Changes in adhesion molecule expression and oxidative burst activity of granulocytes and monocytes during open-heart surgery with cardiopulmonary bypass compared with abdominal surgery

Abstract

Cardiac and major abdominal surgery are associated with granulocytosis in peripheral blood. The purpose of the present study was to describe the granulocyte and monocyte oxidative burst and the expression of adhesion molecules following cardiac surgery with cardiopulmonary bypass and abdominal surgery. The ability to respond with an oxidative burst was measured by means of flow cytometry using 123-dihydrorhodamine. The adhesion molecules CD11a/CD18, CD11c/CD18, CD44 were measured using monoclonal antibodies. Blood samples from eight patients undergoing open-heart surgery were taken before surgery, 1, 5, 10 and 20 min after aortic clamping, and then 1, 5, 10 and 20 min and 1, 2 and 3 h after declamping. Samples from eight patients undergoing abdominal surgery were taken before surgery, at the end of surgery, and 2 and 3 h post-operatively. A decrease in number of granulocytes and monocytes during cardiopulmonary bypass was observed. The percentage of CD11a-positive granulocytes increased from 30% pre-operatively to 75% following cardiopulmonary bypass, while CD44-positive granulocytes increased from 5% to 13%. Despite the extent of the changes, these were not significant. The oxidative burst of the granulocytes and monocytes decreased after declamping to 15% and 27% of initial values in vitro. Several hours after surgery, there was no significant difference between the two groups. These results can be explained by a granulocyte and monocyte refractory response developing subsequent to an increased per-operative oxidative burst activity, and the induction of adhesion molecules on granulocytes associated with the cardiopulmonary bypass and surgery. In conclusion, open-heart surgery with cardiopulmonary bypass was associated with a rapid and pronounced activation of leukocytes which may play a role in reperfusion injury.

Introduction

Cardiac and major abdominal surgery are associated with granulocytosis in peripheral blood, and an increased secretion of hormones and cytokines. In particular, the cytokines IL-1, IL-6, IL-8 and TNF have been implicated in surgical stress ^[1,2]. These cytokines activate the endothelium and the granulocytes, resulting in an accumulation of granulocytes in the area of surgery ^[3]. The expression of adhesion molecules on the surface of granulocytes is essential for their ability to transmigrate through endothelium to underlying tissues ^[4].

The immunological changes which occur during and after cardiac surgery with cardiopulmonary bypass (CPB) are somewhat greater than with other types of major surgery, possibly because of the CPB itself. The blood-surface interaction during CPB may induce a systemic inflammatory reaction with activation of granulocytes and monocytes ^[5,6]. The activated granulocytes and monocytes may be involved in reperfusion injury following CPB. Kawamura et al.^[7] have demonstrated an increased level of elastase immediately after CPB. Animal studies have also revealed an accumulation of granulocytes in lungs and liver following endotoxin-induced inflammation ^[8].

The aim of this study was to assess the oxidative burst activity of granulocytes and monocytes, and to measure the expression of adhesion molecules on granulocytes during and following open-heart surgery with CPB compared with abdominal surgery.

Material and methods

The present study was approved by the Regional Ethics Committee and all patients gave their informed consent. The present authors studied eight patients scheduled for open-heart surgery, and eight patients scheduled for hysterectomy as a result of fibromyoma or excessive menorrhagia. Patients with cancer or endocrine disorders were excluded.

Pre-anaesthetic medication included diazepam (0.2 mg kg⁻¹) in both groups. The patients who underwent open-heart surgery were anaesthetized with fentanyl (50 μg kg⁻¹) and midazolam (0.14 mg kg⁻¹). Pancuronium (0.1 mg kg⁻¹) was given to facilitate tracheal intubation. Ventilation was controlled to maintain normocapnia. The ECG, oesophageal and rectal temperature, arterial blood oxygen saturation, arterial pressure in the radial artery, central venous pressure, and pulmonary artery pressure were monitored continuously. The patients who underwent hysterectomy were anaesthetized with thiomebumal (4 mg kg⁻¹), midazolam (0.07 mg kg⁻¹), fentanyl (5 μg kg⁻¹), pancuronium (0.1 mg kg⁻¹) and 66% N₂O in oxygen. Ventilation was controlled to maintain normocapnia.

Extracorporeal circulation was conducted in the usual fashion with a Polystan (Polystan, Copenhagen, Denmark) roller pump, a Dideco 704 (Dideco, Mirandola, Italy) membrane oxygenator, a cardiotomy reservoir, and a Pall (Pall Biomedical Products, New York, NY, USA) arterial filter. Crystalloid solution was used to prime the extracorporeal circuit. The patients received heparin at a dose of 300 units kg⁻¹ body weight before cannulation. Protamine (1.0 mg 100 units⁻¹ heparin) was administered over 5 min to neutralize heparin post-CPB.

Heparinized blood samples were collected from a central venous catheter, the cardiac patients, and from an intravenous (i.v.) catheter in an antecubital vein, the gynaecological patients.

The granulocyte and monocyte oxidative burst was measured pre-operatively, before aortic clamping, 1, 5, 10 and 20 min after aortic clamping, and 1, 5, 10 and 20 min and 2 and 3 h after aortic declamping in the open-heart surgery group. The adhesion/activation molecules on granulocytes were measured before surgery, before aortic clamping, 10 min after aortic clamping, and 1, 2 and 3 h after aortic declamping.

The adhesion molecules on granulocytes and monocytes were measured pre-operatively, at the end of surgery, and 2 and 3 h post-operatively in the patients who underwent hysterectomy. The oxidative burst was measured at the same time intervals, and additionally, 1, 5 and 10 min after the start of surgery. Similarly, the leukocyte and differential counts were measured at the same time intervals as the adhesion/activation molecules and before declamping in both groups.

Measurement of oxidative burst activity was performed on heparinized venous blood samples by means of flow cytometry ^[11]. Immediately after its withdrawal, 100 μL of blood was added to 3 mL phosphate-buffered saline (PBS) containing 100 mM phorbol 12-myristate 13-acetate (PMA) and 123-dihydrorhodamine (DHR) (Molecular Probes Inc., Eugene, OR, USA), giving a final concentration of 2 μM, in Nunc Minisorp tubes (Life Technologies, Roskilde, Denmark). After 3 min, the reaction was stopped by transfer of the samples to ice. Leukocytes were isolated by ammonium chloride lysis of erythrocytes. Formation of the green fluorescent rhodamine (the oxidation product of 123-DHR) was assessed using a FACScan™ flow-cytometer (Becton-Dickinson, Copenhagen, Denmark) with a filter wavelength of 513 nm. Granulocytes and monocytes were identified on the basis of their forward and side light scatter characteristics. Calculation of mean fluorescence values was performed using the Becton-Dickinson Lysis™ II software program. The granulocyte and monocyte respiratory burst activity of each sample was measured as the mean fluorescence intensity, and expressed as a percentage of the intensity of a reference-sample obtained prior to surgery.

Adhesion/activation molecules on granulocytes and monocytes were also measured using flow cytometry and monoclonal antibodies. Heparinized peripheral blood (100 μL) was incubated with 20 μL of monoclonal antibodies for 15 min in the dark at room temperature. The following antibodies were used: anti-CD11a-FITC (MαH, IgG₁) and anti-CD11c-FITC (MαH, IgG₁) (DAKO, Glostrup, Denmark); and anti-CD44-FITC (MαH, IgG₁) and anti-CD69-FITC (MαH, IgG₁, 1:20) (Becton Dickinson, San Jose, CA USA). Red blood cells were lysed using ammonium chloride. Cells were then washed twice with phosphate-buffered saline and fixed in 1% formalin in PBS before being analysed by flow cytometry (Coulter Elite, Luton, UK). The granulocytes were indentified on the basis of their forward and side scatter characteristics. Cells positive for the specific antigens were scored, based on Simultest control (MIgG₁-FITC; Becton Dickinson).

The total leukocyte count was determined by a Coulter Counter S (Coulter Electronics, Luton, UK) and the differential count was performed automatically (Hematrack, Geometric Data, Model 360, Luton, UK).

Statistical analyses included Friedman's one- way analysis of variance (ANOVA) and Wilcoxon test for the differences within the group compared with pre-operative values. (Values were expressed as mean ± SE.) P<0.05 was considered significant.

Results

The patients scheduled for open-heart surgery were older and weighed more than the gynaecological patients. The mean age of the cardiac patients was 64±3 years, while the mean age of the patients scheduled for hysterectomy was 46±2 years (P<0.05). The mean weights of the patients were 79±6 kg and 71±4 kg, respectively (P<0.05). During open-heart surgery, the CPB time was 87±6 min, whereas the aortic clamping time was 49±6 min. The duration of surgery was 189±11 min for open-heart surgery and 99±10 min for hysterectomy (P<0.05). Five men and three women were included in the open-heart surgery group. However, there were no differences in the measured parameters between men and women within the group.

Measurement of cell numbers and oxidative burst of granulocytes

There was a significant decrease in the number of granulocytes in peripheral blood following aortic clamping and CPB (Fig. 1). The present authors observed a significant increase in cell numbers in peripheral blood following CPB as well as after hysterectomy (P<0.05) (Fig. 1).

Fig. 1: Changes in the number of granulocytes in peripheral blood and the granulocyte oxidative burst in patients undergoing open-heart surgery (○―○) or hysterectomy (□―□); (D-D) values adjusted for haemodilution during CPB (mean ± SE). Asterisks (*) indicate a statistically significant change within the group. Closed symbols indicate statistically significant changes from pre-operative control values.

The oxidative burst reaction to PMA stimulation decreased after aortic clamping to 39±10% of pre-operative values (P<0.05). A further decrease was observed after aortic declamping and reperfusion (15±4% of initial values; P<0.05). A degree of normalization occurred 2-3 h after aortic declamping (44±12% and 29±15%, respectively; P<0.05) (Fig. 1), occurring simultaneously with peripheral blood granulocytosis. The oxidative burst did not decrease during abdominal surgery. There was a tendency for a reduced oxidative burst following the operation decreasing to 41±14% of pre-operative values (P<0.05) 3 h post-operatively. In half of the patients in the CPB group and in three out of eight patients in the hysterectomy group, a bimodal distribution of the positive and negative subpopulations was observed among PMA-stimulated granulocytes.

Measurement of cell numbers and oxidative burst of monocytes

The number of monocytes in peripheral blood decreased from 29±7 × 10⁷ L⁻¹ before surgery to 18±5 × 10⁷ L⁻¹ before aortic clamping (NS). Ten minutes after aortic clamping and just before declamping, the monocytes decreased significantly to 12±3 × 10⁷ L⁻¹ and 5±2 × 10⁷ L⁻¹, respectively (P<0.05), then normalized 2-3 h after declamping (Fig. 3). The number of monocytes decreased from 43±7 × 10⁷ L⁻¹ pre-operatively to 33±1 × 10⁷ L⁻¹ at the end of hysterectomy (NS), then rose to 63±8 × 10⁷ L⁻¹ 3 h after hysterectomy (NS). (Fig. 2).

Fig. 3: Changes in CD11a/CD18-, CD11c/CD18- and CD44-positive granulocytes in open-heart surgery (○―○) and hysterectomy (□―□) (mean ± SE).

Fig. 2: Changes in the number of monocytes in peripheral blood and the monocyte oxidative burst in open-heart surgery (○―○) and hysterectomy (□―□); (D-D) values adjusted for haemodilution during CPB (mean ± SE). Asterisks (*) indicate a statistically significant change within the group. Closed symbols indicate statistically significant changes from pre-operative control values.

The monocyte oxidative burst reaction to PMA stimulation decreased from 86±15% at the start of CPB to 42±5% of pre-operative values 20 min after aortic clamping. There was a further decrease to 27±5% 20 min after declamping (P<0.05), continuing several hours post-operatively to 10±2% 3 h after declamping (P<0.05) (Fig. 2). The oxidative burst reaction of monocytes to PMA stimulation decreased to 83±19% (P<0.05) at the end of hysterectomy, reaching as little as 30±6% 3 h post-operatively.

Expression of granulocyte adhesion molecules

The percentage of granulocytes expressing CD11a/CD18 adhesion molecules on their surface decreased from 30±9% to 24±7% 10 min after aortic clamping. Following CPB, the granulocytes increased to 54±12% and 75±8% 2 and 3 h after aortic declamping (Fig. 3). In the hysterectomy group, CD11a/CD18-positive granulocytes increased from 11±10% pre-operatively to 65±12% 3 h post-operatively (Fig. 3). There was no significant difference within the two groups using Friedman's one-way ANOVA.

The percentage of granulocytes expressing CD11c/CD18 adhesion molecules on their surface also decreased during CPB, but was normalized 3 h after aortic declamping. There were 19±4% cells positive for CD11c/CD18 pre-operatively, decreasing to 5±1% 10 min after aortic declamping, then rising to 15±3% 3 h post aortic declamping (NS). The CD11c/CD18-positive granulocytes increased from 4±0.8% pre-operatively to 8±5% (NS) 3 h following hysterectomy (Fig. 3).

There was no significant change in the percentage of CD44-positive cells, which increased from 5±0.8% pre-operatively to 13±5% 3 h post declamping (Fig. 3).

Discussion

The observed decrease in the number of granulocytes in peripheral blood during the CPB followed by granulocytosis is in agreement with other reports ^[9,10]. The decrease in the number of granulocytes during CPB may be the effect of haemodilution ^[12]. The granulocytosis in peripheral blood following intraabdominal surgery has also been observed in other studies ^[3].

The cardiac patients were older and the procedures lasted longer than in the hysterectomy group. Prolonged operative time induced more pronounced surgical stress, while cell mediated immunity has a more depressed stress response in older patients ^[13]. Therefore, the overall effect of age and duration of operation is probably small in the present study.

Granulocyte infiltration into the myocardium and lungs following CPB plays an important role in ischaemic myocardial damage and reperfusion injury, and may contribute to endothelial damage in the lungs manifested by increased capillary permeability. Bando et al.^[14] observed a better preserved pulmonary function following CPB in dogs when blood leukocytes were removed by a leukocyte filter. In the study by Wilson et al.^[15], it was demonstrated that mechanical leukocyte filtration decreased sequestration in the coronary vascular bed, decreased myocardial creatinine kinase release and improved left ventricular systolic function in neonatal pigs. In humans, CPB induced the release of the protease enzymes lactoferrin, elastase and α-proteinase from granulocytes ^[16]. Furthermore, Stahl et al.^[17] demonstrated that the level of elastase correlated positively with the duration of CPB. Cavarocchi et al.^[18] found that CPB was associated with an increase in H₂O₂ in plasma. Using the H₂O₂-sensitive probe 123-dihydrorhodamine ^[11], the present authors recorded a decrease in granulocyte H₂O₂ production in response to a PMA stimulus. The decrease observed during CPB was even more pronounced following declamping and reperfusion. The present authors suggest that the decreased oxidative burst reaction to PMA stimulation in vitro is the result of the refractory response of the granulocytes and monocytes subsequent to per-operative stimulation. This is in accordance with the findings of Cavarocci et al.^[18]. In contrast, the present authors observed only a non-significant slight decrease in oxidative burst during the intra-abdominal operation.

The partial normalization of the oxidative burst reaction 2 h after declamping may be explained by the influx of newly produced granulocytes from the bone marrow. A decreased oxidative burst reaction to an in vitro PMA stimulus was found 3 h after abdominal and open-heart surgery. This is probably the result of activation of the granulocytes in vivo or an efflux of the most active cells to other tissues or to the area of the operation ^[3,12]. The present results are in accordance with other investigations showing activation and dysfunction of granulocytes in peripheral blood following abdominal surgery as well as open-heart surgery with CPB ^[19,20].

The bimodal distribution with positive and negative subpopulations among PMA-stimulated granulocytes which was occasionally observed in the present study agrees with earlier reports of heterogeneity in oxidative product formation among granulocytes in patients with infection ^[21].

In the present study, the authors observed a decrease in number of monocytes in peripheral blood during abdominal and cardiac surgery. There was a further decrease immediately after CPB was initiated, which reached a very low level at the end of CPB. The decrease in the number of monocytes in peripheral blood during CPB has not been described before to the present authors' knowledge and it cannot be explained by haemodilution. The present authors' observation of monocytosis post-hysterectomy (Fig. 2) is in keeping with the study by Grzelak et al.^[22], who demonstrated a 12.4% increase in the number of monocytes 24 h post-hysterectomy.

The total capacity of the mononuclear phagocyte system is depressed immediately after CPB and major uncomplicated surgery ^[9,23]. Contrary to this, monocytes in peripheral blood are active. Neoptolemos et al.^[24] observed an increase in the migration, phagocytosis and chemiluminescence of monocytes 24 h after abdominal surgery. The monocytes were activated during abdominal surgery and several hours post-operatively, producing free oxygen radicals. This is in accordance with the observed refractory response of monocytes in this study with respect to H₂O₂ production.

The present authors observed a pronounced decrease in the oxidative burst reaction of monocytes to stimulation in vitro immediately after the start of CPB. In the present study, CPB is likely to induce activation of monocytes and produce oxygen free radicals in these cells in vivo. This is in accordance with recent studies by Stefano and Bilfinger ^[25,19], who showed that monocytes were activated during and after CPB. Although the monocytes showed an increased chemokinesis, chemotaxis was reduced, suggesting that these were dysfunctional despite the activated state.

It is often suggested that CPB may induce a systemic inflammatory reaction ^[5,6,9] which includes activation of the complement and coagulation systems, degranulation of granulocytes and the synthesis of various cytokines from monocytes. However, several of these events are also observed to a lesser extent several hours after major surgery.

Two families of adhesion molecules are necessary for the adhesion and migration of leukocytes into tissues, namely the integrin and the selectin family. The CD11a-c/CD18 adhesion molecules are the most important integrins and are largely responsible for adhesion strengthening and migration. As demonstrated in the study by Kappelmayer ^[26], CPB has been shown to increase the expression by these molecules, where upregulation of CD11b/CD18 on the surface of granulocytes was observed in human blood exposed to ECC. It has also been shown recently that monoclonal antibodies against CD11a-c/CD18 molecules protect animals partially against reperfusion injury following myocardial infarction.

A third family, the homing-associated cell adhesion molecules (HCAM), probably also plays a role in cell adhesion. The most important member of this family is CD44, a hyaluronate receptor. The present authors studied the effect of ECC and surgery in vivo on CD11a/CD18, CD11c/CD18 and CD44.

Several hours after open-heart surgery with CPB and abdominal surgery, the present authors observed increases in granulocytes expressing CD11a/CD18 and CD44. This can be the result of an upregulation of these molecules, as observed with CD11b/CD18 in vitro by Kappelmayer et al.^[26].

Changing anaesthetic practice will probably not reduce the activation of leukocytes during CPB. However, the increased activation of leukocytes during CPB demonstrated in the present study underlines the need to develop artificial surfaces with improved biocompatibility.

In conclusion, the present authors suggest that open-heart surgery with CPB induces the activation of both granulocytes and monocytes as measured by their oxidative burst activity. This effect was not seen during, but was seen several hours after abdominal surgery. Thus, open-heart surgery with CPB was associated with a fast and pronounced activation of leukocytes which may play a role in reperfusion injury.

Acknowledgment

The authors wish to thank Dr Michelle Chew for reviewing this manuscript.