• View in gallery

    Agarose gel revealing PCR products obtained by use of the Salmonicida subspecies-specific MIY primer set (512-bp product) for typical isolates only (lanes 2, 4, 6, 8, 10, 12, 14, 16, and 18) and the Salmonicida species-specific AP primer set (421-bp product) for typical and atypical isolates (lanes 3, 5, 7, 9, 11, 13, 15, 17, and 19). Lanes were as follows: 1 and 20, 1-kilobase DNA ladder; 2 and 3, Aeromonas salmonicida subsp salmonicida ATCC 33658; 4 and 5, A salmonicida subsp masoucida ATCC 27013; 6 and 7, A salmonicida subsp achromogenes ATCC 33659; 8 and 9, A salmonicida subsp smithia ATCC 49393; 10 and 11, Maine91 (typical); 12 and 13, 4059 (atypical); 14 and 15, A caviae ATCC 15468; 16 and 17, A veronii ATCC 9071; and 18 and 19, negative control samples. Values on the left represent molecular size in number of bp. Notice that the MIY primer set did not generate a band at 421 bp in lanes 4, 6, 8, and 12, which is as expected for atypical A salmonicida isolates.

  • View in gallery

    Frequency distribution for MICs and diameters of the zone of inhibition for isolates of A salmonicida when tested against oxytetracycline (30 μg; A), ormetoprim-sulfadimethoxine (1.25 and 23.75 μg, respectively; B), oxolinic acid (2 μg; C) and florfenicol (30 μg; D). Epidemiologic cutoff values are indicated for MICs (horizontal dashed lines) and diameters of the zones of inhibition (vertical dashed lines) for each antimicrobial. Notice that there are 2 clusters of isolates for oxytetracycline, ormetoprim-sulfadimethoxine, and oxolinic acid but only 1 cluster of isolates for florfenicol.

  • 1.

    Floyd RF. Incorporating pet fish into your small animal practice. Available at: edis.ifas.ufl.edu/VM108. Accessed Feb 2, 2006.

  • 2.

    Kuehn BM. Veterinarians test the waters of fish medicine. J Am Vet Med Assoc 2002;221:16711672.

  • 3.

    FDA Center for Veterinary Medicine Web site. Animal drugs for minor uses and minor species. Available at: www.fda.gov/cvm/minortoc.htm. Accessed Feb 2, 2006.

    • Search Google Scholar
    • Export Citation
  • 4.

    Miller RA, Walker RD & Baya A, et al. Antimicrobial susceptibility testing of aquatic bacteria: quality control disk diffusion ranges for Escherichia coli ATCC 25922 and Aeromonas salmonicida subsp. salmonicida ATCC 33658 at 22 and 28 degrees C. J Clin Microbiol 2003;41:43184323.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Miller RA, Walker RD & Carson J, et al. Standardization of a broth microdilution susceptibility testing method to determine minimum inhibitory concentrations of aquatic bacteria. Dis Aquat Organ 2005;64:211222.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Clinical and Laboratory Standards Institute. Methods for antimicrobial disk susceptibility testing of bacteria isolated from aquatic animals; proposed guideline M42-P. Wayne, Pa: Clinical and Laboratory Standards Institute, 2005.

    • Search Google Scholar
    • Export Citation
  • 7.

    Clinical and Laboratory Standards Institute. Methods for broth dilution susceptibility testing of bacteria isolated from aquatic animals; proposed guideline M49-P. Wayne, Pa: Clinical and Laboratory Standards Institute, 2005.

    • Search Google Scholar
    • Export Citation
  • 8.

    Kahlmeter G, Brown D. Harmonization of antimicrobial breakpoints in Europe—can it be achieved? Clin Microbiol Newsl 2004;26:187192.

  • 9.

    European Society of Clinical Microbiology and Infectious Diseases Web site. Clinical breakpoints and epidemiological cut-off values. Available at: www.escmid.org/sites/index_f.aspx?par=2.4. Accessed Mar 8, 2006.

    • Search Google Scholar
    • Export Citation
  • 10.

    Miyata M, Inglis V, Aoki T. Rapid identification of Aeromonas salmonicida subspecies salmonicida by the polymerase chain reaction. Aquaculture 1996;141:1324.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Gustafson CE, Thomas CJ, Trust TJ. Detection of Aeromonas salmonicida from fish by using polymerase chain reaction amplification of the virulence surface array protein gene. Appl Environ Microbiol 1992;58:38163825.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Clinical and Laboratory Standards Institute. Development of in vitro susceptibility testing criteria and quality control parameters; approved standard M23-A2. Wayne, Pa: Clinical and Laboratory Standards Institute, 2001.

    • Search Google Scholar
    • Export Citation
  • 13.

    Clinical and Laboratory Standards Institute. Development of in vitro susceptibility testing criteria and quality control parameters for veterinary antimicrobial agents; approved guideline M37-A2. Wayne, Pa: Clinical and Laboratory Standards Institute, 2002.

    • Search Google Scholar
    • Export Citation
  • 14.

    Metzler DM, DeHaan RM. Susceptibility tests of anaerobic bacteria: statistical and clinical considerations. J Infect Dis 1974;130:588594.

  • 15.

    Amsterdam D. Susceptibility testing of antimicrobials in liquid media. In: Lorian V, ed. Antibiotics in laboratory medicine. 4th ed. Baltimore: The Williams & Wilkins Co, 1996;52111.

    • Search Google Scholar
    • Export Citation
  • 16.

    Tsoumas A, Alderman DJ, Rodgers CJ. Aeromonas salmonicida: development of resistance to 4-quinolone antimicrobials. J Fish Dis 1989;12:493507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Smith P, Hiney M. Towards setting breakpoints for oxolinic acid susceptibility of Aeromonas salmonicida using distribution of data generated by standard test protocols. Aquaculture 2005;250:2226.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Reimschuessel R, Stewart L & Squibb E, et al. Fish drug analysis—Phish-Pharm: a searchable database of pharmacokinetics data in fish. AASPJ 2005;7:E288E327.

    • Search Google Scholar
    • Export Citation

Advertisement

Epidemiologic cutoff values for antimicrobial agents against Aeromonas salmonicida isolates determined by frequency distributions of minimal inhibitory concentration and diameter of zone of inhibition data

View More View Less
  • 1 FDA, Center for Veterinary Medicine, Office of Research, Division of Animal Research, 8401 Muirkirk Rd, Laurel, MD 20708.
  • | 2 FDA, Center for Veterinary Medicine, Office of Research, Division of Animal Research, 8401 Muirkirk Rd, Laurel, MD 20708.

Abstract

Objective—To develop epidemiologic cutoff values by use of frequency distributions for susceptibility to 4 antimicrobial agents when tested against a representative population of a major aquaculture pathogen, Aeromonas salmonicida.

Sample Population—217 typical and atypical A salmonicida isolates obtained from 20 states and 12 countries.

Procedures—Species identification of A salmonicida isolates was confirmed by detection of specific nucleotide sequences by use of a PCR assay. Minimal inhibitory concentration (MIC) and diameter of the zone of inhibition for oxytetracycline, ormetoprim-sulfadimethoxine, oxolinic acid, and florfenicol were determined for each isolate in accordance with standardized antimicrobial susceptibility testing methods that have been approved by the Clinical and Laboratory Standards Institute for bacterial isolates from aquatic animals. Susceptibility data were tabulated in a scattergram and analyzed by use of error rate bounding.

Results—Susceptibility tests for oxytetracycline, ormetoprim-sulfadimethoxine, and oxolinic acid revealed 2 distinct populations of bacteria. Isolates tested against florfenicol clustered into a single population. Oxolinic acid susceptibility data revealed higher MICs in the non–United States A salmonicida isolates. Slow-growing (atypical) A salmonicida isolates were generally more susceptible than typical isolates for all antimicrobials, except oxolinic acid.

Conclusions and Clinical Relevance—Use of frequency distributions of susceptibility results to develop epidemiologic cutoff values appears to be applicable to aquatic isolates. Frequency distributions of susceptibility results for A salmonicida revealed clear divisions between isolate susceptibilities. This type of data, considered in conjunction with pharmacokinetic and efficacy data, may be useful for developing clinical breakpoints for use in aquaculture.

Abstract

Objective—To develop epidemiologic cutoff values by use of frequency distributions for susceptibility to 4 antimicrobial agents when tested against a representative population of a major aquaculture pathogen, Aeromonas salmonicida.

Sample Population—217 typical and atypical A salmonicida isolates obtained from 20 states and 12 countries.

Procedures—Species identification of A salmonicida isolates was confirmed by detection of specific nucleotide sequences by use of a PCR assay. Minimal inhibitory concentration (MIC) and diameter of the zone of inhibition for oxytetracycline, ormetoprim-sulfadimethoxine, oxolinic acid, and florfenicol were determined for each isolate in accordance with standardized antimicrobial susceptibility testing methods that have been approved by the Clinical and Laboratory Standards Institute for bacterial isolates from aquatic animals. Susceptibility data were tabulated in a scattergram and analyzed by use of error rate bounding.

Results—Susceptibility tests for oxytetracycline, ormetoprim-sulfadimethoxine, and oxolinic acid revealed 2 distinct populations of bacteria. Isolates tested against florfenicol clustered into a single population. Oxolinic acid susceptibility data revealed higher MICs in the non–United States A salmonicida isolates. Slow-growing (atypical) A salmonicida isolates were generally more susceptible than typical isolates for all antimicrobials, except oxolinic acid.

Conclusions and Clinical Relevance—Use of frequency distributions of susceptibility results to develop epidemiologic cutoff values appears to be applicable to aquatic isolates. Frequency distributions of susceptibility results for A salmonicida revealed clear divisions between isolate susceptibilities. This type of data, considered in conjunction with pharmacokinetic and efficacy data, may be useful for developing clinical breakpoints for use in aquaculture.

Veterinarians are expanding their practices to include exotic species, including fish.1,2 In the United States, only a few antimicrobial agents (including ormetoprim-sulfadimethoxine, oxytetracycline, and florfenicol) are approved for use in fish farmed for food production. Legislation such as the Minor Use and Minor Species Animal Health Act of 2004 is fostering the availability of additional therapeutic agents for use in fish.3 Before such drugs can be used, it is important for clinicians treating fish to become familiar with aquatic bacterial diseases and the susceptibility of those pathogens to various antimicrobial agents. Although, to our knowledge, clinical breakpoints or interpretive criteria (susceptible, intermediate, and resistant) have not been developed for any aquatic pathogen in any aquatic animal species, standardized AST methods for aquatic isolates4,5 should improve a clinician's ability to choose an appropriate antimicrobial agent. Historically, veterinarians and researchers of aquatic diseases have used laboratory-specific clinical breakpoints. These values have had limited application or reliability outside of the regions in which they were generated. These limitations can be attributed to variations among in vitro testing procedures, limited diversity of isolates, and unique environmental conditions that may have affected therapeutic efficacy. Efforts to enhance the probability of therapeutic success when relying on AST results are dependent on interpretive criteria that are as specific as possible for a given bacterial pathogen in a given animal species. The reliability of such interpretive criteria is enhanced when standardized AST methods, such as those published by the CLSI, are used.6,7 Two CLSI guidance documents, M426 and M49,7 provide standardized test conditions for nonfastidious aquatic bacterial isolates and provide details on methods for quality control and quality assurance.

Frequency distributions of MICs can be used to delineate epidemiologic cutoff values (also known as species-specific microbiologic breakpoints), as defined by the European Committee on AST.8,9 These cutoff values can be used to discriminate wild-type (ie, originally susceptible bacterial populations) from non–wild-type (ie, populations with acquired and mutational resistance mechanisms) isolates. These cutoff values are not to be confused with clinical breakpoints, which are used primarily for predicting clinical outcomes.

The purpose of the study reported here was to develop epidemiologic cutoff values by use of frequency distributions of MICs and diameters of zones of inhibition for 217 typical and atypical (slow growing) isolates of Aeromonas salmonicida (causative agents of furunculosis, goldfish ulcer disease, and carp erythrodermatitis) against 3 FDA-approved antimicrobials and 1 antimicrobial commonly used in some European countries. These distributions may be useful in developing clinical breakpoints when combined with data from pharmacokinetic-pharmacodynamic studies in targeted fish species and, if possible, clinical outcome data from fish with furunculosis or outbreaks of associated disease.

Materials and Methods

Sample population—Isolates of A salmonicida were obtained from 16 contributors located in various countries (8 in the United States; 2 in Israel; and 1 each in Canada, the United Kingdom, Switzerland, Spain, Norway, and Finland). Contributors were contacted by the authors and requested to provide typical and atypical A salmonicida isolates from a wide geographic region that included clinical and wild-type strains representing a wide range of susceptibilities.

A total of 217 A salmonicida isolates were used for AST, including 112 isolates from the United States, representing 20 states; 99 isolates from 11 other countries; and 6 isolates from an unknown origin. Strains were originally isolated from 28 fish species. The year of original isolation for the isolates ranged from 1955 to 2004 (median year of original isolation, 1995).

All isolates were stored in tryptic soy broth with 20% glycerol at −80°C and then cultured on tryptic soy agar supplemented with 5% sheep blood at 22°C for 48 hours. After culture, cells were harvested for DNA extraction and AST.

Six isolates (1 from the National Collections of Industrial, Food and Marine Bacteriaa and 5 from the ATCCb-f) served as control isolates in the PCR assays. These isolates were not included in the sample population used for AST.

PCR assay—Genomic DNA was extracted from all bacterial strains by use of a commercially available kit.g Extraction was conducted in accordance with the manufacturer's instructions.

The PCR assays were performed in 0.2-mL thin-walled PCR tubes in a thermal cycler.h Genomic DNA from A salmonicida subsp salmonicida ATCC 33658 was used as a positive control sample for each of the 2 PCR assays (MIY and AP). Nuclease-free water was used as a negative control sample. Template DNA (10 to 100 ng) was added for each reaction, and a 1-kilobase DNA ladderi was used. Products were separated by use of electrophoresis on 1.5% agarose gels and developed with ethidium bromide staining and UV illumination in a gel documentation system.j

The MIY primer set, which is specific for only typical strains of A salmonicida subsp salmonicida,10 was used. The MIY primer set comprises MIY1 (5′–AGCCTCCACGCGCTCACAGC–3′) and MIY2 (5′–AAGAGGCCCCATAGTGTGGG–3′). Each reaction (volume, 25 μL) contained 0.6 units of Taq DNA polymerase, 2.5 μLof10X PCR buffer, 1.5mM MgCl2, 16 pmol of each amplification primer (ie, MIY1 and MIY2), and 0.2mM of each of the 4 deoxynucleotide triphosphates.k-o Reaction mixtures were maintained at 94°C for 2 minutes and amplified for 35 cycles with denaturation at 94°C for 30 seconds, annealing at 68°C for 90 seconds, and elongation at 68°C for 90 seconds. A final extension was performed at 68°C for 3 minutes. Expected size of the PCR product was 512 bp.

The AP primer set, which is specific for all strains of A salmonicida,11 was used. The primer set comprised AP1 (5′–GGCTGATCTCTTCATCCTCACCC–3′) and AP2 (5′–CAGAGTGAAATCTACCAGCGGTGC–3′). Each reaction (volume, 25 μL) contained 0.25 units of Taq DNA polymerase, 2.5 μL of 10X PCR buffer, 2.5mM MgCl2, 8 pmol of each amplification primer (ie, AP1 and AP2), and 0.2mM of each of the 4 deoxynucleotide triphosphates.k-o Reaction mixtures were maintained at 94°C for 2 minutes and amplified for 30 cycles with denaturation at 94°C for 15 seconds, annealing at 57°C for 30 seconds, and elongation at 72°C for 90 seconds. A final extension was performed at 72°C for 3 minutes. Expected size of the PCR product was 421 bp.

Disk diffusion testing— Disk diffusion tests were conducted in accordance with CLSI guidelines.6Escherichia coli ATCC 25922 or A salmonicida subsp salmonicida ATCC 33658, or both, were used as quality-control isolates. All tests were conducted on Mueller-Hinton agar,p with incubation at 22°C for 44 to 48 hours. Disks containing florfenicolq (30 μg), oxolinic acidr (2 μg), oxytetracyclines (30 μg), and ormetoprim-sulfadimethoxinet (1.25 and 23.75 μg of ormetoprim and sulfadimethoxine, respectively) were used. Diameters of the zones of inhibition were measured with a ruler and rounded to the nearest millimeter. Bacterial inocula were standardized and monitored for cell densities in the range of 1 × 108 CFUs/mLto 2 × 108 CFUs/mL.

MIC testing— Broth microdilution tests were conducted in accordance with CLSI guidelines.7Escherichia coli ATCC 25922 or A salmonicida subsp salmonicida ATCC 33658, or both, were used as quality-control organisms. All tests were conducted in 96-well platesu; plates were incubated at 22°C for 44 to 48 hours. Plates contained dehydrated antimicrobial agent in each well and were formatted in 2 identical series of twelve 2-fold dilutions for florfenicol (32 to 0.015 μg/mL), oxolinic acid (4 to 0.002 μg/mL), and oxytetracycline (32 to 0.015 μg/mL) and eleven 2-fold dilutions for ormetoprim-sulfadimethoxine (8/152 to 0.008/0.15 μg/mL). Two wells were used as positive control wells. An autoinoculator unitv was used to place 100 μLof standardized inoculum prepared in cation-adjusted Mueller-Hinton brothw into each well. Bacterial inocula were standardized and monitored for cell densities of approximately 5 × 105 CFUs/mL.

Scattergram analysis—The MIC and corresponding diameter of the zone of inhibition for each isolate were tabulated to generate a frequency distribution for each antimicrobial agent in the form of a scattergram.x As recommended by the CLSI,12,13 an error rate bounding method initially described elsewhere14 was modified to calculate discrepancy rates on the basis of MICs and diameters of zones of inhibition for all A salmonicida isolates, typical A salmonicida isolates, and atypical A salmonicida isolates. Discrepancy rates were calculated for use in selecting epidemiologic cutoff values for the diameters of the zones of inhibition. The MIC50 and MIC90 values were also calculated for all isolates, isolates from the United States, and isolates from other countries.

Results

Analysis of PCR results obtained by use of AP (Salmonicida species specific for typical or atypical isolates) and MIY (Salmonicida subspecies specific for typical isolates) primer sets revealed a pool of isolates consisting of 163 typical and 54 atypical A salmonicida isolates; these results did not include the 6 reference isolates. Of the 163 typical isolates, 110 were from the United States, 49 were from other countries, and 4 were from an unknown origin. Of the 54 atypical isolates, 2 were from the United States, 50 were from other countries, and 2 were from an unknown origin.

Assay of a subset of the population revealed species-specific AP primers yielded positive results for only A salmonicida isolates (Figure 1). One atypical isolate (A salmonicida subsp pectinolytica) yielded negative results in PCR assays for both primers (data not shown). As expected, all atypical A salmonicida isolates yielded negative results in the PCR assay for the MIY primer set. Some background banding was observed, but PCR products with intensely positive results made identification by use of PCR assays unambiguous.

Figure 1—
Figure 1—

Agarose gel revealing PCR products obtained by use of the Salmonicida subspecies-specific MIY primer set (512-bp product) for typical isolates only (lanes 2, 4, 6, 8, 10, 12, 14, 16, and 18) and the Salmonicida species-specific AP primer set (421-bp product) for typical and atypical isolates (lanes 3, 5, 7, 9, 11, 13, 15, 17, and 19). Lanes were as follows: 1 and 20, 1-kilobase DNA ladder; 2 and 3, Aeromonas salmonicida subsp salmonicida ATCC 33658; 4 and 5, A salmonicida subsp masoucida ATCC 27013; 6 and 7, A salmonicida subsp achromogenes ATCC 33659; 8 and 9, A salmonicida subsp smithia ATCC 49393; 10 and 11, Maine91 (typical); 12 and 13, 4059 (atypical); 14 and 15, A caviae ATCC 15468; 16 and 17, A veronii ATCC 9071; and 18 and 19, negative control samples. Values on the left represent molecular size in number of bp. Notice that the MIY primer set did not generate a band at 421 bp in lanes 4, 6, 8, and 12, which is as expected for atypical A salmonicida isolates.

Citation: American Journal of Veterinary Research 67, 11; 10.2460/ajvr.67.11.1837

On the basis of evaluation of scattergrams that contained plots of the MICs versus the diameters of the zones of inhibition for oxytetracycline, ormetoprim-sulfadimethoxine, and oxolinic acid, 2 clearly discernible populations of isolates were observed (wild type [susceptible to antimicrobials; no resistance mechanisms] and non–wild type [acquired and mutational resistance mechanisms]; Figure 2). A wide range of diameters of the zones of inhibition for oxytetracycline (13 to 30 mm) was observed between the 2 populations. Similar separation was evident for ormetoprim-sulfadimethoxine (7 to 19 mm) and oxolinic acid (22 to 34 mm). Distribution of the plotted points for florfenicol revealed a single wild-type population with all isolates having MICs ≤ 2 μg/mL and zone diameters ≥ 34 mm.

Figure 2—
Figure 2—

Frequency distribution for MICs and diameters of the zone of inhibition for isolates of A salmonicida when tested against oxytetracycline (30 μg; A), ormetoprim-sulfadimethoxine (1.25 and 23.75 μg, respectively; B), oxolinic acid (2 μg; C) and florfenicol (30 μg; D). Epidemiologic cutoff values are indicated for MICs (horizontal dashed lines) and diameters of the zones of inhibition (vertical dashed lines) for each antimicrobial. Notice that there are 2 clusters of isolates for oxytetracycline, ormetoprim-sulfadimethoxine, and oxolinic acid but only 1 cluster of isolates for florfenicol.

Citation: American Journal of Veterinary Research 67, 11; 10.2460/ajvr.67.11.1837

Discrepancy rates and error rate bounding were used as recommended by the CLSI12,13 to determine epidemiologic cutoff values (Tables 1 and 2). The epidemiologic cutoff values were adjusted until the number of false wild-type results on disk diffusion tests (very major discrepancies; type I errors) and false non–wild-type results (major discrepancies; type II errors) were held to a minimum. As specified by the CLSI13 for collections of clinical isolates, all rates for major and very major discrepancies were held at < 1.5% and < 3%, respectively. Minor discrepancies (ie, when 1 test result was classified as intermediate and the other was wild type or non–wild type) were also considered in the calculations.

Table 1—

Discrepancy between MICs and diameters of the zones of inhibition for Aeromonas salmoni-cida when tested against various antimicrobials.

AntimicrobialIsolatesMIC rangeNo.Discrepancy*,  
Very majorMajorMinor
OxytetracyclineAll isolates≥ Ihigh + 22170NA0
Ihigh + 1 to Ilow − 1217001 (< 0.01)
≤ Ilow − 2217NA01 (< 0.01)
Typical isolates≥ Ihigh + 21630NA0
Ihigh + 1 to Ilow − 1163001 (0.01)
≤ Ilow − 2163NA01 (0.01)
Atypical isolates≥ Ihigh + 2540NA0
Ihigh + 1 to Ilow − 154000
≤ Ilow − 254NA00
Ormetoprim-sulfadimethoxineAll isolates≥ I + 22171 (< 0.01)NA 
I + 1 to I − 12171 (< 0.01)03 (0.01)
≤ I − 2217NA03 (0.01)
Typical isolates≥ I + 21631 (0.01)NA3 (0.02)
I + 1 to I − 11631 (0.01)03 (0.02)
≤ I − 2163NA04 (0.02)
Atypical isolates≥ I + 2540NA0
I + 1 to I − 154000
≤ I − 254NA00
Oxolinic acidAll isolates≥ Ihigh + 22170NA0
Ihigh + 1 to Ilow −1217000
≤ Ilow − 2217NA00
Typical isolates≥ Ihigh + 21630NA0
Ihigh + 1 to Ilow −1163000
≤ Ilow − 2163NA00
Atypical isolates≥ Ihigh + 2540NA0
Ihigh + 1 to Ilow −154000
≤ Ilow − 254NA00
FlorfenicolAll isolates≥ NWT + 12170NANA
NWT + WT21700NA
≤ WT − 1217NA0NA
Typical isolates≥ NWT + 11630NANA
NWT + WT16300NA
≤ WT − 1163NA0NA
Atypical isolates≥ NWT + 1540NANA
NWT + WT5400NA
≤ WT − 154NA0NA

Very major discrepancies represent the number of false wild-type (WT) results on disk diffusion tests (type I errors), major discrepancies represent the number of false non–wild-type (NWT) results on disk diffusion tests (type II errors), and minor discrepancies represent when 1 test result was classified as intermediate and the other was WT or NWT.

Values reported are number (%).

Ihigh + 1 and Ihigh + 2 represent 1 and 2 dilutions above the highest MIC within the intermediate range, respectively, and Ilow − 1 and Ilow − 2 represent1 and 2 dilutions below the lowest MIC within the intermediate range, respectively. The I + 1 and I + 2 represent 1 and 2 dilutions above the intermediate MIC value, respectively, and I − 1 and I − 2 represent 1 and 2 dilutions below the intermediate MIC value, respectively. The NWT represents results for the population of isolates with acquired and mutational resistance mechanisms. The WT represents results for the population of isolates susceptible to antimicrobials (no resistance mechanisms). The MIC range for florfenicol was defined such that NWT + 1, NWT + WT, and WT − 1 represent 1 dilution above the NWT cutoff value, the NWT cutoff value and WT cutoff value, and 1 dilution below the WT cutoff value, respectively.

NA = Not applicable.

Table 2—

Epidemiologic cutoff values for diameters of the zones of inhibition and MICs for all A salmonicida isolates when tested against various antimicrobials.

AntimicrobialDiameter of zone of inhibition (mm)MIC (μg/mL) Intermediate  
IntermediateIntermediate  
WTrangeNWTWTrangeNWT
Oxytetracycline (30 μg)≥ 2824–27≥ 23≤ 12–4≥ 8
Ormetoprim-sulfadimethoxine (1.25 and 23.75 μg)*≥ 2017–19≤ 16≤ 0.5/9.5*1/19*≥2/38*
Oxolinic acid (2 μg)≥ 3026–29≤ 25≤ 0.120.25–0.5≥ 1
Florfenicol (30 μg)≥ 31NA≤ 30≤ 4NA≥ 8

Values reported are for ormetoprim and sulfadimethoxine, respectively.

See Table 1 for remainder of key.

Analysis of MIC50 and MIC90 values calculated for all isolates, isolates from the United States, and isolates from other countries revealed a pattern only for oxolinic acid (Table 3). Isolates from the United States had considerably lower MICs for oxolinic acid, compared with the MICs for isolates from other countries. Isolates from the United States had slightly higher MICs for oxytetracycline, compared with the MICs for isolates from other countries (Table 4). Isolates from other countries had slightly higher MICs for ormetoprim-sulfadimethoxine, compared with the MICs for isolates from the United States (Table 5). On the basis of geographic origin of the isolates, no difference was observed with regard to MICs for florfenicol (Table 6).

Table 3—

Cumulative percentage of MICs for 217 isolates of A salmonicida (112 isolates obtained from the United States, 99 isolates obtained from other countries, and 6 isolates obtained from an unknown origin) when tested against oxolinic acid.

MIC (μg/mL)All isolatesUnited StatesOther countries
> 4100100100
498.610097.1
295.910091.2*
194.0*10087.3
0.589.910078.4
0.2589.910078.4
0.1289.910078.4
0.0688.098.376.5
0.0382.093.9*,68.6
0.0159.78.710.8
0.0080.00.00.0
0.0040.00.00.0
0.0020.00.00.0
≤ 0.0020.00.00.0

Represents the MIC90.

Represents the MIC50.

Table 4—

Cumulative percentage of MICs for 217 isolates of A salmonicida (112 isolates obtained from the United States, 99 isolates obtained from other countries, and 6 isolates obtained from an unknown origin) when tested against oxytetracycline.

MIC (μg/mL)All isolatesUnited StatesOther countries
> 32100100*100
3292.2*89.695.1*
1681.175.787.3
871.067.075.5
469.163.575.5
269.163.575.5
168.763.574.5
0.568.262.6†74.5
0.2550.7†48.752.9†
0.128.810.46.9
0.060.00.00.0
0.030.00.00.0
0.0150.00.00.0
≤ 0.0150.00.00.0

See Table 3 for remainder of key.

Table 5—

Cumulative percentage of MICs for 217 isolates of A salmonicida (112 isolates obtained from the United States, 99 isolates obtained from other countries, and 6 isolates obtained from an unknown origin) when tested against ormetoprim-sulfadimethoxine.

MIC (μg/mL)*All isolatesUnited StatesOther countries
> 8/152100100100
8/15294.096.591.2
4/7694.096.591.2
2/3894.096.591.2
1/1992.694.890.2
0.5/9.590.393.986.3
0.25/4.886.688.784.3
0.12/2.474.776.572.5
0.06/1.211.16.116.7
0.03/0.60.50.01.0
0.015/0.30.00.00.0
0.008/0.0150.00.00.0
≤ 0.008/0.0150.00.00.0

Values represent concentrations for ormetoprim and sulfadimethoxine, respectively.

Represents the MIC90.

Represents the MIC50.

Table 6—

Cumulative percentage of MICs for 217 isolates of A salmonicida (112 isolates obtained from the United States, 99 isolates obtained from other countries, and 6 isolates obtained from an unknown origin) when tested against florfenicol.

MIC (μg/mL)All isolatesUnited StatesOther countries
> 32100100100
32100100100
16100100100
8100100100
4100100100
2100100100
198.2*98.3*98.0*
0.578.8†72.2†86.3†
0.2523.515.732.4
0.123.70.96.9
0.060.00.00.0
0.030.00.00.0
0.0150.00.00.0
≤ 0.0150.00.00.0

See Table 3 for remainder of key.

Gross observations of values for MICs and diameters of the zones of inhibition revealed that typical A salmonicida isolates had slightly higher MICs for oxytetracycline, ormetoprim-sulfadimethoxine, and florfenicol than were evident for the atypical isolates (data not shown). Slower growth rate, characteristic of atypical A salmonicida isolates, and subsequent increased growth inhibition may help explain this increased susceptibility. In contrast, atypical isolates had noticeably higher MICs for oxolinic acid; however, most of these isolates were from countries in which oxolinic acid is approved for use.

Analysis of frequency distributions of susceptibility results for all 4 antimicrobial agents revealed a wider range of zones of inhibition for most MICs than the range of MICs at specific diameters of zones of inhibition. These noticeable variations may be explained by a decreased robustness of disk diffusion tests for slower growing (atypical A salmonicida) and fastidious organisms.15 However, the distinct separation of wild-type and non–wild-type isolates on the basis of diameters of the zones of inhibition alone should still provide accurate and useful epidemiologic cutoff values for isolates of this pathogen. Susceptibility data revealed that both disk diffusion and broth microdilution testing methods may be used to monitor for the development of antimicrobial resistance in A salmonicida isolates.

On the basis of the epidemiologic cutoff values developed in the study, 6 (2.7%) isolates were classified as non–wild type for oxytetracycline, ormetoprim-sulfadimethoxine, and oxolinic acid; 15 (6.7%) isolates were classified as non–wild type for 2 of these antimicrobials; and 56 (25.1%) isolates were classified as non–wild type for only 1 of these antimicrobials.

Discussion

To our knowledge, the study reported here represents the first large-scale study in which standardized AST methods were used to generate frequency distributions of MICs and diameters of the zones of inhibition for a disease-causing bacterium in aquaculture. As recommended by the CLSI for the development of interpretive criteria, more than 100 clinical and wild-type isolates relevant to the class of antimicrobial and representing multiple geographic locations were tested.

The study reported here relied on donors providing us with isolates from their own stocks; thus, it did not fully represent a random sample of A salmonicida isolates. Also it is possible some isolates used in this study may have been derived from the same bacterial clone. Clonality was not addressed in this study. Nevertheless, the large number and diversity of isolates in terms of location and species of origin should contribute to the credibility of these data.

General recommendations can be made on the basis of the distinct separation (or clustering in the case of florfenicol) of the test population with regard to susceptibility. These epidemiologic cutoff values for isolates of A salmonicida should not be considered in a clinical context because they are based solely on susceptibility distributions determined in vitro. These cutoff values can be used to detect the development of resistance.

Discrepancy between in vitro test results of susceptibility and therapeutic effectiveness is a result of the numerous factors that influence the interactions of antimicrobials and bacteria in vivo. To have clinical application, these cutoff values must subsequently be correlated (and adjusted when necessary) with serum kinetics of the antimicrobial agent when administered at therapeutic doses and, if possible, clinical outcome data. In the United States, such clinical breakpoints for antimicrobials used in humans have been determined by panels of experts who review large data sets. The data provided here should assist in efforts to determine clinical breakpoints for antimicrobials used in aquatic animal medicine.

In 1 study,16 investigators reported the frequency distribution of MICs for 70 isolates of A salmonicida against oxytetracycline and 5 other antimicrobial agents and suggested a susceptible breakpoint of ≤ 1 μg/mL for oxytetracycline. Data reported here reinforces this recommendation that an oxytetracycline cutoff value of 1 μg/mL clearly separates the wild-type from the non–wild-type population (ie, susceptible from resistant). In another study,17 investigators evaluated frequency distributions for MICs and diameters of the zones of inhibition for oxolinic acid against A salmonicida isolates and postulated classifying A salmonicida strains into 3 groups (susceptible, ≤ 0.0625 μg/mL; intermediate, 0.125 to 0.5 μg/mL; and resistant, ≥ 1 μg/mL). Those results are extremely similar to the findings of the study reported here.

When veterinarians are faced with a decision to treat a patient or population, oftentimes the only tools they possess are susceptibility data for the test isolate, recommendations from the supplier of the disk diffusion tests, clinical experience, and information extracted from published reports. Other important considerations are the pharmacokinetic and physiologic differences among species, overall health of the patient or population, and route of administration. Data sets collected by our laboratory group18 summarize the multitude of external factors that can alter the pharmacokinetics of many drugs in piscine patients. Some of these include route of administration, species, temperature, salinity, and disease state. Effects of such variables must also be considered when treating fish.

Clinical aquatic animal medicine is challenging because of a lack of available antimicrobial agents, minimal efficacy data in many cultured fish species, and little information regarding frequency distributions of susceptibility results. The study reported here was an attempt to provide clinicians with some of this much needed data. These data represent a valuable component in the development of interpretive criteria and should be useful as researchers and clinicians move closer to establishing true clinical breakpoints for a major aquatic pathogen, A salmonicida. Additional high-quality in vivo pharmacokinetic-pharmacodynamic and efficacy data will be required to allow clinicians and researchers to make comparisons and correlations with in vitro data on frequency distributions of susceptibility results reported here.

ABBREVIATIONS

AST

Antimicrobial susceptibility testing

CLSI

Clinical and Laboratory Standards Institute

MIC

Minimal inhibitory concentration

ATCC

American Type Culture Collection

MIC50

Concentration of antimicrobial agent required to inhibit 50% of the isolates

MIC90

Concentration of antimicrobial agent required to inhibit 90% of the isolates

a.

Aeromonas salmonicida subsp salmonicida NCIMB 13076, National Collections of Industrial, Food and Marine Bacteria, Aberdeen, UK.

b.

Aeromonas salmonicida subsp salmonicida ATCC 33658, American Type Culture Collection, Manassas, Va.

c.

Aeromonas salmonicida subsp salmonicida ATCC 14174, American Type Culture Collection, Manassas, Va.

d.

Aeromonas salmonicida subsp achromogenes ATCC 33659, American Type Culture Collection, Manassas, Va.

e.

Aeromonas salmonicida subsp masoucida ATCC 27013, American Type Culture Collection, Manassas, Va.

f.

Aeromonas salmonicida subsp smithia ATCC 49393, American Type Culture Collection, Manassas, Va.

g.

Generation capture column kit, Gentra Systems, Minneapolis, Minn.

h.

GeneAmp PCR system 9700, Applied Biosystems, Foster City, Calif.

i.

Ready-Load 1-kilobase DNA ladder, Invitrogen, Carlsbad, Calif.

j.

Bio-Rad Gel Doc 2000, Bio-Rad Laboratories, Hercules, Calif.

k.

Platinum Taq DNA polymerase, Invitrogen, Carlsbad, Calif.

l.

10X PCR buffer, Invitrogen, Carlsbad, Calif.

m.

50mM MgCl2, Invitrogen, Carlsbad, Calif.

n.

MIY and AP primer sets, Invitrogen, Carlsbad, Calif.

o.

10mM dNTPs, Applied Biosystems, Foster City, Calif.

p.

Mueller-Hinton agar, Difco, Sparks, Md.

q.

Florfenicol Sensi-discs, BD Diagnostic Systems, Sparks, Md.

r.

Oxolinic acid Sensi-discs, BD Diagnostic Systems, Sparks, Md.

s.

Oxytetracycline Sensi-discs, BD Diagnostic Systems, Sparks, Md.

t.

Ormetoprim-sulfadimethoxine Sensi-discs, BD Diagnostic Systems, Sparks, Md.

u.

Custom panel format CML1DFRM, Trek Diagnostic Systems, Cleveland, Ohio.

v.

Sensititre autoinoculator, Trek Diagnostic Systems, Cleveland, Ohio.

w.

Cation-adjusted Mueller-Hinton broth, Trek Diagnostic Systems, Cleveland, Ohio.

x.

Microsoft Excel 2003, Microsoft Corp, Redmond, Wash.

References

  • 1.

    Floyd RF. Incorporating pet fish into your small animal practice. Available at: edis.ifas.ufl.edu/VM108. Accessed Feb 2, 2006.

  • 2.

    Kuehn BM. Veterinarians test the waters of fish medicine. J Am Vet Med Assoc 2002;221:16711672.

  • 3.

    FDA Center for Veterinary Medicine Web site. Animal drugs for minor uses and minor species. Available at: www.fda.gov/cvm/minortoc.htm. Accessed Feb 2, 2006.

    • Search Google Scholar
    • Export Citation
  • 4.

    Miller RA, Walker RD & Baya A, et al. Antimicrobial susceptibility testing of aquatic bacteria: quality control disk diffusion ranges for Escherichia coli ATCC 25922 and Aeromonas salmonicida subsp. salmonicida ATCC 33658 at 22 and 28 degrees C. J Clin Microbiol 2003;41:43184323.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Miller RA, Walker RD & Carson J, et al. Standardization of a broth microdilution susceptibility testing method to determine minimum inhibitory concentrations of aquatic bacteria. Dis Aquat Organ 2005;64:211222.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Clinical and Laboratory Standards Institute. Methods for antimicrobial disk susceptibility testing of bacteria isolated from aquatic animals; proposed guideline M42-P. Wayne, Pa: Clinical and Laboratory Standards Institute, 2005.

    • Search Google Scholar
    • Export Citation
  • 7.

    Clinical and Laboratory Standards Institute. Methods for broth dilution susceptibility testing of bacteria isolated from aquatic animals; proposed guideline M49-P. Wayne, Pa: Clinical and Laboratory Standards Institute, 2005.

    • Search Google Scholar
    • Export Citation
  • 8.

    Kahlmeter G, Brown D. Harmonization of antimicrobial breakpoints in Europe—can it be achieved? Clin Microbiol Newsl 2004;26:187192.

  • 9.

    European Society of Clinical Microbiology and Infectious Diseases Web site. Clinical breakpoints and epidemiological cut-off values. Available at: www.escmid.org/sites/index_f.aspx?par=2.4. Accessed Mar 8, 2006.

    • Search Google Scholar
    • Export Citation
  • 10.

    Miyata M, Inglis V, Aoki T. Rapid identification of Aeromonas salmonicida subspecies salmonicida by the polymerase chain reaction. Aquaculture 1996;141:1324.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Gustafson CE, Thomas CJ, Trust TJ. Detection of Aeromonas salmonicida from fish by using polymerase chain reaction amplification of the virulence surface array protein gene. Appl Environ Microbiol 1992;58:38163825.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Clinical and Laboratory Standards Institute. Development of in vitro susceptibility testing criteria and quality control parameters; approved standard M23-A2. Wayne, Pa: Clinical and Laboratory Standards Institute, 2001.

    • Search Google Scholar
    • Export Citation
  • 13.

    Clinical and Laboratory Standards Institute. Development of in vitro susceptibility testing criteria and quality control parameters for veterinary antimicrobial agents; approved guideline M37-A2. Wayne, Pa: Clinical and Laboratory Standards Institute, 2002.

    • Search Google Scholar
    • Export Citation
  • 14.

    Metzler DM, DeHaan RM. Susceptibility tests of anaerobic bacteria: statistical and clinical considerations. J Infect Dis 1974;130:588594.

  • 15.

    Amsterdam D. Susceptibility testing of antimicrobials in liquid media. In: Lorian V, ed. Antibiotics in laboratory medicine. 4th ed. Baltimore: The Williams & Wilkins Co, 1996;52111.

    • Search Google Scholar
    • Export Citation
  • 16.

    Tsoumas A, Alderman DJ, Rodgers CJ. Aeromonas salmonicida: development of resistance to 4-quinolone antimicrobials. J Fish Dis 1989;12:493507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Smith P, Hiney M. Towards setting breakpoints for oxolinic acid susceptibility of Aeromonas salmonicida using distribution of data generated by standard test protocols. Aquaculture 2005;250:2226.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Reimschuessel R, Stewart L & Squibb E, et al. Fish drug analysis—Phish-Pharm: a searchable database of pharmacokinetics data in fish. AASPJ 2005;7:E288E327.

    • Search Google Scholar
    • Export Citation

Contributor Notes

Supported by the Oak Ridge Associated Universities through a grant sponsored by the FDA.

Presented in part at the National Foundation for Infectious Diseases Annual Conference on Antimicrobial Resistance, Bethesda, Md, June 2006, and at the FDA Science Forum, Washington, DC, April 2006.

The authors thank Drs. Ana Baya, Rocco Cipriano, Joy Evered, Hui-Min Hsu, Sharon Landin, Lindsay Oaks, Scott Lapatra, Jessica Boyd, David Bruno, Sarah Burr, Maria Figueras, Bjarnhei Gu∂ˇdmundsdóttir, Vera Lund, Sarah Maurice, Michael Sinyakov, and Satu Viljamaa-Dirks for provision of bacterial isolates.

Address correspondence to Mr. Miller.