Damage-Control Orthopaedic Surgery in Polytrauma: Influence on the Clinical Course and Its Pathogenetic Background Hans-Christoph Pape
Decision Making for the Treatment of Major Fractures: Evidence for the Benefits of Tapered Treatment to According to the Patient Condition The standard of care for major extremity fractures in patients with multiple injuries foresees early definitive fixation of all major fractures. Recently controversial results were gathered from studies [1–7], and observations were made about certain patient sub-groups that developed unexpectedly high complication rates, when early definitive stabilization was performed. It was suggested that potential deleterious effects of internally fixing long-bone fractures in the acute setting, where systemic hypo-perfusion and inflammation occurred, increases susceptibility to end-organ injury. In summary, three factors appeared to play a role that are pathogenetically connected: 1. Additional severe injuries (such as chest trauma) 2. Prolonged surgeries 3. The pre-operative condition [8, 9] Several observational studies and one randomized study have suggested clinical benefits from early stabilization of major long-bone fractures in reducing the incidence of pulmonary complications and mortality. Differences in treatment definition, time cut-offs and type of fixation used, small samples limited to single institutions and inadequate control of confounding variables, all serve to limit the validity and general application of prior findings. One of the problems of previous studies was the lack of focus on those patient sub-groups that did appear to be at high risk for complications and take into
Hans-Christophe Pape, MD, FACS Chairman, University of Aachen, Department of Orthopaedics Pauwelstreet 30 52074 Aachen, Germany e-mail:
[email protected]
account the three factors listed above. Recently, two studies have become available that investigate the effect of fracture fixation depending on the pre-operative clinical condition of the patient and take into account that additional severe injuries and hypo-perfusion may influence the course. 1. A prospective randomized multi-center study summarized 165 multiply-injured patients from ten different centres between 1, Jan 2000 and 28, Feb 2006 [10]. This study is different from other retro- or prospective studies because no patients with isolated fractures were included. Moreover, patients were excluded if they had severe head and chest injuries, as well as exsanguinating conditions. Thereby, a better focus on the potential impact of the fixation of the femoral shaft fracture was obtained. Of the 165 patients in the study, 71 (43.0%) were randomized to the external fixation treatment group and 94 (56.9%) were randomized to the intra-medullary nailing treatment group. Comparisons between the two groups on demographic characteristics and initial injury severity demonstrated a similar pattern, except for a higher head trauma score in the external fixation group when compared with patients undergoing initial intra-medullary nailing of the femur. Analyses documenting differences between patients in a clinically stable or uncertain (borderline) condition in terms of initial injury severity and post-operative outcomes validated the notion that borderline patients have significantly worse injuries and post-operative outcomes than stable patients. In terms of initial injury, borderline patients demonstrated higher scores on the revised trauma index, injury severity index, and head and thorax injury indices in comparison to stable patients. Borderline patients were also more likely to have a bilateral femoral fracture, a hemothorax, and require a blood transfusion within 24 h of admission in comparison to stable patients. In terms of post-operative outcomes, borderline patients spent more hours in the ICU and more hours on ventilation in comparison with
G. Bentley (ed.), European Instructional Lectures. European Instructional Course Lectures 9, DOI: 10.1007/978-3-642-00966-2_8, © 2009 EFORT
67
68
Hans-Christoph Pape
stable patients. Borderline patients were also more likely than stable patients to experience clinical complications such as acute lung injury, systemic inflammatory response, sepsis, and multiple organ failure. The analyses examined the influence of treatment group status on post-operative clinical course and complications separately for stable patients and those in uncertain condition (see Table 1). Again, regression analyses statistically controlled for differences between the two treatment groups in terms of initial injury severity (i.e., revised trauma score, new injury severity index, head injury score) when examining group differences. For patients who presented in stable condition, those who underwent intra-medullary nailing experienced a shorter duration on a ventilator in comparison to those with external fixation. In contrast, borderline patients that underwent initial nailing of the femur had a higher incidence of acute lung injury in comparison to those who underwent external fixation. After adjusting for initial injury severity, the odds of developing acute lung injury were 6.69 times greater in borderline patients who underwent intra-medullary nailing in comparison to those who underwent external fixation. The authors concluded that the type of surgical procedure for fixation of a femoral shaft fracture should be carefully selected, according to the initial assessment of the clinical condition. In patients who present in an uncertain (borderline)
condition, an external fixator should be applied for temporizing purposes. 2. The second study used the National Trauma Databank (NTDB) from the United States [10] between 1, January 2000 and 31, December 2004 [12]. In this study, the inclusion criteria encompassed: (a) The presence of a closed or open fracture(s) of the femoral shaft (b) An injury severity score greater than or equal to 15 (c) Sixteen years of age or older (d) Those who underwent a internal fixation of the femur This study is different from many other studies because it tried to provide clinically relevant categories for treatment beyond a simple dichotomy at 12 or 24 h. Five time periods were selected a priori, based on commonly used cut-points from the literature [11–13]: t0 ≤ 12h, 24 < t2 ≤ 48 h, 12 < t1 ≤ 24 h, 48 < t3 ≤ 120 h, It is also different because the authors hypothesize that an additional physiologic stress from definitive fracture surgery could activate an adverse systemic response leading to end-organ injury and a higher mortality rate. The authors used an inverse-probability-of-treatment-weighted (IPTW) analysis to estimate the risk of mortality for a defined treatment time. Their results document that definitive fixation in all but one (24–48 h) of four delayed treatment categories
Table 1 Treatment group differences in clinical course and complications for patients in stable and uncertain (borderline) condition (data from [58]) Stable condition
Uncertain (borderline) condition
s−I°ExFix (N = 50)
s−I°IMN (N = 71)
Regression analyses
Outcomes
M ± SD
M ± SD
HR
95% C.I.
ICU (h) Ventilator (h)
212.4 ± 167.93 142.2 ± 121.32
133.52 ± 193.49 66.54 ± 108.45
1.06 1.55
%
%
23.8 28.6 9.5 30.2 11.9 0.0
6.5 12.9 6.3 30.8 6.3 0.0
Pneumonia ALI ARDS SIRS Sepsis MOF
b−I°ExFix (N = 21)
b−I°IMN (N = 23)
Regression analyses
p
M ± SD
M ± SD
IRR
95% C.I.
p
0.84−1.86 1.04−2.33
0.290 0.030
476.95 ± 284.50 360.48 ± 245.47
399.14 ± 404.13 313.81 ± 359.77
1.28 1.36
0.67−2.44 0.74−2.53
0.445 0.325
OR
95% C.I.
p
%
%
OR
95% C.I.
p
0.40 0.39 0.73 1.49 0.60
0.11−1.50 0.14−1.08 0.15−3.53 0.62−3.57 0.15−2.36
0.176 0.170 0.700 0.367 0.469
38.9 16.7 11.1 50.0 11.1 16.7
45.0 52.4 16.7 52.6 36.8 22.2
1.00 6.69 2.01 0.73 3.86 0.78
0.22−4.59 1.01−44.08 0.13−31.91 0.17−3.24 0.46−32.52 0.13−4.75
0.995 0.048 0.618 0.684 0.214 0.791
Regression analyses represent the relation between treatment condition (0 = external fixation, 1 = intramedullary nailing) and each outcome after controlling for initial treatment group differences on the revised trauma score, new injury severity score, and AIS head score. Cox regression with robust standard errors was used for number of hours until release from ICU and hours until taken off the ventilator. Logistic regression with robust standard errors was used for binary outcomes ALI acute lung injury; ARD acute respiratory distress; SIR systemic inflammatory response; MOF multiple organ failure; HR hazard ratio; OR odds ratio; C.I. confidence interval
Damage-Control Orthopaedic Surgery in Polytrauma
69
was associated with a significantly lower risk of mortality to about 50% of that expected with early treatment (less than 12 h). Also, patients with serious associated injuries demonstrated greater risk reductions from delayed fixation when compared with those with less serious or no abdominal injury. Their data was consistent with the results of multivariate standardized risk ratios when the early treatment group was used as the standard population Table 2. The authors conclude from this study that a cautious approach to early definitive femoral shaft fracture fixation among multi-trauma patients should be performed and re-inforce this for patients who present with serious associated abdominal injuries. These two studies appear to present strong evidence in favour of a tapered approach toward the multiply-injured patient. They show that a close relationship exists between the duration of surgery, the degree of initial injuries and the state of resuscitation, or the general condition. It appears that current end-points used to guide resuscitation, such as blood pressure, urine output, heart rate, base deficit and serum lactate may underestimate occult tissue hypo-perfusion. The clinical status was graded for decades by using only cardiovascular parameters. General surgeons have then considered several other factors to be important to describe changes observed in exsanguinating injuries caused by penetrating trauma: these include hypothermia, acidosis, and coagulopathy induced by hemorrhagic shock, and have been named the “triade of death” [14]. Orthopaedic surgeons have then tried to adapt these to the population of blunt trauma patients with predominant extremity injuries. As a fourth element, they have added “soft tissue injuries” summarizing all muscular and subcutaneous injuries of the extremities, lung, abdomen,
and pelvis. It should be noted that this fourth entity summarizes a number of clinical diagnoses rather than a single cascade system. These four pathophysiological entities – hemorrhagic shock, hypothermia, coagulopathy, and soft tissue injuries/ inflammation – were named the “four pathophysiological cycles of blunt polytrauma”. Moreover, many authors ask for more sensitive measures of tissue oxygenation measured by polarographic or nearinfrared technologies and markers of inflammation and coagulation that better reveal the physiologic condition of a patient and these are likely to replace simple temporal distinctions. The following paragraph summarizes why there are similarities in the pathogenesis of post-traumatic organ dysfunction and the changes induced by major surgical procedures.
Pathogenesis of Organ Failure Following Trauma Induced by Multiple Fractures, Soft-Tissue Damage and Acute Haemorrhage In those subsets of polytrauma patients who present with severe haemorrhage, severe soft tissue trauma, severe pelvic disruptions or sustained head trauma [11], clinical studies documented that an over-activated immune response pre-determines these complications [12, 15, 16]. The following pathogenetic changes after trauma and the influence of fracture fixation applies. Within the first hours, the most important physiologic changes are induced by local and systemic hypoxia [17]. Blood loss and tissue damage caused by fractures and soft tissue crush injuries induce
Table 2 Comparison of crude, regression, and marginal structural model estimates of effect of treatment time on mortality (data from [59]) Estimate
12–24 h
24–48 h
48–120 h
>120 h
Crude relative risk
0.50 [0.06] (0.18, 1.01) 0.45 [0.03] (0.15, 0.98) 0.47 [0.07] (0.14, 1.11)
1.13 [0.69] (0.58, 2.05) 0.83 [0.49] (0.43, 1.44) 0.94 [0.85] (0.44, 1.76)
1.19 [0.52] (0.67, 2.03) 0.58 [0.03] (0.28, 0.93) 0.58 [0.09] (0.21, 1.09)
1.17 [0.77] (0.24, 2.65) 0.43 [0.03] (0.10, 0.94) 0.43 [0.05] (0.09, 0.94)
IPTWa relative risk Standardized risk ratiob
Point estimates are given with p values in square brackets and 95% confidence intervals in parentheses. All analyses use t0 (£12 h) as the referent treatment group a Inverse-probability-of-treatment-weighted (IPTW) population relative risk was derived using model for treatment assignment controlling for Wm (NISS, GCS, Northeast region, age, arrival time, number of serious extremity/pelvis or head injuries, number of femur fractures, presence of cardiac or cerebrovascular comorbidities, teaching status, and ACS Level 1 designation) b Standardized risk ratio (SRR) using same treatment model as IPTW analysis but modified weights to give the estimated proportionate risk had those subjects treated early (t0) received treatment at a later time
70
generalized hypoxemia in the entire vascular bed of the body [18]. Hypoxemia is the leading cause of damage as it causes all endothelial membranes to alter their shape. Subsequently, the circulating immune system, namely the neutrophil and macrophage defence systems, identify these altered membranes. The first immunologic reaction is the adhesion of neutrophils to the altered endothelial cell walls. They subsequently release their proteolytic enzymes that cause additional damage to the cell walls. This auto-destructive reaction occurs because of a lack of external pathogens that usually are a target for these mediators. Proteolytic enzymes and oxygen radicals are liberated into the bloodstream and aggravate the degree of endothelial damage [19]. Circulating neutrophils also adhere directly to tissue damaged from contusions, which may be located in the extremities, the muscles, or the lungs. Fracture haematoma is known to produce a manifold increase in cytokines, which may subsequently produce systemic effects [20]. Subsequently, the damaged endothelial cell walls, by trying to seal the damaged tissue, induce activation of the coagulatory system. This explains why these patients develop a drop in the platelet count. Further cascade mechanisms, such as the complement system, the prostaglandin system, the specific immune system and others, are activated [21]. Two prominent theories explain the pathogenesis of organ failure: The first concept is based on the on-going inflammatory process. The early neutrophil response described above is followed by a later response initiated by macrophages [22]. While the neutrophils exert their activity inside the vascular bed, macrophages have been well described cross the endothelium and act inside the interstitium [13]. If the reaction is part of an on-going inflammatory reaction, the classical signs of inflammation are clinically measurable. Generalized tissue swelling occurs, including in the extremities, a negative fluid balance is observed, and sometimes vasopressors are required in addition to increased fluid uptake. The subsequent chronic tissue hypoxemia caused by the interstitial swelling later results in the deterioration of several internal organs, a condition known as multiple organ dysfunction. It is represented by rubor (increased microvascular perfusion), calor (fever >38.5°C), tumour (generalized tissue swelling), dolor (pain and requirement for analgesia) and functio laesa(loss of function) of all organs [23, 24]. The main causes of late death following trauma are adult respiratory distress syndrome (ARDS) and multiple organ failure (MOF) and the immunologic response is closely associated with the development of these complications. Longstanding inflammation can convert the neutrophils and macrophages into a state of exhaustion and anergy. Anti-inflammatory mediators are liberated along with pro-
Hans-Christoph Pape
inflammatory mediators and weaken the defence, a state summarized as counter-activity-response syndrome, CARS [25]. Clinically, this period usually develops following the first week after trauma and is associated with an increased risk of infection, a decrease in wound healing capacity and the development of organ dysfunction [26]. Since the largest endothelial space is located in the lungs, which represent a special focus, it explains why the lungs are usually the first organ to deteriorate in the sequence of organ failure (requirement of high PEEP, high FiO2, and inverse inspiration-vs.-expiration ratio) [22, 27, 28]. Postmortem analyses demonstrated an increase in tissue oedema in various organs, and accumulation of inflammatory cells, namely neutrophils and macrophages in these organs [29, 30]. Neutrophils are regarded as mediators of the early phase and the increase in systemic Interleukin 6 levels after trauma is thought to be mediated by neutrophils. A parallel process occurs in the intestine and leads to activation and exhaustion of the stationary immune system, the reticulo-endothelial system: The centralization induced by blood loss causes intestinal hypoxemia and an alteration in intestinal permeability [31], which enables the crossing of intestinal pathogens through intestinal barriers and the permeation of pathogens in surrounding lymphatic tissues. It has been clearly proven that pathogens proceed to the liver through the portal vein break down [32, 33]. The most important effect of the exposure to endotoxins deriving from these pathogens is the induction of an inflammatory stimulus. This systemic inflammation occurs in the lymphatic tissues, followed by the reticulo-endothelial system of the liver. This can act as a buffer system [34] against pathogens and immunologic stimuli. Its protective effect fails when the endotoxin-induced inflammatory stimulus is on-going and overwhelming. Then, pathogens can spill over into the lung and accentuate the pre-existing inflammatory reaction. This spill-over is another reason why the lung is the first organ to show organ dysfunction in severely-injured trauma patients. The pivotal importance of the intestinal theory of posttraumatic organ dysfunction is widely accepted [35].
Interaction Between Acute Surgery After Trauma and the Pathophysiological Responses Induced by Trauma Many clinical studies have revealed that operations cause similar changes of the immunologic response to those induced by acute trauma. Among these, pro-inflammatory cytokines are most specific for trauma patients [2, 36]. Their
Damage-Control Orthopaedic Surgery in Polytrauma
71
levels usually remain elevated for more than 5 days after trauma in patients with a high injury-severity score, and early elevated levels discriminate trauma patients who later develop organ failure [3]. They also have been demonstrated to be closely related with the magnitude of the injury (burden of trauma) and with the operative procedure [37]. The degree of surgery has also been determined by changes in the specific immune system [38–40]. These and other studies reveal that the immuno-modulatory mechanisms after elective surgery and after primary surgery in trauma patients are well described [41]. Also, the inflammatory response induced by femoral nailing is biochemically comparable to that induced by other Orthopaedic operations. Moreover, in polytrauma patients, an additional impact due to primary surgery could be determined that occurs in addition to the one induced by the initial trauma. Clinical studies have clearly documented an exaggerated inflammatory response in which the duration of surgery and the amount of blood and temperature loss [28, 42, 43] play a role [38, 44]. A clinical prospective randomized study has shown that patients in uncertain conditions have a higher incidence of acute lung injuries, if longer surgical procedures are performed [10]. These factors may be induced by major surgery when a pre-activated immune response is present, such as in severe haemorrhage and soft tissue damage. In patients who present with very severe injuries, these factors outweigh the positive effects from the definitive stabilization of fractures described above [45]. Therefore, in these cases many Level I trauma centres have preferred to use a temporizing approach by using external fixation of the femoral shaft in selected patients, and named this approach “damage control orthopaedics” [45, 46]. Prior to the introduction of this approach, the most frequent treatment of major fractures in these situations was traction. The advantage of using an external fixator over traction lies in the fact that the fracture is stabilized which allows the patient to
move for nursing manoeuvres and to sit up in the ICU, which improves pulmonary toilet [47].
Factors for Grading Patient Assessment The damage-control concept of limiting the initial surgical time in patients with severe exsanguinating injuries was used for abdominal shotgun victims in Philadelphia. Packing of the major sources of haemorrhage, followed by ICU treatment and definitive repair in the following days was found to improve survival rates [48, 49]. For Orthopaedic injuries, the potential benefits of the damage control approach are more difficult to detect. Several aspects must be considered, such as slower progress of haemorrhage, existence of severe soft tissue injuries, and longer duration of the Orthopaedic repair. For patients with abdominal trauma the most important clinical factors are blood loss, coagulopathy and loss of temperature (triade-of-death); likewise in Orthopaedic patients, it is important to account for soft tissue injuries as well [50, 51]. Orthopaedic surgeons added a new patient category to the three existing classifications (stable, unstable, in extremis), which was named the “borderline” patient [52] (Table 3). Existing parameters for the classification of the patient condition have been for assessment. The abbreviated injury scale, Moore’s organ injury score and ATLS criteria of severe haemorrhage have all proven to be effective and have therefore been incorporated. Three out of the four criteria should be present to qualify a patient for a specific category. Fortunately, most patients fall into the stable or borderline category and among the latter, resuscitation can lead to improvement of the clinical condition to allow early fracture fixation [53]. Critical parameters can be summarized as follows: soft tissue injuries (major extremity fractures, crush injuries, severe pelvic fractures, lung
Table 3 Inverse-probability-of-treatment-weighted estimates of the relative risk of mortality for treatment delay (> 12 h) by severity of associated injury (data from [59]) Associated injury
Low severity (AIS 110,000 90–100 >1 Normal range 2.5 No data III–IV 2), coagulopathy (platelets
