Drug treatments

Beta2 agonists (inhaled)

In this section:

Key points | Benefits and Harms | Comment

Top

Key points

  • For GRADE evaluation of interventions for COPD, see table.
  • Inhaled beta2 agonists improve lung function and symptoms and reduce exacerbations in stable COPD compared with placebo.
  • It is unclear whether inhaled anticholinergics or inhaled beta2 agonists are the more consistently effective drug class in the treatment of COPD.
Top

Benefits and harms

Short-acting beta2 agonists (short-term treatment) versus placebo:

We found two systematic reviews (search dates 2002[26] and 2004[27]) and one subsequent RCT.[28]

Lung function and exercise capacity

Compared with placebo Short-acting beta2 agonists (short-term treatment) may be more effective at increasing FEV1 in people with stable COPD, but we don't know whether they are more effective at increasing exercise tolerance (low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Lung function

[26]

Systematic review

196 people

6 RCTs in this analysis

FEV1

short-acting beta2 agonists (delivered by metered-dose inhaler)

placebo

Absolute results not reported

WMD 0.14 L

95% CI 0.04 L to 0.25 L

The trials were small and the results heterogeneous

The meta-analysis used post-crossover results, but because the treatment is short acting there is unlikely to be persistence of treatment effects after crossover

not-calculated

short-acting beta2 agonists

[28]

RCT

4-armed trial

209 people with COPD

Other comment

The remaining arms assessed levosalbutamol (levalbuterol) 1.25 mg and racemic salbutamol (albuterol) 2.5 mg

% change in area under the curve (AUC) in FEV1, 6 weeks

10.5% with levosalbutamol 0.63 mg three times daily

1.6% with placebo

Other comment

108 people in this assessment

P <0.003

not-calculated

levosalbutamol 0.63 mg

[28]

RCT

4-armed trial

209 people with COPD

Other comment

The remaining arms assessed levosalbutamol 0.63 mg and racemic salbutamol 2.5 mg

% change in AUC in FEV1, 6 weeks

9.2% with levosalbutamol 1.25 mg three times daily

1.6% with placebo

Other comment

104 people in this analysis

P <0.003

not-calculated

levosalbutamol 1.25 mg

[28]

RCT

4-armed trial

209 people with COPD

Other comment

The remaining arms assessed levosalbutamol 0.63 mg and 1.25 mg

% change in AUC in FEV1, 6 weeks

15.3% with racemic salbutamol 2.5 mg three times daily

1.6% with placebo

Other comment

107 people in this analysis

P <0.003

not-calculated

racemic salbutamol 2.5 mg

Exercise capacity

[26]

Systematic review

188 people

4 RCTs in this analysis

Distance walked

short-acting beta2 agonists (delivered by metered-dose inhaler)

placebo

Absolute results not reported

SMD +0.18 m

95% CI –0.11 m to +0.47 m

The trials were small and the results heterogeneous

The meta-analysis used post-crossover results, but, because the treatment is short acting, there is unlikely to be persistence of treatment effects after crossover

not-significant

not significant

No data from the following reference on this outcome.[27]

COPD exacerbation and worsening of symptoms

Compared with placebo Short-acting beta2 agonists (short-term treatment) may be more effective at improving daily breathlessness scores in people with stable COPD (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Symptom severity

[26]

Systematic review

188 people

4 RCTs in this analysis

Daily breathlessness score

short-acting beta2 agonists (delivered by metered-dose inhaler)

placebo

Absolute results not reported

WMD –1.33

95% CI –1.65 to –1.01

P <0.001

The trials were small and the results heterogeneous

The meta-analysis used post-crossover results, but, because the treatment is short acting, there is unlikely to be persistence of treatment effects after crossover

not-calculated

short-acting beta2 agonists

[28]

RCT

4-armed trial

209 people with COPD

Withdrawals because of COPD exacerbations

2% with levosalbutamol 0.63 mg three times daily

4% with levosalbutamol 1.25 mg three times daily

10% with racemic salbutamol 2.5 mg three times daily

0% with placebo

Absolute numbers not reported

P = 0.01 for racemic salbutamol v placebo

Significance not assessed for either dose of levosalbutamol v placebo

not-calculated

placebo

No data from the following reference on this outcome.[27]

Mortality

No data from the following reference on this outcome.[26][27][28]

Quality of life

No data from the following reference on this outcome.[26][27][28]

Adverse effects
Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Adverse effects

[27]

Systematic review

15,276 people with asthma or COPD

22 RCTs in this analysis

Adverse cardiovascular events

beta2 agonists (short- and long-acting)

placebo

Absolute results not reported

RR 2.54

95% CI 1.59 to 4.05

moderate

placebo

[27]

Systematic review

386 people with asthma or COPD

11 RCTs in this analysis

Increased heart rate

beta2 agonists (single dose of either short- or long-acting)

placebo

Absolute results not reported

WMD 9.12

95% CI 5.32 to 12.92

not-calculated

placebo

[27]

Systematic review

168 people with asthma or COPD

6 RCTs in this analysis

Reduction in serum potassium concentration

beta2 agonists (single dose of either short- or long-acting)

placebo

Absolute results not reported

WMD –0.36

95% CI –0.54 to –0.18

not-calculated

placebo

No data from the following reference on this outcome.[26][28]

Short-acting beta2 agonists (long-term treatment) versus placebo:

We found no systematic review of only long-term treatment with short-acting beta2 agonists versus placebo.

Long-acting beta2 agonists (short-term or long-term treatment) versus placebo:

We found no review on only short-term (follow-up <6 months) or only long-term (>6 months) treatment with long-acting beta2 agonists compared with placebo. We found 4 systematic reviews (search dates 2002,[29] 2005,[30][31] and 2007[32]), 5 additional RCTS,[19][33][34][35][36] and 5 subsequent RCTs that combined data on a range of treatment duration.[37][38][39][40][41] In addition, we found one systematic review that reported on adverse effects.[27]

Mortality

Compared with placebo Long-acting beta2 agonists (short-term or long-term treatment) seem no more effective at reducing mortality (moderate-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Mortality

[32]

Systematic review

8400 people with COPD

13 RCTs in this analysis

All-cause mortality, 1 to 36 months

4.9% with long-acting beta2 agonist

6.5% with placebo

Absolute numbers not reported

RR 1.6

95% CI 0.8 to 2.4

P >0.05

not-significant

not significant

No data from the following reference on this outcome.[29][30][31][19][33][34][35][36][37][38][39][40][41]

Lung function and exercise capacity

Compared with placebo Long-acting beta2 agonists (short-term or long-term treatment) seem more effective at improving lung function, but we don't know whether they are more effective at improving capacity for exercise (endurance time and shuttle walking distance) (moderate-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Lung function

[34]

RCT

657 people with COPD

Change in FEV1 from baseline, 6 months

+5% with formoterol 9 micrograms twice daily

–1.4% with placebo

Other comment

Both groups were allowed to take terbutaline 0.5 mg as needed

AR 6.5%

95% CI 2.5% to 10.7%

P <0.01

not-calculated

formoterol

[37]

RCT

5-armed trial

717 people with COPD

Change in pre-dose FEV1 from baseline, 12 weeks

+16.9% with arformoterol 15 micrograms twice daily

+18.9% with arformoterol 25 micrograms twice daily

+14.9% with arformoterol 50 micrograms once daily

+17.4% with salmeterol 42 micrograms twice daily

+6.0% with placebo

P <0.001 for all active treatments v placebo

not-calculated

long-acting beta2 agonist

[38]

RCT

3-armed trial

163 people with COPD

Change in pre-dose FEV1 from baseline, 28 days

+220 mL with indacaterol 400 micrograms once daily

+210 mL with indacaterol 800 micrograms once daily

Other comment

Placebo value for change in FEV1 not reported

P <0.0001 for either dose of indacaterol v placebo

not-calculated

indacaterol

Exercise capacity

[19]

RCT

3-armed trial

183 people with moderate to severe COPD

In review [29][30]

Other comment

The third arm assessed ipratropium

Increase shuttle walking test (increase in distance from baseline), 12 weeks

20.4 m with formoterol 18 micrograms twice daily

6.0 m with placebo

Other comment

Baseline mean distance was 325 m

Reported as not significant (formoterol v placebo)

P value not reported

not-significant

not significant

[39]

RCT

Crossover design

20 people with clinically stable COPD

Mean difference in endurance shuttle walking test (ESWT), 2.5 hours post treatment

salmeterol 50 micrograms single dose

placebo single dose

Absolute numbers not reported

Other comment

Mean distance walked was 160 m more with salmeterol than with placebo

People received standardised instructions to walk for as long as possible, with a predetermined 20-minute maximum

P = 0.02

not-calculated

salmeterol

[36]

RCT

Crossover design

23 people with moderate to severe COPD (mean FEV1 42% predicted)

In review [29][30]

Difference in peak exercise endurance time, 12 weeks

salmeterol 50 micrograms (inhaled)

placebo

Absolute results not reported

Difference between groups of 96 seconds

P = 0.02

not-calculated

salmeterol

[33]

RCT

Crossover design

5-armed trial

34 people

Other comment

The fifth arm assessed ipratropium (80 micrograms three times daily)

Time to exhaustion, 1 week

10.94 minutes with formoterol 4.5 micrograms

10.78 minutes with formoterol 9 micrograms

10.59 minutes with formoterol 18 micrograms

10.20 minutes with placebo

P <0.0001 (formoterol 4.5 micrograms v placebo)

P <0.01 (formoterol 9 micrograms v placebo)

P <0.05 (formoterol 18 micrograms v placebo)

not-calculated

formoterol

No data from the following reference on this outcome.[31][32][35][40][41]

COPD exacerbation and worsening of symptoms

Compared with placebo Long-acting beta2 agonists (short-term or long-term treatment) seem more effective at reducing the rate of COPD exacerbations and at improving symptoms (assessed by the Chronic Disease Respiratory Questionnaire and Transitional Dyspnoea Index) (moderate-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
COPD exacerbations

[32]

Systematic review

6453 people with COPD

14 RCTs in this analysis

Cumulative incidence of severe COPD exacerbations, 1 to 36 months

7.5% with long-acting beta2 agonist

10.8% with placebo

Absolute numbers not reported

RR 0.80 (random effects model)

95% CI 0.69 to 0.82

small

long-acting beta2 agonist

[37]

RCT

5-armed trial

717 people with COPD

Frequency of COPD exacerbations

19/141 (13.5%) with arformoterol 15 micrograms twice daily

19/143 (13.3%) with arformoterol 25 micrograms twice daily

17/146 (11.6%) with arformoterol 50 micrograms once daily

20/144 (13.9%) with salmeterol 42 micrograms twice daily

24/143 (16.8%) with placebo

Other comment

Double-dummy trial

P >0.05 for all treatments v placebo

not-significant

not significant

[38]

RCT

3-armed trial

163 people with COPD

Number of COPD exacerbations, 28 days

3/68 (4%) with indacaterol 400 micrograms

0/67 (0%) with indacaterol 800 micrograms

2/28 (7%) with placebo

Significance not assessed

Symptom severity

[30]

Systematic review

545 people

2 RCTs in this analysis

Improvement in the Chronic Respiratory Disease Questionnaire (CRQ)

long-acting beta2 agonists

placebo

Absolute results not reported

Other comment

Long-acting beta2 agonists assessed by the review were salmeterol and formoterol

OR 1.71

95% CI 1.21 to 2.42

P = 0.002

small

long-acting beta2 agonists

[30]

Systematic review

736 people

2 RCTs in this analysis

Improvement in transitional dyspnoea index (TDI)

long-acting beta2 agonists

placebo

Absolute results not reported

Other comment

Long-acting beta2 agonists assessed by the review were salmeterol and formoterol

OR 1.70

95% CI 1.25 to 2.31

P = 0.0008

small

long-acting beta2 agonists

No data from the following reference on this outcome.[29][31][19][33][34][35][36][39][40][41]

Quality of life

Compared with placebo Long-acting beta2 agonists (short-term or long-term treatment) may be no more effective at improving quality of life (moderate-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Quality of life

[32]

Systematic review

6453 people with COPD

14 RCTs in this analysis

Mean change in St George's Respiratory Questionnaire, 1 to 36 months

long-acting beta2 agonist (salmeterol or formoterol)

placebo

Absolute numbers not reported

WMD –3.26

95% CI –4.57 to –1.96

not-calculated

long-acting beta2 agonist

No data from the following reference on this outcome.[29][30][31][19][33][34][35][36][37][38][39][40][41]

Adverse effects
Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect size (for ORs and RRs)Favours
Adverse effects

[31]

Systematic review

2404 people with COPD

4 RCTs in this analysis

Mortality attributed to treatment-related respiratory problems

21/1320 (2%) with long-acting beta2 agonist

8/1084 (1%) with placebo

RR 2.47

95% CI 1.12 to 5.45

P = 0.03

moderate

placebo

[30]

Systematic review

5055 people

12 RCTs in this analysis

Proportion of people who withdrew because of an adverse effect

long-acting beta2 agonists

placebo

Absolute results not reported

Other comment

The most common immediate adverse effect is tremor, which is usually worse in the first few days of treatment

OR 0.86

95% CI 0.72 to 1.02

P = 0.09

not-significant

not significant

[34]

RCT

657 people with COPD

Rate of adverse effects (rates/1000 treatment days), 6 months

3.8 with formoterol 9 micrograms twice daily

4.5 with placebo

Other comment

Both groups were allowed to take terbutaline 0.5 mg as needed

No further information on adverse effects was given

Significance not assessed

[35]

RCT

4-armed trial

6184 people with COPD; 6112 people included in efficacy analysis

Other comment

The third arm assessed salmeterol 50 micrograms once daily plus fluticasone 500 micrograms twice daily and the fourth arm assessed fluticasone alone (500 micrograms twice daily) alone

Proportion of people experiencing a drug-related adverse effect

12% with salmeterol (50 micrograms twice daily)

13% with placebo

Absolute numbers not reported

Other comment

3045 people in this analysis; includes people who had discontinued study medication

The most common adverse effect reported was COPD exacerbation

Significance not assessed

[27]

Systematic review

15,276 people with asthma or COPD

22 RCTs in this analysis

Adverse cardiovascular events

beta2 agonists (short- and long-acting)

placebo

Absolute results not reported

RR 2.54

95% CI 1.59 to 4.05

moderate

placebo

[27]

Systematic review

386 people with asthma or COPD

11 RCTs in this analysis

Increased heart rate

beta2 agonists (single dose of either short- or long-acting)

placebo

Absolute results not reported

WMD 9.12

95% CI 5.32 to 12.92

not-calculated

placebo

[40]

RCT

5-armed trial

1465 people with COPD

Heart rate difference from baseline, 12 weeks

–2.4 bpm with arformoterol 15 micrograms twice daily

–0.6 bpm with arformoterol 25 micrograms twice daily

–0.3 bpm with arformoterol 50 micrograms once daily

–0.0 bpm with salmeterol 42 micrograms twice daily

–1.8 bpm with placebo

Other comment

Pooled results of 2 identically designed phase III RCTs

Rates of atrial tachycardia, atrial fibrillation or flutter, and non-sustained and sustained ventricular tachycardia did not increase with long-acting beta2 agonist compared with placebo

P >0.05 for any active treatment v placebo

not-significant

not significant

[41]

RCT

3-armed trial

351 people with COPD

Arrhythmia, 12 weeks

formoterol 20 micrograms nebulised twice daily

formoterol 12 micrograms dry powder twice daily

placebo

Absolute numbers not reported

Other comment

Baseline heart rate and rates of atrial tachycardia, atrial fibrillation, and ventricular tachycardia did not increase with formoterol compared with placebo

P >0.05 for either dose of formoterol v placebo

not-significant

not significant

[27]

Systematic review

168 people with asthma or COPD

6 RCTs in this analysis

Reduction in serum potassium concentration

beta2 agonists (single dose of either short- or long-acting)

placebo

Absolute results not reported

WMD –0.36

95% CI –0.54 to –0.18

not-calculated

placebo

No data from the following reference on this outcome.[29][32][33][36][37][38][19][39]

Beta2 agonists versus inhaled anticholinergics:

See option on inhaled anticholinergics versus beta2 agonists.

Beta2 agonists alone versus inhaled anticholinergics plus beta2 agonists:

See option on inhaled anticholinergics plus beta2 agonists.

Beta2 agonists alone versus inhaled corticosteroids plus beta2 agonists:

See option on inhaled corticosteroids plus beta2 agonists.

Further information on studies

[30]Owing to heterogeneity among studies in reporting of effects on FEV1, the review did not pool data for this outcome. However, the review reported that most RCTs found an improvement in FEV1 with long-acting beta2 agonists compared with placebo. The review reported that RCTs found no significant difference between long-acting beta2 agonists and placebo in effects on exercise as measured by various walking tests, but the review did not pool data for this comparison.

[35]The RCT also carried out a last observation carried forward analysis for the outcome of FEV1. However, the withdrawal rate from the RCT was high and the proportion of people followed up at 3 years for this outcome was 56% (851/1524) in the placebo group, and 63% (960/1521) in the salmeterol alone group. These follow-up rates are below Clinical Evidence reporting criteria of 80%, and so these data are not reported here.

Top

Comment

Clinical guide:

High doses of beta2 agonists can reduce plasma potassium, cause dysrhythmia, and reduce arterial oxygen tension.[42] The risk of adverse events may be higher in people with pre-existing cardiac arrhythmias and hypoxaemia.[43]