Comprehensive study of antimicrobial susceptibility pattern and extended spectrum beta-lactamase (ESBL) prevalence in bacteria isolated from urine samples

Nowadays, increasing extended-spectrum β-lactamase (ESBL)-producing bacteria have become a global concern because of inducing resistance toward most of the antimicrobial classes and making the treatment difficult. In order to achieve an appropriate treatment option, identification of the prevalent species which generate ESBL as well as their antibiotic susceptibility pattern is essential worldwide. Hence, this study aimed to investigate the prevalence of ESBL-producing bacteria and assess their drug susceptibility in Fardis Town, Iran. A total of 21,604 urine samples collected from patients suspected to have urinary tract infection (UTI) were processed in the current study. The antimicrobial susceptibility of the isolates was tested by the disk diffusion method. The ESBL producing bacteria were determined by Double Disc Synergy Test (DDST) procedure. Bacterial growth was detected in 1408 (6.52%) cases. The most common bacterial strains causing UTI were found E. coli (72.16%), followed by K. pneumoniae (10.3%) and S. agalactiae (5.7%). Overall, 398 (28.26%) were ESBL producer. The highest ESBL production was observed in E. coli, followed by Klebsiella species. ESBL producers revealed a higher level of antibiotic resistance compared with non-ESBLs. In conclusion, ESBL production in uropathogens was relatively high. Carbapenems and Aminoglycosides were confirmed as the most effective treatment options for these bacteria.


Materials and methods
Study population and samples processing. In the current descriptive cross-sectional study, 21,604 urine samples were aseptically collected from patients suspected to have UTI who were referred to Fardis Town laboratory located in Alborz province, Iran during one year (2018-2019). Positive bacterial growth was detected in urine samples of 1408 (6.52%) patients. The specimens were cultivated on Blood Agar and Eosin Methylene Blue Agar (EMB) medium (Merck, Germany) and incubated at 37 °C for 24 h. Initially, the colonies were counted. In cultures with bacterial counts of > 10 4 cfu/ml, the specimens were considered as positive, and gramstaining technique was performed. Then, bacterial genus and species were determined by standard biochemical tests.
Antimicrobial susceptibility testing. Antimicrobial susceptibility of the isolates was performed by the standard Kirby-Bauer disk diffusion method on the Mueller-Hinton agar media (Merck, Germany) using commercially available antibiotic disks (Mast, UK). The diameter of inhibition zone was measured for each antibiotic disk, and the results were defined in accordance with the CLSI guidelines 20 .
Phenotypic identification of ESBL-producing strains. Detection of ESBL-producing organisms was performed by Double Disc Synergy Test (DDST) method following CLSI recommendations. In this method, first, a suspension was prepared for each pure bacterial isolate according to the 0.5 McFarland turbidity standard and cultured on Mueller-Hinton agar. Fifteen minutes after bacterial cultures, pairs of antibiotic disks containing Ceftazidime (30 μg) with Ceftazidime/Clavulanic acid (30/10 μg), and Cefotaxime (30 μg) with Cefotaxime/ Clavulanic acid (30/10 μg) were placed on Mueller-Hinton agar medium center to center, at a distance of 20 mm apart from each other. The plates were incubated for 24 h at 37 °C. Then, the diameter of inhibition zone was measured. According to CLSI guidelines, an increase of ≥ 5 mm in the zone diameter around the clavulanic acid combination disks versus the same disks alone confirmed the presence of ESBL producer strains 20 . Ethical considerations. All ethical aspects of this research have been completely observed by the authors.
It was approved by Research and Ethics committee of the Iran University of Medical Sciences. All experiments were performed in accordance with relevant guidelines and regulations in Iran. Informed consent was obtained from all participants or their legal guardians before the study. The patient's demographic characteristics were recorded in a questionnaire and their information remained confidential. Data analysis. Data were analyzed using descriptive statistics (frequency and frequency percent, average and standard deviation), Chi square statistical test, and Fisher's exact test, hierarchical clustering analysis on SPSS software version 22. The confidence limits for statistical tests were considered as 0.95.
The results of urine culture. Positive bacterial growth was detected in urine samples of 1408 (6.52%) patients. Among uropathogens, E. coli (1016 cases, 72.16%) was the most commonly isolated species, followed by K. pneumoniae (144 cases, 10.3%) and S. agalactiae (80 cases, 5.7%). In addition, fungal infection was found in 223 cases. Out of the 1408 positive bacterial cultures, 1255 (89.13%) cases were related to females. As indicated in Table 1, the patients mostly belonged to the age groups of 60-75 years old (283 cases, 20.11%) and 45-60 years old (254 cases, 18%). In both genders, the main infectious strain was E. coli, while the proportion of E. faecalis infection in males was far higher than in females (13.07-2.47%) ( Table 1).
Hierarchical clustering analysis. Using Hierarchical cluster analysis for pattern of antibiotic effect for ESBL-negative/positive bacterial strains, we plotted 11 commonly used antibiotics for three of the bacterial strains producing ESBL (E. coli, K. pneumoniae and K. oxytoca). The goal was to cluster antibiotics with similar efficacy. Clustering was done by the nearest neighbor method and Minkowski measure. Figures 5 and 6 (Dendrogram plots) show the clustering of antibiotics with similar effects for ESBL-negative and ESBL-positive bacterial strains. As depicted in Fig. 5, the two antibiotics Amikacin and Imipenem were most similar in their effect on ESBLnegative bacterial strains.
As depicted in Fig. 6, the antibiotics Ampicillin, Cephalexin, Ceftazidime, and Cefotaxime had the most similarity in their effect on ESBL-positive bacterial strains.

Discussion
Urinary tract infection (UTI) is the second common infectious disease throughout the world caused by a wide range of microbial pathogens 21 . In the present work, from 21,604 suspected UTI patients, 1408 (6.52%) uropathogenic bacteria containing different species of Gram-negative and Gram-positive bacteria were isolated. The   www.nature.com/scientificreports/ majority of the isolates were obtained from females 1255 (89.13%). This finding is supported by other studies reporting a higher rate of UTI prevalence in female patients compared to males 14,22 . These studies suggest that females are more at risk of developing infection by uropathogens which is due to their anatomical structure 23 . In terms of age, it was found that the most frequently uropathogens were related to the age groups of 60-75 (20.11%) and 45-60 (18%) years old. These outcomes agree with previous studies in which the incidence of UTIs was higher among elderly patients 24,25 . The members of Enterobacteriaceae family especially E. coli and Klebsiella spp. are identified as vital causative agents of UTIs as they possess a number of factors including adhesion, pilli, fimbrae, and P1 blood group genotype receptor, which contribute to the attachment of bacteria to the urothelium 26 . In this regard, here the predominant urinary isolates were E. coli 1016 (72.16%) followed by K. pneumoniae 144 (10.3%). These results are in line with the earlier studies conducted by other researchers [27][28][29] . In addition, we found S. agalactiae 80 (5.7%) as the most frequently isolated Gram-positive bacterium involved in UTI.   www.nature.com/scientificreports/ At present, increasing Enterobacteriaceae producing ESBLs is a global healthcare concern because of high antibiotic resistance and the restricted treatment options 30 . In this survey, ESBL production was found in 28.26% (398/1408). In line with this result, some researchers reported the rate of ESBL production as 29% and 30.23% 31,32 . In contrast, an earlier study revealed that 11.75% of uropathogens were ESBL positive 33 which is lower than our result. Elsewhere, the researcher reported 55.4% for ESBL production rate which is larger than ours 34 . These differences may be due to geographical area, time, and the diagnostic technique used 32 .
Among the ESBL producers, the highest rate was observed in E. coli (35.7%), followed by Klebsiella spp. (22.7%) and C. diversus (4.34%). This is in accordance with other studies in which among various ESBL isolates, E. coli species was the most prominent isolates followed by Klebsiella spp. On the other hand, the minimum ESBL isolates were related to Citrobacter spp. [35][36][37] . In contrast, several studies reported the highest ESBL production among K. pneumoniae followed by E. coli which do not match our results 38-40 , since we found E. coli as the predominant ESBL producer.
The antibiotic susceptibility testing by commonly prescribed antibiotics was accomplished for the most frequent pathogens found in our study. In determining the antibiotic susceptibility patterns of E. coli, a high level of sensitivity to Imipenem (99.2%), Amikacin (97.9%), Meropenem (97.2%), and Nitrofurantoin (92.8%) was observed, while the least sensitivity was related to Piperacillin (17%) and Ampicillin (19.1%). Most studies on the antibiotic susceptibility of urinary pathogens around the world have found similar results. For example, in Ahmed et al. 's study, E. coli showed high resistance to Ampicillin and Piperacillin, and low resistance rates against Meropenem, Amikacin, and Nitrofurantoin 41  In the case of P. aeruginosa, we found it 100% sensitive to 10 antibiotics from 13 antibiotics tested. Its minimum sensitivity was observed to Cefalexin (7.1%) and Cefotaxime (15.4%). A similar study by Shah et al. reported that Imipenem, Piperacillin/Tazobactam, and Amikacin with a minimum resistance rate were the most effective antibiotics against P. aeruginosa in UTI patients 46 . In contrast, a study by Abdollahi et al. showed that P. aeruginosa isolated from UTI patients were highly resistant to Ampicillin (100%), Gentamicin (66.7%), and Nalidixic-acid (66.7%) 47 . On the other hand, we found no resistance to Ampicillin and Nalidixic-acid, while resistance to Gentamicin was very low (13.3%).
S. agalactiae isolates in our study were highly sensitive to Linezolid, Ampicillin, Cefalexin, and Nitrofurantoin. In agreement with our findings, Shayanfar et al. found high susceptibility to Ampicillin (96%), Nitrofurantoin (95.5%), Vancomycin (95%), and Norfloxacin (96.5%). They also found the most resistance to Tetracycline www.nature.com/scientificreports/ (81.6%) and Co-trimoxazole (68.9%) 48 , while in our study the most resistance was seen to Co-trimoxazole (98.7%) followed by Tetracycline (50%). Another study conducted by Tayebi et al. showed that all S. agalactiae urinary isolates were sensitive to Linezolid and Vancomycin, 99.6% to Ampicillin and nitrofurantoin, which is in line with our results. Nevertheless, they showed 12% sensitivity to Tetracycline which was lower than ours (50%) 49 .
Considering the results of antibiogram test for K. oxytoca isolates, the greatest sensitivity was related to Imipenem (100%) and Amikacin (92.3%). This agrees with Razzaque et al. 's study which found Imipenem (94.7%) and Amikacin (92.3%) as the most effective antibiotics for K. oxytoca urinary isolates. They also found sensitivity to Nitrofurantoin (84.6%) and Gentamicin (65.5%), which are different to our results 53 .
Depending on the ESBL-positive or negative bacteria, this study suggested that ESBL-producing isolates had higher resistance to some antibiotics tested compared to non-ESBL producers. For example, ESBL-positive isolates of E. coli were more resistant to antibiotics tested except Amikacin, Imipenem, Meropenem, and Nitrofurantoin. ESBL-producing K. pneumoniae presented a higher resistance rate to Ceftazidime, Cefotaxime, Ceftriaxone, Gentamycin, Cefalexin, Nalidixic-acid, and Co-trimoxazole. Also, in C. diversus isolates, ESBL production caused resistance to antimicrobial agents except for Amikacin and Imipenem. Regarding K. oxytoca, it led to resistance to Ceftazidime, Cefotaxime, Norfloxacin, Cefalexin, and Co-trimoxazole.
In line with these outcomes, a study carried out by Poovendran et al. indicated that resistance to Tetracycline, Amikacin, Ampicillin, Tobramycin, and Norfloxacin was comparatively higher among uropathogenic E. coli ESBL-producer than non-ESBL producer. Nevertheless, both groups of isolates were 100% sensitive to Imipenem 54 . Furthermore, Albu et al. found that ESBL-producing E. coli and K. pneumoniae isolates were more resistant to most of the antibiotics tested compared to non-ESBLs, except for Amikacin for E. coli and Imipenem for K. pneumoniae, which had lower resistance rates than non-ESBLs 55 . Further, another study by Abayneh et al. indicated that ESBL-positive bacteria had higher resistance to most of the antimicrobial agents tested, while in both ESBL-producing and non-ESBL-producing isolates, no resistance was observed toward Imipenem and resistance to Amikacin was low 56 . Additionally, we used Hierarchical clustering analysis based on the antibiogram pattern of ESBL-negative/positive E. coli, K. pneumoniae and K. oxytoca strains for clustering antibiotics with similar effects. In the case of ESBL positive strains, the results showed that the effectiveness of Cephalexin, Cefotaxime, and Ceftazidime was similar to that of Ampicillin, an antibiotic which had poor effect, and about the ESBL-negative strains, Amikacin and Imipenem demonstrated the most similarity in their efficacy.
In conclusion, the current study indicated a significant rate of infection with ESBL-producing Gram-negative bacilli among UTI patients. The ESBL production was found predominantly among E. coli followed by Klebsiella spp. An intensifying level of resistance to various classes of antimicrobial agents was observed among ESBL producers compared with non-ESBLs.
Hierarchical cluster analysis on the antibiogram pattern showed that in ESBL-positive bacterial strains, the efficacy of Cephalosporins; Cephalexin, Cefotaxime, and Ceftazidime was similar to that of Ampicillin, an antibiotic that had very little effect. With respect to the results of this research and similar investigations worldwide, Carbapenems and aminoglycosides were confirmed as the best options for antibiotic therapy against the ESBLproducing isolates. On the other hand, Penicillins and Co-trimoxazole are not recommended in the treatment of these bacteria; also, the administration of Cephalosporins should be limited. In conclusion, given the rise of antibiotic resistance and the high prevalence of ESBL production in Gram-negative bacteria, plus the importance of this issue in the field of treatment and public health and the costs associated with it, precise infection control and careful monitoring of antibiotic administration is crucial. Thus, routine screening of ESBL-producing isolates before prescribing antibiotics is recommended to prevent prolonged and inappropriate use of antibiotics and therapeutic failures.