Evidence

Surveillance report 2016 – Chronic obstructive pulmonary disease in over 16s: diagnosis and management (2010) NICE guideline CG101

Commentary on selected new evidence

Commentary on selected new evidence

With advice from topic experts we selected 3 studies for further commentary.

Managing stable COPD – Inhaled combination therapy

We selected a network meta‑analysis by Oba et al. (2016) for a full commentary because it compared dual bronchodilator therapy (LAMA plus LABA) with LAMA or LABA monotherapy. It was considered that these comparisons may affect guideline recommendations.

What the guideline recommends

Section 1.2.2.6 of CG101 states:

"In people with stable COPD who remain breathless or have exacerbations despite using short‑acting bronchodilators as required, offer the following as maintenance therapy:

  • if FEV1 ≥ 50% predicted: either LABA or LAMA

  • if FEV1 < 50% predicted: either LABA with an inhaled corticosteroid (ICS) in a combination inhaler, or LAMA."

Methods

The network meta‑analysis by Oba et al. (2016) pooled data from 23 randomised controlled trials, including 27,172 patients, comparing any LAMA plus LABA combination with LAMAs alone, LABAs alone or placebo.

Published and unpublished studies that enrolled patients with moderate to severe COPD, without an acute or recent exacerbation, were included. Additionally, included studies had to be at least 12 weeks in duration. The LAMA group comprised aclidinium, glycopyrronium, umeclidinium and tiotropium while the LABA group comprised formoterol, indacaterol, olodaterol, salmeterol and vilanterol. The LAMA plus LABA group included of any combination of the aforementioned LAMAs and LABAs. Outcome measures included change from baseline in FEV1 (in litres), St Georges Respiratory Questionnaire (SGRQ) scores and Transitional Dyspnoea Index (TDI) scores, as well as COPD exacerbation, mortality, serious adverse event, and cardiac serious adverse event rates.

Authors performed a network meta‑analysis, using Bayesian Markov chain Monte Carlo approach, as well as a pairwise comparison meta‑analysis.

Results

Change from baseline in trough FEV1 data were available for 4 trials (n=4836 patients) with 12 month follow‑up periods. LAMA plus LABA combinations were ranked as the most effective class in improving FEV1 measurements followed by LAMAs then LABAs. When dual therapy was compared against placebo, the mean difference in FEV1 measurements was 243 ml (95% Credible Interval [CrI]; 139 to 351) in favour of dual therapy. The mean difference in trough FEV1 between the LAMA plus LABA group and the LAMA monotherapy group was 73 ml (95% CrI 43 to 149), in favour of the dual therapy. The mean difference in trough FEV1 between the LAMA plus LABA group and the LABA monotherapy group was 104 ml (95% CrI 84 to 126), in favour of the dual therapy. Authors state that no significant differences in changes in trough FEV1 were observed between drug classes.

Change from baseline data in SGRQ scores was available from 9 trials (n=12,716) at 6 month follow‑up. SGRQ scores range from 0 to 100 with lower scores indicating better quality of life. Dual therapy was ranked highest followed by LABAs and LAMAs in all SGRQ outcomes. LAMA plus LABA combinations reduced SGRQ scores by a mean of 4.1 points (95% CrI 2.3 to 5.9) compared with placebo. When compared with LAMA monotherapy, LAMA plus LABA combinations reduced SGRQ scores by a mean of 1.6 (95% CrI 0.5 to 2.8) points. Finally, dual therapy reduced SGRQ scores by a mean of 1.1 (95% CrI 0.4 to 2.5) compared with LABA monotherapy. Data on SGRQ responders (patients who had at least a 4 point improvement in scores – the minimal clinically important improvement) were available from 12 trials at 6 month follow‑up. A significantly greater proportion of SGRQ responders was reported in the LAMA plus LABA group compared with the LAMA monotherapy group (Odds ratio [OR] 1.23; 95% CrI 1.06 to 1.39) and LABA monotherapy group (OR 1.24; 95% CrI 1.11 to 1.36).

Change from baseline data in TDI scores was available from 8 trials (n=14,568) at 6 month follow‑up. TDI scores range from −9 to 9 with lower scores indicating a more severe dyspnoea. Dual bronchodilation was ranked highest followed by LABAs and LAMAs. At 3 month follow‑up, dual bronchodilation yielded a significant improvement in TDI scores compared with placebo (mean difference 1.17; 95% CrI 0.96 to 1.38), LAMAs (mean difference 0.40; 95% CrI 0.26 to 0.53), and LABAs (mean difference 0.35; 95% CrI 0.24 to 0.47). All observations were statistically significant. Data on TDI responders (patients who had at least a 1 point improvement in scores – the minimal clinically important improvement) were available from 7 trials at 6 month follow‑up. A significantly greater proportion of TDI responders was reported in the LAMA plus LABA group compared with the LAMA monotherapy group (OR 1.34; 95% CrI 1.16 to 1.56) and LABA monotherapy group (OR 1.30; 95% CrI 1.13 to 1.48).

COPD exacerbation data were available from 16 trials (n=18,224) for moderate‑to‑severe exacerbations and 19 trials (n=25,401) for severe exacerbations (follow‑up period was not specified). Dual bronchodilation was associated with significantly fewer moderate‑to‑severe exacerbations compared with placebo (Hazard ratio [HR] 0.66; 95% CrI 0.57 to 0.77) and LABA alone (HR 0.82; 95% CrI 0.73 to 0.93), but not when compared with LAMA alone (HR 0.92; 95% CrI 0.84 to 1.00). No significant differences in the occurrence of severe exacerbations were observed when LAMA plus LABA combinations were compared with placebo, LABAs or LAMAs.

Due to a considerable degree of overlap in CrIs and rankings, authors stated that there were no significant differences in mortality, serious adverse event, and cardiac serious adverse event rates between groups.

Strengths and limitations

Strengths

The network meta‑analysis included a number of strengths:

  • The study was robust in terms of study identification, selection, data extraction and data synthesis.

  • Included studies were consistent in their key inclusion and exclusion criteria. Generally, patients who were over 35 years, with a diagnosis of COPD in accordance with the American Thoracic Society‑European Respiratory society or GOLD guidelines, a smoking history of at least 10 pack‑years, and moderate to severe COPD (FEV1 ranging from 30 to 70% predicted) were included.

  • Validated outcome measure which have previously been used to assess COPD (SGRQ and TDI scores) were assessed.

Limitations

The following limitations were identified:

  • Authors pooled data on different medications within each drug class to make comparisons.

  • The number of included trials varied according to each outcome measure and follow‑up period: for example, trough FEV1 data were available for less than 20% of patients at 12 month follow‑up.

  • The concomitant use of a fixed dose inhaled corticosteroids (ICS) was permitted in most studies. Authors do not provide any details as to what corticosteroids were used.

  • Authors state that there was no evidence of inconsistency within the network but highlight that there may not have been sufficient power to detect inconsistency.

Impact on guideline

The study provides evidence of benefits of dual bronchodilation over LABA or LAMA monotherapy in patients with moderate to severe COPD. Topic experts felt that the new evidence highlights advantages of using dual bronchodilation over monotherapy, in patients with moderate to severe COPD, and suggested that there is a need to review the algorithm for inhaled therapy

Managing stable COPD – Inhaled combination therapy

We selected a network meta‑analysis by Tricco et al. (2015) for a full commentary because it provides comparative evidence of different long‑acting inhaled therapies (LAMA or LABA monotherapy versus dual bronchodilator therapy with or without ICS) that may impact on guideline recommendations.

What the guideline recommends

Section 1.2.2.6 of CG101 states:

"In people with stable COPD who remain breathless or have exacerbations despite using short‑acting bronchodilators as required, offer the following as maintenance therapy:

  • if FEV1 ≥ 50% predicted: either LABA or LAMA

  • if FEV1 < 50% predicted: either LABA with an inhaled corticosteroid (ICS) in a combination inhaler, or LAMA."

Section 1.2.2.7 recommends:

"In people with stable COPD and an FEV1 ≥ 50% who remain breathless or have exacerbations despite maintenance therapy with a LABA:

  • consider LABA+ICS in a combination inhaler

  • consider LAMA in addition to LABA where ICS is declined or not tolerated."

Methods

Tricco et al (2015) performed a network meta‑analysis of 208 randomised controlled trials, including 134,692 patients, which aimed to compare the safety and effectiveness of various combinations of LAMAs, LABAs and ICSs for managing COPD.

Randomised controlled trials including adults with COPD who received long‑acting inhaled agents in any combination compared against each other or placebo, were eligible for inclusion. The majority of patients (61%) had moderate‑to‑severe COPD. Published studies were included regardless of duration of follow‑up, date of dissemination or publication status. Unpublished studies conducted after 2004 were included.

A random‑effects network meta‑analysis was performed because authors assumed that treatment effects were heterogeneous across included studies. A design‑by‑treatment interaction model was used because included studies had varying numbers of study arms. The primary outcome of interest was the proportion of patients with moderate‑to‑severe COPD exacerbations. Secondary outcomes included mortality and the occurrence of pneumonia.

Results

A network meta‑analysis of 20 of the 208 identified randomised controlled trials, including 26,141 patients, showed that the following medications were more effective than placebo in reducing the risk of moderate‑to‑severe COPD exacerbations: tiotropium, salmeterol, indacaterol, formoterol plus budesonide, salmeterol plus fluticasone, glycopyrronium plus indacaterol, tiotropium plus salmeterol plus fluticasone and tiotropium plus formoterol plus budesonide. Of all interventions, tiotropium plus formoterol plus budesonide (OR 0.23; 95% CrI 0.14 to 0.40) and glycopyrronium plus indacaterol (OR 0.48; 95% CrI 0.36 to 0.64) were found to be the most effective agents in reducing the risk of exacerbations when compared with placebo. These results were statistically significant.

A network meta‑analysis of 88 randomised controlled trials, including 97,526 patients, indicated that the ultra LABA, AZD3199 (OR 0.46; 95% CrI 0.02 to 10.32), and the combination of tiotropium and formoterol (OR 0.66; 95% CrI 0.08 to 5.18) were the most effective agents in reducing the risk of mortality compared to placebo. However, these results were not statistically significant. Only fluticasone plus salmeterol significantly decreased the risk of mortality when compared with placebo (OR 0.78; 95% CrI 0.63 to 0.96). Analysis also revealed that formoterol increased the risk of mortality when compared with fluticasone plus salmeterol (OR 1.64; 95% CrI 1.01 to 2.67). Furthermore, fluticasone plus salmeterol resulted in a reduced risk of mortality compared fluticasone alone (OR 0.75; 95% CrI 0.60 to 0.94).

To assess the risk of pneumonia, authors pooled data from 54 randomised controlled trials including 61,551 patients. A treatment hierarchy was obtained using the Surface Under the Cumulative Ranking (SUCRA) curve analysis. This approach allowed for ranking of interventions according to the probability of being the least effective in reducing the risk of pneumonia. The most harmful agents were salmeterol plus fluticasone (89% probability), vilanterol plus fluticasone (88%) and fluticasone alone (82%). Unfortunately, authors did not report OR data from the network meta‑analysis in the manuscript.

Strengths and limitations

Strengths

The network meta‑analysis included a number of strengths:

  • The study was robust in terms of study identification, selection, data extraction and data synthesis.

  • Since the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria have changed over time, a clinician reviewed all of the included studies to ascertain the average COPD severity of patients included in each randomised controlled trial.

  • Before performing the analysis, authors assessed whether any systematic differences were prevalent in the distribution of potential treatment effect modifiers across treatment comparisons in the network.

  • Consistency of the data was extensively assessed. If inconsistency was identified, the loop‑specific model was used to identify local inconsistency within the networks.

Limitations

The following limitations were identified:

  • Only some trials contributed to each meta‑analysis.

  • Many of the included trials were at a high risk of bias for many of the Cochrane risk‑of‑bias criteria.

  • Investigators did not adjust for differences in treatment doses and durations as they were inconsistently reported across included studies. The assumption was made that these factors had an equal impact on the treatment effect.

Impact on guideline

The study provides strong evidence that dual bronchodilator therapy (specifically, glycopyrronium plus indacaterol) is more effective than LAMA or LABA monotherapy and LABA plus ICS therapy in reducing the risk of moderate‑to‑severe COPD exacerbations. Topic experts agreed with the evidence and suggested that dual bronchodilation produces greater improvements in FEV1 and dyspnoea than LABA plus ICS combinations in patients with severe COPD. However, new intelligence suggested that some LABA plus ICS combinations are becoming cheaper due to expiring patents. A review of the clinical and cost effectiveness of inhaled therapies may be needed to identify whether recommendations in the guideline should remain.

Managing stable COPD – Oral prophylactic antibiotic therapy

We selected a systematic review by Ni et al. (2015) for a full commentary because it provides evidence on antibiotic prophylaxis that may impact on guideline recommendations.

What the guideline recommends

Recommendation 1.2.3.12 in CG101 states that 'there is insufficient evidence to recommend prophylactic antibiotic therapy in the management of stable COPD'.

Methods

Ni et al (2015) performed a systematic review of 9 randomised controlled trials, including 1,666 patients, which aimed to evaluate whether prophylactic macrolide therapy prevented acute exacerbations of COPD.

Investigators included randomised controlled trials comparing patients with COPD (severity not specified) who received macrolide prophylaxis and those who did not receive macrolides (controls). Studies were included if they enrolled adults older than 18 years with stable COPD, confirmed by lung function testing (FEV1/FV<70%, FEV1<80% predicted, and an increase in FEV1<12% or <200 millilitres). Prophylactic use of macrolides must have been administered orally, in appropriate daily doses, for a minimum of 3 months.

Primary outcome measures included the number of patients with 1 or more exacerbations and the rate of exacerbations per patient per year. Secondary outcome measures included hospitalisation rates, all‑cause mortality rates, quality of life scores and adverse event rates.

Results

Meta-analysis of 7 studies, including 1,614 patients, revealed that the proportion of patients with 1 or more exacerbations was significantly lower in the macrolide prophylaxis group than the control group (Risk ratio 0.7; 95% Confidence Interval [CI] 0.56 to 0.87; p<0.01; I2=66.43%).

Meta-analysis of 8 studies, including 1,582 patients, revealed that the rate of exacerbations per patient per year was significantly lower in the macrolide prophylaxis group than the control group (Rate ratio 0.58; 95% CI 0.43 to 0.78; p<0.01; I2=67.8%). Subgroup analysis, according to type of macrolide, failed to show that azithromycin reduced the rate of exacerbations when administered for 3 months (Rate ratio 0.46; 95% CI 0. 18 to 1.18; p=0.11; I2 not reported) but significantly reduced the rate of exacerbations when administered for 6 to 12 months (Rate ratio 0.82; 95% CI 0.76 to 0.90; p<0.01; I2 not reported).

No significant differences in the hospitalisation rates and all‑cause mortality rates were observed between groups (p values>0.05).

Meta-analysis of 4 studies, including 1,323 patients, assessed the impact azithromycin on SGRQ scores (scores range from by 0 to 100 with lower scores indicating better quality of life). Results indicated that SGRQ decreased by a mean of 2.12 points in patients who received azithromycin prophylaxis when compared to controls (95%CI 0.79 to 3.44; p=0.002; I2=0%). Authors state that the result was statistically significant but did not reach the threshold for clinical significance (a 4 or more point difference in SGRQ scores).

Meta-analysis of adverse event rates revealed that the occurrence of adverse events (not specified) was marginally higher in the macrolide prophylaxis group (OR 1.55; 95% CI 1.003 to 2.39; p=0.049; I2=15.04%).

Strengths and limitations

Strengths

The systematic review included a number of strengths:

  • Multiple databases were searched and additional approaches were used to identify as many relevant studies as possible.

  • Selection criteria were clear and specific, indicating that appropriate studies would have been included. Consistent diagnostic criteria were used in all included studies.

  • Investigators excluded studies which included patients with bronchiectasis, asthma, cystic fibrosis and other genetic diseases.

Limitations

The following limitations were identified:

  • I2 values for the exacerbation rate meta‑analyses were above 60%; indicating substantial heterogeneity.

  • Authors do not provide any definition of a COPD exacerbation and state that exacerbation definitions were not consistent between included studies. As a result, it is unclear if exacerbation rates were overestimated or underestimated.

  • The Jadad scale was used to assess the quality of RCTs included in the meta‑analysis. This tool is used to assess the quality of RCTs and does not explicitly assess risk of bias.

  • There is no indication that funnel plots were used to assess potential biases. Instead, authors state that Egger regression and subgroup analyses were performed.

Impact on guideline

Current recommendations state that there is insufficient evidence to recommend prophylactic antibiotic therapy in the management of stable COPD. The systematic review indicates some benefits of antibiotic prophylaxis delivered for 6 to 12 months. Topic experts highlighted that guidance on the long‑term use of antibiotics may be useful as macrolides are being used extensively in people with COPD, particularly for those with chronic bronchitis and exacerbations.


This page was last updated: