Introduction Eosinophilic oesophagitis (EoE) is a chronic condition of the oesophagus characterised by a dense eosinophilic infiltrate defined as >15 eosinophils/high power field (eos/hpf). The aim of this study was to capture the prevalence of EoE in a tertiary referral centre in London, to identify factors associated with a positive diagnosis and interrogate optimal response to therapy.

Methods From February 2013 to November 2015, all patients presenting with solid food dysphagia to University College Hospital had a high resolution white light endoscopy. Those with cancer, achalasia, postoperative stricture or endotherapy for Barrett’s were excluded. Endoscopy, histology and clinical data were collected. >15 oes/hpf were defined as positive for EoE. A separate histopathology search identified patients with >15 eos/hpf and no dysphagia. A prospective follow-up was conducted in focusing on therapy and response in those with >15 eos/hpf.

Results Out of the 1566 patients with dysphagia, 524 were excluded for reasons stated. 736/1042 (71%) had oesophageal biopsies. Of those, 67 (9.1%) had more than 15 eos/hpf. Another 14 patients with >15 eos/hpf with symptoms other than dysphagia were identified from histology records, making the total number with eosiniphilia 81. The mean number of biopsies taken in those with >15 eos/hpf (6.3) was greater than those with <15 eos/hpf (5.1; p = 0.003). EoE patients were more likely to be male (70%) and younger (43±16 years) compared to nonEoE (40% male, p < 0.0001; 59±16 years, p < 0.0001). Typical endoscopic features were found in 39 (48%) EoE patients; rings/furrows in 26 (32%) and strictures in 15 (18%). 42/81 (52%) were treated with PPIs only of which 19 (45%) clinically responded. 18 (22%) patients had both PPI and topical steroids (12 had steroids after PPI failure) while 8/81 (10%) had steroids only. Clinically 14/26 (54%) responded optimally to topical steroids, 13 of which had dysphagia. Overall, response to steroids occurred in those with a higher eosinophilia (53 vs 24, p = 0.004) and all 9 with ≥40 eos/hpf had a complete response. Furthermore, typical EoE findings at endoscopy was more likely to be associated with a poor response to PPIs (p < 0.0001).

Conclusion A higher number of biopsies taken raises the diagnostic yield; however still up to 1/3 patients in a modern referral centre have no biopsies taken. EoE should be excluded in those with no dysphagia and refractory reflux symptoms. Although PPIs are provided as first-line therapy, a positive response to steroids is more likely in those with higher numbers eos/hpf, while those with fewer numbers and no endoscopic features could be considered for PPI therapy first. Such findings might be a useful tool to help tailor therapy.

Disclosure of Interest None Declared