Activated protein C–protein C inhibitor complex as a prognostic marker in sepsis

OBJECTIVES
Up- and down-regulation of inflammatory response was described in blood cells from septic patients, according to the stage of sepsis and the cells evaluated. This study aimed to evaluate the Toll-like receptor (TLR) signaling pathway gene expression in peripheral blood mononuclear cells (PBMC) and neutrophils in patients throughout the different stages of sepsis.


DESIGN
Prospective, observational study.


SETTINGS
Two emergency rooms and two intensive care units in one university and one teaching hospital.


PATIENTS AND CONTROLS
A total of 15 septic patients, five with sepsis, five with severe sepsis, and five with septic shock, in addition to five healthy volunteers were enrolled.


INTERVENTIONS
None.


MEASUREMENTS AND MAIN RESULTS
The Human-TLR Signaling Pathway, which comprises 84 genes related to TLR-mediated signal transduction, was evaluated by real time polymerase chain reaction in PBMC and neutrophils obtained from patients and controls. The fold change for each gene (2(-Delta DeltaCt)) was compared between the groups. Genes with fold changes greater than 2 and significant changes in DeltaCT are reported as differently expressed. The fold change ratios in PBMC gene expression between septic patients and healthy controls revealed a dynamic process according to the stage of sepsis, tending toward down-regulation of the TLR signaling pathway in PBMC in the more severe forms of the disease. However, the differential gene expression was restricted to five down-regulated genes in septic shock patients, which are found in the effector and downstream pathways. Neutrophils showed a different pattern of adaptation. Patients with sepsis, severe sepsis, and septic shock presented a broad gene up-regulation, which included all functional groups evaluated and persisted throughout the stages of the disease.


CONCLUSIONS
TLR-signaling pathway genes are differently regulated in PBMC and neutrophils of septic patients, and are dynamically modulated throughout the different stages of sepsis.

Background The induction of chemokine secretion by fibroblasts is crucial for the migration of leukocytes into the parenchyma of the injured lung. Several bacterial products activate the lung's structural as well as immune cells to produce proinflammatory cytokines and chemokines. We report that elastase from Pseudomonas aeruginosa (PE) evokes IL-8 mRNA expression and protein secretion in nonmalignant culture of human lung fibroblasts by activating the receptor for epidermal growth factor (EGFR) and downstream mitogen-activated protein kinases (MAPK) pathway. Methods We utilized western blot analysis to detect phosphorylation of EGFR and signal transduction intermediates. Northern blot and ELISA analyses were used to determine IL-8 RNA expression and cytokine secretion. Results We found that the enzymatically active PE enhances IL-8 mRNA and protein secretion but does not increase IL-10 or TNF expression. PE induces phosphorylation of the EGFR and the extracellular signal-regulated proteins (ERK1/2) of the MAPK pathway. Pretreatment of the cells with neutralizing antibody to EGFR or the EGFR-specific tyrphostin AG1478 markedly attenuated the PE-induced ERK1/2 activation. PE-induced IL-8 expression is also abolished in the presence of the MEK inhibitor U0126, indicating the involvement of ERK1/2 in this process. Conclusion Taken together, the results show PE could modulate lung inflammation by exploiting the EGFR/ERK/NFκB pathway and enhancing IL-8 production by lung fibroblasts. Background Central venous catheter (CVC)-associated bloodstream infections (BSI) remain a major complication in ICUs. The aim of the study was to evaluate the influence of a structured multimodal intervention programme on the CVC-BSI rate of 38 ICUs in Germany.

Methods ICUs of the 'Krankenhaus Infektions Surveillance System'
showing a CVC-BSI rate above the median were asked to implement a 12-month intervention programme starting in April 2007. The intervention included specific evidence-based recommendations for CVC insertion and use, and involved nurses and physicians. Modules were posters, script and advanced training. The modules' content was composed and distributed according to the 'train the trainer' principle by the National Reference Center for Surveillance of Nosocomial Infections. Infection rates were calculated before (January 2005to June 2006) and during the intervention (May 2007to March 2008. Results Thirty-eight ICUs participated in the study. The ICUs had a median of 11 beds and eight ventilator beds. The majority of ICUs (47%) were affiliated to teaching hospitals; 30% were affiliated to university hospitals. The CVC utilization rate before implementation of the intervention was 69.4 CVC-days per 100 patient-days. The pooled mean CVC-BSI rate was 2.9 CVC-BSI per 1,000 CVCdays. A preliminary analysis of the data obtained during the intervention period showed a decrease of the mean CVC-BSI rate in the participating ICUs (2.2 CVC-BSI per 1,000 CVC-days; relative risk = 0.77, 95% confidence interval = 0.63 to 0.94, P = 0.011), whereas the CVC utilization rate remained almost unchanged (69.0 CVC-days per 100 patient-days). Conclusion A structured multimodal intervention programme in addition to ongoing surveillance activities led to a significant decrease of CVC-associated BSI rates.
Background We analysed the clinical utility of the standardised, Conformite Europeanne-certified, multiplex real-time PCR assay for the molecular diagnostics of sepsis that was approved for in vitro diagnostics use (LightCycler SeptiFast assay; Roche Diagnostics, Pleasanton, CA, USA). The SeptiFast assay enables detection of DNA from 25 human pathogens (Gram-positive and Gram-negative bacteria as well as fungi). Materials The study enrolled 50 patients with clinical diagnosis of sepsis that received medical care at the University Hospital for Infectious Diseases, Zagreb and Zagreb University Clinical Center in Croatia. Ten patients were treated at the Department of Haematology following bone marrow or peripheral blood stem cell transplantation; 30 patients were hospitalised in the ICU and 10 patients outside the ICU. All patients enrolled in the study were already receiving empirical antimicrobial therapy at the time of testing.
Methods Peripheral blood samples from the patients were analysed using the LightCycler SeptiFast assay and cultivation. Results Fifteen out of 50 (30%) samples tested positive with the SeptiFast assay for bacterial or fungal DNA. Gram-negative bacteria were detected in 13 of 15 samples (Klebsiella pneumoniae/ oxitoca, n = 6; Escherichia coli, n = 3; Pseudomonas aeruginosa, n = 4). Gram-positive bacterial DNA (Enterococcus faecium, n = 1 and Streptococcus pneumoniae, n = 1) was detected in two patients with polymicrobial sepsis (in combination with K. pneumoniae/oxitoca in both patients). Aspergillus fumigatus DNA was detected in two patients. Six out of 50 samples (12%) were positive by both SeptiFast assay and culture. Additional SeptiFastpositive results (negative by cultivation) were obtained in nine of 50 patients (18%). Four out of 50 samples (12%) tested negative by the SeptiFast assay but were positive by culture. Those results were interpreted as false negative molecular testing. The remaining 31 samples tested negative by both SeptiFast assay and culture. In the group of 10 haematological patients, SeptiFast results were positive for six of the 10 patients (60%), whereas blood cultures were positive in only two out of 10 patients (20%). Conclusion We conclude that the SeptiFast assay is a clinically valuable add-on to conventional culture methods for rapid aetiological diagnosis of sepsis in patients where the empirical antimicrobial therapy has already been started and pretreatment blood cultures were negative.

P9
Eosinophilia as a marker of adrenal insufficiency in critically ill patients with severe septic shock: Background Adequate adrenocortical function is essential to survive critical illness. The number of circulating eosinophils has been proposed as a marker of adrenocortical function. The goal of the present study was to determine whether eosinophilia could serve as a useful and early marker of adrenal insufficiency in critically ill patients with severe septic shock. Methods During a 1-year period, we studied prospectively all 294 patients admitted to our ICU. Sixteen patients (13 male/three female, 5.4% of admissions) with eosinophilia defined as more than 3% of the white blood cell count and severe septic shock, refractory to fluid and vasopressor resuscitation, were included. A high-dose (250 μg, intravenously) corticotropin stimulation test was performed in all included patients.

Results
The mean age was 47.2 ± 18.7 years, the Acute Physiology and Chronic Health Evaluation II score on admission day was 18.6 ± 6.8 and the Sepsis-related Organ Failure Assessment score was 10.3 ± 2.7 on eosinophilia day. The mean eosinophil count was 6.9 ± 3.5% of white blood cells. Eosinophilia was present 1.9 ± 0.9 days (range 8 hours to 4 days) before the onset of septic shock. Multidrug-resistant Gram-negative bacteria in 14 patients, Gram-positive in three patients and fungi in two patients were isolated and considered responsible for sepsis. Baseline cortisol levels were 19.4 ± 8.1 μg/dl and the adrenal response to the corticotropin stimulation test was 8.3 ± 4.9 μg/dl above baseline. Eleven out of 16 patients failed to respond to the corticotropin stimulation test above the critical level of a 9 μg/dl rise, and two out of 16 patients had baseline cortisol concentration <10 μg/dl. A hydrocortisone infusion (300 mg/day) treatment resulted in haemodynamic improvement in 12 out of 16 patients (75%). The 28-day mortality (following the onset of septic shock) was 43.7%. The only independent predictor of death was age (P = 0.027). Conclusion Relative eosinophilia may be considered a useful and early bioassay for adrenocortical function assessment in critically ill patients with severe septic shock and assumed adrenocortical depression.

P10
Polymerase chain reaction detection of sepsis-inducing pathogens in blood using SepsiTest™ Background PCR enables the identification of bacterial DNA in culture-negative samples from patients with suspected infection, allowing the confirmation of, for example, meningitis and septic arthritis. Gross discrepancies in the incidence of positive results between culturing and PCR have been reported, the latter corresponding better to bacterial loads observed by immunofluorescence microscopy and inflammatory response measurements. The goals of PCR assaying of clinical samples for pathogens are improved disease surveillance, early guidance on appropriate antibiotic therapy and patient management. Methods SepsiTest™ is a new PCR test for the presence of bacterial and yeast pathogens in whole blood samples. The test combines sample preparation, the directed extraction of pure pathogen DNA from 1 ml blood, with PCR assays for the universal detection of bacteria and yeasts based on the amplification and monitoring of 16S and 18S rDNA sequences, respectively. Blood from septic patients was extracted and analysed, using SepsiTest™ together with sequencing of amplicons from positive samples and online BLASTN analysis for the identification of pathogens.  (Table 2).

Conclusion
An educational program and a retraining process when linked with SSC implementation were effective in producing process change in the management of severe sepsis, achieving better compliance with currently accepted best practice.

S6
Methods We used a new multiplex PCR-based test, LightCycler ® SeptiFast ® Test MGRADE (Roche Diagnostics, Penzberg, Germany), including software for simultaneous detection of sepsisrelevant microorganisms directly from blood, according to the manufacturer's recommendations. We followed the standard manufacturer's recommended procedure for extraction and compared it with the procedure using the MagNa pure ® compact (Roche Applied Science, Penzberg, Germany) extraction column based on magnetic nanoparticles. Samples for blood culture analysis were drawn from a fresh venipuncture site according to the common guidelines (DGHM). Blood cultures were analyzed using a semi-automated blood system (BacT/ALERT ® ; bioMerieux, Marcy-Etoile, France). VITEK II (bioMerieux) was the system used to phenotypically identify pathogens growing from blood cultures and other microbiological tests (stool, urine, respiratory samples, catheter, etc.).

Results
The results were both retrospective and observational. The quality of PCR results versus other laboratory findings and clinical evaluation, using sensitivity and specificity analysis and kappa evaluation to compare the results obtained, was determined. One hundred determinations were studied comparing both methods for extraction.
the positive troponin group, every other variable was not significantly different, rendering the whole group very homogeneous. Kaplan-Meier survival analysis within 28 days of patient inclusion is shown in Figure 1 as stratified by troponin positivity (>1.0 ng/ml). Troponin-positive patients showed significant increased mortality with a log-rank value of 0.0072.

Conclusion
The elevations of troponin observed were mostly small to modest, reflecting minor cardiac injury, but they nonetheless presaged increased mortality very early in the course of the disease. Others have postulated that increased-troponin patients can probably benefit most from drotrecogin-α administration with mortality reduction, thereby rendering troponin determination mandatory in critically ill septic patients. Troponin should therefore probably be included as an early routine test in the Surviving Sepsis Campaign.

P14
Pulmonary macrophage migration inhibitory factor: an important contributor to age-dependent inflammation in sepsis Methods Saline alone (control) or LTA 1.5 mg/kg and PGN 5 mg/kg were instilled intratracheally into Fischer 344 rats 6, 18, or 24 months old (equating approximately to humans of 18, 45 and 60 years). After 6 hours, the animals were euthanized, blood collected from the left atrium, and the lungs lavaged with saline (bronchoalveolar lavage (BAL)). To further characterize the contributions of MIF following LTA-PGN challenge, we used GeneChips to obtain a description of the global transcriptomic response of lungs from mif +/+ and mif -/mice 6 hours post LTA-PGN. Results There were significant age-related increases in lung compliance and BAL protein content (sham, 0.2 ± 0.01; 6 months, 1.2 ± 1.4; 18 months, 5.2 ± 0.7; 24 months, 4.8 ± 0.4 mg/ml), suggesting a more severe injury in the older age groups. BAL concentrations of GRO-KC (CXCL1) (sham, 0.13 ± 0.03; 6 months, 2.46 ± 0.79; 18 months, 2.89 ± 0.50; 24 months, 4.02 ± 1.54 ng/ml) and IL-6 (sham, 0.1 ± 0.1; 6 months, 5.2 ± 5.6; 18 months, 14.2 ± 2.3; 24 months, 13.0 ± 6.8 ng/ml), and blood MIF concentrations around threefold higher in the older rats suggest an agedependent increase in inflammatory response. Comparative analysis of lung transcriptomes (~8,500 mRNAs) of the mice suggested a larger response in mif -/mice of genes regulated by NFκB. Conclusion These observations suggest, for the first time, a role for MIF-NFκB molecular circuitry modulating cardiopulmonary system responses following pulmonary infection. Since MIF can cause cardiac dysfunction, the increased MIF response during sepsis in the older individual may be responsible for an increased mortality in this patient population.
Since the 1990s, MIF has been known to be a critical mediator of sepsis and septic shock. Circulating concentrations of MIF are elevated in patients with sepsis, and MIF levels are associated with disease severity and fatal outcome. Mouse monoclonal antibodies and rabbit polyclonal antibodies against MIF were shown to be protective in various animal models of sepsis. MIF therefore emerged as an attractive new target for treatment of patients with severe sepsis and septic shock. Methods A diverse panel of human MIF-specific antibodies was generated by selection from a phage display library. This panel was subjected to extensive in vitro testing to identify antibodies that neutralize MIF activity. The antibody showing the highest potential was improved by affinity maturation; that is, by generating modified versions of this antibody by CDR1-2 shuffling and selecting highaffinity variants by phage display. The lead candidate antibody and its affinity maturated variant were tested in experimental mouse models of endotoxic shock and Escherichia coli peritonitis sepsis.

Results
In vitro testing of human anti-MIF antibodies enabled the identification of antibodies that neutralize the activity of MIF in a glucocorticoid overriding activity assay and in a proliferation assay. Antibody Bax94 was designated as lead candidate, and affinity maturation of this antibody led to the generation of BaxA10, a variant with a 10-fold higher affinity for MIF. In an endotoxic shock model, pretreatment of mice with Bax94 and BaxA10 reduced circulating concentrations of TNF (control antibody: 4.9 ± 10.3 ng/ml; Bax94: 0.4 ± 0.4 ng/ml, P <0.01; BaxA10: 0.2 ± 0.5 ng/ml, P <0.0001) and of IL-6 (control antibody: 3.1 ± 1.6 ng/ml; Bax94: 2.1 ± 0.78 ng/ml, P = 0.0003, Bax10: 1.5 ± 0.7 ng/ml, P <0.0001), and increased survival from 21% (control antibody) to 52% (Bax94; P <0.05) and 58% (BaxA10; P <0.05). The protective effect of the two anti-MIF antibodies was also demonstrated in a live E. coli peritonitis sepsis model in which survival rates increased from 10% (control antibody) to 34% (Bax94; P <0.05) and to 56% (BaxA10; P <0.01). Conclusion We have generated fully human anti-MIF antibodies that neutralized the proinflammatory effects of MIF in vitro and that showed significant protective effects in experimental sepsis. In vivo protection tended to correlate with the affinity of the antibody for MIF. There were some differences between groups in comorbidities and surgical procedures for the patients overall; however, these characteristics were more closely matched in the subset of severe sepsis patients. The incidence of renal failure and the ICU and hospital mortalities were similar in the two groups. In multivariate analysis, cumulative doses >33 ml/kg of either HES (odds ratio = 1.85, 95% confidence interval = 1.01 to 3.41, P <0.001) or gelatin (odds ratio = 1.99, 95% confidence interval = 1.05 to 3.79, P = 0.035) were associated with higher risk of renal failure. In severe sepsis, hospital mortality tended to be higher in Group HES than in Group GEL (43% vs. 31%, P = 0.076). Patients with severe sepsis who received a cumulative dose >33 ml/kg of either colloid had a higher incidence of renal failure, which reached statistical significance for the HES group.

P16
Conclusion Moderate cumulative doses of modern HES or gelatin solutions are associated with higher risk for acute renal failure. These solutions should be used with caution, especially in severe sepsis, until their safety can be demonstrated.

P17
Daily blood sampling in septic mice: an optimal and effective monitoring tool Background Daily assessment of circulating immune/inflammatory and organ function parameters is strongly advocated in septic patients. In models of murine sepsis, such monitoring becomes challenging given the limited blood volume available for analysis. We studied the influence of daily versus single sampling in acutely (days 1 to 5) septic mice upon their short/long-term survival, organ function and complete blood count (CBC). We additionally tested the reliability of CBC differential in resuspended cell pellet versus whole blood analysis. Methods Seventy-four female OF-1 mice (18 to 21 g body weight) were subjected to cecal ligation and puncture (CLP). Blood sampling volumes (by facial vein puncture) of 35 μl (n = 40) and 20 μl (n = 34) were tested. The samples were immediately diluted 1:10 to a final volume of either 350 μl (group 1) or 200 μl (group 2). Half of each group was sampled either daily for 5 days or only on day 5 post CLP. For comparison of resuspended versus regular CBCs, 150 μl (of the original 350 μl) was analyzed immediately after sampling. The remaining 200 μl was then spun, plasma removed (180 μl), the cell pellet re-suspended with an equal volume of the diluent and CBC performed. Results Repetitive daily bleeding, regardless of the volume, did not affect either short-term (5 days) or long-term (28 days) CLP mortality. By day 5, changes between groups in the level of circulating IL-6, IL-1 receptor antagonist and organ function/metabolic Critical Care November 2008 Vol 12 Suppl 5 Sepsis 2008 S9 parameters (ALT, LDH, glucose and urea) were identical. In group 1 (35 μl), the red blood cell (RBC) count was reduced by 22% while the hemoglobin (Hb) concentration decreased by 23% (both P <0.05). However, only a minimal decrease of RBC and Hb by 10% and 11%, respectively (both P <0.05), was observed in group 2 (20 μl). In neither group were platelet or white blood cell counts affected by repetitive bleeding. Except for lymphocytes, the comparison of regular and resuspended CBCs displayed a high correlation for all cell types (r >0.9, slope >0.9). On each post-CLP day, the lymphocytes correlation remained moderate, reaching r = 0.6 (slope = 0.6) on average (days 1 to 5). This effect was reproduced when tested in non-CLP OF-1 mice (n = 12) at 1:2 dilution (r = 0.5, slope = 0.7).
Conclusion Although we noted a statistically significant (and inversely proportional) decrease in RBC and Hb after repetitive daily bleeding, its biological impact was probably marginal. Differential blood analysis in resuspended pellet was highly reliable for all (except lymphocytes) cell populations tested. The results indicate that low-volume daily blood sampling allows a multidirectional and minimally invasive monitoring of various immunoinflammatory parameters in acutely septic mice. Background Patients with cirrhosis have higher risk of infection and multiorgan dysfunction syndrome. Increasing evidence relates them to the higher susceptibility to bacterial translocation (BT) by liver dysfunction and hemodynamic changes. Herein, we evaluated the role of BT in acute and chronic portal hypertension (a-PH and c-PH) states without cirrhoses. Methods Forty-eight Wistar rats were distributed in BT, a-PH, c-PH with/without BT and control groups. a-PH (minimal shunting) and c-PH (extensive shunting) were induced by calibrated portal vein stenosis and were submitted to BT on days 2 and 14, respectively, by oroduodenal inoculation (10 ml Escherichia coli R-6, 10 7 or 10 10 colony-forming units (CFU)/ml) and confinement into the small bowel for 2 hours. BT-sham (saline), PH-sham (without portal stenosis), and PH groups were monitored for splachnic (liver, spleen, ileum) and kidney perfusion by laser Doppler, and mesenteric lymph node (MLN), liver, spleen, lung, blood and peritoneal fluid (PF) samples were cultured.

Results
In the BT 10 10 group, the culture was 100% positive at the MLN, liver and spleen (5.3 and 3 log 10 CFU/g, respectively), while the blood, PF and lung were negative. In a-PH animals, the BT 10 10 pattern was 100% to the MLN, liver and spleen (5.4 and 4 log 10 CFU/g, respectively), lung (100%, 3 log 10 , P <0.05) and PF (10%, 0.6 log 10 ). In turn, c-PH-BT10 10 findings were similar to BT10 10 alone but there was an increased translocation to PF (40%, 1 log 10 , P <0.05). On the other hand, for BT 10 7 all cultures were negative, but in PH-BT10 7 translocation occurred to the MLN (a-PH 50%, 1 log 10 CFU/g; c-PH 25%, 0.7 log 10 CFU/g) plus to the PF (a-PH 12.5%, 0.08 log 10 CFU/g; c-PH 25%, 0.16 log 10 CFU/g), evidencing a change in the gut threshold for BT in the PH state. Bacterial challenge in the a-PH state showed that the liver, spleen and kidney go into a hypoperfusion state (-38, -45.2 and -36 Δ%, respectively), in contrast to the ileum hyperperfusion response (+75 Δ%). Similarly, at c-PH the liver and kidney maintained a hypoperfusion state (-17%, -33% Δ%). Based on both saline-PH groups' data, however, the ileum and liver microcirculation functional adaptation was mostly compromised only at the chronic phase. Conclusion Portal hypertension factor without cirrhosis increases and changes the pattern of BT, especially to the lung at a-PH and to the PF at c-PH states, under small bowel Gram-negative bacterial overgrowth conditions. These findings, in addition to the tissue perfusion impairment, might explain the higher susceptibility to infection in portal hypertension diseases.

Background
The gastrointestinal tract has been implicated in sepsis and organ failure by bacterial translocation (BT) and gutimmune system crosstalk with the systemic immunity, but is not yet clearly demonstrated in clinics. Herein we examined the role of gutassociated lymphoid tissue on host inflammatory response by BT as a contributing feature for multiorgan dysfunction in sepsis.

Results
The BT index was not modified by MLFI, but BT alone and sepsis plus BT provoked significant hypoperfusion in all organs plus microcirculation injuries. The MLFI at BTL-G and CL-G completely abrogated tissue hypoperfusion ( Figure 1) and microcirculation injury. The lymph-cell count post BT, sepsis and combined challenges was significantly increased compared with controls although the composition was similar (98% to 99% lymphocytes and 1% to 2% others). The blood-leucocyte count was unchanged in all groups ( Figure 2). TCD3 + over B lymphocytes (CD45RA + ) predominated in both lymph (83%/7%) and blood (68%/3%) (P <0.05), and their percentages did not differ between groups. The CD4 + /CD8 + ratio also did not differ between groups. Naïve cells (CD4 + CD45RC + ) predominated in the TCD4 + population in lymph, and memory cells (CD4 + CD45RC -) in blood. CD8 + CD45RC + cells predominated in both lymph and blood. The proportion of Tregulatory (CD4 + CD8 + /CD25 + ) cells was low in both lymph and blood for all groups (Figures 3 and 4). MLFI completely prevented the death observed in C-G (LD 50 ). Conclusion The gut-associated lymphoid tissue response following bacterial challenge and its crosstalk with the systemic immunity Available online http://ccforum.com/supplements/12/S5 via the lymphatic system is the key factor related to the aggravation of systemic inflammation and death in sepsis. Background Some of the enteric microbiota role is linked to the gut immune system, the angiogenic mechanism and colonization resistance, and its disruption has been correlated to local disease and aggravation of a systemic inflammatory state such as in sepsis and critical illness. Herein we examined the role of diverse sepsis intensities on aerobic and anaerobic facultative Gram-negative microbiota of the small bowel (SB) and large bowel (LB) and subsequent bacterial translocation (BT) potentials in rats. Methods Wistar rats (±200 g) were submitted to varying degrees of monobacterial sepsis (S-7, S-8 and S-9 groups, with 10 7 , 10 8 and 10 9 colony-forming units/ml/100 g body weight of Escherichia coli R6 intravenously, respectively) and samples of the duodenum, jejunum, ileum, cecum, feces, mesenteric lymph nodes (MLN), liver and spleen were harvested at 6, 12 and 24 hours post sepsis and cultured in MacConkey agar medium (n = 6/period/group). Control groups were sham with saline injection and naïve without any procedure (n = 6/period/group). Results The Gram-negative colonization rise within the SB occurred from the proximal to distal compartments and significant overgrowth onset was seen from 6 hours in the SB and 12 hours in the LB in the sepsis groups, suggesting that LB overgrowth was probably due to SB overgrowth. With severe sepsis (S-8, S-9), the overgrowth was more pronounced and remained for a longer period at the ileum and cecum, but at the duodenum and jejunum the peak growth seen in the 6 to 12 hours period returned to normal level at 24 hours. The maximum overgrowth index comparisons between naïve, S-8 and S-9 were (log 10 ): 0.0 versus 2.8 versus 4.7 (duodenum), 2.5 versus 4.2 versus 6.0 (jejunum), 4.0 versus 7.4 versus 7.5 (ileum), and 5.1 versus 8.0 versus 7.8 (cecum), respectively. Spontaneous BT to the MLN occurred only following sepsis (50% at S-8, 5.6% at both S-7 and S-9) and was 100% E. coli. Not only E. coli but all other Gram-negatives were overgrown after sepsis stimulus. S-7 and minor trauma (sham) provoked a transient overgrowth but in lesser intensity and endurance. Conclusion Acute sepsis states induced a significant and transient SB and LB Gram-negative microbiota overgrowth directly proportional to the severity of sepsis. The BT event was dependent on both the sepsis degree and overgrowth factors.

Relationship between sepsis intensity, Gram-negative bacterial overgrowth and bacterial translocation in rats
Critical Care November 2008 Vol 12 Suppl 5 Sepsis 2008

Figure 2 (abstract P19)
Total number of leukocytes in the mesenteric lymph and blood per mm 3 of all groups. *P <0.05.

Figure 1 (abstract P19)
Comparison of the tissue perfusion (jejunum, ileum, liver and kidneys) in Δ% and the mortality index (DL) in all groups. *P <0.05.

Figure 3 (abstract P19)
CD4-positive T-cell subsets (N, naïve; M, memory; and Reg, regulatory) in the mesenteric lymph and blood of all groups by flow cytometry.

Figure 4 (abstract P19)
CD8-positive T-cell subsets (N, naïve; M, memory; and Reg, regulatory) in the mesenteric lymph and blood of all groups by flow cytometry. Results The maximum APC-PCI values were 0.03 to 29 ng/ml, median 0.44 ng/ml. The overall mortality of the 53 sepsis patients was 32% (17/53). The mortality and relative mortality (mortality of activation group/overall mortality) of each activation group are presented in Table 1. A bell-shaped mortality relationship was noted, with high mortalities in both the nonactivated and highly activated groups. Notably, the lowest mortality was recorded in the moderately activated group. Subdividing activation groups by the Acute Physiology and Chronic Health Evaluation (APACHE) II score yielded the highest mortality (5/7 = 71%, relative mortality 227.7%) in the nonactivated subgroup with APACHE II score ≥25, whereas the APACHE II score did not influence mortality in the other activation groups. Minimum PC levels did not correlate with APC-PCI and showed no significant differences between the activation groups. Conclusion Nonactivation of PC in sepsis may represent the failure of an appropriate protective response and is therefore associated with increased mortality, especially when the APACHE II score is elevated. Septic patients without PC activation and a high APACHE II score may be those who are most likely to benefit from APC treatment. PC measurements were not predictive of PC activation as indicated by APC-PCI levels.

P22
Which is the worst factor in the sepsis aggravation: translocated bacterial amount or the gut-associated lymphoid tissue activation? Background The intestinal hypothesis of sepsis has been attributed to bacterial translocation (BT) and the aggravation of sepsis is related to the increased vascular permeability state that potentates the BT index. In the present study we examined the BT index during sepsis with or without mesenteric lymph exclusion. Methods Wistar rats (±200 g) were submitted to the BT process (Escherichia coli R6, 10 ml of 10 10 colony-forming units (CFU)/ml) and nonlethal sepsis (Escherichia cloacae 89, 2 ml of 10 7 CFU/ml) plus BT, with or without mesenteric lymph interruption by mesenteric lymph node resection and lymph duct ligature 5 days prior to the experiments. Samples (blood, spleen and liver) were collected 2 hours after the inoculation and were cultured to recover bacteria of intestinal origin. One-half of the animals/group was observed to the mortality index. The groups (n = 20/group) were: BT group (BT-G); BT with lymphadenectomy (BTL-G); combination (C-G); and combination with lymphadenectomy (CL-G).
Results BT was 100% positive in all groups. The BT index was similar between BT-G, BTL-G and CL-G (P = 0.6) and mortality was not observed in these groups, although a considerable amount of translocated bacteria could be recovered, particularly at the liver and spleen (Figure 1). When BT was added to the sepsis without lymph exclusion (C-G) the BT index was statistically lower (P = 0.04), but 50% (LD 50 ) of mortality occurred within 30 hours ( Figure 1).

Figure 1 (abstract P22)
Conclusion These results showed that, more than the amount of translocated bacteria, the gut-associated lymphoid system activation by the BT process played a pivotal role in the worsening of sepsis. Besides, BT occurred independently of mesenteric lymph interruption, showing that the hematological pathway of BT might be the principal route for bacterial dissemination into the bloodstream. In the present study, we demonstrate a role of inducible (iNOS) and neuronal (nNOS) nitric oxide synthase-derived excessive NO in MRSA-induced cardiovascular morbidity using potent and selective iNOS dimerization inhibitor BBS-2 and selective nNOS inhibitor 7-nitroindazole (7-NI).

Effects of inducible nitric oxide synthase and neuronal nitric oxide synthase inhibition on methicillin-resistant
Methods Ewes were operatively prepared and randomized after a 5-day to 7-day recovery period to the groups: sham (noninjured, nontreated, n = 6); control (injured, nontreated, n = 6); BBS-2 (injured, treated with BBS-2, n = 5); and 7-NI (injured, treated with 7-NI, n = 3). Injury consisted of insufflation of 48 breaths of cotton smoke followed by instillation of 2 to 5 x 10 11 colony-forming units of live MRSA into the airways of the sheep. BBS-2 (100 μg/kg/hour) and 7-NI (1 mg/kg/hour) was given starting from 1 hour to the end of the study (24 hours).
Results Hemodynamic variables were stable in sham animals. Control sheep developed a hyperdynamic circulatory state as evidenced by a significant increase in cardiac output and a severe fall in mean arterial pressure (Table 1). BBS-2 significantly attenuated hypotension 12 and 18 hours (P <0.05) post injury. In contrary, 7-NI reversed the fall in blood pressure at a later time point. Severe fluid retention seen in control animals was reduced by 7-NI, but not by BBS-2. 7-NI also improved oxygenation. Elevated levels of heart tissue 3-nitrotyrosine (marker of peroxynitrite formation) and poly (ADP)ribose (footprint of DNA damage) were attenuated by both iNOS and nNOS inhibition.
Conclusion iNOS inhibition had a partial effect in MRSA sepsisrelated cardiovascular morbidity. However, nNOS inhibition had a stronger effect on both severe hypotension and fluid retention following MRSA sepsis. Since nonspecific NOS inhibition is associated with unwanted side effects, more targeted inhibition of NO using specific NOS inhibitors may be a useful tool against the MRSA sepsis-related menace.

P24
Use of direct Etest in the management of ventilatorassociated pneumonia due to resistant Gram-negative pathogens Methods Chromogenic Mueller-Hinton agar was inoculated with 100 respiratory specimens from patients clinically suspected to have VAP. Vancomycin, cefoxitin, cefotaxime, ceftazidime, piperacillin-tazobactam and meropenem Etest strips were then applied to the inoculated medium strips selected to aid detection of resistant Gram-negative pathogens. In addition, a P. aeruginosa diatab was applied to plates to facilitate identification of this organism. The plates were incubated at 37°C in 5% CO 2 overnight and subsequently interpreted using a prospectively designed protocol to suggest antimicrobial choice. Specimens were processed using UK standard methods in parallel to allow comparison of speed and accuracy.
Results Forty-four out of 100 samples yielded no significant bacterial growth using the standard method. Sixty-three isolates were speciated from the remaining 56 samples (including 37 coliforms and 13 P. aeruginosa). Fifty-four of these samples had detectable growth at day 1 by the Etest method. Of the coliforms,  Background The selection of appropriate antibiotics to treat Gram-negative bacteraemia may be life-saving. Rapid methods of antimicrobial susceptibility testing have sought to guide early antibiotic selection and usage. We sought to evaluate whether a combination of chromogenic agar and six Etest gradient diffusion strips could be used to provide a clinically useful, direct rapid antimicrobial susceptibility test result following 4 hours of incubation. Methods Fifty consecutive Gram-negative blood culture isolates were tested over a 4-month period. Following confirmation of Gram-negative bacilli, 200 μl blood was directly inoculated from the enrichment broth onto a chromogenic MH agar plate. Six Etest strips were directly applied onto the agar using an automated placement device (Simplex C76; Inverness Medical UK (Bio-Stat Division), Stockport, UK). The antibiotics used were cefoxitin, cefotaxime and ceftazidime (to indicate the presence of extendedspectrum and ampC β-lactamase producers), vancomycin (to 'screen' for Gram-positive organisms), and piperacillin-tazobactam and meropenem (based on local prescribing patterns). The plates were incubated at 35 to 37ºC and read at 4, 6 and 24 hours. Minimum-inhibitory concentration values were determined and organisms were categorized as susceptible/resistant according to British Society for Antimicrobial Chemotherapy breakpoints. A presumptive identification and susceptibility profile was obtained at 4 hours, based upon which the investigators recorded a decision as to whether the patients' antibiotics could be escalated, deescalated or remain unchanged. The results were correlated with those at 6 and 24 hours, and with the report issued following routine susceptibility testing, as performed at our institution. Results Forty-five (90%) cultures had Etest susceptibility profiles interpretable at 4 hours. Of the five remaining, three (6%) were read at 6 hours and two (4%) at 24 hours. In three cases there was a mixed growth of organisms. Twelve (24%) organisms had resistance mechanisms identified, of which 10 (83%) were confirmed by our routine antimicrobial susceptibility rest method. At 4 hours, nine (18%) patients were receiving too narrow spectrum antibiotics. Additionally, the investigators felt that antibiotics could have been safely de-escalated in 16 (32%) cases and continued in the remaining 25 (50%) patients. Conclusion Our evaluation of this method has shown that it can provide rapid and reliable antibiotic susceptibility information at 4 hours. This could potentially have a major impact on antibiotic use and may significantly affect patient management. Background Statins have a well-known pleiotrophic effect as antiinflammatory agents, immunomodulators, antioxidants, antithrombotic agents and endothelial stabilizers. Statins, however, have been proposed as a therapeutic tool in septic patients. Objective To investigate possible statin therapeutic uses in the induced sepsis inflammatory process. Methods A prospective, longitudinal, experimental study was performed from November 2007 to March 2008 of 40 consecutive septic patients who were randomly assigned to one of the following groups: treatment group (received 80 mg daily simvastatin for 14 days) or control group (did not receive simvastatin). Inflammatory markers (sedimentation rate (SR), C-reactive protein (CRP), and antitrombin III) were measured on days 0, 5, 10 and 14. Results are expressed as the median (25th-75th interquartile interval) and groups were compared with the Mann-Whitney U test. Results The SR diminished from 34 (21 to 45) to 19 (14 to 23) in the treatment group, versus the control group where it increased from 28 (21 to 40) to 36 (27 to 50), with P <0.01 when both groups were compared. CRP behaved in a similar way, diminishing in the treatment group and increasing in the control group. On day 14, the SR and CRP reached normal values: 4 (2 to 6) and 1 (0 to 2), respectively, in the treatment group versus 22 (19 to 36) and 8 (4 to 14), respectively, in the control group (P <0.001). Antithrombin III increased in both groups, from 33 (28 to 50) to 90 (88 to 98) in the treatment group and from 33 (29 to 50) to 50 (48 to 55) in the control group (P <0.01). The length of stay was longer in the control group: 22 (18 to 26) days versus 15 (14 to 16) days in the treatment group (P <0.01). Conclusion The present study demonstrates that statins are able to decrease the systemic inflammatory response and provide endothelial increased stability properties from the fifth treatment day; reducing the mechanical ventilation time rates, and so on, with the patient's long stay. Background During the course of a bacterial infection, a rapid identification of causative agents is necessary for the determination of effective treatment options. We developed a method based on a modified broad-range PCR and an oligonucleotide microarray for the simultaneous detection and identification of 19 sepsis-causing pathogens at species level, as well as coagulase-negative staphylococci and Enterobacteriaceae at taxon level.
Available online http://ccforum.com/supplements/12/S5 Materials One hundred and fifty-four positive and 15 negative blood culture samples were collected to evaluate the assay performance. For the analysis, DNA was automatically extracted from the samples. Methods The broad-range PCR primer mixture was designed using conservative regions of topoisomerase gene subunits gyrB and ign from various bacteria. The primer design allowed the use of a novel DNA amplification method producing labeled, singlestranded DNA suitable for microarray hybridization. The probes on the microarray were designed against species-specific or taxonspecific variable regions of the gyrB and ign genes flanked by the primers. As a microarray platform for the probes, we applied TubeArray that is a microreaction vial containing a microarray at the bottom. To indicate the detection of antimicrobial resistance, we included mecA-specific primers and probes in the same assay. Furthermore, the software for automated data analysis was provided. Results Comparison of the assay results with the gold standard culturing method revealed sensitivity of 98% and specificity of 93%. When the mecA identification was correlated with Staphylococcus aureus, Staphylococcus epidermidis, and Staphylococcus genus detection by the specific probes, accurate information about the association of the mecA gene with staphylococci was provided. The results from these 15 samples were in line with the antimicrobial susceptibility testing (oxacillin susceptibility testing). Conclusion The results from the method were available 3 hours after the positive blood culture result. Up to 24 samples could be processed simultaneously. The assay therefore provides rapid and reliable data, which can guide antimicrobial treatment decisions in a timely manner. We have further broadened the pathogen panel for the detection of 50 bacterial species; 24 at the species level and at least 26 species at the taxon level. The described panel is now commercially available from Mobidiag Ltd (Helsinki, Finland) under the product name Prove-it™ Bacteria. Background Septic shock is associated with severe cardiac dysfunction, whose mechanisms remain partly undefined. Recent data suggested that it might be triggered by the direct action of microorganisms and their products on the heart itself. We previously showed that flagellin, the protein monomer from bacterial flagella, is a potent activator of NFκB-dependent proinflammatory signaling in cultured cardiomyocytes. In the present study, we investigated whether flagellin might induce such an inflammation in the heart in vivo and contribute to cardiac dysfunction. Methods Mice were injected intravenously with 1 μg flagellin. At selected timepoints (30 minutes to 4 hours), the effects of flagellin were evaluated by its ability to activate NFκB, mitogen-activated protein kinases and downstream signaling. Expression of the flagellin receptor TLR5 was also investigated. Cardiac function was evaluated after 4 hours using a microtip pressure-volume catheter inserted into the left ventricle. Also, human cardiac tissue was obtained from the right atrium in patients undergoing elective coronary artery bypass grafting surgery, to determine the presence of TLR5 in the human heart. Results Cultured cardiomyocytes, as well as hearts from mice and humans, expressed TLR5 protein at a high level. Flagellin activated NFκB and the mitogen-activated protein kinases p38 and JNK in cardiomyocytes in vitro and in vivo, and also upregulated the transcription of TNFα and MIP-2. In vivo, flagellin also induced the recruitment of neutrophils within the heart. Functionally, flagellin induced significant increases in end-systolic and end-diastolic left ventricle volumes, indicating cardiac dilation, and a significant reduction of end-systolic elastance and maximal elastance, indicating depressed myocardial contractility. In contrast, no change in the slope of the end-diastolic pressure-volume relationship was noted. Conclusion Bacterial flagellin induces a prototypical inflammatory response in cardiomyocytes in vitro and in the myocardium in vivo. These effects are associated with a profound alteration of the left ventricle systolic function in vivo, suggesting that flagellin may represent a critical mediator of cardiac dysfunction in septic shock. Background Impaired cardiac function due to reduced myocardial contractility is a typical manifestation of septic shock, whose mechanisms are poorly defined. Experimentally, the administration of endotoxin (lipopolysaccharide (LPS)) to laboratory animals is classically used to study the mechanisms of septic cardiomyopathy. However, most studies evaluating the effects of LPS on the heart in vivo have relied on indirect, load-dependent, indices of cardiac function, and thus could not precisely determine the real consequences of LPS on cardiac contractility. We therefore evaluated the direct effects of LPS on cardiac contractility in mice, using left ventricular (LV) micro-tip pressure-volume (PV) catheters, which provide load-independent measurements of cardiac function, including end-systolic elastance (Ees) and maximal elastance (Emax). Methods Male BALB/c mice received an intraperitoneal injection of Escherichia coli LPS (1, 5, 10, or 20 mg/kg). After 2, 6 or 20 hours, selected groups of mice were anesthetized, intubated and mechanically ventilated. A PV catheter was inserted into the left ventricle through the right carotid artery. LV pressure (end systolic (LVSP) and end-diastolic (LVDP)) and volumes (end systolic (ESV) and end-diastolic (EDV)) were recorded, allowing the calculation of the stroke volume, stroke work, cardiac output and ejection fraction. Ees and Emax were computed from the slope of the end-systolic PV relationships of successive PV loops obtained at rapidly reduced preload, by inferior vena cava compression. Mice were sacrificed at the end of the experiments. Results EDV decreased with LPS, mostly after 6 hours, whereas ESV did not change. LVSP was slightly decreased only after 6 hours, and LVDP was not significantly influenced by LPS. The stroke volume, stroke work, ejection fraction and dp/dt max were reduced at all doses of LPS, mostly after 6 hours and slightly recovered after 20 hours. In spite of an increase in heart rate, the cardiac output decreased, especially after 6 hours and at the high doses (10 and 20 mg/kg) of LPS. Most importantly, both Ees and Emax markedly increased after all doses of LPS, mostly after 2 and 6 hours, and returned back to control values after 20 hours. Conclusion In striking contrast with the usual belief, LPS does not induce direct negative inotropic effects in the mouse, but instead markedly enhances contractility. The alterations in cardiac function induced by LPS are only, and entirely, due to altered loading conditions, which are mainly observed 6 hours after the injection of LPS. Background The development of septic shock is related to the activation of nonspecific (innate) immune responses, triggered by the interactions between molecules released by pathogens and specific cellular receptors in the host, termed Toll-like receptors (TLRs). Flagellin is a 55 kDa protein isolated from the flagellum of Gram-negative bacteria, which may activate such responses through its recognition by TLR5. The tissue distribution of TLR5, as well as the actions of flagellin on various organs in vivo, has not been previously established. We therefore conducted the present study to determine the presence of TLR5 receptor in major organs from mice, and to evaluate whether flagellin could trigger prototypical innate immune responses through activation of NFκB and mitogen-activated protein kinase (MAPK) signaling pathways in these organs. Methods Mice were injected intravenously with 1 μg recombinant Salmonella flagellin. At selected timepoints (30 minutes to 6 hours), the mice were sacrificed and the major organs (lung, liver, gut and kidney) were harvested for expression of the flagellin receptor TLR5; for the activation state of NFκB (monitored by the degree of phosphorylation and degradation of its inhibitor IκBα and by the NFκB-DNA-binding activity); for the activation state of MAPK (monitored by the degree of phosphorylation of JNK, p38 and ERK); for the expression of inflammatory cytokines; and for the activation of apoptotic pathways (monitored by the degree of caspase-3 and poly(ADP-ribose) polymerase cleavage). Plasma was obtained for the measurements of cytokine levels. Results TLR5 protein was constitutively expressed in all organs. The injection of flagellin activated NFκB and MAPKs at 30 minutes, and markedly enhanced the generation of the cytokines TNFα, IL-1β, IL-6, TREM-1, and MIP-2 at 1 hour and 3 hours. Similarly, these cytokines significantly increased in the plasma from 1 hour to 6 hours after flagellin. Flagellin also triggered prototypical apoptotic changes in all organs. Conclusion Bacterial flagellin activates inflammatory signaling and apoptosis in most major organs in vivo, and thus may represent a critical mediator of multiple organ failure during Gram-negative septic shock. Background Acute kidney injury (AKI) occurs in about one-half of the patients who develop septic shock, and the mortality of AKI with sepsis is extremely high. An effective therapeutic intervention is urgently needed. In the present study we tested the ability of a novel tetrapeptide, EA-230, to improve survival and attenuate loss of kidney function in a clinically relevant model of sepsis -cecal ligation and puncture (CLP) in mice. Methods Sepsis was induced in C57BL/6 mice by CLP. Four hours postoperatively, EA-230 was administered. Subsequently, animals were treated twice daily for four consecutive days intraperitoneally. The effects of 20, 30, 40, or 50 mg/kg were compared with those of saline. Survival and renal function were monitored. Inulin clearance and para-aminohippuric acid clearance were used to measure the glomerular filtration rate and renal blood flow. Results All saline-treated control animals died within 5 days of CLP, whereas EA-230 treatment improved survival significantly in a dose-dependent manner. The best result was obtained with 50 mg/kg EA-230 (43.8% survival after 2 weeks). Serum creatinine and blood urea nitrogen increased markedly 24 hours after CLP. EA-230 attenuated the increases in creatinine and blood urea nitrogen significantly in the 30 to 50 mg/kg treatment groups. Furthermore, the glomerular filtration rate and renal blood flow were significantly higher (P <0.05) 36 hours post CLP in EA-230treated mice versus those treated with saline. Conclusion EA-230 is a novel and promising therapeutic agent for preventing AKI in sepsis. Its beneficial effect is associated with an improvement in renal hemodynamics. Nonbacterial infections or noninfectious conditions were diagnosed in 82 patients. The physiological variables recorded were: temperature, heart rate, blood pressure, respiratory rate (RR), oxygen saturation, urine output, and cerebral status. The laboratory variables were: C-reactive protein (CRP), lactate, bicarbonate, creatinine, urea, hemoglobin, white blood cells (WBC), neutrophils, platelets, International Normalized Ratio, D-dimer, albumin, bilirubin, procalcitonin (PCT), IL-6 and lipopolysaccharide binding protein (LBP). Results In a univariate analysis, PCT, IL-6, LBP, CRP, bilirubin and maximum RR during the first 4 hours (RR max 0 to 4 hours) were associated with bacteremia with P <0.001 and CRP, PCT, IL-6, LBP, WBC, neutrophils, RR max 0 to 4 hours and hemoglobin were associated with a bacterial infection with P <0.001. In a multivariate logistic regression, PCT, RR max 0 to 4 hours, bilirubin and CRP each contributed significantly to the accurate prediction of bacteremia. To predict a bacterial infection, CRP, WBC, hemoglobin and RR max 0 to 4 hours contributed significantly. The diagnostic accuracy of these variables was compared with the ability of the physicians caring for the patients to prescribe antibiotics appropriately. Of the 306 patients with bacterial infections requiring antibiotics, 76% had actually received antibiotics within 4 hours of arrival; and of the patients not requiring antibiotics, 54% were not on antibiotics after 4 hours. All variables tested had inferior diagnostic accuracy compared with the clinician. Conclusion We conclude that for the clinician, who evaluates patients with a suspected infection, special attention should be directed to the RR, CRP and WBC but the basic evaluation of the patient's medical history and a thorough clinical examination and assessment of the patient's general condition cannot be replaced by any laboratory parameter. Novel markers such as PCT, IL-6, and LBP seem not to give added value in the emergency room. Background Polymicrobial sepsis induced by cecal ligation and puncture (CLP) causes a massive nitric oxide (NO) synthesis and consequently several physiological alterations in cardiovascular and hormonal systems. Central NO is reported to modulate the secretion of vasopressin. Our aim was to study the central effect of an unspecific NO synthase inhibitor (L-NAME) on the mean arterial pressure (MAP), plasma nitrate levels (pNO), plasma arginine vasopressin concentration (pAVP) and hypothalamic arginine vasopressin mRNA content during polymicrobial sepsis induced by CLP. Methods Male Wistar rats received an intracerebroventricular injection of L-NAME (250 μg) or saline (vehicle) and 30 minutes later they were submitted to CLP or to a sham operation. Animals were decapitated 0, 4, 6, 20 or 24 hours after surgery and blood was collected for pNO and pAVP measurements. The brains were removed and the supraoptic and paraventricular nuclei were punched out for vasopressin mRNA determination by real-time PCR. In another set of animals the MAP was measured each 15 minutes 1 hour before and during 24 hours after surgery with intervals. Results CLP caused an increase in pNO after 6 hours, and in pAVP at 4 and 6 hours, while the MAP decreased during 5 hours after surgery. Hypothalamic vasopressin mRNA showed a tendency to decrease in both nuclei. L-NAME pretreatment increased survival (80% versus 67%), blocked pNO increase and MAP decrease and also resulted in an increase in plasma vasopressin concentration in the initial phase of sepsis (P <0.05). The vasopressin mRNA content increased at 20 and 24 hours in the paraventricular nucleus and only at 24 hours in the supraoptic nucleus. Conclusion These results demonstrate that central NO plays a role in blood pressure and in vasopressin synthesis and release during polymicrobial sepsis in rats. Acknowledgement Financial support from FAPESP. Background Recent studies revealed that vasopressinergic neurons have a high content of LTC4 synthase, a critical enzyme in cys-leukotriene synthesis that may play a role in regulating vasopressin secretion. The present study investigates the role of this enzyme in arginine vasopressin (AVP) release during experimentally induced sepsis. Methods Male Wistar rats received an intracerebroventricular injection of MK 886 (1.0 μg/kg), a leukotriene (LT) synthesis inhibitor, or vehicle, 1 hour before cecal ligation and puncture (CLP) or sham operation. In one group of animals the survival rate was monitored for 5 days. In another group, the animals were decapitated 0, 4, 6, 18 and 24 hours after CLP or sham operation, and blood was collected for hematocrit, serum sodium and nitrate, plasma osmolality, protein and arginine AVP determination. The neurohypophysis was removed for quantification of AVP content, and the hypothalamus was dissected for LTC4 synthase analysis by western blot. Results The mortality rate after CLP was reduced by the central administration of MK 886. The increase in plasma AVP levels and hypothalamus LTC4 synthase content in the initial phase of sepsis was blocked, whereas the decrease in neurohypophyseal AVP content was partially reversed. The increase of serum nitric oxide and hematocrit was reduced, and the decrease in plasma protein and osmolality was not affected by the LT blocker. In the final phase of sepsis, the plasma AVP level and the hypothalamic LTC4 synthase content were at basal levels. The central administration of MK 886 increased the hypothalamic LTC4 synthase content but did not alter the neurohypophysis AVP content and plasma AVP levels observed during this phase. Conclusion These results suggest that the central LTs are involved in the vasopressin release observed during sepsis.

Blocking central leukotrienes synthesis affects vasopressin release during sepsis
Methods A total of 404 adult patients admitted to the Department of Infectious Diseases from the emergency room (ER) for suspected severe infection was studied prospectively. Laboratory variables defining SIRS and severe sepsis were measured on arrival while physiological variables were recorded on arrival in the ER and every 4 hours for 24 hours. Results Bacterial infections requiring antibiotic treatment were diagnosed in 306 patients. One hundred and fifty of these developed severe sepsis during the first 24 hours. Significant bacteremia was detected in 68 patients. In these three groups 26%, 22% and 21%, respectively, failed to meet two or more of the SIRS criteria on arrival in the ER. In the group of patients that did not have an infection nor did not need antibiotic treatment, 63% had SIRS on arrival. SIRS on arrival correlated significantly with bacterial infection and development of severe sepsis, but not with bacteremia. Of the SIRS criteria, only the respiratory rate and white blood count contributed significantly to this finding; the heart rate and temperature did not. Intensive care was required for 14/150 patients (9%) with severe sepsis and for 6/68 (9%) bacteremic patients. Altogether 11/404 (2.7%) patients died within 28 days. Conclusion As a tool for definition of sepsis and selection of patients for clinical sepsis trials, SIRS lacks acceptable sensitivity and specificity in a selected ER population with a high risk of serious infection. Excluding patients with less than two or even three SIRS criteria may exclude a large cohort of patients with sepsis and result in biased enrollment to clinical trials. It may be time to abandon the SIRS criteria as an entry criterion for sepsis trials and to instead focus on more strict definitions of underlying infections in association with sepsis-related hypoperfusion and organ dysfunction. Many of the patients developed severe sepsis within 24 hours, yet only a small proportion required intensive care, putting the term severe sepsis into question. Severe sepsis in the ICU setting is known to be associated with a high mortality, whereas this might not be the case outside the ICU. Background Dendritic cells (DCs) play a key role in the initiation and integration of innate and adaptive immune responses to microbial infection. In contrast to neutrophils, macrophages or lymphocytes, there are virtually no data on the time course of circulating DCs in septic shock. Using a novel specific and sensitive assay, we analyzed the evolution of circulating myeloid DCs (mDCs) and plasmacytoid DCs (pDCs) in septic shock. Methods We enrolled immunocompetent adult patients with septic shock (SS, n = 43) and with shock from other etiologies (NSS, n = 29). Sixteen healthy controls (HC) were also used as reference for mDCs and pDCs. Blood samples (200 μl) were drawn on the day of shock, then after 3 and 7 days. DC analyses were performed using the DC-labelling kit Trucount ® assay (BD Biosciences, San Jose, CA, USA). CD11c + CD123cells (mDCs) and CD11c -CD123 + cells (pDCs) were selected and counted by flow cytometry (FACSCanto™; BD Biosciences). The HLA-DR mean fluorescence index was measured. Results The age, sex ratio, Simplified Acute Physiology Score II, Sepsis-related Organ Failure Assessment score, nosocomial infection (NI) and mortality rates did not statistically differ between SS and NSS patients. On day 1, mDC and pDC counts were significantly lower in both SS and NSS patients as compared with controls (P <0.001). Patients with SS had significantly lower mDC and pDC counts than NSS patients (P <0.001) both at day 1 and day 3. The HLA-DR mean fluorescence index of mDCs and pDCs was lower in SS patients compared with HC (P = 0.005 and P = 0.037, respectively) but did not differ between other groups. Interestingly, 10 out of the 43 SS patients developed NI after a median time of 9 (7.5 to 11) days in the ICU. Whereas mDCs increased in patients without NI, mDC counts remained low at day 7 in patients who developed NI: mDC counts and their relative variation between day 1 and day 7 were significantly lower in patients who developed NI than in those who did not (P <0.05). Logistic regression analysis indicates that a negative mDC relative variation is associated with an increased risk of nosocomial infection with an odds ratio of 22 (2.53 to 191) (P = 0.005). Conclusion Circulating mDC and pDC counts are lower in SS than in NSS as early as day 1. The persistence of a low count of mDCs after SS seems to be associated with the advent of nosocomial infection during the ICU stay. Background Biomarkers may aid in risk triaging of systemic inflammatory response syndrome (SIRS) patients at admittance to hospital and in the monitoring of response to medical intervention. The overall aim is reducing mortality in SIRS and sepsis patients. Methods A prospectively collected cohort of patients with SIRS that were admitted to an emergency department and a department of infectious diseases at a Copenhagen University hospital were studied. Samples obtained daily during hospitalization were measured for soluble urokinase plasminogen activator receptor (suPAR) using the CE/IVD-approved (the product complies with the European Directives for In-Vitro Diagnostics) suPARnostic ® assay and were compared with various other clinical parameters associated with assessing disease severity, including C-reactive protein, procalcitonin, Simplified Acute Physiology Score II, and Sepsisrelated Organ Failure Assessment scores. Survival was assessed using receiver operating curve statistics. Results One hundred and fifty-one patients were included in the study, nine of whom died within 30 days of admission. Admission levels of suPAR were significantly higher in survivors compared with nonsurvivors with an area under the curve of 0.80 and 0.92 when combined with age. Admission levels of procalcitonin and Creactive protein were not significantly different between survivors and nonsurvivors. Simplified Acute Physiology Score II and Sepsisrelated Organ Failure Assessment scores were significant predictors of death in this setting as well. During treatment, survivors showed overall declining suPAR levels ( Figure 1) while continuously elevated suPAR levels were observed in nonsurvivors (Figure 2).

Conclusion
The suPARnostic ® assay provided significant information on risk of mortality following admission. Continuous elevated suPAR levels during treatment were associated with poor clinical outcome. is now a common infection encountered in hospitals and communities. We have previously shown that MRSA causes reactive oxygen and nitrogen (ROS/RNS)-dependent increased vascular permeability, multiorgan system failure and death in our ovine model. Using type II alveolar epithelial cells (A549), we hypothesized that MRSA increases expression of vascular endothelial growth factor (VEGF), a regulator of vascular permeability to water and proteins, and disrupts barrier function by disrupting cytoskeletal integrity. Methods A549 cells were challenged with 10 5 colony-forming units MRSA over a time course of 24 hours and were visualized for markers of ROS/RNS formation (2,7-dichlorodihydrofluorescein), as well as VEGF and actin expression by confocal imaging and western blot analyses. Cellular permeability was measured by quantifying FITC-dextran flow through a monolayer of A549 cells. Results MRSA caused a significant 7.4-fold increase in 2,7-dichlorodihydrofluorescein fluorescence over unchallenged controls. L-NAME, an inhibitor of nitric oxide formation, blocked this response. Western blot analyses confirmed the confocal observations that MRSA caused an 8.15-fold increase in VEGF expression, versus cells that were pre-incubated with L-NAME (3.4-fold). MRSA also induced formation of actin stress fibers and subsequent cellular contraction. In support of these observations, MRSA caused a 405% increase in cellular permeability to FITC-dextran. However, pre-incubation with L-NAME had no effect on MRSA-induced barrier dysfunction. Conclusion MRSA induces VEGF expression in a ROS/RNSdependent manner. MRSA also causes alveolar epithelial cell barrier dysfunction by disrupting the actin cytoskeleton independent of nitric oxide synthase activity. Together, the data suggest that MRSA-increased vascular permeability in the lung may be due, in part, to disruption of the cytoskeletal integrity and increased expression of VEGF, but the overall mechanism involves multiple pathways and requires further study.  suPAR levels among patients who either had severe complications (n = 14) or died (n = 9) within 30 days of hospitalization. Dotted line, mean suPAR for the 23 patients. S19 contribute to septic mortality. Phosphoinositide-3 kinase gamma (PI3Kγ) plays a dominant role in the inflammatory response; however, its role in the pathogenesis of sepsis, specifically SIRS, lung/liver inflammation and damage, apoptosis, coagulation and mortality remains unknown. We hypothesized that mice lacking PI3Kγ or possessing a kinase-dead enzyme will be protected against septic-induced injury. Methods PI3Kγ wild-type (WT), knockout (KO) and kinase dead (KD) mice were randomized to cecal ligation and perforation (CLP)-induced sepsis or a sham laporatomy. After 18 hours, plasma and bronchoalveolar lavage and/or lung and liver tissue were collected. Plasma was assessed for inflammatory mediators and the lung/liver was analysed for pathology score, leukocyte infiltration, inflammatory mediators, edema, apoptosis, coagulation and downstream intracellular signalling of PI3Kγ. A separate cohort of WT and KO mice were used for evaluation of 7-day survival following CLP. Results Systemically, KO and KD mice showed a reduction in five of 22 measured cytokines/chemokines (MIP1a, MIP2, RANTES, MCP1 and IL-10) compared with WT controls. In the lungs, KO and KD mice were significantly protected against septic damage, as observed by decreased pathology scores, edema/permeability, leukocyte infiltration, inflammation (all 22 measured mediators), apoptosis and Akt/mitogen-activated protein kinase activation, compared with WT lungs. Similarly, livers of CLP-exposed KO and KD mice had decreased pathology scores, leukocyte infiltration, apoptosis and coagulation derangements compared with WT controls. Furthermore, Kaplan-Meier analysis of 7-day survival following CLP showed KO mice had significantly reduced mortality compared with WT mice. See Table 1. Conclusion The present study demonstrates that while PI3Kγ has a modest effect on SIRS during sepsis, its kinase activity is pivotal to the successive development of coagulation derangement and lung/liver inflammation and damage, probably through the modification of leukocyte recruitment and apoptosis. Furthermore, PI3Kγ is shown to effect CLP-septic-induced mortality, implying that it may be a possible therapeutic target in sepsis and multiple organ failure. Background Recently, a reciprocal relationship has been known between anti-inflammation and anticoagulation responses to infection. In our previous studies, protein C (PC) deficiency and antithrombin III (AT III) deficiency have been shown in septic patients. Moreover, these AT III deficiencies in sepsis did not relate to their disseminated intravascular coagulation (DIC) status. We hypothesize that PC activity relates to AT III activity in septic patients without DIC status. Materials Fifty ICU patients were included in this study and divided into three groups by primary diagnosis on admission; trauma patients, nonseptic patients, and septic patients. The patients who had already DIC on admission were excluded. Methods Serum PC activity (%) (Diagnostica Stago ® , Tokyo, Japan) and serum AT III activity (%) (Sysmex ® , Kobe, Japan) were measured on admission. PC and AT III activities were compared between three groups. Values are expressed as the median. Data were analyzed by the Kruskal-Wallis test and the Mann-Whitney U test. Pearson's correlation coefficient was used for correlation. P <0.05 was considered statistically significant. Results There were 23 trauma patients, 12 nonseptic patients and 15 septic patients. PC activity was significantly lower in septic patients than in trauma or in nonseptic patients (54.6 versus 85.6, Available online http://ccforum.com/supplements/12/S5 MIP-2 concentration (pg/ml in plasma x 10 4 ) 0 ± 0 0 ± 0 0 ± 0 13 ± 5** 2 ± 1* 2 ± 1* IL-6 concentration (pg/ml in plasma x 10 3 ) 0 ± 0 0 ± 0 0 ± 0 18 ± 5* 19 ± 10* 22 ± 8* Background Granulocytes or polymorphonuclear cells (PMN) represent the majority of leukocytes in peripheral blood. As terminally differentiated cells, they contain few ribosomes and assist innate immunity mainly through phagocytosis and degranulation. Whether or not they can release proinflammatory cytokines such as TNFα and IL-1β has been a controversial issue.
To clarify the role of PMN in this aspect, lipopolysaccharide (LPS)induced cytokine secretion from PMN was analyzed at the singlecell level with the ELISpot technique. Methods PMN and peripheral blood mononuclear cells (PBMC) from healthy human donors were prepared by gradient-based centrifugation to a purity >98%. ELISpot assays were used for detection of a large panel of inflammatory mediators. Cells were stimulated with endotoxin (LPS, 100 ng/ml) for 20 hours and the numbers of secreting cells were quantified with an ELISpot reader. For comparison, cytokine production was also analyzed by ELISA.
In some experiments, PBMC were depleted of monocytes using anti-CD14 magnetic beads. Results Purified PMN secreted IL-8 and MIP-1β and a subpopulation also released TNFα after LPS stimulation. In contrast and different from some earlier reports, we were unable to detect secretion of IL-1β, IL-12, granulocyte-macrophage colony-stimulating factor, IL-6 or IFNγ. Furthermore, granulocytes did not secrete the cytotoxic molecules perforin or granzyme B in response to LPS. Compared with the limited cytokine production by PMN, PBMC secreted significant amounts of all substances investigated and were found to require a 100x lower concentration of LPS than granulocytes to obtain the maximum number of responding cells. In addition, CD14 + monocytes were found to be the primary source of production. Discussion By use of the ELISpot method we could establish the cytokine profiles for both PBMC and PMN based on the frequency and pattern of cytokine secreting cells, rather than the amount of produced cytokine as by ELISA. This way, low levels of contaminating monocytes present in our PMN preparations could be discriminated from the granulocytes. Additionally, we could demonstrate that ELISpot, compared with ELISA, not only provides a more sensitive means of detection but potentially gives biologically more relevant information.
Conclusion LPS-stimulated PMN were shown to secrete IL-8, MIP-1β and TNFα but not IL-1β, IL-6, IL-10, IL-12, granulocyte-macrophage colony-stimulating factor, IFNγ, perforin or granzyme B. Our findings suggest that the ELISpot assay may be a suitable tool in further studies of cellular signaling. Methods The Acute Physiology and Chronic Health Evaluation II score, Serial Organ Failure Assessment (SOFA) at admission and serial 1 to 4 days after admission, demographic characteristics, comorbidity conditions with the Charlson score, Glasgow coma scale, organ dysfunction index, infection site, organism, and laboratory data at admission of 1,026 severe sepsis patients were analysed and evaluated to determine the association with 7-day mortality respectively. To develop a prognostic model, decision tree analysis was carried out with SAS 9.1. Results The 7-day mortality rate was 13.6/100 patients. Age was an independent risk factor, but the highest mortality (25.3%) was seen in the 60 to 69 years age group. The greater the number of organ dysfunctions, the higher the mortality. The underlying conditions were not statistically significant as a risk factor of 7-day mortality except liver diseases (P = 0.0015). The blood pressure, Charlson score, Acute Physiology and Chronic Health Evaluation II score and SOFA score at admission were all significantly associated with mortality. The initial laboratory values of hemoglobin, white blood cells, platelets, fibrinogen, prothrombin time, partial prothrombin time, arterial pH, potassium and albumin at admission were also statistically significant in bivariate analysis. Systemic infection and central nervous system infection showed 26.7% and 25.0% 7-day mortality. In a prognostic model by decision tree analysis, the blood coagulation factors (prothrombin time, platelet) and SOFA at 5 days after admission were the most significant prognostic factors of 7-day mortality. The sensitivity and specificity of this model were 67.5% and 96.8%, respectively.

P42
Conclusion The blood coagulation factors and SOFA were the most significant prognostic factors of 7-day mortality.

P43
Nucleic acid amplification-based pathogen detection in the blood of severe sepsis patients Background Upregulation and downregulation of inflammatory response was described in blood cells from septic patients, according to the stage of sepsis and the cells evaluated. This study aimed to evaluate the Toll-like receptor (TLR) signaling pathway gene expression in peripheral blood mononuclear cells (PBMC) and neutrophils in patients throughout the different stages of sepsis. Methods Septic patients admitted to two emergency rooms and two ICUs in one university and one teaching hospital were enrolled in the study, including five with sepsis, five with severe sepsis and five with septic shock. Five healthy volunteers were enrolled as controls. The Human-TLR Signaling Pathway, which comprises 84 genes related to TLR-mediated signal transduction, was evaluated by real-time PCR in PBMC and neutrophils obtained from patients and controls. Results were expressed as CT and were normalized with the housekeeping gene 18SrRNA (ΔCT). The fold change for each gene (2 -ΔΔCT ) was compared between the groups. Genes with fold changes greater than two and significant changes in ΔCT are reported as differently expressed.

Results
The fold change ratios in PBMC gene expression between septic patients and healthy controls revealed a dynamic process according to the stage of sepsis, tending towards downregulation of the TLR signaling pathway in PBMC in the more severe forms of the disease. In patients with sepsis and severe sepsis, fold-change analyses showed upregulated genes mostly in TLR receptors and adaptor or TLR interacting protein groups. The downregulated genes consisted mostly of downstream pathways and target genes, and they included the NFκB, JNK/p38 pathway, and effectors. However, the differential gene expression was restricted to five downregulated genes in septic shock patients, which are found in the effector and downstream pathways. Neutrophils showed a different pattern of adaptation. Patients with sepsis, severe sepsis and septic shock presented a broad gene upregulation, which included all functional groups evaluated and persisted throughout the stages of the disease. Conclusion TLR-signaling pathway genes are differently regulated in PBMC and neutrophils of septic patients, and are dynamically modulated throughout the different stages of sepsis.

P46
Downregulation of IL-6 and preserved reactive oxygen species production in human monocytes tolerized by lipopolysaccharide and challenged with Toll-like receptor agonists and whole Gram-negative and Gram-positive bacteria Background Tolerance to lipopolysaccharide (LPS) occurs when animals or cells exposed to LPS become hyporesponsive to a subsequent challenge of LPS. This mechanism is believed to be involved in the downregulation of cellular responses observed in patients with severe sepsis and septic shock. The aim of the present investigation was to evaluate the induction of tolerance in monocytes of healthy volunteers, in whole blood, after LPS exposition in vitro, assessed by intracellular cytokine detection and reactive oxygen species (ROS) generation.
Methods Peripheral blood cells were conditioned with small doses of LPS for 18 hours and challenged with different agonists of Tolllike receptors (macrophage-activating lipopeptide-2, flagellin and LPS) and whole Gram-negative (Pseudomonas aeruginosa) and Gram-positive (Staphylococcus aureus) killed bacteria. For detection of intracellular IL-6, samples of whole blood were stimulated for 6 hours. Monocytes were identified by forwardscatter and side-scatter parameters and CD14-positive staining. The samples were stained to verify the intracellular production of IL-6 on monocytes by flow cytometry. For induction of ROS, whole blood was stimulated for 30 minutes with LPS, P. aeruginosa and S. aureus. ROS production was measured by flow cytometry, using 2′,7′-dichlorfluorescein-diacetate detection.

Results
The conditioning with increasing doses of LPS resulted in lower intracellular detection of IL-6 in monocytes after the challenge with LPS. A similar effect was observed with macrophage-activating lipopeptide-2, P. aeruginosa, and S. aureus, but not with flagellin. LPS conditioning with 15 ng/ml LPS, on the other hand, resulted in preserved or increased production of ROS in monocytes after challenge with LPS, P. aeruginosa and S. aureus. Conclusion The phenomenon of tolerance involves a complex regulation, in which the production of proinflammatory cytokines, such as IL-6, is diminished, whereas the production of ROS is preserved or even increased. Background A positive blood culture (BC) is considered the goldstandard method for sepsis diagnosis, although its sensibility is low (10% to 30%) -which demands a better diagnostic tool to limit broad-spectrum antibiotic use in the majority of patients without culture-based sepsis diagnosis. Besides, after microbial invasion, bacteria can remain dead or fragmented in the circulation, thus limiting BC efficiency. Herein we evaluated the PCR diagnostic efficacy under live, dead and fragmented bacteria contents in the bloodstream.
Methods Wistar rats were distributed into three groups (n = 20/ group) based on live, dead and DNA inoculations. Another lipopolysaccharide (LPS) + DNA group (1 mg/kg LPS injection plus 4 hours later DNA injection, n = 10) was designed for DNA detection under an induced inflammatory state. Live, dead and extracted DNA forms of Pseudomonas aeruginosa (ATCC 27853), 10 7 colony-forming units/ml/100 g body weight, were injected into the tail vein of respective groups. Blood samples were collected after 20 minutes (n = 10) and 6 hours (n = 10) from all groups except for the LPS + DNA group (6 hours), and were submitted to a nested PCR assay using general and specific primers. BC was performed only in the live group.

Results
In the live group at 20 minutes the sensibility was 100% by both BC and PCR, and at 6 hours the sensitivity was 60% to BC and 80% to PCR. In the dead group, the PCR sensitivity was 90% at 20 minutes and 50% at 6 hours. In the DNA group, the sensitivity was 90% at 20 minutes and 40% at 6 hours, and in the LPS + DNA group at 6 hours the sensitivity was 40%. Conclusion The sensitivity of the PCR was as effective as BC in 20 minutes and superior in 6 hours. Besides, the PCR assay was able to detect circulating dead bacteria and bacterial DNA in the blood, which is not possible by the BC method. These findings suggest that the live bacteria remains for a short period of time in the bloodstream as compared with dead and DNA bacteria, and a systemic inflammation state seems to not interfere with the PCR assay. Besides, the PCR tool with specific primers can be a useful method for sepsis diagnosis in the negative blood culture conditions as well as in specific bacterial surge events in the ICU, thus improving the antibiotic usage potentials.

P48
A novel multiplex quantitative polymerase chain reaction assay for the early prediction of sepsis in critically ill patients presenting signs of shock and organ dysfunction Background Clinical signs of sepsis often overlap with other symptoms present in patients after trauma, surgery and chemotherapy, amongst others. The clinical utility of existing biomarkers to detect sepsis, such as procalcitonin (PCT) and C-reactive protein (CRP), is limited in patients suffering from shock and organ dysfunction. As a result, tests with superior positive and negative predictive values are mandatory in life-threatening infection. Methods We have analyzed the clinical utility of a new class of transcriptomic biomarkers derived from circulating leukocytes. Prospectively collected whole blood samples from 460 patients admitted to the operative ICU of the University Hospital Jena were used in a microarray/quantitative PCR study to identify sensitive and specific biomarkers. Microarrays comprising 5,308 probes corresponding to 3,704 human genes relevant to inflammation, immune response and related processes were used for analysis. The identification of a signature specific for the discrimination between systemic inflammatory response syndrome and sepsis in patients suffering from shock and organ dysfunction was performed in independent training and test phases. The training set of 96 patients was selected by an independent ICU committee. An algorithm was established combining and transforming the geneexpression signals into a continuous, nondimensional score indicating either infectious or noninfectious causes for organ dysfunction. The resulting classificator was validated in a test set comprising 1,784 ICU days of 364 patients. For each marker, a robust quantitative PCR assay was established.

Results
The final microarray signature could be transferred into a multiplex quantitative PCR format retaining full sensitivity and specificity with a time to result of approximately 5 hours. Moreover, it could be demonstrated that the combination of seven biomarkers possesses the same accuracy compared with the complete biomarker set. The area under the curve in the test group was determined as 0.79 (PCT -0.65, CRP -0.67). Moreover, the quantitative PCR assay determined the onset of sepsis up to 48 hours prior to the clinical diagnosis backed by daily CRP and PCT testing. Conclusion With its high predictive value for the differentiation between infectious and noninfectious causes of shock and organ dysfunction, this new class of biomarkers may help to identify patients with life-threatening infections among patients at risk and to guide therapy (for example, with anti-infective agents).

P49
A comparative study between conventional and antimicrobialfilled central venous catheters Patients who require CVCs for long periods have benefited with the use of impregnated CVCs, because they present long-term use, lower rates of infection, and avoidance of successive punctures and risks of the procedure. In view of the clinical benefits already mentioned, the benefit reached by the use of antiseptic-impregnated catheters compensated the initial extra expensive cost of 40%.