Macrolides are used to treat many respiratory bacterial infections, and the increasing prevalence of macrolide-resistant species is a significant problem. Macrolide resistance is caused by mutations in ribosomal structural elements which alter the binding affinity of macrolides to the ribosome. Resistance can also be attributed to the presence of acquired macrolide resistance genes (AMRGs), including the erm genes that code for ribosomal methylases that alter the binding affinity of the drug, and the efflux-mediating mef genes.

Macrolides are used to treat infections where nontypeable Haemophilus influenzae (NTHi) may be involved. Most NTHi isolates carry an intrinsic efflux mechanism that makes them less susceptible to macrolides than other species, but true resistance is rare and is usually attributed to ribosomal mutations. However, a recent study by Roberts et al. (2011) involving NTHi isolates derived from cystic fibrosis (CF) patients on an azithromycin placebo-controlled trial found at least one AMRG in every isolate, irrespective of the presence of phenotypic resistance.¹

Our aim was to further investigate these AMRGs and macrolide resistance in NTHi to see if these findings could be replicated in our collection. In our study of 186 NTHi isolates from a mix of CF and non-CF patients, with and without prior macrolide exposure, we used highly specific probe-based PCR and found that none of our isolates carried any of the AMRGs found in the Roberts study (ermA, ermB, ermC, ermF, and mefA), but that false positives were frequently seen using the original PCR primers and protocol.

From our collection, only 2 were highly resistant to macrolides by Etest (azithromycin MIC>64); interrogation of whole genome sequences of these isolates revealed mutations in genes for ribosomal structural components that have previously been attributed to macrolide resistance in other species. A further 135 AMRGs were not detected in these isolates.