Key points

Introduction

Chronic obstructive pulmonary disease (COPD) is characterized by chronic respiratory symptoms, such as breathlessness and chronic productive cough, limited exercise tolerance, poor quality of life, and acute exacerbations. COPD is characterized by an abnormal, chronic inflammatory response in the airways to the inhalation of tobacco smoke or other noxious particles. This results in the development of fixed obstruction of the small airways, due to airway wall thickening and fibrosis. Other pathologic features include mucus hypersecretion and emphysema, due to destruction of the lung parenchyma.

Bronchodilators hold the first line and second line of maintenance pharmacologic treatment of COPD for most patients, conferring significant health benefits and limited adverse events. There are 2 classes of long-acting bronchodilators (LABDs), with different pharmacologic targets. Long-acting muscarinic antagonists (LAMAs) inhibit the action of acetylcholine at muscarinic receptors, whereas long-acting beta-2 adrenoreceptor agonists (LABAs) stimulate beta-2 adrenergic receptors, enhancing cyclic adenosine monophosphate pathway signaling. Both LABAs and LAMAs mediate airway smooth muscle relaxation, exerting potent, prolonged bronchodilator effects, that are to some extent additive. Other beneficial effects of LABAs, mediated through their action on beta-adrenoreceptors in various cell types, include an enhanced mucociliary clearance and attenuation of the recruitment of neutrophils, of the release of mast cell mediators, and of the inhibition of smooth muscle cells proliferation. Anti-inflammatory effects also are attributed to LAMAs that appear to inhibit airway neutrophilia, eosinophilia, MUC5AC expression, and fibrosis. ^, These anti-inflammatory effects, however, are supported only by in vitro studies and have not been proved in clinical trials in COPD.

This review (1) summarizes data on the safety and clinical efficacy/effectiveness of LABDs from systematic reviews, randomized controlled trials (RCTs), and real-life studies, and (2) describes their role in COPD management, as defined by current national and international clinical guidelines. Although narrative, this review is based on thorough searches of PubMed, the Cochrane Library, and Google Scholar for relevant RCTs, real-life studies, systematic reviews, and clinical practice guidelines.

Safety and clinical effectiveness of long-acting bronchodilators

Data on the safety and clinical effectiveness of LABAs, LAMAs, and dual combinations of a LABA and a LAMA are summarized in the following paragraphs and in Fig. 1 . In preparation of this review, patient-important effectiveness outcomes, including mortality, frequency of exacerbations, quality of life, symptoms severity, and exercise tolerance, have been prioituized. To a lesser extent, forced expiratory volume in the first second of expiration (FEV ₁ ) has been included, which is a favored surrogate endpoint for regulators. FEV ₁ , a traditional lung function parameter that correlates with disease progression and mortality, is related only weakly to patient-reported outcomes. Where good systematic reviews are used, information has been added on individual large trials with particular impact.

Summary of the impact of bronchodilators on patient-important outcomes.

( Data from Refs. ^{, , ,} )

Minimal clinically important differences (MCIDs) have been defined for several of the outcomes. For patient-reported outcomes, however, it is well established that there is a marked placebo effect and, therefore, treatment effects may be more difficult to evaluate for these outcomes than for traditional physiologic measures. Responder analyses, describing the proportion of patients exceeding the MCIDs during the treatment period, are more appropriate measures and are used in this article.

Reports from the ECLIPSE and other studies have demonstrated patients at risk of exacerbations have distinct clinical characteristics and a worse prognosis. When possible, data are presented in this group of patients separately.

Long-acting beta-2 adrenoreceptor agonist versus placebo

The safety and clinical efficacy of LABA versus placebo were evaluated in a Cochrane review of 26 RCTs, totaling 14,939 participants with moderate or severe airflow limitation. The median treatment period was 6 months and only a small number of the studies were enriched with patients experiencing exacerbations. Overall, LABA monotherapy was found to improve quality of life as measured by the St. George’s Respiratory Questionnaire (SGRQ), but the mean difference (MD −2.32 units [−3.09 to −1.54]) did not exceed the MCID of 4 units. LABAs also were associated with a decreased risk of moderate (odds ratio [OR] 0.73 [0.61–0.87]) and severe (OR 0.73 [0.56–0.95]) exacerbations. No impact was observed on the risk of mortality (OR 0.90 [0.75–1.08]) or severe adverse events (OR 0.97 [0.83–1.14]).

This Cochrane review did not include a responder analysis. Many of the older trials did not include such analyses. Newer trials demonstrated a significant proportion of responders in the LABA group compared with placebo.

TRISTAN (n=733, study duration of 12 months) and TORCH (n = 3,087, 36 months) were the largest double-blind trials to compare salmeterol to placebo, whereas SHINE (n = 584, 6 months) and INVOLVE (n = 867, 12 months) compared formoterol to placebo. Apart from patient-important outcomes, they also evaluated FEV ₁ longitudinally and demonstrated a decelerated FEV ₁ decline rate over time (MD 13.0 mL [4.3–21.7] per year in the TORCH trial). Moreover, TRISTAN and SHINE recruited patients with a history of exacerbations, whereas the remaining trials were not enriched for exacerbations. Their results were consistent with the Cochrane review findings, confirming the efficacy of LABAs in this group of patients.

Newer LABA molecules (indacaterol, vilanterol, and olodaterol) have a duration of action that exceeds 24 hours and for this reason they can be administered once a day and, therefore, have been called ultra-LABAs. Their safety and efficacy profiles are comparable with the older LABAs.

Long-acting muscarinic antagonist versus placebo

Tiotropium was for many years the only commercially available LAMA and, for this reason, it is the most thoroughly evaluated. Data from RCTs comparing tiotropium with placebo were pooled in a recent Cochrane meta-analysis of 22 RCTs with a study duration that varied between 3 months and 4 years (median: 6 months), with an overall population exceeding 23,000 participants. Most RCTs were of good methodological quality and included patients with severe or very severe airflow limitation. Only 2 studies selected patients with a history of exacerbations. Tiotropium was associated with a decrease in the number of patients experiencing 1 or more moderate or severe exacerbations (OR 0.78 [0.70–0.87]) or severe exacerbations (OR 0.85 [0.72–1.00]). It also was associated with an improved quality of life, evaluated by the SGRQ (MD −2.89 units [−3.35 to −2.44]). Although the MD did not exceed the MCID of 4 units, the number of participants that experienced such a positive change was higher in the tiotropium group (OR 1.52 [1.38–1.68]). Tiotropium did not appear to decrease mortality (OR 0.98 [0.86–1.11]). From a safety perspective, it was not associated with increased risk of serious adverse events.

Aclidinium, a newer LAMA compound, was evaluated in another Cochrane review, based on 12 good-quality RCTs, involving 9547 participants with moderate or severe COPD, not necessarily exacerbators, who were followed for 4 weeks to 52 weeks. Compared with placebo, aclidinium was found to decrease the proportion of severe exacerbations (OR 0.64 [0.46–0.88]), whereas there also was a trend over decreased proportion of moderate exacerbations (OR 0.88 [0.74–1.04]). It also increased the number of people who achieved a clinically meaningful improvement in their quality of life (OR 1.49 [1.31–1.7]) but did not appear to affect mortality (OR 0.92 [0.43–1.94]) or the risk of serious adverse events.

Similar results also were found for umeclidinium, which was evaluated in 4 studies of 12 weeks’ to 52 weeks’ duration and good methodological quality, involving 3798 participants with COPD. Umeclidinium decreased the proportion of participants experiencing at least 1 moderate or severe exacerbation (OR 0.61 [0.46–0.80]) and increased the number of people who achieved a clinically meaningful improvement (SGRQ; OR 1.45 [1.16–1.82]), although it was not associated with frequent serious adverse events (OR 1.33 [0.89–2.00]). Umeclidinium did not appear, however, to have an impact on the number of participants experiencing severe exacerbations (OR 0.86 [0.25–2.92]).

Tiotropium did not reduce the rate of decline in FEV ₁ in the UPLIFT trial. The TIE-COPD trial (n = 841) evaluated longitudinal impact of tiotropium on FEV ₁ decline over 24 months, in patients with mild or moderate airflow limitation and demonstrated a decrease in post-bronchodilator FEV ₁ decline rate (22 mLs preserved per year [6–37]; P = .006) and a trend over decreased pre-bronchodilator FEV ₁ decline rate (15 mLs [−1, 31] per year; P = .06). The conflicting results between the 2 trials might be explained by the earlier spirometric stage of the TIE-COPD participants, because FEV ₁ decline is more pronounced during earlier stages of COPD and, therefore, more amenable to therapeutic interventions. The impact of LAMAs on mortality has been evaluated extensively, although none of the available RCTs was powered to identify a potential survival benefit. A focused meta-analyses of 28 RCTs involving greater than 33,500 participants, who were followed between 6 weeks and 4 years, evaluated the impact of tiotropium versus placebo or any control on mortality. Data from all available studies were homogeneous and suggested tiotropium may be associated with a significant decrease in all-cause mortality (relative risk [RR] 0.86 [0.76–0.98]). This positive result was driven mainly from the UPLIFT trial, which recruited approximately 6000 participants, who were followed for 4 years, because all remaining studies had significantly more limited study population, duration, and fatalities. Decreased mortality associated with LAMAs also was observed in several extensive, well-conducted, real-life studies. Currently available evidence, however, cannot confidently confirm this hypothesis.

Long-acting beta-2 agonist versus long-acting muscarinic antagonist

The comparative safety and efficacy of LAMA versus LABA were evaluated in a meta-analysis of 16 good-quality RCTs involving 22,872 participants, who were followed between 12 weeks and 52 weeks. Similar to other RCTs evaluating LABAs or LAMAs as monotherapies, exacerbations history rarely were considered inclusion criteria. LAMAs achieved a greater decrease in the rate of moderate or severe exacerbations (OR 0.84 [0.74–0.94]) and were associated with a lower risk of adverse events (OR 0.92 [0.86–0.97]) compared with LABA. No significant differences were observed in quality-of-life measures, transitional dyspnea index, or the trough FEV ₁ .

The POET-COPD and INVIGORATE trials compared LAMA versus LABA among patients experiencing exacerbations. The POET-COPD study, a 1-year trial involving 7384 patients, compared tiotropium with salmeterol. Compared with salmeterol, tiotropium increased the time to first moderate or severe exacerbation (187 days vs 145 days, respectively; hazard ratio [HR] 0.83 [0.77–0.90]) and the time to first severe exacerbation (HR 0.72 [0.61–0.85]). It also decreased the frequency of moderate or severe (RR 0.89 [0.83–0.96]) and severe (RR 0.73 [0.66–0.82]) exacerbations, compared with salmeterol. The INVIGORATE trial compared tiotropium with the ultra-LABA indacaterol and found the latter to have a noninferior impact on trough FEV ₁ , health status (SGRQ), or the degree of dyspnea (TDI total score). Time to first exacerbation, however, was longer with tiotropium (HR 1.20 [0.73–1.33]) and indacaterol was associated with a higher rate of exacerbations (RR 1.24 [1.12–1.37]).

Long-acting beta-2 agonist/long-acting muscarinic antagonist versus monocomponents

Results of early studies evaluating the combination of tiotropium with a LABA as 2 separate inhalers versus the monocomponents were pooled in a Cochrane review of 10 good-quality RCTs, evaluating 10,894 participants. The addition of a LABA to tiotropium was associated with an increase in the proportion of patients achieving the MCID for SGRQ (OR 1.32 [1.19–1.46]) and a small increase in the trough FEV ₁ [MD 0.06 L [0.05–0.07]). No differences were observed in the rate of severe exacerbations, hospitalization for any cause, or mortality. The addition of tiotropium to a LABA led to an increase in the proportion of patients achieving the MCID for SGRQ (OR 1.53 [1.31–1.79]) and in the trough FEV ₁ (MD 0.07 L [0.06–0.09]) but did not appear to have a significant impact on the rate of severe exacerbations, hospital admission for any cause, or mortality. Newer meta-analyses, involving more RCTs, have confirmed these findings and also suggest that the addition of a LAMA to a LABA might lead to a significant decrease in the proportion of patients experiencing an exacerbation (OR 0.77 [0.62–0.97]).

The addition of a LAMA to a LABA in patients at risk of exacerbations has not been evaluated in any trials, whereas only 3 trials (DYNAGITO, SPARK, and ISRCTN29870041 ) have tested the addition of LABA to a LAMA in this population. In the DYNAGITO trial, comparing the combination of tiotropium with olodaterol versus tiotropium monotherapy in more than 9000 participants, dual bronchodilator therapy was not associated with a decreased rate of moderate or severe exacerbations at the prospectively targeted 0.01 significance level (RR 0.93; 99% CI [0.85–1.02]; P = .05). In addition, it failed to demonstrate a decrease in the frequency of severe exacerbations (RR 0.89; 95% CI [0.78–1.02]) or a prolongation in the time to the first severe and moderate or severe exacerbation. Similar findings were observed in the SPARK trial, where the combination of indacaterol with glycopyrronium was compared with glycopyrronium monotherapy or tiotropium monotherapy in 741 participants. Dual bronchodilation therapy was associated with a small but statistically significant decrease in the rate of moderate or severe exacerbations compared with glycopyrronium (RR 0.88 [0.77–0.99]) and only a trend over decreased exacerbations rate compared with tiotropium (RR 0.90 [0.79–1.02]). Moreover, compared with LAMA monotherapies, dual bronchodilation did not appear to decrease the rate of severe exacerbations. Similar results also were observed in the ISRCTN29870041 study. These findings suggest that a LABA/LAMA combination is likely to have only a modest impact on the frequency of moderate or severe exacerbations compared with LAMA monotherapy.

This observation is validated further by a well-designed real-life study from the Optum research database. This study comparing the impact of dual bronchodilator therapy versus LABA or LAMA monotherapies (mostly LAMA), involved 2572 participants and used propensity score matching to improve the comparability of the participants. Dual therapy was not associated with decreased risk of moderate or severe exacerbations but only with a modestly decreased risk of exacerbations leading to hospitalization.

Early introduction of a LAMA/LABA combination over monocomponents has been evaluated in several RCTs. A post hoc analysis of the ACLIFORM COPD and AUGMENT trials evaluated the introduction of aclidinium/formoterol compared with the monocomponents or placebo in 1056 treatment-naïve patients with COPD and moderate or severe airflow limitation. LABA/LAMA combination was associated with a modest improvement in quality-of-life and symptoms scores that did not exceed the MCID and improved postdose FEV ₁ compared with all comparators. In addition, the combination also improved trough FEV ₁ compared with LABA monotherapy or placebo. The role of LABA/LAMA combination for patients with a significant symptoms burden but no history of frequent exacerbations and no background inhaled corticosteroid (ICS) use was prospectively evaluated in the EMAX trial, which compared the efficacy of the combination of umeclidinium and vilanterol versus umeclidinium and salmeterol monotherapies. LABA/LAMA combination led to an increase in trough FEV ₁ compared with LAMA (MD 66 mL [43–89]) and LABA (MD 141 mL [118–164]) monotherapy. It also was associated with an improved TDI total score compared with LAMA (MD 0.37 [0.06–0.68]) and LABA (MD 0.45 [0.15–0.76]), which did not, however, exceed the MCID. In the TDI responder analysis, LABA/LAMA combination was superior to LAMA (OR 1.43 [1.17–1.75]) and LABA (OR 1.48 [1.21–1.81]). Exacerbations frequency and incidence of severe adverse events were similar in all groups. Overall, in this group of patients, addition of a second bronchodilator may limit the burden of respiratory symptoms to some extent, without posing an additional risk of adverse events.

Long-acting muscarinic antagonist versus long-acting beta-2 agonist/inhaled corticosteroid

Few RCTs have compared the combination of LABA/ICS with a LAMA. Pooled results of 2 RCTs of unclear methodological quality, comparing once-daily administration of LABA/ICS versus LAMA, involving 880 participants, did not reveal any between-treatment difference on the rate of exacerbations, mortality, risk of pneumonia, or quality of life.

Twice-daily administration of fluticasone/salmeterol versus once-daily tiotropium was compared in the INSPIRE trial, a 2-year RCT involving 1323 participants with severe airflow limitation and a clinical history of COPD exacerbations. Although no difference was found on the annual rate of moderate or severe exacerbations (RR 0.97 [0.84–1.12]), LABA/ICS was associated with a decrease in SGRQ total score (MD 2.1 units [0.1–4.0]), with a corresponding increase in the proportion of participants achieving the MCID (OR 1.30 [1.05–1.61]) compared with LAMA. From a safety perspective, LABA/ICS was associated with an increased risk of pneumonia (HR 1.94 [1.19–3.17]), but lower risk of death (RR 0.56 [0.33–0.94]).

These findings need to be interpreted with caution because COPD is a heterogeneous disease and patients have a differential response to ICS. Several studies have demonstrated ICS are effective only for patients with higher blood eosinophil count ^, and this was not accounted for. Accordingly, a post hoc analysis of the INSPIRE trial suggested superiority of LABA/ICS in reducing the rate of moderate or severe exacerbations in patients with blood eosinophils of greater than or equal to 2% of total white cell count at presentation (25% reduction [8% to 40%]) and a trend toward inferiority of LABA/ICS in the remaining patients (18% increase [−8% to 51%]).

Long-acting beta-2 agonist/long-acting muscarinic antagonist versus long-acting beta-2 agonist/inhaled corticosteroid

The Cochrane review comparing LABA/LAMA versus LABA/ICS combinations included 11 RCTs of good methodological quality, comprising 9839 participants. Study duration varied between 6 weeks and 52 weeks. Only 1 of the included studies was enriched in patients experiencing exacerbations. Compared with LABA/ICS, LAMA/LAMA led to a decrease in the proportion of patients who experienced at least 1 moderate or severe exacerbation (OR 0.82 [0.70–0.96]) and the risk of pneumonia (OR 0.57 [0.42–0.79]). More people receiving LABA/LAMA achieved the MCID in SGRQ (OR 1.25 [1.09–1.44]), whereas there were no significant between group differences in the risk of serious adverse events or all-cause mortality.

An important limitation of this review and the included RCTs is that they did not account for the presence of airway eosinophilic inflammation, a trait that is known to be associated with ICS treatment response. A post hoc analysis from the FLAME, ^, the largest of the included RCTs (n = 3362), explored if blood eosinophils could predict treatment response. LABA/LAMA combination was more effective than LABA/ICS in preventing exacerbations (all severities, or moderate or severe) at lower eosinophil counts, whereas there was no treatment difference at higher eosinophil counts. Contrasting findings were reported in a prespecified analysis from the IMPACT trial, where LABA/LAMA were compared with LABA/ICS for 1 year in 6204 participants. ^, LABA/LAMA was more effective than LABA/ICS in reducing the rate of exacerbations among patients with lower blood eosinophil counts, whereas LABA/ICS was more effective among patients with a higher eosinophil count. Differences in the inclusion criteria of the trials may be partly responsible for these differences. Importantly, IMPACT included more patients with 2 exacerbations in the previous year (54% vs 19%, respectively) compared with FLAME; this suggests that the effects of ICS are increased in patients at higher exacerbation risk. Additionally, FLAME excluded patients with any asthmatic characteristics (including those with a current or previous history of asthma, any other respiratory disease or symptoms prior to the age of 40, very high eosinophils, and allergic rhinitis), who are more likely to respond to ICS. On the contrary, IMPACT only excluded subjects with a current diagnosis of asthma. In addition, FLAME included a run-in period of 1 month, whereas all participants were receiving only tiotropium. Patients who potentially could benefit from ICS might have experienced a disease deterioration during this period, potentially leading to a discontinuation and exclusion from the trial.

Impact of bronchodilators on dynamic hyperinflation and exercise performance

Several trials have evaluated the impact of dual bronchodilation on exercise performance. MORACTO 1 and MORACTO 2, two replicate, double-blind, 6-week crossover trials involving 585 patients with moderate or severe airflow limitation, assessed the impact of the combination of tiotropium with olodaterol compared with the monocomponents or placebo. Compared with placebo, exercise endurance time during constant work-rate cycle ergometry was improved by all active treatments: LAMA (MD 65.60 seconds [63.94–67.26]), LABA (MD 50.50 seconds [48.83–52.17]), and LABA/LAMA combination (MD 83.90 seconds [82.25–85.55]). Tiotropium also improved exercise endurance time compared with olodaterol (MD 15.10 seconds [13.44, 16.76]). Finally, LABA/LAMA combination significantly improved the exercise endurance time 5.6% compared with olodaterol (MD 33.4 seconds [31.75–35.05]), but no significant difference was observed between LABA/LAMA and LAMA monotherapy.

Change in the inspiratory capacity from baseline was used to evaluate the impact of treatments on hyperinflation. LABA/LAMA combination was superior to LAMA (MD 87 mL [34–141]), LABA (MD 92 mL [38–145]), and placebo (MD 218 mL [164–271]). LABA and LAMA monotherapies also were superior to placebo.

Two other 12-week trials involving 657 patients with moderate or severe airflow limitation ( NCT01323660 and NCT01328444 ) assessed the impact of the combination of indacaterol with glycopyrronium compared with the monocomponents or placebo on exercise tolerance. Dual bronchodilation initially appeared to significantly improve exercise endurance time compared with placebo. This effect was not maintained at 12 weeks, however, mainly because of a significant treatment effect in the placebo group.

An 8-week study comparing aclidinium/formoterol to placebo, the ACTIVATE trial, also included a behavioral intervention for all participants, aiming at enhancing their exercise endurance. In this trial, LABA/LAMA combination achieved a maintained improvement in exercise tolerance, demonstrated by 55.2 seconds’ ( P = .03) increase in the exercise endurance time, daily activity (treatment difference of 731 steps per day; P = .002), trough FEV ₁ , and different measures of lung hyperinflation. The behavioral intervention might have been the key success factor in the ACTIVATE trial. More specifically, the challenge in these studies often is that patients are deconditioned and, unless exercise is trained and maintained along treatment initiation and/or escalation, it will be difficult to translate any improvement in lung function into an effect on exercise capacity.

Bronchodilators and comorbidities

Although the safety of LABAs and LAMAs has been confirmed in numerous RCTs, there still is ongoing discussion regarding their cardiovascular safety, owing to their mechanisms of action. As a result, cardiovascular safety was evaluated in a meta-analysis involving 43 RCTs, which did not find any association between cardiovascular death and the use of LAMA (RR 0.92 [0.81–1.04]), LABA (RR 0.93 [0.75–1.15]), or LAMA/LABA combination (RR 1.03 [0.38–2.80]). No impact on the frequency of overall cardiovascular adverse events was found either. Use of LABA appeared associated with an increased risk of cardiac failure (RR 1.71 [1.04–2.84]).

Cardiovascular safety of introducing a second LABD for COPD was evaluated in a real-life study from the UK Clinical Practice Research Datalink, involving 62,348 patients receiving single or dual bronchodilators, who were matched using high-dimensional propensity scores. Adding a second LABD was associated with a small increase in the risk of heart failure (HR 1.16 [1.03–1.30]), but it did not have an impact on the risk of myocardial infarction, stroke or any arrhythmia. The increase in the risk of heart failure was confined to the subgroup where patients receiving LABA/LAMA were compared with matched patients receiving LAMA monotherapy (HR 1.28 [1.07–1.54]).

Moreover, cardiovascular safety of LABAs versus LAMAs was compared in 52,884 propensity matched patients with COPD from the same database. There were no significant differences in the risk of myocardial infarction, stroke, or arrhythmia, but there was a trend over increased risk of heart failure associated with the use of LABAs. The LABA population, however, included in these 2 real-life observational studies was overlapping. Therefore, the results of these studies were not independent and could have been driven by an unexpectedly high risk of heart failure in the included LABA population.

These findings were not confirmed, however, in SUMMIT, an extensive RCT of 16,590 participants with COPD and heightened cardiovascular risk, where the combination of fluticasone furoate and vilanterol was compared with the monocomponents and placebo. ^, This RCT, specifically designed to evaluate the cardiovascular safety, did not reveal between-group differences in cardiovascular outcomes, including sudden death, acute coronary syndrome, arrhythmia, cardiac failure, stroke, transient ischemic attack, or any adverse cardiovascular event. More specifically, heart failure was reported in 4% of the participants who received the LABA vilanterol and in 5% of the participants in the placebo group.

Overall, cardiovascular safety of LABA, LAMA, and LABA/LAMA combination is strongly supported by these findings. The potential association between LABA and heart failure found in some studies is more likely the result of type 1 statistical error.

Factors to be considered when choosing molecules and devices

There now are a wide range of different LABD molecules available for clinicians to choose from, administered though different inhaler devices. Although this offers patient choice, allowing individuals to choose a suitable inhaler device, it also poses challenges to clinicians faced with making optimum treatment recommendations. Once a clinician has decided on the class of drug to be prescribed (eg, a LAMA), then the choice of molecule and inhaler device has to be made. This should be done based on considerations of efficacy, safety, once-a-day versus twice-a-day dosing, inhaler device characteristics, and local cost issues.

In general, there is little to differentiate between molecules within a class when considering efficacy and safety. There is evidence that indacaterol has greater efficacy than twice-a-day LABAs. ^{, ,} The reported differences, however, between other molecules within a class are small; for example, a 12-week study comparing tiotropium and umeclidinium reported a 53-mL difference in trough FEV ₁ in favor of the latter but without any treatment differences in symptom scores. Similarly, the combination treatment umeclidinium/vilanterol caused a 52-mL greater trough FEV ₁ increase compared with tiotropium/olodaterol but again with no differences in symptoms.

The profile of bronchodilation over 24 hours caused by once-a-day versus twice-a-day LABDs is different. Accordingly, the choice of LABDs may be tailored to individual patient choice based on requirements for additional bronchodilation in the evening. This may be important particularly for patients who suffer with respiratory symptoms at night or in the early morning.

Perhaps the most important consideration when making a choice within LABD drug class is the selection of inhaler device. In general, pressurized metered dose inhalers require more coordination of operation and inhalation but are not dependent on the inspiratory flow generated. In contrast, breath-activated dry powder inhalers are difficult for some patients to use because they cannot reach the required inspiratory flow rates.

In summary, the choice of drug within LABD class should be driven mainly by tailoring treatment to individual needs regarding inhaler type and the suitability of once-a-day versus twice-a-day delivery. This is influenced by local cost issues.

Place of long-acting bronchodilators in guidelines and the clinical practice

In the strategy document from the Global Initiative for Chronic Obstructive Lung Disease (GOLD), the introduction of a single LABD followed by a dual combination of LABDs is recommended as the initial maintenance therapy for all patients, apart from those experiencing frequent exacerbations, who have raised blood eosinophils. An option to use a combination of LABDs as first line of treatment of patients with GOLD group B or group D with more severe symptoms also is offered. For these patients, early introduction of an ICS also may be considered. Overall, GOLD recommends LABDs as the mainstay of COPD treatment of all patients experiencing symptoms and/or exacerbations.

The Australian and New Zealand guidelines (COPD-X) follow a similar approach, suggesting the use of a LABD in case short-acting bronchodilators are insufficient, with LABA/LAMA combination the next step. A combination of a LABA with an ICS without the addition of a LAMA is suggested as only an option in cases of severe airflow limitation (FEV ₁ <50% predicted), with frequent exacerbations, although the associated risks of pneumonia and inferiority of LABA/ICS in reducing the frequency of exacerbations are highlighted.

The National Institute for Health and Care Excellence (NICE), in the United Kingdom, recommends the introduction of a LABA/LAMA combination for all patients who do not have asthmatic features and experience symptoms or exacerbations, despite using a short-acting bronchodilator as needed. This contrasts with previous guidelines suggesting a step-up approach, starting from a single bronchodilator, and may appear counterintuitive, given that many patients respond well to a single LABD. Given their demonstrated safety, even a modest effect of adding a second bronchodilator is considered acceptable probably because the costs of LAMA monotherapy and dual bronchodilation therapy are similar and NICE recommendations are based on cost-effectiveness analyses. Despite the positive group level data, however, there is a wide variability between individual patients in the magnitude of clinical response to LABDs. Therefore, in real life, not every patient gains benefit from the introduction of a second LABD and it still is not clear which individuals would benefit. Moreover, patients with mild airflow limitation, many of whom are classified in GOLD group A, have not been included in any of the trials evaluating dual bronchodilator combinations; thus, the evidence base supporting treatment decisions is thin in this group. Overall, based on currently available evidence, a stepwise approach might be more appropriate. For patients exerting asthmatic features, NICE recommends the use of LABA/ICS combination. Asthmatic features are defined as a confirmed diagnosis of asthma or atopy, higher blood eosinophil count, a substantial variation of FEV ₁ over time or substantial diurnal variation in the peak expiratory flow. Many of these criteria appear arbitrary and have not been tested in RCTs; only blood eosinophil counts have evidence that supports prediction of ICS effects.

Spanish guidelines (GesEPOC) also suggest the use of single or dual bronchodilators as first-line treatment of the majority of patients with COPD. LABA/ICS combination is recommended as a first-line treatment for those with asthma-COPD overlap.

Summary

The management of COPD should be guided by patients’ symptoms and exacerbations. LABA, LAMA, and their combinations represent safe and effective options for maintenance therapy. Introduction of a single LABD should be considered the first step in the maintenance treatment of all patients, apart from those with more severe symptoms and frequent exacerbations (GOLD group D). Compared with a LABA, a LAMA is more effective in preventing exacerbations, and, for this reason, LABA monotherapy should be avoided in patients experiencing frequent exacerbations. The threshold for adding a second LABD should be low because dual bronchodilator therapy has shown greater effect on lung function and quality of life and an excellent safety profile. The most important consideration when making a choice within a LABD drug class is the selection of inhaler device and clinicians must devote adequate time to demonstrate use of the inhaler device and periodically check and reinforce the correct technique.