Introduction

Uncomplicated urinary tract infections (UTIs) are one of the most common infections for which antibiotics are prescribed, and there is a current trend of increasing resistance to antibiotics typically used to treat these infections [1]. Contributing to this increased resistance is the increased frequency of extended spectrum β-lactamase (ESBL) production in organisms causing UTIs, particularly Escherichia coli and Klebsiella spp [2, 3].

Fosfomycin, trimethoprim/sulfamethoxazole (SXT), and nitrofurantoin are the 3 treatment options recommended for uncomplicated UTIs by the Infectious Diseases Society of America (IDSA) [1]. To safeguard against development of resistance, fluoroquinolones, such as ciprofloxacin and levofloxacin, are not recommended for use in treatment of uncomplicated UTI, despite their effectiveness. The IDSA guidelines also recommend that ampicillin not be used because of the increased prevalence of ESBL-producing organisms in uncomplicated UTIs, and that β-lactam antibiotics are used only when other treatment options are unavailable [1].

Ciprofloxacin and SXT resistance rates among urine isolates from female outpatients increased dramatically from 2003 to 2012, especially among elderly outpatients, highlighting the need for other oral therapeutic alternatives [4]. In a study of 1518 extraintestinal E. coli isolates, including 1034 urinary isolates, recovered from patients in the United States and Germany between 2010 and 2011, investigators observed resistance rates to nitrofurantoin, ciprofloxacin, and SXT of 6.8%, 17.1%, and 26.9%, respectively [5].

Fosfomycin is a bactericidal, cell wall active agent that binds to the MurA enzyme, thereby inhibiting the first step in peptidoglycan formation [6, 7]. It is approved, as fosfomycin tromethamine for oral administration, by the US Food and Drug Administration (FDA) for the treatment of women with uncomplicated UTIs (acute cystitis) caused by E. coli and Enterococcus faecalis and has broad-spectrum activity in vitro against Gram-positive and Gram-negative bacteria [8]. Fosfomycin has recently received renewed interest because of its activity against multidrug-resistant bacteria [7, 9, 10] and the low frequency of cross-resistance to it [8], likely due to its unique mechanism of action. Furthermore, fosfomycin has low and stable resistance rates worldwide (1–3% of E. coli) despite more than 20 years of clinical use in some countries [11, 12]. In cases where fosfomycin resistance does emerge, common mechanisms include mutations affecting transport proteins (e.g., GlpT and UhpT) and mutation of its target, MurA [11].

This study sought to compare the sensitivity of UTI isolates to fosfomycin with sensitivity to antibiotics commonly used for treatment of UTIs, including ciprofloxacin, levofloxacin, nitrofurantoin, and SXT, among US isolates from the 2012 Assessing Worldwide Antimicrobial Resistance Evaluation (AWARE) surveillance program. AWARE is a global surveillance program established to monitor the susceptibility of prespecified pathogens to a number of antimicrobial agents [13]. Medical centers from the nine US census regions contributed to the database, with organisms collected being sent for susceptibility testing to a central laboratory (JMI Laboratories, North Liberty, IA, USA) [13]. Isolates for the present study were selected proportionally to reflect the most commonly isolated species from uncomplicated UTIs.

To test the hypothesis that fosfomycin has activity against typical urinary tract pathogens, the activity against isolates with a resistant phenotype, ESBL-positive E. coli, Klebsiella spp., and Proteus spp. was assessed, as was the activity of fosfomycin against ceftazidime-nonsusceptible P. aeruginosa, Enterobacter spp., and vancomycin-nonsusceptible E. faecalis isolates.

Methods

Isolates and Reagents

Isolates collected during the 2012 AWARE surveillance program were retrospectively obtained from JMI Laboratories. Isolates were selected to obtain similar numbers of each species from each of the nine US census regions, based on their regional prevalence in the AWARE program [14]. Isolates (n = 658) included E. coli (n = 257), Klebsiella spp. (n = 156), Enterobacter spp. (n = 79), P. aeruginosa (n = 60), E. faecalis (n = 54), and Proteus spp. (n = 52). Resistant isolates included E. coli, Klebsiella spp., and P. mirabilis with an ESBL phenotype; ceftazidime-nonsusceptible P. aeruginosa and Enterobacter spp.; and vancomycin-nonsusceptible E. faecalis. The ESBL-phenotype was previously determined by JMI Laboratories and defined according to the Clinical and Laboratory Standards Institute (CLSI) ESBL screening criteria (ceftazidime and/or ceftriaxone and/or aztreonam minimum inhibitory concentration [MIC] value >1 μg/mL) [15]. Ceftazidime non-susceptibility (MICs ≥8 μg/mL) was used to categorize Enterobacter spp. Ceftriaxone and vancomycin susceptibilities were determined by JMI Laboratories using CLSI breakpoints [15]. Genotype determinations were made by JMI Laboratories using Check-points microarray kit CT–101 (Check-Points, Wageningen, The Netherlands) [16, 17]. This kit has the capability to detect bla CTX-M Groups 1, 2, 8 + 25, and 9, bla TEM wild-type (wt) and ESBL, bla SHV wt and ESBL, bla ACC, bla ACT/MIR, bla CMYII, bla DHA, bla FOX, bla KPC, and bla NDM-1 (but not bla OXA-48, bla IMP or bla VIM).

Antibiotics included in this study were fosfomycin (P5396-5G, lot BCBJ9873V; Sigma-Aldrich, St. Louis, MO, USA), ciprofloxacin (1134335, lot J0H307; USP, Rockville, MD, USA), levofloxacin (28266-1G-F, lot 1425507; Sigma-Aldrich, St. Louis, MO, USA), nitrofurantoin (1464001, lot K01060; USP), trimethoprim (Sigma T7883-5G, lot BCBB6418; Sigma-Aldrich), and sulfamethoxazole (Sigma S7507-10G, lot 048K0124; Sigma-Aldrich). Reagents used in this study included Mueller–Hinton agar (MHA; BBL 211438, lot 3217394; BD Diagnostic Systems, Sparks, MD, USA), glucose-6-phosphate (G7879-1G, lot 031M7004V; Sigma-Aldrich), and cation-adjusted Mueller–Hinton broth (CAMHB; BBL 21322; BD Diagnostic Systems).

Agar Dilution

Minimum inhibitory concentration for fosfomycin was determined by agar dilution according to CLSI documents M07-A9 and M100-S24 [15, 18]. MHA was prepared according to manufacturer’s instructions followed by the addition of glucose-6-phosphate to a final concentration of 25 µg/mL. Fosfomycin was added to MHA in 2-fold serial dilutions at concentrations ranging from 0.25 to 256 µg/mL. Approximately 104 colony-forming units (CFU)/spot were added to plates. After incubation at 35 ± 2 °C for 16–20 h, the MIC was recorded as the lowest concentration of antimicrobial agent that completely inhibited visible growth.

Broth Microdilution

Minimum inhibitory concentration values for ciprofloxacin, levofloxacin, nitrofurantoin, and SXT were determined by broth microdilution methods according to CLSI documents M07-A9 and M100-S24 [15, 18], with testing of appropriate quality control strains.

Antibiotic test plates were prepared in-house using a twofold dilution series in 96-well round-bottom plates containing antibiotic (10× solution). Bacterial suspensions were diluted in CAMHB and added to the antibiotic test plates to yield a final culture density of 5 × 105 CFU/mL. After incubation for 18 h at 35 ± 2 °C, the MIC was recorded as the lowest concentration where there was no visible bacterial growth.

Analysis

Minimum inhibitory concentration values were interpreted with respect to CLSI interpretative criteria for E. coli and E. faecalis [fosfomycin breakpoints were MIC ≤64 μg/mL (susceptible), MIC = 128 μg/mL (intermediate), and MIC ≥256 μg/mL (resistant)] [15].

Compliance with Ethics Guidelines

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964, as revised in 2008. Informed consent was not required because isolates used in this study were from a previously conducted surveillance program and no patient-identifiable data were held by the investigators undertaking the current research.

Results

A total of 658 UTI isolates were tested from the nine US census regions representing defined numbers of isolates of selected species collected during the 2012 AWARE surveillance program. Overall, MIC values for fosfomycin ranged from ≤0.25 to >256 µg/mL, and the fosfomycin MIC50 and MIC90 values for all isolates were 4 and 64 µg/mL, respectively (Table 1). A total of 202 isolates had a resistant phenotype (Table 2). Overall, 21.9% (144/658) of isolates were of the ESBL-phenotype (E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis only), 53 (8.1%) were ceftazidime-nonsusceptible E. cloacae and P. aeruginosa isolates and 5 (0.8%) were vancomycin-nonsusceptible E. faecalis isolates.

Table 1 Pathogen characteristics and antibiotic susceptibilities
Full size table
Table 2 Pathogen characteristics and antibiotic susceptibilities for resistant isolates
Full size table

Susceptibility results of the 257 E. coli isolates tested are shown in Table 1. The fosfomycin MIC50 and MIC90 values for E. coli were 1 and 4 µg/mL, respectively. Only one (0.4%) E. coli isolate (an ESBL-phenotype) was resistant to fosfomycin. Susceptibility results of the 76 E. coli isolates with an ESBL phenotype are shown in Table 2.

Summaries of pathogen susceptibilities are provided in Tables 1 and 2. Fosfomycin had MIC50 and MIC90 values of 8 and 32 µg/mL for Klebsiella spp., 4 and 32 µg/mL for Proteus spp., and 16 and 32 µg/mL for Enterobacter spp., respectively (Table 1). Most E. faecalis isolates (94.4%) were susceptible to fosfomycin (Table 1). Fosfomycin MIC50 and MIC90 values against P. aeruginosa were 64 and 128 µg/mL, respectively (Table 1). Among all 202 resistant organisms tested (including those with an ESBL phenotype, ceftazidime-nonsusceptible Gram-negative pathogens, and vancomycin-nonsusceptible E. faecalis), fosfomycin had MIC50 and MIC90 values of 4 and 64 µg/mL, respectively (Table 2). Twenty-four of these isolates (23 K. pneumoniae and 1 E. coli) were carbapenemase producers based on the detection of a KPC gene and on being nonsusceptible to meropenem (data not shown). All but two of these carbapenem resistant isolates had MICs indicative of susceptibility to fosfomycin (based on the CLSI interpretive guidelines for E. coli and E. faecalis, one was intermediate and one resistant). There were only 5 vancomycin-nonsusceptible E. faecalis tested, but all were susceptible to fosfomycin with MIC values ≤64 µg/mL (Table 2). Fosfomycin MIC50 and MIC90 values for Klebsiella spp. with an EBSL-phenotype were 8 and 128 µg/mL and for ceftazidime-nonsusceptible Enterobacter spp. were 16 and 128 µg/mL, respectively (Table 2). By comparison, the fosfomycin MIC50 for ceftazidime-nonsusceptible P. aeruginosa was numerically higher (64 µg/mL), whereas the MIC90 value was similar (128 µg/mL; Table 2).

In total, 17 (2.6%) isolates had MIC values suggesting resistance to fosfomycin (data not shown), including E. coli (ESBL-phenotype; n = 1), ceftazidime-nonsusceptible E. cloacae (n = 3), vancomycin-susceptible E. faecalis (n = 1), K. pneumoniae [n = 10; ESBL-phenotype (n = 5) and non-ESBL-phenotype (n = 5)], non-ESBL-phenotype P. mirabilis (n = 1), and ceftazidime-nonsusceptible P. aeruginosa (n = 1). Among those isolates with an ESBL-phenotype that were fosfomycin-“resistant” (n = 6), there was no common β-lactamase gene detected among all the isolates, and there was no clear correlation between β-lactamase genes present and fosfomycin resistance (data not shown). Three isolates with bla CTX-M genes were resistant to fosfomycin (2 K. pneumoniae and 1 E. coli).

Discussion

Overall, fosfomycin exhibited antibacterial activity against the panel of 658 organisms. With the exception of 1 isolate, all E. coli isolates tested were susceptible to fosfomycin. By comparison, the E. coli susceptibility rates were markedly lower for the other antibiotics tested, a significant finding given that E. coli is the pathogen most commonly found in UTI infections. The fosfomycin susceptibility rate for E. coli isolates with an ESBL-phenotype was markedly higher than any other antibiotic tested. Among the isolates with an ESBL phenotype, bla CTX-M group 1 and non-ESBL bla TEM were the most common β-lactamase genes detected (data not shown).

A study of E. coli isolated from urine of US outpatients from 2000 to 2010 reported resistance rates for ciprofloxacin, nitrofurantoin, and SXT that were slightly lower than those observed in the present study; this likely reflects a continued increase in resistance over the period 2010 to 2012 similar to that observed for the period 2000 to 2010 [20]. A recent study of E. coli isolates from patients with UTI in Canada during the period from 2010 to 2013 demonstrated susceptibility rates similar to the present study for fosfomycin (99.4% vs. 99.6%, respectively) and nitrofurantoin (96.1% vs. 90.3%, respectively) and higher than observed for ciprofloxacin (77.4% vs. 65.4%, respectively), and SXT (74.7% vs. 59.9%, respectively) [21]. US researchers also found E. coli to commonly exhibit susceptibility to fosfomycin (100%, tested using agar dilution) and nitrofurantoin (98.0%) [19].

Compared with the present study, similar susceptibility rates were also reported among ESBL-producing E. coli isolates from patients with UTI in the Canadian study for fosfomycin (100%) and nitrofurantoin (83.3%), with SXT (35.7%), and ciprofloxacin (9.5%) rates being somewhat lower [21]. Similarly, in a small US study, ESBL-producing E. coli isolates were susceptible to fosfomycin (96.7%; MIC values ranged from 0.25 to 4 µg/mL), but less so to nitrofurantoin (76.7%; MIC values of 8–64 µg/mL; susceptibility breakpoint ≤32 µg/mL) [22]. Another study of E. coli isolated from patients with UTI in China from 2004 to 2012 showed that fosfomycin susceptibility rates were high (93.4–99.4%), despite high ESBL-producing phenotype rates of 58.1% among isolates tested; ESBL-producing E. coli susceptibility rates for fosfomycin (93.8%) and nitrofurantoin (86.2%) were also similar to those reported here for isolates with ESBL phenotypes [23]. Over the period 2010 through to 2013, susceptibility rates of E. coli isolates were similarly high for fosfomycin (95.8%) in multidrug-resistant urinary isolates from a Veterans Affairs population in Boston, MA, USA [24].

The reduced susceptibility of cUTI isolates Klebsiella and Enterobacter spp. to nitrofurantoin observed in this study was also observed in a study conducted in Canada in 2001 (susceptibility rates of 69.2% and 63.0%, respectively) [25], and in a recent US study (susceptibility rates of 52.3% and 71.9%, respectively) [19]. Similarly, in the Boston area Veterans Affairs population, 64.0% of Klebsiella spp. were susceptible to nitrofurantoin, but in contrast to the current study slightly more than half of the multidrug-resistant isolates were susceptible to fosfomycin [24]. Ongoing surveillance is needed to assess the activity of nitrofurantoin against contemporary isolates of these organisms.

Overall, the rate of fosfomycin resistance was low in this study, as has been recently reported in general populations in the US (resistance rates of 1.5% and 3.3%) [19, 22], but not in a predominantly male population from the Boston area (overall resistance rate of 21.6%) [24]. There are several mechanisms by which bacteria become resistant to fosfomycin, including amino acid substitutions that modify the fosfomycin target MurA, mutations that affect fosfomycin transporters, and acquisition of genes that encode enzymes that inactivate fosfomycin [11]. Although fosfomycin resistance is not typically associated with β-lactamase production, one study demonstrated that the fosA3 gene, which encodes a fosfomycin-inactivating enzyme, can be associated with bla CTX-M genes in E. coli [26]. This may explain the fosfomycin resistance observed in a few of the isolates assessed in this study. Further testing is needed to determine whether these fosfomycin- resistant pathogens encode the fosA3 gene.

This study shares the limitations inherent to in vitro trials, including the potential limited applicability to the clinical setting. In this case, our study is further limited by relying on a subset of data from the AWARE surveillance program. The data presented here represent a subset of surveillance results of US isolates from the 2012 AWARE surveillance program, and although the data have been stratified by region and by pathogen to reflect the original database, it is possible that the susceptibility patterns of the actual isolates included in our study may have differed from the underlying AWARE data. Furthermore, the isolates that we could assess were limited to those isolates collected through the AWARE surveillance program and duplicate MIC testing was not performed. Nonetheless, the AWARE surveillance program offers a useful dataset from which to assess susceptibility patterns at a point in time. The clinical applicability of our study may be further limited as susceptibility rates may change over time or differ in other regions of the world.

Conclusion

The data reported in this study demonstrate that, compared with the other antibiotics tested, fosfomycin has greater in vitro activity against US UTI isolates, including Enterobacteriaceae with ESBL-phenotypes and other drug-resistant isolates. The susceptibility data for US surveillance isolates in this study should serve as a baseline for benchmarking any future changes in the susceptibility of uropathogens to fosfomycin as it becomes more widely used for the treatment of infections caused by resistant pathogens.