CT Mucus Score and ¹²⁹Xe MRI Ventilation Defects After 2.5 Years' Anti-IL-5Rα in Eosinophilic Asthma.

Abstract

We previously showed in patients with poorly controlled eosinophilic asthma that a single dose of benralizumab resulted in significantly improved Asthma Control Questionnaire (ACQ-6) score and ¹²⁹Xe MRI ventilation defect percent (VDP) 28 days postinjection, and ¹²⁹Xe MRI VDP and CT airway mucus occlusions were shown to independently predict this early ACQ-6 response to benralizumab. Do early VDP responses at 28 days persist, and do FEV₁, fractional exhaled nitric oxide, and mucus plug score improve during a 2.5 year treatment period? Participants with poorly controlled eosinophilic asthma completed spirometry, ACQ-6, and MRI, 28 days, 1 year, and 2.5 years after initiation of treatment with benralizumab; chest CT was acquired at enrollment and 2.5 years later. Of 29 participants evaluated at 28 days post-benralizumab, 16 participants returned for follow-up while on therapy at 1 year, and 13 participants were evaluable while on therapy at 2.5 years post-benralizumab initiation. As compared with 28 days post-benralizumab, ACQ-6 score (2.0 ± 1.4) significantly improved after 1 year (0.5 ± 0.6, P = .02; 95% CI, 0.1-1.1) and 2.5 years (0.5 ± 0.5, P = .03; 95% CI, 0.1-1.1). The mean VDP change at 2.5 years (-4% ± 3%) was greater than the minimal clinically important difference, but not significantly different from VDP measured 28 days post-benralizumab. Mucus score (3 ± 4) was significantly improved at 2.5 years (1 ± 1, P = .03; 95% CI, 0.3-5.5). In six of eight participants with previous occlusions, mucus plugs vanished or substantially diminished 2.5 years later. VDP (P < .001) and mucus score (P < .001) measured at baseline, but not fractional exhaled nitric oxide or FEV₁, independently predicted ACQ-6 score after 2.5 years. In poorly controlled eosinophilic asthma, early MRI VDP responses at 28 days post-benralizumab persisted 2.5 years later, alongside significantly improved mucus scores and asthma control.