Relapse Rates, Time-to-Flare, and Biomarker Predictors Following Omalizumab Gradual De-escalation versus Abrupt Discontinuation in Controlled Chronic Spontaneous Urticaria: A Systematic Review and Meta-Analysis of Randomized Controlled Trials and Observational Cohorts.

Abstract

Background: How best to discontinue omalizumab once chronic spontaneous urticaria (CSU) is well controlled remains unsettled, and whether to taper gradually or to stop abruptly is a recurring source of clinical uncertainty.

Objective: We set out to quantify relapse rates, time-to-flare kinetics, and biomarker predictors following gradual omalizumab de-escalation as opposed to abrupt discontinuation.

Methods: PubMed, Embase, and Cochrane CENTRAL were searched systematically through April 2026. A random-effects proportion meta-analysis (DerSimonian-Laird) spanned 13 independent study arms (N = 1,382 patients contributing to the relapse-proportion analysis, drawn from 5,130 patients reviewed across 11 studies). Shared control groups within multi-arm trials were weighted a priori. Prespecified subgroups contrasted gradual de-escalation-whether achieved by interval extension (IE) or dose reduction (DR)-with abrupt discontinuation, and the certainty of evidence was appraised using GRADE.

Results: Across all studies, the pooled relapse rate was 39.5% (95% CI, 30.1%-49.7%). Gradual de-escalation was associated with a significantly lower relapse rate than abrupt discontinuation (31.2% [95% CI, 20.3%-44.7%] vs. 48.9% [95% CI, 37.0%-61.0%]; Qb = 8.12; P = .004), and dose reduction to 150 mg every 4 weeks yielded the lowest rate of all (23.3%; 95% CI, 9.2%-47.6%; I² = 50.4%). Several baseline features independently predicted relapse: a high total IgE (>90 IU/mL, P = .005; or ≥341 kU/L, Cox P < .001), blood eosinophils >0.15 × 10⁹/L (P = .043), concomitant angioedema (OR, 2.40; P = .004), and disease duration exceeding 2 years (HR, 0.81 for successful drug survival; P = .03). Although Egger's test fell short of conventional significance (reported P = .109), the large negative intercept-coupled with the well-recognized unreliability of this test in proportion meta-analyses-meant that small-study effects could not be excluded with confidence. A sensitivity analysis confined to homogeneous cohorts followed for at least 6 months reaffirmed the superiority of de-escalation (P = .008).

Conclusion: Gradual de-escalation of omalizumab is associated with an almost two-fold reduction in the risk of post-discontinuation relapse relative to abrupt stopping. Baseline total IgE and blood eosinophils emerge as pivotal, endotype-specific biomarkers for judging which patients are ready to stop.