Modulation of the cellular and humoral immune response to pediatric open heart surgery by methylprednisolone

Authors


  • How to cite this article: Bocsi J, Hänzka M-C, Osmancik P, Hambsch J, Dähnert I, Sack U, Bellinghausen W, Schneider P, Janoušek J, Kostelka M, Tárnok A. Modulation of the cellular and humoral immune response to pediatric open heart surgery by methylprednisolone. Cytometry Part B 2011; 80B: 212–220.

Abstract

Background:

With the intention to reduce overshooting immune response, glucocorticoids are frequently administered perioperatively in children undergoing open heart surgery. In a retrospective study we investigated extensively the modulation of the humoral and cellular immune response by methylprednisolone (MP).

Methods:

This study was carried out on blood samples from two groups of children who had undergone surgical correction of atrial or ventricular septal defects, either without (MP, n = 10), or with MP administration (MP+, n = 23, dose median 11 (IQR 10–16) mg kg−1 body weight) before cardiopulmonary bypass (CPB, duration median 42 (IQR 36–65) min). EDTA blood was obtained 24 h preoperatively, after anesthesia, at CPB begin and end, 4, 24, and 48 h after surgery, at discharge and at out-patient follow-up (median 8.2 (IQR 3.3–12.2) months after surgery). Complex blood analysis including clinical chemistry and flow cytometry were performed to monitor humoral immune response, differential blood count, lymphocyte subsets, and the degree of activation of various leukocyte subpopulations.

Results:

The patients' postoperative courses and follow-up were uneventful. Release of IL-6 and IL8 was reduced and that of the anti-inflammatory cytokine IL-10 upregulated by MP. Significant increase of circulating neutrophils and monocytes as inflammatory reaction to surgery and CPB contact was detected in both groups. However, invasion of monocytes to the periphery was delayed with MP. CD4+ and CD8+ T-lymphocyte counts were lower with MP treatment. B-lymphocyte count increased significantly after surgery in MP+ but remained constant in MP group.

Conclusions:

MP treatment partially decreased the pro-inflammatory effect of CPB surgery and induced anti-inflammatory effect on the cellular and humoral level. © 2011 International Clinical Cytometry Society

Cardiopulmonary bypass (CPB) in children is thought to contribute to several adverse reactions following open heart surgery, including effusions, capillary leak syndrome (CLS), and multiorgan failure (1–4). Escalating pro-inflammatory response is thought to be responsible (2, 3, 5) and can induce low cardiac output (6–8). According to type of congenital heart disease and surgery, the incidence of post surgical CLS in children with open heart surgery ranges from 4 to 37% (3).

To reduce systemic inflammatory reaction induced by CPB and to diminish the peri- and postoperative morbidity several pharmacological and technical therapy interventions were developed (9–12). One of these strategies is the pre- or intraoperative application of glucocorticoids (GC). These steroids have an anti-inflammatory action (13), therefore steroid derivatives such as methylprednisolone (MP) or dexamethasone have been widely used during CPB surgery (14, 15). MP also has a protective effect on the vascular endothelium during cardiovascular surgery (16).

Available data suggest a clinical benefit of steroid application, but many details are unknown and conditions of steroid administration are highly variable (15). Evidence from adult trials is conflicting regarding the effects of GC on postoperative outcome such as postoperative course, length of hospital stay and survival (14, 17). It is also unclear whether results from adults can be directly transferred to children (18). Moreover, little is known how MP modulates the immune response in children in combination with the major stress response to surgery and administration of other drugs.

The aim of the study was to analyze the effect of perioperative MP administration on acute cellular and humoral immune reaction in children undergoing simple open heart surgery and comparing it to a similar group of patients without MP administration.

Since about 10 years preoperative administration of MP in these children is an internationally accepted standard procedure. Therefore, the present study is very unique and renders important information that will not be repeatable in new prospective studies for ethical reasons. In addition, even in a large cardiac center the number of pediatric patients is relatively low. Thus, it took many years to recruit a homogeneous patient group. We therefore followed the concept of predictive medicine by Cytomics or cell systems biology (19, 20) and performed, as comprehensively as possible, a laboratory analysis of the drawn samples. We show that the majority of the analyzed parameters rendered no significant differences between the two groups. We believe that these unique and important studies yield a plethora of information that is of value to a broader community.

PATIENTS AND METHODS

Selection of Patients

Our present study is a retrospective non-randomized analysis of children operated at our Heart Center in the years 1995–2000. Humoral and cellular immune response to elective CPB surgery was investigated. All studies were approved by the ethical committee, University of Leipzig. Written informed consent was obtained from the parents of all patients.

The majority of the patients enrolled in our earlier studies received MP (MP+) (21). Only a minority of them did not receive MP (MP). We analyzed the surgical records of all children (n = 198). According to the records, 26 of them did not receive MP before or during surgery. Most of these children underwent correction of an atrial (ASD) or a ventricular septal defect (VSD; total n = 14) while others had diverse congenital heart diseases. To make our patients group as homogenous (and therefore comparable) as possible, we decided to enroll only patients who underwent correction of ASD or VSD. Surgery for more complex congenital heart disease is associated with substantially longer CPB and cross-clamp times carrying prolonged effect of surgical trauma and CPB on the immune system or is known to be associated with latent immunological abnormalities, e.g., Glenn-Fontan surgery (22–24).

Inclusion criteria.

ASD or VSD repair, open heart surgery with CPB, age range 3–16 years. All patients included in the study had normal heart functions.

Exclusions criteria.

Left ventricular ejection fraction < 50%, known chromosomal abnormalities (e.g., trisomy 21) or other genetic defects, minimal-invasive surgery of the congenital heart disease, infections before or during surgery.

This resulted in 10 children not receiving MP (MP) and 23 who received MP (MP+) being enrolled in this study. The MP+ patients received, between anesthesia onset and begin of CPB, 250 mg MP either in a single dose or in two doses of 125 mg each. This resulted in a median dose of 11 mg kg−1 body weight (IQR: 10–16 mg kg−1).

Anesthesia and Antibiotic Regime

General anesthesia and myorelaxation were performed with midazolam, fentanyl, etomidate, propofol, and pancuronium. CPB was performed with a Stockert roller pump (Stockert Instrumente GmbH, Munich, Germany) and hollow-fiber oxygenator (DIDECO, Mirandola, Italy). Priming solution consisted of a crystalloid solution, mannitol (4–6 ml kg−1 body weight), human albumin, and compatible fresh blood. During bypass the hematocrit level was kept at 22–30% and the flow-rate was maintained at 2.7–3.5 l m−2 BSA per min. Hypothermia (28–35°C) was induced by cooling the priming solution in the extracorporeal circuit and circulating blood with heat exchanger. During the cooling period all patients received sodium nitroprussid (0.3–2.0 μg/kg/min) for vasodilatation. Bretschneider's or St. Thomas cardioplegic solutions were used for myocardial perfusion. Antibiotic regime consisted of Cephazolin (50 mg kg−1 body weight) in three separate doses.

During rewarming, also heated humidified inspired gases and intravenous administration of sodium nitroprussid were used. At end of the bypass, patients underwent ultrafiltration. Then, normal flow was re-established and the patient was rewarmed and heparin was neutralized by protamine sulphate. If necessary, catecholamine (dopamine, dobutamine, or epinephrine) was infused before the patient was weaned off the bypass.

Postoperative Care

After surgery all patients were mechanically ventilated. To optimize blood pressure and diuresis catecholamines and diuretics were infused. Laboratory assays including blood count, electrolytes, renal and liver tests, and coagulation analysis were performed immediately after the surgery and 16–24 h later. Echocardiography, chest X-ray, and sonography were performed, whenever necessary.

Blood Samples

Blood samples (∼5 ml patient−1) were drawn 1 day before surgery before any medication (1d-), after onset of anesthesia (anaesth), shortly after CPB onset (mean ± SD, 14 ± 6 min; range, 10–30 min) (CPB1), before CPB termination during reperfusion and rewarming (CPB2), 4–6 h after the end of surgery (4 h+), 1 (1 d) and 2 days (2 d) after surgery and at discharge (>3 d; 9.8 ± 3.6 days postoperatively), and at out-patient follow-up (3 m; time after surgery: 9.1 ± 7.0 months (range: 2–30 months); obtained from 25 patients only (16 MP+/9 MP).

EDTA anticoagulated blood and native blood were collected for flow cytometry and serology, respectively. Samples for serum and plasma were sedimented at 2,800g, and supernatants were frozen at −80°C in aliquots within 1 h after sampling. From the same blood samples, hematocrit values were determined.

Serum Concentration of Cortisol, MP, Complement, Cytokines, and Soluble Adhesion Molecules

Plasma concentration of MP and cortisol were measured by a Cortisol ELISA (IBL, Hamburg, Germany, Fig. 1A). This ELISA had a cross reactivity with MP of 5%. From CPB2 and 4 h+ values the half life of MP was estimated by:

equation image

where t1/2 is the half life, t the time between two concentration measurements, C0 the initial concentration (t = 0, here CPB2), and Ct the concentration after time t (here 4 h+) (25, 26).

Figure 1.

Time course of serum components induced by cardiovascular surgery in the MP+ (closed symbols) and the MP group (open symbols). Data display median values of 23 children with MP and 10 without MP treatment. Error bars indicate IQR. Data at sampling time CPBI and CPBII were corrected for hemodilution as shown in Patients and Methods. A. Cortisol (MP) and methylprednisolon with cortisol (MP+); B. IL-6; C. IL-8; D. sE-selectin; E. IL-10; F. C1-inhibitor. The symbols in the graphs indicate statistically significant differences: * = MP vs. MP+ by Wilcoxon test; + = time point × compared to 1d- (pre-operative value), for MP+ patients by Mann-Whitney U-Test; ∧ = time point × compared to 1d- for MP patients by Mann-Whitney U-Test.

The serum concentrations of the complement components C3, C4, C5, C1-inhibitor, and C3d were determined by radial immunodiffusion (The Binding Site, Heidelberg, Germany) with serum or EDTA-plasma (C3d). Total hemolytic complement CH100 was determined by lysis of antibody coated sheep erythrocytes (The Binding Site). Interleukin (IL) −8, IL-6, soluble-(s)E-selectin, and sL-selectin (all from R&D Systems, Oxon, UK), sICAM-1 (Bender Med Systems), C5a (Enzygnost, Behringwerke AG, Marburg, Germany) were quantified in serum by enzyme linked immune assay (ELISA). The lower limits of detection were 1.1 mg l−1 (C3d), 0.1 μg l−1 (C5a), and 4 pg ml−1 (IL-6 and IL-8) 142 U ml−1 (CH 100), 155 mg l−1 (C3), 58 mg l−1 (C4), 20 mg l−1 (C5), 45 mg l−1 (C1-Inhibitor), 68.75 ng ml−1 (sICAM-1), 8.19 ng ml−1 (sE-selectin) and 62.73 ng ml−1 (sL-selectin).

The measured serum concentration of a biologically active compound is relevant for the response of the patient. However, in order to compare patients with slightly varying hemodilution serum component concentration and absolute cell count values were corrected for hemodilution according to Hambsch et al. (21):

equation image

where [C]korr is the corrected concentration, [C]sample that in the sample, HKsample the hematocrit value in the sample and HK1d- the preoperative HK value. Correction for hemodilution was done for all samples except the out-patient follow-up sample.

Cell Phenotyping and Antigen Expression

Phenotyping of leukocytes was done as detailed elsewhere by the whole blood technique (24, 27). About 40 μl EDTA blood was mixed with a cocktail of directly fluorescence dye-conjugated monoclonal antibodies. Cells were stained for 15 min at room temperature in the dark, after adding 1 ml of a lysing solution (BD Biosciences, San Jose, CA) the samples were mixed and incubated for 10 min at room temperature in the dark. The cells were spun down at 300g, and the supernatant discarded. The cells were washed twice in 1 ml phosphate buffered saline pH = 7.4 (PBS, Sigma-Aldrich Chemie GmbH, Deisenhofen, Germany) and finally resuspended in 200 μl 0.5% paraformaldehyde (Sigma) in PBS. Antibodies for lineage and activation markers were obtained from following providers: BD Biosciences (CD3—T cells, CD4—T-helper/inducer cells, CD8—T-cytotoxic/suppressor cells, CD14—LPS receptor on monocytes, CD45—pan leukocyte antigen, HLA-DR—MHC II, CD11a—LFA-1, CD11b—Mac-1, CD69—early activation antigen), Caltag (Hamburg, Germany), (CD16—Fc-γ Receptor III on neutrophils and NK-subset, CD19—B cells, CD25—IL2-R, CD56—N-CAM on NK subset), Beckman-Coulter Corp. (Hialeah, FL) (CD54—ICAM-1), and DAKO (Glostrup, Denmark) (CD18—β2-integrin).

Background fluorescence was quantified after staining with appropriate control antibodies (BD Biosciences). Antibodies were labeled with the fluorescent dyes fluorescein isothiocyanate (FITC), phycoerythrin (PE), peridinin chlorophyll protein (PerCP™), or allophycocyanin (APC) in cocktails of three- or four-color combinations. Cells were measured on a dual laser flow cytometer (FACSCalibur; BD Biosciences) calibrated with calibration microbeads (Spherotech, Libertyville, IL) for standardization (28, 29).

Flow cytometric data were analyzed with the CellQuest software package (BD Biosciences). The percentage of leukocyte subsets was quantified, and their cell counts per blood volume were calculated based on the differential blood count. For each cell subset the mean fluorescence intensity (MFI) as a measure of antigen expression and activation degree was determined after subtraction of the fluorescence intensity of control antibody-stained cells (28, 29).

Statistical Analysis

Data distribution was tested for normality (Kolmogorov-Smirnov test). Because the majority of the analyzed laboratory parameters were not normally distributed data are presented as median and inter quartile range (IQR) and only nonparametric tests were performed (exception: multivariate ANOVA,). Changes with time within individual groups (MP+ or MP) were analyzed by the Friedman test. The time courses of MP+ and MP groups were compared using multivariate ANOVA (MANOVA). Single time-points within one group were compared to the 1d- values by paired Wilcoxon test, identical time-points between MP+ and MP by nonparametric Mann-Whitney U-test. All statistical analyses were performed using the SPSS program package (SPSS V.8; Knowledge Dynamics, Canyon Lake, TX). Values were regarded significantly different if P < 0.05. In multiple tests the significance levels were corrected according to Bonferroni: P = 0.05/n where n = number of time-points analyzed for a single analyte tested. If not stated otherwise, in the following only data rendering statistically significant differences or changes are discussed.

RESULTS

Clinical Characteristics

Clinical data of the patients are shown in Table 1. Both groups did not differ with respect to age, gender, body weight, and duration of surgery. Intubation time, ICU days and time until discharge from the clinic were identical. Pleural and pericardial effusions were detected in both groups mostly without hemodynamic significance and had disappeared until discharge. The drainage volume did not differ on the day of surgery and the following days between MP+ and MP group. All children were discharged well. Altogether, the postoperative courses could be regarded as normal and uneventful. Post-operatively none of the patients developed symptoms of cardiac insufficiency.

Table 1. Patient Characteristics and Perioperative Data
 MP+-group (n = 23)MP-group (n = 10)Significance
Median (IQR)Median (IQR)MP+ vs. MP
  • NS = not significant (Mann Whitney U-test,

  • a

    Except gender: X2-test).

Age (years)6.5 (5.2 ; 7.4)7.9 (6.5 ; 9.7)NS
Gender (m/f)′10/13′4/6NSa
Weight (kg)21 (15 ; 26)26 (22 ; 27)NS
Height (cm)120 (106 ; 128)128 (121 ; 132)NS
Duration of surgery (min)120 (103 ; 134)110 (106 ; 118)NS
CPB time (min)42 (36 ; 65)51 (42 ; 59)NS
Cross-clamping time (min)21 (14 ; 35)27 (14 ; 32)NS
Hypothermia (°C)32 (30 ; 32)32 (30 ; 32)NS
Duration of Intensive Care (h)43 (29 ; 48)43 (29 ; 48)NS
Total Drainage Volume (ml kg−1 body weight)6.7 (4.34 ; 9.21)5.67 (3.75 ; 7.52)NS
Methylprednisolone (mg)250 (250 ; 250)0 
Methylprednisolone (mg kg−1 body weight)11 (10 ; 16)0 

Laboratory Results

From all analyzed immunological parameters only 11 showed significantly different values either according to MANOVA test and/or at identical time-points according to Wilcoxon test. Table 2 summarizes the results for MANOVA and Friedmann test for these parameters.

Table 2. P-Values After MANOVA and Friedman Tests
ParameterMANOVAFriedman
MP+MP−
IL6<0.001<0.001<0.001
IL8NS<0.001<0.001
sE-Selectin<0.001<0.001<0.001
IL10NS<0.001<0.001
C1-inhibitorNS<0.001<0.001
Neutrophil cell countNS<0.001<0.001
Monocyte cell count<0.001<0.001<0.001
NK cell countNS<0.001<0.001
CD4 cell count0.002<0.001<0.001
CD8 cell countNS<0.001<0.001
B-cell countNS<0.001<0.001

Serum MP and Cortisol

In the MP group the postoperative cortisol level increased from base line level of 100 ng ml−1 up to 250 ng ml−1 at 4h+ (Fig. 1A). All patients from the MP+ group had during and up to 4 h after surgery about 100× higher cortisol ELISA values than the MP patients. Maximal values were obtained at CPB2 (median: 11,940 μg l−1) followed by the 4h+ concentration (median 3,200 μg l−1). At 1d+ serum levels did not significantly differ from those of MP patients. From these values half life of MP in the blood was calculated as 2.62 h (median, IQR: 2.53–2.93 h).

Humoral Parameters

Serum concentration of the following compounds did not show significant MP related changes: IL2 receptor, TNF-α, Histamine, IL-1β, IL-12p70, sL-selectin, sP-selectin, sICAM-1, CH-100, C3, C3d, C3d/C3-ratio, C4, C5, C5a, C5a/C5 ratio, hematocrit, erythrocyte, and platelet counts. Their time courses were found to be similar with and without MP (MANOVA) and there were no significant differences at identical time-points between MP+ and MP. This is in agreement with our earlier reports (21, 30, 31).

Proinflammatory Cytokines and Endothelium Activation

IL-6 and IL-8 concentrations (Figs. 1B and 1C) were significantly affected by surgery in both groups (both P < 0.001, Friedman) and the time courses between the MP+ and MP group were different (IL-6: P < 0.001, IL-8: P = 0.014; MANOVA). The concentrations were higher without MP administration for both cytokines (significant at 4h+). The concentration of sE-selectin (Fig. 1D) was in both groups also significantly altered by surgery (P < 0.001) and the time courses differed significantly between MP+ and MP (P = 0.001; MANOVA). sE-selectin serum level was under MP administration perioperatively reduced.

Antiinflammatory Parameters

IL-10 and C1-inhibitor serum concentrations (Figs. 1E and 1F) were significantly affected by surgery (P < 0.001 for both). Interestingly, the time courses of IL-10 did not significantly differ between MP+ and MP) whereas those of C1-inhibitor were different (P = 0.019). IL-10 levels increased substantially (two to three fold) at CPB2 in both groups. However, in the MP+ group the IL-10 levels were at 4h+ still increased whereas in the MP group values returned to baseline. C1-inhibitor concentration slightly decreased perioperatively and increased until discharge above baseline to similar amplitude. This increase was slightly faster in the presence of MP.

Cellular Immune Response

Innate immunsystem.

Cell counts of cells from the innate immune system (neutrophilic granulocytes, monocytes and NK cells; Figs. 2A–2C) were group independently modified by surgery (all P < 0.001). However, time courses of MP+ and MP- were only significantly different for monocytes (P < 0.001). The data indicate that with MP time course for monocytes is slightly delayed but amplitude values remain similar. Neutrophil count strongly increased during CPB contact and remained highly increased from CPB2 up to 2d after surgery. Monocyte cell count decreased in both groups after anesthesia until CPB1 and was postoperatively two- to three-fold above baseline. Maximum monocyte count was reached with MP at 1d but already at 4h+ without MP. NK-cells counts increased at CPB2 but were not significantly affected by MP. Eosinophilic and basophilic granulocyte cell counts did not show MP-related changes.

Figure 2.

Time course of peripheral blood leukocyte subset cell counts in the MP+ (closed symbols) and the MP group (open symbols). Data display median values of 23 children with MP and 10 without MP treatment. Error bars indicate IQR. Data at sampling time CPBI and CPBII were corrected for hemodilution as shown in Patients and Methods. A. Neutrophilic granulocytes; B. Monocytes; C. Natural killer cells; D. CD4+ T-lymphocytes; E. CD8+ T-lymphocytes; F. B-lymphocytes. Symbol legend explanations in Figure 1 caption.

Adaptive immune system.

The time courses of CD4+ and CD8+ T-lymphocyte counts (Figs. 2D and 2E) were significantly modified for both groups by surgery (all P < 0.001) and the kinetics differed between MP+ and MP for CD4+ (P = 0.002) but not for CD8+. After a short decrease at anesthesia CD4+ and CD8+ counts increased to preoperative levels at CPB2 both for MP+ and MP followed by a significant cell loss. Cell counts were lowest at 4h+ in the MP+ group and at 1d in the MP group. This drop was followed by a recovery reaching baseline values at around discharge.

Interestingly, B cell counts were only significantly affected by surgery in the MP+ (P < 0.001) but not in the MP group (Fig. 2F). Therefore, time course of MP+ patients significantly differed from that of MP (P = 0.013). In the MP+-group cell counts decreased from anesthesia on until 4h+ and increased thereafter to a maximum at 2d. B cell counts of MP+ patients were in the 3-m sample significantly below 1d-.

Level of cell activation.

It was tested if the surgery induced alterations of the expression level (MFI) of cell surface activation markers and adhesion molecules were modified by MP. Although significant changes of the MFI of activation markers (CD25, CD69, HLA DR) and adhesion molecules (CD11a, CD11b, CD18) were induced by surgery no MP induced modifications were observed on the leukocyte subsets shown in Figures 2A–2F. Time courses were similar to earlier reports from our group (21, 31).

DISCUSSION

Our results show that humoral (complement, cytokines, soluble adhesion factors) and cellular immune reactions (cell count changes, cell activation level) were not substantially affected by single dose MP treatment. Only slightly reduced pro-inflammatory and enhanced anti-inflammatory response with a short delay of the cellular immune response were detected.

To the best of our knowledge, so far no study has compared in children such a broad scale of immunological reactions to CPB surgery with and without MP administration in a very homogeneous group of patients.

In agreement with earlier findings, the systemic inflammatory reactions to CPB resemble to sepsis with increased pro- and anti-inflammatory cytokine release and reduction of the activity of the cellular innate and adaptive immune system (1, 32–34). To reduce the proinflammatory reaction the majority of surgical corrections for congenital heart disease are performed under high-dose GC administration (15). However, very little is known about immunomodulatory effects of GC in pediatric cardiac surgery.

We could show that MP ameliorates the proinflammatory (IL-6, IL-8) and supports the anti-inflammatory (IL-10) part of the immune response. These results corroborate earlier findings in children with and without GC administration (35–37). Bronicki et al. (36) reported reduced proinflammatory (IL-6, TNFα) response by 1–10 mg kg−1 BW GC but no effect on C3a levels and neutrophil count. Gessler et al. (35) could not find significant changes of C-reactive protein, IL-8 and leukocyte count induced by 30 mg kg−1 BW MP but Lindberg et al. (37) found reduction of CRP release when applying 1 mg kg−1 GC. CPB surgery substantially affects the immune response with rapid IL-10 and slower IL-6 increase (30). This response is further escalated by GC. MP seems also to have a protective effect on the endothelium as indicated by reduced sE-selectin serum levels. Increased IL-10 levels at the end of surgery modulate the subsequent IL-6 and IL-8 secretion but MP could directly reduce proinflammatory cytokine secretion without induction of IL-10 as shown in experiments using a simulated extracorporeal circuit (38).

Mobilization of monocytes but not of NK cells and granulocytes was modulated by MP. Although the cell count amplitudes did not differ in both groups monocyte time courses were delayed with MP by around one day. This difference may be due to the lowered levels of IL-6 and IL-8. Grosek et al. (39, 40) showed that in children HLA-DR expression on monocytes is dramatically reduced during cardiac surgery with GC. This is in agreement with our present and earlier studies (21, 30, 31). However, MP did not affect the expression of all analyzed leukocyte cell surface activation markers including HLA-DR.

Importantly, complement activation that is a prominent response to CPB (30) was in agreement with others (36), not influenced by MP. Therefore the levels of anaphylatoxins, the most potent chemokines for neutrophils and monocytes, remain similar. This may be a reason why the mobilization of NK cells, that is a characteristic reaction to CPB (33, 34, 41) and of neutrophils was unaffected by MP. Accordingly, MFI values of activation markers on leukocytes were not affected.

Regarding the response of the cellular adaptive immune system our results clearly show lower T-lymphocyte and T-cell subpopulation counts with MP shortly after surgery. Typically, after surgical trauma T-cell numbers drop dramatically as a result of increased migration into the periphery, particularly into the lymphatic organs (34). Unfortunately, no studies exist on effects of GCs on T-lymphocytes in vivo during CPB surgery. Grosek et al. (40) showed that in children with normal cardiac output the postoperative drop of T-cell counts is more pronounced than in those with low cardiac output. Therefore, enhanced reduction with MP may be interpreted as a beneficial response.

T-cell activity was according to surface activation marker expression not affected by MP. However, GCs have an in vitro inhibiting effect on cytokine production (42) as well as reduction of IL-2-receptor signal-transduction pathway on these cells (43). The parallel reduction of the T-cell function and their IL2-production promotes probability of infections and postoperative complications (41). Further functional assay would be needed to test alterations in the biological activity of T-cells.

MP-Induced B-Cell Loss in the Follow-Up

The most prominent cellular observation was the postoperative (1d to >3d) mobilization of B-lymphocytes by MP. Intra- and extravasion of B-cells is in the absence of MP in equilibrium during and after CPB contact. By MP treatment B-cells are mobilized and increase postoperatively in numbers. The trigger for their release from peripheral organs is unclear but may be due to T-cell-dependent activation or by enhanced cytokine secretion. Although, activation antigen expression analysis did not reveal phenotypic differences between MP and MP+ group this significant increase could be a consequence of MP boosted B-cell activation (43).

B-cells were the only cell type where a significant change of cell count in the out-patient follow-up sample was found. In MP+ B-cells counts were at 3m significantly lower than at −1d, but no difference was found for MP. The 3m samples were obtained at average 9 months after the surgery. At an age similar to our study group normal maturation leads to a decrease in B-cell count (44, 45). Therefore, this decrease may be due to normal age related changes. To clarify this issue, we used the cell calculator provided by Huenecke et al. (45) to determine theoretically expected B-cell numbers in out-patient follow-up samples based on preoperative B-cell counts. Measured counts were significantly below expected counts by about 30%. No such changes were found for all other major cell types analyzed. It is difficult to conclude from this observation if in children single high-dose MP treatment leads to long-lasting B-cell depletion. Further, follow-up and controlled studies would be needed to confirm our finding.

Influence of MP on Postoperative Clinical Course

Our study with a small number of patients does not allow definite clinical conclusions particularly because the clinical course of these patients had never been a matter of concern and was not the aim of the present study. However, results obtained in double-blind trials with children (36, 37) and in adult cohorts (43) could not demonstrate obvious clinical benefits regarding postoperative clinical course, intubation time, stay on ICU, duration of hospitalization, as well as complications including pleural - or pericardial effusions or drainage volume. Anti-inflammatory effects of MP may be clinically beneficial in more complex lesions with longer CPB runs and exaggerated systemic inflammatory response as shown for adults (46).

Influence of MP Dose

CPB-induced trauma leads to postoperative release of cortisol as also shown by others (47). GC values were in the MP+ group approximately two orders of magnitude higher due to MP administration. MP was cleaved with a half life that was in the range of that for adults (3.2–5.5 h) (48). Already 24 h after surgery there were no differences of cortisol levels between MP and MP+ therefore MP was only very shortly present.

The findings on dose dependent modulation of the immune response by MP are contradictory. Varan et al. (49) did not find differences in the release of proinflammatory cytokines and neutrophil counts in children receiving 2 or 30 mg kg−1 MP. Schroeder et al. (50), however, found a decreased proinflammatory response (IL-6, sE-selection, sICAM-1) and an elevated anti-inflammatory response (Il-10) in the 2 × 30 mg kg−1 dose vs. 30 mg kg−1 dose group. This seems to be in agreement with Bourbon et al. (51) who reported for adults that reduction of the proinflammatory immune response is more pronounced under 10 mg kg−1 MP than under 5 mg kg−1 MP dosage.

We tested dose dependent effects of MP by splitting MP+ patients into a high and a low dose group (< or > median, 10.7 mg kg−1; high dose: median 16.1, IQR 15.4–16.8, 11.3–18.5 mg kg−1 BW, n = 11; low dose: median 9.6, IQR 9.5–9.6, 4.2–10.7 mg kg−1 BW, n = 12). Differences between these groups were not significant. Although our study group is too small to give a clear answer on dose dependence it may indicate effects that need further attention.

MP doses commonly used were derived from septic shock studies in adults which demonstrated that at least 30 mg kg−1 was required to influence patient survival (52). Although both CPB induced complications and septic shock are related to escalating proinflammatory syndrome significant differences exist so that high-dose GC may not be appropriate during CPB. Furthermore, dosages in adults may also not be appropriate for children. Therefore, routine application of GC for CPB surgery has been critically discussed and selective use for selected patient population has been recommended (53, 54).

Study Limitations

Owing to the cardiac defects and the small number of patients enrolled it is difficult to judge clinical relevance associated with MP treatment. Also the finding of absent observable changes in cell activity has to be further tested by in vitro functional assays. The “traditional” way of data-analysis may leave system wide changes induced by MP undiscovered. Therefore, more complex data pattern analysis needs to be applied to test for cell systems wide immunomodulatory effects in future (55, 56). Furthermore, polychromatic cytometry (57) or mathematical combination of our four-color panel data (58) may render more detailed cell subsets with specific MP related responses.

CONCLUSIONS

MP carries a partial shift of short-term immune response to CPB surgery while promoting some anti-inflammatory processes and simultaneously decreasing levels of pro-inflammatory cytokines. At the applied MP concentration the humoral and cellular response of the innate and the adaptive immune system were dose independent. The immunological changes observed might play a role in a more complex clinical context such as neonatal cardiac surgery or surgical procedures with longer CPB runs and pronounced systemic inflammatory response.

Acknowledgements

The authors thank Jacqueline Richter for excellent technical help and Dr. Anja Mittag for careful reading of this manuscript.

Ancillary