Evidence review

This evidence review focuses on 2 published randomised controlled trials (RCTs) that provide published evidence for indacaterol/glycopyrronium in managing patient-oriented symptoms of chronic obstructive pulmonary disease (COPD):

  • SPARK (Wedzicha et al. 2013) investigated whether indacaterol/glycopyrronium prevented more exacerbations than glycopyrronium or another a long-acting muscarinic antagonist (LAMA) (tiotropium) alone in people with severe or very severe COPD.

  • BLAZE (Mahler et al. 2013) compared the effects of indacaterol/glycopyrronium on dyspnoea (breathlessness) with the effects of placebo and tiotropium in people with moderate to severe COPD.

Two other published 26-week RCTs, which investigated lung function measures as primary outcomes (forced expired volume in 1 second [FEV1]) are also discussed:

  • SHINE (Bateman et al. 2013) compared indacaterol/glycopyrronium with indacaterol, glycopyrronium, tiotropium and placebo in people with moderate to severe COPD.

  • ILLUMINATE (Vogelmeier et al. 2012) compared indacaterol/glycopyrronium with fluticasone/salmeterol, a combined inhaled corticosteroid/long-acting beta2-agonist (ICS/LABA), in people with moderate to severe COPD.

ENLIGHTEN (Dahl et al. 2013), a 52-week, placebo-controlled safety study, which also included some efficacy end points as secondary outcomes, is also summarised.

SPARK (Wedzicha et al. 2013)

  • Design: RCT in 362 centres in 27 countries, including the UK, comparing the effects of indacaterol/glycopyrronium with glycopyrronium (the primary objective) and tiotropium (the key secondary objective) on exacerbations in people with severe or very severe COPD. Treatment with indacaterol/glycopyrronium and glycopyrronium was double-blind; tiotropium was used open-label because double-blind treatment was not available. Allocation was concealed. The study included a 14-day run-in period and a 64-week treatment period, extended to 76 weeks after a mid-study reassessment of sample size to ensure the study had sufficient statistical power.

  • Population: 2224 adults aged 40 years or over (mean age around 63 years) who were current or ex-smokers with a smoking history of at least 10 pack-years. Patients had severe or very severe airflow limitation (stages III to IV of the 2008 Global Initiative for Chronic Obstructive Lung Disease [GOLD] criteria); post-bronchodilator FEV1 less than 50% predicted; and FEV1/forced vital capacity [FVC] ratio less than 0.70 at screening. Patients were required to have a documented history of at least 1 exacerbation in the previous 12 months that needed treatment with systemic corticosteroids or antibiotics, or both. Long-acting bronchodilators were discontinued before screening but continued use of ICS was permitted during the study, with patients using ICS combination inhalers switched to the same or equivalent dose and regimen of single-component ICS.

  • Intervention and comparison: patients were randomised in approximately equal numbers to treatment with:

    • indacaterol/glycopyrronium 110/50 micrograms once daily (using the Breezhaler; delivered doses 85/43 micrograms, as per the approved product)

    • glycopyrronium 50 micrograms once daily (using the Breezhaler)

    • tiotropium 18 micrograms once daily (using the Spiriva Handihaler).

  • Outcome: the primary end point (indacaterol/glycopyrronium compared with glycopyrronium) and the key secondary efficacy end points (indacaterol/glycopyrronium compared with tiotropium) were the annualised rates of moderate or severe COPD exacerbations (number of events per patient per year). Moderate exacerbations were defined as worsening symptoms of COPD needing treatment with systemic corticosteroids and/or antibiotics. Severe exacerbations were defined as worsening symptoms of COPD that needed hospital admission or emergency treatment. Other secondary end points included annualised rates of exacerbations by degree of severity, use of rescue salbutamol and adverse events (summarised in table 2). Measures of lung function, health status scores (as measured by the St. George's Respiratory Questionnaire [SGRQ]) and adverse events were also assessed.

Table 2 Summary of SPARK : Wedzicha et al. (2013)

Indacaterol/

glycopyrronium

(110/50 micrograms once daily)

Glycopyrronium

(50 micrograms once daily)

Tiotropium

(18 micrograms once daily)

Analysis

Randomised

n=741

n=741

n=742

Efficacya

n=729

n=739

n=737

Primary outcome and key secondary outcome

Annualised rate of moderate or severe exacerbations

0.84 (95% CI 0.75 to 0.94)

0.95 (95% CI 0.85 to 1.06)

0.93 (95% CI 0.83 to 1.04)

Indacaterol/glycopyrronium versus glycopyrronium: RR 0.88 (95% CI 0.77 to 0.99); p=0.038 (12% reduction; NNT=8b)

Indacaterol/glycopyrronium versus tiotropium: RR 0.90 (95% CI 0.79 to 1.02); p=0.096 (NS)

Other selected secondary outcomes

Annualised rate of severe exacerbations

0.09 (95% CI 0.07 to 0.13)

0.12 (95% CI 0.09 to 0.16)

0.08 (95% CI 0.06 to 0.11)

Indacaterol/glycopyrronium versus glycopyrronium: RR 0.81 (95% CI 0.60 to 1.10); p=0.18 (NS)

Indacaterol/glycopyrronium versus tiotropium: RR 1.16 (95% CI 0.84 to 1.61); p=0.36 (NS)

Annualised rate of mild exacerbations

2.51 (95% CI 2.25 to 2.80)

2.96 (95% CI 2.66 to 3.29)

2.98 (95% CI 2.68 to 3.32)

Indacaterol/glycopyrronium versus glycopyrronium: RR 0.85 (95% CI 0.75 to 0.96); p=0.0072

Indacaterol/glycopyrronium versus tiotropium: RR 0.84 (95% CI 0.75 to 0.95); p=0.0052

Annualised rate of all exacerbations

3.44 (95% CI 3.15 to 3.75)

4.04 (95% CI 3.71 to 4.40)

4.02 (95% CI 3.69 to 4.38)

Indacaterol/glycopyrronium versus glycopyrronium: RR 0.85 (95% CI 0.77 to 0.94); p=0.0012

Indacaterol/glycopyrronium versus tiotropium: RR 0.86 (95% CI 0.78 to 0.94); p=0.0017

Rescue salbutamol use (change from mean baseline value puffs/day [SE])

−2.3 (0.13)

−1.5 (1.3)

−1.5 (1.3)

Baseline values ranged from 5.5 to 5.7 puffs/day across groups

LSM differences:

indacaterol/glycopyrronium versus glycopyrronium: −0.81; p<0.0001;

indacaterol/glycopyrronium versus tiotropium: −0.76; p<0.0001

Safetyc

n=729

n=740

n=737

Patients reporting serious adverse events

22.9% (167/729)

24.2% (179/740)

22.4% (165/737)

Statistical significance not reported

Patients reporting worsening COPD as a serious adverse event

14.7% (107/729)

15.7% (116/740)

11.8% (87/737)

Statistical significance not reported

Patients reporting adverse events leading to discontinuation

8.1% (59/729)

9.1% (67/740)

6.4% (47/737)

Statistical significance not reported

Patients discontinuing treatment because of unsatisfactory therapeutic effect

2.5% (18/729)

4.3% (32/740)

5.2% (38/737)

Statistical significance not reported

Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; LSM, least square mean; NNT, number needed to treat; NS, not significant; p, p value; RR, relative risk; SE, standard error.

a All randomised patients who received at least 1 dose of study drug, analysed according to the treatment they were randomly assigned to receive. This definition was later modified to exclude 9 patients from 1 site that did not meet good clinical practice standards.

b NNT quoted from the European public assessment report for indacaterol/glycopyrronium.

c All patients receiving a least 1 dose of study drug whether randomly assigned to a treatment group or not, according to the treatment they received, excluding the 9 patients from 1 site that did not meet good clinical practice standards.

BLAZE (Mahler et al. 2013)

  • Design: randomised, double-blind, double-dummy, placebo-controlled 3-period crossover study undertaken at 42 centres in Belgium, Canada, Germany, Spain and the UK, comparing the effects of indacaterol/glycopyrronium with placebo (primary objective) and tiotropium (secondary objective) on dyspnoea over 6 weeks.

  • Population: 247 adults aged 40 years or over (mean age 62.8 years) who were current or ex-smokers with a smoking history of at least 10 pack-years. Patients had moderate or severe stable COPD (stage II or III according to 2009 GOLD criteria); post-bronchodilator FEV1 at screening of at least 30% and less than 80% predicted; post-bronchodilator FEV1/FVC less than 0.70; and a modified Medical Research Council (mMRC) dyspnoea scale grade of at least 2. Most patients (70%) did not report any exacerbations in the previous year. Use of long-acting bronchodilators or short-acting muscarinic antagonists (SAMAs) was not permitted during the study, but treatment with ICS was continued, with patients using combined LABA/ICS switched to equivalent ICS monotherapy.

  • Intervention and comparison: after a 14-day screening period, patients were randomised to 1 of 6 treatment sequences, to receive each of the following 3 treatments for a block of 6 weeks, with each 6-week treatment separated by a 2-week washout period (it is unclear whether allocation was concealed):

    • indacaterol/glycopyrronium 110/50 micrograms once daily (using the Breezhaler)

    • tiotropium 18 micrograms once daily (using Spiriva Handihaler)

    • placebo (matching double-dummy placebos were used for both active treatments to maintain study blinding).

  • Outcome: the primary (indacaterol/glycopyrronium compared with placebo) and key secondary efficacy end points (indacaterol/glycopyrronium compared with tiotropium) were the improvements in patient-reported dyspnoea scores after 6 weeks' treatment, as assessed using a computerised version of the Baseline and Transition Dyspnoea Indices (BDI/TDI). These results are presented in table 3, as are selected safety outcomes. Other secondary outcomes included measures of lung function, use of rescue medications, night-time awakenings due to symptoms, and symptom scores.

Table 3 Summary of BLAZE : Mahler et al. (2013)

Indacaterol/

glycopyrronium

(110/50 micrograms, once daily)

Tiotropium

(18 micrograms, once daily)

Placebo

Analysis

Randomised

247 patients were randomised to 1 of the 6 crossover treatment sequences with overall exposure to treatment as shown below

Efficacya

n=223

n=220

n=218

Primary outcome and key secondary outcome

Dyspnoea (TDI total score after 6 weeks' treatmentb)

0.88

0.39

−0.49

LSM differences:

indacaterol/glycopyrronium versus placebo: 1.37 (95% CI 0.95 to 1.79); p<0.001

Indacaterol/glycopyrronium versus tiotropium: 0.49 (95% CI 0.07 to 0.91); p=0.021

Safetyc

n=223

n=220

n=218

Patients reporting adverse eventsd

35.0% (78/223)

35.5% (78/220)

39.4% (86/218)

Statistical significance not reported

Patients reporting serious adverse eventsd

2.7% (6/223)

2.7% (6/220)

2.3% (5/218)

Statistical significance not reported

Discontinuations due to adverse eventsd

4.9% (11/223)

5.5% (12/220)

4.1% (9/218)

Statistical significance not reported

Abbreviations: CI, confidence interval; LSM, least squares mean; p, p value; TDI, Transition Dyspnoea Index.

a All randomised patients who received at least 1 dose of study drug.

b Change in dyspnoea is measured by the TDI total score, which assesses change from the baseline dyspnoea index score for 3 domains. A score of −3 indicates major deterioration in a domain, a score of +3 indicates major improvement. Domain scores are then added together to give the total score. A total score of 1 unit on the TDI is considered to be the minimal clinically important difference.

c All patients who received at least 1 dose of study drug, whether or not they were randomised.

d Adverse events, serious adverse events and discontinuations occurring in the corresponding 6-week drug treatment period.

Additional studies

The results of these studies are briefly summarised in the clinical effectiveness section of this evidence summary.

SHINE (Bateman et al. 2013)

  • Design: 26-week, multicentre, placebo-controlled RCT, comparing the effects of indacaterol/glycopyrronium with indacaterol, glycopyrronium and placebo (all administered double-blind) and tiotropium (administered open-label) on trough FEV1.

  • Population: 2144 adults recruited from centres in Europe, North and South America, Asia, Australia and South Africa who were aged 40 years or over (mean age about 64 years) and were current or ex-smokers with a smoking history of at least 10 pack-years. Patients had moderate or severe stable COPD (stage II or III according to GOLD 2008 criteria); post-bronchodilator FEV1 at screening of at least 30% and less than 80% predicted; and post-bronchodilator FEV1/FVC less than 0.70. Patients using fixed-dose ICS/LABA at study entry were switched to an equivalent dose of ICS monotherapy. Most patients (75%) did not report any exacerbations in the previous year.

  • Intervention and comparison: participants were randomised 2:2:2:2:1 to the following 5 treatments (all delivered using the Breezhaler except tiotropium, which was delivered using the Spiriva Handihaler), taken once daily in the morning:

    • indacaterol/glycopyrronium 110/50 micrograms (n=475)

    • indacaterol 150 micrograms (the lower of the 2 licensed doses for single-component indacaterol; n=477)

    • glycopyrronium 50 micrograms (n=475)

    • tiotropium 18 micrograms (n=483)

    • placebo (n=234).

  • Outcomes: the primary outcome was trough FEV1 (measured at 23.25 and 23.75 hours post-dose) at week 26 for indacaterol/glycopyrronium compared with its individual components. Secondary outcomes included trough FEV1 compared with placebo and tiotropium, changes in dyspnoea and health status scores, use of rescue medication, other lung function end points, patient-reported symptoms and adverse events.

ILLUMINATE (Vogelmeier et al. 2012)

  • Design: 26-week, multicentre, double-blind, double-dummy RCT, comparing the effects of indacaterol/glycopyrronium and fluticasone/salmeterol on lung function.

  • Population: 523 adults from 93 centres in 10 countries who were aged 40 years or over (mean age 63.3 years) and were current or ex-smokers with a smoking history of at least 10 pack-years. Patients had moderate or severe stable COPD (stage II or III according to GOLD 2009 criteria); post-bronchodilator FEV1 between 40% and 80% predicted; and post-bronchodilator FEV1/FVC less than 0.70 at screening. Patients with a history of a COPD exacerbation needing treatment with antibiotics, systemic corticosteroids or hospitalisation in the previous 12 months were excluded. Long-acting COPD maintenance therapy (LAMAs, LABAs, ICS and LABA/ICS) was withdrawn in a washout period.

  • Intervention and comparison: participants were randomised in a 1:1 ratio to:

    • indacaterol/glycopyrronium 110/50 micrograms once daily (using the Breezhaler; n=258)

    • fluticasone/salmeterol 500/50 micrograms twice daily (using the Seretide 500 Accuhaler; n=264)

  • Outcomes: the primary outcome was the standardised area under the curve from 0 to 12 hours post-dose for FEV1 (FEV1 AUC0-12h) at week 26. Secondary outcomes included other measures of lung function, changes in dyspnoea and health status scores, use of rescue medication and adverse events.

ENLIGHTEN (Dahl et al. 2013)

  • Design: 52-week, multicentre, double-blind, placebo-controlled RCT, comparing the safety and efficacy of indacaterol/glycopyrronium with placebo.

  • Population: 339 adults from centres in Europe, Canada, Asia and South Africa who were aged 40 years or over (mean age 62.6 years) and were current or ex-smokers with a smoking history of at least 10 pack-years. Patients had moderate or severe stable COPD (stage II or III according to GOLD 2008 criteria); post-bronchodilator FEV1 at least 30% and less than 80% predicted; and post-bronchodilator FEV1/FVC less than 0.70 at screening). Use of LABAs, LAMAs and SAMAs was not permitted during the study but ICS use was maintained.

  • Intervention and comparison: participants were randomised in a 2:1 ratio to once-daily treatment with :

    • indacaterol/glycopyrronium 110/50 micrograms (n=226)

    • placebo (n=113).

  • Outcomes: the primary outcome was the frequency of treatment-emergent adverse events. Secondary outcomes included lung function, use of rescue medication and symptom scores.

Clinical effectiveness

Exacerbations

In SPARK, indacaterol/glycopyrronium statistically significantly reduced the annualised rate of moderate to severe exacerbations (the primary outcome) in people with severe or very severe chronic obstructive pulmonary disease (COPD) by 12% compared with glycopyrronium (relative risk [RR] 0.88, 95% confidence interval [CI] 0.77 to 0.99, p=0.038). A non-significant reduction of 10% was seen in this outcome between indacaterol/glycopyrronium and open-label tiotropium (RR 0.90, 95% CI 0.79 to 1.02, p=0.096).

In the European public assessment report for indacaterol/glycopyrronium, the 12% reduction was considered to be 'very small' and not supportive of the manufacturer's requested indication of 'exacerbation reduction'. The full NICE guideline on COPD considers a relative reduction in the risk of exacerbations of 20% or more to be clinically important.

The annualised rate of severe exacerbations (defined as worsening symptoms needing admission to hospital or emergency treatment) was not statistically significantly reduced with indacaterol/glycopyrronium compared with either glycopyrronium or tiotropium (see table 2 for details).

Dyspnoea

In BLAZE, after 6 weeks' treatment, indacaterol/glycopyrronium statistically significantly improved dyspnoea scores (the primary outcome) in people with moderate or severe COPD compared with placebo (Transition Dyspnoea Index [TDI] total score +0.88 compared with −0.49, p<0.001). The mean difference (but not the lower limit of the 95% CI value) exceeded the 1 point improvement considered to be clinically important (least squares mean [LSM] difference 1.37, 95% CI 0.95 to 1.79; p<0.001). The dyspnoea score for indacaterol/glycopyrronium was also statistically significantly higher than for tiotropium (TDI score +0.88 compared with +0.39; LSM difference 0.49, 95% CI 0.07 to 0.91; p=0.021). However, this difference is unlikely to be clinically important.

In SHINE, dyspnoea (TDI total) scores were statistically significantly improved with indacaterol/glycopyrronium at week 26 compared with placebo (LSM difference 1.09, p<0.001) and open-label tiotropium (LSM difference 0.51, p=0.007), but not compared with single-component indacaterol or glycopyrronium. Improvements in dyspnoea scores were seen with indacaterol/glycopyrronium in the ILLUMINATE study, compared with fluticasone/salmeterol (LSM difference at week 26 0.76, p=0.0031).

Health status

In SPARK, health status (St. George's Respiratory Questionnaire [SGRQ] total) scores were statistically significantly lower (improved) with indacaterol/glycopyrronium at all time points assessed (12, 24, 38, 52 and 64 weeks). LSM differences ranged from −1.7 to −3.1 compared with glycopyrronium and open-label tiotropium (p<0.05 in all comparisons). According to the full NICE guideline on COPD, a difference of 4 units is considered to be the minimum clinically important difference in the SGRQ total score.

In SHINE, SGRQ total scores at week 26 were statistically significantly lower (improved) with indacaterol/glycopyrronium compared with placebo (LSM difference −3.01; p=0.002) and open-label tiotropium (LSM difference −2.13; p=0.009). There were no significant improvements with the other treatments compared with placebo. In ILLUMINATE, there was no significant difference in health status scores between indacaterol/glycopyrronium and fluticasone/salmeterol.

Use of rescue salbutamol

In SPARK, use of rescue salbutamol was significantly reduced by 2.3 puffs/day in people receiving indacaterol/glycopyrronium compared with 1.5 puffs/day in people receiving glycopyrronium or open-label tiotropium, a difference of approximately 0.8 puffs/day (p<0.0001).

In BLAZE, people taking indacaterol/glycopyrronium used significantly less rescue medication and had a significantly higher percentage of days with no rescue medication use compared with those taking placebo (p<0.001 for both) or tiotropium (p=0.002 and p<0.001 respectively).

In the SHINE, ILLUMINATE and ENLIGHTEN studies, treatment with indacaterol/glycopyrronium was associated with small but statistically significant reductions in the use of rescue salbutamol compared with placebo and active comparators.

Night-time awakenings and daytime symptoms

In BLAZE, the percentage of nights with no awakenings over 6 weeks was significantly higher for indacaterol/glycopyrronium compared with placebo (p<0.001). The percentage of days with no daytime symptoms was also significantly higher with indacaterol/glycopyrronium compared with placebo (p=0.001). However, for both assessments, indacaterol/glycopyrronium was not statistically significantly better than tiotropium.

In ENLIGHTEN, compared with placebo, people receiving indacaterol/glycopyrronium reported an increased percentage of days with no daytime symptoms (p=0.012), and days able to perform usual daily activities (p=0.028). The number of nights with no night-time awakenings did not differ between the groups.

Lung function

In SPARK, trough forced expired volume in 1 second (FEV1) was statistically significantly higher with indacaterol/glycopyrronium at all assessments (continued to 64 weeks) compared with glycopyrronium (differences 70 to 80 ml; p<0.0001) and open-label tiotropium (differences 60 to 80 ml; p<0.0001). The full NICE guideline on COPD considers a change in FEV1 of 100 ml or more to be clinically important.

In BLAZE, indacaterol/glycopyrronium provided statistically significant improvements in lung function, with higher post-dose FEV1, standardised area under the curve from 0 to 4 hours (FEV1 AUC0-4h) compared with placebo and tiotropium at day 1 and week 6 (all p<0.001)

In SHINE (1 of the 2 studies investigating lung function as a primary outcome), indacaterol/glycopyrronium statistically significantly improved trough FEV1 compared with indacaterol, glycopyrronium, open-label tiotropium and placebo at week 26 (LSM differences 70 ml, 90 ml, 80 ml and 200 ml respectively; p<0.001 in all comparisons). In the other study, ILLUMINATE, FEV1 AUC0-12h at week 26 was significantly higher with indacaterol/glycopyrronium compared with salmeterol/fluticasone (LSM difference 138 ml, 95% CI 100 ml to 176 ml; p<0.0001).

In ENLIGHTEN, pre-dose FEV1 at week 52 was increased by 189 ml with indacaterol/glycopyrronium compared with placebo (p<0.001).

Safety

The summary of product characteristics for Ultibro Breezhaler reports that up to 15 months' treatment with indacaterol/glycopyrronium showed similar adverse reactions to those observed when people were treated with each drug individually. The safety profile of indacaterol/glycopyrronium is characterised by typical anticholinergic and beta-adrenergic symptoms. Other most common adverse reactions (reported in at least 3% of people and also greater than placebo) are cough and oropharyngeal pain (including throat irritation).

In the 52-week safety study, ENLIGHTEN, statistical analyses of the primary outcome (incidence of all adverse events) in the indacaterol/glycopyrronium and placebo groups were not presented, although the overall incidence does appear to be similar (130/225 [57.8%] and 64/113 [56.6%] respectively). Thirteen people receiving indacaterol/glycopyrronium (5.8%) had an adverse event leading to discontinuation of the study drug compared with 7 people receiving placebo (6.2%; p value not reported).

In ENLIGHTEN, some respiratory-related adverse events occurred in greater numbers of people in the indacaterol/glycopyrronium treatment group compared with the placebo group, including cough (8.0% compared with 6.2%) and lower respiratory tract infections (6.7% compared with 3.5%; p values not reported). Significantly more people in the indacaterol/glycopyrronium group experienced pneumonia (8 compared with 0; p=0.074). In additional post-hoc analyses to investigate these differences, rates of pneumonia and respiratory-related serious adverse events were found not to be statistically significantly different. The authors suggest that the numerical difference for some adverse events may, in part, be explained by demographic imbalances at baseline, with greater numbers of people in the indacaterol/glycopyrronium group with severe COPD and using ICS. The European public assessment report for indacaterol/glycopyrronium indicates this imbalance relates to stratification of people based only on their smoking status, in accordance with regulatory guidance on clinical trials for COPD treatments. Also, a higher discontinuation rate was seen in the placebo group leading to a healthier population in that group.

In ILLUMINATE, the overall incidences of adverse events and serious adverse events were similar for indacaterol/glycopyrronium and fluticasone/salmeterol treatment groups (55.4% compared with 60.2%, and 5.0% compared with 5.3%, respectively; p values not reported). Reports of pneumonia, an adverse event associated with long-term ICS, were numerically higher in the fluticasone/salmeterol treatment group (4/159 compared with 0/143), although no statistical comparisons were provided.

Evidence strengths and limitations

SPARK demonstrated that indacaterol/glycopyrronium statistically significantly reduced the annualised rate of all exacerbations compared with glycopyrronium and open-label tiotropium alone, and of moderate to severe exacerbations compared with glycopyrronium. However, the relative reductions were all below the 20% considered in the full NICE guideline on COPD to be clinically important. In addition, the rate of severe exacerbations was not reduced compared with either active comparator. The European Medicines Agency concluded that the benefits of indacaterol/glycopyrronium in terms of reducing exacerbations were not sufficiently proven to grant a license for this indication.

Although there is evidence of a statistically significant improvement compared with placebo in the SHINE study, the clinical significance of improvements in trough FEV1 for indacaterol/glycopyrronium compared with glycopyrronium alone, open-label tiotropium or the lower licensed dose of indacaterol (150 micrograms) is unclear because the differences did not meet the 100 ml difference considered in the full NICE guideline on COPD to be clinically important. Similar results were seen in SPARK.

Interpretation of both SPARK and SHINE, within the context of the NICE pathway on COPD, is complicated because many participants (75% in SPARK and 57% in SHINE) used ICS. This meant that many people in the indacaterol/glycopyrronium group received triple therapy (LAMA/LABA plus ICS), and were compared with a LAMA treatment arm in which a substantial number of people received dual therapy with a LAMA and an ICS. LAMA/ICS is not recommended in the NICE clinical guideline on COPD because of the paucity of evidence. Furthermore, no single-component ICS inhalers are licensed for treating COPD in the UK.

According to the European public assessment report for indacaterol/glycopyrronium, the improvement in FEV1 AUC0-12h of approximately 140 ml seen in ILLUMINATE is not unexpected because indacaterol/glycopyrronium contains 2 bronchodilators compared with 1 in fluticasone/salmeterol. In addition, the eligibility criteria for ILLUMINATE permitted enrolment of people with less severe COPD (post-bronchodilator FEV1 between 40% and 80% predicted) than that for which fluticasone/salmeterol is currently licensed in the UK (pre-bronchodilator FEV1 less than 60% predicted).

Although statistically significant improvements in several other outcomes have been reported for indacaterol/glycopyrronium (for example, dyspnoea scores, health status scores and use of rescue salbutamol) compared with some active comparators, the differences appear small and are of uncertain clinical importance. Nevertheless, the European Medicines Agency considered that, taking the overall results and the safety profile of glycopyrronium/indacaterol into account, a first-line indication for relieving symptoms in people with COPD is justified.

Patients are currently being recruited for a 52-week study comparing the effects of indacaterol/glycopyrronium and fluticasone/salmeterol on exacerbations in people with moderate to very severe COPD (ClinicalTrials.gov NCT01782326). In addition to important patient-oriented outcome data, this study is likely to provide better longer-term comparative safety data for the 2 treatments.