurgINTEGRINS AND CARDIOPULMONARY BYPASS 2000;69:1192–7

Pulmonary dysfunction and adult respiratory distress syn-drome after cardiopulmonary bypass [Current Review]. AnnThorac Surg 1999;68:1107–15.

12. Hill GE, Alonso A, Spurzem JR, Stammers AH, Robbins RA.Aprotinin and methylprednisolone equally blunt cardiopul-monary bypass-induced inflammation in humans. J ThoracCardiovasc Surg 1995;110:1658–62.

13. Alonso A, Whitten CW, Hill GE. Pump prime only aprotinininhibits cardiopulmonary bypass-induced neutrophil CD11bup-regulation. Ann Thorac Surg 1999;67:392–5.

14. Rubio MA, Sotillos M, Jochems G, Alvarez V, Corbi AL.Monocyte activation: rapid induction of alpha 1/beta 1(VLA-1) integrin expression by lipopolysaccharide and in-terferon-gamma. Eur J Immunol 1995;25:2701–5.

15. Namie S, Ozono Y, Harada T, Hara K. Expression andfunction of fibronectin receptors on peripheral mononuclearcells in IgA nephropathy. Nephrol Dial Transplant 1995;10:1342–7.

16. Bohnsack JF, Zhou X-N. Divalent cation substitution revealsCD18- and very late antigen-dependent pathways that me-diate human neutrophil adherence to fibronectin. J Immunol1992;149:1340–7.

17. Bohnsack JF. CD11/CD18-independent neutrophil adher-ence to laminin is mediated by the integrin VLA-6. Blood1992;79:1545–52.

18. Faull RJ, Kovach NL, Harlan JM, Ginsberg MH. Affinitymodulation of integrin alpha 5 beta 1: regulation of thefunctional response by soluble fibronectin. J Cell Biol 1993;121:155–62.

19. Weber C, Alon R, Moser B, Springer TA. Sequential regula-

tion of alpha 4 beta 1 and alpha 5 beta 1 integrin avidity byCC chemokines in monocytes: implications for transendo-thelial chemotaxis. J Cell Biol 1996;134:1063–73.

20. Yauch RL, Felsenfeld DP, Kraeft SK, Chen LB, Sheetz MP,Hemler ME. Mutational evidence for control of cell adhesionthrough integrin diffusion/clustering, independent of ligandbinding. J Exp Med 1997;186:1347–55.

21. Luscinskas FW, Kansas GS, Ding H, et al. Monocyte rolling,arrest and spreading on IL-4-activated vascular endotheliumunder flow is mediated via sequential action of L-selectin,beta 1-integrins, and beta 2-integrins. J Cell Biol 1994;125:1417–27.

22. Gerszten RE, Lim YC, Ding HT, et al. Adhesion of monocytesto vascular cell adhesion molecule-1-transduced humanendothelial cells: implications for atherogenesis. Circ Res1998;82:871–8.

23. Lobb RR, Hemler ME. The pathophysiologic role of a4integrins in vivo. J Clin Invest 1994;94:1722–8.

24. Jakubowski A, Rosa MD, Bixler S, Lobb R, Burkly LC.Vascular cell adhesion molecule (VCAM)-Ig fusion proteindefines distinct affinity states of the very late antigen-4(VLA-4) receptor. Cell Adhes Commun 1995;3:131–42.

25. Stewart M, Thiel M, Hogg N. Leukocyte integrins. Curr OpinCell Biol 1995;7:690–6.

26. Albelda SM, Smith CW, Ward PA. Adhesion molecules andinflammatory injury. FASEB J 1994;8:504–12.

27. Gott JP, Cooper WA, Schmidt FE Jr, et al. Modifying risk forextracorporeal circulation: trial of four antiinflammatorystrategies. Ann Thorac Surg 1998;66:747–54.

1197Ann Thorac Surg ASIMAKOPOULOS ET AL2000;69:1192–7 INTEGRINS AND CARDIOPULMONARY BYPASS