The concept of cardiopulmonary bypass (CPB)-related "whole body inflammatory response," which involves activation of several mediator cascades, has been appreciated for many years . The immune system normally serves to protect the host from offenders; however, an ongoing (overreacting) immunologic response may have detrimental sequelae resulting in "postperfusion syndrome" . One important trigger for this response appears to be the exposure of blood to the nonendothelized synthetic surfaces of the CPB equipment . Thus, "postperfusion syndrome" is defined as a noninfectious "whole body inflammatory response" directed against nonendothelialized material of the CPB circuit surfaces. Polymorphonuclear granulocytes (PMNs) appear to be of major importance in this process [2,3]. The role of PMNs in this situation is complex, and includes rolling of PMNs along the endothelium, adherence to the endothelium, and transendothelial migration in the tissue . After attachment to the endothelial cells, neutrophils get activated and release several inflammatory mediators, toxic oxygen metabolites, and phospholipase products, which are involved in an increased permeability of the endothelium [5,6]. Attached neutrophils then migrate transendothelially and further contribute to organ inflammation. This complex process is initiated and maintained by interactions of circulating leukocytes and by interactions of PMNs with the endothelium via cell- and organ-specific adhesion molecules [7,8]. Three different families of adhesion receptors are known: 1) the immunoglobulin superfamily (e.g., vascular cell adhesion molecule-1 [VCAM-1] and intercellular adhesion molecule-1 [ICAM-1], 2) the integrin family, and 3) the selectins (E-selectin = endothelial leukocyte adhesion molecule; P-selectin = granule membrane protein 140). Initial interactions are mediated by members of the selectin family to (loosely) associate the leukocytes with the endothelium, followed by firm adhesion requiring members of the integrin and immunoglobulin family .
In addition to membrane-bound adhesion molecules, soluble subforms of these molecules can be found in the circulating blood under various circumstances [9,10]. In adult septic patients, they have been judged as markers of endothelial activation or damage and even have been described as markers of outcome .
Changes of circulating soluble adhesion molecules have only rarely been documented in adults undergoing cardiac surgery . Even less information is available in pediatric cardiac surgery patients. In this descriptive study, we investigated whether plasma levels of soluble circulating adhesion molecules in pediatric cardiac surgery patients were different from those in adult cardiac surgery patients.
Fifteen consecutive children undergoing corrective operation of congenital heart disease were included in this prospectively designed study. Reoperations were excluded from the study; none of the children studied underwent deep hypothermic circulatory arrest, and none showed signs of infection preoperatively (white cell count in all children <8000/m3, temperature <38.0 degrees C). Fifteen consecutive adults scheduled for primary coronary artery bypass grafting and valve replacement were also studied. Diabetes mellitus and renal and liver insufficiency were defined as excluding criteria. None of the patients had taken aspirin, other cyclooxygenase inhibitors, or steroids within the last 10 days before the operation. Informed consent was obtained from the children's parents and the adults according to the protocol of the Ethic Study Board of the hospital.
Anesthesia was comparable for all patients and consisted of weight-related doses of fentanyl, midazolam, and pancuronium bromide. All children were mechanically ventilated as needed according to monitoring of pulse oximetry, end-expiratory carbon dioxide, and blood gas analyses. Volatile anesthetics were not used in any patient. In the children, cefuroxine (100 mg/kg three times a day) was given for 2 days; the adults received cefamandol (25 mg/kg three times a day) for 2 days. All antibiotics were started before onset of the operation after induction of anesthesia.
Bovine heparin (300 IU/kg) was administered 10 min before the start of CPB to achieve anticoagulation. Half of the initial dose was given after 60 min of CPB to keep activated clotting time (ACT) >400 s within the entire bypass period. In pediatric cardiac surgery, a COBE VPCMLplus Trademark membrane oxygenator (Cobe, Lakewood, CO) was used, and a flow of 2.2 L centered dot min-1 centered dot m2 was maintained during CPB. Priming of the extracoporeal circuit consisted of 500 mL of Ringer's solution, 250 mL of 5% human albumin, and 50 mL of 20% human albumin. Packed red blood cells (PRBCs; not older than 3 days) were added to the prime according to the children's weight and preoperative hemoglobin (Hgb) value. When necessary, Ringer's solution was added to maintain circuit volumes. When the Hgb value fell below 7 g/dL, PRBCs were given. After the end of CPB, the blood remaining in the extracorporeal oxygenation equipment was salvaged by a centrifugation device (Cell Saver, Hemonetics, Braintree, MA), and this autologous blood was retransfused in the postbypass period. Lowest temperature during CPB in the children ranged from 30 to 33 degrees C (32.7 +/- 1.2 degrees C). Before weaning from bypass, temperature was increased to 36 degrees C.
In adult cardiac surgery, CPB was performed using a membrane oxygenator (Monolyth Trademark; Sorin, Turino, Italy). Blood was returned to the circuit using a two-stage cannula (monoatrial cannulation technique). This technique implies that all fluids (cardioplegic solution, suction, cooling) are returned to the extracorporeal circuit. Within 20 min after start of CPB, blood of the circuit was also concentrated by a centrifugation device (Cell Saver) with the aim to adjust the Hgb level between 8 and 9 g/dL. The residual blood remaining in the circuit after CPB was salvaged by the same technique and retransfused after CPB. Rectal temperature during adult CPB ranged from 33 degrees to 35 degrees C (33.9 +/- 0.5 degrees C).
After termination of CPB, in all patients heparin was neutralized by protamine chloride in a ratio of 1:1 with regard to the initially given dose of heparin. Postoperatively, homologous blood (PRBCs) and blood derivatives (fresh frozen plasma, platelet concentrates) were given by physicians who were not involved in the study. In the pediatric cardiac surgery group, PRBCs were given when Hgb was less than 10 g/dL; fresh frozen plasma was indicated when bleeding exceeded 5 mL centered dot kg-1 centered dot h-1 and the platelet count was >50,000/mL (ACT <200 s). Platelet concentrates were administered when bleeding exceeded 5 mL centered dot kg-1 centered dot h-1 and the platelet count was <50,000/mL (ACT <200 s).
Hgb, hematocrit, neutrophil count, ACT (using kaolin as activator and a Hemochron Registered Trademark system [International Technidyne Corp., Edison, NJ]), and blood gas variables were measured from arterial blood samples. Additionally, soluble circulating endothelial leukocyte adhesion molecule-1 (sELAM-1), intercellular adhesion molecule-1 (sICAM-1), and vascular cell adhesion molecule-1 (sVCAM-1) were measured using commercially available double-antibody enzyme-linked immunosorbent assay kits (British Bio-technology Products, Abington, UK) according to the instructions of the manufacturer. Blood samples were drawn into sodium citrate-containing tubes and immediately spun at 3000g for 10 min; the serum was removed and stored at -70 degrees C. All measurements were performed within 2 mo after blood sampling. All data from enzyme-linked immunosorbent assays represent the mean from duplicate measurements.
In six healthy volunteers (ages ranging from 22 to 25 yr), adhesion molecules were additionally measured to define "normal" values). In six children <5 yr undergoing noncardiac major surgery, plasma levels of sELAM-1, sICAM-1, and sVCAM-1 were measured prior to surgery to define normal values in children. Informed consent was also obtained from the adult patients or the parents.
In all cardiac surgery patients, a volume correction for all adhesion molecules was made to eliminate hemodilution effects : Equation 1 where phct = plasma hematocrit (100% - hct%), to = value at baseline, and ti = value at the time of sampling. The actual measured plasma level of the soluble adhesion molecule was multiplied by this correction factor.
All patients were operated on in the morning by the same surgical team. All measurements were carried out after induction of anesthesia ("baseline" values), 20 min after start of CPB (after hemoconcentration), at the end of surgery, and on the postoperative days 1 and 2 (at 9:00 AM).
Mean values and SDS were calculated for all parameters. One-way and two-way analysis of variance with repeated measures was used to determine the effects of group, time, and group-time interaction for each measured parameter. Scheffe's test was additionally carried out. Analyses of covariance and regression analyses were used (distinct for each group) to assess a relationship between two parameters (e.g., analyses of covariance: duration of CPB and plasma levels of one adhesion molecule; regression analyses: temperature and plasma levels of one adhesion molecule). P values less than 0.05 were considered significant.
Biometric profile and data from cardiopulmonary bypass (CPB) are listed in Table 1. Characteristics of the underlying disease of the children and type of operation in the adults are listed in Table 2. One child died of multiple organ failure during the stay on the pediatric intensive care unit, whereas none of the adults died Table 1. Neutrophil count was different only on postoperative day 2, with a significantly higher count in the children Table 3. At baseline, PaO2/FIO2 was significantly lower in the children than in the adults Table 3. Ten of the 15 children were cyanotic at baseline, showing a PaO2/FIO2 of less than 100 mm Hg after induction of anesthesia.
Normal values of circulation adhesion molecules derived from the healthy adult volunteers were defined as follows: sELAM-1: normal range, 35 to 55 ng/mL (mean, 45 ng/mL); sICAM-1: normal range, 220 to 280 ng/mL (mean, 210 ng/mL); and sVCAM-1: normal range, 400 to 680 ng/mL (mean, 540 ng/mL). In the noncardiac children, sELAM-1 ranged from 43 to 64 ng/mL (mean, 48 ng/mL), sICAM-1 ranged from 185 to 295 ng/mL (mean, 245 ng/mL), and sVCAM-1 ranged from 450 to 605 ng/mL (mean, 515 ng/mL).
At baseline, the plasma level of sELAM-1 was significantly higher in the children (88.8 +/- 13.8 ng/mL) than in the adults (30.7 +/- 8.0 ng/mL) and was beyond the normal range only in the children Figure 1. In the adults, sELAM-1 plasma concentrations remained almost unchanged in the postbypass period, whereas in the children sELAM-1 was significantly lower during and after CPB as well as in the postoperative period (45.2 +/- 12.2 ng/mL at postoperative day 2).
The sICAM-1 plasma concentration Figure 2 in the children was also increased beyond normal values at baseline (349 +/- 27 ng/mL) and was significantly different from the sICAM-1 level in the adults (229 +/- 29 ng/mL). In the children, it was significantly lower at the end of the investigation period (254 +/- 30 ng/mL on postoperative day 2) than at baseline. In the adults, sICAM-1 plasma concentrations remained almost unchanged within the entire study period.
The circulating sVCAM-1 plasma level in the children (564 +/- 49 ng/mL) was higher than that in the adults (352 +/- 23 ng/mL) at the beginning of the study Figure 3. In the children, it significantly decreased during CPB until the end of surgery and remained lower than baseline values in the postoperative period (postoperative day 2: 450 +/- 33 ng/mL). In the adults, sVCAM-1 plasma concentrations did not change significantly throughout the entire study period.
A significant inverse relationship between plasma levels of soluble adhesion molecules (distinct for sELAM-1, sICAM-1, and sVCAM-1) and PaO2/FIO2 ratio at baseline was found (regression analysis of baseline values: r = -0.78; P < 0.01). No significant relationship between plasma levels of any of the soluble adhesion molecules and any of the other laboratory (e.g., neutrophil count) and clinical (e.g., temperature) data was seen (analyses of covariance and regression analyses).
The ability of neutrophils to move toward an area of infection and/or inflammation may be affected by CPB, and their ability to ingest and kill bacteria may also be altered [3,14]. Silva et al.  showed a decrease in phagocytic function during CPB that recovered as early as the first postoperative day, thus implicating a transient abnormal phagocytic capacity of granulocytes.
Adhesion molecules appear to play a central role in initiating and modifying the complex process of "rolling," attachment, and transendothelial migration of PMNs. Expression of membrane-bound ICAM-1, ELAM-1, and VCAM-1 is known to be up-regulated secondary to inflammatory response [16,17]. ICAM-1 and VCAM-1 seem to be of particular importance for attachment and transendothelial migration of leukocytes . Unlike other adhesion molecules, such as VCAM-1 and ICAM-1, ELAM-1 is not constitutively present in normal endothelium , and thus it appears to be a useful marker for inflammation .
A growing body of evidence is now accumulating to suggest that soluble forms of these adhesion molecules found in the circulating blood may reflect endothelial activation and damage or even may predict outcome [9,18-20]. The increase of plasma levels of soluble adhesion molecules may result from either increased expression and subsequent increased release into the circulating blood, proteolytic cleavage of membrane-bound forms secondary to endothelial damage, or both [9,10].
The role of these circulating forms of adhesion molecules, however, is not fully defined: sE-selectins appear to be biologically active, being able to mediate the adhesion of neutrophils ; sELAM-1 selectively binds to cells that (normally) bind to cell-surface ELAM-1 (e.g., neutrophil molecule CD11/CD18) . A recombinant soluble form of ELAM-1 was shown to be fully functional as an adhesion protein . Soluble circulating ICAM-1 is also biologically active, retaining the ability to bind to LFA-1 (a leukocyte integrin) . It has been suggested that circulating sICAM-1 may regulate cell adhesion by promoting de-adhesion . It may compete with membrane-bound ICAM-1 for leukocyte adhesion molecules, thus preventing attachment (or even promoting deattachment) of white cells and thus preventing neutrophil-induced tissue damage.
In the present study, plasma levels of all measured soluble adhesion molecules were significantly higher at baseline in the pediatric, compared to the adult, cardiac surgery patients. Concentration of soluble adhesion molecules in the circulating blood may not be as important as local levels of adhesion molecules, particularly at sites of inflammation. Thus it might be assumed that in these children even higher levels of these adhesion molecules can be found in the microcirculation. sELAM-1 and sICAM-1 plasma concentrations exceeded even normal values in these children. The etiology of the elevated plasma levels of these adhesion molecules can only be speculated upon at present. Direct or indirect effects of hypoxia, microcirculatory abnormalities , release of proteinases from activated white cells , activation of mediator systems (e.g., the complement or coagulation system ), and activation of platelets all may contribute to an increased expression of membrane-bound adhesion molecules and the release of soluble subforms into the circulating blood. Leukocyte adhesion and emigration take place almost exclusively in the venular network because shear rates in the venules are rather low. When flow is reduced in arterioles to levels similar to those in the venules, leukocytes still roll much better on venules than on arterioles , indicating preferential expression of adhesion molecules by endothelial cells of venules . Microcirculatory deficits can be assumed in most children suffering from congenital heart disease because of impaired macrohemodynamics and an elevated hematocrit, which may both result in a reduced blood flow in the microcapillary network. Hypoxia and low-flow state are considered to be important stimuli for activating the inflammatory cascade [22,27]. In the present study, PaO2/FIO2 was significantly lower in the children at baseline (10 of the 15 children showed a PaO2/FIO2 was less than 100 mm Hg). There was a significant inverse relationship between elevated plasma levels of adhesion molecules and PaO (2/FIO)2 in the children. Endothelial function abnormalities induced by microcirculatory disturbances may result in an increased expression of membrane-bound adhesion molecules and then, by proteolytic cleavage, result in elevated levels of circulating adhesion molecules. The reduced plasma levels of all measured soluble adhesion molecules during and after CPB may reflect improved (macro- and micro-) circulation and tissue oxygenation after correction of congenital heart disease and associated hypoxemia.
Our data showing a reduction in plasma levels of all adhesion molecules in the post-CPB period in our children appear to disagree with those in a sophisticated paper from Kilbridge et al.  in which messenger RNA (mRNA) encoding the adhesion molecules E-selectin and ICAM-1 from samples of cardiac tissue from infants (<2 yr) undergoing CPB were measured. Samples were obtained before and at the conclusion of bypass. CPB resulted in a significant increase of mRNA (mean increase, 2.1- to 3.5-fold). Some of the children, however, did not show a relevant increase of E-selectin and ICAM-1 mRNA (e.g., a 0.6- to 1.1-fold increase). CPB in this study was very nonhomogeneous (12 of 20 patients underwent deep hyopthermic circulatory arrest; CBP times ranged from 42 to 229 min), and it is doubtful that the two studies can be compared. The extreme conditions of CPB in Kilbridge et al.'s study may be associated with a more pronounced increase in circulating endotoxin and various inflammatory mediators (e.g., interleukin-6, tumor necrosis factor ) with the risk of subsequent increased expression of endothelial adhesion molecules. Extrapolating from their results, Kilbridge et al. speculated on the benefits of eventual use of antiadhesion molecule therapy in the clinical setting. This may, however, be a two-edged sword, because altering the host response by modifying the function of adhesion molecules may attenuate the injury caused by inappropriate behavior of the host defense cells. On the other hand, this may leave the host without an effective defense mechanism. Thus, whether specific antibodies directed at leukocyte and endothelial cell adhesion molecules, which block the endothelial binding of neutrophils and prevent the endothelial damage, are a useful therapeutic strategy has to be proven in future studies.
According to the findings of Rothlein et al. , it is an open question whether the elevated plasma levels of adhesion molecules in our children only indicate endothelial activation (or damage) or whether they may have beneficial, protective effects in this situation: soluble adhesion molecules bind to circulating PMNs and hinder the binding of PMNs to the endothelium, thus preventing the neutrophil-mediated endothelial injury .
Our results concerning the course of plasma levels of soluble adhesion molecules in adults undergoing cardiac surgery are in accordance with a study of Gillinov et al. : the sICAM-1 plasma level was approximately 200 ng/mL before CPB and remained almost unchanged when a membrane oxygenator was used. In a group in which CPB was carried out with a bubble oxygenator, sICAM-1 significantly increased 24 h after CPB (from 175 to approximately 290 ng/mL), implicating greater endothelial damage with the bubble than with the membrane oxygenator . In a recent study, Holdright et al.  investigated plasma levels of von Willebrand factor antigen (vWF Ag), which is considered to be a marker of endothelial injury. The plasma level of vWF Ag significantly increased from 0.75 +/- 0.11 IU/mL (prior to the operation) to 0.95 +/- 0.12 IU/mL (prior to termination of CPB) without, however, exceeding normal values (ranging from 0.49 to 1.44 IU/mL). Whether the increase by 0.20 IU/mL is important and whether these changes indicate relevant endothelial damage must be questioned. Since vWF Ag plasma levels were always within the normal range, it seems more likely that with modern techniques of CPB marked endothelial damage is unlikely in most patients.
In summary, circulating adhesion molecules may serve as markers of endothelial activation or damage. The actual role of soluble adhesion molecules such as sICAM-1, sELAM-1, and sVCAM-1 has not been fully elucidated. In children scheduled for pediatric cardiac surgery, plasma levels of all measured circulating adhesion molecules were significantly higher before CPB than in adult cardiac surgery patients. Whether this may be regarded as a marker of activation or damage of the endothelium already prior to CPB, or as a protective reaction to prevent PMN-related tissue destruction, warrants further controlled studies. In both pediatric and adult cardiac surgery patients, CPB appears to possess only a moderate influence on levels of soluble adhesion molecules.
1. Kirklin JK, Westaby S, Blackstone EH. Complement and the damaging effects of cardiopulmonary bypass. J Thorac Cardiovasc Surg 1983;86:845-52.
2. Knudsen F, Andersen LW. Immunological aspects of cardiopulmonary bypass. J Cardiothorac Anesth 1990;4:245-58.
3. Westaby S. Aspects of biocompatibility in cardiopulmonary bypass. Crit Rev Biocompat 1987;3:193-234.
4. Paulson JC. Selectin/carbohydrate-mediated adhesion of leukocytes. In: Harlan JM, Liu DY, eds. Adhesion: its role in inflammatory disease. New York: WH Freeman and Co, 1991:19-42.
5. Royston D, Fleming JS, Desai JB, et al. Increased production of peroxidation products associated with cardiac operations. J Thorac Cardiovasc Surg 1986;91:759-66.
6. Butler J, Rocker GM, Westaby S. Inflammatory response to cardiopulmonary bypass. Ann Thorac Surg 1993;55:552-9.
7. Center DM, Berman JS, Kornfield H, et al. Mechanisms of lymphocyte accumulation in pulmonary disease. Chest 1993;103(Suppl 3):88S-91S.
8. Springer TA. Adhesion receptors of the immune system. Nature 1990;346:425-34.
9. Gearing AJH, Newman W. Circulating adhesion molecules in disease. Immunol Today 1993;14:506-12.
10. Rothlein R, Mainolfi EA, Czajkowski M, Marlin SD. A form of circulating ICAM-1 in human serum. J Immunol 1991;147:3788-93.
11. Cowley HC, Heney D, Gearing AJH, et al. Increased circulating adhesion molecule concentrations in patients with septic inflammatory response: a prospective cohort study. Crit Care Med 1994;22:651-7.
12. 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.
13. Feindt P, Volkmer I, Seyffert UT, Haack H. The role of protein C as an inhibitor of blood clotting during extracorporeal circulation. Thorac Cardiovasc Surg 1991;39:338-43.
14. Kress HG, Gehrsitz P, Elert O. Predictive value of skin testing, neutrophil migration and C-reactive protein for postoperative infections in cardiopulmonary bypass patients. Acta Anaesthesiol Scand 1987;31:397-404.
15. Silva J, Hoeksema H, Fekety FR. Transient defects on phagocytic functions during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1974;67:174-83.
16. Cotran RS, Pober HS. Endothelial activation. Its role in inflammatory immune reactions. In: Simionescu N, Simionescu M, eds. Endothelial cell biology. New York: Plenum, 1988:335-47.
17. Osborn L. Leukocyte adhesion to endothelium in inflammation. Cell 1990;62:3-6.
18. Seth R, Raymond FD, Makgoba MW. Circulating ICAM-1 isoforms: diagnostic prospects for inflammatory and immune disorders. Lancet 1991;338:83-4.
19. Kayal S, Guerot E, Rezgui MN, et al. Elevated levels of circulating ICAM-1 and E-selectin in the plasma of patients with severe sepsis syndrome. Intensive Care Med 1994;20(Suppl 1):S26.
20. Ballantyne CE, Mainolfi JB, Young NT. Prognostic value of increased levels of circulating intercellular adhesion molecules-1 after heart transplantation [abstract]. Clin Res 1992;39:285a.
21. Lobb R, Chi-Rosso G, Leone DR, et al. Expression and functional characterization of a soluble form of endothelial-leukocyte adhesion molecule 1. J Immunol 1991;147:124-9.
22. Chaudry IH, Ayala A. Immunological aspects of hemorrhage. Austin, TX: Medical Intelligence Unit, RG Landes Company, 1992.
23. Yamazaki M, Asakura H, Sato T, et al. Changes in plasma levels of thrombomodulin during hemodialysis. Blood Coagul Fibrinolysis 1992;3:113-7.
24. Sakurabashi I, Itou H, Ishii S. Studies on granulocyte elastase in plasma. Jpn J Clin Hematol 1985;33:1113-8.
25. Ley K, Gaethgens P. Endothelial, not hemodynamic, differences are responsible for preferential leukocyte rolling in rat mesenteric venules. Circ Res 1991;69:1034-41.
26. McEver RP. Leukocyte interactions mediated by selectins. Thromb Haemost 1991;66:80-7.
27. Chaudry IH, Ayala A. Immune consequences of hypovolemic shock and resuscitation. Curr Opin Anesth 1993;6:385-92.
28. Kilbridge PM, Mayer JE, Newburger JW, et al. Induction of intracellular adhesion molecule-1 and E-selectin mRNA in heart and skeletal muscle of pediatric patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1994;107:1183-92.
29. Butler J, Chong GL, Baigrie RJ, et al. Cytokine responses to cardiopulmonary bypass with membrane and bubble oxygenation. Ann Thorac Surg 1992;53:833-8.
30. Harlan JM. Neutrophil-mediated vascular injury. Acta Med Scand 1987;15(Suppl 7):123-9.
31. Holdright DR, Hunt BJ, Parratt R, et al. The effects of cardiopulmonary bypass on systemic and coronary levels of von Willebrandt factor. Eur Cardiothorac Surg 1995;9:18-21.