Resistance Levels and Epidemiology of Non-Fermenting Gram-Negative Bacteria in Urinary Tract Infections of Inpatients and Outpatients (RENFUTI): A 10-Year Epidemiological Snapshot

Background: Urinary tract infections (UTIs) are one of the most common infections in the human medicine, both among outpatients and inpatients. There is an increasing appreciation for the pathogenic role of non-fermenting Gram-negative bacteria (NFGNBs) in UTIs, particularly in the presence of underlying illnesses. Methods: The study was carried out using data regarding a 10-year period (2008–2017). The antimicrobial susceptibility testing was performed using the disk diffusion method, E-tests, and broth microdilution. Results: NFGNB represented 3.46% ± 0.93% for the outpatients, while 6.43% ± 0.81% of all positive urine samples for the inpatients (p < 0.001). In both groups, Pseudomonas spp. (78.7% compared to 85.1%) and Acinetobacter spp. (19.6% compared to 10.9%), were the most prevalent. The Acinetobacter resistance levels were significantly higher in inpatients isolates (p values ranging between 0.046 and <0.001), while the differences in the resistance levels of Pseudomonas was not as pronounced. The β-lactam-resistance levels were between 15–25% and 12–28% for the Acinetobacter and Pseudomonas spp., respectively. 4.71% of Acinetobacter and 1.67% of Pseudomonas were extensively drug resistant (XDR); no colistin-resistant isolates were recovered. Conclusions: Increasing resistance levels of the Acinetobacter spp. from 2013 onward, but not in the case of the Pseudomonas spp. Although rare, the drug resistant NFGNB in UTIs present a concerning therapeutic challenge to clinicians with few therapeutic options left.


Introduction
Urinary tract infections (UTIs) are the second most common type of infections in the human medicine in the United States and Europe and the third most common (following respiratory tract infections and gastrointestinal infections) infectious pathologies worldwide, representing an important factor of morbidity and mortality, both among outpatients and hospitalized patients (in the latter group, they may represent 25-50% of infections overall) [1][2][3]. UTIs are a considerable economic burden for healthcare institutions and national economies; additionally, they also have a substantial economic impact, as they result in lost working days [4,5]. In fact, the annual cost of UTIs in the US has been estimated to be more than 3.5 billion US dollars [6]. The principal causes of uncomplicated

Antibiotic Resistance Levels Among Urinary Non-fermenting Gram-negative Bacteria
The resistance levels of Acinetobacter and Pseudomonas isolates against the relevant antibiotics are presented in Tables 1 and 2, respectively. To identify temporal developments in the resistance levels, the 10-year study period was divided into two five-year periods (2008-2012 and 2013-2017, respectively). The level of resistance in the Acinetobacter species was significantly higher in (p values ranging between 0.046 and <0.001) in the isolates originating from inpatients, in both study periods (excluding SMX/TMP resistance in the second half of the study period), the ratio of resistant isolates was 3-10-times higher between 2008-2012, while 3-5-times higher during 2013-2017. The differences in the resistance levels of Pseudomonas spp. was not as pronounced: While in the first part of the study period, there was a significant difference among the inpatient/outpatient isolates (p values ranging between 0.033-0.045; excluding amikacin resistance), this difference was shown only for gentamicin (p = 0.043), imipenem (p = 0.036), and meropenem (p = 0.029) in the second half of the study period; the ratio of the resistant isolates was 1.

Antibiotic Resistance Levels Among Urinary Non-fermenting Gram-negative Bacteria
The resistance levels of Acinetobacter and Pseudomonas isolates against the relevant antibiotics are presented in Tables 1 and 2, respectively. To identify temporal developments in the resistance levels, the 10-year study period was divided into two five-year periods (2008-2012 and 2013-2017, respectively). The level of resistance in the Acinetobacter species was significantly higher in (p values ranging between 0.046 and <0.001) in the isolates originating from inpatients, in both study periods (excluding SMX/TMP resistance in the second half of the study period), the ratio of resistant isolates was 3-10-times higher between 2008-2012, while 3-5-times higher during 2013-2017. The differences in the resistance levels of Pseudomonas spp. was not as pronounced: While in the first part of the study period, there was a significant difference among the inpatient/outpatient isolates (p values ranging between 0.033-0.045; excluding amikacin resistance), this difference was shown only for gentamicin (p = 0.043), imipenem (p = 0.036), and meropenem (p = 0.029) in the second half of the study period; the ratio of the resistant isolates was 1. Based on the susceptibility-patterns of the individual isolates, 9.66% of the Acinetobacter spp. and 8.54% of the Pseudomonas spp. were multidrug resistant (MDR), while 4.71% of the Acinetobacter spp. and 1.67% of the Pseudomonas spp. were extensively drug resistant (XDR), during the 10-year period overall. No colistin-resistant Acinetobacter or Pseudomonas isolates were recovered from the urinary isolates during the study period. Resistance trends of the urinary S. maltophilia were the following: Among eight outpatient isolates, six were susceptible to SMX/TMP, five to levofloxacin, four to colistin and two to amikacin in the inpatient group, among 16 isolates, 12 were susceptible to SMX/TMP, 10 to levofloxacin, eight to amikacin and seven to colistin.   Colistin 0% (n = 2) 0% (n = 3) -0% (n = 10) 0% (n = 12) a Comparison of resistance levels among isolates originating from outpatients and inpatients; Statistical analyses were performed using the Student's t-test; p values < 0.05 were considered statistically significant, n.s.: Not significant.
The epidemiological characteristics of this region, regarding other Gram-negative urinary pathogens has already been described previously: E. coli was the most prevalent (~57% for outpatients and~42% for inpatients), followed by Klebsiella spp. (~8% compared to~13%) [60], Proteus-Morganella-Providencia species (~5% compared to~7%) [61], and the CES group [Citrobacter-Enterobacter-Serratia species] (~3% compared to~3%) [62]. Thus, it can be concluded that NFGNB in the UTIs should not be neglected as important pathogens from an epidemiological standpoint, as their recorded prevalence was higher than of microorganisms in the CES group, and it was on par with members of the Proteae tribe [39]. Interestingly, the abovementioned group of bacteria are often grouped together by clinicians as "SPACE" pathogens (Serratia, Sseudomonas, Scinetobacter, Sitrobacter and Snterobacter spp.), as all of these bacteria possess AmpC-type β-lactamases in their chromosomes [63,64]. In our present study, there was a marked increase detected in the resistance levels of the Acinetobacter spp. in the second half of the study period (from 2013 onward), while this trend was not as pronounced in the case of the Pseudomonas spp., the β-lactam-resistance levels were between 15-25% among the Acinetobacter species, while for the Pseudomonas spp., the β-lactam-resistance levels were 12-28% and the aminoglycoside resistance was 13-25%. The increase in the ratio of resistant NFGNB isolates severely limits the therapeutic options available for clinicians in the infections, which is especially true for vulnerable patient populations (e.g., neonates, children, pregnant women) as some of the possible alternative drugs (fluoroquinolones, aminoglycosides) are contraindicated due to their debilitating side effects or teratogenicity [36,41,44,49,65]. In some cases, physicians have no choice but to use agents with pronounced toxicities (e.g., colistin), or newer agents with significantly higher prices (e.g., ceftazidime-avibactam, delafloxacin) [66,67]. The introduction of such novel antimicrobial drugs in the last decade may temporarily prevent the situation of untreatable infections, however, it is unknown when will they become a part of mainstream therapeutic protocols, due to financial considerations [66,68]. In addition to underlying patient factors and drug hypersensitivity, national/institutional drug availability and the local resistance profile of urinary pathogens should influence the choice of antibiotic therapy [69][70][71].
The purpose of the present study was to report on the importance of non-fermenting Gram-negative bacteria in urinary tract infections at the southern region of Hungary over a long surveillance period (10 years), in a clear and concise fashion. To the best of our knowledge, this is the longest-spanning and most detailed study originating from Hungary. The data in this study may aid the creation of a national surveillance system for urinary tract pathogens and to ascertain the relevance of non-fermenters as important uropathogens. Some limitations of this study should be noted: The retrospective design and the inability to access the medical records of the individual patients affected by these infections hindered the authors from assessing the correlation of the relevant risk factors and underlying pathologies with the NFGNB UTIs. The selection bias is a characteristic of such epidemiological studies, as most of these reports are originated from tertiary-care centers, corresponding to patients with more severe conditions or underlying illnesses [72]. Lastly, the molecular characterization of resistance determinants in the mentioned isolates was not performed, non-susceptibility was characterized by phenotypic methods only.

Study Location and Design, Data Collection
The present retrospective microbiological study was carried out using data collected, corresponding to the time period between 1 January 2008-31 December 2017, at the Institute of Clinical Microbiology, University of Szeged. This clinical microbiology laboratory serves the Albert Szent-Györgyi Clinical Center, which is an 1820-bed primary-and tertiary-care teaching hospital in the Southern Great Plain of Hungary (population: 401,500 people; 2017) [73]. Data collection was performed electronically, in the records of the laboratory information system (LIS), corresponding to urine samples positive for the NFGNB, based on the criteria below.
Samples with clinically significant colony counts for NFGNB (>10 5 CFU/mL; however, this was subject to interpretation by the senior clinical microbiologists, based on the information provided on the clinical request forms for the microbiological analysis and international guidelines) that were positive for the nitrite and leukocyte-esterase tests were included in the data analysis. Only the first isolate per patient was included in the study; however, isolates with different antibiotic-susceptibility patterns from the same patient were considered as different individual isolates. To evaluate the demographic characteristics of these infections, patient data was also collected, which was limited to sex, age at the sample submission, and inpatient/outpatient status. The study was deemed exempt from ethics review by the Institutional Review Board, and informed consent was not required as data anonymity was maintained.

Identification of Isolates
Ten microliters of each uncentrifuged urine sample was cultured on UriSelect chromogenic agar (Bio-Rad, Berkeley, CA, USA) and blood agar (bioMérieux, Marcy-l'Étoile, Lyon, France) plates with a calibrated loop, according to the manufacturer's instructions, and incubated at 37 • C for 24-48 h, aerobically. In the period between 2008-2012, presumptive, biochemical reaction-based methods and VITEK 2 Compact ID/AST (bioMérieux, Marcy-l'Étoile, France) were used for bacterial identification; from 2013 onward, the MALDI-TOF MS (Bruker Daltonik Gmbh., Billerica, MA., USA) was introduced to the workflow of the Department of Bacteriology. Mass spectrometry was performed by the Microflex MALDI Biotyper (Bruker Daltonics, Germany) instrument, using the MALDI Biotyper RTC 3.1 software (Bruker Daltonics, Germany) and the MALDI Biotyper Library 3.1 for the spectrum analysis. The sample preparation, methodology, and the technical details of the MALDI-TOF MS measurements were described elsewhere [74].

Susceptibility Testing of Relevant Isolates
Antimicrobial susceptibility testing for the Pseudomonas and Acinetobacter species was performed using the Kirby-Bauer disk diffusion method and E-tests (Liofilchem, Abruzzo, Italy) on the Mueller-Hinton agar (MHA) plates in the case of piperacillin-tazobactam, ceftazidime, cefepime, imipenem, meropenem, ciprofloxacin, levofloxacin, gentamicin, tobramycin, amikacin, and sulfamethoxazole-trimethoprim (SMX/TMP), taking into account the intrinsic resistance mechanisms of the NFGNB and the local antibiotic utilization data [44,75]. In addition, for the verification of discrepant results, the VITEK 2 Compact ID/AST (bioMérieux, Marcy-l'Étoile, France) was also utilized. Colistin susceptibility was performed using the broth microdilution method in a cation-adjusted Mueller-Hinton broth (MERLIN Diagnostik). Colistin susceptibility testing was not routinely performed, only per request of the clinicians. Susceptibility testing for the S. maltophilia was performed for sulfamethoxazole-trimethoprim, levofloxacin, colistin, amikacin, and tigecycline, according to a protocol previously described [57]. The interpretation of the results was based on EUCAST breakpoints (http://www.eucast.org). The S. aureus ATCC 29213, E. faecalis ATCC 29212, Proteus mirabilis ATCC 35659, E. coli ATCC 25922, P. aeruginosa ATCC 27853, A. baumannii ATCC 19606, and S. maltophilia ATCC 13637 were used as quality control strains. Intermediate results were grouped with and reported as resistant. Classification of the isolates as a multidrug resistant (MDR) or extensively drug resistant (XDR) was based on the EUCAST Expert Rules [76].

Statistical Analyses
Statistical analyses, including the descriptive analysis (means or medians with ranges and percentages to characterize data) and statistical tests (Student's t-test [for data on resistance levels] and Mann-Whitney U test [for epidemiological data]) were performed with the SPSS software version 24 (IBM SPSS Statistics for Windows 24.0, IBM Corp., Armonk, NY, USA,). The normality of variables was tested using Shapiro-Wilk tests [for epidemiological and resistance data]. p values <0.05 were considered statistically significant.

Conclusions
Urinary tract infections are principally caused by members of the Enterobacterales (E. coli, Klebsiella spp., CES species and Proteae), non-fermenting Gram-negative bacteria are emerging as important causative agents of UTIs, primarily affecting elderly, hospitalized patients (characterized by co-morbidities, catheterization), both in high-and low-income countries. The emergence of drug resistance in these pathogens should be closely monitored, due to their proclivity to becoming MDR and their plasticity in drug resistance mechanisms. The present report aims to summarize the results of a long-term surveillance study of resistance levels in NFGNB originating from urine samples. Although the levels of extensively drug resistant isolates was relatively low in the southern region of Hungary (<5%), an increase in the levels of non-susceptibility to the respective antibiotics (especially in case of Acinetobacter spp.) was shown. For public health purposes, the continuous surveillance of resistance trends in these pathogens (both in urinary tract infections and from invasive samples) is of utmost importance.
Author Contributions: M.G. conceived and designed the study. G.T. performed the data collection and analysis, wrote and revised the full paper. K.B. wrote and revised the full paper.