AMAZES

Published

Gibson PG, et al. "Effect of azithromycin on asthma exacerbations and quality of life in adults with persistent uncontrolled asthma (AMAZES): A randomised, double-blind, placebo-controlled trial". Lancet. 2017. 390(10095):659-668.
PubMed • Full text

Clinical Question

In adult patients with symptomatic asthma who are currently using an inhaled corticosteroid or a long-acting bronchodilator, does oral azithromycin decrease the frequency of asthma exacerbations?

Bottom Line

Among adults with persistent symptomatic asthma despite inhaled corticosteroid or long-acting bronchodilator (~98% of participants were on both), azithromycin 3x/week for 48 weeks reduced the rate of asthma exacerbations and improved quality of life.

Major Points

Optimal medical therapy for adults with asthma includes inhaled corticosteroids and bronchodilators, but despite these interventions some patients have persistent symptoms. This led to the hypothesis that the antimicrobial and anti-inflammatory effects of azithromycin may reduce exacerbations and improve quality of life in these patients. Clinical data in the mid-2010s was too limited to inform the use of azithromycin among patients with persistent asthma.

Published in 2017, AMAZES enrolled 420 patients with persistent symptomatic asthma currently using inhaled corticosteroids or a long-acting bronchodilator (~98% were on both ICS and LABA). A 2-week run-in period excluded patients with unstable symptoms or who were not adherent to their medication regimen. The primary outcome was the number of moderate or severe exacerbations and asthma-related quality of life over the 48-week study period. The study found that azithromycin reduced the rate of moderate and severe asthma exacerbations over the study period (1.86 vs. 1.07 exacerbations per year) with fewer patients in the azithromycin group experiencing at least one exacerbation (44% vs. 61%). As expected, this difference was driven in part by fewer clinically diagnosed infections in patients treated with azithromycin. This benefit was observed in both eosinophilic and non-eosinophilic subtypes, an important distinction as the smaller (N=54) 2013 AZISAST trial only observed benefit with azithromycin in the non-eosinophilic subtype. On subgroup analysis compared the outcome by presence of pathogenic bacteria identified on induced sputum. There was a benefit for both bacteria-positive and bacteria-negative groups, but those with bacteria-positive sputum experienced a greater reduction in the primary outcome with azithromycin.

To assess if that azithromycin led to more resistant bacterial colonizations, sputum cultures were taken at prespecified time points throughout the study. End-of study cultures demonstrated a non-significant, greater proportion of resistant organisms in azithromycin-treated patients. (Participants of AZISAST assigned to azithromycin had more resistant organisms at the end of study.) There was also a modest improvement in quality of life among patients receiving azithromycin. The authors conclude that a subset of patients with persistent asthma symptoms despite inhaled corticosteroid and bronchodilator therapy may benefit from the additional of azithromycin.

The study's generalizability is somewhat limited, since smokers and patients with suboptimal adherence were excluded. Additionally the small sample size limits the integrity of secondary analyses such as the emergence of azithromycin-resistant organisms during therapy and the occurrence of QTc prolongation and cardiac arrhythmias while receiving azithromycin. It's laudable that the authors studied quality of life, but the improvement with azithromycin was modest. Ultimately it seems prudent to offer azithromycin to a carefully selected subset of patients rather than recommending it to all patients with persistent asthmatic symptoms despite optimal inhaler therapy.

Guidelines

As of June 2020, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter randomized, double-blind, placebo controlled parallel group trial.
• N=420
  • Azithromycin group (n=213)
  • Placebo group (n=207)
• Setting: 8 sites, presumably in Australia
• Enrollment: 2009-2015
• Follow up: 48 weeks
• Analysis: Intention-to-treat
• Primary outcome: Total number of asthma exacerbations (severe and moderate)

Population

Inclusion criteria

• ≥18 years of age with asthma defined by variable airflow from bronchodilator response, airway hyperresponsiveness, or increased peak flow variability.
• Symptomatic despite treatment with maintenance inhaled corticosteroids or long-acting bronchodilators
• Clinically stable with no recent infections, exacerbations, or changes in maintenance medications in the past 4 weeks
• Non-smoker

Exclusion Criteria

• Prior smoking more than 10 pack-years of smoking if carbon monoxide diffusing capacity was less than 70% of predicted value.
• Hearing impairment
• Prolonged QTc interval

Baseline Characteristics

From the placebo group.

• Demographics: Age 60 years, 58% female
• Medical comorbidities: Atopy 80%, former smoker 39% (7.5 PYH)
• Asthma details: Age at symptom onset 13, age diagnosed 20, ACQ6 score 1.55, AQLQ score 5.35, ED visits in prior year 0, unscheduled MD visits 1, PO steroid courses 1
  • Medications: ICS low dose 2%, ICS moderate dose 13%, ICS high dose 85%, LABA 99%, leukotriene agent 3%, LAMA 16%, slow release theophylline 3%, PO steroid 3%
• Anthropometrics: BMI 29
• Pre-B2 spirometry: FEV1 74%, FVC 83%, FEV₁/FVC 68%
• Sputum phenotype: Eosinophilic 46%, neutrophilic 15%, paucigranulocytic 33%, mixed 5%
• Serum eosinophils: 0.28x10⁹/L

Interventions

• After a screening visit, there was a 2 week run-in to confirm adherence based upon inhaler dose counters, diary records, questionnaires. Those who had adherence of >80% of doses and had stability in their symptoms (determined by change in ACQ6 <0.5 during run-in) were randomized to a group:
  • Azithromycin at 500 mg po 3x/week daily for 48 weeks
  • Placebo
• Induced sputum was collected prior to randomization at 5 of the sites and at the end of treatment.
• Safety included liver function testing and an ECG at screening, at 6 weeks, and at the end of treatment; those with QTc >480 ms were withdrawn.

Outcomes

Presented as placebo vs. azithromycin.

Primary Outcomes

Total number of asthma exacerbations (severe and moderate)

Severe exacerbations: Hospitalization due to asthma, ER visit requiring systemic corticosteroid, or >3 days of systemic corticosteroids of >10 mg or temporary increase in > 10mg for > 3 days.

Moderate exacerbations: ER visit not requiring systemic corticosteroids, increase in beta-2 agonist for at least 2 days, and increase in inhaled corticosteroids or antibiotics with a deterioration of asthma symptoms or both.

1.86 vs. 1.07 exacerbations/PY (IRR 0.59; 95% CI 0.47-0.74)

Absolute difference -0.46 (95% CI -0.79 to -0.14; P<0.0001)

Asthma quality of life at end of treatment

Comparing AQLQ mean scores between groups at the end of treatment

5.55 vs. 5.73 (adjusted mean 0.36; 95% CI 0.21 to 0.52; P=0.001)

Secondary Outcomes

Asthma control

Comparing ACQ6 mean scores between groups at the end of treatment

1.31 vs. 1.21 (adjusted mean -0.20; 95% CI -0.34 to -0.05)

Lung Function (pre-bronchodilator spirometry)

FEV₁

2.18 vs. 2.06 (adjusted mean -0.06; 95% CI -0.12 to -0.001)

Visual analogue for symptoms at end of treatment

Nasal symptoms: 3.46 vs. 2.95 (adjusted mean -0.87; 95% CI -1.72 to -0.03)

Breathlessness: 3.31 vs. 2.95 (adjusted mean -0.49; 95% CI -1.18 vs. 0.20)

Wheeze: 2.30 vs. 2.02 (adjusted mean -0.11; 95% CI -1.15 to 0.94)

Sputum production: 2.83 vs. 2.16 (adjusted mean -0.62; 95% CI -1.23 to -0.002)

Cough: 2.99 vs. 2.45 (adjusted mean -0.73; 95% CI -1.42 to -0.04)

Adverse Events

Serious

13% vs. 8%

Cardiac: 3% vs. 2%

Respiratory tract infections treated with antibiotics

31% vs. 20%

Treatment discontinuation due to adverse event

A more complete list is found in Table 4

5% vs. 7%

Nausea: 10% vs. 15%

Diarrhea: 19% vs. 34%

Subgroup Analyses

The primary outcome was similar by eosinophilic vs. non-eosinophilic subtype, ICS dose, frequency of exacerbations, and cough and sputum symptomatology.

Presence of bacterial pathogen on induced sputum

Bacteria-negative: IRR 0.61 (95% CI 0.52 to 0.72)

Bacteria-positive: IRR 0.39 (95% CI 0.22 to 0.69)

Interaction P<0.05

Criticisms

• Small population group was observed for study when compared to the amount of people who were screened for the study
• 20% of selected group withdrew from study
• Mean Adherence to trial was 80%, 20% of the population study was not adherent
• Study excluded patients who smoke, patients with hearing impairments and patients with a prolonged QT interval
• Too short to assess for antibiotic resistance development

Funding

National Health and Medical Research Council of Australia, John Hunter Hospital Charitable Trust