Standardized porcine unilateral femoral nailing is associated with changes in PMN activation status, rather than aberrant systemic PMN prevalence

Trauma patients treated by intramedullary nailing (IMN) for a femoral shaft fracture are at increased risk to develop inflammatory complications such as in severe cases acute respiratory distress syndrome (ARDS) [1]. Despite improvements in trauma care, no changes in ARDS incidences in trauma cases occurred over the last three decades [2]. Nevertheless, a large number of studies indicate an association between intramedullary nailing of fractures and the occurrence of ARDS. Albeit underlying pathomechanisms are still poorly understood [3]. Previous trauma studies mainly determined the effect of humoral immune processes on ARDS development [1‐6]. Plasma interleukin-6 (IL-6) levels are considered as a reliable marker for injury severity and it has been shown that femoral shaft fractures are associated with increased plasma IL-6 levels [6]. However, despite being investigated thoroughly, data on interleukins lack prognostic value for trauma conditions [7, 8]. To predict outcome and to gain insight in the pathomechanisms of orthopedic trauma related inflammatory complication, it is key to also investigate cellular immunological parameters. Cytokines, including interleukin-6, are involved in regulation of essential innate effector cells, namely polymorphonuclear leukocytes (PMNs) [7]. Previously, it has been demonstrated that post-trauma inflammation leads to mobilization and activation of circulatory PMNs. Activated circulatory PMNs migrate into the tissue compartment, including highly vascularized lung-tissue. Upon activation, tissue residing PMNs release radical oxygen species (ROS) and proteases, which can cause collateral damage to parenchymal cells and thereby contribute to organ dysfunction [4, 7]. Due to standardization issues in clinical trauma studies, the impact of IMN of long-bone fractures on PMN homeostasis remains unclear. We aimed to evaluate the impact of IMN of a femur fracture on appearance and activation status of the circulatory PMN population, in a standardized long-term large animal experiment, as it has previously been demonstrated that experimental IMN of a unilateral femur fracture is associated with enhanced pulmonary PMN pooling [9]. A porcine model was utilized due to the close similarity of anatomical, physiological, genetical and immunological properties to human [10]. We hypothesized that standardized IMN of a unilateral femur fracture in pigs is associated with increased appearance of PMNs in circulation and enhanced activation status of these cells.

Experiments were executed in accordance with the German legislation governing animal studies following The Principles of Laboratory Animal Care [11]. The study was approved by the regional authority: Landesamt für Natur-, Umwelt- und Verbraucherschutz, LANUV NRW, Germany under application number: AZ 84-02.04.2014.A265. For these experiments, male German landrace pigs were included with a body weight between at 30 ± 5 kg. This study presents partial results obtained from a large animal porcine multiple trauma model. The model has been previously described in detail by Horst et al. [12].

All experimental animals survived the 72 h observation period. Furthermore, a mid-shaft fracture and adequate fracture fixation was confirmed by macroscopic analysis after termination in all intervention animals. No complications were diagnosed during the observation period.

Orthopedic trauma and intramedullary nailing are associated with increased incidences of inflammatory complications [1]. Polymorphonuclear leukocytes are believed to play an important role in the development of these complications after trauma. Clinical trauma studies, however, lack standardization due to (i) heterogeneous insult condition types and severity, (ii) suboptimal timing of sampling/procedures, (iii) variation in patient characteristics (genetic background, co-morbidities, physiology, age, gender, medication). Therefore, the current standardized large animal long-term observation study was executed, and the results may be summarized as follows:

This indicates that a unilateral femur fracture and subsequent intramedullary femoral nailing should be considered as a substantial trigger for the innate cellular immune system.

In addition, we previously managed to demonstrate with this porcine monotrauma model that, compared with control conditions, IMN of an isolated femur fracture is associated with increased pulmonary PMN occurrence after 72 h [9]. This indicates that early alterations of circulatory PMN activation status (current study) precede pulmonary innate immune cell accumulation [9]. It is tempting to speculate that monitoring of circulatory PMN alterations could assist in the prediction of remote organ inflammation.

Post-insult variations in circulatory PMN numbers over time in our study mimic both clinical and experimental findings [17‐19]. Furthermore, in line with a rodent trauma study, we found altered circulatory PMN presence over time in both the control (A/MV) and intervention study groups. This suggests that general anesthesia and mechanical ventilation shape circulatory PMN kinetics and that the additional impact of trauma/surgery on changes in circulatory PMN presence is limited [20]. Despite the presumed limited additional impact of a unilateral fracture and subsequent fracture fixation on circulatory PMN numbers, the current study does reveal that cell surface expression profiles of relevant receptors are largely affected by this trigger (FF + IMN). This makes it tempting to hypothesize that the interplay between (i) trauma anesthesia/mechanical ventilation settings and (ii) fracture (fixation standards) dictates the innate immune response in these trauma patients. All these factors and their interplay should therefore be considered as potential targets for immune modulation and this requires further study.

Nowadays, cell surface receptor expression, in contrast to immune cell functionality, can easily be determined by modern automatized flow systems [21]. Determination of these parameters in a controlled setting is of great relevance and may form the basis for future routine diagnostic, prognostic and therapeutic investigations in the clinical setting. Therefore the current study focusses on PMN activation status by studying well-established PMN markers with appropriately defined biological function [22], rather than ex vivo characterization of PMN functionality.

CD11b is a well-established activation marker and in vitro studies demonstrate that cell activation is associated with neutrophil-CD11b upregulation [23, 24]. Clinical studies on systemic inflammation further showed that circulatory neutrophil-CD11b expression rise upon insult [25‐27]. Like the findings of the current study, Johansson et al. demonstrated increased PMN-CD11b expression in burn patients [25] and others demonstrated an early rise of neutrophil-CD11b cell surface expression in trauma patients [26]. In addition, a similar pattern of Mac-1 expression kinetics was seen by Botha et al. [27]. They showed that neutrophil-CD11b expression levels in trauma patients peaked between 6 and 18 h after admission [27]. In the current porcine experiment, PMN-CD11b expression reached peak levels later, namely after 24 h. We believe that this discrepancy may mainly be due to differences in trauma load between both studies. Trauma victims in the study from Botha et al. had ISS scores over 20 [27], whereas our porcine trauma model represents a monotrauma model with an ISS of 9. This might imply that trauma severity is associated with PMN-integrin expression.

A study from Bathia et al. further found that PMN-CD11b expression started to decrease and return to homeostatic levels, as late as 72 h after severe trauma [28]. A similar pattern of restoration was observed in our experimental study. Timing of definitive fracture fixation for orthopedic trauma is a topic of considerable debate. Operative intervention is an additional trigger for the immune system and thereby leads to an upsurge of the post-trauma inflammatory response [29]. As integrins are considered as essential receptors in PMN adhesion and migration processes, changes in circulatory PMN-CD11b expression may potentially be utilized to guide timing of definitive fracture fixation. Especially as homogeneous blood PMN-CD11b response was found in all previously mentioned trauma studies and our experimental data. This response entails peak PMN-Mac-1 expression levels within the first 24 h and restoration to homeostatic levels within 72 h post-insult. Therefore, it is tempting to speculate that postponing IMN for the first 72 h in the critically ill is beneficial as integrin levels are allowed to normalize first, which is in line with clinical studies.

L-selectin plays an critical role in neutrophil transmigration as this receptor binds to its ligands on vascular endothelium and thereby facilitates slowing down, or ‘rolling’ of neutrophils [30]. Neutrophil activation is typically characterized by shedding of L-selectin [31] and higher trauma load is associated with more profound decreased neutrophil L-selectin expression [32]. In the current study, L-selectin expression on PMNs did not significantly change over time after insult. A similar observation was made in a study by Mommson et al. on patients undergoing elective surgery of the lower extremities [33] and in a cohort of burn patients [25]. Striking alterations in blood neutrophil L-selectin expression over time, however, have been shown in a study performed by Maekawa et al. on more severely injured trauma patients [34]. These discrepancies might be partly explained by differences in trauma load and the intensity of the post-insult inflammatory response and are thereby in line with the observations from Seekamp et al. [32].

Fcγ-receptors are involved in activation of neutrophils as they bind to immunoglobins (IgG) either in aggregates or attached to pathogens. Binding induces phagocytosis or promotes oxidative burst [35]. Interestingly, expression of FcγIII (CD16) on blood PMNs increased over time in our model. This contradicts kinetics in burn injuries [25, 36] and polytrauma patients studied by White-Owen et al. in which overall PMN-CD16 expression dropped after trauma and remained decreased for more than 48 h [37]. We believe that our porcine monotrauma insult-condition did not cause an immune response that was intense enough to evoke the mobilization of novel subsets, and more specifically CD16^low- PMNs in mobilization [38, 39]. Consequently, altered blood PMN-pool constitution-related reduction of PMN-CD16 expression did not occur in our model. The paradoxical increase of PMN-CD16 expression in our trauma study might reflect a specific activation profile of the blood neutrophil pool as regulation of the PMN-CD16-receptor in trauma is affected by different processes including altered PMN-phenotype appearance, delayed apoptosis, potentially modified balance between receptor shedding/biosynthesis and mobilization of the PMNs´ internal pool of pre-formed CD16 [38‐41]. However most likely an interplay between these mechanisms occurs and, in our view, PMN-CD16 receptor dynamics upon trauma should be a focus of upcoming studies.

Our findings regarding PMN-FcγII (CD32) expression changes are in accordance with a study from Visser et al. Both studies showed no evident alterations in PMN-CD32 expression within the first 24 h after trauma [42].

The current study demonstrates that a more profound blood PMN-CD11b/CD16-response was found in the group subjected to orthopedic trauma/IMN than in animals exposed to anesthesia and mechanical ventilation only. Thereby our experimental study reveals that orthopedic trauma and subsequent IMN elicit a specific pattern of enhanced PMN-CD11b/Mac-1 cell surface expression. This may be an important factor in the association between increased occurrence of inflammatory complications and intramedullary nailing in orthopedic trauma. It is tempting to speculate that IMN for long-bone fractures induces a specific systemic PMN activation profile due to increased release of DAMPs, local cytokines, pro-inflammatory bone particles or angiogenic/growth factors [5‐7, 43, 44]. Upcoming experimental studies should focus on the identification of involved pathways.

In contrast to studies on polytrauma [45, 46] we did not observe impaired PMN-responsiveness as LPS-induced upregulation of CD11b on PMNs was retained after controlled monotrauma and fracture fixation. Interestingly, there even seems to be a trend towards intensified early PMN-responsiveness upon trauma and subsequent nailing as peak-PMN activation status, reflected by the amplitude of PMN-CD11b expression after LPS-stimulation, is significantly greater after trauma than under homeostatic conditions. This phenomenon is in contrast with observations in polytrauma studies. As in these clinical investigations, impaired PMN-responsiveness, and the occurrence of refractory neutrophils in circulation are linked with severe trauma and complicated courses. [45‐47].

The specific impact of instrumentation on early PMN-receptor alterations was studied as well. In short, additional samples were collected prior to fracture fixation and 2 h after fracture fixation. In line with Hietbrink et al. [26] no differences in circulatory PMN-integrin/selectin/Fcγ- receptor expression were seen between pre- and post-instrumentation conditions. (Data are displayed in Supplementary file 2).

The current standardized porcine trauma model, and human trauma studies are (except for CD16-data) remarkably similar regarding PMN receptor dynamics trends. Given these similarities, the current porcine trauma model is very suitable for proof-of-principle interventions with novel therapeutic strategies for trauma. Integral translational effects are summarized in Supplementary file 3.

Limitations of this study include its sample size, however relevant effects reached statistical significance and, therefore, we conclude that there was no need to include/sacrifice more experimental animals. We further decided not to include additional study conditions without clinical importance or translational value. So, unconventional fracture fixation techniques for cardiopulmonary compensated monotrauma scenarios (such as external fixation/plating/splinting) or IMN without fracture induction were not investigated. Finally, as porcine immune cells, in contrast to rodent models [48], do allow for adequate analysis of the key Fcγ-receptor family, we decided to focus with our antibody panel on these specific markers plus integrins and selectins, rather than adding antibodies aimed to investigate other circulatory immune cell populations.