Plasma from patients undergoing coronary artery bypass graft surgery does not activate endothelial cells under shear stress in vitro.

BACKGROUND: Cardiac surgery with cardiopulmonary bypass (CPB) is commonly associated with acute kidney injury, and microvascular endothelial inflammation is a potential underlying mechanism. We hypothesized that pro-inflammatory components of plasma from patients who underwent coronary artery bypass graft surgery with CPB induce endothelial adhesion molecule expression when incorporating altered shear stress in the in vitro model.
METHODS: The clinical characteristics and markers of systemic inflammation and kidney injury were analyzed pre and postoperatively in 29 patients undergoing coronary artery bypass grafting with CPB. The effects of tumor necrosis factor (TNF)-α and patient plasma on the expression of endothelial inflammation and adhesion markers were analyzed in vitro.
RESULTS: Plasma TNF-α was elevated 6 h postoperation (median: 7.3 pg/ml (range: 2.5-94.8 pg/ml)). Neutrophil gelatinase-associated lipocalin in plasma peaked 6 h (99.8 ng/ml (52.6-359.1 ng/ml)) and in urine 24 h postoperation (1.6 ng/mg (0.2-6.4 ng/mg)). Urinary kidney injury molecule-1 concentration peaked 24 h postoperation (0.5 ng/mg (0.2-1.2 ng/mg). In vitro, the expression of E-selectin was induced by 20 pg/ml TNF-α. In addition, the expression of interleukin-8, intercellular adhesion molecule-1, and vascular cell adhesion molecule-1 was induced by 100 pg/ml TNF-α. Compared to healthy control plasma exposure, postoperative plasma did not increase the expression of markers of endothelial inflammation and adhesion under shear stress in vitro.
CONCLUSION: Patients undergoing CPB surgery showed mild systemic inflammation and kidney injury. However, the plasma components did not stimulate endothelial inflammation and adhesion molecule expression in vitro. Copyright:

INTRODUCTION

Cardiopulmonary bypass (CPB) allows for coronary artery bypass graft (CABG) surgery on the nonbeating heart. The use of CPB is associated with systemic inflammation, a prevailing response to major surgery.[12]During CPB, systemic inflammation is induced due to contact between blood components and the surface of the extracorporeal circuit, changes in shear stress, hemodilution, and ischemia of the heart and other organs.[3]Systemic inflammation is associated with postoperative organ failure, which results in reduced benefits for the patient after cardiac surgery.[2]

Postoperative organ failure can affect all organs and is associated with increased mortality, particularly when multiple organs are affected.[45]Following CABG surgery, acute kidney injury (AKI) has an incidence of up to 30%.[6]AKI is defined as an abrupt loss of kidney function measured by the changes in urine output and serum creatinine concentrations.[7]The integration of kidney injury markers, such as neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule (KIM)-1 into the risk stratification of patients has assisted in the early diagnosis of patients suspected of AKI following cardiac surgery.[8]Moreover, work by Lassnigg et al. and Haase et al. strongly suggested that even minor kidney injury, although not recognized by AKI classification criteria, is also linked to increased morbidity in cardiac surgery patients.[910]Therefore, it is important to identify mechanisms initiating and maintaining kidney injury and to develop preventive and therapeutic approaches for patients that develop AKI after undergoing CABG surgery with CPB.

Pro-inflammatory activation of the microvascular endothelium has been observed in AKI in the animal models of cardiac surgery using CPB and is likely involved in the initial adverse renal response to CPB in patients.[111]Endothelial pro-inflammatory activation involves the orchestrated expression of the adhesion molecules E-selectin, intercellular adhesion molecule (ICAM)-1, and vascular cell adhesion molecule (VCAM)-1. Leukocytes express complementary sialyl-lewisX, integrins, and other ligands and adhere to the activated endothelium. Consequently, leukocytes infiltrate the underlying tissue whereby inflammation is maintained and propagated. The activation of the endothelium can be triggered by alterations in shear stress as observed during shock and subsequent reperfusion.[12]Contrary to laminar flow, altered shear stress renders the endothelium susceptible to pro-inflammatory plasma components such as lipopolysaccharide and tumor necrosis factor (TNF)-α, which subsequently evoke an inflammatory response. Endothelial interactions with cells circulating in the blood, which are activated by the use of CPB, are another potential source for endothelial activation. However, in the perioperative context, it remains unclear which of these factors induce the pro-inflammatory response of endothelial cells in specific organ injury.[1314]

In the present study using existing patient material, we hypothesized that pro-inflammatory components of plasma from patients who underwent CABG surgery with CPB induce endothelial adhesion molecule expression when incorporating altered shear stress in the in vitro model.

METHODS

Patients

Twenty-nine patients who underwent CABG surgery with CPB were selected for this study. These patients were the CPB subgroup of a prospective randomized controlled trial comparing the outcome of CABG surgery with or without CPB.[1516]CABG with non-pulsatile CPB was used. Blood and urine samples were obtained at the arrival of the patient in the operating room (”preoperative”), at 6 h (”6 h postoperation [6 h postoperative]”) and 24 h postoperation (”24 h postoperative”). Arterial blood samples were collected in ethylenediaminetetraacetic acid and centrifuged at 1000 rpm for 10 min at 4°C (Thermo Fisher Scientific, Waltham, MA, USA). The plasma was aliquoted and stored at-80°C for further analysis. The original study protocol was approved by the local Medical Ethical Committee and written consent was given by each patient. The study is registered on ClinicalTrials.gov (NCT01347827).

Tumor necrosis factor-α quantification in plasma

Concentrations of TNF-α were measured in plasma using a commercially available multiplex immunoassay (human CVD panel 3, EMD Millipore, Corporation, Billerica, MA, USA) and were previously published for both cohorts with CABG surgery, with or without CPB.[16]

Renal injury marker measurement in plasma and urine

Concentrations of the renal injury markers NGAL and KIM-1 were measured in plasma (NGAL) and urine (NGAL and KIM-1) samples by ELISA (Quantikine ELISA kits DLCN20, DKM100, R and D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions. NGAL and KIM-1 concentrations were corrected for urine creatinine concentration (ng/ml per mg creatinine), assessed with the Creatinine Parameter Assay Kit (KGE005, R&D Systems).

Human umbilical vein endothelial cell culture

Human umbilical vein endothelial cell (HUVEC) (#C2519A, Lonza, Breda, The Netherlands) was cultured in EGM-2 BulletKit Medium (CC-3202, Lonza), in a humidified incubator at 37°C, in 5% CO2. In all experiments, confluent HUVEC was used at passage 5.

In vitro human umbilical vein endothelial cell stimulation with tumor necrosis factor-α

To determine the effect of TNF-α on the pro-inflammatory response of endothelial cells, HUVEC was cultured under static conditions in 12-well plates (Corning, Amsterdam, The Netherlands) and starved in fetal calf serum (FCS)-free EGM-2 BulletKit medium for 1 h. As control plasma, pooled plasma from five healthy humans was purchased from Innovative Research (IPLAK2E, Le-Perray-en-Yvelines, France). HUVEC was exposed to FCS-free medium containing 20% control plasma, 1 IU/ml heparin (Leo Pharma, Amsterdam, The Netherlands) and TNF-α (0.25–100 pg/ml, Boehringer Ingelheim, Ingelheim am Rhein, Germany) for 3 h. Thereafter, the medium was replaced by EGM-2 BulletKit medium for 1 h. The TNF-α concentrations used in vitro are comparable to concentration ranges reported for patient plasma during and after CABG surgery with CPB.[16]

Experimental setup to study HUVEC exposure to patient plasma and shear stress

An in vitro model was used to analyze the effects of combined exposure to plasma and shear stress on HUVEC behavior [Figure 1]. μ-Slides I0.4 Luer (80176, ibidi, Gräfelfing, Germany) were used as previously described.[17]Confluent monolayers of HUVEC were starved for 1 h in FCS-free medium and subsequently incubated with FCS-free medium containing 20% control or patient plasma and 1 IU/ml heparin for 3 h. Thereafter, cells were exposed to 20 dyn/cm2 shear stress for 1 h, to mimic the change of shear stress inherent to the CABG surgery with CPB procedure.

Figure 1

Experimental setup to analyze the effects of perioperative patient plasma on endothelial cells. Cells were exposed to fetal calf serum-free medium for 1 h, and subsequently to plasma (20% plasma in fetal calf serum-free medium) from three perioperative time points (preoperative, 6 h and 24 h postoperative) for 3 h. Thereafter, cells were exposed to shear stress (20 dyn/cm2) for 1 h and analyzed for mRNA expression indicative of changes in endothelial integrity (housekeeping), inflammation, and adhesion behavior

Gene expression analysis by quantitative reverse transcriptase-polymerase chain reaction

Cells were harvested in TRIzol™ Reagent (Thermo Fisher Scientific) and total RNA was isolated according to the manufacturer's instructions. RNA concentrations (OD260) were measured by NanoDrop® ND-1000 US-Vis spectrophotometer (NanoDrop Technologies, Rockland, DE, USA). cDNA synthesis was performed using random hexamer primers (Promega, Leiden, The Netherlands) and SuperScript III (Invitrogen, Breda, The Netherlands). The assay-on-demand primers/probe sets E-selectin Hs00174057_m1, GAPDH Hs99999905_m1, ICAM-1 h00164932_m1, interleukin (IL)-6 h00174131_m1, IL-8 h00174103_m1, platelet endothelial cell adhesion molecule (PECAM)-1 h00169777_m1, VCAM-1 h00365485_m1 and vascular endothelial (VE)-cadherin Hs00174344_m1 (TaqMan® Gene Expression, Thermo Fisher Scientific) were used. Quantitative reverse transcriptase-PCR was carried out on the ViiA™ 7 real-time PCR System (Thermo Fisher Scientific) with the following settings: 15 min 95°C, followed by 40 two-step cycles of 15 s at 95°C, and 60 s at 60°C. Average cycle threshold values (CT) were obtained for sample duplicates and normalized to the housekeeping gene GAPDH. The expression relative to GAPDH was calculated by 2-ΔCT. Fold change in the expression was calculated as quotient of 2-ΔCT value of the gene of interest and the 2-ΔCT value of the control plasma.

Statistical analyses

For experiments incorporating patient-derived material nonparametric testing was applied, therefore, whisker-box plots were used with whiskers from 10 to 90 percentiles and individual points represent values below and above the cutoffs. Due to lacking data points, the data were analyzed with the Skillings-Mack test by using the R package “Skillings. Mack: The Skillings-Mack Test Statistic for Block Designs with Missing Observations” by Patchanok Srisuradetchai.[18]Using the Durbin test, post hoc comparisons were done with a Bonferroni correction. Data of TNF-α stimulated endothelial cells were analyzed by the one-way analysis of variance with a Šidák correction. Appropriate post hoc comparisons were done and reported only when the P values of the statistical tests were significant. Differences were regarded as significant when P < 0.05. Graph plotting and one-way analysis of variance with a Šidák correction analyses were performed using GraphPad Prism software 8.2.1 (GraphPad Prism Software Inc., San Diego, California, USA).

RESULTS

Patient characteristics

The average age of the 29 patients was 63 years [range 47–77, Table 1]. Ninety percent were male, which is in line with the fact that more males than females undergo CABG surgery.[19]The total procedure time was 188 ± 31 min and the mean CPB time was 82 ± 23 min, with average aortic cross-clamp times of 53 ± 15 min. As indicated by the preoperative European System for Cardiac Operative Risk Evaluation (EuroSCORE) I of 2.3% and the post hoc calculated EuroSCORE II of 1.1%, this cohort had a low preoperative risk of postoperative mortality after cardiac surgery which corresponded to no observed patient mortality within 30 days postoperative.[1120]

Table 1

Perioperative patient characteristics

	Units	Numbers
Preoperative characteristics
Male/female	n	26/3
Age (years)	Mean (range)	63 (47-77)
Body mass index (kg/m2)	Mean±SD	28.1±4.0
Smoking	n (%)	19 (66)
EuroSCORE I	Mean±SD	2.3±1.7
Post-hoc calculated EuroSCORE II	Mean±SD	1.1±0.6
Medical and surgical history
Cardiovascular disease
Coronary disease	n (%)	29 (100)
Hypertension	n (%)	10 (35)
Percutaneous transluminal angioplasty	n (%)	2 (7)
Myocardial infarct	n (%)	0
Pericarditis/myocarditis	n (%)	0
Venous thrombosis	n (%)	1 (3.4)
Respiratory disease
Chronic obstructive pulmonary disease	n (%)	2 (6.9)
Emphysema	n (%)	1 (3.4)
Pneumonia	n (%)	1 (3.4)
Left ventricular ejection fraction
<40%	n (%)	2 (7)
40%-55%	n (%)	10 (35)
55%-70%	n (%)	17 (59)
Hemoglobin (mmol/L)	Mean±SD	9.1±0.8
Operative characteristics
Cardiopulmonary bypass time (min)	Mean±SD	82±23
Aortic cross-clamp time (min)	Mean±SD	53±15
Total operation time (min)	Mean±SD	188±31
Grafts (n)	Median (range)	3 (2-5)
Hemodilution factor	Mean±SD	0.7±0.1
Postoperative characteristics
Hemoglobin (mmol/L)	Mean±SD	6.2±1.1
Intensive care unit length of stay (days)	Median (range)	1 (0.5-12)
Hospital length of stay (days)	Median (range)	9 (3-29)
Reoperation	n (%)	3 (10)

EuroSCORE: European system for cardiac operative risk evaluation, SD: Standard deviation

Pro-inflammatory and renal injury markers

A temporary increase of plasma TNF-α 6 h postoperative was observed [Figure 2a]. The median plasma TNF-α concentrations were 3.7 pg/ml (range, 0.3–11.7 pg/ml) preoperative, 7.3 pg/ml (2.5–94.8 pg/ml) at 6 h (h), and 4.6 pg/ml (1.0–17.7 pg/ml) at 24 h postoperative. At 6 h postoperative, concentrations were higher than preoperative (P < 0.001) and 24 h postoperative (P < 0.001).

Figure 2

Markers of systemic inflammation and renal injury in plasma and urine from patients undergoing coronary artery bypass graft surgery with the use of cardiopulmonary bypass. Plasma and urine were obtained at three perioperative time points. Measurements of tumor necrosis factor-α in plasma (a), neutrophil gelatinase-associated lipocalin in plasma (b) and in urine (c), and kidney injury molecule-1 in urine (d) are depicted (n= 29). Urine marker measurements were corrected for urine creatinine levels. The whisker-box plots are shown with 10th to 90th percentile and points represent individual values below and above the 10th to 90th percentile cutoffs. Post hoc comparisonPvalues are shown in the graph

The median plasma NGAL concentrations were 72.8 ng/ml (range 29.3–199.6 ng/ml) preoperative, 120.8 ng/ml (52.6–359.1 ng/ml) at 6 h, and 99.8 ng/ml (45.4–272.1 ng/ml) at 24 h postoperative [Figure 2b]. The median plasma NGAL concentrations were increased postoperatively. All time points were significantly different from each other, P = 0.033 for start vs 24 h postoperative and the others P < 0.001. The median urine NGAL concentrations were 0.8 ng/mg (range, 0.2–7.1 ng/mg) preoperative, 0.8 ng/mg (0.2–8.5 ng/mg) at 6 h and 1.6 ng/mg (0.2–6.4 ng/mg) at 24 h postoperative [Figure 2c]. At 24 h postoperative, concentrations were higher than preoperative (P = 0.002) and 6 h postoperative (P = 0.019). The median urine KIM-1 concentrations were 0.1 ng/mg (range, 0.1–0.3 ng/mg) preopeartive, 0.3 ng/mg (0.1–1.2 ng/mg) at 6 h, and 0.5 ng/mg (0.2–1.2 ng/mg) at 24 h postoperative [Figure 2d]. The median KIM-1 concentrations at all time points were significantly different (P < 0.001) from each other. Collectively, these data show postoperatively elevated marker concentrations of inflammation and kidney damage.

In vitro effect of low TNF-α concentrations on endothelial inflammatory response

HUVEC was exposed to varying concentrations of TNF-α in vitro. The concentrations of TNF-α used to stimulate the HUVEC were based on the patient cohort's range of plasma TNF-α (0.3-94.8 pg/ml) [Figure 3]. In response to TNF-α exposure, mRNA expression of the (housekeeping) endothelial integrity markers VE-cadherin and PECAM-1 did not change, except for an upregulation seen for PECAM-1 mRNA at 5 pg/ml TNF-α (P = 0.009). The pro-inflammatory cytokine IL-6 expression varied with TNF-α, but was not significantly different from the unstimulated condition. A TNF-α concentration-dependent increase in mRNA expression was seen for the chemokine IL-8 and the endothelial adhesion molecules E-selectin, ICAM-1 and VCAM-1. Compared to the unstimulated control, mRNA expression of E-selectin at 20 pg/ml TNF-α (P = 0.002) and of IL-8, E-selectin, ICAM-1 and VCAM-1 at 100 pg/ml TNF-α (P ≤ 0.001) were upregulated compared to unstimulated control. Concentrations of TNF-α used in this experiment were comparable to TNF-α concentrations in patient plasma and were sufficient to activate HUVEC in vitro in the experimental set-up employed.

Figure 3

Changes in inflammation and adhesion of human umbilical vein endothelial cell in response to tumor necrosis factor-α. Under static conditions, Human umbilical vein endothelial cell were stimulated with tumor necrosis factor-α (0.25–100 pg/ml) for 3 h. mRNA fold changes compared to the untreated control (dotted line at 1) of VE-cadherin (a) and platelet endothelial cell adhesion molecule-1 (b), interleukin-6 (c) and interleukin-8 (d), E-selectin (e), intercellular adhesion molecule-1 (f), and vascular cell adhesion molecule-1 (g) are shown. Data are depicted as the mean of four biological replicates ± standard deviation. *Control versus tumor necrosis factor-α stimulated. Post hoc comparisonPvalues are shown in the graph

Effect of patient plasma followed by altered shear stress on endothelial behavior in vitro

We next studied the effect CABG plasma has on HUVEC expression of endothelial integrity (housekeeping), endothelial pro-inflammatory cytokines, and endothelial adhesion molecules in combination with shear stress [Figure 4]. Expressions of the integrity markers VE-cadherin and PECAM-1, or the cytokine IL-6 and the adhesion molecule ICAM-1 were not observed. Preopeartive plasma induced a 1.5-to 2-fold increase in the mRNA expressions of IL-8 (median 1.6, range 0.5–3.4), E-selectin (1.4, 0.4–8.3), and VCAM-1 (1.3, 0.5–7.0) compared to control plasma. Patients' plasma obtained 6 h postoperative resulted in lower mRNA expression of IL-8 (median 1.1, range 0.3–4.6; P = 0.007), E-selectin (1.1, 0.2–7.0; P = 0.007), and VCAM-1 (1.0, 0.3–4.8; P = 0.014) than preoperative plasma but similar to control plasma. Patients' plasma obtained 24 h postoperative resulted in lower mRNA expression of IL-8 (median 0.8, range 0.3–2.6; P < 0.001), E-selectin (0.8, 0.2–6.8; P < 0.001), and VCAM-1 (0.7, 0.3-2.7; P < 0.001) than preoperative plasma. Collectively, these data indicate that plasma from CABG-CPB patients obtained at 6 h and 24 h postoperative in combination with shear stress did not induce an increased endothelial adhesion molecule expression in HUVEC.

Figure 4

Effect of patient plasma and shear stress on inflammation and adhesion markers of Human umbilical vein endothelial cell. mRNA expression of VE-cadherin (a) and platelet endothelial cell adhesion molecule-1 (b), interleukin-6 (c), interleukin-8 (d), E-selectin (e), intercellular adhesion molecule-1 (f), and vascular cell adhesion molecule-1 (g) are shown as fold changes normalized to the healthy plasma control (dotted line at 1). The whisker-box plots are shown with 10th to 90th percentile and points represent individual values below and above the 10th to 90th percentile cutoffs. Post hoc comparison P values are shown in the graph

DISCUSSION

Cardiac surgery with CPB is commonly associated with AKI and microvascular endothelial inflammation is a potential underlying mechanism.[212223]We hypothesized that pro-inflammatory components of plasma from patients who underwent CABG surgery with CPB induce endothelial adhesion molecule expression when incorporating altered shear stress in the in vitro model. Although the postoperative patient plasma contained elevated concentrations of markers of inflammation and kidney injury, pro-inflammatory endothelial activation, represented by induction of IL-6, IL-8, E-selectin, ICAM-1, and VCAM-1, was not observed following the exposure of endothelial cells to postoperative plasma and altered shear stress.

In a previous analysis of this patient cohort, the pro-inflammatory cytokines TNF-α and IL-6, as well as the anti-inflammatory cytokine IL-10 and myeloperoxidase temporarily increased in plasma postoperatively, while endothelial-derived soluble adhesion molecules E-selectin, VCAM-1, and ICAM-1 did not.[16]This observation is partially not in line with previous reports that described a postoperative rise in both cytokines and soluble adhesion molecule markers.[2425]This discrepancy indicated that the microvasculature in patients with a low preoperative risk of postoperative mortality undergoing CABG surgery with CPB might not present with a pronounced pro-inflammatory phenotype. This was supported by the fact that only a small number of patients (5/29) had plasma TNF-α concentrations exceeding 20 pg/ml at a single time point, namely at 6 h postoperative. In patients undergoing CABG surgery with CPB, TNF-α concentrations above 20 pg/ml were associated with worse outcome compared to lower concentrations.[26]We demonstrated that TNF-α in concentrations similar to those observed in our cohort have caused a concentration-dependent increase in endothelial adhesion molecules under static conditions in vitro. Contrary to our expectations, the presence of cytokines such as TNF-α in patient plasma did not suffice to activate endothelial cells under shear stress in vitro, which is in line with previous findings under static conditions.[27]Interestingly, out of all time points preoperative plasma induced the highest expression of IL-8, E-selectin, and VCAM-1. This was unexpected as the pro-inflammatory cytokines measured in plasma where lower than those at 6 h postoperative. We can only speculate whether preoperative plasma contained unidentified pro-inflammatory mediators or whether postoperative plasma contained unidentified anti-inflammatory mediators that balanced out the measured pro-inflammatory marker concentrations. Furthermore, reasons for the absence of postoperatively increased adhesion molecules in response to plasma and the only temporary and moderate rise in cytokines may be related to advances in the anesthetic and surgical field. The examples of these advances are improved biocompatibility of extracorporeal circuits and hemodynamic management, as well as blood ultrafiltration for cytokine removal and shorter durations of CPB.[2829]Other reasons may be that the presence of anti-inflammatory mediators such as interleukin-10 in plasma, which also peaked at 6 h postoperative, may have counterbalanced a pro-inflammatory endothelial response.[16]This hypothesis can be studied in future studies using blocking antibodies against IL-10 and other plasma cytokines to see whether an anti-pro inflammatory rebalancing takes place during cardiac surgery. Moreover, cellular whole blood components such as leukocytes may be accountable for postoperative pro-inflammatory endothelial responses.[30]

The effect of CABG surgery with CPB on organs was investigated in terms of kidney injury markers, as AKI is a common CPB-associated complication.[31]The patients did not fulfil classical “Kidney Disease: Improving Global Outcome” AKI criteria, but a rise of NGAL and KIM-1 suggested mild kidney injury at 6 h postoperative, and markers remained elevated at 24 h postoperative in urinary measurements. Both NGAL and KIM-1 are detectable in urine within 2–6 h after renal injury and originate from renal tubules.[31]In addition, NGAL has extra-renal sources such as various inflammatory cells. Furthermore, the systemically measured concentrations only give limited information about organ-specific expression, and hence, it can only be speculated whether the local renal microvasculature expresses more pronounced signs of pro-inflammatory endothelial response in CABG patients, as observed in the rat models of cardiac surgery using CPB.[11]An animal model, which mimics human cardiac surgery with CPB and provides perioperative plasma and organ biopsies, might give insight into these mechanisms in future research.[3233]

Based on our current findings, we cannot conclude that microvascular pro-inflammatory activation by plasma components is the mechanism underlying the observed postoperative increase in kidney injury markers NGAL and KIM-1 in plasma and urine. Alternatively, endothelial hyperpermeability facilitating microvascular leakage and systemic hypotension leading to ischemia/reperfusion injury are noxious stimuli that have been suggested to cause postoperative AKI.[34353637]

This study has a number of limitations. First, the in vitro endothelial cells were exposed to 40% of the plasma that the patient's vasculature was in contact with, due to the experimental requirement to dilute plasma in cell culture medium. This limitation is inherent to an in vitro model, which is assessing only a fraction of factors and also excluded leukocyte-endothelial interactions and soluble factors of the complement and coagulation systems. Second, other pro-inflammatory plasma factors known to be released during CABG surgery such as the hypochlorous acid or thioredoxin-interacting protein were not studied. Third, the plasma samples collected at specific time points merely represent circulating plasmatic components at a particular time point, rather than the kinetics of plasma components throughout the perioperative time period. Only the inclusion of more time points or a large animal in vivo model will provide more information on events in the organ microvasculature during CABG with CPB. Moreover, there is a trend toward the use of pulsatile extracorporeal circulation in patients as this may preserve microcirculatory function better.[38]Since our study was conducted with nonpulsatile CPB, the in vitro model also incorporated nonpulsatile shear stress. Further, more information on baseline health status and comorbidities would have given a better insight into the preoperative patient characteristics. However, this cohort had a low preoperative risk of postoperative mortality. Further studies may look at patients with a higher risk for perioperative complications, for which additional information on health status and comorbidities would be necessary.

CONCLUSION

CABG surgery with CPB patients with low preoperative risk of postoperative mortality postoperatively showed signs of mild systemic inflammation and kidney damage. Plasmatic components do not seem to be accountable for the underlying postoperative microvascular endothelial adhesion molecule expression.

Research quality and ethics statement

This study was approved by the Institutional Review Board / Ethics Committee at the University Medical Center Groningen (Approval # METc 2011/045; Approval date April 19, 2011). Additionally, the project was registered at Clinicaltrials.gov (Identifier NCT01347827). The authors followed the applicable EQUATOR Network (http://www.equator-network.org/) guidelines, specifically the CONSORT 2010 Statement, during the conduct of this research project.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.