Two patients underwent a diagnostic ER during the treatment protocol of slightly elevated BE islands in order to avoid having RFA performed on possibly invading cancers (thus not to supplement the efficacy of RFA). Histology of both ER specimens showed only LGIN. No fatal or severe complications occurred. Four patients (15% [95% CI, 4%-35%]) developed complications after ER or RFA, which were graded as moderate. One patient developed delayed bleeding 6 days after ER. This patient received blood transfusion and was treated successfully with endoscopic hemostatic

therapy (adrenaline injection, bipolar probe coagulation, and clip placement). Two patients had unplanned admissions: one patient was admitted for observation after a superficial laceration that showed no transmural leakage on the swallowing contrast examination. However, selleck kinase inhibitor this click here 80-year-old patient became delirious and, as a result, the admission was prolonged; another patient was admitted 3 days after the RFA procedure because of pain, nausea, and vomiting that resolved with conservative treatment. Because both admissions were for >4 days, these complications were graded as moderate. The fourth patient with a moderate complication had a relative stenosis after ER and developed symptoms of dysphagia after RFA, which resolved

after two dilatations. In 7 patients (27% [95% CI, 12%-48%]), a superficial laceration was observed during the circumferential ablation procedure. Six of these superficial lacerations remained asymptomatic, did not require intervention, and were therefore not considered to be complications. However, one patient was admitted for observation (see

previously), because this was the first laceration we observed during our RFA experience. This patient, again, did not experience symptoms attributable to the laceration. Lacerations were located either at the level of the reflux stenosis (n = 4) or at the level of the ER scar (n = 3). In 4 of the 7 patients, the laceration was noted after the first circumferential ablation pass, and the second pass was either therefore not performed (n = 1) or was modified by the use of a balloon with a smaller Thiamet G diameter (n = 1) or by skipping the zone containing the laceration during the second RFA pass (n = 2). All patients were able to continue the RFA according to the protocol 2 to 3 months later. Patients who achieved CR-neoplasia and CR-IM were followed-up for a mean (± SD) duration of 29 ± 9.1 months (21 ± 11.7 months since last treatment session). None of the 20 patients developed neoplasia during follow-up, thus 100% (95% CI, 82%-100%) continued to have CR-neoplasia status. Two patients had small islands of BE during follow-up.