Outcomes following the discontinuation of biologic therapy in patients with severe asthma

抄録

Background: Biologic therapies are pivotal in managing severe asthma. Despite their efficacy, some patients discontinue biologics, with varying outcomes. Predictors of successful versus unsuccessful discontinuation remain poorly defined. This study aimed to identify clinical factors associated with post-discontinuation outcomes in a real-world practice.

Methods: This retrospective cohort included adults with severe asthma who had received biologics for at least 12 months and subsequently discontinued therapy for a minimum of three consecutive months. We assessed the effects of baseline blood eosinophilia (≥300 cells/μL), residual sputum symptoms during biologic therapy, treatment responsiveness, biologic class, and discontinuation reasons on post-discontinuation asthma exacerbation rates using multivariable Cox models.

Results: A total of 118 patients were analyzed. The Kaplan–Meier analysis estimated a 65 % exacerbation-free probability at 12 months after discontinuation. Factors associated with successful discontinuation included a robust clinical response and absence of exacerbations before cessation. Conversely, baseline eosinophilia, residual sputum symptoms during biologics, and discontinuation due to inadequate therapeutic response or financial burden were associated with post-discontinuation exacerbations. In class-stratified restricted models, persistent sputum remained significantly associated with post-discontinuation exacerbations after stopping anti-IL-5 therapies, while baseline eosinophilia was associated with post-discontinuation exacerbations after stopping anti-IgE or anti-IL-4Rα. Among patients with sputum symptoms or poor-response discontinuation, the overall frequency of exacerbations declined after discontinuation.

Conclusions: Baseline eosinophilia, persistent sputum during therapy, and discontinuation prompted by poor response or cost may serve as risk factors for post-discontinuation exacerbations; however, risk is phenotype- and class-dependent. Careful patient selection and monitoring are essential when considering the discontinuation of biologic treatment.