Why Was the Azithromycin “for Life” Trial Necessary?
Kathryn Maitland, A. Sarah Walker
N Engl J Med 2024;391:760-762
DOI: 10.1056/NEJMe2407000
In 2009, a cluster-randomized trial showed a 49% reduction in all-cause mortality among Ethiopian children 1 to 9 years of age in the year after mass distribution of a single dose of azithromycin for trachoma prevention — a reduction that was sustained over the following 26 months.1 This remarkable finding led to further multinational cluster-randomized trials examining mass distribution of azithromycin for children 1 to 59 months of age in trachoma-endemic areas of Niger, Tanzania, and Malawi (the Macrolides Oraux pour Réduire les Décès avec un Oeil sur la Résistance [MORDOR] trial)2 and such distribution combined with seasonal malaria prophylaxis in children 3 to 59 months of age in Mali and Burkina Faso (the SMCAZ trial).3 The results of the MORDOR trial showed substantial heterogeneity with respect to the effect of mass distribution on mortality among the countries involved, with the greatest mortality reductions in Niger (possibly related to site-specific background mortality4) and in the youngest children (1 to 5 months of age). The SMCAZ trial showed no additional benefit with respect to childhood survival, although gastrointestinal infections, upper respiratory infections, and nonmalarial febrile illnesses were reduced. A systematic review of evidence conducted by the World Health Organization (WHO), which balanced the benefits of treatment with the potential harms of antimicrobial resistance and adverse events, strongly recommended against mass distribution of azithromycin (though the quality of evidence was low). A conditional recommendation was made for azithromycin distribution only for infants (1 to 11 months of age) to prevent death in regions where all the following conditions were met: baseline mortality of more than 60 deaths per 1000 population among infants or of more than 80 deaths per 1000 population among children younger than 5 years of age, continuous monitoring of infant and childhood mortality and antimicrobial resistance, and continued implementation of concurrent interventions to prevent childhood death.5 These conditions set the bar very high for future implementation.
In the Azithromycin “for life” trial (AVENIR [Azithromycine pour la Vie des Enfants au Niger: Implementation et Recherche]), the results of which are reported by O’Brien et al. in this issue of the Journal6 communities in Niger with varying infant and childhood mortality were randomly assigned, in a 1:1:1 ratio, to twice-yearly azithromycin distribution to infants 1 to 11 months of age and placebo to children 12 to 59 months of age (the infant azithromycin group), twice-yearly azithromycin distribution to children 1 to 59 months of age (the child azithromycin group), or twice-yearly placebo distribution to children 1 to 59 months of age (the placebo group). Seasonal malaria chemoprevention coverage was reported to be at least 80% in the trial area during the time that the AVENIR trial was conducted. In an elegant response-adaptive design, communities that were enrolled after the first year of the trial were randomly assigned to one of the three groups on the basis of the probability that communities would be assigned to the group with the lowest mortality among children 1 to 59 months of age (thus maximizing community and individual participant benefits). Children’s vital status was ascertained through a twice-yearly door-to-door population census, which minimized missing data.
Mortality was significantly lower in communities in the child azithromycin group than in communities in the placebo group among children 1 to 59 months of age (14% lower mortality with azithromycin; 95% confidence interval [CI], 7 to 22) and among children 12 to 59 months of age (13% lower mortality with azithromycin; 95% CI, 4 to 21). This finding was not replicated among infants 1 to 11 months of age in the infant azithromycin group as compared with those in the placebo group (6% lower mortality with azithromycin; 95% CI, −8 to 19). A secondary analysis showed an indirect benefit for infants 1 to 11 months of age when all children 1 to 59 months of age received azithromycin (17% lower mortality among infants in the child azithromycin group than among infants in the infant azithromycin group; 95% CI, 4 to 28), but when distribution was restricted to infants, neither infants nor children received benefit. These results provide strong evidence of the indirect effects on community-level mortality of changing the epidemiology of the underlying pathogens that are carried across age groups. The effect size of 14% lower mortality among children 1 to 59 months of age was similar to the findings of the MORDOR trial,2 even when azithromycin was provided alongside seasonal antimalarial prophylaxis. Furthermore, the effect was similar across sites with varying childhood mortality. These results challenge most of the current recommendations for mass distribution of azithromycin except for one — the specter of increased antimicrobial resistance.7
The effect of mass distribution of azithromycin on antimicrobial resistance as a result of this trial has yet to be reported. In the MORDOR trial, analysis of the determinants of macrolide and nonmacrolide resistance as measured by means of rectal swabs collected at baseline and at 36 months and 48 months (after the sixth and eighth mass distribution, respectively) showed that determinants of resistance in the intervention communities were more than 7 times as high as those in the placebo communities at 36 and 48 months.8 Although this class of drug has few clinical indications in children (indicated only for specific enteric infections,)9 an increase in resistance to beta-lactam antibiotics by 2.1 times is a concern, since this antibiotic class is widely used in the countries involved in these trials. Further deliberations for guideline recommendations weighing the risks and benefits of mass distribution of azithromycin for communities and individual persons should not overlook the disease it was originally intended to control — infection with Chlamydia trachomatis. Trachoma has the highest global prevalence in the poorest and most rural areas of the Sahel. The most affected communities bear the brunt of this terrible disease, including high levels of severe visual impairment and irreversible blindness, with an enormous economic burden in terms of lost productivity.10 Trachoma prevention programs that provide azithromycin distribution to the whole population, in which strict prevalence conditions are met, will continue to provide indirect survival benefits to children.
