Species specificity in the adherence of staphylococci to canine and human corneocytes: a preliminary study

Staphylococcal pyoderma is rarely contagious between dogs and humans, or humans and dogs. This study investigated the hypothesis that there are species differences in adherence of Staphylococcus intermedius (the most common isolate from dogs) and Staphylococcus aureus (the most common isolate from humans) to canine and human corneocytes. Sheets of corneocytes were collected from the ventral abdomen of 10 dogs and the medial forearm of 10 humans (all healthy and without any history or physical signs of skin disease) using double-sided tape. Staphylococcus intermedius from a case of canine bacterial pyoderma and a human strain of S. aureus were prepared in phosphate buffered saline (PBS) and applied in duplicate, respectively, to the canine and human corneocyte-covered tapes using PBS as negative control. After incubation, rinsing, and staining with crystal violet, quantification of the adherent bacteria was carried out blindly by computerized image analysis. Staphylococcus intermedius was found to adhere significantly more to canine corneocytes than S. aureus (P = 0.0006), whereas S. aureus showed greater adherence to human corneocytes than S. intermedius (P < 0.0001). In addition, the pattern of adherence differed between the two organisms, with S. intermedius adhering to the entire surface and S. aureus adhering mainly to the periphery of both canine and human corneocytes. Preference for adherence to these two hosts may explain, in part, why S. intermedius and S. aureus are uncommonly isolated from human and canine skin infections, respectively.     

A comparison of adherence by four strains of Staphylococcus intermedius and Staphylococcus hominis to canine corneocytes collected from normal dogs and dogs suffering from atopic dermatitis

The aim of this study was to compare the adherence of four strains of Staphylococcus intermedius and a single strain of Staphylococcus hominis to corneocytes from both normal dogs and dogs suffering from atopic dermatitis. Cells from the skin surface, corneocytes, were collected from 10 normal dogs and 10 dogs suffering from atopic dermatitis. Four strains of S. intermedius, three isolated from canine pyoderma skin lesions (strains A, B and C), and one isolated form from canine synovial membrane sample from a case of septic arthritis (strain D) were compared. S. hominis, which is not normally associated with canine disease, was also evaluated for its ability to adhere to canine corneocytes. S. hominis did not adhere to canine corneocytes. All four strains of S. intermedius adhered well to canine corneocytes collected from both normal and atopic dogs. All strains of S. intermedius showed statistically greater adherence to corneocytes collected from atopic dogs compared with those collected from normal dogs. It was concluded that the adherence assay employed here showed that S. hominis does not adhere to canine corneocytes, S. intermedius adheres preferentially to atopic corneocytes.     

Adherence of Staphylococcus intermedius to canine corneocytes in vitro

This study investigated the in vitro adherence of Staphylococcus intermedius to canine corneocytes, collected from a healthy dog using double-sided adhesive tape. Adherence was shown to depend on duration (P < 0.001) and temperature of incubation (P < 0.001) and the concentration of bacteria (P < 0.001). Isolates of S. intermedius from lesions of pyoderma were not generally more adherent to healthy canine skin than were isolates from healthy dogs. Significant differences in adherence were demonstrated between individual isolates within both groups (P < 0.001). The study suggests that among S. intermedius there is no correlation between virulence and adherence to canine corneocytes in vitro. The finding may be important for the potential use of avirulent variants of S. intermedius as antagonistic strains against canine pyoderma. However, more studies are needed to compare the adherence of the isolates to skin cells obtained from dogs with diseases predisposed to pyoderma.     

Adherence by Staphylococcus intermedius to canine corneocytes: a preliminary study comparing noninflamed and inflamed atopic canine skin

The adherence by three strains of Staphylococcus intermedius to corneocytes collected from healthy dogs was compared to the adherence to corneocytes collected from the inflamed (erythematous) and noninflamed (normal appearing) skin of dogs suffering from atopic dermatitis. All three strains of S. intermedius adhered in greater numbers to corneocytes from both inflamed and noninflamed atopic skin than to corneocytes from healthy dogs. Adherence was greatest to corneocytes from inflamed atopic skin but one strain showed no statistical difference for adherence to inflamed and noninflamed atopic skin. These findings suggest that S. intermedius adheres extensively to both inflamed and noninflamed canine atopic skin. This may be important in the colonization of atopic skin by this microorganism. Strain variation in the ability of S. intermedius to adhere to canine atopic corneocytes is probable.     

Reduced in vitro adherence of Staphylococcus species to feline corneocytes compared to canine and human corneocytes

It is apparent that in-contact humans and animals exchange commensal staphylococci. Previous in vitro studies, however, indicate that staphylococci preferentially adhere to corneocytes from host species. This study compared adherence of meticillin-sensitive and -resistant Staphylococcus aureus (MSSA/MRSA), S. intermedius, S. felis and S. hominis to feline, canine and human corneocytes acquired from 10 healthy subjects using adhesive tape discs. Adherent bacteria were counted using an image processing and analysis programme. Mean adherence of MSSA (P = 0.0009), MRSA (P = 0.0162) and S. intermedius (P = 0.0117), but not S. felis or S. hominis, to feline corneocytes was significantly lower than that to canine and human corneocytes. All the isolates had similar adherence to both human and canine corneocytes. S. felis was the most adherent species to feline corneocytes followed by S. intermedius, and then MSSA, MRSA and S. hominis. For dogs and humans, S. intermedius and S. felis were the most adherent, followed by MRSA and MSSA, and then S. hominis. These results do not reveal any preferential adherence of staphylococci to canine or human corneocytes. Poor adherence to feline corneocytes could suggest that cats are relatively resistant to pyoderma and cross-species transmission of staphylococci.     

Adherence by Staphylococcus intermedius to canine keratinocytes in atopic dermatitis

The adherence of Staphylococcus intermedius to canine keratinocytes in normal dogs was compared to that in dogs suffering from atopic dermatitis, primary seborrhoea and bacterial pyoderma. Statistically significant greater adherence by S. intermedius to keratinocytes occurred in atopic dogs and dogs suffering from pyoderma when compared with the normal group (P < 0.01) and dogs suffering from primary seborrhoea (P < 0.05). This is similar to the results of a study of human atopic dermatitis by Cole and Silverberg (1986) who demonstrated increased adherence by S. aureus to keratinocytes from atopic dermatitis patients when compared with adherence to keratinocytes in a variety of non-atopic dermatoses. This increased adherence by pathogenic staphylococci to keratinocytes may in part explain the high incidence of staphylococcal pyoderma seen in both canine and human patients suffering from atopic dermatitis.     

Use of computerized image analysis to quantify staphylococcal adhesion to canine corneocytes: does breed and body site have any relevance to the pathogenesis of pyoderma?

An optimized system of computerized image analysis was used to investigate variations in the adherence of Staphylococcus intermedius to canine corneocytes from four different breed groups and six different anatomical sites. S. intermedius showed significantly greater adherence to the head and neck compared with the dorsum, but adherence to the limb, axilla and groin did not differ from other sites. Furthermore, there was significantly greater adherence of S. intermedius to corneocytes from the dorsum, forelimb, axilla and groin of Boxers and Bull Terriers than Spaniels and Hounds. S. intermedius, and also Pseudomonas aeruginosa, exhibited abundant adherence, which was significantly greater than Staphylococcus aureus, Streptococcus canis, Klebsiella pneumoniae and Escherichia coli. In addition, S. intermedius adherence demonstrated a sigmoid dose-response curve with increasing bacterial concentration. These results suggest that S. intermedius adheres to canine corneocytes by a specific receptor-ligand interaction and adheres to the skin of some breeds more avidly than others. However, variations in adherence between body regions would not account for the predilection sites of canine bacterial pyoderma.     

Evaluation of aerobic bacteriologic culture of epidermal collarette specimens in dogs with superficial pyoderma

OBJECTIVE: To evaluate a method of aerobic bacteriologic culture of epidermal collarette specimens from dogs with superficial pyoderma and compare results with those for aerobic bacteriologic culture of abdominal skin specimens in healthy dogs. DESIGN: Prospective study. ANIMALS: 22 dogs with epidermal collarettes and 24 healthy dogs. PROCEDURE: Dry sterile cotton swabs were rolled across epidermal collarettes or hairless areas of abdominal skin in healthy dogs and submitted for aerobic bacteriologic culture. Hemolytic colonies of gram-positive-staining cocci were tested for catalase production, and if results were positive, a coagulase test was performed. Colonies with coagulase activity were tested for the ability to ferment mannitol. Antimicrobial susceptibility testing was performed on all Staphylococcus spp that were isolated. RESULTS: S. intermedius was isolated from collarettes in 18 of 22 dogs with superficial pyoderma but not from healthy dogs. Estimated sensitivity and specificity of the culture method were 81.8% and 100%, respectively. There were no significant differences in the ability to culture S. intermedius, the number of S. intermedius isolates without resistance to antimicrobials, and the number of S. intermedius isolates resistant to penicillin G when comparing dogs with superficial pyoderma for the first time and dogs with recurrent pyoderma, dogs that did or did not receive concurrent antimicrobials, and dogs with and without underlying allergic disease. CONCLUSIONS AND CLINICAL RELEVANCE: Bacteriologic culture of epidermal collarette specimens was a simple and reliable method for identification of S. intermedius in dogs with superficial pyoderma, regardless of history of pyoderma or current antimicrobial use.     

Staphylococcal colonization of mucosal and lesional skin sites in atopic and healthy dogs

Staphylococcal colonization was compared in healthy dogs and in dogs with atopic dermatitis. Bacterial swabs were collected from the nasal mucosa, ear and perineum of 43 healthy and 24 atopic dogs and also from potentially infected skin lesions of the atopic dogs. Coagulase positive staphylococcal isolates were identified to the species level. At the time of this study Staphylococcus intermedius was considered a single species but has since been recognized as comprising at least three species with canine isolates believed to belong to Staphylococcus pseudintermedius . Of atopic dogs, 87.5% were colonized with S. intermedius compared to only 37.2% of healthy dogs. The ear was the only carriage site that showed any significant difference in S. intermedius isolation between healthy and atopic dogs. The perineum represented the most frequently colonized mucosal site for both groups. Sampling the nasal mucosa alone identified 71.4% of atopic and 37.5% of healthy S. intermedius carriers. Inclusion of a perineal swab identified 100% of atopic and 93.8% of healthy carriers. S. intermedius was isolated from all the lesional sites sampled from atopic dogs. Significantly fewer dogs were colonized by Staphylococcus aureus than S. intermedius , and there was no significant difference between S. aureus colonization of atopic and healthy dogs. S. aureus was not recovered from any lesions in atopic dogs. The results show that S. intermedius carriage is more prevalent in atopic dogs compared to healthy dogs and that to identify staphylococcal carriers both the nasal mucosa and the perineum should be sampled.     

Aspects of nasal, oropharyngeal and anal carriage of Staphylococcus intermedius in normal dogs and dogs with pyoderma

Aspects of carriage of Staphylococcus intermedius were studied in three groups of dogs. In 150 normal dogs presenting for annual booster inoculations, carriage rates for S . intermedius in the anterior nares (36%) and on the anal ring (36%) were recorded. In 44 dogs presented for routine elective surgery, carriage of S . intermedius was higher on the caudal nares (41%) than on the anterior nares (34%), oropharynx (25%) or anal ring (30%) allowing an assessment of the accuracy of basing nasal carriage rates on swabbing the anterior nares alone. A new finding was that animals from multidog households had significantly higher ( P < 0.005) nasal and anal carriage rates than dogs from single-dog households. In 33 dogs with superficial pyoderma, carrier rates were similar to those reported by others and significantly higher than in the other two groups with regard to both nasal and anal carriage ( P < 0.05 and P < 0.005, respectively).     

Heterogeneity of Staphylococcus pseudintermedius isolates from atopic and healthy dogs

Staphylococcus pseudintermedius is part of the normal canine flora but frequently causes pyoderma in canine atopic dermatitis (AD). This study aimed to determine whether particular S. pseudintermedius strains were associated with AD and/or pyoderma. Ninety‐six S. pseudintermedius isolates from the ear, nares, perineum and lesions of 21 atopic and 16 healthy dogs were lysed with proteinase K and digested with 40 U SmaI. Restriction products were separated using pulsed‐field gel electrophoresis (PFGE) with an Oxford S. aureus control and lambda‐ladder DNA concatomer markers. A dendrogram was constructed by the unweighted pair group method. All isolates showed a ≥56% similarity coefficient. Nine distinct PFGE clusters were identified, as follows: five from both atopic and healthy dogs; three from atopic dogs only; and one from healthy dogs only. Nine clusters were isolated from the nares, eight from the perineum, five from the ears and six from pyoderma lesions. There were no significant differences in the frequency of isolation from atopic or healthy skin, body sites or infected lesions for any of the clusters. Two of six healthy dogs and 18 of 20 atopic dogs with multiple isolates had closely related isolates (less than three band differences) at more than one sampling site. Isolates from pyoderma lesions were closely related to at least one mucosal isolate in 11 of 16 dogs. Staphylococcus pseudintermedius isolates appear to be heterogeneous, and colonization or infection of atopic skin was not associated with any particular strain or cluster of strains.     

Transmission of multiple antimicrobial-resistant Staphylococcus intermedius between dogs affected by deep pyoderma and their owners

The occurrence of antimicrobial-resistant Staphylococcus intermedius strains was investigated in 13 dogs affected by deep pyoderma, their owners and 13 individuals without daily contact with dogs (control group). A total of 90 canine and 33 human S. intermedius isolates were typed by pulsed field gel electrophoresis (PFGE) to determine their possible identity. The occurrence of S. intermedius in dog-owners was significantly higher compared with the control group (Fisher's exact test, P=0.03), with S. intermedius being detected in seven dog-owners and in one individual not exposed to dogs. The results of the PFGE analysis showed that six out of 13 (46%) owners carried strains identical to those isolated from their dogs. Strains detected in both dogs and humans were resistant up to five different antimicrobial classes, including penicillins, fusidic acid, macrolides/lincosamides, tetracycline and chloramphenicol. Based on the results of this study, owners of dogs affected by deep pyoderma often carry multiple antimicrobial-resistant strains of S. intermedius occurring in their dogs. Independent of the direction and modalities of transmission, this finding raises questions concerning the possible transfer of resistance genes from canine S. intermedius to human pathogenic staphylococci.     

Serum and skin IgA concentrations in normal and atopic dogs

Objective To compare serum and skin surface IgA concentrations from atopic and normal dogs.
Procedure IgA concentrations in sera and skin washings of 20 clinically normal dogs that had no history of pruritus or skin disease were compared to those obtained in 20 dogs with a diagnosis of atopy determined by history, clinical examination and positive intradermal skin test.
Results There was no significant difference in the mean serum IgA concentration in normal dogs (252 ± 187 mg/L) versus atopic animals (314 ± 327). When skin washings from all sites in both groups were compared, atopic dogs had significantly greater concentrations of IgA in their skin washings than normal dogs as evaluated by an enzyme-linked immunoassay (P < 0.001). However, there was no significant difference between the individual sites of the skin washings of atopic and normal dogs.
Conclusion IgA concentrations of skin washings in atopic dogs were greater than in normal dogs. Further investigations need to determine if the greater concentrations were caused by nonspecific inflammation or by secretion of allergen-specific IgA onto the skin surface.     

Aspects of the humoral immune response to Staphylococcus intermedius in dogs with superficial pyoderma,deep pyoderma and anal furunculosis

Bacterial infection (pyoderma) of the canine skin is largely caused by Staphylococcus intermedius and may be a superficial or deep infection. Pyoderma may be a primary, idiopathic disease or secondary to a range of other dermatological disorders. In this study, the serum concentrations of IgG, IgA, antistaphylococcal IgG and antistaphylococcal IgA were measured by ELISA in normal dogs (n = 22), dogs with idiopathic deep pyoderma (n = 22), atopic dermatitis and superficial pyoderma (n = 24), atopic dermatitis without pyoderma (n = 25), flea bite dermatitis with superficial pyoderma (n = 8), pustular demodicosis (n = 8) and German shepherd dogs with anal furunculosis (n = 28). The serum IgG was significantly increased in dogs with atopy and superficial pyoderma (p < 0.001), and lower than normal in dogs with idiopathic deep pyoderma (p < 0.015). The concentration of serum IgA was significantly lower than normal in dogs with atopy uncomplicated by pyoderma (p < 0.015). The concentration of antistaphylococcal IgG in all clinical sera was significantly elevated (p < 0.001) when compared to normal dogs but concentrations of antistaphylococcal IgA were no greater than in normal dogs. Western blotting analysis for determination of the specificity of serum IgG antistaphylococcal antibody revealed that there were nine major epitopes. Discriminant analysis demonstrated that particular combinations of these epitopes were recognised more frequently by sera from dogs in different clinical groups.     

Molecular characterization of Staphylococcus intermedius carriage by healthy dogs and comparison of antimicrobial susceptibility patterns to isolates from dogs with pyoderma

We conducted an epidemiological study of Staphylococcus intermedius using arbitrarily primed PCR (AP-PCR) and antibiograms. One hundred and twenty-five S. intermedius isolates were recovered from the oral cavity and/or cranial hair coat of healthy dogs enrolled in a pet therapy program. Commensal S. intermedius was cultured from 32% of the oral cavity cultures and 13% of the cranial hair coat cultures. We characterized the colonization of the dogs as transient, intermittent, or persistent. For dogs characterized as persistently colonized, 73% of the isolates came from the oral cavity. These isolates were also genotyped by AP-PCR. A single major AP-PCR type was observed in 91% of the dogs (n=22); minor variations were frequently observed in these major types. Antibiograms of these commensal isolates were compared to antibiograms from 97 historical clinical isolates (1988-1992) obtained from cases of canine pyoderma. Resistance was most often observed to penicillin (64% and 55%) and tetracycline (38% and 38%) among the commensal and clinical isolates, respectively. The commensal isolates were significantly less resistant to erythromycin, clarithromycin, clindamycin, and trimethoprim/sulfamethoxazole. Our data suggests that differences in both genotype and antimicrobial susceptibility phenotypes exist among S. intermedius strains isolated from different anatomic sites from the same dog and supports the opportunistic nature of S. intermedius in canine infections.     

The influence of topical unsaturated fatty acids and essential oils on normal and atopic dogs

Seven dogs with atopic dermatitis and six normal dogs were treated with a spot-on product containing essential oils and unsaturated fatty acids q 7 days for 8 wk. Seven additional atopic dogs received a daily spray containing similar ingredients to the spot-on. In all dogs, transepidermal water loss (TEWL) was measured before and after treatment using a closed chamber device. In atopic dogs, a validated lesion score (canine atopic dermatitis extent and severity index, CADESI) was determined and pruritus was assessed with a visual analog scale before and after treatment. The mean CADESI scores in atopic dogs decreased with the spot-on (P=0.0043) and with the spray (P=0.0366). Similarly, the pruritus scores decreased with the spot-on (P=0.266) and with the spray (P=0.0177). There was a significant difference between the TEWL values of healthy and atopic dogs on the abdomen (P=0.0181) and back (P=0.0123). TEWL decreased significantly on the back after treatment with the spray (P=0.016), but not on the abdomen (P=0.078). Adverse effects were not observed. The results of this pilot study indicate that topical fatty acids and essential oils are a useful treatment option for canine atopic dermatitis.     

P-6 Double-blinded comparative study of the efficacy of azithromycin and cephalexin in canine pyoderma

Bacterial pyoderma is one of the most frequent skin diseases in dogs. Recurrent pyoderma is often secondary to an underlying skin disease, but no epidemiological study has been published on the subject to the authors' knowledge. This study was designed to prospectively evaluate the causes responsible for recurrent pyoderma in dogs. Dogs presenting with a history of more than three episodes of skin infection in the last year were included in the study. For each case, epidemiological and clinical data were collected. Pyoderma was confirmed by the clinical signs, the demonstration of bacteria on microscopic examination of cytological smears and a positive culture. Each animal was treated with an appropriate course of antibiotics until resolution of signs of pyoderma. Depending upon the presence of pruritus, appropriate diagnostic tests were performed: skin scrapings, acaricidal trial, flea treatment, elimination diet, intradermal testing, biopsies, endocrinological tests, leishmaniasis and ehrlichiosis serology, antinuclear antibody testing. Thirty dogs (14 males and 16 females) of 19 different breeds, aged from 1 to 12 years (mean 4.9 years) were included. Diagnosis was folliculitis (44%), folliculitis and furunculosis (20%), furunculosis (20%), cellulitis (10%). Staphylococcus intermedius was isolated in 97% of cases. The following underlying diseases were identified: atopic dermatitis (60%), food allergy (7%), flea allergy (7%), hypothyroidism (7%), hyperestrogenism (4%), demodicosis (4%), and zinc-responsive dermatosis (4%). In two dogs, no underlying cause could be identified. Atopic dermatitis is the most common disease associated with recurrent pyoderma in dogs.
Funding: Pfizer Animal Health.     

Immunoglobulin G responses in 21 dogs with skin diseases to antigens from different isolates of Staphylococcus intermedius

The aim of this study was to characterise the immunoglobulin G (IgG) response in 21 dogs with or without pyoderma to antigens from six isolates of Staphylococcus intermedius. The staphylococcal proteins were separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis, transferred electrophoretically on to a membrane and subjected to immunoblotting with the dogs' serum. Gels containing separated proteins from the six isolates revealed 29 to 33 distinct bands with molecular weights ranging from 20 to 230 kDa. All the dogs' sera contained IgG that recognised 12 to 24 bands (mean 17), regardless of whether the dogs had pyoderma. The recognised proteins had molecular weights ranging from 20 to 198 kDa but the majority had molecular weights below 75 kDa. The most intense band in all six isolates had a molecular weight of 28 to 29 kDa. The antibody responses to the six isolates were essentially similar except that there were significantly more bands in the response to isolate 2 than to isolate 6, and occasional differences in the intensity of individual bands. All 21 dogs mounted an IgG response to multiple antigens in S intermedius, which differed only marginally between the six isolates. This lack of variation provides evidence that the host's response to different isolates of S intermedius is not a major factor in canine pyoderma.     

Isolation of Staphylococcus schleiferi from healthy dogs and dogs with otitis, pyoderma, or both

OBJECTIVE: To determine the frequency of isolation and susceptibility patterns of Staphylococcus schleiferi from healthy dogs and dogs with otitis, pyoderma, or both that had or had not received antimicrobial treatment. DESIGN: Prospective study. ANIMALS: 50 dogs. PROCEDURE: Dogs were allocated to 1 of 4 groups: healthy dogs (n=13), dogs without otitis but with pyoderma (10), dogs with otitis but without pyoderma (11), and dogs with otitis and pyoderma (16). Bacteriologic culture of ear swab specimens was performed in all dogs. Bacteriologic culture of skin swab specimens was also performed in dogs with concurrent pyoderma. Isolates were identified as S schleiferi subsp schleiferi or S schleiferi subsp coagulans on the basis of growth and biochemical characteristics. RESULTS: S schleiferi was not isolated from any dogs with pyoderma only. Staphylococcus schleiferi subsp schleiferi was isolated from the ears of 2 healthy dogs, and the skin and ears of 2 dogs and the skin of 1 dog with otitis and pyoderma. Staphylococcus schleiferi subsp coagulans was isolated from the ears of 3 dogs with otitis only, and the ears of 6 dogs and the skin of 2 dogs with otitis and pyoderma. One of the S schleiferi subsp schleiferi isolates from ears, 2 of the S schleiferi subsp coagulans isolates from ears, and 1 of the S schleiferi subsp coagulans isolates from the skin were resistant to methicillin. One methicillin-resistant isolate from the ears and 1 from the skin were also resistant to fluoroquinolones. CONCLUSIONS AND CLINICAL RELEVANCE: S schleiferi subsp schleiferi was detected in healthy dogs and dogs with otitis and pyoderma. Methicillin-resistant and -susceptible S schleiferi subsp schleiferi and S schleiferi subsp coagulans were detected as the predominant organisms in dogs with otitis.     

Macroscopic,cytological and bacteriological evaluation of anal sac content in normal dogs and in dogs with selected dermatological diseases

Abstract Macroscopic and cytological aspects of anal sac content were evaluated in 40 normal dogs and 10 dogs each with pyoderma, Malassezia dermatitis associated with atopic dermatitis and uncomplicated atopic dermatitis. Bacteria isolated from anal sacs were compared with those from abdominal skin and hair in 20 normal dogs and 10 dogs with pyoderma. There was no difference between the groups in anal sac dimension, or in the colour, consistency or presence of granules in their content. Extracellular bacteria were found in higher numbers in diseased animals, whereas intracellular bacteria were observed in 40% of dogs with pyoderma and in only 2.5% of normal dogs. Malassezia spp. were present in 15.7% of dogs, with no difference between groups. Neutrophils were observed in 12.5% of normal dogs, 30% of dogs with Malassezia dermatitis with underlying atopic dermatitis and in 70 and 80% of dogs with pyoderma and uncomplicated atopic dermatitis, respectively. Seven bacterial species were isolated from anal sacs, with no difference between normal dogs and dogs with pyoderma. In five normal animals and in four dogs with pyoderma the same bacterial strains were isolated from anal sacs and from abdominal skin and hair.