Prevalence of Antibiotic Resistance in Vibrio cholerae: A Meta-analysis

Background:Cholera could cause severe watery diarrhea and even death, but in patients diagnosed with severe dehydration, treatment with an appropriate antimicrobial agent could decrease the volume of diarrhea and shorten the duration of V. cholerae shedding. Unfortunately, due to the high antibiotic’s consumption, antibiotic resistance in V. cholerae strains is progressively increasing worldwide. The present systematic review and meta-analysis study aimed to investigate the incidence of antibiotic resistance in V. cholerae strains worldwide. Methods: The incidence of antibiotic resistance in V. cholerae strains was assessed by conducting a systematic review and searching international databases including PubMed and Google Scholar for articles published from Jan 1, 1990 to Sept 30, 2016 using related keywords. Studies were selected to be included in this systematic review according to a predened eligibility criteria. International databases were searched for articles evaluating V. cholerae strains resistance rates to antibiotics. Pooled estimates of antibiotics resistance and 95% condence intervals (CIs) were categorized based on WHO regions. Results:V. cholerae was the most common bacterium showing high resistance rates to various antibiotics including furazolidone (83%), trimethoprim-sulfamethoxazole (67%), nitrofurantoin (66%), streptomycin (64%), and nalidixic acid (58%). Conclusions: According to this meta-analysis results, the high resistance of V. cholerae strains to antibiotics could be considered as a global public health threat. Therefore, it is recommended to select appropriate antibiotic treatment regimens for each region based on antibiotic resistance patterns of local strains in that region.


Background
Cholera is an acute diarrheal disease caused by toxigenic strains of V. cholerae bacterium, belonging to O1 or O139 serogroups. So far, the world has witnessed seven cholera pandemics occurring during the 19th and 20th centuries. The rst one occurred in 1817, and the next pandemics began in 1829, 1852, 1863, 1881,1889, and 1961, with the last continuing until now (1). In late 1992, an outbreak of V. cholerae O139 was reported in India and Southern Bangladesh. This epidemic continued until 1993. Since then, other outbreaks of V. cholerae O139 have been reported in Pakistan, Nepal, China, Thailand, Kazakhstan, Afghanistan, and Malaysia, and some imported cases have also been reported in the United Kingdom and the United States.
If the cholera outbreaks due to this new V. cholerae O139 serogroup are constantly recurred so that they could in uence more countries around the world, this may be considered as the eighth pandemic (2).
The disease is currently considered as a primary public health concern in developing countries, and in some regions like Southern Asian, parts of African, and Latin American countries, the disease is in its endemic stage (1). It has the potential to be signi cantly disseminated among populations during epidemics or to affect endemic areas during regular seasons. In cholera-endemic areas, travelers could be infected by V. cholerae strains and act as carriers of this bacterium and spread it to non-endemic areas, although the risk is low (3). Cholera could cause severe watery diarrhea and even death, while those with mild to moderate symptoms could be treated successfully by taking uids and salt orally or by injection.
The early antibiotic therapies which were recommended for cholera patients during the 1940s and 1960s and used effectively included streptomycin and chloramphenicol. Antibiotics such as tetracycline and furazolidone are considered as suitable substitutes for tetracycline in the treatment of children. Sulfamethoxazole-trimethoprim (SXT) was recommended to be used for cholera patients' treatment during the 1970s. Azithromycin and cipro oxacin are also recommended to be administered for cholera patients in order to decrease the duration of diarrhea, excretion of V. cholera, and the severity of the disease (4).
In cholera control strategies recommended by WHO, chemoprophylaxis along with antibiotics has not been suggested to be used for cholera treatment because in addition to not affecting the spread of the disease, it could induce adverse effects through contributing to the development of drug resistance. In regions where cholera is endemic, the Strategic Advisory Group of Experts (SAGE) on immunization suggests vaccination in combination with other prevention and control strategies (5). V. cholerae virulence and antimicrobial resistance grew during the seventh pandemic due to the emergence of a new variant cholera biotype called the atypical biotype (O1 El Tor). This atypical biotype diffuses mobile genetic elements, which are capable of self-transmission and integration into chromosomes, facilitating rapid spread and constant acquisition of resistance genes. It is also associated with more virulence and widespread drug resistance (6). Antimicrobial resistance (AMR) in V. cholerae strains is progressively increasing worldwide. Antimicrobial resistance is widespread in V. cholerae strains and currently threatens the effective treatment and control of cholera, especially in low and middle-income countries (6).
Recent studies have suggested that transposons, plasmids, mobile gene cassettes, and integrons are responsible for the emergence of MDR and XDR V. cholerae strains through inducing rapid and broad spread of genetic information between different species. Horizontal gene transfer (HGT) is a process that allows the pathogen to acquire exogenous DNA and develop drug resistance (7).
Studies have demonstrated that the prevalence rate of antibiotic-resistant V. cholerae strains is increasing worldwide. There is no general and accurate estimation about the prevalence rate of drug resistant V. cholerae strains. This systematic review and meta-analysis study aimed to investigate the incidence and pattern of antibiotic resistance in V. cholerae isolates worldwide, the results of which may be served as a yardstick for future control programs and interventions.

Search Strategy and Eligibility Criteria
Search engines in electronic databases, including PubMed and Google Scholar, were searched for articles assessing the incidence of antibiotic resistance in V. cholerae strains worldwide. No limitation was applied regarding the study type, time, or population. Only studies published in English were included.
Literature published was analytically searched using the following key terms: Vibrio cholerae, antibacterial, antimicrobial, antibiotic, drug, resistance, and resistant.

Data Extraction
At rst, title and abstract screening process was performed to select relevant studies, then full texts of the eligible articles were reviewed for detailed evaluation. Duplicate articles were documented and deleted before selecting nal eligible articles. Moreover, the present study was a meta-analysis; therefore, ethical approval was not required. Data were extracted from the included studies and classi ed based on author name, publication year, country, samples size (total number of tested samples), the number of resistant samples, sample type (clinical or environmental), methodology (antibiotic susceptibility test methodology), and resistance mechanism (refer the "Appendix 1").

Data Analysis
A meta-analysis of proportions was conducted by pooling prevalence data using a random-effect model.
Pooled estimates were classi ed according to study country or continent. Given the different geographic origins, it was assumed that there was a signi cant heterogeneity between the studies. Between-study variation was measured using the I2 statistic. An I2 value of more than 75% was considered as a signi cant heterogeneity. All statistical analyses were accomplished using R statistical software, Version 14.2 (StataCorp LLC, College Station, TX). The prevalence rate of antibiotic resistance in samples was reported by 95% con dence intervals (CIs). A p value of < .05 was considered as a statistically signi cant publication bias.

Results
In this study, a total of 5742 articles were screened based on their title and abstract, of which 84 studies were selected for full text evaluation and nal analysis (  (Table 1). The highest number of studies was accomplished in India (n = 20) compared to those conducted in Iran (n = 11), Thailand (n = 11), Nepal (n = 7), Ghana (n = 5), and Kenya (n = 3), as well as China, Haiti, Pakistan, Vietnam, Mozambique, and Indonesia (each n = 2), and Brazil, Austria, Slovakia, Uganda, Zambia, Guinea, German, Cote d'Ivoire, Sierra Leone, Ethiopia, Lima-Peru, Cameroon, Maryland, Senegal, and Namibia (each n = 1). Pooled antibiotic resistance prevalence rates by country and continent are shown in (Table 2,  3).  ND 0 Sierra Leone ND ND 0     According to the present study results, resistance to furazolidone was identi ed as the most common antibiotic resistance pattern worldwide. The forest plot in (Fig. 2) displays odd ratio and weight for each study. The prevalence rate of furazolidone resistance has previously been reported to vary from 2% in America to 86% in Africa. The incidence of furazolidone resistance tended to be more in developing countries, such as Sierra Leone, Nepal, India, and Iran, than in other countries.
According to the present study results, resistance to trimethoprim-sulfamethoxazole was 67%, to nitrofurantoin was 66%, and to streptomycin was 64%, and the highest rates of resistance to these antibiotics were reported in Africa. The highest rate of nalidixic acid resistance was reported in Zambia and Ivory Coast; nalidixic acid resistance (81%) was the highest in Africa. In the current study, the prevalence rates of tetracyclines and doxycycline resistant V. cholerae strains were found to be 13 and 6%, respectively; high resistance rates were found against tetracyclines (70%) and doxycycline (34.5%) antibiotics in Mozambique.
In this study, the pooled prevalence rates of cipro oxacin and nor oxacin-resistant V. cholerae strains were found to be 2 and 4% in most WHO regions with the highest rate in Haiti and Zambia. Also, the prevalence rate of azithromycin resistance was found to be 4% among V. cholerae strains. This metaanalysis study results showed that resistance rate of V. cholerae strains to other beta-lactam antibiotics was variable; and resistance rates to ampicillin (39%), amoxicillin (24%), ceftriaxone (4%), cefotaxime (1%), cefuroxime (4%), and other aminoglycosides such as gentamicin (2%) were found (8-10). Also V. cholerae resistance to chloramphenicol was found to be (12%) moderately high in Ethiopia (94%), Sierra Leone (93%), and Mozambique (83%).

Discussion
This study was a comprehensive systematic review and meta-analysis, investigating antibiotic resistance in V. cholerae strains. Unfortunately, V. cholerae strains antibiotic resistance patterns are not well elucidated. Antibiotic therapy could not be used alone to treat the disease, to control and reduce the disease duration by approximately 50%, and to reduce bacterial excretion in the feces (11). Oral or intravenous administration of uids comprising of sodium chloride, glucose, trisodium citrate, and potassium chloride is the primary option that could be used to treat cholera and protect a patient from dehydration (WHO, 2002). However, only patients with severe dehydration need an appropriate antibiotic therapy (WHO, 2017). The incidence of antibiotic resistance is considered as a serious challenge threatening the e ciency of almost all antimicrobial agents commonly administered to remedy or prevent this infectious disease (12).
The choice of an appropriate antibiotic should be made by taking into account drug resistance patterns of local strains of V. cholerae O1 or O139. Repeated antibiotic therapy also imposes destructive and harmful ecological effects on the community and further increases the global burden of antibiotic resistance, especially in low and lower-middle income countries. This negatively affects control programs and threatens active treatment of cholera (6).
Treatment of a whole community with an antibiotic is not recommended by WHO. To further complicate the epidemiology of antibiotic resistant cholera, it is su cient to mention that there is a wide range of antibiotics to which V. cholerae could be resistant. If V. cholerae strain is thought to be sensitive, some antimicrobial agents may be effective in its treatment, including doxycycline, cipro oxacin, and azithromycin (13). Currently, some V. cholerae strains have emerged in Haiti, which are sensitive to tetracycline (a proxy for doxycycline) and azithromycin, while resistant to nalidixic acid, sul soxazole, and trimethoprim-sulfamethoxazole (14).
In this study, the antibiotic resistance pattern of V. cholerae strains was comprehensively evaluated globally, and its high resistance to various antibiotics in several regions of the world was shown. This study results demonstrated that resistance to furazolidone, trimethoprim-sulfamethoxazole, nitrofurantoin, streptomycin, nalidixic acid, and ampicillin was the most common antibiotic resistance pattern worldwide (Table 2,3).
The use of doxycycline in combination with oral rehydration solutions is suggested in some cholera cases; also, tetracyclines are extensively used against V. cholerae infection. Erythromycin is another effective option used to reduce cholera symptoms in children and pregnant women (15). In the present study, V. cholerae resistance rate to erythromycin antibiotic was 26% in Africa and developing countries, such as Mozambique and Sierra Leone. Furthermore, due to low resistance rate of V. cholerae strains to macrolide antibiotics, especially azithromycin, they are considered as the drugs of choice for cholera treatment in children and adults.
Chloramphenicol is another effective option which acts through inhibiting protein synthesis and is commonly prescribed for cholera therapy. The use of chloramphenicol has been restricted in some country such as India in the past due to the availability of more effective antibiotics with less side effects (16). Signi cant increase in furazolidone resistance might be attributed to the increased consumption of this antibiotic in countries with lower level of socioeconomic development. Generally, the choice of an appropriate antibiotic depends on drug resistance pattern of local strains. Inappropriate antibiotic prescription, over-the-counter availability of antibiotics without valid prescription, and consumption of inappropriate or partial antibiotic regimens could be the reasons for the emergence of antibiotic resistance crisis. In this regards, antimicrobials resistance is on the rise, and a recent concern is the development of antimicrobial resistance in V. cholerae strains in endemic areas (17).
This study showed that the pooled prevalence rates of V. cholerae strains which were resistant to trimethoprim-sulfamethoxazole, erythromycin, nalidixic acid, ampicillin, streptomycin, and ceftriaxone (ampicillin); Namibia, Senegal, and Nepal (streptomycin); and Mozambique (ceftriaxone). These differences in antibiotic resistance rates could be observed even within countries. It could be justi ed by taking into account that antibiotic resistance depends on multi factors in each geographical area (2).
Signi cant differences in antibiotic resistance rates from 0 to 100% in most WHO regions might be attributed to the lack of a uniform consumption pattern and access to the same antibiotics in different countries. The use of low-price antibiotics for treating this infection in developing countries, over-thecounter availability of these antibiotics, and their widespread consumption in these areas and countries might explain these outcomes. Furthermore, uncertainty about the effectiveness of simple rehydration solutions for mild and moderate dehydration could be considered as another cause of antibiotic resistance (18). Antibiotic resistance rate tended to be higher in African and Asian regions, especially in low and lower-middle income developing countries. More cholera outbreaks in these areas are the reasons for the increase in antibiotic resistance. Research results have displayed that susceptibility pattern of V. cholerae O1 strains to antimicrobials has altered over time, and that the spread of resistance to antibiotics commonly used for cholera treatment is on the rise in Africa and Asia (19). Resistance to azithromycin was higher in European countries, such as Slovakia, than in other countries. These ndings collectively suggest that pattern of antibiotic consumption in these countries should be according to antibiotic resistance pattern.
V. cholerae is an environmental pathogen that could obtain resistance genes through direct contact with inherently resistant organisms carrying resistant genes on mobile genetic elements. This meta-analysis showed that 20% of samples studied were environmental. Antibiotic resistance mechanisms including e ux pumps, spontaneous chromosomal mutation or the development of genetic resistance via the exchange of conjugative plasmids, transposons, integrons, or self-transmissible chromosomally integrating SXT element could be accounted for the mechanisms of antibiotic resistance action in V. cholerae strains.
Antibiotic resistance genes could be exchanged between V. cholerae strains and other bacteria as well as commensals or enteric microorganisms in the human intestine, raising doubt about the reliability of several of these outcomes. The results showed that resistance patterns uctuate may be induced via stable plasmids and plasmid-mediated mechanisms. Few studies investigated (21.4%) antibiotic resistance genes, but other studies did not investigate or discuss about the mechanisms of V. cholerae resistance or these alarming results (15). However, it is recommended that more investigations be carried out in order to determine the exact mechanisms of resistance action.
There are several limitations in this study. First, publications in some countries were very unusual. Between-study variation among the included studies was signi cant. Second, studies conducted on bacterial infections in few countries employed heterogeneous methodologies. A certain challenge was the lack of a standardized panel of antimicrobials against V. cholerae strains tested, thereby making it di cult to combine the results of these studies. The majority of studies performed only disc susceptibility testing without further MIC and ESBL testing or determination of MDR and XDR V. cholerae strains due to differences in patterns of economic development, antimicrobial consumption, and transmission of V. cholerae genotypes. High-quality antibiotic susceptibility testing is considered as a signi cant public health method; identifying AMR contributes to the update of local treatment guidelines and prevents the use of ineffective antibiotics. These issues are vital as our antimicrobial facilities are limited, especially in low-income epidemic areas.

Conclusions
This meta-analysis study provides an overview of the global incidence of antibiotic resistance in V. cholerae strains, exhibiting high resistance rates against various antibiotics. Therefore, to improve the e cacy of V. cholerae infection treatment, the following measures need to be taken, including exploring practical mechanisms of antibiotic resistance and dominant antibiotic-resistant elements, monitoring the incidence of antibiotic resistance in V. cholerae strains, implementing surveillance networks at the local and national levels, and designing new noninvasive methods suitable for clinical practice. Given the complexity of antibiotic-resistant V. cholerae epidemiology, strategies including antibiotic rotation, the use of previous clinical experiences, and modeling of studies to control antibiotic consumption pattern are recommended to be taken into account in future guidelines. However, given the threat imposed by antibiotic-resistant V. cholerae strains on public health, it seems more prudent to use vaccination as an alternative way to reduce antibiotic selection pressure.

Declarations
Ethics approval and consent to participate: Not applicable Consent for publication: Study selection