Program and Abstracts Thirty-first Annual Meeting of the Surgical Infection Society Palm Beach,Florida May 11–14,2011

O1. Efferent Vagal Nerve Stimulation Abates Acute Lung Injury in Post-Hemorrhagic Shock

Background: Trauma/Hemorrhagic shock (T/HS) induced gut injury is known to initiate a dysfunctional inflammatory response leading to secondary lung injury. Our group has shown that vagal nerve stimulation (VNS) protects the intestinal epithelial integrity, attenuates inflammatory cytokine formation, and protects the lung after a thermal insult. We hypothesize that VNS will protect the lung from injury following trauma/ hemorrhagic shock.

Methods: Male Balb/c mice were subjected to a trauma/hemorrhagic shock model (mean arterial pressure of 35 mmHg for one hour), with and without right cervical vagal nerve stimulation. A cohort of animals underwent abdominal vagotomy (Vx) at the gastro-esophageal junction prior to vagal nerve stimulation and trauma/hemorrhagic shock. Lung histology (H&E), myeloperoxidase (MPO) and ICAM-1 immuno-staining, and MPO enzymatic assay were analyzed 24 hours post-shock in all groups. Using NF-kβ-luciferase transgenic mice, NF-kβ-DNA binding was also analyzed by photon emission analysis (Ivis Spectrum) at 4hrs.

Results: T/HS caused acute lung injury at 24 hours characterized by: 1) pulmonary edema, alveolar hemorrhage and intra alveolar hyaline membrane formation, 2) increased number of MPO positive stained cells and MPO enzymatic activity, and 3) increased ICAM-1 expression in lung endothelium. Vagal nerve stimulation abated these lung injury patterns. Lung tissue analysis at 24hrs post VNS + T/HS showed: 1) H&E histology compared to Sham, 2) marked decrease in PMN infiltration demonstrated by MPO immune staining and MPO enzymatic activity, and 3) decreased ICAM-I expression. Lungs at 4hrs post VNS + T/HS demonstrate decreased DNA binding of inflammatory protein NF-kB; in the images below, red colors represent more intensity and more luminescence expressed as photons/sec/cm2/sr (see figure). Abdominal vagotomy blunted the protective effects of VNS in all experiments, with results comparable to the T/HS group.

OPEN IN VIEWER

Conclusion: Vagal nerve stimulation is effective in protecting against acute lung injury post-hemorrhagic shock. The protective properties of vagal nerve stimulation seem to be applicable in more than one shock model, providing promising potential for clinical applicability.

O2. Outcomes of Optimized Fluid Resuscitation after Severe Burn Injury Utilizing Hemodynamic Measurements by Transpulmonary Thermo-Dilution

Introduction: One of the crucial factors in the successful treatment of severe burns is appropriate fluid resuscitation during the initial phase following injury. At present, the calculation of fluid resuscitation is based on calculations of body weight, burn size and urinary outputs as accurate assessment systems of hydration status are not available. Optimizing the guidance of fluid administration during the critical initial phase would have beneficial effects on the outcome of burned patients. The aim of this study was to assess the benefit of utilizing transpulmonary thermo-dilution by a Pulse Contour Cardiac Output (PiCCO) system for optimized resuscitation.

Methods: A cohort of seventy-six severely burned pediatric patients with burns over 30% total body surface area who received adjusted fluid resuscitation by PiCCO (P) measurements were compared to 76 conventionally monitored patients (C) with similar injury characteristics. Clinical hemodynamic measurements, organ function (DENVER2 score), and biomarkers were recorded prospectively for the first 20 days after burn injury. Statistical analysis was performed using student's t-test and chi-square test as appropriate with significance set at p < 0.05.

Results: Both patient groups were similar in demographic and injury characteristics. Patients who received PiCCO-adjusted resuscitation showed a remarkable reduced incidence of sepsis (P: 9%, C: 17%) and mortality (P: 16%, C: 25%), as well as a significantly lower incidence of cardiac and renal failure during acute hospitalization. Patients in the P group received significantly less fluids (p < 0.05) at similar urinary output, resulting in a significantly lower positive fluid balance (p < 0.05). The CVP in the P group was maintained in a significantly more narrow range (p < 0.05), combined with a significantly lower heart rate at similar systemic blood pressure levels (p < 0.05).

Conclusions: Fluid resuscitation guided by transpulmonary thermo-dilution contributes to superior fluid management and balance during the critical initial phase after burn injury and results in beneficial effects on morbidity and mortality.

O3. Interferon Regulatory Factor-1(IRF-1) Mediates Splenocytes and Macrophages Apoptotic and Autophagic Responses in Endotoxemia

Objectives: The pathogenesis of sepsis is complex and unfortunately poorly understood. Recent studies indicate that an immunosuppressive status due to immune effecter cells apoptosis may contribute to sepsis related mortality. Another cellular process, autophagy, is also activated in sepsis and is primarily believed to play a protective role in the disease progression. Apoptosis and autophagy share a number of common signaling mediators. However, the upstream regulator of these two cellular processes is still incompletely defined. Interferon regulatory factor-1 (IRF-1) functions as pro-apoptotic signal. The function of IRF-1 in regulating autophagy has not been shown. In this study, we hypothesized that IRF-1 may participate in the regulation of both apoptotic and autophagic responses in lymphocytes and macrophages in the model of endotoxemia.

Methods: Male IRF-1 knockout (KO) mice and wild type C57BL/6 countertypes received LPS (20 mg/kg) intraperitoneal injection. Control mice received sterilized PBS. Lung was harvested for histopathology assessment. Spleen was harvested 16 h following LPS administration and analyzed by flow cytometry, western blot (WB), TUNEL staining, immunohistochemistry (IHC) and transmission electron microscope (TEM). In vitro, splenocytes and peritoneal macrophages were harvested and stimulated with LPS, and then were analyzed by WB and immunofluorescent microscope (IM). Raw 264.7 cell was infected by Ad IRF-1 and stimulated with SNAP, then were analyzed by WB, high content screening and caspase-3 activity.

Results: In vivo, IRF-1 KO mice were significantly protected from endotoxin induced mortality (p < 0.05) and acute lung injury. This protection was associated with less splenocytes apoptosis as demonstrated by decreased Annexin V positive cells, fewer TUNEL positive cells, less cleaved caspase-3 staining and decreased apoptotic bodies. Conversely, both splenocytes and peritoneal macrophages from IRF-1 KO mice exhibited evidence of increased autophagic flux as measured by LC3 II activation and p62, increased LC3 spots and increased mitophagy. In vitro, LPS induced IRF-1 activation occurred in a TLR4 dependent, TRIF dependent, MyD88 independent manner. To further explore the mechanism of how IRF-1 modulates autophagy and apoptosis, murine macrophages cell line Raw 264.7 cells were chosen. IRF-1 overexpression inhibits autophagy through the activation of mTOR/p70S6 signaling pathway. On the other hand, IRF-1 participates in both extrinsic and intrinsic apoptotic pathway. Furthermore, iNOS/NO is well demonstrated downstream of IRF-1. SNAP, a NO donor, can mimics the effects seen in IRF-1 WT immune cells, promoting extrinsic/intrinsic apoptotic pathway and inhibiting autophagy.

Conclusion: IRF-1 and NO play a pivotal role in the pathogenesis of endotoxemia by impacting two major mechanisms of cell death/survival, apoptosis and autophagy. IRF-1/NO appears to shift the balance to apoptosis, while IRF-1 depletion results in the opposite effect. Understanding the regulation of IRF-1 activity may serve as a target to improve sepsis related outcomes.

O4. Vascular Catheter Cultures for Suspected Catheter-Related Bloodstream Infection in the ICU: A Tradition whose Time has Passed?

Introduction: Catheter-related bloodstream infection (CR-BSI) associated with central venous catheters is estimated to occur in 80,000 intensive care unit (ICU) patients each year in the USA. While traditional epidemiological criteria require matching catheter and blood culture (cx) organisms, more recent criteria only require bacteremia and the absence of an identified alternative source. Clinically, bacteremia from any known pathogen necessitates antimicrobial therapy, regardless of catheter cx results. We sought to expand our previous database and explore the clinical utility of vascular catheter cxs in critically ill patients.

Methods: We retrospectively evaluated all positive (>15 colony forming units by semiquantitative roll-plate technique) vascular catheter tip cultures (CTCs) from 2006 through 2009 from eight mixed medical-surgical and specialty ICUs in two academic hospitals. CR-BSI was defined as a matching positive blood cx obtained up to four days before or two days after catheter removal. Catheter removal was categorized as either empiric (catheter pulled before blood cx reported positive) or clinical (catheter pulled after blood cx reported positive). The time interval from catheter removal and blood cx draw (empiric) or blood cx positivity (clinical) was determined. Antimicrobial appropriateness was assessed.

Results: 1391 CTCs were obtained during the four year study period. 588 (42%) were negative, 335 (24%) were colonized (<15/roll), and 468 (34%) were positive. There were 143 cases of CR-BSI, representing 10% of all CTCs and 31% of positive CTCs.

In 130 (91%) of 143 cases of CR-BSI, the positive blood culture was drawn before or within 24 hrs of the catheter's removal and dictated antimicrobial therapy. In only 13 (9%) of 143 cases of CR-BSI did the catheter's removal and culture significantly (>1 day) precede the positive blood culture. 55% of CTCs were obtained empirically and with a significantly shorter time interval than those removed clinically. Antimicrobial therapy was deemed appropriate in 87% of CR-BSI cases. 32% of positive CTCs had > two organisms identified. The microbiology laboratory costs to process all CTCs, including organism identification, sensitivity testing and labor were approximately $75,300 and required 600 microbiology technician hours.

Conclusion: In our mixed ICU patient population, 34% of vascular catheters removed were culture positive. Only 31% of positive CTCs were associated with CR-BSI. 91% of all CR-BSIs were identified by bacteremia either prior to or coinciding with catheter removal. Antecedent or concomitant bacteremia, and not the CTC, dictated antimicrobial therapy. Therefore, routine culturing of vascular catheters in cases of suspected CR-BSI in the ICU is of negligible clinical (treatment) utility. Our data suggest that the traditional practice of obtaining CTCs in the ICU should be reconsidered.

Background: Recent literature has shown that in sepsis and hemorrhagic shock there is an associated increase in apoptosis of immune cells in animal models as well as human patients and some studies have correlated this with adverse outcomes in critically ill septic patients. However, to date, the potential trigger inducing this phenomenon is unknown.

Hypothesis: 1) Immune cell apoptosis that occurs after shock is triggered by factors carried in the mesenteric lymphatic system. 2) Immune cell apoptosis that occurs after shock is mediated through the TLR4 receptor pathway.

Methods: In experiment #1, male SD rats were subjected to trauma (laparotomy) plus hemorrhagic shock (30-35 mmHg for 90 minutes; T/HS) or T/HS plus mesenteric lymph duct ligation (LDL). A third group, trauma-sham shock (T/SS) was used as controls. Animals were sacrificed at 3 and 24 hours post shock, spleens and thymi were harvested and assessed for apoptosis by TUNEL and Caspase-3 immunohistochemistry (IHC). In experiment #2, male C57bl mice were challenged with T/HS or T/SS lymph (0.03 ml/gm over 3 hours) followed by sacrifice. The spleen was harvested and assessed for apoptosis via TUNEL and IHC. In experiment #3, TLR4 WT or KO animals were subjected to actual T/HS or received T/HS or T/SS lymph. Animals were sacrificed at 3 hours and spleens were harvested and assessed for apoptosis using TUNEL and IHC.

Results: Due to spatial limitations only the 3-hour TUNEL data will be presented, although similar statistical observations were seen with Caspase-3 and at 24 hours. Splenic and thymic apoptosis was 2-3 fold higher in rats subjected to T/HS (11.07 + 2.8 cells/hpf, 27.94 + 2.41 cells/hpf) than T/SS (5.78 + 0.49 cells/hpf, 8.42 + 1.92 cells/hpf). T/HS + LDL prevented the apoptotic response (6.51 + 1.04 cells/hpf, 11.82 + 2.70 cells/hpf), p < 0.01. Injection of T/HS lymph into C57bl mice increased splenic TUNEL levels compared to mice injected with T/SS (36.55 + 9.04 cells/hpf vs. 20.12 + 5.71 cells/hpf), p < 0.001. This indicated that T/HS lymph is sufficient by itself to cause apoptosis.

Lastly, we found that splenic apoptosis was decreased in the TLR4 KO mice compared to their WT littermates after actual T/HS or after being challenged with T/HS lymph (Table)

Conclusion: 1) T/HS leads to the development of immune cell apoptosis and LDL abrogates those effects suggesting that factors carried in the mesenteric lymphatics after shock contribute to immune cell apoptosis and may be a direct trigger. In addition, splenic apoptosis can be recreated with T/HS lymph injection which further supports this hypothesis. 2) TLR4 KO animals did not show increased apoptosis with T/HS or T/HS lymph injection suggesting that immune cell apoptosis is mediated through the TLR4 pathway.

Introduction: In intra-abdominal infections, empiric antibiotics are started until definitive treatment can be employed. Whether antifungal treatment is included is often decided prior to culture availability and based on poorly delineated clinical characteristics. The efficacy of antifungal therapy in these patients is unproven. We hypothesized that isolation of fungus from intra-abdominal infections can be predicted by the organ of origin and hospitalization status at time of infection, that mortality is increased in these patients, and that early antifungal therapy improves patient survival.

Methods: All intra-abdominal infections in patients treated between December 1996 and September 2010 in a single large tertiary care hospital were identified and classified by presence of fungus by culture results. Logistic regression was used to estimate the probability of fungal infection based on patient demographics, comorbidities, nosocomial nature, previous operation, and the organ of origin. The discriminatory function of the statistical model was measured using the C statistic. The independent contribution of each covariate to the predictive model was assessed using the Wald Chi-square test, and the effects of individual covariates were expressed as odds ratios. All-cause in-hospital mortality was assessed according to presence or absence of fungus by Kaplan-Meier analysis. Logistic regression with propensity score matching by treatment group was performed to further assess contributors to mortality.

Results: We identified 3050 infections. Demographic information is shown (table). The predictive model for fungal infection demonstrated adequate performance (c-statistic 0.692), and identified duodenal origin (OR 2.58, CI 1.40-4.76), nosocomial infection (1.88, 1.38–2.57), and previous operation (1.58, 1.23–2.04) as the most significant contributing variables. All-cause in-hospital mortality was not affected by isolation of fungi. Crude mortality was higher for patients who received antifungal therapy for all infections (16.1% vs. 6.2%, p < 0.0001) and mixed bacterial/fungal infections (18.9% vs. 7.2%, p = 0.01), but not for fungi alone (17.5% vs. 0%, p = 0.28). After propensity score matching (n = 1344, c-statistic 0.882), the variables associated with mortality were age (1.05, 1.03–1.07), liver disease (32.28, 6.38–163.30), Crohn's disease (0.12, 0.02–0.64), and transfusion (3.80, 2.27–6.36), but not presence of fungus (1.40, 0.77–2.53) or treatment with antifungals (1.20, 0.80–1.81).

Conclusion: Duodenal source, nosocomial origin, and prior surgery are risk factors for isolation of fungi from an intra-abdominal infection. Patient mortality is not dependent on the presence of fungus. We were unable to demonstrate that treatment with antifungal medications contributed to improved outcomes in any group. Patients with higher mortality likely have more severe illness unrelated to the presence or treatment of fungi.

O7. P. Aeruginosa Expresses a Lethal Phenotype in Response to a Delicate Balance between Local Phosphate Concentration and Exposure to Kappa Opioids

Introduction: From the standpoint of colonizing microbes, the intestinal tract of a critically ill host represents hostile, nutrient scarce environment in which their virulence circuitry can be activated by secreted host factors (i.e cytokines, ischemic end-products, opioids) and physico-chemical “cues” (i.e low phosphate). Using Pseudomonas aeruginosa as the model pathogen, we created nutrient depleted conditions and determined whether this pathogen processes multiple cues within the local environment (i.e phosphate and the specific kappa-opioid U-50,488) to induce a lethal phenotype from within the digestive tube of the nematode C. elegans.

Methods: C. elegans were allowed to feed on P. aeruginosa pre-grown in nutrient poor media (0.1xTY tryptone, 1 g/L, yeast extract 0.5 g/L) supplemented with either 0.1 mM or 25 mM inorganic phosphate and 200 μM U-50,488 and followed for mortality. Genome wide transcriptional analysis, direct virulence factor production (pyocyanin, pyoverdin) and transmission electron microscopy (TEM) were assessed and correlated to lethality. Based on the above analysis, various mutants of P. aeruginosa were screened in reiterative C. elegans killing assays to identify the key genes and pathways involved in lethality and verified using complementary bioluminescent reporter strains of P. aeruginosa.

Results: Growth in nutrient poor media prevented P. aeruginosa from reaching high cell density (i.e quorum sensing activation) and therefore all experiments were performed at low cell density (OD 600 nm = 0.2–0.3). Under these conditions C. elegans displayed high mortality (∼80% at 24 hrs) when P. aeruginosa was exposed to U-50,488, low mortality (∼20%) in the absence of U50,488, and low (∼10%) mortality when exposed to U50,488 but under high phosphate conditions (25 mM) (n = 50/group, p < 0.01). TEM revealed active formation of flagella and the production of secreted microbial exoproducts (vesicles) during exposure to U-50488 and low phosphate conditions which was prevented when phosphate conditions remained high (i.e 25 mM). Direct assay revealed increased production of pyocyanin and pyoverdin under similar media conditions (p < 0.001). Microarray analysis revealed significant and robust expression of genes encoding multiple virulence factors including pyocyanin, cyanide, proteases, siderophores, and antibiotic resistance Mex pumps among others during low phosphate and U-50488 exposure. Mutational analysis identified PqsE (pqs operon) as a key regulator involved in the development of a lethal phenotype. Use of a bioluminescent reporter to measure the expression of the pqs operon revealed that U-50,488 significantly increased pqs expression in response to exogenous administration of a cognate quorum sensing (QS) signaling molecule HHQ suggesting that kappa opioids enhance P. aeruginosa sensitivity to their own QS signals in a phosphate dependent manner.

Conclusion: These results demonstrate multi-component context dependent virulence activation in P. aeruginosa whereby additive cues can induce a lethal phenotype. Host factors secreted in response to stress such as opioids are directly processed by key regulators of the quorum sensing systems resulting in enhanced virulence. This effect appears to be abrogated during phosphate abundance and highlights the important of understanding bacterial information processing at the molecular level.

O8. The Chasm Between CDC Diagnosis and Bronchoalveolar Lavage Diagnosis of Vap May Affect Payments for Critically Ill Trauma Patients

Background: The chasm between CDC diagnosis and Bronchoalveolar lavage diagnosis of VAP may affect payments for critically ill trauma patients.

Background: Ventilator Associated Pneumonia (VAP) remains a significant cause of morbidity and mortality in trauma patients but despite this, there is no gold standard for the diagnosis of VAP. A significant discrepancy exists between diagnosis of VAP made by Infection Control (IC) Departments relying on CDC criteria and clinical diagnosis based on quantitative broncheoalveolar lavage (BAL). Trauma patients typically have significant pathology on chest radiographs secondary to their traumatic insult that confounds diagnosing VAP. The heavy reliance of the IC definition of VAP upon changing pulmonary infiltrates or worsening chest radiographs potentially leads to higher false positive rates in trauma patients. Conversely, the absence of terms suggestive of pneumonia on a radiology report trigger leads to under-diagnosing VAP by IC's methods. In light of potential non-payment for secondary diagnoses of VAP due to the Center for Medicare/Medicaid Service's “Never Events”, physicians may be faced with the impossible choice of treating their patient or ruining their institution's financial health.

Materials and Methods: this is a retrospective chart review of all intubated trauma patients over the last 3 years. The Infection Control (IC) database, which relies upon the CDC criteria to diagnose VAP, was compared to the clinical practice of the trauma team which relies on quantitative BALs to diagnose pneumonia. Patients with suspected VAP underwent BAL, and this was considered positive for pneumonia if the microbiology data returned with > / = 100,000 colony forming units per ml. The IC database was reviewed to assess why patients with negative BAL were assigned a diagnosis of VAP by IC and the clinical chart was reviewed to assess why BAL positive patients were considered negative for VAP by IC practioners.

Results: Over the study period, 48 patients were diagnosed with pneumonia by IC. 13 of these 48 patients (27%) had a negative BAL, never underwent BAL, or were never treated for pneumonia by the trauma team. Among these 13 patients, the most common reasons cited in the IC database for diagnosing pneumonia were “radiology report stating evolving infiltrate suggestive of pneumonia” (13/13 patients) and “worsening ventilator settings” (8/13 patients). Over the same period, 154 patients were diagnosed with pneumonia via BAL by the trauma team. 119 of these 154 patients (77%) were considered negative for pneumonia by IC. The predominant reasons for IC not diagnosing pneumonia included a lack of significant change in radiology reports, and under-appreciation of increasing ventilator requirements (change of mode of ventilation, rising PEEP and FiO2 requirements).

Conclusions: There is a clear discrepancy between VAP diagnosed by IC and the trauma team. This discrepancy may have several untoward effects including improper denial of payment, or adverse quality indicators for VAP in Trauma Intensive Care Units (TICU). The CDC and IC practioners need to recognize that VAP is much bigger problem in TICUs and should acknowledge that VAP can never be “Never Event” and adjust their definitions and expectations accordingly.

O9. TGF Beta Expression is Required for Enterobacter Sakazakii-Induced Epithelial Injury in a Mouse Model of Necrotizing Enterocolitis

Introduction: Necrotizing enterocolitis (NEC) is the leading cause of gastrointestinal emergency in neonates. Bacterial translocation through the injured mucosal barrier plays a detrimental role in the pathogenesis of NEC. Enterobacter sakazakii (ES) is a gram-negative opportunistic pathogen that has been implicated in clinical outbreaks of NEC. We have developed an animal model of NEC based on enteral ES feeding to newborn, breast-fed mouse pups. We have discovered that ES infection increases levels of pro-inflammatory cytokines as well as TGFβ in the serum, and they remain elevated until the animal succumbs to the disease. Bacterial virulence factor outer membrane protein A (OmpA), ubiquitous to all gram-negative bacteria, is necessary for the pathogenesis of ES induced injury. Based on our preliminary findings, we hypothesize that intestinal epithelial injury caused by ES is mediated by up regulation of mucosal TGFβ.

Methods: C57Bl6 3-day old mouse pups were infected with 10^3 CFUs of OmpA + /- ES. We used a novel technology for creating knock down animals using TGFβ siRNA in vivo. Wild type animals received 4 doses of TGFβ siRNA by intra- peritoneal injections. The suppression of TGFβ expression in enterocytes was confirmed in animals using flow cytometry and PCR. Tissue cytokine level was measured from fresh homogenized intestines using ELISA. TGFβ expression in CaCo2 cells was knocked down (KD) using siRNA. Apoptosis was measured in these cells post-infection using Annexin/7 AAD staining by Image Stream analysis. Tight junction function and barrier permeability was measured in transfected CaCo2 monolayers seeded on transwells. Monolayers were infected with OmpA + /- ES for 4 hours. Trans-epithelial electrical resistance (TEER) was measured at one-hour time intervals after infection using a voltmeter. Barrier permeability was also measured using macromolecule Horse Radish peroxidase (HRP). Results were confirmed in primary mouse intestinal epithelial cells. Results: TGFβ levels are significantly increased in the intestine of the animals infected with OmpA + ES. (Figure 1A, p < 0.0001) Knock down of TGFβ expression in vivo protects against ES induced NEC. The TGFβ KD animals survive the infection and show normal intestinal histology (Figure 1B-D). Intestinal epithelial cells in KD animals show no apoptosis compared to positive controls (Figure 1E). TGFβ KD in vitro prevents OmpA + ES-induced apoptosis. Monolayer integrity and permeability remain intact despite infection with OmpA + ES in the absence of TGFβ expression.

OPEN IN VIEWER

Conclusion: ES infection results in significant up-regulation of TGFβ expression in the intestinal epithelium. OmpA expression in ES is necessary for up-regulation TGFβ. KD of TGFβ expression protects the animals from ES-induced NEC by preventing tight junction dysfunction and apoptosis. Blocking TGFβ activity may serve as a potential therapeutic target to prevent ES-induced NEC.

O10. Toll Like Receptor 2 Signaling Prevents the Development of Necrotizing Enterocolitis through Negative Regulation of the Innate Immune Receptor TLR4

Introduction: Necrotizing enterocolitis (NEC) is a devastating intestinal disorder characterized by an uncontrolled and exuberant inflammatory response to bacterial colonization of the gastrointestinal tract of the premature infant. Despite a longstanding appreciation that bacterial-enterocyte signaling is required for NEC development, the mechanisms by which enteric microbes lead to NEC, and the factors that control the degree of bacterial-mediated signaling and inflammation in the newborn intestine, remain largely unknown.

Background: Our lab has recently shown that activation of the innate immune receptor Toll-like receptor 4 (TLR4) on the intestinal epithelium by gram-negative bacterial lipopolysaccharide (LPS) leads to NEC through increased enterocyte apoptosis and reduced mucosal healing. Surprisingly, recent studies have shown that gram-positive bacteria can protect against NEC development, suggesting the intriguing possibility that the receptor for gram-positive bacteria, ie Toll-like receptor 2 (TLR2), may exert anti-inflammatory properties in the gut, potentially through inhibition of TLR4. We now hypothesize that TLR2 signaling protects against NEC development through inhibitory effects on TLR4 signaling or expression.

Methods: NEC was induced in newborn wild-type, TLR2- and TLR4-deficient mice through a combination of formula gavage and hypoxia, and the extent of NEC development was assessed by evaluation of weight loss, mucosal cytokines, ileal histology and enterocyte apoptosis. To activate TLR2 and TLR4 in vivo or in vitro, mice or cultured IEC-6 enterocytes were treated with the specific ligands Pam3CSK4 or LPS respectively, and the degree of NFkB activation was assessed by immuno-confocal microscopy. To evaluate potential interactions between TLR2 and TLR4, IEC-6-enterocytes were treated with lentiviral particles containing short hairpin TLR4 RNA for 48 hours, which completely and stably restricted TLR4 expression, and the extent of TLR2 expression was determined.

Results: Compared to wild-type mice, TLR4 mutant mice had significantly reduced NEC severity, confirming the importance of TLR4 in NEC pathogenesis from our prior work. Strikingly, TLR2 deficient mice showed significantly increased NEC severity compared to wild-type mice, suggesting increased TLR4 signaling after TLR2 inhibition. In support of this possibility, injection of LPS resulted in significantly increased enterocyte apoptosis in the ilea in TLR2-deficient mice compared with wild-type counterparts, while colonization of the intestinal tract resulted in increased inflammatory signaling at baseline in TLR2-deficient mice compared to controls. In seeking to determine the pathways involved, the expression of TLR4 was significantly increased in the intestinal mucosa in TLR2-deficient mice. Strikingly, TLR4-deficient enterocytes and mice were found to have a significant increase in the expression of TLR2 and increased TLR2 signaling, confirming the reciprocal inhibitory relationship between these two innate immune receptors.

Conclusion: These data provide evidence for a novel and unexpected reciprocal relationship between TLR2 and TLR4, and demonstrate that TLR2 exerts a restraining influence on the extent of TLR4 signaling in the newborn intestine. These findings provide insight into the molecular events leading to NEC, and identify TLR2 as a possible therapeutic target in this devastating disease.

O11. Insulin Restores Apoptosis of Trauma Neutrophils (PMN) by Suppressing the Nicotinamide Phosphoribosyl Transferase (NAMPT) Activity of Pre-B Cell Colony-Enhancing Factor (PBEF)

Introduction: Hyperglycemia in association with insulin resistance is a prominent feature of acute critical illness. The role of insulin therapy remains controversial. PBEF is a highly conserved 52 kDa protein having pleiotropic activities as an inflammatory cytokine, an enzyme in a salvage pathway of NAD biosynthesis (Nampt), and a ligand for the insulin receptor (IR). We have previously shown PBEF expression to be increased in neutrophils (PMN) from septic patients, and necessary for the inflammation-induced inhibition of apoptosis. We sought to clarify the contribution of the various biologic activities of PBEF to delayed PMN apoptosis in trauma.

Methods: Circulating PMN from healthy volunteers or trauma patients (ISS > 16) were isolated by density gradient centrifugation. PMN-like HL-60 cells were used in some studies. Apoptosis was quantified as the nuclear uptake of propidium iodide in permeabilized cells. Wild-type myc-tagged PBEF or mutant PBEF in which dimerization and Nampt activity was blocked through mutation of S199 to A199 was expressed in E. coli. NAD was measured colorimetrically using thiazolyl blue and quantified in pmol/L against a standard curve.

Results: Rates of constitutive apoptosis were reduced in circulating PMN from trauma patients (23.5 ± 5.9 vs. 54.2 ± 7.5%, N = 6; p < 0.001). Transfection of HL-60 cells with wild-type, but not S199A mutant PBEF enhanced NAD generation in HL-60 cells (31.7 ± 8.5 vs 14.3 ± 3.8 pmol, p < 0.05). Rates of PMN apoptosis in healthy volunteers were 54.2 ± 7.5% at 24 hours, and were not altered by in vitro exposure to insulin (20 μU/ml). Incubation of trauma patient cells with insulin restored rates of apoptosis to control levels, an effect that could be overcome by exposure to wild-type, but not S199A mutant PBEF (Table). Co-immunoprecipitation studies showed a physical interaction between wild-type PBEF and the β chain of the IR in both PMN and HL-60 cells, however this interaction that was not seen with mutant S199A PBEF.

Conclusion: The Nampt activity of PBEF is necessary for its anti-apoptotic effects in PMN, and appears to be contingent on interactions between PBEF and the IR. Exogenous insulin competitively inhibits the Nampt activity of PBEF in trauma PMN. Since PMN contribute to tissue injury in trauma and sepsis, therapies targeting the PBEF:IR axis may attenuate the severity of PMN-mediated tissue injury.

Introduction: Our goal is to develop techniques for rapid, non-invasive, inexpensive detection of ventilator-associated pneumonia and its pathogens. Exhaled breath is known to contain bacterial and viral particles which correlate with pulmonary infection in animal models. In mechanically-ventilated patients the in-line disposable hygroscopic condenser humidifier (HCH) filter collects exhaled aerosols, potentially containing organisms indicative of evolving pneumonia. We have refined and applied rapid, inexpensive genetic assays (real-time polymerase chain reaction, PCR) to identify and quantify bacteria in these filters (Figure). However, the extent to which HCH filter pathogens correlate with clinically relevant infections has not been evaluated.

OPEN IN VIEWER

Hypothesis: Pathogens in ventilator filters, identified and quantified using PCR, correspond to quantitative broncho-alveolar lavage (BAL) cultures.

Methods: A pilot study was conducted, following controlled laboratory experiments to establish bacteriologic properties of HCH filters and calibrate counts. Seventeen filters from 14 surgical ICU patients were obtained. Organisms were detected and quantified in these filters using PCR, and correlated with BAL performed for clinical indications.

Results: Control experiments: Bacteria quantities remain stable in HCH filters, and counts and identity of pathogens can be reliably determined via PCR. Pilot study: Quantitative PCR confirmed absence of pneumonia in all cases (see table, italics indicate quantitative results > 104) and identified predominant organism in 4/5 patients with pneumonia. Overall, observed correlations were highly significant (P < .0025, exact binomial test).

Conclusions: This pilot study illustrates: 1) Pathogenic pulmonary bacteria are present in ventilator HCH filters; 2) These bacteria can be identified and quantified using PCR much faster than with BAL cultures; and 3) PCR findings correlate with established clinical standards for pneumonia diagnosis. Ongoing improvements to our techniques, and larger clinical trials currently underway, may enable rapid, inexpensive, and non-invasive diagnosis of pneumonia.

O13. Post-Operative MRSA Conversion and Infection in MRSA-Negative Patients: Associated Factors

Background: Hospital-acquired methicillin-resistant Staphylococcus aureus (MRSA) is associated with morbid, invasive infections in hospitalized patients, and has been implicated in nearly every type of nosocomial infection. We aimed at analyzing the risk factors for patient conversion from MRSA-negative pre-operatively to MRSA-positive post-operatively.

Methods: We retrospectively reviewed all patients at the VA-Boston HCS who underwent clean or clean-contaminated surgical procedures between the years of 2008 and 2009 and also had documented pre-operative nasal PCR testing for MRSA. We abstracted post-operative MRSA microbiologic testing results, MRSA infections, surgical site infections (SSI), surgical prophylaxis data, and SSI risk index as calculated using the VA Surgical Quality Improvement Project (VASQIP) database variables. All patients who had a negative nasal MRSA PCR in the 31 day pre-operative period and did not have any positive MRSA clinical cultures in the one year pre-operative period were defined as MRSA-negative. These patients were classified as converters to MRSA-positive if they had at least one documented positive nasal MRSA PCR swab, clinical culture, nosocomial infection, or SSI within 31 days post-operatively.

Results: Among 4,238 eligible patients, 1,550 (36.6%) qualified as MRSA-negative pre-operatively. Eighty-three of these patients (5.4%) converted to MRSA-positive, and 1,467 (94.6%) remained MRSA-negative post-operatively. On multivariate logistic regression analysis, chlorhexidine-alcohol skin preparation was the only protective variable (OR 0.514; 95% CI 0.297;0.889) against conversion. Non-significant risk factors for conversion were Tier 1 surgery (or surgery with a high likelihood for inclusion in a targeted decolonization program) (OR 1.158; 95% CI 0.708;1.894), and number of days elapsed between nasal sampling for MRSA and operation (OR 1.001; 95% CI 0.958;1.045). Vancomycin surgical prophylaxis (OR 6.919; 95% CI 4.103;11.666), SSI risk index (OR 2.200; 95% CI 1.312;3.688), and age at time of surgery (OR 1.032; 95% CI 1.011;1.053) were significantly associated with conversion to MRSA-positive post-operatively.

Conclusion: In MRSA-negative pre-operative patients, vancomycin prophylaxis, SSI risk index, and age were significant risk factors for conversion to MRSA-positive post-operatively. Chlorhexidine-alcohol skin preparation had a protective effect against conversion. Further evaluation of the modifiable risk factors for hospital acquisition of MRSA in the post-operative period is warranted.

O14. Calcium/Calmodulin-Dependent Kinases (CAMK) Regulate Organ Dysfunction Durine Murine Endotoxemia

Introduction: Autophagy, an ancient cytoplasmic homeostasis process regulating cellular biomass quantity, quality, and distribution, endows cells with the ability to sacrifice and reuse portions of their cytoplasm to support vital functions during periods of stress. Recent data support a fundamental role in innate immunity as an anti-microbial effector of Toll-like receptors. The calcium/calmodulin-dependent protein kinase cascade (CaMK), a family of serine/threonine kinases responsive to intracellular Ca²⁺ concentration [Ca²⁺]_i, is also integral to the immune response. In light of the calcium dependency of autophagy, we hypothesized that the CaMK regulate autophagy during sepsis.

Methods: Mice were subjected to LPS (5mg/kg, IP). CaMK activity was inhibited using the broad CaMK inhibitor KN93 (12mg/kg, IP) or its less functional analogue KN92 (12mg/kg, IP) administered 1 hour prior to experimentation. We also utilized mice genetically deficient in the expression of CaMKIV, one of two downstream effector kinases in the cascade. After LPS, blood was collected by cardiac puncture and organs were harvested. Autophagy of total cellular protein was assessed by immunoblot, using antibodies targeting LC3b. Serum cytokine concentration of IL-6, IL-10 and TNFα were analyzed by ELISA. We assessed renal function by assaying serum blood urea nitrogen (BUN) or the more GFR-specific cystatin.

Results: Endotoxemia induced marked elevations in systemic cytokine concentrations, of which IL-6 was inhibited by KN93 (2366 vs. 748, p = 0.12). Autophagy was induced in the kidney as evidenced by the induction of LC3b, which was also attenuated by KN93. The attenuation in inflammation and autophagy with KN93 correlated with reduced renal injury as evidenced by reduced BUN (95 vs. 72, p = 0.12) and cystatin (891 vs. 540, p = 0.12). To explore the specific CaMK mediating these events, we utilized CaMKIV-/- mice. Similar to KN93, CaMKIV-/- mice exhibited reduced IL-6 (3395 vs, 1664, p = 0.08) but also elevated IL-10 concentrations (245 vs. 383, p = 0.08). They demonstrated reduced autophagy in the kidney, but worse renal injury as evidenced by elevated BUN (80 vs. 96, p = 0.08) and cystatin (773 vs. 1014).

Conclusions: These results suggest that the CaMK, and specifically CaMKIV, regulate inflammation and autophagy during endotoxemia. However, though broadly inhibiting the entire family reduced renal dysfunction, specific inhibition of one downstream CaMK species, CaMKIV, worsened the degree of renal dysfunction. Further studies are needed to determine the regulatory roles of other CaMK species (i.e. CaMKK, CaMKI) in the autophagic and inflammatory response underlying the renal dysfunction of sepsis.

O15. Modulation of the Mucus Component of Intestinal Barrier Defense in Vitro

Introduction: An increased incidence of gut-related as well as other nosocomial infections such as pneumonia in ICU patients have been associated with the use of Histamine2 (H2)receptor antagonists including cimetidine. Important factors in gut barrier function include the intestinal mucus layer and secretory IgA (SIgA) related immune protection. The effect of H2 blockers on the collaboration between the intestinal mucus layer and SIgA in gut barrier function is unknown. We compared the barrier protection of mucus and mucin producing intestinal epithelial cells (IEC) as well as the effect of cimetidine and its effect on mucus production in an in vitro model.

Methods: HT-29 MTX (a mucus producing IEC) monolayers were established in a two-chamber cell culture system. Dimeric IgA was added to the basal media of IEC monolayers to allow binding of IgA to the polyimmunoglobulin receptor on the basal side of IEC followed by uptake and delivery of SIgA to the IEC apical surface (transcytosis). Cimetidine was added to the media of a subset of the HT-29 MTX intestinal cells and E. coli (EC) to the apical media of HT-29 MTX cell monolayers. Either fluorescein-labelled E. coli (FITC-EC) or unlabelled EC were added to subsets to quantify bacterial adherence (60 minute co-culture) and bacterial passage (240 minutes) thru IEC monolayer's respectively. Mucus content was indexed by O-linked oligosaccharide chain (OSC) content and Western blot analysis.

Results: Mean ± S.D., N = 3 for all groups.

Conclusion: Bacterial trapping in the intestinal mucus layer enhances SIgA interaction with luminal bacteria and contributes to intestinal barrier integrity. Loss of the intestinal mucus layer associated with cimetidine administration impairs barrier function in this model. Use of cimetidine may contribute to systemic inflammation and remote organ dysfunction in at risk patients especially with subsequent insults to gut homeostasis.

Introduction: Pseudomonas aeruginosa is an opportunistic, gram-negative pathogen associated with many hospital-acquired infections and disease states. In particular, P. aeruginosa has been identified as a critical factor in the pathogenesis of neonatal necrotizing enterocolitis (NEC). NEC presents more frequently in infants fed a formula-based diet – a finding that may be based on the monosaccharide content of this diet class. Thus, we hypothesized that P. aeruginosa would differentially express virulence genes when exposed to monosaccharides present in formula versus those present in human milk.

Methods: Using the results of a metabolomics study on infant diets and their resulting fecal samples, we identified several monosaccharides that distinguished milk from formula diets. Of these diet-distinguishing monosaccharides, four were found to be metabolized by P. aeruginosa. We subsequently grew P. aeruginosa in tryptic soy broth (TSB) supplemented with these four monosaccharides and used qRT-PCR to measure the expression of 59 major P. aeruginosa virulence genes. Results were standardized to an external control of P. aeruginosa grown in TSB alone.

Results: P. aeruginosa did not differentially respond to monosaccharides after six hours of growth. After 24 hours, however, P. aeruginosa grown in arabinose (present in formula), xylose (present in human milk), and galactose (present in feces from milk-fed infants) displayed a significant increase in virulence gene expression in all categories of genes tested. In contrast, P. aeruginosa grown in mannose (also present in the feces from milk-fed infants) displayed a significant decrease in virulence gene expression when compared to the other experimental conditions.

Figure: Numerical values represent the log of the fold-difference in gene expression between the experimental group and the external control. The table was conditionally formatted so the intensity of red correlates with the magnitude of gene up-regulation and the intensity of green correlates with the magnitude of gene down-regulation.

OPEN IN VIEWER

Conclusion: These preliminary studies demonstrate the importance of nutrient content on the relative expression of virulence genes in pathogens that commonly colonize the gut of infants. Understanding the effect of current dietary formulas on virulence gene expression across such gut-colonizing pathogens may present a new approach to elucidate the differences between human milk and formula in the development of NEC.

O17. Small Postoperative Changes in Serum Creatinine are Associated with Postoperative Sepsis and Surgical Infections

Background: Postoperative mild and moderate AKI, characterized by small changes in serum creatinine (sCr), is common and is associated with increased short and long-term mortality. Small changes in sCr may be predictive of postoperative sepsis and surgical infections.

Methods: A retrospective, single center study of 40533 adult surgical patients admitted to a tertiary academic center from 2000–2008 for ≥2 days. Maximum change in sCr from baseline sCr (BsCr) was calculated for each patient. AKI was defined using modified RIFLE classification as postoperative change in sCr ≥ 0.3 mg/dl. Area under the curve (AUC) was used to determine the predictive ability of a maximum change in sCr for postoperative sepsis and surgical infections. The level of change from BsCr that best predicts outcome was determined by maximizing sensitivity and specificity along the ROC curve.

Results: The prevalence of postoperative sepsis and surgical infections was 5% and 5.5%, respectively. Patients with AKI were more likely to develop both sepsis (OR 20, 95% CI 15-25) and SI (OR 2.2, 95% CI 2.1-2.5) compared to patients without AKI. Multivariate logistic analysis revealed that age, maximum change in sCr, number of comorbidities, baseline GFR, Medicaid insurance status, weekend admission and type of surgical service were independently associated with the occurrence of sepsis (c statistic 0.622). Maximum change in sCr from baseline in the first seven postoperative days was predictive of sepsis (AUC range 0.84). ROC analysis indicated that a change of 0.3 mg/dl or greater from BsCr is predictive of sepsis (sensitivity range 0.8–0.9, specificity range 0.76–0.85).

Conclusion: Postoperative sCr change as low as 0.3 mg/dl is predictive of postoperative sepsis.

OPEN IN VIEWER

Figure 1.

Change in sCr in the first seven days and sepsis.

Background: Strict glucose control with intensive insulin therapy has been reported to reduce mortality and morbidity in critically ill adults. The results of these trials lead to widespread support and implementation of intensive insulin therapy protocols. However, considerable controversy has developed over the benefit of tight glucose control and the subsequent risk of hypoglycemia.

Methods: Retrospective review of prospectively collected data on all patients admitted to a Surgical Intensive Care Unit at a large, urban, level 1 trauma center from January 2005 until August 2009. Inclusive criteria consisted of an ICU admission, with hospital length of stay of at least 48-hours, and a minimum of 8-Point of Care (POC) glucose readings. Data were collected via the Clinical Database Warehouse and Trauma Registry (TRACS). Each subjects glucose values were averaged over the length of stay. Each glucose value was designated as either hypoglycemic (<80 mg/dL), or not hypoglycemic (>80 mg/dL), and percent hypoglycemia was calculated as number of hypoglycemic events / total glucose readings. Hospital-acquired infections (HAI), including urinary tract infection, blood stream infections, and ventilator-associated pneumonia were assigned by an independent Infection Control Committee.

Results: Of the 1,561 patients included, 1,039 (66.6%) were admitted with a trauma diagnosis, and 873 (56%) had at least one hypoglycemic event. These hypoglycemic patients were 5-times more likely to develop a HAI when compared to patients with no hypoglycemic events (18.0% v. 3.4%, p < 0.01). Additionally, they were older (53.7 years v. 49.9 years, p < 0.01), more injured (19.5 average ISS v 16.7 average ISS, p < 0.01), and with a higher co-morbidity of diabetes (12.5% v 7.1%, p < 0.01). Multivariate analysis demonstrated that hypoglycemia was and independent risk factor for the development of HAI (see table). HAIs were associated with a significant increase in ICU LOS (21.4 days v. 6.3 days, p < 0.01) and hospital LOS (31.1 days v 14.1 days, p < 0.01).

Conclusion: We demonstrate that in the critically ill surgical patient, any degree of hypoglycemia is an independent risk factor for the development of hospital acquired infections. Furthermore, HAI leads to a significant increase in both hospital and ICU length of stay.

Introduction: Appropriate, early antibiotic therapy is associated with improved outcomes, but the optimum timing for initiation of antibiotics is unknown. Two strategies exist: commencement of antimicrobials as soon as infection is suspected or initiation of treatment only after objective data (i.e., culture, x-ray, etc) confirms infection. We hypothesized that a more aggressive approach would improve patient mortality without affecting ICU-wide infection rates or ICU flora.

Methods: This two-year study included all patients admitted from 9/08 to 8/10 to the surgical ICU at a single tertiary care hospital. During year one, aggressive treatment was employed, where patients suspected of having an infection based on clinical grounds (fever, change in white blood cell count, or other clinical change consistent with new infection) had blood cultures sent and antibiotics started (Aggressive period). During the second year, a conservative strategy was used, with antimicrobials started only after objective findings confirmed infection, i.e., positive gram stain/culture, radiologic study, etc (Conservative period). Univariate (chi square and t-test) and multivariate analysis (logistic regression) were used.

Results: Admissions for the first and second year were 762 and 721, respectively. 247 ICU-acquired infections occurred during the aggressive year and 237 during the conservative year (Table). For ICU-acquired infections, the initiation of antibiotics was significantly delayed in the conservative year (34.8 ± 2.4 vs 20.9 ± 1.8 hours from culture to initiation of antibiotics, p < 0.0001, and 28.1 ± 2.4 vs 9.9 ± 1.1 hours from fever to initiation of antibiotics, p < 0.001); however, duration of antibiotics was shorter (12.5 ± 0.7 vs 17.7 ± 1.8 days, p = 0.008). Resistant Gram positive cocci (MRSA and VRE) were isolated more frequently in the aggressive period than the conservative period (42 versus 24 infections, respectively). Rates of initial adequate empiric antimicrobial use were similar (67.4% and 70.4% for aggressive and conservative, respectively). All cause mortality on a per infection basis was lower during the conservative period for all infections, pneumonia, and abdominal infections. 27/101 (26.7%) patients with ICU-acquired infections died during the aggressive period while 13/100 (13.0%) died during the conservative period (p = 0.015 for the difference). After adjusting for age, gender, location of acquired infection (home, ICU or ward), involvement in trauma, APACHE II score and site of infection, conservative treatment resulted in a 76% reduction in mortality risk (OR = 0.244, 95%CI (0.089 – 0.667)).

Conclusion: Waiting for objective data to diagnose infection prior to the initiation of antimicrobials does not worsen mortality. In fact, a conservative approach to antimicrobial initiation may lead to better outcomes and minimize antimicrobial use in uninfected patients.

O20. Methicillin-Resistant Staphlococcus Aureus as a Causative Pathogen in Early Ventilator-Associated Pneumonia: Cause for Concern?

Introduction: Staphylococcus aureus (SA) remains a common cause of ventilator-associated pneumonia (VAP) in trauma patients. The prevalence of community acquired methicillin-resistant Staphylococcus aureus (MRSA) has increased over the last decade. The purpose of this study was to evaluate the incidence and trend of MRSA VAP over a 6-year period and to examine the adequacy of our current empiric antibiotic regimen.

Methods: All patients with VAP (≥105 CFU/mL in BAL effluent) over a 6-year period were identified from the VAP database maintained in our Level I trauma center. All patients received empiric therapy based on duration of ICU stay (ampicillin/sulbactam (<7 days) or cefepime plus vancomycin (>7 days) or equivalent in penicillin allergic patients). Patients were stratified by gender, age, severity of shock (24-hour transfusions) and injury severity (GCS and ISS). Outcomes (ventilator days, ICU days, hospital length of stay (LOS) and mortality) were compared between patients with early MRSA VAP and those with all other causes of early VAP. Linear regression was used to determine differences in incidence over the six year time period. Multi-variable logistic regression was then performed to determine whether early MRSA was an independent predictor of mortality in patients with early VAP.

Results: 727 patients with 997 VAP episodes (1.4 episodes per patient) were identified. 512 episodes were early VAP of which 63 were MRSA. The incidence of SA in early VAP, MRSA by all early SA VAP, and MRSA by early VAP are shown in the table. Linear regression shows that the incidence of early SA VAP (slope = − 0.911, p = 0.490) is stable, but the incidence of MRSA in all early SA VAP (slope = 3.95, p = 0.0154) and the incidence of MRSA in early VAP are increasing (slope = 1.35, p = 0.0028). There was no statistical difference in gender, age, 24-hour transfusion requirements, ISS, or GCS in those patients that got early MRSA compared to other organisms. There was no difference in mortality (17.5 vs.18.7%, p = 0.81), ventilator days (18.5 vs. 18.1, p = 0.89), ICU days (23.8 vs. 23.1, p = 0.80), or LOS (34.9 vs. 33.7, p = 0.76) comparing MRSA with other early VAP episodes. Multivariable logistic regression performed on all pts with early pneumonia failed to show early MRSA as an independent predictor of mortality (OR = 0.815, p = 0.59) after adjusting for age, ISS, and 24-hour transfusion requirements.

Conclusions: The incidence of early MRSA as a causative pathogen of VAP and in all SA early pneumonias has increased significantly in trauma patients. While MRSA is being seen more frequently within seven days of admission, outcomes and resource utilization from episodes of inappropriate empiric antibiotic treatment are unchanged from other early pneumonias. The current empiric antibiotic algorithms for VAP are adequate and no changes are warranted.

O21. A Prospective, Protocolized Study Evaluating the Reliability of Sputum Cultures Obtained after Administration of Antibiotics in Injured Patients

Introduction: Appropriate utilization of antibiotics in critically ill patients involves tailoring antibiotics to culture results; however, the culture results must be reliable. It is generally believed that administration of antibiotics before obtaining cultures can have a significant effect on culture results; however, this has never been prospectively evaluated in injured patients. We hypothesized that the antimicrobials would significantly reduce the reliability of cultures obtained between 1 and 24 hours after antibiotic administration.

Methods: Patients admitted to a Level I trauma center were prospectively evaluated. Patients were eligible for the study if they were mechanically ventilated and were suspected to have pneumonia. Informed consent and IRB approval were obtained. After enrollment, sputum cultures were obtained and broad spectrum antibiotics were started. Sputum cultures were then repeated at 1 hour, 6 hours, 12 hours, and 24 hours after completion of the first dose of antibiotics. Cultures were incubated in standard broth and were allowed to grow until a final report was generated by the microbiology laboratory (48-72 hours). Patients whose initial culture was positive were included in the analysis. The culture results before antibiotics were administered were compared with the culture results at each subsequent time point. Additionally, patient demographics and hospital course were noted. Categorical data were analyzed with chi square tests and continuous variables were analyzed with t-tests.

Results: There were 21 patients included in the study. The average patient age was 49.4 years. The average injury severity score was 27.7. The average ICU and hospital lengths of stay were 20.2 days and 24.7 days, respectively. All of the organisms that were grown from the pre-antibiotic cultures were also grown in the cultures obtained 1 hour after antibiotics were given. A significant number of these organisms were unable to be grown in subsequent cultures. The rate of negative cultures increased to 21%, 32%, and 42% among the 6 hour, 12 hour, and 24 hour groups (p < 0.01 for each group). Gram positive organisms accounted for 42.9% of infections, with S. aureus being the most common organism. There was a significantly reduced effect of antibiotics on subsequent S. aureus cultures. All cultures positive for S. aureus prior to antibiotic administration remained positive at each subsequent time point, despite the adequacy of initial antibiotic selection.

Conclusion: Antibiotics have a significant effect on culture results. This effect is most pronounced in gram negative organisms and is seen in cultures obtained more than 1 hour after antibiotics are given. As a result, cultures obtained more than one hour after antibiotics are administered cannot be used to tailor antibiotics in injured patients with suspected infections.

O22. Fresh Frozen Plasma Transfusion is an Independent Risk Factor for Bacteremia in Critically Ill Traumatic Brain Injury Patients

Introduction: Traumatic brain injury patients are at high risk for FFP transfusion and infection due to complex injury and prolonged ICU length of stay. To our knowledge, few studies however, have evaluated the impact of FFP transfusion on infection in traumatic brain injury (TBI) patients. The objective of this study was to evaluate the impact of FFP transfusion as an independent predictor of infection in critically injured TBI patients.

Methods: Prospective data were collected on 630 consecutive critically injured trauma patients diagnosed with traumatic brain injury (confirmed by CT scan) over a 4 year period. Data collected included demographic information including age, gender, race, mechanism of injury, Injury Severity Score (ISS) and head Abbreviated Injury Score (hAIS). The admission coagulation profile of each patient in the study was recorded in addition to the total number of units of FFP and packed red blood cells (pRBC) transfused in the 1st 24 hours of hospital admission. Clinical outcomes including mortality, ICU days, nosocomial infection, and hospital length of stay were also evaluated.

Results: Seventy-four percent of patients were male (N = 469) and the majority were Caucasian (N = 399). Mean age, ISS, and HAIS were 43 ± 21 years, 31 ± 12, and 4 ± 1, respectively. Of the 632 TBI patients, 252 (40%) patients received FFP transfusions in the 1st 24 hours of admission. Mean number of units of FFP transfused in the 1st 24 hours of admission were 2 ± 5 units (Range 0 - 64 units).The mean number of ICU days and hospital days were 13 ± 10 days and 16 ± 12 days. The overall mortality was 18.5% with an infection rate of 63%. The 3 most common infections were urinary tract infections (25%), pneumonia (22%), and bacteremia (19%). Patients who received FFP had a significant greater incidence of bacteremia (24%) compared to those who did not receive FFP (15%) in the 1st 24 hours (p = 0.005). Logistic regression analysis, controlling for age, gender, race, mechanism of injury ISS, HAIS, and baseline coagulation profile, show that FFP transfusion in the 1st 24 hours of admission was found to be a independent predictor of bacteremia (OR 1.88, p-value = 0.010). There was no significant difference in pneumonia and UTI rates between the two groups.

Conclusion: Fresh frozen plasma transfusion is an independent risk factor for bacteremia in patients with traumatic brain injury. Further research regarding the potential etiology of this finding in TBI patients is warranted.

O23. Systemic Endotoxemia following Multiple Trauma is Associated with Early Shock and Predicts Subsequent Mortality

Background: Multiple trauma is associated with innate immune activation and inflammatory morbidity, even in the absence of superimposed infection. The potential etiologic role of bacteria or bacterial endotoxin (LPS) translocating from the gut has been controversial. We undertook a serial evaluation of circulating LPS levels in a cohort of trauma victims.

Methods: We performed a prospective observational cohort study of patients sustaining severe multiple trauma (ISS > 16). We measured LPS levels on days 0, 1, 3, and 5 using the Endotoxin Activity (EA) Assay® - a whole blood assay based on LPS-primed neutrophil chemiluminescence that quantifies endotoxin on a scale from 0 to 1.0 activity (EA) units. EA levels were stratified as low (<0.40 EA units), intermediate (≥0.40-0.59 EA units), or high (≥0.60 EA units). We collected data on demographics, resuscitation, length of stay (LOS), and mortality. Statistical analyses used ANOVA, Fisher's exact or Student t tests, as appropriate, with an α level of p < 0.05.

Results: We recruited 22 patients having a mean age of 43 ± 23 years and mean ISS of 40 ± 15. Most (82%) were male, and 17/22 had sustained blunt trauma. We analyzed only those patients having a minimum of three EAA values (N = 20). Their median ICU LOS was 9 days (2–29) and their duration of mechanical ventilation, 9 days (1–27). Four patients died (20%) at 10 ± 1 days post admission. The mean initial (Day 0) EA value was low, but increased significantly over time (Table; ANOVA for trend p = 0.0013). Maximal EA values were reached at an average of 80 ± 40 hours; four (20%) patients recorded high EA levels, while in 12 (60%) the maximal levels were intermediate. EA levels exceeded 0.4 on at least 2 occasions in all 4 non-survivors, but in only 3/16 survivors (p = 0.007). Both maximal EA levels (0.65 ± 0.15 vs 0.45 ± 0.11; P = 0.009), and average EA across Day 0 to Day 5 (0.48 ± 0.06 vs 0.32 ± 0.10; P = 0.008) were higher in non-survivors than in survivors. The average EA level correlated significantly with the presence of shock at admission (r2 = 0.2; P = 0.04).

Conclusion: Endotoxemia is not present immediately following trauma, but develops over time during the early post-injury period. Its presence predicts adverse outcome. The association of endotoxemia with systemic hypoperfusion suggests that the GI tract is the source of circulating endotoxin in the victim of multiple trauma.

Introduction: While the negative impact of infectious complications (IC) after elective surgery during the index hospitalization is well established, the long-term ramifications of hospital acquired post-operative infections are not well studied. This analysis evaluated the impact of a hospital acquired IC after open abdominal vascular surgery on readmission rates as well as 30-day and 90-day mortality after discharge.

Methods: Data from all hospitals in the US that performed elective open abdominal vascular surgeries in Medicare population from 2005 to 2007 were extracted from the national MedPAR database. The cohort included all open abdominal vascular surgeries including aortic, iliac, and visceral procedures. Infectious complications evaluated included: pneumonia (PNA), urinary tract infection (UTI), postoperative sepsis (sepsis), surgical site infection (SSI), and Clostridium difficile (CDiff). For comparison, patients were categorized from their index procedure as either developing an IC during their initial hospitalization (Index + INF) or not developing an IC (No INF). Longitudinal evaluation of rates of 30-day readmission, 30-day infectious complications, and 30-day and 90-day mortality after the initial discharge was performed and these rates were compared between groups.

Results: 29,549 open abdominal vascular procedures were identified. 4,016 (13.6%) patients developed an IC during their index hospitalization: PNA (5.1%), UTI (2.7%), sepsis (1.6%), SSI (1.4%), and CDiff (0.6%). Additionally, 1.13% of patients developed PNA, UTI, SSI or CDiff complicated by sepsis. Hospital mortality during the initial hospitalization was 13.7% (Index + INF) vs. 4.0 (No INF), p < 0.0002. Infectious processes (PNA, UTI, SSI, and CDiff) which were complicated with the development of sepsis significantly increased in-hospital mortality compared to an IC alone (39.1% vs. 5.5%, p < 0.0001). Mortality 30-days and 90-days after the initial discharge from the index procedure was significantly higher for Index + INF compared to No INF (4.4% vs. 1.2% and 8.6% vs. 2.6%, respectively, p < 0.0002). The greatest 30-day mortality rates after discharge were found after CDiff + sepsis (30%), PNA + sepsis (12.6%), and postoperative sepsis alone (8.6%). The same rank was found for 90-day mortality: 30%, 22.5%, and 13.8%. Overall, readmission was higher for Index + INF compared to No INF (33.7% vs. 21.5%, p < 0.0002). Rates of 30-day readmission after index IC ranged between 32.0% and 50.0%.

Conclusion: For Medicare beneficiaries undergoing elective open abdominal vascular procedures, the development of any infectious complication significantly increased not only hospital mortality, but also the 30-day and 90-day mortality after discharge from the hospital. Index infectious complications were also associated with increased 30-day readmission after discharge. Hospital acquired infections have a profound late effect on outcomes after discharge, and future programs targeting high risk patients may improve long term survival and minimize readmissions.

O25. Gut Microflora is Significantly Altered After 72 Hours of Formula Feeding

Introduction: Infants exposed to formula feeding are at increased risk for intestinal diseases with a bacterial etiology (e.g. necrotizing enterocolitis). However, the impact of diet on infant gut bacterial colonization has not been fully elucidated. Formula feeding may alter the normal gut flora in such a way that confers this heightened vulnerability to disease. Here, we utilize 16S bacterial DNA pyrosequencing, a culture independent technique, to quantitatively analyze the colonic microbiota of mice fed formula as compared to mice fed maternal milk to test the hypothesis that diet impacts the relative abundance of bacterial species in the gut.

Methods: C3HeB/FeJ pups were born vaginally and remained housed with the mother until DOL7. On DOL7 half of each litter was assigned to formula feeding (liquid puppy formula via orogastric catheter 4x/day) while the remaining pups were allowed exposure to maternal feeding. All pups were euthanized at 72 h, and colonic tissue was collected for DNA extraction. After colonic microbial DNA was extracted, sequences encoding the 16 S ribosomal subunit were amplified and sequenced on the Roche GS-FLX pyrosequencing platform. Taxonomy was assigned to sequences using the Ribosomal Database Project classifier tool. Statistical tests for differentially abundant 16S-based taxonomic annotations between populations were performed using Metastats methodology to compute nonparametric p-values.

Results: After only 72 h of differential feeding, 16S pyrosequencing demonstrated that Firmicutes (p < 0.001) was the dominant phylum in the maternal fed pups while Proteobacteria (p < 0.001) was the dominant phylum in formula fed animals. At the genus level, colonic microflora of formula fed animals had significantly more Streptococcus (p < 0.001) than that of maternal fed animals. However, the colonic microflora of maternal fed animals had a greater abundance of Lactobacillus (p < 0.001) and Enterococcus (p < 0.001) than that of formula fed animals.

Conclusion: Overall, 72 h of differential feeding resulted in marked alterations in the relative abundance of bacterial species in the gut. Firmicutes was the dominant phylum in maternal fed animals, while Proteobacteria was the dominant phylum in formula fed animals. Interestingly, while Lactobacillus and Enterococcus were more abundant in animals fed maternal milk, the gut microflora of formula fed animals was dominated by Streptococcus. Presumably such alterations in gut bacterial community structure could be responsible for differential disease development in infants fed formula as compared to breast milk. The finding that these changes were observed after only 72 h of differential feeding demonstrates the plasticity of the gut in that gut microflora can undergo relatively rapid repopulation based upon environmental conditions. Future work will evaluate changes in host and microbial gene expression in response to differential feeding.

O26. Parallel Microrna-155 and TNF-&#945; Production could Explain Microbial Tolerance in a Novel Peritonitis Model

Introduction: We observed persistent peritoneal bacteria despite a transient early innate immune response to intra-peritoneal (IP) Klebsiella pneumoniae. Pretreatment with lipopolysaccharide (LPS) prior to peritonitis induced a tolerant pattern of pro-inflammatory cytokine protein production but not at the messenger RNA (mRNA) level. MicroRNAs (miRNA)s regulate inflammatory cytokines and may explain this paradox.

Methods: 24 hours after pretreatment with IP LPS (0.01 mg/g) or saline, C57BL/6 mice were given 103 colony forming units of K. pneumoniae IP. Gentamicin was given 4 hours prior to infection and BID thereafter. Peritoneal exudate cells (PEC)s were obtained through peritoneal lavage and total RNA was isolated (n = 3) at 4, 24, and 48 hours following infection. TNF-α levels were detected using ELISA. MicroRNA (miRNA) expression levels were detected using SA Biosciences mouse immuno-pathology RT2 miRNA PCR array (MAM-104A). mRNA detection was done using SA Biosciences RT2 Profiler PCR array mouse Toll-like receptor signaling pathway (PAMM_018A). Data were analyzed by Ingenuity Pathway Inc. analysis (IPA). Student's t-tests were used to determine significance of fold changes.

Results: TNF-α production increased at 4hours and was significantly decreased at 24 and 48 hours in the LPS as compared to the PBS pre-treated group (p < 0.05). In concordance, miR-155 expression at 48 hours was greater in the PBS and LPS pre-treated groups (p < 0.05). There was no significant difference in TNF-α mRNA expression between groups at 4, 24, and 48 hours suggesting no tolerance at the mRNA level (p > 0.05) IPA analysis of the mRNA data showed that 6 main canonical pathways were constantly dysregulated, and in the same significance order, in both the saline and LPS group at 4, 24, and 48 hours. These are: Toll-like receptor and NF-KB signaling, hepatic cholestasis, interleukin-6 and LPS mediated MAPK signaling pathways, and pattern recognition receptors of bacteria pathway.

Conclusion: Peritonitis increased PEC gene expression associated with sepsis and a pro-inflammatory response, which was further augmented by LPS pretreatment over 24 hours only. Prior exposure to LPS did not induce PEC gene tolerance to subsequent infection with Klebsiella at the mRNA level. Up-regulated pathways were common to the early innate response, and macrophage reprogramming occurs at the genetic level. Lack of bacterial clearance in this model is due to an inadequate innate response which resolves despite incomplete clearance of intra-peritoneal bacteria. miR-155 production parallels TNF-α and not TNF mRNA production and suggests a mechanism for tolerance in this model and persistent peritoneal infection.

Introduction: Intestinal inflammatory disorders such as IBD or necrotizing enterocolitis may result from the failure to down regulate innate immune responses in the gut. Mitogen-Activated Kinase Phosphatase, MKP-1, is a critical regulator of epithelial innate immunity. To gain insight into MKP-1-dependent regulation, we sought to elucidate the mechanisms of negative regulation of MKP-1 by proteolysis.

Methods: ERK phosphorylation sites of MKP-1 were obliterated using site-directed mutagenesis. Wild type and mutant MKP-1 were ectopically expressed in human embryonic kidney HEK293 or rat intestinal epithelial IEC-6 cells. Half-life of MKP-1 proteins was determined using cycloheximide chase assay. ERK activity was blocked using the specific inhibitor, U0126.

Results: Half-life of wild type MKP-1 protein was about 30 min. Complete deletion of the C-terminal non-catalytic domain extended the half-life to over 4 h. Obliteration of ERK phosphorylation sites within the C-terminal non-catalytic domain, or pharmacologic inhibition of ERK extended the half-life of MKP-1 to only about 1 h.

Conclusion: The C-terminal non-catalytic domain of MKP-1 serves as a rapid degradation signal. ERK-dependent phosphorylation within this domain plays only a minor role in targeting MKP-1 for proteolytic degradation. Rapid degradation of MKP-1 is important for the dynamic regulation of the innate immune responses by this phosphatase.

OPEN IN VIEWER

O28. Calcium/Calmodulin-Dependent Protein Kinaseiα Mediates LPS-Induced Autophagy in Macrophages VIA Camkiα, AMPK and APG7 Complex

Introduction: Autophagy, originally described as a stress response to nutrient deprivation, now is emerging as the chief rout for bulk degradation of aberrant organelle, protein aggregates, and invading foreign materials to keep cells homeostasis. It has been shown to be an important effecter arm of innate immunity. CaMKI is an ubiquitously expressed cytoplasmic enzyme that mediates calcium signaling in macrophages (Mϕ) responding to LPS. In light of the calcium dependency of autophagy, we hypothesized that CaMKI positively regulates autophagy in Mϕ responding to LPS.

Methods: In vitro: RAW 264.7 cells and murine peritoneal Mϕ were treated with LPS (100 ng/ml). We utilized the following inhibitors: KN62, Compound-C (AMPK inhibitor), AICAR (AMPK activator), and Rapamycin (mTOR inhibitor). We utilized a constitutively active CaMKI plasmid (CaMKI.293), its kinase-inactive mutant (CaMKI.293K49A), and siRNA CaMKIα to manipulate CaMKIα expression. Autophagy was assessed by immunoblot, immunfluorescent and electron microscopy. CaMKIα activity was measured by immunoblot. Cytokines IL-1ɑ, IL-6, IL-10 and TNF-α in medium released from RAW 264.7 were analyzed with ELISA after the RAW 264.7 was treated with LPS for 16 hours and siRNA APG5, APG7 and LC3B transfection. In vivo: Mice were subjected to cecal ligation and puncture or sham operations after 72 hours with RNAi of CaMKIα, LC3B in tissues were analyzed with immunoblot.

Results: LPS induced autophagy as evidenced by increased APG7, LC3B and APG5-12 conjugation, that was inhibited with CaMKIα siRNA, KN62, and Compound-C. LPS also induced p-AMPKα, p-mTOR, p-p70S6K, and p-S6, all of which were inhibited withsiRNA CaMKIα. Over expression of CaMKI transfection with CaMKI.293 induced APG7, LC3B, APG5-12, p-AMPKT, p-mTOR, p-p70S6K, and p-S6 to a greater extent than it's less active CaMKI-293K49A. The AMPK inhibitor, compound C, inhibited autophagy consequent to either LPS or CaMKI-293. There are complex between CaMKI, AMPKa and APG7. CaMK1 and AMPK complex were not influenced by APG7 siRNA, but there is no immunoprecipitation complex between APG7 and AMPKa after CaMKIα siRNA or compound-C pretreatment. Autophagy genes LC3B, APG5, and APG7 knockdown have different effects on cytokine releasing from RAW 264.7 after LPS treatment, they all reduced IL-1α release, siRNA LC3B increased IL-6, TNF-α release. In vivo, siRNA CaMKI inhibited expression of autophagy protein LC3B in mice subjected to CLP.

Conclusions: These results suggest that CaMKIα regulates autophagy in Mϕ responding to LPS, which depends upon complex between CaMKIα, AMPK and APG7. CaMKIα also directly regulates the mTOR (TORC1) pathway, suggesting that its regulation of autophagy occur independent of mTOR inhibition. Autophagy play a very important role in cytokine release in macrophages responding LPS. These events appear to be operant in vivo.

O29. The Importance of HEME-Oxygenase 1 in Mitochondrial Biogenesis in Sepsis

Introduction: In sepsis there are various cellular pathways that are induced to prevent permanent organ injury and allow for the restoration of cellular homeostasis and metabolism. Central to these adaptive responses is the dynamic regulation of mitochondria to prevent bioenergetic failure. Damaged mitochondria must be removed and replaced with viable mitochondria to produce ATP and maintain basic cellular activities. Heme oxygenase-1 (HO-1) and its catalytic product carbon monoxide are known to be essential signaling molecules to limit inflammation, regulate cellular bioenergetics, and restore cellular homeostasis. Our objective was to test the hypothesis that mitochondrial biogenesis is necessary to prevent organ injury from sepsis, furthermore, to demonstrate that these processes are regulated by HO-1.

Methods: Primary mouse hepatocytes (MHC) with and without tin protophoryin treatment (SnPP) (50 uM), a non-specific inhibitor of heme-oxygenase were stimulated with lipopolysaccharide (LPS; 100 ng/mL) at various time points and mitochondrial function was examined by measuring the oxygen consumption rate, extracellular acidification rate, or mitochondrial membrane potential with JC-1 staining. ATP was also quantified as a measure of mitochondrial function in MHC with and without HO-1 siRNA transfection and LPS stimulation. Mitochondrial biogenesis was examined by immunofluorscene for PGC-1 alpha and NRF1, two important mitochondrial biogenesis transcription factors.

Cecal ligation and puncture was performed in C57Bl/6 mice with and without the administration of SnPP (50 ug/kg; IP). Immunofluorscene was done to evaluate for markers of mitochondrial biogenesis.

Results: Oxygen consumption rate (OCR) in MHC over time decreased until the 12 hour time point and then normalized by 24 hours. This normalization of mitochondrial function does not occur in those MHC which were pretreated with SnPP, in fact mitochondrial OCR continued to decrease over time. The decrease in the mitochondrial function after SnPP treatment was confirmed by looking at mitochondrial membrane potential with JC-1 staining. Looking at the ATP content, LPS stimulation lead to a decrease in ATP content compared to those without LPS stimulation. However, after inhibition HO-1 with siRNA there was an even more significant decrease in ATP content.

When looking at mitochondrial biogenesis transcription factors over time after LPS stimulation in MHC, there was a time dependent increase in PGC-1 expression in those MHC where HO-1 was functional. However, after the inhibition of HO-1 this induction was abrogated.

In the in vivo model of sepsis immunofluorscene confirmed the above in vitro data of the induction of PGC-1 alpha in those mice were HO-1 activity was not pharmacologically inhibited.

Conclusions: The induction of HO-1 after a septic insult is a protective mechanism to replace dysfunction mitochondria with functional mitochondria and to return cellular energetic homeostasis.

O30. Novel Alternative Splice Variant of the Human Glucocorticoid Receptor Augments the Response to Hydrocortisone

Background: Since the initial trials evaluating steroid use in septic shock, little insight has developed as to why some patients respond to steroids and others do not. Prior clinical studies have identified five novel alternative splice variants of the human glucocorticoid receptor (hGR) as a contributor to a patient's inability to respond to steroid therapy in certain diseases. In order to further evaluate the role of hGR variations in humans, we hypothesized that glucocorticoid receptor polymorphisms influence the variability in the response to injury and may lead to individualized tailoring of steroid therapy.

Methods: Total RNA was isolated from ninety-seven healthy human volunteer blood samples and surveyed for the glucocorticoid receptor gene. A novel isoform termed hGR AS-1 resulting from an alternative splicing event and retention of intron H was identified. An hGR isoform matching the National Center for Biotechnology Information reference (hGR NCBI) was used for comparison. Transactivation potentials of TSA201 cells transfected with these hGR isoforms were measured using a reporter assay in the presence of equivalent doses of 0.9% saline or hydrocortisone. In order to further elucidate the mechanism of hGR AS-1, a deletion clone of hGR AS-1 was created in which the 3′ untranslated region (3′UTR) was removed and tested with hydrocortisone. Student's t-test was used for statistical analysis.

Results: Along with a retained intron, hGR AS-1 contains a frameshift mutation resulting in early termination of the coding sequence and a truncated protein of 118 amino acids (compared to the hGR NCBI of 777 amino acids). Isoform hGR AS-1 had decreased activity compared to hGR NCBI without hydrocortisone stimulation (p < 0.01). Most interestingly, a more than 10 fold increase in transactivation potential occurred in hGR AS-1 compared to hGR NCBI at a dose of 1 uM of hydrocortisone (p < 0.001) (Figure 1). A peak response was identified at 0.01 uM for hGR NCBI, whereas the peak response occurred at a higher dose of 1 uM for hGR AS-1. When the 3′untranslated region (3′UTR) was deleted from hGR AS-1, the resultant deletion clone no longer demonstrated a significant increase in transactivation potential when stimulated with hydrocortisone (p < 0.01). Western blot confirmed the protein lengths.

OPEN IN VIEWER

Figure 1:

Transactivation potential after hydrocortisone stimulation of hGR AS-1 and hGR NCBI. A greater than 10 fold increase in activity occurs between hGR AS-1+ Hydrocortisone and hGR NCBI + Hydrocortisone at 1μM (*p < 0.001).

Conclusion: A novel glucocorticoid receptor alternative splice variant isoform caused by an intron retention exists in healthy humans. hGR AS-1demonstrates decreased transactivation potential at baseline, but displays an impressive hyperactive transactivation potential in response to exogenous hydrocortisone in a dose dependent fashion. The 3′UTR appears to be a key mediator of this increased transactivation potential. Additional characterization of such unique isoforms in humans may help explain the differential response to steroids in inflammatory states and allow for tailored steroid therapy.

Introduction: Sepsis and septic shock are the leading causes of death in critically ill patients. The gastrointestinal tract is an important defense barrier for the body; in critically ill patients it can become a major pathogenic source of bacteria and inflammatory mediators leading to a septic response. Acute intestinal ischemia/reperfusion (AII/R) is an adaptive response to shock; mortality ranges from 60–90% and have not changed since the 1940s. The high mortality rate for AII/R is due to the severity of the disease and, more importantly, the failure of timely diagnosis.

Background: Metabolomics is the study of low molecular weight compounds generated by cellular metabolism. The metabolite profile of an organism can vary as a result of environment, activity, or disease. To date, there have been no studies applying this technology to the area of AII/R induced lung injury. The aim of this investigation is to correlate the development of AII/R with specific changes in metabolites which will aid in the diagnosis and understanding the metabolite fingerprint associated with local and distant organ injury.

Methods: We have established a model in 129/SvEv mice where AII/R results in acute lung injury (ALI) in a time dependent manner. Mice were subjected to intestinal ischemia by occlusion of the superior mesenteric artery. A time course of intestinal ischemia from 0.5 to 1 h was performed and followed by reperfusion for 2 h. Urine metabolomic profiles using nuclear magnetic resonance (NMR) spectroscopy were used to find differences in metabolites during the time course of AII/R. Principal component analysis (PCA) and partial least-squares-discriminant analysis (PLS-DA) were conducted using SIMCA P11.0 (Umetrics, Umea, Sweden).

Results: Metabolite concentrations were derived from analysis of NMR spectra from urine samples from mice exposed AII/R. Altogether, 57 compounds were identified in the spectra (e.g. alcohols, amino acids, fatty acids, lipids). PCA and PLS-DA of the data derived from measured urinary metabolite concentration demonstrate large differences in metabolomic profile could be delineated based on length of ischemia (Fig A and B). Interestingly, lactate was not an important metabolite driving the differences between these groups.

OPEN IN VIEWER

Conclusion: This study illustrates that urinary metabolites concentration may be effective at distinguishing AII/R progression in an animal model. The novel use of metabolomic studies to generate an AII/R metabolite profile is a much needed top-down approach to bridge existing knowledge gaps. The discovery of a fingerprint profile of AII/R will be a major advancement in the diagnosis, treatment, and prevention of systemic injury in critically ill patients.

O32. No Associated Increase in ARDS or Sepsis with Beta-Blocker use in Adult Burn Patients: A Multi-Center Study

Introduction: The recently published POISE trial suggested that beta-blockade with metoprolol in patients older than 45 years of age with atherosclerotic heart disease increased mortality and other complications. Our group has successfully used propranolol to decrease the hypermetabolic response in both pediatric and adult burn patients. Here we examined data collected as part of a larger federally-funded multi-center program to determine the morbidity and mortality associated with beta-blockade in adult burn patients.

Methods: Patients admitted within 96 hours of injury and with burn size > 20% TBSA were consented to participate in a larger federally-funded study. Data from 251 patients (132 on beta-blockers (dose and agent not identified) and 119 controls) was down-loaded from the trial database. Two-tailed student t-tests were used to compare data from the two groups.

Results: No differences were found between the two groups for incidences of ARDS, sepsis, cardiac events, vascular events, bradycardia, or hypotension. Although patients on beta-blockers were more likely to develop pneumonia or blood stream infections, they were less likely to die from these infections. Survival was increased in the beta-blocker group (88% vs. 76%, p < 0.05).

Conclusion: In this preliminary evaluation, beta-blocker use in hospitalized adult burn patients improved survival and was not associated with increased cardiac or pulmonary events. As this trial was not a randomized prospective trial to test the efficacy or safety of a single beta-blocking agent, we do not know which beta-blockers were used nor the doses administered. However, the findings of this trial support the need for further demonstration of safety and efficacy testing of particular beta-blockade agents in adult burn patients.

O33. Bacteremia and VAP: A Marker for Contemporaneous Extra-Pulmonic Infection

Introduction: Ventilator Associated Pneumonia (VAP) is a well known complication in severely injured patients. A subset of patients with VAP develops an associated bacteremia (B-VAP), and the risk factors, microbiology, morbidity and mortality are not well described. The goal of this study was to examine the incidence, predictors, and outcome of B-VAP in adult patients admitted to a level-I urban trauma center.

Methods: Retrospective review of trauma patients who developed VAP or B-VAP from 1/07-12/09. VAP was defined as a documented VAP along with positive respiratory cultures (BAL with ≥ 104 organisms or tracheal aspirate with moderate-many organisms and PMNs). Recurrent VAP was defined as positive respiratory cultures after ≥ 7 days of antibiotic therapy. B-VAP was defined as organism cultured from blood matching the lung pathogen. Demographic data, injury severity, transfusion and microbiology were reviewed. Outcome included ICU and hospital length of stay, ventilator days and survival. A student t-test, chi-square test or logistic regression was used as appropriate.

Results: 4,018 adult patients were admitted during a 36-month period. 206 patients were diagnosed with VAP (5%), and 13% had an associated bacteremia. Mean time from admission to the development of VAP was 5 days (95% CI 4.6-5.8). In our population, initial Gram-negative (G-) VAP were twice as common as Gram-positive (G+) and bacteremia was more common in G- vs. G + VAP (16% vs. 10%). Patients who had B-VAP received significantly more units of PRBC transfusion (see table). Of 206 patients, 70 (34%) had recurrent pneumonias—patients with B-VAP had higher rates of recurrent pneumonia (54% vs. 31%). Patients with B-VAP compared to VAP alone also had higher rates of simultaneous non-pulmonary infections (69% vs. 38%). Patients with B-VAP had increased days of mechanical ventilatory support, increased number of ICU days (ICU LOS) and hospital length of stay (H LOS) compared to patients with VAP alone (see table). Patients with B-VAP trended towards a lower survival, but B-VAP was not an independent predictor of mortality.

Conclusion: The number of blood transfusions received is the most significant risk factor for developing B-VAP and B-VAP is associated with a higher morbidity than VAP alone. Over two-thirds of patients with B-VAP have a contemporaneous infection and more than half develop recurrent VAP. Trauma patients with B-VAP may benefit from increased surveillance for additional concomitant infections or more aggressive antimicrobial coverage.

O34. Bladder Pressure Measurements and UTI in Trauma Patients

Introduction: In an effort to decrease urinary tract infections (UTI) in our ICU we evaluated the role of bladder pressure measurements (BPM) as a possible risk factor. We demonstrated that BPM using an open technique was an independent predictor of UTI and correlated with a higher mortality rate in critically injured patients. The purpose of this trial was to determine if changing to a closed technique for BPM would eliminate it as a risk for UTI in trauma patients.

Methods: Data was collected prospectively from 1/06 until 12/09 by a dedicated epidemiology nurse and combined with data from the trauma registry at our Level 1 trauma center. All trauma patients admitted to the surgical trauma ICU (STICU) with and without UTI's were compared for demographics, injury severity (ISS), and epidemiologic data including urinary drainage catheter (UDC) use and bladder pressure measurements (BPM). CDC definitions were used to diagnose UTI. BPM were performed when there was a concern for intra abdominal hypertension. A closed system was used in which the UDC remained connected to the bag and 45cc of saline was injected through a two way valved sideport with subsequent measurements through the sideport.

Results: A total of 1641 patients were included in the trial. The UTI group was sicker, with longer LOS and UDC days compared to the no UTI group (Table 1). There were 285 patients with no UDC and 1 UTI (0.35%), 1292 patients with UDC but no BPM with 37 having a UTI (2.86%) and 64 patients who had UDC and BPM of whom 8 had UTI's (12.50%) (p < 0.0001). The BPM group had more UDC days (15.6 days ± 16.0 BPM vs. 5.4 days ± 7.3 no BPM, p < 0.0001) yet no differences in UTI rates/1000 UDC days (5.7 no BPM vs. 8.0 BPM, p = 0.5291). Subset evaluation of patients with BPM found that those patients who also had UTIs had more BPM obtained compared to those with no UTIs (22 ± 16.1 UTI vs. 9.34 ± 8.34 no UTI, p = 0.0147 yet there was no difference in UTI rates in those patients who had an elevated level (>20) compared to those with only normal measurements (p = 0.1771). Logistic regression demonstrated only UDC days (not BPM or ISS) to be a predictor of UTI (1.125, 1.097–1.154, p < 0.0001) while ISS (1.083, 1.063-1.104, p < 0.0001) and age (1.051, 1.037–1.065, p < 0.0001) were the only predictors of mortality.

Conclusion: Although BPM patients have more UTI's compared to those patients without BPM it is not an independent predictor of UTI using the closed system thereby proving the hypothesis. Regardless of BPM or ISS, patients with prolonged UDC use are at higher risk for UTI as are those patients who have more BPM obtained although this was not an independent predictor. These findings emphasize the judicious use of BPM using a closed system and more importantly the need for early removal of these catheters. Hence patients should be evaluated everyday for the need for BPM and UDC and should have discontinuation of both whenever possible.

Background: Hypophosphatemia is known to occur after a surgical stress in humans. The low phosphorus is due to a combination of negligible dietary intake, increased renal excretion, and the movement of phosphorus intra-cellularly due to glucose/insulin administration. In addition, in murine models, low phosphate levels have been shown to increase the virulence of pathogens common in post surgical infections (Pseudomonas and Candida) and therefore should lead to an increase in the number of infections.

Methods: All adult patients who underwent an operation while admitted to a general surgery service (minimally invasive, colorectal, endocrine, surgical oncology, thoracic, renal transplant, vascular and trauma) from 2005 to 2010 and had at least one phosphate level drawn were included in the study. Patients were noted to have a postoperative infection if cultures from the wound, blood, urine, bronchial-alveolar lavage, catheter tip, or stool for C. difficile were positive. The lowest recorded serum phosphorus level was then used to predict the hazard of the development of a post-operative infection within 30 days of the operation. The institutional review board of Rhode Island Hospital approved this study.

Results: Over the five years, 6,493 patients were included in the analysis. 5,310 of these patients did not have a culture proven infection. The average lowest phosphate was 2.56 in the infection group and 2.78 in the group with out an infection (p < 0.0001). In a proportional hazards regression model to predict the development of a postoperative infection using phosphorus, each increase in phosphorus reduced the risk to develop an infection by 0.892 (Pr > ChiSq = 0.0019). When the patients were divided by type of surgical case (clean, clean-contaminated, contaminated/dirty) similar results were seen. In the proportional hazards model to predict infection, each increase in phosphorus reduced the risk by 0.842 in the clean-contaminated cases (Pr > ChiSq = 0.0242) and 0.864 in the contaminated/dirty cases (Pr > ChiSq = 0.0260). In the clean cases, the ability to predict infection by phosphorus level was lost.

Conclusion: Low levels of phosphorus have been previously shown to increase the virulence of common pathogens in murine models. Here we show that postoperative hypophosphatemia increases the risk of infection in humans. This study translates research initially done in mice and proves its applicability to humans. Based upon these findings we advocate for routine monitoring of phosphorus levels in post surgical patients as well as regular phosphorus replacement to achieve high-normal levels and to consider phosphorus supplementation pre-op.-.

O36. Does Elevated Glycosylated Hemoglobin Predict Post-Operative Infectious Complications?

Background: Previous studies have suggested an association between elevated pre-operative glycosylated hemoglobin (HbA1C) and complications following cardiac and bariatric surgical procedures. Recent literature has suggested that an elevated glycosylated hemoglobin level is a risk indicator for surgical site infection (SSI) and other adverse outcomes in patients with or without diabetes.

Methods: This was an IRB approved research project. A power analysis was performed to determine the necessary sample size to determine if there would be a statistical difference in a group of patients with an expected post-operative infection rate of 3%. 1610 consecutive patients had serum HbA1C levels obtained as part of their routine pre-operative surgical testing. The records from diabetic or non-diabetic patients undergoing elective surgery from January to August 2010 were analyzed. Patients whose surgery was cancelled in order to optimize their pre-operative glucose were excluded from this analysis. Post-operative outcomes were considered adverse if the patient developed a surgical site infection, deep wound abscess, urinary tract infection, or pneumonia within 30 days of their operation.

Results: 687 males and 923 females with a mean age of 46.9 years underwent elective surgery. Overall, 50 patients had an infectious complication (3.1%). There were 29 surgical site infections (1.8%), 13 urinary tract infections (0.8%), 4 pneumonias (0.2%) and 4 deep wound abscesses (0.2%). 26 percent of the patients that had an infectious complication had a preoperative HbA1C of greater than 6.0 percent; compared to 20 percent of uncomplicated post-operative patients (P = 0.293). The mean HbA1C in the group of patients that had infection was 5.9% compared to 5.8% in the group of patients that did not have post-operative infection (P = 0.4). There was no correlation between age, gender and the development of post-operative infectious complication.

Conclusion: Pre-operative measurement of HbA1C levels may be useful in the identifications of patients with occult diabetes but does not appear to predict the development of post-operative infectious complications in this population of patients undergoing elective surgery.

O37. Increased Morbidity and Mortality Associated with Post Liver Transplant Intra-Abdominal Infections Caused by Candida Species and Vancomycin-Resistant E. Faecium

Introduction: Intra-abdominal infection continues to be an important problem in surgical practice and is increasingly associated with the isolation of fungi and resistant pathogens.

Background: For liver transplantation, intra-abdominal infection presents a special challenge due to the complexity of the surgery and the need for immunosupression. In addition, transplant patients are at high risk for infections with fungi and vancomycin-resistant enterococci (VRE). We hypothesized that fungi, including, Candida species and unspeciated yeasts, and VRE would be common pathogens after liver transplantation and would be associated with significantly higher mortality, particularly when isolated together.

Methods: Review of all liver transplants in a single institution as well as episodes of intra-abdominal infection following liver transplantation. Specific isolates were analyzed and differences in outcomes were compared using Chi square analysis and Student's t-tests. Results: In our institution we performed 640 liver transplants from December of 1996 to October 2010; these were complicated by 410 episodes of intra-abdominal infection following transplantation. 306 occurred in males and 104 in females. Enterococcus faecium (100 isolates/24.4% of cases, including 65 VRE) and Candida species including unspeciated yeast (90 isolates/22.0% of cases) were the most common pathogens cultured. Demographics and outcomes, including hospital length of stay (LOS) and mortality, are given in the table below.

Conclusion: We conclude VRE and Candida species causing intra-abdominal infections result in significant morbidity and mortality for liver transplant patients, and their combination may be even more lethal. Empiric antimicrobial therapy with activity against VRE and yeast should be considered in all liver transplant patients suspected of having a new intra-abdominal infection. Whether these pathogens should also be targeted with perioperative antimicrobial prophylaxis deserves further study.

Introduction: Methicillin-resistant Staphylococcus aureus (MRSA) remains a common and problematic hospital pathogen. Previous reports have documented that nasal MRSA colonization increases the risk of developing MRSA in a subsequent infection. Ventilator associated pneumonia (VAP) remains a persistent problem in intensive care units. However, little study has been undertaken exclusively in surgical patients to evaluate the direct association between nasal MRSA colonization and subsequently having MRSA as the causative organism of pneumonia.

Methods: This is a retrospective chart review of surgical patients admitted over 3 years. Patients were included if they were aged 18 years and older, admitted to a general surgery or trauma service and underwent nasal MRSA screening. Patients who were clinically suspected of having pneumonia underwent bronchoalveolar lavage (BAL) with samples sent for culture and sensitivity. Pneumonia was defined as either ≥ 2+ organisms on qualitative culture, or a count of at least 10,000 organisms/ml on quantitative culture. Institutional review board approval was obtained for this study.

Results: Over the study period 4281 patients were included and of these 386 (9%) were positive for MRSA in the nares. Furthermore 4% (169/4281) of patients included developed pneumonia. Patients that were MRSA nares positive were twice as likely to develop pneumonia (7.5%, 29/386) compared to patients who were MRSA nares negative (3.6%, 140/3895, p = < 0.001). MRSA nares screening samples were also compared with the BAL microbiology in patients who developed pneumonia. In patients that developed pneumonia and were MRSA nares positive, 51.7% (15/29, p = < 0.001) had MRSA pneumonia. In patients that developed MRSA pneumonia, 83.3% (15/18, p = < 0.001) had been MRSA nares positive. In patients that developed pneumonia and were MRSA nares negative, only 2.1% (3/140, p = < 0.001) had MRSA pneumonia.

Conclusion: This is a unique description of a direct association between nasal MRSA carriage and subsequent MRSA pneumonia in an exclusively surgical population. Nasal MRSA colonization increases the risk of both developing pneumonia and specifically MRSA pneumonia. Furthermore, the current oral care alone may be inadequate to minimize MRSA pneumonia. We propose that all patients with anticipated intubation for greater than 24 hours should undergo nasal MRSA screening, and if positive, subsequent treatment with mupirocin. Finally, in patients who are MRSA nares negative, the empiric antibiotic regime could be narrowed to exclude MRSA coverage.

O39. Ventilator Free Days as a Predictor for Pneumonia: Is it Accurate?

Background: Recently, the use of ventilator free days VFDs28 (based on 28 days) in predicting outcomes in critically ill patients has been advocated in favor of total ventilator days. Our objectives of this study were to compare number of ventilator days to VFDs (28 day) versus actual ventilator free days in predicting pneumonia in critically injured trauma patients.

Methods: Prospective data were collected on 703 patients who were ventilated between 2001 and 2008 at a level one trauma center. Data collected included gender, race, ISS score, mechanism of injury (MOI). Primary outcomes include the diagnosis of pneumonia, number of ventilator days, number of ventilator free days (VFD), mortality, and hospital length of stay (HLOS). Three separate analyses were performed to predict pneumonia: 1: total number of ventilator days 2:number of VFDs using the 28 day criteria, 3: number of actual VFDs defined as (HLOS-x), where x is the number of days on ventilation. Univariate and logistic regression models were utilized in this model.

Results: 703 patients were enrolled in the study. The mean age and ISS were 40 ± 19 and 28 ± 13 respectively. The majority of patients were male (78%) and MOI was blunt (82%). The average HLOS and ILOS were 15 ± 16 and 13 ± 14. The mortality rate was 13.9%. The diagnosis of pneumonia was made in 146 patients. 91 patients (13%) were in the hospital longer than 28 days. Ventilator days were found to be a significant predictor of developing pneumonia (p < 0.001). Using the 28 day criteria of VFDs, the amount of time off of the ventilator decreased the likelihood of developing pneumonia (p < 0.001). Actual VFDs was not associated with a significant decrease in pneumonia (p = 0.234).

Conclusion: Actual ventilator free days was not found to be a useful predictor of outcome. Although VFDs was predictive of pneumonia, it did not account for 13% of our patient population. A new model is needed to better predict outcome in critically ill patients who are in the hospital beyond the 28 day study period.

O40. The Effect of Adopting a Clinical Pathway for the Management of Pediatric Appendicitis and Comparison to the Pediatric Health Information System Database

Introduction: Health care reform drives hospitals to adopt policies to reflect evidence-based medicine and optimize provision of care and utilization of resources. To achieve these goals, our children's hospital implemented a protocol for the management of appendicitis utilizing a single-agent antibiotic regimen and a uniform clinical pathway for discharge in 2007. We analyzed the effects of our protocol by comparing our outcomes before and after protocol implementation to those reported in the Pediatric Health Information Systems (PHIS) database.

Methods: Data were collected from medical record review of patients with perforated appendicitis at our institution between 2004 and 2009. A clinical pathway for the management of appendicitis was initiated in 2007 using a uniform single-agent antibiotic regimen (ertapenem) and detailed discharge criteria (afebrile for 24 hours, tolerating oral intake, adequate pain control, and decreasing white blood cell count.) Additionally, data were collected from the PHIS database which represents 41 free-standing children's hospitals from 2004–2009. Hospital length of stay (LOS) and thirty day readmission rates were calculated and compared pre- and post- protocol implementation. Additionally, our data were compared to those from the PHIS database pre- and post- protocol initiation. Mean values were compared using Student's t-test, and categorical data were compared using chi-square analysis.

Results: Between 2004 and 2009, there were 544 patients with the diagnosis of perforated appendicitis admitted to our institution, and the PHIS database registered 23,276 perforated appendicitis patients. After implementation of the protocol, the average LOS decreased from 6.4 ± 5.3 days to 4.5 ± 2.9 days (p = 0.0001). Prior to implementation of the protocol, our average LOS, though not statistically significant, was greater than the reported national average (6.4 ± 5.3 days versus 6.1 ± 4.3 days, p = 0.267), and following implementation is now lower than the national average (4.5 ± 2.9 days versus 6.0 ± 4.4 days, p = 0.0001). The number of hospital readmissions within thirty days did not change significantly following the initiation of the protocol (8.1% vs. 11.2%, p = 0.3). Additionally, the pre- and post-protocol readmission rates were comparable to the PHIS data.

Conclusion: Adopting a clinical pathway using a single-agent antibiotic regimen and uniform discharge criteria has led to decreased LOS for children with perforated appendicitis when compared to pre-protocol outcomes as well as national administrative data. Furthermore, the protocol implementation has not adversely affected patient outcomes as measured by thirty day readmission rates. We recommend the development and utilization of a clinical protocol including a single-agent antibiotic regimen and uniform discharge criteria for the management of perforated appendicitis in children to optimize patient care.

OPEN IN VIEWER

O41. Preventing Injury-Associated Sepsis by Redirecting the Immune System

Introduction: Severe injury disrupts immune system homeostasis and predisposes trauma patients to infections and complications. In this study, we wished to test whether early intervention with Rapamycin - a drug that is used to prevent host rejection in organ transplantation – might help restore immune system homeostasis by changing the clinical trajectory of the injury response. Since Rapamycin is known to enhance the activation and expansion of CD4 + regulatory T cells (Tregs), an immune cell that is activated by injury and is counter-inflammatory, we hypothesized that Rapamycin may help counteract the pro-inflammatory response to burn injury in mice.

Methods: Mice underwent sham or burn injury under anesthesia and were treated with Rapamycin (1 mg/kg) or vehicle by intraperitoneal injection. Sepsis survival was evaluated by the cecal ligation and puncture (CLP) method at day 7 after sham or burn injury. Immunophenotyping was accomplished by FACS analysis of blood, spleen, and lymph node cells and by profiling changes in cytokine production by spleen cells that were stimulated with a variety of innate and adaptive immune cell stimuli.

Results: The timing for Rapamycin treatment was assessed by measuring Treg expansion at 7 days after injury in mice that were treated at 2 hours, 1 day, or 3 days post-injury. We observed that treatment at 2 hours after burn injury caused significant expansion of Tregs as compared to day 1 or day 3. Given this finding, we used 2 hour post-injury treatment to test whether Rapamycin might protect mice from two-hit sepsis mortality induced by CLP. We found that burn-injured mice treated with Rapamycin survived CLP sepsis significantly better than vehicle treated mice. Interestingly, there was no significant change in CLP mortality between sham and Rapamycin-treated sham mice. Immunophenotyping studies showed that Rapamycin treatment did not significantly alter the immune cell subset percentages in peripheral immune compartments and that innate responses were not markedly altered. However, we found that Th2-type cytokine production was significantly reduced by Rapamycin treatment. The results of in vitro studies indicated that Rapamycin potently blocked T cell cytokine responses suggesting that T cells represent the primary cell type that may be affected by Rapamycin treatment. We did not observe significant effects of Rapamycin treatment on innate immune cell responses to Toll-like receptor stimuli.

Conclusion: These findings support a potential for using Rapamycin as an early interventional drug treatment to restore immune system homeostasis and to reduce the morbidity and mortality of sepsis following major injury. Furthermore, these data suggest that Rapamycin may act by altering adaptive immune system activation following burn injury.

O42. Cholinergic Regulation of ANP Signaling During Sepsis

Introduction: Atrial natriuretic peptide (ANP) is a hormone secreted primarily by cardiac myocytes that functions in regulation of blood volume and pressure. Evidence indicates that ANP also has potent immunomodulatory effects. ANP suppresses proinflammatory cytokine production and inhibits NFκB activation in tissue macrophages. ANP release from the spleen can cause hypotension and shock in murine endotoxemia. Clinical studies show that concentrations of circulating ANP correlate directly with morbidity and mortality in patients with lethal sepsis.

The cholinergic anti-inflammatory pathway is an endogenous, brain-to-immune pathway that inhibits pro-inflammatory cytokine production, prevents shock, and suppresses lethal organ damage in experimental sepsis. We found that this vagus nerve-mediated pathway specifically targets immune cells in spleen. Interruption of the hard-wired neural connection between the vagus nerve and the spleen abolishes the protective effects of the cholinergic anti-inflammatory pathway.

In light of these findings, we hypothesized that activation of the cholinergic pathway may regulate systemic cytokine production via modulation of ANP in spleen. Here we studied whether administration of the cholinergic agonist nicotine regulates systemic and splenic ANP and TNF production in murine models of systemic inflammation and sepsis.

Methods: Male 8–12 week old BALB/c mice received vehicle or endotoxin (15 mg/kg, i.p., E. coli LPS 0111:B4; Sigma-Aldrich). Plasma and organs were collected after 60 min. Nicotine (2 mg/kg, i.p., Invitrogen) or vehicle was administered 1–2 hrs prior to endotoxin. TNF and ANP concentrations were measured using ELISA (R&D Systems, Phoenix Pharmaceuticals). BALB/c mice underwent cecal ligation and puncture and splenectomy as described previously (Huston et al. 2006. J Exp Med 203; 1623–28). Plasma was collected after 24 hrs. One-way ANOVA using the Bonferroni correction or the Student's t test was used to compare mean values between multiple or two groups, respectively. p-values < 0.05 were considered statistically significant.

Results: Nicotine significantly reduces systemic and splenic TNF concentrations in endotoxemic mice. Nicotine also significantly reduces systemic ANP concentrations, but significantly increases ANP concentrations in spleen. Splenectomy significantly reduces systemic TNF concentrations in endotoxemic mice. There is a trend towards decreased ANP concentrations in splenectomized, endotoxemic mice (p < 0.07). In splenectomized mice, nicotine paradoxically increases systemic TNF, and fails to reduce systemic ANP concentrations. Splenectomy significantly improves survival and decreases systemic ANP during CLP sepsis.

Conclusions: The cholinergic anti-inflammatory pathway reduces systemic ANP and TNF concentrations, but only in the presence of the spleen. Nicotine reduces TNF concentrations, but increases ANP concentrations, in spleen. Splenectomy decreases ANP concentrations and improves survival in CLP sepsis. These findings suggest that the cholinergic pathway may inhibit systemic cytokine production by limiting splenic ANP release and enhancing ANP signaling in spleen. Further studies are needed to determine if specific ANP signaling in spleen is essential to the protective effects of the cholinergic anti-inflammatory pathway.

O43. The Role of Antibiotics in Open Fractures Revisited: Characteristics of Staphylococcus Aureus (SA) and Susceptibility Profile

Background: The benefits of antibiotic prophylaxis in open fractures have been clearly demonstrated in several randomized controlled trials, but the choice of antibiotic remains controversial since the current guidelines do not recommend agents with methicillin resistant Staphylococcus aureus (MRSA) activity.

Methods: To develop preliminary data on the incidence of S. aureus (SA) colonization and surgical site infections (SSI) in patients with open fractures, a randomized prospective clinical trial was performed to compare the safety and benefits of adding MRSA coverage to standard antibiotic prophylaxis. Between April 2009 and November 2010, all consecutive adult patients with an open fracture who met study criteria were included. Patients were randomly allocated to receive either cefazolin alone (control arm) or vancomycin and cefazolin (experimental arm) from presentation to the ER until 24 hours after the surgery. Screening for SA carriage was performed with nares swabs and open fracture wound swabs. Patients underwent prospective assessment for the development of SSI within 30 days of the initial surgery.

Results: Over a 20-month period, 92 patients with an open fracture were randomized to one of the prophylactic antibiotic regimens (46 subjects in each study arm). Study demographics and baseline characteristics were balanced between the two groups. 21.7% of the subjects were identified as colonized with SA. The prevalence of methicillin sensitive SA (MSSA) and MRSA among the sample were 18.4% and 3.2% respectively. None of the open fracture wounds were found to be colonized with SA. Overall 8 of 92 patients (8.7%) developed a SSI, 4 of 46 in each study group. Selected characteristics of patients who developed a SSI are summarized in Table 1. No adverse events in either of the study arms were reported during the study period.

Conclusions: Preliminary findings demonstrate that addition of vancomycin to standard antibiotic prophylaxis is safe. SA was the most common cause of SSI after open fractures. SA colonization in orthopedic trauma patients is similar to the general population and methicillin resistance was shown in 3.2% of these patients. Pathogens inherently resistant to cephalosporins (i.e. Enterococcus) were also causative pathogens in two SSIs, one of which was susceptible to vancomycin. Further data are needed to determine whether the addition of vancomycin to standard antibiotic prophylaxis prevents the development of SSI compared to standard regimens.

Introduction: One of the most common life-threatening complications in long-term immunosuppressive therapy populations is infection. A blunted inflammatory response is seen in the transplantation patient following induction of immunosuppression, and a high rate of infectious complications ensues. Whether the infectious morbidity is due entirely to iatrogenic immunosuppression or also depends on the immunomodulation induced by the presence of an allograft (the “alloresponse”) is unclear. We hypothesized that the transplant population, as compared to other non-transplant immunocompromised surgical patients, has a greater morbidity and mortality from infection, possibly due to the alloresponse.

Methods: Using a prospectively-collected dataset tracking surgical infectious complications over a 15-year time period, all non-viral infections were identified for solid organ transplant recipients and non-transplant immunocompromised surgical patients, defined as fulfilling the definition of immunosuppression by APACHE II criteria (chronic low-dose corticosteroids, acute high dose corticosteroids, etc.). Total infections, mortality following infection, crude mortality, length of stay, and APACHE II scores were obtained. Fisher's exact test and t-tests were used to compare the two groups.

Results: A total of 795 transplant and 412 non-transplant, immunocompromised patients with infections were identified. The most common indications for immunosuppression in non-transplant patients were inflammatory bowel diseases, rheumatoid arthritis/autoimmune diseases, and COPD/asthma. Transplant recipients had a total of 1781 infections, a mean of 2.24 infections per patient, and non-transplant patients had 659 infections, a mean of 1.60 infections per patient. APACHE II scores were 16.1 ± 6.4 and 15.3 ± 7.1 for transplant and non-transplant patients, respectively (p = 0.039). No significant differences were found between sites of infection or causative pathogens, with the exception that transplant patients had a higher percentage of infections caused by resistant organisms, most notably vancomycin-resistant enterococci. Hospital length of stay after the diagnosis of infection was 18.5 ± 23.0 days for transplant and 23.9 ± 35.8 days for non-transplant patients (p = 0.002). All-cause, in-hospital mortality following an infection for the transplant recipients was 9.6% and the non-transplant patients was 10.2% (p = 0.76).

Conclusion: Despite the fact that transplant recipients, when compared to non-transplant, immunocompromised surgical patients, have a greater severity of illness, more infections with resistant pathogens, and more infections per patient, their overall survival is similar. These findings suggest that the transplant patients' susceptibility to contracting an infection may be higher than other immunosuppressed patients but the ultimate ability to clear the infection is unchanged. Whether these findings are due to the alloresponse or level of immunosuppression remains unclear.

O45. The Relationship of Stress-Induced Hyperglycemia and Infectious Complications in Non-Diabetic Orthopaedic Trauma Patients

Introduction: Infection following orthopaedic surgery is a significant concern for the treating surgeon. The purpose of this investigation was to evaluate the effect of stress-induced hyperglycemia on infectious complications in a population of orthopaedic trauma patients.

Methods: Retrospective study of isolated orthopaedic trauma patients admitted to an academic Level-1 trauma center from January 2004-October 2009. Inclusion criteria: age ≥18 years, Intensive Care Unit (ICU) stay ≥1 day, orthopaedic injury requiring operative intervention, and more than 1 documented blood glucose (BG) value. Patients with a history of diabetes mellitus, corticosteroid use, immunologic disease, or Abbreviated Injury Scale scores other than Extremity were excluded. Demographic, injury severity, units of red blood cell transfusion, blood glucose (BG), and infectious complication data were obtained. Duration of ICU and hospital length of stay (LOS) was recorded. Infections considered were pneumonia, urinary tract infection (UTI), surgical site infection (SSI), and bacteremia. The Hyperglycemic Index (HGI) was calculated for each patient and reflects the mean BG level above 108 mg/dL. HGI was determined from BG values obtained prior to the diagnosis of infection. HGI was considered separately as a continuous variable and a dichotomous variable > 1.72 (equivalent to BG of 140 mg/dL). Multivariable logistic regression testing the effect of blood glucose on SSI was evaluated, after controlling for potential confounding variables. Significance was considered for p-value < 0.05.

Results: 187 patients were identified. All were managed with a tight glycemic protocol (target BG: 80-110 mg/dL). An average of 21.5 BG values was obtained for each patient. Mean age was 47.7 ± 23.2 years and 121 patients (64.7%) were male. Mean ICU and hospital LOS was 4.0 ± 4.9 and 10.0 ± 8.1 days, respectively. ISS was higher in patients with an infection (10.6 ± 3.7 vs. 9.5 ± 2.8; p = 0.03), but not clinically significant. Infections were recorded in 43 patients (23.0%) and SSI's specifically documented in 16 (8.6%). Open fractures were not associated with SSI (50% vs. 51.2%; p = 0.9). By univariate analysis there was no difference in admission BG or HGI and infection. However, there was a significant difference in HGI when considering SSI alone (2.1 ± 1.7 vs. 1.2 ± 1.1; p = 0.003). HGI > 1.72 was associated with SSI's (50% vs. 19.3%, p = 0.005). Patients with an SSI received a greater amount of blood transfusions (14.9 ± 12.1 vs. 4.9 ± 7.6, p < 0.001). No patient was diagnosed with a separate infection (i.e. pneumonia, UTI, bacteremia) prior to SSI. There was no difference in ISS among patients with an SSI (11.1 ± 4.0 vs. 9.6 ± 3.0; p = 0.06). Multivariable regression testing HGI as a continuous variable demonstrated a significant relationship (OR: 1.8, 95% CI: 1.2-2.6) with SSI after adjusting for blood transfusions (OR: 1.1, 95% CI: 1.1-1.2).

Conclusion: Stress-induced hyperglycemia is a common event and a significant risk factor for SSI's in non-diabetic orthopaedic trauma patients. Blood glucose monitoring should be considered in non-diabetic orthopaedic trauma patients; however, indications and implications for tight glycemic control in these patients cannot be determined from this study.

O46. Role of Feeding-Induced Oxidative Stress and TLR-Response on Cellular Population Dynamics in the Pathogenesis of Necrotizing Enterocolitis: Insights from an Agent-Based Model

Background: Necrotizing enterocolitis (NEC) is a leading gastrointestinal cause of death in neonates. Risk factors include: prematurity, initiation of enteral feeding, oxidative stress and bacterial effects. We propose that the initial condition in the pathogenesis of NEC is immaturity of the ability of neonatal gut epithelial cells (NGEC) to manage oxidative stress (OS). Reduction in the distribution of OS tolerance across the cellular population is accentuated by enteral feeding, leading to a shift from apoptotic to inflammatory pathways, alterations in gut microenvironment, subsequent bacterial virulence activation, exacerbation of mucosal inflammation, and eventual necrosis. Toll -like receptors (TLRs) play an important role in this process and we propose that danger/pathogen associated molecular pattern molecules (DAMP/PAMPs) acting through TLRs tip the population dynamics of susceptible NGECs to a trajectory towards NEC. The multi-factorial complexity of these interacting processes requires characterization with computational modeling and agent-based modeling has been used to dynamically represent hypotheses in this fashion. We present an agent-based model (ABM) of NGEC to visualize the dynamics of the pathogenesis of NEC.

Methods: An ABM of the neonatal gut was created using NetLogo. Rules for NGEC computational agents, derived from the literature, incorporate pathways for OS, p53, apoptosis, tight junction metabolism, TLR-4, nitric oxide (NO), NF-kB and high mobility group box protein-1 (HMGB-1). Generically modeled bacteria were able to respond to environmental cues, activate virulence, produce PAMPs, adhere to and destroy host cells. Both DAMP/PAMPs could activate TLR-4. Experimental runs included baseline parameter sweeps of OS management capacity, subsequent response to feeding, and alterations in bacterial load and virulence potential.

Results: The ABM reproduced baseline behavior of nutrient uptake and production and clearance of OS. Reductions in either stress detection or clearance produced excess OS, which led to senescence, apoptosis or inflammation depending OS levels. Reduced capacity to manage OS led to increased generation of NO and subsequent disruption of tight junctions. Increased prevalence in the inflammatory phenotype led to a shift in NGEC population dynamics that enhanced bacterial virulence, and generated a forward feedback loop leading to further inflammation and necrosis.

Conclusion: NEC is a complex process of cascading systems failure. Our ABM incorporates known components in the pathogenesis of NEC and demonstrates that decreased capacity to manage OS can lead to the apoptosis and inflammation of individual cells and unstable population dynamics. The ABM demonstrates the importance of DAMP/PAMP/TLR in shifting NGEC population dynamics towards the pathogenic phenotype and further suggests that the pathogenesis of NEC is not necessitated by a particular microbe, but instead by a microbial population with greater virulizing potential in response to microenvironmental cues. By identifying conditions and components that increase the susceptibility of NGECs to NEC we may be able to suggest cellular processes and putative pathways/genes that are impaired in premature infants. This type of dynamic knowledge representation can be integrated into existing research programs and future research endeavors.

O47. The Role of HIF-1 in the Pathogenesis of Liver Injury During Sepsis

Introduction: A well documented observation during severe bacterial infections leading to sepsis is the extensive loss of innate and adaptive immune cells via apoptosis. Interestingly, sepsis is also associated with increased apoptosis of liver cells. Moreover, various strategies used to inhibit apoptosis result in improved hepatic bacterial clearance, a reduction in liver damage and improved survival during sepsis. However, the mechanisms mediating these observations remain unclear. We sought to investigate a potential candidate, Hypoxia-inducible factor-1 (HIF-1), a transcription factor which influences cell survival and apoptosis pathways, during sepsis.

Methods: We created a mouse model of sepsis by injecting a fecal solution into the peritoneal cavity (FIP) of mice. At time points: 0, 3, 6, 12, 18 and 24 hours following FIP or a saline-sham control, mice (n = 4 per group/time point) were euthanized and the livers were harvested for analysis. RNA was extracted from the liver for quantitative reverse-transcriptase PCR (qRT-PCR) to determine expression of HIF-1α, FasL, bcl-2 (anti-apoptotic) and Bax (pro-apoptotic). Liver tissue was also formalin-fixed, paraffin-embedded, and stained with hematoxlyn and eosin to score for inflammation, cellular infiltrate and necrosis. Tissue sections also underwent TUNEL staining to detect apoptosis.

Results: Mice treated with FIP had a median mortality rate of 75% at 24 hours and significant liver injury compared to the sham controls. Using qRT-PCR, we were determined that liver HIF-1α (the regulatory subunit of HIF-1) gene expression was significantly increased over time when compared to sham-control mice (Figure 1). In situ TUNEL staining for apoptotic cells, demonstrated increased apoptotic cells within the liver of the FIP treated mice. To further characterize this apoptosis, we measured FasL, Bcl-2 and Bax gene expression in the liver. In the FIP mice, FasL gene expression increased over time and Bcl-2/Bax ratios decreased compared to the sham control mice confirming the TUNEL data and a state of increased apoptosis in the liver after induction of sepsis using FIP.

OPEN IN VIEWER

Conclusion: In sepsis, HIF-1α gene expression increased over time and was significantly increased compared with sham controls. The increased HIF-1α expression in the sepsis mice was paralleled by increased liver damage, increased number of apoptotic cells and increased pro-apoptotic gene expression compared to sham controls. In summary, during polymicrobial sepsis, HIF-1α expression in the liver is increased over time and is associated with increased levels of apoptosis of cells within the liver.

O48. Initiations of Quality Improvement Measures can Lead to Decreasing Infectious Complications in Emergency Surgical Patients

Introduction: Over the past decade, there has been a significant focus on reducing postoperative infections at both the national and hospital level. Despite focus on quality improvement in surgery, studies continue to show that emergency status contributes significantly to increased infectious complications and mortality of general surgery operations. We looked to see if the initiation of national and hospital wide quality measures at reducing infectious complications at our institution over the past several years have led to decreased morbidity and mortality in emergency general surgery (EGS) patients.

Background: Since 2007 several national and hospital wide quality and infection control initiative have been initiated at our institution. Some of these include, increasing compliance with The Surgical Care Improvement Program (SCIP), Implementation of Ventilator Associated Pneumonia (VAP) Bundle and urinary catheter guidelines.

Methods: We conducted a retrospective study of all American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) EGS operations performed at our tertiary medical center from 1/1/2002-12/31/2009. We compared the EGS operations performed between 2002–2006 (Group 1) to those of 2007–2009 (Group 2).

Results: Overall there has been a trend towards lower infection rates and mortality in the recent years. There has been a significant reduction in the total infection rate in the recent year (25.7% vs. 13%). In particular, there has been a significant reduction in rates of sepsis, septic shock, pneumonia, and UTI (Table 1).

Conclusion: Initiation of several national and hospital wide quality and infection control initiatives at our institution over the past several years have led to decreasing rates of sepsis, septic shock, pneumonia and UTI in our emergency surgical patients. There is also an overall downward trend in SSI rates and mortality.

Background: BK nephropathy (BKN) is a serious complication after renal transplantation. In Jan. 2006, we initiated monthly serum BK PCR screening. Patients detected to have BK viremia underwent an immediate reduction in immunosuppression consisting of a discontinuation or 50% dose reduction of cellcept (CC). Serum PCR was repeated monthly and any recipient experiencing graft dysfunction underwent transplant biopsy.

Methods: Between 1/1/06 and 12/31/08, 174 renal transplants were performed at our center. Immunotherapy consisted of tacrolimus, CC and either sirolimus or prednisone. PCR became positive for BK in 29 (17%) patients (Group A). Mean time to positive PCR post-transplant was 6 months. Following CC dose reduction, 27/29 group A patients became negative for BK detection by PCR The mean time to a negative PCR was 4 months. Transplant dysfunction subsequently developed in only one Group A patient (3%), 4 months after discontinuation of CC therapy. Transplant biopsy in this case showed acute rejection and the graft eventually failed at 37 months due to chronic nephropathy. No group A patient developed BKN. Long-term outcomes in Group A patients, mean follow-up of 42 months, were compared to that of 29 concurrent matched renal transplant recipients who never developed BK viremia and thus never underwent a reduction in immunosuppression, (GroupB)

Results: Demographics were similar between the groups with respect to: age, gender, ethnicity, donor type, retransplant status, and follow-up. Long-term survival rates and renal function were similar between Group A and B patients as shown in the table below.

Conclusion: Monthly screening for BK virus by PCR for early detection, with associated pre-emptive immunosuppressive therapy reduction, effectively prevents BKN. In addition, this screening protocol is associated with a low rate of acute rejection and preservation of excellent long-term renal function.

O50. Infectious Outcomes in Living Donor Kidney Transplant Recipients following Multimodal Preoperative Desensitization to Overcome Humoral Barriers

Introduction: Preexisting humoral barriers significantly challenge transplantation of living donor kidneys (LDK) into highly sensitized ABO and HLA incompatible recipients. Aggressive conditioning of recipients' immune systems is required pre-transplantation. We hypothesized that pre-conditioning would result in higher infection rates post-transplant.

Methods: A matched case-control design was utilized to compare desensitized patients (DS) to non-desensitized (NDS) living donor kidney transplant recipients. Each desensitized patient was hand -matched with 2 or 3 non-desensitized patients. Patients were matched closely for cause of end stage renal disease (ESRD) by general categories (congenital, immune-related, diabetes mellitus (DM) type 1, DM type 2, hypertension) and age. Desensitization included Rituximab (anti-B cell antibody) and mycophenolate mofetil 6 to 8 weeks preoperatively followed by intravenous immunoglobulin (IVIG) and plasmapharesis 2 week preoperatively. Both desensitized and non-desensitized patients received induction therapy (rabbit anti-human thymocyte globulin) followed by maintenance immunosuppression (tacrolimus, mycophenolate mofetil, and prednisone). Primary outcomes included infections (opportunistic, local, systemic) within 12 months of transplant, while secondary outcomes included patient and graft survival, and acute rejection. Univariate and log-rank statistical analysis was completed.

Results: Twenty-five highly sensitized recipients underwent desensitization and LDK transplantation. Mean age for DS patients was 42.3 ± 18.4 years versus 40.8 ± 18.7 years in NDS patients (p = 0.7). Eighty eight percent (22/25) of DS patient matches were similar for diagnosis and 19/25 (76%) were within 5 years of the age of the controls. Patient survival was 100%. Graft survival was 96% in both groups (24/25 and 51/53). Average follow-up was 36.5 ± 14.7 months (desensitized) versus 25.5 ± 17.5 months (non-desensitized). Mean 3-month and 12 month creatinine was 1.16 ± 0.2 mg/dL and 1.2 ± 0.3 mg/dL in DS versus 1.3 ± 0.4 mg/dL and 1.4 ± 0.5 mg/dL in NDS patients. Thirty six percent of desensitized patients had one or more infections versus 28% in non-desensitized (p = 0.6). The average number of infections per infected patient was 1.6 (DS) versus 1.4 (NDS). No difference was observed in opportunistic infections (pathogens unlikely to cause infection in immune-competent hosts, viruses) between DS (16%) and NDS patients (13%, p = 0.7). The number of total local infections (single organ) was significantly greater in DS patients (56% vs. 30%, p = 0.04). No difference was found in number of systemic infections (disseminated) (p = 0.4).

Conclusion: Preoperative desensitization in highly sensitized living donor kidney transplant recipients is associated with a similar number of opportunistic and systemic infections compared to NDS patients. Total number of local infections was significantly greater in the desensitized, however. With careful monitoring for infectious complications, preoperative desensitization is a valuable tool enabling highly sensitized patients to be transplanted. Future studies are required to evaluate these observed outcomes.

O51. Comparing A Locally-Derived Versus Guideline-Based Approach to Treatment of Hospital Acquired Pneumonia in the Trauma Intensive Care Unit

Background: Multidrug resistant (MDR) respiratory pathogens are an increasingly difficult problem in critically ill patients with pneumonia. MDR organisms restrict antibiotic treatment options, and are associated with poorer outcomes. This makes the selection of appropriate initial antibiotics for empiric coverage of pneumonia difficult. Given the high frequency of MDR pathogens in our surgical intensive care unit (SICU) of over 50%, we sought to compare differences in the antibiotic sensitivity profiles and the microbiology of early versus late pathogens, and determine the best empiric approach to initial antibiotic coverage in the SICU.

Methods: Retrospective analysis of continuous patients admitted to the SICU at a tertiary medical center over a one and a half year period with a diagnosis of first hospital-acquired (HAP) and ventilator-associated (VAP) pneumonia. Local microbiology and sensitivity profiles were analyzed. Risk factors for MDR pathogens were based on American Thoracic Society (ATS) guidelines, which define patients at risk based on time since admission (early being <5 days; late being ≥5 days) and other standardized criteria.

Results: A total of 64 first pneumonias were caused by 90 pathogens over the 1.5 year period. There were 25 (28%) early pathogens and 65 (82%) late. 24 of 25 (96%) early HAP/VAP pathogens fulfilled MDR risk criteria; 10 (40%) were true MDR organisms. Of the 24 early at risk patients, ATS guidelines recommend broad spectrum therapy; in our cohort, susceptibility varied: 72% with ceftriaxone/vancomycin (cftri/vanc), 92% with ciprofloxacin/vancomycin (cipro/vanc), 99% with cefepime/vancomycin (cefe/vanc), and 96% with piperacillin-tazobactam/vancomycin (p-t/vanc). Examining late HAP/VAP, there were 35 (54%) MDR organisms. ATS recommends broad-spectrum empiric coverage for all late HAP/VAP; susceptibilities were much poorer: 58% with cftri/vanc, 71% with cipro/vanc, 74% with cefe/vanc, and 71% with p-t/vanc. Addition of carbepenems or aminoglycosides to these late regimens failed to improve adequate coverage above 78%.

Conclusions: Late HAP/VAP is significantly more difficult to empirically treat compared to early pneumonia in the SICU, with the best available coverage only reaching no greater than 78%. This is concerning, as over 80% of SICU HAP/VAP occurs on day 5 or later, with 54% MDR organisms. We were unable to identify any single additional antimicrobial that significantly improves this coverage. Guideline driven treatment of SICU patients with early pneumonia yielded excellent coverage of >90%.

O52. TLR4 Protective Adjuvant Effect to Murine Polymicrobial Sepsis is Trif, but not MYD88, Dependent

Introduction: We have previously shown that pretreatment of murine neonates with toll-like receptor (TLR) 4 ligand, lipopolysaccharide (LPS), is able to significantly improve survival to polymicrobial sepsis. Although this effect was associated with increased activation/recruitment of innate immune effector cells, the mechanism is unknown. As TLR4 utilizes both TLR adaptor proteins, MyD88 and TRIF, and MyD88 but not TRIF has been linked to survival to polymicrobial sepsis in adult studies, we hypothesized that MyD88 signaling is critical for the observed TLR4 agonist protective adjuvant effect to neonatal sepsis.

Methods: Splenocytes from C57BL/6J (wild-type, WT), MyD88-/-, and TRIF-/- 5–7 days old were harvested and either stimulated via PMA to assay for neutrophil reactive oxygen species (ROS) production or stimulated in vitro with LPS for 24 hours and subsequently assayed for neutrophil phagocytic ability. For survival studies, WT, MyD88-/-, and TRIF-/- were pretreated with LPS (1 μg/g body weight) intraperitoneally (ip) followed by challenge with a cecal slurry ip 24 hours later. Mice were followed for survival for 7 days. A second cohort of neonates were euthanized at 0, 4, and 24 hours post LPS and assayed for serum concentrations of inflammatory and anti-inflammatory cyto/chemokines via Luminex assay.

Results: Surprisingly, though no differences in PMA elicited neutrophil ROS production or burst kinetics were noted between any of the groups, TRIF-/- neonates had significantly fewer phagocytic neutrophils compared to WT or MyD88-/- animals following LPS stimulation (71% vs 92% or 87%, respectively; p < 0.002). Similarly, in contrast to WT or MyD88-/- neonates, TRIF-/- neonates were not protected against polymicrobial sepsis following TLR4 adjuvant administration. This was associated with significantly lower serum concentrations of several cyto/chemokines in TRIF-/- neonates compared to WT neonates following LPS administration, including IL-1β, IFN-γ, TNF-α, and CXCL10.

Conclusion: These data provide a novel role for TRIF but not MyD88 signaling in both neonatal neutrophil phagocytic function following LPS stimulation and TLR4 adjuvant induced protection to polymicrobial neonatal sepsis.

O53. Hand Sanitizer Dispensers and Associated Hospital Acquired Infections - Friend or Fomite?

Introduction: Hand hygiene is the key weapon in preventing hospital-acquired infection (HAI). Waterless alcohol-based hand sanitizers are an increasingly popular method of hand hygiene. Hand sanitizer dispensers (HSD) have become ubiquitous in the hospital. Whether HSD may themselves harbor pathogens or act as potential fomites hasn't been previously reported. The goal of this study was to determine if HSD could become contaminated with pathogens that cause HAI.

OPEN IN VIEWER

Methods: The study took place in the SICU of an urban teaching hospital. The wall-mounted manual HSD (Purell FMX-12 Dispenser, Johnson & Johnson, New Brunswick, NJ) were located in the individual patient bays and in a centralized charting area. The dispensers delivered an alcohol solution containing 62% ethyl alcohol (Purell Instant Hand Sanitizer Foam, Johnson & Johnson, New Brunswick, NJ). Sterile cotton tipped swabs pre-moistened with sterile saline were used to sample three surfaces of each HSD unit: the dispenser's lever, the rear portion adjacent to the wall, and the area surrounding the dispensing nozzle. The swabs were subsequently cultured to trypticase soy blood agar (BBL, Becton Dickinson, Sparks, MD). All cultures were incubated at 35°C for 24 to 72 hours with observations at 24 hour intervals. At each observation, bacterial morphotypes were semi-quantified and presumptively identified using the Gram stain and select screening methods. Isolates determined to be of particular interest were identified using Vitek 2 (bioMerieux, St. Louis, MO). Methicillin resistance for the Staphylococcus aureus isolates was determined from a cefoxitin disk using the Kirby Bauer disk diffusion method

Results: All 17 HSD cultured positive for one or more bacterial species. Distribution of contamination included the lever in 82%, wall in 35%, and nozzle in 23% HSD. Multiple bacteria were obtained in 50% of contaminated sites. Growth ranged from 1 to 50 colonies, with a median of 5 and mean of 8.3. Gram-positive cocci predominated, with the majority being coagulase negative staphylococci in 16 (94%) HSD. Methicillin- sensitive S. aureus and Micrococcus grew from 2 (12%) and 1 (6%) HSD respectively. Gram-positive bacilli, including Bacillus species, diphtheroids, and aerobic actinomycetes, were together isolated from 6 (35%) HSD. Gram-negative bacilli, including non-lactose fermenting nonenterics and lactose-fermenting enterics, were together isolated from 5 (29%) HSD.

Discussion: This study showed that 100% of waterless alcohol-based HSD from a SICU were contaminated with bacteria. Colonization was greatest on the lever, likely due to direct hand contact. These bacteria were largely commensal skin flora, gram-positive cocci and gram-positive bacilli. In addition, 24% of HSD grew lactose-fermenting enterics, which may develop drug resistance and are major pathogens of HAI. The passage of pathogens from HSD to health care workers' hands could result in subsequent HAI if proper hand hygiene is ignored. The spread of these pathogens can be decreased by routine cleaning of HSD, correct sanitizer application, and use of gloves prior to patient contact.

O54. Performance of Two Surgical Site Infection Risk Stratification Models for Predicting Infection Risk in Publicly Reported Data from a Safety Net Hospital

Introduction: Since 2007, the Colorado Hospital-Acquired Infections Disclosure Act has required hospitals and ambulatory surgery centers to report selected surgical site infections (SSI) to the CDC's National Healthcare Surveillance Network (NHSN) to be analyzed and risk adjusted. The Colorado Department of Public Health and Environment uses that information to issue comparisons of facility-specific, risk-stratified infection rates to the public.

Background:. In order to provide reliable information to the public, risk adjustment models must be reliable for accurately comparing a facility's performance against that of other procedures with similar risk. However, risk adjustment models currently used for public reporting have not been widely validated. We sought to verify the NHSN basic risk index and standardized infection ratio (SIR) for their abilities to predict SSI in publicly reported herniorrhaphy, hysterectomy, and arthroplasty procedures from an academic public health safety net institution in Denver, CO.

Methods: Cases of arthroplasty (hip and knee) between January 2007 and October 2010; herniorrhaphy and hysterectomy (vaginal and abdominal) between January 2009 and October 2010 were entered into NHSN as mandated by State law. These were retrieved from the NHSN system database and duplicate patients were excluded. Cases were defined as SSI according to the NHSN definitions. The NHSN basic risk index and SIR was calculated for each procedure based on the following variables: wound class, age, anesthesia, ASA, duration of surgery, arthroplasty type, hospital beds, trauma, endoscope, and gender. Association between NHSN basic risk index and SSI were identified using the Cochran-Mantel-Haenszel test for categorical variables. Association between NHSN SIR and SSI were identified using the Wilcoxon two-sample test for continuous variables.

Results: A total of 1337 publicly reported cases were included (488 arthroplasty; 616 herniorrhaphy; 233 hysterectomy), of which 31 were defined as SSI (13 arthroplasty; 8 herniorrhaphy; 10 hysterectomy). The NHSN basic risk index did not correlate with SSI when assessed in aggregate (p = 0.19), after arthroplasty (p = 0.66), or after hysterectomy (p = 0.74); only SSI after herniorrhaphy showed a correlation with the basic risk index (p = 0.02). In contrast, the SIR correlated with SSI when assessed in aggregate (p < 0.001) and after herniorrhaphy (p = 0.001) and hysterectomy (p = 0.03); SIR did not correlate with SSI after arthroplasty (p = 0.09).

Conclusion: Neither basic risk index nor SIR currently used for risk stratification in publicly reported procedures in CO adequately predicted SSI at this institution. SIR predicted infection risk better than the basic risk index for the procedures that we examined. The SIR uses logistic regression modeling and takes into account more variables and procedure-specific risk factors. We hypothesize that improved risk adjustment is due to consideration of these extra risk factors. Given its inferior risk stratification performance, the basic risk index should no longer be used when analyzing SSI data for public reporting. Although superior, the SIR still is not broadly applicable to all procedures and settings, and further refinement is needed before it can be promoted as a tool to compare risk stratified rates of SSI from arthroplasty procedures at safety net institutions.

O55. The Effect of Prolonged Cortisol Excess on Systemic and Cellular Responses to Endotoxin

Introduction: Prolonged illnesses are often associated with persistent cortisol levels. Thus, the aim of the current study was to create a model of chronic stress through prolonged cortisol infusion. Using this model, we sought to determine how persistent endocrine stress will impact the host inflammatory response to an acute inflammatory challenge induced by in vivo endotoxin (LPS) administration. We hypothesized that an extended period of prior cortisol excess would modulate both systemic and cellular markers of the host inflammatory response.

Methods: Healthy volunteers were randomized to receive either cortisol (3μg/kg/mim) (n = 3) or saline (n = 4) infusion for 24 h prior to receiving a bolus dose of LPS (1 ng/kg, CC-RE, Lot #2). Heart rate variability (HRV) data, vital signs, and blood samples were obtained at multiple time points before and after LPS infusion. HRV metrics were serially determined with standard technologies (CardioPro). We determined peripheral blood leukocytes' (PBL) energy levels (ATP) using lysed cells. Statistical analysis was by repeated measures ANOVA.

Results: Prior to endotoxin challenge, subjects receiving cortisol exhibited diminished nocturnal variability (p < 0.05) of an HRV metric of overall host adaptability, SDNN (standard deviation of normal-to-normal intervals), without an apparent influence on efferent vagal activity, RMSSD (square root of the mean squared differences of successive normal-to-normal intervals). No changes in cellular bioenergetics were observed in subjects infused with cortisol alone for 24 hours. Following endotoxin challenge, sustained prior cortisol infusion significantly attenuated the endotoxin-induced decrease of SDNN (p < 0.05) and RMSSD (p < 0.02). In vivo endotoxin elicited a transient decrease in ATP levels and a parallel transient increase in autophagy. These perturbations were not observed in PBL from subjects infused with cortisol prior to in vivo endotoxin challenge.

Conclusion: Despite similar systemic manifestations in vital signs exhibited by both groups following endotoxin, cortisol did attenuate the endotoxin-induced dysregulation of autonomic function. In addition, a prolonged period of hyper-cortisolemia appeared to influence the normal circadian cycling of some HRV parameters. The effect of an antecedent cortisol infusion on cellular bioenergetics was profound, in that it prevented the endotoxin-induced decline in ATP levels, as well as the induction of autophagy in PBL. It remains to be determined whether the net effect of the metabolic responses is protective or detrimental to the host.

O56. Acute Kidney Injury is Associated with Nosocomial Infections and Surgical Site Infections after Trauma

Background: Infections and sepsis are important determinants of mortality in patients with trauma. Mild and moderate acute kidney injury (AKI) occurs in 26% of trauma patients and is associated with increased mortality. We studied the effect of posttraumatic AKI on the occurrence of infectious complications after trauma.

Methods: We performed a secondary analysis of the “Inflammation and the Host Response to Injury” (GlueGrant) database to include adult blunt trauma patients without history of kidney disease. AKI was defined by the RIFLE (Risk, Injury, Failure, Loss, and End-stage Kidney) classification, which requires a 50% increase in sCr and stratifies patients into three severity stages: risk, injury, and failure. Association between all stages of AKI and infectious complications was analyzed using a multivariable logistic regression analysis.

Results: The prevalence of surgical site infection (SSI) and nosocomial infections (NCI) after trauma was 16% and 49%, respectively. Patients with AKI had 2.3 times the odds of having SSI and 2.0 times the odds of having NCI compared to patients without AKI. Among NCI, the odds for pneumonia (1.96), bloodstream infections (3.07) and intravenous catheter-related infections (2.45) were higher for AKI patients. Multivariate logistic analysis revealed that male sex, AKI, hyperglycemia and transfusion of > 6 units of packed red blood cells (PRBC) in the first 24 hours were independently associated with the occurrence of NCI (c statistic 0.612). On the other hand, AKI and transfusion of > 6 units PRBC and African American ethnicity were associated with SSI (c statistic 0.628).

Conclusion: Post-traumatic AKI is an independent risk factor for infectious complications after trauma.

OPEN IN VIEWER

P1. Integrated Safety Summary of Ceftaroline Fosamil: Overall and for Patients with Complicated Skin and Skin Structure Infection (CSSSI)

Introduction: Ceftaroline fosamil (CPT) is a broad-spectrum, parenteral cephalosporin prodrug approved for treatment of acute bacterial skin and skin structure infections (ABSSSI) and moderate to severe community-acquired bacterial pneumonia (CABP). We report an integrated safety summary of 4 CPT phase III trials, with a separate safety analysis of the 2 CANVAS studies.

Methods: In the phase III CANVAS and FOCUS trials, adult patients with complicated skin and skin structure infections and community acquired pneumonia, respectively, were randomized (1:1) to receive either 600 mg CPT q12h or 1 g vancomycin plus 1 g aztreonam (V/A) q12h × 5–14 days (CANVAS) or ceftriaxone 1 g q24h × 5–7 days (FOCUS, plus clarithromycin 500 mg q12h × 1 day in one trial). Treatment-emergent adverse events (TEAEs) occurring from first dose until the test-of-cure visit and serious adverse events (SAEs) occurring up to the late follow-up visit or within 30 days after the last dose were recorded. Laboratory tests were evaluated.

Results: The mean ± SD durations of treatment in the CANVAS trials were 8.4 ± 3.1 days (CPT) and 8.4 ± 3.3 days (V/A). Incidences of TEAEs in CANVAS 1 and 2 combined were similar to those observed in the integrated analysis (Table). Common TEAEs were similar between the CANVAS trials and integrated analysis, except for higher incidences of pruritis, pruritis generalized, and rash in both treatment groups of the CANVAS trials. The majority of TEAEs in both CPT and V/A groups were mild or moderate, and unrelated to study drug. There were no clinically meaningful differences in laboratory test results between the CANVAS trials and the integrated analysis or treatment groups, except for higher incidence of direct Coombs' seroconversion in the CPT group. There was no clinical evidence of hemolysis in any treatment group. SAEs and deaths were lower in the CANVAS trials. In the FOCUS trials, 1 death in each group was possibly related to study drug. No deaths in CANVAS were related to study drug.

Conclusion: This integrated safety summary of phase III clinical trial data indicates that CPT is well tolerated with a safety profile similar to other cephalosporins.

Carlisle Erica M. MD Zaborina Olga PhD Liu Donald MD, PhD Alverdy John MD University of Chicago Medical Center

P2. A Novel Model to Study the Combined Effects of Microbes and Oxidants on the Development of Necrotizing Enterocolitis

Introduction: Necrotizing enterocolitis (NEC) affects 5–10% of preterm infants born weighing less than 1500 g and occasionally critically ill adults. Although NEC pathogenesis is multi-factorial there appears to be a correlation between oxidative stress and the development of NEC that involves bacteria and their exoproducts. While traditional animal models of NEC can create histologic mucosal necrosis that appears to mimic “pre-NEC,” it is rare to see full thickness transmural necrosis in animals as seen intra-operatively in humans. Here we describe a new model of full thickness intestinal necrosis following intraluminal administration of the endogenous oxidant monochloramine into isolated loops of mouse intestine. The aim of this study was to determine what role, if any bacteria have in this model.

Methods: 3wk old C57/Bl6 mice underwent creation of an intestinal loop via suture to isolate a 4 cm segment of ileum while preserving its blood supply. 200μl of the monocholoramine derivative Chloramine T (0.5% v/w) or 200μl of normal saline was injected into the loop. To expose mice to pathogens that might colonize premature babies, separated groups of mice underwent the loop prep followed by simultaneous intracecal injection of 100μl of Pseudomonas aeruginosa or 100μl of normal saline as a control. The isolated intestine was returned to the abdominal cavity and the incision closed. Mice were sacrificed 24 h later, the degree of necrosis assessed and then intestinal segments were harvested and analyzed histologically and microbiologically.

Results: The majority of mice survived to the time of sacrifice (saline 78%; 63% chloramine). Of animals that survived to 24 h, no gross necrosis was observed in the saline group, but 59% of animals in the Chloramine T group had gross, transmural necrosis (p < 0.001). Relative bacterial species abundance was calculated for the following groups: non-instrumented controls, saline injected isolated loops, Chloramine injected loops WITHOUT necrosis, and Chloramine injected loops WITH necrosis (Figure 1). Lactobacillus was present in 60% of non-instrumented controls but completely absent in all animals who underwent isolated loop construction. Methicillin-sensitive Staphylococcus aureus, Pseudomonas aeruginosa, and Enterococcus faecium were present only in animals that underwent isolated loop construction. Groups of animals that underwent cecal injection of P. aeruginosa following loop construction cultured positive for P. aeruginosa in the isolated loop despite suture ligation of proximal and distal loop ends suggesting endogenous homing of P. aeruginosa to the isolated loop.

OPEN IN VIEWER

Conclusion: Intraluminal chloramine causes full thickness intestinal necrosis in 3wk old mice. Surgical injury results in loss of protective lactobacillus species. Although bacterial quantity and species were not determinants of the development of necrosis, this model will allow for more in-depth metagenomic interrogation of host pathogen signaling within a confined space that consistently produces transmural necrosis similar to that observed during the human condition.

Buchanan Patrick J BS Kim Tad MD Lottenburg Larry MD Efron Philip MD Ang Darwin MD, PhD, MPH University of Florida

P4. Necrotizing Soft Tissue Infection Secondary to Candida Albicans in an Immunocompetent Patient

Background: Necrotizing soft tissue infections secondary to a Candidal infection are exceedingly rare with only a handful of case reports published. Patients at increased risk of developing a fungal induced necrotizing fasciitis are those who are immunocompromised, have poorly controlled diabetes, or have peripheral vascular disease. Physical findings of this disease include fever, erythema and necrosis. We present a case of an immunocompetent patient who did not exhibit these risk factors.

Case Presentation: 55 year-old Caucasian male admitted for thoracoabdominal injury secondary to gun shot wounds. Left tube thoracostomy was performed for pneumothorax and a left subclavian 9 French non antibiotic coated introducer was placed. Exploratory laparotomy revealed penetrating injuries to the greater curvature of the stomach, neck of the pancreas, and right kidney. Wedge gastrectomy, distal pancreatectomy, repair of the superior mesenteric artery, splenectomy, and open abdomen vacuum dressing were performed. The abdomen was closed following two second-look laparotomies. Four days following closure, the patient developed significant leukocytosis; peak white blood count of 44. Blood cultures were positive for Candida tropicalis and glabrata; systemic high dose diflucan was administered. The left subclavian central venous line was withdrawn and gross purulence was noted at the insertion site which cultured Candida albicans. The erythema, and induration did not resolve despite antifungal therapy and local drainage (Fig 1). The patient underwent emergent operative exploration for deep space infection of the neck. The neck was widely debrided from the deltoid muscle laterally, pectoralis major inferiorly, sternocleidomastoid muscle superiorly, and sternoclavicular joint medially. Pathologic analysis of debrided tissue revealed acute cellulitis and myositis with numerous fungi identified on a GMS stain. Diflucan was discontinued and anidulafungin 100 mg IV q24 h was started. Subsequent debridements were performed and pathologic analysis of debrided tissue revealed necrotic granulation tissue without fungal organisms. The patient was eventually discharged home with vacuum dressing.

OPEN IN VIEWER

Fig. 1.

Multiseparated absence of left lower neck and left supraclavicular region.

Discussion: Previous case studies have shown that poorly controlled diabetes is a risk factor in immunocompetent patients for the development of Candidal necrotizing fasciitis. However, this patient did not have diabetes and his blood glucose was well-controlled. The source of Candida was likely secondary to the gastric injury with seeding into the blood stream. He was on postoperative antibiotics and antifungal for 48 hours. This case highlights a new patient scenario for necrotizing fungal infection.

Gómez Gonzalo MD Arroyave María C MD Penagos Diego MD García María C MD Martínez Juan D MD Matallana Rogelio MD Vanegas Felipe MD Mendoza María C MD Díaz Sergio MD CES University, Medellin, Colombia P5. Video-Assisted Single Port Appendectomy: Zero Surgical Site Infections

Background: In 1980 Kurt Semm reported the first laparoscopic appendectomy in Switzerland. This procedure was performed with 3 entry ports. Since then, minimally invasive surgery has presented constant development. Today, the single port technique is rapidly evolving. Relation between appendectomy and surgical site infection (SSI) is well known in literature. SSI rate is lower than 10% for non-perforated appendicitis, 15–25% when perforated appendix is found and higher than 35% in patients with peritonitis. Laparoscopic approach shows significant reduction in wound infection in several surgical entities. This study presents a series of cases of single port appendectomy (SPA) with no SSI.

Methods: This was a prospective study. We recruited a total of 68 patients undergoing laparoscopic SPA from June to December 2009. Our port is a surgical glove attached to a 360° wound retraction device (Alexis®). All programmed patients with clinical impression of acute appendicitis will have the appendix removed no matter the operative findings; even with macroscopic appendix normality and another confirmed cause of acute abdomen the appendix will be removed per protocol. Exclusion criteria were pregnancy, generalized peritonitis and contraindications for pneumoperitoneum. Patients were given Ampicillin/Sulbactam 3 gr. IV; antibiotic prophylaxis therapy was infused 30 minutes before the procedure. All patients were followed-up at 8th and 30th day after surgery.

Results: A total of 68 patients underwent single port laparoscopic appendectomy (46 female, 22). Mean age was 29.3 years old (range 11–65). Abdominal pain presentation was longer than 13 hours in 62 cases (91%). The procedure was completed through laparoscopy in 66 patients (97%). Conversion cause was generalized peritonitis in both cases. Operative findings reported 26 edematous appendix (38.2%), 25 fibrinopurulent (36.7%), 6 gangrenous (8.8%), 2 perforated (2.9%), 2 generalized peritonitis (2.9%) and 7 did not have appendicitis on intrasurgical impression (10.2%) We identified the cause of abdominal pain in 6 of these 7 normal appendix cases: hemoperitoneum was found in 3 patients, colon adhesions in 1 case, ruptured ovarian follicles and uterine adenomyosis, in one case respectively. Appendix was removed in all patients as considered per protocol. Postsurgical ileus was the only complication found in our series, one case with successful medical treatment. Microscopic analysis reported acute suppurated appendicitis in 64 cases (94.1%) 3 of these cases with associated peritonitis; one case of appendicular gangrene (1.5%) and 3 cases without appendicitis on histopathological analysis (4.4%). Mean operative time was 29.8 minutes (range 11–65). Postoperative length of stay was less than 24 hours (mean 20 hours). Seven cases were ambulatory treated (discharged within 8 hours after procedure). There were no surgical site infections. Follow up was complete in all patients.

Conclusion: Single port laparoscopic appendectomy is feasible and safe with a very low complication rate. The outstanding zero percent infection rates found in our patients should bring the interest of surgeons to this new procedure.

OPEN IN VIEWER

Amini Amir MD Baker Aaron MS, MD Tung Kelly MD Green Tajia BS Lim Debora BA, Cho Kiho DVM, PhD Greenhalgh David MD, FACS University of California–Davis Medical Center

P34. Three Single Nucleotide Polymorphisms of Human Glucocorticoid Receptor Result in Enhanced Activity

Introduction: The glucocorticoid receptor is integrally involved in the endogenous inflammatory response. Given this receptor's unique ability to modulate the immune response, steroids have become important therapy in diseases of inflammation. Prior studies have identified glucocorticoid receptor polymorphisms to alter glucocorticoid sensitivity in certain disease states, sometimes rendering steroid therapy ineffective.

Objective: Identify isoforms of the human glucocorticoid receptor (hGR) within a healthy human population, and determine if they contribute to variability in steroid response.

Methods: Through the survey of 97 healthy human adults, our lab identified a novel glucocorticoid isoform (hGR NS-1). This isoform has three unique single nucleotide polymorphisms (σNΠσ) and displays a hyperactive transactivation potential when compared to the reference National Center for Biotechnology Information human glucocorticoid receptor (hGR NCBI). In this study, we compared the functional activity of hGR NS-1 with that of an hGR derivative matching NCBI reference at baseline and after administration of exogenous hydrocortisone. These receptors were cloned into expression vectors and transfected into TSA201 cells. These cells were then τρɛατɛδ with 4 different graded concentrations of hydrocortisone; 0.000001μM, 0.00001μM, 0.001μM, and 1 uM. Luciferase activity was measured 24 hours later to measure transactivation potential. The enhanced activity of hGR NS-1 was also analyzed through cotransfection of this isoform with hGR NCBI and hGR DL-1. hGR DL-1 is a novel truncated and hypofunctional isoform also identified by our lab.

Results: At baseline, hGR NS-1 demonstrated a greater than 3-fold incr ease in functional activity when compared to hGR NCBI. In addition, the vehicle activity of hGR NS-1 was greater than hGR NCBI stimulated with hydrocortisone at all 4 doses (p < 0.0001). When treated with hydrocortisone doses of 0.000001μM, 0.00001μM, and 0.001μM, hGR NS-1 demonstrated significantly greater transactivation potential when compared to hGR NS-1 vehicle (p < 0.0001) (Figure 1). A peak response occu rred at a hydrocortisone concentration of 0.001μM. Interestingly, transactivation potential of hGR NS-1 decreased significantly compared to baseline activity at a dose of 1 μM (p < 0.001). When the hyperactive isoform hGR NS-1 was cotransfected with either hGR DL-1 or hGR NCBI, enhanced transactivation potential occurred in comparison to hGR NCBI (p < 0.01) or hGR DL-1 alone (p < 0.00001), but decreased when compared to hGR NS-1 alone (p < 0.001).

OPEN IN VIEWER

Figure 1.

Transactivation potentials of hGR NCBI and hGR NS-1 at baseline and after stimulation with hydrocortisone.

Conclusion: This experiment demonstrates the hyperactive function of a novel glucocorticoid receptor isoform at baseline and after exogenous hydrocortisone treatment. It also suggests that this response to hydrocortisone treatment is dose dependent. The difference in response to steroid stimulation between this isoform and reference hGR may help explain variability of response to steroid therapy in patients.

Williams Monica L BS Fried Aaron J Castle Shannon MD Grishin Anatoly PhD Ford Henri R MD, MHA Childrens Hospital Los Angeles P35. Diversity of Postnatal Microbiota in Experimental Necrotizing Enterocolitis

Background: Necrotizing enterocolitis (NEC) is one of the most devastating gastrointestinal disorders that affect 1-5% of all pre-mature infants. Although it is commonly believed that the development of NEC requires colonization of the GI tract with opportunistic pathogens, little is known about the microbial flora associated with NEC. The aim of this study was to characterize bacterial populations of the small intestine in the rat model of NEC.

Methods: Newborn Sprague-Dawley rats (Harlan) from the same litter were fed sterile formula or allowed to breast feed. On day 4 of life, 1 cm of terminal ileum was aseptically excised and flushed with 300 μl sterile PBS. Serially diluted samples were plated on blood agar. Bacterial colonies were counted and classified into groups according to their appearance, typical colony size, and type of hemolysis. The species of representative colonies were identified by sequencing the V6-V8 fragment of the 16S rRNA gene and searching the DNA database for matching sequences using the BLAST algorithm.

Results:

OPEN IN VIEWER

Conclusion: Despite common origin and environment, different animals show dramatic differences in the composition of culturable bacteria populating their small intestines during the early postnatal period. These differences may account for differential susceptibility to NEC.

Muir Mark MD Furmaga Wieslaw MD Mikhailov Valery Jonas Rachelle RN Louden Christopher MS Cohn Stephen MD Michalek Joel PhD University of Texas Health Science Center at San Antonio

P36. Heat Shock Protein 27 Associated with Poor Outcomes in Hemorrhagic Shock

Objectives: We sought to determine the relationship between serum levels of Heat Shock Protein 27 (HSP27) and clinical outcomes in trauma patients with hemorrhagic shock.

Methods: Serum was collected from a convenience sample of twenty subjects participating in a study of trauma patients with hemorrhagic shock. Samples were collected at admission and three hours later, and HSP27 levels were measured with ELISA by individuals blinded to subject identity and clinical course. Total serum protein levels were also measured, and corrected HSP27 levels were calculated as the ratio of HSP27 to total protein (pg HSP27/mg protein).

Results: Subjects who died (N = 5) had significantly higher initial HSP27 levels than those who did not (median [IQR]: 23.7 [16.7, 24.9] pg/mg vs. 4.2 [2.9, 14.3] pg/mg, p = 0.003). Subjects who developed multiple organ dysfunction syndrome (MODS; N = 2), defined as MODS score ≥ 6 during days 3-30, had significantly higher HSP27 levels at three hours than those who did not (median [IQR]: 14.1 [12.9, 15.3] pg/mg vs. 2.4 [1.2, 5.6] pg/mg, p = 0.026). Three-hour HSP27 values greater than 10 pg/mg appeared predictive (accuracy = 100%) of the development of MODS.

OPEN IN VIEWER

Conclusions: In trauma patients with hemorrhagic shock, elevated HSP27 levels were strongly predictive of the subsequent development of organ dysfunction and death.