European Journal of Cardio-thoracic Surgery 32 (2007) 611—616 www.elsevier.com/locate/ejcts

Endothelial cell dysfunction after coronary artery bypass grafting with extracorporeal circulation in patients with type 2 diabetes mellitus Karla Lehle 1,*, Ju ¨rgen G. Preuner 1, Anja Vogt, Leopold Rupprecht, Andreas Keyser, Reinhard Kobuch, Christof Schmid, Dietrich E. Birnbaum Department of Cardiothoracic Surgery, University Hospital Regensburg, 93042 Regensburg, Germany Received 30 April 2007; received in revised form 21 June 2007; accepted 22 June 2007; Available online 27 July 2007

Abstract Objective: Type 2 diabetes mellitus is a well-known risk factor in patients with severe coronary artery disease undergoing coronary artery bypass grafting (CABG). The aim of the study was to analyze the endothelial dysfunction in these patients by evaluating postoperative soluble inflammatory cytokines. Methods: Patients undergoing CABG without (n = 15, group A) and with (n = 14, group B) diabetes mellitus were analyzed for their release of E-selectin, interleukin-6 (IL-6), and tumor necrosis factor (TNF) up to 3 days postoperatively. A pharmacokinetic quantitative kinetic evaluation (Kinetica 2000) of maximum concentrations (cmax), time to reach cmax (tmax), area under the curve (AUC0—inf), and terminal elimination half time (t1/2) was performed using a non-compartmental model. Results: There was no difference in preoperative plasma concentrations of the cytokines and in the postoperative kinetic analyses of TNF when comparing both groups. However, the release of IL-6 was restricted with cmax of 1055  543 pg/ml for group B versus 2112  1532 pg/ml for group A ( p  0.05), paralleled by a decrease in the absolute amount (AUC0—inf) of IL-6. The t1/2 remained unaffected (13.9  6.6 h and 12.7  4.6 h, respectively). The AUC0—inf of E-selectin decreased by a factor of 1.7 ( p  0.05) with unchanged cmax but reduced t1/2 (12.9  10 h for group B vs 33.1  20.4 h for group A; p  0.01) referring to an augmented endothelial uptake and degradation of E-selectin. Conclusions: CABG with extracorporeal circulation could be used to verify a specific endothelial dysfunction in diabetic patients characterized by an impaired release of IL-6 and an increased turnover of E-selectin. # 2007 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. Keywords: Endothelial dysfunction; Diabetes mellitus; Coronary artery bypass grafting; Cytokines

1. Introduction

2. Patients and methods

Diabetes mellitus is a well-known risk factor in patients with severe coronary artery disease undergoing coronary artery bypass grafting (CABG) [1—3]. The reason for this heightened risk is unclear. It has been assumed that the greater surgical risk associated with diabetes mellitus might be a consequence of an endothelial dysfunction and an altered inflammatory response to CPB, characterized by alteration in the secretion of cytokines (interleukin-6, tumor necrosis factor) or of adhesion molecules [4—9]. The aim of the present study was to analyze endothelial dysfunction and inflammatory response after CABG employing cardiopulmonary bypass (CPB) in patients with coronary artery disease and type 2 diabetes mellitus during extracorporeal circulation (ECC).

2.1. Patients

* Corresponding author. Address: Department of Cardiothoracic Surgery, University Hospital Regensburg, Franz-Josef-Strauss-Allee 11, D-93053 Regensburg, Germany. Tel.: +49 941 9449901; fax: +49 941 9449902. E-mail address: [email protected] (K. Lehle). 1 Both authors contributed equally to this work.

Two hundred and seventy-four consecutive patients (100%) undergoing elective CABG with CPB were screened. A subgroup of n = 75 (27%) with cold crystalloid cardioplegic solution (Janostil, Fresenius) was selected and, according to the presence of diabetes type 2 (n = 14; 19%), included in the study. A control group (n = 61; 81%) was selected and matched for age (n = 15; 20%). The restriction of patients to one particular cardioplegic solution during ECC introduced an unbalance with respect to gender (Table 1). Informed consent was obtained from all patients. Demographic data and surgical parameters are shown in Table 1. Patients with diabetic type 2 were on oral anti-diabetic therapy for at least 1.5 years. In one patient, 7 months before CABG, a change from oral anti-diabetics to insulin administration was necessary. None of the patients suffered from a complicated diabetes (i.e. the presence of polyneuropathy, renal insufficiency, reduced vision, and/or leg and foot ulcers). Patients with type 1 diabetes, as well as those with diffuse arteriosclerosis, preoperative signs of infection (white cell

1010-7940/$ — see front matter # 2007 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.ejcts.2007.06.027

612

K. Lehle et al. / European Journal of Cardio-thoracic Surgery 32 (2007) 611—616

Table 1 Demographic data and surgical parameters Control group (n = 15)

Diabetic group (n = 14)

p-Value

Male/female ratio Age (years) Weight (kg) Height (cm) Body surface (m2) Body mass index (kg/m2) Ejection fraction (%) No. of vessels bypassed ACVB IMA

13/2 59.4  4.5 76.3  8.0 171  10 1.9  0.2 26.7  2.9 69  15 4.0  1.1 3.2  1.2 0.9  0.7

5/9 60.0  3.8 80.6  11.6 167  8 1.9  0.2 28.9  3.1 63  10 3.6  1.1 2.7  0.9 1.1  0.6

0.71 0.33 0.45 0.51 0.053 0.22 0.31 0.36 0.74

Duration (min) CPB Ischemia Reperfusion

96.5  27.3 61.4  22.7 31.1  7.0

88.0  25.5 55.3  17.9 27.3  8.9

0.40 0.43 0.21

Minimal temperature (8C) ICU (median 75%) (days) Hospital stay (days)

31.9  0.9 0.95 (1.1) 9.9  4.3

32.1  0.9 1.4 (1.9) 8.6  1.4

0.56 0.71 0.76

The stay in intensive care unit (ICU) is given as median (75% confidence intervals). All other data are mean  standard deviation. ACVB, aortocoronary venous bypass; IMA, internal mammary artery; CPB, cardiopulmonary bypass.

count 8000/m3), and an ejection fraction 40% were excluded from the study. Additional risk factors including hypertension, hyperlipidemia, hyperuricemia, hyperthyroidism, chronic bronchitis, asthma, and nicotine abuse were equally spread in both patient groups. Biochemical characteristics of both groups are shown in Table 2. Table 2 Biochemical characteristics Control group (n = 15)

Diabetic group (n = 14)

p-Value

5.7  0.7 4.4  0.9 10.7  4.5 0.3  0.24 188  65 1.8  0.6 5.8  0.8 3.6  0.6 1.3  0.4 81.5  14.9 5.9  1.7

8.0  1.7 7.0  1.0 14.8  3.8 1.6  1.5 228  110 2.0  0.6 5.0  1.2 3.1  1.1 1.1  0.2 76.6  20.9 6.7  2.6

0.001 0.001 0.001 0.003 0.24 0.43 0.13 0.20 0.12 0.45 0.34

CRP (mg/l) Preop Mean (0—7th day po)

4.8  5.3 151  63

4.6  2.0 160  72

0.29 0.09

Leucocytes (1/nl) Preop Mean (0—7th day po) Maximum po

7.7  2.0 11.1  2.8 14.5  3.9

7.0  2.2 9.8  3.0 12.6  4.5

0.47 0.22 0.21

CK (U/l) Preop Mean (0—7th day po)

33.4  12.3 156  126

44.1  29.3 158  104

0.30 0.66

CK-MB (U/l) Mean (0—7th day po)

14.2  10.9

8.3  4.1

0.07

HbA1c (%) FPG (mmol/l) preop cmax (glucose) (mmol/l) po tmax (glucose) (days) po AUC0—3 days [(mg/dl)/day] po Total triglyceride (mmol/l) Total cholesterol (mmol/l) LDL-cholesterol (mmol/l) HDL-cholesterol (mmol/l) Serum creatinine (mcmol/l) Urea (mmol/l)

Mean  standard deviation. HbA1c, glycosylated hemoglobin; FPG, fasting plasma glucose; po, postoperatively; preop, preoperatively; CK, creatine kinase; CK-MB, MB isoenzyme of CK; CRP, C-reactive protein; AUC0—3 days, area under the curve (time course from day 0 to 3 po); cmax, peak plasma concentrations; tmax, time to reach cmax.

Anesthesia was comparable for all patients and consisted of intravenous etomidate and fentanyl application. Pancuronium was used for neuromuscular blockage. None of the patients received aprotinin. Heparin (150 IU kg 1) was administered before cannulation of the aorta and right atrium. A Sto ¨ckertW W heart-lung machine with a Maxima Plus CB3380 hollow fiber oxygenator (Medtronic, Anaheim, CA, USA) was used for nonpulsatile extracorporeal circulation. The latter was primed with 1.3 l electrolyte solution (Janosteril, Fresenius) and 250 ml mannit solution (20%). Cold crystalloid cardioplegia was applied for cardiac preservation. Residual blood remaining in the circuit after CPB was salvaged with a cell saver and retransfused. Transfusion of red blood cells was necessary in six patients in group A (450  250 ml) and in three patients in group B (567  57 ml). 2.2. Serial analyses Serial blood samples were withdrawn from the radial artery catheter or peripheral line before induction of anesthesia, at the end of ECC, on administration to the intensive care unit, as well as 6, 12, 18, 24, 48, and 72 h after termination of CPB. Blood samples were collected into sterile vacuum tubes with ethylenediaminetetraacidic acid and in serum-monovettes. Samples of plasma and serum were frozen for analysis of TNF, IL-6, and E-selectin using enzymelinked immunosorbent assay (Beckman Coulter, Hamburg; R&D, Wiesbaden). Plasma albumin was measured with a calorimetric assay (bromcresol green method, MPR3, Boehringer Mannheim, Mannheim), while hemoglobin content was analyzed with an automated blood gas analysis system (ABL510, Radiometer, Copenhagen). For all parameters, a volume correction was made to eliminate hemodilution effects according to Diago et al. [10]. The correction factor was calculated as the ratio of c/c(ti), with c being the concentration of hemoglobin (or albumin) preoperatively and c(ti) being the concentration at the time of sampling. In 13 healthy volunteers (age 24—35 years) TNF, IL-6, and Eselectin were measured as a reference. After weaning from ECC, the hemoglobin level dropped to about 40% and remained low throughout the 3-day observation period ( p  0.001) in both patient groups (Fig. 1).

Fig. 1. Hemoglobin blood levels of non-diabetic (filled circles) and diabetic (open circles) patients undergoing CPB. Plasma samples were obtained from peripheral artery before, at the end of CPB (t = 0), and after CPB (1—120 h). Hemoglobin decreased significantly ( p  0.001) at the beginning of the operation. Results are expressed as median value with 25% percentile (negative error bars) and 75% percentile (positive error bars).

K. Lehle et al. / European Journal of Cardio-thoracic Surgery 32 (2007) 611—616

Similar observations were made for plasma albumin concentrations. During CPB, plasma albumin decreased significantly (20.4  3.6 mg/ml vs 43.3  3.5 mg/ml; p  0.001), and only began to normalize about 3 days after surgery (35.9  1.4 mg/ml). 2.3. Kinetic analysis For quantitative kinetic analysis of the release of IL-6, TNF, and soluble E-selectin (sE-selectin) maximum concentrations (cmax), time to reach cmax (tmax), area under the curve (AUC), and terminal elimination half time (t1/2) were assessed using a non-compartmental model (trapezoidal rule) (Kinetica 2000, version 3.0) [11]. The parameter levels obtained before CPB were set at zero to be able to correlate the absolute amount of IL-6, TNF, and soluble E-selectin with the AUC. Fasting plasma glucose concentrations (FPG) were assessed to determine the extent of the diabetic disease [12] (Table 2). Blood samples were drawn immediately prior to starting CPB. Incremental plasma glucose excursion was defined as the area under the 3-day glucose excursion curve (AUC0—3 days) (Table 2). It represents the time-weighted excursion intensity and its influence of both the duration of the excursion and the overall glucose levels during the excursion. The calculation did exclude negative areas (areas below baseline), as well as FPG and 100 mg/dl (5.6 mmol/l) for patients without and with diabetes mellitus, respectively. The incremental AUC0—3 days values were compared between patients for maximal peak glucose (cmax) and time to reach cmax (tmax). 2.4. Statistical analysis Statistical analysis was accomplished using SigmaStat2.0 (Jandel Scientific, Erkrath) software. Results were expressed as mean  standard deviation (SD). The Mann—Whitney rank sum test was used to compare two consecutive data points. Differences were considered significant for p  0.05.

3. Results 3.1. Clinical findings Demographic data and operative details were similar between the groups as shown in Table 1. There was no significant difference in blood loss, blood product transfusion requirement, postoperative myocardial infarction or stroke, as well as in the new onset of atrial fibrillation. No reoperation for bleeding or early bypass occlusion and no prolonged postoperative intensive care were necessary. The diabetic patients of group A demonstrated a significantly increased fasting plasma glucose level of 7.0 mmol/l as compared to 4.4 mmol/l for the non-diabetic control patients. Mean HbA1c level in the diabetic patients was 8% (Table 2). Initiation of postoperative anti-diabetic therapy with insulin was triggered by a plasma glucose level of 200 mg/dl (11.1 mmol/l). The actual plasma glucose level determined after insulin administration (2.0  1.8 days post-op) amounted to 250  21 mg/dl (13.9  1.2 mmol/l). In all

613

Table 3 Postoperative anti-diabetic pharmacotherapy

Without anti-diabetic therapy Insulin therapya, b Glibenclamide b Metformin b Acarbose b

Group

Patients

Control Diabetic Control Diabetic Diabetic Diabetic Diabetic

14/15 6/14 1/15 6/14 2/14 3/14 2/14

Days on therapy

Dose per day

0.5 1.7  1.5 2.1 and 2.5 3.3  3.1 1.1 and 3.5

10 IU (i.m.) 90  55 IU Both: 7 mg 1292  1380 mg Both: 100 mg

Data are given as mean  SD. a Continuous infusion, intravenous bolus, and intramuscular (i.m.) application of insulin were triggered by a plasma glucose level of 200 mg/dl (11.1 mmol/l). b Data included single therapy and a combination therapy of different oral anti-diabetics or of insulin and oral anti-diabetics.

control patients plasma glucose levels increased with the beginning of CPB. Therefore, in one patient of the control group an intramuscular (i.m.) application of insulin was administered (Table 3). Oral anti-diabetic therapy of the patients was restarted 2—3 days post-op after leaving intensive care unit (Table 3). 3.2. Soluble mediators before CPB There was no significant difference in the preoperative serum concentration of TNF, soluble E-selectin, and IL-6 between the two patient groups (Fig. 2). Healthy volunteers showed comparable concentrations for TNF and IL-6, whereas the concentrations for E-selectin of volunteers were significantly decreased (Fig. 2). 3.3. Soluble mediators peri- and postoperatively In both groups, all studied mediators increased significantly during cardiopulmonary bypass grafting ( p  0.001 at cmax; Fig. 3). TNF reached its peak value 4.1  4.1 h and 4.7  3.7 h for groups A and B ( p = 0.599), respectively, after termination of CPB (tmax). The maximum concentrations (cmax) noted were 219  312 pg/ml and 217  275 pg/ml ( p = 1.0), respectively. Already 24 h after termination of CPB, TNF concentration dropped to baseline (Fig. 3). The absolute amount (AUC) of TNF did not differ between the groups ( p = 0.729). IL-6 levels increased also during and after surgery in both groups, with a significantly higher peak (cmax) in group B as compared to group A. Peak values were reached at 9.2  5.5

Fig. 2. TNF, IL-6, and soluble E-selectin blood levels of healthy volunteers (gray bars), non-diabetic (black bars), and diabetic (open bars) patients before CPB. Results are expressed as mean with standard deviation.

614

K. Lehle et al. / European Journal of Cardio-thoracic Surgery 32 (2007) 611—616

Fig. 5. Kinetic analysis of soluble E-selectin in non-diabetic (filled bars) and diabetic (open bars) patients undergoing CPB. Serum samples were obtained from peripheral artery before, at the end of CPB (t = 0), and after CPB (1— 72 h). A quantitative analysis of E-selectin as a response to the CPB and/or ECC was done with Kinetica 2000. AUC, area under curve; cmax, maximum peak concentration; tmax, the time at cmax; and t1/2, terminal elimination half time. Results are expressed as mean with standard deviation.

and 8.8  2.9 h (groups A and B; p = 0.982) following termination of CPB. The underlying pharmacological noncompartmental model showed no difference in the elimination half time (t1/2), but a 30% decrease of both the absolute amount (AUC) and the maximal release of IL-6 in the diabetic patient group (Figs. 3 and 4). The values remained increased up to 3 days after termination of CPB ( p  0.001). Serum concentrations of soluble E-selectin peaked 10.9  4.7 and 11.3  3.9 h (group A and group B; p = 0.645) (Fig. 3). The absolute amount (AUC0—inf) of soluble E-selectin in the 3-day follow-up was significantly reduced in the diabetic group ( p = 0.02) accompanied by a reduction of the terminal elimination half time of about 60% ( p = 0.001) (Fig. 5).

4. Discussion Fig. 3. TNF, IL-6, and soluble sE-selectin blood levels of non-diabetic (filled circles) and diabetic (open circles) patients undergoing CPB preoperatively, at the end of CPB (t = 0) and 1—72 h postoperatively with correction for hemodilution. Results are expressed as median value with 25% percentile (negative error bars) and 75% percentile (positive error bars).

Fig. 4. Kinetic analysis of IL-6 in non-diabetic (filled bars) and diabetic (open bars) patients undergoing CPB. Plasma samples were obtained from peripheral artery before, at the end of CPB, and after CPB. A quantitative analysis of IL-6 as a response to the CPB and/or extracorporeal circulation was done with Kinetica 2000. AUC, area under curve; cmax, maximum peak concentration; tmax, the time at cmax; and t1/2, terminal elimination half time. Results are expressed as mean with standard deviation.

The inflammatory process, i.e. the release of cytokines and soluble adhesion molecules during and after CPB, is well documented in clinical studies [13,14]. However, diabetic patients were mostly excluded in these investigations for an assumed endothelial dysfunction. Here, we analyzed for the first time the altered endothelial inflammatory response in diabetic patients following CPB. Our data demonstrated not only an increased production of cytokines (IL-6, TNF) and of soluble adhesion molecules (E-selectin) during and after CPB but also a qualitatively different inflammatory reaction in patients with type 2 diabetes mellitus. CPB induced an increase in plasma TNF in patients with and without diabetes that lasted for the duration of the study. Peak levels for TNF were noted at 2 h after surgery, i.e. prior to the maximum concentrations of IL-6 and E-selectin. These results are in line with other studies [15]. The effect of TNF to implicate in the pathogenesis of myocardial ischemia and reperfusion injury [16,17] was independent of the diabetic disease. Due to the wide interindividual variability, postoperative serum TNF levels alone do not allow us to draw exact conclusions with regard to endothelial function. The IL-6 levels peaked 4 h postoperatively and decreased over the next 48 h in both study groups, which is evidence for a strong pro-inflammatory response. Even 3 days after the

K. Lehle et al. / European Journal of Cardio-thoracic Surgery 32 (2007) 611—616

termination of CPB, IL-6 concentrations remained elevated in both groups compared to baseline levels. Steinberg et al. had reported that IL-6 levels increased after protamine administration reached a maximum at 3 h after bypass and remained above the baseline levels [18]. Similar time— response curves have been reported by Franke et al. [19]. The kinetic analysis over a 3-day period in our patient cohort proved the impaired release of IL-6 in diabetic patients, and also demonstrated that the elimination of IL-6 was not affected. Apart from IL-6, the release of IL-8 and TNF is impaired in patients with diabetes undergoing elective CPB as well [20]. The pathophysiology of this inflammatory response in diabetic patients is still unclear [20]. An increased oxidative stress and complement activation in diabetic patients was discussed by Matata and Galinanes [20]. Differences in gene expression profiles of diabetic and nondiabetic patients undergoing CPB and cardioplegic arrest were discussed [21]. Therefore, an altered regulation of the genes of the various inflammatory cytokines and their signal transduction pathways is also likely in diabetic patients. This is an area that clearly requires further investigation. E-selectin is specifically localized on endothelial cells and hence a more sensitive marker of endothelial activation than other adhesion molecules that have a wider tissue distribution [22]. Soluble isoforms of E-selectin have been reported to be elevated in a number of pathological conditions including diabetes [8,9]. Our data confirm these observations as we also noted increased levels of soluble E-selectin before CPB. Interestingly, no differences in E-selectin levels were found between diabetic and non-diabetic patients with coronary artery disease. As the accompanying risk factors (e.g. nicotine abuse, hypertension, etc.) were evenly distributed between both patient groups, one may only speculate why E-selectin levels were increased prior to surgery. The most interesting finding was that the inflammatory response after CPB with regard to E-selectin differed between the groups. In diabetic patients, the contact of blood with the artificial surfaces of the heart-lung machine resulted in a decreased release of E-selectin into the plasma compartment. While the time to reach maximum concentration was comparable with other studies [20], we found a faster clearance of soluble E-selectin levels in diabetic patients. In contrast, Matata and Galinanes described a prolonged resting time of E-selectin up to 48 h after CPB in patients with diabetes mellitus [20]. While we calculated the terminal elimination half time over a time period of 3 days following surgery, Matata and Galinanes observed unchanged elevated peak values up to 2 days postoperatively only [20]. A reduced plasma level of E-selectin is not indicative of parallel changes in the density and expression of the molecule on the endothelial surface, which was not different in aorta, and internal mammary artery of diabetic and non-diabetic patients [23]. Instead, it represents a protective mechanism that allows the clearance of these molecules from the cell surfaces, limiting leukocyte attachment and cell injury [24]. We speculate that CPB initiates an increase of the shedding of E-selectin and an augmented endothelial uptake and degradation of this molecule in diabetic patients. A clear estimation of the degree of the endothelial dysfunction after CPB in the described in vivo system is difficult since the underlying pathophysiological mechanisms

615

are multifactorial. Microarray analysis of tissue samples from diabetic and non-diabetic patients undergoing CPB showed different gene expression profiles: 28 genes showed an upregulation in the diabetic group exclusively (including inflammatory/transcription activators) after CPB [21]. However, the diabetic endothelial dysfunction could not be exactly localized. In this context, the pro-inflammatory basic state of the ‘diabetic endothelial cell’ was approved in vitro in a disease-specific endothelial cell culture system derived from patients with diabetes mellitus [25]. We speculate that under in vivo conditions after CPB the ‘diabetic endothelial cells’ showed a reduced metabolic activity. In both patient groups, elevated overall glucose levels were observed during the early postoperative period. As the increased glucose levels were similar, the incremental plasma glucose excursions (AUC0—3 days), which would influence the kinetics of soluble mediators, could be excluded as an objectionable parameter. The presence of insulin and/or oral anti-diabetics had no influence on the kinetic profile of TNF, IL-6, and E-selectin (data not shown). Hemodilution effects issue from the need of priming solution for ECC. So far, only a few studies corrected the concentrations of soluble inflammatory markers with regard to hemodilution [10,15]. Our data demonstrate a significant reduction of parameter levels at the end of CPB after quantification of the dilution factor using hemoglobin or albumin concentration in the respective plasma sample. A future perspective might be the application of minimal extracorporeal circulation for diabetic patients to reduce the maximum stimulating effects after CPB. In summary, CAD patients with diabetes mellitus demonstrate a specific endothelial dysfunction during CABG with extracorporeal circulation characterized by an impaired release of IL-6 and an increased turnover of E-selectin.

Acknowledgement The authors gratefully acknowledge the excellent technical assistance of K. Bielenberg.

References [1] Pintar T, Collard CD. The systemic inflammatory response to cardiopulmonary bypass. Anesthesiol Clin North America 2003;21(3):453—64. [2] Cohen Y, Raz I, Merin G, Mozes B. Comparison of factors associated with 30-day mortality after coronary artery bypass grafting in patients with versus without diabetes mellitus. Israeli Coronary Artery Bypass (ISCAB) Study Consortium. Am J Cardiol 1998;81(1):7—11. [3] Bucerius J, Gummert JF, Walther T, Doll N, Barten MJ, Falk V, Mohr FW. Diabetes in patients undergoing coronary artery bypass grafting. Impact on perioperative outcome. Z Kardiol 2005;94(9):575—82. [4] Calles-Escandon J, Cipolla M. Diabetes and endothelial dysfunction: a clinical perspective. Endocr Rev 2001;22(1):36—52. [5] Pickup JC, Mattock MB, Chusney GD, Burt D. NIDDM as a disease of the innate immune system: association of acute-phase reactants and interleukin-6 with metabolic syndrome X. Diabetologia 1997;40(11):1286—92. [6] Eppihimer JM. The role of leukocyte-endothelial cell adhesion in cardiovascular disease. Pathophysiology 1998;5(3):167—84. [7] McCarty MF. Interleukin-6 as a central mediator of cardiovascular risk associated with chronic inflammation, smoking, diabetes, and visceral obesity: down-regulation with essential fatty acids, ethanol and pentoxifylline. Med Hypotheses 1999;52(5):465—77.

616

K. Lehle et al. / European Journal of Cardio-thoracic Surgery 32 (2007) 611—616

[8] Albertini JP, Valensi P, Lormeau B, Aurousseau MH, Ferriere F, Attali JR, Gattegno L. Elevated concentrations of soluble E-selectin and vascular cell adhesion molecule-1 in NIDDM. Effect of intensive insulin treatment. Diabetes Care 1998;21(6):1008—13. [9] Matsumoto K, Sera Y, Nakamura H, Ueki Y, Miyake S. Serum concentrations of soluble adhesion molecules are related to degree of hyperglycemia and insulin resistance in patients with type 2 diabetes mellitus. Diabetes Res Clin Pract 2002;55(2):131—8. [10] Diago MC, Garcia-Unzueta MT, Marcano G, Merino J, Salas E, Amado JA. Serum soluble selectins in patients undergoing cardiopulmonary bypass. Relationship with circulating blood cells and inflammation-related cytokines. Acta Anaesthesiol Scand 1997;41(6):725—30. [11] Matthews JN, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. BMJ 1990;300:230—5. [12] American Diabetes Association (ADA): standard of medical care in diabetes — 2006. Diabetes Care 2006;29(Suppl. 1):S4—42. [13] Siminiak T, Smielecki J, Dye JF, Balinski M, El-Gendi H, Wysocki H, Sheridan DJ. Increased release of the soluble form of the adhesion molecules L-selectin and ICAM-1 but not E-selectin during attacks of angina pectoris. Heart Vessels 1998;13(4):189—94. [14] Kalawski R, Majewski M, Kaszkowiak E, Wysocki H, Siminiak T. Transcardiac release of soluble adhesion molecules during coronary artery bypass grafting: effects of crystalloid and blood cardioplegia. Chest 2003;123(5):1355—60. [15] Roth-Isigkeit A, Borstel TV, Seyfarth M, Schmucker P. Perioperative serum levels of tumour-necrosis-factor alpha (TNF-alpha), IL-1 beta, IL-6, IL-10 and soluble IL-2 receptor in patients undergoing cardiac surgery with cardiopulmonary bypass without and with correction for haemodilution. Clin Exp Immunol 1999;118(2):242—6. [16] Meldrum DR. Tumor necrosis factor in the heart. Am J Physiol 1998; 274(3Pt 2):R577—95.

[17] Cain BS, Meldrum DR, Dinarello CA, Meng X, Joo KS, Banerjee A, Harken AH. Tumor necrosis factor-alpha and interleukin-1beta synergistically depress human myocardial function. Crit Care Med 1999;27(7):1309—18. [18] Steinberg JB, Kapelanski DP, Olson JD, Weiler JM. Cytokine and complement levels in patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1993;106(6):1008—16. [19] Franke A, Lante W, Fackeldey V, Becker HP, Kurig E, Zoller LG, Weinhold C, Markewitz A. Pro-inflammatory cytokines after different kinds of cardiothoracic surgical procedures: is what we see what we know? Eur J Cardiothorac Surg 2005;28(4):569—75. [20] Matata BM, Galinanes M. Cardiopulmonary bypass exacerbates oxidative stress but does not increase proinflammatory cytokine release in patients with diabetes compared with patients without diabetes: regulatory effects of exogenous nitric oxide. J Thorac Cardiovasc Surg 2000;120 (1):1—11. [21] Voisine P, Ruel M, Khan TA, Bianchi C, Xu SH, Kohane I, Libermann TA, Otu H, Saltiel AR, Sellke FW. Differences in gene expression profiles of diabetic and nondiabetic patients undergoing cardiopulmonary bypass and cardioplegic arrest. Circulation 2004;110(11 Suppl. 1):II280—6. [22] Gearing AJ, Newman W. Circulating adhesion molecules in disease. Immunol Today 1993;14(10):506—12. [23] Ribau JC, Hadcock SJ, Teoh K, DeReske M, Richardson M. Endothelial adhesion molecule expression is enhanced in the aorta and internal mammary artery of diabetic patients. J Surg Res 1999;85(2):225—33. [24] Wyble CW, Hynes KL, Kuchibhotla J, Marcus BC, Hallahan D, Gewertz BL. TNF-alpha and IL-1 upregulate membrane-bound and soluble E-selectin through a common pathway. J Surg Res 1997;73(2):107—12. [25] Lehle K, Haubner F, Mu ¨nzel D, Birnbaum DE, Preuner JG. Development of a disease-specific model to evaluate the endothelial dysfunction in patients with diabetes mellitus. Biochem Biophys Res Commun 2007;357:308—13.