Globally, more than 2.5 million neonatal deaths occur every year, and a third of these deaths are caused by severe bacterial diseases, which often present as sepsis. In sub-Saharan Africa, neonatal sepsis is caused primarily by Gram-positive bacteria with Staphylococcus aureus often associated with these infections.
Oral azithromycin (AZI) prophylaxis has been shown to be a promising intervention. AZI given during pregnancy has been used in clinical trials in low- and middle-income countries to reduce infant mortality and preterm delivery.  Intra-partum AZI has also been shown to reduce maternal and neonatal infections as well as colonization.

However, of concern is that AZI interventions are also associated with an increase in the prevalence of macrolide-resistant bacterial species, even after the administration of a single dose.

In a double-blinded placebo-controlled trial (PregnAnZI-1) conducted in The Gambia, a single dose of oral AZI administered to women in labour decreased S. aureus colonization in women and their babies, but was also associated with a short term increased prevalence of S. aureus AZI resistant strains.

The primary aims of this current investigation were to use genomics to characterize the population of AZI resistant S. aureus recovered from mothers and babies, and to further identify the genetic mechanisms responsible for AZI resistance in this population.

The investigation revealed seven different S. aureus sequence types (STs) among mothers and babies with ST5 dominant followed by ST15.  Among AZI exposed mothers and babies, ST15 was most prevalent whereas ST5 was prevalent among unexposed individuals. The public health importance of our findings is that AZI intervention appears to select for methicillin sensitive S. aureus (MSSA) ST15 that frequently habour the Panton Valentine Leukocidin (PVL) and enterotoxin A genes compared to ST5 that rarely carry these virulence genes.

AZI resistant S. aureus were found to carry either an erm(C) (n=9, 20%) or msr(A) (n=36, 80%) gene associated with macrolide resistance.  The two macrolide resistance genes are associated with different phenotypes: carriage of msr(A) among staphylococci is associated with phenotypic resistance to 14-membered (clarithromycin, dirithromycin and erythromycin) or 15-membered (azithromycin) ring macrolides and streptogramin A, but susceptibility to 16-membered ring macrolides. Carriage of erm genes in staphylococci is associated with a broader phenotypic resistance depending on whether the gene is inducible or constitutively expressed. The former, often referred to as iMLSB, mediates resistance to 14- and 15-membered macrolides and streptogramin B, but susceptibility to 16-membered macrolides and lincosamides, with a risk of constitutive expression arising in vivo. The latter, denoted by cMLSB, mediates resistance to all macrolide, lincosamides and streptogramin B antibiotics.

Abdoulie Bojang, molecular microbiologist and PhD candidate at the Medical Research Council Unit The Gambia at the London School of Hygiene and Tropical Medicine, and first author of the paper said, “Genomic investigation of resistant bacterial population is of paramount importance as it reveals not only the impact of AZI intervention on the bacteria at the genomic level, but also the prevalence of circulating bacteria types and the antibiotic susceptibility pattern as well as the potential to cause disease”.

We have used comprehensive genomic analyses to reveal the dynamics of AZI resistant S. aureus colonization in mothers and babies after AZI treatment or placebo. Plasmid-encoded msr(A) in ST5 MSSA was the most common clone, being responsible for most AZI resistance. These results provide important information for greater understanding of the ecological impact of azithromycin prophylaxis on staphylococcal populations in Western Africa.

Dr Anna Roca, PI of the PregnAnZI trial and Abdoulie’s PhD supervisor said, “The ongoing work on antibiotic resistance linked to the intra-partum azithromycin trials is very important to assess the overall impact of this potential intervention. In addition, it is very encouraging seeing a Gambian PhD student conducting cutting edge research on a topic as important as AMR”.

This research was undertaken in Professor Ben Howden lab in Melbourne, Australia. It was jointly supported by the MRC Unit The Gambia at LSHTM PhD fellowship and Doherty Applied Microbial Genomics, Department of Microbiology & Immunology, University of Melbourne at The Peter Doherty Institute for Infection & Immunity.

The results have been published in the Journal of Antimicrobial Chemotherapy(JAC): https://www.ncbi.nlm.nih.gov/pubmed/31424550