Materials and Methods

Animals—In this prospective study, dogs examined by the clinical dermatology service at The Ohio State University College of Veterinary Medicine Teaching Hospital between August 2004 and August 2005 were evaluated for inclusion into the study if they had lesions consistent with superficial bacterial folliculitis.⁸ Inclusion criteria were the presence of ≥ 3 pustules on dermatologic examination and the presence of coccoid bacteria upon cytologic evaluation of a specimen collected from a lesion (papule, pustule, epidermal collarette, scale, or under a crust). Dogs were excluded if they had been treated with topical antimicrobial agents or had received systemic antimicrobial agents by any route within the 7 days preceding examination. Approval by the Institutional Review Committee on the Care and Use of Client-owned Animals at The Ohio State University was obtained prior to enrolling patients into this study. Informed consent and permission for sample collection were obtained from the owner. The age, breed, sex, and neuter status were recorded for each dog, and the history of prior antimicrobial use (topical and systemic) was obtained for each dog by interviewing owners and examining medical records.

Bacterial culture, morphology, and catalase and coagulase tests—Six sites, 3 pustules and 3 carriage sites (anus, nonlesional axillary skin, and nasal mucosa), were sampled by use of a dry sterile swab technique.² The 6 samples collected from each dog were transported to the microbiology laboratory within 1 hour and inoculated immediately onto culture plates. Standard methods for bacterial culture were used.²¹ Swab specimens were plated onto trypticase soy agar with 5% sheep blood^a and MacConkey medium^b and incubated at 35°C for 18 to 24 hours. A representative colony of each different staphylococcal morphology type was isolated.²² For each isolate, colony morphology and Gram stain characteristics were recorded. By use of standard methods, catalase production with 3% hydrogen peroxide and coagulase activity with rabbit plasma^c were determined.

Antimicrobial susceptibility testing—Antimicrobial susceptibility testing was performed^d for each staphylococcal isolate by use of the standardized disk diffusion (Kirby-Bauer) method, and isolates were identified and reported as susceptible, resistant, or intermediate in susceptibility to the antimicrobial agents tested on the basis of zone diameter interpretation criteria.^22-25 The following antimicrobial agents were evaluated: cefpodoxime, cephalothin, amoxicillin-clavulanic acid, chloramphenicol, trimethoprim-sulfamethoxazole, enrofloxacin, marbofloxacin, doxycycline, erythromycin, lincomycin hydrochloride, polymyxin B sulfate, and oxacillin. These agents were selected for this study, as they are commercially available, orally administered antimicrobial agents commonly used for the treatment of superficial bacterial folliculitis in dogs or agents used to differentiate among staphylococcal species. If an isolate was resistant or intermediate in susceptibility to oxacillin on the basis of disk diffusion testing, an OSA screening test was performed to confirm susceptibility or resistance to oxacillin in vitro.^26-28 For the purposes of this study, if an isolate was determined to be oxacillin resistant on the basis of results of OSA screening, the susceptibility of the coagulase-positive staphylococcal isolates to the β-lactam antimicrobial agents (cefpodoxime, cephalothin, and amoxicillin-clavulanic acid) was reported on the basis of the actual measured zone diameters on the disk diffusion test.

Each unique antimicrobial susceptibility phenotype identified during the study was assigned a pattern number, and one of these was assigned to each isolate on the basis of the results of the disk diffusion test. After antimicrobial susceptibility testing, isolates were transferred to brain-heart infusion medium^e and stored in an −80°C freezer until speciation and PFGE were performed.

Species identification—A commercial identification system^f was used to speciate each staphylococcal isolate as previously described.² Isolates that could not be speciated with this test were classified as coagulasepositive Staphylococcus spp on the basis of colony morphology and coagulase test results.

PFGE—Pulsed-field gel electrophoresis was performed^g on the DNA from all isolates for strain typing as previously described.² Random genetic events including point mutations, insertions, and deletions alter PFGE patterns. Patterns without fragment differences are considered to be genetically the same strain and genotypically related. One random genetic event may lead to 2 or 3 band differences between related isolates.²⁹ For the present study, isolates with < 4 band differences were considered to be the same strain and genotypically related (same PFGE pattern) and isolates with ≥ 4 band differences were considered to be different strains and not genotypically related (different PFGE pattern).

Statistical analysis—Descriptive statistics were used to characterize the population of dogs included, the isolates obtained, and the antimicrobial susceptibility phenotypes identified. Electrophoretic patterns and antimicrobial susceptibility phenotypes for isolates from pustules and carriage sites within individual dogs were analyzed to determine the percentage of dogs that had coagulase-positive staphylococcal isolates with the same PFGE pattern and the same susceptibility phenotype, different PFGE patterns and different susceptibility phenotypes, the same PFGE pattern and different susceptibility phenotypes, and different PFGE patterns and the same susceptibility phenotype. For relevant proportions, exact binomial 95% CIs and P values were calculated for testing a null hypothesis that the distribution of the PFGE and susceptibility patterns is random (ie, a proportion of dogs with the same PFGE and susceptibility pattern is 0.5). Analyses were performed with commercial software.^h Values of P < 0.05 were considered significant.

Results

Animals—Forty dogs met the inclusion criteria and were included in this study. Dogs ranged from 3 months to 15 years of age (mean, 4.6 years). Eighteen neutered males, 9 sexually intact males, 10 spayed females, and 3 sexually intact females were included in this study. Ten mixed-breed dogs and 21 purebred dogs were represented, with Labrador Retrievers (n = 7), Golden Retrievers (5), Boxers (3), German Shepherd Dogs (2), Lhasa Apsos (2), and Cairn Terriers (2) represented more than once. One dog each of the following breeds was included: Border Collie, Boston Terrier, English Bulldog, Cocker Spaniel, Newfoundland, Pug, Scottish Terrier, Shar Pei, and Chihuahua.

The history of antimicrobial use was considered incomplete as a result of poor owner recollection and incomplete documentation in the available medical records. Therefore, this information was not evaluated.

Gram stain, catalase, and coagulase test results—Six sites, 3 pustules and 3 carriage sites (anus, nonlesional axillary skin, and nasal mucosa), were sampled in each of 40 dogs, resulting in the submission of 240 samples for bacterial culture. From these, 246 isolates were obtained. Staphylococcal isolates were not cultured from 1 of 3 pustules in 2 dogs or from any carriage site in 2 dogs. All 246 isolates were gram positive and catalase positive; 224 were coagulase positive. One coagulase-positive isolate from the anus was lost during storage at −80°C after bacterial culture, morphologic description, Gram staining, and catalase and coagulase testing. This isolate and the 22 coagulase-negative isolates were not considered further in this study. Thus, 223 coagulase-positive staphylococcal isolates were included; 203 (91.0%) were S intermedius, 5 (2.2%) were Staphylococcus aureus, and 15 (6.7%) were classified as coagulase-positive Staphylococcus spp.

Oxacillin resistance—Fifteen of 223 (6.7%) coagulase-positive staphylococcal isolates from 3 dogs were resistant (5 isolates) or intermediate (10 isolates) in susceptibility to oxacillin with the disk diffusion test. Four (1 S intermedius isolate, 2 coagulase-positive Staphylococcus spp isolates, and 1 S aureus isolate) of these 15 isolates from 2 dogs were determined to be resistant to oxacillin on the basis of the OSA screening test. Thus, 1.8% (4/223 isolates) of the coagulasepositive staphylococcal isolates obtained from these 40 dogs with superficial bacterial folliculitis had in vitro oxacillin resistance.

Antimicrobial susceptibility phenotypes—Seventeen susceptibility phenotypes were identified for the 223 coagulase-positive staphylococcal isolates obtained from pustules and carriage sites: 11 for the 203 S intermedius isolates, 2 for the 5 S aureus isolates, and 6 for the 15 coagulase-positive Staphylococcus spp isolates. The most frequent susceptibility phenotype for S intermedius and coagulase-positive Staphylococcus spp isolates was pattern 1 (susceptible to all antimicrobial agents), and the most frequent susceptibility phenotype for S aureus isolates was pattern 5 (resistant only to polymyxin B sulfate; Table 1).

Coagulase-positive isolates from pustules— Eighty-four of the 128 (65.6%) coagulase-positive isolates from pustules were susceptible to all antimicrobial agents, and only 19 (14.8%) were resistant or intermediate in susceptibility to ≥ 2 antimicrobial agents. Eighty of 117 (68.4%) S intermedius isolates from pustules were susceptible to all antimicrobial agents, all 3 S aureus isolates from pustules were resistant to 1 antimicrobial agent, and 4 of the 8 coagulase-positive Staphylococcus spp isolates from pustules were susceptible to all antimicrobial agents.

All 128 (100%) isolates from pustules were susceptible to cephalothin and amoxicillin-clavulanic acid. One hundred fourteen of 128 (89.1%) isolates from pustules were susceptible to erythromycin and lincomycin hydrochloride, whereas only 103 of 128 (80.5%) isolates from pustules were susceptible to trimethoprimsulfamethoxazole (Table 2).

Fourteen of 117 (12.0%) S intermedius isolates from pustules were resistant to erythromycin and lincomycin hydrochloride, whereas 25 of 117 (21.4%) were resistant or intermediate in susceptibility to trimethoprim-sulfamethoxazole. All 3 S aureus isolates from pustules were resistant to polymyxin B sulfate. Four of 8 coagulase-positive Staphylococcus spp isolates from pustules were resistant or intermediate in susceptibility to cefpodoxime, 3 were intermediate in susceptibility to enrofloxacin, 1 was intermediate in susceptibility to marbofloxacin, and 1 was resistant to oxacillin.

Coagulase-positive isolates from carriage sites—Fifty-three of 95 (55.8%) coagulase-positive staphylococcal isolates from carriage sites were susceptible to all antimicrobial agents, and only 15 (15.8%) were resistant or intermediate in susceptibility to ≥ 2 antimicrobial agents. Forty-eight of the 86 (55.8%) S intermedius isolates from carriage sites were susceptible to all antimicrobial agents. One of 2 S aureus isolates from a carriage site was resistant to 1 antimicrobial agent, and the second was resistant to 6 antimicrobial agents. Five of the 7 coagulase-positive Staphylococcus spp isolates from carriage sites were susceptible to all antimicrobial agents tested.

All 95 (100%) isolates from carriage sites were susceptible to cephalothin, and 94 (98.9%) were susceptible to amoxicillin-clavulanic acid. Eighty-two of 95 (86.3%) isolates were susceptible to erythromycin and lincomycin hydrochloride, whereas only 70 of 95 (73.7%) isolates were susceptible to trimethoprim-sulfamethoxazole (Table 2).

Twelve of the 86 (14.0%) S intermedius isolates from carriage sites were resistant to erythromycin and lincomycin hydrochloride, and 25 (29.1%) were resistant or intermediate in susceptibility to trimethoprimsulfamethoxazole. Both S aureus isolates from carriage sites were resistant to polymyxin B sulfate, and 1 was resistant to erythromycin and lincomycin hydrochloride. One of the 7 coagulase-positive Staphylococcus spp isolates from carriage sites was intermediate in susceptibility to cefpodoxime, 1 was intermediate in susceptibility to enrofloxacin, 1 was intermediate in susceptibility to marbofloxacin, and 1 was resistant to oxacillin.

Untypable isolates and PFGE patterns—Of the 223 coagulase-positive isolates submitted for PFGE, a PFGE pattern could not be obtained for 7 isolates; 5 of these untypable isolates included all the isolates cultured from 1 of the 40 dogs. Isolates from this dog were excluded from PFGE analyses. From 39 dogs, 216 coagulase-positive (196 S intermedius, 5 S aureus, and 15 Staphylococcus spp) isolates were successfully typed by use of PFGE.

PFGE pattern for coagulase-positive isolates from pustules—Because 3 pustules were submitted for bacterial culture from each dog and multiple isolates were obtained from some pustules in some dogs, it was possible for an individual dog to be included in > 1 group. In 30 of 39 (76.9%; 95% CI, 60.7% to 88.9%; P < 0.001) dogs, isolates from all pustules had the same PFGE pattern and the same susceptibility phenotype. Seven of 39 (17.9%) dogs had isolates from 2 of 3 pustules with the same PFGE pattern and the same susceptibility phenotype. Thus, 37 of 39 (94.9%; 95% CI, 82.7% to 99.4%; P < 0.001) dogs had isolates with the same PFGE pattern from ≥ 2 pustules, and these isolates had the same susceptibility phenotype.

Three of 39 (7.7%) dogs had isolates from pustules with different PFGE patterns and different susceptibility phenotypes. Furthermore, 2 of 39 (5.1%) dogs had isolates from pustules with the same PFGE pattern and different antimicrobial susceptibility phenotypes, whereas 7 of 39 (17.9%) dogs had isolates from pustules with different PFGE patterns and the same susceptibility phenotype.

In all 39 dogs, the susceptibility phenotype for isolates from ≥ 2 pustules was the same, regardless of PFGE pattern. In 5 of the 39 dogs, the susceptibility phenotype for some isolates from pustules varied, regardless of PFGE pattern.

PFGE for coagulase-positive isolates from carriage sites—Three of 39 dogs were excluded, as coagulase-positive staphylococci were not isolated from any carriage site. Three other dogs were excluded, as only 1 coagulase-positive staphylococcal isolate from 1 of 3 carriage sites was cultured or successfully genotyped with PFGE. Therefore, 33 of 39 dogs were evaluated. Because samples from 3 carriage sites were submitted for bacterial culture from each dog and multiple isolates were obtained from some carriage sites in some dogs, it was possible for an individual dog to be included in > 1 group.

In 12 of 33 (36.4%; 95% CI, 20.4% to 54.9%; P = 0.163) dogs, all isolates from carriage sites had the same PFGE pattern and the same susceptibility phenotype. Nine of 33 (27.3%) dogs had isolates from 2 carriage sites with the same PFGE pattern and the same susceptibility phenotype. Thus, 21 of 33 (63.6%; 95% CI, 45.1% to 79.6%; P = 0.163) dogs had isolates with the same PFGE pattern from ≥ 2 carriage sites, and these isolates had the same susceptibility phenotype.

Nine of 33 (27.3%) dogs had isolates from carriage sites with different PFGE patterns and different susceptibility phenotypes. Furthermore, 3 of 33 (9.1%) dogs had isolates from carriage sites with the same PFGE pattern and different susceptibility phenotypes, whereas 10 of 33 (30.3%) dogs had isolates from carriage sites with different PFGE patterns and the same susceptibility phenotype.

In 31 of 33 (93.9%; 95% CI, 79.8% to 99.3%; P <0.001) dogs, the susceptibility phenotype for isolates from ≥ 2 carriage sites was the same, regardless of the PFGE pattern. In 11 of 33 (33.3%; 95% CI, 18.0% to 51.8%; P = 0.080) dogs, the susceptibility phenotype for some isolates from carriage sites varied, regardless of PFGE pattern.

Discussion

Results of our study indicated that in a substantial number of dogs with superficial bacterial folliculitis, coagulase-positive staphylococcal isolates from multiple pustules that are genotypically the same strain have the same antimicrobial susceptibility phenotype. Results of our study are consistent with those of a human study³⁰ in which S aureus isolates were evaluated and found that susceptibility phenotypes were the same among isolates with the same genotype. In addition, in our study, coagulase-positive staphylococcal isolates from pustules and carriage sites in 7 of 39 (17.9%) and 10 of 33 (30.3%) dogs, respectively, that were genotypically different strains had the same antimicrobial susceptibility phenotype. This was most likely attributable to the fact that most of the isolates in our study were susceptible to all or most of the tested antimicrobial agents. Thus, an unrelated strain could have the same susceptibility phenotype. Therefore, what is clinically relevant from the results of our study is that ≥ 2 pustule isolates from all (n = 39) dogs and ≥ 2 carriage site isolates from 31 of 33 (93.9%) dogs had the same susceptibility phenotype, regardless of strain type. Thus, in dogs with superficial bacterial folliculitis, selection of systemic antimicrobial agents for the treatment of staphylococcal bacterial infection should not require strain typing.

However, in our study, we identified isolates from pustules and carriage sites in 7.7% and 27.3% of dogs, respectively, that were different strains with different susceptibility phenotypes. It is interesting to mention that this occurred more frequently for isolates from carriage sites than for isolates from pustules. The importance of identifying different strains of coagulase-positive staphylococci with different susceptibility phenotypes from carriage sites is not presently known, as samples from carriage sites are not typically submitted for bacterial culture. However, as 7.7% of the dogs had pustule isolates of different strains and with different susceptibility phenotypes, it is important to recognize that it is possible for dogs with superficial bacterial folliculitis to have > 1 strain of staphylococci causing clinical signs, and the susceptibility phenotypes for these isolates may be different. In addition, few of the dogs in our study had coagulase-positive staphylococcal isolates from pustules and carriage sites that were the same strain but had different susceptibility phenotypes. There are multiple mechanisms (eg, gene mutation or the acquisition of genetic material) that result in the alteration of microbial DNA and development of antimicrobial resistance among bacteria. Such genome changes may remain or be lost by an organism over time.³¹ Therefore, it is possible for isolates that are the same strain to have different susceptibility phenotypes.³⁰ On the basis of the results of our study, most of the time, treatment of superficial bacterial folliculitis will be successful based on empiric treatment or bacterial culture and susceptibility testing of 1 lesional isolate. On occasion, if a clinical response is not achieved, bacterial culture and susceptibility testing should be performed and clinicians should consider obtaining samples from > 1 pustule to optimize the selection of an antimicrobial agent for the treatment of superficial bacterial folliculitis in dogs.

Results of our study indicated that antimicrobial resistance to antimicrobial agents commonly used for the treatment of superficial bacterial folliculitis in dogs caused by coagulase-positive staphylococci, specifically S intermedius, is uncommon in the geographic region in which our study was performed. All isolates in our study were susceptible to cephalothin, and with the exception of 1 S aureus isolate from a carriage site, all were susceptible to amoxicillin-clavulanic acid. Numerous studies,^5,32–40 with results similar to those of our study, have been preformed to evaluate the incidence of resistance to cephalosporins and amoxicillin-clavulanic acid among S intermedius isolates obtained from dogs with superficial bacterial folliculitis.

Potentiated sulfonamides are often recommended for the empiric treatment of dogs with superficial bacterial folliculitis.^41,42 However, in our study, 21.4% of S intermedius isolates from pustules and 29.1% of S intermedius isolates from carriage sites were resistant or intermediate in susceptibility to trimethoprim-sulfamethoxazole. This finding is in contrast to those of previous studies^4,32,33,43 with lower reported percentages of S intermedius isolates from dogs with superficial bacterial folliculitis that were resistant to potentiated sulfonamides, ranging from 0% to 9%. Results of our study are, however, consistent with the results of 3 studies^5,34,37 with reported higher percentages of S intermedius isolates from dogs with superficial bacterial folliculitis that were resistant to potentiated sulfonamides, ranging from 14.3% to 33.2%. Because of the variability in the reported resistance of S intermedius isolates from dogs with superficial bacterial folliculitis to trimethoprim-potentiated sulfonamides, antimicrobial agents in this class appear to be poor empiric choices for the treatment of superficial bacterial folliculitis in dogs. The use of a trimethoprim-potentiated sulfonamide for the treatment of superficial bacterial folliculitis in dogs should be based on bacterial culture and antimicrobial susceptibility testing.

In our study, 12.0% of S intermedius isolates from pustules and 14.0% of S intermedius isolates from carriage sites were resistant to erythromycin and lincomycin hydrochloride. This was not unexpected, as previous studies^{5,32-34,37,44} have reported similar percentages of coagulase-positive staphylococcal isolates obtained from dogs with superficial bacterial folliculitis that were resistant to these antimicrobial agents. As such agents are narrow-spectrum antimicrobials with demonstrated efficacy against S intermedius, they are often recommended for initial empiric treatment in dogs with superficial bacterial folliculitis.^45,46 However, their use is limited by the development of resistance and the common occurrence of cross-resistance between macrolides and lincosamides.^47,48 Results of our study support the need for antimicrobial susceptibility testing if a poor clinical response to empiric treatment is found with either of these antimicrobial agents.

In our study, all S intermedius and S aureus isolates were susceptible to enrofloxacin and marbofloxacin. This lack of resistance to fluoroquinolones was consistent with findings in previous studies,^38,49 and recent studies^32,50,51 have reported only a low incidence of resistance to these antimicrobial agents. Thus, it was encouraging to document the lack of fluoroquinolone resistance among a large number of S intermedius isolates obtained from dogs with superficial bacterial folliculitis in our study.

In our study, 15 isolates were classified as coagulase-positive Staphylococcus spp and the susceptibility phenotype of these isolates varied from the S intermedius isolates. It is important to mention that these isolates were the only isolates that were resistant or intermediate in susceptibility to enrofloxacin or marbofloxacin, and the pattern of resistance was similar for isolates from both pustules and carriage sites. Furthermore, these isolates represented 5 of the 8 isolates that were resistant to cefpodoxime and 2 of the 4 isolates that were resistant to oxacillin. Interestingly, none of these isolates were resistant or intermediate in susceptibility to erythromycin, lincomycin hydrochloride, or trimethoprim-sulfamethoxazole, the most common antimicrobial agents to which S intermedius isolates were resistant or intermediate in susceptibility in our study.

For S aureus isolates from humans, polymyxin B sulfate resistance may differentiate S aureus isolates from other coagulase-positive staphylococcal species.²² To our knowledge, no study has specifically evaluated polymyxin B sulfate resistance by use of the disk diffusion test among S aureus isolates obtained from dogs. In our study, only the 5 S aureus isolates (which include all S aureus isolates in our study) were resistant to polymyxin B sulfate, suggesting that polymyxin B sulfate resistance may be used as an adjunctive test to differentiate S aureus isolates obtained from dogs from other coagulase-positive staphylococci.

In our study, 15 isolates required verification of oxacillin resistance or susceptibility with the OSA screening test. As a result, only 1.8% of the coagulase-positive staphylococcal isolates from these dogs with superficial bacterial folliculitis were resistant to oxacillin. Previous studies^{5,27,32,37,52,53} have also reported a lack of, or a low occurrence of, oxacillin resistance among coagulase-positive staphylococci obtained from dogs. If the OSA screening test had not been performed in our study, 11 additional isolates would have been reported as resistant or intermediate in susceptibility to oxacillin. For the purposes of identifying susceptibility phenotypes in our study, disk diffusion test results were reported on the basis of the actual measured zone diameters. However, clinical veterinary microbiology laboratories report oxacillin-resistant staphylococci as resistant to cephalosporins and amoxicillin-clavulanic acid, even if the organism is susceptible on the disk diffusion test. If oxacillin resistance is not verified with confirmatory tests, staphylococci may be reported as resistant to some of the most effective antimicrobial agents available for the treatment of superficial bacterial folliculitis in dogs.

Fifteen coagulase-positive isolates could not be identified with the commercial identification used and were classified as Staphylococcus spp. The deficiencies of this test in the speciation of these 15 staphylococcal isolates were recently discussed.²

To our knowledge, ours is the first study to investigate the relationship between PFGE pattern and susceptibility phenotype for coagulase-positive staphylococcal isolates obtained from carriage sites in dogs with superficial bacterial folliculitis. If it is determined that carriage sites are the source of staphylococcal lesions, these results will provide data for studies designed to investigate the effectiveness of antimicrobial treatment of carriage sites in dogs with superficial bacterial folliculitis.