This study demonstrated that two mechanisms are involved in macrolide resistance among S. pneumoniae
isolates from Lebanon, namely, efflux pump mediated resistance and ribosomal modification due to adenine-dimethylase, with dominance of the latter. This observation is not concordant with that seen in some European countries like France, Spain, and Poland where macrolide resistance due to efflux pumps is almost exclusive [18
]. This mechanism is also mostly prevalent in the USA and some other European countries like Greece and Germany [18
]. Moreover, a high rate of dual resistance was detected in our isolates where 14 of 44 (32%) of the isolates carried both genes.
Some of our clinical isolates were found to be both erm
(B) and mef
negative, suggesting the possibility of one of the newly described resistance mechanisms, such as mutations in the 23S rRNA or alteration of the ribosomal proteins L4 and L22 [20
] requiring further investigation. The high (79.5%) prevalence rate of penicillin resistance among our erythromycin resistant isolates denotes that in our community, the evolution of erythromycin resistance is driven, possibly by the spread of penicillin resistant clones since isolates with the same serotypes manifested resistance to both penicillin and macrolides. Moreover, the high rate of macrolide resistance was accompanied by a high rate of tetracycline resistance (77%) indicating a possible association to the conjugative transposon Tn1545 that confers resistance to tetracycline via the tet
(M) gene in addition to resistance to macrolides [22
There is increasing evidence that macrolide resistance may result in clinical failure. Studies worldwide have shown that the frequency of this resistance might be related to the level of macrolide consumption [23
]. The same rationale may apply to Lebanon where overuse of macrolides may contribute to the observed increase in resistance to these antimicrobial agents in S. pneumoniae
As already mentioned, the majority of macrolide resistant isolates in our community belonged to seven different serotypes: 19F, 23, 2, 14, 19A, 12F, and 9V/9A. These results are concordant with the most common serotypes found in Asia, Turkey, and Saudi Arabia [25
]. The serotypes of isolates recovered from invasive pneumococcal infection cases were 19F, 14, 23, 19A, 9V/9A, and 12F with serotype 19F being the most common. Nine of 14 (64.28%) isolates carrying both the resistance genes were serotype 19F (Table ). This might suggest that isolates belonging to serotype 19F are highly resistant to macrolides. In addition, the serotypes 12F and 14 were associated with both dual resistance and invasiveness. Serotype 2, one of the most prevalent serotypes, seems to be non-invasive since it was found among the upper and lower respiratory tract specimens, but not among specimens from sterile sites.
Distribution of macrolide resistant genotypes versus serotypes among S. pneumoniae isolates
It is noteworthy that the pool of serotypes found among macrolide resistant isolates was completely different from the pool of serotypes found for susceptible isolates with the exception of serotype 9V/9A which was common among macrolide resistant and macrolide susceptible isolates. Interestingly, serotype 19F, which was the major serotype found among resistant and invasive isolates, was not found among the susceptible ones. This denotes that a few, specific serotypes are responsible for macrolide resistance among our S. pneumoniae isolates.
Alarmingly, some of the most prevalent macrolide resistant serotypes recovered, including serotypes 2 and 12, are not covered by PCV7, PCV10 or PCV13 vaccines [12
] and thus, constitute a risk for dissemination in the community. Notably the remaining macrolide resistant serotypes that were recovered are covered by PCV-7, PCV-10 and/or PCV-13, including the highly prevalent serotypes 19F, 23 and 14 [12
]. In Lebanon, pneumococcal conjugate vaccines are available in the private sector but are not included in the Expanded Program of Immunization. The estimated PCV vaccine coverage is 10-15% for children under 5 years of age. Most of our isolates with PCV covered serotypes were obtained from unvaccinated subjects.
In conclusion, though the macrolide resistance observed in our S. pneumoniae isolates is mostly due to target site modification by the methylase encoded by the erm(B) gene, many isolates also demonstrated efflux pump mediated resistance to macrolides. Moreover, a number of isolates were invasive and macrolide resistant, with only a few belonging to serotypes covered by currently available vaccines. Understanding the mechanisms of resistance to macrolides may be helpful in choosing the correct treatment regimen in certain situations but is definitely important in the development of new antimicrobial agents. Moreover, continued surveillance for changes in serotype distribution is necessary, especially after the introduction of new vaccines.