Article Text

Download PDFPDF

Putative role of TNF-α, interleukin-8 and ICAM-1 as indicators of an early inflammatory reaction after burn: a morphological and immunohistochemical study of lung tissue of fire victims
  1. Joachim Boehm1,
  2. Kai Fischer1,2,
  3. Michael Bohnert2
  1. 1Institute of Pathology, University Hospital Freiburg, Freiburg, Germany
  2. 2Institute of Legal Medicine, University Hospital Freiburg, Freiburg, Germany
  1. Correspondence to Dr Joachim Boehm, Postfach 214, Freiburg 79002, Germany; joachim.boehm{at}


Aims To investigate incipient inflammatory changes as first steps in the development of a systematic inflammatory response syndrome in the lungs of burn victims.

Methods At the Institute of Legal Medicine of the University Hospital of Freiburg a collection of 40 forensic autopsy cases of burn victims was established that had died within 1 h after fire exposure. This group was compared with a total of 48 autopsy cases in three control groups (postmortem burns vs deaths from haemorrhagic shock vs railway suicide deaths). In all cases, immunohistochemical studies of lung tissue probes were performed to detect the presence of pro-inflammatory mediators using antibodies against tumour necrosis factor α (TNF-α), interleukin-8 (IL-8) and inter-cellular adhesion molecule 1 (ICAM-1).

Results The lungs of burn victims showed a significantly higher extent of intra-alveolar oedema than the other groups. Immunohistochemically, macrophages in all groups mostly showed a distinct expression of TNF-α, but not of IL-8 or ICAM-1. Interestingly, intravascular erythrocytes often showed a positivity of TNF-α that was strongest in the group of burn victims and differed significantly from all the control groups.

Conclusions In burn victims with short survival times of ≤1 h after fire exposure, the immunohistochemical expression profiles of TNF-α, IL-8 and ICAM-1 in the lungs were not altered enough to prove an instant inflammatory reaction in these cases. Nevertheless, the positive reaction of TNF-α in erythrocytes of burn victims may indicate the beginning of a non-specific immune response to fire-induced inhalation trauma.

  • Lung
  • fire
  • SIRS
  • TNF-α
  • immunohistochemistry
  • cytokines
  • forensic pathology
  • inflammation
View Full Text

Statistics from


In burn victims, inhalation trauma is one of the most frequent causes of death. Damage of lung tissue due to combustion aerosols and subsequent pneumonia often leads to progressive lung failure. As complications, dysfunction of other organs (eg, hypermetabolic syndrome, haemodynamic instability, generalised oedema or sepsis) may develop. This syndromal status is termed ‘burn disease’.1 From an immunological point of view, burn disease resembles septic shock, or ‘systematic inflammatory response syndrome’ (SIRS).2 3 As a rule, burn disease shows a biphasic course of symptoms. In an initial hyperdynamic phase, toxic metabolites that arise in the body during the burn process induce an activation of macrophages, monocytes and neutrophils. This provokes an uncontrolled release of pro-inflammatory mediators like tumour necrosis factor α (TNF-α), complement C3 and C5a, interleukin-1 (IL-1), IL-6 and IL-8 as well as γ-interferon.4 5 In a second hypodynamic phase, anti-inflammatory mediators like IL-10, IL-11 and IL-12 predominate.6

The pathophysiology of burn disease has been intensively studied, mostly with regard to clinical aspects.1 There have been a number of papers about the pathological findings that result within survival times of at least several hours after exposure to fire.2 3 7–10 Only recently, we were able to demonstrate similar morphological changes in the lung parenchyma of burn deaths with much shorter survival times. These early findings were interpreted as first steps towards an SIRS syndrome.11 So far, it has remained unclear whether other pathophysiological mechanisms beyond the activation of cell adhesion molecules are necessary for the development of SIRS in these lethal cases with short survival times. Therefore, the aim of the present study was to investigate the presence of the inflammatory mediators TNF-α, IL-8 and inter-cellular adhesion molecule 1 (ICAM-1) in the lung parenchyma of acute fire fatalities, and in control groups with postmortem burn or not exposed to fire.

TNF-α is one of the most important mediators of systemic inflammation and is present in a variety of cells and tissues. It is released within a few minutes after local or systemic tissue damage. As a potent activator of neutrophils and macrophages, TNF-α may activate the coagulation cascade, and induce fever and a systemic vasodilatation leading to hypotonia. TNF-α may also stimulate the secretion of acute phase proteins in the liver, and the release of other pro-inflammatory mediators, especially ICAM-1, and IL-1 and IL-6. IL-8 acts as an activator of neutrophils, inducing chemotaxis and enhancing the expression of adhesion molecules. ICAM-1 may be found in macrophages, lymphocytes and endothelial cells. It induces the binding of leucocytes to endothelial cells and stimulates cell migration. All three mediators under investigation are supposed to play an important role in the morphological changes in the bronchial system after thermal trauma.10 12–14

Materials and methods

Among the forensic autopsy cases investigated at the Institute of Legal Medicine of the Freiburg University Hospital between 1996 and 2006, four distinct groups were collected (table 1): (1) fire fatalities (FF); (2) postmortem burns (PMB); (3) deaths by haemorrhagic shock (HS); and (4) railway suicide victims (RS).

Table 1

Specific features of the four death groups studied

The FF group consisted of persons who had died of massive intravital exposure to fire with short survival times of less than 1 h, but without additional mechanical trauma to the lungs. The forensic findings proved that burn shock had been the cause of death in all group members. Their CO-Hb concentrations were raised (>10%) in 29 cases, and were 18% on average. With the exception of two cases, all fire victims showed burns of the mouth, pharynx and larynx. Ten cases also showed burns of the trachea. The PMB group comprised cases with a defined, not burn-related cause of death and lack of vital reactions to fire exposure that happened only postmortem. In all cases of the HS group, the survival time was <1 h. There was no mechanical traumatisation of the lungs. The small RS group was formed of five persons who committed suicide by railway run-over. Death had occurred instantly due to severe polytrauma, and often included mechanical injuries of the lungs.

In each case, lung tissue probes of all lobes including bronchi and pleura were taken at autopsy. The samples were fixed in formalin, and embedded in paraffin. Tissue sections of 4 μm in thickness were stained with H&E. The microscopic changes of the lung tissue, especially the quantity of macrophages, the presence of intra-alveolar oedema, or acute hyperaemia were detected by a pathologist (JB) under blinded conditions. The immunohistochemical stainings were performed with antibodies against TNF-α (sc-52746, dilution 1:50), IL-8 (sc-107, 1:50) and ICAM-1 (sc-73321, 1:800). All antibodies were provided by Santa Cruz Biotechnology (Santa Cruz, California, USA). Antigen retrieval was carried out using the PT-module and a buffer at pH 9 (DAKO S 2367) following incubation with the antibodies for 30 min. For visualisation of bound antibodies we used the labelled streptavidin-biotin method, DAKO K 5005, AP/RED. The expression intensity of all 264 slides stained immunohistochemically was semiquantitatively graded by JB. A staining reaction was ‘negative’ (0) when there was no, and ‘mild’ (1+) when there was only a weak, but perceptible cytoplasmic positivity. A ‘moderate’ grade (2++) signified a distinct positive staining, and ‘strong’ (3+++) meant that the positive cells and structures were clearly conspicuous even against an unspecific background.

The differences between the FF group and the HS group were statistically analysed. For statistical comparison the Wilcoxon–Mann–Whitney test and the Kruskal–Wallis test were used, with an error probability of p<0.05 being considered as significant.


Among the H&E-based morphological parameters under evaluation (table 2), the number of intra-alveolar macrophages did not differ significantly between the FF cases and the control groups. Intra-alveolar oedema was generally stronger in deaths related to fire exposure (intra-vital or postmortem) than in the other groups, and was different on a highly significant level (p<0.009) between the FF and HS groups. Acute blood congestion was strongest in HS cases (p<0.026). In no case were inflammatory cell infiltrates present. TNF-α staining was positive mostly in macrophages (figure 1A, table 3), and often also in erythrocytes that were present in blood vessels (figure 1B). The positivity of TNF-α in erythrocytes was found in all cases of the FF group, in which 10–40% (mean 24.1%) of the red cells showed a diffuse cytoplasmic TNF-α staining. In the PMB group, TNF-α positivity of erythrocytes was present in all cases except one and was less frequent (range 0–30%, mean 13.6%), whereas in the HS and RS groups, only a minority of the cases showed a weak TNF-α positivity (HS: 0–30%, mean 4.8%; RS: 0–10%, mean 3.0%). TNF-α immunostaining of macrophages was strongest in FF and was able to discriminate this group from the PMB group on a highly significant level (p<0.001, figure 2). The expression patterns of IL-8 in the groups of deaths showed non-significant differences, and was slightly stronger in both fire-related groups (FF and PMB) than in HS. Sometimes, an expression of IL-8 was also found in the epithelia of minor bronchial glands and of pneumocytes (figure 1C). In all groups, only an inconstant and weak staining was found with ICAM-1 (figure 1D).

Table 2

Grade of morphological changes in the four different autopsy groups; mean values of semiquantitative evaluation according to the H&E staining

Figure 1

Staining pattern in the group of fire fatalities. (A) Tumour necrosis factor α (TNF-α) expression of intra-alveolar macrophages (×400). Inset: technical control of TNF-α staining in a lymph node with sarcoidosis. Note positivity in multinucleated giant cells (×400). (B) TNF-α positivity in some 30% of erythrocytes (and in a few leucocytes that have larger size and nuclei) in a blood vessel. For comparison: serial section stained with H&E (both ×400). (C) Groups of intra-alveolar macrophages showing interleukin-8 positivity (×400). Inset: technical control of IL-8 staining in a lymph node with sarcoidosis. Note positivity in multinucleated giant cells (×400). (D) Group of intra-alveolar macrophages showing weak expression of inter-cellular adhesion molecule 1 (ICAM-1). There is strong diffuse ICAM-1 positivity within the alveolar walls, primarily along basement membranes and in endothelial cells of capillaries (×400). Inset: technical control of ICAM-1 staining in a lymph node with sarcoidosis. Note diffuse positivity of epithelioid cells surrounding multinucleated giant cells (×400).

Table 3

Immunohistochemical marker expression in lung macrophages (in %)

Figure 2

Mean rate of tumour necrosis factor α positivity of erythrocytes in the different death groups (in %).


The effects of an inhalation of gaseous products of combustion on lung function have been extensively studied in animal models.8 10 15–18 These studies showed an increase of lymph flow owing to damages to the endothelia, and a higher permeability for proteins through the walls of blood vessels, leading to interstitial oedema. These effects became manifest within several hours after the damaging event.19 Within the first 24 h after combustion gas inhalation, the morphological changes of the alveolar epithelia were only discreet, but functionally relevant with a deterioration of the alveolar clearance and of the barrier function leading to an increased permeability for proteins.12 The main pathway for the development of these changes seemed to be the accumulation of activated neutrophils and macrophages.14 20 21

Immunohistochemical studies of burn victims have shown an activation of the adhesion molecules P-selectin (CD62P), PECAM-1 (CD31), von-Willebrand factor,11 heat-shock-protein 7022 and fibronectin.23 These findings suggested that there may be incipient systematic inflammatory reactions even very shortly after burn trauma. According to that we found that intra-alveolar oedema was significantly stronger and more frequent in burn victims than in fatalities of haemorrhagic shock. Resuming these previous studies we investigated the expression profiles of TNF-α, IL-8 and ICAM-1 as pro-inflammatory mediators that are known to play a role in inhalation trauma.10 12–14

An interesting finding in our study was the positivity of TNF-α in erythrocytes, which was on a highly significant level more often present in fatalities exposed to fire, and also significantly more frequent in FF than in PMB cases. We suggest that erythrocytes in vessels that are located close to the airways might be damaged by the thermal trauma, thus becoming susceptible for permeation or adherence of TNF-α. Damaged erythrocytes are known to interact with monocytes in the activation of cytokine responses.24 Otherwise, our immunohistochemical studies did not show significant differences between burn victims and deaths unrelated to fire. Although TNF-α may be activated within a few minutes after tissue damage,14 the survival times of <1 h of the cohorts may have been too short to reach the phase of leucocyte immigration necessary for further morphological changes. Thus, microscopically no reactive inflammatory cell infiltrates were found in our cases. Moreover, it is possible that not only the inhalation of heat and combustion gas, but additional factors like endotoxins20 25 26 may be necessary to increase TNF-α expression in the lung. Since ICAM-1 becomes activated via TNF-α, the absence of ICAM-1 expression in our cases is plausible, as it should become manifest only at a later stage of the inflammatory process.

So far, the role of TNF-α in the development of lung parenchymal changes after burn trauma is not clear. Cox et al found an increase of TNF-α in bronchial glands of sheep within 8 h after combustion gas inhalation and burn trauma.10 Hales et al were not able to prove any influence of TNF-α in smoke-induced microvascular lung injury in sheep, but found an increase of TNF-α in lymph fluid of the lung after injection of endotoxin.25 Some other authors27 28 were also able to demonstrate an increase of TNF-α in burn disease in men, but this might have been due to bacterial superinfections.20 26

The mild increase of IL-8 expression in our FF group may reflect a non-specific immune response of the lung tissue to burn trauma. Increased alveolar concentrations of IL-8 were also described in cases with lung failure after mechanical traumatisation.28 Wright found that the inhalation of combustion gas alone did not induce an increase of alveolar IL-8.20 Raised concentrations of IL-8 were found in the serum of patients 2–4.5 days after burn trauma.29 In this cohort, the extent of increase of IL-8 of patients was highly dependent on the presence or absence of sepsis. This observation, too, suggests that up-regulation of IL-8 (as it is the case with TNF-α und ICAM-1) may not be caused by thermal trauma or combustion gas inhalation alone but may also need additional factors.

In conclusion, our findings show that TNF-α, ICAM-1 and IL-8 are not able to prove the vitality status of burn victims during or within 1 h after exposure to fire. We believe that these inflammatory mediators may need more time (>1 h) after burn trauma to become immunohistochemically detectable, or that other, so far unknown pathogenetic factors in addition to combustion gas inhalation are necessary for their expression. Further studies on larger cohorts are needed to clarify the exact role of inflammatory mediators in the initial phase of burn disease.

Take-home messages

  • In fatal burn victims with only short survival times (<1 h), the immunohistochemical expression profiles of tumour necrosis factor α (TNF-α), inter-cellular adhesion molecule 1 and interleukin-8 do not clearly discriminate between cases of intravital and postmortem exposure to fire.

  • The phenomenon of a positive TNF-α reaction in erythrocytes very shortly after fire exposure may be the first morphological sign of a developing burn disease and needs to be elucidated.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.