Clinical, Histologic, and Safety Outcomes With Long-term Maintenance Therapies for Eosinophilic Esophagitis: A Systematic Review and Meta-analysis.

BACKGROUND & AIMS: Our aim was to evaluate the outcomes of maintenance treatments for eosinophilic esophagitis (EoE) among observational studies (OSs) and randomized controlled trials (RCTs).

MATERIALS AND

METHODS

Studies reporting histologic success of maintenance therapy ≥48 weeks were included. The primary outcome was histologic success rate (defined as <15/<6 eosinophils/high-power field). Risk ratios (RRs) for histologic success of maintenance therapy vs placebo or induction and drug safety were also assessed. Random effects meta-analyses with heterogeneity measured with I2 were performed. RCTs and OSs were analyzed separately.

RESULTS

In RCTs, histologic <15 eosinophils/high-power field rates were 86% (95% confidence interval [CI], 71%-96%) for corticosteroids and 79% (95% CI, 69%-87%) for biologics. Dupilumab alone accounted for 82% (95% CI, 72%-89%), whereas small molecules yielded 28%. Biologics showed higher <6 eosinophils/high-power field rates compared with corticosteroids (70% vs 59%). Clinical success was 58% (95% CI, 31%-83%) for corticosteroids and 59% (95% CI, 34%-82%) for biologics. Budesonide showed common-effect adjusted RR of 7.87 (95% CI, 4.19-14.77) of maintaining histologic remission over therapy discontinuation. In OSs, proton pump inhibitors showed 64% (95% CI, 43%-83%) histologic and 80% (95% CI, 53%-97%) clinical success, whereas corticosteroids achieved 49% (95% CI, 30%-68%) and 51% (95% CI, 18%-83%) rates, respectively. Therapy de-escalation was not associated with histologic relapse (RR, 1.04; 95% CI, 0.72-1.51). Long-term safety was confirmed with 3% (95% CI, 1%-6%) severe adverse events in RCTs and 5% (95% CI, 2%-9%) in OSs. Treatment withdrawal rates were low (10% for RCTs, 4% for OSs). Moderate to substantial heterogeneity was observed for most outcomes.

CONCLUSIONS

Maintenance therapies prevent histologic relapse in the long term, without clear disadvantage of dose de-escalation from induction to maintenance phase. Low adverse events and withdrawal rates confirm long-term treatment is well-tolerated.