Abstract
Keywords
Introduction
The reactions of the human body to traumatic tissue injuries and bacterial infections are similar. The reason of this might be that antigen structures of mitochondria released during injury and the genetic structures of the pathogens attacking the host are very similar. The inflammatory patterns caused by tissue injuries are called DAMPs (damage/danger associated molecular patterns), the reactions caused by microorganisms are PAMPs (pathogen associated molecular patterns).
Physiological DAMPs are located intracellularly, mainly in the nucleus or in the cytoplasm. Due to traumatic events these can express on the cell surface or can get out to the extracellular space [20]. The extracellular functions can be e.g. the stimulation of the inflammatory reaction, the emergency alert of the host. The molecules generating DAMPs include heat-shock proteins, purine-metabolites (ATP, adenosine, uric acid) or high-mobility group box 1 (HMGB-1) [17, 23]. HMGB-1 can form through traumatic, i.e. not programmed cell death, its main function is to enhance the expression of some inflammatory cytokines (e.g.: Interleukin [IL]-1, IL-6, IL-8 or Tumour Necrosis Factor [TNF]-α) and adhesion molecules (e.g.: Intercellular Adhesion Molecule [ICAM]-1, Vascular Cell Adhesion Molecule [VCAM]-1) [1, 10].
The members of the PAMPs are ancient molecular patterns which cannot be found on the surface of eukaryotic cells. Toll-like receptors (TLR) and pattern recognition receptors (PRR) can bear a part in the identification [2]. The molecules that trigger PAMPs work through the elimination of the resting state of the TLRs. This causes the death of the pathogen or infected cells, generating an immune reaction, like the activation of T cells.
The innate immune system recognizes molecules causing DAMPs and PAMPs expression resulting in a rather similar response. According to Bianchi [2] both traumatic and infectious harms effect in tissue and cell damages, therefore, they trigger affinitive reactions. Ensue from this in clinical practice both infection and polytrauma or combustion can effect in similar Systematic Inflammatory Response Syndrome (SIRS). The dysfunction of the non-traumatized organs or – in severe cases – multiple organ failure (MOF) appear due to the escalating SIRS. We can define the criteria of SIRS by the consensus conference of the American College of Chest Physicians/Society of Care Medicine in 1991 and since 2016 we can say that sepsis is a life threatening condition which evolves from the infectious injuries affecting the tissues and organs of the body [22]. Unfortunately, bacterial infection or sepsis is extremely hard to diagnose in patients who have explicit SIRS. In everyday clinical practice the mostly used markers to diagnose and separate SIRS and sepsis are C-Reactive Protein (CRP) and procalcitonin (PCT). TNF-α can be used not routinely as well as IL-1, IL-6, IL-8 and IL-10. IL-1 has proinflammatory effect and plays role in the B-lymphocyte regulation just like in the induction of the acute phase proteins. Macrophages produce IL-6, -8 and -10 molecules which play important roles in the post combustion physiological process [9, 25]. Numerous clinical researches have shown that the level of cytokines released during tissue injury correlate with the severity of injury, the developing MOF and mortality [11]. Frink and associates found correlation between IL-6 levels and the organ dysfunction and mortality caused by polytrauma [12]. Csontos and associates found elevated anti-inflammatory IL-10 levels in burned patients. In their research the early appear of significant anti-inflammatory response (IL-10 elevation) was a good predictor of sepsis or mortality [8]. The importance of in time recognized infection is that the survival of sepsis can be improved by fast and adequate antibiotic treatment in the early hours [21].
The available diagnostic procedures cannot identify perfectly the time of the emerging bacterial infection or sepsis. It would be important to find such a marker or a threshold, perhaps a tendency for the existing markers that can predict the clinical state tending to infection or threatening infection. Infection generates humoral and cellular reactions. The first step of the cellular response is an aspecific immune function, when the neutrophil leukocytes become activated which results in increase of adherence, chemotaxis to the interstitial space. Free radical production, phagocytosis and water uptake are also enhanced. All these changes in their function can be measured by sophisticated methods but the water uptake reduces their specific gravity and it can be measured easily. Previous studies found that anticoagulated whole blood samples from healthy individuals contains zero to 10% more leukocytes than originally in the upper half of the blood column (due to upward flotation i.e. “antisedimentation”) after one hour gravity sedimentation [3, 7]. Leukocytes’ water uptake associated with their activation leads to further decrease in their specific gravity, therefore, antisedimentating ratio will increase above 15% [3, 7]. This ratio is called leukocyte antisedimentation rate (LAR) which can be measured by whole blood gravity sedimentation [4–6]. In severe burns and polytrauma SIRS starts before admitting to the intensive care units (ICU) because of extensive tissue necrosis. The “classic” symptoms (tachycardia, tachypnoe, leukocytosis, fever) and – in a false positive way – the increased PCT and CRP serum levels confirm this. The evolving bacterial infection must be diagnosed in this state, in the first few days of the treatment. This task is difficult since we can’t relay on exact parameters because the existing SIRS. A tool is needed which is more specific and is indicative of emerging earlyinfection.
Patients and methods
Legal ethical aspects and informed consent
Our study’s protocol was based on the ethical directives of the 2003 Helsinki Declaration. After receiving permission from the Regional Research and Ethical Committee of the University of Pécs (4422/2012) an informed consent was obtained from the patient or from his/her relatives.
Inclusion and exclusion criteria
Polytrauma patients who had an Injury Severity Score (ISS) 16 or above and patients who had 20% or more total burned surface area (TBSA) were selected to our study. Exclusion criteria were Baux index (Age + Percent Burn + 17 * (Inhalational Injury, 1 = yes, 0 = no) [16] above 100, age under 18 years, New York Heart Association (NYHA) IV grade heart failure, patient with chronic haemodialysis treatment, liver cirrhosis or portal hypertension, chronic use of steroid drugs, malignant haematological disease and immune suppressive therapy affecting the normal immune response (radio-, chemotherapy etc.). Healthy volunteers with similar age and sex to the observed population were included to the control group. Primary and secondary outcomes were critical care survival and development of septic complications, respectively.
Location and study design
Our prospective, observational study was conducted on acute polytrauma and burn victims admitted to our 10-bed ICU at the University of Pécs, between March 2013 and September 2015. The patients were treated regarding to the latest Advanced Trauma Life Support (ATLS), American Burn Association (ABA) and Sepsis Guidelines [18, 22]. Blood samples were drawn immediately after their admission to ICU (T1) and on the following days till discharge from ICU. The kinetics of parameters were analysed for five days (T1-T5). Due to previous observations [13–14] during the first 5 days after trauma, SIRS reaches its peak and starts to cease in case of uncomplicated events. We assumed that these five days are far enough to monitor the acute changes in LAR, serum CRP and PCTlevels.
Diagnosis of sepsis
When obvious clinical sings of sepsis were observed, blood sampling continued for an additional three days’ time. The diagnosis of sepsis was based on latest guidelines [22]. Day 0 (D0) for sepsis was regarded when the declaration of sepsis was made by clinical signs and microbiological sample was taken and antimicrobial therapy was introduced. Of course blood samples were also drawn before sepsis was diagnosed and on the previous 3 days (D-3 to D-1).
Measurement of LAR
The first step was to draw arterial blood to test-tube containing sodium-citrate (Vacutainer, Becton Dickinson, Meylan, France). After one hour of blood sedimentation the leukocyte count of the total blood column’s upper part (U) and lower part (L) were measured (Fig. 1). For the calculation the equation of LAR = 100×(U–L)/(U+L) was used. This equation represents the number of the leukocytes, in the percentage of the original leukocyte number, which passed upwards the half line of the test tube after one hour of blood sedimentation [7]. With multiple measurement of the same sample the coefficient of variation for LAR measurement estimated 3.2%.
Serum CRP and PCT levels were measured by the Department of Laboratory Medicine, University of Pécs as the part of the patients’ daily routine examination. Serum CRP is the conventional laboratory marker widely used in the everyday clinical practice for the diagnostics of inflammatory processes (reference level of our lab below 5 mg/l). Serum PCT is a commonly used the marker in the identification of sepsis (reference level of our lab below 0.5 ng/ml).
Statistics
Statistical Package for the Social Sciences (SPSS) statistics software, version 22.0 (IBM Corporation, USA) was used for statistical analysis. Data were expressed as median and inter-quartile range (IQR standard 25th–75th percentile) as distribution was not normal according to Kolmogorov-Smirnov test. Kruskal-Wallis test followed by Mann Whitney U test were used for interday analysis.
Patients with healthy controls, survivor (SU) group with non-survivor (NSU) group were compared using Mann-Whitney U test. Jonckheere-Terpstra test was used to detect significant trends across the study period. The incidence of the septic complications in SU compared to NSU was analysed with Fisher’s exact test. Values of p < 0.05 were considered significant.
Results
Demographic data
Sixteen burned (Abbreviated Burn Severity Index, ABSI: 7 (5–8)) and twenty polytraumatized (ISS: 29 (22–34)) were involved in the study. Nine of the patients were women and twenty-seven were men. Twenty-six of them survived ICU treatment (ten burned and sixteen polytraumatized). Their demographic data are shown in Table 1. We found significant differences between the SU and the NSU group in age (p < 0.05) and the extension of the burn surface (p < 0.05). There was also a significant difference in age in burn and polytraumatized patients (p < 0.05). In the same time ISS showed no difference in SU and NSU group in the polytraumatized group. Septic complications occurred in eleven patients (six of them were burned and five were polytrauma victims) in the first two weeks in ICU.
Changes in LAR, CRP and PCT levels
LAR levels increased by time (p < 0.001) and became statistically significant on T2 (p < 0.05), T3 (p < 0.05), T4 (p < 0.001) and T5 (p < 0.001) (Fig. 2a). There was no statistically significant difference between healthy control group and the whole traumatized group on T1. The CRP levels showed an increasing tendency (p < 0.001). Its levels were above the upper level of the normal range from T2 (Fig. 2b). During the observed period PCT levels did not show any statistically significant changes (Fig. 2c) and its median did not exceed the upper limit of the lab.
Differences of LAR kinetics between survivors and non-survivors
In SU group LAR levels increased significantly (p < 0.05) from T1 and reached its peak value on T4. In NSU group a significant elevation was detected from T3 (p < 0.05) only, reached its peak on T4. Compared SU to NSU LAR levels were significantly higher on T1 (p < 0.01) and T2 (p < 0.01). Compared to healthy controls LAR levels were significantly higher on T1 (p < 0.05), T2 (p < 0.001), T3 (p < 0.001), T4 (p < 0.001) and T5 (p < 0.001) in SU group. Compared NSU group to healthy controls a non-significant decrease could be observed on T1 and T2. In NSU group LAR levels were significant higher on T3 (p < 0.05), T4 (p < 0.05) and T5 (p < 0.01) compared to the healthy controls(Fig. 3a).
Differences of serum CRP levels between survivors and non-survivors
CRP levels of SU group were above the reference range of the lab on T1 while NSU group were below it. Compared SU group to NSU group CRP levels were significant on T1 (p < 0.05). CRP levels showed an increasing tendency thereafter in both groups (SU (p < 0.05), NSU (p < 0.05)). The median CRP levels exceeded the upper lab limit from T2 in both groups (p < 0.05) with a peak level in SU group on T3 (p < 0.01) and in NSU group on T4 (p < 0.01). CRP levels started to decrease in NSU group on T5, but it was not statistically significant compared to the peak value. CRP levels were higher in NSU group than in SU group on T4, this difference was significant (p < 0.05) (Fig. 3b).
Differences of serum PCT levels between survivors and non-survivors
Median values of PCT were below the reference level given by the laboratory in SU group during the whole observation period whereas they were slightly above the reference level in NSU group from T3. Significant tendencies were not found neither in SU group nor in NSU group. PCT levels did not show any significant differences between SU group and NSU group during the observation period(Fig. 3c).
Incidence of the septic complications in survivors and non survivors
The incidence of sepsis between SU group and NSU group showed a statistically significant difference (p < 0.05). In SU group five, while in in NSU group six patients become septic in the first two weeks. The average onset time point of sepsis was on T5 (T2-T8) in SU group whereas in NSU group it was on T3 (T2-T3). Compared SU group to NSU group the time difference was notsignificant.
Changes of LAR, serum CRP and PCT before and after clinical diagnosis of sepsis
LAR (Fig. 4a) and serum CRP levels (Fig. 4b) showed a decreasing tendency in eleven patients who become septic in three days (D-3 to D-1) before clinical diagnosis of sepsis (D0). Compared LAR on D-3 to D-1 and D0 LAR levels were significantly lower on D-1 and D0 (p < 0.05). The drop of CRP levels were significant (p < 0.05) too on D-1. Serum PCT levels did not show any specific kinetic before the clinical diagnosis of sepsis. Its levels were increased significantly (p < 0.05) from D0only (Fig. 4c).
Discussion
The main findings of the study: the patients who failed to increase LAR levels from Day 2 had higher risk for septic complications and consequent mortality. a sudden drop of LAR levels may predict the onset of sepsis earlier than the other clinical signs.
In the everyday intensive care LAR is not a frequently measured parameter, therefore, we have little reference about the characteristics of LAR in traumatized patients. According to our knowledge it is the first paper examining LAR kinetics in polytrauma and burned victims. In our study leukocyte activation was found due to injury (mechanical or thermal). Compared SU to NSU and healthy controls this change was statistically significant form T1. These findings are differing from Molnar and associates results [15] who detected early (within 6 hours from admission) LAR elevation in patients suffering from ischemic stroke. In our study the first blood sample was taken at the patient’s arrival to our ICU. This very early blood sampling can explain the missing LAR elevation on T1. Bogar and associates [5] found significantly higher values of LAR on the second postoperative day in patients admitted to ICU after oesophagus tumour resection and in whom respiratory insufficiency developed on the 3rd postoperative day. The mortality rate of these patients was higher too. Molnar and associates found a weak positive correlation between the severity of stroke and the LAR values. In this regard LAR might have a predictive value [15]. Our results showed significant differences between SU group and NSU group in the kinetics of LAR. SU group showed a significant elevation of LAR from D1 while in NSU group this tendency was present from D3 only. We assumed two possibilities behind this. The first one is the decreased immune response. These findings are identical with the above mentioned results of Csontos and associates [8] who found an elevated anti-inflammatory cytokine expression profile (elevated IL-10 levels) in non-surviving burned patients compared to survivors. On the other hand as we sampled circulating leukocytes, it is possible, that the activated leukocytes had passed the capillary wall – so maybe we detect the inverse of the real processes. Because LAR mirrors the cytokine kinetics the first possibility is more feasible. Bogar and associates found statistically significant differences in the median values of LAR among patients treated at the ICU [3] due to bacteremic or non-bacteremic fever. Molnar and associates found that in patients with septic complications the lack of increase in LAR was found on the second day and it was the sign of worse neurological outcome [15]. In NSU group a decrease of LAR levels was seen on T1 and T2 in our study. The significant drop in LAR levels on the day before and on the day when sepsis was supports our hypothesis together with the above mentioned studies that LAR can predict septic complications. Our findings are supported with the fact that in the observation period the PCT levels did not showed any kinetics before the developmentof sepsis.
CRP is a part of the acute phase proteins, its increased production can be seen in inflammatory conditions. As serum CRP levels reflect to the inflammatory processes, we may presumed that not the extravasation of the leukocytes rather the late onset of the inflammatory processes or temporary inhibition of the inflammatory response before the development of sepsis might be responsible for the observed changes in the LAR kinetics. In patients whom sepsis was developed we saw a decreasing tendency before the onset of sepsis, which was statistically significant too.
The serum PCT levels had no prognostic value in the prediction of the development of sepsis in our study. PCT become significantly elevated at the onset of sepsis. According to the references PCT is a good marker of guiding and termination of antibiotic treatment.
Limitations of our study
Eleven patients who became septic after burn injury are not enough to observe their kinetics around the development of sepsis – further patients need to be involved. In our case we faced to the problem with anyone who is dealing with sepsis at the time we have no adequate laboratory parameter, radiological finding or any other sign that confirm the presence of infection, so in our practice we can rely on clinical signs only. The microbiological results for declaration of sepsis were not useful too, due to time delay, sample contamination or the possibility of false negative findings that may not be ruled out.
Conflict of interest
The authors have read the journal’s policy and declare the following: no potential conflict of interest relevant to this article can be reported.
Dedication
The authors dedicate the present scientific contribution to the 650th anniversary of the foundation of the University of Pécs, Hungary.
Footnotes
Acknowledgments
We wish to thank for the cooperation of the University of Pécs provided by the following units: Department of Anesthesiology and Intensive Care, 1st Department of Internal Medicine, Department of Laboratory Medicine, Department of Burn Care Surgery, Department of Traumatology and Hand Surgery.