Abstract
Rationale
AIRFLOW-3 was a 1:1 randomized, double-blind, sham-controlled trial of the D’Nerva targeted lung denervation (TLD) system in patients with chronic obstructive pulmonary disease (COPD).
Objective
To evaluate the impact of TLD on COPD exacerbations compared with optimal medical treatment.
Methods
AIRFLOW-3 patients were symptomatic (COPD Assessment Test score ⩾10) with moderate to very severe airflow obstruction (FEV1 ≥25% but ⩽80% predicted) and Global Initiative for Chronic Obstructive Lung Disease E status (at least two moderate or one severe exacerbation during the previous 12 mo). The primary endpoint was a comparison of time to the first moderate or severe COPD exacerbation through 12 months between the treatment (TLD plus optimal medical treatment) and sham control groups (sham procedure plus optimal medical treatment). Secondary endpoints included the rate of severe COPD exacerbations, change in quality of life (St. George’s Respiratory Questionnaire for COPD and Short Form–36 questionnaire), change in lung function (FVC, FEV1, residual volume), and change in COPD Assessment Test score.
Measurements and Main Results
388 patients were randomized at a 1:1 ratio at 32 sites. There was no difference between treatment and the sham procedure in terms of the probability of participants having a moderate or severe COPD exacerbation (hazard ratio, 1.268; 95% confidence interval, 0.988–1.628). At 1 year, the TLD group had less dyspnea (>1-point improvement in Transitional Dyspnea Index, 35.4 vs. 24.1%; P = 0.021) compared with the sham group. Post hoc analyses suggests that failure to reach the primary endpoint was driven by an insufficient number of patients exhibiting an airway-predominant phenotype (lung hyperinflation without significant emphysema).
Conclusions
AIRFLOW-3 failed to meet its primary endpoint. However, post hoc analyses identified a responder profile; a prospective multicenter randomized controlled trial is being designed to confirm these findings.
Clinical trial registered with www.clinicaltrials.gov (NCT 03639051).
Keywords: COPD, targeted lung denervation, acetylcholine, anticholinergic, bronchoscopy
At a Glance Commentary
Scientific Knowledge on the Subject
Targeted lung denervation (TLD) is an emerging therapy for patients with moderate to severe chronic obstructive pulmonary disease (COPD) that is symptomatic despite optimal medical care. Previous clinical studies have demonstrated the potential for TLD to preserve lung function, provide symptom relief, and reduce COPD exacerbations.
What This Study Adds to the Field
The present study was a large randomized, double-blind, sham-controlled trial evaluating the impact of TLD on the reduction of COPD exacerbations in patients with moderate to severe COPD with a history of COPD exacerbations. The study failed its primary endpoint but demonstrated preservation of lung function and reduction of dyspnea in TLD-treated patients. A subgroup of TLD responders was described for future evaluation in a clinical trial.
Many patients with chronic obstructive pulmonary disease (COPD) remain symptomatic despite optimal medical treatment (1–5). The human lung is extensively innervated by pulmonary branches of the vagus nerve, which impact lung physiology in health and disease (6–8). In COPD, the autonomic nervous system becomes dysregulated. Airway smooth muscle tone, which is under vagal control, drives excessive bronchoconstriction and hyperinflation that presents as dyspnea in patients with obstructive lung disease (9). Blunting the impact of the parasympathetic nervous system in patients with COPD has been a focus of pharmacotherapy in COPD (i.e., inhaled long-acting muscarinic antagonists and long-acting β-antagonists).
Targeted lung denervation (TLD) is a single outpatient procedure performed under general anesthesia. Both lungs receive circumferential radiofrequency (RF) ablation of the pulmonary nerves at the juncture of the mainstem bronchi to permanently disrupt parasympathetic nerve signaling and reduce neural hyperactivity (10). Prior studies demonstrated the ability of the D’Nerva lung denervation system to successfully denervate the lungs and diminish parasympathetic signaling to and from the lungs (11). Across multiple clinical trials, patients undergoing D’Nerva lung denervation exhibited durable reductions in COPD exacerbations and stabilization of lung-function and quality-of-life measures (12, 13).
AIRFLOW-3 (clincaltrials.gov identifier NCT 03639051) was a prospective randomized multicenter trial designed to evaluate the impact of TLD on exacerbation reduction in patients with a history of exacerbations (GOLD class E) and symptom burden (COPD Assessment Test [CAT] score ⩾10) despite optimal medical treatment (14). Some of the results of these studies have been previously reported in the form of an abstract (15).
Methods
Trial Design
AIRFLOW-3 trial was conducted under a U.S. Food and Drug Administration– approved Investigational Device Exemption for the D’Nerva lung denervation system. A total of 388 patients were randomized between September 2019 and August 2023. The study was approved by each site’s institutional review board or ethics committee, and all participants provided written informed consent.
The study design was prospective, multicenter, 1:1–randomized, sham-controlled, and double-blinded (subjects and follow-up assessors). The primary endpoint was defined as the time to the first moderate or severe COPD exacerbation through 12 months. COPD exacerbations were defined as a complex of respiratory events and/or symptoms (increase or new onset) of more than one of the following: cough, sputum, wheezing, dyspnea, or chest tightness with at least one symptom lasting ≥3 days requiring treatment with antibiotics and/or systemic steroids (moderate exacerbations) and/or hospitalization (severe exacerbations) (16).
Participants
Participants eligible for randomization were ⩾40 years of age with a smoking history of ≥10 pack-years, a diagnosis of COPD with moderate to severe airflow obstruction (postbronchodilator FEV1 ≥25% but ⩽80% and FEV1/FVC ratio <70%), and a documented history of at least two moderate or one severe COPD exacerbation in the 12 months before consent while receiving optimal medical treatment (minimum of long-acting dual bronchodilation inhaled therapy with or without inhaled corticosteroids [ICSs] per GOLD guidance [17]). Patients were excluded from randomization if a chest computed tomography scan revealed the presence of whole-lung emphysema >50% (density mask threshold of <−950 HU; VIDA Diagnostics), symptomatic gastric motility disorder (Gastroparesis Cardinal Symptom Index score of ⩾18 [18]), or diagnosis of a dominant non-COPD lung disease (e.g., cystic fibrosis, asthma). See data supplement for the full list of inclusion and exclusion criteria.
Participants were randomized at 19 sites in the United States and 12 sites in Europe and the United Kingdom (see data supplement).
Intervention
After consent, all baseline tests and assessments were completed per protocol, and patients were scheduled for the study bronchoscopy (see data supplement). Randomization occurred after the patient was under general anesthesia in preparation for the study bronchoscopy. Participants were randomized (1:1 ratio) into two groups: TLD with continuation of optimal medical treatment (TLD group) or sham bronchoscopy with continuation of optimal medical treatment (sham group).
All TLD and sham procedures were performed by a physician. Participants randomized to the sham group underwent a sham procedure with the D’Nerva system (the D’Nerva catheter was manipulated into the main stem bronchi and the balloon was inflated, but no fluoroscopy or RF energy was delivered). The TLD group underwent active treatment with the D’Nerva system (fluoroscopy and RF energy was delivered). TLD was performed as previously described (14).
Participants were receiving protocol-directed medical treatment at baseline and were encouraged to not change treatment during the 12 months of primary endpoint follow-up unless there was a medical need. All medication changes were recorded through 12 months.
Randomized patients were contacted by phone at 7, 14, and 21 days after bronchoscopy and then monthly through 12 months. Office visits occurred at 1, 6, and 12 months.
Primary Outcome
The primary endpoint was defined as the time to the first occurrence of a moderate or severe COPD exacerbation between the date of randomization and the date of a subject’s first primary endpoint event or to the close date of the 12-month visit window for subjects who did not experience a primary endpoint event. The statistical test for the primary endpoint was a log-rank test comparing the survival distribution of the time to first event for the primary endpoint between the TLD and sham groups. A blinded independent clinical events committee was responsible for adjudicating every COPD exacerbation to confirm the event met the definition of a primary endpoint event.
Secondary Outcomes
Secondary endpoints included the time to the first severe COPD exacerbation (0–12 mo);12-month absolute change in St. George’s Respiratory Questionnaire for COPD (SGRQ-C) score, FVC, FEV1, residual volume (RV), or Short Form–36 score; 12-month Transitional Dyspnea Index (TDI); and proportion of patients achieving a minimal clinically important difference (MCID) at 12 months on the CAT. Additional prespecified outcomes included 12-month change in inspiratory capacity and total lung capacity, 12-month proportion of patients achieving an MCID in dyspnea scores (modified Medical Research Council (mMRC) dyspnea scale and TDI), quality-of-life metrics (SGRQ-C and CAT), and annualized rate of moderate and severe COPD exacerbations (0–12 mo and 3–12 mo). See data supplement for a complete list of prespecified secondary and additional outcomes and associated statistical analysis.
Sample Size
Sample size for the study was based on the results from the AIRFLOW-2 clinical trial (19). Assuming a 1:1 randomization allocation and percentages of subjects with primary endpoint events through 12 months of 65% for the sham group and 48.75% for the TLD group, a sample size of 375 was shown to provide >90% power based on a one-sided 0.025 α-level log-rank test. A total of 388 patients were randomized between the arms of the study.
Randomization and Blinding
Randomization was stratified based on investigational site, patient participation in a pulmonary rehabilitation maintenance program, and baseline use of an ICS. A study blinding plan was implemented at each site to ensure participant and outcome assessor double-blinding was maintained throughout the 12-month follow-up period. An unblinded index procedure team conducted study procedures and were excluded from subsequent contact with patients through 12 months of follow-up. A blinded outcome assessor team began contact with the participant after the index procedure and performed all follow-up activities through 12 months. At the close of the 12-month follow-up visit window, and after completing the blinding questionnaire, randomized subjects were notified of their treatment status.
Statistical Analyses
A P value <0.05 was considered statistically significant for two-sided tests. Descriptive statistics included means, standard deviations, and 95% confidence intervals (CIs). Time-to-first-event analyses were performed as a comparison between study groups of the survival distributions for events based on a log-rank test. Continuous variables were compared between study groups based on a linear model for change of the outcome of interest adjusted for its baseline value. Categorical variables were compared between study groups based on an exact test for binomial proportions. All statistical analyses were performed using SAS version 9.4 (SAS Institute). Primary analyses for study objectives (including the primary endpoint, secondary endpoints, and additional prespecified endpoints) were conducted on an intent-to-treat (ITT) basis, for which subjects were analyzed according to their randomized assignment irrespective of the treatment received.
The post hoc analysis was informed by a mechanistic understanding of the role of nerves in pulmonary physiology (9, 20–24), as well as physiologic outcomes from AIRFLOW-3 and other prior trials of D’Nerva lung denervation (12, 25, 26). A subgroup analysis was conducted to assess the influence of baseline parameters on 12-month outcomes of lung function. Subgroup correlations between baseline parameters and 12-month change in lung function were evaluated using Pearson’s correlation coefficient. Two-sample Student t tests were used to assess statistical differences of change in lung function at 12 months between the TLD and sham groups within each subgroup. Statistical analysis outside of the primary endpoint presented and discussed is viewed as hypothesis-generating (i.e., P values should be considered nominal).
Results
Demographic and Procedure Details
A total of 388 patients met the inclusion and exclusion criteria and were randomized (TLD group, 198 patients; sham group, 190 patients). A Consolidated Standards of Reporting Trials diagram is shown in Figure 1. Both groups were well matched for all baseline demographic and clinical characteristics (Table 1). Total bronchoscopy time for the active procedure (113.6 ± 62.4 min; n = 198) was longer than for the sham procedure (80.8 ± 21.6 min; n = 190). Other procedure characteristics are listed in Table E6 in the data supplement and were not different between groups. A total of 92% of patients in both groups returned home on the same day as the procedure. Subject retention was high, with 95% (371 of 388) retained through 1 year.
Figure 1.
Consolidated Standards of Reporting Trials flow chart. AE = adverse event; TLD = targeted lung denervation.
Table 1.
Baseline Demographic and Clinical Characteristics
| Characteristic | TLD | Sham |
|---|---|---|
| Age, yr | ||
| Mean ± SD (n) | 67.8 ± 7.2 (198) | 67.7 ± 7.0 (190) |
| Median (range) | 68.0 (45.0–89.0) | 69.0 (42.0–84.0) |
| Female sex | 111/198 (56.1%) | 103/190 (54.2%) |
| Body mass index, kg/m2 | ||
| Mean ± SD (n) | 26.9 ± 4.6 (198) | 27.1 ± 4.6 (190) |
| Median (range) | 27.0 (4.0–35.0) | 27.0 (18.0–35.0) |
| Pack-years smoking | ||
| Mean ± SD (n) | 47.6 ± 21.5 (198) | 45.8 ± 22.5 (190) |
| Median (range) | 44.0 (10.0–201.0) | 43.0 (10.0–128.0) |
| Lung function (postbronchodilator) | ||
| FEV1, L | ||
| Mean ± SD (n) | 1.07 ± 0.37 (198) | 1.07 ± 0.39 (190) |
| Median (range) | 0.98 (0.55–2.71) | 1.00 (0.49–2.59) |
| FEV1, % predicted | ||
| Mean ± SD (n) | 41.6 ± 11.9 (198) | 40.5 ± 11.8 (190) |
| Median (range) | 39.0 (25.0–80.0) | 38.0 (25.0–78.0) |
| FEV1/FVC | ||
| Mean ± SD (n) | 0.39 ± 0.10 (198) | 0.39 ± 0.11 (190) |
| Median (range) | 0.37 (0.18–0.68) | 0.37 (0.18–0.70) |
| GOLD stage (postbronchodilator FEV1) | ||
| I | 2/198 (1.0%) | 0/190 (0.0%) |
| II | 44/198 (22.2%) | 30/190 (15.8%) |
| III | 127/198 (64.1%) | 133/190 (70.0%) |
| IV | 25/198 (12.6%) | 27/190 (14.2%) |
| % Emphysema score −950 HU | ||
| Mean ± SD (n) | 22.1 ± 12.1 (198) | 21.5 ± 11.1 (190) |
| Median (range) | 22.4 (0.2–47.3) | 20.9 (0.5–48.3) |
| Patient-reported outcomes | ||
| CAT score | ||
| Mean ± SD (n) | 23.5 ± 6.2 (198) | 23.8 ± 6.5 (190) |
| Median (range) | 24.0 (10.0–39.0) | 24.0 (10.0–40.0) |
| SGRQ-C, total score | ||
| Mean ± SD (n) | 59.0 ± 14.6 (198) | 59.8 ± 16.0 (190) |
| Median (range) | 60.0 (19.4–91.4) | 61.2 (19.5–92.3) |
| GCSI score | ||
| Mean ± SD (n) | 0.65 ± 0.61 (198) | 0.60 ± 0.59 (190) |
| Median (range) | 0.50 (0.00–2.25) | 0.50 (0.00–2.25) |
| mMRC score | ||
| Mean ± SD (n) | 2.58 ± 0.84 (198) | 2.58 ± 0.99 (190) |
| Median (range) | 3.00 (0.00–4.00) | 3.00 (0.00–4.00) |
| Baseline dyspnea index | ||
| Mean ± SD (n) | 4.6 ± 2.2 (194) | 4.5 ± 2.1 (187) |
| Median (range) | 5.0 (0.0–12.0) | 4.0 (0.0–10.0) |
| Prior-year exacerbation history | ||
| Mean no. of moderate/severe exacerbations | ||
| Mean ± SD (n) | 2.9 ± 1.5 (198) | 3.0 ± 1.5 (190) |
| Median (range) | 3.0 (1.0–12.0) | 3.0 (1.0–10.0) |
| Mean no. severe (hospitalized) exacerbations | ||
| Mean ± SD (n) | 0.6 ± 0.9 (198) | 0.7 ± 0.9 (190) |
| Median (range) | 0.0 (0.0–7.0) | 0.0 (0.0–5.0) |
| Baseline COPD maintenance medication use | ||
| LABA monotherapy | 0/198 | 0/190 |
| LAMA monotherapy | 0/198 | 0/190 |
| LABA + LAMA | 14/198 (7.1%) | 12/190 (6.3%) |
| LABA + ICS* | 1/198 (0.5%) | 1/190 (0.5%) |
| LAMA + ICS | 1/198 (0.5%) | 0/190 |
| LABA + LAMA + ICS | 182/198 (91.9%) | 177/190 (93.2%) |
| Other COPD maintenance medications | 174/198 (87.9%) | 161/190 (84.7%) |
Definition of abbreviations: CAT = COPD Assessment Test; COPD = chronic obstructive pulmonary disease; GCSI = Gastroparesis Cardinal Symptom Index; GOLD = Global Initiative for Chronic Obstructive Lung Disease; ICS = inhaled corticosteroid; LABA = long-acting β-antagonist; LAMA = long-acting muscarinic antagonist; mMRC = modified Medical Research Council; SGRQ-C = St. George’s Respiratory Questionnaire for COPD; TLD = targeted lung denervation.
The patient was receiving LAMA/LABA/inhaled corticosteroid (triple therapy) through most of the 1 year before study enrollment and had two moderate and two severe (qualifying) exacerbations while receiving triple therapy. At the time of consent, the patient switched to inhaled corticosteroid/LABA and a short-acting β-antagonist. Medical records provided to the sponsor by the study site indicated a medication change due to drug intolerance, satisfying the inclusion criteria.
Primary Endpoint Analysis
The comparison of the survival distribution of the time-to-first-primary-endpoint event for the TLD and sham groups resulted in a hazard ratio of 1.27 (95% CI, 0.988–1.628 per ITT analysis), which was not statistically different between groups (Figure 2A).
Figure 2.
Moderate and severe chronic obstructive pulmonary disease (COPD) exacerbation (intention-to-treat population). (A) The time-to-first-event analysis with a primary endpoint window of 0–12 months of follow-up. (B) The time-to-first-event analysis with a primary endpoint window of 3–12 months of follow-up. (C) Annualized rate of moderate and severe COPD exacerbations with an endpoint window of 0–12 months of follow up. (D) Annualized rate of moderate and severe COPD exacerbations with an endpoint window of 3–12 months of follow-up. Error bars indicate 95% confidence intervals.
A Cox proportional hazards model was used to calculate the hazard ratio for moderate or severe COPD exacerbations from the date of randomization to 12 months after the procedure. The model included individual study site, prior participation in a pulmonary rehabilitation program, and baseline use of ICS as strata to control for confounding effects. The assumption of proportional hazards for the primary analysis was tested and found to be invalid (cumulative sums of martingale residuals test, P = 0.013). As such, the hazard ratio for treatment type was recalculated using the prespecified 3–12-month periprocedural window with the same strata described previously. Excluding the 3-month periprocedural window resulted in a valid model fit (P = 0.994) (Figure 2B). Overall, no significant difference was indicated between treatment types in either model (3–12 mo: hazard ratio, 0.793; 95% CI, 0.534–1.177). The variability of moderate or severe COPD exacerbations (primary endpoint events) within 3 months of the procedure in the treatment group are shown in Figure E1.
Secondary Outcomes
The analysis of time to the first severe COPD exacerbation (0–12 mo) was not statistically different, with a hazard ratio of 1.62 (95% CI, 0.98–2.69) (Figure 3A). As with the primary endpoint, the 3–12-month window satisfied the model assumptions and showed no statistical difference between study groups, with a hazard ratio of 1.12 (95% CI, 0.59–2.11) (Figure 3B). The mean change from baseline to 12 months in SGRQ-C in the TLD group was −3.1 ± 14.1 (n = 179; P = 0.004 vs. baseline), compared with −2.1 ± 13.4 (n = 176; P = 0.036 vs. baseline) in the sham group, and was not different between groups (P = 0.420). At 12 months after the procedure, the blinded sham group exhibited a statistical decline across key secondary measures of lung function: FVC (−0.07 ± 0.45 L vs. baseline [n = 173]; P = 0.039), FEV1 (−0.05 ± 0.17 L vs. baseline [n = 173]; P < 0.001), and RV (0.16 ± 0.88 L vs. baseline [n = 170]; P = 0.017). Conversely, the blinded TLD group exhibited no decline in lung function over 12 months of follow-up: FVC (−0.01 ± 0.41 L vs. baseline [n = 176]; P = 0.720), FEV1 (−0.01 ± 0.15 L vs. baseline [n = 176]; P = 0.281), and RV (−0.02 ± 1.03 L vs. baseline [n = 168]; P = 0.791). There were no statistical differences between the TLD and sham groups at 12 months with respect to these lung function parameters (FVC, P = 0.228; FEV1, P = 0.068; RV, P = 0.094; Table 2). See data supplement for a post hoc interaction analysis of pulmonary function changes versus baseline. No difference was observed between groups in the mean change from baseline of the CAT or Short Form–36 scores (Table 3).
Figure 3.
Severe chronic obstructive pulmonary disease (COPD) exacerbation (intention-to-treat population). (A) The time-to-first-event analysis with a primary endpoint window of 0–12 months of follow-up. (B) The time-to-first-event analysis with a primary endpoint window of 3–12 months of follow-up. (C) Annualized rate of severe COPD exacerbations with an endpoint window of 0–12 months of follow-up. (D) Annualized rate of severe COPD exacerbations with an endpoint window of 3–12 months of follow-up. Error bars indicate 95% confidence intervals.
Table 2.
Lung Function Data
| Baseline | 12 Months | Change vs. Baseline | P Value | |
|---|---|---|---|---|
| FVC, L | ||||
| TLD | 2.82 ± 0.91 (198) | 2.79 ± 0.87 (176) | −0.01 ± 0.41 (176) | 0.720 |
| Sham | 2.82 ± 0.86 (190) | 2.79 ± 0.88 (173) | −0.07 ± 0.45 (173) | 0.039 |
| P value* | — | — | 0.228 | — |
| FEV1, L | ||||
| TLD | 1.07 ± 0.37 (198) | 1.05 ± 0.36 (176) | −0.01 ± 0.15 (176) | 0.281 |
| Sham | 1.07 ± 0.39 (190) | 1.04 ± 0.40 (173) | −0.05 ± 0.17 (173) | <0.001 |
| P value* | — | — | 0.068 | — |
| Residual volume, L | ||||
| TLD | 3.94 ± 1.15 (197) | 3.90 ± 1.18 (169) | −0.02 ± 1.03 (168) | 0.791 |
| Sham | 3.86 ± 1.07 (190) | 4.01 ± 1.15 (170) | 0.16 ± 0.88 (170) | 0.017 |
| P value* | — | — | 0.094 | — |
| Inspiratory capacity, L | ||||
| TLD | 1.94 ± 0.54 (197) | 1.93 ± 0.55 (170) | −0.02 ± 0.40 (169) | 0.427 |
| Sham | 2.01 ± 0.64 (190) | 1.95 ± 0.65 (170) | −0.09 ± 0.38 (170) | 0.001 |
| P value* | — | — | 0.217 | — |
| TLC, L | ||||
| TLD | 6.82 ± 1.51 (197) | 6.76 ± 1.54 (169) | −0.05 ± 0.94 (168) | 0.505 |
| Sham | 6.80 ± 1.43 (190) | 6.92 ± 1.49 (170) | 0.10 ± 0.88 (170) | 0.140 |
| P value* | — | — | 0.110 | — |
Definition of abbreviations: TLC = total lung capacity; TLD = targeted lung denervation.
Data presented as mean ± SD (n).
P values for change vs. baseline are based on a linear model adjusted for baseline; P values for responders are based on a χ2 test.
Table 3.
Patient-reported Quality-of-Life Outcomes
| Outcome | Baseline | 12 Months | Change vs. Baseline | P Value |
|---|---|---|---|---|
| SF-36 mental component score | ||||
| TLD | 46.5 ± 12.1 (198) | 46.6 ± 12.2 (166) | 0.8 ± 10.7 (166) | 0.320 |
| Sham | 45.1 ± 12.8 (190) | 43.7 ± 13.5 (161) | −0.6 ± 12.5 (161) | 0.558 |
| P value* | — | — | 0.087 | — |
| SF-36 physical component score | ||||
| TLD | 35.0 ± 6.6 (198) | 35.6 ± 8.0 (166) | 0.6 ± 7.0 (166) | 0.298 |
| Sham | 35.3 ± 7.7 (190) | 35.4 ± 7.3 (161) | 0.1 ± 7.5 (161) | 0.890 |
| P value* | — | — | 0.629 | — |
| SGRQ-C | ||||
| TLD | 59.0 ± 14.6 (198) | 55.9 ± 16.5 (179) | −3.1 ± 14.1 (179) | 0.004 |
| Sham | 59.8 ± 16.0 (190) | 57.5 ± 18.0 (176) | −2.1 ± 13.4 (176) | 0.036 |
| P value* | — | — | 0.420 | — |
| ⩾4-point decrease | — | — | 79/179 (44.1%) | — |
| TLD | — | — | 76/176 (43.2%) | — |
| Sham | — | — | 0.856 | — |
| COPD Assessment Test | ||||
| TLD | 23.5 ± 6.2 (198) | 21.4 ± 7.1 (182) | −2.2 ± 6.8 (182) | <0.001 |
| Sham | 23.8 ± 6.5 (190) | 22.6 ± 7.6 (178) | −1.2 ± 7.3 (178) | 0.033 |
| P value* | — | — | 0.104 | — |
| ⩾2-point decrease | ||||
| TLD | — | — | 101/182 (55.5%) | — |
| Sham | — | — | 83/178 (46.6%) | — |
| P value* | — | — | 0.092 | — |
Definition of abbreviations: COPD = chronic obstructive pulmonary disease; SF = short form; SGRQ-C = St. George’s Respiratory Questionnaire for COPD; TLD = targeted lung denervation.
Data presented as mean ± SD (n) where applicable.
P values for change vs. baseline are based on a linear model adjusted for baseline; P values for responders are based on a χ2 test.
In terms of impact on the sensation of dyspnea, the mean TDI score at 12 months approached significance between groups (P = 0.09; Table 3). The percentage of patients with an MCID change in TDI and mMRC score at 12 months was statistically greater in the TLD group (35%) compared with the sham group (24%) (Table 4). Annualized rates of moderate and severe COPD exacerbations and of severe COPD exacerbations alone are shown in Figures 2 and 3C and 3D. No statistical differences were found between groups.
Table 4.
Dyspnea Data
| Baseline | 12 Months | Change vs. Baseline | P Value | |
|---|---|---|---|---|
| TDI* | ||||
| TLD | — | — | −0.43 ± 4.15 (178) | 0.171 |
| Sham | — | — | −1.14 ± 3.75 (174) | <0.001 |
| P value† | — | — | 0.090 | — |
| ⩾1-point increase | ||||
| TLD | — | — | 63/178 (35.4%) | — |
| Sham | — | — | 42/174 (24.1%) | — |
| P value† | — | — | 0.021 | — |
| mMRC | ||||
| TLD | 2.58 ± 0.84 (198) | 2.30 ± 0.93 (181) | −0.25 ± 0.96 (181) | <0.001 |
| Sham | 2.58 ± 0.99 (190) | 2.56 ± 1.04 (178) | −0.02 ± 0.93 (178) | 0.809 |
| P value† | — | — | 0.006 | — |
| ⩾1-point decrease | ||||
| TLD | — | — | 64/181 (35.4%) | — |
| Sham | — | — | 46/178 (25.8%) | — |
| P value† | — | — | 0.051 | — |
Definition of abbreviations: mMRC = modified Medical Research Council; TDI = Transition Dyspnea Index; TLD = targeted lung denervation.
Data presented as mean ± SD (n) where applicable.
TDI is collected only at the 12-month follow-up visit as change. P value based on ANOVA.
P values for change vs. baseline are based on a linear model adjusted for baseline; P values for responders are based on a χ2 test.
Post Hoc Analysis
A post hoc analysis was performed to investigate the failure of the TLD therapy to significantly impact physiological outcomes in AIRFLOW-3. Based on the impact of lung denervation on airway mechanics (20, 27, 28), and previous studies of TLD (12, 25, 26), it was expected that denervated patients would experience a reduction in airway tone, reduction in gas trapping (i.e., decrease in RV), and a consequent improvement in airflow (i.e., increase in FVC and FEV1).
An analysis of 12-month changes in postbronchodilator FEV1 between the TLD and sham groups was performed by stratifying for baseline covariates of gas trapping (postbronchodilator RV percent predicted) and whole-lung emphysema (percent lobar attenuation area <−950 HU). The findings revealed a negative correlation between the extent of baseline emphysema and 12-month improvement in FEV1 (Pearson correlation coefficient of −0.21 for TLD). An increase in the magnitude of FEV1 improvement in the TLD group versus the sham group was observed with greater degrees of baseline hyperinflation and lower levels of quantified lung emphysema (Figure 4A). This analysis was also applied to COPD exacerbation rates; greater reductions in exacerbation frequency were observed with increasing baseline hyperinflation levels and decreasing baseline emphysema (Figure 4B). These findings were independent of patient sex, age, blood eosinophilia, and study site or region.
Figure 4.
AIRFLOW-3 post hoc responder analysis: (A) difference in change in FEV1 between targeted lung denervation (TLD) treatment and sham groups at 12 months after randomization as a function of baseline levels of hyperinflation (residual volume percent predicted [RV% pred]) and emphysema (percent low-attenuation area <−950 HU). (B) Percent reduction in moderate and severe chronic obstructive pulmonary disease (exacerbations in the TLD treatment group vs. the sham group as a function of baseline levels of hyperinflation [RV% pred] and emphysema [percent low-attenuation area <−950 HU; *P < 0.05, TLD group vs. sham group). AECOPD = acute exacerbation of chronic obstructive pulmonary disease.
Adverse Events
The serious adverse event (SAE) rate through 1 year of follow-up was similar between treatment groups: 37.4% (TLD) and 34.2% (sham) (Table 5). All-cause mortality was similar between groups: 4.5% (TLD) and 4.2% (sham). The rates of respiratory-related death were 2.0% in the TLD arm and 3.7% in the sham arm. Within the first 3 months of the procedure, the SAE rate was higher in the TLD group compared with the sham group (21.7% and 10.5%, respectively) (Table 6). Tables E7 and E8 list all SAEs from 3 to 12 months and all follow-up. The between-group SAE differences are driven by higher rates of periprocedural (≤3 mo) respiratory and gastrointestinal (GI) SAEs in the TLD group. All periprocedural respiratory SAEs (14% in the TLD group and 7% in the sham group) resolved with standard care. The only exceptions were two in the sham group (a death discussed later and a COPD exacerbation) and two in the TLD group (a death and a bronchial esophageal fistula discussed later).
Table 5.
Serious Adverse Events at 0–12 Months Occurring in ≥2.5% of Subjects in Either Group
| MedDRA System Organ Class and Lowest-Level Term | TLD |
Sham |
||
|---|---|---|---|---|
| Total Events | Subjects with Events | Total Events | Subjects with Events | |
| All events | 149 | 74/198 (37.4%) | 114 | 65/190 (34.2%) |
| Respiratory, thoracic, and mediastinal disorders | 84 | 52/198 (26.3%) | 54 | 40/190 (21.1%) |
| COPD exacerbation | 68 | 43/198 (21.7%) | 38 | 30/190 (15.8%) |
| Respiratory failure | 5 | 5/198 (2.5%) | 7 | 6/190 (3.2%) |
| Pneumothorax | 1 | 1/198 (0.5%) | 5 | 5/190 (2.6%) |
| Gastrointestinal disorders | 16 | 13/198 (6.6%) | 7 | 7/190 (3.7%) |
| Gastrointestinal motility disorder | 5 | 5/198 (2.5%) | 0 | 0/190 |
| Cardiac disorders | 12 | 11/198 (5.6%) | 7 | 7/190 (3.7%) |
| Dysrhythmias | 6 | 5/198 (2.5%) | 2 | 2/190 (1.1%) |
| Infections and infestations | 19 | 16/198 (8.1%) | 22 | 18/190 (9.5%) |
| Pneumonia | 11 | 10/198 (5.1%) | 18 | 15/190 (7.9%) |
| Injury, poisoning, and procedural complications | 5 | 5/198 (2.5%) | 5 | 5/190 (2.6%) |
| Neoplasms benign, malignant, and unspecified (including cysts, polyps) | 2 | 2/198 (1.0%) | 5 | 5/190 (2.6%) |
| All-cause mortality | — | 9/198 (4.5%) | — | 8/190 (4.2%) |
| Respiratory-related mortality | — | 4/198 (2.0%) | — | 7/190 (3.7%) |
Definition of abbreviations: COPD = chronic obstructive pulmonary disease; MedDRA = Medical Dictionary for Regulatory Activities; TLD = targeted lung denervation.
Table 6.
Serious Adverse Events at 0–3 Months Occurring in ≥2.5% of Subjects in Either Group
| MedDRA System Organ Class and Lowest-Level Term | TLD |
Sham |
||
|---|---|---|---|---|
| Total Events | Subjects with Events | Total Events | Subjects with Events | |
| All events | 61 | 43/198 (21.7%) | 27 | 20/190 (10.5%) |
| Respiratory, thoracic, and mediastinal disorders | 31 | 28/198 (14.1%) | 14 | 13/190 (6.8%) |
| COPD exacerbation | 25 | 23/198 (11.6%) | 8 | 7/190 (3.7%) |
| Gastrointestinal disorders | 11 | 10/198 (5.1%) | 2 | 2/190 (1.1%) |
| Gastrointestinal motility disorder | 5 | 5/198 (2.5%) | 0 | 0/190 |
| Cardiac disorders | 8 | 8/198 (4.0%) | 2 | 2/190 (1.1%) |
| All-cause mortality | — | 1/198 (0.5%) | — | 1/190 (0.5%) |
| Respiratory related mortality | — | 1/198 (0.5%) | — | 1/190 (0.5%) |
For definition of abbreviations, see Table 5.
Periprocedural respiratory SAE durations were 27 ± 28 days in the TLD group and 35 ± 43 days in the sham group. The average duration of stay in the hospital was similar between groups: 5.7 ± 5.2 days (TLD) and 6.1 ± 6.5 days (sham). One patient from each group was treated with mechanical ventilation; the ventilated patients are the patients who died discussed later. The presence of COPD symptoms (cough, sputum production, wheeze, dyspnea, and chest tightness) during each event was captured and was similar between the TLD and sham groups (Table E9).
An analysis of baseline characteristics’ impact on the incidence of periprocedural respiratory SAEs was performed (Table E10). Each group exhibited a positive correlation between baseline history of prior-year severe COPD exacerbation and the incidence of periprocedural respiratory SAE (0.28 vs. 0.14). The TLD group exhibited a positive correlation between baseline levels of emphysema and periprocedural respiratory SAEs that was different from the sham group (0.19 vs. −0.09).
Ten patients in the TLD group (5.1%) experienced a GI SAE in the first 3 months after the procedure, versus two patients (1.1%) in the sham group (Table 6). GI events are a known complication of the TLD procedure. In 7 of the 10 patients, the GI SAE resolved during follow-up. Among the remaining three, two patients’ Gastroparesis Cardinal Symptom Index score had returned to normal before study exit. One patient had symptom improvement over time, but with increased Gastroparesis Cardinal Symptom Index score until study exit 22 months after the procedure.
In the TLD group, one patient experienced a bronchoesophageal fistula 16 days after the procedure (event rate of 0.5% for the full study), resulting in implementation of an esophageal cooling device. In 60 active treatment recipients, after implementation of esophageal cooling, no bronchial esophageal fistulas occurred. A reduced rate of related respiratory SAEs was observed in the cohort of active-treatment patients with esophageal cooling compared with active-treatment patients without esophageal cooling (Table E11).
Both treatment groups had one patient death within the first 3 months after the procedure. In the sham arm, the patient death was secondary to respiratory failure. In the TLD treatment group, the death occurred 23 days after the procedure and was secondary to the onset of prolonged bronchospasm.
Discussion
TLD with the D’Nerva lung denervation system is an emerging treatment for obstructive lung disease. It has been evaluated across four clinical trials, of which two were randomized, double-blind, sham-controlled studies involving more than 525 treated patients with as long as 3 years of follow-up (IPS-1 and 2 and AIRFLOW-1, -2, and -3) (19, 26, 29, 30). The present study (AIRFLOW-3) failed to meet the prespecified primary endpoint of reduction in time to the first moderate to severe exacerbation. However, prespecified secondary and other endpoints indicate a physiological improvement from TLD, which is consistent with previous trials.
The primary endpoint analysis in this trial was impacted by a higher rate of primary endpoint events in the TLD group during the first 3 months after the procedure. Previous clinical studies of bronchoscopic interventions have shown a higher rate of COPD exacerbations in the periprocedural period (31–33). In the AIRFLOW-2 trial, an increased rate of periprocedural COPD exacerbations was not observed in the treatment arm (19).
However, 24% of the AIRFLOW-2 population had a prior-year history of hospitalization, whereas 42% of AIRFLOW-3 patients had one or more prior-year COPD hospitalizations. A correlation between prior-year history of multiple respiratory hospitalizations and periprocedural respiratory SAEs was observed in both AIRFLOW-3 treatment groups. Additionally, there was a correlation between higher levels of baseline emphysema and periprocedural respiratory SAEs in the TLD group but not in the sham group.
The SAE profile for the TLD treatment group was similar to that for the sham group through the full 12 months of follow-up. There are trends for more GI disorders and COPD exacerbations that were driven by events in the first 3 months after the procedure. The periprocedural COPD exacerbations were discussed earlier. GI disorders are a known risk associated with thermal ablation in the mediastinal space as a result of the proximity to the branches of the vagal nerve that go on to innervate the GI tract (34). The occurrence of GI events after TLD has improved versus previous TLD trials as a result of improvements in the anatomical understanding of the location of these sensitive branches and procedural improvements (19, 30). Mortality rates were similar between treatment groups.
Preservation of lung function over time has been consistently observed in patients with COPD following TLD. In the COPD population enrolled in the present trial, a natural FEV1 decline of approximately 50 ml per year is expected (35) and was observed in the blinded sham group (P < 0.001). In contrast, FEV1 in the blinded TLD treatment group remained unchanged at 1 year (P = 0.281). Previous studies of TLD-treated patients have shown a similar pattern of FEV1 preservation over longer periods of follow-up (12, 25, 26, 29). Lung denervation is known to decrease airway smooth muscle tone, which reduces airflow limitation and decreases air trapping, thereby resulting in increases in FVC and FEV1 (20, 22, 28, 36).
Over time and with disease progression, RV is expected to increase in patients with COPD (37). In AIRFLOW-3, the sham group exhibited a significant increase in RV over 1 year of follow-up, whereas the TLD-treated group exhibited no change in RV, also consistent with observations across previous TLD trials (12, 19, 26).
TLD-treated patients had improved dyspnea at 12 months compared with sham-treated patients across both measures of dyspnea used in the trial (mMRC scale, TDI). Dyspnea improvement was also observed in the treatment groups of previous clinical TLD studies (11, 19, 26, 29). The higher percentage of patients in the treatment arm achieving the MCID in TDI (35% vs. 24%) and mMRC score (36% vs. 26%) underscores the potential beneficial impact of the intervention.
Lung denervation causes airway smooth muscle to dilate, leading to decreases in airflow resistance (20, 22, 28, 36). Fessler and Permutt demonstrated mathematically that the removal of airway tone reduces RV and improves FEV1 (38). These fundamental mechanics in combination with the recurring signal of RV reduction and FEV1 improvement observed across the entire AIRFLOW-3 data set led to the identification of patients whose hyperinflation is driven by airway-dominant pathology (persistent hyperinflation with less emphysema). Patients with this phenotype showed a response to TLD with a decrease in RV, an increase in FEV1, and fewer COPD exacerbations.
The AIRFLOW-3 clinical trial is one of the largest COPD device trials ever conducted. Although it failed to meet its primary endpoint, preservation of lung function and improved dyspnea was observed in the ITT analysis. These findings led to a post hoc analysis that identified a TLD treatment responder phenotype consistent with the mechanistic understanding of lung denervation in patients with COPD. A prospective clinical study is being designed to confirm the benefits of TLD treatment in patients with COPD with airflow obstruction and hyperinflation associated with an airway-predominant phenotype.
Supplemental Materials
Footnotes
Author Contributions: P.L.S. and G.J.C. had full access to all the data in the study, take responsibility for the integrity of the data and the accuracy of the data analysis, and had authority over manuscript preparations and the decision to submit the manuscript for publication. Study concept and design: D.-J.S., P.L.S., G.J.C., and F.C.S. Acquisition, analysis and interpretation of data: all authors. Drafting of manuscript: P.L.S. and G.J.C. Critical revision of the manuscript for important intellectual content: all authors. Study supervision, patient recruitment and follow-up: all authors. All authors contributed to the conduct of the study, patient recruitment, and revised the manuscript.
A data supplement for this article is available via the Supplements tab at the top of the online article.
Artificial Intelligence Disclaimer: No artificial intelligence tools were used in writing this manuscript.
Originally Published in Press as DOI: 10.1164/rccm.202502-0404OC on September 8, 2025
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1. Albert P, Agusti A, Edwards L, Tal-Singer R, Yates J, Bakke P. et al. Bronchodilator responsiveness as a phenotypic characteristic of established chronic obstructive pulmonary disease. Thorax . 2012;67:701–708. doi: 10.1136/thoraxjnl-2011-201458. [DOI] [PubMed] [Google Scholar]
- 2. Klooster K, ten Hacken NH, Hartman JE, Kerstjens HA, van Rikxoort EM, Slebos DJ. Endobronchial valves for emphysema without interlobar collateral ventilation. N Engl J Med . 2015;373:2325–2335. doi: 10.1056/NEJMoa1507807. [DOI] [PubMed] [Google Scholar]
- 3. Lipson DA, Barnhart F, Brealey N, Brooks J, Criner GJ, Day NC. et al. IMPACT Investigators. Once-daily single-inhaler triple versus dual therapy in patients with COPD. N Engl J Med . 2018;378:1671–1680. doi: 10.1056/NEJMoa1713901. [DOI] [PubMed] [Google Scholar]
- 4. Pavord ID, Chanez P, Criner GJ, Kerstjens HAM, Korn S, Lugogo N. et al. Mepolizumab for eosinophilic chronic obstructive pulmonary disease. N Engl J Med . 2017;377:1613–1629. doi: 10.1056/NEJMoa1708208. [DOI] [PubMed] [Google Scholar]
- 5. Chapman KR, Hurst JR, Frent SM, Larbig M, Fogel R, Guerin T. et al. Long-term triple therapy de-escalation to indacaterol/glycopyrronium in patients with chronic obstructive pulmonary disease (SUNSET): a randomized, double-blind, triple-dummy clinical trial. Am J Respir Crit Care Med . 2018;198:329–339. doi: 10.1164/rccm.201803-0405OC. [DOI] [PubMed] [Google Scholar]
- 6. Howarth PH, Springall DR, Redington AE, Djukanovic R, Holgate ST, Polak JM. Neuropeptide-containing nerves in endobronchial biopsies from asthmatic and nonasthmatic subjects. Am J Respir Cell Mol Biol . 1995;13:288–296. doi: 10.1165/ajrcmb.13.3.7654385. [DOI] [PubMed] [Google Scholar]
- 7. Sparrow MP, Weichselbaum M, McCray PB. Development of the innervation and airway smooth muscle in human fetal lung. Am J Respir Cell Mol Biol . 1999;20:550–560. doi: 10.1165/ajrcmb.20.4.3385. [DOI] [PubMed] [Google Scholar]
- 8. Spencer H, Leof D. The innervation of the human lung. J Anat . 1964;98:599–609. [PMC free article] [PubMed] [Google Scholar]
- 9. Canning BJ. Reflex regulation of airway smooth muscle tone. J Appl Physiol (1985) . 2006;101:971–985. doi: 10.1152/japplphysiol.00313.2006. [DOI] [PubMed] [Google Scholar]
- 10. Mayse ML, Norman HS, Peterson AD, Rouw KT, Johnson PJ. Targeted lung denervation in sheep: durability of denervation and long-term histologic effects on bronchial wall and peribronchial structures. Respir Res . 2020;21:117. doi: 10.1186/s12931-020-01383-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Valipour A, Mayse ML, Peterson AD, Johnson PJ, Rouw KT, Asadi S. et al. Respiratory sinus arrhythmia attenuation via targeted lung denervation in sheep and humans. Respiration . 2019;98:434–439. doi: 10.1159/000501623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Pison C, Shah PL, Slebos DJ, Ninane V, Janssens W, Perez T. et al. Safety of denervation following targeted lung denervation therapy for COPD: AIRFLOW-1 3-year outcomes. Respir Res . 2021;22:62. doi: 10.1186/s12931-021-01664-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Valipour A, Shah PL, Herth FJ, Pison C, Schumann C, Hubner RH. et al. AIRFLOW-2 Trial Study Group. Two-year outcomes for the double-blind, randomized, sham-controlled study of targeted lung denervation in patients with moderate to severe COPD: AIRFLOW-2. Int J Chron Obstruct Pulmon Dis . 2020;15:2807–2816. doi: 10.2147/COPD.S267409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Slebos DJ, Degano B, Valipour A, Shah PL, Deslee G, Sciurba FC. et al. AIRFLOW-3 Trial Study Group. Design for a multicenter, randomized, sham-controlled study to evaluate safety and efficacy after treatment with the Nuvaira(R) lung denervation system in subjects with chronic obstructive pulmonary disease (AIRFLOW-3) BMC Pulm Med . 2020;20:41. doi: 10.1186/s12890-020-1058-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Criner GJ, Slebos D-J, Sue RD, Bhatt SP, Ghattas C, Strange C. et al. Results from AIRFLOW-3: randomized, double blind, sham controlled study of targeted lung denervation in patients with COPD [abstract] Am J Respir Crit Care Med . 2025;211(suppl):A3112. doi: 10.1164/rccm.202502-0404OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Vogelmeier C, Hederer B, Glaab T, Schmidt H, Rutten-van Molken MP, Beeh KM. et al. POET-COPD Investigators. Tiotropium versus salmeterol for the prevention of exacerbations of COPD. N Engl J Med . 2011;364:1093–1103. doi: 10.1056/NEJMoa1008378. [DOI] [PubMed] [Google Scholar]
- 17. Singh D, Agusti A, Anzueto A, Barnes PJ, Bourbeau J, Celli BR. et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive lung disease: the GOLD science committee report 2019. Eur Respir J . 2019;53:1900164. doi: 10.1183/13993003.00164-2019. [DOI] [PubMed] [Google Scholar]
- 18. Revicki DA, Rentz AM, Dubois D, Kahrilas P, Stanghellini V, Talley NJ. et al. Gastroparesis Cardinal Symptom Index (GCSI): development and validation of a patient reported assessment of severity of gastroparesis symptoms. Qual Life Res . 2004;13:833–844. doi: 10.1023/B:QURE.0000021689.86296.e4. [DOI] [PubMed] [Google Scholar]
- 19. Slebos DJ, Shah PL, Herth FJF, Pison C, Schumann C, Hubner RH. et al. AIRFLOW-2 Study Group. Safety and adverse events after targeted lung denervation for symptomatic moderate to severe chronic obstructive pulmonary disease (AIRFLOW). A multicenter randomized controlled clinical trial. Am J Respir Crit Care Med . 2019;200:1477–1486. doi: 10.1164/rccm.201903-0624OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Colebatch HJ, Halmagyi DF. Effect of vagotomy and vagal stimulation on lung mechanics and circulation. J Appl Physiol . 1963;18:881–887. doi: 10.1152/jappl.1963.18.5.881. [DOI] [PubMed] [Google Scholar]
- 21. Cuthbert AW, Murthy M, Darlington APS. Neural control of submucosal gland and apical membrane secretions in airways. Physiol Rep . 2015;3:e12398. doi: 10.14814/phy2.12398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Hummel JP, Mayse ML, Dimmer S, Johnson PJ. Physiologic and histopathologic effects of targeted lung denervation in an animal model. J Appl Physiol (1985) . 2019;126:67–76. doi: 10.1152/japplphysiol.00565.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Molfino NA, Slutsky AS, Julià-Serdà G, Hoffstein V, Szalai JP, Chapman KR. et al. Assessment of airway tone in asthma: comparison between double lung transplant patients and healthy subjects. Am Rev Respir Dis . 1993;148:1238–1243. doi: 10.1164/ajrccm/148.5.1238. [DOI] [PubMed] [Google Scholar]
- 24. Rienhoff WF, Jr, Gay LN. Treatment of intractable bronchial asthma by bilateral resection of the posterior pulmonary plexus. Arch Surg . 1938;37:456–469. [Google Scholar]
- 25. Hartman JE, Conway F, Degano B, Augustijn SWS, Herth FJF, Mayr A. et al. Rate of lung function decline slows in the 3 years after targeted lung denervation in COPD. Respir Med . 2021;188:106604. doi: 10.1016/j.rmed.2021.106604. [DOI] [PubMed] [Google Scholar]
- 26. Slebos DJ, Klooster K, Koegelenberg CF, Theron J, Styen D, Valipour A. et al. Targeted lung denervation for moderate to severe COPD: a pilot study. Thorax . 2015;70:411–419. doi: 10.1136/thoraxjnl-2014-206146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Bradley GW, Hale T, Pimble J, Rowlandson R, Noble MI. Effect of vagotomy on the breathing pattern and exercise ability in emphysematous patients. Clin Sci (Lond) . 1982;62:311–319. doi: 10.1042/cs0620311. [DOI] [PubMed] [Google Scholar]
- 28. Karczewski W, Widdicombe JG. The role of the vagus nerves in the respiratory and circulatory responses to intravenous histamine and phenyl diguanide in rabbits. J Physiol . 1969;201:271–291. doi: 10.1113/jphysiol.1969.sp008755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Valipour A, Asadi S, Pison C, Jondot M, Kessler R, Benneddif K. et al. Long-term safety of bilateral targeted lung denervation in patients with COPD. Int J Chron Obstruct Pulmon Dis . 2018;13:2163–2172. doi: 10.2147/COPD.S158748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Valipour A, Shah PL, Pison C, Ninane V, Janssens W, Perez T. et al. AIRFLOW-1 Study Group. Safety and dose study of targeted lung denervation in moderate/severe COPD patients. Respiration . 2019;98:329–339. doi: 10.1159/000500463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Criner GJ, Sue R, Wright S, Dransfield M, Rivas-Perez H, Wiese T. et al. LIBERATE Study Group. A multicenter randomized controlled trial of Zephyr endobronchial valve treatment in heterogeneous emphysema (LIBERATE) Am J Respir Crit Care Med . 2018;198:1151–1164. doi: 10.1164/rccm.201803-0590OC. [DOI] [PubMed] [Google Scholar]
- 32. Herth FJ, Valipour A, Shah PL, Eberhardt R, Grah C, Egan J. et al. Segmental volume reduction using thermal vapour ablation in patients with severe emphysema: 6-month results of the multicentre, parallel-group, open-label, randomised controlled STEP-UP trial. Lancet Respir Med . 2016;4:185–193. doi: 10.1016/S2213-2600(16)00045-X. [DOI] [PubMed] [Google Scholar]
- 33. Sciurba FC, Criner GJ, Strange C, Shah PL, Michaud G, Connolly TA. et al. RENEW Study Research Group. Effect of endobronchial coils vs usual care on exercise tolerance in patients with severe emphysema: the RENEW randomized clinical trial. JAMA . 2016;315:2178–2189. doi: 10.1001/jama.2016.6261. [DOI] [PubMed] [Google Scholar]
- 34. Lakkireddy D, Reddy YM, Atkins D, Rajasingh J, Kanmanthareddy A, Olyaee M. et al. Effect of atrial fibrillation ablation on gastric motility: the atrial fibrillation gut study. Circ Arrhythm Electrophysiol . 2015;8:531–536. doi: 10.1161/CIRCEP.114.002508. [DOI] [PubMed] [Google Scholar]
- 35. Tantucci C, Modina D. Lung function decline in COPD. Int J Chron Obstruct Pulmon Dis . 2012;7:95–99. doi: 10.2147/COPD.S27480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Olsen CR, Colebatch HJH, Mebel PE, Nadel JA, Staub NC. Motor control of pulmonary airways studied by nerve stimulation. J Appl Physiol . 1965;20:202–208. [Google Scholar]
- 37. Park J, Lee CH, Lee YJ, Park JS, Cho YJ, Lee JH. et al. Longitudinal changes in lung hyperinflation in COPD. Int J Chron Obstruct Pulmon Dis . 2017;12:501–508. doi: 10.2147/COPD.S122909. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Fessler HE, Permutt S. Lung volume reduction surgery and airflow limitation. Am J Respir Crit Care Med . 1998;157:715–722. doi: 10.1164/ajrccm.157.3.9608004. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.