Serotypes and Antibiotic Susceptibility of Streptococcus pneumoniae Isolates from Invasive Pneumococcal Disease and Asymptomatic Carriage in a Pre-vaccination Period, in Algeria

In Algeria, few data is available concerning the distribution of pneumococcal serotypes and respective antibiotic resistance for the current pre-vaccination period, which is a public health concern. We identified the most frequent Streptococcus pneumoniae serogroup/types implicated in invasive pneumococcal disease (IPD; n = 80) and carriage (n = 138) in Algerian children younger than 5 years old. Serogroup/types of 78 IPD isolates were identified by capsular typing using a sequential multiplex PCR. Overall, serotypes 14, 19F, 6B, 23F, 18C, 1, 5, 7F, 19A, and 3 (55% of PCV7 serotypes, 71.3% of PCV10, and 90% of PCV13) were identified. Additionally, 7.5% of the non-vaccine serotypes 6C, 9N/L, 20, 24F, 35B, and 35F, were observed. In the case of S. pneumoniae asymptomatic children carriers, the most common serogroup/types were 6B, 14, 19F, 23F, 4, 9V/A, 1, 19A, 6A, and 3 (42.7% of PCV7 serotypes, 44.2% of PCV10, and 58% of PCV13). For 6.1% of the cases co-colonization was detected. Serotypes 14, 1, 5, and 19A were more implicated in IPD (p < 0.01), whereas serotype 6A was exclusively isolated from carriers (p < 0.01). Deaths associated with IPD were related to serotypes 19A, 14, 18C, and one non-typeable isolate. Among IPD related to vaccine serotypes, the rates of penicillin non-susceptible isolates were higher in no meningitis cases (80%) than in meningitis (66.7%), with serotypes 14, 19A, 19F, and 23F presenting the highest MIC levels (>2μg/ml). Resistance to cefotaxime was higher in isolates from meningitis (40.5%); however, resistance to erythromycin and co-trimoxazole (>40%) was more pronounced in no-meningeal forms. Overall, our results showed that PCV13 conjugate vaccine would cover up to 90% of the circulating isolates associated with IPD in Algeria, highlighting the importance of monitoring the frequency of S. pneumoniae serogroups/types during pre- and post-vaccination periods.


INTRODUCTION
Streptococcus pneumoniae remains the leading cause of bacterial infection among children worldwide, being the most common cause of bacterial pneumonia, and an important cause of meningitis and bacteremia. Approximately 800,000 deaths per year occur among children as a result of pneumococcal infection (Johnson et al., 2010). The management of pneumococcal infections has been aggravated by the rapid worldwide increase of resistance to penicillin and other antibiotics, mostly related to the misuse of these drugs in respiratory pathogens (Song et al., 2012;Ginsburg et al., 2013).
Despite the diversity of capsular types, comprising at least 98 distinct serotypes, only some variants are associated with invasive pneumococcal disease (IPD; Caierão et al., 2014;Richter et al., 2014).
Pneumococcal disease is frequently preceded by asymptomatic nasopharyngeal colonization, which can be quite high in early childhood (Bogaert et al., 2004;Miernyk et al., 2011;Satzke et al., 2013). It is generally agreed that most serotypes recovered from IPD are also frequently identified in colonized healthy children (Bogaert et al., 2004;Dias and Caniça, 2007). However, serotype prevalence can change according to age, geography, time, and antibiotic resistance, among other factors (Finland and Barnes, 1977;Hausdorff et al., 2005;Ingels et al., 2012). IPD has also been related to recent respiratory viral infection (Weinberger et al., 2014).
In some countries, the use of effective pneumococcal conjugate vaccines (PCVs) during infancy have contributed to reduce morbidity and mortality associated to IPD, as well as nasopharyngeal colonization by vaccine serotypes (Ghaffar et al., 2004;Millar et al., 2006;Tan, 2012). Despite the availability of these vaccines, pneumococcal infections remain a global problem due to the replacement of vaccine by non-vaccine serotypes, mostly associated with the emergence of multidrug resistant serotypes, such as the serotype 19A (Dias and Caniça, 2007;Dagan et al., 2009;Richter et al., 2014). Thus, due to the widespread phenomena of serotype replacement, PCVs have been substituted by higher valence pneumococcal vaccines (from PCV7 and PCV10 to PCV13) across the world, according with the recommendations of the Centers for Disease Control and Prevention (CDC, 2013).
The monitoring of antibiotic resistance trends and serotype distribution in the pre-and post-vaccination periods is essential to assess the dynamic change of epidemiology. This way, the impact of vaccines and antibiotic use control programmes should be evaluated across countries.
In this study, we analyzed the frequency of serotypes associated to IPD and S. pneumoniae asymptomatic carriage among children younger than 5 years old in Algeria, before the introduction of the pneumococcal vaccine, and correlated the isolates antibiotic susceptibility with vaccine serotypes .

Bacterial Isolates
An overall of 218 S. pneumoniae isolates recovered from children <5 years old were collected and studied regarding the serotype.
Briefly, 80 IPD isolates [children <1 year old (n = 44), 1-2 years old (n = 20), and 3-5 years old (n = 16)] were collected between the 1st January 2010 and the 31st December 2014. The isolates were recovered from routine microbiological cultures at Laboratory of Clinical Microbiology, at CHU Mustapha Bacha (LCM/CHU), Algiers, Algeria, and at cooperation laboratories in the north (n = 50), west (n = 12), south (n = 2), and northeast (n = 16) regions of the country that sent samples and/or isolates to LCM/CHU. Isolates were included if they were obtained from consecutive blood, and cerebrospinal, pleural, peritoneal, ascitic, bone, and joint fluids from patients with symptoms compatible with IPD. Only one isolate per patient was considered .
In addition, all nasopharyngeal cultures (n = 130) recovered at LCM/CHU between 2011 and in 2012, from asymptomatic children <5 years old were also included for serotyping evaluation. Each age group of <1 year, 1-2 years, and 3-5 years included 91, 20, and 27 S. pneumoniae isolates, respectively. Only one isolate per child was considered, except for eight cultures corresponding to carriers co-colonized with two pneumococcal serogroups/types (in a total of 138 isolates). Authorization for carriage study was approved by "The Direction de la Santé et de la Population de la Wilaya d'Alger, " in Algeria.

DNA Extraction and Multiplex PCRs
To extract DNA from clinical and control S. pneumoniae isolates, several colonies were picked from the culture plates. DNA was extracted by the heat lysis method and stored at −20 • C until further analysis. The serogroups, and whenever possible the serotypes, of 80 S. pneumoniae from IPD and of 138 recovered in asymptomatic carriers were determined by sequential multiplex groups (SMGs), as previously described ; Figure 1).

Quellung Serotyping
In addition to multiplex PCR , serogroups/types of all the 80 invasive isolates were determined at LCM/CHU by Quellung reaction with specific type antiserum [Statens-Serum Institute]. Concerning the isolates recovered from carriers, only those determined as serotype 6A/B or as non-typeable by the multiplex PCR method, were characterized by Quellung reaction (Figure 1).

PCR Detection of Autolysin (lytA) Gene
A primer pair 5 ′ -TCCAGCCTGTAGCCATTTCG-3 ′ and 5 ′ -GCGGTTGAACTGATTGAAAG-3 ′ that specifically targeted a 472 bp internal region of the autolysin (lytA) gene was used to the identification of cps-negative S. pneumoniae. The amplification conditions were: initial denaturation at 94 • C for 5 min., followed by 30 cycles of denaturation at 94 • C for 30 s, annealing at 56 • C for 30 s and extension at 72 • C for 30 s, and a final extension at 72 • C for 5 min. Positive and negative controls were included in each PCR reaction.

Antibiotic Susceptibility Testing
Susceptibility testing of 72 IPD vaccine serotype isolates and of 8 IPD non-vaccine serotype was carried out by an agar disk diffusion method for four antibiotics (erythromycin, clindamycin, co-trimoxazole, tetracycline), and minimal inhibitory concentration (MIC) was determined using E-test method (AB Biodisk) for four β-lactam antibiotics (penicillin, amoxicillin, cefotaxime, and imipenem). Testing conditions and susceptibility interpretation followed the standards proposed by the Clinical and Laboratory Standards Institute (CLSI, 2014). S. pneumoniae ATCC 49619 was used as the quality control strain.

Statistical Analysis
OpenEpi software, version 3.03a (Dean et al., 2015), was used for statistical analysis. Fisher exact test was used to assess differences between IPD and carriers groups. One-tail P ≤ 0.05 were considered to be statistically significant.

IPD Serogroup/Types
Between 2010 and 2014, a total of 207 episodes of IPD were registered at LCM/CHU. Among those, 80 (38.6%) corresponded to children <5 years old and were retained for this study. These 80 IPD cases (Figure 2) comprised 56.3% of occurrences in males and 43.7% in females, with 64 cases (80%) being reported to infants (≤2 years old). Overall, the serotypes were identified in 78 isolates, while two were non-typeable (Table 1).
From no meningitis infections (n = 25), 80% of the isolates were penicillin non-susceptible with 12% of high level of resistance (MIC = 4-8 µg/ml). Intermediate resistance to amoxicillin and cefotaxime was 9.1 and 8.3%, respectively. High level of resistance to amoxicillin was observed in 4.5% isolates (MIC 8 µg/ml).

DISCUSSION
S. pneumoniae remains the leading cause of bacterial infection among children worldwide, including numerous cases of invasive disease associated high morbidity and mortality rates (Harboe et al., 2010;Adegbola et al., 2014). The pneumococcal FIGURE 4 | Rate of serogroup/type cumulative coverage (C %) in each conjugate vaccine (PCV7, PCV10, PCV13) for S. pneumoniae recovered from invasive pneumococcal disease (IPD) and asymptomatic carriers less than 5 years old, from Algeria. NT, non-typeable. epidemiology regarding capsular types and antibiotic resistance varies geographically and temporally in terms of origin of the isolates (infections or carriage), clinical presentation, pathogenicity (Hausdorff et al., 2005;Harboe et al., 2010;Ingels et al., 2012;Geno et al., 2015), and the methods used for serotyping (Turner et al., 2011;Song et al., 2012;Geno et al., 2015). Although antibiotic susceptibility and serotype data remains insufficient in many countries such as Algeria, its collection and evaluation is essential for the treatment of pneumococcal infections and for the usage of conjugate vaccines (Ramdani-Bouguessa and Rahal, 2003;Tali-Maamar et al., 2012;Ziane et al., 2012;Ramdani-Bouguessa et al., 2015).
Thus, the main goal of this study was to analyze the frequency of serotypes associated to IPD and S. pneumoniae asymptomatic carriage in Algeria, in children <5 years old before the introduction of the pneumococcal vaccine. To accomplish our purpose, we used the more recent IPD samples studied in the country (Ramdani-Bouguessa and Rahal, 2003;Tali-Maamar et al., 2012;Ziane et al., 2012;Ramdani-Bouguessa et al., 2015). It is worth mentioning that a molecular approach was here applied for the first time to isolates recovered in this country in the context of a monitoring survey of S. pneumoniae serogroup/types .
In this study, meningitis represented 60% of the clinical presentation of IPD, with 79.2% registered in children younger than 2 years old. Pleuropneumonia was reported in 22.5% of the total IPD cases, with the majority being manifested in infants (88.8%, 16/18). Differently from bacteremia, mastoiditis, and abdominal infections, the bone and joint infections were noted exclusively in infants with up to 1 year old. Indeed, IPD occurs mostly in children under the age of 5 years, especially within the subgroup of those under 2 years of age (Tan, 2012). The prevalence of S. pneumoniae carriage in healthy children younger than 5 years old can fluctuate between 20 and 93.4% in low income countries, being superior than what is stated for lower-middle income countries (from 6.5 to 69.8%), as reported by Adegbola et al. (2014). In addition, the density of pneumococcal nasopharyngeal carriage seems to decrease in higher age groups, with the children being more competent than adults in the transmission of pneumococci (Roca et al., 2012).
Mortality rate of pneumococcal invasive diseases may range from 10 to 30%, according with the studies of Pebody et al. (2006), andBravo (2009). Although it has been reported that the highest S. pneumoniae-associated morbidity and mortality rates are in Africa and Asia (Johnson et al., 2010;Turner et al., 2011), in our study the case fatality was 8.8%, mainly comprising meningitis related with vaccine serotypes (85.7%).
Among IPD isolates, serotypes 14, 19A, 19F, 6B, 1, and 5 were the most frequent, which shows some differences from data described for other periods in Algeria. For instance, the serogroups/types 1 and 5 were the most common for the period of 1996-2000(Ramdani-Bouguessa and Rahal, 2003, the serogroups/types 14, 19F, 23F, and 6B for the periods of 2001(Tali-Maamar et al., 2012Hecini-Hannachi et al., 2014), and serogroups/types 14, 19F, 6B, 1, and 19A for the period of (Ramdani-Bouguessa et al., 2015. In Tunisia, the most prevalent serogroups for IPD in children were nearly the same: 19, 14, 23, and 4 (Charfi et al., 2012). Internationally, among the 98 known pneumococcal serotypes 11 of them account for more than 70% of IPD, in children younger than 5 years old, with serotypes 1, 5, 6A, 6B, 14, 19F, 23F being the most common (Johnson et al., 2010). Indeed, it has been described that a restricted number of serotypes is in the origin of the majority of the IPD cases worldwide. To summarize, since the report of the first studies concerning the distribution of IPD serotypes in Algeria, there was an emergence of serotype 19A, which is one of most common causes of invasive disease in developed countries in children (Geno et al., 2015). Furthermore, the studied isolates expressing serotype 19A were assigned to ST276 (M. Caniça, personal communication), which is one of the most predominant sequence-types within this serotype (Reinert et al., 2010;Ramos et al., 2014). Some of the S. pneumoniae serotypes are more prone to successfully colonize the nasopharynx, being in advantage to cause invasive disease. The carriage of S. pneumoniae may play an important role in the pathogenesis of IPD and in the transmission of this bacterium (Roca et al., 2012). Indeed, vaccination often brings a decrease in the reduction of vaccine serotype S. pneumoniae isolates, and a raise in carriage of nonvaccine serotype isolates (Dias and Caniça, 2007).
In this study, it is essential not only to emphasize the differences in serotype distribution (in IPD and asymptomatic carriage), but also to consider the presence of non-vaccine types (such as 15, 35B, and 34), particularly in carriage, despite the unavailability of pneumococcal vaccines in Algeria (Figure 2). Thus, when comparing serotype rate for IPD and nasopharyngeal carriage, serotypes 14, 1, 5, and 19A were more implicated in IPD than in carriage (p < 0.01), whereas serotype 6A was exclusively isolated from carriers (p < 0.01; Table 2). Non-typeable isolates by PCR methods should be typed by Quellung reaction in order to monitor and detect the emergence of S. pneumoniae serotype variants, and if this test is negative, the isolates should be sequenced for the cps locus to characterize variants (Bentley et al., 2006). In fact, the cpsA gene is common within most encapsulated S. pneumoniae, being a highly conserved region of the cps locus in all known pneumococcal cps operons (Bentley et al., 2006;Pai et al., 2006;Jin et al., 2009). However, the absence of cpsA amplification in the multiplex PCR scheme has been reported among rough strains, and pneumococci with mutated capsular genes, or without the cps locus (Pai et al., 2006;Ahn et al., 2012;Richter et al., 2013). Previous studies have also reported absence of cpsA gene by PCR-based serotyping of S. pneumoniae in 1-3% of the cases, particularly among serotypes 38 and 25F (da Gloria Carvalho et al., 2010;Jourdain et al., 2011). In this study, 4.1% of the isolates lacked the cpsA gene.
Among the cpsA negative isolates, the detection of the lytA gene resolved the S. pneumoniae specie (Moreno et al., 2005). Thus, we propose the scheme in Figure 1 for pneumococcal serotyping that will extend the application of multiplex PCR in laboratories, and concentrate the use of conventional method in reference laboratories only.
Similarly to other countries (Flasche et al., 2011;Ahn et al., 2012;Steens et al., 2013;von Gottberg et al., 2014), we noticed in Algeria an increase of serotypes covered by the different PCVs, particularly among IPD (Figure 4) Concerning the antibiotic resistance of S. pneumoniae from vaccine serotypes, the rates obtained were higher in no meningitis infections probably due to the large use of these antibiotics in the treatment of respiratory or urinary tract infections in Algeria. Tetracycline resistance rate might be explained by a lesser use of this antibiotic.
The fluctuations of S. pneumoniae circulating serotypes and its relation with antibiotic resistance and the PCVs coverage, reinforces the importance of S. pneumoniae serogroup/type identification and studies of antibiotic susceptibility, in preand post-vaccination periods, particularly in countries with few data, such as Algeria. These will help guide the treatment and will motivate the implementation of strategies for prevention of pneumococcal disease.

AUTHOR CONTRIBUTIONS
HZ designed the study, performed experiments, analyzed the data, and wrote the manuscript. VM analyzed the data and reviewed the manuscript. EF performed experiments and reviewed the manuscript. IM performed experiments and reviewed the manuscript. SB performed experiments. MT reviewed the manuscript. MC designed the study, analyzed the data and reviewed the manuscript. All authors read and approved the final manuscript.