Federal Scientific Clinical Center of Miners’ Health Protection,

Leninsk-Kuznetsky, Russia

 

Infectious complications are one of the main causes of mortality in critically ill patients who experienced shock and blood loss [1, 2, 3, 4]. Opportunistic pathogenic gram-positive and/or gram-negative microflora is the common infectious agent [4, 5, 6]. In case of massive bacterial contamination or non-effective local response the infection dissemination appears with clinical manifestation of generalized intravascular inflammation ‒ systemic inflammatory response syndrome (SIRS) [7].

At the present time many researchers and practicing physicians perform diagnostics of septic complications with definitions and criteria by Bone R.C. (1992) [8] which were approved at American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference (ACCP/SCCM) [9]. The conception by Bone R.C. was revised once and again, however, without significant changes. In 2001 another revision took place, with adding other signs, symptoms and laboratory values (C-reactive protein, procalcitonin, glucose, lactate) indicating possible sepsis [10]. However the offered sepsis biomarkers are non-specific. For example, procalcitonin increases without infection, for example, after severe injury, with massive cell death [11]. It causes some difficulties for early sepsis diagnostics in polytrauma and determines interest for searching other markers.

Previously, we studied the possibility of the original criteria by Bone R.C. (1992) for sepsis diagnostics in patients with polytrauma [12]. We showed that amount of SIRS signs had important predictive significance for infection development, clinical outcome and duration of treatment for critically ill patients. The important role of gram-negative opportunistic strains in development of infectious complications was found. Gram-negative microflora influences on the macroorganism by means of external membrane component ‒ lipopolysaccharide-binding protein (LPS-BP), which releases as acute phase reactant to appearance of lipopolysaccharide in the blood, and its level indicates intensity of bacterial contamination.

In the sepsis criteria, 2001 revision, increased lactate level is offered as tissue hypoperfusion indicator. Some authors consider this metabolite as indicator of inflammation [13] or metabolic stress [14]. Blood lactate monitoring is widely used in clinical practice, but its diagnostic significance for critically ill patients is not determined to the full degree [15].

Study objective ‒ to assess clinical and predictive significance of criteria for systemic inflammatory response syndrome (SIRS), levels of inflammatory markers, LPS-BP and lactate in the blood of critically ill patients.

 

MATERIALS AND METHODS

Clinical examination included 1,565 patients with polytrauma (men ‒ 70.3 %, women ‒ 29.7 %, mean age of 42.2 ± 2.23) who were admitted to Clinical Center of Miners’ Health Protection within 2 hours after trauma in 2003-2013 (table 1).

Table 1
Characteristics of patients

Note: 1 – arithmetic average ± error of mean; 2 – median (interquartile range: 25 %–75 %);
2 – median (interquartile range: 25 %–75 %).

Traumatic shock of degrees II-III (APACHE III > 80) was diagnosed in all patients on admission. The supposed blood loss was 1,200-2,500 ml (20-50 % of circulating blood volume (CBV)). The individual estimation of blood loss was performed according to the sum of external and cavitary blood loss with consideration of approximate blood loss in fractures. The inclusion criteria were age of 16-65, severe multiple or concomitant injuries (ISS > 30) and supposed blood loss of 20 % of CBV.

The control group included 15 virtually healthy persons at the age of 20-50.

The research program was realized with microbiological and laboratory techniques at days 1-3, 5-7, 8-10, 11-14 and 17-21 after admission to intensive care unit. Each day the data was registered including microbiologic, clinical infections and antibiotics administration. For identification of bacterial infections inoculation of different biomaterials was performed according to the active order #535 by USSR Ministry of Health from 22.04.1985. The identification of microorganisms was performed with Vitek 2 bacteriologic analyzer (bioMerieux, France). Serum procalcitonin (PCL) and C-reactive protein (CRP) were assessed with Cobas 6000 SWA modular platform (Switzerland). The serum levels of LPS-BP, tumor necrosis factor alpha (TNF-α) and interleukin-6 (IL-6) were assessed with IMMULITE ONE immunochemiluminiscent analyzer (USA) using DPC reagents (USA). Lactate levels in the whole venous blood were assessed with Roche Omnis S analyzer of critical states (Germany).

At the end of follow-up period (day 21) all patients were retrospectively distributed to the groups according to maximal manifestation of SIRS and in compliance with ACCP/SCCM criteria [9]: SIRS (n = 575), local infection (n = 360), sepsis (n = 270), severe sepsis (n = 120), septic shock (n = 60). Classification was performed by two physicians who did not participate in treatment of the patients. SIRS was diagnosed at presence of more than one of the following symptoms: body temperature < 36.0 °C or > 38.0 °C, respiratory rate > 20 per minute or PaCO₂ < 32 mm Hg, heart rate > 90, leukocyte count < 4,000/ml or > 12,000/ml or > 10 % of immature forms. A case was considered as infection after identification of infection source and microbiologic confirmation, as well as after identification of microorganisms in normally sterile tissues. In presence of supposed or confirmed infection with at least two SIRS criteria sepsis was diagnosed, and severe sepsis ‒ in case of signs of acute dysfunction of one or multiple organ systems. Septic shock was classified as severe sepsis with signs of tissue or organ hypoperfusion and arterial hypotony, which was not removed with infusion therapy.        

The statistical preparation of the data was performed with IBM SPSS Statistics 20. The quantitative variables are presented as me (LQ-UQ), where Me ‒ median, (LQ-UQ) ‒ interquartile range (LQ 25 %, UQ 75 % quartiles). The qualitative signs are presented as absolute and relative (%) values. Kruskal-Wallis non-parametric test was used for comparison of several independent groups according to quantitative signs. Intergroup differences were identified with Freedman’s one-way ANOVA test. Dannet’s test of multiple comparisons was performed in case of identification of differences (comparison group was SIRS group (in the first case) or basic data (in the second case)). The description of qualitative signs was realized by means of calculation of absolute and relative frequency. For comparison of qualitative values Fisher’s exact test and χ²-test were used. Logistic regression was used for identification of relationships. ROC-curve was analyzed for evaluation of diagnostic information capacity of the tests. The differences were statistically significant with p < 0.05.

 

RESULTS AND DISCUSSION

The analysis of clinical manifestations of systemic inflammatory response syndrome showed presence of two-three SIRS criteria (increased respiratory and heart rates, leukocytosis) in 83.8 % of the patients with polytrauma on admission and during the following 3 days.

There were no significant differences in mortality rates according to individual signs of SIRS.

Only 28 % of the patients with polytrauma had no complications. Acute respiratory distress syndrome (ARDS) was observed in 25 % of the patients. In 47 % of the patients the course of traumatic disease was complicated with different purulent septic complications. The favoring factors were circulation disorders, severe metabolism disorders, single-step formation of significant number of injured tissues initiating inflammatory response, insufficient immune protection.

The bacteriologic examination of different biomaterials did not show any microflora growth during 3 days. At days 5-7, 783 patients showed signs of infection joining in view of respiratory organ inflammation (purulent bronchitis (25.22 %), pneumonia (14.3 %)), wound purulence (18.8 %), urinary tract inflammation (12.6 %), different types of necrosis (6.7 %), peritonitis (6.5 %), osteomyelitis (6 %), meningoencephalitis, meningitis (6 %), sepsis (3.9 %) which were accompanied by identification of microflora in diagnostically significant titer. Microorganisms were identified in view of monocultures (34 %) and associations (66 %). Moreover, gram-positive and gram-negative microflora with equal frequency of isolation rate was identified both in the patients with local infection and generalized infectious process (table 2). At days 8-10 inoculations of different biomaterials did not find microflora growth in the patients with local infection, but in the patients with sepsis the positive results with prevailing gram-negative bacteria were noted in 100 % of the cases. During ICU stay the severe sepsis course was characterized with gradual replacement of antibiotic sensitive microflora to hospital multiresistant: gram-negative Pseudomonas aeruginosa, Acinetobacter spp, extended spectrum beta lactamases and Klebsiella ðneumoniae positive strains, in association with methicillin-resistant S. aureus (MRSA). It is consistent with the data from other researchers [1, 5, 17, 18, 19].

Table 2
The amount of positive inoculations (abs.) in patients with polytrauma in different periods of observation and in dependance on severity of septic complications

At days 17-21 microflora growth was noted only in the group with septic shock, in 100 % of the cases in view of microbial associations (P. aeruginosa and Acinetobacter spp, in 66.6 % ‒ in association with K. ðneumoniae, in 33.4 % ‒ with MRSA).

The results of LPS-BP examination showed its significant increase during the first 3 days of observation compared to healthy individuals (22.0 (19.5-23.2 ug/ml). It can suppose translocation of endogenous microflora as result of severe injury (table 3). During the following observation the levels of LPS-BP gradually decreased in the SIRS group. In the local infection group the decrease in serum levels of LPS-BP was noted beginning from days 11-14, whereas the patients of the septic group did not show significant changes in LPS-BP levels, and it was possibly associated with ongoing bacterial aggression [3].

Table 3
The dynamics of lactate (mmol/L) and LPS-BP (mcg/ml) in the blood of patients with polytrauma (Me (LQ-UQ))

Note: * – statistically significant differences compared to days 1-3; +  – statistically significant differences compared to SIRS group, p < 0.001

Increased levels of LPS-BP in the patients of septic groups could be conditioned by massive translocation of endogenous microflora and attachment of secondary infection [3]. Possibly, they reflect excessive LPS stimulation of bacteria of monocytic macrophage cells. In the future it could lead to disruption of autoregulatory mechanisms and favor infectious process generalization [20]. So, it is known that LPS of gram-negative bacteria induces release of proinflammatory cytokines by monocytes, macrophages and neutrophils. Proinflammatory cytokines present typical signs of intoxication in view of fever, disorders of hemodynamics and hemostasis, and later ‒ internal organ dysfunction and insufficiency. It is known that LPS play the role in bacterial suppression of immune system. But LPS-BP can induce multiple protective reactions in the macroorganism which acquire pathologic characteristics and can result in endotoxic shock [21]. The proinflammatory effect of plasma LPS-BP is conditioned by modulation of LPS bioactivity. It is provided by means of formation of LPS-BP-LPS complexes and translocation to other LPS binding proteins.

Because of the fact of identification of LPS-BP hyperproduction (its degree was associated with severity of developing sepsis syndrome) in the critically ill patients with polytrauma at days 1-3, the possibility of using of this value as early biomarker of sepsis was evaluated. For this purpose the ROC-curve was constructed. It demonstrates relation of likelihood for positive results of the test (Fig. 1). The ROC-curve analysis showed the cutoff values of LPS-BP levels (at days 1-3) for early prediction of sepsis ‒ 237 ug/ml. The relatively high diagnostic efficiency of the test for sepsis diagnostics was shown: the area under the ROC-curve was 0.867 (95 % CI: 0.792-0.943). The diagnostic sensitivity of cutoff value of LPS-BP was 100 %, the specificity ‒ 76 %. The frequency of occurrence of LPS-BP diagnostic levels was 100 % in the blood from the patients of the septic groups at days 1-3 and 5-7, whereas the fact of contamination with gram-negative microflora was only in 58 % of the patients at days 5-7.

The serum levels of inflammatory markers of procalcitonin, CRP and interleukin-6 increased with increasing severity of developing septic states. However there were no statistically significant differences in TNF-α in the blood serum (Fig. 1).

Figure 1. The levels of inflammation markers in the blood of the patients with polytrauma, considering sepsis severity.

PCT – procalcitonin, CRP – C-reactive protein, IL-6 – interleunin-6, TFN-α – tumor necrosis factor-α.

0 – no SIRS, 1 – SIRS, 2 – local infection, 3 – sepsis, 4 – septic shock.

The results are presented as median, interquartile range and confidence interval (Me, 25-75 %, 95 % CI). 

We performed the investigation of clinical significance of cutoff values of inflammatory markers with consideration of severity of septic states. The ROC-curves were constructed which reflected relation of likehood for positive results and demonstrated sensitivity and specificity of the tests for diagnostics of septic states.

The ROC-curve analysis allowed establishing the areas for LPS-BP, CRP and PCL about 1. It supports high sensitivity of these tests. The high diagnostic specificity is shown for tests of LPS-BP and CRP, as well as presence of ≥ 3 SIRS criteria. The diagnostic sensitivity and specificity for sepsis diagnostics was: LPS-BP (335 ug/ml) – 84 % and 88 % (ROC-curve: 0.88; p < 0.001); CRP (> 26 mg/dl) – 80 % (ROC-curve: 0.81; p < 0.001). For PCL (> 0.35 ng/ml) the diagnostic sensitivity was 90 %, the specificity ‒ 43 % (ROC-curve: 0.707; p = 0.006) (table 4). The amount of SIRS signs had important predictive significance for infection development, clinical outcome and duration of treatment (diagnostic sensitivity and specificity ‒ 58 % and 90 %, area under ROC-curve: 0.816) (table 4).

Table 4
The assessment of predictive value and relationship of inflammation indices

Note: OR – odds ratio: CI – confidence interval

The examination of blood lactate did not show significant intergroup differences in the patients with polytrauma on admission to the hospital (table 2). Later blood lactate level decreased in the patients of the SIRS and local infection groups, whereas it increased in the septic groups and achieved the values of lactate acidosis (> 4 mmol/L) in the patients with septic shock. The received data reflects expression of tissue hypoperfusion and severe disorders of energy production [22, 23, 24]. The strong correlation relationship was found between lactate levels and LPS-BP (Spearman’s rank correlation coefficient ρ = 0.76, ð < 0.001) that testified a possibility of lactate using for estimation of SIRS severity in polytrauma.             

   The increase in blood lactate level was associated with increasing mortality in the patients with polytrauma. So, lactate level < 2.5 mmol was associated with mortality of 5.4 % (95 % CI, 4.5-6.2 %), lactate level of 2.5-4.0 mmol ‒ 6.4 % (95 % CI, 5.1-7.8 %), > 4 mmol – 18.8 % (95 % CI, 15.7-19.9 %).

Therefore, hyperproduction of LPS-BP, PCL, CRP and IL-6 is observed in critically ill patients with polytrauma beginning from the first day after trauma. Its expression is related to severity of developing sepsis. The maximal increase in serum LPS-BP levels is observed at days 1-3 and 5-7 in patients with polytrauma, whereas microbiological examinations of different biomaterials give positive results only at days 5-7 in 58 % of patients with evident infection. The received data testifies the high diagnostic and predictive value of LPS-BP, CRP and the possibility for use as early biomarkers of purulent septic complications. Lactate monitoring allows estimating severity of developing SIRS: increased levels (compared to basic values) show generalization of infectious process.

 

CONCLUSION:

1. In 72 % of critically ill patients with polytrauma the posttraumatic period is associated with development of infectious complications, in 47 % sepsis is diagnosed at days 8-10, and its severe course is characterized with gradual attachment of multiple resistant opportunistic gram-negative microflora (Ps. Aeruginosa and Acinetobacter spp., in 66.6 % in association with Kl. Pneumoniae, in 33.4 % ‒ with MRSA).

2. The hyperproduction of inflammatory markers (PCL, CRP, IL-6) defines a possibility for using these values for diagnostic and predictive aims in patients with polytrauma from the first day of observation.

3. High diagnostic sensitivity (100 %) of LPS-BP cutoff value (in early posttraumatic period: days 1-3 and 5-7) allows its using as an early biomarker of purulent septic complications conditioned by gram-negative microflora.

4. Increase in blood lactate level correlating with serum LPS-BP levels testifies generalization of inflammatory process and can be used as additional diagnostic criterion of sepsis in patients with polytrauma at days 5-7.

5. Estimation of blood lactate level is an independent predictive factor of mortality after polytrauma.