Introduction
Streptococcus pyogenes, or Lancefield group A streptococcus (GAS), is an important pathogen implicated in a wide variety of human infections. The species is associated with both non-invasive diseases, such as acute pharyngitis, an infection for which it is the most common bacterial agent, and invasive infections, such as skin and soft-tissue infections, necrotizing fasciitis, bacteremia, sepsis, and toxic shock syndrome [
1].
GAS remains sensitive to β-lactams, which is the drug class of choice in the treatment of most streptococcal infections because of its narrow spectrum of action and its efficacy in the prevention of post-streptococcal sequelae, such as rheumatic fever [
1]. Macrolides have been recommended for patients allergic to β-lactams, and clindamycin is the preferred antibiotic in the treatment of patients with serious soft-tissue infections because of its ability to inhibit the production of several streptococcal virulence factors [
1]. Resistance to erythromycin and related antibiotics has represented an important cause of concern [
2,
3] and is mainly associated with two mechanisms. The first is expressed by
mef genes, such as
mef(A), encoding for an efflux pump, which confers resistance to 14- and 15-membered ring macrolides and susceptibility to clindamycin (M phenotype) [
4]. The second mechanism involves
erm genes, including
erm(A) [subclass
erm(TR)] and
erm(B), which encode methylases targeting 23S rRNA [
5]. The modification is associated with a decreased binding of all macrolides, lincosamides, and streptogramin type B to their targets on the ribosomal RNA (MLS
B phenotype), and it can be either induced (iMLS
B phenotype) or constitutive (cMLS
B phenotype). Other less common mechanisms of macrolide resistance are associated with mutations in the 23S rRNA gene sequence and/or alterations in riboproteins L4 and L22 [
5]. Furthermore, an association between erythromycin resistance and cell invasiveness has been observed [
6].
The major factor thought to be influencing the prevalence of macrolide resistance is macrolide consumption [
7]. In addition, GAS clones showing
emm types strongly associated with erythromycin resistance may contribute to the overall prevalence of macrolide resistance [
8]. In this respect, Italy has always been highly ranked in the list of countries for macrolide resistance rates, ranging between 16 % and 36 % in Central Italy [
9‐
11].
The aim of the present study was to examine the prevalence and phenotypic and genotypic characteristics of macrolide resistance in pharyngeal GAS isolates causing pharyngitis collected from children during two respiratory seasons (2012 and 2013) in Central Italy and compare our results with data published worldwide. Antimicrobial resistance phenotypes and genotypes were defined; the overall genotypes of strains were determined by emm typing and pulsed-field gel electrophoresis (PFGE).
Discussion
The macrolide resistance rates vary considerably among GAS strains from different countries and over time between <3 % to >26 % [
2,
16‐
20]. In Europe from 2005 onwards, while in some regions macrolide resistance rates continued to remain high with an increasing trend, such as Greece [
20], in others, such as Spain, Portugal, France, and Germany, a significant decrease of macrolide resistance rates was reported [
14,
21‐
23].
In Italy, based on regional studies mainly, macrolide resistance rates steadily increased from 9 % in 1992 to 53 % in 1997 [
24]. Over the period 2000–2009, the rates continued to remain high in Central Italy, varying between 16 % and 36 % [
10,
11]. In those years, Italy was among the regions with the highest levels of erythromycin resistance in Europe. According to the present study, we witnessed, for the first time, a decline in macrolide resistance rates in Central Italy, among GAS isolates over the period 2012–2013, down to 7.4 %.
erm(B) was the predominant macrolide resistance gene found, followed by
mef(A).
erm(B) was responsible for the increase in macrolide resistance rates observed in Italy during the period 1992–1997 and remained prevalent between 2000 and 2003 [
10,
24]. In the present study, the
erm(A) gene was rarely found, as also reported in Belgium [
25].
In this study, among erythromycin-resistant isolates,
emm types 12, 4, and 11 predominated, accounting for 66 % of all resistant strains. In Italy, while
emm types 12 and 4 were the most prevalent types among erythromycin-resistant isolates in previous studies,
emm11 was only rarely found, even among susceptible strains [
10,
26]. Associations between
emm types and macrolide resistance genes resulted to be the same as to those found previously, with only rare exceptions, thus suggesting that, in Italy, few successful clones are associated with macrolide resistance [
9,
10,
26]. With only one exception, represented by
emm type 4 with two different PFGE types, each
emm type was specifically associated with a distinct PFGE type. The most prevalent
emm types found among our GAS strains represent also those types frequently detected in different geographical areas [
2,
20‐
22]. An
mef(A)-positive/
emm4 clone has been frequently found associated with macrolide resistance in GAS [
8,
20,
22], as well as an
erm(B)-positive/
emm11 clone that seemed to increase in prevalence in some countries [
14,
21,
22,
27]. The finding of specific associations between
emm type and macrolide resistance genotype and the fact that some
emm types are never or rarely found in resistant isolates are suggestive of the limited transfer of macrolide resistance determinants within GAS. Nevertheless, rare
emm type/resistance gene associations have been found in our study. It is the case of the second most prevalent detected clone, the
erm(B)-positive/
emm12/PFGE 3, whereas
emm12 isolates resistant to macrolides have been previously found associated with
mef(A) [
2,
21,
25], and the case of three
emm1 isolates with
erm(B), whereas
emm1-resistant isolates are generally
mef(A)-positive [
22,
27]. The uncommon
emm types/macrolide resistance associations observed in this study could reflect circulation of different clonal lineages in this geographic area.
The finding of 17 out of 22 (77.3 %) erythromycin- and tetracycline-resistant isolates that carried both
erm(B) and
tet(M) could suggest that these isolates may carry conjugative transposons belonging to the Tn916 family, such as Tn3872, Tn6002, Tn6003, Tn1545, and Tn2010, where
erm(B) and
tet(M) are genetically associated [
13]. These
erm(B)/
tet(M) isolates belonged to six different
emm types, the predominant being
emm types 12 and 11 (seven isolates each). While the association between
emm type 11 and
erm(B)/
tet(M)-mediated co-resistance has already been reported [
27],
emm12 isolates carrying both
erm(B) and
tet(M) are very rare, due to the unusual association of this
emm type with
erm(B), as stated above. Three
emm2 isolates had
mef(A) and
tet(O), suggesting that they may carry the transferable chimeric prophage Φm46.1 [
28]. The two erythromycin- and tetracycline-resistant isolates of
emm type 77 harbored
erm(A) and
tet(O).
High-level macrolide consumption, especially long-acting macrolides, have been directly associated with an increase of macrolide resistance, due to the antibiotic selective pressure, which could favor the spread of specific macrolide-resistant GAS clones, with changes in macrolide resistance rates and genotypes [
19]. In some countries, reduction in macrolide consumption proved to be an important factor responsible for the decline in macrolide resistance rates [
22,
23,
25]. In Italy, macrolide consumption remained relatively stable during the last 15 years (1999–2013) [
29], although available data refer to the general population and not to the pediatric group only. Also, the consumption of tetracyclines, which have been associated with the increase of macrolide resistance, remained stable [
29]. Although we do not have information on antibiotic usage in the patients of this study, the observed decline in macrolide resistance rates seems to not correlate to general macrolide consumption. A recent study from Slovenia indicated even the opposite; that is, a boost in resistance rates among non-invasive GAS, despite a decrease in macrolide use [
30]. Thus, besides the presumed association with antibiotic use, other underlying mechanisms influencing the development and spread of antibiotic-resistant GAS isolates are to be considered, such as natural fluctuations in the prevalence of resistant clones and low fitness costs of some erythromycin-resistant clones [
25]. We found that the major
emm types in the overall GAS pharyngeal population circulating in Italy during 2012–2013 were the same as those reported in previous Italian studies, specifically
emm types 4, 12, and 89, although with differences in their relative frequencies [
10,
26]. Some of the most frequent
emm types were also the most prevalent among erythromycin-resistant strains. PFGE was not able to differentiate resistant and susceptible isolates belonging to the same
emm type, and this observation seems to indicate that the decrease of macrolide resistance would not be due to a decrement of specific GAS macrolide-resistant clones within a given
emm type.
This study documents a decline in macrolide resistance rates in Italy, where macrolide resistance has been documented to be high in the past, and it provides useful comparative data for future epidemiological studies across erythromycin-resistant GAS populations.