Introduction Endoscopic resection and ablation treatment is the standard of care for barretts oesophagus and related early neoplasia. To evaluate the real world experience with EMR and RFA treatment at a single centre outside of clinical trials.

Method Data was evaluated from a prospectively maintain charts of patients undergoing RFA/EMR treatment for Barretts and Early Oesophageal Cancer (EOC) (T1a and T1b). All patients with Barretts dysplasia referred through MDT from 2010–2015 were included in the study.

Results A total of 91 patients The mean age of the patients undergoing RFA/EMR is 72±8, (Range: 47–92) years with predominance of male population (85%). The most common reason for referral was HGD (58%) followed by LGD (17%), T1a (16%), T1b (7%) and squamous dysplasia (2 and ).

EMR was performed in in 75/91 (82%) patients in the study mostly in OEC patients (100%) followed by HGD (78%) and LGD (45%) patients. The disease was upstaged in 20% of study patients with 12 patients referred with HGD upstaged to T1a and 3 patients to T1b. Upstaging to HGD was observed in 3 patients with LGD. Only 3 patients had their disease down staged (x2: T1a to HGD and x1: HGD to LGD).

RFA was performed in majority of the patients in our study (86.4%, n=76) with median RFA sessions of 2 (Range:1–6). The proportion of patients having RFA performed with final diagnosis of LGD, HGD, T1a and T1b was 92%, 95% and 91% and 30% respectively.

A complete response (CR) defined as complete endoscopic and histological eradication of barretts oesophagus after HALO procedure was seen in 88% patients who completed the study (n=68). The excluded patients included those who did not opt for further RFA (n=6) sessions and patients which died of unrelated deaths (n=2). The CR is 100% in patients with LGD and T1a and 79% in HGD group.

Out of 9 patients with T1b cancer, 7 had oesophagectomy done with none of the patient having cancer in resected specimen or had disease recurrence including 1 patient who had T1N1 cancer. CR was mostly seen in patients with prior EMR than who did not (79% vs. 21%). In all the non-CR (n=6), 80% had Barretts>6 cms and 100% of incomplete responders had HGD.

Nine patients (10%) had complications from HALO procedure with 6 patients requiring endoscopic dilatation of post RFA oesophageal stricture after CR and 2 having ulceration at previous RFA site precluding further RFA. Post RFA stricture was seen mostly in LGD patients than in HGD patients (23% vs. 5%) and 80% of these patients had a prior EMR. 1 patient had a suspected oesophageal perforation after a rescue EMR which was successfully sealed by OVESCO clip. There was no procedure related deaths.

Conclusion RFA is a safe and effective procedure for management of early oesophageal neoplasia.

Disclosure of Interest None Declared