scale to investigate whether this has a definite improvement in patient experience and management. Disclosure of Interest None Declared PTU-161 SYSTEMATIC REVIEW AND META-ANALYSIS ON COMPLICATIONS FOLLOWING OESOPHAGEAL DILATATION FOR BENIGN OESOPHAGEAL STRICTURES- PRELIMINARY RESULTS SV Venkatachalapathy * , N Burr, V Subramanian, S Everett. Gastroenterology, Leeds Teaching Hospitals NHS Trust, Leeds, UK 10.1136/gutjnl-2016-312388.246 Introduction The incidence of benign oesophageal strictures is 0.5% in patients with dyspeptic symptoms. 1 It affects the quality of life by causing dysphagia, regurgitation and in severe cases weight loss. The first line of management is bal- loon or bougie dilatation. There is ambiguity about the com- plication rates associated with this procedure. We therefore did a systematic review and metanalysis on complications (bleeding and perforation) associated with endoscopic dilatation. Methods We searched several electronic databases including Pubmed for full journal articles published after 1990 reporting on the use of endoscopic dilatation using bougies or balloons in the treatment of beingn oesophageal strictures. We hand searched the reference lists of all retrieved articles. Cohort or prospective studies involving 10 or more adult patients were included in the analysis. Studies on corrosive/caustic strictures and radiological non-endoscopy guided diltations were excluded. We calculated the pooled proportion of patients who had a complication (perforation or bleed) to therapy in the selected studies. Heterogeneity between the studies was assessed using the I 2 statistic. Results Our search identified 32 studies that were included in the final analysis (26 cohort studies and 6 randomised control trials). There were 11 studies that reported on balloon, 8 on bougie and 13 studies reported on both balloon and bougie dilatations. There were 18104 patients, 7195 balloon dilations and 15,936 bougie dilations. There were 7711 (42.5%) males and 7305 (40.3%) females. The pooled rate of perforation was 0.5% (95% CI, 0.3–0.8, I 2 –7.5%) and 0.3% (95% CI, 0.2–0.5, I 2 –41.1%) for balloon and bougie respectively. The rate of bleeding was 0.6% (95% CI 0.4–1.1, I 2 –17.1%), and 0.3% (95% CI, 0.2–0.8, I 2 –60.6%) for balloon and bougie dilatations respectively. Conclusion This large meta analysis on 18104 patients shows that the risk of perforation and bleeding is low and compara- ble in both endoscopic guided balloon and bougie dilatations. The rates are lower than the commonly accepted figure of 1% and should be reassuring to both patients and endoscopists. REFERENCE 1 Breslin NP, Thomson ABR, Bailey RJ, et al. Gastric cancer and other en- doscopic diagnoses in patients with benign dyspepsia. Gut 2000;46. Disclosure of Interest None Declared PWE-001 TRANSCUTANEOUS CERVICAL VAGAL NERVE STIMULATION EXERTS AN ANTI-TNF-ALPHA EFFECT IN HEALTHY HUMANS 1 C Brock, 2 B Brock, 2 HJ Moller, 1 AM Drewes, 1,3 4 AD Farmer * . 1 MechSense, Aalborg University Hospital, Aalborg; 2 Department of Clinical Biochemistry, Aarhus University Hospitals, Aarhus, Denmark; 3 Department of Gastroenterology, University Hospitals of North Midlands, Stoke on Trent; 4 Wingate Institute of Neurogastroenterology, Barts & The London School of Medicine, Whitechapel, UK 10.1136/gutjnl-2016-312388.247 Introduction The vagus nerve is the main neural substrate of the parasympathetic nervous system and has a role in modu- lating inflammation through the cholinergic anti-inflammatory pathway, via inhibition of the production of pro-inflammatory cytokines at the level of both the spleen and the intestinal muscularis. Animal studies have demonstrated that in vagotom- ised mice, electrical vagal nerve stimulation (VNS), applied dis- tal to the severance, ameliorates the pro-inflammatory cytokine response to lipopolysaccharide. The cervical vagus nerve is located directly under the skin, making it a suitable target for transcutaneous non-invasive VNS (n-VNS). In this pilot study, we sought to evaluate the effect of non- n-VNS on pro-inflammatory cytokines and cardiometrically derived autonomic parameters in humans. Methods In healthy volunteers, heart rate, blood pressure and validated sympathetic, (cardiac sympathetic index) and vagal, (cardiac vagal tone (CVT)) indices were measured directly before, and 24 hours after, 2 minutes of n-VNS applied bilat- erally. Venous blood was also drawn and assayed for pro- inflammatory cytokines (tumour necrosis factor-alpha (TNF-a) and interferon-gamma (IFN-g)) and the anti-inflammatory cytokine (interleukin-10) directly prior to, and 24 hours after n-VNS. Results 20 healthy volunteers (13 females, median age 34 years, range 23–55) all tolerated the n-VNS. Table 1 details the changes in the recorded parameters. There was a negative correlation between change in CVT and change TNF-a (r s = 0.45, p < 0.05). Abstract PWE-001 Table 1 Prior to n-VNS (mean and standard deviation) 24 hours post n-VNS (mean and standard deviation) P value Heart rate (bpm) 72.9 ± 9.1 72.5 ± 7.8 0.78 Systolic blood pressure (mmHg) 129.9 ± 14.3 130.2 ± 17.2 0.88 Diastolic blood pressure (mmHg) 77.7 ± 7.5 77.5 ± 9.9 0.85 Cardiac sympathetic index 2.5 ± 0.5 2.6 ± 0.6 0.22 Cardiac vagal tone (linear vagal scale) 8.4 ± 4.5 9.9 ± 5.5 0.02 TNF-a (pg/ml) 2.0 ± 0.4 1.8 ± 0.5 0.03 IFN-g (pg/ml) 5.5 ± 4.5 5.2 ± 3.9 0.54 IL-10 (pg/mL) 0.54 ± 0.9 0.59 ± 1 0.33 Conclusion These results, for the first time in humans, provide preliminary evidence for an anti-TNF-a in response to n-VNS, potentially mediated by an increase cardiac vagal tone. These data, warrant further investigation in immune mediated Abstracts Gut 2016;65(1):A1–A310 A137