Cancer burden on piecemeal endoscopic resection of early adenocarcinoma in Barrett's oesophagus correlates with the risk of neoplastic recurrence

Background and study aims Endoscopic resection (ER) is curative for early-stage oesophageal adenocarcinoma (OAC) without high-risk features. Piecemeal endoscopic mucosal resection (pEMR) prevents assessment of lateral margins, complicating risk estimation for neoplastic recurrence. We investigated risk factors for residual and recurrent OAC post-pEMR. Methods We performed a longitudinal study of two independent patient cohorts: the test cohort, who underwent piecemeal or en-bloc ER, (n=138) and the validation cohort, treated with pEMR only (n=89). Inclusion criteria were: OAC stage T1a or low-risk T1b, no lympho-vascular invasion, and R0 resection. The primary outcome was residual OAC at first post-ER endoscopy, and secondary outcomes were residual high-grade dysplasia (HGD), recurrence of neoplasia at any post-ER endoscopy, and remission of neoplasia and metaplasia at most recent endoscopy. Results In the test cohort, the incidence of HGD recurrence was higher in patients treated with pEMR versus en-bloc ER (p=0.021). The percentage of pEMR specimens with OAC was an independent risk factor for residual OAC at the first post-pEMR endoscopy (OR for a 10% increase=1.21, CI=1-1.46, p=0.044). A 50% cut-off of involved pEMR specimens was optimal to predict residual OAC (specificity=0.69, sensitivity=0.63). Rates of residual (p=0.02) and recurrent (p=0.0024) OAC were higher when >50% of pEMR specimens were involved by OAC. In the validation cohort, recurrent OAC was also more frequent when cancer burden was >50% (p=0.013). Conclusions High OAC burden on pEMR specimens correlates with the risk of residual OAC. Post-pEMR site check before endoscopic ablation is recommended if more than 50% of pEMR specimens show OAC.