Mo1274 MID TO LONG TERM EFFICACY OF ENDOSCOPIC THERAPY IN NEOPLASTIC BARETT’S ESOPHAGUS: A SYSTEMATIC REVIEW AND META ANALYSIS

Abstract

Although endoscopic therapy is recommended in patients with neoplastic BE, there are limited data on long term follow up. Electronic literature search was performed in Pubmed, Embase, Google Scholar and Cochrane for eligible studies meeting following criteria: 1) BE patients with high grade dysplasia (HGD) and/or early cancer who underwent any form of therapy including ablation+/- endoscopic mucosal resection; 2) defined efficacy end-point: complete eradication of intestinal metaplasia (CEIM) defined as confirmation of absence of IM on biopsies post therapy; 3) follow up for treatment endpoints clearly defined as starting after study completion; 4) follow up > 2 years; 5) post treatment surveillance protocol clearly reported. Case series, review articles, editorials, abstracts and studies not meeting these were excluded. Primary outcome was long term efficacy (CEIM) at end of follow up. Secondary outcomes were: pooled rates of CEIM at completion of initial therapy and any recurrence (IM, HGD, cancer) at end of follow up period. Pooled rates with 95% CI and forest plots were created with assessment of heterogeneity (I2) using statistical software R. Meta-regression was performed to examine for co-variates if significant heterogeneity found. Using the focused inclusion criteria, a total of 12 studies with 1958 patients who completed follow up were eligible for analysis; 82.7 % males, average age 62.6 (+/- 4.8) years. Pooled CEIM at study completion was 87.8% (95% CI 82.2%-93.4%) whereas durable remission of CEIM was 72.6% (95% CI 61.7%-83.6%) at an average follow up of 3.5 years (Figure 1). Pooled difference between CEIM at therapy completion and at end of follow up was 14.1% (95% CI 5%-23.2%). Long term follow up (>5 years) information was only available from 2 studies including 112 patients: CEIM 48.2%. At 3.5 years of follow up, pooled recurrence rates for IM and HGD/cancer were 5.77 (95% CI 3.6 to 8) percent and 1.42 (95% CI 0.06-2.3) percent per person years, respectively. Meta-regression did not show a significant effect of BE length on mid term efficacy (p=0.9) whereas duration of follow up did impact the results (p<0.05). There was significant heterogeneity found (I2=96%) in primary outcome; lack of information on salvage therapy from majority of studies precluded further analysis of heterogeneity. There was no significant evidence of publication bias. Endoscopic therapy can provide mid term durable remission of BE in approximately 75% of the treated patients. True long term follow up data are not available currently and such studies are needed to determine the true efficacy, recurrence patterns, stratified surveillance and the need for salvage therapy in neoplastic BE patients.Proportion of patients with CEIM over 2-3 year follow up, 3-5 year follow up and more than 5 year follow upView Large Image Figure ViewerDownload Hi-res image Download (PPT)