Functional respiratory imaging reveals early structural and functional airway responses to benralizumab in severe eosinophilic asthma

Abstract

Background

Benralizumab improves outcomes in severe eosinophilic asthma (SEA). We aimed to evaluate early effects of benralizumab on airway pathophysiology using functional respiratory imaging (FRI).

Methods

In this prospective, single-arm study, 20 adults with SEA underwent CT at baseline, 4 and 12 weeks. Coprimary endpoints were changes in specific airway volume (siVaw) and resistance (siRaw) at functional residual capacity (FRC) and total lung capacity (TLC) at 12 weeks. Imaging outcomes included mucus score/volume and air-trapping.

Results

At week 12, siVaw increased at FRC by 0.16 mL/L (27%; 95% CI 0.01% to 0.31%; p = 0.02) and at TLC by 0.14 mL/L (14%; 95% CI 0.03% to 0.25%; p = 0.02); siRaw at TLC decreased by −0.41 kPa·s/L (−24%; 95% CI −0.80% to −0.01%; p = 0.04). siRaw at FRC increased (2.99 kPa·s/L; 95% CI −4.57 to 10.54; p=0.42), with a significant relative change (293%; p=0.03) that lost significance after excluding outliers. siVaw at FRC increased early (week 4: 0.19 mL/L; 33%; 95% CI 0.03% to 0.34%; p = 0.01). At week 12, the mucus score decreased by 0.41 (95% CI −0.67 to −0.14; p<0.01), and air-trapping by 5.5% (95% CI −8.8 to −2.1; p < 0.01). Changes in siVaw and mucus correlated with improvements in forced expiratory volume in 1 s (r = 0.66 and −0.77; both p<0.01) and Saint George’s Respiratory Questionnaire (mucus: r = 0.45; p = 0.05).

Conclusions

Benralizumab treatment in SEA was associated with early improvements in airway volume, resistance, mucus and air-trapping, correlating with better lung function and quality of life. These findings support FRI to monitor early treatment effects.

Trial registration number

NL-OMON26915.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Benralizumab reduces exacerbations and improves lung function in patients with severe eosinophilic asthma. However, early effects on airway geometry, resistance, mucus plugging and air-trapping—and their relationship to clinical outcomes—are incompletely understood.

WHAT THIS STUDY ADDS

• This study shows that treatment with benralizumab is associated with early improvements in airway volume, resistance, mucus score and air-trapping, as measured by functional respiratory imaging. These improvements correlated with better lung function and quality of life.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• By revealing early structural and functional airway changes, this study supports the use of imaging biomarkers to better understand and potentially monitor biologic treatment response early in severe eosinophilic asthma.

Introduction

Severe asthma, defined as asthma that requires treatment with high-dose inhaled corticosteroids (ICS) plus a second controller and/or systemic corticosteroids, accounts for 4%–10% of all patients with asthma.^{1 2} These patients contribute to asthma-related morbidity and associated healthcare costs to a great extent.³ In recent years, biologic therapies have emerged for the treatment of severe asthma.⁴ Benralizumab is a monoclonal afucosylated antibody directed against the Interleukin-5 receptor (IL-5-R). In patients with severe eosinophilic asthma (SEA), benralizumab reduces exacerbations, has oral corticosteroid (OCS) sparing properties and improves forced expiratory volume in 1 s (FEV₁), symptoms and quality of life.5,8 It exerts its effect through eosinophil depletion⁹; however, its impact on airway pathophysiology remains incompletely understood. Previously, a single dose of benralizumab improved ventilation as measured by (129)Xe MRI¹⁰ in patients with SEA with persistent effects after 2.5 years. In another study, a reduction in mucus, measured with CT scans, after 2.5 years of benralizumab treatment was found in a subset of patients.¹¹

Building on early improvements in ventilation observed via MRI and the suggestion of persistent/late mucus reduction on CT, we hypothesised that functional respiratory imaging (FRI) could detect early changes in airway geometry, function and mucus volume in patients with SEA following treatment with benralizumab. FRI is based on CT and computational fluid dynamics (CFD) and has shown to be a marker of airway dysfunction in asthma, chronic obstructive lung disease (COPD) and cystic fibrosis.12,15 We further hypothesised that changes in FRI parameters would correlate with improvements in asthma control, quality of life, lung function (spirometry, body plethysmography, oscillometry), and 6 min walk test (6MWT). The aim of this study was to describe changes in airway geometry and mucus and explore their associations with clinical parameters following benralizumab treatment.

Methods

Study design

This study was a prospective, open-label, single-arm, observational study, with a follow-up period of 12 weeks. It was prospectively registered in the International Clinical Trials Registry Platform (NL-OMON26915).

Subjects

Adults with SEA were recruited from the outpatient clinic at the department of Respiratory Medicine of Maastricht University Medical Centre+, a tertiary academic referral centre for severe asthma. Severe asthma was diagnosed according to ERS/ATS guidelines¹ and was an essential inclusion criterion. In addition, patients were required to have a prebronchodilator FEV₁ <80% predicted and a history of 2 or more exacerbations in the preceding year. Further inclusion criteria included treatment with high-dose (>400 ug/day beclomethasone dipropionate equivalent) ICS and long-acting beta-2 agonists (LABA) for ≥3 months, and ≥12 months of medium-dose ICS before visit 1. Prednisolone maintenance therapy ≤10 mg/day was permitted. Blood eosinophil count (BEC) >300/μL or >150/μL when using OCS maintenance therapy was required. Exclusion criteria were the existence of other lung diseases, pregnancy, active smoking within 6 months, smoking history ≥15 pack-years, active malignancy or prior use of anti-IL-5 or anti-IL-5-R antagonists. Previous omalizumab use was permitted, after a wash-out period of at least 28 days.

Study procedures

Screening and baseline visit (visit 1) could take place within a maximum of 30 days apart. All visits were scheduled in the morning, after withholding morning doses of inhalation medication. Benralizumab 30 mg was administered subcutaneously at visit 1, and after 4 and 8 weeks. CT scans were done at baseline, and at weeks 4 and 12. An overview of study procedures is given in online supplemental figure 1. At baseline, comorbidities were recorded. BEC was measured at baseline, after 4 and 12 weeks at a local laboratory; values below <100 cells/µL were reported as ‘<100’. Allergic sensitisation was assessed using the Phadiatop test (Thermo Fisher Scientific, Uppsala, Sweden). Safety was monitored throughout the study.

Anthropometrics, lung function and 6MWT

Body height and weight were measured prior to other study assessments. Spirometry was performed at baseline, week 4 and week 12 using Vyaire Vyntus One (Vyaire Medical, USA) according to American Thoracic Society (ATS)/ European Respiratory Society (ERS) standards.¹⁶ Body plethysmography and impulse oscillometry were performed at baseline and week 12 using the Vyntus Body and Masterscreen systems, respectively (Vyaire Medical, USA). Fraction of expired nitric oxide (FeNO) was measured at baseline and week 12 using NIOX Vero (NIOX, USA) according to the manufacturer’s instructions. Global Lung function Initiative-2012 reference equations¹⁷ were used for predicted values. Exercise tolerance was evaluated at baseline, week 4 and week 12 by a 6MWT following ATS/ERS guidelines.¹⁸

Patient-reported outcome measures

Six-point Asthma Control Questionnaire (ACQ-6)¹⁹ was administered on paper, as well as the Saint George’s Respiratory Questionnaire (SGRQ)²⁰ at each visit (baseline, weeks 4, 8 and 12). ACQ-6 score ranges from 0 to 6 points, with higher scores indicating worse asthma control and a score >1.5 points reflecting uncontrolled asthma. SGRQ scores range from 0 to 100 points, where higher scores represent worse quality of life.

CT

Low-dose high-resolution CT scans were acquired at total lung capacity (TLC) and functional residual capacity (FRC) during breath-hold at baseline, week 4 and week 12. To ascertain the scans were done at the correct lung volumes, real-time spirometry with Spirostick (Geratherm Medical AG, Germany) was performed. All CT scans were acquired on a 192 slice CT scanner (Siemens Medical Solutions, Germany), with a slice thickness of 0.6 mm and 100 kV tube voltage. Detailed CT settings are reported in the online supplemental materials.

Functional respiratory imaging

The coprimary outcomes were changes in specific airway volume (siVaw) and specific airway resistance (siRaw) at both FRC and TLC at week 12. FRI (Fluidda, Belgium), including segmentation of the airways and use of CFD has been described previously.¹² The airways were segmented up to the point where intraluminal and alveolar air could no longer be distinguished, typically at the 7th–10th bifurcation. Airway volumes were then measured, and siVaw was calculated by dividing airway volume by lung volume and presented as ml/L. Likewise, airway resistance (iRaw) was determined using CFD by solving Reynolds-averaged Navier–Stokes equations, with flow distributions adjusted to match CT-based segmentations. siRaw was calculated by multiplying iRaw by lobar volume, normalising resistance across patients and presented as kPa*s/L. Mucus score (MS) was adapted from Dunican et al²¹ and ranged from 0 to 18 (instead of 20), based on the number of bronchopulmonary segments with a visible mucus plug, defined as a complete occlusion of the airway at TLC. FRI-based air-trapping was defined as the percentage of voxels with Hounsfield Units between −1024 and −850 at FRC. CT-analysing technicians were blinded for clinical parameters and study time point (baseline vs follow-up).

Statistical analyses

Sample size estimations for the coprimary outcome were determined based on the change in siVaw and derived from previous studies, aiming for a power of 90% and a significance level of 0.05 (two-tailed). In one study,²² patients with partially controlled asthma switched from fine to extra-fine beclomethasone/formoterol, resulting in a mean increase in siVaw of 0.255±0.280 mL/L, with a Cohen’s d effect size of 0.91 and a required sample size of 15. Another study¹⁴ investigated the addition of roflumilast to standard care in COPD, revealing a mean effect on siVaw of 0.717±0.861 mL/L, with a Cohen’s d effect size of 0.833 and a required sample size of 18. Based on these findings, a sample size of 20 patients was chosen for the present trial. Statistical analyses were performed using R (V.4.40) and SPSS (V.24.0, IBM). Continuous variables were tested for normality of distribution using Q-Q plots, skewness and kurtosis statistics, and Shapiro-Wilk test. FRI parameters were analysed using a linear mixed-effect model to account for the repeated measurements per zone within subjects. The dependent variable was the absolute value of the parameter at a certain visit. The model included fixed effects for visit, zone and the interactions between visit and zone. Zone was included as a random slope by subject. For the variables included in these models, no missing values occurred. Score-based parameters were analysed using ordered logistic mixed-effect regression. For non-FRI parameters, we employed paired t-tests for parametric data and paired Wilcoxon tests for non-parametric data to assess changes between visits. When multiple visits were compared, analysis of variance or Friedman test with Bonferroni correction was applied where appropriate. Spearman’s rank correlations between changes in FRI parameters and lung function or patient reported outcome measures (PROMs) were computed.

Patient and public involvement

No patient or public partners were involved in the design, conduct or dissemination of this research. Due to the exploratory nature of this early-stage study, direct patient and public involvement was not feasible.

Results

31 patients were assessed for eligibility, of whom 7 declined participation, and 4 patients did not meet inclusion criteria: 1 patient exceeded 15 pack-years of smoking, 2 patients had an FEV₁ >80%, 1 patient had BEC<300/µL. 20 subjects were enrolled between November 2021 and May 2023 and all completed the study. Baseline characteristics are presented in table 1.

Table 1. Baseline characteristics of the study population.

	Number (%), mean (STD) or median (IQR)
Demographics
Age* (years)	59.3 (7.6)
Female sex, n (%)	12 (60%)
BMI* (kg/m²)	28.5 (4.9)
Ex-smoker, n (%)	10 (50%)
Smoking history (pack-years)	4.3 (2, 5.5)
Asthma control
Exacerbations in previous 12 months	3 (2, 4)
Any hospitalisation in previous 12 months, n (%)	4 (20%)
ACQ-6 score*, points	2.82 (0.89)
SGRQ total score*, points	60.8 (21)
Lung function
FEV₁ before bronchodilation* (L)	1.86 (0.54)
FEV₁* (% predicted)	63 (13.2)
FEV₁/FVC ratio before bronchodilation*	0.61 (0.11)
Residual volume* (L)	3.05 (2.13, 3.54)
Residual volume* (% predicted)	142 (29)
Phenotype
Blood eosinophils (per µL)	415 (335, 765)
FeNO (ppb)	36 (21, 69)
Allergic sensitisation, n (%)	9 (45%)
Late-onset asthma (>12 years), n (%)	15 (75%)
Medication Use
ICS/LABA, n (%)	20 (100%)
LAMA, n (%)	13 (65%)
ICS dose BDP equivalent* (µg)	800 (65)
Oral corticosteroids, n (%)	5 (25%)
Prednisolone daily dose (mg)	5 (2.5, 7.5)
Comorbidities
Chronic rhinosinusitis, n (%)	13 (65%)
Nasal polyps, n (%)	5 (25%)
Cardiovascular disease, n (%)	2 (10%)
Diabetes mellitus, n (%)	2 (10%)

^*Variables marked with an asterisk are normally distributed.

ACQ-6, Asthma Control Questionnaire (6-item version); BDP, beclomethasone dipropionate; BMI, body mass index; FeNO, fractional exhaled nitric oxide; FEV₁, forced expiratory volume in 1 s; FVC, forced vital capacity; ICS, inhaled corticosteroid; LABA, long-acting beta-agonist; LAMA, long-acting muscarinic antagonist; N, number of participants; SGRQ, St. George’s Respiratory Questionnaire.

In short, 60% was female with a mean age of 59.3 years. Half were ex-smokers with a median smoking history of 4.3 pack-years. Most had poorly controlled asthma (median ACQ-6 score 2.82), with frequent exacerbations (median 3/year) and hospitalisations in the last year in 20% signifying that the population consisted of patients with severe asthma and a high burden of disease. Lung function was impaired (mean FEV₁ 63% of predicted), and air-trapping was present (mean residual volume (RV) 142% of predicted). BEC was elevated (median 415 /µL), and 75% had late-onset asthma (onset above the age of 12).²³ All patients used high-dose ICS/LABA, with 25% on maintenance OCS treatment.

FRI parameters

At week 4, siVaw at FRC significantly increased compared with baseline (least square mean (LSM) difference: 0.19 mL/L; 95% CI 0.03 to 0.34; 33%; p=0.01), with sustained improvement at week 12 (0.16 mL/L; 95% CI 0.01 to 0.31; 27%; p=0.02). siVaw at TLC did not change at week 4 (0.08 mL/L; 95% CI −0.13 to 0.30; 8%; p=0.17) but increased at week 12 (0.14 mL/L; 95% CI 0.03 to 0.25; 14%; p=0.02). siRaw at TLC decreased significantly at week 4 (−0.74 kPa·s/L; 95% CI −1.15 to −0.34; p<0.01) and week 12 (−0.41 kPa·s/L; 95% CI −0.80 to −0.01; p=0.04). The corresponding percent change was not significant (7%, 95% CI −28% to 42%; p=0.67) at week 4 or at week 12 (24%, 95% CI −10% to 58%; p=0.16). In contrast, siRaw at FRC showed a non-significant increase at week 12 (2.99 kPa·s/L; 95% CI −4.57 to 10.54; p=0.42), whereas the relative change was significant (293%; 95% CI 25% to 562%; p=0.03). After excluding three outliers (>3 SD from the mean), the change in siRaw at FRC was no longer significant (0.04 kPa·s/L; 95% CI −0.88 to 0.96; p=0.92). At baseline, the mean MS was 12.6±4.8 (median 15, range 8–17), and the mean mucus volume was 0.49±0.49 mL (median 0.28, range 0.1–0.8 mL). MS declined significantly at week 12 (−0.41; 95% CI −0.67 to −0.14; p<0.01), but not at week 4 (−0.21; 95% CI −0.48 to 0.06; p= 0.11). Mucus volume significantly decreased at both week 4 (−0.04 mL; 95% CI −0.06 to −0.01; p<0.01) and week 12 (−0.05 mL; 95% CI −0.07 to −0.03; p<0.01). Air-trapping at FRC decreased by 5.47 percentage points (95% CI −8.84 to −2.10; p<0.01). No significant changes were observed in total airway wall volume (iVAWW). Table 2 and figure 1 present the changes in FRI parameters from baseline to week 12. Figure 2 illustrates representative FRI changes in a single participant.

Table 2. Functional respiratory imaging parameters.

Outcome	Week 4 LSM difference	95% CI (lower, upper)	P value	Week 12 LSM difference	95% CI (lower, upper)	P value
siVaw at FRC (mL/L)	0.19	0.03 to 0.34	0.02	0.16	0.01 to 0.31	0.04
siVaw at FRC (%)	33.0	11 to 55	0.01	27.0	5 to 49	0.02
siVaw at TLC (mL/L)	0.08	−0.13 to 0.30	0.44	0.14	−0.08 to 0.35	0.20
siVaw at TLC (%)	8.0	−4 to 19	0.17	14.0	3 to 25	0.02
siRaw at FRC (kPa·s/L)	−3.32	−11.09 to 4.44	0.38	2.99	−4.57 to 10.54	0.42
siRaw at FRC (%)	15.0	−261 to 291	0.91	293.0	25 to 562	0.03
siRaw at TLC (kPa·s/L)	−0.74	−1.15 to –0.34	<0.01	−0.41	−0.80 to –0.01	0.04
siRaw at TLC (%)	7.0	−28 to 42	0.67	24.0	−10 to 58	0.16
Mucus score (points)	−0.21	−0.48 to 0.06	0.11	−0.41	−0.67 to –0.14	<0.01
Mucus volume (mL)	−0.04	−0.06 to –0.01	<0.01	−0.05	−0.07 to –0.03	<0.01
Air-trapping at FRC (percentage point)	−1.38	−4.80 to 2.04	0.41	−5.47	−8.84 to –2.10	<0.01
iVAWW (mL)	−0.08	−0.48 to 0.33	0.69	−0.18	−0.58 to 0.22	0.37
iVLobe at FRC (%)	−0.64	−5.3 to 4.2	0.78	−5.74	-10.4 to–1.13	0.02
iVLobe at TLC (%)	2.3	−0.67 to 3.32	0.18	0.9	−1.08 to 2.88	0.35

Bold indicates significance P<0.05.

% change, relative percentage change from baseline; FRC, functional residual capacity; iVAWW, airway wall volume; iVLobe, lobar volume expressed as a percentage of total lung volume; LSM, least squares mean difference from baseline; siRaw, specific image-based airway resistance; siVaw, specific image-based airway volume; TLC, total lung capacity.

Figure 1. Least square mean change in specific airway volume (siVaw) in mL, siVaw percentage change, mucus volume and mucus score. FRC, functional residual capacity; TLC, total lung capacity.

Figure 2. Functional imaging results. Top panel: Mucus (red). Middle panel: Change in siVaw at FRC from baseline. Bottom panel: Air-trapping (blue). FRC, functional residual capacity; siVAW, specific airway volume.

Other secondary outcome measures

Significant improvements were observed in lung function (table 3).

Table 3. Non-FRI parameters.

	Baseline		Week 4		Week 12
	Mean±SD	Median (IQR)	Mean±SD	Median (IQR)	Mean±SD	Median (IQR)	P value
Patient-reported outcomes
ACQ-6	2.8±1.4	2.8 (2.1, 3.5)	2.05±1.3	1.8 (0.9, 3.1)	1.6±0.9	1.5 (0.3, 3.0)	<0.001
SGRQ	61±21	66 (43, 74)	46±25	50 (20, 68)	34±29	35 (4, 58)	<0.001
Spirometry
FEV₁ (% predicted)	62±14	62 (53, 75)	67±16	69 (54, 83)	73±14	72 (61, 85)	<0.001
FVC (% predicted)	79±12	83 (67, 89)	83±13	80 (73, 93)	87±13	89 (82, 92)	0.004
FEV₁/FVC (%)	61±11	61 (52, 69)	62±11	63 (55, 70)	64±10	63 (58, 70)	0.124
Body plethysmography
FRC (% predicted)	130±22	129 (115, 149)			125±24	130 (114, 140)	0.028
RV (% predicted)	142±29	140 (112,172)			124±31	119 (99, 151)	<0.001
IC (% predicted)	101±18	98 (88, 110)			106±21	101 (88, 124)	0.101
sRaw (% predicted)	235±120	201 (147, 227)			205±75	181 (152, 252)	0.172
sGaw (% predicted)	51±21	54 (35,62)			55±19	56 (40, 66)	0.491
Oscillometry
R5 (kPa·s/L)	0.72±0.32	0.62 (0.51,0.78)			0.6±0.26	0.55 (0.43, 0.82)	0.025
R5 (% predicted)	216±65	203 (166, 252)			188±58	180 (143, 205)	0.006
R5-R20 (kPa·s/L)	0.3±0.25	0.21 (0.14, 0.35)			0.22±0.18	0.17 (0.07,0.33)	0.014
X5 (kPa·s/L)	−0.34±0.3	−0.26 (-0.4,–0.21)			−0.3±0.2	−0.25 (-0.4,–0.14)	0.313
X5 (% predicted)	267±143	229 (154, 361)			220±177	177 (129, 281)	0.135
AX (kPa/L)	2.98±1.73	1.73 (1.0, 4.5)			2.3±2.3	1.41 (0.39, 3.28)	0.1
Inflammatory marker
FeNO (ppb)	48±38	39 (21,69)			50±41	29 (17, 90)	0.78
Exercise capacity
6 min walk test (m)	454±114	445 (389, 510)	472±113	484 (395, 540)	472±129	473 (386, 565)	0.258

Bold indicates significance P<0.05.

ACQ-6, Asthma Control Questionnaire (6-item version); AX, area of reactance; FeNO, fractional exhaled nitric oxide; FEV₁, forced expiratory volume in 1 s; FRC, functional residual capacity; FRI, functional respiratory imaging; FVC, forced vital capacity; IC, inspiratory capacity; ppb, parts per billion; R5, resistance at 5 Hz; RV, residual volume; sGaw, specific airway conductance; SGRQ, St. George’s Respiratory Questionnaire; sRaw, specific airway resistance; X5, reactance at 5 Hz.

FEV₁ increased from 62%±14% predicted to 73%±14% at week 12 (p<0.001), and FVC from 79%±12% to 87%±13% predicted (p=0.004). RV decreased from 142%±29% to 124%±31% predicted (p<0.001), and FRC declined from 130%±22% to 125%±24% predicted (p=0.028). Oscillometric measurement of small airway disease R5-R20 also improved (decrease from 0.30±0.25 to 0.22±0.18 kPa·s/L (p=0.014)), as did total airway resistance: R5 decreased from 0.72±0.32 to 0.60±0.26 kPa·s/L (p=0.025). Patient-reported outcomes improved as well with ACQ-6 scores decreasing from 2.8±1.4 to 1.6±0.9 (p<0.001), and SGRQ scores from 61±21 to 34±29 (p<0.001). BEC was reduced to <100/µL in all patients by weeks 4 and 12.

Correlations between changes in FRI and other outcomes

Correlations between changes in FRI parameters and changes in other outcomes are shown in online supplemental table 1. Change in siVaw at TLC at week 12 correlated with change in lung function, particularly FEV₁ (r=0.66; p<0.01), RV (r=−0.58; p<0.01), airway resistance (sRawpred r=−0.60; p<0.01), and conductance (sGawpred r=0.64; p<0.01). Strong correlations were observed between the change in mucus volume at week 12 and improvements in lung function, including FEV₁ (r=−0.77; p<0.01), FEF_25–75 (r=−0.68; p<0.01) and RV (r=0.63; p<0.01). FRI-derived improvements correlated with better quality of life and physical capacity: at week 12, reduced mucus volume was linked to improved SGRQ (r=0.45; p=0.05) and 6MWT (r=−0.57; p<0.01). Less air-trapping also correlated with 6MWT (r=−0.49; p=0.03), as did change in siVaw (r=0.46; p=0.04). Furthermore, reduction of iVAWW correlated with improved ACQ-6 (r=0.49; p=0.03) and SGRQ (r=0.47; p=0.04). Online supplemental figure 2 shows that patients with reduced mucus volume at week 12 had greater improvements in FEV₁, FEF25–75, SGRQ scores and 6MWT than those with increased mucus (n=4).

Adverse events

Most common adverse events were: cough (30%), fatigue (20%), dyspnoea (20%), headache (15%) and exacerbation of asthma (n=15%). No safety concerns were observed.

Discussion

Benralizumab treatment in patients with SEA was associated with significant structural and functional airway improvements, as assessed by FRI at the 12-week primary outcome. This study demonstrates that FRI can capture early airway changes following benralizumab treatment, supporting its use as a tool to monitor biologic treatments. The most consistent changes were observed in siVaw, which increased at both FRC and TLC. While siRaw improved at TLC, changes at FRC were inconsistent and likely influenced by outliers. Mucus burden—assessed by MS and volume—also declined, alongside reduced air-trapping. Importantly, these imaging changes were associated with clinically meaningful improvements in lung function, quality of life and exercise capacity. Notably, several changes were already evident at week 4, suggesting that structural and functional airway responses may begin early. While the VESTIGE study²⁴ applied FRI to assess dupilumab, our study is among the first to use FRI to evaluate benralizumab in SEA. Previous imaging studies with benralizumab have shown reductions in mucus and air-trapping, but did not assess changes in airway volume or resistance.^{10 11 25} Our findings extend this evidence by demonstrating early, clinically relevant FRI-detected changes from week 4 onward.

The observed improvements in siVaw at FRC might be particularly meaningful as airway closure near FRC contributes to airflow limitation, impaired gas exchange, airway hyperresponsiveness and exacerbation risk in asthma.26,28 Although the underlying mechanisms differ, the 27% increase in siVaw observed here is comparable to that seen with bronchodilators in COPD (11%, 23%).¹³ It was, however, smaller than the 226% increase reported in moderate-to-severe asthma patients treated with dupilumab,²⁴ though that estimate was imprecise and not statistically different from placebo. Contrary to our hypothesis, siRaw did not improve consistently. Since Raw is calculated by dividing the pressure drop by the flow rate, near-zero flow results in extremely high values. This suggests that siRaw at FRC may not be a reliable FRI marker in cases of pronounced mucus plugging. Similar to our findings, siRaw did not consistently improve in patients treated with dupilumab in the aforementioned VESTIGE study.²⁴

We observed a reduction in MS after benralizumab treatment, suggesting that part of its clinical benefit in SEA may be mediated by mucus clearance. This effect was supported by moderate-to-strong correlations between reductions in mucus volume and improvements in FEV₁, SGRQ and 6MWT. However, the response was heterogeneous: a small subgroup showed increased mucus volume and no improvement in these clinical outcomes, and no consistent clinical pattern distinguishing this subgroup was identified. Mechanistically, decreased mucus may contribute to improved airway volume and less air-trapping, potentially enhancing lung function and quality of life. The magnitude of MS reduction in our study (LSM −0.41) was modest compared with previous studies with benralizumab (mean −2¹¹; median −4.5²⁵), dupilumab (LSM −4.9²⁴; median −4)²⁹ or tezepelumab (mean −1.7).³⁰ This may be below a clinically meaningful change; however, a minimal clinically important difference for MS in SEA has not yet been established, and the clinical interpretation of changes of this magnitude remains uncertain. Several factors may explain the modest effect. First, our use of linear mixed-effects modelling likely produced more conservative estimates by accounting for within-subject variability.³¹ Second, the 12-week treatment duration may have been insufficient to capture the full effect of benralizumab on mucus; larger reductions have been reported in longer-term studies.¹¹ Third, our cohort had more severe disease, with higher baseline MS (12.5 vs 3.5 in a tezepelumab trial,³⁰ universal high-dose ICS use and 25% on maintenance OCS). This may have attenuated the eosinophil-driven component of mucus and left a relatively larger non-eosinophilic and structural component that is less amenable to IL-5 receptor blockade. Indeed, benralizumab does not target IL-13 or IL-18, cytokines implicated in mucus hypersecretion.^{32 33} Consistent with our imaging findings, improvements in RV, FRC and oscillometry further support an effect of benralizumab on small airways. Taken together, these multimodal findings strengthen the evidence that benralizumab improves distal airway function, even within a 12-week time frame. Although no overall improvement in 6MWT was observed, changes in walk distance correlated with improvements in siVaw, mucus volume and air-trapping. Heterogeneous responses may reflect factors such as deconditioning, steroid-induced myopathy or fatigue, highlighting the need for comprehensive management, including pulmonary rehabilitation, beyond anti-inflammatory therapy.34,36

This study has several strengths. First, the studied population consisted of a well-defined SEA sample, closely reflecting clinical practice. Second, the prospective nature and complete and extensive assessments of all participants, including spirometry-gated CT scans, body plethysmography and oscillometry, add to the validity of the findings. The most notable limitation of this study is the lack of a usual care control group, which limits causal interpretation of the observed effects. Alternative explanations for the reduction in mucus, such as natural disease variation, cannot be ruled out. However, trials have shown that MS typically remains unchanged or worsens in the absence of biological treatment.^{24 30} Given the severe phenotype and stable FeNO, a study-entry effect or improved adherence as explanations for mucus reduction seems unlikely. Additionally, external validity is limited to SEA characterised by severely elevated MS. Moreover, information on the type of airway inflammation would have helped clarify the mucus pathophysiology, such as whether it is IL-5 or IL-13 driven. Furthermore, there was no evaluation of size-based plug phenotypes³⁷: small numbers in this study would prevent drawing conclusions from such a post hoc analysis. Although the results have potential in aiding early benralizumab responder identification, the sample size limits the creation of a prediction model. Lastly, the relatively short follow-up period may be a limitation, as the maximum treatment effect may not have been captured.

Future studies on biologics in severe asthma should integrate imaging with biomarkers of inflammation and airway remodelling. The positive impact of biologics in patients with asthma suggests their potential in halting remodelling and achieving remission if initiated earlier.³⁸ The early improvements observed here provide encouraging insights for future imaging-based trials aimed at preventing airway remodelling in patients with less advanced disease.

In conclusion, benralizumab treatment in patients with SEA was associated with early improvements in airway volume, resistance, mucus burden and air-trapping. These structural changes, particularly in airway volume and mucus, were associated with improvements in lung function, while reductions in mucus also correlated with better quality of life. Beyond confirming its clinical efficacy, these findings offer new insight into the mechanism of action of benralizumab and support the use of FRI as a research tool to assess early responses to biologic therapy.

The funder had no role in the design of the study; the collection, analysis or interpretation of data; the writing of the manuscript; or the decision to submit the manuscript for publication.

Data availability statement

Data are available on reasonable request.