Abstract
Sections of 14 skin biopsies of cats with plasmacytic pododermatitis and a clinical follow-up of 12–36 months were stained with a polyclonal anti-Mycobacterium bovis (Bacille Calmette-Guerin = BCG) antibody cross-reactive to a broad spectrum of fungi and bacteria. All sections were negative for organisms within the actual footpad tissue with the anti-BCG antibody stains. Polymerase chain reaction (PCR) assays that amplify the DNA of Bartonella spp., Ehrlichia spp., Anaplasma phagocytophilum, Chlamydophila felis, Mycoplasma spp., Toxoplasma gondii, and feline herpesvirus 1 (FHV-1) were applied to tissue digests. DNA of those pathogens assessed was not amplified from tissue.
Keywords
Introduction
Feline plasmacytic pododermatitis is characterized by soft and frequently painless swelling of multiple foot pads, typically with no other associated dermatitis. Ulceration and secondary infection, pain, lameness, and lymphadenopathy may occur. 3 Plasmacytic pododermatitis is diagnosed tentatively by history, clinical examination, and cytology; and the diagnosis is confirmed histopathologically. 13, 41, 43 The cause and pathogenesis of feline plasmacytic pododermatitis are unknown. Response to immunosuppressive therapy, particularly glucocorticoids at high doses and gold salts, 23, 29, 36 suggests an immune-mediated pathogenesis. However, other factors are probably involved, as the disease is localized and also responds to surgical intervention. 14, 44 Doxycycline monohydrate has been reported to produce partial or complete clinical remission in more than half of the cases. 3, 35 Tetracyclines have been shown to have immunomodulatory effects in various species. 2, 12, 26 They are used to treat immune-mediated skin disease in human 6 and veterinary medicine, 3, 34, 35, 42 and the reported response of feline plasmacytic pododermatitis to doxycycline is possibly due to these immunomodulatory effects. However, an infectious etiology has also been suggested for plasmacytic pododermatitis. Doxycycline has activity against a number of feline pathogens including Ehrlichia spp. Anaplasma phagocytophilum, some Mycobacteria spp., Mycoplasma spp., Chlamydophila felis, and Bartonella spp. 5, 7, 9, 22, 24, 28, 30 The purpose of this study was to evaluate biopsy specimens of feline plasmacytic pododermatitis for possible infectious organisms.
Materials and Methods
Skin biopsy specimens from cats sent to the Central Veterinary Diagnostic Laboratory in Melbourne, Australia, in a 24-month period were evaluated. Cats with a clinical presentation of plasmacytic pododermatitis (swollen, spongy foot pads) and consistent histopathologic criteria (a severe cellular infiltration in the dermis of the foot pad dominated by plasma cells) were included in the study. These 14 specimens came from a previous prospective study. 3
For immunohistochemical evaluation, mounted tissue sections were deparaffinized in xylene and then hydrated up to 70% alcohol. Quenching of endogenous peroxidase was performed by slide immersion in 0.3% H2O2/methanol for 30 minutes at room temperature (25°C). Slides were rinsed with a phosphate buffered saline (PBS) buffer/Tween 20. Antigen retrieval was achieved by slide immersions in 0.4% pepsin in pH 3.3 acidulated water for 15 minutes at 37°C. After antigen retrieval, 10% normal goat serum was applied for 20 minutes at 25°C to block nonspecific binding. Slides were stained with polyclonal anti-Mycobacterium bovis (Bacille Calmette-Guerin = BCG) antibody (Dako, Carpenteria, CA) using an streptavidin-horseradish peroxidase technique as previously described. 4 Counterstaining was achieved with Mayer's haematoxylin (Sigma, St. Louis, MO). For negative control tissues, PBS was substituted for the primary antibody. Positive control sections were bovine intestine infected with Mycobacterium paratuberculosis.
Approximately 25 mg of formalin-fixed tissue was sectioned from each skin specimen, using separate instruments to avoid potential DNA contamination. Tissue sections were placed into polymerase chain reaction (PCR)-ready, microcentrifuge tubes (Qiagen Inc., Valencia, CA), and DNA extraction was performed using a commercially available kit (DNeasy Tissue Kit, Qiagen Inc.). For all assays, water was digested as an internal negative control and assayed on each PCR assay run. The positive controls included DNA digests from fresh brain tissue from a Toxoplasma gondii–infected mouse and DNA of Chlamydophila felis, Ehrlichia canis, A phagocytophilum, Mycoplasma spp., Bartonella henselae, and feline herpesvirus 1 (FHV-1) grown in culture. The T. gondii PCR assay, 8, 18 Bartonella spp. PCR assay, 21 C. felis PCR assay, 39 Ehrlichia spp. PCR assay, 7, 25 A phagocytophilum assay, 7, 24, 25 and FHV-1 PCR assay 41 were performed as previously described. The amplified products were observed by electrophoresis in a 1.5% agarose gel containing ethidium bromide.
Results
Fourteen cats were included in the study. In all sections, plasma cells were the dominant cell type, and neutrophils, lymphocytes, and macrophages were present to a variable degree. All sections were negative for organisms within the footpad tissue with the anti-BCG antibody stains. Two sections contained a few cocci focally within the stratum corneum without associated epidermitis. None of the PCR assays were positive for DNA of the pathogens in tissue.
Discussion
All 14 cats included in this study had clinical and histologic evidence of classic plasmacytic pododermatitis and, so we believe, were an adequate sample set for study. 3, 13, 43 While many believe that the syndrome has an immune-mediated pathogenesis, infectious diseases have also been suspected. For example, a possible link between feline plasmacytic pododermatitis and concurrent feline immunodeficiency virus (FIV) infection was suggested in one study. 38 Further evidence was provided by a study showing a concurrent FIV infection in 50% of the cats with plasmacytic pododermatitis. 14 FIV testing was not performed in most of the cats included in the previously reported clinical part of this study; the 2 cats tested were positive and negative respectively. 3 In a more recent study, 9/9 cats were feline leukemia virus–negative; 5/9, FIV negative; and 4/4, Leishmania negative. 35
Bacterial and fungal cultures of affected tissue are recommended to rule out infectious granulomas in cats with pododermatitis and are typically negative. 37 However, in 2 recent studies of cats with plasmacytic pododermatitis, there was a clinical response to doxycycline monohydrate orally. 3, 35 Doxycycline has several effects on the immune system such as inhibition of chemotaxis of human polymorphonuclear neutrophils, 10 suppression of phagocytosis, 11, 31 IL-1 secretion of stimulated thymocytes, 19 and lymphocyte proliferation. 20 If feline plasmacytic pododermatitis is indeed an immune-mediated disease, as suggested by the response to glucocorticoids and gold salts, the reported response of affected cats to doxycycline may be due to these immunomodulating effects. However, a dermatitis that responds to local surgical intervention is unusual for a classic immune-mediated disease. In addition, doxycycline is an antibiotic agent effective against a number of infectious agents of cats including most hemoplasma spp., Ehrlichia spp., Bartonella spp., Chlamydophila spp., Mycoplasma spp., and Rickettsia spp. 27 Several of these agents are common in cats in some parts of the world and conceivably may be associated with this disease syndrome. These organisms all have special culture requirements and can be difficult to identify histologically and so could have been missed in previous studies that utilized routine techniques.
To the authors' knowledge, no studies attempting to identify infectious organisms with immunohistochemistry or PCR have so far been published. In this study, sections were stained with HE as well as polyclonal antibodies against BCG. BCG stain was chosen as a sensitive screening stain for micro-organisms, as it was found to be superior to other histochemical stains identifying Mycobacteria and Nocardia organisms and comparable to appropriate histochemical stains against Actinobacillus, Streptococcus, Staphylococcus, Dermatophilus, Blastomyces, Coccidioides, Histoplasma organisms, dermatophytes, and Malassezia and Sporothrix organisms. 4 For all 14 specimens, the stain with the anti-BCG antibody was negative, providing evidence against the presence of such microorganisms in the specimens and their involvement in the pathogenesis of the disease.
However, 2 alternatives need to be considered with this technique. First, the length of formalin fixation prior to processing the skin specimens may interfere with the ability of the antibody to label micro-organisms. 15 For optimal tissue staining, tissue fixation should not exceed 48 hours; longer fixation leads generally to decreased staining intensity due to loss of cellular antigens via cross-linking of reactive end groups of protein chains. 34 Although specimens were collected by daily courier and processed immediately after arrival in the laboratory, it is unknown if some were kept in formalin by referring veterinarians for some time before submission for histopathology. Second, it is possible that a microorganism is responsible for triggering the disease but is subsequently eliminated by the immune response. Influx of inflammatory cells and release of inflammatory mediators may continue beyond the elimination of the pathogen and so plasmacytic pododermatitis may represent an aberrant localized immune response to a pathogen trigger.
The organisms evaluated by PCR assay assessment were selected for several reasons. Bartonella spp., Ehrlichia spp., C felis, Mycoplasma spp., and potentially T gondii have clinical responses when tetracycline derivatives are administered. In addition, both T gondii 1 and FHV-1 16 are known to cause skin disease in cats. Last, Ehrlichia spp. infection is thought to cause vasculitis in cats, 7 and Bartonella spp. infects endothelial cells, commonly inducing skin lesions like bacillary angiomatosis in immune-suppressed people. 5 The failure to amplify DNA of these organisms from the tissue samples can be interpreted in several ways. It is most likely that none of the agents were present in the tissues and so were not likely to be involved in the pathogenesis of disease in these 14 cats. However, it is also possible that the organisms are not distributed equally throughout the tissues and so were not present in the sections digested for PCR assay amplification. In addition, while we previously showed that formalin fixation does not adversely affect the sensitivity of the T gondii PCR assay used here, 8 the same experiments have not been performed with the other PCR assays used. Thus, it is possible that the results of the PCR assays could have been falsely negative because of lack of sensitivity. It is also possible that an infectious agent is associated with the initiation of pododermatitis in cats and is cleared during the resultant inflammatory response, making it more difficult to detect. This has been noted in people with bartonellosis 32 and in sea turtles with herpesvirus-associated fibropapillomatosis. 17 Indeed, many of these patients showed clinical signs for months before histopathologic specimens were taken.
In future prospective studies of this syndrome in cats, it would be optimal to use fresh tissue samples and attempt to correlate histopathologic lesions to results of serum antibody tests and PCR assays performed on tissues and blood concurrently. In addition, all samples collected in this study were from Australia. Because of different distributions of some infectious agents around the world, tissues collected from cats in other countries should be assessed concurrently.
We believe the results of this study support the hypothesis that plasmacytic dermatitis in most cats is an idiopathic immune-mediated disease. However, because many cats respond favorably to doxycycline therapy and the numbers of cases assessed for infectious agents by special stains and PCR assays are small, it is recommended for clinicians to perform an antimicrobial trial with doxycycline monohydrate prior to administration of immunosuppressive agents.