Antimicrobial Susceptibilities and Laboratory Profiles of Escherichia coli, Klebsiella pneumoniae, and Proteus mirabilis Isolates as Agents of Urinary Tract Infection in Lebanon: Paving the Way for Better Diagnostics

Background: Urinary tract infections (UTIs) are major healthcare problems that are usually treated empirically. However, antimicrobial resistance has been increasing across many settings. This study aims to elucidate the antibiotic resistance profiles of three common uropathogens, Escherichia coli (E. coli), Klebsiella pneumoniae (K. pneumoniae), and Proteus mirabilis (P. mirabilis) and compare between extended spectrum beta-lactamase (ESBL) and non-ESBL strains among Lebanese patients. Methods: This retrospective study was conducted at multiple tertiary healthcare centers in South Lebanon, between January and September 2017, including 551 patients of all age groups. Demographic, clinical, and laboratory data of patients were collected and analyzed statistically. Results: The prevalence of UTI in Lebanon was highest in adults between 19 and 64 years (44%). E. coli was the predominant uropathogenic organism (67.1%) followed by K. pneumoniae (10%) and P. mirabilis (3.7%). ESBL represented 32.9% of the UTI agents. The three common uropathogens studied were found to be most susceptible to imipenem (100%) and meropenem (100%). Interestingly, 115 (25.1%) out of the 458 E. coli isolates were resistant to more than eight antibiotics while 107 (23.4%) were susceptible to all antibiotics studied. Conclusions: Our study underlined the importance of adequate antimicrobial prescription for UTIs in Lebanon to avoid multidrug resistance.


Introduction
Urinary tract infections (UTIs) are major healthcare problems commonly acquired in both community-community-associated urinary tract infections (CAUTIs)-and hospital-hospital-Germany) revealed that the sample size n = 551 was enough to attain a statistical power of at least 80% with alpha error of 5%, balanced on each side, and effect size set to 5%.

Data Collection
Patient data was also collected in a special form designed for this study. These data included patients' demographics (age, gender, and date of collection), type of hospital admission (outpatient or inpatient), urine cultures and antibiograms, and patients' laboratory findings (White Blood Cell-WBC, C-reactive Protein-CRP, Blood Urea Nitrogen-BUN, creatinine, and urine pH).

Ethics Approval and Consent to Participate and Publication
The study was conducted under the Research Ethics Committee approval of the Lebanese University. In accordance with the Declaration of Helsinki, all patients enrolled in this study provided written informed consent for both participation and publication of identifying information. Ethical clearance was taken as per the norms and in accordance with relevant guidelines and regulations of the Lebanese University and the tertiary healthcare centers included. This study will be done in a manner that ensures the confidentiality of patients. In addition, chart review was carried out by CITI (Collaborative Institutional Training Initiative) certified researchers. All data collected were de-identified and stored at the principal investigator's office.

Determination of Extended-Spectrum β-Lactamases (ESBL)
ESBL production was assessed by the medical laboratory team of the healthcare centers. Strain ATCC 25,922 was used as a quality control strain for antibiotic susceptibility testing. For initial ESBL screening of each common uropathogen, isolates showing resistance to a third-generation cephalosporin antibiotic (30 µg) using the disk diffusion method were identified as potential ESBL producers. The double-disc synergy test (DDST) was carried out for the phenotypic confirmation of ESBL production. For this test, a third-generation cephalosporin (ceftriaxone) antibiotic disc (30 µg) was placed 25 mm away from a combination disc containing this last in addition to clavulanic acid (20/10 µg). According to clinical and laboratory standards institute (CLSI) guidelines [25], the zone of inhibition between the combination disc and the third-generation cephalosporin disc differed by ≥5 mm, the strain was identified as ESBL-producing.

Statistical Analysis
Quantitative variables were tested for normality distribution using the Kolmogorov-Smirnov test. Descriptive statistics were carried out and reported as frequencies and percentages for categorical variables and median and interquartile range (quartile 1-quartile 3) for continuous ones. After tabulating the patients' demographic factors, baseline comparisons between the three studied groups (E. coli, K. pneumoniae, and P. mirabilis) were performed using the Jonckheere-Terpstra test for continuous variables. A chi-squared test was used to evaluate any significant difference between the categorical variables. Mann-Whitney test were used to assess differences between ESBL and non-ESBL groups for laboratory data. Statistical Package for Social Science software (SPSS, Inc., version 24.0, Armonk, NY, USA) was used for conducting the statistical analyses. The level of significance was set at p < 0.05 for all statistical analyses.

Demographic Characteristics of Patients
During the study period (from January until September 2017), 3080 patients presented to the urology departments of the multiple tertiary healthcare centers in South Lebanon involved in our study. Out of this population, 682 (22.1%) met our inclusion criteria.
Regarding gender and age distributions, females had the highest incidence rates with 85% of the total population, with adults between 19 and 64 years old being the predominant group (44.0%). The prevalence of bacterial UTI was lowest in adolescents between 12 and 18 years old (3.7%). The number of non-hospitalized patients outnumbered approximately twice those hospitalized for UTI (data not shown). Regarding the monthly distribution, the highest incidence of UTI was in January (15.4%), and the lowest was in June (6.2%).

Prevalence of Common Uropathogens
Out of 682 isolates from patients with UTI, E. coli was the predominant uropathogenic organism (67.1%) followed by K. pneumoniae (10.0%) and P. mirabilis (3.7%). The other less common uropathogens constituted 19.2% of the cases and were not included in the analysis within our study. The prevalence of ESBL was 32.9%, and E. coli was the major ESBL-producing organism (35.2%) followed by K. pneumoniae (27.7%) ( Table 1).  Table 2 shows no significant association between the distribution of different uropathogenic bacteria and the demographic factors (different age groups, gender, and type of hospital admission). Monthly distribution of uropathogens reveals no difference between the three uropathogens as regarding their frequency of occurrence in specific months ( Figure 1). With respect to the age groups, E. coli and K. pneumoniae primarily occur in adults and elderly whereas P. mirabilis is mainly found in children. Females are always more prone to contract UTIs regardless of the uropathogen than males as expected.

Correlation between Different Uropathogens and Demographic Parameters
During the study period (from January until September 2017), 3080 patients presented to the urology departments of the multiple tertiary healthcare centers in South Lebanon involved in our study. Out of this population, 682 (22.1%) met our inclusion criteria.
Regarding gender and age distributions, females had the highest incidence rates with 85% of the total population, with adults between 19 and 64 years old being the predominant group (44.0%). The prevalence of bacterial UTI was lowest in adolescents between 12 and 18 years old (3.7%). The number of non-hospitalized patients outnumbered approximately twice those hospitalized for UTI (data not shown). Regarding the monthly distribution, the highest incidence of UTI was in January (15.4%), and the lowest was in June (6.2%).

Prevalence of Common Uropathogens
Out of 682 isolates from patients with UTI, E. coli was the predominant uropathogenic organism (67.1%) followed by K. pneumoniae (10.0%) and P. mirabilis (3.7%). The other less common uropathogens constituted 19.2% of the cases and were not included in the analysis within our study. The prevalence of ESBL was 32.9%, and E. coli was the major ESBL-producing organism (35.2%) followed by K. pneumoniae (27.7%) ( Table 1).  Table 2 shows no significant association between the distribution of different uropathogenic bacteria and the demographic factors (different age groups, gender, and type of hospital admission). Monthly distribution of uropathogens reveals no difference between the three uropathogens as regarding their frequency of occurrence in specific months ( Figure 1). With respect to the age groups, E. coli and K. pneumoniae primarily occur in adults and elderly whereas P. mirabilis is mainly found in children. Females are always more prone to contract UTIs regardless of the uropathogen than males as expected.

Distribution of ESBL-Producing Organisms among In-and Out-Patients
As expected, ESBL-producing organisms almost equally occur in in-patients (51.1%) and out-patients (48.9%) indicating that ESBL is both a hospital-and community-acquired infection, while non-ESBL is more common among out-patients (65.4%) than in-patients (34.6%) with a statistically significant different difference between the two groups (p < 0.001) ( Table 3). Table 3. Non-ESBL vs. ESBL producing organisms among in-and out-patients. A chi-squared test was used to evaluate any significant difference between ESBL and non-ESBL groups for the hospital admission. The level of statistical significance was set at p < 0.05.

Laboratory Analysis Profile of the Three Different Studied Uropathogens
When comparing blood and urine laboratory analysis related to each of the three studied uropathogens in Table 4, data showed highest median urine pH levels (6.5; Q 1 -Q 3 : 5.5-7.5) in patients with UTI caused by P. mirabilis (p = 0.480) and highest median white blood cells (WBCs) levels in those with UTI caused by E. coli (10.81; Q 1 -Q 3 : 8.25-16.27; p = 0.020). Results are consistent with what has been previously published where urine with E. coli or K. pneumoniae demonstrated the most acidic pH while P. mirabilis exhibited the least acidic pH [26][27][28]. The pathophysiological mechanism behind this is that P. mirabilis produces urease which hydrolyzes urea in urine into CO 2 and ammonia, both of which cause urine pH to increase [27,29]. As for the lower WBC count in the K. pneumoniae group, an explanation could be that WBCs have been shown in a study on rats to be reduced in the first few days post-infection and only increase after seven days of infection [30]. In addition, since urease-producing uropathogens such as P. mirabilis often convert normally acidic urine into an alkaline state which can lyse WBCs, the WBC count is reduced during an infection [31,32].

Comparison of Laboratory Analysis Profiles between ESBL and Non-ESBL Producing Organisms
Laboratory data revealed that median WBC, C-reactive protein (CRP), blood urea nitrogen (BUN), and creatinine levels were higher in patients with UTI caused by ESBL-producing organism as compared to patients with UTI caused by non-ESBL producing organism (p = 0.015, p = 0.288, p = 0.269, and p = 0.063, respectively) ( Table 5). Table 4. Laboratory data related to each of the three studied uropathogens. Jonckheere-Terpstra test was used to compare the medians of the laboratory data of patients among the three uropathogen groups (E. coli, K. pneumoniae, and P. mirabilis). The level of statistical significance was set at p < 0.05.

Antibiotic Susceptibility Profiles of the Common Uropathogens
Based on CLSI guidelines, all E. coli, K. pneumoniae, and P. mirabilis were sensitive to carbapenems and 24.7%, 44.8%, and 80% were sensitive to cephalothin antibiotic, respectively. Cefoxitin and fosfomycin showed high levels of sensitivity ranging between 85% and 97%. Susceptibility to aminoglycosides (including gentamicin and amikacin), piperacillin/tazobactam, ceftazidime, and cefepime ranged from 72% to 92% for E. coli, K. pneumoniae and P. mirabilis (Table 6).

Prevalence of MDR Strains among UTI Isolates
Upon studying the antibiotic resistance profiles of the three common uropathogens, we found that out of the 458 E. coli isolates, two were resistant to 15 different antibiotics, 115 (25.1%) were resistant to more than eight antibiotics, while 107 (23.4%) isolates were susceptible to all current antibiotics studied. On the other hand, only one K. pneumoniae isolate was resistant to 16 different antibiotics, 19 (27.9%) were resistant to more than eight antibiotics, and 21 (30.9%) were susceptible to all studied antibiotics. Remarkably, one isolate of P. mirabilis was resistant to 12 antibiotics (Data not shown).

Discussion
The emergence of MDR organisms has been recently acknowledged as an important global concern, and infections caused by ESBL-producing MDR organisms are specifically being perceived as major threats to both physicians and patients [33]. Our study focused on assessing the epidemiology of UTI among Lebanese patients attending tertiary healthcare centers in South Lebanon over a period of 9 months and evaluating the profiles of antimicrobial susceptibility to three common uropathogens, namely E. coli, K. pneumoniae, and P. mirabilis.
Our results showed that the highest incidence of UTI was during the Fall season, mostly in January, suggesting that weather conditions may be implicated in increasing the risk of UTIs and affecting the outcome of patients as well. A study by Elo et al. showed comparable results to ours [34]. In terms of demography, females had higher prevalence of UTIs owing to the anatomical and physical factors, particularly their short urethras and the small distance between the urinary system from one side and the genital/intestinal system from the other side [10]. In our study, the number of out-patients outnumbered those hospitalized for UTI, as the vast majority of cases represented uncomplicated cystitis that do not necessitate hospitalization, a fact that was reinforced by Schmiemann et al. [35].
In this study, E. coli was the most commonly detected isolate (67.1%) followed by K. pneumoniae, and P. mirabilis, representing the three mostly reported uropathogens as shown in other publications [36,37]. Our study results are consistent with results from a previous study done in Lebanon revealing that E. coli has the highest incidence (60.53-73.98%) followed by K. pneumoniae (5.32-8.33%) [37] and another study demonstrating most frequent UTI isolates as E. coli (70%), K. pneumoniae (14%), and Pseudomonas (5%) [38]. Similar results were also obtained in antimicrobial resistance epidemiological survey on cystitis (ARESC) study where E. coli was the most common isolate with a prevalence of 76.7% followed by Enterococcus faecalis, Staphylococcus saprophyticus, K. pneumoniae, and P. mirabilis [39]. In addition, the average ESBL production was found to be 32.9% during our study period and again E. coli was the major ESBL-producing organism (35.2%) followed by K. pneumoniae (27.7%). These data were consistent with a study conducted by Chamoun et al., revealing ESBL production rates in E. coli and Klebsiella spp. of 32.3% and 29.2%, respectively [16].
Our results showed that the rate of ESBL-producing organisms was significantly higher among in-patients compared to out-patients (p < 0.001). This pattern could be explained by the extensive misuse, overuse, and abuse of different antibiotics by healthcare workers or patients themselves who tend to take antibiotics without medical consultation and prescriptions [40].
Laboratory data showed that urine pH levels were highest in patients with UTI caused by P. mirabilis, an expected outcome as this organism produces urease and the direct result of urease activity is an increase in local pH level [41]. In addition, the WBC counts were higher in patients with UTI caused by E. coli. The results shown in this study were similar to the results obtained by Peltola et al., where the total WBC counts were increased in E. coli infections [42]. When it comes to the association between the laboratory results and the antimicrobial resistance among clinical bacterial isolates, conflicting reports were noted. Our study showed a clear correlation between WBC, BUN, creatinine levels and ESBL-producing organisms. However, Park et al. showed no significant association [43], which indicates the need for further studies to clarify this point and to potentially set guidelines for future management.
As for antimicrobial susceptibility, the three common uropathogens studied herein were found to be most susceptible to imipenem (100%) and meropenem (100%). This is consistent with a study conducted by Daoud et al. in which imipenem remained the only antibiotic with 100% of susceptibility [10]. It is well appreciated that rates of uropathogens with high resistance towards antibiotics have increased in recent years [44]. Indeed, in our study, some uropathogens were found to be resistant to various antibiotics. Therefore, continuous surveillance and monitoring of MDR organisms is highly needed and recommended to reduce their prevalence and prevent treatment failure.
We acknowledge that our study had several limitations, one of which is related to its short duration and missing data pertaining to some of the collected information. However, there is a scarcity of data regarding the epidemiology of UTI and the profiles of antibiotic susceptibility in South Lebanon, which constituted the main strength of this study.

Conclusions
As a conclusion, this study revealed the predominance of E. coli among other uropathogens among Lebanese patients with UTIs. It also underlined the threat of ESBL-producing bacteria which constituted around a third of the studied population. These results highlighted the importance of an adequate antimicrobial prescription for UTIs-clearance of the infection, avoiding recurrence, and being aware of multidrug resistance should all be priorities by physicians when treating patients with UTIs. This study offers a "terrain" to set new guidelines in the management of UTI; further studies are still needed in order to stratify these infections based on severity and to adjust the management accordingly.