Prevalence of and risk factors for isolation of meticillin-resistant Staphylococcus spp. from dogs with pyoderma in northern California, USA
Background – Canine pyodermas associated with meticillin-resistant Staphylococcus spp. (MRS) have increased in prevalence over the past decade.
Hypothesis/Objectives – To compare the prevalence of MRS isolation from dogs with superficial pyoderma at a primary care clinic (PCC) and those at a tertiary care facility (VMTH) in California, USA, and identify associated risk factors.
Animals – Client-owned dogs from the VMTH (80 dogs) and the PCC (30 dogs).
Methods – Aerobic bacterial culture and antibiotic susceptibility were performed on swab specimens collected from dogs, and meticillin resistance was determined using microdilution methods according to Clinical and Laboratory Standards Institute guidelines. A mecA gene PCR assay was used to confirm meticillin resistance when possible.
Results – Of 89 staphylococcal isolates from the VMTH, 34 (38.2%) were meticillin resistant. In 31 dogs, pyoderma persisted, and one or more follow-up isolates were obtained. The species isolated and drug susceptibility changed unpredictably during treatment. Of 33 PCC isolates, nine (27.3%) were meticillin resistant. Multiple drug resistance was identified in 41 of 53 (77.3%) MRS isolates from the VMTH and five of nine from the PCC. The sensitivity and specificity of PCR for the detection of meticillin resistance was 34 of 39 (87%) and 86 of 87 (99%), respectively. Risk factors for meticillin resistance for both sites were antibiotic treatment within the last year (P = 0.001), and for VMTH, hospitalization of dogs within the last year (P = 0.001).
Conclusions and clinical importance – The prevalence of meticillin resistance was not different between VMTH and PCC isolates (P = 0.29). Previous antimicrobial therapy was an important risk factor for the isolation of MRS at both sites.
In the past decade, skin infections caused by meticillin-resistant coagulase-positive Staphylococcus spp., particularly meticillin-resistant Staphylococcus pseudintermedius (MRSP), have become increasingly widespread in dogs.1–4 Meticillin resistance results from presence of the mecA gene, which encodes penicillin binding protein 2a and confers resistance to all β-lactam antimicrobial drugs. The mecA gene exists on a large mobile genetic element (the staphylococcal cassette chromosome or SCC) and is passed on through clonal spread of mecA-positive bacteria.5
Prevalence rates ranging from 0.58 to 30% of meticillin-resistant Staphylococcus spp. isolated from dogs have been reported; however, comparison of these studies is difficult because of differences in the populations sampled (healthy versus diseased, hospitalized versus outpatients) and in the sites of specimen collection.3,6-10 Risk factors for MRSP infection in dogs identified thus far include previous antimicrobial therapy, previous hospitalization, living in an urban environment and older age of the affected animal.8
Together with the increase in the prevalence of MRSP infections, there has been an increase in the number of all multidrug-resistant staphylococcal infections in dogs. Multidrug resistance (MDR) is typically defined as resistance to three or more drug classes.11,12 Staphylococci are capable of development of resistance not only to β-lactam antimicrobials but also to many other classes of antimicrobials.1,13,14 The emergence of MDR in veterinary medicine has led to a limited range of treatment options, with increases in morbidity, mortality and cost of treatment.
Both MRSP and meticillin-resistant Staphylococcus aureus (MRSA) can cause pyoderma in dogs, but infections with MRSP are more prevalent.15,16 As MRSA prefers to colonize humans, canine pyoderma caused by MRSA is of greater zoonotic concern than pyoderma caused by MRSP.17 Nevertheless, reports of MRSP infections in humans have been documented in the literature.18–22
The purpose of this study was to compare the prevalence of meticillin-resistant Staphylococcus (MRS) isolation from dogs with first-time or recurrent pyoderma from a primary care facility with those from a tertiary care facility, and to identify risk factors that might be associated with isolation of MRS. We hypothesized that a higher prevalence of MRS would be found in dogs seen at a tertiary care facility than in dogs seen at a primary care facility.
Materials and methods
All dogs enrolled in the study were required to have both clinical and cytological evidence of superficial pyoderma. Pyoderma could be newly diagnosed or recurrent. Clinical abnormalities consistent with superficial pyoderma included epidermal collarettes, crusts, papules or pustules. All dogs included in the study were client-owned animals, and the study design was approved by the William R. Pritchard Veterinary Medical Teaching Hospital clinical trial review board at the University of California, Davis. Dogs with evidence of superficial pyoderma were not excluded from the study for any reason, including concurrent illness or current/previous medical therapy such as corticosteroids, antihistamines and antimicrobials. At the time of the initial appointment, owners were required to fill out a questionnaire regarding their dog’s age, breed, sex, history of current and previous antibiotic administration, concurrent systemic and/or dermatological disease, previous hospitalizations, previous surgeries, bathing frequency (at home or by a professional groomer), and whether there were other pets in the household that were receiving antibiotics. In regards to humans sharing the household with the dog, the following factors were noted: whether any humans in the household were receiving or had a history of receiving antibiotics, and whether there were immunocompromised individuals, school-age athletes, healthcare workers or humans with a diagnosis of MRSA infections.
Swab specimens were collected prospectively for aerobic bacterial culture and antibiotic susceptibility testing from dogs diagnosed with superficial pyoderma at their first appointment with the Dermatology Service of the William R. Pritchard Veterinary Medical Teaching Hospital, University of California, Davis (VMTH-UCD) between August 2010 and June 2011. Over the same time period, specimens were also collected from dogs diagnosed with superficial pyoderma at their first appointment by general practice veterinarians at a large general private practice in northern California. A sterile cotton swab was used to collect specimens from lesions consistent with superficial pyoderma.23 Specimens were collected from pustules, beneath crusts and/or at the margin of epidermal collarettes. The same brand of cotton swab (BD BBL™ CultureSwab™ Plus Amies Gel without Charcoal, Double Swab; Franklin Lakes, NJ, USA) was used for collection at both the VMTH and the primary care clinic. Samples from the primary care clinic and the VMTH were submitted to the VMTH microbiology diagnostic laboratory within 24 h of collection.
Dogs seen at the VMTH-UCD were managed for the underlying cause(s) of superficial pyoderma, and systemic antimicrobial drug treatment was initiated for 30 days based on results of the bacterial culture and susceptibility. All dogs from the VMTH were treated with systemic antimicrobial therapy with the exception of one dog, which was treated with topical therapy alone. Some dogs were also treated with topical antimicrobial therapy in the formulation of shampoo, spray and/or wipe. Follow-up examination and, if clinical and cytological evidence of pyoderma persisted, skin aerobic bacterial culture and susceptibility testing, were recommended on a monthly basis after initiation of antimicrobial therapy. Whether and at what time reevaluation occurred were dependent on owner compliance with these recommendations. For dogs that did have follow-up cultures performed, the time of follow-up after the initial visit and the results of culture and susceptibility were recorded.
Swabs were inoculated within 24 h of collection onto 5% (v/v) sheep blood agar and MacConkey agar (Hardy Diagnostics, Santa Maria, CA, USA) and incubated at 37°C for 24-48 h. Primary identification of staphylococci was based on colony morphology, Gram staining and the conventional catalase test. Isolates were further identified using a panel of conventional biochemical tests including tube coagulase, haemolysis and acetoin production (maltose, mannitol, trehalose and arginine dihydrolase).24 Staphylococcus schleiferi ssp. coagulans was differentiated from other coagulase-positive staphylococci, using standard biochemical methods, including the Voges–Proskauer, tube and slide coagulase assays.25 Coagulase-negative staphylococci (CNS), including S. schleiferi ssp. schleiferi, were not identified to the species level. Coagulase-positive staphylococci with biochemical reaction patterns that did not clearly classify them as S. intermedius, S. aureus, S. pseudintermedius or S. schleiferi were designated as ‘S. intermedius group’ (SIG) organisms.
Antimicrobial susceptibility testing
An automated microdilution method (Sensititre® Autoinoculator; Trek Diagnostic Systems, Clevland, OH, USA) was used for minimal inhibitory concentration (MIC) determination for 22 antimicrobial drugs according to Clinical and Laboratory Standards Institute (CLSI) guideines.26 The panel of antimicrobials tested included amikacin, amoxicillin-clavulanic acid, ampicillin, cefazolin, cefovecin, cefoxitin, cefpodoxime, ceftiofur, chloramphenicol, clindamycin, doxycycline, enrofloxacin, erythromycin, gentamicin, imipenem, marbofloxacin, oxacillin + 2% NaCl, penicillin, rifampicin, ticarcillin–clavulanic acid, ticarcillin and trimethoprim sulfamethoxazole. The isolates were characterized as susceptible, intermediate or resistant on the basis of cut-off MIC values published by the CLSI.26 Isolates were classified as meticillin resistant if MIC values for oxacillin + 2% NaCl were above the MIC breakpoint published by the CLSI for that antimicrobial. Multidrug resistance was characterized as resistance to three or more different antimicrobial classes.11,12
mecA gene PCR assay
Quantitative PCR assay design
Primers and the probe used for mecA (penicillin binding protein 2a gene) were published previously.27 Briefly, the forward primer sequence was 5′-CATTGATCGCAACGTTCAATTT-3′, the reverse primer sequence was 5′-TGGTCTTTC TGCATTCCTGGA-3′, and the probe was modified as a minor groove binder probe (Applied Biosystems, Life Technologies Corporation, Carlsbad, CA, USA) with the sequence 5′-FAM-AGTTAGATTGGGATCATAGC-BHQ-3′. The quantitative PCR systems were validated using twofold dilutions of genomic DNA that contained the mecA gene. The dilutions were analysed in triplicate, and a standard curve was plotted against the dilutions. The slope of the standard curve was used to calculate amplification efficiencies using the formula E =101/-s – 1, where E is efficiency and s is slope of the curve with assay efficiency at 97.4%. All PCRs were carried out with a positive control. Water was used as a negative control.
Sample collection and automated nucleic acid preparation
Colonies were selected randomly and were scraped into 200 μL of PBS and vortexed for 20 s. Total nucleic acid was extracted from 200 μL of cell suspension using an automated nucleic acid extraction system (QiaX-tractor; Qiagen, Valencia, CA, USA) according to the manufacturer’s recommendations.
The PCR assays contained a final concentration of 400 nM for each primer and 80 nM for the probe. The PCR was performed using 1 μL bacterial genomic DNA, 4 μL water and a commercially available PCR mastermix (TaqMan Universal PCR Mastermix; Applied Biosystems, Carlsbad, CA, USA) containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 5 mM MgCl2, 2.5 mM deoxynucleotide triphosphates, 0.625 units AmpliTaq Gold DNA polymerase per reaction and 0.25 units AmpErase uracil-N-glycosylase per reaction in a final volume of 12 μL. The samples were placed in 384-well plates, amplified in an automated fluorometer (AB PRISM 7900 HT FAST; Applied Biosystems, Foster City, CA, USA) and run with the mecA gene assay along with a panbacterial assay designed to detect bacterial 16S DNA for extraction quality control (PanBakt assay; University of California, Davis molecular core facility, Davis, CA, USA). This assay utilizes two forward primers (5′-AACTCAAAGGAATTGACGGGG-3′ and 5′-AAACTCAAATGAATTGACGGGG-3′). The reverse primer sequence was 5′-GCTCGTTGCGGGACTTA-3′, and the minor groove binder probe 5′-FAM-TGTCGTCAGCTCGTG-BHQ-3′. Standard amplification conditions were used, as follows: 2 min at 50°C, 10 min at 95°C, 40 cycles of 15 s at 95°C and 60 s at 60°C. Fluorescence signals were collected during the annealing temperature and cycle threshold (Ct) values exported with a threshold of 0.1 and baseline values of 3-10. Any Ct value <40 was considered positive.
A two-tailed Fisher’s exact test was used to compare prevalence between groups for categorical variables. The chi-squared test was used for comparisons that involved more than two groups. A P value of <0.05 was considered significant.