Are Circulating Adhesion Molecules Specifically Changed in Cardiac Surgical Patients?

ORIGINAL ARTICLES: CARDIOVASCULAR

Are Circulating Adhesion Molecules Specifically Changed in Cardiac Surgical Patients? Joachim Boldt, MD, Bernd Kumle, MD, Michael Papsdorf, and Gunter Hempelmann, MD Department of Anesthesiology and Intensive Care Medicine, Justus-Liebig-University Giessen, Giessen, Germany

Background. Soluble adhesion molecules are considered to be markers of inflammation, endothelial activation, or damage. This study was designed to assess whether adhesion molecules are specifically altered in patients undergoing cardiac surgical procedures. Methods. Three groups of 20 patients each were prospectively studied: patients undergoing elective coronary artery bypass grafting; patients scheduled for a Whipple pancreatoduodenectomy; and patients undergoing elective pneumonectomy for lung cancer. Plasma levels of soluble adhesion molecules (endothelial leukocyte adhesion molecule-1, intercellular adhesion molecule-1, vascular cell adhesion molecule-1, and granule membrane protein 140) were measured from arterial blood samples after induction of anesthesia (baseline), at the end of the operation, 2 hours and 5 hours after operation, and on the morning of the first postoperative day. Results. Duration of operation was longest in the group having a Whipple operation (289 6 50 minutes) and did not differ between the other two groups. Plasma levels of all measured adhesion molecules at baseline were within normal ranges. After cardiopulmonary bypass, levels of adhesion molecules were significantly increased in the cardiac surgical patients (soluble endo-

thelial leukocyte adhesion molecule-1, from 38 6 11 ng/mL at baseline to 68 6 12 ng/mL; soluble intercellular adhesion molecule-1, from 241 6 50 ng/mL to 498 6 78 ng/mL; and granule membrane protein 140, from 69 6 12 ng/mL to 150 6 25 ng/mL). On the morning of the first postoperative day, all levels had returned to baseline except that of soluble vascular cell adhesion molecule-1, which was still elevated (p < 0.05). In both the other groups, concentrations of adhesion molecules remained almost unchanged. Conclusions. Cardiac operation was associated with increased plasma levels of soluble adhesion molecules, a finding indicating endothelial activation or dysfunction. In contrast, in patients undergoing complex, long-lasting abdominal or lung operations, soluble adhesion molecules remained unchanged. Activation of proinflammatory cascades, ischemia/reperfusion phenomenon, and microcirculatory dysfunction appear to be the most likely reasons for this difference between groups. Whether modulation of adhesion molecules may influence organ function after cardiopulmonary bypass remains to be elucidated in further studies. (Ann Thorac Surg 1998;65:608 –14) © 1998 by The Society of Thoracic Surgeons

E

ICAM-3]), the integrin family (eg, lymphocyte functionassociated antigen), and the selectins {E-selectin, or endothelial-leukocyte adhesion molecule (ELAM-1), Lselectin (eg, leukocyte-endothelial cell adhesion molecule), and P-selectin (eg, granule membrane protein 140 [GMP 140])}. Soluble isoforms of some adhesion molecules can be found in the circulating blood under various circumstances [5]. An increase in circulating adhesion molecules results either from increased expression of activated endothelial cells or from increased proteolytic cleavage of endothelial-bound forms secondary to endothelial cell damage [6]. Thus, increased concentrations of soluble adhesion molecules may serve as markers for activated or damaged endothelium [5, 6]. Triggered by appropriate stimulation such as infection or surgical injury, production of cytokines can increase markedly and thereby possibly damage endothelial function. Cardiac operations using cardiopulmonary bypass (CPB) have been reported to be associated with release of various cytokines [7], activation of the coagulation cascade, and increased plasma levels of circulating adhesion

ndothelial damage appears to be the common end point of a complex pathophysiologic process that may result in development of multiple-organ dysfunction syndrome [1]. Polymorphonuclear cells have been implicated to be of major importance in this process. The leukocytes roll on, adhere tightly, spread, and finally migrate into tissues [2]. Adhering to the endothelium, they subsequently release toxic agents and thus cause endothelial damage. Stimulated by several factors (eg, cytokines and hypoxia), the endothelial surface expresses several distinct counterreceptors that are essential for polymorphonuclear cell adherence [2, 3]. Different ligand molecules (adhesion molecules) have been identified as mediators in the interaction between the endothelial cells and the leukocyte subpopulation [4]. These adhesion molecules include the immunoglobulin superfamily (eg, vascular cell adhesion molecule-1 [VCAM-1] and intercellular adhesion molecules [ICAM-1, ICAM-2, and Accepted for publication Aug 16, 1997. Address reprint requests to Dr Boldt, Department of Anesthesiology and Intensive Care Medicine, Klinikum der Stadt Ludwigshafen, Bremserstr 79, D-67063 Ludwigshafen, Germany.

© 1998 by The Society of Thoracic Surgeons Published by Elsevier Science Inc

0003-4975/98/$19.00 PII S0003-4975(97)01306-4

Ann Thorac Surg 1998;65:608 –14

molecules [8, 9]. Depending on the magnitude of the operation, an increase in cytokines (eg, interleukin-6 and TNF-alpha) has been observed also in noncardiac surgical procedures [10]. Anesthesia appears to modulate release of some cytokines including interleukin-6 and interleukin-1-beta [11], thus possibly influencing endothelial function and subsequently, concentration of circulating adhesion molecules. The present study was designed to evaluate whether plasma levels of circulating adhesion molecules in patients undergoing a cardiac surgical procedure were specifically altered and whether they differed from those in patients undergoing major noncardiac thoracic or abdominal surgical interventions.

Material and Methods Patient Groups Three groups of patients were prospectively studied: 20 consecutive patients undergoing elective first-time coronary artery bypass grafting (CABG); 20 consecutive patients with pancreatic cancer or other periampullar cancers undergoing a Whipple pancreatoduodenectomy; and 20 consecutive patients scheduled for elective pneumonectomy for lung cancer. Informed consent was obtained from each patient according to the protocol of the human ethics committee of the hospital. Exclusion criteria were renal insufficiency (creatinine level .1.5 mg/dL) and liver dysfunction (aspartate aminotransferase .40 U/L and alanine aminotransferase .40 U/L).

Operative Techniques Induction and maintenance of anesthesia were similar for all groups and consisted of weight-related doses of fentanyl, midazolam, and pancuronium bromide. All patients with lung cancer were intubated with a doublelumen endotracheal tube to provide atelectasis for the lung operation (one-lung ventilation). Controlled mechanical ventilation was adjusted to keep arterial oxygen tension between 100 and 150 mm Hg and arterial carbon dioxide tension between 35 and 45 mm Hg. All patients were extubated in the intensive care unit when temperature was greater than 36°C and hemodynamic, laboratory, and respiratory variables had been stable for at least 30 minutes. Low-molecular weight hydroxyethyl starch solution, Ringer’s solution, or both were given to keep pulmonary capillary wedge pressure between 10 and 15 mm Hg. Banked packed red blood cells were transfused when hemoglobin levels were lower than 9 g/dL. Epinephrine was given when mean arterial blood pressure was lower than 60 mm Hg despite sufficient volume loading. Norepinephrine was administered when systemic vascular resistance was less than 600 dyne z s21 z cm25 and mean arterial pressure, lower than 60 mm Hg. The entire perioperative therapeutic management (anesthetic drugs, volume replacement, and pharmacologic support) of each patient was carried out by anesthesiologists and intensivists who were not involved in the study. In the cardiac surgical patients (CABG group), CPB

BOLDT ET AL ADHESION MOLECULES AND MAJOR OPERATIONS

609

was performed with a capillary oxygenator (Monolyth; Sorin, Turino, Italy). Five minutes before the start of CPB, heparin sodium, 300 IU/kg, was administered to achieve anticoagulation. Activated clotting time (kaolin as activator) was monitored using a Hemochron system (International Technidyne Corporation, Edison, NJ). The extracorporeal circuit was primed with 1,000 mL of Ringer’s solution, 1,000 mL of dextrose 5%, and 250 mL of albumin 5%; aprotinin was not used. A nonpulsatile flow of 2.4 L z min21 z m22 was maintained throughout CPB. Rectal temperature was kept normothermic (.35°C), and Bretschneider cardioplegic solution was infused for myocardial preservation. Within 20 minutes after the start of CPB, the perfusate was concentrated using a hemofiltration device (HF-80; Fresenius, Bad Homburg, Germany) to adjust the hemoglobin level between 8 and 10 g/dL. When necessary to guarantee filling of the circuit, Ringer’s solution was added. When the hemoglobin value dropped to less than 7 g/dL, packed red blood cells were added to the perfusate. After the patient was separated from CPB, the residual blood in the circuit was salvaged by the hemofiltration device, and the autologous blood was transfused until the end of the operation. To reverse the effects of heparin, protamine sulfate was given in a 1:1 ratio to the initial heparin dose.

Measured Variables and Data Points In all patients, hemodynamic monitoring was performed using a pulmonary artery catheter. Arterial blood samples were withdrawn from an indwelling arterial cannula into an EDTA (ethylenediaminetetraacetic acid)– containing tube. After immediate centrifugation (15 minutes at 600 g), the plasma samples were stored at 270°C. Circulating (soluble [s]) ELAM-1 (sELAM-1) (normal range from our laboratory, 30 to 53 ng/mL; mean value, 42 ng/mL), VCAM-1 (sVCAM-1) (normal range from our laboratory, 450 to 700 ng/mL; mean value, 530 ng/mL), and ICAM-1 (sICAM-1) (normal range from our laboratory, 190 to 300 ng/mL; mean value, 210 ng/mL) were measured using commercial standardized enzymelinked immunosorbent assay kits (British Bio-technology Products, Abingdon, UK). The tests are based on simultaneous reaction of the adhesion molecule to two monoclonal antibodies directed against different epitopes on the adhesion molecule (ICAM, ELAM, and VCAM). All assays are standardized against purified forms of recombinant ICAM, ELAM, or VCAM. Sensitivity (minimal detectable dose) for sELAM-1 is lower than 1.0 ng/mL; for sVCAM-1, lower than 2.0 ng/mL; and for sICAM-1, less than 0.35 ng/mL. In addition, soluble GMP 140 (normal range, 50 to 115 ng/mL; mean value, 90 ng/mL) was measured by commercial enzyme immunoassay kit (TaKaRa, Shiga, Japan). Normal values were derived from measurements from healthy volunteers. All results from enzyme-linked immunosorbent assay measurements represent the means from duplicate measurements. Standard laboratory variables were also documented in all patients. All measurements were performed after induction of anesthesia (baseline), at the end of operation, 2 hours and 5 hours after operation, and on the morning of the first postoperative day.

610

BOLDT ET AL ADHESION MOLECULES AND MAJOR OPERATIONS

Ann Thorac Surg 1998;65:608 –14

Table 1. Demographic Data and Data From Perioperative Perioda Variable Age (y) Weight (kg) Sex (M/F) Duration (min) Operation CPB Cross-clamping No. of deaths PRBC (total units for all patients during study) FFP (total units during study) Catecholamine support Epinephrine (.5 mg/min) (no. of patients) Range (mg/min) Norepinephrine (.5 mg/min) (no. of patients) Range (mg/min) a

CABG (n 5 20)

Whipple (n 5 20)

Pneumonectomy (n 5 20)

66.3 6 10.5 76.1 6 12.2 12/8

64.1 6 12.4 75.2 6 12.6 14/6

63.9 6 12.2 79.6 6 13.3 13/7

198.7 6 23.5 101.8 6 14.4 76.3 6 10.3 0 13 6 5

289.4 6 49.9b ... ... 1 32 6 12b

201.3 6 25.3 ... ... 1 15 6 6

863

20 6 7b

662

6b

3

2

5–30 2

5–10 2

5 1

5–7

3– 8

3–5

Where applicable, data are shown as the mean 6 the standard deviation.

CABG 5 coronary artery bypass grafting;

b

CPB 5 cardiopulmonary bypass;

Statistical Analysis

Significance: p , 0.05 versus the other groups. FFP 5 fresh frozen plasma;

PRBC 5 packed red blood cells.

not differ in respect to age, sex, and survival. Time of operation was longest in the group having a Whipple operation (289 6 50 minutes); the duration was not different between the patients having CABG and those having pneumonectomy. Standard coagulation variables did not differ between the three groups (Table 2). Hemoglobin concentration, C-reactive protein, and plasma lactate levels also were not significantly different between the groups throughout the study period (Table 3). Neutrophil count and temperature increased slightly (p 5 0.05) in the CABG group and the Whipple group (see Table 3). Hemodynamic data were almost comparable between the three groups (Table 4). The arterial oxygen

All data are expressed as the mean 6 the standard deviation. One-way analysis of variance and two-way analysis of variance for repeated measures of the same variable (analysis of variance followed by Scheffe´ test) were used for statistical analyses. The Mann-Whitney U test and the x2 test were used when appropriate. A p value of less than 0.05 was considered significant.

Results A summary of group data and groups of data from the perioperative period is shown in Table 1. The groups did Table 2. Summary of Data on Coagulation Variablesa,b

Variable Platelet count (103/mL) AT III (%) (normal, 70 –110) Fibrinogen (g/L) (normal, 1.5– 4.0) aPTT (s) (normal, 20 –30)

a

Group

Baseline

After Operation

CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy

188 6 44 199 6 33 211 6 31 79.3 6 5.8 82.1 6 8.9 82.1 6 9.1 3.22 6 0.4 3.18 6 0.5 3.69 6 0.4 30.1 6 4.4 28.7 6 3.8 29.1 6 4.4

132 6 32 162 6 40 190 6 33 72.2 6 12.1 71.1 6 9.2 79.2 6 10.1 3.19 6 0.7 3.64 6 0.5 3.92 6 0.4 32.2 6 3.1 32.3 6 4.5 34.4 6 4.7

Data are shown as the mean 6 the standard deviation.

aPTT 5 activated partial thromboplastin time;

b

2 Hours After Operation

5 Hours After Operation

Postoperative Day 1

138 6 32 135 6 35 184 6 31 70.2 6 10.2 72.6 6 11.1 78.5 6 9.3 4.09 6 0.6 3.92 6 0.5 4.05 6 0.4 39.1 6 3.1 39.4 6 4.2 36.8 6 3.9

139 6 33 132 6 42 192 6 32 76.2 6 11.2 77.7 6 10.1 82.2 6 9.1 3.76 6 0.5 3.88 6 0.3 4.04 6 0.5 36.7 6 4.4 38.3 6 3.9 33.1 6 4.0

149 6 42 139 6 34 199 6 35 78.3 6 11.2 80.9 6 10.1 81.1 6 10.2 3.87 6 0.6 4.09 6 0.5 4.11 6 0.3 32.3 6 4.6 39.7 6 3.8 33.1 6 4.2

There were no significant differences between groups.

AT III 5 antithrombin III;

CABG 5 coronary artery bypass grafting.

Ann Thorac Surg 1998;65:608 –14

BOLDT ET AL ADHESION MOLECULES AND MAJOR OPERATIONS

611

Table 3. Temperature, Hemoglobin, Neutrophil Count, C-Reactive Protein, and Plasma Lactate Levelsa

Variable Temperature (°C)

Hemoglobin (g/dL)

Neutrophil count (103/mL) CRP (mg/L)

Lactate (mg/dL)

a

Group

Baseline

After Operation

CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy

36.1 6 0.5 36.2 6 0.4 36.3 6 0.3 12.8 6 0.4 12.3 6 0.9 13.6 6 0.7 8.2 6 2.2 7.0 6 2.0 7.1 6 2.6 1.4 6 0.5 1.1 6 0.4 1.2 6 0.3 1.1 6 0.5 1.0 6 0.3 1.2 6 0.4

35.8 6 0.7 35.2 6 0.4 35.9 6 0.5 10.1 6 0.7 9.8 6 0.4 11.0 6 1.2 9.7 6 3.1 8.4 6 2.6 8.5 6 1.9 4.8 6 2.5 4.9 6 3.8 4.6 6 0.5 1.2 6 0.3 1.6 6 0.4 1.3 6 0.4

Data are shown as the mean 6 the standard deviation.

CABG 5 coronary artery bypass grafting;

b

2 Hours After Operation

5 Hours After Operation

Postoperative Day 1

37.4 6 0.4 36.0 6 0.5 36.1 6 0.4 10.3 6 0.5 10.6 6 0.6 10.9 6 0.7 10.1 6 3.5 8.5 6 2.4 9.4 6 2.9 6.1 6 2.8 5.9 6 3.4 4.8 6 0.5 1.3 6 0.5 1.7 6 0.5 1.4 6 0.5

37.8 6 0.4 36.3 6 0.5 37.8 6 0.4 10.3 6 0.5 10.8 6 0.7 11.1 6 0.9 12.9 6 4.2b 12.1 6 4.2b 10.9 6 2.7 7.6 6 3.0 8.4 6 3.6 3.9 6 0.8 1.2 6 0.4 1.8 6 0.5 1.4 6 0.2

38.0 6 0.5b 37.9 6 0.4b 37.6 6 0.6 10.4 6 0.6 10.7 6 0.8 11.2 6 1.1 12.5 6 3.3b 12.3 6 2.4b 11.9 6 3.2 4.9 6 3.7 4.5 6 3.5 3.6 6 0.9 1.2 6 0.5 1.7 6 0.3 1.6 6 0.5

Significance: p , 0.05 versus baseline.

CAP 5 C-reactive protein.

tension to inspired oxygen fraction ratio was temporarily reduced in the cardiac surgical patients but had returned to baseline by the morning of the first postoperative day (see Table 4). Plasma levels of sELAM-1 increased significantly only in the CABG group (from 38 6 11 ng/mL at baseline to 68 6 12 ng/mL 5 hours after operation) and had not decreased to baseline by the first postoperative day (Fig 1). In this same group, plasma concentrations of sICAM-1 increased significantly from 241 6 50 ng/mL at baseline to 498 6 78 ng/mL 2 hours after operation (Fig 2). They also increased slightly in the Whipple group (from 241 6

69 ng/mL to 378 6 75 ng/mL; p 5 0.05) but remained almost unchanged in the pneumonectomy group. At baseline, all groups showed comparable normal plasma sVCAM-1 levels (Fig 3). In the CABG group, sVCAM-1 levels increased significantly and were higher than in the other two groups at the end of the study period. Plasma concentrations of sGMP-140 increased over the normal value only in the cardiac surgical group (from 69 6 12 ng/mL to 150 6 25 ng/mL) but had returned to normal on the morning of the first postoperative day (Fig 4). Levels of sGMP-140 remained unchanged in the other two groups.

Table 4. Summary of Hemodynamic Dataa

Variable MAP (mm Hg)

HR beats/min (min21) CI (L z min21 z m22)

SVRI (dyne z s z cm25 z m22) PaO2/FIo2 (mm Hg)

a

Group

Baseline

After Operation

CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy CABG Whipple Pneumonectomy

78 6 10 81 6 12 80 6 11 79 6 15 80 6 12 78 6 11 3.11 6 0.4 3.05 6 0.5 3.15 6 0.6 1,513 6 170 1,689 6 248 1,666 6 165 393 6 83 392 6 67 371 6 54

72 6 7 80 6 10 81 6 12 96 6 14 91 6 12 91 6 16 3.22 6 0.7 3.26 6 0.6 2.97 6 0.5 1,690 6 208 1,564 6 290 1,665 6 217 289 6 79b,c 381 6 66 295 6 70c

Data are shown as the mean 6 the standard deviation.

b

2 Hours After Operation

5 Hours After Operation

Postoperative Day 1

80 6 8 81 6 11 85 6 11 102 6 13 90 6 12 92 6 12 3.18 6 0.4 3.43 6 0.5 2.93 6 0.5 1,588 6 189 1,684 6 241 1,666 6 199 313 6 77b,c 377 6 65 311 6 67c

85 6 11 86 6 10 85 6 8 92 6 13 92 6 12 93 6 11 3.16 6 0.5 3.66 6 0.7 3.03 6 0.5 1,784 6 269 1,681 6 230 1,777 6 239 359 6 86 364 6 61 319 6 66c

82 6 10 86 6 12 83 6 15 84 6 14 87 6 15 89 6 12 3.03 6 0.6 3.33 6 0.5 2.98 6 0.6 1,516 6 236 1,740 6 277 1,726 6 211 422 6 62 402 6 54 329 6 62c

Significance: p , 0.05 versus Whipple group.

c

Significance: p , 0.05 versus baseline.

CABG 5 coronary artery bypass grafting; CI 5 cardiac index; HR 5 heart rate; MAP 5 mean arterial pressure; oxygen tension to inspired oxygen fraction ratio; SVRI 5 systemic vascular resistance index.

PaO2/FIo2 5 arterial

612

BOLDT ET AL ADHESION MOLECULES AND MAJOR OPERATIONS

Fig 1. Changes in plasma levels of circulating (soluble) endothelial leukocyte adhesion molecule-1 (sELAM-1) (normal range, 30 to 53 ng/mL; mean value, 42 ng/mL). (p.o. 5 postoperative; ** 5 p ,0.05 versus pneumonectomy group; † 5 p ,0.05 versus baseline.)

Comment The endothelium modulates vascular tone, controls local blood flow, influences the rate of leakage, modulates the accumulation and extravasation of leukocytes into the tissue, and markedly influences leukocyte activation [1]. The endothelial–leukocyte interactions, which are dependent on the presence of various adhesion molecules, appear to be a process that is highly involved in the possible development of organ failure [2, 6]. Enhanced neutrophil-endothelial binding with subsequent neutrophil migration takes place in adult respiratory distress

Fig 2. Changes in plasma levels of circulating (soluble) intercellular adhesion molecule-1 (sICAM-1) (normal range, 190 to 300 ng/mL; mean value, 210 ng/mL). (p.o. 5 postoperative; * 5 p ,0.05 versus other two groups; † 5 p ,0.05 versus baseline.)

Ann Thorac Surg 1998;65:608 –14

Fig 3. Changes in plasma levels of circulating (soluble) vascular cell adhesion molecule-1 (sVCAM-1) (normal range, 450 to 700 ng/mL; mean value, 530 ng/mL). (p.o. 5 postoperative; * 5 p ,0.05 versus other two groups; † 5 p ,0.05 versus baseline.)

syndrome, reperfusion injury to the myocardium, hemorrhagic shock, and CPB [8, 12, 13]. E-selectin (ELAM-1) is a 95- to 115-kD glycoprotein found exclusively on cytokine-activated endothelial cells [6]; thus, E-selectin appears to be a specific marker of endothelial cell activation [3]. This glycoprotein peaks after 4 to 6 hours of cytokine stimulation. Intercellular adhesion molecule-1 is not produced exclusively by endothelial cells. Its expression is markedly increased by cytokine stimulation (eg, interleukin-1 and TNF) [6] and peaks after 12 hours. Expression of VCAM-1 is also induced by cytokines and is sustained for more than 24 hours [2, 6]. P-selectin (GMP 140) is a rapidly expressed

Fig 4. Changes in plasma levels of circulating (soluble) P-selectin (granule membrane protein 140 [sGMP-140] (normal range, 50 to 115 ng/mL; mean value, 90 ng/mL). (p.o. 5 postoperative; * 5 p ,0.05 versus baseline; † 5 p ,0.05 versus other two groups.)

Ann Thorac Surg 1998;65:608 –14

surface adhesion molecule. It has been identified on activated platelets and in Weibel-Palade bodies of human endothelial cells [14]. Thrombin, histamine, and oxygen free radicals can mobilize P-selectin to the endothelial surface within minutes [12, 15]. Acute inflammation is associated with increased expression of E-selectin and P-selectin, whereas in more chronic inflammation processes, VCAM-1 expression predominates [2]. Soluble isoforms of the adhesion molecules modulate the inflammatory process in many different ways: by acting as a chemotaxin, blocking neutrophil activation, or competing with a membrane-bound form of cell-to-cell adhesion. One major finding in the present study was that a cardiac surgical procedure, ie, CABG, resulted in a significantly higher increase in plasma levels of all soluble adhesion molecules than did a major noncardiac surgical intervention (Whipple operation or pneumonectomy), thus implying more pronounced endothelial cell alterations in the cardiac surgical population than in the other patients. On the basis of our data, we can only speculate on the reasons for this dissimilarity. It has become increasingly clear that different predisposing conditions can result in a similar inflammatory response. Activation of the complement/coagulation cascade, macrocirculatory and microcirculatory dysfunction, and release of endotoxins are known triggers of proinflammatory cytokine production resulting in endothelial dysfunction with subsequent release of soluble adhesion molecules [4, 5, 16]. In the present study, the extent of the surgical procedure was similar in all groups, the biometric data were comparable, the anesthesia regimen was identical, and the duration of operation was longest in the Whipple group. Nevertheless, plasma levels of soluble adhesion molecules were highest in the cardiac surgical patients. It is widely accepted that several immunoregulatory mediators are produced by CPB [7]. Cardiopulmonary bypass also induces an increase in the percentage of circulating platelets that express GMP-140 [17]. This might explain why sGMP-140 was increased beyond normal only in the cardiac surgical patients and why it is most likely caused by exposure of blood to the nonphysiologic surfaces of the CPB system. In addition, there is convincing evidence that ischemia plus reperfusion may be an even more important trigger for cytokine release during CPB [18], and this release enhances the risk of subsequent endothelial dysfunction. The hemodynamic data were not significantly different between the three groups. There is a balance of forces promoting leukocyte or endothelial adhesion and hemodynamic dispersal forces. Blood flow–induced shear forces play an important role in disrupting the adhesive interaction between neutrophils and endothelium [19]. It can be assumed that the microcirculation of patients undergoing a cardiac operation is affected more (especially during CPB) than that of patients having a noncardiac surgical procedure. The hypothesis that the intestinal organs become ischemic during CPB is supported by a study by Fiddian-Green [20]. This ischemia may result in endothelial cell alterations either directly through

BOLDT ET AL ADHESION MOLECULES AND MAJOR OPERATIONS

613

microcirculatory abnormalities or indirectly by fluid translocation phenomena, which are associated with release of endotoxins and proinflammatory cytokines. The importance of circulating adhesion molecules is a much debated issue. Plasma levels of sICAM-1 have been reported to be of prognostic value after heart transplantation [21] and in patients in the intensive care unit [22]. Sessler and associates [23] compared plasma sICAM-1 levels in patients with sepsis, severe sepsis, and septic shock. Soluble ICAM-1 concentrations were markedly higher in patients with fatal sepsis (1,793 ng/mL 6 1,026 ng/mL) than in patients who survived (786 6 454 ng/mL). Plasma concentrations of sICAM-1 greater than 1,000 ng/mL were associated with a higher rate of death. Cowley and coworkers [16] found elevated plasma levels of soluble adhesion molecules in patients with systemic inflammatory response and organ dysfunction. Particularly, elevated soluble E-selectin (sELAM-1) was correlated with the development of multiple-organ dysfunction and death. Two studies [13, 20] have shed light on an interesting therapeutic aspect of adhesion molecules in cardiac surgery: by blocking neutrophil adhesion molecules during reperfusion, the authors [13, 24] achieved complete prevention of myocardial stunning, contracture, low reflow, and edema after heart transplantation. None of the patients in the present study sustained high fever or organ dysfunction or needed prolonged intensive care therapy. The arterial oxygen tension to inspired oxygen fraction ratio was significantly reduced in the immediate postoperative period in the cardiac surgical patients compared with the Whipple group; this finding indicates that pulmonary function was affected more profoundly in the former group. It has been shown that the endothelium may become increasingly permeable during CPB, and this results in an increase in extravascular lung water in the period after CPB [25]. The fact that plasma levels of circulating adhesion molecules were increased only temporarily in the patients having elective CABG may imply that endothelial activation or dysfunction was also of only short duration. The extent of the increase in circulating adhesion molecules in our cardiac surgical patients was much lower than in patients dying of severe sepsis. Thus, it can be assumed that the extent of endothelial cell alterations was limited only in our elective first-time CABG patients. In patients undergoing more complex, long-lasting cardiac surgical procedures, this elevation of circulating adhesion molecules may be much higher and sustained, which would increase the risk of the development of organ dysfunction. In summary, endothelial function abnormalities can be triggered or modified by circulatory, respiratory, or metabolic disorders. A cardiac surgical procedure, ie, CABG, resulted in a significant increase in the plasma levels of all soluble adhesion molecules measured. Although duration of the operation and extent of the surgical procedure were similar or greater in patients undergoing a Whipple operation or pneumonectomy, plasma levels of circulation adhesion molecules were not markedly changed in these patients. Most likely, the interactions of blood with the CPB equipment, the ischemia/reperfusion phenomenon, or both are responsible for the differences

614

BOLDT ET AL ADHESION MOLECULES AND MAJOR OPERATIONS

between cardiac and noncardiac surgical patients. Whether monitoring soluble adhesion molecules will be a valuable tool for the early diagnosis of endothelial damage and the prognosis for the risk of development of multiple-organ failure must be elucidated in further controlled studies.

References 1. Kirkpatrik CJ, Klosterhalfen B, Hauptmann S. The role of the endothelium in multiple organ failure. In: Vincent JL, ed. Yearbook of intensive care and emergency medicine. New York: Springer, 1992:14–24. 2. Adams DH, Shaw S. Leucocyte-endothelial interactions and regulation of leucocyte migration. Lancet 1994;343:831– 6. 3. Bevilacqua MP, Nelson RM. Selectins. J Clin Invest 1993;91: 379– 87. 4. Springer TA. Adhesion receptors of the immune system. Nature 1990;346:425–34. 5. Gearing AJH, Newman W. Circulating adhesion molecules in disease. Immunol Today 1993;14:506–12. 6. Ley K. Leukocyte adhesion to vascular endothelium. J Reconstr Microsurg 1992;8:495–503. 7. Casey LC. Role of cytokines in the pathogenesis of cardiopulmonary-induced multisystem organ failure. Ann Thorac Surg 1993;56:S92– 6. 8. Gillinov AM, Bator JM, Zehr KJ, et al. Neutrophil adhesion molecule expression during cardiopulmonary bypass with bubble and membrane oxygenators. Ann Thorac Surg 1993; 56:847–53. 9. Boldt J, Osmer C, Linke LC, Go¨rlach G, Hempelmann G. Hypothermic versus normothermic cardiopulmonary bypass: influence on circulating adhesion molecules. J Cardiothorac Vasc Anesth 1996;10:342–7. 10. Cruickshank AM, Fraser WD, Burns HJG, van Damme J, Shenkin A. Response of serum interleukin-6 in patients undergoing elective surgery of varying severity. Clin Sci 1990;79:161–5. 11. Crouzier A, Mu¨ller JE, Quittkat D, Sydow M, Wuttke W, Kettler D. Effect of anaesthesia on the cytokine response to abdominal surgery. Br J Anaesth 1994;72:280–5. 12. Mulligan MS, Polley MJ, Bayer RJ, Nunn MF, Paulson JC,

Ann Thorac Surg 1998;65:608 –14

13.

14. 15. 16.

17. 18.

19. 20. 21. 22. 23. 24.

25.

Ward P. Neutrophil-dependent acute lung injury. J Clin Invest 1992;90:1600–7. Byrne JG, Smith WJ, Murphy MP, Couper GS, Appleyard RF, Cohn LH. Complete prevention of myocardial stunning, contracture, low-reflow, and edema after heart transplantation by blocking neutrophil adhesion molecules during reperfusion. J Thorac Cardiovasc Surg 1992;104:1589–96. Bevilacqua MP. Endothelial-leukocyte adhesion molecules. Annu Rev Immunol 1993;11:767– 84. Skinner MP, Likas CM, Burns GF, Chesterman CN, Berndt MC. GMP-140 binding to neutrophils is inhibited by sulfated glycans. J Biol Chem 1991;26:5371– 4. Cowley HC, Heney D, Gearing AJH, Hemingway I, Webster NR. Increased circulating adhesion molecule concentrations in patients with septic inflammatory response: a prospective cohort study. Crit Care Med 1994;22:651–7. Rinder CS, Bonan JL, Rinder HM, Mathew J, Hines R, Smith BR. Cardiopulmonary bypass induces leucocyte-platelet adhesion. Blood 1992;5:1201–5. Kawamura T, Wakusawa R, Okada K, Inada S. Elevation of cytokines during open heart surgery with cardiopulmonary bypass: participation of interleukin-8 and 6 in reperfusion injury. Can J Anaesth 1993;40:1016–21. Mayrovitz H, Wiedeman M, Tuma R. Factors influencing leukocyte adherence in microvessels. Thromb Haemost 1977; 38:823–30. Fiddian-Green RG. Association between intramucosal acidosis in the gut and organ failure. Crit Care Med 1993;21: S103–7. Ballantyne CE, Mainolfi JB, Young NT. Prognostic value of increased levels of circulating intercellular adhesion molecules-1 after heart transplantation. Clin Res 1992;39:285a. Endo S, Inada K, Kasai T, et al. Levels of soluble adhesion molecules and cytokines in patients with septic multiple organ failure. J Inflamm 1995;46:212–9. Sessler CN, Windsor A, Watson L, Schwartz M, Fowler AA. Soluble ICAM-1 receptor levels are increased in human septic shock. Chest 1993;104(Suppl):11S. Verrier ED, Shen I. Potential role of neutrophil antiadhesion therapy in myocardial stunning, myocardial infarction, and organ dysfunction after cardiopulmonary bypass. J Cardiac Surg 1993;8:309–12. Boldt J, von Bormann B, Kling D, Hempelmann G. Age and cardiac surgery—influence on extravascular lung water. Chest 1987;91:185–9.