Introduction

Severe injury causes immune alterations with disturbed cellular and humeral immune response marked by a rapid decrease of human leukocyte antigen (HLA)-DR expression on monocyte as well as B- and T-cells [1, 2, 3]. The synthesis of cytokines such as tumour necrosis factor α (TNFα) or interleukin-6 (IL-6) after ex vivo lipopolysaccharide (LPS) stimulation is also down-modulated early after trauma [3, 4, 5]. Possibly, the anti-inflammatory effects of the cytokine interleukin-10 (IL-10) contribute to the downregulation of several immune functions after trauma, since this factor has been shown to be elevated after trauma [6]. In addition, IL-10 has been identified as factor decreasing HLA-DR surface expression on monocytes in patients with sepsis [7]. This hyporesponsiveness of the immune system after severe injury renders the traumatised patients susceptible to infectious challenges.

Surgical trauma causes similar changes of the immune response to those observed after accidental injury. Thus, decreased HLA-DR expression on monocytes have been observed after elective neurosurgery [8], cardiac surgery with cardiopulmonary bypass [9], after conventional but not laparoscopic cholecystectomy [10], and partial colectomy [11]. Surgical interventions, therefore, represent an additional challenge for the immune system that can superimpose on the trauma-induced disturbance of the immune response. Several studies demonstrated an inflammatory response caused by surgical trauma marked by elevated serum IL-6 and C-reactive protein (CRP) levels with a correlation between the degree of injury and the amount of circulating IL-6 [11, 12, 13].

The timing of secondary surgery after multiple injury can influence the incidence of multiple organ failure after trauma. A retrospective analysis of more than 4,000 severely injured patients revealed a higher incidence of multiple organ failure in patients sustained to surgery during day 2 and 4 than those with surgery after the 6th day following trauma [14]. Multiply injured patients operated on during a period of increased inflammatory response developed postoperative organ dysfunction more often than those with a normal homeostasis, as has been shown in a prospective study using CRP, elastase of polymorphonuclear granulocytes (PMN), and platelet counts as markers for the inflammatory response [14]. The role of secondary surgery after trauma as an additional burden on the immune system has been discussed in this and other papers. However, cellular immune functions after primary compared to secondary surgery after trauma have not been analysed in more detail yet. The treatment regime in the present study hospital consists of a primary damage control surgery and a definitive fracture stabilisation in the second or third week after trauma. The status of the immune response at the time of surgery and the impact of surgical intervention on the immune response in this phase after trauma have not been investigated. Therefore, in the present study we analysed HLA-DR expression on monocytes and ex vivo TNFα synthesis of blood cultures after trauma and after secondary surgery in multiply injured patients.

Material and Methods

Patients

This is a prospective study of severely injured patients with an Injury Severity Score (ISS) greater than 25 points, excluding predominantly blunt brain injury. The study group consisted of 16 consecutive patients with severe trauma admitted to the University Hospital of Essen, Department of Trauma Surgery, immediately after accident. Criteria for inclusion of patients were age (>18 and <80 years); ISS >25 points, and primary admission to the surgical intensive care unit within 8 h after accident. The specific threshold for the ISS was chosen, because a significant increase of posttraumatic multiple organ failure has been reported in patients with this degree of injury [15]. All of the patients’ medical histories were free of pre-existing immunological disorders, systemic steroid medication or known cancer disease. Sepsis was defined according to the consensus conference criteria published by Bone et al. [16]. Organ failure was evaluated on a daily basis applying the Sepsis-related Organ Failure assessment (SOFA)-Score [17]. Multiple organ failure (MOF) was defined as SOFA-Score >2 points of two organ systems persisting for more than 2 days. Blood from 12 healthy volunteers were collected at 0800 hours. The study and the additional blood sampling were approved by the ethic committee of the university of Essen, Germany.

Treatment protocol/surgical management

All patients received standardised emergency room management and intensive care treatment. After admission primary surgery was performed in the case of uncontrollable haemorrhage and for stabilisation of unstable fractures which was usually managed by external fixation. Spine injury were only primarily operated in the case of progressive neurologic deficit. As part of the general management at the study hospital definitive internal fracture stabilisation or plastic reconstruction were performed when platelet counts were above 80,000/mm3, CRP below 11 mg/dl or decreasing more than 20% within 2 days, and not more than one organ system with an SOFA-Score above 2 points, which was generally not the case before the 7th day after trauma. Almost all patients at the time point of surgery met these criteria except those with surgical focus clearance.

Classification of surgical interventions

Three different types of operation were defined: surgical intervention with the aim of clearing a septic focus (n=5), operation associated with a minor surgical trauma, such as osteosynthesis of peripheral joint fractures, secondary wound closures (n=9), and major surgery such as unreamed intramedullary nailing of long bone fractures or osteosynthesis of pelvic girdle and spine fractures (n=10).

Blood sampling

Arterial blood (heparinised blood samples, Sarstedt 5 ml and 10 ml heparin monovettes) was collected on the first day after hospital admission (day 1) and every Monday and Thursday after the accident at 08.00 hours until discharge from the intensive care unit. At the same time points serum was collected and stored at –20 °C until cytokine detection.

Whole blood stimulation

One millilitre of whole blood was diluted 1:1 with 1 ml RPMI 1640 medium containing 100 U/ml penicillin, and 100 µg/ml streptomycin, and glutamine and incubated in 12-well flat bottom tissue culture plates at 37 °C in an atmosphere of 5% CO2 in air. Blood cultures were set up as duplicates and incubated with 10 ng/ml lipopolysaccharide (LPS from Salmonella friedenau, kindly provided by H. Brade, Borstel, Germany) for 14 h before supernatants were collected after centrifugation for TNFα detection by means of enzyme-linked immunosorbent (ELISA).

ELISA

One hundred microlitres of the supernatant were used in the ELISA for TNFα detection. TNFα was determined using an ELISA purchased from Beckmann Coulter Company (Marseille, France). The lower detection limit was 15 pg/ml. IL-10 was detected in serum samples by means of ELISA. One hundred microlitres of serum sample were used for each detection. Each sample was analysed as duplicate. A commercially available ELISA from Beckmann Coulter Company (Marseille, France) was applied. The lower detection limit was 7.8 pg/ml.

Flow cytometry

In separate heparinised blood samples phenotyping of leukocytes was conducted by two-colour flow cytometry using whole blood lysis technique and monoclonal antibodies. Antibodies used were: phyctoerythrin conjugated (PE) CD14, FITC coupled CD45, and HLA-DR, PE/FITC conjugated paired isotype control antibodies (all antibodies Becton Dickinson, Heidelberg, Germany). To 20 µl of antibody pairs 100 µl of heparinised blood was added and incubated for 20 min. The erythrocytes were lysed with 2 ml FACS lysing solution (Becton Dickinson, Heidelberg, Germany), centrifuged, and washed twice with 2 ml PBS. Measurement of stained cells was performed with a FACscan (Becton Dickinson) counting 20,000 cells for each measurement. Flow cytometry defined the monocyte population using scatter characteristics and CD45/CD14 fluorescence staining. This monocyte population was analysed for HLA-DR expression, which is expressed on all monocytes. Flow cytometric analysis was conducted using a linear format to measure channel fluorescence intensities as numerals and to calculate mean fluorescence intensity (MFI) values. Paired isotype controls were run with each samples revealing always less than 2% unspecific binding. Ten thousand cells were computed in list mode and analysed using the FACScan research software (Becton Dickinson, Heidelberg, Germany). The results are expressed as HLA-DR molecules on the cell surface of monocytes, given as MFI-values.

Statistics

Values are expressed as boxplots with representation of median, 75th and 90th percentile or as mean±standard deviation as indicated in the figure legend. Pre- and postoperative values were compared by Wilcoxon test, and analysis of differences between trauma patients and healthy donors were evaluated by Mann Whitney U-test. For multiple group analysis the Wilcoxon test was applied which is independent of normality distribution of the data. All tests were calculated two-sided. Data analysis and statistics were performed using the Statistical Package for Social Sciences (SPSS) version 11.0 for Windows (SPSS Headquarters, Chicago, Ill., USA). The statistical procedure is indicated in the figure legend. A P-value <0.05 was considered as statistically significant.

Results

Severely injured patients primarily admitted to the local trauma department and the surgical intensive care unit were monitored regarding the immunologic status by measuring HLA-DR expression on monocytes by means of flow cytometry and the ex vivo TNFα synthesis of endotoxin-stimulated blood. Sixteen of these fulfilled the inclusion criteria with a mean ISS of 39±9 points and mean age of 46±17 years. Erythrocyte concentrates (2,220±852 ml) and fresh frozen plasma (4,018±2750 ml) were transfused within the first 48 h in these patients. Ninety-five point one percent developed sepsis and 52.9% a multiple organ failure (MOF) as defined in the method section. Two patients died due to persisting MOF, and a third died because of cardiac failure after secondary surgery. The patients’ data are summarised in Table 1.

Table 1 Patients’ data. Patients characteristics, clinical data, injury pattern according to the abbreviated injury scale (AIS), and Injury Severity Score (ISS) are expressed as mean±standard deviation. For definition of sepsis and multiple organ failure (MOF) see Methods section
Full size table

Whereas 62.5% of the patients required primary surgery during the first 6 h after trauma, 37.5% could be admitted to the surgical intensive care unit directly after the emergency room treatment and completion of the diagnostic work-up. Patients with or without primary surgery did not differ significantly in ISS (42±10 with primary surgery vs 35±7 without primary surgery). Secondary surgery consisted of unreamed intramedullary nailing of long bone fractures (n=5), osteosynthesis of the pelvic girdle (n=4), internal spine stabilisation (n=2), osteosynthesis of peripheral fractures (n=6), plastic reconstructive surgery (n=3),and septic focus clearance (n=5).

Healthy donors had a mean age of 36.6±13.8 (mean ± SD) years and included six males and six females. When compared to healthy donors the patients revealed significantly diminished HLA-DR expression on monocytes and ex vivo TNFα synthesis (Fig. 1). The HLA-DR expression on monocytes on the day after admittance to the intensive care unit was depressed to the same degree in patients with and those without surgery at the day of initial trauma (441±62 MFI HLA-DR with primary surgery vs 429±40 MFI HLA-DR without primary surgery). The TNFα-producing capacity of whole blood was lower in patients’ samples after primary surgery (234±285 with primary surgery vs 101±88 ng/ml TNFα without primary surgery ), although this difference did not reach level of significance.

Fig. 1A,B
figure 1

A HLA-DR expression on monocytes and B ex vivo TNFα synthesison day 1 after trauma and before surgery and HLA-DR expression on monocytes and ex vivo TNFα synthesis after secondary surgery. Whole blood cultures drawn from severely injured patients on the 1st day after trauma (day 1; n=16), 1–2 days before secondary surgery (pre-; 24 events in 16 patients), 1–2 days after secondary surgery (post-; 24 events in 16 patients), 3–4 days after secondary surgery (21 events in 16 patients), and from healthy donors (n=12) were in vitro stimulated with 10 ng/ml LPS for 20 h. LPS-induced TNFα synthesis is expressed in pg/ml in the supernatant (boxplots with median, 75th and 90th percentile). Dots represent values out of 90th percentile, extreme values are shown as numbers. In the same blood samples HLA-DR expression on monocytes was analysed by flow cytometry and is expressed as MFI (median, 75th and 90th percentile ). ** indicates statistical difference to healthy volunteers with a P-value<0.01 in a Mann Whitney U-test. ## indicates statistical difference to preoperative values with a P-value <0.01 (Wilcoxon test). && shows statistical difference between day 1 and preoperative samples with a P-value <0.01(Wilcoxon test)

Full size image

Definitive surgical treatment was generally performed after the second week after trauma, on average 17.5±10.9 days after trauma. Shortly before surgery the ex vivo TNFα synthesis was significantly higher than on day 1 after trauma and no longer significantly different from that of healthy donors. HLA-DR expression on monocyte before surgery was still lower than that of healthy donors, but significantly higher than on day 1 after injury (Fig. 1).

When performed in the second week after trauma surgical interventions (including all operations) did not cause a significant change in HLA-DR expression on monocytes or ex vivo LPS-induced TNFα synthesis (Fig. 1). Three to five days after secondary surgery the HLA-DR expression on monocytes significantly increased in comparison to preoperative values, indicating a normalisation of the immune function (Fig. 1A).

A more detailed analysis of the influence of secondary surgery on the immune system after trauma was performed by taking the character of the operation into account. Three categories of surgical intervention defined as described in the method section demonstrated different immunological response patterns. Surgical intervention performed to clear septic foci consisted of drainage of perineal abscess (2×), infected hematoma after acetabulum osteosynthesis, purulent sinusitis, and intraabdominal abscess after splenectomy. Focus clearance generally improved immune functions reflected by an increase in HLA-DR expression on monocytes within 1–3 days after surgery. Ex vivo TNFα synthesis of whole blood after LPS stimulation was also slightly increased after focus clearance (Fig. 2). Minor surgery had no influence on the immune system in the second week after trauma: HLA-DR expression on monocytes as well as ex vivo LPS-stimulated TNFα production even tended to rise towards normal levels despite of the surgical intervention (Fig. 3). In contrast, more traumatising major surgery significantly reduced HLA-DR expression on monocytes. Ex vivo TNFα synthesis 1–2 days after the operation was not significantly decreased (Fig. 4). However, these disturbances of the immune response were reversible, since both ex vivo TNFα-producing capacity and HLA-DR expression on monocytes reached preoperative levels within 3–5 days after surgery (Fig. 4).

Fig. 2
figure 2

HLA-DR expression on monocytes and ex vivo TNFα synthesis after septic focus clearance. Heparinised blood cultures were drawn from severely injured patients on the day before surgical septic focus clearance (five events in four patients) and 1–2 days after surgery. Data are expressed as median, 75th and 90th percentile. For details see Fig. 1. No statistical evaluation was performed because of small sample numbers

Full size image
Fig. 3
figure 3

HLA-DR expression on monocytes and ex vivo TNFα synthesis after minor secondary surgery. Heparinised blood cultures were drawn from severely injured patients on the day before minor secondary surgery (nine events in eight patients) and 1–2 days after surgery. Data are expressed as median, 75th and 90th percentile. For details see Fig. 1. No statistical difference to preoperative values was found (Wilcoxon test)

Full size image
Fig. 4
figure 4

HLA-DR expression on monocytes and ex vivo TNFα synthesis after major secondary surgery. Heparinised blood cultures were drawn from severely injured patients on the day before major secondary surgery (ten events in nine patients) and 1–2 days and 3–4 days (eight events in seven patients) after surgery. Data are expressed as median, 75th and 90th percentile. For details see Fig. 1. # indicates statistical difference to preoperative values with a P-value <0.05 (Wilcoxon test). Statistical difference between blood sample drawn 3–4 days after surgery and other time points were not found

Full size image

Serum IL-10 levels were found to be elevated initially after injury as well as at the time point of secondary surgery (Table 2). However, secondary surgery itself lowered serum IL-10 levels in parallel to a reduction in HLA-DR expression and TNFα-producing capacity (Table 2). The lowest IL-10 serum levels were found in patients after clearance of septic foci.

Table 2 Serum IL-10, ex vivo LPS-induced TNFα, and monocyte HLA-DR expression of human volunteers and patients at day 1 after trauma, preperative and postoperative. Serum IL-10 (pg/ml), ex vivo LPS-induced TNFα production (pg/ml), and HLA-DR expression on monocytes (MFI) of human volunteers and patients at day 1 after trauma, preoperative (pre), 1 –2 days postoperative (post), and 3 – 4 days postoperative (post2) are shown. Values are expressed as mean and standard deviation (SD) as well as median with 75th percentile in parenthesis. (n.d. below detection limit of 7.8 pg/ml)
Full size table

Discussion

The present study shows that major secondary surgery after severe trauma disturbs the immune response of these patients. HLA-DR expression on monocytes and the ex vivo TNFα synthesis of LPS-stimulated whole blood were used as marker for the status of the immune response. Both markers have been extensively studied in severely injured patients. The degree of depression of HLA-DR expression of monocytes and the ex vivo TNFα synthesis has been reported to correlate with the degree of injury [18, 19]. In patients with sepsis the same markers have a high negative prognostic value [20]. Low HLA-DR expression after major elective resectional surgery have been shown to be associated with the development of sepsis after surgery [21] underlining the importance of this marker of the immune function.

In this study, secondary surgery in severely injured patients was generally performed not before the second week after trauma. Primary surgery was only performed in a minimally traumatising manner as a damage control procedure. Since there was no difference in HLA-DR expression after admittance on the ICU between the patients with and those without primary surgery, one could speculate that the surgical stress of a damage control surgery in the initial phase after trauma does not further decrease immune functions. The TNFα-producing capacity tends to be lower, although not significantly, in patients with primary damage control surgery. Generally, there exists a growing body of evidence that underlines the impact of additional surgical stress in the initial treatment of severely inured patients. A large retrospective study with more than 4,000 severely injured patients revealed a higher incidence of posttraumatic multiple organ failure (MOF) in patients with secondary surgery on day 2–4 compared to patients with definitive surgical treatment on day 6–8 after trauma [14]. Other studies demonstrated a high predictive value of elevated CRP serum levels for multiple organ failure after secondary surgery in severely injured patients [22]. In our study we observed a still suppressed HLA-DR expression and ex vivo TNFα producing capacity at the time point of surgery even with a delayed treatment regime. A downmodulation of HLA-DR expression on monocytes for a period of 2 weeks after injury is in line with other studies of this marker in trauma patients [23]. At the time point of definitive surgical treatment minor surgery did not influence the analysed immune functions, whereas more traumatising surgery as intramedullary nailing still caused a depression of HLA-DR expression on monocytes and a reduction of the ex vivo LPS-induced TNFα synthesis. Corresponding surgery-induced changes of markers of the inflammatory response such as circulating IL-6, interleukin-8 (IL-8) and elastase have already been reported in trauma patients [24]. In this study with severely injured patients, operations on the pelvic girdle and femur caused higher levels of circulating IL-6, IL-8, and elastase compared to maxillofacial surgery, for example. Obviously, more traumatizing secondary surgery trauma in severely injured patients causes both a systemic inflammatory response marked by elevated systemic levels of, for example, IL-6 and a parallel anti-inflammation as HLA-DR expression or cytokine-producing capacity.

Suppression of cytokine-producing capacity after accidental and surgical trauma has been found to be associated with the appearance of a cytokine-inhibitory activity in the serum of these patients [4]. While some studies suggested the involvement of the anti-inflammatory activity mediators IL-10 or transforming growth factor (TGF) ß as the dominant cytokine-suppressive mechanism [25, 26], other studies could exclude IL-10 and TGFß as the only reason for the diminished cytokine-producing capacity after injury or surgery [27, 28]. Decrease HLA-DR surface expression on monocytes by serum from patients with sepsis has also been contributed to the action of IL-10 [7]. In the present study the decreased ex vivo TNFα-producing capacity and HLA-DR expression on monocytes after major secondary surgery was not accompanied by an increase in serum IL-10, which supports the involvement of additional mechanisms.

It remains to be elucidated whether this depression of immune functions represents a necessary counter regulatory mechanism or renders the patients more susceptible to infectious complications. Studies with immune modulating substances such as interferon γ, interleukin-12 (IL-12), granulocyte colony-stimulating factor (G-CSF) or granulocyte-macrophage colony-stimulating factor (GM-CSF) could answer these questions. Preoperative immune modulation with G-CSF in patients undergoing esophagectomy could reduce the rate of infections [29, 30]. A study with low-dose GM-CSF given perioperatively in patients operated on for primary colorectal carcinoma showed that this factor could counteract surgery-induced depression of HLA-DR expression on monocytes [11]. However, further studies with immune modulating substances and detailed analysis of immune functions need to be performed.

Surgery for the clearance of septic foci seems to improve immune functions rapidly. The reduction of the bacterial burden does not only support the immune system to definitively eliminate the remaining infectious agents, but also improves the immune response itself directly, which underlines the importance of a basic surgical principle from an immunologic point of view.

Summarising, we found that delayed secondary minor surgery in severely injured patients caused no change in immune functions, while major surgery suppresses the immune functions even if performed in stable patients 2 weeks after trauma. Septic focus clearance, as expected, improves the immune response in terms of HLA-DR expression and ex vivo TNFα synthesis.