ABSTRACT
A systematic review and meta-analysis of antibiotic-resistant Gram-positive bacteria in Africa, showing the molecular epidemiology of resistant species from animal, human and environmental sources, is lacking. Thus, the current burden, type, and sources of Gram-positive bacterial resistance and their dissemination routes from farm to fork is absent. To fill this One Health information gap, we systematically searched PubMed, Web of Science and African Journals Online for English research articles reporting on the resistance mechanisms and clonality of resistant Gram-positive bacteria in Africa within 2007 to 2018. The review and all statistical analysis were undertaken with 130 included articles.
From our analyses, the same resistant Gram-positive bacterial clones, resistance genes, and mobile genetic elements (MGEs) are circulating in humans, animals and the environment. The resistance genes, mecA, erm(B), erm(C), tet(M), tet(K), tet(L), vanB, vanA, vanC, and tet(O), were found in isolates from humans, animals and the environment. Commonest clones and mobile genetic elements identified from all three sample sources included Staphylococcus aureus ST5 (n=208 isolates), ST 8 (n=116 isolates), ST 80 (n=123 isolates) and ST 88 (n=105 isolates), and IS16 (n=18 isolates), Tn916 (n=60 isolates) and SCCmec (n=202 isolates). Resistance to penicillin (n=4 224 isolates, 76.2%), erythromycin (n=3 552 isolates, 62.6%), ampicillin (n=1 507 isolates, 54.0%), sulfamethoxazole/trimethoprim (n=2 261 isolates, 46.0%), tetracycline (n=3 054 isolates, 42.1%), vancomycin (n=1 281 isolates, 41.2%), streptomycin (n=1 198 isolates, 37.0%), rifampicin (n=2 645 isolates, 33.1%), ciprofloxacin (n=1 394 isolates, 30.5%), clindamycin (n=1 256 isolates, 29.9%) and gentamicin (n=1 502 isolates, 27.3%) (p-value <0.0001) were commonest.
Mean resistance rates of 14.2% to 98.5% were recorded in 20 countries within the study period, which were mediated by clonal, polyclonal and horizontal transmission of resistance genes. A One Health approach to research, surveillance, molecular epidemiology, and antibiotic stewardship to contain ABR should be prioritized.
HIGHTLIGHTS
Gram-positive bacteria (GPB) isolated from human, animal and environmental samples were of the same clones and/or shared common resistance genes and mobile genetic (MGEs).
Multidrug resistant (MDR) clones such as S. aureus ST5 and E. faecium ST80 were isolated from human, animal and environmental sources.
mecA, erm(B), erm(C) tet(M/K/L), and vanA/B/C were common in GPB, including VRSA.
Mean drug resistance rates of isolates from humans, animals and the environment were respectively 62.0% (95% CI: 54.7 - 69.3%), 68.2% (95% CI: 58.0 -78.4%) and 84.6% (95% CI: 69.9 – 99.31%) (P-value < 0.0001).
SCCmec, IS16, and Tn916 mobilized mecA, erm(B) and tet(M) respectively across various GPB species isolated from animals, humans, and the environment.
A One Health approach to studying antibiotic resistance mechanisms and molecular epidemiology of GPB is warranted.
1. INTRODUCTION
Antibiotic resistance, a threat to public health
Limited research and surveillance data in Africa makes it impossible to track and monitor the true burden of antibiotic resistance (ABR) 1, particularly the distribution and dissemination of resistance genes between humans, animals and the environment. According to a recent WHO report, the potential for ABR to lead to higher mortalities and morbidities in low- and middle-income countries such as Africa may even be greater as a result of the higher burden of bacterial infections, limited diagnostic capacity and lower access to second-line antibiotics1,2. This makes it imperative to have a One Health analysis that describes the burden and epidemiology of resistance genes in bacteria isolated from humans, animals and the environment 3.
In a recent review, Gram-positive bacteria (GPB) were responsible for a high proportion of infections among children and showed a high level of resistance to WHO-recommended drugs in Africa 4. In some African regions, as many as 80% of Staphylococcus aureus infections are methicillin-resistant S. aureus (MRSA), which show resistance to most standard licensed drugs including quinolones and peptides 25. Although Enterococcus spp. are mostly not as virulent as S. aureus, their multidrug resistance (MDR) propensities restrict drug options for clinicians 7. Patients infected with MRSA are estimated to be 64% more likely to demise than those infected with methicillin-susceptible S. aureus (MSSA) 6.
Reviews addressing GPB in Africa have reported on increasing rates of ABR from blood-stream infections, pneumonia, urinary tract infections and meningitis caused by Streptococcus agalactiae, S. aureus, Streptococcus pneumoniae and Enterococcus faecium in both children and adults. Sepsis due to S. agalactiae accounts for about 26% of all neonatal deaths and 10% maternal deaths in Sub-Saharan Africa 8 However, the potential dissemination of these resistant strains from farm (environment and animals) to fork (humans), are less described.
Sources and anthropogenic activities driving resistance
High-level ABR has been reported in humans, animals and the environment, with indiscriminate antibiotic use being fingered as a major contributor in Africa. Resistance genes have been detected in surface water fed with runoff effluents from farms utilizing antibiotics, hospitals, and sewage processing plants as well as in ground water 9-11. Furthermore, genes mediating resistance to last-resort GPB antibiotics such as vancomycin have been recovered from raw milk and animal products, pigs, wild animals (buffalo, zebra and cattle), waste water, effluents and patients, implicating veterinary and agricultural use of antibiotics as potential sources of resistance genes in humans 12-14. These reports suggest that a larger share of the antibiotics that end up polluting the environment and communities emanate from livestock production 15-17. This interconnectivity between animals, humans and the environment, explains the need to adopt a One Health research policy.
Several studies have reported high rates of MDR among GPB isolates from humans, animals and the environment in Africa, mainly as a result of overuse, underuse and wrong choice of antibiotics 18-24. Different factors have been implicated in the high rate of ABR to the limited drugs in Africa. These include: unrestricted access to antibiotics over-the-counter without prescription such as selling on the streets; inadequate hygienic practices; uncontrolled usage of antibiotics as growth promoters in food animals production; wrong diagnosis and prescription, off-label use and errors in dosage regimens; use of untreated poultry and cattle manure to fertilize agriculture lands; extensive use of broad-spectrum antibiotics in poultry production; and inefficient chlorination of hospital wastewater effluents before discharge into the environment10,18,22,25-29. Additionally, inadequate knowledge of animals’ diseases, misdiagnosis and poor antibiotic handling practices in animal production add up to the overall burden of ABR in Africa 17.
Molecular ABR mechanisms
Selective pressures exerted by various antibiotics used in human and veterinary medicine, as well as in agriculture, have resulted in the emergence and dissemination of numerous mechanisms of resistance in GPB in Africa. Commonly reported mechanisms include blaZ, erm(B), mecA, tet(M), vanB and vanC 30-33. These resistance genes have been found to be associated with mobile genetic elements (MGEs) such as transposons, conjugative plasmids, integrons, and insertion sequences, which are capable of mobilizing resistance genes across a wide spectrum of bacterial species 34,35. SCCmec, Tn916 and IS16 are notable MGEs that carry major ABR determinants in Africa and are transmissible between clones of the same or different bacteria species by a conjugative mechanism. These MGEs have the potential to thus spread resistance genes from environmental and animal bacterial hosts to human pathogens in Africa; they have therefore been analysed herein 36-38.
Purpose of this review
Excellent reviews addressing antimicrobial resistance in some GPB and Gram-Negative ones in Africa have been published 4,39-44 However, reviews discussing the molecular epidemiology and mechanisms of ABR in GPB such as Staphylococcus spp., Streptococcus spp. and Enterococcus spp. in Africa in the context of resistance rates, resistance mechanisms (and MGEs), clonality, and geographical distribution from a One Health perspective are non-existent, to the best of our knowledge. This review sought to fill this gap by analyzing the burden, types, and molecular epidemiology of resistant GPB from a One Health context.
1.1 Search strategy and inclusion criteria
English research articles published within the last ten years (01/01/2007 to 07/08/ 2018) and indexed in PubMed, Web of Science and African Journals Online were searched with the following keywords: “Enterococcus”, and “Streptococcus”, “Staphylococcus”, in permutations and combinations with “resistance AND Africa”. Studies which did not identify the underlying ABR mechanisms/genes as well as the clonality of antibiotic-resistant GPB were excluded. Thus, studies that only reported on antibiotic sensitivity testing (AST) results or undertook ABR surveillance studies without further molecular tests to characterize the ABR mechanisms and/or clonality of the isolates were excluded (Figure 1). In all, 248 studies were excluded because they only had MIC data (See Supplementary data 1). All searches were undertaken independently by both authors in triplicates to ensure replication of the results.
Data extracted from the articles included year of study, country, GPB species, clones, sample sources, sample size/number of isolates, number of resistant isolates, resistance genes and MGEs and antibiotics to which the strains were resistant (Tables 1-6; Supplementary data 2). The mean rate of ABR among GPB per country and in Africa was determined to identify countries with the highest or lowest levels of resistance in Africa (Table 5). As well, the antibiotics to which the isolates were most resistant were determined to evaluate their correlation with the detected/reported resistance mechanisms (Table 6).
The resistance mechanisms, as well as MGEs involved in the transmission of resistance genes per species or clone, were determined to assess the means of resistance transfer i.e., horizontal or vertical (through clonal expansion), per specimen sources (animal, human, and environment) (Figures 2a & 2b). The distribution of clones, resistance genes, and MGEs were considered to identify countries with most resistant clones, resistance genes, and their associated MGEs (Figure 3a).
1.2 Statistical analysis
The data was analyzed using Microsoft Excel® 2017 and Graph pad prism™ 6 (GraphPad Software, San Diego, CA, USA) (Supplementary data 2). Calculation for the statistical significance of the data was determined using the kolmogorov-smirnov test (with Dallal - wilkinson-Lilliefors p-value) and/or column statistics or one sample t-test, and the confidence intervals determined at 95%. The p-values were two tailed with a Gaussian approximation. A p-value of <0.05 was considered as statistically significant. Only studies that provided the required information were used in the analysis. In all, 130 articles were used for the data analysis (Fig. 1).
2. RESULTS AND DISCUSSION
Of the 1,486 articles returned from the systematic literature search from PubMed, Web of Science and African Journals Online, 130 studies representing 20 out of 54 African countries were included in this review and data analysis (Fig. 1). A total of 249 papers were excluded because they only had MIC data. Tunisia (n=33 studies) recorded the highest number of studies followed by South Africa (n=21 studies), Egypt (n=21 studies), Nigeria (n=13 studies) and Algeria (n=7 studies), Angola (n=6 studies), Uganda (n=5 studies), Democratic Republic of the Congo (n=3 studies), Ghana (n=3 studies), Kenya (n=3 studies), São Tomé and Príncipe (n=3 studies), Gabon (n=2 studies), Tanzania (n=2 studies), Cape Verde (n=1 study), Libya (n=1 study), Namibia (n=1 study), Senegal (n=1 study) and Sudan (n=1 study). Majority of the included studies were undertaken in Northern Africa (n=65 studies, 50%), Southern Africa (n=35 studies, 26.9%) and West Africa (n=18 studies, 13.9%). Different rates of resistance to antibiotics were reported in different countries in Africa (Tables 2-5; Supplementary data 1).
A meta-analysis of published literature confirmed the presence of a high mean rate of drug resistance in GPB isolated from humans (62.0%, 95% CI: 54.7 – 69.3%), animals (68.2%, 95% CI: 58.0 -78.4%) and the environment (84.6%, 95% CI: 69.9 – 99.3%) (P-value<0.0001) in Africa, albeit many studies that did not address the molecular mechanisms of resistance in GPB were excluded. Obviously, the mean rate of resistance would have been higher had all research articles using only phenotypic methods to describe ABR in GPB been included (Supplementary data 1). Interestingly, although a lesser number of GPB were isolated from environmental sources, they expressed higher ABR than those from humans and animals; hence, the higher mean resistance rate of 84.6%. This also underscores the fact that there is increasing ABR genes in the environment, obviously due to antibiotic pollution from human activity. Evidently, ABR is high among GPB in certain regions in Africa (Figures 3a & 3b) (Table 5) and underpins the need to up the ante against this menace through increased molecular surveillance research, education of clinical microbiologists on ABR, and antibiotic stewardship.
Studies describing detailed molecular mechanisms of GPB resistance and molecular epidemiology in Africa are few, making it difficult to paint a vivid comprehensive picture of ABR in Africa. However, this review shows that S. aureus ST5, E. faecium ST18, ST80 and ST910, E. faecalis and S. agalactiae harbouring mecA, tet and erm genes, were commonly found in humans, animals and the environment, particularly in Northern, Western, and Southern Africa. Thus, careful use of β-lactams, tetracyclines, and macrolides is warranted to prevent further selection and dissemination of these resistance genes and resistant clones. Furthermore, it will be prudent for countries within these regions to review their recommended antibiotic regimens, guidelines/protocols for infections caused by these species.
erm(B) and tet(M) were found in S. aureus, Enterococcus spp. and Streptococcus spp., with erm(B), tet(M) and vanA genes being mobilized by Tn916 and IS16, indicating horizontal transfer within same clones, different clones and species. The discovery of same clones and resistance genes in specimens from humans, animals and the environment suggest a possible transmission of these clones between humans, animals and the environment, corroborating the need for a One Health approach to infection control and management of antibiotic-resistant infections. Further molecular epidemiological surveillance in the above-mentioned states is crucial to forestall further spread of these resistant pathogenic clones both within their borders and from their borders to other countries.
Resistance rates per countries and MDR GPB species
High mean resistance rates were reported in Sudan (98.5%), South Africa (82.7%) Nigeria (71.2%), Egypt (70.5%), Angola (66.2%), Tunisia (66.8%), Ghana (65.1%), Algeria (62.2%) etc. (Table 5). Crosscontamination of multi-drug resistant bacteria between patients and the environment accounted for the high rate of resistance in Algeria 45-49. The high rate of ABR in Tunisia was attributed to cross contamination between hospital patients and hospital environment, immune deficiency 50, overconsumption of antibiotics, heavy consumption of sheep meat, which is a reservoir of MRSA, and high consumptions of antibiotics in animal feed 51,52. In Egypt, inappropriate antibiotic prescription practices 29, inadequate hygienic handling and processing of food 12, and close contact with pet dogs accounted for the high resistance 53. The high rate of drug resistance in Nigeria has been attributed to the exchange of resistance genes between farm animals or their products and man 54,55, existence of MRSA in clinical and community settings 56, uncontrolled usage of antibiotics 57 and the presence of efflux pumps in coa gulase - negative staphylococcus strains 58. Expansion of resistant clones 59, variability of hospital acquired MRSA clones 60, consumption of unpasteurized milk or inefficient thermal processing of milk 21, shedding of resistant clones from animals to the environment and heavy consumption of antibiotics to treat TB due to high HIV burden 61, were incriminated for the high-level resistance in South Africa.
Staphylococcus spp. (S. aureus, S. haemolyticus and S. saprophyticus); Streptococcus spp. (S. pyogenes and S. agalactia), and Enterococcus spp. (E. faecium, E faecalis, E. hirae, E. durans, and E. gallinarum) were the antibiotic-resistant GPB widely distributed in Northern, Southern, Western and Central Africa. The high number of tet(M/L/K), erm(A/B/C), aph(3’)-lll and vanA/B/C in Staphylococcus spp., Enterococcus spp., and Streptococcus spp. reported in Tunisia, South Africa, Nigeria, Algeria and Egypt accounted for the high rate of resistance to tetracycline, erythromycin, kanamycin and vancomycin (Figure 3a). Such resistant GPB are known to compromise the safety of invasive medical procedures such as organ transplants, orthopedic surgery, and cancer treatment. In addition, infections such as sepsis, endocarditis, deep wound infections, pneumonia, meningitis and urinary tract infections caused by these resistant pathogens are becoming increasingly fatal due to limited treatment options 62,63. The abuse of antibiotics as growth promoters, prophylaxis, and metaphylaxis in food animals in these countries have been implicated in the selection of resistant bacteria that can pass on to humans through food consumption, direct contact with animals and the environment, as well as trade of animals and food products between countries 64.
Approximately 26, 385 GPB were isolated from humans (n=83 studies), animals (n=32 studies) and the environment (n=14 studies) (Tables 1-4), with mean rates of ABR varying from 14.2% to 98.5% across the 20 included countries (Tables 2-5). The antibiotics to which the isolates were most resistant to were penicillin (n=4 224 isolates, 76.2%), erythromycin (n=3 552 isolates, 62.6%), ampicillin (n=1 507 isolates, 53.9%), sulfamethoxazole/trimethoprim (n=2 261 isolates, 46.0%), tetracycline (n=3 054 isolates, 42.1%), vancomycin (n=1 281 isolates, 41.2%), streptomycin (n=1 198 isolates, 37.0%), rifampicin (n=2 645 isolates, 33.1%), ciprofloxacin (n=1 394 isolates, 30.5%), clindamycin (n=1 256 isolates, 29.9), and gentamicin (n=1 502 isolates, 27.3%) (p-value <0.0001) (Tables 2-4 & 6).Countries with high number of studies such as Tunisia, South Africa, Egypt and Nigeria recorded high number of ABR (Table 5) and high number of mecA, erm(B), tet(M), drfG and vanB resistance genes (Figure 3a). Vancomycin resistance was reported in seven studies each for animals and the environment, and 12 studies in Humans. Vancomycin-resistant Enterococcus spp. (n=102 isolates) and vancomycin-resistant Staphylococcus spp. (n=258 isolates) were reported in humans, animals and the environment (Tables 2-4; Figures 2). Vancomycin-resistant Staphylococcus aureus (VRSA) was reported in animals (n=238 isolates), the environment (n=15 isolates) and humans (n=5 isolates). A similar situation occurred with vancomycin-resistant E. faecium, which was isolated from the environment (n=306 isolates), animals (n= 671 isolates) and humans (n=26 isolates) (Supplementary data 1).
Antibiotic-resistant S. aureus ST5, E. faecium (ST18, ST80 and ST910) and E. faecalis harbouring mecA, erm(B), erm(C), tet(M), tet(K), tet(L) and vanB were isolated from humans, animals and the environment, albeit in higher proportion in humans and animals than the environment (Tables 2-4). For instance, Farhat et al. (2014) 46, van Rensburg et al. (20 1 2) 59 and De Boeck et al. (2015) 65 in Algeria, South Africa and Democratic Republic of Congo respectively, reported on resistant S. aureus ST5 in humans whilst Fall et al. (2012) 66 reported on the same clone (S. aureus ST5) in pigs from Senegal. Further, Mariem et al. (2013) 24 isolated the same clone (S. aureus ST5) from the environment in Tunisia, suggesting that this clone is widely distributed in Africa in humans, animals and environment. It is currently not clear whether this clone first emerged from humans, animals or the environment, but its presence in all three spheres shows the possibility of resistant species and clones being disseminated between animals, humans and the environment. Notably, S. aureus ST5 is among the frequently reported clones in Asia 67 and recent evidence suggest that it has spread from hospitals into communities, resulting in community-acquired MRSA 68.
Similarly, Lochan et al. (2016) 30 in South Africa, Dziri et al. (2016) 20 and Elhani et al. (2014) 69 in Tunisia isolated resistant E. faecium ST80 from humans. For the first time, E. faecium ST80 was isolated from environmental samples in a hospital in Tunisia by Elhani et al. (2013) 69 and Dziri et al. (2016) 70. Transmission of this resistant clone to animals is possible, although not yet reported. This implies that these resistant species and clones are circulating between humans and the environment, underpinning the broad host range and transmissibility of these strains between humans and the environment.
mecA was the predominant resistance gene, which corresponded with the higher penicillin resistance recorded (Figure 2aii). MRSA strains were the most commonly isolated strains (≥ 2,350) 71-74. This is consistent with the global report of increasing prevalence of MRSA 75,76. MRSA harbours the mecA gene, which is carried by the SCCmec MGE, and mediates resistance to multiple β-lactam antibiotics77. From this review, MRSA showed resistance to eleven different antibiotic classes: aminoglycosides (gentamicin, tobramycin), β-lactams (penicillin, ampicillin, oxacillin, cefoxitin), fluoroquinolones (ciprofloxacin, levofloxacin, ofloxacin), glycopeptides (vancomycin), lincosamide (clindamycin), macrolides (erythromycin), phenicols (chloramphenicol), rifamycins (rifampicin), streptogramins (pristinamycin), sulfonamides (trimethoprim/sulfamethoxazole), and tetracyclines (tetracycline). MRSA is thus a worrying public health threat as some strains have evolved resistance to almost all licensed drugs (26). Vancomycin-resistant Enterococci (VREs) (≥ 594), which were reported in Northern and South Africa, also pose a serious threat to public health as they are resistant to vancomycin, a glycopeptide that is reserved for fatal or life-threatening Gram-positive infections, and other important antibiotics such as ampicillin, erythromycin, fluoroquinolones (ciprofloxacin, levofloxacin), gentamicin, rifampicin, streptomycin, trimethoprim/sulfamethoxazole and tetracycline. In this study, enterococcus isolates had a resistance rate of 60.1% (95%, CI=32.2 -87.9) (p-value = 0.0005) to vancomycin (Table 6). Multidrug resistance in VREs increases VRE-associated mortality rates, which is likely to increase to 75% compared with 45% from susceptible strains 13,80. As well, evolution of macrolide resistance (42.0%, 95% CI: 12.02 – 72.1) (p-value = 0.0129) in drug-resistant streptococci is limiting treatment options and resulting in high mortalities 81-83. In this study, MRSA, VRE and drug-resistant streptococci remain major public health threats, calling for measures to contain ABR. Novel antibiotics such as linezolid, synercid, and daptomycin should be used empirically whilst awaiting susceptibility results. The empirical therapy can be changed or maintained based on the susceptibility report 84.
Resistance rates of species per animals, humans and the environment
The rates of ABR in isolates recovered from the environment was highest, followed by isolates from animal sources. Among environmental isolates, 91.2% (95%, CI=78.8–103.6) were resistant to penicillin, 82% (95%, CI=40.6–123.4) were resistant to sulfamethoxazole/trimethoprim, 68.5% (95%, CI=24.1–100) were resistant to ampicillin, 60.8% (95%, CI=25.0–96.6) were resistant to vancomycin, 56.9% (95%, CI=-40.7–73.2) were resistant to erythromycin, 54.5% (95%, CI=29.49–79.5) were resistant to ciprofloxacin, and 51.3% (95%, CI=21.3–100) were resistant to clindamycin (Table 6). Among animal isolates, 71.8% (95%, CI=54.9–88.73) were resistant to penicillin, 58.9% (95%, CI=36.1–81.7) were resistant to clindamycin, 58.5% (95%, CI=37.6 –79.4) were resistant to ampicillin, 49.6% (95%,CI=30.1–69.1) were resistant to trimethoprim/sulfamethoxazole, 42.3% (95% CI=17.7–67.0) were resistant to vancomycin, 47.6% (95% CI=34.0–61.2) were resistant to erythromycin, and 38.8% (95% CI=21.3–56.3)(p-value = 0.15) were ciprofloxacin resistant (Table 6; Supplementary file 1).
The rates of resistance were much lower in humans for most of the antibiotics used (Tables 2-4). Among the various species, Enterococcus spp. and Staphylococcus spp. recorded high rates of resistance for most antibiotics (Figure 3b). Streptococcus spp. reported low rates of resistance except for tetracycline to which it recorded a high rate of 55.13% (95%, CI=20.63.18–89.64) (p-value = 0.006). Resistance to vancomycin was not reported in any Streptococcus spp. Isolate (Table 6).
Enterococcus spp, mainly E. faecium and E. faecalis, recorded a resistance rate of 98.5% (95%, CI=94.5–102.6)(p-value = 0.0001) to clindamycin, 81.6% (95%, CI=52.1–110)(p-value = 0.0008) to trimethoprim/sulfamethoxazole, 64.0% (95%, CI=50.0–78.1)(p-value=0.0001) to erythromycin, 60.1% (95%, CI=32.2–87.9)(p-value = 0.0005) to vancomycin, 57.3% (95%, CI=24 -90.7)(p-value=0.0057) to penicillin, 51.7% (95%, CI=35.8–67.6)(p-value=0.0001) to tetracycline, 49.9% (95% CI=31.3–68.5)(p-value=0.0001) to ciprofloxacin, 48.9% (95% CI=20.6–77.2)(p-value=0.004) to kanamycin, 47.1% (95% CI=26.7–67.7)(p-value=0.0006) to ampicillin, 40.8% (95% CI=24.3–57.4)(p-value=0.0001) to streptomycin and 34.0% (95% CI=19.7–48.4)(p-value=0.0002) to gentamicin (Table 6).
S. aureus showed high resistance (79.6%) to penicillin (95% CI=69.7–89.5)(p-value = 0.0001), 67.8% to erythromycin (95% CI=11.5–147.0)(p-value = 0.0917), 55.5% to ampicillin (95% CI=44.50–88.5)(p-value = 0.0001), 39.3% to trimethoprim/sulfamethoxazole (95% CI=39.3–47.8)(p-value = 0.0001), 36.9% to tetracycline (95% CI=29.3–44.5(p-value = 0.0001), 35.8 to streptomycin (95% CI=14.7–57.0)(p-value = 0.004), 33.6% to rifampicin (95% CI=20.1–47.03)(p-value = 0.0001), 24.0% to clindamycin (95% CI=14.9–33.1)(p-value = 0.0001), 23.9% to ciprofloxacin (95% CI=17.6-30.2)(p-value= 0.0001), 22.7% to vancomycin (95% CI=4.3–41.2)(p-value = 0.0212) and 22.2% to vancomycin (95% CI=15.7–28.3)(p-value = 0.0001) (Table 6).
Resistance mechanisms, clones, and MGEs
Few studies identified the clones and MGEs in the resistant isolates. Of the 130 included studies, 32 identified the clones whilst 22 described the MGEs, which were used in the statistical analysis. The most dominant gene detected in Africa, which was widespread and responsible for resistance in GPB, was mecA (n=3 547), followed by erm(B) (n=1 268), vanC1/2/3 (n=971), tet(M) (n=720), blaZ (≥565), dfrG (n=422), vanB ≥451), aph(3’)-IIIa (≥170) and aac(6’)-aph(2’)(≥268) (p-value = 0.0011) (Fig. 2a).
Figure 2b represents MGEs per clone. S. aureus clones ST5, ST8, ST 80 and ST88 were highly associated with mecA. Resistant S. aureus, E. faecium and E. faecalis clones such as S. aureus ST5, and E. faecium clones ST18, ST80, and ST16 were widely distributed in humans, animals and the environment. Similarly, mecA, erm(B), erm(C), tet(M), tet(K), tet(L), vanB, vanA, vanC and tet(O) were reported in isolates from humans, animals and the environment (Table 1).
IS16 and Tn916 were found with the resistance genes erm(B) and tet(M) in E. faecium (ST18, ST80 and ST910), S. agalactiae (ST612, ST616 and ST617), E. faecalis and S. pyogenes (emm18, emm42, emm76 and emm118) isolated from humans, animals and the environment (Tables 2-4; Figure 2b). tet(M) was associated with Tn916 transposon in tetracycline-resistant S. agalactiae 85 and S. pyogenes 81 in humans in Tunisia. Fischer et al. (2013) also reported the association between Tn916 and tet(M) in tetracycline-resistant S. agalactiae in camel in Kenya 86. Similarly, IS 16 was found in vancomycin-resistant E. faecium (ST80, ST180 and ST910) in humans and the environment in Tunisia 69,70. Investigations into the association between MGEs and resistance genes were limited by few studies (n=22 studies) on MGEs.
From Tables 2-4, majority of the resistance genes namely, mecA, erm(B), tet (M), vanA etc. were responsible for drug resistance to antibiotics such as aminoglycosides (gentamicin, streptomycin, kanamycin), β-lactams (penicillins, cephalosporins), fluoroquinolones (ciprofloxacin), macrolide (erythromycin), sulfamethoxazole/trimethoprim, tetracycline and glycopeptides (vancomycin). These resistance genes were widely distributed in Northern Africa (Tunisia, Algeria, Egypt, Morocco, and Libya) and Southern Africa (South Africa and Namibia). All the three different MGEs (Tn916, SCCmec and IS16) were reported in Tunisia, with two being reported in Kenya (SCCmec and Tn916). IS16 was only reported in an E. faecium infection in Tunisia (Figure 3) whilst mecA was mostly associated with SSCmec. erm(B) and tet(M) were highly associated with Tn916 and IS16.
In Africa, different studies have reported SCCmec-borne mecA in S. aureus in humans, animals and the environment 23,47,60,66,87 besides the discovery of IS16 and Tn916 in the environment of erm(B) and tet(M) genes in Enterococcus and Streptococcus. These reports show that MGEs are mediating the dissemination of these (and possibly other) resistance genes across different GPB clones and species. MGEs-mediated mobilization of various resistance genes in different GPB clones and species in humans, animals and the environment (Tables 1-4; Figure 2b) calls for prompt measures to contain ABR as the situation may worsen if additional resistance genes are acquired by the MGEs. Resistance genes on MGEs can be horizontally transferred to susceptible cells or vertically transferred to daughter clones 37,88,89, which can easily spread these resistance genes to susceptible pathogens. The higher number of resistant Gram-positive cocci and mean resistance rate in Tunisia may be due to the presence of these three MGEs in this region 69,70,81,90
Molecular epidemiology of antibiotic-resistant GPB
Staphylococcus spp. (S. aureus, S. haemolyticus and S. saprophyticus)
North Africa: Algeria, Egypt, Morocco, Tunisia, Libya
Algeria
S. aureus was recovered from two different studies in Algeria. In assessing the nasal carriage of S. aureus in patients with medical conditions including pneumonia, urinary tract infections, osteoarthritis, heart diseases, diabetes and chronic kidney disease, Djoudi et al. (2014) isolated MRSA 46. They also found nasal carriage of S. aureus to be significantly associated with cancer and previous hospitalization of patients with kidney failure due to immunological suppression and hemodialysis. The nine MRSA isolates, i.e. ST80 (n=4), ST5 (n=2), ST22 (n=2) and ST535 (n=1), harboured mecA and were resistant to tobramycin (n=6), gentamicin (n=1), trimethoprim/sulfamethoxazole (n=2), tetracycline (n=3) and erythromycin (n=1). MRSA ST80 is a well-known and frequent etiological agent of infections in North Africa and Middle-East countries91,92. Typing of 64 MRSA isolated from human pus (n=47), venous catheters (n=7), tracheal aspirates (n=4), punction fluids (n=3), blood (n=2) and urine (n=1) in 64 Algerian patients revealed that 50 were hospital acquired (HA-MRSA) and 14 community acquired (CA-MRSA), which were all resistant to cefoxitin and oxacillin 47. mecA, mobilized by SCCmec, was the only detected mechanism of resistance.
Egypt
MRSA have been respectively isolated in five animal-based and two human-based studies in Egypt between 2011 to 2017. Hashem et.al (2013) isolated 94 S. aureus strains from blood and wounds in which 45 were MRSA while 25 were fluoroquinolone-resistant 29. Mutations such as C2402T, T2409C, T2460G, T1497C, and A1578G in gyrase enzymes, which leads to fluoroquinolones’ target-site alterations, were implicated in resistance to fluoroquinolones (ciprofloxacin, levofloxacin, ofloxacin). The high rate of fluoroquinolone resistance (55.56%) among MRSA infections is rather concerning as patients unable to tolerate vancomycin are treated with other antibiotics such as fluoroquinolones. Vancomycin is often reserved as a last-resort therapy for MRSA infections due to their high resistance to several antibiotics.
Multidrug resistance to drugs such as gentamicin, ampicillin, amoxicillin, cefepime, tetracycline and chloramphenicol in MRSA is mediated by diverse resistance mechanisms including impermeability effects and efflux pumps. Unrestricted access to antibiotics and inappropriate prescriptions were responsible for the high rates of drug resistance in this study 29. In a similar study, MRSA was isolated from patients suffering from surgical wound infections, diabetic foot, abscess and burns. Although mecA was the only mechanism of resistance, the isolates were multiple-resistant to several antibiotics belonging to the β-lactams, aminoglycosides, fluoroquinolones, macrolides, lincosamides, tetracyclines and glycopeptides, indicating other mechanisms of resistance 93. It therefore implies that administration of such antibiotics will not relieve patients from S. aureus infections. The high rate of S. aureus isolation confirms it to be the most prevalent Gram-positive pathogen isolated from soft tissue and wound infections.
Al-Ashmawy et. al. detected a high rate of MRSA (53%) in milk and dairy products believed to originate from human contamination rather than contamination from animals. Besides being resistant to β-lactams and other antibiotics, thirty-six of the isolates were resistant to vancomycin known to be effective in treating MRSA infections 12, making milk and dairy products a significant source of multidrug-resistant and toxigenic S. aureus infections. The occurrence of MRSA in pets such as dogs admitted in a veterinary clinic 53 may confirm a possible route in the community transmission of this pathogen, which is emerging as a veterinary pathogen of public health importance.
In 2017, Osman and colleagues detected Staphylococcus spp. in imported beef meat. Sixteen of these isolates were MDR and showed resistance to different groups of antibiotics due to resistance mechanisms such as mecA, and mutations in gyrA and gyrB. Indeed, MRSA has made methicillin and other β-lactams antibiotics clinically useless as a result of their high MDR 94. Imported meat acts as a transmission vector for MRSA and is worrisome as Staphylococcus spp. are among the most common foodborne pathogens causing food poisoning outbreaks worldwide. Of 133 S. aureus recovered from animal origin, more than 70% were MDR and 30 were MRSA, exhibiting high resistance to clindamycin, co-trimoxazole, tetracycline, oxacillin, cefoxitin, ceftriaxone and erythromycin; four of the isolates were resistant to vancomycin 23. The isolates showed the maximum sensitivity to imipenem, chloramphenicol and rifamycin, which is consistent with similar reports in China and Pakistan 95,96, indicating their effectiveness in treating S. aureus infections.
MRSA was isolated from chicken products mainly due to poor hygienic handling processes, posing a risk to public health in 2016. The mean S. aureus count in the chicken products were beyond the permissible limits of the Egyptian organization for Standardization and Quality Control (EOSQC 2005), coupled with resistance to different antibiotics classes; thus, retail chicken products could constitute a high health risk to human consumers 28
Morocco
In a study to assess S. aureus carriage among end-stage renal diseases patients undergoing hemodialysis, 42.9% were carriers, of which only one was MRSA. The methicillin-susceptible S. aureus (MSSA) was resistant to many of the local antibiotics, thus limiting the successful treatment of MSSA infections. Moreover 81.8% of the MSSA were penicillin-resistant. The male gender and age 30 or below were identified as risk factors of S. aureus nasal carriage (P-value < 0.001) 27. Periodic monitoring of patients with hemodialysis is crucial as they are at increased risk of S. aureus infection due to periodic hospitalization, immunosuppression and high invasive vascular interventions.
Tunisia
Resistant S. aureus was isolated from the environment, animals and humans between 2011 to 2017. Ben Said, et al. recovered 12 MSSA from wastewater samples that were resistant to penicillin (n=12 isolates), erythromycin (n=7 isolates), tetracycline (n=1 isolate) and clindamycin (n=1 isolate) due to the presence of blaZ (n=7), msr(A) (n= 7) and tet(K)(n=1). These resistant strains were of ST3245(n=7) and ST15(n=1) 18, which have been also reported in animals and humans. In an investigation to evaluate the prevalence of coagulase-negative Staphylococcus (CoNS) in the hospital environment, MDR S. haemolyticus and S. saprophyticus were the most dominant. Methicillin resistance was detected in S. haemolyticus, S. epidermidis and S. saprophyticus. These isolates were resistant to erythromycin, tetracycline, gentamicin, kanamycin, tobramycin and streptomycin due to the presence of msrA (32), erm(C) (8), tet(K) and tet(M), aac(6’)-Ie-aph(2”)-Ia (16),), aph(3’)-IIIa(19), ant(4’)-Ia (n=14) and ant(6 ‘)-Ia (3) 97. The high prevalence of MDR Staphyloccoci spp. isolates may result from transmission between the staff, patients and the environment. Strict infection controls are needed as infections caused by CoNS are common cause of death, particularly in low-birth-weight children, and are opportunistic infections in immunocompromised patients 98.
Moreover, nasal swab from sheep detected five MRSA (mecA=5), which were all of ST153 and carried blaZ, ant(6)-Ia, aph(30)-IIIa, erm(C), tet(K), and fusB genes that respectively encoded resistance to penicillin, streptomycin, kanamycin, erythromycin, tetracycline and fusidic acid. This study shows that the nares of healthy sheep could act as reservoirs of MRSA 99.
Between 2011 to 2012, 99 MRSA strains were detected from nasal swabs, blood, catheter, wounds, pleural puncture and abscess, among which 39 were tetracycline resistant. These isolates were resistant to aminoglycosides, fluoroquinolones, macrolides and lincosamide, with mechanisms of resistance including mecA (n=24) tet(K) (n=6) tet(L) (n=1) and/or tet(M) (n=18), erm(A)(n=14), aph(2’)-acc(6’) (n=13). Identified drug-resistant strains included ST247 (n=12), ST239 (n=6), ST728 (n=2), ST241 (n=1), ST398 (n=1), ST5 (n=1) and ST641 (n=1) 50. For the first time, clonal lineage ST398, which has been reported in pigs from several studies in USA, South America, Asia and Canada 100-103, was found in human MRSA isolates in Africa in a nasal swab of a 74-year old patient.
Additionally, 69 MRSA strains were isolated from hospital-acquired and community-acquired infections. Although mecA (n=59) was the only mechanism of resistance identified, the isolates were resistant to aminoglycosides, tetracycline, fluoroquinolones, macrolides and rifampicin. The resistant clones were ST1 (n=2), ST5 (n=5), ST22 (n=1), ST80 (n=41), ST97 (n=2), ST153 (n=2), ST239 (n=4), ST241 (n=3), ST247 (n=3), ST256 (n=1), ST1819 (n=3) and ST1440 (n=1) 24.
Mezghani Maalej and colleagues (2012) isolated five pristinamycin-resistant S. aureus strains from patients with skin infections. These isolates were MDR (Table 2), being the first detection of resistance to streptogramins due to vat(B) and vga(B) resistance genes 104, which emerged due to selective pressure from the use of pristinamycin. Thirty-six methicillin-resistant S. haemolyticus (MRSHae) were isolated from neutropenic patients (suffering from febrile neutropenia) with hematological cancer between 2002 and 2004. These MDR isolates carried SCCmec-borne mecA (Table 2) 105, which agrees with a report on S. haemolyticus’MDR capacity, particularly in immunocompromised patients106,107,
Libya
Due to the high risk of MRSA colonization developing into infections in children, nasal samples were collected from children inpatients, their mothers, healthcare workers and outpatients’ workers, which yielded a MRSA nasal carriage rate of 8.3%, 11%, 12.3% and 2.2% respectively in Libya 108. Thus, nasal carriage of MRSA is common in inpatients children, their mothers and health workers in Libya and could be a source of MRSA infections.
West Africa: Ghana, Nigeria, Senegal
Ghana
Among 308 staphylococcus isolates collected across Northern, Central and Southern Ghana in 2013, low prevalence of antibiotic resistance was reported except for penicillin (97%), tetracycline (42%) and erythromycin (6%) 109. Moreover, mecA was detected in only nine isolates, implying the presence of other β-lactam resistance mechanisms. The MRSA clones included ST8 (n=1), ST72 (n=1), ST88 (n=2), ST239 (n=1), ST250 (n=2), ST789 (n=1), and ST2021 (n=1). In a similar study that characterized 30 MRSA isolates resistant to tetracycline, fluoroquinolones and macrolides, tet(M) (n=13), tet(K) (n=10), aphA3 (n=7), aacA-aphD (n=5) and erm(C) (n=4) were detected. Similar and different resistant clones, viz. ST88 (n=8), ST8 (n=5), and ST247 (n=4) were detected 110, indicating high MRSA clonal diversity in Ghana. These studies show a high rate of resistance to non-β lactams that further complicate MRSA treatment. Furthermore, the isolation of USA300 and other epidemic multidrug-resistant MRSA clones calls for MRSA surveillance and adequate control measures.
Nigeria
Five different studies reported drug-resistant S. aureus from several human anatomical sites such as throat swabs, soft skin and tissue infection, urinary tract and respiratory infections, wound, vagina, otitis, conjunctivitis, septicemia and bronchitis. Of a total ≥602 isolates, ≥433 were resistant to several antibiotic classes (Table 1). Of note, 429 of the ≥433 drug-resistant isolates were all resistant to cotrimoxazole or trimethoprim/sulfamethoxazole (SXT). Mechanisms of resistance included mecA (≥54), blaZ (n=284), dfrA (≥5) and dfrG (≥152). S. aureus-resistant clones ST8, ST14, ST37, ST39, ST88, ST152, ST241, and ST772 were present. Colonized persons, including immune-compromised individuals, facilitated the spread of S. aureus and MRSA ST8 identified as ubiquitous in various geographic areas of Nigeria. High utilization of co-trimoxazole or SXT because of low cost and easy obtainability through lenient medication regulations were implicated for the high resistance 56. Besides S. aureus, S. haemolyticus was the major species isolated, and is considered as the second most detected and clinically important Staphylococci spp., particularly in immunocompromised patients 111. All the S. haemolyticus isolates detected were resistant to at least three antibiotics classes (Tables 2-4) 112.
Moreover, O. Ayepola et al. (2015) reported a higher rate of 20.8% S. aureus from UTIs than the reported ranges in Africa (6.3-13.9%), and far exceed the rate reported from Europe and Brazil (1.1%) 113. None of the isolates exhibited resistance to vancomycin, linezolid, daptomycin and mupirocin; indicating their usefulness in treating S. aureus infections. Co-trimoxazole, which was previously clinically valuable in treating MRSA infections, demonstrated the highest level of resistance, hence it’s not recommendable56,57,90,112. In a study to examine the genetic mechanism(s) of resistance in CoNS in faecal samples, all the 53 islolated CoNS were Penicillin V-resistant and between three to 19 exhibited multidrug resistance (Table 2); mecA (n=15), erm(C), tet(M) (n=4) and tet(K) (n=6) were identified 112. CoNS isolates from faeces carrying tetracycline, macrolides and aminoglycosides resistance genes may transfer them inter- and intra-species, disseminating MDR in Staphylococcus.
Senegal
A low prevalence of MRSA (10.5%) was reported in Senegalese pigs compared to those reported in developed countries. This might be due to a lesser veterinary antibiotic use as growth promoters and/or for therapy. However, all the isolates were resistant to penicillin, 27 were resistant to co-trimoxazole and 16 were resistant to tetracycline 66. Five of the MRSA were of ST5 66, evincing the spread of this clone in animals, humans 46,59, and the environment 24; the importance of this clone as a cause of human infections is well-established 68.
Cape verde
In Cape Verde, a low prevalence of 5.6% (6/107) MRSA nasal carriage was documented in 2015. The predominant MRSA clones was ST5 (n=3), ST8 (n=1) and ST88 (n=2). These isolates showed significant level of resistance to erythromycin (ERY), sulphamethoxazole-trimethoprim (SXT) and penicillin G (PEN) 114.
Central Africa: Gabon, D.R. Congo
Gabon
In Gabon, S. aureus isolated from colonized persons, blood, as well as soft and skin tissue infections resulted in 49% (104/212) resistance to trimethoprim: dfrA (n=1), dfrG (n=100), dfrK+G (n=1), dfrB (n=2), and mecA (n=1) were detected in the isolates 55. Thus, dfrG is obviously the most abundant and common trimethoprim resistance mechanism in Africa, refuting dfrB mutation as the main mechanism of resistance to trimethoprim 115-117.
D.R. Congo (DRC)
A total of 215 (79.3%) drug-resistant S. aureus isolates were collected between 2015 to 2017 from nasal swab and bloodstream infections in the D. R. Congo; 70 isolates were MRSA. Other major resistance genes mediating resistance to trimethoprim/sulfamethoxazole, aminoglycoside, macrolides, tetracycline, penicillin, and chloramphenicol were dfrG (≥120), tet(K) (≥98), and femA (≥98). MRSA showed high-level resistance to β-lactams, aminoglycoside, macrolides and tetracycline. The pathogen caused severe infections such as pneumonia, meningitis, complicated urinary tract infections, gynaecological infections and peritonitis. S. aureus ST8 (≥47) was the dominant clone, followed by ST152 (≥17), ST5 (≥2) and ST88 (≥2). In DRC, MRSA ST8 outnumbers the African MRSA clone ST88, which is dominant in Africa. The high-level oxacillin resistance in DRC was associated with a mutation in femA (Y195F) whist high-level trimethoprim resistance was due to the detection of dfrG, which is consistent with trimethoprim resistance in Africa and Asia. In Africa, SXT or cotrimoxazole is frequently administered as prophylactic to immuno-suppressed patients such as HIV/AIDS patients to prevent opportunistic infections such as Pneumocystis carinii pneumonia, toxoplasmosis and bacterial pneumonia 118 Hence, prophylactic use of SXT in HIV patients may impact resistance. Additionally, there was high-level MDR among MRSA, which is a great concern as microbiological laboratories/facilities and second-line antibiotics are rare in DRC. Moreover, the detection of nasal carriage among healthcare workers’ demands strict infection controls and surveillance 65,119,120.
East Africa: Kenya, Tanzania
Kenya
In contrast to earlier studies done in Kenya, Omuse and colleagues (2016) detected a wide genetic diversity of MRSA and well-established epidemic MRSA clones among clinical isolates. MRSA clonal complexes 5, 22 and 30, implicated in several outbreaks were described. These clones included ST5 (n=1 isolates), ST8 (n=2 isolates), ST22 (n=4 isolates), ST88 (n=1 isolates), ST241 (n=12 isolates), ST239 (n=2 isolates) and ST789 (n=1 isolates). Approximately 41% of the MRSA in the study were MDR (Table 2), showing resistance to clindamycin, erythromycin and SXT 87. Detection of these clones in referral hospitals in Kenya calls for implementation of strict infection control measures to reduce the high morbidities and mortalities associated with HA-MRSA infections.
Tanzania
In a study to investigate the molecular epidemiology of trimethoprim resistance in MSSA causing skin and soft tissues infections, dfrG was detected in all 32-trimethoprim resistant isolates. Other reported trimethoprim resistance mechanisms such as dfrA, dfrB and dfrK were missing, confirming dfrG as the main trimethoprim resistance mechanism in Sub-Sahara Africa 55.
Uganda
A MRSA carriage of 56.1% (23/41) was detected in milk from pastoral communities in Uganda, exactly 70% of which were tetracycline-resistant. MRSA clones ST97 and ST1 were identified. Furthermore, over 90% of the isolates carried genes encoding enterotoxin that causes food-borne diseases. The weak veterinary delivery system and the high dependency on animals and animal products for food in Uganda was implicated for the high prevalence of MRSA 121.
S. aureus isolates, including 24 MRSA and 40 MSSA, were isolated from patients with surgical site infections (SSI). The MRSA isolates were MDR (including resistance to oxacillin, gentamicin, ciprofloxacin and chloramphenicol) compared to the MSSA. Inducible clindamycin resistance was found in 17.2% of the isolates, mostly in MRSA. In a multivariate analysis, inducible clindamycin resistance and cancer were identified as independent predictors of MRSA-SSI 122.
Southern Africa: Angola, Malawi, Mozambique, Namibia, South Africa
Angola
Conceicão et al (2014) reported a nasal S. aureus carriage of 23.7% (n=128 isolates), out of which 58.1% (n=77 isolates) were MRSA. Fifty-seven of the MRSA clones were of ST5, followed by ST88 (n=9), ST8 (n=5) and ST72 (n=3). This study represents the first description of the spread of MRSA ST5 in Africa. All the 77 MRSA strains were resistant to SXT, cefoxitin (FOX) and PEN 123. In a study to identify oxacillin-susceptible mecA-positive S. aureus (OS-MRSA) for the first time in Africa, a prevalence of 17.7% was detected among healthy healthcare workers in Angola and São Tome′ & Principe, making them potential OS-MRSA reservoirs 124. OS-MRSA have been reported worldwide in humans, animals and food animals 125-128. The OS-MRSA isolates expressed MDR (Table 2) and belonged to ST88 (n=15 isolates) and ST8 (n=9 isolates). In sub-Saharan Africa, the identification of clinically important S. aureus is heavily based on phenotypic agar-screening and oxacillin disc-diffusion methods.
Mozambique
The prevalence of HA-MRSA and CA-MRSA in Mozambique was found to be 15.1% and 1%, respectively. MRSA showed high-level resistance to penicillin, cefoxitin, gentamicin, ciprofloxacin, erythromycin, SXT, chloramphenicol and tetracycline, compared to MSSA. Additionally, inducible macrolide–lincosamide–streptogramin B (MLSB) resistance was 41.7% and 10.7% in hospital-acquired S. aureus (HA-SA) and community-acquired S. aureus (CA-SA) isolates respectively 129, further limiting therapeutic options for S. aureus infections. This study, which is the first to detect the emergence of HA-MRSA within post-operative abdominal wounds and burn wounds in Mozambique, reported that patients with infected burn wounds had a significantly longer hospitalization than patients with post-operated abdominal wounds. Efforts to prevent the transmission of MDR HA-SA, such as education on proper hand-washing techniques, are urgently needed.
Namibia
The dominant resistance gene mediating trimethoprim resistance in MRSA and MSSA in Namibia was dfrG. This is similar to reports in other Africa countries 55. Moreover, dfrG was frequently detected in S. aureus from SSTIs in travelers returning from other African countries, suggesting that dfrG can be transmitted into populations with low antifolate resistance such as North America and Europe 130,131.
South Africa
Thirty MDR S. aureus were recovered between April 2015 to April 2016 from ten beaches in the Eastern Cape Province, South Africa (Table 2). Notably, the isolates harbored mecA, femA, rpoB, blaZ, erm(B) and tet(M) 11, making marine environments and public beaches potential depositaries of MDR S. aureus that can be transmitted to animals and humans. Further, the 50% resistance to vancomycin recorded is concerning to global health due to its role as a last-resort antibiotic for treating MRSA infections.
S. aureus was detected in raw and pasteurized milk at an isolation rate of 75% and 29% respectively, due to inefficient thermal processing and post-process contamination. A high proportion (60%-100%) of these isolates showed resistance to aminoglycosides, β-lactams, vancomycin, tetracycline and erythromycin, albeit only 19 mecA genes were present 21. Evidently, raw and pasteurized milk can harbour MDR S. aureus, exposing consumers to colonization and/or infections. Again, Staphylococcus spp., including S.aureus, S. haemolyticus, S. xylosus and S. capitis were isolated from healthy pigs and cattle, of which between 75 to 100% were resistant to penicillin G, tetracycline, sulfamethoxazole and nalidixic acids, due to their use as growth promoters; mecA and mphC were identified. Additionally, 12% of the isolates were resistant to vancomycin and erythromycin, evincing the important role of animals in the dissemination of resistance determinants and the importance of commensals to public health 61.
Van Rensburg et al. 59 detected 43.4% (1432/3298 isolates) and 3.1% (328/10448 isolates) rifampicin resistance rate among MRSA and MSSA respectively. Similar studies in South Africa have also reported of high rifampicin resistance in MRSA132, 133, obviously due to frequent use of rifampicin among tuberculosis patients, who are highly prevalent in South Africa. MRSA ST5 and ST612 were detected while H481Y/N and I527M mutations in rpoB were associated with high-level rifampicin resistance, similar to reports in Italy 134. Additionally, novel H481N, I527M, K579R mutations were also detected.
Three studies reported a prevalence of 29.1% 135, 45.44% 60 and 100% 136 MRSA recovered from humans, expressing resistance to macrolides, tetracycline, aminoglycoside, cotrimoxazole and rifampicin. MRSA ST612, ST239, ST36 and ST5 were the dominant strains similar to other findings in Australia and Europe137. The study showed that S. aureus bacteremia is common and account for high mortality in South Africa. For instance, in a study by Perovic et al., 135 202 patients died from S. aureus bacteremia infections, with HIV patients being more likely to acquire HA-MRSA. The isolates were however susceptible to glycopeptides, fluoroquinolones, linezoid, tigecycline, fosfomycin and fusidic acid, confirming their clinical usefulness in treating MRSA infections. In a recent study, a high prevalence and genetic diversity of multi-drug efflux (MDE) resistance genes were found in clinical S. aureus isolates, including 81 MRSA and 16 MSSA 138. norA, norB, mepA, tet(38), sepA, mdeA, imrs and sdrM were present in at least 86% of the isolates, predicting resistance to broad-spectrum biocides and fluoroquinolones, which is disturbing. Efforts to develop efflux pump inhibitors can mitigate such resistance mechanisms.
Sao Tome & Principe
MRSA prevalence of 26.9% 139 and 25.5% 114 was reported in nasal swabs in 2014 and 2015, respectively, in Sao Tome & Principe. Additionally, a high prevalence of oxacillin-susceptible mecA-positive S. aureus was reported in the same study in Sao Tome & Principe and Angola 124. The most dominant MRSA clone was ST8 (n=25 isolates), followed by ST5 (n=13 isolates) and ST80 (n=13 isolates). High genetic variability was found in the MSSA strains. Both MRSA and MSSA showed different levels of resistance to SXT, ERY, CIP and TET; however, all the MRSA isolates were resistant to cefoxitin.
Streptococcus spp. (S. pyogenes, S. pneumoniae and S. agalactiae)
Drug resistant Streptococcus spp. including S. agalactiae and S. pyogenes have been identified in Northern, Eastern and Southern Africa. S. pyogenes were reported in only humans whilst S. agalactiae was reported in both animals (camels) and humans with a high rate of resistance to tetracycline and erythromycin.
North Africa: Algeria, Egypt, Morocco, Tunisia, Libya
Algeria
A sole study has so far detected 44 tetracycline (100%, 44/44 isolates)- and erythromycin-resistant (43.18%, 19/44 isolates) S. agalactiae from vaginal swabs; tet(M); and erm(B) respectively mediated this resistance. A high diversity of resistant clones viz., ST1, ST19, ST10, ST158, ST166, ST233, ST460, ST521 and ST677 were detected 45, which have been reported worldwide for causing life-threatening invasive diseases such a meningitis and sepsis140,141.
Egypt
Similarly, Shabayek et al. (2014) detected 98% and between 14-17% S. agalactiae resistance to tetracycline and macrolides respectively. tet(M) was detected in all the 98 tetracycline-resistant isolates whilst erm(B) and erm(A) mediated erythromycin resistance. Efflux pump genes such as tet(K) (n=12 isolates), tet(L) (n=1 isolates) and mefA/E (n=1 isolates) were also found 32, which reflects the increasing reports of S. agalactiae resistance to tetracycline and macrolides 142. This study also showed that vancomycin and fluoroquinolones are effective replacement for erythromycin and clindamycin, and for patients allergic to penicillin. Although penicillin is the antibiotic of choice for treating S. agalactiae infections, reports of penicillin resistance in USA and China calls for increased surveillance in Africa 142.
Tunisia
S. agalactiae
From January 2007 to December 2009, 226 S. agalactiae were isolated from female genitals and gastric fluid of infected newborns. Of these, 97.35% (220/226 isolates), 40% (90/226 isolates) and 19.1% (43/226 isolates) were resistant to tetracycline, erythromycin and rifampicin respectively. Additionally, seven isolates were resistant to aminoglycoside (gentamycin and streptomycin) and chloramphenicol. tet(M) (n=205 isolates), encoding a ribosomal protection protein, which protect the ribosome from the action of tetracycline, was the main tetracycline resistance mechanism, and was significantly associated with Tn916 (p-value = 0.0002). Other resistance genes including erm(B) (n=79 isolates) and tet(O) (n=50 isolates) were detected. All isolates were however susceptible to β-lactams and quinupristin-dalfopristin 85. Between 2005 and 2007, 160 erythromycin-resistant S. agalactiae were isolated from humans, with a high resistance rate of 84.3% (135/160 isolates) to the constitutive macrolides-lincosamides, streptogramines B (MLSB) 143.
S. pyogenes
Hraoui et al., (2011) reported a low macrolide resistance rate (5%, 5/103) and a high tetracycline resistance rate (70%, 72/103) among human isolates, with tet(M), associated with Tn916, being responsible for tetracycline resistance 144. Increase tetracycline use in food animals was implicated in this instance, leading to selection and dissemination of resistance genes from animals to human. Macrolide resistance was only detected in seven isolates, which is corroborated by the findings of Ksia et al. (2010), who detected low-level macrolides resistance among Children 145.
East Africa: Kenya, Tanzania
Kenya
S. agalactiae
In the horn of Africa, camel plays a significant role in the survival of humans by providing milk, meat and transportation. In 2013, Fischer et al. detected 36% (37/92) tetracycline resistance in S. agalactiae isolates from camels’ wound infections and mastitis that was mainly mediated by a Tn916-borne tet(M). ST616 (n=22) was the major resistant clone, followed by ST612 and ST617 146. Shifting from tetracycline to other antibiotics is evidently necessary for effective treatment outcomes in camel infections in Kenya.
Southern Africa: Angola, Malawi, Mozambique, Namibia, South Africa
South Africa
S. agalactiae
A S. agalactiae colonization rate of 30.9% was detected from vaginal and rectal swabs of pregnant women. Similar to other reports in Africa, a high rate of tetracycline (94.5%, 120/128 isolates) and macrolide (21.1%, 27/128) resistance was documented. All the isolates were however sensitive to penicillin, ampicillin, vancomycin and gentamicin. Macrolide and clindamycin resistance were associated with erm(B) and mefA genes 147. The study highlights the need for research on treatment options for patients allergic to penicillin due to high-level resistance in alternative drugs such as macrolides and lincosamides.
Enterococcus spp. (E. faecium, E. faecalis, E. hirae, E. durans, E. gallinarum)
North Africa: Algeria, Egypt, Morocco, Tunisia, Libya
Algeria
The first study to molecularly characterize Enterococcus spp. from urinary tract and wound infections in Algeria revealed a high rate of resistance to erythromycin (86.4%, 108/125 isolates), tetracycline (82.4, 103/125 isolates), levofloxacin (71.2%, 89/125 isolates) and gentamicin (54.4, 68/125 isolates). Only 3.2% (4/125 isolates) were VRE, confirming glycopeptides as ideal antibiotics for treating Enterococcus infections. A mortality rate of 10% was reported due to infections caused by Enterococcus. E. faecium, E. faecalis and E. gallinarum were the main Enterococcus isolated. Majority of these isolates were from females (53%). erm(B) (≥92) and vanC1(≥4) were the main mechanisms of resistance. A high genetic diversity among strains was seen in E. faecium and E. faecalis, with E. faecium ST78 being the dominant resistant strain 148, which is also prevalent in Asian (Japan, Taiwan, China and Korea) and European (Italy and Germany) countries 149-151. A novel ST317 (n=33) clone was predominant among the E. faecalis isolates. Rational use of antibiotics, as well as close monitoring of the epidemiology of the strains are crucial.
Egypt
In a similar study to characterize E. faecium and E. faecalis from patients, 82% of the isolates were MDR, showing high-level resistance to aminoglycosides, β-lactams and tetracycline. VanA was detected in two E. faecium isolates, all of which were resistant to all antibiotics tested. Bioinformatic (sequence) analysis revealed that vanA was transmitted horizontally to S. aureus, showing the importance of horizontal gene transfer in ABR and subsequent management of enterococci infections such as bacteremia, endocarditis and urinary tract infections 152.
Tunisia
Antimicrobial-resistant Enterococcus was found in faeces of pet and camel, irrigation water from farm environments, food vegetables, hospital environments, animal meat and patients in Tunisia 19,22,31,51,52,69. High-level resistance to vancomycin, macrolides, aminoglycosides, β-lactams and tetracycline was detected in the environment, animals and humans with majority of the isolates being E. faecium, followed by E. faecalis. tet(M), tet(L), erm(B), ant (6)-la, vanA and aph(3’)-llla were the major resistance mechanisms, with IS16 being the main MGE disseminating the resistance genes. E. faecium ST80, ST910 and ST16 were the dominant resistant clones in Tunisia. The studies show that meat, animals, pets, hospital environment and wastewater used for farm irrigation play a crucial role in the spread of antibiotic resistant Enterococcus.
West Africa: Cape Verde, Ghana, Nigeria, Senegal
Nigeria
Enterococcus spp. isolated from poultry and cattle as well as their manure demonstrated high-level resistance to tetracycline, erythromycin, gentamicin, ampicillin and streptomycin. Sixty isolates were MDR, showing resistance to three or more antimicrobials 153. The rate of MDR is a reflection of the substantial use of broad-spectrum antibiotics in Nigeria, raising major public health concerns as practices such as the use of untreated poultry and cattle manure for fertilizing agricultural soils, particularly vegetables, are a common practice in Africa. This could transfer MDR Enterococci to humans, and cause serious nosocomial infections including endocarditis, bacteremia and urinary tract infections that can result in high morbidities and mortalities.
Ngbede et al. (2017) recently characterized 63 ampicillin- and 37 gentamicin-resistant E. faecium from vegetables, soil, farms, animal and manure 25. Approximately 95% (35/37 isolates) and 8% (5/63 isolates) of the aminoglycoside- and ampicillin-resistant clones were recognized as high-level aminoglycosides- and ampicillin-resistant E. faecium respectively. Modifying enzymes’ genes such as aac(6’)-Ie-aph(2”)- Ia), aph(2’)-1c,aph(3’)-llla, and ant(4’)-la accounted for the aminoglycoside resistance.
East Africa: Kenya and Tanzania
Tanzania
In a study to determine if cattle co-grazing with wild life influence ABR, ABR in wild animals such as buffalo, zebra and wildebeest was higher than in cattle, although wildlife is periodically treated with antibiotics. Ten VRE and ampicillin-resistant Enterococcus were found in the wild animals but not cattle. Additionally, Enterococcus isolates from wildlife were highly resistant to tetracycline, rifampicin, macrolides, aminoglycosides and cotrimoxazole 14. tet(W) and sull were the resistance genes identified in the isolates. The practice of co-grazing possibly resulted in transmission of ABR genes from livestock to wildlife. The high presence of ABR bacteria in wildlife was likely due to contact with more environmental surfaces that have been contaminated with human, birds or animal excreta. Result from this study demonstrates the presence of ABR Enterococci in wild animals without antibiotic pressure.
Southern Africa: Angola, Malawi, Mozambique, Namibia, South Africa
South Africa
Multiple antibiotic-resistant Enterococci were isolated from borehole water, waste water, pigs and humans in South Africa. Notably, a very high-level vancomycin, aminoglycoside, β-lactam, macrolides and fluoroquinolones resistance was detected among the Enterococci isolates compared to other countries. erm(B) (≥300 isolates), vanC 2/3(162 isolates), vanB (≥138 isolates), vanC (≥120 isolates), strA (≥120 isolates) were the major resistance genes. The vancomycin-resistant isolates were from patients with haematological malignancies, bacteremia, pigs, wastewater and underground water 9,10,26,30. Inefficient chlorination to kill bacteria accounted for the high resistance rates in the final effluents’ discharge into the environment. Hospital wastewater is therefore a major source of MDR Enterococcus. Sub-therapeutic antibiotic usage in animal feed also accounted for the emergence of ABR in pigs whilst the construction of boreholes near pit toilets resulted in high enterococcal isolation and resistance rates in South Africa.
Experimental procedures used in included studies
The studies included in this review basically used the following experimental procedures. Transport media such as stuart agar, cary-blair medium, and gel transport swabs with charcoal were used to transport the samples to the laboratory 53,65. Cotton swabs were used to swab sample specimens, tissues, surfaces, fluids, etc. and cultured on nutrient agar, blood agar, tryptone soya agar, mannitol salt-phenol red agar, brain-heart infusion broth, Slanetz-Bartley mannitol salt agar, and Edwards agar media prior to identifying the 24-hour colonies using Gram-staining and different biochemical tests such as catalase and coagulase tests, latex coagulase test and DNase agar test. Subsequently, antimicrobial susceptibility testing (AST) using disc diffusion (Kirby-Bauer method or E-test) on Mueller Hinton agar plates and a 0.5 McFarland bacterial inoculum was performed. Antibiotics such as ampicillin (AMP), amoxicillin (AMX), amikacin (AMK), ampicillin-Sulbactam (SAM), amoxicillin-clavulanic acid (AMC), azithromycin (AZI), apramycin (APR), chloramphenicol (CHL), cefoxitin (FOX), ceftazidime (CFZ), clarithromycin (CLR), ciprofloxacin (CIP), cefuroxime (CXM), clindamycin (CLI), cephalexin(LEX), cefoperazone (CFP), cefepime (FEP), cefotaxime (CTX), ceftaroline (CPT), cephalothin (CET), cloxacillin (CLX), doxycycline (DOX), erythromycin (ERY), fusidic acid (FUS), fosfomycin (Fof), gatifloxacin (GAT), gentamicin (GEN), imipenem (IPM), kanamycin (KAN), levofloxacin (LVX), linezolid (LZD), lincomycin (LIN), meropenem (MER), mupirocin (MUP), minocycline (MIC), moxifloxacin (MXF), methicillin (MET), metronidazole (MTZ), nitrofurantoin (NIT), norfloxacin (Nor), nalidixic acid (NAL), netilmicin (NEL), oxacillin (OXA), ofloxacin (OFX), perfloxacin (PF), penicillin (PEN), pristinamycin (PRI), rifampicin (RIF), streptomycin (STR), streptogramin B (SB), sulfamethoxazole (SMZ), tetracycline (TET), teicoplanin (TEC), telithromycin (TEL), tobramycin (TOB), trimethoprim-sulfamethoxazole (SXT), and vancomycin (VAN) were mostly used for the AST. Polymerase chain reaction (PCR) was used to detect the antimicrobial resistance genes and clones (i.e. molecular typing) of the isolates.
3. CONCLUSION AND STUDY LIMITATIONS
We report of high rate of ABR among GPB in several African countries, mediated largely by S. aureus ST5, ST8, and ST80, Enterococcus faecium and Enterococcus faecalis strains, SCCmec, Tn916 and IS16 MGEs are a major threat to clinical medicine, the economy and socio-economic development. This calls for national as well as international rules and regulations to contain resistance. Heavy consumption of antibiotics in animal feed, exchange of resistance genes between animals and food animal products to man, uncontrolled and inappropriate antibiotics prescription practices, inadequate hygienic handling and processing of food, close contact with pet dogs, shedding of resistant clones from animals to humans and the environment, as well as high consumption of antibiotics in humans, particularly in HIV patients, account for the high rate of ABR in Africa.
Effective surveillance and monitoring of antimicrobial drug usage and licensing, banning or restricting the prescription of reserved, expired and substandard drugs, periodic monitoring of pharmacies and veterinary shops, and antibiotic stewardship are recommended measures to contain ABR. Improving animal health through hygienic practices on farms, avoiding prophylactic or growth-promoting antibiotic usage in veterinary medicine, integrative efforts between human and veterinary medicine as well as environmental health are urgently needed to contain ABR. Implementation of these policies will decrease the high rate of ABR in Africa, reduce longer hospital stays and the resort to expensive but toxic antibiotic alternatives, with a concomitant reduction in morbidity and mortality rates. Few studies reporting on the molecular determinants of ABR in GPB in Africa limited the study to 130 articles. Among these, only few studies reported on MGEs and resistant clones.
Role of Funding Source:
Not applicable.
Contributors:
JOS conceived, designed and supervised the study, analysed and vetted the results, wrote the paper, edited and formatted it for publication. EM co-conceived and co-designed the study, gathered and analysed the data and drafted the paper. Both authors approved the final version for submission.
Funding:
None
Declaration of interests:
The authors declare no conflict of interest.
Figure 1. PRISMA-adapted flow chart showing included and excluded articles. All search were conducted on PubMed, Web of Science and African Journals Online, and a final number of 130 articles were used for the quantitative analysis.
Supplementary data 1
List of excluded articles on the basis of only phenotypic (antibiotic sensitivity) tests.
Supplementary data 2
Raw data and analysis of extracted information from included articles.
Acknowledgments:
None
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.
- 41.
- 42.
- 43.
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.
- 73.
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.
- 79.
- 80.↵
- 81.↵
- 82.
- 83.
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.
- 102.
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
- 116.
- 117.↵
- 118.↵
- 119.↵
- 120.↵
- 121.↵
- 122.↵
- 123.↵
- 124.↵
- 125.↵
- 126.
- 127.
- 128.↵
- 129.↵
- 130.↵
- 131.↵
- 132.↵
- 133.
- 134.↵
- 135.↵
- 136.↵
- 137.↵
- 138.↵
- 139.↵
- 140.↵
- 141.↵
- 142.↵
- 143.↵
- 144.↵
- 145.↵
- 146.↵
- 147.↵
- 148.↵
- 149.↵
- 150.
- 151.↵
- 153.↵
- 154.
- 155.
- 156.
- 157.
- 158.
- 159.
- 160.
- 161.
- 162.
- 163.
- 164.
- 165.
- 166.
- 167.
- 168.
- 169.
- 170.
- 171.
- 172.
- 173.
- 175.
- 176.
- 177.
- 179.
- 180.
- 181.
- 182.
- 183.
- 184.
- 185.
- 186.
- 187.
- 188.
- 189.
- 190.
- 191.
- 192.
- 193.
- 194.
- 195.
- 196.
- 197.
- 198.
- 199.
- 200.
- 201.
- 202.
- 203.
- 204.
- 205.
- 206.