nan Thickett

S96 Simvastatin as an adjuvant therapy for infection and sepsis–in-vitro and in-vivo studies suggest pre-emptive / early therapy in the elderly

Ageing is associated with increased episodes of sepsis and poorer outcomes. Statins are associated with improved outcomes during infection. We aimed to characterise the impact of age and acute severe infection on key neutrophil functions, assess whether physiologically relevant doses of simvastatin altered neutrophil functions and if benefits were seen, when during a septic episode statins could be utilised. Methods Neutrophils from healthy volunteers and patients with lower respiratory tract infections (LRTI), pneumonia and sepsis were assessed for migratory accuracy, phagocytosis and neutrophil extracellular trap production before and after in-vitro treatment with simvastatin. Healthy elderly volunteers received 80mg simvastatin or placebo in a cross over double-blind randomised controlled trial and neutrophil functions were assessed. Data presented is for migration. Results Neutrophils from healthy subjects (n = 70, aged 21–94) demonstrated preserved neutrophil movement) (R2 = -0.48, p < 0.0001) towards chemoattractants (data shown for IL-8). Neutrophil chemotaxis decreased after 60yrs (comparing <35 to >65yrs: mean difference (MD)1.25μm/min, p = 0.02). There was a progressive decrease in neutrophil chemotaxis in old patients with a LRTI, pneumonia and severe sepsis (MD compared to healthy control; LRTI (n = 10), 0.7μm/min, p = 0.04; pneumonia (n = 5), MD1.1μm/min, p = 0.02; sepsis (n = 22) MD1.6μm/min, p = 0.01) with “septic neutrophils” unable to mount targeted chemotaxis. Improvements to baseline were seen following recovery. In-vitro treatment of neutrophils from healthy older people with simvastatin (1µM) restored “old” neutrophil chemotaxis to that of “young” cells. Simvastatin also restored neutrophil migration from old patients with LRTI and pneumonia to baseline but not in patients with sepsis. Two weeks of oral simvastatin 80mg once daily therapy in healthy old volunteers (Age>65,n = 20) increased the accuracy of neutrophil migration (MD1.68μm/min, p = 0.02) replicating benchwork. Conclusions “Elderly” neutrophil function is compromised in health, and deteriorates during infective episodes, in accordance with the severity of the insult. Migratory accuracy can be improved with simvastatin therapy however neutrophil function in sepsis patients cannot be modulated during short term in-vitro therapy. Our data suggest statin therapy might be a preventative or an early adjuvant intervention rather than a treatment in established sepsis. We are testing whether simvastatin 80mg for seven days modifies neutrophil responses in elderly patients with pneumonia and sepsis (SNOOPI Trial). Abstract S96 Figure 1. Simavastatin 80mg once daily for 14 days improves directional migration (chemostaxis) of neutrophils from healthy elderly volunteers towards IL-8. *Student’s t-Test

T2 Vitamin D supplementation reduces perioperative systemic and alveolar inflammation in patients undergoing oesophagectomy: Results of the Vindaloo Trial

Vitamin D deficiency is associated with increased risk of ARDS post-oesophagectomy. We recruited patients to a double-blind, randomised controlled trial of high dose Vitamin D supplementation 3–14 days pre-oesophagectomy. 79 patients were randomised to receive placebo or 300,000 IU oral Vitamin D liquid 3–14 days prior to oesophagectomy. Blood samples were collected pre-dose, post-dose (pre-op) and post-op and analysed for 25-OH and 1,25-dOH Vitamin D, inflammatory cells and cytokines. Broncho-alveolar lavage fluid was collected at the end of the operation. PICCO biomarkers of alveolar capillary damage (EVLWI and PVPI) were measured pre- and post-op. Pre-operative supplementation with Vitamin D was well tolerated with no SUSARs and significantly increased circulating 25-OH and 1,25-OH Vitamin D (p < 0.0001). This was associated with reduced systemic inflammation (IL-6 (p = 0.02) and IL-8 (p = 0.002)) and an increase in circulatory Treg (p = 0.027). Changes in PICCO biomarkers were lower in supplemented patients suggesting lower perioperative alveolar oedema (EVLWI p = 0.05, PVPI p = 0.04). This did not result in a significant difference in oxygenation at 24 h. Post-op, systemic and alveolar alarmin (IL-1B, IL-6, IL-8) response was similar in treated and untreated patients but the systemic release of IL-1ra (p = 0.046), sTNFR-1 (p = 0.05) and s-TNFR-2 (p = 0.02) were elevated in treated patients. There was also evidence of decreased alveolar macrophage efferocytosis in patients with Vitamin D deficiency (p = 0.003). Clinical diagnoses of ARDS were significantly lower in this cohort than in previous cohorts, but the study was not powered to detect that outcome. Mortality post-operative was not significantly different at 30 or 90 days but there is a significant difference after 300 days of follow-up (placebo 33% mortality, Vitamin D 8% mortality p = 0.033). In conclusion, vitamin D supplementation was a safe, well tolerated pre-operative intervention that reduced systemic inflammation and biomarkers of alveolar oedema. With evidence of enhanced anti-inflammatory mechanism that may have influenced longer term post-operative survival, Vitamin D deficiency should be identified and treated in patients at risk of ARDS.

S98 Vitamin D deficiency increases bacterial load in a murine model of sepsis-induced lung injury

Introduction We have previously shown that patients with Acute Lung Injury (ALI) are severely vitamin D deficient. Several studies have reported a high prevalence of vitamin D deficiency in critically ill patients with sepsis, associated with increased morbidity and mortality but whether this is cause or effect is unknown. Bacteraemic sepsis is more common in the winter months when vitamin D levels are lowest. The purpose of this study was to investigate the local and systemic effects of vitamin D deficiency in a murine model of sepsis-induced lung injury where we can predictably time the initiating insult. Methods We fed 8 wild-type C57Bl/6 mice a diet completely devoid of vitamin D for 6 weeks to induce severe vitamin D deficiency (9 nmol/l) and compared to 7 mice fed a vitamin D sufficient diet (42 nmol/l). Caecal ligation and puncture (CLP) was used to establish sepsis. Animals were culled 16h after CLP and blood, peritoneal lavage fluid (PLF) and bronchoalveloar lavage fluid (BALF) were collected. Cell infiltrates were assessed by flow cytometry. Fluid protein levels were measured and tissue protein permeability index (PPI) was calculated as the ratio between fluid and serum protein. Bacterial load was evaluated as colony-forming units (CFU) after 24h incubation on appropriate media. Results Vitamin D deficient mice had increased bacterial load in BALF, blood and PLF compared to dietary sufficient mice. BALF protein permeability index was higher in deficient compared to sufficient mice but there was no difference in cell numbers recruited to the lung. PLF protein permeability index was also increased in the deficient group compared to sufficient mice with an associated significant increase in neutrophils recruited to the peritoneum. (See Table 1) Abstract S98 Table 1. Differences between dietary deficient and sufficient mice post CLP induced sepsis. Data is expressed as median values. Sufficient (n = 7) Deficient (n = 8) p-value Bacteria (CFU x103) BALFBloodPLF 0.510.288.22 2.5166.1336.6 0.0380.0190.005 PPI (x1000) BALFPLF 1.8227.2 3.3046.9 0.00030.05 Neutrophil Number BALFPLF (x106) 64.12.11 27.94.60 0.1830.04 Conclusion Vitamin D deficiency significantly increases the bacterial load both systemically, locally and within the lung in a murine model of peritonitis. This is associated with an increase in tissue permeability locally and within the lung. These data support pre-existing vitamin D deficiency as a determinant of the severity of bacteraemic sepsis and may account for some of the seasonal variations observed in the incidence of sepsis.

S122 Effects of vaped e-cigarette liquid condensate upon human alveolar macrophage function. to vape or not to vape that is the question?

Introduction and objectives Electronic cigarette usage or “vaping” has risen exponentially in recent years in smokers and ex-smokers. Published data suggests that vaping e-cigarette liquid (ECL) may not be as benign as propounded by e-cigarette companies which are increasingly owned by “big tobacco”. Much of the current literature has focused on the effect of non-vaporised ECL – such studies do not fully reflect the exposure of the user, as the process of vaping causes chemical changes in ECL. To investigate the effect of unvaped ECL and vaped e-cig condensate (ECVC) using our novel system, with and without nicotine, on alveolar macrophage (AM) viability and immune responses. Methods We developed a novel method to produce ECVC to allow direct comparison with unvaped ECL. Nicotine concentration as assessed by GFID was 31 mg/ml in ECL and 26 mg/ml in ECVC. AMs were obtained from lung resection tissue and treated with ECVC/ECL ± nicotine. Cell viability was assessed by cell titre aqueous assay, apoptosis, necrosis and markers of macrophage phenotype (CD68, CD80, CD163, CD206) were assessed by flow cytometry. IL-8 release by AMs was assessed by ELISA. Results AM culture with ECL or ECVC resulted in dose dependent reduction in cell viability. ECVC was cytotoxic at lower concentrations than ECL (0.8% ECVC vs 5 %ECL, n = 6). 24 hour culture with 1% ECVC resulted in a 5fold increase in AM apoptosis and 2 fold increase in necrosis compared with 1%ECL (p = 0.079, n = 5). Nicotine containing ECVC caused more apoptosis vs nicotine free ECVC (27.2% vs 13.4%, (p = 0.0079,n = 4). Culture with 0.6%ECVC significantly increased supernatant levels of IL-8 compared with 1% ECL (p = 0.015, n = 4). ECVC was also found to affect macrophage phenotype, showing both nicotine dependent/independent regulations of markers of macrophage m1/m2 polarisation (CD80 p = 0.0357, CD163 p = 0.0179, CD206 p = 0.0357, n = 6). Conclusions Our novel system creates ECVC which is sterile, minimises loss of nicotine and prevents dilution of the vapour. Vaped E-cigarette condensate is significantly more toxic to AMs than non-vaped e cigarette liquid. Furthermore, ECVC with nicotine is significantly more toxic than ECVC without Nicotine. Effects shown on inflammatory cytokine production and markers of macrophage polarisation indicate both nicotine dependent and independent effects of ECVC on alveolar macrophages.

S15 Changes in perioperative ards with time: a comparison of two trials

Introduction and objective The BALTI-Prevention Trial1 translational sub-study (recruitment completed in 2011) and VINDALOO Trial2 (recruitment completed in 2015) both used oesophagectomy as a model for investigation of the pharmacological prevention of ARDS. The VINDALOO trial showed a lower ARDS incidence independent of the agents evaluated. Our objective was to characterise this difference. Methods Databases from both trials were available and additional information was obtained retrospectively from hospital records. Analysis was performed using appropriate statistical tests. Results There was a substantially higher ARDS incidence in BALTI-P compared to VINDALOO (RR 1.17 (1.08–1.28)) (Figure 1). There were more never (21 versus 7) and fewer current (12 versus 25) smokers in the VINDALOO group (p = 0.003). Perioperative risk scores did not differ between groups (P-POSSUM mortality p = 0.98, P-POSSUM morbidity p = 0.87, O-POSSUM p = 0.58), nor did use of neoadjuvant chemotherapy (p = 1.0). The incidence of serious complications (Clavien-Dindo score of four or more), was not different between the two groups (RR 1.15 (0.94–1.40)). Univariate analysis of the combined groups showed risk factors for ARDS not using regional anaesthesia (OR 3.2 (1.0–9.8) p = 0.038) and current smoking (OR 3.0 (1.1–8.3) p = 0.038). Abstract S15 Figure 1 ARDS incidence was higher in the BALTI-Prevention compared to the VINDALOO trial (p = 0.004) Conclusion Smoking is an environmental risk factor for ARDS and changing patterns in smoking behaviour appear important in these perioperative cohorts. Other studies have demonstrated that aggregates of small but important changes in care can cumulatively make a significant difference to risk of nosocomial ARDS and our data suggests that this may have resulted in a reduction in perioperative ARDS in this group. Our study has demonstrated that ARDS incidence post oesophagectomy has fallen. In future, trials utilising oesophagectomy as a model of ARDS should consider selection of a defined higher-risk sub-set of patients. References Perkins G, et al. The beta agonist lung injury trial prevention. A randomised controlled trial. Am J Respir Crit Care Med 2014;189(6):674–683. Parekh D, et al. Vitamin D to prevent acute lung injury following oesophagectomy (VINDALOO): study protocol for a randomised placebo controlled trial. Trials 2013;14(100). doi:10.1186/1745-6215-1114-1100.

S67 Vitamin D deficiency drives pulmonary inflammation in a human model of the acute respiratory distress syndrome induced by inhaled lipopolysaccharide in healthy volunteers

The acute respiratory distress syndrome (ARDS) is characterised by exaggerated alveolar inflammation. Vitamin D deficiency in an LPS induced murine model of ARDS results in exaggerated alveolar inflammation. However the role of vitamin D deficiency in pulmonary inflammation in humans is unclear. We hypothesised that in healthy volunteers with vitamin D deficiency, pulmonary inflammation would be increased following LPS inhalation. Methods Healthy volunteers inhaled 50 micrograms of LPS and six hours later underwent bronchoalveolar lavage for measurement of cytokines. Plasma was collected at baseline and one day post LPS inhalation for measurement of vitamin D. Results 28 participants were included. The mean age of volunteers was 26.2 +/- 5.5 years. All 28 patients were vitamin D deficient (plasma levels <50 nmol/l), with 89% (25/28) patients having severe vitamin D deficiency (<25 nmol/l). Vitamin D levels were significantly higher after LPS inhalation (p < 0.002). Levels of IL-1β in BALF were significantly higher in those with severe deficiency than those with mild/moderate deficiency (Figure 1; p = 0.04). Levels of IL-6, IL-8 or TNF-α did not differ between groups.Abstract S67 Figure 1 Bronchoalveolar lavage fluid (BALF) levels of IL-1 beta were significantly elevated in volunteers with severe plasma vitamin D deficiency (<25 nmol/l) compared to those with mild or moderate deficiency (25–50 nmol/l) Conclusions Vitamin D deficiency was highly prevalent in this population of healthy volunteers. The rise in vitamin D levels post LPS exposure may represent mobilisation of vitamin D from fat stores during inflammation though vitamin D metabolism and kinetics are complex and may differ in healthy volunteers and the critically ill. Severe deficiency correlated with increased alveolar inflammation.

P164 Smoking at the time of curative-intent lung cancer surgery increases perioperative complications: is there a role for electronic cigarettes?

Introduction Smoking is a risk factor for postoperative pulmonary complications (PPCs) following curative-intent surgery for lung cancer. Risk modification is via smoking cessation; the role that electronic cigarettes (e-cigarettes) have in preoperative tobacco replacement is a debated topic. Aims Investigate the impact of smoking on postoperative outcome including long-term survival. Assess current smoking habits and attitudes towards preoperative smoking cessation, with emphasis on e-cigarette use. Methods A prospective observational study was carried out on all patients following curative-intent lung cancer resection in a regional thoracic centre over 4 years. Preoperative smoking status was self-reported by all patients. PPCs were assessed daily in hospital using the Melbourne group scale.1 Other data included patient demographics, hospital length of stay (LOS), intensive treatment unit (ITU) admission and mortality data. To assess smoking habits, a questionnaire was given to 105 patients attending the preoperative assessment unit. Results Of 460 patients, 24% were current smokers, 12% ex-smokers 6 weeks duration, and 11% never smoked Compared to never smokers, current smokers had significantly longer hospital LOS in days (9, CI 7–11 vs. 6, CI 4–8; p < 0.001), higher frequency of PPCs (22% vs 2%, p = 0.001) and ITU admissions (14% vs. 0%; p < 0.005). Compared to never smokers, the trend was for reduced survival in current smokers from 1–3 years, but the survival lines converged after this (median follow-up 30 vs. 31 months; p = 0.31). The questionnaire found 24/105 patients were smokers, of these 80% patients had previously tried to quit but only 38% had been specifically approached by health-care professionals about smoking cessation. When asked if they would consider stopping smoking immediately if supplied an e-cigarette, 54% said yes. Conclusions Preoperatively, 1 in 4 patients continue to smoke; the majority have attempted to quit and failed. Current smokers have higher postoperative morbidity with no significant survival difference within our follow-up period. Current methods of preoperative smoking cessation in this population are ineffective; patients appear willing to use e-cigarettes. Further research in this field is urgently needed. Reference 1 Agostini P, et al. Thorax 2010;65:815–18

P147 Current smokers face increased risk of Acute Lung Injury post oesophagectomy compared to former smokers- implications for therapy and trial design?

Introduction Patients undergoing oesophagectomy have ~25% risk of developing post-operative Acute Lung Injury (ALI). A recent meta-analysis showed that smoking cessation prior to an operation reduces the risk of respiratory complications [1]. We hypothesised that current smokers would have an increased risk of ALI post-oesophagectomy compared with former smokers. Methods We analysed data from 14 current smokers and 37 former smokers recruited to the translational sub-study of the BALTI prevention trial. Extravascular Lung Water Index (EVLWI) and Pulmonary Vascular Permeability Index (PVPI) were measured using PICCO. Plasma and Bronchoalveolar lavage fluid (BALF) cytokine levels were measured by ELISA. Results Current smokers were significantly younger and had a lower BMI than former smokers. Current smokers were more likely to develop post-operative ALI which required ventilation but there was no difference in the rate of respiratory infection. PVPI was significantly higher immediately post op and on day 1 post op and EVLWI was significantly higher on day 1 post op in current smokers (see table). Plasma levels of FAS ligand were significantly lower in current smokers pre-op, post-op and on day 1 post-op. Pre-op plasma levels of sICAM1 were significantly higher in current smokers but there was no difference in post-op levels. Plasma levels of IL-17 were lower pre-op, post-op and on day 1, although only the post-op difference reached significance. In BAL, levels of VEGF were significantly lower and levels of IL1ra and TNFR1 were significantly higher in current smokers (see table). Conclusion Current smokers have a higher risk of Acute Lung Injury following oesophagectomy than former smokers. This finding is supported by higher post-operative levels of extravascular lung water and pulmonary vascular permeability and differences in both plasma and BAL cytokines in current smokers. Our data highlights the importance of encouraging patients to stop smoking prior to major surgery as well as the need to control for smoking history in trials aiming to ameliorate lung injury in this patient group. References Wong, J. et al. Short-term preoperative smoking cessation and postoperative complications: a systematic review and meta-analysis. Can J Anaesth, 2012. 59(3): 268–79 Abstract P147 Table 1. Differences between current smokers and former smokers undergoing oesophagectomy. Data is expressed as median values except where specified. Current Smokers (n = 14) Former Smokers (n = 37) P-value Age (Years) 48 65 <0.001 BMI (kg/m2) 22 26 0.002 Developed ALI - n (%) 6 (43) 6 (16) 0.045 Developed respiratory infection - n (%) 5 (35) 12 (32) 0.824 EVLWI (ml/kg) Pre-op 8.3 7.8 0.520 Post-op 11.0 9.3 0.191 Day 1 9.8 7.9 0.040 PVPI Pre-op 2.13 1.74 0.151 Post-op 2.43 1.97 0.050 Day 1 2.09 1.69 0.008 Fas Ligand(pg/ml) Pre-op 21.85 36.78 0.001 Post-op 17.62 30.73 0.001 Day 1 9.05 18.38 <0.001 sICAM-1 (ng/ml) Pre-op 103.52 56.95 0.005 Post-op 51.38 43.01 0.199 Day 1 100.38 88.00 0.226 IL-17A (pg/ml) Pre-op 8.42 55.66 0.088 Post-op 2.64 36.84 0.043 Day 1 8.26 83.57 0.084 VEGF (pg/ml) BAL 94.05 153.40 0.016 IL1ra (pg/ml) BAL 50.99 9.83 0.007 sTNFR1 (pg/ml) BAL 270.5 168.5 0.042