Prevalence of Class 1 Integron and Antibiotic Resistance Pattern in Pseudomonas aeruginosa Isolated from Iranian Clinical Specimens; a Systematic Review and Meta-Analysis

Background The role of integrons has been shown in the horizontal transmission of antibiotic resistance genes among bacterial isolates especially Gram-negative microorganisms in clinical settings. Objectives The aim of this study was to systematically review the prevalence of class 1 integrons and antibiotic resistance in Pseudomonas aeruginosa isolates of clinical samples of Iranian patients. Methods The Web of Science, PubMed, Scopus, and Science Direct databases were searched using preferred keywords based on the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines. The cross-sectional studies addressing the frequency of class 1 integrons and antibiotic‑resistance in P. aeruginosa isolates from clinical samples of Iranian patients published from 1 January 2000 until 31 December 2018 were included. Meta-analysis was performed using Comprehensive Meta-Analysis¬ (CMA) software. The random effects model was used for meta-analysis. The Cochran’s Q and I2 tests were applied for statistical analyses. Publication bias was assessed using Funnel plot and Egger’s linear regression test. Results Out of 911 studies retrieved in the initial search, 17 articles met the eligibility standards for being included in the meta-analysis. The Egger’s linear regression test indicated no publication bias (P = 0.95). The combined prevalence of class 1 integrons in P. aeruginosa isolates was obtained as 58.9% (95% CI: 46.3–70.4%). The highest rate of combined antibiotic resistance was related to Carbenicillin with a resistance rate of 79.9%. On the other side, the most effective antibiotic against P. aeruginosa was Polymyxin B with the resistance rate of 0%. The pooled prevalence of multi-drug resistant (MDR) P. aeruginosa isolates was 55% (¬95% CI: 33.8-75.3%). Conclusions Our findings indicated the high prevalence of class 1 integrons and antibiotic resistance among P. aeruginosa isolates of Iranian patients’ clinical samples. Also, the prevalence of MDR P. aeruginosa isolates was noticeable requiring prompt action. aeruginosa drug susceptibility, Pseudomonas aeruginosa

. Although the antibiotics are the main therapeutics to control the infections caused by this microorganism, the inherent resistance of P. aeruginosa to antibiotics is increasing [ 2].
The selective pressure of inappropriate antibiotic consumers on one hand, and the increasing use of antibiotics on the other hand are probably the main causes for the development of multidrug-resistant (MDR) P. aeruginosa in hospital settings [ 3]. MDR P. aeruginosa isolates are those showing resistant against three classes of conventional antibiotics particularly aminoglycosides, carbapenems, and fluoroquinolones [ 4].
Infections caused by MDR-Pseudomonas aeruginosa strainspresent clinically significant challenges.
The empirical antibiotic therapy for MDR P. aeruginosa has represented poor outcomes with high morbidities, mortalities, long hospital stays and high economic and therapeutic burden on both health systems and the patients [ 5].
According to the reports of the European Centre for Disease Prevention and Control(CDC), P.
aeruginosa comprises 9% of all hospital-based infections presenting the fourth most common hospital pathogen in Europe [ 6]. Also, CDC reported similar findings in the United States with the frequency of about 7% for P. aeruginosa infections in hospital settings [ 7]. A survey in Spain in 2016 reported a higher prevalence about 13% for this microorganism in ICU units of hospitals [ 8]. In European countries, the prevalence of MDR P. aeruginosa isolates was reported nearly 30% [ 9]. The rate of MDR P. aeruginosa isolates in Iran; however, has been reported between 30-100% [ 10].
Some possible antimicrobial resistance mechanisms in P. aeruginosa include the production of beta-4 lactamases, overexpression of efflux pumps, down regulation of outer membrane porins, production of AmpC or loss of OprD, genetic mutations, and finally expression of integrons on plasmids and transposons [ 6,11]. The integrons are specialized genetic structures by which bacteria can acquire resistance genes through horizontal transmission [ 12].
Integrons can acquire external drug resistance gene cassettes and integrate them by site-specific recombination. In clinical settings, integrons facilitate rapid horizontal transmission of antibiotic resistance genes among bacterial isolates especially Gram-negative microorganisms [ 13,14]. Generally, integrons comprise of an integrase gene, two conserved sequences called sul1 and int1, and a variable region harboring gene cassettes between the two conserved fragments [ 15].
Based on the structure of integrase gene, several classes of integrons have been recognized. Three main classes of integrons (i.e. class 1, 2, and 3) have been identified in Gram-negative bacteria including Enterobacteriaceae, and Pseudomonas. Among these, class 1 integrons are the most frequent in these microorganisms. Class 1 integrons usually carry one or several gene cassettes conferring resistance to a broad spectrum of antibacterial agents such as β-lactams, aminoglycosides and fluoroquinolones [ 16,17].
It is necessary to comprehend the prevalence of MDR P. aeruginosa and the mechanisms associated with antibiotic resistance (such as the presence of integrons) to eradicate infections caused by this organism. Regarding the significance of P. aeruginosa in hospital acquired infections, and the lack of a comprehensive study on the prevalence of class 1 integrons in P. aeruginosa isolates and antimicrobial resistance patterns of this organism in Iran, we aimed to investigate the prevalence of class 1 integrons and antibiotic resistance patterns of P. aeruginosa recovered from clinical samples of Iranian patients.

Results
A total of 911 relevant articles were obtained in the primary literature search (Figure 1). Out of these, 456 duplicate studies were excluded. Also, 90 records with irrelevant titles were deleted. After reading the abstracts of 365 remained papers, 259 were excluded owing to justified reasons. Then, 106 full text studies assessed. From these, 89 studies were omitted due to either lack of data accessibility, missing data, or not reporting the frequency of class 1 integrons. Finally, 17 articles were included in the meta-analysis ( Figure 1).
The combined prevalence of class 1 integrons in P. aeruginosa isolated from clinical specimens of Iranian patients varied from 13.3% to 99.1% ( Figure 2

Overall effects
There was a statistically significant heterogeneity among the included studies (Q2=341.7, I 2 =95.3, t=1.2, P=0.00). Accordingly, the random effects model was applied to combine the prevalence of class 1 integrons in P. aeruginosa isolates. The combined prevalence of class 1 integrons was obtained as 58.9% (95% CI: 46.3-70.4%) in Iranian patients' clinical specimens ( Table 2). The publication bias was checked using Funnel plot. Concerning possible asymmetrical data distribution in the selected studies, the Egger's linear regression test was used to further evaluate any publication bias. Nevertheless, the results of the Egger's linear regression test revealed no publication bias (P=0.95). The pooled prevalence of MDR P. aeruginosa isolates was obtained as 55% (95% CI: 33.8-75.3%). Subgroups analysis revealed that the highest combined antibiotic resistance belonged to Carbenicillin following by Cloxacillin and Cefotaxime with respective resistance rates of 79.9%, 77.4%, and 76.6%. On the other hand, the most effective antibiotics against P. aeruginosa were Polymyxin B and Colistin each with resistance rate of 0% (Table 3).

Discussion 6
The non-lactose fermentative Gram-negative pathogens, especially P. aeruginosa, are emerging causes of nosocomial infections in patients hospitalized in ICU and Burn units [ 19].
In the present study, the combined prevalence of class 1 integrons in P. aeruginosa isolates was obtained as 58.9% ranging from 13.3% to 99.1%. About 64.7% of the included studies described a correlation between the presence of class 1 integrons and antibiotic resistance among P. aeruginosa In the current study, the most effective antibiotics against P. aeruginosa isolates were Polymyxin B and Colistin each with susceptibility rate of 100%. This is while most of P. aeruginosa isolates represented the high resistance toward Carbenicillin (79.9%), Cloxacillin (77.4%), and Cefotaxime 8 (76.6%). Polymyxin B and colistin are among the most important anti-pseudomonal antibiotics with the highest effects against MDR P. aeruginosa isolates. However, both of these antibiotics have been associated with side effects and toxicities [ 6]. In fact, the restricted prescription of Polymyxins because of their toxicity is probably the most important reason for the high susceptibility rate (100%) of P. aeruginosa isolatesin exposure to these antibiotics [ 28]. Accordingly, a report from Spain in 2015 revealed a high rate of combined resistance to three or more frequently prescribed antimicrobial agents (piperacillin-tazobactam, Ceftazidime, fluoroquinolones, aminoglycosides and Carbapenems) among P. aeruginosa isolates [ 29]. As well, polymyxins have shown the highest antibacterial activity among XDR P. aeruginosa isolates [ 9] which is in agreement with our results.
To the best of our knowledge, fluoroquinolones (such as Ciprofloxacin) are among the most effective available antibiotics for the treatment of P.aeruginosa infections, particularly urinary tract infections [ 30]. In this study; however, more than 50% of P. aeruginosa isolates showed resistance against Ciprofloxacin. According to the studies conducted in Latin America and Europe, 25 It is noteworthy that hospital acquired infections caused by MDR bacterial isolates are relatively frequent in developing countries such as Iran in comparison with developed nations such as the USA 9 and most European countries. This difference can be explained by the effective programs implemented to prevent and control nosocomial hospital infections in developed countries [ 34]. This is while the hospital-based committees for controlling infections are either unavailable or their members are inexperienced and untrained in developing countries. Therefore, it is advisable to recruit clinical microbiologists for effective management of nosocomial infections in hospitals [ 35,36].
Overall, the prevalence of MDR P. aeruginosa isolates was high in clinical samples obtained from Iranian patients especially those hospitalized in critical care units (i.e. ICU and Burn). Furthermore, a high penetrance of class 1 integrons was noted in the MDR P. aeruginosa isolates delineating their association with antibiotic resistance in these bacteria. Timely reporting of antibiotic resistance patterns in these bacteria is recommended to prescribe appropriate antibiotics. Also, it is recommended to develop hospital-based infection control committees and educate their members regarding the nosocomial infections control programs.
In this review, we included only the articles published in English which is a limitation of our study.
Also, we did not include unpublished literatures in the present review.

Conclusions
Our findings indicated the high prevalence of class 1 integrons and antibiotic resistance among P. aeruginosa isolates of Iranian patients' clinical samples. Also, the prevalence of MDR P. aeruginosa isolates was noticeable requiring prompt action.

Materials And Methods Search strategy
The studies addressing the prevalence of class 1 integrons and antibiotic resistance pattern of P. aeruginosa isolatesfrom Iranian patients' clinical specimens published from January 2000 to the end of 2018 were included. The search was carried out in Web of Science, Cochrane Library, Scopus, PubMed, and Google Scholar databases. The search was restricted to the English original articles reporting either the prevalence or incidence of P. aeruginosa. The following keywords from Medical Subject Headings or titles or abstracts were used with the help of Boolean operators (and, or): P. aeruginosa, burned patients, drug susceptibility, and drug resistance. The keywords of Integrons, Pseudomonas aeruginosa or P. aeruginosa, prevalence/incidence and distribution of INT1 in Pseudomonas aeruginosa, Iran, Int1, antibiotic resistance, and local studies were used alone or in combination to conduct a complete search antibiotic resistance.

Inclusion and exclusion criteria
All the original articles reporting the prevalence, incidence and distribution of integrons in P. aeruginosa isolated from Iranian patients' clinical specimens were included. Only those studies that performed antibiotic susceptibility test based on CLSI guidelines were considered. Foreign studies (i.e. not performed in Iran) and those studies that did not follow CLSI directions in performing antibiotic susceptibility test were excluded. Also, narrative and systematic reviews, meta-analyses, editorials, prospective studies, congress abstracts, case reports, and letters to the editors were omitted. Studies published in languages other than English, articles only available in abstract form and also duplicate publications were excluded.

Data extraction
The intended data including first author's name, year of publication, location of study, sample size, the frequencies of MDR and class 1 integrons, hospital wards, and correlation between the presence of integrons and antibiotic resistance were recorded in a data extraction form designed by the researchers.

Qualification of the studies
The strengthening the reporting of observational studies in epidemiology (STROBE) checklist was utilized for qualifying the methodology of the included studies [ 18]. Based on the qualification criteria, the studies were categorized as high, medium or low quality.

Statistical analysis
The data was analyzed using Comprehensive Meta-Analysis software (Version 3.3.070). The prevalence of class 1 integrons in P. aeruginosa isolated from Iranian patients' clinical samples was reported with 95% confidence interval (CI). The random effects model was used for meta-analysis.

11
The statistical heterogeneity among the studies was determined by Cochrane Q and I 2 tests. To evaluate possible publication bias, the funnel plot and quantitative Egger weighted regression test were applied. P value<0.05 was considered as the statistical significance cut off for detecting any publication bias.