Airway mucus and ¹²⁹Xe MRI ventilation after single inhaler triple therapy in asthma

Study participants and design

This study evaluated the effects of FF/UMEC/VI (200/62.5/25 µg) on patients who were poorly controlled on ICS/LABA (with no prospectively planned control group). Participants with Step 4 or 5 asthma, according to the Global Initiative for Asthma (GINA) [28], provided written informed consent to a protocol approved by an ethics board (Ethics Board Identifier: 00000940) and Health Canada (NCT04651777). Males and females aged 18–75 years with forced expiratory volume in 1 s (FEV₁) ≥35 and ≤80% predicted were enrolled, with documented treatment (at least 6 months) of low-to-medium-dose ICS/LABA and a history of poor control (Asthma Control Questionnaire (ACQ-6) score ≥1.5 [29]). Participants attended a baseline visit, a 6-week, an optional 12-week and an open-extension 1-year follow-up visit (NCT02351141).

Participants stopped previous ICS/LABA asthma medication on the day prior to the first visit. Asthma medications were withheld on the morning prior to all visits [30].

At the baseline visit, complete blood count, chest CT, exhaled nitric oxide fraction (FENO), ACQ-6 [31], Asthma Quality of Life Questionnaire (AQLQ) [32] and St. George's Respiratory Questionnaire (SGRQ) [33] were completed. Spirometry, plethysmography, airwave oscillometry and ¹²⁹Xe MRI were performed pre-bronchodilator (pre-BD) and 15 min post-bronchodilator (post-BD) inhalation of 4×100 μg Novo-Salbutamol Hydrofluoroalkane (Teva Novopharm Ltd., Toronto, Canada) using an AeroChamber (Trudell Medical International, London, Canada). After training to use a proprietary multi-dose dry powder inhaler, participants inhaled the first dose of FF/UMEC/VI (200/62.5/25 µg) under observation, with instructions to continue once every morning until 6 weeks±5 days after baseline or an optional 12 weeks±5 days after baseline. All baseline measurements were repeated at the 6- and 12-week visits. Study doses were provided at the baseline visit and again during the 6-week visit for those participants who decided to continue to the 12-week visit. An open-extension follow-up visit, planned for 1 year (52±16 weeks) after baseline, followed the study procedures for the 6-week visit, with the addition of chest CT. During the period between the participants’ last interventional visit (6 weeks or 12 weeks) and the 1-year visit, asthma therapy decisions were based on patients’ access to funded medication and the clinical judgement of their referring pulmonologist. All exacerbations and oral corticosteroid (OCS) use between baseline and the 1-year visit were reported.

Study end-points

The primary outcome measurement was ¹²⁹Xe MRI VDP at 6 weeks. Secondary outcome measurements included: FENO, spirometry measurements, oscillometry measurements of small-airway resistance (R_5–19) and ACQ-6 score at 6 weeks. Exploratory outcomes included VDP and CT airway measurements, including airway mucus at the open-extension 1-year visit.

ICS dose equivalence

We calculated the equivalent ICS dose prescribed before starting the study and FF/UMEC/VI (200/62.5/25 µg) based on the GINA 2023 [28] categories of low-, moderate- and high-dose ICS (e.g. 200 µg FF delivered via dry powder inhaler as equivalent to 320 µg mometasone furoate via inhalation powder capsule). Based on this, we converted low-, moderate- and high-dose ICS prescribed prior to the baseline visit and the 6-week and 1-year visits to an equivalent low-, moderate- and high-dose of budesonide.

Pulmonary function tests and questionnaires

FENO [34] was measured using a NIOX VERO system (Circassia Pharmaceuticals Inc., Morrisville, NC, USA). Airwave oscillometry [35] was performed using a tremoflo C-100 Airwave Oscillometry System (Thorasys Thoracic Medical Systems Inc., Montreal, Canada) to measure airway resistance (R) and reactance (X) between 5 and 37 Hz. Spirometry [36] and plethysmography [37] were performed to measure FEV₁, forced vital capacity (FVC), residual volume and total lung capacity using a whole-body plethysmograph (MedGraphics Corporation, St. Paul, MN, USA). ACQ-6, AQLQ and SGRQ were self-administered under supervision.

MRI acquisition and analysis

¹²⁹Xe gas was polarised to 30–55% (Model 9820, Polarean, Durham, NC, USA). Anatomic (¹H) and functional (¹²⁹Xe) MRI were acquired at 3.0 Tesla (Discovery MR750; GE Healthcare, Milwaukee, WI, USA) [10]. ¹H MRI was acquired using a fast-spoiled gradient-recalled echo (FGRE) sequence (partial-echo acquisition, total acquisition time 8 s; repetition time/echo time 4.7 ms/1.2 ms; flip angle 30°; field-of-view 40×40 cm²; bandwidth 24.4 kHz, matrix 128×80 zero-padded to 128×128; partial-echo percentage 62.5%; 15–17×15 mm slices). ¹²⁹Xe MRI was acquired using a three-dimensional FGRE sequence (total acquisition time 14 s; repetition time/echo time 6.7 ms/1.5 ms; variable flip angle; field-of-view 40×40 cm²; bandwidth 15.63 kHz; 128×128 matrix; 14×15 mm slices). Participants were instructed to inhale 1.0 L of gas (1.0 L N₂ for ¹H MRI, 400 mL hyperpolarised ¹²⁹Xe+600 mL ⁴He for ¹²⁹Xe MRI) from functional residual capacity (FRC) to ensure volume-matched images. Quantitative MRI analysis was performed as previously described [38] for the calculation of VDP.

Chest CT acquisition and analysis

Within 30 min of MRI, chest CT was acquired post-BD after inhalation of 1.0 L N₂ from FRC using a 320-slice Aquilion ONE/PRISM Edition (Canon Medical Systems Canada Ltd, Markham, Canada). CT parameters included: 80×0.5 mm collimation; 120 kilovoltage peak; 210 mA; tube rotation time 0.5 s; pitch factor 0.95; soft tissue reconstruction kernel FC08; slice thickness 0.5 mm; field-of-view 40 cm². The ImPACT CT patient dosimetry calculator (based on United Kingdom Health Protection Agency NRBP-SR250) estimated the mean total effective dose as 1.8 mSv.

CT images were analysed using VIDAVision (VIDA Diagnostics Inc. Coralville, IA, USA), as previously described [15, 17]. Anatomically equivalent segmental, subsegmental and sub-subsegmental airways for all airway paths were used to generate airway wall area (WA), wall thickness (WT), lumen area (LA) and total airway count (TAC). Pulmonary vessels were segmented using Chest Imaging Platform (Brigham and Women's Hospital, Boston, MA, USA) which automatically generated total blood volume and the volume of small vessels with cross-sectional area <5 mm² (BV₅) and large vessels with cross-sectional area >10 mm² (BV₁₀). As previously described [18], CT mucus occlusions were identified to generate mucus score. Mucus occlusions were identified within airways as a complete occlusion more radio-dense than the lumen in the sagittal, axial and coronal planes. Blinded, trained observers (S. Tcherner and E. Durom) scored mucus plugs in the 20 airway segments to generate a whole lung mucus score for each participant, as previously described [19]. Mucus scoring depends on reader experience and reproducibility; in this study, variability was mitigated via dual-observer consensus as previously demonstrated [18, 39].

Statistics

SPSS Statistics software version 28.0 (IBM, Armonk, NY, USA) was used for statistical analyses. Data were tested for normality using Shapiro-Wilk tests and nonparametric tests were performed for non-normal data. Differences were evaluated using repeated-measures analysis of variance (ANOVA). Participants were dichotomised based on median mucus score (score <2 and ≥2) and blood eosinophil count (<300 cells·µL⁻¹ or ≥300 cells·µL⁻¹) [40]. Results for subgroups were tested for significance using repeated-measures ANOVA or the Wilcoxon signed-rank test. Linear regression was used to evaluate the significant predictors of the change in VDP, FEV₁ and R_5–19 at 6 weeks, and displayed using forest plots (standardised β). Results were considered statistically significant when the probability of making a type I error was <5% (p<0.05).

Participant characteristics

The CONSORT diagram shows that 31 participants (P01–P31) provided written informed consent and completed the baseline visit (figure 1). Three participants withdrew prior to the 6-week visit including participant P22 who withdrew following an emergency department admission for presumed COVID-19 infection. Participant P26 withdrew due to cough when inhaling the study drug; a third participant (P31) cited being too busy and withdrew. Of the 28 participants who completed the 6-week visit, 11 chose not to attend the optional 12-week visit. Fifteen participants attended the open-extension, 1-year long-term follow-up visit which was 62±16 weeks (range 40–85 weeks) after the baseline visit.

There were no serious adverse events during the interventional study (baseline to 12 weeks); 17 mild adverse events and one moderate adverse event (emergency department admission) were reported in 16 different participants. All adverse events resolved and did not result in long-term effects. Two adverse events were judged to be related to ¹²⁹Xe MRI (light-headedness during breath-hold) and two adverse events were judged related to the study drug (dry mouth, cough during inhalation). For the long-term extension, exacerbations and OCS were reported. There was a single exacerbation in the open-extension period in the subgroup who did not attend the 1-year visit and four exacerbations and/or OCS use in three participants who attended the 1-year visit (p=0.2).

Table 1 summarises baseline demographics and study measurements for the 31 participants who attended the baseline visit (24 female, 54±15 years), the 28 participants (21 female, 54±15 years) who attended both the baseline and 6-week visits and the 15 patients who attended the open-extension 1-year follow-up visit (11 female, 56±14 years). These three subgroups were not different at baseline. Table S1 provides a by-participant list of asthma medications taken prior to the study; the mean ICS dose equivalent at baseline was 1200±400 µg budesonide as compared with 900±300 µg budesonide at the 6-week visit and 1100±500 µg budesonide at the 1-year visit.

Primary and secondary end-points

Figure 2 shows the effect of FF/UMEC/VI on the primary end-point, MRI VDP, in three representative participants at the baseline, 6-week and 1-year visits. Pre-BD ventilation defects observed at the baseline visit were visibly and quantitatively improved at 6-weeks (P10: ΔVDP, −3%; P29: ΔVDP, −3%; P33: ΔVDP, −7%). These VDP improvements persisted or continued to improve at the 1-year visit relative to the baseline visit (P10: ΔVDP, −6%; P29: ΔVDP, −4%; P33: ΔVDP, −8%). Participants P10 and P33 received FF/UMEC/VI from the baseline to the 1-year visit, while participant P29 used a dual-inhaler ICS/LABA + single-inhaler LAMA during the open-extension period, because of a lack of funded access.

Table 2 shows that pre-BD VDP (p=0.02), pre- (p=0.008) and post-BD (p=0.005) FEV₁, pre-BD R₅, (p=0.03) pre- (p=0.001) and post-BD (p=0.02) R_5–19, AQLQ (p<0.001) and SGRQ (p=0.001) were significantly different across timepoints. In addition, airway mucus score (p=0.01), LA (p=0.03) and TAC (p=0.03) were all significantly different between the baseline and 1-year visits.

Spaghetti and box and whisker plots in figure 3 show these results for MRI VDP, pre-BD FEV₁ % pred, FEV₁/FVC, R_5–19, ACQ-6, AQLQ as well as mucus score and TAC. Table S2 provides results for all participants stratified by concurrent biologic and other therapies; there were no significant measurement differences at 6 weeks and 1 year for these treatment subgroups.

Figure 4 shows that the change in VDP at 6 weeks was only significant in subgroups with augmented mucus score and blood eosinophils (high type 2 inflammation) whereas the improvement in peripheral airways resistance (R_5–19) and all airways resistance (R₅) was independent of high type 2 inflammation. Figure 5 provides a forest plot of predictors of the 6-week change in VDP, FEV₁ and R_5–19. Predictors of improved VDP included baseline R_5–19 and AQLQ, while predictors of FEV₁ included blood eosinophils, R_5–19, ACQ-6 and AQLQ. Improved R_5–19 was predicted by VDP, FEV₁ and blood eosinophils.

Exploratory end-points

CT measurements of airway morphology and mucus occlusions were examined as exploratory end-points. Figure 6 shows five representative participant examples of mucus occlusions which resolved at the 1-year visit, and a Sankey plot summarising the changes for all participants who attended the baseline and the 1-year visits (n=15). In total, 48 of 58 mucus plugs at baseline had resolved at the 1-year follow-up (83%). There were 15 newly identified mucus occlusions, with a net difference of 33 fewer mucus plugs at the 1-year visit (25 plugs at 1 year compared to 58 plugs at baseline, 43%). Four participants who attended the 1-year visit were also receiving biologic therapy, which is understood to influence airway mucus plugs [15, 17]. However, the change in mucus score between baseline and 1-year was significant in all 15 participants (Δ mucus score −2±2, p=0.02) and the subgroup (n=11) not receiving biologic therapy (Δ mucus score −2±2, p=0.01). For three of the four participants on biologic therapy, this was prescribed >6 months prior to the study start. A by-participant listing in Table S3 shows changes in mucus occlusions at 1 year and asthma therapy prescribed during the period in between the treatment study and the open-extension long-term follow-up visit; 13 of 15 participants remained on some form of triple therapy (dual inhaler, single inhaler) during the follow-up period. The spaghetti plot in figure 3 for mucus score shows that two participants (P06 and P10) experienced a net increase of 1 for mucus score and all other participants reported a net decrease in mucus score. Participant P06 was on dual-inhaler triple therapy+mepolizumab which had been continuous during the entire study period, while participant P10 was on FF/UMEC/VI for the entire study. The participant (P13) with the greatest decrease in mucus (Δ mucus score, −10) initiated benralizumab in addition to FF/UMEC/VI at week 21, which was 31-weeks prior to the 1-year visit. Importantly, this participant also showed substantial pre-BD VDP (baseline, 18%; 6 weeks, 1%) and ACQ-6 (baseline, 4.33; 6 weeks, 1.33) improvement in response to FF/UMEC/VI at 6-weeks, prior to initiation of benralizumab, and this improvement persisted for the remainder of the study.