Functional immune monitoring in severely injured patients—A pilot study

After severe trauma, the resulting excessive inflammatory response is countered by compensatory anti‐inflammatory mechanisms. The systemic inflammatory response to trauma enhanced by inappropriately timed surgical second hits may be detrimental for the patient. On the other hand, overwhelming anti‐inflammatory mechanisms may put patients at increased risk from secondary local and systemic infections. The ensuing sepsis and organ dysfunction due to immune dysregulation remain the leading causes of death after injury. To date, there are no clinically applicable techniques to monitor the pro‐/anti‐inflammatory immune status of the patients and the remaining ability to react to microbial stimuli. Therefore, in the present study, we used a highly standardized and easy‐to‐use system to draw peripheral whole blood from polytraumatized patients (ISS ≥ 32, n = 7) and to challenge it with bacterial lipopolysaccharide. Secreted cytokines were compared with those in samples from healthy volunteers. We observed a significant decrease in the release of monocyte‐derived mediators. Surprisingly, we detected stable or even increased concentrations of cytokines related to T cell maturation and function. For clinical practicability, we reduced the incubation time before supernatants were collected. Even after an abbreviated stimulation period, a stable release of almost all analysed parameters in patient blood could be detected. In conclusion, the data are indicative of a clinically well‐applicable approach to monitor the immune status in severely injured patients in a short time. This may be used to optimize the timing of necessary surgical interventions to avoid a boost of proinflammation and reduce risk of secondary infections.


| INTRODUCTION
Severe injury and the subsequent development of a systemic inflammatory response syndrome (SIRS) and sepsis often lead to multiple organ dysfunction and failure, resulting in a substantial morbidity in trauma victims. [1][2][3][4][5] It is a long-known paradox that excessive immune activation due to vast amounts of endogenous cellular components released after massive tissue trauma and exogenous pathogen-derived danger molecules can be associated with a compensatory anti-inflammatory response which may render the patient susceptible to infections. [6][7][8][9] The ensuing sepsis and organ dysfunction due to immune dysregulation remain the leading causes of death after injury. To date, there are no clinically applicable techniques to monitor the pro-/anti-inflammatory immune status of the patients and the remaining ability to react to microbial stimuli. Therefore, in the present study, we used a highly standardized and easy-to-use system to draw peripheral whole blood from polytraumatized patients (ISS ≥ 32, n = 7) and to challenge it with bacterial lipopolysaccharide. Secreted cytokines were compared with those in samples from healthy volunteers. We observed a significant decrease in the release of monocytederived mediators. Surprisingly, we detected stable or even increased concentrations of cytokines related to T cell maturation and function. For clinical practicability, we reduced the incubation time before supernatants were collected. Even after an abbreviated stimulation period, a stable release of almost all analysed parameters in patient blood could be detected. In conclusion, the data are indicative of a clinically well-applicable approach to monitor the immune status in severely injured patients in a short time. This may be used to optimize the timing of necessary surgical interventions to avoid a boost of proinflammation and reduce risk of secondary infections. Therefore, close monitoring of the immune function and the eventually occurring alterations in immune status after trauma, possibly also as a remodelling process due to local hyperfunction of the tissue-associated immune system, would be crucial in order to avoid opportunistic infections that can lead to septic complications. 3,5,10 Currently, in contrast to specific parameters that allow monitoring of organ function, the state of the immune system can only be controlled in a static, snapshot-like manner, not allowing conclusions on remaining potency to react to infectious entities.
In this regard, it also seems vital to use a standardized stimulus in order to monitor and assess immune reaction. Interleukin (IL)-6, a highly popular biomarker and mediator of inflammation, is released in response to more than 20 damage-and pathogen-associated molecular patterns 11,12 that are often present in the patients' circulation in varying amounts after trauma. This underlines the importance of employing one defined molecule as stimulating agent ex vivo in order to master the complex immune response. The resulting information could be of great use especially during the critical early phase after injury when it is imperative to carefully plan necessary surgical interventions which represent 'second hits'. 13,14 Such an additional surgical 'trauma load' could be applied at a time of a balanced immune status in the trauma patient, to prevent either post-surgery infection or enhanced SIRS. Therefore, it is essential to develop fast testing methods; the so far recommended protocol of stimulating whole blood for at least 24 hours does not allow a close bedside monitoring. For future transfer to widespread clinical application, incubation times will have to be reduced markedly to get results within hours.
The present study was performed to find out whether (a) a whole blood ex vivo system can be used to monitor the proand anti-inflammatory immune responsiveness of trauma patients, and (b) whether a 24 hours standard incubation time of this ex vivo system can be reduced to a 4 hours incubation period without loss of discriminatory power.

| Clinical study
The clinical study was performed at the University Hospital Ulm in accordance with the Declaration of Helsinki and its later modifications. The study protocol was approved by the local Ethics Committee of the University of Ulm, approval number 244/11 (ClinicalTrials.gov identifier: NCT00710411; https ://clini caltr ials.gov/ct2/show/NCT00 710411). Inclusion criteria were an Injury Severity Score (ISS) ≥32 (estimated after whole-body computed tomography in the emergency room and checked retrospectively) and age ≥18 years; patients were excluded if one or more of the following criteria were fulfilled: life expectancy <24 hours, participation in other trials, cardiopulmonary reanimation on the accident scene or dying immediately after hospital admission, known or suspected pregnancy, radio-or chemotherapy within the last 3 months. Healthy volunteers served as controls. Patients and healthy volunteers were included between 15 March 2014 and 15 December 2014. All samples were collected with informed written consent of patients or their legal representatives and volunteers.

| Samples and assays
Polytrauma patients were analysed 4 hours, 24 hours and 5 days after admission. 1 mL of venous blood was drawn into TruCulture ® blood collection tubes prefilled by the manufacturer under standardized conditions with 2 mL medium, with unfractionated heparin as an anticoagulant at a final concentration of 50 IU/mL and with or without 100 ng/mL lipopolysaccharide (LPS, from Escherichia coli, O55:B5; HOT Screen GmbH). Using these tubes allow direct blood withdrawal into the culture system without any pipetting steps by the clinician which strongly reduces the risk of contamination and intra-individual variation. 15 Tubes were incubated at 37°C for 24 hours; in order to analyse whether a shorter incubation time with better clinical applicability was sufficient for considerable cytokine production, a second set of tubes ±LPS taken at the second time point (24 hours after trauma) was incubated for only 4 hours. After incubation, the cellular sediment was separated from the supernatant using a valve, and supernatants were stored at −80°C until further analysis. After inclusion of all patients and volunteers, samples were analysed using a multiplexed sandwich immunoassay on a Bio-Plex ® 200 platform (both BioRad) according to the manufacturer's recommendations. Since most samples were out of range for the central pro-inflammatory marker IL-6, they were re-analysed using the IL-6 Quantikine kit (BD Biosciences).

| Statistics
Results after LPS stimulation were compared using one-way analysis of variance followed by Student-Newman-Keuls post hoc test. SigmaPlot (Version 11.0, Systat Software) was used as analytic software. For the comparison of incubation for 24 hours and 4 hours, stimulated and unstimulated samples were compared using paired t testing. Pearson Product Moment Correlation was employed to assess associations between cytokine production upon LPS stimulus with clinical parameters. A P < .05 was considered as significant.

| Patients
Seven patients with a median ISS of 41 (34-47) and five healthy volunteers were included in the study. The median Glasgow Coma Scale at the time of hospital admittance was three. A median of two units of blood were transfused on the first day after injury. Further clinical parameters are summarized in Table 1; individual patient characteristics are included in Table S1.

| Endogenous cytokine secretion
In contrast to healthy volunteers who did not have detectable endogenous IL-6 concentrations, polytrauma patients displayed IL-6 concentrations of 21 pg/mL (12-66 pg/mL) 4 hours after trauma which increased to around 274 pg/mL (40-866 pg/mL) and 39 pg/mL (0-1054 pg/mL) 24 hours and 5 days after injury, respectively ( Figure 1A). Similarly, endogenous secretion of the early pro-inflammatory cytokines tumour necrosis factor and IL-1β, produced mainly by activated monocytes and macrophages, was elevated in unstimulated samples from patients especially 24 hours and 5 days after trauma ( Figure 1B,C).

| Central pro-inflammatory cytokines with unaltered or decreased LPS response after trauma
In order to evaluate how the reaction to a defined endotoxin stimulus was altered after severe trauma, peripheral whole blood was stimulated with LPS. Blood samples from healthy volunteers and trauma patients displayed highly increased IL-6 concentrations in response to the LPS stimulus, but the amounts secreted by leucocytes after trauma did not differ significantly from those of the healthy controls at all time points ( Figure 1A). Regarding tumour necrosis factor and IL-1β, patients displayed significantly lower concentrations in response to LPS compared to the healthy individuals ( Figure 1B,C). Interestingly, the same was true for the IL-1β/IL-1RA ratio; all patients displayed a significantly decreased ratio after LPS stimulus compared to healthy controls ( Figure 1D). Furthermore, TNF concentrations secreted in response to LPS (the difference between stimulated and unstimulated sample) in patient blood drawn 4 hours after injury correlated with the length of stay on the intensive care unit (Pearson correlation coefficient r = .954, P = .0031).

| Mediators unaltered by polytrauma after LPS challenge
Endogenous concentrations of interferon-γ, an important activator of macrophages secreted by lymphocytes, IL-10 produced by monocytes and lymphocytes, IL-12, secreted by lymphoid and myeloid cells, IL-17 (T helper cells), eotaxin (also known as eosinophil chemotactic protein), basic fibroblast growth factor, an inducer of angiogenesis, and granulocyte-macrophage colony-stimulating factor as the central regulator of myeloid proliferation and maturation were all visibly higher in patients especially 24 hours and 5 days after trauma. However, the amounts released by leucocytes from patients after severe trauma in response to LPS incubation remained very similar compared to healthy volunteers (Figure 2A-E, G, H). Endogenous granulocyte colony-stimulating factor as a specific activator of neutrophil proliferation remained largely unaffected by trauma and concentrations produced after LPS stimulus were even slightly decreased in patients ( Figure 2F).

| Cytokines increased after trauma and LPS stimulus
When comparing patients with healthy volunteers, secretion of several cytokines was significantly increased upon stimulus, especially 24 hours after trauma. Concentrations of IL-2, IL-4, IL-5 and IL-9, all secreted mostly by T cells, as well as IL-7, a stimulator of lymphocyte maturation, were unaltered or only slightly elevated 4 hours after injury compared to controls, but increased significantly in blood drawn 24 hours after polytrauma. Concentrations in stimulated blood from patients remained high until day 5 ( Figure 3A-E). IL-13, also secreted mainly by T helper cells, did not change over time ( Figure 3F). Endogenous levels of the inducer of natural killer cell proliferation IL-15 were increased in some patients 24 hours and 5 days after injury, and the secretion in response to endotoxin challenge was significantly higher in PT patients at all time points ( Figure 3G). Similarly, monocyte chemoattractant protein 1 was higher in some patients after 24 hours and 5 days, and the secretion after LPS incubation was significantly elevated 4 hours and 24 hours after trauma ( Figure 3H).

| Shortened incubation time to improve clinical applicability
In order to test whether blood stimulation for 24 hours is necessary to detect the response to LPS stimulation, we drew a second set of tubes from patients 24 hours after trauma and performed a shortened incubation time of 4 hours before supernatants were collected. As shown in Figure 4, the median concentrations of secreted cytokines in response to LPS were partly lower, but clearly detectable and showed a pattern greatly similar to the results after incubation for 24 hours. A comparison of stimulated and unstimulated samples after different incubation is shown in Table S2.

| DISCUSSION
This study was performed with two aims: (a) to define the alterations in immune reaction to a defined microbial stimulus after severe trauma, and (b) to test the applicability of a standardized whole blood model for clinical monitoring of the immune function. To our knowledge, this is the first study performing standardized functional immune monitoring in polytraumatized patients using whole blood to reflect We were able to confirm that pro-inflammatory cytokines predominantly secreted by monocytes and macrophages, but also by lymphocytes, are reduced when whole blood is challenged with a pathogen-associated molecular pattern early after severe trauma. 3,4,9,[16][17][18] However, it was surprising that 24 hours after injury, the TNF response was not as profoundly impaired as at the other time points. Whether these alterations are due to systemic suppression of the inflammatory response or rather represent a remodelling process simultaneous to tissue-specific hyperactivation of immune cells remains an open question. In contrast to previous publications and a recently published study in septic patients, 19 we did not observe an inhibition of T cell function. In our cohort, cytokines secreted by or involved in the activation of T cells, such as interferon-γ, IL-2, IL-4 and IL-9 were either unaltered or even increased in blood from polytrauma patients. Endogenous production of IL-10 with its dual anti-and pro-inflammatory features was also increased, but amounts secreted after LPS challenge were comparable to healthy controls.
The present study has some limitations. The use of heparin as an anticoagulant may have interfered with complement activation by thrombin. 20,21 However, the high amounts of released inflammatory mediators suggest that this inhibition is only of minor importance, if at all, in the employed model. Furthermore, cytokine release is a biomarker of immune responsiveness; other aspects of cellular immune function such as chemotactic and phagocytic activity which may also have profound effects on organ dysfunction and ability to clear infection are not captured using this technique. As a major limitation, we were only able to include seven patients during a 9-month period due to the high preset ISS limit since our study was aimed at most severely injured patients. In addition, the assays used are currently associated with a high cost/ sample ratio which limited extensive testing such as adding a shorter incubation time of samples taken on the first day after injury. We were also unable to assess the extent of subtle, but clinically meaningful alterations in T cell function. 9,[22][23][24] It was all the more unexpected that despite the small cohort, we were able to detect several significant differences between the groups.
Several previous studies have investigated the influence of shock, (surgical) trauma and sepsis on the ability of peripheral leucocytes to produce pro-inflammatory mediators upon LPS stimulus. 18,[25][26][27][28][29][30][31][32][33][34] However, those studies either used isolated cells, disregarding the role of fluid-phase mediators, granulocytes, and platelets, or employed stimulation protocols which are not feasible in clinical routine. Using our approach has the advantage of stimulation under highly standardized conditions and rapid assessment of the complex immune response in whole blood. A larger patient cohort including those with less severe injuries as planned by the REALISM study group (NCT02638779) may allow the identification of the markers with best sensitivity and specificity. However, the group is performing stimulation of whole blood for 24 hours 35 ; considering the substantial amounts of mediators we were able to detect after 4 hours of incubation, shorter incubation times are both feasible and advisable. This would enable a bedside monitoring similar to blood gas analysis and monitoring of the coagulation system during surgeries and in intensive care units. Shorter stimulation and clinical routine point-of-care cytokine measurements of a small number of parameters with established thresholds could offer a protocol to determine within less than 24 hours whether secondary surgery is recommended or not. Apart from trauma victims, the described approach may also be highly useful for the surveillance of sepsis or other intensive care patients and their stratification for clinical trials. 36 After validation in a larger cohort, our approach can be of high clinical interest as a practical tool for monitoring the immune status in severely injured patients in order to better time necessary procedures and avoid lethal post-injury complications such as excessive inflammatory reactions or lifethreatening infections. F I G U R E 4 Incubation of blood samples for 4 h instead of 24 h results in a similar release of inflammatory mediators. Samples taken at 24 h after injury were incubated for 24 h before cytokine determination. A second set of tubes taken 24 h after injury was incubated for 4 h and compared to the 24 h incubation time. Concentrations of cytokines were measured in supernatants (plasma) of whole blood incubated with (LPS) or without LPS (Ctrl). Data are shown as group median; n = 7 patients