Hyporesponsiveness of T cell subsets after cardiac surgery: a product of altered cell function or merely a result of absolute cell count changes in peripheral blood?

An appropriate TH cell response plays a key role in immune reaction after surgical trauma. Several studies have shown that T helper (TH) cell function is altered after cardiac surgery with cardiopulmonary bypass (CPB) [1,2]. Initially, both the synthesis of pro-inflammatory TH1 cytokines (IFN-gamma and IL-2) and the release of anti-inflammatory TH2 cytokines (IL-4 and IL-10) are suppressed [3]. However, there is an ongoing discussion as to whether this altered cytokine synthesis in cell culture is mediated by a functional impairment of lymphocyte subsets or simply by a reduction in cell numbers [4–6]. The initial decrease in the specific immune response is followed by an increase in TH2 cell activity and a delayed recovery of TH1 cell function. This process is known as the TH1/TH2 shift and is considered responsible for the reduced competence of the specific and the non-specific immune system in patients with sepsis and patients after cardiac surgery with CPB. This TH1/TH2 shift may thus predispose to postoperative complications and/or a poor outcome [7,4,5,8]. Mechanical alteration, non-specific activation of lymphocytes by extracorporeal circulation as well as dilution effects are discussed as possible causes of the change in T cell function and the decrease in cytokine synthesis in cell culture after cardiac surgery with CPB. A number of experimental studies have shown that incomplete T lymphocyte activation can lead to functional impairment (anergy) and even induce apoptosis [4,9–11]. A better understanding of the interaction between extracorporeal circulation and circulating lymphocytes might help reduce postoperative alterations in the specific and the non-specific immune system. For this reason, we conducted a prospective study involving 40 patients who underwent elective cardiac surgery with CPB. We investigated lymphocyte subpopulations, lymphocyte activation in whole blood, and the synthesis of TH1 cytokines (IFN-gamma and IL-2) and TH2 cytokines (IL-4 and IL-5) in stimulated whole blood cultures (Figs. 1–4). Absolute leukocyte, neutrophil and lymphocyte counts and haemoglobin levels (mean ± SEM) in whole blood samples obtained from study group patients preoperatively and on the first five postoperative days after elective cardiac surgery (d0: preoperatively; dx: immediately after surgery; d1: first postoperative day; d3: third postoperative day, d5: fifth postoperative day). The grey bars show preoperative values of the study group patients (mean ± SEM). p < 0.05 versus d0 and p < 0.001 versus d0. The results are expressed as mean ± standard error of the mean. IFN-gamma, IL-2, IL-4 and IL-5 levels (mean ± SEM) in the supernatants of stimulated whole blood cultures (SEB 1μg/mL and LPS 1μg/mL) obtained from study group patients preoperatively and on the first five postoperative days after elective cardiac surgery (d0: preoperatively; dx: immediately after surgery; d1: first postoperative day; d3: third postoperative day, d5: fifth postoperative day). The grey bars show preoperative values of the study group patients (mean ± SEM). White dots: uncorrected cytokine level; black dots: concentration corrected to preoperative absolute TH cell count; p < 0.05 versus d0 and p < 0.001 versus d0. The results are expressed as mean ± standard error of the mean. After having given their written informed consent, 40 patients undergoing elective cardiac surgery were included in this study, which was approved by the local ethics committee. All surgical procedures began at 8:30 a.m. Patients with immune, renal, or central nervous system dysfunction as well as patients with congestive heart failure, exogenous hormone therapy, malnutrition, diabetes, malignancy, infection or inflammation were not enrolled in the study. Clinical characteristics of the patients are given in Table 1. Clinical data of patients (n = 40) We used the same cardiopulmonary bypass and perioperative management procedures as described previously [3]. Briefly, anaesthesia was induced with etomidate (200–300 μg/kg) and fentanyl (20–30 μg/kg). After endotracheal intubation and muscle relaxation with pancuronium (100–200 μg/kg), anaesthesia was maintained using fentanyl, midazolam and isoflurane (0.4–1.5%). The cardiopulmonary bypass equipment included non-pulsatile roller pumps and membrane oxygenators (Affinity®, Avecor, Bellshill, Scotland). The pump was primed with a standard electrolyte solution containing 5000 IU heparin, 1000 mL Ringer's lactate, 500 mL NaCl 0.9%, 500 mL of a commercial serum protein solution (Biseko®, Biotest, Dreieich, Germany) and 250 mL of a 15% mannitol solution (Osmofundin 15%®, Braun Melsungen, Melsungen, Germany). Heparin (300 IU/kg) was administered just before vascular cannulation. After the institution of cardiopulmonary bypass at a flow rate of 2.4–3 L/m2/min, the aorta was cross-clamped and a cold crystalloid cardioplegic solution was injected. After CPB, protamine was infused. All patients received 2,000,000 IU of aprotinin before the onset and 1,000,000 IU at the end of CPB to preserve platelet function. Cefazolin (3× 1.5 g) was used for perioperative antibiotic prophylaxis. All study patients underwent cardiac surgery at 8:30 a.m. in order to minimise any time-dependent variation of mediator synthesis. Blood samples were taken preoperatively (d0), immediately after surgery (dx), and on the 1st (d1), 3rd (d3) and 5th (d5) postoperative days from arterial or venous lines. Previous tests have shown that the blood collection site had no significant effect on the expression of the epitopes investigated or on the composition of the samples. The blood samples were collected in ethylene diamine tetraacetic acid (EDTA) tubes (Vacutainer®, Becton Dickinson, Heidelberg, Germany) for blood cell counting and lymphocyte analysis or in heparinised cell preparation tubes (CPT®, Becton Dickinson). Absolute leukocyte counts and haemoglobin levels were determined using a commercially available blood cell counter (MAX M®, Coulter, Krefeld, Germany) according to the supplier's instructions. This method was validated through interlaboratory proficiency tests in which the laboratory participated. Lymphocyte phenotyping was performed using a MultiTEST® system together with a FACSCalibur flow cytometer (Becton Dickinson) according to the supplier's instructions. 50 μL of EDTA whole blood samples were stained with the following antibodies (Antigen/Clone/Fluorescent dye): CD45/2D1 (Hle-1)/peridinin chlorophyll protein (PerCP); CD3/SK7/fluorescein isothiocyanate (FITC); CD4/SK3/allophytocyanin (APC); CD8/SK1/APC; CD16/B73.1/phytoerythrin (PE); CD56/NCAM 16.2/PE; HLA-DR/L 243/APC. The CD45+ population was enumerated using a CD45 versus side scatter plot. These lymphocytes were further assayed. The following antigens were used to identify lymphocyte subsets: CD45+: leukocytes; CD3+: mature T lymphocytes; CD3+/CD4+: T helper/inducer lymphocytes; CD3+/CD8+: T suppressor/cytotoxic lymphocytes; CD3-/CD16+/CD56+: natural killer lymphocytes; CD4+/HLA-DR+: activated T helper/inducer lymphocytes; CD4+/CD38+/HLA-DR+: activated mature T helper/inducer lymphocytes; CD8+/HLA-DR+: activated T suppressor/cytotoxic lymphocytes; CD8+/CD38+/HLA-DR+: activated mature T suppressor/cytotoxic lymphocytes. Using TruCOUNT® tubes, absolute numbers and percentages of HLA-DR-positive cells were determined. All flow cytometry reagents were obtained from Becton Dickinson, Heidelberg, Germany. This method was validated through interlaboratory proficiency tests in which the laboratory participated. The heparinised blood samples were diluted 1:2.5 with RPMI medium containing 10% heat-inactivated fetal calf serum and 2% penicillin-streptomycin. The whole blood (WB) cultures were stimulated with lipopolysaccharide (LPS, 1 μg/mL) and staphylococcal enterotoxin B (SEB, 1 μg/mL) to induce the production of IL-4, IL-5, IL-2 and IFN-gamma. The cultures were incubated at 37 °C in a 6% carbon dioxide atmosphere. After 24 h, supernatants were collected and stored at −80 °C until assayed. All reagents were obtained from SIGMA®, Deisenhofen, Germany. In a pilot study (data not shown), no cytokine production was detected in unstimulated cultures at the time points investigated. For this reason, we did not investigate unstimulated samples. IL-2, IFN-gamma, IL-5, and IL-4 were measured using the cytometric bead array (CBA) kit from Becton Dickinson. This test has a sensitivity of 5 pg/mL. If the concentration exceeded the upper limit of the assay range (5000 pg/mL), the supernatant was diluted in assay diluent. Operative and postoperative data are given in Table 2. There was no operative morbidity or mortality. Of the 40 patients, 24 underwent coronary bypass grafting. In 14 patients, coronary artery bypass grafting was performed together with valve replacement (aortic valve: n = 10, mitral valve: n = 3, both: n = 1). In two patients, a Bentall procedure was performed. Operative data of patients (n = 40) The absolute numbers of leukocytes and neutrophil granulocytes increased during the entire postoperative course. Maximum values were observed on dx and d1. Although the number of lymphocytes increased immediately after surgery (p < 0.05), it dropped significantly below preoperative levels on d1. The values then increased again and returned to preoperative levels on d3. Haemoglobin levels were significantly reduced during the entire postoperative course. The number of T lymphocytes in peripheral blood increased immediately after surgery and decreased to a minimum on d1. Although the number increased again, preoperative baseline levels were not reached until d5. T helper cells showed a similar pattern. The increase on dx, however, did not reach a statistical significance of p < 0.05. Unlike the number of T helper cells, the absolute number of T suppressor cells significantly increased immediately after surgery. The number decreased to a minimum on d1 and returned to baseline levels on d3. The absolute number of natural killer cells postoperatively increased on dx. The peripheral blood concentration of these cells then decreased but was not significantly reduced. Lymphocyte activation was assessed by monitoring HLA-DR expression on CD4+ and CD4+/CD38+ cells (mature T helper lymphocytes) and on CD8+ and CD8+/CD38+ cells (mature T suppressor lymphocytes). Both cell populations showed no increase in HLA-DR expression before and immediately after surgery. A significant increase in T helper lymphocyte activation in contrast to a minor and statistically non-significant increase in T suppressor cell activation was observed on the first postoperative day. The activation levels returned to preoperative baseline values on d3 and d5. The release of TH1 cytokines (IL-2 and IFN-gamma) and TH2 cytokines (IL-4 and IL-5) was significantly reduced in the immediate postoperative period (dx) and reached a minimum on d1. IL-5 synthesis returned to preoperative levels on d3. IFN-gamma, IL-2 and IL-4 levels also increased in the supernatants of stimulated whole blood cultures but remained decreased until d5. When the cytokine concentration was corrected to the absolute TH lymphocyte count that had been determined preoperatively in cell culture, IFN-gamma showed the same course as for the uncorrected levels. Whereas the corrected IL-2 and IL-5 levels returned to preoperative values on d1, IL-4 release did not return to baseline levels until d3. In our study we were able to show in a representative population of elective cardiac surgery patients that the activity of the non-specific immune system increased in the immediate postoperative period and did not return to the baseline level until the fifth postoperative day. In addition, we proved that the concentration of lymphocytes and especially T lymphocytes did not decrease immediately after surgery (dx) but on d1, whereas CD4+ lymphocytes showed the strongest activation at this time point (d1). Nevertheless, IFN-gamma synthesis remains depressed even when compared with a standardised number of lymphocytes. Cardiac surgery with CPB thus initiates the mobilisation of a sufficient number of cells of the specific immune system that are able to compensate the effects of haemodilution in the immediate postoperative period. In this context, the type of surgical procedure appears to play only a minor role since there were no significant differences between the groups (data not shown). For this reason, the entire prospectively recruited population of patients was included in the analysis. In addition, the patients were not compared with healthy controls because cellular immune function changes with ageing [12–15]. We therefore compared our measurement results with preoperative baseline values in order to document the effects of surgical trauma in this special patient population (elderly cardiac patients). Several studies investigated how the cells of the specific immune system change after cardiac surgery with CPB and reported contradictory findings [16,17,18,2,5,1]. They analysed either small patient populations or failed to include functional parameters such as cytokine synthesis. For this reason, the hypothesis of a TH1/TH2 shift is still under debate. Haemodilution effects or changes in cell concentrations are often considered responsible for immune system alterations. Our study, partially contradicts these findings. We found that the number of T lymphocytes increased immediately after surgery. The absolute counts of T lymphocytes and TH cells, however, then decreased to a minimum on d1. Incomplete lymphocyte activation [19,20] during extracorporeal circulation and the subsequent induction of anergy or apoptosis, as described by Pellegrini et al. for trauma patients and by other authors [9–11,21], appear to be unlikely in the case of the patients investigated in this study, since we found no T lymphocyte activation in the immediate postoperative period. This finding is in some contrast to other studies, which reported activation of lymphocytes after cardiac surgery with CPB [14,22,23]. Some differences in results may be attributed to the difference in methodology. In addition, our observation of a decrease in cytokine synthesis in the immediate postoperative period might indicate a mechanical alteration of lymphocytes and subsequent apoptosis induction as reported by Ankersmit et al. [21] for patients with a left-ventricular assist device (LAVD). Alternatively, it is possible that T lymphocytes decrease in number in the postoperative course because they home to the lymphoid tissues from where they had been released immediately after surgery. In this case, cytokine synthesis should be unchanged in stimulated whole blood cultures when cytokine concentrations are corrected to the absolute T cell count. In our study, this was true for IL-2, IL-5 and IL-4 but not for IFN-gamma. Further research is therefore needed to better understand why there is a decrease in T cells. The results of this study suggest that functional alteration appears to be a more likely cause than haemodilution or cell distribution effects. In spite of a decrease in the absolute count, TH cells are more strongly activated on d1 than at all other time points in the postoperative period. Accordingly, these cells are active and able at this early stage to release both IL-2 and IL-5. Assuming that there is a functional alteration of TH cells on dx as discussed above, it is attractive to relate the documented recovery of cytokine synthesis to newly generated TH lymphocytes. This is supported by the fact that IL-2 is released by precursor T lymphocytes. By contrast, IFN-gamma is secreted by more mature TH1 cells and its synthesis remains depressed. When cytokine levels are corrected to the preoperative absolute TH cell count, IL-2 and IL-4 return to baseline. This suggests an increased regeneration of TH cells. IFN-gamma release, however, remains reduced. We believe this finding indicates a TH1/TH2 shift since the release of IL-2 cannot be interpreted as indicating full activity of TH1 cells. Accordingly, there is a stronger activity of TH2 cells, which stimulate the proliferation of B lymphocytes and antibody synthesis by increasing IL-4 and IL-5 secretion and thus the serum levels of these cytokines [3]. We were able to document the effects of this increase in TH2 cell activity in a similar patient population both in vitro and in vivo [24]. Despite the significant changes that we were able to demonstrate, the design of our study entails a few problems that might limit the validity of our results. The patients in our study underwent different surgical procedures and some received red blood cell concentrates or medications (aprotinin) that may have an immunomodulating effect. The transferability of our results is therefore reduced. As described above, the system was primed with Biseko®. This is a 5% serum protein solution that is prepared from pooled human plasma. It contains no clotting factors but does have a wide variety of enzyme inhibitors. The extent to which this solution influences T cell function has not yet been investigated. We used it in our tests in order to inhibit the non-specific activity of released enzymes [25–28]. In addition, a population of only 40 patients limits the generalisability of the results. Likewise, we must admit that the determination of TH cell cytokine concentrations in the supernatant of stimulated whole blood cultures provides only a surrogate parameter for an evaluation of the function of the various cells. Our analysis of leukocyte and lymphocyte subpopulations in whole blood cultures nevertheless enabled us to accurately document relevant changes in cell counts. Despite the limitations of this study, our results allow us to draw the following conclusions. Both the number of immune cells of the specific immune system (lymphocytes) and the number of immune cells of the non-specific immune system (leukocytes, neutrophil granulocytes) are not reduced but are significantly elevated in the immediate postoperative period. At this stage, haemodilution has no detectable effect on the concentration of these cells. The function of specific immune cells, especially TH lymphocytes, is severely suppressed, beginning on d1. This is proven by a decreased release of TH1 cytokines (IL-2 and IFN-gamma) and TH2 cytokines (IL-4 and IL-5) in stimulated whole blood cultures. This functional impairment is apparently not preceded by T cell activation during CPB since there was no increase in HLA-DR expression on CD4+ or CD8+ lymphocytes in the immediate postoperative period. In spite of an increase in TH cell counts on the first postoperative day (d1), cytokine synthesis is reduced to a minimum. When cytokine synthesis is corrected to the preoperative absolute number of TH lymphocytes in culture, there is strong evidence of an increased TH2 cell activity in the postoperative period. This finding supports the hypothesis of a TH1/TH2 shift. The authors gratefully acknowledge the excellent technical assistance provided by B. Isenberg, L. Herter and Y. Wyss. This research was conducted with financial support from the Bundeswehr Medical Service. Absolute T lymphocyte, TH cell, T suppressor cell and natural killer cell counts (mean ± SEM) in whole blood samples obtained from study group patients preoperatively and on the first five postoperative days after elective cardiac surgery (d0: preoperatively; dx: immediately after surgery; d1: first postoperative day; d3: third postoperative day, d5: fifth postoperative day). The grey bars show preoperative values of the study group patients (mean ± SEM). p < 0.05 versus d0 and p < 0.001 versus d0. The results are expressed as mean ± standard error of the mean. Percentages of HLA-DR expressing CD4+ cells and HLA-DR expressing CD8+ cells as well as percentages of HLA-DR expressing CD4+ CD38+ cells (mature T helper cells) and HLA-DR expressing CD8+ CD38+ cells (mature T suppressor cells). The results are expressed as mean ± SEM. Whole blood samples were obtained from study group patients and analysed by flow cytometry preoperatively and on the first five postoperative days after elective cardiac surgery (d0: preoperatively; dx: immediately after surgery; d1: first postoperative day; d3: third postoperative day, d5: fifth postoperative day). The grey bars show preoperative values of the study group patients (mean ± SEM). p < 0.05 versus d0 and p < 0.001 versus d0. The results are expressed as mean ± standard error of the mean.