ABSTRACT
Fluoroquinolones (e.g., ciprofloxacin) have become a mainstay for treating severe Salmonella infections in adults. Fluoroquinolone resistance in Salmonella is mostly due to mutations in the topoisomerase genes, but plasmid-mediated quinolone resistance (PMQR) mechanisms have also been described. In 2012, the Clinical and Laboratory Standards Institute (CLSI) revised the ciprofloxacin interpretive criteria (breakpoints) for disk diffusion and MIC test methods for Salmonella. In 2013, the CLSI published MIC breakpoints for Salmonella to levofloxacin and ofloxacin, but breakpoints for assigning disk diffusion results to susceptible (S), intermediate (I), and resistant (R) categories are still needed. In this study, the MICs and inhibition zone diameters for nalidixic acid, ciprofloxacin, levofloxacin, and ofloxacin were determined for 100 clinical isolates of nontyphi Salmonella with or without resistance mechanisms. We confirmed that the new levofloxacin MIC breakpoints resulted in the highest category agreement (94%) when plotted against the ciprofloxacin MICs and that the new ofloxacin MIC breakpoints resulted in 92% category agreement between ofloxacin and ciprofloxacin. By applying the new MIC breakpoints in the MIC zone scattergrams for levofloxacin and ofloxacin, the following disk diffusion breakpoints generated the least number of errors: ≥28 mm (S), 19 to 27 mm (I), and ≤18 mm (R) for levofloxacin and ≥25 mm (S), 16 to 24 mm (I), and ≤15 mm (R) for ofloxacin. Neither the levofloxacin nor the ofloxacin disk yielded good separation of isolates with and without resistance mechanisms. Further studies will be needed to develop a disk diffusion assay that efficiently detects all isolates with acquired resistance to fluoroquinolones.
INTRODUCTION
Salmonellosis caused by nontyphoidal Salmonella continues to be an important public health problem in the United States and worldwide (1, 2). While most enteric infections are self-limiting and do not require any treatment, antimicrobial agents are critical for treating immunocompromised patients and patients with invasive infections (3). Unfortunately, the increased prevalence of antimicrobial resistance is limiting the use of traditional antimicrobial agents such as ampicillin, chloramphenicol, and trimethoprim-sulfonamide combinations. Instead, fluoroquinolones, such as ciprofloxacin, have become a mainstay for treating severe Salmonella infections in adults (3). However, strains of Salmonella with increased levels of resistance to fluoroquinolones and treatment failures due to decreased susceptibility to ciprofloxacin have been reported (4–7). Importantly, isolates with ciprofloxacin MICs as low as 0.125 μg/ml have been associated with treatment failures (4).
Several fluoroquinolone resistance mechanisms have been described for Enterobacteriaceae (8). For a long time, chromosomal mutations in the quinolone resistance-determining region (QRDR) of the topoisomerase genes gyrA, gyrB, parC, and parE, and chromosomally encoded efflux mechanisms, were thought to be the only mechanisms of resistance. However, in the late 1990s, a plasmid-mediated mechanism called qnrA was described, and, since then, a variety of other plasmid-mediated mechanisms have been discovered, including different qnr variants, aac(6′)-Ib-cr, qepA, and oqxAB (8, 9). The plasmid-mediated quinolone resistance (PMQR) mechanisms typically confer decreased susceptibility to ciprofloxacin (MICs, 0.125 to 1.0 μg/ml) and a modest increase in nalidixic acid MICs (8 to 32 μg/ml) (8, 10, 11). The clinical impact of plasmid-mediated mechanisms is not yet known; there are currently no studies describing how well patients with such infections respond to fluoroquinolone treatment. However, as stated above, treatment failures have been associated with isolates displaying ciprofloxacin MICs as low as 0.125 μg/ml. In addition, patients infected with Salmonella serotype Typhi isolates displaying decreased susceptibility to ciprofloxacin (MICs, 0.125 to 1 μg/ml) have been shown to experience longer times to fever clearance and more frequent treatment failures than patients infected with Salmonella serotype Typhi isolates fully susceptible to ciprofloxacin (MICs, <0.125 μg/ml) (12). Given these results, it is reasonable to assume that isolates showing decreased susceptibility to ciprofloxacin due to PMQR mechanisms might influence treatment.
Another clinically relevant observation is that plasmid-mediated quinolone resistance mechanisms seem to provide a favorable background for the selection of chromosomal QRDR mutations and high-level fluoroquinolone resistance (13). In an in vitro experiment by Robicsek et al., a qnrA-positive isolate of Enterobacter cloacae was shown to acquire chromosomal QRDR mutations 100 times faster than a qnrA-lacking variant of the same isolate (14). The same experimental phenomenon was demonstrated with strains of Escherichia coli (9, 15). The in vivo selection of fluoroquinolone resistance in qnr-positive Enterobacteriaceae has also been confirmed; Poirel and colleagues reported on a qnrA-positive urinary tract isolate of E. coli (ciprofloxacin MIC, 0.5 mg/liter) that acquired ciprofloxacin resistance (MIC, ≥32 mg/liter) during treatment with norfloxacin (16).
The facts that PMQR mechanisms increase fluoroquinolone MICs and facilitate the acquisition of QRDR mutations make PMQR-harboring isolates of Salmonella clinically relevant (4, 12, 13). It is therefore important that the decreased susceptibility to fluoroquinolones these isolates show be detected during routine antimicrobial susceptibility testing. Prior to 2012, the Clinical and Laboratory Standards Institute (CLSI) provided MIC and disk diffusion breakpoints for ciprofloxacin, levofloxacin, and ofloxacin that were common for all Enterobacteriaceae (17). However, in 2012, the interpretive criteria for extraintestinal isolates of Salmonella and ciprofloxacin were reevaluated and new Salmonella-specific breakpoints were implemented (susceptible [S], ≤0.064; intermediate [I], 0.12 to 0.5; and resistant [R] ≥1.0 μg/ml) (18). The new recommendations lowered the resistance breakpoint for ciprofloxacin and extraintestinal Salmonella from ≥4 μg/ml to ≥1.0 μg/ml, and isolates with MICs ranging from 0.125 to 0.5 μg/ml were classified as intermediate (18). Similarly, new Salmonella-specific breakpoints for interpreting ciprofloxacin disk diffusion zone diameters were implemented, and the resistance breakpoint was changed from ≤15 mm to ≤20 mm (18). Thus, with the new breakpoints, isolates with first-step QRDR mutations or plasmid-mediated mechanisms displaying ciprofloxacin MICs of 0.125 to 0.5 μg/ml will be classified as intermediate. This is valuable information for clinicians who now have the possibility of either adjusting the dosage and duration of a fluoroquinolone treatment or prescribing another drug to patients infected with such isolates; with the old breakpoints, these isolates would have remained undetected and would have been classified along with fully susceptible isolates free of QRDR mutations and/or PMQR mechanisms (19). Salmonella-specific MIC breakpoints for other fluoroquinolones, such as levofloxacin and ofloxacin, have been evaluated by the CLSI and were published in January 2013 (20). The new MIC breakpoints for both levofloxacin and ofloxacin are ≤0.125 (S), 0.25 to 1 (I), and ≥2 (R). The zone diameter breakpoints for levofloxacin and ofloxacin are yet to be determined.
In the present study, we evaluated the categorical agreement among MICs for ciprofloxacin, levofloxacin, and ofloxacin based on the new Salmonella breakpoints published by the CLSI in 2013. In addition, the MIC and disk diffusion data were used to determine tentative zone diameter breakpoints for levofloxacin and ofloxacin. Finally, we sought to determine how well the ciprofloxacin, levofloxacin, and ofloxacin disk diffusion methods detect all isolates with acquired resistance mechanisms, including isolates with PMQR mechanisms.
MATERIALS AND METHODS
Bacterial isolates.The isolates used in this study were collected through the U.S. National Antimicrobial Resistance Monitoring System (NARMS), a national surveillance program that monitors antimicrobial resistance trends among Salmonella and other enteric pathogens collected from humans in the United States (21). One hundred isolates of nontyphi Salmonella with or without different quinolone resistance mechanisms were included: 40 isolates with no topoisomerase mutations or PMQR mechanisms, 40 isolates with one or more mutations in the topoisomerase genes, and 20 isolates with PMQR mechanisms and no topoisomerase mutations (11, 22). For the two latter groups, the presence or absence of PMQR genes and topoisomerase mutations was investigated by PCR and sequencing (11, 22–24).
Antimicrobial susceptibility testing.MICs were determined by broth microdilution using a frozen fluoroquinolone reference panel and cation-adjusted Mueller-Hinton broth with N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (TES) buffer (Sensititre; Trek Diagnostics, Westlake, OH). Four antimicrobial agents were evaluated: nalidixic acid (0.016 to 32 μg/ml), ciprofloxacin (0.016 to 16 μg/ml), levofloxacin (0.016 to 32 μg/ml), and ofloxacin (0.016 to 32 μg/ml). Antimicrobial susceptibility testing was performed according to the manufacturers' instructions and interpreted using CLSI guidelines. Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 served as quality control strains.
Antimicrobial susceptibilities were further tested by using the disk diffusion method described in CLSI document M02-A11 (25). The following antimicrobial disks were tested on Mueller-Hinton II agar: nalidixic acid (30 μg), ciprofloxacin (5 μg), levofloxacin (5 μg), and ofloxacin (5 μg). Escherichia coli ATCC 25922 was used as a quality control strain. The inhibition zone diameters were read according to the CLSI recommendations and interpreted using CLSI guidelines, when available (20).
PCR amplification and sequencing of topoisomerase genes.The presence of topoisomerase mutations was determined by PCR and DNA sequencing. For each isolate screened, genomic DNA was prepared by lysing the bacteria at 95°C and collecting the supernatant following centrifugation. The gyrA/B and parC/E genes were amplified using previously described primers (23, 24). Amplicons were visualized on 1.5% agarose gels stained with GelRed nucleic acid gel stain (Biotium, Inc., Hayward, CA). Nucleic acid sequencing was performed by using a 3730 DNA analyzer (Applied Biosystems, Foster City, CA).
Data analysis.For each antimicrobial agent, a scattergram was established with zone diameters on the x axis and MICs on the y axis. In addition, levofloxacin and ofloxacin MICs were plotted against ciprofloxacin MICs. The zone diameter breakpoints were selected using the modified error rate-bounded method and adjusted until the number of very major (false-susceptible), major (false-resistant), and minor (when one of the test results is intermediate and the other susceptible or resistant) errors had been minimized (26).
In order to assess the appropriateness of the breakpoints, discrepancy rates were calculated according to CLSI recommendations (26). Because of the inherent ±1 dilution variation in MIC testing and the ±3-mm variation intrinsic to disk diffusion testing (27), it is appropriate to allow greater discrepancy rates for isolates with MICs at the intermediate ±1 dilution step (high intermediate [IHigh] + 1 to low intermediate [ILow] − 1) than for isolates with MICs ≥2 dilution steps above (≥IHigh + 2) or below (ILow − 2) the intermediate MICs (26). Therefore, for each antimicrobial agent, very major, major, and minor discrepancy rates were calculated for each of these MIC groups. Obtained rates were compared to acceptable discrepancy rates presented in CLSI document M23-A3 (26). For the zone size interpretative criteria, the following acceptable discrepancy rates have been established by the CLSI: for IHigh + 1 to ILow − 1, <10% very major, <10% major, and <40% minor discrepancies; for ≥IHigh + 2, <2% very major and <5% minor discrepancies; and for ILow − 2, <2% major and <5% minor discrepancies (26).
RESULTS
Genetic characteristics of bacterial isolates.Among the 40 isolates with topoisomerase mutations, 18 contained a single mutation in codon 83 of gyrA (Ser83Phe), and 16 contained a single mutation in codon 87 of gyrA (Asp87Tyr, Asp87Asn, or Asp87Gly). Six isolates contained two or more topoisomerase mutations in the gyr and par genes. Among the 20 isolates with plasmid-mediated mechanisms, qnrS1 was the most common. None of the PMQR isolates harbored any topoisomerase mutations associated with quinolone resistance. The genetic characteristics of all isolates with plasmid-mediated mechanisms are presented in Table 1.
Characteristics of 20 non-Typhi Salmonella isolates with PMQR mechanisms
Nalidixic acid.The nalidixic acid MIC zone diameter scattergram is presented in Fig. 1. All 40 isolates lacking a quinolone resistance mechanism (topoisomerase mutations or PMQR) displayed nalidixic acid-susceptible MICs (≤16 μg/ml), whereas all 40 isolates with confirmed topoisomerase mutations displayed nalidixic acid-resistant MICs (≥32 μg/ml). The PMQR isolates displayed nalidixic acid MICs of 16 (n = 17) or ≥32 (n = 3) μg/ml and were thus defined as either susceptible or resistant. Among the three PMQR isolates that exhibited nalidixic acid-resistant MICs (MICs, ≥32 μg/ml), two contained a qnrS1 gene and one a qnrB5 gene (Table 1). By disk diffusion, all 40 isolates lacking a quinolone resistance mechanism displayed a zone diameter of ≥19 mm (susceptible), while all 40 isolates with confirmed topoisomerase mutations exhibited a zone diameter of 6 mm (resistant). The isolates with a PMQR mechanism were defined as either nalidixic acid-resistant (n = 3) or intermediate (n = 17) by disk diffusion showing zone diameters of ≤17 mm.
Nalidixic acid and ciprofloxacin MICs versus zone inhibition diameters for 100 isolates of nontyphi Salmonella. The horizontal and vertical lines represent MICs and zone diameter breakpoints recommended by the Clinical and Laboratory Standards Institute (CLSI). The numbers in brackets represent results for isolates with topoisomerase mutations, and the results for isolates with plasmid-mediated resistance mechanisms are highlighted in light gray.
Ciprofloxacin.The MIC zone diameter scattergram for ciprofloxacin is presented in Fig. 1. The ciprofloxacin MICs of the isolates lacking a resistance mechanism ranged from ≤0.016 to 0.064 μg/ml, whereas the ciprofloxacin MICs for the 40 isolates with topoisomerase mutations ranged from 0.125 to ≥16 μg/ml. Salmonella isolates with a PMQR mechanism showed ciprofloxacin MICs ranging from 0.25 to 1 μg/ml. The disk diffusion testing of isolates lacking a resistance mechanism yielded zone diameters of 30 to 38 mm. The zone diameters observed for isolates with topoisomerase mutations ranged from 6 to 30 mm. The 20 isolates harboring a PMQR mechanism displayed zone diameters of 21 to 27 mm. Interpretation of the data based on the disk diffusion and MIC breakpoints for ciprofloxacin and Salmonella published by the CLSI in 2012 gave no very major or major errors. Twelve minor errors were detected.
Levofloxacin.The MIC zone scattergram for levofloxacin is presented in Fig. 2. The 40 isolates with no resistance mechanisms displayed levofloxacin MICs of 0.032 to 0.125 μg/ml, whereas the 40 isolates with topoisomerase mutations displayed MICs of 0.25 to 16 μg/ml. The PMQR isolates, including one isolate harboring aac(6′)-Ib-cr, showed levofloxacin MICs of 0.25 to 1 μg/ml. By disk diffusion, the isolates with no resistance mechanisms showed zone diameters of 28 to 36 mm and the isolates with topoisomerase mutations showed zone diameters of 6 to 28 mm. The zone diameters recorded for the PMQR isolates ranged between 20 to 25 mm. The aac(6′)-Ib-cr-positive isolate showed a zone size of 23 mm.
Levofloxacin and ofloxacin MICs versus zone inhibition diameters for 100 isolates of nontyphi Salmonella. The horizontal solid lines indicate established CLSI breakpoints, and the vertical dashed lines represent proposed zone diameter breakpoints for levofloxacin and ofloxacin. The numbers in brackets represent results for isolates with topoisomerase mutations, and the results for isolates with plasmid-mediated resistance mechanisms are highlighted in light gray.
The newly established levofloxacin MIC breakpoints were evaluated by plotting levofloxacin MICs against ciprofloxacin MICs (Fig. 3). This comparison yielded a 94% category agreement. By evaluating the levofloxacin MIC zone scattergram, the following zone diameter breakpoints resulted in the lowest number of errors: ≥28 mm (S), 19 to 27 mm (I), and ≤18 mm (R). The discrepancy rates for this combination of MIC and zone diameter breakpoints are presented in Table 2. All discrepancy rates were acceptable according to CLSI recommendations; no very major or major discrepancies were detected. Two minor discrepancies were detected.
Levofloxacin versus ciprofloxacin MICs for 100 isolates of nontyphi Salmonella. The solid lines indicate the CLSI established MIC breakpoints.
MIC zone diameter discrepancy rates for three fluoroquinolones and 100 isolates of non-Typhi Salmonellaa
Ofloxacin.The 40 isolates with no resistance mechanisms displayed ofloxacin MICs of 0.064 to 0.25 μg/ml, whereas the 40 isolates with topoisomerase mutations showed MICs of 0.5 to ≥32 μg/ml. The PMQR isolates, including the isolate harboring aac(6′)-Ib-cr, displayed ofloxacin MICs of 0.5 to 2 μg/ml. By disk diffusion, the isolates lacking a resistance mechanism displayed zones between 25 and 31 mm. Isolates with topoisomerase mutations showed zones ranging between 6 and 24 mm. The PMQR isolates displayed zones sizes of 17 to 23 mm. The aac(6′)-Ib-cr-positive isolate showed a zone size of 20 mm.
The new ofloxacin MIC breakpoints were evaluated by plotting ofloxacin MICs against ciprofloxacin MICs (Fig. 4). The newly established MIC breakpoints resulted in 92% category agreement (8 minor errors) between ofloxacin and ciprofloxacin.
Ofloxacin versus ciprofloxacin MICs for 100 isolates of nontyphi Salmonella. The solid lines indicate the CLSI established MIC breakpoints.
The MIC versus zone diameter scattergram for ofloxacin is presented in Fig. 2. Based on the ofloxacin MIC breakpoints published by the CLSI in 2013, the following zone diameter breakpoints resulted in the lowest number of errors: ≥25 mm (S), 16 to 24 mm (I), and ≤15 mm (R). Twelve minor discrepancies were detected. All discrepancy rates were within CLSI accepted limits.
DISCUSSION
In January 2012, the CLSI introduced MIC and zone diameter breakpoints specific for Salmonella and ciprofloxacin (18). The MIC breakpoints for two other fluoroquinolones, levofloxacin and ofloxacin, have been evaluated by the CLSI and were published in January 2013 (20). In this study, we performed quinolone and fluoroquinolone susceptibility testing on 100 isolates of nontyphi Salmonella in order to evaluate the categorical agreement among MICs for different fluoroquinolones based on the new breakpoints.
The primary purpose of clinical breakpoints is to predict clinical outcomes and not to detect resistance mechanisms. However, in the case of Salmonella and fluoroquinolones, the accurate detection of isolates with acquired resistance becomes important since treatment failures have been reported for isolates with ciprofloxacin MICs as low as 0.125 μg/ml. In the present study, we determined appropriate zone diameter breakpoints for levofloxacin and ofloxacin and also sought to determine how well the ciprofloxacin, levofloxacin, and ofloxacin disk diffusion methods may detect isolates with an acquired resistance mechanism.
Disk diffusion testing using a 30-μg nalidixic acid disk is a well-established method to detect isolates with topoisomerase mutations (28, 29). Our results confirm that this method correctly assigns isolates to the appropriate category. All 40 isolates with topoisomerase mutations were categorized as nalidixic acid resistant by disk diffusion, and among the isolates lacking topoisomerase mutations, all were identified as susceptible by disk diffusion. In the present study, the PMQR isolates were categorized as either intermediate or resistant by nalidixic acid disk diffusion. Although the disk diffusion assay allowed the separation of susceptible and resistant populations, other studies have reported conflicting data and found the nalidixic acid disk to be not appropriate for the detection of PMQR mechanisms (29).
The newly established CLSI MIC and zone diameter breakpoints for ciprofloxacin and Salmonella were evaluated in an MIC zone diameter scattergram. Based on the new MIC and zone diameter breakpoints published by the CLSI in 2012, all isolates with topoisomerase mutations were identified as resistant or intermediate by broth microdilution and disk diffusion testing. Twelve minor errors were detected; 6 isolates that were defined as resistant by MIC testing were classified as intermediate by disk diffusion, and among the isolates displaying susceptible MICs, 6 were classified as intermediate by disk diffusion testing. According to CLSI recommendations, the minor discrepancy rate for the ≤ILow − 2 population (15.4%) was not within approved limits (<5%). The reason for the high number of discrepancies in the ≤ILow − 2 population becomes apparent when looking at the scattergram; the zone diameter limit of the susceptible range (31 mm) published by the CLSI in 2012 actually dissects the population defined as wild-type based on the lack of resistance mechanisms. When data points for a large proportion of isolates are close to an approved breakpoint, a high percentage of minor discrepancies are expected to occur.
Results and data from this study and others support the new MIC breakpoints for Salmonella and levofloxacin (S ≤ 0.125, I = 0.25 to 1, and R ≥ 2 μg/ml) published in 2013. In the present study, the new breakpoints resulted in the highest category agreement (94%) when plotted against ciprofloxacin MICs. In addition, the susceptibility breakpoint for levofloxacin (S ≤ 0.125 μg/ml) is supported by the work of Booker et al. (30). They used an in vitro infection model and Monte Carlo simulations to determine pharmacokinetic-pharmacodynamic (PK-PD)-based susceptibility breakpoints for levofloxacin. Based on an oral dosing regimen of 500 mg once daily for levofloxacin, the PK-PD-derived breakpoint for levofloxacin is 0.25 μg/ml. Thus, according to this model, isolates with levofloxacin MICs >0.25 μg/ml were predicted to result in poor clinical outcomes. Finally, results of the present study confirmed that the new MIC interpretive criteria will classify all isolates harboring topoisomerase mutations or PMQR mechanisms as intermediate or resistant.
The newly published MIC breakpoints for Salmonella and ofloxacin (S ≤ 0.125, I = 0.25 to 1, and R ≥ 2) resulted in 92% category agreement between ofloxacin and ciprofloxacin. According to our data, an ofloxacin MIC breakpoint that defines the susceptible population as ≤0.25 μg/ml and an intermediate category of 0.5 to 1 μg/ml would allow for a slightly higher category agreement (93%) among the drugs and also between the ofloxacin MIC and zone diameter data. However, studies have reported that low-level fluoroquinolone resistance (ofloxacin MIC, ≥0.25 μg/ml) is associated with a poor clinical response to fluoroquinolone therapy (31, 32). Therefore, a breakpoint that defines susceptible as ≤0.125 μg/ml is a more conservative and clinically appropriate breakpoint for ofloxacin and Salmonella. There is currently no PK-PD-based susceptibility breakpoint described for ofloxacin.
For practical and economic reasons, many laboratories choose to perform antimicrobial susceptibility testing using the disk diffusion method, and in some resource-limited areas, disk diffusion might be the only method available. In this study, tentative disk diffusion breakpoints for levofloxacin and ofloxacin were determined from the MIC zone scattergrams using a modified error rate-bounded method. Based on the new MIC interpretive criteria published by the CLSI in 2013 (20), the following disk diffusion breakpoints generated the least number of errors: ≥28 mm (S), 19 to 27 mm (I), and ≤18 mm (R) for levofloxacin and ≥25 mm (S), 16 to 24 mm (I), and ≤15 mm (R) for ofloxacin. Our data for ofloxacin and the suggested zone diameter breakpoint of ≥25 mm as susceptible are in close agreement with previously published data by Parry et al., who investigated the ofloxacin susceptibility of 816 isolates of Salmonella serotype Typhi (31). However, although the proposed zone diameter breakpoints for levofloxacin and ofloxacin give an acceptable category agreement between MIC and disk diffusion data, these disk diffusion assays are not optimal for the detection of isolates with an acquired resistance mechanism (topoisomerase mutations or PMQR). The levofloxacin disk yielded an overlap in the zone sizes between the isolates lacking a resistance mechanism and the isolates with acquired resistance mechanisms, and although the ofloxacin disk did not result in an overlap between these populations, the separation of zone diameters between the isolates that lacked resistance mechanisms and those that had acquired resistance mechanisms was not good enough to ensure reproducible detection of all isolates with acquired resistance mechanisms. Thus, neither the levofloxacin 5-μg disk nor the ofloxacin 5-μg disk is optimal for detecting isolates with topoisomerase mutations or PMQR mechanisms. Although the ciprofloxacin 5-μg disk did allow for all isolates with an acquired resistance mechanism to be classified as intermediate or resistant, there was substantial overlap in the zone diameters recorded for fully susceptible isolates lacking a resistance mechanism and those with an acquired resistance mechanism. Even though such an overlap might cause reproducibility issues, using the ciprofloxacin 5-μg disk is currently the best way of detecting all isolates with PMQR mechanisms. This is because there are plasmid-mediated mechanisms, such as aac(6′)-Ib-cr and qepA, that only affect ciprofloxacin and norfloxacin (8). However, in our particular study, the aac(6′)-Ib-cr-positive isolate displayed slightly elevated MICs not only to ciprofloxacin (1.0 mg/liter) but also to levofloxacin and ofloxacin (0.5 and 1.0 mg/liter, respectively). Since this isolate was nalidixic acid susceptible and did not harbor any QRDR mutations, one possible explanation for the elevated levofloxacin and ofloxacin MIC values is the presence of efflux mechanisms.
Since variation in zone size is expected with the disk diffusion method, a disk that generates a greater separation between isolates with and without resistance mechanisms is desirable. Results from a previous study indicated that a ciprofloxacin 1-μg disk might give a better separation of isolates with and without resistance mechanisms (29). Additional studies should confirm this and focus on the evaluation of other fluoroquinolone disks in order to find a method that will efficiently detect all isolates with resistance mechanisms. Until such a disk diffusion assay has been developed, Salmonella susceptibility to fluoroquinolones by disk diffusion testing is best determined by performing a combination of tests: the nalidixic acid 30-μg disk will accurately detect all isolates harboring QRDR mutations and the ciprofloxacin 5-μg disk currently represents the best option to detect isolates with plasmid-mediated mechanisms. For any nontyphoidal isolate with intermediate or resistant results with the ciprofloxacin 5-μg disk, the presence of plasmid-mediated mechanisms should be considered. As an alternative to disk diffusion, all isolates with acquired resistance mechanisms can be detected by determining the MIC to ciprofloxacin; isolates with an MIC of ≥0.125 μg/ml are likely to contain a resistance mechanism.
In summary, we present MIC and disk diffusion data for Salmonella and nalidixic acid, ciprofloxacin, levofloxacin, and ofloxacin and suggest tentative disk diffusion breakpoints for levofloxacin and ofloxacin. Since our data were obtained from a single laboratory and interlaboratory variation in susceptibility testing is expected, the suggested breakpoints should be validated by other laboratories. Further studies should focus on the development of a disk diffusion assay that will facilitate more sensitive and reproducible detection of all Salmonella isolates with acquired resistance to fluoroquinolones.
FOOTNOTES
- Received 26 September 2013.
- Returned for modification 24 October 2013.
- Accepted 23 December 2013.
- Accepted manuscript posted online 3 January 2014.
- Copyright © 2014, American Society for Microbiology. All Rights Reserved.
