Comments
Abstract
Rationale: The BLOCK COPD (β-Blockers for the Prevention of Acute Exacerbations of Chronic Obstructive Pulmonary Disease) study found that metoprolol was associated with a higher risk of severe exacerbation.
Objectives: To determine the mechanism underlying these results, we compared changes in lung function over the course of the study between treatment groups and evaluated whether baseline bronchodilator response or early reduction in forced expiratory volume in 1 second (FEV1) or forced vital capacity (FVC) was associated with exacerbation risk.
Methods: We compared changes in lung function (FEV1 and FVC) over the treatment period between treatment groups using linear mixed-effect models. Cox proportional hazards models were used to evaluate the association between baseline bronchodilator responsiveness (FEV1, FVC, and combined FEV1 and FVC), early post-randomization (14 d) change in lung function, and the interaction between treatment assignment and these measures with risk of any or severe or very severe exacerbations. Negative binomial models were used to evaluate the relationship between bronchodilator responsiveness, the interaction between bronchodilator responsiveness and treatment assignment, and exacerbation rate.
Results: Over the 336-day treatment period, individuals in the metoprolol group had a significantly greater decrease in logarithmic FEV1 from baseline to visit on Day 28 than individuals in the placebo group. Individuals in the metoprolol group had a significantly greater decrease in FVC from baseline to visits on Days 14 and 28, and also a significantly greater decrease in logarithmic FVC from baseline to visits on Days 42 and 112 than individuals in the placebo group. There were no associations between early lung function reduction or interactions between lung function reduction and treatment assignment and time to any or severe or very severe exacerbations. There were no interactions between treatment arm and baseline bronchodilator responsiveness measures on risk or rate of exacerbations. However, those with baseline FVC bronchodilator responsiveness had a higher rate of severe or very severe exacerbations (adjusted rate ratio, 1.62; 95% confidence interval, 1.04–2.48).
Conclusions: Metoprolol was associated with reduced lung function during the early part of the treatment period, but these effects were modest and did not persist. Early lung function reduction and baseline bronchodilator responsiveness did not interact with the treatment arm to predict exacerbations; however, baseline FVC bronchodilator responsiveness was associated with a 60% higher rate of severe or very severe exacerbations.
Clinical trial registered with www.clinicaltrials.gov (NCT 02587351).
Despite observational studies suggesting that β-blockers may have a beneficial effect in preventing exacerbations, the randomized BLOCK COPD (β-Blockers for the Prevention of Acute Exacerbations of Chronic Obstructive Pulmonary Disease) study found that metoprolol was not associated with any benefit and, in fact, increased the risk of severe exacerbations requiring hospitalization (1–3). The trial was stopped early because of futility for the primary outcome (any exacerbation) and safety concerns. There is no definitive explanation for the higher risk of severe exacerbation seen in the metoprolol group; however, there have been long-standing concerns about the potential adverse respiratory effects of β-blockers. Although cardioselective β-blockers, including metoprolol, have a much greater affinity for β-1 receptors, there is some affinity for β-2 receptors, which may lead to lower lung function or may differentially affect those with bronchodilator responsiveness (BDR) (4, 5).
A meta-analysis of studies reporting the effects of cardioselective β-blockers on lung function over periods ranging up to 16 weeks found no significant effects on forced expiratory volume in 1 second (FEV1) (6); however, many of these studies were small and did not enroll patients with chronic obstructive pulmonary disease (COPD) at high risk for exacerbations. The BLOCK COPD study examined changes in FEV1 percent predicted from baseline during the 336-day treatment period and also found no difference between metoprolol and placebo, although metoprolol was associated with a worsening in COPD Assessment Test score and greater dyspnea (3). Differences in other spirometric measures between treatment groups were not examined, nor were the associations between early changes in lung function after randomization and the risk of subsequent exacerbations.
Although there is consensus that BDR in patients with COPD is common, there are conflicting data on whether its presence can predict exacerbations. Most studies have found that BDR in FEV1 (flow BDR) and/or BDR in forced vital capacity (FVC) (volume BDR) do not predict exacerbations after controlling for baseline post-bronchodilator FEV1 (7–9). However, recent COPDGene (Genetic Epidemiology of COPD) studies showed that BDR in FEV1 and FVC are differentially associated with clinical and radiographic features (10, 11).
In this study, we aimed to 1) determine if lung function changes during receipt of metoprolol therapy were different from lung function changes in individuals receiving placebo, 2) determine if early lung function reduction at 14 days after randomization was associated with subsequent exacerbation risk, and 3) determine if BDR at baseline was associated with exacerbation risk. We hypothesized that metoprolol therapy would lead to a reduction in lung function compared with placebo and that early reduction in lung function resulting from metoprolol would be associated with higher exacerbation risk. Because of the potential for metoprolol to exhibit β-2 blockade, we also hypothesized that those with baseline BDR who were treated with metoprolol would have a higher risk of exacerbation.
This is a post hoc analysis of the longitudinal BLOCK COPD trial, which was a prospective, placebo-controlled, double-blind, randomized trial of 532 participants aged 40–85 years across 26 centers, with post-bronchodilator spirometry–confirmed COPD (defined as FEV1/FVC < 0.70), who received either extended-release metoprolol (n = 268) or placebo (n = 264). In the trial, the treatment group received metoprolol at a starting dose of 50 mg/d. On the basis of vital signs, spirometry, and side effects, metoprolol or matching placebo was titrated to a final dose of 25 mg, 50 mg, or 100 mg daily. This adjusted dose was continued until study completion. The main outcome of interest was time to first exacerbation during the treatment period. All participants had either a history of exacerbation in the prior year or were prescribed supplemental oxygen for use at home. Patients who had a proven indication for a β-blocker use, including heart failure with an ejection fraction less than 40% or a history of myocardial infarction or revascularization within the prior 36 months, were excluded.
Pre- and post-bronchodilator spirometry was performed only at the screening visit; post-bronchodilator spirometry alone was performed at visits on Day 14, Day 28, Day 42, Day 112, and Day 336. For the analysis evaluating whether FEV1 and FVC decline in the first 14 days were predictive of exacerbations, we excluded those who had early (<14 d) exacerbations, who ended the study before the Day 14 visit, or who did not have Day 14 spirometry measures (n = 239 in metoprolol group and n = 244 in placebo group were included in the analysis). For the analysis evaluating the relationship between BDR and risk of exacerbation, participants were included if they had pre- and post-bronchodilator spirometry measured at baseline (n = 264 in metoprolol group and n = 261 in placebo group). Post-bronchodilator spirometry was performed 15–30 minutes after two puffs of albuterol administration via inhaler. FEV1 BDR was defined as an increase in FEV1 of ⩾12% and ⩾200 ml after bronchodilator administration. FVC BDR was defined as an increase in FVC of ⩾12% and ⩾200 ml after bronchodilator administration. Combined BDR was defined as an increase in both FEV1 and FVC of ⩾12% and ⩾200 ml after bronchodilator administration (10, 12).
The primary outcomes included 1) change in absolute FEV1 and FVC over the treatment period (336 d), 2) time to exacerbation (any exacerbation and severe or very severe exacerbations), and 3) rates of exacerbation (any exacerbation and severe or very severe exacerbation). Any exacerbation included mild, moderate, severe, and very severe exacerbations. An exacerbation was defined as an increase or a new onset of two of the following: shortness of breath, cough, sputum production, wheezing, or chest tightness treated with antibiotics and/or systemic steroids for at least 3 days. A mild exacerbation was an exacerbation that was treated at home with or without contacting healthcare personnel. A moderate exacerbation involved a visit to the emergency department, whereas a severe exacerbation involved a hospitalization. A very severe exacerbation involved treatment with invasive mechanical ventilation. Rates of exacerbation were defined as events per person-year.
Linear mixed-effect models with patient-specific random intercepts were used to compare changes in lung function (absolute FEV1 and FVC) from baseline to subsequent visits between treatment groups. Lung function measurements were analyzed using both the original measurements and their logarithms. The choice of log transformation was based on optimal Box-Cox λ values of 0 for FEV1 and 0.1 for FVC (13). These two analytic approaches (original vs. log-transformed measures) provide complementary but distinct comparisons. Analyses using the log-transformed measures also test whether changes in lung function differed proportionately (relative to baseline values) between treatment groups. Close-out measurements from participants who ended the study early were analyzed as if they were measured at the next scheduled study visit.
Unadjusted and adjusted Cox proportional hazards models were used to evaluate whether baseline BDR (FEV1, FVC, and combined FEV1 and FVC) or change in absolute FEV1 or FVC during the first 14 days of treatment were associated with subsequent risk of any or severe or very severe exacerbation. To evaluate if those with BDR were adversely impacted by metoprolol or if metoprolol had a differential effect based on early changes in lung function, we tested the interaction between treatment assignment and the primary predictor (BDR measures or 14-d change in FEV1 or FVC) in adjusted Cox proportional hazards models. The interaction term was removed from the model if the association was not significant (P > 0.05). Unadjusted and adjusted negative binomial models were used to evaluate the association between baseline BDR and exacerbation rate for any and severe or very severe exacerbations.
Adjusted models controlled for age, sex, race (Black vs. not Black), smoking status, body mass index, long-acting β-agonist use, post-bronchodilator baseline FEV1 percent predicted, and treatment assignment. In addition, adjusted Cox proportional hazards models were stratified by study center. P values were not adjusted for multiple comparisons.
The distribution of FEV1 and FVC at each visit is displayed in Figure E1 in the data supplement. We did not detect an association between treatment assignment and change in absolute FEV1 at any time point during the 336-day follow-up period (Figure 1A). However, when FEV1 was logarithmically transformed, metoprolol was associated with a 0.039 (95% confidence interval [CI], 0.069–0.009) greater decrease in log FEV1 from baseline to the visit on Day 28 than in the placebo group (Figure 1B). We did not detect a significant difference in change in log FEV1 from baseline to the remaining time points (visits on Days 14, 42, 112, or 336) between the treatment groups.
Figure 1. (A) Changes in forced expiratory volume in 1 second (FEV1) (in milliliters) with 95% confidence intervals, by assignment to metoprolol or placebo. (B) Changes in logarithmically transformed FEV1 (in milliliters with 95% confidence intervals, by assignment to metoprolol or placebo. BL = baseline.
[More] [Minimize]When comparing changes in absolute FVC over the follow-up period, we found that individuals in the metoprolol group experienced a significant decrease in absolute FVC from baseline to the visit on Day 14 and the visit on Day 28 compared with the placebo group (Figure 2A). After logarithmic transformation of FVC, we also detected differences between treatment arms in change from baseline to the visit on Day 42 and the visit on Day 112 (Figure 2B). We did not detect differences in the change in absolute FVC or log-transformed FVC from baseline to the visit on Day 336 between treatment groups.
Figure 2. (A) Changes in forced vital capacity (FVC) (in milliliters) with 95% confidence intervals, by assignment to metoprolol or placebo. (B) Changes in logarithmically transformed FVC (in milliliters) with 95% confidence intervals, by assignment to metoprolol or placebo. BL = baseline.
[More] [Minimize]We did not detect interactions between treatment assignment and 14-day change in absolute FEV1 or 14-day change in absolute FVC with time to any or severe or very severe subsequent exacerbation (P > 0.4) (Table E1). We also did not detect associations between 14-day change in absolute FEV1 or FVC and time to any or severe or very severe subsequent exacerbation in unadjusted models or in models accounting for treatment assignment and other potentially confounding variables (Table 1).
| 14-d Change in FEV1 | 14-d Change in FVC | |||||||
|---|---|---|---|---|---|---|---|---|
| Unadjusted HR* (95% CI) | P Value | Adjusted HR* (95% CI) | P Value | Unadjusted HR* (95% CI) | P Value | Adjusted HR* (95% CI) | P Value | |
| Any exacerbation | 1.079 (0.54–2.159) | 0.8 | 0.92 (0.408–2.073) | 0.8 | 0.861 (0.606–1.225) | 0.4 | 0.925 (0.633–1.352) | 0.7 |
| Severe exacerbation | 1.046 (0.304–3.598) | 0.9 | 0.802 (0.186–3.452) | 0.8 | 0.704 (0.394–1.257) | 0.2 | 0.854 (0.463–1.576) | 0.6 |
We did not detect an interaction between treatment assignment and any BDR measure (FEV1, FVC, or combined) on time to any or severe or very severe exacerbation (P > 0.1) (Table E2). In unadjusted analysis, individuals who had an FEV1 BDR had a lower risk of any exacerbation (hazard ratio [HR], 0.66; 95% CI, 0.45–0.97; P = 0.04) (Table 2). However, this relationship was no longer significant after adjusting for covariates (HR, 0.85; 95% CI, 0.56–1.29). We did not detect an association between FVC BDR or combined BDR and time to any exacerbation in either unadjusted or adjusted analyses (P ⩾ 0.2).
| BDR | Number with BDR | Any Exacerbation | Severe Exacerbation | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Unadjusted HR* (95% CI) | P Value | Adjusted HR* (95% CI) | P Value | Unadjusted HR* (95% CI) | P Value | Adjusted HR* (95% CI) | P Value | ||
| FEV1 12% and 200 ml | 63 | 0.66 (0.45–0.97) | 0.04 | 0.85 (0.56–1.29) | 0.4 | 0.71 (0.37–1.37) | 0.3 | 0.91 (0.44–1.87) | 0.8 |
| FVC 12% and 200 ml | 111 | 1.01 (0.77–1.34) | 0.9 | 1.07 (0.80–1.43) | 0.7 | 1.58 (1.04–2.39) | 0.03 | 1.54 (0.99–2.40) | 0.06 |
| Both FVC and FEV1 12% and 200 ml | 40 | 0.75 (0.47–1.18) | 0.2 | 0.92 (0.57–1.48) | 0.7 | 1.02 (0.51–2.01) | ⩾0.9 | 1.20 (0.56–2.61) | 0.6 |
In unadjusted analysis, we found that individuals who had an FVC BDR had a higher risk of severe or very severe exacerbation (HR, 1.58; 95% CI, 1.04–2.39; P = 0.03), which was only slightly attenuated after adjusting for other covariates (HR, 1.54; 95% CI, 0.99–2.40; P = 0.06) (Table 2). We did not detect an association between FEV1 BDR or combined BDR and time to severe or very severe exacerbation (P ⩾ 0.3).
We did not detect an interaction between treatment assignment and any BDR measure (FEV1, FVC, or combined) on rate of any or severe or very severe exacerbation (P > 0.3) (Table E3). FEV1 BDR and combined BDR were both associated with lower rate of any exacerbation in unadjusted analyses (rate ratio [RR], 0.61; 95% CI, 0.43–0.87; P = 0.007 for individuals with FEV1 BDR compared with individuals not meeting the FEV1 BDR threshold; RR, 0.61; 95% CI, 0.39–0.92; P = 0.02 for individuals with combined BDR compared with individuals not meeting combined BDR thresholds). However, the association was not significant after adjusting for covariates (P ⩾ 0.1) (Table 3). We did not find an association between FVC BDR and rate of any exacerbation in unadjusted or adjusted analyses (P ⩾ 0.5).
| BDR | Number with BDR | Any Exacerbation | Severe Exacerbation | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Unadjusted RR (95% CI) | P Value | Adjusted RR (95% CI) | P Value | Unadjusted RR (95% CI) | P Value | Adjusted RR (95% CI) | P Value | ||
| FEV1 12% and 200 ml | 63 | 0.61 (0.43–0.87) | 0.007 | 0.76 (0.52–1.10) | 0.1 | 0.51 (0.24–1.02) | 0.07 | 0.85 (0.38–1.76) | 0.7 |
| FVC 12% and 200 ml | 111 | 1.02 (0.80–1.31) | 0.8 | 1.09 (0.85–1.39) | 0.5 | 1.58 (1.01–2.45) | 0.04 | 1.62 (1.04–2.48) | 0.03 |
| Both FVC and FEV1 12% and 200 ml | 40 | 0.61 (0.39–0.92) | 0.02 | 0.76 (0.48–1.17) | 0.2 | 0.70 (0.30–1.50) | 0.4 | 1.13 (0.48–2.47) | 0.8 |
We did not find a strong association between either FEV1 BDR or combined BDR and rate of severe or very severe exacerbation in unadjusted or adjusted analyses (P ⩾ 0.07). FVC BDR was associated with a higher rate of severe or very severe exacerbation in unadjusted (RR, 1.58; 95% CI, 1.01–2.45; P = 0.04 for individuals with FVC BDR compared with individuals not meeting the FVC BDR threshold) and adjusted analyses (RR, 1.62; 95% CI, 1.04–2.48; P = 0.03) (Table 3).
Although the original BLOCK trial did not find that metoprolol use impacted FEV1 percent predicted values during the study duration, the present analysis suggests that metoprolol use was associated with a statistically significant worsening of other spirometric measures at several time points. When examining the effect of early lung function reduction (FEV1 or FVC) or baseline BDR (FEV1, FVC, or combined), we did not find a difference in the relationship between these measures and risk of exacerbation for individuals assigned to metoprolol versus those assigned to placebo. However, in analyses adjusted for treatment assignment and other covariates, FVC BDR was associated with a higher rate of severe or very severe exacerbations.
The BLOCK COPD study allowed us to examine changes in FEV1 and FVC during a period up to 336 days in people with COPD randomized to metoprolol versus placebo, which has not been reported previously. We evaluated both the absolute change in the measurement and the proportional changes relative to baseline (using a log-transformed value) to get a comprehensive evaluation of changes over time. We did not detect a difference between treatment assignment and either absolute or relative change in FEV1, except for one time point (baseline to the Day 28 visit) when FEV1 was logarithmically transformed. This one–time point finding would likely not be statistically significant after accounting for the data-driven transformation approach and multiple testing. However, we did detect significant changes in FVC, both absolute and relative to baseline, based on treatment assignment at early time points in the study (specifically the Day 14 and Day 28 visits). Although these mean differences are less than the minimal clinically important difference of 100 ml (14), they suggest that these adverse changes in lung function could play a role in worsening symptoms. The decrease in FVC may be due to increased air trapping, which could explain the increase in symptoms, including shortness of breath and quality of life as measured by the COPD Assessment Test, that was seen in the metoprolol group (3).
Our study did not find an association between change in lung function, either FEV1 or FVC, in the first 14 days and time to any or severe or very severe exacerbation. We also did not detect interactions with treatment. With the caveat that those who experienced exacerbations before 14 days were excluded and the study was underpowered because of the trial’s early termination, our results suggest that more severe exacerbations seen in the metoprolol group may not be attributed to early lung function reduction. Some have suggested monitoring lung function after the initiation of β-blockers (15); however, the results of the present study call into question the utility of this approach as a safety tool to determine whether β-blockers should be continued.
Our study did not find an association between FEV1 BDR or combined BDR and either time to exacerbation or rate of exacerbation after adjusting for covariates. However, FVC BDR was significantly associated with a higher rate of severe or very severe exacerbation (approximately 60% higher rate than among those without FVC BDR) after controlling for demographic and clinical covariates, including baseline FEV1 percent predicted. The lack of association with FVC BDR and any exacerbation may be owing to the shortcomings of the current definition of an exacerbation and classification of its severity, which relies solely on patient-reported symptoms and on location of treatment (16). In severe exacerbations, which are managed in a hospital, diagnoses are made on the basis of cumulative assessment of vital signs, laboratory work, imaging, and physical examination, thus increasing the probability that other diagnoses are reliably ruled out. In contrast, mild exacerbations that are treated at home or over the phone may be subject to a higher degree of heterogeneity. Our findings contribute to prior literature that evaluated associations between FVC BDR and exacerbations after controlling for FEV1 (7, 9). One such study by Janson and colleagues (9) that did not find an association between FVC BDR and exacerbations included a composite analysis of three cohorts from large population-based studies. These studies did not focus solely on high-risk COPD populations as in the BLOCK trial and used an alternative definition for FVC BDR, which may explain the difference in results. A study by Barjaktarevic and colleagues (7) used a similar 12% and 200-ml increase in FVC to define BDR, and in their univariate analyses of isolated FVC BDR, they did note an increased incidence rate ratio for exacerbations and reduced 3-year survival in FVC BDR responders versus nonresponders. However, the exacerbation and mortality relationships diminished in their multivariate analysis, which included baseline post-bronchodilator FEV1 percent predicted. Notably, we did not detect an interaction between BDR and the treatment assignment on exacerbation risk or rate, which argues against the notion that metoprolol would have a differential clinical impact on those with baseline FVC BDR. However, FVC BDR, which is commonly associated with severe COPD, may be a useful marker to determine an individual’s risk of severe exacerbation because a change in volume and degree of hyperinflation can impact the degree of dyspnea (17–20). In contrast to the lack of association seen with early lung function reduction and exacerbation risk, we suspect that the inherent airway reactivity in those with baseline BDR may be responsible for the association with severe exacerbation risk.
The results of this study should be interpreted with the consideration of its limitations. Given the timing of spirometry measurements in the BLOCK COPD study, we were unable to evaluate the impact of early lung function reduction on exacerbations occurring before 14 days after randomization. Our study sought to explain the increased harm seen in the metoprolol group during the BLOCK COPD trial; however, the early termination of the trial minimized its power to examine risk factors associated with increased severe exacerbations (a secondary outcome). In addition, given the exploratory nature of this post hoc analysis, some of the identified trends could be spurious findings due to type I error. Finally, to define BDR, we used the American Thoracic Society criteria of an increase of ⩾12% and ⩾200 ml; however, we did not assess other definitions of BDR (11, 21) and did not ask participants to withhold bronchodilator medications before the baseline visit.
Our study did not find an explanation for a higher risk of severe exacerbations in the metoprolol group seen in the BLOCK COPD trial. Although there were some changes in FEV1 and FVC in the metoprolol group early in the study, neither early lung function decline nor BDR interacted with treatment assignment in predicting either time to or rate of exacerbation. Some of the spirometric changes seen in the metoprolol group may explain the increased symptoms seen in the original trial. Our results suggest that FVC BDR may impact the rate of exacerbations, independent of metoprolol use, and may have a role in determining clinical phenotypes.
| 1. | Bhatt SP, Wells JM, Kinney GL, Washko GR Jr, Budoff M, Kim Y-I, et al.; COPDGene Investigators. β-Blockers are associated with a reduction in COPD exacerbations. Thorax 2016;71:8–14. |
| 2. | Du Q, Sun Y, Ding N, Lu L, Chen Y. Beta-blockers reduced the risk of mortality and exacerbation in patients with COPD: a meta-analysis of observational studies. PLoS One 2014;9:e113048. |
| 3. | Dransfield MT, Voelker H, Bhatt SP, Brenner K, Casaburi R, Come CE, et al.; BLOCK COPD Trial Group. Metoprolol for the prevention of acute exacerbations of COPD. N Engl J Med 2019;381:2304–2314. |
| 4. | Wellstein A, Palm D, Belz GG, Butzer R, Polsak R, Pett B. Reduction of exercise tachycardia in man after propranolol, atenolol and bisoprolol in comparison to beta-adrenoceptor occupancy. Eur Heart J 1987;8:3–8. |
| 5. | Baker JG. The selectivity of β-adrenoceptor antagonists at the human β1, β2 and β3 adrenoceptors. Br J Pharmacol 2005;144:317–322. |
| 6. | Salpeter S, Ormiston T, Salpeter E. Cardioselective beta-blockers for chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2005;2005:CD003566. |
| 7. | Barjaktarevic IZ, Buhr RG, Wang X, Hu S, Couper D, Anderson W, et al.; NHLBI SubPopulations and InteRmediate Outcome Measures In COPD Study (SPIROMICS). Clinical significance of bronchodilator responsiveness evaluated by forced vital capacity in COPD: SPIROMICS cohort analysis. Int J Chron Obstruct Pulmon Dis 2019;14:2927–2938. |
| 8. | 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. |
| 9. | Janson C, Malinovschi A, Amaral AFS, Accordini S, Bousquet J, Buist AS, et al. Bronchodilator reversibility in asthma and COPD: findings from three large population studies. Eur Respir J 2019;54:1900561. |
| 10. | Fortis S, Comellas A, Make BJ, Hersh CP, Bodduluri S, Georgopoulos D, et al. Combined forced expiratory volume in 1 second and forced vital capacity bronchodilator response, exacerbations, and mortality in chronic obstructive pulmonary disease. Ann Am Thorac Soc 2019;16:826–835. |
| 11. | Hansen JE, Dilektasli AG, Porszasz J, Stringer WW, Pak Y, Rossiter HB, et al. A new bronchodilator response grading strategy identifies distinct patient populations. Ann Am Thorac Soc 2019;16:1504–1517. |
| 12. | Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A, et al.; ATS/ERS Task Force. Standardisation of spirometry. Eur Respir J 2005;26:319–338. |
| 13. | Box GEP, Cox DR. An analysis of transformations. J R Stat Soc B 1964;26:211–252. |
| 14. | Donohue JF. Minimal clinically important differences in COPD lung function. COPD 2005;2:111–124. |
| 15. | Lipworth B, Wedzicha J, Devereux G, Vestbo J, Dransfield MT. Beta-blockers in COPD: time for reappraisal. Eur Respir J 2016;48:880–888. |
| 16. | Celli BR, Fabbri LM, Aaron SD, Agusti A, Brook R, Criner GJ, et al. An updated definition and severity classification of chronic obstructive pulmonary disease exacerbations: the Rome proposal. Am J Respir Crit Care Med 2021;204:1251–1258. |
| 17. | Ouaalaya EH, Falque L, Dupis JM, Sabatini M, Bernady A, Nguyen L, et al. The determinants of dyspnoea evaluated by the mMRC scale: the French Palomb cohort. Respir Med Res 2021;79:100803. |
| 18. | Alter P, Orszag J, Kellerer C, Kahnert K, Speicher T, Watz H, et al. Prediction of air trapping or pulmonary hyperinflation by forced spirometry in COPD patients: results from COSYCONET. ERJ Open Res 2020;6:00092-2020. |
| 19. | Tan WC, Bourbeau J, Hernandez P, Chapman KR, Cowie R, FitzGerald JM, et al.; CanCOLD Collaborative Research Group. Bronchodilator responsiveness and reported respiratory symptoms in an adult population. PLoS One 2013;8:e58932. |
| 20. | Quanjer PH, Ruppel GL, Langhammer A, Krishna A, Mertens F, Johannessen A, et al. Bronchodilator response in FVC is larger and more relevant than in FEV1 in severe airflow obstruction. Chest 2017;151:1088–1098. |
| 21. | American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis 1991;144:1202–1218. |
Supported by National Heart, Lung, and Blood Institute grant K23HL153672.
Author Contributions: T.M.P., E.S.H., J.C., H.V., S.X.L., S.C.L., S.P.B., D.M.M., T.M., K.M.K., S.F., D.K., and M.T.D. provided substantial contributions to the conception or design of the work or the acquisition, analysis, or interpretation of data for the work. T.M.P. was responsible for drafting of the work, and all authors contributed to revising it critically for important intellectual content. All authors provided final approval of the version to be published and agreed to be accountable for all aspects of the work, including ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
This article has a data supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
Author disclosures are available with the text of this article at www.atsjournals.org.
