Bronchiectasis unrelated to cystic fibrosis (CF) is an heterogenous condition with many initiating events. It is now considered more common than thought previously, especially in developing countries , in socioeconomically disadvantaged Indigenous communities in affluent countries  and as a co-morbidity of other chronic pulmonary diseases, such as asthma [3, 4] and chronic obstructive pulmonary disease (COPD) . Bronchiectasis affects all age groups, including infants . Although the consequences of bronchiectasis are predominantly related to respiratory morbidity, there are also independent cardiovascular effects, such as coronary artery disease [7, 8], which are likely to worsen with the systemic inflammation arising from chronic pulmonary infection.
Over the past 15 to 20 years, it has been increasingly recognized that non-CF bronchiectasis is a major contributor to chronic respiratory morbidity [9, 10] and mortality [11, 12] worldwide [2, 13]. In our recently completed multicenter study of 346 Australian children newly referred for chronic cough and managed using a standardized protocol , 31 (9%) had radiologically proven bronchiectasis . In the USA, the average number of bronchiectasis-associated hospitalizations increased by 2 to 3% per year between 1993 and 2006, and during this period, the average age-adjusted annual hospitalization rate was 16.5 per 100,000 of the population . In some populations from affluent countries, the prevalence of bronchiectasis is one of the highest reported in the world, such as in Central Australia, where 1 in 68 Indigenous children are affected . Similarly, in New Zealand, the estimated national prevalence rates of bronchiectasis are much higher in Maori (1 in 2300) and Pacific Island (1 in 625) children than in New Zealand European (1 in 7440) children aged under 15 years . The minimum national incidence of non-CF bronchiectasis before the age of 15 years is 1 in 1,700 births, compared with 1 in 3,179 births for a diagnosis of CF in New Zealand . In the European Union and the USA, the prevalence of CF is 7.4 to 7.9 per 100,000 (whole population data) . Indeed, there are far more patients with bronchiectasis than patients with CF attending respiratory services globally . In the US, about 30,000  people have CF, whereas over 110,000 people have non-CF bronchiectasis . Moreover, the latter is likely to be an underestimate, as many cases of non-CF bronchiectasis are misdiagnosed. or co-exist with asthma [3, 4, 18] and COPD .
Effective clinical management reduces both short-term and long-term morbidity (and probably mortality) associated with bronchiectasis [1, 21–23]. Cohort data have shown that about 80% of newly diagnosed adults (non-smokers) with bronchiectasis reported symptoms dating back to childhood, and that the duration of chronic cough (the most common symptom of bronchiectasis ) was related (r = −0.51, P<0.001 in non-smokers) to lung function at diagnosis . Arguably, appropriate overall management and treatment of exacerbations (leading to a reduction in persistent symptoms) potentially prevents or reduces deterioration of chronic respiratory disease . Our study and a London-based retrospective study both found that with appropriate treatment in specialized centers, lung function improves and can be maintained [22, 23]. However, despite substantial improvements, those with poor lung function at diagnosis were still likely to have poor lung function 5 years later . We also found that the only significant predictor of decline in forced expiratory volume in 1 second (FEV1) was the frequency of hospitalized exacerbations, and that FEV1 percentage predicted declined by 1.95% with each hospitalized exacerbation . In addition to the biologic effects of respiratory exacerbations, these episodes also impair quality of life (QOL) and well-being. as shown by deteriorating scores for QOL and on the Depression, Anxiety and Stress Scale (DASS) during exacerbations . Taken together, as airway injury in children is superimposed upon the physiological changes involving lung growth and development [28, 29], we speculate that early and effective management of bronchiectasis exacerbations in children may lead to reduced hospitalizations, better QOL, and improved future adult lung function.
Although most respiratory physicians will treat acute exacerbations intensively with antibiotics and airway clearance [24, 30], some exacerbations are caused by viral infections, and may not require antibiotic therapy. However, it is possible that viral-bacterial interactions in the airways could promote or prolong endobronchial bacterial infection, which, with the accompanying inflammatory cascade, is a risk for additional lung injury [26, 31]. To address whether antibiotics are superior to placebo at providing short-term clinical benefits, the first component of our Bronchiectasis Exacerbation Study (BEST), is a multicenter randomized controlled trial (RCT; BEST-1) designed to answer the question: ‘Does azithromycin or amoxycillin-clavulanate, compared with placebo, improve the resolution of respiratory exacerbations on day 14 of treatment?’ .
The second component of BEST (BEST-2) seeks to determine whether the two antibiotics used in BEST-1 are equivalent at achieving resolution of an acute respiratory exacerbation in children with bronchiectasis, compared with placebo. Based on available, but limited, data on the microbiology of lower-airway secretions in children with non-CF bronchiectasis , amoxycillin-clavulanate is currently recommended as the first-line empirical oral-antibiotic treatment for non-severe bronchiectasis exacerbations in children . However, amoxycillin-clavulanate requires dosing two to three times per day, and causes gastrointestinal symptoms in many children. Oral azithromycin is attractive as an alternative first-line therapy because of its long half-life, markedly reduced dosing schedule, and good safety profile in children . Although azithromycin can be used 3 days/week  or even once weekly , our proposed study uses a daily dose in order to maintain effective blinding of the medications. However, if daily azithromycin is shown to be equivalent to amoxicillin-clavulanate, then extended-interval azithromycin dosage schedules could be trialed in appropriate clinical settings where reduced dosing frequency is particularly appealing, such as for patients likely to have poor adherence.
As well as having anti-bacterial activity against most respiratory bacterial pathogens associated with non-CF bronchiectasis , azithromycin has additional anti-microbial properties that could prove beneficial. These include its unique bactericidal activity against intracellular strains of non-typeable Haemophilus influenzae, inhibitory effects upon biofilm formation , and anti-viral properties . Furthermore, as a member of the macrolide class of antibiotics, it also has immunomodulatory and anti-inflammatory functions . Given the emerging importance of both airway inflammation and biofilm development in the pathogenesis of acute and chronic respiratory disease [26, 42], azithromycin may be a valuable intervention. Nevertheless, no RCTs have yet been carried out for azithromycin to help determine its role in treating acute exacerbations of bronchiectasis. Despite the lack of high-level evidence, azithromycin is sometimes used to treat adults with bronchiectasis exacerbations, perhaps as an extension of its reported positive effects on acute exacerbations of COPD . By contrast, two RCTs of maintenance azithromycin therapy have been conducted in adults with non-CF bronchiectasis. One was a small, non-blinded, double cross-over RCT, where 6 months of azithromycin therapy was shown to improve lung function and reduce exacerbation frequency . More recently, a 6-month, parallel, double-blind, placebo-controlled trial in New Zealand adults with bronchiectasis reported that azithromycin decreased exacerbation events, but did not alter lung function or QOL measures .
Aims of the study
In the second phase of BEST (BEST-2) the primary question will be: ‘Is daily oral azithromycin non-inferior (within a 20% margin) to oral amoxycillin-clavulanate at achieving resolution of exacerbations by day 21 of treatment?’
The secondary aims are similar to those in BEST-1 , and are to: 1) determine the effect of azithromycin or amoxycillin-clavulanate on QOL, systemic inflammation, time to next respiratory exacerbation, and duration of exacerbations; 2) examine factors that predict response to the two antibiotics, including respiratory pathogens (viruses, bacteria, macrolide-resistant bacteria) present in respiratory secretions, and systemic markers of inflammation; and 3) describe, by using sensitive molecular detection techniques, the point prevalence and diversity of respiratory viruses and Mycoplasma pneumoniae and Chlamydiales species during exacerbations, compared with the findings at enrolment when the children are clinically stable.
The study will test the primary hypothesis that oral azithromycin is non-inferior (within a 20% margin) to oral amoxycillin-clavulanate at achieving resolution of respiratory exacerbations by day 21 of treatment in children with non-CF bronchiectasis.