Small Airways Disease Affects Aerosol Deposition in Children with Severe Asthma: A Functional Respiratory Imaging Study

Abstract

Background:

Small airways disease (SAD) in severe asthma (SA) is associated with high disease burden. Effective treatment of SAD could improve disease control. Reduced end-expiratory flows (forced expiratory flow [FEF]_25–75 and FEF₇₅) are considered sensitive indicators of SAD. Inhaled medication should be delivered to the smaller peripheral airways to treat SAD effectively. Aerosol deposition is affected by structural airway changes. Little is known about the effect of SAD on aerosol delivery to the smaller peripheral airways. Functional respiratory imaging (FRI) is a validated technique using 3D reconstructed chest computed tomography (CT) and computational fluid dynamics to predict aerosol deposition in the airways.

Aim:

This study aims to compare central and peripheral (= small airways) deposition between children with SA and SAD and children with SA without SAD, with different inhaler devices and inhalation profiles.

Methods:

FRI was used to predict the deposition of beclomethasone/formoterol dry powder inhaler (DPI), beclomethasone/formoterol pressurized metered dose inhaler with valved holding chamber (pMDI/VHC), and salbutamol pMDI/VHC for different device-specific inhalation profiles in chest-CT of 20 children with SA (10 with and 10 without SAD). SAD was defined as FEF_25–75 and FEF₇₅ z-score < −1.645 and forced vital capacity (FVC) z-score > −1.645. No SAD was defined as forced expiratory volume (FEV)₁, FEF_25–75, FEF₇₅, and FVC z-score > −1.645. The intrathoracic, central, and peripheral airways depositions were determined. Primary outcome was difference in central-to-peripheral (C:P) deposition ratio between children with SAD and without SAD.

Results:

Central deposition was significantly higher (∼3.5%) and peripheral deposition was lower (2.9%) for all inhaler devices and inhalation profiles in children with SAD compared with children without SAD. As a result C:P ratios were significantly higher for all inhaler devices and inhalation profiles, except for beclomethasone administered through DPI (p = .073), in children with SAD compared with children without SAD.

Conclusion:

Children with SA and SAD have higher C:P ratios, that is, higher central and lower peripheral aerosol deposition, than children without SAD. The intrathoracic, central, and peripheral deposition of beclomethasone/formoterol using DPI was lower than using pMDI/VHC.

Introduction

The role of the small airways (<2 mm) in asthma pathophysiology is being increasingly recognized.¹ Structural and functional changes in the small airways are often observed.^2,3 The prevalence of small airways dysfunction as a result of small airways disease (SAD) in patients with asthma varies between 50% and 90%, depending on asthma severity and the measures used to assess small airways function.^3,4

SAD in patients with severe asthma (SA) has a high prevalence.³ SA is a type of asthma that, despite high doses of inhaled medication and treatment of modifiable factors, such as adherence to treatment, comorbidities, and incorrect inhaler technique, is not well controlled.⁵ Asthma exacerbations and hospital admissions pose a heavy burden on these patients.

Spirometry is the most frequently used tool to monitor lung disease. Forced end-expiratory flows, such as the forced expiratory flow between 25% and 75% (FEF_25–75) of vital capacity (VC) and at 75% (FEF₇₅) of VC, are considered to be indicators of small airways function.⁶ Reduced end-expiratory flows in patients with asthma are considered to reflect small airway dysfunction as a result of SAD. Patients with asthma and reduced end-expiratory flows have a higher risk of losing asthma control, having structural airway abnormalities, needing to use systemic corticosteroids, and having exacerbations.^1,7–9

Inhalation therapy with corticosteroids and bronchodilators is the cornerstone of asthma treatment.¹⁰ In order to treat SAD effectively, aerosolized medication should be deposited into the small airways. Currently, inhaled corticosteroids (ICS) and bronchodilators are being administered using a pressurized metered-dose inhaler with a valved holding chamber (pMDI/VHC), dry powder inhaler (DPI), or nebulizer.¹¹ Deposition of aerosols in the small airways, also known as peripheral airways, is dependent on particle-related factors (such as size, density, and shape), patient-related-factors (quality of inhalation maneuver, diameter of the airways, and structural airways abnormalities), and breathing patterns.¹² As smaller-sized particles are more likely to penetrate deeper into the lungs, several inhalers nowadays contain aerosols with a mass median aerodynamic diameter (MMAD) between 1 and 2 μm and a high fine particle fraction.¹³ In order to deposit these smaller- and medium-sized particles in the peripheral regions of the lung, a technically suitable inhalation maneuver is needed. Each type of inhaler requires an inhaler-specific inhalation maneuver, such as a slow and deep inhalation for a pMDI/VHC and a forceful and deep inhalation for a DPI.¹⁴ The quality of these different inhalation maneuvers can markedly influence aerosol deposition.¹⁵

The presence of structural airway abnormalities has been shown to negatively affect aerosol deposition in the lungs.¹² In children with cystic fibrosis, lobar structural lung abnormalities were inversely correlated with lobar deposition, indicating that more diseased lobes receive less inhaled medication.¹⁶ As end-expiratory flows, both FEF_25–75 and FEF₇₅, are considered to be indicators of small airways (or peripheral airways) function, we wondered whether children with asthma with reduced FEF_25–75 and FEF₇₅ have different central and peripheral airways deposition patterns in comparison with children with asthma and normal end-expiratory flows.

Aerosol deposition from inhalers can be predicted using functional respiratory imaging (FRI). FRI is a well-validated technique that has been extensively described previously.^16–19 This methodology uses patient-specific airway models obtained from high-resolution volumetric computed tomography (CT) scans and computational fluid dynamic (CFD) simulations to predict aerosol behavior in the lungs. The use of this technique has been validated in patients with asthma using single photon emission CT¹⁸and has shown close agreement with scintigraphy results.^20,21 The biggest advantage of FRI is that the effect of various model parameters can be assessed without the need for the patient to undergo additional evaluations.

The aim of this study was to compare central and peripheral deposition of ICS and bronchodilators, using different inhaler devices and inhalation maneuvers, between children with SA with SAD and without SAD. We hypothesized that children with SA and SAD have higher central and lower peripheral airways deposition in comparison with children with SA without SAD. Furthermore, we hypothesized that a slow and deep inhalation maneuver using an inhaler device with small particles will improve aerosol deposition in the peripheral regions of the lung.

Materials and Methods

Study design

In this in silico study, we used FRI to study the intrathoracic, central, and peripheral airway deposition of different inhaler devices and different inhalation maneuvers in children with SA, with and without SAD. The spirometer-controlled chest CT scans used for this study were retrospectively selected from a cohort of children with SA who underwent chest CT as part of clinical care between 2007 and 2019 at the pediatric pulmonology department of the Erasmus MC Sophia Children’s Hospital (Rotterdam).⁹ The local medical ethics review committee reviewed the research protocol and provided a waiver (MEC-2019–0554).²² Informed consent for the use of the chest CT scans and clinical data was obtained before data collection was initiated (Fig. 1).

FIG. 1.

Flowchart of the study design. CT, computed tomography; DPI, dry powder inhaler; pMDI + VHC, pressurized metered dose inhaler with valved holding chamber; SAD, small airways disease.

Outcome measures

The primary outcome of this study was the central-to-peripheral deposition ratio (C:P) for the different inhaler devices and inhalation maneuvers for both the SAD group and the control group. The C:P ratio is a measure that defines the aerosol distribution over the larger central airways and the smaller peripheral airways.^19,23 It is determined by dividing the dose deposited in the central airways by the dose deposited in the peripheral airways $:\hspace{0.17em}\left(\frac{\mathit{Dose}\hspace{0.17em}\mathit{deposited}\hspace{0.17em}in\hspace{0.17em}\mathit{central}\hspace{0.17em}\mathit{airways}}{\mathit{Dose}\hspace{0.17em}\mathit{deposited}\hspace{0.17em}in\hspace{0.17em}\mathit{peripheral}\hspace{0.17em}\mathit{airways}}\right)$. As secondary outcome measures, we investigated the differences in intrathoracic, central, and peripheral airways deposition for the different inhaler devices and inhalation maneuvers. Central airways were defined as airways with a diameter of >2 mm, and peripheral airways (or small airways) were defined as those having a diameter of ≤2 mm.

Functional respiratory imaging

Patient-specific airway models

For the patient-specific 3D airway models, we selected 20 spirometer-controlled chest CT scans from a larger cohort of children with SA.⁹ All children were receiving medium-to-high doses of ICS with at least one other controller, such as long-acting beta-agonists (LABA) or leukotriene receptor antagonists. Despite optimization of modifiable factors and inhaled treatment, these children had persistent asthma symptoms and/or frequent exacerbations. In addition, the diagnosis of SA was confirmed by their treating physician. Inclusion and exclusion criteria are listed in Table 1. For the SAD group, we randomly selected 10 patients with FEF_25–75 of forced vital capacity (FVC) and at FEF₇₅ of FVC z-score below −1.645, which is considered to be the lower limit of normal (LLN). For the control group, we selected age- and gender-matched patients with forced expiratory volume in 1 second (FEV₁), FEF_25–75, FEF₇₅ and FVC z-score > LLN.

Table 1.

Inclusion and Exclusion Criteria

Inclusion criteria
Severe asthma diagnosis according to GINA guidelines,24 confirmed by treating physician
Age 5–17 years
CT scan requirements
Spirometer-controlled CT
Good CT quality
Slice thickness <1.5 mm
Control group (no SAD)
Age- and gender-matched controls
FEF25–75, FEF75, FEV1, and FVC > LLN
SAD group
FEF25–75 and FEF75 z-score < LLN (−1.645)
FVC and FEV1 > LLN
Exclusion criteria
Known pulmonary, thoracic, or cardiac abnormality prior to CT
Preterm birth in medical history
Active smoking at the time of CT

CT, computed tomography; FEF, forced expiratory flow; FEV, forced expiratory volume; FVC, forced vital capacity; LLN, lower limit of normal; SAD, small airways disease.

After the chest CT scans were selected, patient-specific 3D models of the airways and lungs were reconstructed from the inspiratory chest CT images using commercially available, validated software packages (Mimics 20.0 and 3-Matic 12.0; Materialise NV, Belgium). After automatic segmentation, the airway tree segmentation was manually checked, and missing branches were added by a FLUIDDA technician. The segmentation was performed up to the point where no distinction can be made between the intra-luminal and alveolar air, being up to the airways of 2 mm in diameter.¹⁹ After the segmentation process was completed, the airway tree is reconstructed into a 3D airway model. Patient-specific internal airflow distribution is determined by calculating the change in lung volume between inspiration and expiration. After the 3D airway models are complete, flow and particle simulations can be applied.

Inhaler device characteristics

For this study three different inhaler devices were studied: beclomethasone/formoterol (Foster^®; Chiesi, Italy), pMDI with a valved holding chamber (AeroChamber Plus Flow-Vu^®; Trudell, Canada) (pMDI/VHC), salbutamol (Ventolin^®; GlaxoSmithKline, UK), pMDI/VHC and beclomethasone/formoterol (Foster Nexthaler; Chiesi) DPI. The combination inhalers have nominal (or labeled) dose of 100 μg beclomethasone and 6 μg formoterol per actuation. The salbutamol pMDI/VHC has a nominal dose of 100 μg per actuation. Beclomethasone/formoterol and salbutamol pMDI/VHC are both registered for the treatment of asthma in children. The beclomethasone/formoterol DPI is registered for use from the age of 12 years and has shown to be an effective delivery device for treatment of adults with asthma. It has the added benefit over pMDIs that DPIs are more practical to use and are less harmful for the environment because no greenhouse gasses are emitted.^25,26

The Foster NEXThaler is not licensed for children under 12 years of age. Nevertheless, we included this device in the study as proof of concept. The MMAD, geometrical standard deviation, fine particle fraction, delivered dose, and nominal dose of the individual compounds of different inhaler devices can be found in Table 2. The nominal dose is defined as the mass of drug that is available within the aerosol generator per actuation (metered dose). The delivered dose is the mass of the drug emitted per actuation that is available for inhalation at the mouthpiece. The delivered dose differs from the nominal dose, because part of the nominal dose stays behind in the device.²⁷

Table 2.

Inhaler Device Characteristics

Device	API	MMAD (µm)	GSD	ND (µg)	DD (µg)	FPF (% of DD)
pMDI + VHC
Beclomethasone/Formoterol (foster®)	Beclomethasone	1.4	1.9	100	57	96.5
	Formoterol	1.4	1.8	6	3.2	99.4
Salbutamol (Ventolin®)	Salbutamol	2.8	1.6	100	37.8	90.7
DPI
Beclomethasone/Formoterol (NEXThaler®)	Beclomethasone	1.2	2.4	100	81.9	59.7
	Formoterol	1.6	1.9	6	5	58.6

This table provides particle characteristics for the different inhaler devices used in this study.

API, active pharmaceutical ingredient; DD, delivered dose; DPI, dry powder inhaler; FPF, fine particle fraction; GSD, geometric standard deviation; MMAD, mass median aerodynamic diameter; ND, nominal dose; pMDI + VHC, pressurized metered dose inhaler.

Inhalation profiles

Three different generic breathing profiles were created for this study. For the pMDI/VHC, we created one inhalation profile that resembles a slow and deep inhalation and one that resembles an inhalation maneuver similar to tidal breathing. For the DPI we created a profile that resembles a forceful and deep breath, as is recommended by the package leaflet.¹⁴ The three inhalation profiles were derived from available in-house (FLUIDDA NV) inhalation profiles for adults. These inhalation profiles were scaled down using age- and height-appropriate values for inspiratory flow, inspiratory time, and inhaled volume. For both the slow and deep inhalation profile and the forceful and deep inhalation profile, a 5-second breath hold was added. For the three different inhalation profiles, simulations were performed over one breathing cycle. Simulations were stopped whenever more than 99.99% of particles were deposited. An overview of the inhalation profiles is provided in Table 3.

Table 3.

Inhalation Profiles

Inhalation profile	Mean flow rate (L/min)	Max flow rate (L/min)	Inhalation time (seconds)	Inhaled volume (L)
pMDI + VHC
Slow and deep	25	30	2.4	1
Tidal breathing	14.15	31.5	1.29	0.42
DPI
Forceful and deep	40	60	1.5	1

This table shows the characteristics of the different inhalation profiles that were used in this study.

The slow and deep inhalation profile was derived from the European Respiratory Society and the International Society for Aerosols in Medicine Task Force guidelines on how to breath optimally through a pMDI/VHC. These guidelines suggest a slow inhalation of 4–5 seconds, with a flow rate of 30 L/min. As this guideline was developed for adults, we considered that an ideal inhalation for children would involve a maximum inspiration rate of 30 L/min and an inhaled volume of 1 L. The inhaled volume was derived from in vivo inhalation profiles measured with inhalers, in a similar age group as the study group.^28,29 Using these parameters in combination with in-house inhalation profiles, we developed the current inhalation profile (Fig. 2).

FIG. 2.

Inhalation flows. This figure depicts the different inhalation flows. The bold line reflects the slow and deep inhalation profile, the dashed line reflects the tidal breathing inhalation profile, and the dotted line reflects the forceful and deep inhalation profile.

For the tidal breathing inhalation profile, we used reference equations based on the mean age of the population (rounded off to a whole year), the mean height (145 cm), and a male child to estimate the respiration rate (19 respirations/min) and inspiration time of 1.29 seconds.³⁰ The mean weight of the patient population (41.6 kg) was used to calculate the tidal volume (10 mL/kg).¹⁶ For convenience purposes we chose 420 mL as tidal volume instead of 416 mL. The breathing profile had an inspiration–expiration ratio of 1:2 and a sinusoidal shape (Fig. 2).

For the forceful and deep inspiration profile, we created an idealized but realistic breathing profile based on in-house data (FLUIDDA NV) and findings in the literature. The maximum and mean inspiratory flows from devices that had similar inspiratory flow resistance were chosen because there is no literature available on the inspiratory flow for children using the NEXThaler.³¹ Based on the literature, a rapid increase to an ideal maximum inspiratory flow rate of 60 L/min, inhalation time of 1.5 seconds, and volume of 1 L was chosen.^28,29

CFD simulations

Aerosol deposition in the 3D airway models was divided into two regions: the extrathoracic region and the intrathoracic region. The extrathoracic airways include the mouth and upper airways. Matching of the upper airway model from available in-house pediatric upper airway models with the trachea was performed using the diameter of the trachea as the most important criterion. The intrathoracic airways start at the beginning of the trachea (around the sternum) and end further downstream. The combined dose deposited in the extrathoracic and the intrathoracic airways is similar to the delivered dose. The intrathoracic airways can further be divided into central airways (airways with a diameter >2 mm) and small or peripheral airways (airways with a diameter ≤2 mm) present in the peripheral regions of the lung. For the CFD analysis, a particle was considered to deposit in the peripheral airways when the particle exits the model of the central airways.

Statistical analyses

Patient characteristics and the simulated aerosol deposition data are presented as descriptive statistics. Intrathoracic, central, and peripheral deposition data are presented as percentage (%) of nominal dose. The C:P was calculated by dividing the deposition in the central airways by the deposition in the peripheral airways. Descriptive statistics are reported as mean with standard deviation or median with interquartile range when data are respectively normally or non-normally distributed. Normality testing was performed using the Kolmogorov–Smirnov test. The unpaired Student’s t-test was used to test for differences in C:P, extrathoracic, central, and peripheral deposition between the control and the SAD group. One-way analysis of variance (ANOVA) with repeated measures was performed to test for differences for intrathoracic, central, and peripheral deposition between the three different inhalation profiles in the control and the SAD group combined. The assumption of sphericity was tested using Mauchly’s test of sphericity. The sphericity assumption concerns the equality of variances of the differences between all possible pairs of related groups. If the sphericity assumption was violated, epsilon (ε) was calculated according to Greenhouse and Geisser to correct the one-way repeated measures ANOVA.^32,33 One-to-one comparisons between two different inhalation profiles (e.g., slow and deep vs. tidal breathing) were made using the paired t-test. We assumed a significance level of α = 0.05. Statistical analyses were performed with IBM SPSS statistics 28 (USA) and GraphPad Prism 8 (USA).

Results

Patient characteristics

Patient characteristics are presented in Table 4.

Table 4.

Patient Characteristics

	Control (N = 10)	SAD (N = 10)
Age in years, mean ± SD	10.5 ± 3.0	10.5 ± 3.5
Male/female, n	8/2	8/2
Height in cm, mean ± SD	145.1 ± 16.7	144.1 ± 20.1
Weight in kg, mean ± SD	43.9 ± 18.3	39.4 ± 14.2
BMI-for-age, z-score mean ± SD	0.8 ± 1.8	0.35 ± 1.1
Sensitization to inhaled allergens, n (%)	5 (50%)	8 (80%)
Medication, (mean daily dose), n
Beclomethasone/formoterol (Foster)	2 (600/36)	4 (600/36)
Beclomethasone and formoterol separate	2 (400/48)	2 (500/24)
Beclomethasone and salmeterol separate	—	1 (600/100)
Fluticasone/salmeterol (Seretide)	4 (750/100)	—
Ciclesonide and formoterol	2 (400/48)	1 (640/48)
Ciclesonide and salmeterol	—	1 (960/50)
Budesonide and formoterol	—	1 (1000/24)
Leukotriene receptor antagonists	5	7
Mepolizumab	0	1
Spirometry
FVC %predicted, median (IQR) FVC z-score, mean ± SD	101.4% (97.1–113.0) 0.3 ± 0.7	106.1% (97.9–114.6) 0.7 ± 1.2
FEV1 %predicted, median (IQR) FEV1 z-score, mean ± SD	106.0% (98.3–112.3) 0.5 ± 0.7	86.5% (82.4–93.5) −1.0 ± 0.7
FEV1/FVC %predicted, median (IQR) FEV1/FVC, mean z-score ± SD	101.2% (97.6–104.3) 0.1 ± 0.6	83.3% (75.0–85.7) −2.3 ± 0.5
FEF25–75 %predicted, median (IQR) FEF25–75, z-score, mean ± SD	107.2% (99.3–116.9) 0.2 ± 0.6	51.9% (48.2–57.0) −2.4 ± 0.4
FEF75 %predicted, median (IQR) FEF75 z-score, mean ± SD	112.5% (100.9–129.0) 0.3 ± 0.6	44.7% (34.9–50.0) −2.2 ± 0.4

This table shows the patient characteristics of all patients included in the study. In addition, this table shows the patient characteristics subdivided for severe asthma patients without SAD (control) and severe asthma patients with SAD.

BMI, body mass index; IQR, interquartile range; SD, standard deviation.

Control group versus SAD

Intrathoracic deposition was similar between the control group and the SAD group for all inhalation devices and inhalation profiles (Table 5). The C:P ratios were significantly higher for the SAD group in comparison with the control group for all inhaler devices and medications, except for the forceful and deep inhalation of beclomethasone with the DPI (p = 0.073) (Fig. 3 and Supplementary Table S1). The difference in C:P ratios is the result of ∼3.5% higher central deposition and lower peripheral deposition (Fig. 4 and Table 5). Figure 5 provides a visualization of the deposition of beclomethasone in a patient from the control group and the SAD group. Data on the extrathoracic deposition can be found in Supplementary Table S2.

Table 5.

Intrathoracic, Central, and Peripheral Deposition

Device	Inhalation profile	Intrathoracic			Central			Peripheral
		Control	SAD	p	Control	SAD	p	Control	SAD	p
Beclomethasone
pMDI + VHC	Slow and deep	46.54 (±1.59)	47.08 (±1.09)	0.383	22.95 (±2.85)	26.88 (±3.47)	0.013	23.59 (±4.09)	20.21 (±3.61)	0.660
	Tidal breathing	46.58 (±1.55)	47.19 (±1.13)	0.327	22.72 (±2.85)	26.76 (±3.57)	0.006	23.85 (±4.14)	20.43 (±3.56)	0.031
DPI	Forceful and deep	33.61 (±2.50)	34.50 (±1.86)	0.380	17.91 (±2.76)	20.74 (±3.05)	0.043	15.70 (±2.99)	13.76 (±2.53)	0.133
Formoterol
pMDI + VHC	Slow and deep	44.72 (±1.49)	45.25 (±1.04)	0.365	21.88 (±2.77)	25.64 (±3.37)	0.014	22.84 (±3.94)	19.61 (±3.48)	0.068
	Tidal breathing	44.75 (±1.45)	45.35 (±1.08)	0.310	21.65 (±2.78)	25.52 (±3.46)	0.006	23.09 (±3.98)	19.82 (±3.43)	0.065
DPI	Forceful and deep	34.11 (±2.73)	35.07 (±2.03)	0.385	18.79 (±3.02)	22.04 (±3.41)	0.037	15.32 (±3.20)	13.03 ± 2.80	0.106
Salbutamol
pMDI + VHC	Slow and deep	30.05 (±1.58)	30.35 (±1.01)	0.680	17.59 (±1.87)	20.87 (±2.77)	0.006	12.46 (±2.98)	9.47 (±3.02)	0.039
	Tidal breathing	30.17 (±1.56)	30.52 (±0.97)	0.553	17.46 (±1.87)	20.81 (±2.90)	0.006	12.71 (±3.04)	9.70 (±3.03)	0.040

This table shows the mean (and standard deviation) percentage of the nominal dose that is deposited in the intrathoracic region, the central airways, and the peripheral airways for different inhaler devices and inhalation profiles. The results are divided for patients without SAD (control) and with SAD. Bold font indicates statistical significance (p < 0.05).

FIG. 3.

Central-to-peripheral deposition ratios. This figure shows the central-to-peripheral deposition ratios (C:P) for the different inhalation profiles and inhaler devices between patients without small airway disease (control) and with small airways disease (SAD). In the graph the circles (•) represent the slow and deep inhalation profile, the squares (▪) resemble the tidal breathing inhalation profile, and the triangles(▲) represent the forceful and deep inhalation profile. In the figures, the mean and standard deviation are depicted.

FIG. 4.

Intrathoracic, central, and peripheral deposition. This figure shows the intrathoracic, central, and peripheral deposition for the different inhalation profiles and inhaler devices between patients without small airway disease (control) and with small airways disease (SAD). In the graph the circles (•) represent the slow and deep inhalation profile, the squares (▪) resemble the tidal breathing inhalation profile, and the triangles(▲) represent the forceful and deep inhalation profile. In the figures, the mean and standard deviation are depicted.

FIG. 5.

Visualization of the deposition of beclomethasone administered through a pressurized metered dose inhaler with valved holding chamber while tidal breathing in (A) a patient from the control group and (B) a patient from the small airways disease group. The higher the intensity, the higher the deposition of beclomethasone in that region.

Inhalation profiles and inhalation devices

For the comparison of intrathoracic, central, and peripheral deposition of ICS and LABA between the three different inhalation profiles, the sphericity assumption was violated and a Greenhouse–Geisser correction was applied (Supplementary Table S3). Intrathoracic, central, and peripheral deposition of ICS and LABA was significantly different between the three inhalation profiles (p < 0.001). One-to-one comparison of the inhalation profiles showed that intrathoracic, central, and peripheral deposition of ICS and LABA was significantly lower for the deep and forceful inhalation in comparison with both the slow and deep and tidal breathing inhalation profiles (exact p-values are presented in Supplementary Table S4). Intrathoracic deposition was similar for ICS and LABA between the slow and deep and tidal breathing inhalation profiles (Supplementary Table S4). We observed significant differences in central and peripheral deposition between the slow and deep and tidal breathing inhalation profiles. For salbutamol, both intrathoracic and peripheral depositions were significantly different between the slow and deep and tidal breathing inhalation profiles.

Discussion

In this study, we have shown that regional aerosol deposition is different in children with SA and SAD compared with children with SA without SAD. The most important finding of this study is that children with SAD have higher C:P ratios than patients without SAD. The increase in C:P is the result of higher central deposition and lower peripheral deposition in the SAD group. To our knowledge, this is the first study that shows that aerosol deposition in children with SA is affected by SAD.

In general, structural abnormalities have been shown to negatively affect aerosol deposition.¹² In patients with SA, aerosol deposition in the central airways is increased and reduced in the smaller peripheral airways.^34,35 This increase in central airways deposition is likely to be the result of pathophysiological processes involved in asthma that lead to a decrease in the internal airway cross-sectional area of central airways. A reduction in airway cross-sectional area may elicit changes in flow regime that affect the deposition of aerosols in the lungs. The hypothesis that a decrease in airways cross section plays an important role in increased central deposition is supported by studies that show that induced airway narrowing by bronchoprovocation increases deposition in the central airways.^34,36

In the small airways of patients with SA, a large variety of structural airway changes has been observed. These structural alterations involve changes to the epithelium, extracellular matrix composition, airway smooth muscle mass, and/or airway-parenchymal tethering forces.² These small airways changes are considered to reduce the airway cross-sectional area and, as a result, reduce the flow of air through them. Reduced end-expiratory flows are considered to be sensitive indicators of small airways abnormalities.⁶ The limitation of flow through the small airways could result in enhanced central airways deposition and decreased peripheral airways deposition. The finding that C:P ratios in children with SA and SAD are higher than those in children without SAD supports this hypothesis.

Previous studies have shown that particle size is an important determinant of aerosol deposition in the lungs. Particles that are larger are mostly affected by the reduction in airway cross section, as they are more likely to deposit in the central airways through inertial impaction at high turbulent airflows.^17,37 In our study, we observed that C:Ps were highest for the inhaler device with the larger sized particles, being salbutamol pMDI/VHC. Patients with SA and SAD may therefore benefit more from treatment with smaller sized particles.

Besides particle size, the inhalation maneuver is an important factor that determines the extent and site of drug deposition in the lungs. Previous studies in children with asthma and in healthy adults have shown that a slow and deep inhalation followed by a breath-hold is superior over tidal breathing in terms of whole lung deposition.^38,39 Surprisingly, in our study we did not observe a difference in intrathoracic deposition between the slow and deep inhalation and tidal breathing inhalation profile. However, we did observe differences in central and peripheral deposition between the two inhalation profiles. CFD models show that during inhalation the majority of aerosols deposit in the first few seconds of the inhalation. The similarity in slope of the two inhalation profiles could therefore well explain why we did not observe a difference in intrathoracic deposition between these two inhalation profiles. Future studies including inhaler-specific and patient-specific inhalation profiles may demonstrate whether these inhalation maneuvers are interchangeable.

In our study, a forceful and deep inhalation using a DPI resulted in lower intrathoracic deposition and central and peripheral deposition, especially if you consider that the delivered dose of the DPI was much higher than the delivered dose of the pMDI/VHC. For the forceful and deep inhalation profile, we utilized the same inhaler device and inhalation parameters that were used previously in a FRI study in adults with asthma by Watz et al.¹⁹ Intrathoracic deposition in our study was similar to what was found in that study. However, the C:P for both ICS and LABA was approximately two times lower (0.60 vs. 1.20) than what we found in children with SA without SAD.¹⁹ The discrepancy in findings between these studies can have several explanations. First, the airways of adults are larger than those of children. Particles are therefore more likely to bypass the larger airways in adults in comparison to children. Second, as asthma severity worsens, the presence of structural airway abnormalities increases. As described earlier, structural airway abnormalities may augment central airways deposition. Our study included children with SA, while the study by Watz et al. included adults with various degrees of asthma. The discrepancy in findings could therefore well be the result of differences in asthma severity in combination with the presence of smaller airways.

Moreover, the beclomethasone/formoterol DPI is not registered for children under the age of 12; therefore there are no breathing profiles available for children for this device. As a solution, we chose age-appropriate inhalation profile characteristics based on instructions for use and other DPIs registered for children with similar internal resistance.^28,29 The use of these inhalation profiles could have resulted in the poor performance of the DPI in comparison with the pMDI/VHC. However, previous studies have shown the benefit of a slow and deep inhalation on aerosol deposition; we believe, therefore, that the inhalation maneuver itself is the most important explanation for the poorer performance of the DPI.³⁹

Limitations

This study has some limitations. The first and most important limitation is the lack of inhaler-specific and patient-specific inhalation profiles for the different inhalation profiles in this study. Ideally, we would have included individually recorded inhalation profiles of all patients who were included in this study. However, the CT scans that were used in this study are part of a larger cohort of CTs from children with SA that were retrospectively collected.⁹ It was therefore not possible to record and include patient-specific inhalation profiles. To overcome this limitation, we chose to scale down available in-house inhalation profiles for adults using age- and height-appropriate values for inspiratory flow, inspiratory time, and inhaled volume. Despite our efforts to use realistic inhalation profiles, it could be that they are not completely representative of the real-life situation. Regardless of the inhalation profiles, we have observed significant differences in C:P between the control and the SAD group.

A second limitation of this study is that it contains a relatively small sample of chest CT scans of children with SA. We chose to include 20 randomly selected CTs as this was similar to the number of CTs included in previous studies.^19,23 The small sample size per group may have resulted in a reduced power to detect small differences and might explain why we were not able to detect a significant difference between the control and SAD group for peripheral airway deposition. Despite the small sample size, we were able to detect a significant difference between the SAD group and the control group in C:P and central airways deposition.

A third limitation, or an aspect that warrants careful consideration, is that in our study we used a single type VHC. It is well known that VHCs are known to be noninterchangeable. Therefore, our findings may not be extrapolated for different VHCs.

Finally, a limitation of the study is related to the FRI technique itself. Despite that FRI is a well-validated technique, it is not possible for FRI models to fully replicate the real-life situation, as it does not include the complex physics associated with aerosol generation.^17,40 Despite this limitation, FRI offers a number of advantages of conventional techniques to study aerosol deposition such as the ability to perform multiple simulation on the same patients and the ability to determine deposition at specific anatomical sites, and there is no additional exposure to radiation.

Implications

The presence of SAD in patients with SA is associated with poor asthma control, the use of systemic corticosteroids, and asthma exacerbations. The small airways are therefore an important target for treatment. The results of our study suggest that children with SAD have ∼3.5% higher central airways deposition and 2.9% lower peripheral airways deposition than children without SAD. This seems like a marginal difference; however, for a patient who uses a daily dose of 600/36 μg beclomethasone/formoterol, this would imply that per day patients with SAD receive ∼21/1.3 μg beclomethasone/formoterol more in the central airways and 17.4/1.0 μg beclomethasone/formoterol less in the smaller peripheral airways. As the total cross-sectional area increases exponentially with every consecutive generation from ∼2.5 cm² at the trachea to ∼2 m² at the small airways, the daily decrease in peripheral deposition may have a substantial effect.⁴¹ Previous studies have shown the clinical benefit of increased small airways deposition on asthma control, markers of inflammation, and small airways function.^42,43 Chronic underdosing of inhaled corticosteroids and bronchodilators in patients with SA and SAD may result in inadequate control of inflammation, risk of progressive airway remodeling, and increased bronchoconstriction. As the effectivity of inhaled corticosteroids is dependent on multiple factors such as the dose, pulmonary residence time, and potency (which is further determined by the receptor binding affinity),⁴⁴ the exact effect of this decrease in deposition is difficult to determine.

The goal of asthma treatment is to achieve optimal asthma control while minimizing the risk of side effects. The effective dose is therefore tailored to each patient’s specific needs. As this study included patients receiving high dose of inhaled medication, it is questionable whether further increasing the daily dose is beneficial in these patients. In an era where precision medicine has become very important in the management of SA, it would be valuable to investigate what the clinical implications are of these differences in deposition. As this study is only conducted in children with SA, future studies should investigate whether our findings are generalizable for the larger asthma population and other pulmonary diseases. Additionally, including a combination of SAD parameters, such as impulse oscillometry or multiple breath washout, in future studies may enhance the detection of SAD and help determine its effect on aerosol deposition.³

To conclude, we found that children with SA and SAD have higher C:P ratio for aerosol deposition than children with SA without SAD. The difference is bigger if the aerosols consist of larger particles or when a DPI with a forceful inhalation is used compared to tidal or slow and deep inhalation with a pMDI/VHC. These results further underline the importance of optimizing delivery of medication to the peripheral or small airways.

Abbreviations Used

3D

Three-dimensional

ANOVA

One-way analysis of variance

CFD

Computational fluid dynamic

C:P

Central-to-peripheral deposition ratio

CT

Computed tomography

DP I

Dry powder inhaler

FEV₁

Forced expiratory volume in 1 second

FRI

Functional respiratory imaging

FVC

Forced vital capacity

ICS

Inhaled corticosteroids

LABA

long-acting beta agonists

LLN

Lower limit of normal

MMAD

mass median aerodynamic diameter

PMDI/VHC

Pressurized metered dose inhaler with valved holding chamber

SAD

Small airways disease

SA

severe asthma

FEF_25–75

Forced expiratory flow between 25% and 75% of vital capacity

FEF₇₅

Forced expiratory flow at 75% of vital capacity

VC

Vital capacity

Authors’ Contributions

W.B.v.d.B.: Conceptualization, methodology, formal analysis, investigation, data curation, and writing—original draft. E.J.R.: Conceptualization, methodology, and writing—reviewing and editing. N.M.T.: Software and writing—reviewing and editing. H.A.W.M.T.: Conceptualization, methodology, resources, writing—reviewing and editing, and funding acquisition. H.M.J.: Conceptualization, methodology, resources, writing—reviewing and editing, and supervision.

Author Disclosure Statement

E.J.R. and N.M.T. have nothing to disclose. W.B.v.d.B. reports an unconditional research grant for a PhD trajectory from Vectura Group PLC during the conduct of the study (contract was signed in 2018 for the duration of 4 years). H.A.W.M.T. is Emeritus Professor at the Erasmus MC and chief medical officer at Thirona BV and reports personal fees from Thirona BV and Vertex during the conduct of the study; Erasmus MC has a patent licensed for the Perth-Rotterdam Annotated Grid Morphometric Analysis for CF scoring system. The Erasmus MC–Sophia Children’s Hospital has in the past 3 years received unconditional research grants for investigator-initiated studies from Novartis, Vectura Group PLC Grant (contract signed in 2018), and Insmed and research grants from Inovative Medicines Initiative, Cystic Fibrosis Foundation, European Cystic Fibrosis Society, and the Sophia Foundation. H.M.J. reports grants and other support from Vertex Pharmaceuticals outside the submitted work.

Funding Information

We received funding for a PhD trajectory that included this research. Unconditional research grant from Vectura Group Plc. (contract signed in 2018).