Multi-Drug Resistant Staphylococcus aureus Carriage in Abattoir Workers in Busia, Kenya

Abattoir workers have been identified as high-risk for livestock-associated Staphylococcus aureus carriage. This study investigated S. aureus carriage in abattoir workers in Western Kenya. Nasal swabs were collected once from participants between February-November 2012. S. aureus was isolated using bacterial culture and antibiotic susceptibility testing performed using the VITEK 2 instrument and disc diffusion methods. Isolates underwent whole genome sequencing and Multi Locus Sequence Types were derived from these data. S. aureus (n = 126) was isolated from 118/737 (16.0%) participants. Carriage was higher in HIV-positive (24/89, 27.0%) than HIV–negative participants (94/648, 14.5%; p = 0.003). There were 23 sequence types (STs) identified, and half of the isolates were ST152 (34.1%) or ST8 (15.1%). Many isolates carried the Panton-Valentine leucocidin toxin gene (42.9%). Only three isolates were methicillin resistant S. aureus (MRSA) (3/126, 2.4%) and the prevalence of MRSA carriage was 0.4% (3/737). All MRSA were ST88. Isolates from HIV-positive participants (37.0%) were more frequently resistant to sulfamethoxazole/trimethoprim compared to isolates from HIV-negative participants (6.1%; p < 0.001). Similarly, trimethoprim resistance genes were more frequently detected in isolates from HIV-positive (81.5%) compared to HIV-negative participants (60.6%; p = 0.044). S. aureus in abattoir workers were representative of major sequence types in Africa, with a high proportion being toxigenic isolates. HIV-positive individuals were more frequently colonized by antimicrobial resistant S. aureus which may be explained by prophylactic antimicrobial use.


Introduction
Staphylococcus aureus is a common commensal of the skin, thought to persistently colonize approximately one third of the human population [1]. Nasal carriage of S. aureus is dataset to understand the effect of prophylactic use of sulfamethoxazole/trimethoprim in HIV-positive individuals on the emergence of MSSA/MRSA antibiotic resistance and associated virulence by comparing HIV-positive and HIV-negative workers. Genotypic characterization of virulent strains and their antibiotic resistance profiles will contribute to understanding the potential sources and transmission routes of S. aureus in this setting. This information will be valuable to Kenya's National Policy for the prevention and containment of AMR [39].

Description of Study Population
A total of 737/738 abattoir workers, recruited between February and November 2012, consented to a blood sample and a single nasal swab which was cultured for the presence of S. aureus. One participant declined to provide a blood sample and was excluded.
Highly resistant MDR S. aureus that were resistant to at least five antibiotic classes were isolated from one HIV-positive and three HIV-negative participants.

Discussion
This study described the nasal carriage of S. aureus in abattoir workers in western Kenya. The overall nasal carriage of S. aureus in this population was 16.0% where 15.6% were MSSA and 0.4% were MRSA. Humans are asymptomatically colonized with nasal S. aureus in the range of between 20-30% [40] with MRSA colonization varying between studies and dictated by the methodology used [41]. The carriage of S. aureus in this population was lower than expected but consistent with another study conducted in abattoir workers in Nigeria where the prevalence of S. aureus carriage was 13.5% [42].
The study population was made up of two groups: HIV-positive abattoir workers and HIV-negative abattoir workers which had an impact on the phenotypes and genotypes of S. aureus isolates in this population. The prevalence of HIV infection in this population (12%) was higher than the national average (5%). The reasons for the increased HIV positivity in this population may be related to the sociodemographic group but this was not explored further in this study. The prevalence of S. aureus nasal carriage of HIV-positive abattoir workers was significantly higher, 27.0%, when compared to HIV-negative abattoir workers, 14.5%. HIV infection is considered a risk factor for S. aureus colonization [22]. This difference in nasal carriage has been reported in previous studies in Africa. In Lagos, Nigeria, HIVpositive study participants were more likely to be colonised with S. aureus (33%) compared to HIV-negative participants (21%) [43]. These findings in the sub-Saharan region, suggest that HIV individuals are predisposed to S. aureus nasal colonization. Nasal colonization can lead to opportunistic infection in immunocompromised people, and the infection can be life-threatening if not treated promptly [2,44]. This highlights the need to monitor AMR in this population to determine treatment options and improve antimicrobial stewardship.
There was a high proportion of toxigenic strains of S. aureus carrying the pvl gene (42.9%) and the tsst-1 gene (11.9%). This is consistent with other studies conducted in sub-Saharan Africa where the carriage of pvl genes (33%) is higher than that reported in Europe (5%) [22]. The majority of pvl positive strains were ST152-MSSA, which is one of the predominant pvl-positive clones in Africa [22,36,45]. The carriage of pvl-positive S. aureus puts abattoir workers at risk of opportunistic deep skin and soft tissue infections [36,46,47]. The carriage of S. aureus pvl-positive strains was not significantly different between HIV-positive abattoir workers (37.0%) and HIV-negative abattoir workers (44.4%) which is consistent with reports from Nigeria, where the proportion of S. aureus pvl-positive strains were evenly distributed between isolates from HIV-positive and HIV-negative individuals [43].
There was an inverse relationship between pvl carriage and AMR, with 26.5% of MDR isolates being pvl positive compared to non-MDR isolates (48.9%) (p = 0.025). This differs from other studies in the region where pvl carriage has been associated with MDR. An association between pvl carriage and sulfamethoxazole/trimethoprim resistance has been observed in Gabon and Nigeria among HIV-positive individuals [43,46].
The prevalence of MRSA carriage (0.4%) identified in this study was low compared to studies conducted with abattoir workers in Europe (5.6%) [48], and the USA 3.6% [16]. However, our results are consistent with other reports of MRSA carriage in Kenya (0.8%) [26] but much lower than studies of MRSA cultured from hospital patients in Kenya (53.4%) [49]. This may suggest that MRSA infection in Kenya is predominantly linked to the hospital environment rather than acquired from the community. The three MRSA isolates belonged to ST88, which is referred to as the "African" community-associated (CA-MRSA) clone [50] but is not the most reported MRSA sequence type in Kenya, which is ST239 [24]. ST88 has been reported in pigs and workers in a Nigerian abattoir and may indicate an animal source [51]. Antimicrobial use in animals is not regulated in Kenya and the most frequently used antibiotics in animals in western Kenya are oxytetracycline and penicillinstreptomycin [52]. Further work to understand MRSA carriage in animals, as well as spread in the human population and environment, is required.
There was high genotype-phenotype concordance between resistance genes detected and antimicrobial susceptibility test (AST) results for most antimicrobials as has been previously reported [25]. However, there were two mecA positive MRSA strains that were susceptible to oxacillin and one of these was also susceptible to cefoxitin. This may be due to a misclassification error, although the sensitivity of the VITEK 2 instrument for detecting oxacillin resistance is high (97.5%) [53]. Alternatively, this may indicate oxacillin susceptible mecA MRSA (OS-MRSA) strains are circulating in this environment. OS-MRSA strains have previously been identified in other parts of Africa mainly associated with ST88 as described here [54]. Information regarding the presence of the mecC gene in these isolates was not available. The presence of OS-MRSA in this setting may complicate treatment options that are based solely on AST results, since OS-MRSA may be misidentified as MSSA, and these isolates can develop β-lactam resistance following antibiotic therapy [55].
Sulfamethoxazole/trimethoprim is an effective antibiotic combination in the treatment and prevention of bacterial infections in people who are HIV positive and has been used to treat Pneumocystis jiroveci pneumonia and other bacterial infections in severely immunocompromised HIV-positive individuals. The prophylactic use of sulfamethoxazole/trimethoprim in all HIV-positive individuals regardless of CD4 counts, especially in regions having high prevalence of malaria and/or severe bacterial infections, such as sub-Saharan region [56], may have resulted in the high prevalence of sulfamethoxazole/trimethoprim resistant S. aureus in Africa [43,46]. This, coupled with the extensive use of penicillin and tetracycline for use in animal production for both therapeutic and non-therapeutic purposes in Africa [12], has created a favorable environment for the emergence of multidrug resistance S. aureus through antibiotic related selective pressure [57,58]. These MDR S. aureus reduce the treatment options for effective treatment for HIV-positive individuals with opportunistic infections. With the presence of MDR S. aureus that are resistant to additional multiple resistant combinations, including erythromycin, clindamycin, ciprofloxacin and gentamicin, the treatment options will be further diminished and become more expensive. The improved availability of antiretroviral therapy (ART) in sub-Saharan Africa, has led to reduction of severely immune compromised HIV-positive individuals and fewer cases of serious Pneumocystis jiroveci pneumonia and opportunistic infections among HIV-positive individuals [59]. Improved health outcomes for HIV-positive individuals through access to ART may lead to reduced prophylactic use of sulfmethoxazole/trimethoprim in this region. This will aid in the preservation of therapeutic advantages of this affordable drug in treatment and prevention of bacterial infections.
Multi-drug resistant S. aureus strains colonizing HIV-positive abattoir workers may be a risk for SSTI since 25% of workers reported being injured at work [60]. MDR S. aureus could also be transmitted to the community directly or through the meat supply chain to consumers [4]. In addition, if the respective abattoir workers are hospitalized, the strains can be spread to hospital staff and compromised inpatients, threatening effective treatment of resultant infections [61]. Thus, MDR S. aureus-colonized abattoir workers pose a public health problem in hospitals, community, and food industry in Busia County, which can extend to other parts of Kenya and neighboring countries of Kenya since Busia is a border town.
There is a need for increased monitoring of antibiotic usage and surveillance measures for AMR bacteria in both animals and humans in this region already burdened by HIV/AIDS infection and where there is rapidly increasing demand for meat products caused by population growth and urbanization [12]. Additionally, there is a need for strategies to promote the prudent use of antimicrobials and antimicrobial stewardship as described by the global strategy [62]. It is particularly important to strategize on the appropriate use of sulfamethoxazole/trimethoprim, tetracyclines and penicillin in human and animal healthcare and food production in sub-Saharan region, since there were high proportions of resistant isolates to these antimicrobials. It has been demonstrated that a reduction of antibiotic consumption leads to decreased prevalence of antimicrobial resistance [63]. This can be done through national action plans for the prevention and containment of AMR with contributions from human and animal health agencies [39].
There was a delay between the collection of samples and the publication of these findings which may limit the usefulness of these results. However, given the lack of available data regarding the circulating MSSA and MRSA strains and antimicrobial resistance profiles of S. aureus in this population the data is a valuable contribution to knowledge regarding AMR in the region and may prove a useful baseline for comparison to future studies. Data regarding the circulating MSSA and MRSA strains in livestock was not available at the time of this publication. This information would have been useful to understand the potential for transmission of isolates between livestock and workers and is a data gap that should be targeted in future research.

Study Site
The study area was a 45 km radius from Busia town including most of Busia County, and parts of Bungoma, Siaya and Kakamega Counties spanning the 3200 km 2 of the Lake Victoria ecosystem. The predominant industry in the study area is subsistence agriculture [64]. A census of all abattoirs (n = 156) was conducted in the study area in 2012. Fourteen abattoirs declined to participate. Participants were recruited from 142 abattoirs, 84 ruminants and 58 porcine [60].

Study Population
A total of 738 abattoir workers were recruited into the study between February and November 2012 from a total of 1005 workers (73.3%) in the selected slaughterhouses. In abattoirs with 12 workers or less all consenting workers were recruited and in abattoirs with more than 12 workers a random selection of twelve workers were recruited [60].

Data and Sample Collection
All participants were informed of the project objectives and protocol by a clinical officer who collected signed informed consent. Data was collected using a structured questionnaire regarding demographic details, health events and recent antimicrobial use. Nasal samples were collected by rotating a sterile swab five times in both anterior nares, from consenting abattoir workers. The swabs were inoculated in tryptone soya broth with 6% salt and transported in cool boxes to the laboratory for culturing.
Blood was collected by a clinical officer into 4 mL EDTA vacutainers using a butterfly catheter (BD Vacutainer ® Safety-Lok™ blood collection set). Samples were transported in cool boxes to the laboratory in Busia. Whole blood samples were stored frozen at −40 • C until transportation to the International Livestock Research Institute (ILRI) laboratory in Nairobi for long term storage at −80 • C. HIV testing was performed on whole blood using the SD Bioline HIV 1/2 Fast 3.0 test strips (Standard Diagnostics Inc., Suwon-si, South Korea).
Swabs were streaked onto Mannitol Salt agar and incubated at 37 • C in air overnight. Suspect S. aureus isolates, those fermenting mannitol and producing yellow colonies, were stocked in tryptone soya broth with 10% glycerol and stored at −40 • C and transported on dry ice to Nairobi. Presumptive S. aureus isolates were further cultured onto mannitol salt agar (MSA) and sub-cultured to obtain pure culture. The S. aureus isolates were identified using Gram reaction (Gram-positive cocci in clumps), catalase, coagulase (tube method using rabbit plasma) and DNase tests. Initially, a long sweep of the colonies was done to allow preservation of genetic diversity of nasal carriage of the participant. During whole genome sequencing, some samples were shown to consist of mixed isolates and these samples were recultured to select single colonies for sequencing and antimicrobial susceptibility testing.
Multi-drug resistant S. aureus were defined as isolates that were resistant to three or more different antimicrobials using the VITEK results [67][68][69]. Isolates resistant to 5 or more antimicrobials were described as highly multi-drug resistant.

Statistical Analyses
Statistical analysis was performed using the chi-squared test. A p-value <0.05 was considered an indication of significant difference.

Conclusions
This study identifies the circulating MSSA and MRSA strains in a population occupationally exposed to livestock in rural western Kenya. This gives an improved understanding of the epidemiology of S. aureus particularly the strains, sources, and risk groups in a setting that has not been previously studied. More importantly, the study indicates the levels of AMR and prevalence of toxigenic genes in S. aureus isolates, which is particularly important in this community with high prevalence of immunocompromised individuals. This information can contribute to developing measures for the prevention and containment of AMR.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/antibiotics11121726/s1, Table S1: Sequence types of MSSA and MRSA strains isolated from nasal of HIV positive and negative abattoir workers Busia County.