Streptococcus pneumoniae serotype specific anti-microbial susceptibility profiles among PCV-10 vaccinated and unvaccinated children attending Gertrude’s Children’s Hospital: a cross-sectional study [version 1; peer review: 2 approved with reservations]

Background: The spread of antimicrobial resistance threatens effective control and treatment of pneumococcal disease worldwide. In Kenya, an estimated one in every five children dies from pneumococcal disease every year. Of these, ≥50% are attributable to antibiotic resistance. Consequently, the WHO has recommended that continuous regional surveillance be done to detect early resistance to available antibiotics and make necessary changes. We therefore investigated antimicrobial susceptibility patterns of Streptococcus pneumoniae among PCV-10 vaccinated and unvaccinated children ≤5 years old at Gertrude's Children’s Hospital. Methods: A 0.5 McFarland standard of freshly subcultured organisms were inoculated on Mueller–Hinton plates with 5% sheep blood agar. A standard disk dispenser was used to dispense various antibiotic disks on the Mueller–Hinton agar plate. Incubation was done overnight (20-24 hours) at 37oC in 5% CO2 and clearance zones read using a Vanier caliber. Antimicrobials tested included vancomycin (30μg, ≥17mm); erythromycin (15μg, ≥21mm); clindamycin (2μg, ≥19mm); oxacillin (1μg, ≥19mm) and ceftriaxone (1μg, ≥30mm). Results: Thirty nine (92.86%) Streptococcus pneumoniae isolates were susceptible to erythromycin; 39 (92.86%) were susceptible to vancomycin; eight (19.86%) Streptococcus pneumoniae isolates were susceptible to oxacillin, while 34 (80.95%) were non-susceptible; 40 (95.24%) isolates were susceptible to clindamycin; and 24 (57.86%) isolates were susceptible to ceftriaxone, while 18 (42.86%) were nonOpen Peer Review


Introduction
Globally, there is a growing concern over the burden of antimicrobial resistance (AMR) because it is presently estimated to account for more than 700,000 deaths per year worldwide (WHO, 2014). Hypothetically, studies have reported that if no appropriate measures are taken to arrest this unfortunate trend, AMR may cost up to 10 million lives and around US $100 trillion per year by 2050 (de Kraker et al., 2016). What is more appalling is that, contrary to some other health trepidations, AMR is a problem that concerns every country regardless of its level of income and development as resistant pathogens do not respect borders and jurisdictions (Cole, 2016).
Diverse forms of pneumococcal disease are rated among the highest vaccine preventable etiologies of child morbidity and mortality globally (Lowth, 2015). This is true despite availability of effective vaccines and antibiotic agents; it is therefore a matter of grave concern that Streptococcus pneumoniae is developing resistance to available antimicrobial agents. Paucity of quality and verifiable data on the pneumococci antimicrobial susceptibility profiles is a precursor for inadequate treatment guidelines, especially for countries that are still largely developing (Assamala, 2014). The gap in public health capacity is also an issue, given the changing resistance machineries and the emergence of multidrug-resistant Streptococcus pneumoniae that can only be detected through systematic screening in quality assured microbiology laboratories (O'Neill, 2016).
Overtime, there has been unswerving universal increase in antibiotic resistance by Streptococcus pneumoniae isolates, a trend that has made treatment options more challenging (Henriques-Normark & Tuomanen, 2013). The bacterium's polysaccharide capsule is a major virulence factor (Mitchell & Mitchell, 2010). The bacterial serotype is determined on the basis of the capsule's antigenic features, which have also been expansively correlated with AMR patterns. Although Streptococcus pneumoniae antimicrobial resistance is a wide-reaching concern, low and middle income countries (LMICSs) are vulnerable to exceptional difficulties because substitute and/or effective therapeutic choices are either unattainable or too expensive for them to afford (Reardon, 2014). Experiential prescription of antibiotics for treatment of various forms of pneumococcal disease without data on antimicrobial susceptibility profiles is a very common practice, especially in Africa (Feldman & Anderson, 2016). Purchase of antibiotics over the counter and failing to comply with rules of antibiotic consumption and/or use significantly augments occurrence of AMR in Kenya (Donkor & Badoe, 2014). Unregulated and illicit hawking of major antibiotics at bus-stops by unlicensed persons and back-street manufacturing of counterfeit antimicrobial agents contribute majorly to pneumococci AMR (Laxminarayan et al., 2013).
The end result of 'abuse' of antimicrobial agents meant for treatment of Streptococcus pneumoniae infections is apparent; the pneumococcus has connivingly developed resistance to novel antibiotic agents (Ventola, 2015). Since the rates of Streptococcus pneumoniae AMR vary on the basis of geographic location, time, age, and site of infection, AMR surveillance must be incessant and guidelines be derived from local AMR data (Feldman & Anderson, 2016). As AMR impinges on the quality of patient care, concerted efforts must be made to understand the best way to circumvent it. To treat various forms of pneumococcal disease, antibiotics are the most commonly prescribed drugs in hospitals. Judicious use of antimicrobial agents is therefore one of the major avenues that can be employed to arrest the worrying trend of AMR.
An audit of Streptococcus pneumoniae antibiotic susceptibility patterns is imperative in the execution of a lucid and pragmatic antibiotic strategy. We therefore sought to analyze antimicrobial resistance profiles to commonly used antibiotics meant for treatment of pneumococcal disease among children ≤5 years in Nairobi County.

Methods
This was a descriptive cross-sectional study. Streptococcus pneumoniae anti-microbial susceptibility patterns among PCV-10 vaccinated and unvaccinated children below five years of age were measured between May 2017 to February 2018.

Ethical statement
Ethical approval for this study was given by the Kenyatta University Ethics Review Committee {KU/ERC/APPROVAL/VOL-UME.1 (12)} and permission to carry out the study obtained from the National Commission of Science Technology and Innovation (NACOSTI/P/17/65428/15801). Gertrude's Children's Hospital (GCH) research committee issued permission for the study to be conducted at the GCH outpatient clinic (GCH/ ERB/VOLMMXVII/121). Biological mothers and/or legal guardians to eligible study subjects were approached to sign a written informed consent to allow their children participate in the study; this was done after reading and having the study explained to them. Participation in the study was completely on voluntary terms.

Study location
This study was conducted among children attending GCH, Nairobi County. GCH is the leading stand-alone hospital in East and Central Africa specializing in pediatric care. The hospital is accredited by the Joint Commission International (JCA). Streptococcus pneumoniae laboratory isolation, disk diffusion for AMR profiling and isolate stocking was done at the GCH main laboratory, while Streptococcus pneumoniae serotyping was done at KEMRI Wellcome Trust, Kilifi.

Study population
To take part in this study, biological mothers and/or legal guardians were approached at the GCH outpatient clinic when they brought their children for medical attention. For inclusion in this study, the subject was required to: be at most five years of age; be a resident of either Nairobi or Kiambu Counties; have been vaccinated with either three doses of PCV-10 or none (any child who had received less than three doses was considered ineligible); not be consuming any antibiotics at the time of the study; be clinically diagnosed by a physician at the site as having some form of pneumococcal disease; and the parent or guardian needed to voluntarily sign an informed consent form. Subjects who had known HIV/AIDS infection or had any immunosuppressive conditions were excluded from the study.

Sample size determination
To determine the minimum sample size, the Fisher's formula (1998) was used, with a prevalence rate of 16% (Tuti et al., 2017) being the current occurrence rate of pneumococcal disease in Kenya. Where: n = desired minimal sample size; z = standard normal deviation = 1.96 (from the tailed normal table); p = prevalence rate; and m = desired degree of accuracy at 95% confidence level = 0.05. n = 1.96 2 x 0.16 (1-0.16) / 0.05 2 = 206; therefore, n = 206.

Sociodemographic characteristics
A standardized questionnaire was used to collect data on the subjects' sociodemographic features. The characteristics assessed included whether the child attended daycare or not, exposure to both cigarette and cooking smoke, breastfeeding frequency during the first six months after birth (classified as: none, meaning the child was never breastfed at all; moderate, meaning the child partially breastfed and partially given other food; and exclusive, meaning the child was entirely dependent on breast-milk), child's age and PCV-10 immunization status.
Isolation and identification of Streptococcus pneumoniae Nasopharyngeal swabs collected using flocked cotton swabs (Copan) were the specimen of choice for this study. All swabs were initially suspended in Amies Media and transported to the laboratory within three hours of collection. Each swab was inoculated onto a selective gentamicin blood agar (GBA) enriched with 5% sheep blood. The plates were incubated at 37°C in 5% CO 2 and examined at 16-24 hours and then again at 40-48 hours for growth of Streptococcus pneumoniae.
Isolates were identified as Streptococcus pneumoniae on the basis of colony morphology (mucoid, draughtsman's appearance, α-haemolysis), susceptibility to optochin (ethyl hydrocupreine hydrochloride) and, where necessary, solubility in bile salts (desoxycholate salts). The isolation of a single colony indicated nasopharyngeal carriage.
Optochin sensitivity test All colonies on GBA plates demonstrating Streptococcus pneumoniae characteristics were evenly inoculated on plain blood agar (BA) plates. This was done using a sterile inoculating loop. An optochin disk was aseptically placed on the preparation using a sterile antibiotic dispenser and incubated for 24-48 hours at 37°C in 5% CO 2 atmosphere. After 24-48 hours, a clearance zone of ≥14mm around the optochin disk was a confirmation that the organism was Streptococcus pneumoniae. Plates with clearance zones <14mm but ≥6mm were further subjected to a bile solubility test. Bile soluble stocks confirmed that the organism was Streptococcus pneumoniae.
Bile solubility test Using a sterile cotton inoculating swab, Streptococcus pneumoniae colonies from the blood agar plate were added to 1.0ml of 0.85% saline solution to achieve 1.0 McFarland standard turbidity. The cell suspension was divided into two equal volumes of 0.5ml/tube. Then, 0.5 ml of 2% sodium desoxycholate solution (bile salts) were added to one tube and 0.5 ml of 0.85% saline solution to another tube. The two tubes were gently mixed and incubated for up to two hours at 37°C in 5% CO 2 . The tubes were vortexed while observing for any clearance of turbidity within 10 minutes and subsequently checked until two hours later. Any clearance in the tube containing bile salts and not in the saline tube confirmed that the organism was Streptococcus pneumoniae.

Quellung reaction
Capsular serotyping was done using the Quellung reaction test according to (Habib et al., 2014). Frozen vials containing Streptococcus pneumoniae stocks stored at -70°C were thawed at room temperature for about 30 minutes. Next, 10µl of the stored Streptococcus pneumoniae cells were suspended in 50µl PBS and gently vortexed. Subsequently, 10µl of the suspended cells were added on to a glass slide and mixed with 10µl pooled antisera (Statens Serum Institute, cat. No.16744). The glass slide was swirled gently while observing for any agglutination reaction until a positive reaction was observed with various pooled antisera. The process was repeated with individual groups under various antisera pools. After that, 10µl of the suspended cells in PBS were added to a glass slide and mixed with various Streptococcus pneumoniae serotype-specific antisera included in the antisera pools that gave a positive reaction. This was done until a positive reaction with the particular serotype specific antisera was observed.
Those serotypes that did not belong to any pool were typed directly until a positive agglutination reaction was observed. The cells/PBS/serotype-specific antisera mixture on the glass slide were covered with a cover slip and observed under a phase contrast microscope with a ×400 objective lens with oil emulsion.

Disk diffusion method
Antimicrobial susceptibility testing (AST) was performed using the Kirby-Bauer disc diffusion method according to Clinical and Laboratory Standards Institute (CLSI) guideline (Hegstad et al., 2014). A 0.5 McFarland standard of freshly subcultured pneumococci organisms was inoculated on a 150 mm Mueller-Hinton plate with 5 % sheep blood (MH-BA). A standard disk dispenser was used to dispense the various antibiotic disks on the MH-BA agar (CLSI, 2004). Incubation was done overnight (20-24 hours) at 35°C in 5% CO 2 and various clearance zones read with the aid of a vernier caliber. Antimicrobial agents tested included vancomycin (30µg, ≥17mm), erythromycin (15µg, ≥21mm), clindamycin (2µg, ≥19mm), oxacillin (1µg, ≥19mm) and ceftriaxone (1µg, ≥30mm). Because of ambiguity caused by poorly resolved concentration gradient around disks containing cephalosporins, this method may not give accurate results for ceftriaxone.

Statistical analysis
Data were entered, cleaned and analyzed on SPSS version 22. Descriptive statistics were used to summarize socio-demographic and antibiotic susceptibility patterns of Streptococcus pneumoniae isolates.
Adjusted logistic regression was used to establish the strength of relationship between selected risk factors and susceptibility patterns of the selected antibiotic agents studied.
Compared with males, Streptococcus pneumoniae isolated from female subjects were more likely to be resistant to erythromycin (OR 1.94; CI 0.23, 16.10) but odds of being resistant to vancomycin, oxacillin, clindamycin and ceftriaxone decreased when the subject was female: OR 0.14 (95% CI 0.01, 2.81), OR 0.51 (95% CI 0.11, 2.26), OR 0.51 (95% CI 0.11, 2.26), OR 0.2 (95% CI 0.01, 4.43) and OR 0.56 (95% CI 0.17, 1.86), respectively. The odds of being resistant to erythromycin, vancomycin, oxacillin, clindamycin and ceftriaxone decreased when the subject was aged 13-24 months, 25-36 months and 37-48 months old, but the odds increased when the subject was aged 49-60 months old. The odds of being resistant to erythromycin, vancomycin, oxacillin and clindamycin increased when the subjects had consumed antibiotics two weeks prior to the study. However, the odds of being resistant to ceftriaxone decreased. The odds of being resistant to the tested range of antimicrobial agents decreased when the subjects had had none, moderate and exclusive breastfeeding types with the exception of oxacillin whose odds of resistance increased when the subjects had had moderate breastfeeding. Odds of being resistant to the tested antimicrobial agents decreased when the subjects had received a full dose of PCV-10, except for vancomycin and ceftriaxone whose odds of resistance increased when the subjects had received a full dose of PCV-10. The odds of being resistant to the tested antimicrobial agents significantly increased when the subjects had been attending daycare centers (Table 4).

Discussion
Antimicrobial resistance is a major global public health hazard in the waiting lounge. Infections that are resistant to antibiotics are often hard to treat, the cost of managing them is much higher and they generally pose more stress on families' finances. Its envisaged that by 2050, general global mortality due to antimicrobial resistance will be at 10 million people annually (de Kraker et al., 2016); Streptococcus pneumoniae is likely to be one of the major culprits, the need for its earlier containment has never been more urgent.
pneumococcal infections. This, coupled up with failure to consume the correct regimen, results in antibiotic selection pressure, which consequently brings forth resistant strains that are then horizontally spread across the population (Neu, 1992). This would explain the unexpected high resistance levels of Streptococcus pneumoniae serotypes to oxacillin. These results are consistent with those of Bronzwaer et al. (2002), who reported increased (53%, n=115) antibiotic failure, especially to different derivatives of penicillins.
High levels of sensitivity and tolerance to clindamycin and ceftriaxone agree with findings by scientists at the St. Jude Children's Research Hospital, who demonstrated that clindamycin and azithromycin, which kill bacteria by inhibiting protein synthesis, are more effective than a standard first-line treatment with the beta-lactam antibiotic ampicillin, which causes the bacteria to lyse, or burst (Sutcliffe et al., 1996).
Clindamycin is especially recommended for Streptococcus pneumoniae treatment because of its spectrum of activity includes group A streptococci (Puopolo et al., 2007). Usually, it is reserved for cases of Streptococcus pneumoniae with high resistance to penicillins and/or cases of subjects being allergic to first line antibiotic therapies (Karlström et al., 2009). The high level of ceftriaxone tolerant Streptococcus pneumoniae (CT-Streptococcus pneumoniae) observed in this study is concerning. This is because broad spectrum cephalosporins like ceftriaxone are recommended for invasive pneumococcal disease in cases where other first line antibiotics have failed (Molyneux et al., 2011). Resistance would only be expected in cases where the subjects had been inappropriately exposed to other cephalosporins as the first line of therapy (Von Mollendorf et al., 2014). Resistance to penicillins has also been reported to have a nexus with resistance to cephalosporins as they both target the beta-lactam ring of the bacteria (Drawz & Bonomo, 2010). Irregular and non-prescribed purchase of antibiotics at local pharmacies and shops may be the other culprits behind these unfortunate findings.
Pneumococcal resistance to ceftriaxone was found to be high (up to 66-67%) in Estonia (Tadesse et al., 2017) and Romania, while the rates were lower in Norway (4.6%) (Steens et al., 2015), the Netherlands (12.9%) (Voets et al., 2012) and the Czech Republic (15.7 %) (Zemlickova et al., 2010). Streptococcus pneumoniae has over 90 serotypes belonging to different groups according to their capsular polysaccharides and the degree of cross-reactivity (Hathaway et al., 2012). The various Streptococcus pneumoniae serotypes behave differently when exposed Odds ratio (OR)=1: exposure does not affect odds of outcome; OR>1: exposure associated with higher odds of outcome; OR<1: exposure associated with lower rates of outcome. Confidence intervals (CI) are used to estimate the precision of the OR. The higher the CI the lower the precision of the OR and the lower the CI the higher the precision of the OR. P-value is a measure of the significance level of the exposure.
to different antibiotic agents (Reinert, 2009). In this study, seven serotype 28F strains out of eight were tolerant to oxacillin, while only three were sensitive to ceftriaxone. Oxacillin, which is a penicillin, is normally given as a first line of treatment for the various forms of pneumococcal disease infections and ceftriaxone is a broad spectrum antibiotic normally given after penicillin has failed (File, 2006). The mode of resistance to the two antimicrobial agents is shared to some extent as both involve the use of a four atom beta-lactam ring (Kong et al., 2010). As such, resistance in one is likely to influence resistance in another. Inappropriate exposure to one or both of these agents prior to the collection of the study samples may explain the high resistance levels in both.
Most strains of serotypes 6A, 3, 23A, 23B, 20, 11A and 17 were oxacillin and ceftriaxone tolerant but highly susceptible to erythromycin, vancomycin and clindamycin. These results agree with findings from a previous study by Liñares et al. (2010), who demonstrated Streptococcus pneumoniae serotype susceptibility patterns to different antibiotics and intimated that serotype 28F and 6A are susceptible to erythromycin, vancomycin and clindamycin but highly resistant to penicillin and ceftriaxone. Serotypes 19A, 7C, 13, 15B, 21, 35B, 35F, 39, 48, and the untypeable Streptococcus pneumoniae were highly intolerant to erythromycin, clindamycin, vancomycin, oxacillin and ceftriaxone. While most Streptococcus pneumoniae serotypes in this study were susceptible to majority of the antimicrobials studied, resistance to oxacillin and ceftriaxone was clinically significant and largely worrying. A similar study examined antibiotic susceptibility of non-invasive Streptococcus pneumoniae in children aged less than six years with signs of an acute respiratory tract infection in Moscow, Russia. Non-susceptibility to penicillin, erythromycin and MDR was found in 28%, 26% and 22% of pneumococcal strains, respectively (Mayanskiy et al., 2014).
The slightly lower rates of Streptococcus pneumoniae resistance may be due to the enrolment of both vaccinated unvaccinated children and differences in the enactment and implementation of the antibiotic use/consumption policies. Although none of the risk factors had a significant influence on the Streptococcus pneumoniae nasopharyngeal carriage, odds demonstrated that age, gender, recent consumption of antibiotics by subjects, attendance at daycare centers, breast-feeding type and vaccination status had a bearing on the effectiveness of antibiotics. For instance, being female and young increased the subject's susceptibility to erythromycin tolerance, while decreasing susceptibility to vancomycin, clindamycin, ceftriaxone and oxacillin. Children exposed to antibiotics within the two weeks prior to sample collection and PCV-10 unvaccinated children had increased odds of being resistant to all the antibiotics studied. Breastfeeding type, attendance at day-care centers and PCV-10 vaccination status play a major role in antibiotic susceptibility according to the findings of this study.
With the exception of oxacillin, the odds of Streptococcus pneumoniae being resistant to clindamycin, vancomycin, erythromycin, ceftriaxone decreased when subjects had been moderately or exclusively breastfed. According to Lynch & Zhanel (2009), due to the likelihood of horizontal transfer of Streptococcus pneumoniae resistant genes within a community, subjects who attended daycare centers and had not received PCV-10 immunization had increased chances of being resistant to all the sampled antibiotics. These results are consistent with findings from Lynch & Zhanel (2010) who postulated that age, PCV7 vaccination, attendance at day care or school, previous respiratory infection and non-susceptibility to penicillin were significantly or insignificantly associated with antibiotic resistance. While some of the results may not have strong research significance, they pose significant clinical concerns.

Data availability
This project contains the following underlying data: Name Extended data This project contains the following extended data: Questionnaire

Introduction
Paragraph 1-Repetition of globally and worldwide in the same sentence. The focus of this paragraph need to be antimicrobial resistance.
○ Paragraph 2 -Rewrite to introduce Streptococcus pneumonia as the causative agent of pneumococcal disease.
○ Paragraph 3 -Should focus briefly on the mechanisms of antimicrobial resistance.
○ Paragraph 4 -The problem and uniqueness of the study.

Methods
This was a descriptive cross-sectional study conducted among children attending Gertrude's Children's Hospital (GCH) in Nairobi.

○
Ethical statement: This was obtained and approved from ….. "not given…". This statement need to be rewritten to reflect: Written consent was sought from each child guardian/parent with each consent form indicating clearly the study purpose, any anticipated consequences of the research, the anticipated uses of the data, possible benefits of the study and possible harm, confidentiality of the data and the option to withdraw their children participation at any given time.

Results
Confirm if the 4 kids were smokers, and if smokers what was the criteria used.

○
Consume antibiotics does not sound medically -better word to use is take antibiotics.

○
Have headings for the several parts of the results focusing on: Socio-Demographic characteristics -a brief summary (Table 1).
○ Antimicrobial susceptibility patterns of Streptococcus pneumoniae among PCV-10 vaccinated children ≤5 years (N =?) by selected antibiotic agents. Table 3 need to be split to show vaccinated children, then describe the results in text.
○ Antimicrobial susceptibility patterns of Streptococcus pneumoniae among PCV-10 unvaccinated children ≤5 years (N=?) by selected antibiotic agents. Table 3 need to be split to show unvaccinated children, then describe the results in text.