Cephalexin susceptibility breakpoint for veterinary isolates: Clinical Laboratory Standards Institute revision

Abstract

The Clinical and Laboratory Standards Institute (CLSI) uses cephalothin as the class representative for testing veterinary isolates for susceptibility to other first-generation cephalosporins, including cephalexin. We examined replacing cephalothin with cephalexin because cephalexin is used more often clinically. Bacterial isolates were obtained from dogs and cats from a national surveillance program. CLSI testing methods were used to determine the MIC for 4 cephalosporins used in veterinary medicine. Cephalexin clinical breakpoints for canine isolates were established by using published pharmacokinetic data and Monte Carlo simulations to calculate the probability of target attainment (PTA). For 1,112 Staphylococcus pseudintermedius isolates, the mode, MIC₅₀, and MIC₉₀ were 1, 2, and 64 µg/mL, respectively, for cephalexin, and ≤0.06, 0.12, and 2 µg/mL for cephalothin. Susceptibility of S. pseudintermedius from 2011 to 2014 did not change for the 4 cephalosporins tested. Only 4.3% of the penicillin-binding protein 2a–positive S. pseudintermedius isolates had MIC values ≤2 µg/mL for cephalexin, but 66.3% of these isolates had MIC values ≤2 µg/mL for cephalothin. There were also discrepancies between cephalexin and cephalothin for other bacteria tested, but the largest difference was for S. pseudintermedius, with a MIC difference of 4 doubling dilutions. Cephalexin interpretive categories (breakpoints) of ≤2 μg/mL (susceptible), 4 μg/mL (intermediate), and ≥8 μg/mL (resistant) were established for isolates obtained from dogs. Cephalothin should not be used for susceptibility testing of cephalexin for veterinary bacterial pathogens, and canine-specific breakpoints should be used for testing susceptibility. Breakpoints determined using the methods described herein for the interpretive categories will be added to future CLSI tables to reflect this recommendation.

Introduction

There are several cephalosporin commercial formulations in the United States and other countries. Cefadroxil, cefpodoxime, and ceftiofur are approved in the United States for veterinary medicine, and cefpodoxime and ceftiofur have minimum inhibitory concentration (MIC) breakpoints and zone diameter interpretations that have been established and published by the Veterinary Antimicrobial Susceptibility Testing subcommittee (VAST) of the Clinical Laboratory Standards Institute (CLSI).⁸,⁹ These published standards provide veterinary microbiology laboratories with the information necessary to report the susceptibility of veterinary isolates. When clinical breakpoints are not available for a veterinary species, laboratories may report the susceptibility test result on the basis of a human standard. When human standards are used, there is a footnote in the CLSI document to indicate that interpretive criteria and MIC breakpoints are taken from the human CLSI document. However, the human breakpoint may not be accurate because veterinary and human susceptibility testing breakpoints are often different.⁹

The 2015⁹ and earlier editions of the CLSI veterinary tables did not list cephalexin for primary testing, and recommended use of cephalothin susceptibility tests to predict the other first-generation cephalosporins. The tables that listed cephalothin for testing in dogs and cats included the footnote: “The results of cephalothin susceptibility tests are used to predict susceptibility to the first-generation cephalosporins, such as cephapirin, cephalexin, and cefadroxil.”⁹ Other cephalosporins (ceftiofur and cefpodoxime) have their own breakpoints and are listed in the CLSI tables for dogs and other animals.⁹

The practice of using cephalothin as the class representative for susceptibility testing of other oral cephalosporins for people started in the 1970s because of a decision by the U.S. Food and Drug Administration (FDA).¹⁶ This decision by the FDA was eventually adopted by CLSI after studies³,²¹ concluded that cephalothin was a reasonable representative for other oral first-generation cephalosporins. However, there has been a call to re-evaluate the use of cephalothin as the class representative for oral cephalosporins.¹⁶ When cephalothin MIC results were compared to cephalexin MIC results, there were high minor and major error rates, suggesting that cephalothin was a poor predictor of cephalexin susceptibility.¹⁶ The authors concluded that CLSI should reconsider the recommendation to use cephalothin susceptibility as the surrogate predictor of susceptibility to cephalexin.¹⁶

The CLSI VAST subcommittee followed up on this recommendation. As a result of the analysis described herein, cephalexin will be listed in future versions of the veterinary CLSI document VET08 (formerly called VET01-S and scheduled for publication in 2018). We provide the framework and data that led to this change in listing for cephalexin in CLSI veterinary documents.

We examined microbiological and pharmacokinetic data from dogs to establish susceptibility testing interpretive categories and clinical breakpoints for cephalexin in dogs. Prior to this study, there were no cephalexin breakpoints available for testing bacteria isolated from dogs, even though this is a common oral cephalosporin. We also compared the MIC distributions for cephalexin to cephalothin. Cephalothin has been used in previous CLSI documents as a test for all first-generation cephalosporins, except for cefazolin. Because cephalexin is more commonly used clinically in dogs, the goal of the CLSI subcommittee was to generate sufficient data that cephalothin could be replaced with cephalexin in CLSI tables for susceptibility testing. We also examined susceptibility data (MIC values) spanning 2011–2014 to determine if there are any observable trends in susceptibility during that time. Although our primary objective was to evaluate the testing of Staphylococcus pseudintermedius isolated from dogs, other bacteria were also included in our study because cephalosporin susceptibility of these isolates from dogs is also performed in veterinary laboratories. The other organisms included were Escherichia coli (dog and cat), Staphylococcus aureus (dog), β-hemolytic Streptococcus spp. (dog), Pasteurella multocida (cat), and Proteus mirabilis (dog).

CLSI guidelines¹⁰ require 3 criteria to establish a breakpoint: 1) pharmacokinetic data in the test species, 2) MIC distribution data for the pathogen(s), and 3) evidence of clinical efficacy at the dose analyzed. Furthermore, pharmacokinetic–pharmacodynamic (PK-PD) targets must be known in order to derive a PK-PD cutoff value for this analysis.

Materials and methods

Evidence of efficacy

A literature search was conducted to identify published evidence of efficacy of cephalexin in canine patients.

Pharmacokinetic data and protein binding

Pharmacokinetic data were compiled through a literature search. The data were evaluated for appropriate methods and analysis and summarized in tables. Because pharmacokinetic data were available from several studies, it was necessary to calculate an overall mean value and standard deviation using a weighted least square means approach that accounted for both between-study and within-study variation. Protein binding data were obtained from previous studies.

Doses examined

To perform the PK-PD modeling and simulations for cephalexin, it was necessary to identify a clinically acceptable dose for oral administration to dogs. The literature was searched, and recommendations were examined from a FDA-approved formulation (Rilexine chewable tablets for dogs, Virbac, Ft. Worth, Texas).

Collection of bacterial isolates

The MIC data were collected as part of a surveillance program conducted by Zoetis (Zoetis Veterinary Medicine Research & Development, Kalamazoo, MI. Study A671Z-US-15-079, Project 7AMFQA7079). The S. pseudintermedius isolates used in our study were obtained from laboratories in the United States and Canada that were enrolled in the Zoetis Companion Animal Susceptibility Surveillance Program during 2011–2014. These test sites identified and provided pathogens from naturally occurring canine and feline urinary tract and skin and/or soft tissue infection cases submitted by primary care and/or general care practices. Laboratories sent no more than a single isolate per animal and no more than one isolate of each bacterial species from a single household. Identifications were confirmed by the Zoetis microbiology laboratory, and all S. pseudintermedius isolates from the 2011 (n = 199), 2012 (n = 303), and 2013 (n = 265) programs were held in the Zoetis Animal Health Development Research Culture Collection. Isolates were subcultured onto sterile media, incubated for 18–24 h, checked visually for purity, and suspended in trypticase soy broth supplemented with 10–15% glycerol to make a heavy suspension. A 1-mL aliquot of the suspension was frozen at ~–70°C and was sent frozen on dry ice to Microbial Research (MRI; Ft. Collins, CO) for MIC testing. The S. pseudintermedius isolates from the 2014 program (n = 345) were stored at MRI and were therefore not transferred. Other bacterial species were collected during the same surveillance program and analyzed to provide a comparison between cephalothin and cephalexin. The other bacterial species included were E. coli (n = 143), S. aureus (12), β-hemolytic Streptococcus spp. (27), and Proteus mirabilis (67).

Testing for penicillin-binding protein 2a

All staphylococci submitted to the Zoetis Companion Animal Susceptibility Surveillance Program had been tested for the presence of penicillin-binding protein 2a (PBP2a) using an immunochromatographic membrane monoclonal antibody assay (Alere PBP2a culture colony test, Alere, Scarborough, ME). Samples of bacterial cultures were evaluated following the test manufacturer’s instructions. Although this test has been validated for PBP2a detection for isolates of S. aureus from human patients, the Zoetis microbiology laboratory has observed 100% correlation (>500 isolates tested using both methods, unpublished data) of this test with the presence of the mecA gene when tested using a PCR method.²²

Quality control determinations

Prior to our study, cephalothin was used as the class representative for testing veterinary isolates for susceptibility to other first-generation cephalosporins. Cephalexin testing using CLSI standards was not possible because there were no quality control (QC) ranges accepted and published by CLSI. Therefore, to accomplish our study, it was necessary to establish QC ranges for cephalexin. The tests were coordinated and conducted by a commercial partner (Clinical Microbiology Institute, Wilsonville, OR). Eight laboratories were used for the QC analysis, one more than needed to meet CLSI test requirements.¹⁰ American Type Culture Collection (ATCC) control strains for testing were S. aureus ATCC 29213 and E. coli ATCC 25922. Statistical analysis of the QC data was performed as described by CLSI ¹⁰ to identify the QC ranges. Cephalothin was also tested using these procedures as a comparison.

Minimal inhibitory concentration determinations

MRI conducted all MIC tests for our study. The laboratory strictly adhered to CLSI standard methods.⁸ Drugs tested included amoxicillin–clavulanic acid, cephalothin, cefpodoxime, cefadroxil, cephalexin, and oxacillin. Prior to testing, bacterial isolates were subcultured onto fresh trypticase soy agar plates. Plates were subcultured a second time and incubated overnight at 35 ± 2°C in ambient atmosphere. Purity was checked, and the overnight growth was used for MIC determinations. Cephalexin MICs for all isolates were determined using a broth microdilution method.⁸ Direct colony suspensions were used when testing cephalexin against all organisms. Panels were read manually.

The ATCC QC strains, described above, were used each day of testing. Each run was validated by confirming that the QC results met the following CLSI-approved criteria for the respective MIC results: S. aureus ATCC 29213 = 1–8 µg/mL; E. coli ATCC 25922 = 4–16 µg/mL. MIC values were determined using cephalexin panels prepared in a custom panel (Sensititre custom panel CMP1ZLEX, TREK Diagnostic Systems, Thermo Fisher Scientific, Oakwood Village, OH). These panels included a cephalexin concentration range of 0.06–64 µg/mL.

Cephalexin MIC values of ≤2 µg/mL for PBP2a-positive isolates or MIC values of ≥4 µg/mL for PBP2a-negative isolates were verified by repeat testing. When cephalexin MIC results were within one doubling dilution of the first test, the first result was used for analysis. On those occasions where results differed by more than one doubling dilution, the isolate was tested a third time and the last test result was used for analysis unless the results were inconsistent with both of the first 2 tests. When an isolate tested differently 3 times, the result was not included in the data analysis.

E. coli (dog and cat), S. aureus (dog), β-hemolytic Streptococcus spp. (dog), P. multocida (cat), and P. mirabilis (dog) were tested using CLSI methods⁸ and a MIC range of 0.06–32 µg/mL for cephalothin and cephalexin. In addition to the examination of MIC values from the Zoetis surveillance program, we also compared to MIC data generated for cephalexin in other published studies.⁴,^15,19,24

Pharmacokinetic simulations

Pharmacokinetic data were used to construct simulations of plasma concentration necessary to meet PK-PD targets. The pharmacokinetic data were entered into the Monte Carlo simulation program (Oracle Crystal Ball v.11.1.2.3.500, Oracle, Redwood Shores, CA). Monte Carlo simulations generated the probability of plasma drug concentrations reaching a time above MIC for >50% of the time interval for MIC values of 0.25–64 µg/mL. Variation (standard deviation [SD]) in the assumption variables was taken from the pharmacokinetic analysis.

Simulations were performed to calculate the probability of target attainment (PTA). The target selected was 50% time above the MIC for a 24-h interval based on previous recommendations.²⁷ Target attainment (% time > MIC) was calculated for a 24-h interval using the following formula:

$$\%\mathrm{time}>\mathrm{MIC}=ln(\mathrm{dose}/[\mathrm{VD}\times \mathrm{MIC}])\times (\mathrm{T}\mathrm{½}/\ln 2)\times (100/\mathrm{DI})$$

Where ln is the natural logarithm, VD is the apparent volume of distribution (listed as VD/F for a non-IV dose), T½ is the terminal half-life, and DI is the dose interval. Monte Carlo simulations were generated for 1,000 trials. Data entered for forecasting were the values for the above formula—MIC, half-life, dose intervals, and dose—as well as the variability of the data (SDs of the parameters) and were allowed to vary independently in the simulations assuming a log-normal distribution. Protein binding is also incorporated into the analysis because the concentration is intended to apply only to the free drug (protein unbound) fraction. Therefore, because cephalexin has 16–25% plasma protein binding in dogs, a fraction unbound (fu) value of 0.84–0.74 was used in the analysis.¹⁸

Results

Evidence of efficacy

Oral cephalexin has been effective for treating canine pyoderma at doses of 22–35 mg/kg orally every 12 h, as well as 15 mg/kg every 12 h, and 30 mg/kg once daily according to the published literature²⁵ or conference papers (Guaguere E, et al. Cephalexin in the treatment of canine pyoderma: comparison of two dose rates [abstract]. Proc Eur Soc Vet Dermatol 1996:82; Maynard L, et al. Clinical efficacy of cephalexin administered once or twice daily by oral route in the treatment of pyoderma in dogs [abstract]. Eur Soc Vet Dermatol and Eur Coll Vet Dermatol. Vet Dermatol 2003;14:238; Maynard L, et al. Clinical efficacy of cephalexin administered by oral route at two dosages in the treatment of deep pyoderma in dogs [abstract]. Eur Soc Vet Dermatol and Eur Coll Vet Dermatol. Vet Dermatol 2003;14:238). In clinical infections in dogs treated with cephalexin (26–39 mg/kg twice daily), a 90% response for skin, soft tissue, urinary tract, and respiratory infections was reported.⁶ Cephalexin and cefadroxil were equally effective for treatment of bacterial pyoderma in dogs at doses of 22–35 mg/kg orally every 12 h.¹² When 157 dogs with bacterial pyoderma were treated with oral cephalexin, there was 93.9% success at a dose of 26 mg/kg twice daily.⁷ In addition to these studies, cephalexin has been FDA-approved in dogs for treatment of skin infections (Rilexine) at a dose of 22 mg/kg twice daily for 28 d for the treatment of secondary superficial bacterial pyoderma in dogs caused by susceptible strains of S. pseudintermedius.

Pharmacokinetic data and protein binding

Pharmacokinetic data were obtained from published studies and relevant pharmacokinetic values for cephalexin considered for this analysis (Table 1).⁴,⁵,¹¹,¹⁴,¹⁸,²⁰,²³,²⁸ An overall mean value and SD that accounted for both between-study and within-study variation was calculated (Table 2). Protein binding of cephalexin was found to be 16–26% in dogs, depending on the concentration as determined by the ultrafiltration method.¹⁸

Table 1.

Pharmacokinetic data for cephalexin in dogs.

Clearance/F (mL/kg/min)	VD/F (L/kg)	Half-life (h)	CMAX (µg/mL)	Reference
NR	NR	2.6	27.8	4
2.67	NR	1.78	NR	11
3.54 (0.31)	0.52 (0.042)	1.7 (0.08)	18.6 (1.7)	23
2.75 (0.68)	1.57 (0.56)	6.52 (1.49)	21.9 (4.4)	28
NR	0.81	2.48 (0.29)	20.3 (1.7)	5
2.8 (0.50)	1.162 (0.431)	4.7 (1.1)	31.5 (11.5)	18
5.11 (0.70)	0.78 (0.15)	1.79 (0.26)	18.77 (2.83)	20
2.67	NR	2.16 (0.3)	19.5 (5.6)	14

Results are presented as a mean value, with standard deviation in brackets. Oral absorption was reported to be 91% and 57% in 2 studies.⁵,²⁸ C_MAX = peak plasma drug concentration; Clearance/F = clearance per fraction absorbed; NR = not reported in relevant study; VD/F = volume of distribution per fraction absorbed.

Table 2.

Overall mean and standard deviation for cephalexin compiled from 8 pharmacokinetic studies reported in Table 1.

Property	Total no. of observations	Mean	Standard deviation
Clearance/F (mL/kg/min)	52	3.14	0.87
Volume of distribution/F (L/kg)	33	0.92	0.48
Half-life (h)	62	2.74	1.60
Peak concentration (µg/mL)	56	19.52	6.90

Clearance/F = clearance per fraction absorbed; volume of distribution/F = volume of distribution per fraction absorbed.

Dosages examined

Clinically accepted doses were obtained from published studies, conference proceedings, and product labels. Published studies demonstrating efficacy used dosages of 22–35 mg/kg orally every 12 h,¹,⁶,¹² as well as 15 mg/kg every 12 h, and 30 mg/kg once daily²⁵ (Guaguere et al., 1996; Maynard et al., 2003). Current clinical practice for cephalexin is to administer 25–30 mg/kg orally twice daily.¹⁷ The dose for the approved product (Rilexine) is 22 mg/kg orally twice daily. The dose used for our analysis was 25 mg/kg orally twice daily.

QC ranges

All cephalexin QC results for S. aureus ATCC 29213 and E. coli ATCC 25922 were within the range accepted by CLSI⁹: 1–8 µg/mL and 4–16 µg/mL, respectively. The corresponding QC ranges for cephalothin for these ATCC strains was 0.125–0.5 µg/mL for S. aureus, and 4–16 µg/mL for E. coli.

MICs

The cephalexin MIC data generated at MRI were captured manually using an MRI data capture form, entered into a spreadsheet, and reported electronically to Zoetis. The cephalexin MIC values generated were then compared to cephalothin, cefpodoxime, and cefovecin MIC values previously generated on these isolates using an identical protocol. The MIC mode, MIC₅₀, MIC₇₀, MIC₈₀, and MIC₉₀ values for cephalexin, cephalothin, cefovecin, and cefpodoxime were determined for S. pseudintermedius isolates from dogs (Table 3). The population frequency distribution was determined (Supplementary Table 1). Additionally, the cephalexin and cephalothin MIC distributions were compared to PBP2a results (Fig. 1, Supplementary Table 2). For PBP2a-positive isolates, the majority (84.7%) had cephalexin MIC values greater than or equal to the resistant breakpoint of 8 µg/mL, with only 4.3% being less than or equal to the susceptible breakpoint of 2 µg/mL. On the other hand, 66.3% of PBP2a isolates had cephalothin MIC values less than or equal to the susceptible breakpoint of 2 µg/mL. All (100% for cephalothin) or almost all (99.2% for cephalexin) of the PBP2a-negative isolates were less than or equal to the susceptible breakpoint of 2 µg/mL.

Table 3.

MIC₅₀, MIC₉₀, and mode values for cephalothin and cephalexin.

Pathogen	n	Animal species	Site	MIC50		MIC90		Mode
				CEP	CLX	CEP	CLX	CEP	CLX
Pasteurella multocida	5	Cat	Skin	NA	NA	NA	NA	0.25	2
Staphylococcus pseudintermedius	39	Cat	Skin	0.12	2	8	>32	≤0.06	1
	1,073	Dog	Skin	0.12	2	2	>32	≤0.06	1
Staphylococcus aureus	12	Dog	Skin	0.5	8	32	>32	0.5	>32
Β-hemolytic Streptococcus species	27	Dog	Skin	0.25	0.25	0.25	0.5	0.25	0.25
Staphylococcus pseudintermedius	57	Dog	Urine	≤0.06	1	0.12	2	≤0.06	1
Proteus mirabilis	67	Dog	Urine	4	16	8	16	4	16
Escherichia coli	172	Dog	Urine	16	8	>32	>32	16	8
	71	Cat	Urine	16	8	32	16	8/16	8

All values are shown in µg/mL. CEP = cephalothin; CLX = cephalexin; n = number of isolates; NA = data not available.

Figure 1.

Cephalexin and cephalothin Staphylococcus pseudintermedius minimum inhibitory concentration (MIC) population distributions plotted with corresponding penicillin-binding protein 2a result (n = 1,112). These results correspond to data in Supplementary Tables 2 and 3. Red dashed lines represent breakpoints for susceptible (≤2 µg/mL) and resistant (≥8 µg/mL).

The listing of the yearly MIC frequency distributions for each cephalosporin tested show that there were no detectable shifts in the S. pseudintermedius population over time (2011–2014) for any of the 4 cephalosporins (Supplementary Table 1). The overall 2011–2014 MIC parameters (mode, MIC₅₀, MIC₇₀, MIC₈₀, and MIC₉₀) for the 2 third-generation cephalosporins (cefovecin and cefpodoxime) are nearly identical (Supplemental Table 1). The most striking difference in MIC parameters for the first-generation cephalosporins (cephalothin and cephalexin), regardless of year of collection, was a difference of 4 doubling dilutions, with cephalexin MICs being higher (Table 3, Supplementary Table 1, Fig. 1).

In addition to the large differences for testing S. pseudintermedius, there are also important differences in MIC between cephalothin and cephalexin for testing S. aureus and P. multocida (Table 3, Supplementary Table 3). Smaller differences, or identical results, were observed for testing E. coli, P. mirabilis, and β-hemolytic Streptococcus species (Table 3, Supplementary Table 3). Thus, for some bacteria isolated from dogs and cats, there are wide discrepancies in MIC values between the 2 first-generation cephalosporins, cephalexin and cephalothin.

PK-PD analysis and Monte Carlo simulations

Monte Carlo simulations generated a PTA table (Table 4). A PTA of 90% certainty (probability) is accepted to indicate likelihood of attaining the PK-PD target and achieving therapeutic success.¹³,²⁶ This PTA was achieved at a MIC of ≤2 µg/mL. The PTA for a MIC of 4 and 8 µg/mL was 73% and 47%, respectively.

Table 4.

Probability of target attainment (PTA) for administration of cephalexin oral to dogs using Monte Carlo simulations.

Drug and dose regimen	PTA from Monte Carlo simulation for the indicated MIC values
	0.25	0.5	1	2	4	8	16	32	64
Cephalexin in dogs (25 mg/kg oral) every 12 h	98.7%	97.93%	94.8%	89.6%	72.9%	47.05%	14.03%	1.74%	0%

Value listed for each minimum inhibitory concentration (MIC; µg/mL) is the probability (certainty) that a target of 50% time above MIC (time > MIC) can be achieved.

Discussion

CLSI has recommended the results of cephalothin susceptibility tests to predict in vitro susceptibility to the other first-generation cephalosporins, such as cephapirin, cephalexin, and cefadroxil.⁹ This recommendation will be replaced in a new document scheduled for publication in 2018 based on our analysis presented herein. Our study provides the basis for changes in susceptibility testing standards (QC ranges) and clinical breakpoints for cephalexin for dogs. It is appropriate to replace cephalothin with cephalexin in the tables because cephalexin is the oral cephalosporin administered most commonly to dogs.

All 3 CLSI criteria that are needed for determining a clinical breakpoint¹⁰ were met. Sufficient evidence was presented to indicate that the clinically accepted oral cephalexin dose of 25–30 mg/kg twice daily to dogs is effective. Monte Carlo simulations and calculation of the PTA showed that breakpoints of ≤2 µg/mL (susceptible), 4 µg/mL (intermediate), and ≥8 µg/mL (resistant) agree with MIC distribution for S. pseudintermedius isolates from dogs. Examination of previously published data spanning several years of collection shows that the distributions reported herein agree with earlier established patterns of MIC wild-type distributions.⁴,¹⁵,¹⁹,²⁴ Most wild-type S. pseudintermedius strains in our study had cephalexin MIC values ≤2 µg/mL. This breakpoint can also be applied to Streptococcus spp. isolates from dogs (with most isolates susceptible) and E. coli (with most isolates testing resistant) when testing pathogens from dogs.

In addition to establishing these new standards to be published in future CLSI documents, we have revealed problems with relying on the previous practice of using cephalothin for testing. For some bacteria—particularly S. pseudintermedius, the most common staphylococcal isolate from dogs—the MIC value is much higher for cephalexin than for cephalothin. It is important for laboratories to recognize this finding and use cephalexin for clinical testing isolates from dogs instead of cephalothin. Discrepancies were also observed in another study,¹⁶ but in a different direction; that study reported high major error and minor error rates between cephalexin and cephalothin for testing human E. coli urinary tract isolates. As well, their data showed that MIC values for these E. coli were generally higher for cephalothin than cephalexin.¹⁶

We noted no substantial difference between cephalexin and cephalothin for PBP2a-negative isolates; however, comparison of cephalothin and cephalexin MICs for PBP2a-positive isolates looked quite different. Although neither cephalexin nor cephalothin are typically used to detect methicillin resistance in staphylococci,² cephalexin MICs correlate well with the presence of PBP2a among these isolates of S. pseudintermedius, whereas cephalothin MICs do not. Although the majority of cephalothin MIC values were below the breakpoint for PBP2a-positive isolates, there are no data available regarding clinical efficacy of cephalothin for treating PBP2a-positive S. pseudintermedius infections in dogs. According to CLSI standards²,⁸ if an isolate is resistant to oxacillin, it should be reported as resistant to all β-lactam antibiotics, including cephalexin and cephalothin, regardless of the MIC value.

Based on our analysis, and because cephalexin is used clinically in dogs instead of cephalothin, the CLSI committee unanimously agreed to replace cephalothin with cephalexin for testing canine isolates. The statement that “cephalothin should be tested as the class representative for first-generation cephalosporins” will be removed from future VET01 documents and VET08 supplements. Cefazolin can be tested separately because it has different activity and its own breakpoint already listed.⁹ Despite new FDA approvals of cefpodoxime (2004), cefovecin (2008), and cephalexin (2012) for treatment of dogs in the United States, the data from our study did not reveal any increased trends in resistance during a 4-y collection period.