Summary of findings
Description of the condition
Bronchial smooth muscle hypertrophy, hyperplasia and contraction play a central role in bronchial obstruction in patients with asthma. In fact, the first treatment step in such patients, particularly when they are symptomatic, is the use of bronchodilators (beta
Despite these efforts, however, asthma remains a poorly controlled disease in a small but important minority of patients, even when they are compliant; as such, it is still a common reason for emergency department visits. Furthermore, between 3% and 6% of patients with asthma respond poorly to treatment (Torrego 2010), including oral corticosteroid therapy, and continue to experience symptoms and a diminished quality of life. Care for patients with more severe disease is a major public health burden, which makes it clear that new treatments are required to improve the prognosis of this group of patients (Moore 2006).
Description of the intervention
Bronchial thermoplasty is an innovative procedure that consists of the delivery of controlled radiofrequency-generated heat via a catheter inserted into the bronchial tree through a flexible bronchoscope. Once the bronchoscope has been placed into the airway, the catheter is advanced to a bronchial segment and is expanded so that it is in contact with bronchial mucosa (Mayse 2007). Then radiofrequency energy is applied through the catheter, across the contact points. Treatments are applied consecutively so that the entire bronchopulmonary segment is treated. Three separate bronchoscopic treatment sessions are required to treat all of the airways (except for the right middle lobe).
How the intervention might work
The mechanism of action of bronchial thermoplasty is not definitively established, but experimental evidence in animal models (Brown 2005; Danek 2004) suggests that it might work by reducing airway smooth muscle bulk, thereby reducing the effect of smooth muscle bronchoconstriction.
As a consequence of this mechanism of action, this treatment could reduce asthma symptoms and exacerbations, resulting in improved asthma control and quality of life.
Why it is important to do this review
Although several recent reviews have discussed the efficacy of bronchial thermoplasty (Cayetano 2011; Wahidi 2012), this review will provide a more rigorous summary of available evidence regarding the efficacy of this intervention, focusing on the trade-off between relevant patient outcomes and safety, in the short term and over the long term (Thomson 2011).
To determine the efficacy and safety of bronchial thermoplasty in adults with bronchial asthma.
Criteria for considering studies for this review
Types of studies
We included randomised controlled clinical trials that compared bronchial thermoplasty versus conventional treatment, with or without the use of a placebo, as treatment for adult participants with moderate or severe persistent asthma.
Types of participants
We included adult participants with moderate or severe persistent asthma according to the Global Initiative for Asthma (GINA) criteria (GINA 2011).
Types of interventions
We included trials assessing bronchoscopically delivered thermoplasty adjuvant to conventional treatment for participants with moderate to severe persistent asthma.
Eligible trials included a control group of participants with moderate to severe persistent asthma who received conventional treatment or preferably a control sham bronchoscopy.
Types of outcome measures
- Health-related quality of life (HRQOL) evaluated through asthma-specific or generic questionnaires.
- Asthma-specific (e.g. Asthma Quality of Life Questionnaire (AQLQ; Juniper 1994)); considering a change in the score of 0.5 points the threshold of clinical relevance).
- Respiratory-specific (e.g. St. George's Respiratory Questionnaire; considering a change in the score of four points the threshold of clinical relevance (Jones 2002)).
- Generic (e.g. Short Form (SF)-36 Questionnaire on health status; Nottingham Health Profile (NHP); World Health Organization (WHO) instrument on health-related quality of life (WHOQOL-100)).
- Asthma control evaluated through validated questionnaires (e.g. Asthma Control Questionnaire (ACQ; Juniper 1999)); considering a change in the score of 0.5 points the threshold of clinical relevance.
- Number of exacerbations, defined as any of the following.
- Exacerbations requiring hospital or intensive care unit admissions.
- Exacerbations requiring emergency department visits or unscheduled healthcare visits.
- Exacerbations resulting in a course of oral or systemic corticosteroids, or an increase in the regular required doses.
- Serious adverse events, defined as any of the following.
- Fatal events.
- Hospital admission.
- Risk of death at the time of event.
- A permanent or significant disability.
- Lung function tests.
- Bronchial hyperresponsiveness evaluated through non-specific (direct muscle contraction) bronchial stimulation tests.
- Peak expiratory flow (PEF).
- Forced expiratory flow during the first second (FEV
- Regular medication needs for asthma control with inhaled corticosteroids +/- long-acting beta
2-agonist, according to recommendations in international guidelines (GINA 2011).
- Use of rescue medication with:
- the addition of a short-acting beta
2-agonist relief medication to the combination of long-acting beta 2-agonist and inhaled corticosteroids; or
- the as-needed use of combined budesonide plus formoterol.
- Asthma symptom-free days.
- Days missed from work or school.
- Adverse events
Search methods for identification of studies
The Cochrane Airways Group Trials Search Co-ordinator searched the Group's Specialised Register of Trials (CAGR) up to January 2014. This Register of Trials is derived from systematic searches of bibliographic databases, including the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, CINAHL, AMED and PsycINFO, and from handsearching of respiratory journals and meeting abstracts (see Appendix 1 for further details).
We searched records in the CAGR coded as asthma, using the following terms:
thermoplast* or bronchoscop* or ((surger* or surgical) and bronchi*) or thermal* or catheter or Alair or Asthmatx
We also conducted a search of ClinicalTrials.gov. We searched all databases from their inception to January 2014, with no restriction on language of publication.
Searching other resources
We checked reference lists of all primary studies and review articles for additional references.
Data collection and analysis
Selection of studies
Two independent evaluators (AT and IS) screened the titles and abstracts identified through the electronic searches to identify studies to include in the review. We discussed disagreements and consulted with a third review author (VP or JJY-N).
Data extraction and management
Two independent evaluators (AT and JJY-N) read all reports in detail and summarised the pertinent details on a standard data extraction sheet (which included study design methodology; number and description of participants; characteristics of intervention; type and method of outcome measurement; and evaluation of methodology). We discussed disagreements and reached agreement by consensus with a third review author, who checked the data extraction for accuracy (IS).
Assessment of risk of bias in included studies
Two review authors (IS and AT) independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We solved disagreements by discussion, or by involving a third assessor. We assessed risk of bias according to the following domains.
Sequence generation (selection bias): For each included study, we described in detail the methodology used to generate the allocation sequence, and we evaluated the methodology to determine whether it can produce comparable groups. We assessed sequence generation as low risk of bias (any truly random process, e.g. random number table, computer random number generator); high risk of bias (any non-random process, e.g. odd or even date of birth, hospital or clinic record number); or unclear risk of bias.
Allocation concealment (selection bias): For each included study, we described in detail the methodology used to conceal the allocation sequence, and we evaluated the methodology to determine whether intervention allocation could have been foreseen in advance, noted during recruitment or changed after assignment. We evaluated allocation concealment as low risk of bias (e.g. telephone or central randomisation, consecutively numbered sealed opaque envelopes); high risk of bias (open random allocation, unsealed or non-opaque envelopes, alternation, date of birth); or unclear risk of bias.
Blinding (performance bias): For each included study, we described the methodology used, if any, to blind study participants and personnel from knowing the intervention that a participant received. We collected from the included studies information on whether the intended blinding was effective. When blinding was not possible, we assessed whether lack of blinding was likely to have introduced bias. We assessed blinding for participants and for outcome assessors separately. We evaluated blinding as low risk of bias, high risk of bias or unclear risk of bias for participants; and as low risk of bias, high risk of bias or unclear risk of bias for outcome evaluators.
Incomplete outcome data (attrition bias through withdrawals, dropouts or protocol deviations): For each included study, we provided a description of completeness of data, including attrition and exclusions from analysis, as well as an assessment of the reasons for attrition or data exclusion (if available). We extracted from trials the numbers lost through attrition and exclusion, as well as the numbers of participants included in the analysis at each stage (compared with the total number of randomly assigned participants).
Selective outcome reporting: We assessed selective outcome reporting for each included study. We evaluated selective outcome reporting as low risk of bias (when it is clear that all of the study’s prespecified outcomes and all expected outcomes of interest to the review have been reported); high risk of bias (when not all of the study's prespecified outcomes have been reported, one or more reported primary outcome(s) were not prespecified, outcomes of interest are reported incompletely and so cannot be used or study fails to include results of a key outcome that would have been expected to have been reported); or unclear risk of bias.
Overall risk of bias: We classified all studies according to the following criteria: low risk of bias (all individual items were at 'low risk of bias'); moderate risk of bias (one or more individual item(s) was at 'unclear risk of bias', and the remaining were at 'low risk of bias'); or high risk of bias (one or more individual item(s) was at 'high risk of bias').
Measures of treatment effect
When possible, we assessed treatment effect through mean differences (MDs) or standardised mean differences (SMDs) for continuous outcomes, and through risk ratios (RRs) for dichotomous outcomes. We present all measures with 95% confidence intervals (CIs). We performed all statistical analyses using Review Manager 5.1 (RevMan 2011).
We extracted data at the end of follow-up for continuous outcomes. When trials reported data in terms of changes from baseline, we converted these into final scores. When standard deviations (SDs) at the end of follow-up were not reported, we imputed these from baseline data.
Unit of analysis issues
For most outcomes, the unit of analysis was the participant (i.e. for dichotomous outcomes such as hospitalisations, we used the number of participants who had an admission to the hospital while experiencing one or more event(s) to avoid double counting of participants). We could not analyse data using this approach for some outcomes that reported data as rates (e.g. for outcomes such as exacerbations, trials reported data as the number of exacerbations per participant per year).
Dealing with missing data
We based our data extraction and main analysis on the available data analysis in each of the papers. If a paper presented both intention-to-treat and per-protocol data, we used the former in the analyses.
Assessment of heterogeneity
We assessed clinical heterogeneity by examining the trials in terms of participant characteristics, interventions, controls and definitions of results. We evaluated statistical heterogeneity through the I
Assessment of reporting biases
We performed meta-analyses using a random-effects model and using the inverse variance method. We present forest plots for each result when we were able to extract data. We did not perform pooled analyses for outcomes reported as rates.
Subgroup analysis and investigation of heterogeneity
The characteristics of included trials did not allow us to carry out the planned subgroup analysis according to severity of asthma among participants classified as having severe asthma versus moderate asthma. In future updates of the review, we plan to use the classification reported in the included trials and, when possible, to classify severity on the basis of the intensity of treatment required to achieve good asthma control (GINA 2011; Taylor 2008).
The characteristics of included trials did not allow us to carry out the planned sensitivity analysis. In future updates of the review, we will try to perform sensitivity analyses to determine the effects of the risk of bias from included studies and the meta-analysis model on treatment effect estimates.
Description of studies
Results of the search
The search of the Cochrane Airways Group Specialised Register of Trials (CAGR) yielded 206 records. Revision of titles and abstracts led to the exclusion of 183 references. We reviewed in detail 23 publications, resulting in the inclusion of five reports and 16 abstract communications with results related to just three randomised clinical trials (AIR; AIR 2; RISA).
The two excluded references discussed in scientific meetings clinical issues related to design and results of the included trials and did not include relevant data for the review (references summarised in the Characteristics of excluded studies table). Figure 1 shows a flowchart of study selection following PRISMA guidance.
|Figure 1. Study flow diagram.|
(CAGR: Cochrane Airways Group Specialised Register of Trials).
We included three randomised controlled trials in this review. Two compared bronchial thermoplasty versus usual care (AIR; RISA), and the other compared bronchial thermoplasty versus a sham bronchoscopic procedure (AIR 2). For details of the included studies, see Characteristics of included studies.
The AIR trial (NCT00214526) recruited 112 participants with moderate to severe asthma (pre-dose FEV
The RISA trial (NCT00214539) limited its sample to 34 participants with symptomatic, severe asthma (pre-dose FEV
Finally, the AIR 2 trial (NCT00231114) randomly assigned 288 participants with symptomatic, persistent asthma (pre-dose FEV
Intervention and comparison groups
The AIR trial randomly assigned participants to receive bronchial thermoplasty plus medical management with inhaled corticosteroids and LABA (N = 56) or medical management alone (control group; N = 56). Before treatment with the intervention of interest was provided, participants received maintenance therapy for four weeks, with inhaled corticosteroids and LABA for the first two weeks and with LABA withheld for two additional weeks. Maintenance therapy was resumed during the treatment period. Participants allocated to bronchial thermoplasty had three bronchoscopy procedures performed at least three weeks apart with the Alair system (Asthmatx). Participants were followed up at six weeks and at three, six and 12 months.
The RISA trial also randomly assigned participants to receive bronchial thermoplasty plus medical management with inhaled corticosteroids and LABA (N = 17) or medical management alone (N = 17). Before treatment with bronchial thermoplasty, participants were entered into a two-week run-in period. During the treatment period, participants allocated to bronchial thermoplasty had three bronchoscopy procedures performed with the Alair system at three-weekly intervals. During these initial phases, all participants maintained their asthma medications. After recovery, participants entered into a 16-week steroid stable phase (weeks six to 22), followed by a 14-week steroid wean phase (weeks 22 to 36) and a final period of 16 weeks with reduced steroid treatment (weeks 36 to 52).
The AIR 2 trial allocated participants in a 2:1 ratio to receive bronchial thermoplasty (N = 190) or a sham intervention (N = 98) to ensure blinding of participants and outcome assessors. Both groups of participants continued their conventional medical treatment with inhaled corticosteroids and LABA. As in the other included trials, participants underwent three bronchoscopy procedures performed with the Alair system at three-weekly intervals. Participants in the control group underwent three sham bronchoscopies at the same time intervals as those treated with bronchial thermoplasty. After the last bronchoscopy, participants were evaluated after six weeks after the last procedure, which was defined as the end of the treatment period. The post-treatment period extended from six to 52 weeks after the last procedure, with completion of assessments at three, six, nine and 12 months.
Outcomes of interest
The primary outcome in the AIR trial was the difference between compared groups in the change in rates of mild asthma exacerbations between baseline and follow-up. The authors analysed only exacerbations recorded in daily diaries that occurred during the two-week period of abstinence from LABA, before treatment and then at 3, 6 and 12 months. This outcome measurement therefore was different from that in the other two studies. Secondary outcomes were changes in PEF, use of rescue medication, number of symptom-free days ascertained from a symptom score and changes on ACQ and AQLQ questionnaires. Adverse events were registered from telephone interviews with participants and from data in their diaries and were classified as respiratory or non-respiratory.
Participants who completed the 12 months of follow-up in this trial were invited to participate in an extension study to evaluate the long-term safety of the procedure over five years (NCT00448812). More participants who had received bronchial thermoplasty during the study agreed to participate in this post-treatment study (45/52; 87%), although fewer did so in the control arm, and data were available for only three years (24/49; 49%) (Thomson 2011).
The RISA trial was designed to assess the safety of bronchial thermoplasty, with respiratory adverse events per participant as the primary outcome of interest. Although the main publication of the trial (Pavord 2007) did not specify primary or secondary outcomes, the protocol of the trial nominated the following secondary outcomes: use of maintenance medications, use of rescue medications, asthma symptom score, symptom-free days, pre- and post-bronchodilator FEV
The AIR 2 trial was designed to assess the safety and effectiveness of the procedure and established the difference between study groups in the change in AQLQ scores from baseline as the primary outcome of interest. Secondary outcomes included changes in ACQ scores, symptom-free days, numbers and days of rescue medication used and pulmonary function (FEV
We excluded two scientific meeting abstracts that discussed issues related to the three included trials (Hales 2010; Wechsler 2009). We have included the reasons for exclusion of these abstracts in the Characteristics of excluded studies table.
Risk of bias in included studies
Overall, the included trials had no major limitations in design or execution; with the exception of blinding, assessed domains were judged to be at low risk of bias (see Figure 2).
|Figure 2. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
The AIR and RISA trials had a low risk of selection bias, with a centrally generated computer randomisation sequence in blocks of four participants provided in sealed envelopes to each participating centre.
The AIR 2 trial randomly assigned participants in a 2:1 ratio and stratified them by AQLQ score, percentage of symptom-free days and site, according to a sequence that was computer generated using minimisation. The publication of trial results did not describe the efforts made to ensure concealment of the randomisation sequence.
Both AIR and RISA trials had an open design; therefore both investigators and participants were aware of whether they had received bronchial thermoplasty. Outcomes in the AIR trial were collected by participants in diaries monitored by the research staff at meetings, consultations or telephone interviews. The RISA trial followed the same procedure to assess outcomes. This resulted in high risks of performance and detection bias, particularly for the subjective outcomes of quality of life and asthma control.
The AIR 2 trial ensured blinding of participants by comparing bronchial thermoplasty versus a sham procedure in which participants allocated to the control group received three bronchoscopies that involved sedation and a sham procedure that mimicked the active-arm procedure. A catheter was deployed into the airways through a bronchoscope, an electrode array was expanded and a radiofrequency controller was activated, simulating indistinguishable audio and visual signals, but no radiofrequency energy was delivered. Duration and time intervals for treatments and assessments in active and control arms were the same. Participants in both groups were unable to guess their treatment assignment after the first bronchoscopy, but in the second procedure, a larger proportion of participants in the bronchial thermoplasty group guessed their assigned treatment. Outcome assessment was performed by a blinded assessment team. The staff that performed bronchoscopies was unblinded to treatment assignment.
Incomplete outcome data
The AIR trial was designed to perform an intention-to-treat analysis with no imputation of missing data. Researchers randomly assigned 112 participants (56 to each group, with baseline data reported for 109 participants), and at first follow-up contact at three months, data on four participants in the bronchial thermoplasty group and eight in the control group were lost (three and four participants, respectively, withdrew consent; one and four participants, respectively, were lost to follow-up). At 12 months, complete assessment data were available for 52 participants in the bronchial thermoplasty group and for 49 in the medical management group.
The RISA trial did not provide details about the plan to perform analysis of results but reported clearly the numbers of participants randomly assigned and withdrawn from the study. Researchers randomly assigned 34 participants (17 to each group), but two participants in the bronchial thermoplasty group withdrew from the study before they received treatment, in keeping with a recommendation from the study Data and Safety Monitoring Board.
The AIR 2 trial described a minimum recruitment goal of 225 participants (150 in the bronchial thermoplasty group and 75 in the sham group) and planned to analyse its results on an intention-to-treat basis, defined as participants randomly assigned who received at least a single bronchoscopy. Missing data for secondary outcomes were imputed using the last observation carried forward approach. The trial reached the randomisation target of 288 participants (190 in the bronchial thermoplasty group and 98 in the sham group), and at 12 months of follow-up, only nine active participants and one control participant were lost to follow-up, but all were included in the final analysis.
The AIR trial described data for all outcomes described in the trial protocol (NCT00214526). Although most results were reported in graphical form, making it difficult to extract data for analysis, accurate data were available from electronic appendices. In the RISA and AIR 2 trials, all outcomes described in the study protocol (NCT00214539, NCT00231114) were reported.
Effects of interventions
We outline the main results of the review and the quality of evidence in Summary of findings for the main comparison.
Health-related quality of life
A pooled analysis showed a clinically small but statistically significant increase in AQLQ scores at 12 months of follow-up among participants who received bronchial thermoplasty (three trials, 429 participants; MD 0.28, 95% CI 0.07 to 0.50; Analysis 1.1). This difference was driven by effect estimates from the two trials that compared bronchial thermoplasty versus medical management, as the trial that compared bronchial thermoplasty versus a sham intervention did not achieve a statistically significant difference in final AQLQ scores.
The AIR trial reported an improvement at 12 months in the quality of life of participants treated with bronchial thermoplasty. This was explained for a statistically significant improvement in AQLQ score in the bronchial thermoplasty group from baseline to follow-up (from 4.91 ± 1.23 to 6.18 ± 0.88) compared with the moderate change seen in the control group (from 5.15 ± 1.19 to 5.72 ± 1.11). The difference between groups at 12 months of follow-up in final AQLQ scores was 0.46 points.
The RISA did not report final scores for quality of life measures. At the end of the steroid stable phase (22 weeks of follow-up), the trial reported statistically significant improvement in the quality of life of participants who received bronchial thermoplasty compared with those who received medical treatment. The improvement from baseline in AQLQ scores was greater in the bronchial thermoplasty group (1.21 ± 1.05) than in the control group (0.15 ± 0.75). A greater proportion of participants receiving bronchial thermoplasty experienced improvement greater than the minimum clinically important increase in scores of 0.5 (77% vs 35%), and fewer experienced deterioration of 0.5 or more points (8% vs 18%).
Results from AIR 2 suggest a large placebo effect in the studies without a sham intervention. The 12-month follow-up of the AIR 2 trial reported similar changes in AQLQ scores both for participants who received bronchial thermoplasty (from 4.30 ± 1.17 to 5.66 ± 1.06) and for participants treated with sham bronchoscopies (from 4.32 ± 1.21 to 5.48 ± 1.15). The improvement from baseline on the AQLQ was slightly greater in the bronchial thermoplasty group (1.35 ± 1.10 compared with 1.16 ± 1.23) but did not reach the posterior probability of superiority (PPS) planned (96.0% observed vs 96.4% planned). Significantly more participants in the bronchial thermoplasty group showed an improvement of 0.5 or greater in AQLQ scores compared with the sham group (78.9% vs 64.3%; PPS = 0.996).
No significant difference in symptomatic control as measured by the ACQ was reported (three trials, 429 participants; MD -0.15, 95% CI -0.40 to 0.10; Analysis 1.2).
The AIR trial reported improvement in ACQ at 12 months, with a significantly greater reduction in the scores of participants who received bronchial thermoplasty (from 2.50 ± 0.92 to 1.32 ± 0.85) compared with the reduction between participants in the control group (from 2.16 ± 0.86 to 1.69 ± 0.99). However, it should also be noted that the subjective changes experienced by participants were clinically relevant irrespective of the group to which they were allocated (Juniper 1994), suggesting a strong Hawthorne effect.
The RISA reflected an improvement in symptom-based ACQ scores with a mean decrease of -1.04 ± 1.03 in bronchial thermoplasty compared with a decrease of -0.13 ± 1.00 in controls. At the end of the reduced steroid phase (52 weeks of follow-up), bronchial thermoplasty participants also showed statistically significant improvement from baseline compared with controls in both AQLQ (1.53 ± 0.79 vs 0.42 ± 0.82) and ACQ scores (-0.99 ± 0.83 vs -0.22 ± 0.78). However the baseline difference between the groups was -0.76 for the ACQ.
Changes from baseline observed in AIR 2 in ACQ scores (bronchial thermoplasty from 2.13 ± 0.87 to 1.31 ± 0.94 vs sham 2.09 ± 0.90 to 1.32 ± 0.91) were not statistically significantly different between groups (change of -0.82 ± 0.95 vs -0.77 ± 1.08; PPS = 0.638).
Number of exacerbations
The AIR trial reported the number of mild exacerbations per participant per week, but the exacerbations were counted only from two-week periods in which LABA was withdrawn from both groups at three, six and 12 months. The change in the mean frequency of exacerbations per participant/wk was statistically significantly different between groups at 12 months of follow-up (bronchial thermoplasty -0.16 ± 0.37 vs control 0.04 ± 0.29). Participants in the bronchial thermoplasty group showed a decrease from 0.35 ± 0.32 exacerbations per participant/wk at baseline to 0.18 ± 0.31 exacerbations per participant/wk at 12 months of follow-up, compared with an increase in the control group from baseline 0.28 ± 0.31 exacerbations per participant/wk to 0.31 ± 0.46 exacerbations per participant/wk at 12 months of follow-up. This difference between groups was statistically significant. The trial authors reported that the difference between groups was statistically significant also at three months of follow-up, but these data were reported graphically only and so do not allow performance of an accurate extraction of the rates for each group. In contrast, the trial showed no significant differences in data on the number of severe exacerbations per participant per week, with participants in both groups experiencing a decrease at 12 months (baseline bronchial thermoplasty 0.07 ± 0.18 vs control 0.09 ± 0.31, and at 12 months bronchial thermoplasty 0.01 ± 0.08 vs control 0.06 ± 0.24).
The rate of severe exacerbations per participant per year was significantly lower among participants who received bronchial thermoplasty compared with controls after 12 months of follow-up in the AIR 2 trial (0.48 ± 0.067 vs 0.70 ± 0.122 exacerbations per patient-year, respectively). In the bronchial thermoplasty group, a significantly lower proportion of participants experienced severe exacerbations compared with controls (26.3% of participants (50/190) vs 39.8% (39/98), respectively).
Serious adverse events
Data from included trials showed a significant increase in risk of hospitalisation due to respiratory adverse events among participants treated with thermoplasty during the treatment period (three trials, 429 participants; RR 3.50, 95% CI 1.26 to 9.68; I
During the treatment period in the AIR trial, hospitalisations for adverse respiratory events were more frequent among participants who received bronchial thermoplasty (4 participants with a total of 6 hospitalisations; 4 related to an asthma exacerbation) compared with participants who received medical treatment alone (2 participants with 2 hospitalisations). Hospitalisations in the follow-up period were similar between groups (3 admissions per group). The trial did not register any deaths during the study period. It is interesting to note that over two to three years of follow-up, during the extension period of the trial, more participants who had received bronchial thermoplasty required hospitalisation for respiratory symptoms than participants in the control group, but the difference was not statistically significant. The trialists reported that over five years of follow-up, the number of participants who received bronchial thermoplasty and required hospitalisation and the number of hospitalisations required for a respiratory adverse event did not get worse compared with the first 12 months of follow-up within the formal trial (Thomson 2011).
The RISA trial reported no hospitalisations due to respiratory adverse events among participants receiving medical management, although a quarter of participants from the bronchial thermoplasty group were hospitalised. Four participants accounted for a total of seven hospital admissions, five of which related to an asthma exacerbation. During the follow-up period, the groups did not differ in the numbers of participants who required hospitalisation (one participant in the control group had four admissions, and three participants treated with bronchial thermoplasty had five admissions).
In the AIR 2 trial, 16 participants in the bronchial thermoplasty group required a total of 19 hospitalisations for respiratory symptoms during the treatment period, with 12 admissions in 10 participants related to worsening of asthma. In the sham group, only two participants had two hospital admissions—all related to worsening of asthma. In contrast, in the post-treatment period, five participants in the thermoplasty group had six admissions to hospital compared with 12 hospitalisations among four participants who received sham bronchoscopies. However, the rate of visits to the emergency department was significantly and markedly higher among participants in the sham group (0.43 visits/y vs 0.07 visits/y). Over the study period, the proportion of participants with a respiratory-related visit to the emergency department was 15.3% of those receiving sham treatment and 8.4% of those treated with thermoplasty, representing an absolute reduction of 7%.
Lung function tests
During the first 22 weeks, participants treated with bronchial thermoplasty in the RISA trial reported greater improvement in change from baseline in pre-bronchodilator FEV
In the AIR trial, participants treated with bronchial thermoplasty experienced significant improvement from baseline in morning PEF (369.4 ± 97.9 to 397.4 ± 100.7 L/min) compared with the control group (394.0 ± 98.2 to 395.4 ± 88.6 L/min) at one-year follow-up, but of course, they ended up remarkably similar in this. The study authors reported no statistically significant differences in pre-bronchodilator FEV
A report on long-term follow-up results for the AIR trial showed stability in FEV
The AIR 2 trial found no statistical differences in morning PEF nor in FEV
Doses of regular medication for asthma control
In the RISA trial, no statistically significant differences were noted between groups in the numbers of participants who had been receiving oral corticosteroids and were able to completely wean off oral corticosteroids at 52 weeks of follow-up (four of eight participants in the bronchial thermoplasty group vs one of seven participants in the control group). The mean percentage reduction in oral steroid doses was large but was not significant between active and control groups (63.5 ± 45.4% vs 26.21 ± 40.7%); it was neither marked nor significant for inhaled corticosteroids (28.6 ± 30.4% vs 20.0 ± 32.9%, respectively).
Use of rescue medication
A pooled analysis of the results from included trials showed no significant difference between bronchial thermoplasty and control groups in reduction in the use of rescue medication over 12 months (3 trials, 429 participants; MD -0.68 puffs/wk, 95% CI -3.63 to 2.28; Analysis 1.5).
Participants who received bronchial thermoplasty in the AIR trial did experience a significant reduction in needed doses of rescue medication at 12 months of follow-up. Participants in the bronchial thermoplasty group reduced their short-acting bronchodilator doses from 19.8 ± 17.2 puffs per week at baseline to 10.9 ± 15.0 puffs per week at 12 months, compared with the reduction in doses in the control group from 16.0 ± 18.8 puffs per week at baseline to 14.8 ± 21.2 puffs per week at 12 months of follow-up.
The requirement for rescue medication was also significantly reduced among participants receiving bronchial thermoplasty compared with controls in the RISA trial. The reduction in short-acting bronchodilator doses per week was greater among participants in the bronchial thermoplasty group than among those who received only medical management both in the steroid stable phase (up to week 22 of follow-up; 26.6 ± 40.1 fewer puffs/wk vs 1.5 ± 11.7 fewer puffs/wk) and in the final assessment after 52 weeks of follow-up (25.6 ± 31.2 fewer puffs/wk vs 6.1 ± 12.4 fewer puffs/wk).
Although after one year of follow-up, participants in the AIR 2 trial reduced the requirement for rescue medication, no significant difference was noted between compared groups (bronchial thermoplasty: from 13.4 ± 19.17 puffs/wk to 7.4 ± 15.01 puffs/wk vs sham group: from 11.8 ± 11.24 puffs/wk to 7.5 ± 12.60 puffs/wk), nor was a significant difference observed in the number of days that rescue medication was used.
Asthma symptom-free days
In the AIR trial, participants who received bronchial thermoplasty experienced a significant increase in days that were free of symptoms at 12 months compared with participants who received only medical treatment. The increase in symptom-free days for participants in the bronchial thermoplasty group was 41% ± 40% at 12 months of follow-up. For participants in the control group, this increase was more moderate, at 17% ± 40%. However, it should be noted that symptom diaries at 12 months were assessed over a four-week period, during which LABA had been withdrawn.
Days missed from work or school
The AIR 2 trial reported that the decrease in severe exacerbations experienced by participants in the bronchial thermoplasty group resulted in fewer days lost from work or other activities because of asthma compared with participants who received the sham intervention (1.32 ± 0.36 days/y vs 3.92 ± 1.55 days/y).
During the six weeks that participants were treated with bronchial thermoplasty in the AIR trial, they experienced a greater rate of respiratory adverse events than participants in the control group (407 events vs 106 events), with the most common being dyspnoea, wheezing and cough. Most adverse events were mild (69%), but one-third of participants experienced moderate adverse events. In the post-treatment period, the rate of adverse events was similar between groups. During the extension period of the study, when data for control participants were available at two and three years of follow-up, the rate of respiratory adverse events did not differ significantly between groups (Thomson 2011).
The greater increase in respiratory adverse events with bronchial thermoplasty was also noted in the RISA trial. During the treatment period, the number of respiratory adverse events was 136 in bronchial thermoplasty participants compared with 57 among participants treated with medical management, with again the most frequently observed adverse events being wheezing, cough and, in this case, chest discomfort. Most adverse events were mild (49%) or moderate (41%). Although during the post-treatment period more than half of participants had wheezing or dyspnoea, no difference between groups was noted.
The AIR 2 trial reported that more participants in the bronchial thermoplasty group experienced respiratory adverse events during the treatment period than those receiving the sham procedure (85% vs 76%), and adverse events were mild to moderate in most cases. More participants in the bronchial thermoplasty group than in the sham group experienced a composite of multiple symptoms related to asthma (52 vs 39); the most common were wheezing, chest discomfort, cough and chest pain.
For all trials included in the review, most adverse events experienced by bronchial thermoplasty participants during the treatment period occurred within one day after the bronchoscopy but were resolved within seven days.
At present, bronchial thermoplasty is an add-on treatment for severe asthma patients who fail to adequately control their disease with conventional pharmacological treatments. This review has identified three trials of thermoplasty that were generally free of major sources of bias and that have reported data on a range of clinically relevant endpoints for asthma, including validated asthma-related questionnaires on symptoms and quality of life, asthma medication use, lung function measurements, exacerbation rates, use of relief medications, asthma-free days and hospital attendances/admissions (AIR; AIR 2; RISA).
Summary of main results
Bronchial thermoplasty has shown a modest benefit in quality of life in patients with persistent asthma, with the possibility of a control group 'Hawthorne' effect. The most meaningful improvements with thermoplasty seemed to be seen in a reduction in asthma exacerbation rates and hospital emergency department attendances, although rather limited data were collected on the latter and a lot of variability was noted between studies in the former. The limited quantity of long-term data suggests that after five years, patients treated with bronchial thermoplasty may maintain their reduction in asthma exacerbations.
Bronchial thermoplasty was associated with a high proportion of participants with respiratory adverse events in all studies, although only during the treatment period, and most of these events were mild or moderate. They appeared in the first 24 hours and resolved by one week post-treatment. The most common symptoms after these procedures were dyspnoea, wheezing, cough, chest discomfort and mucus production. A proportion of participants required hospitalisation because of these adverse events, including, for a case of haemoptysis, respiratory infection, asthma worsening and atelectasis. Longer-term follow-up reports have not shown adverse events related to thermoplasty (Pavord 2011; Thomson 2011). No fatalities or other severe disabilities have been reported after bronchial thermoplasty. Taken together, these results show that despite transient mild to moderate respiratory worsening after the bronchoscopic procedure, bronchial thermoplasty seems to be a generally safe treatment.
Five years of observational follow-up data from the RISA and AIR 2 trials have been published recently, providing data for participants who received bronchial thermoplasty during the trial period (but not for the control groups). Participants treated with thermoplasty in the RISA trial showed a non–statistically significant decrease in the number of visits to the emergency department during the five-year follow-up period (mean of 0.12 visits/participant/y) compared with the year before receiving the intervention (0.36 visits/participant/y). These participants also showed a decrease in the number of hospitalisations during follow-up (overall rate of 0.23 hospitalisations/participant/y during five years) compared with hospitalisation experienced before the intervention (0.71 hospitalisations/participant/y) (Pavord 2013). At five years of follow-up, participants treated with bronchial thermoplasty in the AIR 2 trial showed a maintained decrease in the rate of severe exacerbations in comparison with the year before treatment. In this trial, the proportions of participants who had severe exacerbations in each follow-up year compared with the first year after thermoplasty were not significantly different. The reduction in the proportion of participants experiencing severe exacerbations in the year after thermoplasty (30.9%) compared with the 12 months before thermoplasty (51.6%) was maintained during the follow-up period (average decrease of 44%) (Wechsler 2013). Results from the US Food and Drug Administration (FDA) post-approval study, with estimated completion during 2018, should bring new data on the durability of the efficacy and safety of the procedure (NCT01350336).
Overall completeness and applicability of evidence
The proposed mechanism of action for bronchial thermoplasty is a reduction in bronchial smooth muscle mass, which may be expected to be associated especially with improvement in pulmonary function. It is surprising to note that the available trials showed no unequivocal benefit for pulmonary function outcomes. This could be explained by the fact that the procedure is limited to central airways accessible by flexible bronchoscopy. Smaller airways are untreated, and it is known that they are an important contributor to airway resistance in asthma patients, especially those < 2 mm in diameter.
Available evidence does not allow fully firm conclusions about the benefits of bronchial thermoplasty because of the lack of information on some important issues. First of all, no accurate data are available regarding the mechanism of action of bronchial thermoplasty in asthma patients. Smooth muscle reduction, the proposed effect of radiofrequency on the airway wall, was observed in animal models and in a small sample of study participants without asthma in a feasibility study (Miller 2005). Although controlled data about the thermoplasty mechanism of action in humans are lacking, reports on a limited series of participants treated with bronchial thermoplasty have been published with inconclusive results (Doeing 2013; Gordon 2013).
Second, taking into account that severe asthma is a very complex condition with a high health and economic burden, the lack of information on long-term cost-effectiveness of the procedure is a limitation to the applicability of the results. Similarly, thermoplasty trials have not provided sufficient information regarding potential asthma medication step-down after treatment, with only the RISA trial showing a limited number of participants who could reduce their steroid therapy appreciably without losing asthma control. Small case series of patients treated out of clinical trials have reported successful weaning off for other severe asthma-related treatments after bronchial thermoplasty (Bricknell 2012). One point that is striking about two of the studies that we have reviewed in detail is the improvement seen in control participants, especially in subjective outcomes, perhaps related to better compliance with medication whilst participating in a study. Alternatively, it may be that participants are recruited to such studies whilst especially symptomatic, and subsequent improvements are at least partially a regression back towards the mean.
The included trials have shown a proportion of non-responder participants to bronchial thermoplasty. The AIR 2 trial showed that 21% of participants treated with thermoplasty did not achieve clinically meaningful improvement in quality of life. Possible explanations for this result could include several factors. Severe asthma applies to a heterogeneous group of patients with different inflammatory disease characteristics. Thus, trial inclusion criteria have used only pragmatic definitions of severe asthma based basically on medication requirements, and no information was obtained about details of asthma phenotype, such as neutrophil, eosinophil or pauci-cellular predominance in sputum. Moreover, differences in study design, number of radiofrequency activations and bronchoscopic management among centres and variable pharmacological treatments for asthma may help to explain differences in the results of thermoplasty.
Quality of the evidence
The available trials were designed adequately, had an adequate allocation of participants and had a complete follow-up. The only source of bias is the open design of two of the trials (AIR; RISA), which compared the procedure with medical management only. In these trials, a placebo effect cannot be discarded mainly for subjective outcomes such as quality of life, which is critical for judging the benefits of this intervention. Despite this, the remaining trial ensured blinding of participants by comparing bronchial thermoplasty versus a sham procedure in participants allocated to the control group (AIR 2). Even then, however, a substantial beneficial placebo effect was seen in the control group.
A marked baseline clinical imbalance is evident in the RISA trial, in which participants treated with bronchial thermoplasty showed more symptoms at baseline, which poses an additional source of potential bias and concern for interpretation. This trial reported its results in terms of changes from baseline, omitting the absolute outcome data at the end of follow-up. For the purposes of our pooled analyses, we converted the data from this trial into final scores and imputed baseline standard deviation into our estimates. As a result of this, our estimates could be biased by the direction of the baseline imbalances.
Despite some differences between the trial design (including different comparators in the AIR 2 trial) and the characteristics of included participants (e.g. 100% participants in the RISA trial had severe persistent asthma compared to 57% and 87% in AIR and AIR 2 respectively), the heterogeneity analyses did not reveal these differences as a source of relevant inconsistency.
Potential biases in the review process
Although we tried to follow rigorous standards in this review and performed an exhaustive search that allowed us to include all trials that have assessed the effects of bronchial thermoplasty to date, no review process is completely free of bias. A potential limitation is that the results of our review rely exclusively on data reported in the trial results publications in biomedical journals and in communications to scientific meetings, and we did not formally contact the trial authors. Interpretation of trial results was not always as transparent as expected, as was the case for the AIR trial, whose main publication reported its main results in graphical form only. In future updates of the review, we will try to obtain additional information directly from trialists to avoid this limitation.
Implications for practice
Bronchial thermoplasty is a novel procedure tested for moderate to severe asthma that provides modest clinical benefit for quality of life but with increased risk of admission to hospital during treatment, which diminishes subsequently. The clearest advantage is related to a reduction in exacerbations and hospitalisations post-treatment. The benefit in symptoms appeared to be greater in the study that recruited exclusively severe patients on the basis of symptoms, although unfortunately, the main outcome measure was adverse events (RISA), and this study did not include a sham treatment arm. It would be advisable to follow up on the outcomes from participants treated with this procedure by collecting data via independent clinical registries.
Implications for research
Further research is needed in response to several issues, especially regarding the mechanism of action of thermoplasty in asthma patients, identification of patients and asthma phenotypes that could benefit to the greatest extent, with particular focus on the effect of the intervention in patients with the most severe disease, lower FEV
The authors would like to acknowledge Elizabeth Stovold for designing and updating the search of this review. Also thanks to Emma Welsh and Chris Cates for support and editorial assistance provided during development of the review.
Data and analyses
- Top of page
- Summary of findings [Explanations]
- Authors' conclusions
- Data and analyses
- Contributions of authors
- Declarations of interest
- Sources of support
- Differences between protocol and review
- Index terms
Appendix 1. Sources and search methods for the Cochrane Airways Group Specialised Register (CAGR)
Electronic searches: core databases
Handsearches: core respiratory conference abstracts
MEDLINE search strategy used to identify trials for the CAGR
1. exp Asthma/
3. (antiasthma$ or anti-asthma$).mp.
4. Respiratory Sounds/
6. Bronchial Spasm/
8. (bronch$ adj3 spasm$).mp.
10. exp Bronchoconstriction/
11. (bronch$ adj3 constrict$).mp.
12. Bronchial Hyperreactivity/
13. Respiratory Hypersensitivity/
14. ((bronchial$ or respiratory or airway$ or lung$) adj3 (hypersensitiv$ or hyperreactiv$ or allerg$ or insufficiency)).mp.
15. ((dust or mite$) adj3 (allerg$ or hypersensitiv$)).mp.
17. exp Aspergillosis, Allergic Bronchopulmonary/
18. lung diseases, fungal/
20. 18 and 19
21. (bronchopulmonar$ adj3 aspergillosis).mp.
22. 17 or 20 or 21
23. 16 or 22
24. Lung Diseases, Obstructive/
25. exp Pulmonary Disease, Chronic Obstructive/
27. (chronic$ adj3 bronchiti$).mp.
28. (obstruct$ adj3 (pulmonary or lung$ or airway$ or airflow$ or bronch$ or respirat$)).mp.
34. exp Bronchiectasis/
38. (ciliary adj3 dyskinesia).mp.
39. (bronchial$ adj3 dilat$).mp.
41. exp Sleep Apnea Syndromes/
42. (sleep$ adj3 (apnea$ or apnoea$)).mp.
43. (hypopnoea$ or hypopnoea$).mp.
48. Lung Diseases, Interstitial/
49. Pulmonary Fibrosis/
50. Sarcoidosis, Pulmonary/
51. (interstitial$ adj3 (lung$ or disease$ or pneumon$)).mp.
52. ((pulmonary$ or lung$ or alveoli$) adj3 (fibros$ or fibrot$)).mp.
53. ((pulmonary$ or lung$) adj3 (sarcoid$ or granulom$)).mp.
55. 23 or 33 or 40 or 47 or 54
Filter to identify RCTs
1. exp "clinical trial [publication type]"/
2. (randomised or randomised).ab,ti.
11. 9 not (9 and 10)
12. 8 not 11
The MEDLINE strategy and the RCT filter are adapted to identify trials in other electronic databases.
Contributions of authors
AT, IS: eligibility
AT, IS, JJY-N, AMM: data extraction
AT, IS: risk of bias assessment
IS: summary of findings table
IS, MRF: data analysis
AT, IS, AMM: data interpretation and drafting of the review
All review authors have contributed to the conception of the review and have provided relevant contributions to its discussion and conclusions.
Haydn Walters was the assigned editor for this review. Haydn commented critically and edited the review text.
Declarations of interest
Dr A Torrego is a member of a Scientific Steering Committee of an international registry of patients treated with bronchial thermoplasty. He works at the Respiratory Department of the Hospital de la Santa Creu I Sant Pau, and since October 12 performs bronchial thermoplasty in patients with severe asthma.
Sources of support
- Iberoamerican Cochrane Centre, Spain.
- Programa Enlaza-Mundos. Alcaldía de Medellín 2011, Colombia.Postgraduate Training Fellowship
Differences between protocol and review
Although the protocol of the review stated that adverse events would be reported in terms of serious complications derived from bronchial thermoplasty, the included trials reported data about relevant adverse events during the treatment period, and so we decided to include them in that form. The protocol stated that we would contact authors of included trials to ask them about other published or unpublished studies, but we did not contact them formally, as we had stated in the protocol we would do.
Asthma control was added as a primary outcome and adverse events was added as a secondary outcome.
Medical Subject Headings (MeSH)
Asthma [*therapy]; Bronchi [*surgery]; Disease Progression; Emergency Service, Hospital [utilization]; Hospitalization [statistics & numerical data]; Hypertrophy [therapy]; Muscle, Smooth [pathology]; Pulsed Radiofrequency Treatment [adverse effects; *methods]; Quality of Life; Randomized Controlled Trials as Topic; Time Factors
MeSH check words
* Indicates the major publication for the study