Design: investigator-initiated, double-blind, placebo-controlled randomised trial with bronchoscopy and ex vivo rhinovirus challenge (AZIMUNE)
Population: 40 adults with uncontrolled asthma, 16 per arm in the analysis
Primary outcome: rhinovirus-induced epithelial interferon-beta and interferon-lambda, and alarmin concentrations, at 12 weeks
Effect: increases from baseline in interferon-beta (p = 0.047) and interferon-lambda (p = 0.013) with azithromycin and not placebo; IL-33 reduced (p = 0.001); TSLP unchanged

AZIMUNE randomised 40 adults with uncontrolled asthma to azithromycin 500 mg three times weekly or placebo for 12 weeks, with bronchoscopy at baseline and week 12. Bronchial epithelial cells were cultured and infected with rhinovirus outside the body, and the interferon response measured.

Azithromycin increased rhinovirus-induced interferon-beta and interferon-lambda from baseline, and lowered interleukin-33, while thymic stromal lymphopoietin was unchanged. No equivalent change occurred on placebo.

The value of this is explanatory rather than practice-changing. Azithromycin's ability to reduce asthma exacerbations is already established from larger clinical trials; what has been missing is why, with proposals ranging from antibacterial to anti-inflammatory to motility effects. This points at a specific mechanism — the impaired epithelial interferon response that is characteristic of asthmatic airways, partially restored.

The dual action is the interesting part. Enhancing antiviral defence while lowering an alarmin that drives type 2 inflammation are two different things, and both run in the direction of fewer virus-triggered exacerbations.

What this does not do is change who should be offered maintenance azithromycin. Clinical outcomes improved only numerically here, in 40 patients, and the balance against macrolide resistance and QT effects rests on the larger clinical trials, not on this.

Do not widen maintenance azithromycin use on the basis of this; the indication rests on the clinical trials
Where azithromycin is used long term, check an ECG for QT interval and review hearing
Consider local macrolide resistance rates in the decision; in India they are high and the population cost is real
Reserve it for uncontrolled asthma despite optimised inhaled therapy and adherence, as before
Recognise virus-triggered exacerbation patterns in the history; they are the phenotype this mechanism predicts will benefit

Don't overread it

The endpoints were laboratory measures on cultured cells; clinical outcomes did not differ significantly from placebo in this trial.

The statistics, in plain English

The significant comparisons here are within-group changes from baseline in the azithromycin arm (p = 0.047 for interferon-beta, 0.013 for interferon-lambda, 0.001 for IL-33) rather than between-group differences against placebo — a weaker form of evidence, because a within-group change can reflect anything that happened over 12 weeks. With 16 patients analysed per arm, the trial had power for laboratory endpoints and none at all for clinical ones, which is why the clinical improvements are described as numerical.