Atypical Dermatophilus Congolensis Infection in a Three-Year-Old Pony

Abstract

Dermatophilosis caused by Dermatophilus congolensis causes exudative dermatitis in a variety of species. The infection is generally limited to the cutaneous tissues, where infection is acquired from the environment and bacteria penetrate into keratinized epithelium through epithelial disruption. A 3-year-old pony filly was examined for enlarging mandibular lymph nodes during the preceding 10 months. Biopsy of the node revealed mixed and granulomatous inflammation and thick, filamentous Gram-positive bacteria. Dermatophilus congolensis, confirmed by biochemical testing and sequencing of the ribosomal RNA gene, was cultured from a sample aspirated from the infected site. Dermatophilus congolensis should be considered as a possible etiologic agent associated with lymphadenopathy and granulomatous inflammation in the horse.

Dermatophilus congolensis is a Gram-positive coccobacillary actinomycete that causes an exudative dermatitis in a variety of species, most notably in ruminants and horses, although rare infections occur in cats, dogs, and humans. 1,2,12 Infection is generally restricted to the skin and subcutaneous tissue, although uncommon involvement of lymphoid tissue in ruminants and cats has been described. 3,4,12

A 3-year-old female palomino pony was referred to the William R. Pritchard Veterinary Teaching Hospital at the University of California, Davis (Davis, CA) with the primary complaint of progressively enlarging bilateral mandibular lymphadenopathy during the previous 10 months. Intermandibular swelling was noticed the winter of the previous year (10 months before referral), and a biopsy of the affected area revealed inflamed lymphoid tissue. Prior to referral, the first premolar teeth had erupted and were removed; significant hooks on cheek teeth were treated with dental floatation. The pony was also reported by the owner to be lethargic and appeared pruritic, as evidenced by biting at or rubbing her sides. The filly had been regularly vaccinated and had most recently been dewormed with pyrantel pamoate 3 weeks prior to referral.

The filly was slightly depressed on examination. Her temperature and heart rate were within normal limits. A slightly elevated respiratory rate of 25 breaths per minute was detected, although there were no abnormal lung sounds on thoracic auscultation. Large, firm, nonpainful multilobulated masses in the intermandibular space were palpable. The only other abnormalities on physical examination were small areas of superficial cutaneous excoriations over the flanks, thorax, and gluteal musculature. Dark, loose fecal material was observed after normal defecation.

A complete blood count was unremarkable, and plasma fibrinogen was 300 mg/dl. A serum biochemistry panel was within normal limits, with the exceptions of slightly elevated alkaline phosphatase (359 IU/l; reference [ref.] range: 86–285 IU/l) and triglycerides (0.56 mmol/l; ref. range: 0.02–0.46 mmol/l) and a slightly decreased direct bilirubin (2 mmol/l; ref. range: 3–10 μmol/l). The biochemical abnormalities were attributed to the age and breed.

A lymph node aspirate and biopsy were performed on the affected tissue; a small amount of inspissated pus was observed during the biopsy procedure. Histopathology of the lymph node revealed densely collagenous fibrovascular tissue with an absence of normal lymph node morphology. A mixed inflammatory infiltrate comprising abundant eosinophils, plump fibroblasts, and fewer plasma cells was observed. Histiocytes, multinucleated giant cells, and occasional clusters of neutrophils were present at the cell-rich periphery of the biopsy. Some histiocytes contained round, lightly eosinophilic, granular-to-globoid cytoplasmic material. A Brown and Brenn Gram stain demonstrated thick, long filamentous bacteria with divisions suggesting multicellularity. A Ziehl–Nielsen acid-fast stain failed to reveal acid-fast organisms.

The inspissated pus was submitted for bacterial isolation and identification. A direct Gram stain of the sample demonstrated numerous neutrophils, but no organisms were identified. Abscess material was inoculated onto 5% defibrinated sheep blood a and MacConkey a agars and incubated at 35°C in 5% CO₂. Anaerobic culture was performed using Brucella 5% sheep blood agar b and incubation at 35°C anaerobically. Growth was apparent at 48 hr on blood agar incubated aerobically; there was no growth anaerobically. Small numbers of pinpoint to 0.25-mm rough, dry, yellow, raised colonies that were adherent to the agar were noted, with complete hemolysis. With further incubation, differences in colony morphology were observed, with some colonies becoming more mucoid. Isolated colonies were subcultured to ensure purity after variation in colony morphology was observed; all appeared to be the same organism.

A Gram stain revealed Gram-positive coccobacilli and irregular Gram-positive rods, which were sometimes in thick, filamentous, multicellular arrangements. The organism was Kinyoun and Ziehl–Neelsen acid-fast negative. Motile zoospores were observed under microscopic examination of a wet preparation of an inoculated brain–heart infusion broth. These findings were most consistent with Dermatophilus sp., but additional biochemical and molecular testing was performed to confirm the identification. Tubed media c used for biochemical testing were incubated for at least 7 days before yielding a negative result, except gelatin, which was held 10 days before a positive result was observed. Phenotypic characterization showed the isolate to produce catalase and soluble hemolysin in blood broth; it was unable to reduce nitrate to nitrite or nitrogen gas. It also produced urease and digested gelatin and Loeffler coagulated horse serum. The isolate fermented glucose and maltose but not lactose, salicin, sucrose, mannitol, dulcitol, xylose, or sorbitol.

To confirm the isolate as D. congolensis, 16S ribosomal RNA (rRNA) gene sequencing was performed. 11 Partial sequences of the rRNA gene showed the current isolate had 98% and 99% identity to this region in the D. congolensis type strain DSM 44180 in the forward (nucleotides 376–930 of accession no. AJ243918) and reverse (nucleotides 665–1224) directions, respectively. No antimicrobial susceptibility testing was performed because the isolate was slow growing and no procedural or interpretive standards exist for this organism.

The skin over the enlarged lymph nodes was prepared as for surgery, blocked with a local infusion of 2% mepivacaine hydrochloride, d lanced for drainage, and flushed with 0.85% saline. The filly was treated with oral doxycycline (10 mg/kg) twice daily and wound flushing. During the next month, the drainage stopped, and the skin healed, although swelling in the region remained. Unfortunately, the filly was euthanized because of behavioral difficulties and was disposed without a necropsy; the persistence and/or extent of the Dermatophilus infection was not determined.

The majority of D. congolensis infections are cutaneous and characterized by severe encrustation, purulent discharge, and hair loss. Nearly all infections are confined to the epidermis. Zoospores penetrate the superficial epithelium to infect the deeper epidermis. Repeated cycles of bacterial penetration into the epithelium, neutrophilic infiltration, and development of parakeratotic keratinocytes above the stratum spinosum result in the typical crusts observed. 1

Although Dermatophilus certainly has pathogenic properties, disruption of the epidermis by arthropods, trauma, or moisture is thought to be an important factor leading to cutaneous lesions. 1,7,12 Interestingly, in the horse in the current study, no known trauma or other cutaneous lesions around the head or cranial cervical regions were observed on physical examination or reported by the owner. The superficial epidermal defects observed on physical examination were not typical of dermatophilosis and were most likely due to self-trauma associated with pruritis. Possible routes of infection in the pony were cutaneous trauma or oral mucosal penetration associated with the poor dental condition. Given the longstanding duration of lymph node enlargement, the exact route of infection in the current case could not be determined. It is possible that the isolation of D. congolensis resulted from superficial cutaneous contamination of the sample during biopsy; however, microscopic examination revealed the presence of bacteria within the inflamed lymphoid tissue.

Although cutaneous D. congolensis infections are not unusual, atypical dermatophilosis, characterized by granulomatous inflammation and localization in tissues other than the epithelium, has rarely been described in horses. 9,10 In a case series reported on horses in Uganda, dermatophilosis and abscessation of lymph nodes were associated with severe tick infestation. 10 However, in the same report from Uganda, many cutaneous lesions were present, and identification of D. congolensis was made solely on the basis of microscopic morphology on a Giemsa-stained cutaneous crust; culture of a lymph node failed to isolate Dermatophilus. The only other report of noncutaneous dermatophilosis in a horse was in an aborted fetus. 9 The route of infection in the case of the fetus was most likely ascension via the cervix.

Atypical dermatophilosis has been described in several species, including sheep, cattle, deer, and a cat. 3,4,6,8 Frequently, this manifestation is associated with cutaneous lesions, suggesting that the initial site of infection was the skin, and involvement of deeper organs, such as the subcutis, lymph nodes, kidney, or liver resulted from local extension, lymphatic drainage, or hematogenous spread. 3,4,6,8

Host age, genetic factors, and immunosuppression related to tick infestation have been associated with increased susceptibility to dermatophilosis in ruminants. 1,5,7 The filly in the current report had no indications of immunocompromise, and she was not an Arabian, which is the horse breed associated with severe combined immunodeficiency. However, no specific testing was performed to evaluate immune function.

In summary, to the authors' knowledge, this is the first report of isolation, confirmed by phenotypic testing and sequencing, of D. congolensis from the lymph node of a horse. Dermatophilus should be considered a possible etiologic agent associated with granulomatous inflammation in the horse even if no cutaneous lesions are observed.