Abstract
Objective: Due to the combination of local trauma, extracorporeal circulation (ECC), and pulmonary and myocardial reperfusion, cardiac surgery leads to substantial changes in the immune system and possibly to post-operative complications. Procedures without ECC, however, have failed to demonstrate clear advantages. We hypothesized that ECC is far less important in this context than the reperfusion/reventilation of the lung parenchyma and the surgical trauma. We therefore conducted a prospective observational study to compare immune reactions after cardiac operations with those after thoracic surgery. Methods: Serum levels of pro-inflammatory interleukin (IL)-6, IL-8, tumor necrosis factor (TNF)-alpha as well as C-reactive protein (CRP), lipoprotein-binding protein (LBP) and procalcitonin (PCT) were measured pre-operatively (d0), at the end of the operation (dx), 6h after the operation (dx+), on the 1st (d1), 3rd (d3), and 5th (d5) post-operative days in 108 patients (pts) undergoing elective coronary artery bypass grafting (CAB) with ECC (n=42, CPB CAB), off-pump coronary artery bypass surgery (n=24, OP CAB) without ECC or thoracic surgery (n=42, TS). Results: After cardiac surgery (CS), IL-6 and IL-8 increased and reached a maximum on dx+. IL-6 returned to baseline values at d3, whereas IL-8 remained elevated until d5. No difference was found between OP CAB and CPB CAB patients. In the TS patients, IL-6 increased later (dx+) and absolute levels were lower than in the CS patients. No increase in IL-8 was noted in the TS patients. Due to the high variation in the results obtained in all three groups, there was no significant change in TNF-alpha. A comparison of TS, OP CAB, and CPB CAB revealed that the CS patients had higher levels on d0, dx, d3, and d5. Serum levels of CRP, LBP, and IL-2R increased from dx+ to d5 in all groups and reached maximum values on d3. Whereas we found no difference in CRP and IL-2R between the groups, LBP levels were significantly higher from dx+ to d3 after OP CAB. PCT was elevated from dx+ to d3 in all pts. Similar levels were noted for the TS and OP CAB patients. The CPB CAB patients showed the highest levels. Conclusions: Surgical trauma and reperfusion injury appear to represent the predominant factors resulting in immunologic changes after cardiac surgery. Cardiopulmonary bypass (CPB) may be less important for immune response and acute-phase reactions than previously suspected. In addition, our data indicate a relationship between IL-6 synthesis and the degree of surgical trauma. IL-8 appears to be elevated only after cardiac surgery whereas PCT liberation depended on the use of ECC.
1 Introduction
It is well accepted that the combination of local trauma, CPB as well as pulmonary and myocardial reperfusion result in activating the innate and the adaptive immune system (IS). The extent of this immune reaction (IR) has been correlated to clinical and operative data as well as the incidence of post-operative complications [1].
In an attempt to reduce the immunological reaction, procedures avoiding cardiopulmonary bypass (e.g. off-pump coronary bypass surgery—OP CAB) or lung injury (e.g. Drew–Anderson technique) have been studied in recent years [2]. It was somehow surprising that these studies were unable to clearly demonstrate a diminished activation of the IS. Several investigations [3,4] indicate a decreased IR following OP CAB surgery. However, these differences in IR have not been shown to result in a clinically relevant benefit. In addition, several studies comparing different kinds of trauma mechanisms have shown different time courses of pro- and anti-inflammatory cytokine synthesis as well as acute-phase proteins. Taken together, these studies indicate that any substantial injury results in IS changes that may be specific to the type of injury, the organ system(s) injured or a combination of both. It was thus the objective of this study to assess how the IR is influenced by CPB on the one hand and the incision and reperfusion injuries on the other. For this purpose, pro-inflammatory cytokine synthesis (IL-6, IL-8, TNF-alpha), acute-phase reactions (CRP, LBP, PCT), and lymphocyte proliferation (IL-2R) were assessed in patients either undergoing CS via sternotomy for CAB with CPB (CPB CAB) or without CPB (OP CAB) and in patients undergoing thoracic surgery (TS) via thoracotomy. We hypothesized that these types of surgery result in similar immune reactions in patients with pulmonary reperfusion during reventilation and in (CPB CAB) patients with pulmonary and myocardial reperfusion injury and CPB-induced immune alterations.
2 Patients, materials and methods
2.1 Patients
The study, which was approved by the local ethics committee, included 108 patients (n=42 CPB CAB, n=24 OP CAB, n=42 TS). Only patients without known local or systemic infection or inflammation (fever, leucocytosis, tachycardia or tachypnea), immune or central nervous system dysfunctions as well as congestive heart failure, exogenous hormone therapy or malnutrition were enrolled in this study.
2.2 Clinical management
CPB CAB. We used the same cardiopulmonary bypass and perioperative management procedures described previously [5]. Anesthesia was induced with etomidate (200–300μg/kg) and fentanyl (20–30μg/kg). After muscle relaxation with pancuronium (100–200μg/kg) and endotracheal intubation, anesthesia 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 5000IU heparin, 1000mL Ringer's lactate, 500mL NaCl 0.9%, 500mL of a solution containing human serum protein (Biseko®, Biotest, Dreieich, Germany), and 250mL of a 15% mannitol solution (Osmofundin 15%®, Braun Melsungen, Melsungen, Germany). Heparin (300IU/kg) was administered just before vascular cannulation. After the institution of cardiopulmonary bypass at a flow rate of 2.4–3L/m2 per min, the aorta was cross-clamped and a cold crystalloid cardioplegic solution (St Thomas' Hospital solution or Bretschneider's cardioplegic solution) was injected. After CPB, protamine was infused. All patients received 2,000,000IU of aprotinin before the onset and 1,000,000IU at the end of CPB to preserve platelet function. Cefuroxime (3×1.5g) was given for perioperative antibiotic prophylaxis.
OP CAB. We used the same perioperative management procedures described for CPB CAB patients. This applies to the induction of anesthesia, endotracheal intubation and the maintenance of anesthesia.
Once the pericardium was opened, an initial heparin dose of 1mg/kg was administered. Intravenous heparin was then used to maintain an activated clotting time (ACT) of more than 250s until the anastomoses were created. An Octopus stabilizer system (Medtronic Inc., Minneapolis, MN) was used during the operation. After incision of the coronary artery, a stent was inserted and an anastomosis was performed. Following revascularization, the heparin effect was reversed with protamine sulphate (at a ratio of 1:1). Additional doses of protamine were administered until the ACT had returned to baseline or below. Only in the presence of significant intraoperative bleeding did we use a cell separator to collect and reinfuse blood during surgery. All patients received 2,000,000IU of aprotinin before the onset and 1,000,000IU at the end of CPB to preserve platelet function. Cefuroxime (3×1.5g) was given for perioperative antibiotic prophylaxis.
TS. In patients undergoing thoracic surgery, propofol (2–4mg/kg) and remifentanil (1–2μg/kg) were used to induce anesthesia. After muscle relaxation with pancuronium (100–200μg/kg) and one-lung ventilation using a double-lumen endotracheal tube, anesthesia was maintained by a continuous infusion of propofol (0.1–0.2mg/kg per min) and remifentanil (0.25μg/kg per min). All patients underwent posterolateral thoracotomy.
2.3 Sampling
Blood samples were taken pre-operatively (d0), directly (dx), 4–6h (dx+) after the operation, and on the 1st (d1), 3rd (d3) and 5th (d5) post-operative days at 8:00 a.m.
2.4 Parameters
IL-6, IL-8, and TNF-alpha were measured in order to assess the pro-inflammatory immune reaction. CRP, LBP, and PCT were used as parameters of the acute-phase response. Synthesis patterns of IL-2R were identified in order to study the proliferative response to lymphocyte activation and thus the adaptive specific immune response.
Pro-inflammatory cytokines (IL-6, IL-8, TNF-alpha), IL-2R and LBP were measured in serum samples using the commercially available Immulite® system (DPC-Biermann, Bad Nauheim, Germany), which is a luminometric immunoassay with a sensitivity of 5pg/mL for each parameter.
Serum CRP levels were measured using a nephelometer (Behring BN II Analyzer, Dade Behring, Marburg, Germany) and the N High Sensitivity Reagent Kit (sensitivity: 0.175mg/L, Dade Behring, Marburg, Germany).
Serum PCT concentrations were assayed using the commercially available BeriLux® Analyzer 250 system (Dade Behring, Marburg, Germany), which is an immunoluminometric assay (PCT LUMItest®, Brahms AG, Hennigsdorf, Germany) with a sensitivity of 0.1ng/mL.
The results were not corrected for hemodilution, since we intended to document in vivo levels. In addition, organ function and thus the occurrence of post-operative complications appear to depend first and foremost on actual effective concentrations. It should also be borne in mind that the perioperative use of packed red cells or human albumin considerably limits the possibility to correct serum levels on the basis of hemoglobin or total protein levels.
2.5 Statistical analysis
Data were collected and analyzed using standard computer software (Statview 5.0, Abacus concepts, Berkeley, CA). The analysis of variance (ANOVA) test was used for intragroup (time points) and intergroup (surgical procedures) comparisons. When the ANOVA test revealed significant group differences, Fisher's PLSD was used for pairwise comparisons. If ANOVA assumptions were violated, unpaired Student's t-tests were used for intragroup or intergroup comparisons. A P-value of less than 0.05 was considered significant. The results are expressed as mean±standard error of the mean, unless otherwise indicated. Since we used relatively long measurement intervals, neither Cmax over time nor the area under the curve was included in the analysis. In our opinion, accurate and comparable data regarding these measures require shorter measurement intervals.
3 Results
3.1 Clinical results
Operative and post-operative data are provided in Tables 1 and 22. The thoracic surgery patients were younger than the CS patients. It should also be noted that the CPB CAB patients showed a significantly lower pre-operative left ventricular ejection fraction than the OP CAB patients. Thoracic surgery was associated with shorter operating time and less hemodilution. In addition, the thoracic surgery patients did not require packed red cell transfusions. The OP CAB patients received significantly less transfusions than the CPB CAB patients. On average, our patients were discharged from the intensive care unit on the first post-operative day despite the aforementioned differences. There was no operative mortality, and no patient experienced serious post-operative complications.
OP CAB (19♂, 5♀). Of the 24 patients who underwent elective cardiac surgery without CPB, 23 had one-vessel or two-vessel coronary artery bypass grafting (CABG) using the left internal mammary artery (LIMA). Three distal anastomoses were performed in one patient, two were made in nine patients and one distal anastomosis was performed in 13 patients. One of the 24 patients had three bypasses.
CPB CAB (24♂, 18♀). Of the 42 patients who underwent elective cardiac surgery with CPB, 27 had CABG using the left internal mammary artery. Six of these patients received only a LIMA graft. Nine patients received a LIMA bypass and an autologous vein graft. Three distal anastomoses were performed in seven patients. Five patients received four bypasses. Due to insufficient arterial flow in the LIMA, 5 patients received only autologous vein grafts (one distal anastomosis in four cases, two distal anastomoses in eight cases and three distal anastomoses in three cases).
TS (34♂, 8♀). Twenty patients underwent lobectomy. These patients presented with a malignant tumor (adenocarcinoma: n=5, squamous carcinoma: n=10, carcinoid tumor: n=1 and metastatic disease: n=2) or required partial removal of a lung after inflammatory disease (tuberculosis: n=1, bronchiectasia: n=1). Wedge resection was performed in 22 cases on the basis of the following indications: metastatic disease (n=4), squamous carcinoma (extended disease) (n=3), granuloma/hamartoma (n=9), lymphoma (n=2), actinomycosis (n=1), pericardial cyst (n=1), and emphysema (n=2).
3.2 Pro-inflammatory cytokines (IL-6, IL-8, TNF-alpha)
Cardiac surgery. Serum levels of IL-6 and IL-8 increased significantly on dx and dx+. IL-6 reached a maximum on dx+ and remained elevated until d1. IL-8 peaked earlier (on dx) and returned towards pre-operative levels on d1. These values remained elevated on d3 and d5, but only the OP CAB patients showed values that were significantly increased during the entire post-operative period. TNF-alpha started from elevated levels and increased further on dx. It then decreased on dx+ and increased again on d3. None of these changes reached statistical significance in the CPB CAB group, as far as interindividual differences were concerned. Due to lower baseline values in the OP CAB group, these patients reached significantly elevated levels on dx, dx+, and d5 (Fig. 1).
OP CAB versus CPB CAB. The same courses were observed for IL-6, IL-8, and TNF-alpha values. We found no significant differences in post-operative pro-inflammatory cytokines between these two study groups. The CPB CAB patients, however, showed significantly higher pre-operative values than the OP CAB patients (Fig. 1).
Thoracic surgery. IL-6 increased on dx+ and d1. At all other points in time as well as for IL-8, we were unable to detect significant changes from pre-operative levels during the investigation period. TNF-alpha levels increased to higher levels on dx and dx+ (Fig. 1).
Thoracic surgery versus cardiac surgery. IL-6 levels following CS were higher on dx and dx+. In contrast to the unchanged synthesis pattern of IL-8 following TS, post-operative IL-8 levels were higher in the OP CAB and CPB CAB patients until d5. The CS patients had elevated TNF-alpha levels on d0. In the post-operative follow-up, the CPB CAB patients presented higher TNF-alpha levels than the TS patients on dx, d3, and d5 (Fig. 1).
3.3 Acute-phase proteins/parameters (CRP, LBP, PCT)
Cardiac surgery. Elevated CRP and LBP serum levels were observed from dx+ to d5. Maximum values were reached on d3. PCT serum levels increased from dx+ to a maximum on d1 and pre-operative values were reached on d5 (Fig. 2).
OP CAB versus CPB CAB. We found no significant difference in CRP levels between these two study groups. The OP CAB patients showed significantly higher LBP levels from dx to d3. By contrast, the PCT values in the CPB CAB group were higher on dx+ and d1 (Fig. 2).
Thoracic surgery. Serum levels of CRP, LBP, and PCT following TS were similar to the synthesis patterns described for these mediators after CS (Fig. 2).
Thoracic surgery versus cardiac surgery. The TS patients presented higher pre-operative CRP and LBP serum levels. A comparison of TS and CPB CAB patients showed no significant differences. The only exception was on dx when the CPB CAB patients had significantly lower CRP and LBP levels. From dx+ to d3, the OP CAB group showed higher levels than the TS patients. There were no differences in PCT release between the two groups. The highest levels were measured for CPB CAB patients on dx+ and d1 (Fig. 2).
3.4 Lymphocyte proliferation following activation (IL-2R)
Cardiac surgery. Serum levels of IL-2R increased significantly on d1, peaked on d3 and remained elevated until d5 (Fig. 3).
OP CAB versus CPB CAB. A direct comparison between these two groups showed that the OP CAB patients had higher levels than the CPB CAB patients on d0, dx, and d1.
Thoracic surgery. Post-operative IL-2R levels increased on d1 and peaked on d5 (Fig. 3).
Thoracic surgery versus cardiac surgery. The TS patients had higher IL-2R levels on dx than the CPB CAB patients. At the later time points, there were no significant differences (Fig. 3).
4 Discussion
We hypothesized that different kinds of cardio-thoracic surgical procedures with or without ECC induce similar post-operative immune responses. We therefore conducted a prospective observational study and compared a number of parameters of the pro-inflammatory response, acute-phase reactions and the specific immunity in patients who underwent elective CAB with or without CPB and in patients who underwent elective thoracic surgery for a non-specific mass.
Our study clearly demonstrated that post-CS IR was first and foremost influenced by the surgical trauma. IL-6, IL-8, CRP, and IL-2R showed similar synthesis patterns both in the presence or absence of CPB. In addition, we found that absolute levels of acute-phase proteins (CRP, LBP) and specific immune function (IL-2R) were as high after TS as they were after CPB CAB, although the pro-inflammatory stimulus (IL-6) was weaker in the latter group and there was no increase in IL-8 synthesis. At the same time, PCT values were considerably increased in CPB CAB patients. How then should we interpret these findings?
The IL-6 levels that began to increase on dx in all study groups are evidence of a strong pro-inflammatory response that peaked on dx+. There were no differences between the two cardiac surgery groups. This contradicts other studies suggesting that the myocardium is one of the major sources of post-CS IL-6 after cardioplegic arrest and positively correlated IL-6 levels with clinical and operative data as well as with the incidence of post-operative complications [6,7]. We found that the increase in post-TS IL-6 levels occurred later but was marked as well. The delay in post-TS IL-6 release might be attributable to the different anesthetic procedures [8]. Whereas propofol was given to the TS patients, midazolam was used as a hypnotic agent for the CS patients. Galley et al. [9], however, were able to show that midazolam and propofol have similar inhibitory effects on the synthesis of pro-inflammatory cytokines. In contrast to studies focusing on CS, our data appear to indicate that the surgical approach and the lungs participate in and contribute to the acute immune response (IR) via IL-6 liberation, whereas ECC plays only a minor role in IL-6 secretion.
Unlike IL-6, serum IL-8 was found to increase only after cardiac surgery. IL-8 levels were as high in the OP CAB group as they were in the CPB CAB group. They peaked on dx+ and remained slightly elevated until d5. IL-8 plays a major role in the recruitment and activation of leucocytes during post-operative (multi-)organ dysfunction syndrome (MODS) [6,10]. The selective increase in IL-8 following CS with or without CPB thus requires further attention. Several studies have focused on IL-8 synthesis in coronary sinus blood or arterial blood samples after CS with or without CPB [11,12]. Due to different findings, it remained unclear whether the myocardium or the lungs were the main source of IL-8 secretion. Our data indicate that the latter hypothesis is unlikely and that an activation of immunocompetent cells synthesizing IL-8 during ECC—as shown by Sbrana et al. [13,14] and Naldini et al.—is unlikely to be the cause of elevated IL-8 levels following CPB CAB.
TNF-alpha, which is an initiator of the IR to surgical trauma 2, showed higher pre-operative values in the CS group than in the TS group, but decreased after CS. Our results are in line with studies showing that the gene expression of TNF-alpha and its receptors is increased during ischemic heart failure [15,16]. This could be the cause of the pre-operative increase in TNF-alpha levels especially in the CPB CAB group since these patients had a significantly lower left ventricular ejection fraction and since the TS patients were younger and showed no clinical signs of coronary heart disease or cardiac insufficiency (ischemic heart disease). There was, however, wide interindividual variation in post-operative TNF-alpha serum levels in all groups of patients. It is therefore impossible to draw any conclusions on the basis of the TNF-alpha results.
Irrespective of qualitative differences, the post-operative pro-inflammatory IR described above gives evidence of a uniform acute-phase reaction and specific immune reaction. This is confirmed by the CRP and IL-2R levels, which were elevated in all three groups of patients from d1 to d5 and showed a peak on d3. This was surprising considering that the gene expression of CRP and LBP in hepatocytes is increased by IL-6 [17,18] and therefore should have led to different courses as documented for IL-6. Furthermore, the TS patients showed higher pre-operative levels of CRP and LBP. Although there were no clinical signs of infection in this group prior to surgery, the histologically confirmed diagnoses appear to be associated with an increased expression of CRP. This limits the comparability of the study groups.
It was also interesting to note that the CPB CAB patients showed a decrease in CRP and LBP levels immediately after surgery. This could be the result of protein adsorption during ECC since a direct comparison of OP CAB and CPB CAB suggests that hemodilution-induced effects are unlikely.
Unlike CRP, LBP acts as a cofactor that facilitates the binding of LPS to monocytes in the presence of sCD14 [19]. It is possible that this interaction is the reason for the significant decrease in post-operative LBP levels in the CPB CAB group. Whereas the CRP and LBP synthesis patterns were similar at the first time points examined, we found an increase in LBP levels in the OP CAB group on d1 and d3. It is possible that the release of LPS during ECC is associated with a decrease in LBP, which suggests lower levels in the CPB CAB group. Our data, however, neither prove nor disprove this hypothesis.
The differences in PCT synthesis may be the result of bacterial translocation from the gastrointestinal tract and the associated release of LPS. PCT is becoming more and more clinically relevant because it is known to be markedly elevated in the serum of patients under stressful conditions (infection, sepsis, burns, etc.) [6,7,20,21]. Its cellular origin or possible PCT-inducing stimulatory factors have not yet been completely clarified [22]. In our study, PCT levels were elevated post-operatively in all study groups. The highest levels were observed in the CPB CAB group. If we bear in mind that IL-6 and IL-8 after CS demonstrated an earlier as well as higher increase than after TS, it is attractive to speculate that PCT synthesis might be related to absolute levels of IL-6 or IL-8. In addition, the fact that the CPB CAB group showed higher post-operative levels than the OP CAB group suggests a selective increase of PCT by ECC. This hypothesis is corroborated by other studies [6,7,20,21] showing comparable results.
5 Limitations of the study
The small number of patients may be the reason why some results showed no statistical significance. Nevertheless, our findings were rather homogeneous and uniform. We therefore do not know whether a larger patient population would have significantly changed our results.
The anesthetic agents that were used for the induction and maintenance of anesthesia in the cardiac surgery groups were different from those used in the thoracic surgery group and may have caused different effects on cytokine release.
The serum levels of the parameters studied were not corrected for hemodilution since we intended to document in vivo levels. In addition, perioperative hemodilution was similar in all CS patients and transfusion requirements were low. Hemodilution therefore appears to play only a minor role in our study.
Our patients showed an uneventful clinical course and needed neither additional therapy nor additional surveillance. It would therefore be useful to conduct a study on patients with post-operative complications on the basis of clearly defined inclusion criteria.
6 Summary
In summary, we could add the following aspects to the current understanding of mediators and the immune response in patients who underwent different kinds of elective cardio-thoracic surgical procedures with an uneventful clinical course:
CRP and LBP do not depend on the type of trauma and represent a non-specific, late acute-phase response. These findings are paralleled by the synthesis pattern of IL-2R. Pre-operative inflammation can influence these parameters, which seem to react similarly to different kinds of cardiac and cardio-thoracic surgical procedures.
IL-6 appears to represent a marker of the degree and duration of local and systemic surgical trauma and is unrelated to the presence or absence of ECC.
Since IL-8 was found to be selectively elevated after both kinds of CS, the myocardium and not the lungs appears to be the major source of IL-8 liberation after CS.
PCT showed fast changes in synthesis patterns after CS and was strongly related to the presence of CPB. The hypothesis that bacterial translocation or direct stimulation of PCT-liberating cells during ECC may be the major source of PCT should be tested in further studies.
(a–c) Levels of the pro-inflammatory cytokines (displayed as mean±SEM) IL-6, IL-8, and TNF-alpha in serum samples of patients undergoing cardiac surgery with CPB (n=42; CPB CAB), without CPB (n=24; OP CAB) or thoracic surgery (n=42; TS) during the first five post-operative days (d0—before the operation; dx—at the end of the operation; dx+—4–6h after the operation; d1—first post-operative day; d3—third post-operative day; d5—fifth post-operative day). ( indicates P≪0.05 versus d0; indicates P≪0.001 versus d0; in intergroup comparisons at the various time points * indicates P≪0.05 and ** indicates P≪0.001).
(a–c) Levels of the acute-phase proteins (displayed as mean±SEM) CRP, LBP, and PCT in serum samples of patients undergoing cardiac surgery with CPB (n=42; CPB CAB), without CPB (n=24; OP CAB) or thoracic surgery (n=42; TS) during the first five post-operative days (d0—before the operation; dx—at the end of the operation; dx+—4–6h after the operation; d1—first post-operative day; d3—third post-operative day; d5—fifth post-operative day). ( indicates P≪0.05 versus d0, indicates P≪0.001 versus d0; in intergroup comparisons at the various time points * indicates P≪0.05 and ** indicates P≪0.001).
Levels of IL-2R (displayed as mean±SEM) in serum samples of patients undergoing cardiac surgery with CPB (n=42; CPB CAB), without CPB (n=24; OP CAB) or thoracic surgery (n=42; TS) during the first five post-operative days (d0—before the operation; dx—at the end of the operation; dx+—4–6h after the operation; d1—first post-operative day; d3—third post-operative day; d5—fifth post-operative day). ( indicates P≪0.05 versus d0, indicates P≪0.001 versus d0; in intergroup comparisons at the various time points * indicates P≪0.05 and ** indicates P≪0.001).
Clinical data of patients
Operative data of patients
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.
