Sepsis 2018

Robert A. Balk, MD, went to the University of Missouri at Kansas City 6 year Medical School and received his BA in Biology in 1976 and the MD degree in 1978. He completed his internship and residency in internal medicine at the University of Missouri – Kansas City and went to the University of Arkansas For Medical Sciences for Fellowship training in Pulmonary and Critical Care Medicine. After spending one and one-half years on the teaching faculty of the University of Arkansas for Medical Sciences, Dr. Balk moved to Chicago to join the Section of Pulmonary and Critical Care Medicine at Rush University Medical Center and Rush Medical College. Currently, Dr. Balk is the J. Bailey Carter, MD Professor of Medicine at Rush Medical College and the Director of the Division of Pulmonary and Critical Care Medicine at Rush University Medical Center in Chicago, IL. Dr. Balk is the Associate Chief Medical Officer for Critical Care and he is also the Director of Adult Cystic Fibrosis Program. He also serves as the medical director of the Pulmonary Rehabilitation Program at Rush Oak Park Hospital.

Background Nangibotide (LR12) peptide is a specific TREM-1 inhibitor. In preclinical septic shock models, nangibotide was able to restore appropriate inflammatory response, vascular function, and improved survival. In phase I, nangibotide was found to be safe and well tolerated up to the highest dose (6 mg/kg/h). Materials and methods International, multi-center phase IIa randomized, double-blind, twostage, placebo-controlled study (NCT03158948). Main inclusion criteria were septic shock according to Sepsis 3 definition and nangibotide to be initiated within 24 hours of shock onset. Patients were randomized to receive either placebo, 0.3, 1 or 3 mg/kg/h of nangibotide. Stage-1 investigated ascending doses. In stage-2 patients were randomized to complete 12 patients in each group. Study drug was infused until end of vasopressors + 12h or up to 5 days. Safety data were reviewed by an independent Data Safety Monitoring Board (DSMB). Primary endpoint was safety and tolerability. Patient follow-up period was 90 days. Results 50 patients were randomized and 49 treated (1 patient died before dosing). All groups were well balanced in terms of baseline characteristics, except for APACHE II score which tend to be nonsignificantly lower in placebo group. Primary infection source was 40% abdominal, 50% pulmonary and 10% urinary. Nangibotide was safe and well tolerated in all groups. During the trial, the DSMB did not raise any safety concern. The number of SAEs/AEs and the number of patients with SAEs/AEs was comparable between all groups (Table 1). Most frequent AEs were atrial fibrillation, anemia, pleural effusion and thrombocytopenia. Ventilator and vasopressors free days alive were similar in all groups ( Table 2). All-cause mortality at day-28 was 14% (5/37) in pooled nangibotide groups and 25% (3/12) in placebo group. In the subgroup with sTREM-1 levels above median, the day-5 mortality was calculated as 40% (2/5) and 20% (4/20) in placebo and nangibotide groups respectively. A trend toward a decrease in circulating levels of endothelium injury markers was observed in nangibotide-treated patients. Conclusion Nangibotide was shown to be safe and well tolerated in septic shock patients. Although this small exploratory study was not powered to conclude on efficacy, a signal with non significant lower mortality was observed in the nangibotide group. These results support the need of a larger study to demonstrate the role of nangibotide in the treatment of septic shock.
Background Serum lactate is an important marker of tissue perfusion. Impaired hepatic function especially among cirrhosis patients could decrease lactate elimination and result in higher lactate level during septic shock. Whether serum lactate level could be used to predict outcome of cirrhosis patients with septic shock had not been identified. This study was thus aimed to evaluate the utility of serum lactate to predict mortality, comparing between cirrhosis and non-cirrhosis septic shock patients.

Background
Thailand is moving toward aged society. It is interesting whether octogenarian, elderly aged 80 years old or older, with septic shock has different clinical characteristics and outcomes as compared with younger patients. We report here the information from our patient cohort. Materials and methods This is a retrospective study, included adult septic shock patients who were admitted at Siriraj hospital between April 2011 and December 2017. Patients' baseline information, severity score, treatment modality and outcomes were retrieved. Octogenarian was defined by the age of 80 years or older at the day of admission. All patients received septic shock resuscitation, following standard guideline. 1 The primary outcome was 30 days mortality. Results A total of 782 septic shock patients were enrolled. Of these 133 were 80-year-old or older. When compared with younger patients, octogenarian had lower APACHE II score and higher comorbidities which included: hypertension, coronary artery disease, previously stroke, cirrhosis, and chronic kidney disease. Pneumonia was the leading site of infection in both groups but occurred in higher proportion in octogenarian (46.6% vs 29.7%, P 80 years old as the independent predictor of death within 30 days, as well as other parameters as shown in Table 1.
Conclusions Octogenarian with septic shock had lower disease severity but higher comorbidities. This group had higher mortality, both at 30 day and at discharge.

Background
Efficacy of low dose hydrocortisone administrated during septic shock resuscitation had been long debated. Despite recent two large randomized controlled trials, steroids benefit is still inconclusive [1,2]. The aim of this study is to evaluate the effect of hydrocortisone on resuscitation outcomes, comparing between hydrocortisone and placebo. Materials and methods This is a retrospective study, included participants who were enrolled in the previously reported randomized controlled trial of hydrocortisone treatment in early sepsis-associated acute respiratory distress syndrome [3]. The patients who met criteria of septic shock, according to Sepsis-3 definition [4] were included in sub-group analysis. The treatment group received hydrocortisone 50 mg intravenously every 6 hours for 7 days, while the control group received normal saline in identical volume. The primary outcome was rate of survival without organ support at 28 days after septic shock diagnosis.

Results
Of 197 patients enrolled in the previous study, 154 patients met septic shock criteria. Seventy-eight patients were randomized to receive hydrocortisone and 76 patients were in the placebo group. The patients' age, gender, underlying conditions, source of infection, baseline arterial pressure and severity score were not different. Mean APACHE II score was 22.3+5.8, mean SOFA score was 11.6+3.1 and mean initial lactate was 4.0+2.4 mmol/L. The rate of survival with no organ support at 28 days was non-significantly higher in hydrocortisone group (61.5% vs 48.7%, relative risk 1.28 [95%CI 0.93-1.76]; P=0.12). The secondary outcomes, as shown in Table 1, were not different; except vasopressor dependent day which was shorter in hydrocortisone group (4.8+3.0 vs. 6.8+5.7 days; P=0.008). As for adverse events, hyperglycemia occurred more in hydrocortisone group (83.8% vs. 68.9%; P=0.03) while the rate of re-infection, cardiac arrhythmia and gastrointestinal bleeding was the same. Conclusions Hydrocortisone administration during septic shock resuscitation associated with shorter vasopressor dependent day without any benefit in other outcomes.

Conclusions
We developed a simplified THAI-ICU score which outperformed the standard severity score and can be considered as a powerful tool to predict hospital mortality. The simplicity of THAI-ICU score will increase the possibility to feasibly apply in resource limited setting.

Background
In 2004 a novel suicidal antimicrobial strategy of human neutrophils was described, named NETosis (NET -Neutrophil Extracellular Trap) [1]. NETosis is the particular case of ETosis (ET -Extracellular Trap), which is specific for a variety of human phagocytes. During ETosis phagocytes are transformed, resulting in subcellular structures formation, such as so-called extracellular DNA-traps (DTs), consisted of nuclear DNA with immobilized antimicrobial proteins and enzymes. DTs efficiently bind and kill microorganisms, however they are toxic to host cells. DTs play important role in thromboses, inflammatory and autoimmune diseases [2]. DTs are known to be important in the pathogenesis of sepsis [3]. At the same time standard methods for determination of ETosis-transformed phagocytes in human blood are not developed yet [4]. New method [5] was proposed for ETosistransformed phagocytes determination in blood of patients with sepsis, based on DTs calculation in standardized thin blood smears. However, additional studies are needed to confirm its clinical usefulness. The aim of this study is to evaluate the connection between increased DTs level (in blood smears from patients with sepsis) with outcome of the disease. Materials and methods DTs levels were determined in blood smears of 51 patients in intensive care unit with verified sepsis. 55 healthy donors were in control group. Peripheral venous blood was collected into tubes containing EDTA, standardized thin blood smears were prepared using 2 μl of blood, and stained with Giemsa stain. Levels of DTs were calculated using automated microscopy system MECOS-C2 (Moscow, Russia). Blood of the patients was studied several times during staying in intensive care unit; the maximal DTs levels during all time of observation were analyzed.

Results
Among 51 patients with sepsis 20 died and 31 recovered. The most informative indicator was the maximal DTs level during all time of observation. In group of all patients this indicator was in average    The differences between the groups were statistically significant according to Mann-Whitney U-test: donors vs septic patients -p=0.000026*; recovered vs died patients -p=0.01*** (Fig. 1). Conclusions In average, maximal DTs level was significantly higher in patients with sepsis than in healthy donors. Besides, maximal DTs level was significantly higher in died patients than in recovered patients. These results allow to conclude, that high level of DTs in patients with sepsis correlates with lethal outcome.

P12
Monitoring the activity of antithrombin in sepsis allows predicting the likelihood of a clinical outcome of the disease Ivan V Redkin, Andrey F Lopatin, Vera V Samoylenko, Yury V Skripkin, Valery V Likhvantsev Moscow Regional Scientific Research Clinical Institute, Russia, Moscow Materials and methods A retrospective, observational, cohort study of the dynamics of inflammatory response and hemostasis in patients with sepsis was performed. 83 patients were included in the study. The diagnosis of sepsis is based on the presence of a foci of infection and SOFA more than 4, confirmed by isolates of pathogens from the blood. On the outcome of the disease, two groups of patients were formed: 1 -survivors (n = 41), 2 -dead (n-= 42). Laboratory data of patients from both groups were retrospectively studied: 1) at the time of diagnosis of sepsis and 2) the 5th day of the disease. The tests included the determination of blood leukocytes, young forms of neutrophils, lymphocytes, total plasma protein and albumin, plasma creatinine, C-reactive protein, procalcitonin (PCT), platelets, fibrinogen concentration, activity of antithrombin, D-dimer level. In dead patients, pathohistological studies were performed to identify microthrombosis of the vascular bed.

Results
At the stage of sepsis diagnosis in the groups of statistically significant differences by the studied parameters was not revealed. An exception was albumin and PCT, the values of which could be taken to predict the likelihood of an adverse outcome. Albumin (Odds ratio -2.6 [1.0595 -6.5038], P = 0.0371); PCT (Odds ratio -9.2500 [2.7763 -30.8184], P = 0.0003). However, the ROC analysis data do not allow us to substantiate their recommendation for the prognosis of the outcome of the disease. Albumin Cut-off ≤26 g/l, sensitivity/ specificity -70%/53%; PCT -Cut-off > 16.2 ng/ml: sensitivity/specificity -48%/90. After 5 days (stage 2 of the study), the level of activity of antithrombin was significant for the prognosis of the outcome of sepsis (Odds ratio -26,4 [8,02 -8,6,86]); P <0.0001. ROC analysis of the level of activity of antithrombin as a predictor of lethal outcome revealed a high sensitivity (79%) and specificity (88%) for the value (Cut-off) ≤61%. Data of pathohistological studies of deceased patients revealed microcirculatory thrombosis of the vascular bed in 73.6%.

Conclusions
The obtained results show that at the stage of development of sepsis there is a high heterogeneity of indices of inflammatory reaction and hemostasis. The revealed regularity of the dynamics activity of antithrombin in patients with sepsis indicates one of the leading pathophysiological links in the multiple organ failure and allows us to recommend an antithrombin activity test for predicting the dynamics and probability of the outcome of the disease.

P13
Open label prospective randomised control study of high cut point level of procalcitonin guided antibiotic therapeutic protocol in surgical critically ill patients  Background Sepsis is a generalized infectious disease, has no propensity to selfdestruct [1]. At present, the idea of the importance of bacteremia in sepsis is not systematized. There are data based on the frequency of bacterial isolation from the blood of patients with sepsis, which, depending on the severity of the patients, varies from 20 to 80% [2,3]. Materials and methods Diagnosis of 96 cases of sepsis with positive hemoculture in patients was carried out. The study included 58 men and 38 women aged 18 to 78 years. All patients were divided into groups depending on the underlying disease. Blood culture from all patients was carried out on nutrient media. The sensitivity of the isolated pathogens to antibacterial drugs was determined by diffusion into agar using discs.

Results
In just 3 years of observation, 155 cases of positive bacteremia, accompanied by an inflammatory reaction syndrome, were detected. In total, 89 strains of Gram-positive pathogens were isolated in monocultures, 27 Gram-negative pathogens in the culture, 19 mixed cultures of Gram-positive and Gram-negative microorganisms, 8 mixed cultures of Gram-positive microorganisms, 5 mixed cultures of Gramnegative microorganisms and 7 mixed cultures of bacteria and fungi. Gram-negative bacteria predominantly prevailed, constituted about 70% of all isolated strains, Gram-negative bacteria -about 29%, the frequency of fungal secretion was low. Among the representatives of Gram-positive bacteria, staphylococci were most often isolated, with S.aureus and coagulase-negative staphylococci standing out at approximately the same frequency. Among the representatives of Gram-negative bacteria, enterobacteria were most often isolated (17.7%), non-fermenting bacteria were isolated twice less often (8%). From the point of view the choice of empirical regime of antibacterial therapy (ABT), fundamental importance was causative agent of sepsis belongs to hospital or non-hospital microflora. Effective ABT allows to reliably decrease the level of mortality in patients with sepsis. The effectiveness of sepsis treatment in different clinical groups of patients varies from 59-82%, according to different authors. On average, the efficiency of adequate ABT, according to our data, was 40%. The lethality of patients receiving adequate ABT, according to our data, is significantly lower than those receiving inadequate ABT (p <0.01). Timely correction of ABT, based on the sensitivity of the isolated pathogen, also significantly reduces lethality (p <0.01).

Conclusions
The initial regimens of ABT patients with sepsis, severe sepsis and septic shock do not correlate with the degree of severity of the systemic inflammatory reaction syndrome.

Background
In present the problem of treatment purulent-inflammatory diseases and is currently relevant to clinical surgery [1]. Despite significant advances, associated with the expansion deepening of knowledge about etiology, pathogenesis, clinical manifestations of surgical infection based on the modern advances in immunology, microbiology, biochemistry, reduction in number of patients and the severity of purulent surgical diseases is not observed [2]. Proper use of vacuum system can significantly reduce the wound regeneration and the cost of treatment for this category of patients. Vacuum therapy improves the course of all stages of the wound process [3,4].

Materials and methods
The results of treatment of 54 patients with severe surgical soft tissue infection at the age from 18 to 75 years who were treated at the department of purulent surgery were presented. In the treatment of 25 patients, system that created negative pressure was used in the wound in a constant or alternate regimen. The use of vacuum system and therapy with negative pressure was carried out after radical necrectomy. The control group consisted of 29 patients, in which the treatment was not included vacuum therapy. Keeping wounds in the control series was carried out in a semi-open manner using frequent dressings, local application of traditional antiseptics and ointment antibacterial compositions.

Results
In basic group in 19 patients from 25 in wounds identified a number of pathogens and opportunistic microorganisms. The analysis of the results of cytological control in the comparison groups allowed to establish that in the first 3-6 days after the operation, in all types of post-operative wound management, the necrotic type of the cytogram was dominated. The decrease in the level of bacterial insemination of the wound tissues below the critical (103 CFU/g) during NPWT therapy was achieved on average by the 3rd day against 7 days in traditional methods of local wound healing (P <0.05). In all patients, where was used vacuum system, pronounced positive effect of treatment was mostly noted, what optimized the timing of cleansing and healing of wounds. In the control group until the 15th day of observation, complete closure of the defect was noted only in 3 patients (16.7%).

Conclusions
The usage of wound healing technology by negative pressure promotes mechanical elimination from purulent foci of a large number of microbial bodies and products of tissue decay, which slows healing of the wounds, reduces interstitial edema of tissues, improves their lymph and blood circulation.

Background
The excessive release of pro-inflammatory and anti-inflammatory cytokines enables the latter to act as mediators for hemodynamic alterations, metabolic acidosis, and multi-organ failure in severe sepsis.
Recently, the oXiris® hemofilter that comprises an AN69 core membrane, polyethyleneimine, and which is grafted with heparin, has been introduced as a novel hemofilter membrane to mitigate inflammatory response during sepsis-associated acute kidney injury (AKI) that requires renal replacement therapy (RRT).

Materials and methods
In the present case series, we retrospectively evaluated the impact of the oXiris hemofilter on hemodynamics and clinically relevant outcome parameters in critically ill patients with septic shock who require continuous RRT and those with at least two organ dysfunctions.

Results
Thirty-five patients were enrolled. There was a nonsignificant trend for improvement of hemodynamic status as shown by increased mean arterial pressure (MAP), decreased norepinephrine dose, inotropic score, and vasopressor dependency index. Blood lactate and base excess also showed significant improvement. oXiris treatment was safe with no device-related adverse events. Conclusions Using oXiris hemofilter was feasible, well-tolerated, and should encourage the conduction of randomized controlled trials to evaluate the potential benefits of this therapeutic option.
Background Sepsis with unidentified pathogen is common and makes targeted therapy challenging. It is therefore likely that these patients receive less effective therapy than other patients with sepsis. Materials and methods All patients with suspected sepsis, taken care of by the medical emergency team in the Emergency department of a Norwegian secondary and tertiary care hospital, were prospectively included in this study. The patients were re-evaluated for sepsis after the hospitalization, sepsis diagnosed if the patient fulfilled both of the following criteria: I) Probable infection II) New or deteriorated organ dysfunction corresponding to a Sequential Organ Failure Assessment (SOFA) Score ≥2 in organ systems not primarily infected. Respiratory distress in relation to respiratory tract infection was for instance not included in the SOFA-calculation. Results A total of 487 patients were included from May 2017 to May 2018. Three patients were excluded due to incomplete registrations. 201 (42%) were assessed to have had true sepsis. A pathogen probably responsible for the septic episode was identified in 146 (73%). The overall 30 days mortality for sepsis was 17 %; 12% in patients with an identified pathogen opposed to 31 % for patients without an identified pathogen (p=0.001, see Table 1 below). High rates of lower respiratory tract infection were found among sepsis patients without identified pathogen, compared to patients with identified microbe (56% vs 29%, p<0.001). Moreover, septic patients with focus in the lower respiratory tract without identified pathogen had a significantly higher mortality rate (52% vs 21%, respectively; p=0.007). Conclusion Unidentified pathogen is a common feature in patients with lower respiratory tract infections and in our cohort also associated with very high mortality. Efforts should be made to improve early identification of pathogens in such patients. Leptospirosis is the lethal tropical infectious disease, which shown significant mortality in severe cases. Hemodynamic instability, acute kidney injury, acute liver injury and acute respiratory failure were among the common presentation required ICU admission. Female, acidemia, lung injury and coma were related to ICU mortality in this cohort. Background Surviving Sepsis Campaign guidelines recommend resuscitation with at least 30ml/kg intravenous (IV) fluid in patients with sepsis and hypoperfusion [1]. Trials in developing countries have found higher mortality with larger resuscitation fluid volumes [2,3]. A clinical trial of a smaller volume of fluid with earlier commencement of vasopressors is warranted, but it is unknown if this intervention is feasible and clinically acceptable in industrialised countries. Materials and methods The REstricted Fluid REsuscitation in Sepsis-associated Hypotension (REFRESH) trial (ACTRN126160000006448), a prospective, randomised, open-label, clinical feasibility trial was conducted in the emergency department (ED) of eight Australian hospitals between October 2016 and March 2018. The protocol has previously been published [4]. Inclusion criteria were suspected infection and hypotension (systolic blood pressure <100mmHg despite at least 1000ml IV fluid). Details of the trial intervention are summarised in Fig. 1. The primary outcome was total fluid administered within the first six hours from presentation. A range of process-of-care, feasibility and clinical outcomes were assessed including vasopressor requirement, protocol adherence, organ failure, and mortality at 90 days post randomisation. Results There were 99 participants (49 standard and 50 restricted volume) in the intention-to-treat analysis (Fig. 2). Baseline participant characteristics are shown in Table 1. Fluid volume (total for 24 hours from randomisation) and vasopressor use are summarised in Table 2. A significantly lower volume was administered over the first six hours in the restricted volume group (median 30ml/kg vs 43ml/kg, p<0.001) along with earlier commencement of vasopressors (median 1h vs 2h post randomisation, p=0.001). There was no significant difference in the rate of ICU admission, ventilation and acute kidney injury (including dialysis) between the groups. At 90 day follow up 3 of 47 (6%) in the standard arm and 4 of 48 (8%) in the restricted arm had died (P=1.0). Conclusions A restricted fluid and early vasopressor regimen among ED patients with suspected sepsis and hypotension led to a significant reduction in fluid volume given over the first 6 hours and at 24 hours. Median dose and duration of vasopressors were lower in the restricted volume group, but this was not statistically significant. There were no differences in clinical outcomes, however the study was not designed to detect these. The results demonstrate the feasibility and clinical acceptability of the trial intervention in the Australian ED setting. A large randomised trial to assess the effect on clinical outcomes should be pursued.    Blood PCT was > 2 ng/ml, EAA did not exceed 0.5. In all cases, hemosorption using the CytoSorb column was included in the treatment package. Indications for its use was IL-6 blood concentration of more than 800 pg/ml, as an indicator of the severity of hypercytokinemia, which, as previously identified, correlated with a high risk of of generalized inflammation progression and MOF. Hemosorption was performed by the Multifiltrate machine (Fresenius Medical Care, Germany) for a period of 12 -24 hours. Vascular access was v.subclavia/v.femoralis. Perfusion rate was 150 ml/min. Heparinization was performed using unfractionated heparin. PCT, CRP, IL-6, IL-10 in blood serum were studied. Measurements were taken before hemosorption, 1 and 6 hours after its completion. The statistical processing of the results was performed using "Statistica 6.0" program with the calculation of the Student's reliability criterion (t). The difference was considered significant for p<0.05.

Results
After hemosorption using the CytoSorb column, in 8 out of 11 cases the regress of generalized inflammation activity was noted. PCT reduction was recorded from 27.3 ± 6.7 ng/ml to 4.3 ± 1.1 ng/ml; CRP -from 467.2 ± 98.6 mg/ml to 105.1 ± 54.3 mg/ml; IL-6 -from 1461.3 ± 102.3 pg/ml to 346.5 ± 91.1 pg/ml; IL-10 -from 46,1 ± 4,3 pg/ml to 2,1 ± 0,4 pg/ml; blood lactatefrom 5.9 ± 1.7 mg/ml to 1.9 ± 1.2 mg/ml (p<0.05). The SOFA index in these patients decreased by 4.1 ± 1.2 points (p<0.05). In 3 patients inflammatory reaction activity and organ disorders progression was diagnosed, which required the use of other extracorporeal detoxification methods. Conclusions Extracorporeal removal of excessively produced inflammatory mediators is an effective and safe auxiliary "tool" in the treatment of sepsis in cancer patients. However, in order to determine the criteria for indications, timing and tactics of combined use with other methods of extracorporeal detoxification, further in-depth studies are needed.

P23
Simple Clinical Score: Predicts mortality in sepsis to early warning score of the emergency room The second objective is to compare variables of SCS predictive abilities of different parameters lead to mortality. Firstly, we separate our dataset to trained and test data. The trained data consists of 75% of the dataset (167 patient records) and the test data consist of 25% of the dataset (58 patient records). We create a logistic regression model from trained data using the patient's status 'dead' as outcome, with all the variables as independent predictors. From the created model, Table 3 show the important variables for predicting deadly outcome. Oxygen saturation, prior to current illness, spent some part of daytime in bed, coma without intoxication or overdose, respiratory rate (per min) and abnormal ECG are the most important variable for predicting deadly outcome, respectively. We measure the performance of our model using the sensitivity and specificity analysis that are shown sensitivity of 86.7% and specificity of 92.9% for SCS predicts mortality. We test our model by prediction and comparing with the train data. The result is plotted as the Receiver Operator Characteristic (ROC) curve in Fig. 2. We also calculate the Area Under the Curve (AUC) as AUC = 0.6954657 meaning that the prediction accuracy is quite poor which could be due to the limitation of our dataset. Conclusions SCS are not sufficiently accurate to predict mortality in patients with sepsis, which is quite poor quality. However, the outcomes of this analysis known the important variable identification tool for deteriorating patients to early warning sign for detection and prevent adverse events in the patient with sepsis.

Results
There are totally 17 patients that admitted in ICU were consulted through HoloLens in the period of this study. The topics of consultation were problems in ventilator settings (9 patients), change of symptoms and laboratory results notification (7 patients) and making a decision for ventilator support (1 patient). The questionnaire was used to evaluate the satisfaction of hologram technology in 3 aspects (the system, the utility and the development team) by 1 CVT physician, 1 intensivist and 14 ICU nurses who used HoloLens for the consulting system in the ICU. The mean (SD) of the satisfaction of hologram technology in 3 aspects were 4.37 (0.62), 4.50(0.63) and 3.87(0.50) respectively (score out of 5). When compared with the old-fashioned consultation through LINE application, HoloLens did not reach expectations from the user's perspective.

Conclusions
The HoloLens introduces novel opportunities that highly skilled staff trained in critical care able to deliver timely, quality care service to patients admitted to ICU in the areas that lack healthcare professional. In addition, it can be used as another effective way of bedside training for non-ICU physicians and nurses in ICU with limited experience. However, there are some limitations of Holo-Lens that should be developed to be better for more convenient use in the future.

Background
Rapid diagnosis is essential in the case of sepsis. Current methods for diagnosis of sepsis have insufficient specificity causing delay in clinical intervention. Neutrophil CD64 index, measured by Flow cytometry, shows great promise as a sensitive and specific marker of neonatal sepsis, but its use is severely constrained due to lack of access to high quality Flow cytometry. Our hypothesis was that use of immunoassays for simultaneous measurement of the total of two proteins, neutrophil CD64 and a neutrophil-specific protein in whole blood, which together yields the nCD64 index as measured by Flow Cytometry.

Methods
We are recruiting adult ICU patients (n=50) with clinically suspected sepsis (Alfred ICU) and age-matched controls from the ICU (n=50), and healthy individuals (n=50). Thus far 40 patients with clinically suspected sepsis, 18 ICU controls and 50 healthy volunteers have been recruited. Levels of a selected neutrophil specific marker, neutrophil elastase (NE), and neutrophil activation marker (CD64) are measured by commercial sandwich ELISA kits. Samples were also tested using Leuko64 TM assay, procalcitonin electrochemiluminescence immunoassay (PCT ECLIA) and further examined by flow cytometry with intracellular staining, and microscopy. Results Analysis of total CD64 and NE in whole blood demonstrated non-linear correlation in healthy controls, allowing us to define two separate methods for assigning assay cut-offs based on either (1) total CD64 (Fig. 1A), or (2) elevated CD64 relative to NE for samples with NE < 2.5 μg/ml (Fig. 1B) Background Rapid and accurate profiling of infection-causing pathogens is crucial for critically septic patient management. During our previous investigations, we evaluated both the performance of some syndromic tests [1,2] and the most successful sepsis biomarkers [3]. Despite advances in molecular diagnostic techniques, blood culture analysis remains the gold standard for diagnosing sepsis. A rapid molecular noncultural method was recently developed for identification of Candida and bacterial bloodstream infections from whole blood samples. The aim of this study was to evaluate the performance of an automated assay that utilize proprietary T2 Magnetic Resonance technology to enable faster and potentially accurate identification for pathogens that cause sepsis directly from whole blood clinical samples. Materials and methods Thirty-six whole blood samples, obtained from critical patients admitted to "Magna Graecia" Hospital of Catanzaro (Italy), were analyzed by innovative technology of Magnetic Resonance (T2 Biosystems, Inc., USA) and compared to a conventional laboratory method (Vitek2, bioMèrieux, France). T2MR combines the nuclear magnetic resonance and PCR molecular assay to directly detect six bacteria species (Enterococcus faecium, S. aureus, A. baumannii, E. coli, P. aeruginosa and K. pneumoniae ) and the five most common Candida species (C. albicans, C. tropicalis, C. glabrata, C. parapsilosis, C. krusei) directly from freshly obtained whole blood samples, in approximately 3.5-5 h. Specifically, T2 system lyses red blood cells, concentrates microbial cells and cellular debris by mechanical bead-beating, amplifies the DNA using a thermostable polymerase and target-specific primers. The amplified DNA is detected by hybridization to the superparamagnetic nanoparticles, which causes large changes in the sample's T2MR signal. A synthetic heterologous DNA target used as an internal control, is processed with each clinical specimen, to monitor the integrity of the T2MR results. The limit of detection of T2MR is 1 colony-forming unit (CFU)/mL.

Results
Of the overall 36 whole blood samples tested for candida and bacteria panels, four specimens were not included in the analysis because the internal control was invalid, as well as the samples whose blood cultures were missing (seven specimens). Data obtained by T2MR was confirmed in 100% (11/11 cases) for T2Candida panel when compared to Vitek2 method as well as antigen mannan. Concerning T2Bacteria panel, the assay was able to identify 12/14 (86%) specimens when compared to conventional laboratory method respectively. Furthermore, in one case, the novel T2MR assay was capable to detect A. baumannii species after 3.5 h, while the positivity of the conventional blood culture was obtained only six days before.

Conclusions
The data of this study demonstrated that T2MR represents a highly promising molecular diagnostic method for the rapid identification of the most important pathogens causing-sepsis. The use of such technology may warrant a place in the diagnostic pathway of septic syndrome. Our data confirmed that presepsin is a valuable prognostic biomarker for the studied patients (p=0.0169). However, PCT did not behave as an affordable prognostic marker. The Treg cytokine, IL-10 seemed to exhibit a trend toward increased values in dead vs. alive patients. Th1 cytokines did not show any significant difference between dead and alive patients at admission time. The Th2 cytokine, IL-4 followed the expected behavior of a lack of any significance in such very early sepsis phase.

Conclusions
In conclusion lack of significance between different groups of cytokines evaluated, further support the value of presepsin as an excellent prognostic biomarker. Although a trend toward significancy was found in dead vs. alive patients for IL-10, reflecting the basal higher level of such immunosuppressive mediator even in early sepsis pathophysiology. An unexpected result was the lack of significant difference between the same groups regarding Th1 cytokines, which should be quite higher in such early sepsis stage, particularly in nonsurvivors. Both presepsin and cytokine prognostic role warrants to be assessed in larger cohorts of septic patients.

P28
To (2) CYP1A1-overexpression and CYP1A1-loss-enzymatic-function macrophages were adaptively transferred to wild type mice challenged by LPS and CLP to observe the survival and illustrate the molecular mechanism.
(3) The level of CYP1A1 in human peripheral blood monocytes (PBMCs), isolated from healthy and septic individuals, were also detected by qRT-PCR and LSCM. Results CYP1A1 was highly expressed in mice peritoneal macrophages challenged by both LPS and heat-killed E.coli. Moreover, CYP1A1 and inflammatory factors were aberrant highly expressed in LPS challenged peritoneal macrophages from Ahr-/-mice compared with that from Ahr+/+ mice. Using CYP1A1-overexpression RAW264.7 cells (CYP1A1/ RAW), we found that CYP1A1 could intensify macrophages inflammatory responses by up-regulating TNF-α and IL-6 through enhancing JNK/AP-1(c-fos and c-jun) phosphorylation rather than NF-κB (p65 and p50) phosphorylation. We discovered that CYP1A1/RAW secreted more 12(S)-HETE, while 12(S)-HETE antibody could attenuate intensified inflammatory responses caused by CYP1A1 overexpression. Administrating 12(S)-HETE alone into normal RAW264.7 cells could also reinforce inflammatory responses even without LPS stimulation. Furthermore, we showed that loss-of-enzymatic-function of CYP1A1 in macrophages impeded LPS induced JNK/AP-1 phosphorylation and inflammatory factors secretion due to impaired generation of 12(S)-HETE. Mice transferred with CYP1A1-overexpression but not with CYP1A1-loss-enzymatic-function macrophages were highly susceptible to LPS and CLP challenge. In addition, the CYP1A1 expression and 12(S)-HETE concentration levels were elevated in PBMCs from septic patients and had a positive correlation with Sequential Organ Failure Assessment (SOFA) score. Conclusions Altogether, these results revealed a novel signaling axis, namely CYP1A1-12(S)-HETE-JNK-AP-1, regulating macrophages inflammatory responses in sepsis, which could be a promising target for preventing and treating sepsis.

Results
Our study showed that 12(S)-HETE production in LPS-activated macrophages was increased and its plasma level was elevated in septic mice induced by LPS or faeces (i.p) respectively. When added into macrophages culture system without LPS stimulation, 12(S)-HETE significantly enhanced the production of TNF-α, IL-6 in a dosedependent manner. In vivo administration of 12(S)-HETE (i.p) also strengthened pro-inflammatory response and organs injury in septic mice, which is evidenced by increased TNF-α, IL-6 levels and more serious lung and liver pathological injury and lower survival rate. When compared with healthy control, the plasma levels of 12(S)-HETE, TNF-α, IL-6 were elevated in septic patients, and 12(S)-HETE concentrations of septic patients showed a significant positive correlation with SOFA score, (r=0.850, P<0.002).

Conclusion
It is suggested that 12(S)-HETE possesses pro-inflammatory effect and is positively correlated to SOFA score in septic patients, which might be a potential biomarker for the assessment of sepsis outcomes and the possible therapeutic target for the treatment of sepsis.

Background
There are still limited data on the role of human endogenous bacteriophages in the prevention and treatment of infectious complications and sepsis in critically ill patients. The aim of this study was to investigate into the influence of endogenous bacteriophages on the mortality is intensive care unit patients with bacteremia and sepsis. Materials and methods 25 intensive care unit patients with severe combined trauma were enrolled in this observational single-center study (8 women, 17 men; 48.3±18.3 y.o.). 28 blood samples were collected. Blood cultures were performed using an automated Bactec-9050 blood culture analyzer (Becton Dickinson, USA). Identification of the isolated microorganisms was carried out using an automatic analyzer WalkAway-40 (Beckman Coulter, USA). In order to isolate endogenous bacteriophages patients were sampled for blood and urine cultures. Proceeding of bacteriophages was performed by means of traditional virological methods. Bacteriophages extracted from the lysis zones after spot testing were examined on an electron microscope JEOL-1011 (Japan). Statistics was done by Microsoft Excel and Statistica 6.0, X2 criterion was used.

Results
Out of 25 patient's bacteriophages were isolated in 10. Blood microbiology was: in 4 cases Klebsiella pneumoniae, in 3 -Acinetobacter sp., in 2 -Staphylococcus spp., 1 case of S. aureus and Enterococcus faecalis. In the group of patients who had their own endogenous bacteriophages 4 people died (40%). At the same time 2 patients had their own phages identical to the isolated microorganisms (Staphylococcus aureus and S. epidermidis) which were then eliminated. Later on these patients isolated strains of K. pneumonia to which endogenous phages were absent and patients died due to this infection. Out of 15 patients who did not have endogenous bacteriophages 9 (60%) died (p <0.002). Blood cultures in these 15 patients were: 5 cases -S. aureus, 2 -Staphylococcus sp., 5 -K. pneumonia, 2 -Acinetobacter sp. and in 1-Enterobacter aerogenes.

Conclusions
Endogenous bacteriophages influence the mortality of trauma patients with bacteremia and sepsis: mortality was significantly higher in the absence of endogenous bacteriophages (p <0.002).

Background
The intestinal epithelium is a continuous single-cell layer separating the sterile bloodstream from the gut microbes. Accumulating evidence suggests that this barrier could be substantially impaired during severe systemic inflammation. Importantly, the severity of derangement of intestinal permeability in critically ill patients predicts subsequent development of multiple organ failure and its regulation by host factors remains unclear. Finding ways to maintain the intestinal epithelial integrity is one of the major directions to improve the clinical outcome of sepsis Materials and methods We performed cecal-ligation and puncture on C57BL/6 mice to induce polymicrobial sepsis. Animals undertook either sham surgery with saline injection, CLP with N,N-dimethylsingosine (a potent inhibitor of sphingosine kinase 1), or CLP with saline injection. At 20h, the animals were orally fed with 4kDa fluorescein-dextran (FITC). At 24h after surgery, we collected blood and harvested ileal tissues for molecular characterization, including transcriptomic analysis.

Results
Of septic mice, intestine exhibited remarkable pathological changes, including increased epithelial apoptosis, reduced expression of tight junction proteins, and elevated serum FITC level. Transcriptomic profiling revealed that sphingosine kinase 1 increased more than six-fold after sepsis-induction. Pathway enrichment analysis revealed significant pathways pertinent to pro-inflammatory response. Animals with pharmacological inhibition using N,N-dimethylsphingosine had lower level of serum FITC, less severe sepsis symptoms and lower total bacterial DNA in the blood.

Conclusions
Inhibition of sphingosine kinase 1 may be a therapeutic strategy for preventing gut barrier dysfunction in sepsis. Further preclinical studies will be needed to confirm the beneficial effect in prior to clinical trials.
Background Dysregulated inflammatory responses can lead to host-driven sequelae such as sepsis, as well as an immune response not optimally tailored for bacterial clearance. Heterogeneity in the host response to invasive bacterial infection suggests that specific biomarker signatures could be utilized to differentiate patients prone to severe disease, thereby facilitating earlier implementation of more aggressive therapies. To further elucidate the inflammatory correlates of poor clinical outcomes in bacteremic patients, we evaluated the association between a large panel of blood proteins at initial presentation of bacteremia and disease severity outcomes.

Materials and methods
We conducted an observational study (n=32) to evaluate the prognostic value of 64 circulating protein biomarkers for mortality or persistent bacteremia (defined as positive blood cultures for >5 days) in patients with S. aureus bloodstream infections. Prioritized biomarkers were evaluated in a case-control study of 156 patients with S. aureus or Gram-negative bloodstreams infections using serum samples collected within 1-3 days of empiric antibiotic therapy.

Results
We identified 8 candidate biomarkers of inflammation and endothelial activation that were prognostic for mortality and relevant to both  Background Autophagy plays a central role in controlling innate immune responses by degradation of proteins. Autophagic factors like p62/ SQSTM1 directly interact with molecules of the Toll-like receptor 4 (TLR4) pathway [1], which is activated in macrophages in response to lipopolysaccharide (LPS) [2]. In this signaling cascade Pellino3 serves as an IRAK E3 ligase and scaffold protein [3]. Its autophagydependent degradation impairs the hyperinflammatory phase during LPS challenge by inhibiting TLR4 signaling and IL1b expression [4]. Based on these in vitro data, we want to translate this setting in an in vivo cecal ligation and puncture (CLP) mouse model to substantiate the modulation of Pellino3 expression during sepsis as a new strategy for the development of a therapy approach. Materials and methods We used cecal ligation and puncture (CLP) to induce polymicrobial sepsis in wildtype mice to monitor Pellino3 expression. 6h/24h/48h following sepsis induction peritoneal macrophages and peritoneal fluid were harvested. Pellino3 and Il1b mRNA were analysed by RT-PCR. Pellino3 as well as LC3, p62 and pro-IL1b protein expression was elucidated by western analysis. IL1b levels in the peritoneal fluid were determined by cytometric bead array (CBA) via FACS analysis.

Results
We observed an initial increase of Pellino3 protein expression in peritoneal macrophages of wildtype mice after 24 h following CLP. This is accompanied by an increase of LC3-II, suggesting autophagy inhibition, and an increase in pro-IL1b. Concomitantly, IL1b secretion in the peritoneal fluid is elevated. 48h after CLP Pellino3 is degraded, most likely due to increased autophagy as shown by simultaneously degradation of LC3-II. In lysates of peritoneal macrophages pro-IL1b protein expression is decreased and consequently in supernatants of these cells IL1b is attenuated. Conclusions Our in vivo data confirmed our in vitro findings of a correlation of Pellino3 and its degradation, impairing IL1b expression and secretion. The proof that autophagic processes are involved in Pellino3 degradation is still elusive in the mouse in vivo model. Therefore, experiments in autophagy-deficient mice using a p62-knockout model are planned for the near future. Moreover, the induction of autophagy at different timepoints during sepsis progression will help to confirm the mechanism. Since an overwhelming proinflammatory response is a main problem during sepsis, targeting Pellino3 by considering the knowledge of autophagy induction should be approached and extended in the future as a new therapy regime.

Background
Peroxisome proliferator-activated receptor gamma (PPARγ) has been shown to play a crucial role in immunosuppression during sepsis. Activated PPARγ is upregulated in T cells of septic patients, sensitizing these cells to PPARγ dependent apoptotic depletion [1]. In the mouse cecal ligation and puncture (CLP) sepsis model, both T cellspecific PPARγ gene knockout and systemic pharmacological PPARγ antagonism by GW9662 improved survival [2,3]. As GW9662 it is not suitable for therapeutic use in humans, we developed and improved a new class of selective PPARγ modulators (SPPARγMs) for use in the treatment of sepsis in humans [4]. Materials and methods To design improved SPPARγMs, initial three-dimensional computational modeling was performed. Potential candidates were synthesized and kinetically characterized in vitro using PPARγ-dependent transactivation assays in human cell lines. Cytotoxic effects of promising compounds were tested in cell viability assays. The intracellular accumulation of suitable compounds in the same cell lines was assessed by mass spectrometry. To verify the potential modulating effects of compounds on PPARγ target genes, gene expression was studied using quantitative real-time polymerase chain reaction (qPCR). Based on their in vitro properties, promising compounds were further investigated for their in vivo pharmacokinetic properties. Compounds fulfilling in silico, in vitro and in vivo requirements were finally tested in the mouse CLP sepsis model.

Results
In an initial screen, T-10017 was identified, from structurally related chemical compounds, as a potential lead compound for a new class of competitive, reversible PPARγ antagonists. In the mouse CLP sepsis model, administration of T-10017 nanoparticles led after 48 h to a significant improvement in the survival of septic mice, compared with control mice, treated with unloaded nanoparticles. Starting from T-10017, we further optimized the compound properties by chemical modification, including agonistic and antagonist characteristics, solubility and stability, resulting in two derivatives with markedly improved in vitro and in vivo properties.

Conclusions
Compounds from our new class of SPPARγMs are significantly effective in the treatment of murine sepsis. These compounds competitively inhibit PPARγ dependent T cell depletion, help to maintain immune competence, facilitate the controlled therapeutic regulation of hypoinflammation and thus, reduce severity and mortality of sepsis.

P38
The process of septic shock is attenuated by the intravenous administration of peptide VSAK: A FDG-PET study Ismael Luna-Reyes 1 , Eréndira

Background
The development of an uncontrolled inflammatory response is considered the main responsible of the most deleterious effects during severe sepsis and septic shock, that finally can lead to the condition known as Multiorgan Dysfunction Syndrome (MODS). The inflammatory response is triggered by pathogen derived molecules, mainly lipopolysaccharides (LPS). LPS and other pathogen molecules are released in early stages of sepsis and recognized by Pattern Recognition Receptors (PRR), directly involved in the initiation of the inflammatory response. In previous work, we have demonstrated in vitro and in vivo that VSAK, a peptide derived from the last eighteen aminoacids of the carboxy-end segment of CETPI, considered an isoform of the cholesteryl-ester transfer protein (CETP) [1], is able to bind LPS and therefore, neutralize their harmful effects [2]. The present study provides additional evidence supporting the protective effects of VSAK administration in rabbits induced to a septic-shock-like state employing Positron Emission Tomography (PET).

Materials and methods
During the experiment, we used ten male Dutch dwarf rabbits provided by a local certified farm. Rabbits were randomly assigned to four experimental groups: 1) Control, 2) VSAK treatment, 3) LPS treatment, and 4) LPS+VSAK treatment. Immediately after, rabbits were administered 2-[18F]fluoro-2-deoxy-D-glucose (FDG) as the radiotracer and monitored by PET. Each scan was performed during 70 minutes and PET data analysis carried out using spherical areas to define Volume of Interest (VOI) in order to reduce slight variations associated with differences in animal size. Results Figure 1 shows PET images from four representative rabbits, corresponding to each experimental group. In general, a marked decrease of glucose uptake can be observed in the LPS group. Results obtained from the normalized Standard Uptake Value (SUV) taking into account a spherical area of 5.5 mm in all four experimental groups are shown in Fig. 2. In all cases, SUV estimated in the first ten minutes was considered as 100% SUV and used to normalize individual data sets. A negative correlation between time and %SUV is observed in all experimental groups. The negative correlation is slightly more pronounced in the LPS group, decreasing up to 73.5% ±2.7 SEM at 70 minutes, while in the control group (83.43% ± 1.3 SEM), VSAK (83.99% ± 6.55 SEM) and VSAK-LPS (89.15% ± 2.09 SEM).

Conclusions
This study supports the potential in vivo use of peptide VSAK in order to neutralize the harmful effects of LPS and therefore avoid a metabolic dysfunction pattern during sepsis and septic shock [3]. Background Sepsis patients commonly present with features including progressive subcutaneous and body cavity oedema, typically caused by permeabilization of the vascular endothelial monolayer. Strict endothelial barrier control is mediated by structures located at cell-to-cell contact sites known as endothelial tight junctions and adherens junctions. Of these junctions, VE-Cadherin is arguably the most important functional component. Previously we have demonstrated that bacterial attachment to the endothelium decreases VE-cadherin expression resulting in increased vascular permeability [1][2][3]. MicroRNAs (miR-NAs) play a key role in maintaining normal endothelial cell function and their dysregulation has been linked to a number of clinically important diseases including cancer, myocardial infarction, neurodegeneration and infection [4]. This study aims to investigate the role of miRNAs in controlling VE-cadherin during infection, and to uncover their influence on barrier integrity and permeabilization of the vascular endothelial monolayer. Materials and methods Human endothelial cells were subjected to shear forces (10dynes/ cm2), mimicking physiological conditions experienced in the vasculature [1][2][3]. Via S. aureus administration, miRNA alterations were uncovered and determined by Taqman array cards, generating miRNA profiles of both uninfected and infected cells (RQ = 2-ΔΔCt). miRNA candidates of interest were selected bioinformatically and confirmed via individual miRNA assays. Potential mRNA targets for the candidate miRNA chosen were established bioinformatically and confirmed by Next Generation RNAseq, western blots, flow cytometry,  Background Sepsis is associated with severe vascular endothelial cell dysfunction which rapidly leads to dysregulation of normal systemic hemostasis and vascular reactivity triggering widespread tissue oedema. Although this process is considered central to the progression to organ failure during sepsis there are no therapeutic options available to treat or prevent endothelial dysfunction. Using an ex-vivo model of sepsis, we recently demonstrated that inhibition of major vascular endothelial cell integrin aVb3 prevented endothelial dysfunction induced by a number of sepsis pathogens by stabilising the vascular endothelium [1,2,3]. Aims This study investigated the therapeutic usefulness of blocking aVb3 to control vascular endothelial cell dysfunction and stabilize the endothelium in an Escherichia coli induced rat pneumonia model of sepsis.

Materials and methods
Escherichia coli (CFT073) was delivered directly into the lungs, via a 14G cannula, of anaesthetized Sprague Dawley rats. Cilengitide, a specific aVb3 inhibitor, was administered intravenously (IV). Following progression to sepsis animals were sacrificed, organs were harvested and homogenised. Bacterial load in the spleen, liver, kidney, lungs and blood was assessed by viable count assay. Biomarker lactate was measured by ELISA. Vascular permeability was assessed using Evans blue dye.

Results
All animals were sacrificed at 24hrs. Bacterial counts (CFU/ml) in the blood (Fig. 1) and all major organs including the liver ( Fig. 2A), kidney (Fig. 2B), spleen (Fig. 2C) and lung (Fig. 2D) were significantly lower in animals treated with cilengitide (control vs cilengitide treated, P<0.05). We next investigated if cilengitide was stabilizing the vascular endothelium thus preventing the dissemination of bacteria from the lung into the bloodstream. To test this, control noninfected animals were injected with Evan blue dye. These animals did not leak dye from their blood vessels as they maintain vascular integrity. In contrast animals that developed sepsis from the pneumonia displayed significant leakage from their blood vessels. Most strikingly, animals that developed pneumonia and were treated with cilengitide, failed to leak dye from their blood vessels suggesting that their endothelium was stabilised by the cilengitide thus reducing vascular permeability (Fig. 3) (untreated vs cilengitide treated, P<0.01). Finally and most critically, there was a significant reduction in the serum biomarker lactate in the cilengitide treated animals versus control animals (P<0.01) (Fig. 4), suggesting recovery.

Conclusions
Our results suggest that stabilizing the vascular endothelium by targeting the major endothelial cell integrin aVb3, may prevent the dissemination of bacteria to the blood and all major organs thus preventing progression to sepsis.

Background
In patients with sepsis general platelet hypo-responsiveness to exogenous agonists is associated with mitochondrial dysfunction [1], but the mechanisms of this are not yet known. As shown previously, aromatic microbial metabolites (AMM) may influence on some mitochondrial functions [2]. The aim of this work is to evaluate the potential effect of aromatic metabolites on platelets aggregation in whole blood. Materials and methods We evaluated platelet aggregation (native and ASPItest) using the Mulitplate platelet function analyzer (Roche), a total of 112 tests in whole blood of healthy volunteers (n = 11). Additives (20 μl) of AMM were selected in clinically significant dosage as shown early in patients with bacterial infection and sepsis [3]. To compare the data, the Sign Test was used for the median.

Results
There were no statistically significant dose-dependent effects of AMM on native platelet aggregation. A significant decrease of median ASPItest under the influence of HPAA 15 μM and HPLA 30 μM were identified on 27% and 19,5% respectively, that presumably allows to regard these acids as cyclooxygenase (COX) and IIb / IIIa inhibitors antagonists. Complex mixtures had a more pronounced antiaggregation effect, especially the mixture II-C, comparable to the impact of LPS. Conclusions Aromatic microbial metabolites have the potential to affect platelet aggregation in vitro, which requires further study.

Background
Sepsis is a syndrome defined as life-threatening organ dysfunction caused by a dysregulated host response triggered by an infection [1].
To date, numerous studies have captured aspects of the host response by measuring biomarkers pertinent to sepsis in human subjects [2,3]. Historically, the aim of these studies has been to demonstrate potential diagnostic or prognostic value of a set of biomarkers as opposed to specifically characterizing the degree of dysregulation of the host response. In our study, we sought to quantify the degree of dysregulation of the host response in a clinically relevant population in an intuitive and unbiased manner. Materials and methods In a cohort of 252 patients (Table 1) suspected of infection, we measured 15 biomarkers spanning various aspects of the host response as a function of time per patient. We specifically hypothesized that a multi-dimensional measure of time-dependent abnormality (termed the 'personalized immune score') could serve as a useful surrogate for quantifying dysregulation. We realize this may not capture the spirit of "dysregulation" but believe it could be a useful step forward toward that goal. We quantified multi-dimensional time-dependent abnormality through the formula ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ðx−yÞ T S −1 ðx−yÞ q , where x referred to a set/subset of 15 biomarker measurements at a given time point for a given patient, S referred to the covariance matrix of all biomarker measurements, and y referred to a set/subset of 15 biomarker measurements at discharge for a given patient (Fig. 1). We tested the association of relevant features/variables with respect to the following sepsis-related outcomes: ICD9/10 codes (sepsis/severe sepsis/septic shock), the sepsis-3 definition, clinically significant bacteremia, and hospital length of stay.

Results
We constructed three variants of the personalized immune score: one using all the biomarkers, one using only 'low signal' biomarkers, and one using only 'high signal' biomarkers. We assessed the associations of each of the variants of the personalized immune score and the respective biomarkers that composed each of the immune scores with respect to our outcomes of interest ( Fig. 2A-C). We also compared the optimized personalized immune score with a patient's SIRS and SOFA score. (Fig.  2D). Raw p-values for each entry in the heatmap are reported in Tables 2, 3, 4 and 5, with the worst p-value (termed the 'bottleneck p-value') highlighted in red. In Fig. 3, we visualized the optimized personalized immune score as a function of time stratified by outcome. Overall, there was a greater number of stronger associations between each outcome and all variants of the personalized immune score compared to that of any individual biomarker or EMR feature. Furthermore, the optimized immune score yielded stronger associations than any individual biomarker for three out of the four sepsisrelated outcomes. Finally, when examining the bottleneck p-value in each table, all variants of the personalized immune score achieved the lowest value. Conclusion This study shows that a multi-dimensional measure of timedependent abnormality of biomarkers that span various aspects of the host response could be a useful metric in evaluating the health state of a septic patient. Ultimately, we believe our work lays a foundation for quantifying the dysregulation of the host response in an unbiased manner.      Background Improved prognostic and risk-stratification tools for patients with sepsis are needed both for improved clinical resource allocation and for prognostic enrichment in clinical trials [1]. We recently described a 12-mRNA host-response signature that, when measured quantitatively in whole blood, can accurately predict 28-day mortality [2]. In particular, we previously showed that the 12-gene 'Stanford' score has prognostic power both alone (AUROC mean 0.89, range 0.68-1.0 across 9 validation cohorts) and in combination with standard riskstratification tools (such as SOFA) when linearly combined (mean boost in AUROC 0.10). Materials and methods Patient with sepsis according to Sepsis-2 criteria were enrolled at admission across multiple Greek ICUs. Whole blood was drawn into PAXgene RNA tubes at admission and frozen. For a pilot subset of 29 patients, RNA extraction (Qiagen QIAcube) was followed by measuring the 12-mRNAs using NanoString nCounter with a custom codeset. We combined expression levels of the 12 mRNAs into a single score as previously described [2]. Scores were used to directly calculate AUROCs, and also combined with clinical risk scores using linear regression trained locally.

Conclusions
The 12-mRNA host-response signature continues to show validity for the prognosis of 28-day mortality in patients with sepsis in independent validation. Its ability to improve on standard prognostic scores such as qSOFA and SOFA is reconfirmed. A larger set of samples from the described cohort will be analyzed to increase the statistical power and overall validity of the results.

P45
Cost effectiveness of emergency department use of a novel multi-mRNA assay for diagnosis and risk assessment of acute respiratory tract infections and   for a hypothetical cohort of patients. Selected input outcomes were based on review of peer-review articles. The 30-day outcomes considered in the study were expected cost per person presenting with ARTI or suspected sepsis to an ED, incremental cost per life-year saved, antibiotics days and hospital length of stay (HLOS).

Results
In the base-case scenario, across a cohort of 1,000, the HostDx Sepsis arm saves 10 lives/1,000 person-years and results in expected costs of $1,404/person less than the standard of care arm, reduced total antibiotics exposure by 1,097 days and 341 days of hospital stays comparing to the standard of care arm. Standard of care relied on empirical treatment, clinical judgment and multiple testing for bacterial and viral infections, with accuracy for bacterial, viral and mortality risk differentiation lower than that of HostDx Sepsis. In one-and two-way sensitivity analysis, model results were most sensitive to the HLOS, followed by the HostDx Sepsis test price, and length of antibiotics treatment.

Conclusions
The HostDx Sepsis arm demonstrated clinical utility and cost effectiveness versus the current standard of care arm. Improved care is reflected by fewer unnecessary antibiotic prescriptions and side effects, fewer admissions and shorter HLOS. Interventional studies are necessary to compare actual clinical and costs outcomes between current practice and that which incorporates HostDx Sepsis.

Background
The main biomarkers for diagnosis and prognosis of sepsis are the following three categories: acute phase proteins, vascular related biomarkers, cytokines. However, it is still controversial whether these large numbers of biomarkers improves prediction of sepsis in trauma population. The aim of this study was to explore the predictive efficiency of new model based on creatinine, prothrombin time and procalcitonin in severe trauma patients. Materials and methods This was a retrospective observational analysis (reviewing the medical records) of 593 patients with severe trauma hospitalized in ICU of the Daping Hospital of Army Medical University from January 1, 2012 to December 31, 2017. Two models were established: Model-1, a logistic regression model with sepsis as the response value and mechanical ventilation, shock and APACHE-II at ICU admission as explanatory variables; and Model-2, the same variables as in Model-1, with creatinine, prothrombin time and procalcitonin added as explanatory variables. Data were processed with t test, chi-square test, and receiver operating characteristic (ROC) curves were determined for both two models. The discrimination indexes (C-index and Nagelkerke's R2) and calibration curves for predicted sepsis versus observed sepsis were calculated for each model.

Results
There were 18527 trauma patients admitted at Daping Hospital of Army Medical University from January 1, 2012 to December 31, 2017.
The study population of 593 cases with severe trauma hospitalized in ICU were ≥16 years old, incoming ICU within 24 hours after injury, the length of ICU stay ≥48 hours, ISS≥16, perfect clinical data, and without coexisting illness. 252 cases (42.50%) of 593 patients examined sepsis in hospital. In Model-2, creatinine, prothrombin time and procalcitonin were each a significant independent risk factor for sepsis diagnosis. The LR test comparing between Model-1 and Model-2 indicated significant improvement in Model-2 (P<0.001). Model-2 resulted in better values than Model-1 for discrimination indexes (Cindex: 0.989 VS 0.941, Nagelkerke's R2: 0.694 VS 0.570). ROC curves for Model-1 and Model-2 were 0.793 and 0.879 (P<0.001). Calibration improved the model fit: the mean absolute errors were 0.027 and 0.005 and the mean squared errors were 0.00069 and 0.00020 for Model-1 and Model-2, respectively.

Conclusions
In conclusion, the ability of a new model to predict sepsis in severe trauma patients improved following the addition of creatinine, prothrombin time and procalcitonin measured on ICU arrival within 24 hours as explanatory variables.
could potentially be applied in hospital of low-middle income country.   Background Sepsis causes morbidity and mortality. In the UK, sepsis causes 250 000 cases and 44 000 deaths annually [1]. Better diagnosis and treatment could save 11,000 lives and £160 million each year [2]. Sepsis 6 are interventions [1,[3][4] which can reduce mortality by 50% if all components are implemented within 60 minutes [5]. Local audit of patients with suspected and confirmed sepsis reveals low compliance with these interventions. We sought to develop a teaching strategy to improve sepsis recognition and management. Traditional classroom-based teaching (TCBT) methods and an online module were developed and staff sepsis training evaluated. These teaching methods included a case-study approach encompassing all facets of sepsis recognition and management. We sought to evaluate the acceptability and effectiveness of each intervention. Materials and methods Current local, national and international guidelines informed development and subject matter of both TCBT and an online module. This module is available to all hospital employees via local intranet and mobile devices (Fig. 1). Both teaching methods were departmentally peer-reviewed and trialled in pilot clinical areas. Defined outcomes were specified: • Acceptability and effectiveness of TCBT and online teaching • Definition of sepsis, septic shock, sepsis red flags, Sepsis 6 components, desired time to management of sepsis • Individuals' confidence in sepsis management in a post-teaching survey • Ongoing collection of qualitative feedback Trust clinical staff members will be allocated either face-to-face training and/or an online module and asked to evaluate their experience and recall of key components Results 1. TCBT-pre and post teaching results were recorded. There was a marked improvement in all areas specified including sepsis knowledge and confidence in sepsis management (Fig. 2). 2. Online module was well received by staff and was considered an effective and enjoyable way to learn. Data collection is ongoing, including the specified outcomes.

Conclusions
We have demonstrated benefits to staff sepsis knowledge from TCBT. Advantages of online teaching include availability, Fig. 1 (abstract P49). Seahorse assay parameters in healthy controls, low and high risk infection groups, measured in oxygen consumption rate (OCR). OCR represents cellular metabolism through mitochondrial respiration. Basal denotes baseline energetic demand of the cell; maximal respiration shows the maximum rate of respiration the cell can achieve; spare respiratory capacity denotes the capability of the cell to respond to energetic demand; ATP production shows ATP produced by the mitochondria to meet the cellular energetic need. p value (2-way ANOVA): ns >0.05, *0.01-0.05, ****<0.0001. Notably, this reduced mitochondrial respiration was not accompanied by a compensatory increase in glycolysis (Fig. 2) Fig. 2 (abstract P49). Extracellular acidification rate (ECAR)-time plot as measured by Seahorse assay in healthy controls, low and high risk infection group. ECAR represents cellular metabolism through glycolysis. Seeing no difference in total ROS production across the three groups ( Fig. 3), indicating that oxidative stress did not play a role in reducing mitochondrial respiration in PBMC of high risk infection groups. Fig. 3 (abstract P49). Total cellular ROS as measured by DCFDA with flow cytometry analysis (p = 0.4038, Kruskal-Wallis test). Fewer number of samples in low risk infection group was due to inadequate cell number. Apoptosis, a common sequela of reduced mitochondrial function, did not differ between groups; however, the high-risk group did show some signs of early apoptosis (Fig. 4). minimal disruption to clinical care, a standardised approach and recordable outcomes. Online learning modules and TCBT can improve sepsis care and management. High-quality online modules can deliver sepsis training and ensure all staff have access to learning resources. Our data provides insights into the learning experience. Future work will compare TCBT and online teaching.

P52 Withdrawn
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.