(iii) An update on the systemic response to trauma

Background 

Trauma care has evolved dramatically in recent decades. Less than a century ago, the systemic consequences of major injury were recognized but doubts still existed about whether shock was due to wound contamination or blood loss. In the appalling circumstances of the First World War it is understandable that such difficulties arose. Nevertheless, innovations included the use of blood transfusion and early stabilization of fractures non-operatively, using the Thomas splint.1 This simple intervention transformed the mortality rate associated with open femoral fractures, from 80% to 20%. It is hard to imagine a more graphic example of a simple clinical measure applied early to stabilize a fracture with such dramatic yet unforeseeable benefits.

In the post-war era, the advent of modern intensive care hinged upon the ability to support patients through periods of reversible organ failure. During this period, non-operative fracture management was the mainstay of treatment for the poly-trauma patient. However, Riska reported a series of patients in whom operative fracture stabilization brought about dramatic simplification of the patients Intensive Treatment Unit (ITU) care and heralded the opportunity to reduce the length of ITU stay, hasten rehabilitation and possibly avoid the risks of recumbency.2

Such observations began a shift in practice in favour of early operative stabilization of major fractures soon after life saving surgery.3, 4, 5 Several case series were supplemented by a randomized controlled trial of early versus delayed femoral fracture stabilization. This trial demonstrated dramatic benefits for those patients randomized to early fracture stabilization (<24 h of injury).6 This strategy, known as Early Total Care (ETC), came to dominate the orthopaedic management of the poly-traumatized patient.

Cellular and humoral elements of the systemic inflammatory response to trauma 

If Bone's findings6 were truly translatable to the trauma population as a whole, then ARDS/MOF should have become rarities. This was not the case. A greater understanding of the consequences of the systemic response to injury was clearly needed. Simple clinical parameters (such as C-reactive protein-CRP) were noted to be raised within a few days of injury and generally paralleled increasing injury severity. The realization that there was considerable cross-over between cascades of mediator interactions (such as between clotting, inflammation, tissue hypoxia and reperfusion) was a vital step.7 Hypoxia and tissue factor expression following injury initiate a host of local responses not conforming to the traditional concept of cascades, but rather webs of interactions involving the coagulation, inflammation and thrombolytic pathways. The inter-related problems of acidosis, hypothermia and coagulopathy (the lethal triad) spurred on the development of Damage Control Surgery (DCS), which aimed to surgically control life-threatening haemorrhage as rapidly as possible.8, 9 Rapid, effective emergency treatment aimed at preventing early death was also noted to decrease the third peak of deaths due to ARDS/MOF.

Following its original description,10 it was soon noted that the development of ARDS was often just the first manifestation of MOF. Early work was somewhat hampered by a lack of consistency in definitions of ARDS and acute lung injury (ALI). Similarly MOF, whilst clinically easy to recognize, was difficult to quantify. Consensus definitions have since been established and it is now possible to stratify the severity of SIRS and organ dysfunction in a meaningful and reproducible manner.11, 12

ARDS was found to be associated with systemic capillary leakage, which preceded the development of the syndrome and is sustained during its course.13, 14 Thus ARDS is much more than simply a pulmonary condition. Interest focused upon the polymorphonuclear neutrophil (PMN), the “rapid-responder” leukocyte whose migration and activation in the lung were associated with lung injury and with the severity of ARDS.

It soon became clear that the interaction of PMN with endothelial cells was a vital part of the evolving systemic inflammatory response. Far from being passive bystanders between which PMN migrate, the endothelial cells are intimately involved in the direction of PMN movement and indeed in their activation.15 In health the majority of mature PMN's are marginated on the walls of blood vessels and can be rapidly mobilized into the circulation when needed. These PMN's are in a resting state, expressing a limited array of receptors and adhesion molecules, and with all of their cytotoxic apparatus (enzymes for superoxide generation and proteolysis) safely locked away in cytoplasmic granules. PMN's can be activated directly from the resting state, which results in the sequential expression and shedding of adhesion molecules (with complimentary expression on endothelial cells), which allows first rolling, then firm adhesion and finally trans-endothelial migration (Table 1). Once out of the blood vessel, the PMN then seeks to phagocytose and destroy any hazardous material which it encounters.

Table 1.

	Phase of PMN/endothelial interaction	PMN expression	Endothelial cell expression
	Rolling	L-selectin	E-selectin
	Firm adhesion and migration	β2-Integrins CD11b/CD18	Intercellular/endothelial leukocyte/vascular adhesion molecules (ICAM/ELAM and VCAM)

PMN/endothelial interaction depends upon the sequential expression and shedding of adhesion molecules (with complimentary expression on endothelial cells) which allows first rolling, then firm adhesion and finally trans-endothelial migration.

However it is clear that a primed state also exists. Here, exposure to a pro-inflammatory environment stimulates PMN's to gear-up, ready for action. Enhanced adhesion molecule and receptor expression are seen, which results in enhanced migration in response to chemotactic signals. Furthermore the PMN's destructive potential once activated, is enhanced, with greater production of superoxide and proteolytic enzymes, with extracellular degranulation and death of the PMN's. Local tissue injury is inevitable.

Priming is driven by injury, hypoxia and acidosis.16, 17, 18, 19 The exact mediators implicated appear to have a built-in redundancy. Platelet activation factor (PAF) appears to be important in the first few hours after injury. However, in experimental studies as circulating IL-8 levels begin to rise blockade of PAF is ineffective. The recognition that events at a cellular level could proceed in a stepwise fashion lent credibility to the development of the concept of the 1-and 2-hit hypothesis (Figure 1).20

View Large Image Download to PowerPoint

Figure 1 One and two hit hypothesis

In the face of major trauma similar local responses within the lung may be particularly important in the development of early ARDS, as an apparent 1-hit phenomenon. Pulmonary macrophages produce IL-8 rapidly when subject to hypoxia and this is exacerbated further by a period of subsequent hyperoxia.21 Elevated levels of IL-8 have been described in bronchoalveolar lavage of those injured patients studied shortly after arrival in hospital and at greatest risk of the development of ARDS. The capillary system within the lung is complex and tortuous, obliging PMN's to squeeze along in contact with the endothelial cells.22 Coupled with priming for IL-8 mediated transmigration,23 sequestration within the lung results.

Great interest has focused upon the pro-inflammatory environment in the hope of identifying a simple, reliable marker of the risk of development of post-injury ARDS/MOF. The complexity of the webs of cellular and humoral interactions, with cross-over between clotting and the inflammatory response, has already been alluded to. Increased levels of circulating cytokines (IL-8 and IL-6) are well described. Such cytokines can be regarded as both mediators and markers of the severity of the systemic inflammatory response. IL-6 has been associated with increased risk of ARDS/MOF24 but has not been widely adopted as a clinical marker of risk. Indeed the biology of IL-6 is complex, as its function is intimately related to the presence of its soluble receptor (sIL-6R). In other inflammatory conditions, rising sIL-6R heralds the termination of the acute inflammatory response.25, 26 The importance of sIL-6R remains to be described in trauma.

The initial phase of inflammation is followed by relative immuno-suppression, in which the circulating PMN's are relatively unresponsive. The other leukocyte lineage implicated in this phase are the monocytes. In local inflammation, monocytic infiltration heralds the termination of the acute inflammatory process. Following major injury, monocyte HLA-DR expression declines. HLA-DR expression is vital to the efficacy of acquired immune response. Failure of monocyte expression to recover is associated with septic complications and the risk of death.27 Increased anti-inflammatory cytokine expression is also seen. These observations have led to the concept of a Counter-regulatory Anti-inflammatory Response Syndrome (CARS).28 However, it is entirely possible that these phenomena are simply pathological too.

Protein-C warrants particular attention. Tissue hypoxia and hypoperfusion drive protein-C mediated acute coagulopathy of trauma (ACT).29 The thrombin/thrombomodulin complex potentiates tissue factor mediated protein-C activation (to APC). The activated form exerts a direct anti-coagulant effect. Once bound with protein-S the complex inactivates factors Va & VIIIa. ACT is associated with increased mortality and organ dysfunction. However, protein-C is pleiotropic and has also been shown to have cytoprotective effects.30 Directly anti-inflammatory effects have been described for both leukocytes and endothelial cells. These include down-regulation of adhesion molecule expression, inhibition of inflammatory mediator release and endothelial barrier protection by stabilizing the cytoskeleton. Some benefit has been reported in the therapeutic use of APC in sepsis, but despite the above properties, until a cytoprotective analogue of APC is developed its therapeutic use in trauma will remain hypothetical.

Clinical relevance 

Data from several series have yielded strong evidence of a vulnerable window after life saving surgery in which complex reconstructive surgery risks causing a physiological deterioration.31 Similar findings have been observed in series of fracture patients undergoing definitive stabilization while still having raised serum lactate levels despite restoration of haemodynamic normality.32 Such persisting lactic acidosis has been called occult hypoperfusion and is associated with an increased risk of septic complications.

Detailed studies of the inflammatory response to fracture surgery, even in those with isolated fractures, tended to yield results indicating worsening of inflammation/immuno-suppression following surgery, despite clinical findings that such patients invariably did very well.33, 34, 35 However, specific concern began to emerge relating to pulmonary function following femoral nailing within 24 h of injury.36 Evidence began to accumulate that for some patients this was acting as a clinically important second hit.

Techniques similar to DCS had been employed for years for trauma victims with exsanguinating pelvic injuries, with rapid recognition and immediate steps to control haemorrhage in the form of binder and then external fixator application.37 With the concerns over ETC in some patient groups, interest grew in applying these techniques to other major fractures to confer the benefits of early fracture stabilization, without the risks of major surgery, an approach known as Damage Control Orthopaedics (DCO).38 The following phase of extended resuscitation allows ventilation, perfusion, coagulation and core temperature to recover, and so the tide turns in the patient's favour.

Conversion to definitive fixation is no longer associated with worsening of pro-inflammatory changes described previously. Indeed several series indicate that such strategies clearly have benefits in terms of the extent and duration of the patient's Systemic Inflammatory Response Syndrome (SIRS).39 Furthermore, concerns over increased significant complications (wound infections etc) have not been proven.

Summary of current clinical strategies 

An understanding of the systemic inflammatory response to injury is only of benefit if it can be translated into a refinement in clinical practice. Such strategies remain in a state of evolution. Haemorrhage and resulting tissue hypoxia drive the acute coagulopathy of trauma (ACT). A significant proportion of patients arriving at hospital with major injuries have ACT at presentation. It is believed that hypoxia drives this by the generation of activated protein-C (APC) which has potent anti-coagulant properties.29 The mortality for these patients is high and is associated with subsequent ARDS/MOF. The importance of haemorrhage control is self-evident, but the benefits of rapid, effective treatment go far beyond the prevention of death due to exsanguination. Concerns over the limitations and perceived risks of damage control have faded in the face of improved survival.

Damage Control Orthopaedics 

The concept of DCS was extrapolated into musculo-skeletal trauma, essentially building upon the established management pathways of severe pelvic injuries. Percutaneous temporary external fixation is used to reduce and stabilize major fractures, minimizing blood loss and any risk of intra-operative embolization.38, 42 The patient in extremis, or who is unstable, is simple to identify. The criteria of the ‘borderline patient’ have been well described (Table 2). However, many would also include occult hypoperfusion and any significant coagulopathy of hypothermia.

Table 2. “The borderline patient”

	Polytrauma + ISS > 20 + thoracic trauma (AIS > 2)
	Polytrauma + abdominal/pelvic trauma and haemodynamic shock (initial BP < 90 mmHg)
	ISS > 40
	Bilateral lung contusions on X-ray
	Initial mean pulmonary arterial pressure > 24 mmHg
	Pulmonary artery pressure increase during IM nailing > 6 mmHg
	Occult hypoperfusion (OH) serum lactate < 2.5 mmol/l despite haemodynamic normality

ISS – Injury Severity Score, AIS – Abbreviated Injury Scale.

Conversion to definitive fixation is not recommended between 2 and 5 days after injury, the so-called “no-go zone”. This period corresponds to the vulnerable window in which conversion would carry the greatest risk of a second hit.

The future 

The function of haemopoetic bone marrow is central to the process but poorly understood.45 The effects of shock and tissue injury are difficult to study in the bone marrow, as this represents a relatively inaccessible part of the inflammatory system. Modulation of initial inflammatory and secondary immuno-suppressive phases may prove best addressed by stopping faulty production of leukocytes rather than attempting to modify faulty cells once released.

Summary 

Evolution of clinical care always seems to be ahead of our understanding of the basic science underpinning these developments. Keen clinical observation remains at the heart of driving forwards improvements in care. Such observation hinges upon clinicians of all specialties having a clear and shared grasp of the pathophysiology of the systemic inflammatory response to injury. The pioneering work done in the military surgery sphere in developing DCR has provided a vital step forwards. The challenge for civilian practice is to catch up with the standards in military surgery and for academic medicine to tease apart the vital components of care to enhance their further development. At present, manipulation of the systemic inflammatory response to injury is only possible by the indirect means offered by DCR.