Endoscopic treatment is an established treatment of Zenker’s diverticulum. However, evidence on endoscopic retreatment of ZD after prior treatment failure is limited. We aimed to compare long-term outcomes of endoscopic treatment of ZD in patients with and without previous endoscopic or surgical treatments.

This is a retrospective analysis of a prospectively maintained cohort of patients who underwent endoscopic treatment for ZD at a single high-volume center between March 2019 and March 2025. Patients who had completed at least 6 months of follow-up were included. Baseline data included demographics, symptom duration, and septum size. Procedural data included type of treatment (flexible endoscopic septotomy (FES) vs peroral endoscopic septotomy (POES), sedation, procedural time, adverse events and length of hospitalization. The primary outcome was clinical failure (defined as Kothari-Haber Dysphagia Score ≥ 2) during overall follow-up. The effect of previous treatment on clinical failure was estimated using inverse-probability-of-treatment weighting (IPTW) with a Cox model, accounting for relevant influencing factors. A sensitivity analysis used 1:1 propensity-score matching (PSM) followed by Cox regression was performed.

250 patients underwent endoscopic treatment for ZD during the study period. 224 patients (median age 74 years; 70.5% male; median ASA score 2; median septum 20 mm; median symptom duration 12 months) with at least 6 months of follow-up were included in the analysis. 50 patients (22.3%) had undergone previous treatment. No difference between the two groups in terms of baseline characteristics and procedural data (median procedural time 20 min; adverse events 3.1%) were reported. Over a median follow-up of 36 months, clinical failure was comparable in the two groups (10.4% naïve vs 14.0% previous; p=0.53). A conventional multivariable Cox model stratified by technique (FES vs POES) found prior failure treatment did not significantly predict clinical failure (HR 1.06; 95%CI 0.46–2.42; p=0.887). In the IPTW–Cox analysis, previous treatment was not associated with clinical failure (HR 1.34; 95%CI 0.56–3.22; p=0.508), with good post-weighting balance across covariates. The PSM–Cox sensitivity analysis yielded concordant results (HR 1.35; 95%CI 0.61–2.97; p=0.46). Among previously treated patients, a subgroup analysis by prior modality (endoscopic vs surgical) showed no association with clinical failure (endoscopic: HR 0.78, 95% CI 0.28–2.23, p=0.65; surgical: HR 1.63, 95% CI 0.63–4.22, p=0.31).

Endoscopic treatment for ZD after prior intervention failure remains feasible, safe, and effective, achieving clinical results comparable to treatment-naïve patients even in the long-term. These findings support offering endoscopic treatment as the first-line therapy for ZD irrespective of previous failure history.