Macrolides, including azithromycin, possess antibacterial, anti-inflammatory, immunomodulatory and potentially anti-viral properties. The prolonged half-life of azithromycin allows convenient once-daily dosing, leading to it becoming the most commonly prescribed macrolide for acute respiratory infections. Azithromycin is also used increasingly in a broad range of chronic pulmonary disorders characterised by neutrophilic inflammation, such as cystic fibrosis, bronchiectasis, chronic obstructive pulmonary disease and poorly-controlled asthma. Nevertheless, evidence supporting macrolides in these conditions remains inadequate.
Early viral lower respiratory infections and their interactions with the respiratory microbiota have been linked to developing asthma. Three randomised controlled trials (RCTs) reported short-term azithromycin in children with acute respiratory infections reduced the risk of hospitalisation and delayed recurrent wheezing. In contrast, four RCTs of azithromycin in infants with bronchiolitis found it did not provide clinical benefits acutely or for 6-months afterwards. Similarly, in children with pneumonia, macrolides did not shorten hospital stay, even if atypical pathogens were identified.
Long-term azithromycin potentially confers benefit at every level of the vicious cycle hypothesis of chronic lower airway inflammatory diseases driven by infection. RCTs in cystic fibrosis and bronchiectasis report on average a 50% reduction in exacerbations when azithromycin is administered for 6-24 months, although whether this is sustained is unknown.
Azithromycin’s favourable pharmacokinetics also results in prolonged subinhibitory concentrations at carriage sites and selection of antibiotic resistant strains. Poor adherence increases the risk of dysbiosis and macrolide-resistance. While antibiotic-resistance is usually reversible once treatment is ceased, macrolide-resistant Staphylococcus aureus strains persist. The impact of macrolide-resistance at an individual and community level is uncertain. Macrolide-resistance in Haemophilus influenzae is associated with treatment failure in otitis media, while the effect on pneumococcal bacteraemic pneumonia is inconsistent, and whether it influences the course of chronic lower airway infections is unknown, although selecting macrolide-resistant mycobacteria in these patients is concerning.
Macrolides highlight the challenges of translating in-vitro observations into clinical trials and practice, while trying to avoid ‘therapeutic creep’ and unintended ‘off-target’ effects.