Discussion
AOM is a disease with worldwide prevalence having broad disease burden and may require prolonged treatment courses because at least a third of children have two or more episodes of AOM (recurrent AOM) in the first three years of life [
14]. Reliable epidemiological data on etiology and burden of AOM are important as the data help clinicians with the selection of appropriate empiric antibiotic therapy for pediatric AOM and for public health policy decision-making.
In this retrospective study, we found that
S. pneumoniae and
S. aureus were the most predominant etiologic agents causing AOM, being isolated in 36.4 and 16.2% of the specimens of children with AOM, respectively. Most of the
S. aureus was MDR and resistant to erythromycin, clindamycin and tetracycline. The first two antibiotics (erythromycin and clindamycin) were the most frequent medicines prescribed by Chinese pediatricians for infectious diseases [
4]. Historically, the major bacteria responsible for most cases of AOM were
S. pneumoniae and
H. influenzae [
15]. The etiology of the pathogenic bacteria does not appear to have changed significantly over time. Since the prevalence and the main causal agents of AOM varied by geographic location, we observed a different epidemiology and etiology from previous studies [
1,
3,
5] which revealed that the most causal agents of AOM were
S. pneumoniae and
H. influenzae, but our study was in line with one study conducted in southern China which demonstrated that the major pathogens causing AOM were
S. pneumoniae and
S. aureus, which accounted for 47.2 and 18.5% of the specimens isolated from AOM patients, respectively [
2]. It has been reported that in the era of universal pneumococcal conjugate vaccine (PCV) immunization, that
H. influenzae may become the predominant pathogen of AOM, suggesting that the introduction of PCV7 can change the relative prevalence of main causal agents [
16]. The same result was observed in Saudi children, after the introduction of pneumococcal vaccines in the routine immunization schedule,
S. aureus has become the most predominant contributor to AOM [
17]. The determinants of why
S. aureus has become the second most common causal agent of AOM in China is poorly understood. In China, as the
H. influenza b vaccine and PCV were self-paid and did not enter into the Chinese Expanded Program on Immunizations (EPI), we didn’t see the changes of pathogen patterns distributed in the AOM disease for the vaccination of
H. influenza b vaccine and PCV. The low coverage of PCV7 and antibiotic overuse and abuse in China can partly explain this disparity [
2]. In this region of the world,
S. aureus should be considered and targeted with appropriate therapy if initial therapies targeting
S. pneumoniae fail to lead to clinical improvement, especially if culture is not available.
Antibiotic resistance has become an important public health problem in mainland China. Restriction of β-lactam use in MRSA infections required use of other types of antibiotic options for treatment. However, except for resistance to all kind of β-lactam antibiotics, the MRSA isolates found in our study developed a high resistant rate to non-β-lactam antibiotics, especially to erythromycin, clindamycin and tetracycline. We found that the resistance rate to clindamycin in MSSA is even higher than in MRSA isolates. It was reported that both erythromycin and clindamycin have been common prescribed antibiotics for
S. aureus infection [
18]. A high resistance rate was also reported in mainland China [
19], which indicated that the high antibiotic resistance rate of
S. aureus is a common public health problem in China and that the two antibiotics were not the priority options for the empiric antibiotic therapy in pediatric infections. It was previously reported that in the macrolide resistance isolates, there were 59.4, 24.3, and 21.6% of which carried
ermA,
ermC and both
ermA and
ermC gene, respectively. Our study was consistent with a previous report that showed that of resistant
S. aureus isolates, 37.7% had
ermA, 26.6% had
ermC and 18.6% had both
ermA and
ermC genes [
20], but different from a study conducted in Turkey which showed that 50% of
ermA positive isolates also carried the
ermC gene [
10].
As a pathogen with extremely high prevalence,
S. aureus causes various clinical infections such as skin and soft tissue infections (SSTIs) [
19], invasive community-acquired MRSA (CA-MRSA) infections [
21], and pneumonia [
22]. Few studies about MRSA and MSSA isolates contemporaneously circulating among age-specific groups of children attending otolaryngology clinics have been examined [
23,
24]. According to the previous report [
11], hospital-acquired MRSA (HA-MRSA) is usually detected with SCC
mec type I, II and III, while CA-MRSA is usually detected with SCC
mec type IV, IVa and V. In this study, all of the MRSA isolates carried SCC
mec IV, IVa and V, which confirmed that these MRSA isolates belonged to CA-MRSA. Twelve
Spa types and seven ST types clustered into 7 clonal complex (CCs) among MSSA and 3 CCs among MRSA were observed in our study, indicating that there is great genetic diversity in
S. aureus isolated from AOM patients. MSSA isolates with a genetic background (ST30-
t037, ST45-
t1081 and ST59-
t437) was common to MRSA clones in this study suggesting that these MSSA isolates might have the potential to become CA-MRSA clones once acquisition of the
mecA gene occurs [
21].
Despite the high prevalence, only a few epidemic clones have been identified in China [
25‐
27]. Previous studies throughout mainland China found that ST59- SCC
mec-IVa/V strains were the most common strains causing CA-MRSA infections among children [
25‐
28]. Our study also confirmed that the predominant sequence type of MRSA isolated from AOM children was ST59, which accounted for 75% of all the MRSA isolates. The previous report of ST59 was detected from a few MSSA isolates and in a single MRSA isolate in the United States, a large proportion of ST59 emerging in Taiwan was reported in 2004 and ST59-MRSA was called Taiwan clone [
29]. ST59 was not only predominant in Shanghai [
30], Guangzhou [
31], and Taiwan [
29], but also served as prevailing strains in Hongkong [
32] and Vietnam [
33]. The Asian Network for Surveillance of Resistant Pathogens (ANSORP) study conducted in 17 hospitals from Asian countries demonstrated that the predominant clones of CA-MRSA isolates were ST59-MRSA-SCCmec type IV-spa type t437 [
34]. These findings suggested that ST59 is currently spreading between adjacent regions and supporting its dominance in the Asian region as a whole [
33]. It is widely assumed that the CA-MSSA isolates acquiring the resistance gene
mecA would become the major sources of CA-MRSA. In our study, we observed that ST59-MSSA was the predominant sequence type in the MSSA group, accounting for 28% of all MSSA isolates, which indicated that the
S. aureus isolates undergoing genetic variations have great capacities for environmental adaption. The similar genetic background of ST59 between MRSA and MSSA isolates was also observed in ST30 and ST45 in our study.
ST45 was the second prevailing ST in our study, accounting for 20% of MSSA and 16.7% of MRSA isolates. It was reported that clonal complex 45 (CC45) is common throughout European countries such as Germany and the Netherlands and Belgium [
35]. The ST45-SCC
mec-I
V/V-t437 clone is well known as the Berlin clone. The Berlin clone was first observed in 1993, and its emergence was attributed to acquisition of
mecA by a community clone of MSSA [
36]. The ST45 now spread in Hongkong [
35] and mainland China [
37], including in western China where our study was conducted. It was speculated that CC45 strains may be more transmissible among health care settings and hospitals [
35].
One of the interesting findings was that ST398-MSSA was found in this study. ST398 is considered as a livestock-associated pathogen mainly affecting people in contact with major animal reservoirs [
38]. It is noteworthy that this AOM case with isolates of ST398 reported no direct livestock-associated risk factors, although many reports documented that persons living in places of high livestock density were found to have a greater chance of livestock-associated CC398 carriage even if they lacked direct contact with animals [
39,
40]. CC398 may now be sporadic and distributed in China including areas such as Shanghai [
30] and Liuzhou. This study finding suggests the probability of CC398 transmission via human contact instead of animal contact [
41].
Panton-Valentine leukocidin (PVL) is a bicomponent toxin that can cause the lysis of leucocytes and it is a main virulence factor of
S. aureus which is responsible for severe invasive disease such as necrotizing pneumonia [
30]. An important finding in this study was the high detection rate of the
pvl gene in
S. aureus isolates, with significant differences between the MRSA and the MSSA groups. Our result was consistent with previous reports indicating that the
pvl gene was more common in MRSA isolates than in MSSA isolates [
42]. Several studies found that the proportion of
pvl positivity was approximately 27–40% among
S.aureus isolates detected from children in mainland China [
30]. In the current study, the
pvl gene was found in ST30, ST45 and ST59 clones. It was reported that CA-MRSA ST59 isolates had significantly more pronounced virulence than the geographically matched HA-MRSA clones ST239 in various animal models, including the
pvl gene [
43]. The CC59 was predominant among
pvl positive CA-MRSA in mainland China [
30], for example, Li et al. [
44] reported 55.5% of CC59 MRSA isolates to be
pvl positive in China, while we detected 66.7% of CC59 MRSA isolates with
pvl positive in AOM disease.
There are some limitations to our study. First of all, the single-center design and the small number of AOM patients may limit the generalizability of our study results. Secondly, the AOM cases in this study may not accurately represent all AOM cases as we swabbed spontaneous ear pus drainage from the deep ear canal and the external auditory canal to culture organisms, the results of which may or may not have included the true middle ear pathogen. S.aureus may have been a leading cause of AOM, but as we swabbed the ear canal, this may lead to detection of some colonizing agents such as S.aureus. Lastly, a retrospective review of medical records for identifying patients presented to an Otolaryngology clinic may have potentially decreased the generalizability of the results, as some children may have had more severe disease which were referred to a surgeon, as opposed to a primary care provider.