Abstract
Background
Bronchial thermoplasty (BT) is a treatment for patients with poorly controlled, severe asthma. However, predictors of treatment response to BT are defined poorly.
Research Question
Do baseline radiographic and clinical characteristics exist that predict response to BT?
Study Design and Methods
We conducted a longitudinal prospective cohort study of participants with severe asthma receiving BT across eight academic medical centers. Participants received three separate BT treatments and were monitored at 3-month intervals for 1 year after BT. Similar to prior studies, a positive response to BT was defined as either improvement in Asthma Control Test results of ≥ 3 or Asthma Quality of Life Questionnaire of ≥ 0.5. Regression analyses were used to evaluate the association between pretreatment clinical and quantitative CT scan measures with subsequent BT response.
Results
From 2006 through 2017, 88 participants received BT, with 70 participants (79.5%) identified as responders by Asthma Control Test or Asthma Quality of Life Questionnaire criteria. Responders were less likely to undergo an asthma-related ICU admission in the prior year (3% vs 25%; P = .01). On baseline quantitative CT imaging, BT responders showed less air trapping percentage (OR, 0.90; 95% CI, 0.82-0.99; P = .03), a greater Jacobian determinant (OR, 1.49; 95% CI, 1.05-2.11), greater SD of the Jacobian determinant (OR, 1.84; 95% CI, 1.04-3.26), and greater anisotropic deformation index (OR, 3.06; 95% CI, 1.06-8.86).
Interpretation
To our knowledge, this is the largest study to evaluate baseline quantitative CT imaging and clinical characteristics associated with BT response. Our results show that preservation of normal lung expansion, indicated by less air trapping, a greater magnitude of isotropic expansion, and greater within-lung spatial variation on quantitative CT imaging, were predictors of future BT response.
Trial Registry
ClinicalTrials.gov; No.: NCT01185275; URL: www.clinicaltrials.gov
Key Words: air trapping, bronchial thermoplasty, lung deformation, quantitative CT, severe asthma
FOR EDITORIAL COMMENT, SEE PAGE 755
Take-home Points.
Study Question: Do baseline radiographic and clinical characteristics exist that predict response to bronchial thermoplasty?
Results: Responders are more likely to have fewer ICU admissions and greater evidence of lung expansibility by quantitative CT scan metrics.
Interpretation: Quantitative CT imaging could be a useful tool to identify patients with severe asthma who are more likely to respond to bronchial thermoplasty.
Asthma continues to cause significant morbidity and mortality in the United States with an estimated annual cost of care of > $80 billion US dollars.1 Severe asthma accounts for nearly 80% of that cost.2 The European Respiratory Society and American Thoracic Society consensus guidelines define severe asthma based on the requirement for treatment with a high-dose inhaled corticosteroid plus a second controller, with or without systemic corticosteroids, to maintain control or patients who, despite this therapy, have suboptimally controlled disease.3 Bronchial thermoplasty (BT), which received US Food and Drug Administration approval in 2010 for the treatment of severe asthma, works by delivering radiofrequency thermal energy directly to the airway wall via bronchoscopy.4 This therapy effectively reduces airway smooth muscle mass and is believed to attenuate the effect of airway remodeling in people with severe asthma.5 BT has been shown to result in significant improvement in multiple measures of asthma control including quality of life, symptom-free days, and number of exacerbations.6, 7, 8, 9
Because the clinical response to BT can be variable and often is associated transiently with worsening asthma symptoms related to the procedure, a better understanding of the optimal patient phenotype for BT is critical.10 Prior studies have found that better baseline asthma control, greater number of BT activations during the procedure,11,12 as well as greater markers of type 2 inflammation, including greater IgE or blood eosinophil levels,13 greater mast cell interferon-α expression, and more mucosal eosinophils and mucosal IL-33 positive cells, each are associated with response to BT.14 However, each of these studies was limited by a small sample size.15 Furthermore, sputum and mucosal assessments of inflammation are not performed routinely in clinical settings.
Quantitative CT imaging of the chest has been used increasingly to characterize better the pathophysiologic features of asthma.16 Prior studies have demonstrated that quantitative CT scan measurements of airway morphometry, lung density, and lung expansion correlate with disease severity, asthma symptoms, and airway remodeling.17,18 Patients who receive BT demonstrate changes in quantitative CT scan findings, including improved air trapping, lung density, and airway wall thickness.19 However, it is not clear whether quantitative CT imaging can predict future response to BT. Therefore, we conducted a prospective cohort study of patients who underwent BT to determine whether baseline clinical and quantitative CT scan characteristics were associated with future positive response to BT. We hypothesized that a combination of quantitative CT scan and clinical features could predict clinical response to BT, as defined by improvement in the Asthma Control Test (ACT) or Asthma Quality of Life Questionnaire (AQLQ) scores.
Study Design and Methods
Study Design and Included Participants
This was a longitudinal cohort study of patients with severe asthma who received three separate BT treatments across eight academic tertiary care hospitals. The sites included in this study were (in order of greatest to least number of enrollees): Washington University in St. Louis, University of Chicago, Cleveland Clinic, Creighton University, Louisiana State University, National Jewish Health, University of Alabama, and University of Arizona. All participants provided written informed consent, and this study was approved by the institutional review board at all participating sites.
We enrolled 72 participants (Fig 1) who gave consent and completed three BT treatments as part of the Bronchial Thermoplasty Study (ClinicalTrials.gov Identifier: NCT01185275). In addition, 16 patients in the separate Bronchial Thermoplasty Registry were enrolled. The Bronchial Thermoplasty Registry comprised patients who received three BT treatments at Washington University in St. Louis and whose disease was considered too severe to be included in the Bronchial Thermoplasty Study. Participants in both cohorts were included in analyses if they had undergone at least 3 months of follow-up after the last BT treatment to assess successful BT response. All participants were required to have a clinical diagnosis of severe asthma as defined by American Thoracic Society and European Respiratory Society criteria and were receiving high-dose inhaled corticosteroid or long-acting beta-agonist therapy.3
Figure 1.
Study flow diagram. BT = bronchial thermoplasty; LSU = LSU Health Sciences Center; qCT = quantitative CT; QOL = quality of life; U of Alabama = University of Alabama; U of Arizona = University of Arizona; U of Chicago = University of Chicago; Wash U = Washington University.
Clinical Characteristics
We performed pulmonary function testing before and after bronchodilator administration and methacholine bronchoprovocation testing as needed according to American Thoracic Society guidelines to confirm the diagnosis of asthma. Maximum response to bronchodilation with short-acting beta-agonists was assessed after the administration of 360 μg, 540 μg, and 720 μg of albuterol by the measurement of FEV1 and FVC. Participants completed the AQLQ and ACT before starting BT treatment and after each follow-up visit until 12 months after the last BT treatment.
Smoking history, inhaled corticosteroid use (in beclomethasone equivalents, micrograms per day), and oral corticosteroid use (in prednisone equivalents, milligrams per day) were collected at baseline and each follow-up visit. Finally, we quantified the number of exacerbations after at least 90 days from the last BT treatment and the time to first exacerbation. Exacerbations were defined as receipt of an oral corticosteroid burst, ED visit, urgent care visit, or hospitalization for asthma-related symptoms.
Quantitative CT Imaging Analyses
Baseline CT images obtained before BT were gathered during breath holds at total lung capacity and functional residual capacity. All CT scans subsequently were analyzed using Apollo version 2.0 software (VIDA Diagnostics, Inc.). The CT scan integrity as well as the automated airway segmentations were confirmed to be appropriate by at least two trained imaging analysts. After airway segmentation confirmation, the imaging software generates quantitative CT scan calculations.
Airway-specific analyses were calculated at segmental airways across five airway pathways (RB1, RB4, RB10, LB1, and LB10). Lung density measurements included: hyperinflation percentage (percentage of voxels at ≤ –950 Hounsfield units at total lung capacity), air trapping percentage (percentage of voxels at ≤ –856 Hounsfield units at functional residual capacity), as well as disease probability measures (DPMs).20,21 Indexes of lung deformation were derived by voxel-by-voxel image coregistration of quantitative CT scans at total lung capacity and functional residual capacity. These measurements serve as a surrogate for lung function at the regional level and reflect the local volume change that occurs at a particular point from expiration to inspiration.22 The Jacobian determinant is a measurement of air volume change, and the anisotropic deformation index (ADI) is a measurement of the magnitude of directional preference in air volume change.23,24
Primary Outcome
A positive BT response was defined as either an improvement in the ACT score of ≥ 3 or the AQLQ score of ≥ 0.5 at any of the interval follow-up time points after BT. This definition for positive BT response was chosen because these values reflect the minimally clinically important differences for these questionnaires and because this is the most commonly used definition of a positive BT response in prior studies.6,7,11, 12, 13,25,26
Statistical Analysis
Data were analyzed using SPSS version 27.0 software (IBM Corporation) and SAS for Windows version 9.4 software (SAS Institute). We analyzed demographic, biological, physiologic and radiographic characteristics of participants to identify clinical predictors of response or nonresponse to BT. We compared baseline characteristics of responders and nonresponders using t tests, Wilcoxon signed-rank tests, and χ2 tests as appropriate. For variables that were nonnormally distributed, statistical analysis was conducted on the log-transformed data. Univariable and multivariable logistic regression analyses were used to evaluate the association between clinical characteristics and quantitative CT scan measures with subsequent response to BT. All multivariable models included BMI, sex, and number of follow-up time points as covariates because these were deemed to be potentially important confounders in the relationship between baseline quantitative CT imaging and response to BT. Two models (looking at DPM analyses) did not control for sex because of complete separation of data when sex was included as a covariate. We also conducted a multivariable Cox proportional hazards analysis to examine time to first exacerbation.
Results
Clinical Characteristics of BT Responders and BT Nonresponders
As shown in Figure 1, 88 participants (72 from the BT study and 16 from the BT registry) underwent at least one documented 3-month follow-up visit and thus were included in our analyses to assess for BT response. The mean ± SD age of the cohort was 48 ± 11.8 years at the time of enrollment, and participants predominately identified as female and White. Most participants required oral corticosteroids in the year before enrollment (56%) and used reliever beta-agonists on a daily or near-daily basis (92%). Given the time that the study took place, most participants were not currently and had never used an asthma-specific biologic therapy (67% had never used anti-IgE therapy, 96% had never used an anti-IL-5[Rα], anti-IL-4Rα, or anti-thymic stromal lymphopoietin agent).
Seventy patients (79.5%) met criteria for a positive BT response, whereas 18 patients (20.5%) were considered nonresponders (Table 1). The mean follow-up time for the entire cohort was 385 ± 131 days, with no difference in follow-up time between responders and nonresponders (376 days vs 419 days; P = .22). Fourteen participants were classified as intermittent responders (positive response at some point). This is in line with previously published data on the durability of BT response.27 Blood eosinophil levels, serum IgE levels, and exhaled nitric oxide readings were not different between the two groups, although analyses were limited by small sample size. BT responders underwent significantly fewer ICU admissions in the prior 12 months than nonresponders (3% vs 25%; OR, 0.12; 95% CI, 0.02-0.75; P = .02). Finally, responders showed a trend toward more activations across all BT procedures than nonresponders (mean, 307 activations vs 257 activations; OR, 1.01; 95% CI, 1.00-1.01; P = .08).
Table 1.
Baseline Characteristics of BT Responders and Nonresponders
| Characteristic | BT Responder (n = 70) | BT Nonresponder (n = 18) | P Valuea | OR (95% CI) |
|---|---|---|---|---|
| Demographics | ||||
| Age at study enrollment, y | 48.6 ± 12.6 | 48.0 ± 11.0 | .85 | 1.02 (0.98-1.07) |
| Sex, % female | 49 (70) | 13 (72) | .85 | 1.46 (0.43-4.90) |
| Race, % non-White | 11 (16) | 6 (33) | .10 | 2.83 (0.83-9.69) |
| BMI, kg/m2 | 35.1 ± 8.6 | 35.4 ± 10.4 | .92 | 0.99 (0.93-1.05) |
| Asthma characteristics | ||||
| Asthma duration, y | 25.8 ± 16.6 | 23.0 ± 15.3 | .55 | 1.01 (0.98-1.05) |
| Allergy status, % atopic | 34 (89) | 5/6 (83) | .54 | 0.53 (0.05-6.14) |
| > 3 oral corticosteroid courses in year before enrollment, % yes | 29 (47) | 7 (41) | .68 | 1.06 (0.34-3.35) |
| Hospitalized for asthma in prior year, % yes | 19 (45) | 8 (67) | .19 | 0.43 (0.11-1.74) |
| ICU admission for asthma in prior year, % yes | 2 (3) | 4 (25) | .01b | 0.12 (0.02-0.75) |
| ACT score | 11.1 ± 4.0 | 11.7 ± 4.7 | .57 | 0.97 (0.85-1.11) |
| AQLQ score | 3.2 ± 1.0 | 3.8 ± 1.3 | .03b | 0.62 (0.38-1.0) |
| PC20, mg/mL | 1.5 ± 3.1 | 0.6 ± 0.7 | .51 | 2.15 (0.30-15.51) |
| Use of beta-agonist inhaler on a daily or near-daily basis, % yes | 58/61 (95) | 12/15 (80) | .09 | 6.92 (1.05-45.5) |
| Ever used oral corticosteroids, % yes | 37/67 (55) | 11/18 (61) | .65 | 1.25 (0.39-4.03) |
| Any anti-IgE therapy, % yes | 23/67 (34) | 5/18 (28) | .60 | 1.11 (0.34-3.71) |
| Ever used anti-IL-5 therapy, % yes | 3/43 (7) | 0/13 (0) | > .99 | NA |
| Blood eosinophils, absolute, K/mm3 | 0.26 ± 0.19 | 0.35 ± 0.43 | .61 | 0.39 (0.02-8.19) |
| Serum IgE, IU/mL | 297.2 ± 660.4 | 138.8 ± 161.9 | .47 | 1.17 (0.79-1.72) |
| Feno, ppb | 47.3 ± 52.4 | 65.4 ± 64.9 | .21 | 0.65 (0.25-1.66) |
| Spirometry findings | ||||
| FEV1, % predicted | 68.36 ± 20.63 | 63.55 ± 20.71 | .40 | 1.01 (0.98-1.04) |
| FVC, % predicted | 80.35 ± 16.31 | 76 ± 20.21 | .36 | 1.01 (0.98-1.05) |
| BT characteristics | ||||
| Procedure 1, no. of activations, 0-300 | 98.7 ± 43 | 77.1 ± 28.2 | .05 | 1.01 (1.0-1.03) |
| Procedure 2, no. of activations, 0-300 | 92.0 ± 38.8 | 82.3 ± 34.1 | .34 | 1.01 (0.99-1.02) |
| Procedure 3, no. of activations, 0-300 | 117.8 ± 37.4 | 97.3 ± 39.9 | .05 | 1.01 (1.0-1.03) |
| Total no. of activations across all BT procedures | 306.6 ± 108.1 | 256.7 ± 90.1 | .08 | 1.00 (1.0-1.01) |
ACT = Asthma Control Test; AQLQ = Asthma Quality of Life Questionnaire; BT = bronchial thermoplasty; Feno = fractional expired nitric oxide; PC20 = provocation concentration causing a 20% fall in FEV1; ppb = parts per 1 billion.
P value from t test or χ2 analysis, as appropriate.
P < .05.
Quantitative CT Scan Characteristics of BT Responders and Nonresponders
Table 2 describes the baseline quantitative CT scan characteristics of BT responders and nonresponders. A trend toward lower DPM air trapping percentage was found in responders, although this did not reach statistical significance (percentage of DPM air trapping mean ± SD, 22.7 ± 14.7% vs 35.2 ± 19.2%; OR, 0.96; 95% CI, 0.91-1.01; P = .05). The mean Jacobian determinant was higher in BT responders (mean ± SD, 2.00 ± 0.50 vs 1.53 ± 0.36; OR, 1.32; 95% CI, 1.00-1.74; P = .02). The within-lung SD of Jacobian determinant was higher in BT responders (mean ± SD, 0.62 ± 0.3 vs 0.37 ± 0.16; OR, 1.61; 95% CI, 0.99-2.63; P = .023). Finally, mean ADI (mean ± SD, 0.50 ± 0.16 vs 0.37 ± 0.10; OR, 2.22; 95% CI, 0.973-5.07; P = .036) and within-lung SD of ADI (mean ± SD, 0.36 ± 0.1 vs 0.28 ± 0.08; OR, 2.91; 95% CI, 0.95-8.96; P = .037) also were increased among responders. This is depicted in Figure 2.
Table 2.
Quantitative CT Scan Characteristics of BT Responders and Nonresponders
| Baseline Quantitative CT Scan Characteristics | Responders (n = 40) | Nonresponders (n = 9) | P Valuea | OR (95% CI) |
|---|---|---|---|---|
| Airway measurements | ||||
| WA, %, mean ± SD (No). | 62.8 ± 3.1 (40) | 64.8 ± 3.1 (9) | .08 | 0.82 (0.58-1.15) |
| WT, %, mean ± SD (No.) | 17.2 ± 1.2 (40) | 17.9 ± 1.2 (9) | .10 | 0.65 (0.32-1.29) |
| Lumen area, mm2, mean ± SD (No.) | 17.6 ± 4.4 (40) | 14.8 ± 6.6 (9) | .12 | 1.12 (0.93-1.34) |
| Eccentricity, mean ± SD (No.) | 0.8 ± 0.05 (40) | 0.7 ± 0.08 (9) | .40 | 1.41 (0.40-4.93) |
| Lung density measurements | ||||
| < 950 HU, %, mean ± SD (No.) | 4.0 ± 3.8 (40) | 2.7 ± 2.1 (9) | .32 | 1.37 (0.88-2.14) |
| Air trapping, %, mean ± SD (No.) | 10.7 ± 12.1 (39) | 18.0 ± 23.8 (8) | .42 | 0.99 (0.94-1.04) |
| fSAD DPM, %, mean ± SD (No.) | 22.7 ± 14.7 (39) | 35.2 ± 19.2 (7) | .05 | 0.96 (0.91-1.01) |
| Hyperinflation DPM %, mean ± SD (No.) | 2.9 ± 5.6 (39) | 1.9 ± 3.4 (7) | .63 | 1.09 (0.82-1.46) |
| Lung deformation measurements | ||||
| Jacobian, mean ± SD (No.) | 2.0 ± 0.5 (39) | 1.5 ± 0.4 (8) | .02b | 1.32 (1.00-1.74) |
| JacobianSD, mean ± SD ( No.) | 0.6 ± 0.3 (39) | 0.4 ± 0.2 (8) | .02b | 1.61 (0.99-2.63) |
| ADI, mean ± SD (No.) | 0.5 ± 0.2 (39) | 0.4 ± 0.1 (8) | .04b | 2.2 (0.97-5.07) |
| ADI SD, mean ± SD (No.) | 0.4 ± 0.1 (37) | 0.3 ± 0.1 (8) | .04b | 2.91 (0.95-8.96) |
ADI = anisotropic deformation index; BT = bronchial thermoplasty; DPM = disease probability mapping; fSAD = functional small airways disease; HU = Hounsfield unit; WA = wall area; WT = wall thickness.
t test or χ2 test, as appropriate. Lung density and lung deformation measurements were not calculated if expiratory images were not obtained. Two patients with extreme outliers for ADI SD were excluded.
P < .05.
Figure 2.
A, B, Sample quantitatively analyzed CT scans showing coronal and sagittal views of a BT responder (A) and BT nonresponder (B). The top row for each part shows disease probability mapping values tinted on a red-to-yellow scale based on hyperinflation (red) or air trapping (yellow) predominance. Furthermore, in these images, the degree of normality of the tissue at the voxel level is indicated by the transparency wherein normal pixels have no coloration (ie, purely gray scale). The bottom row for each section demonstrates lung deformation values tinted on a red-to-yellow scale based on the anisotropic deformation index (ADI) value, with red demonstrating an ADI value of 0.0 (perfectly isotropic deformation) and yellow having an ADI of 1.0. In this section, the transparency is indicated by the Jacobian determinant value wherein a Jacobian determinant of 1.0 has pure coloration (no gray scale) and a Jacobian determinant of 2.5 has no coloration (ie, purely gray scale). Images were made in collaboration with VIDA Imaging (VIDA Diagnostics, Inc.). BT = bronchial thermoplasty.
Multivariable Predictors of BT Responders and Nonresponders Including Baseline Clinical and Quantitative CT Scan Characteristics
Table 3 demonstrates the predictors of BT response on multivariable analyses after controlling for sex, BMI, and number of follow-up time points. After adjusting for these potential confounders, several quantitative CT scan measurements predicted a positive response to BT. Patients with higher baseline air trapping percentage were less likely to be responders (OR, 0.90; 95% CI, 0.82-0.99; P = .03). Those with a greater baseline mean Jacobian determinant, indicating greater baseline lung deformation, were more likely to show a BT response (OR, 1.49; 95% CI, 1.05-2.11; P = .03). Similarly, participants who had a larger SD of the Jacobian determinant, indicating greater heterogeneity in lung deformation, were more like to show a BT response (OR, 1.84; 95% CI, 1.04-3.26; P = .04). Findings were similar for mean ADI (OR, 3.06; 95% CI, 1.06-8.86; P = .04), and a similar trend was observed for SD of ADI (OR, 4.06; 95% CI, 0.96-17.23; P = .06), although the result did not reach statistical significance. This indicates that patients who have preserved lung expansion are more likely to respond to BT. In the multivariable analysis, we added BT activation count as a covariate (e-Table 1) and eosinophil count (e-Table 2) individually to the model. No significant difference in BT response was found with each of these covariates. Also, no difference in responder status was found across the study sites (e-Table 3). Sensitivity analyses excluding intermittent responders were conducted and found that air trapping percentage and Jacobian determinant still were associated significantly with BT response (e-Table 4).
Table 3.
BT Response After Controlling for Baseline Clinical and Quantitative CT Scan Characteristics
| Baseline Quantitative CT Scan Characteristics | OR | P Value |
|---|---|---|
| Airway measurements | ||
| WA, % | 0.80 (0.57-1.13) | .20 |
| WT, % | 0.68 (0.33-1.40) | .29 |
| Lumen area | 1.17 (0.94-1.46) | .15 |
| Eccentricity | 2.54 (0.56-11.54) | .23 |
| Lung density measurements | ||
| < 950 HU, % | 1.25 (0.73-2.13) | .41 |
| Air trapping, % | 0.90 (0.82-0.99) | .03a |
| fSAD DPM, %b | 0.94 (0.89-1.01) | .07 |
| Hyperinflation DPM, %b | 1.00 (0.78-1.27) | .97 |
| Lung deformation measurements | ||
| Jacobian mean | 1.49 (1.05-2.11) | .03a |
| Jacobian SD | 1.84 (1.04-3.26) | .04a |
| ADI mean | 3.06 (1.06-8.86) | .04a |
| ADI SD | 4.06 (0.96-17.23) | .06 |
BMI, sex, and number of follow-up time points were included as covariates because these were deemed to be important confounders in the relationship between baseline quantitative CT imaging and response to BT. ADI = anisotropic deformation index; BT = bronchial thermoplasty; DPM = disease probability mapping; fSAD = functional small airways disease; HU = Hounsfield unit; WA = wall area; WT = wall thickness.
P < .05.
Model did not control for sex because of complete separation of data when including sex as a covariate.
Exploratory Analyses of Positive BT Response Defined as ACT or AQLQ Improvement or Time to Asthma Exacerbation
We repeated analyses using ACT, AQLQ, or time to first exacerbation to define response. The number of responders meeting AQLQ criteria only (n = 69) was almost exactly the same as the overall number of responders (n = 70), whereas only 58 participants showed a positive response based on ACT score alone. Univariable analysis of CT scan characteristics associated with response based on the AQLQ score alone was similar to response based on either ACT score or AQLQ score. Using ACT score only, the Jacobian determinant and Jacobian SD still were predictive of a positive BT response (e-Table 5). We also conducted a multivariable Cox proportional hazards analysis of time to first exacerbation (e-Table 6) and found no significant associations between the CT scan metrics and time to exacerbation.
Discussion
This study was, to our knowledge, the largest study to date that evaluated if baseline clinical and quantitative CT characteristics could predict a future positive response to BT. We found that patients with higher values for the indexes of lung expansion (Jacobian determinant and ADI) were more likely to respond to BT. This suggests that those with most severe asthma are less likely to respond to BT, potentially because of abnormal lung mechanics. The diminished expansibility of BT nonresponders' lungs may be the result of more severe air trapping and hyperinflation, which was also higher in nonresponders. The Jacobian determinant has been shown to reflect lung function and has been studied extensively in obstructive lung disease, showing an inverse correlation with air trapping (ie, patients with more lung expansion as defined by the Jacobian determinant show less air trapping).28
Our findings suggest that a point may exist at which the lungs of patients with severe asthma become too remodeled and are no longer capable of a degree of normal expansibility, making BT less effective. The mechanism behind why BT is more likely to benefit those with higher lung deformation is unclear without pathologic correlation. Given the mechanism of action of BT in reducing airway smooth muscle,13 it is likely those with less distensibility have less smooth muscle or other factors contributing to reduced responsiveness. Our quantitative CT scan findings suggest a mechanical property of certain patients with severe asthma that is associated with BT response. We know that the phenotypes of individuals with severe asthma,29 and thus additional unmeasured confounders could be associated with both the baseline quantitative CT scan changes as well as a positive BT response.
The concept of lung deformation stems from the idea that two CT images can be used to recreate functional lung assessments of motion as opposed to static assessments of air trapping.28 These metrics previously were studied in patients with asthma and COPD. Bodduluri et al30 demonstrated that the Jacobian determinant was associated significantly with 6-min walk distance; lung function; and BMI, airflow obstruction, dyspnea, and exercise capacity index. Low Jacobian levels also have been associated with future asthma exacerbations and subsequent lung function decline.24 Our study demonstrated that baseline indexes of lung deformation may be useful not only in predicting clinical outcomes, but also in determining the likelihood of a positive response to specific treatments.
A handful of prospective trials predating ours also have attempted to evaluate clinical factors associated with BT response. Goorsenberg et al13 used ACT and AQLQ scoring 6 months after BT to assess response and found that higher IgE levels and eosinophil levels were associated with BT response. Langton et al12 used a change in ACQ score at 6 months to define response and found that a higher ACQ score was associated with BT response. Wang et al31 also used ACQ score improvement at 6 months to define response and found that a higher ACQ score, greater FEV1 % predicted, and higher number of activations all predicted response. Similar to these studies, we also used a change in asthma control using the ACQ score to define a positive BT response.
Our study differs substantially from prior work in that we used quantitative CT imaging to examine not only airway morphometry and lung density, but also lung expansion, to predict BT response. Quantitative CT imaging can be carried out at any facility that has the capacity to do inspiratory and expiratory CT imaging; previous publications have demonstrated that the training is relatively straightforward with a few key instructions to the patient during the CT scan.32 Quantitative CT imaging is not performed often, but it can be deployed widely, and studies like this show the value of having both inspiratory and expiratory data at minimal additional cost and radiation exposure.
Our study has several strengths. First, this was a multicenter trial examining a well-characterized group of patients undergoing BT over at least a 10-year period. Our study was also enriched by the inclusion of BT Registry patients, which comprises sicker patients. Finally, we performed quantitative CT scan analyses using standardized protocols with regular quality checks to ensure the accuracy of the imaging data. To our knowledge, this is the largest attempt at characterization of a BT responder phenotype using quantitative CT scan analysis that we anticipate will be available to clinicians in the near future.
Some of the limitations of this study include our definition of response. Definitions of response to BT have been heterogenous and range from improvement in asthma-related quality of life, decrement in exacerbations, decreased inhaled or oral corticosteroid use, or a combination thereof.11,12,14 Another limitation is that few participants received biologic therapy before or during the course of the study, which differs from real-world practice with severe asthma today. At the time of our study (2006-2017), omalizumab was the biologic therapy most commonly used. Participants were selected for the presence of bronchial hyperreactivity as in previous BT trials.6,7,13 Patients predominantly were White and female, which limits the generalizability of our results.11,12 Finally, our analyses were limited by our small sample size, which made controlling for multiple comparisons difficult. Although our results are encouraging, the limited availability of quantitative CT scan analysis likely limits its widespread clinical use outside of academic medical centers.
The National Asthma Education and Prevention Program was the first to address BT formally in their 2020 guidelines and conditionally recommended against BT, but acknowledged that it has a role in registries and clinical trials that track the long-term safety and effectiveness of BT.33 Boston Scientific recently discontinued manufacturing the BT catheter. Although the BT catheter may have limited usefulness based on market forces, we believe BT still has clear clinical usefulness for carefully selected patients with severe asthma, especially when they are not a candidate for biologic therapy or fail to respond. This study is another step to help clinicians and researchers better define patients who are more likely to benefit from BT.
Interpretation
In conclusion, we demonstrated that some quantitative CT scan metrics of lung expansion are associated with a positive response to BT. Our work suggests that as the airways of patients with severe asthma become more remodeled, a threshold exists at which BT is less likely to be effective. Further work is needed to identify that threshold and its application in clinical decision-making better.
Funding/Support
Research reported in this publication was supported in part by the National Center for Advancing Translational Sciences, National Institutes of Health [Grants U10-HL109257, U01HL146002, and KL2 TR002346 (J. G. K.)].
Financial/Nonfinancial Disclosures
The authors have reported to CHEST the following: J. G. K. reports personal fees and nonfinancial support from Genentech and Sanofi. C. S. H. has received speaking fees from VIDA, Boehringer Ingelheim, Polarean. A. S. has received consulting fees from Entana Pharmaceuticals. S. P. is an employee and stock option holder of VIDA. D. K. H. is a consultant for Boston Scientific. J. T. is on the Astra Zeneca Study steering committee. M. E. W. has received consulting, advisory, or speaking honoraria from Amgen, AstraZeneca, Avalo Therapeutics, Boehringer Ingelheim, Cerecor, Cohero Health, Cytoreason, Eli Lilly, Equillium, Glaxosmithkline, Incyte, Kinaset, Novartis, Om Pharma, Overtone Therapeutics/Foresite Labs, Phylaxis, Pulmatrix, Rapt Therapeutics, Regeneron, Restorbio, Roche/Genentech, Sanofi/Genzyme, Sentien, Sound Biologics, Tetherex Pharmaceuticals, Teva, and Upstream Bio. M. C. reports institutional grant funding from the National Institutes of Health, American Lung Association, Patient Centered Outcomes Research Institute, AstraZeneca, GSK, Novartis, Pulmatrix, Sanofi-Aventis, and Shionogi; consulting fees from Genentech, Teva, Sanofi-Aventis, Merck, Novartis, Arrowhead OM Pharma, and Allakos; payment for speaker’s bureau activities for Amgen, AstraZeneca, Genentech, GSK, Regeneron, Sanofi-Aventis, and Teva; and royalties from Elsevier. None declared (M. S., D. L., C. W. G., T. K., M. C. M., J. B., K. B. S., S. E., Z. D. L. E. M., R. T., A. S., X. S., J. C.).
Acknowledgments
Author contributions: M. S. and M. C. are the guarantors of the paper, taking responsibility for the integrity of the work from inception to publication. M. C. conceived of and designed the study in its entirety. D. L. and C. G. performed the primary statistical analysis and made significant contributions to the design of the study and interpretation of the data. S. P. was responsible for the interpretation of the imaging analysis. M. S. prepared the first draft of the manuscript. All authors revised the draft critically for intellectual content. All authors provided approval of the final manuscript version.
Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Additional information: The e-Tables are available online under “Supplementary Data.”
Supplementary Data
References
- 1.Nurmagambetov T., Kuwahara R., Garbe P. The economic burden of asthma in the United States, 2008-2013. Ann Am Thorac Soc. 2018;15(3):348–356. doi: 10.1513/AnnalsATS.201703-259OC. [DOI] [PubMed] [Google Scholar]
- 2.Moore W.C., Bleecker E.R., Curran-Everett D., et al. Characterization of the severe asthma phenotype by the National Heart, Lung, and Blood Institute’s Severe Asthma Research Program. J Allergy Clin Immunol. 2007;119(2):405–413. doi: 10.1016/j.jaci.2006.11.639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chung K.F., Wenzel S.E., Brozek J.L., et al. International ERS/ATS guidelines on definition, evaluation and treatment of severe asthma. Eur Respir J. 2014;43(2):343–373. doi: 10.1183/09031936.00202013. [DOI] [PubMed] [Google Scholar]
- 4.Doeing D.C., Mahajan A.K., White S.R., Naureckas E.T., Krishnan J.A., Hogarth D.K. Safety and feasibility of bronchial thermoplasty in asthma patients with very severe fixed airflow obstruction: a case series. J Asthma. 2013;50(2):215–218. doi: 10.3109/02770903.2012.751997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pretolani M., Dombret M.C., Thabut G., et al. Reduction of airway smooth muscle mass by bronchial thermoplasty in patients with severe asthma. Am J Respir Crit Care Med. 2014;190(12):1452–1454. doi: 10.1164/rccm.201407-1374LE. [DOI] [PubMed] [Google Scholar]
- 6.Cox G., Thomson N.C., Rubin A.S., et al. Asthma control during the year after bronchial thermoplasty. N Engl J Med. 2007;356(13):1327–1337. doi: 10.1056/NEJMoa064707. [DOI] [PubMed] [Google Scholar]
- 7.Castro M., Rubin A.S., Laviolette M., et al. Effectiveness and safety of bronchial thermoplasty in the treatment of severe asthma: a multicenter, randomized, double-blind, sham-controlled clinical trial. Am J Respir Crit Care Med. 2010;181(2):116–124. doi: 10.1164/rccm.200903-0354OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Thomson N.C., Rubin A.S., Niven R.M., et al. Long-term (5 year) safety of bronchial thermoplasty: Asthma Intervention Research (AIR) trial. BMC Pulm Med. 2011;11:8. doi: 10.1186/1471-2466-11-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chupp G., Laviolette M., Cohn L., et al. Long-term outcomes of bronchial thermoplasty in subjects with severe asthma: a comparison of 3-year follow-up results from two prospective multicentre studies. Eur Respir J. 2017;50(2):01140–02017. doi: 10.1183/13993003.00017-2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tan L.D., Yoneda K.Y., Louie S., Hogarth D.K., Castro M. Bronchial thermoplasty: a decade of experience: state of the art. J Allergy Clin Immunol Pract. 2019;7(1):71–80. doi: 10.1016/j.jaip.2018.08.017. [DOI] [PubMed] [Google Scholar]
- 11.Langton D., Sha J., Ing A., Fielding D., Thien F., Plummer V. Bronchial thermoplasty: activations predict response. Respir Res. 2017;18(1):134. doi: 10.1186/s12931-017-0617-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Langton D., Wang W., Sha J., et al. Predicting the response to bronchial thermoplasty. J Allergy Clin Immunol Pract. 2020;8(4):1253–1260 e2. doi: 10.1016/j.jaip.2019.10.034. [DOI] [PubMed] [Google Scholar]
- 13.Goorsenberg A.W.M., d’Hooghe J.N.S., Srikanthan K., et al. Bronchial thermoplasty induced airway smooth muscle reduction and clinical response in severe asthma. The TASMA randomized trial. Am J Respir Crit Care Med. 2021;203(2):175–184. doi: 10.1164/rccm.201911-2298OC. [DOI] [PubMed] [Google Scholar]
- 14.Ladjemi M.Z., Di Candia L., Heddebaut N., et al. Clinical and histopathologic predictors of therapeutic response to bronchial thermoplasty in severe refractory asthma. J Allergy Clin Immunol. 2021;148(5):1227–1235.e6. doi: 10.1016/j.jaci.2020.12.642. [DOI] [PubMed] [Google Scholar]
- 15.Svenningsen S., Cox G., Nair P. Eosinophilia and response to bronchial thermoplasty. Am J Respir Crit Care Med. 2021;203(1):148. doi: 10.1164/rccm.202008-3221LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Trivedi A., Hall C., Hoffman E.A., Woods J.C., Gierada D.S., Castro M. Using imaging as a biomarker for asthma. J Allergy Clin Immunol. 2017;139(1):1–10. doi: 10.1016/j.jaci.2016.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Busacker A., Newell J.D., Jr., Keefe T., et al. A multivariate analysis of risk factors for the air-trapping asthmatic phenotype as measured by quantitative CT analysis. Chest. 2009;135(1):48–56. doi: 10.1378/chest.08-0049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Choi S., Hoffman E.A., Wenzel S.E., et al. Quantitative assessment of multiscale structural and functional alterations in asthmatic populations. J Appl Physiol (1985) 2015;118(10):1286–1298. doi: 10.1152/japplphysiol.01094.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zanon M., Strieder D.L., Rubin A.S., et al. Use of MDCT to assess the results of bronchial thermoplasty. AJR Am J Roentgenol. 2017;209(4):752–756. doi: 10.2214/AJR.17.18027. [DOI] [PubMed] [Google Scholar]
- 20.Pompe E., van Rikxoort E.M., Schmidt M., et al. Parametric response mapping adds value to current computed tomography biomarkers in diagnosing chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2015;191(9):1084–1086. doi: 10.1164/rccm.201411-2105LE. [DOI] [PubMed] [Google Scholar]
- 21.Kirby M., Yin Y., Tschirren J., et al. A novel method of estimating small airway disease using inspiratory-to-expiratory computed tomography. Respiration. 2017;94(4):336–345. doi: 10.1159/000478865. [DOI] [PubMed] [Google Scholar]
- 22.Choi S., Hoffman E.A., Wenzel S.E., et al. Registration-based assessment of regional lung function via volumetric CT images of normal subjects vs. severe asthmatics. J Appl Physiol (1985) 2013;115(5):730–742. doi: 10.1152/japplphysiol.00113.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Choi S., Haghighi B., Choi J., et al. Differentiation of quantitative CT imaging phenotypes in asthma versus COPD. BMJ Open Respir Res. 2017;4(1) doi: 10.1136/bmjresp-2017-000252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Krings J.G., Goss C.W., Lew D., et al. Quantitative CT metrics are associated with longitudinal lung function decline and future asthma exacerbations: results from SARP-3. J Allergy Clin Immunol. 2021;148(3):752–762. doi: 10.1016/j.jaci.2021.01.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Schatz M., Kosinski M., Yarlas A.S., Hanlon J., Watson M.E., Jhingran P. The minimally important difference of the Asthma Control Test. J Allergy Clin Immunol. 2009;124(4):719–723 e1. doi: 10.1016/j.jaci.2009.06.053. [DOI] [PubMed] [Google Scholar]
- 26.Juniper E.F., Guyatt G.H., Willan A., Griffith L.E. Determining a minimal important change in a disease-specific Quality of Life Questionnaire. J Clin Epidemiol. 1994;47(1):81–87. doi: 10.1016/0895-4356(94)90036-1. [DOI] [PubMed] [Google Scholar]
- 27.Wechsler M.E., Laviolette M., Rubin A.S., et al. Bronchial thermoplasty: long-term safety and effectiveness in patients with severe persistent asthma. J Allergy Clin Immunol. 2013;132(6):1295–1302. doi: 10.1016/j.jaci.2013.08.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Young H.M., Eddy R.L., Parraga G. MRI and CT lung biomarkers: towards an in vivo understanding of lung biomechanics. Clin Biomech (Bristol, Avon) 2019;66:107–122. doi: 10.1016/j.clinbiomech.2017.09.016. [DOI] [PubMed] [Google Scholar]
- 29.Schoettler N., Strek M.E. Recent advances in severe asthma: from phenotypes to personalized medicine. Chest. 2020;157(3):516–528. doi: 10.1016/j.chest.2019.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bodduluri S., Bhatt S.P., Hoffman E.A., et al. Biomechanical CT metrics are associated with patient outcomes in COPD. Thorax. 2017;72(5):409–414. doi: 10.1136/thoraxjnl-2016-209544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Wang T., Long F., Huang Z., et al. Correlation of activation site and number with the clinical response to bronchial thermoplasty. J Asthma Allergy. 2022;15:437–452. doi: 10.2147/JAA.S357037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sieren J.P., Newell J.D., Jr., Barr R.G., et al. SPIROMICS protocol for multicenter quantitative computed tomography to phenotype the Lungs. Am J Respir Crit Care Med. 2016;194(7):794–806. doi: 10.1164/rccm.201506-1208PP. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Expert Panel Working Group of the National Heart, Lung, and Blood Institute (NHLBI) administered and coordinated National Asthma Education and Prevention Program Coordinating Committee. Cloutier M.M., Baptist A.P., et al. 2020 focused updates to the asthma management guidelines: a report from the National Asthma Education and Prevention Program Coordinating Committee Expert Panel Working Group. J Allergy Clin Immunol. 2020;146(6):1217–1270. doi: 10.1016/j.jaci.2020.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.