PWE-131 Endoscopically treatable t1 oesophageal adenocarcinoma diagnosed by emr – what is the value of ct, pet and eus?

Introduction Endoscopic Therapy for T1a oesophageal adenocarcinoma (OAC), and some T1b with favourable pathological features (sm1, no LVI, G1 or G2 differentiation), is now the recommended standard of care by BSG. Many endotherapy centres have large geographic catchment areas, so accurate staging for these early cancers is crucial to avoid unnecessary visits for patients. Our aim was to assess the utility of staging techniques to define whether endotherapy is possible. Method All referrals for endoscopic therapy to tertiary centre (St Thomas’ Hospital), who subsequently underwent EMR and had confirmed T1 OAC, were assessed. All patients were referred with baseline CT CAP. PET scans were scheduled to assess accuracy of nodal disease following confirmation of OAC. EUS was undertaken prior to EMR when intramucosal cancer or greater was diagnosed on biopsy locally. Results Forty seven patients had confirmed T1 OAC on EMR, undertaken by 3 expert endoscopists at STH between April 2012 – November 2016. Age range 46–86 years old, 38/47 male. Barrett’s oesophagus was present in all, median length C2M4. Original histology was upstaged by EMR in 24 (51%) patients from dysplasia only, including 7/47 with LGD. CT staging of T stage was inaccurate in 57%, under-staging in 19 patients (40%) and over-staging in 8 (17%). CT over staged nodal disease >N0 in 9% vs. PET, all were T1a. Only 1/39 PET scans showed possible nodal disease but no involvement was found after primary surgery. 20 patients had EUS and 7/20 had nodal disease but in 6 of these PET was negative for nodes. No Tx/T1 disease on baseline CT was upstaged by EUS. Pathological staging and treatment Detailed pathological staging is shown in table 1. 25 subjects had T1a, of which 23 had no LVI. 22 subjects had T1b, of which 15 had no LVI or T1b sm1 (or any LVI) n=23 n=15 n=9 Baseline Histology (>dysplasia) 13 (57%) 11 (73%) 7 (78%) CT (>T1) 5/18 (28%) 2/14 (14%) 1/9 (10%) CT (node accuracy vs. PET) 10/13 12/12 8/8 PET (node negative) 17/18 13/13 8/8 EUS (T and N) 6/9 4/6 3/5 Conclusion In this retrospective study we have shown that baseline histology and CT are inaccurate for staging T1 disease, compared to EMR histology. Endoscopically treatable T1 disease was over staged by CT in 22%. Decisions on referral to Endotherapy centre for potential EMR should not be based on CT and histology alone. PET may prove a useful adjunct to confirm node negative disease but more follow up data is required. Disclosure of Interest None Declared