FELINE PANNICULITIS DUE TO RAPIDLY-GROWING MYCOBACTERIA
Rapidly growing mycobacteria (RGM) are a heterogeneous group of organisms that produce colonies on synthetic media within seven-days when cultured at 24oC to 45oC. They are distributed ubiquitously in nature and can be isolated from soil, dirt and bodies of water (including tap water). RGM include the M fortuitum group (including M fortuitum, M peregrinum and the 3rd biovariant complex), the M chelonae/abscessus group (including M chelonae and M abscessus), the M smegmatis group (including M smegmatis sensu stricto, M goodii and M wolinkskyi) and a variety of other species. The taxonomy of this group has been revised recently and because of this, the word 'group' is used when referring to isolates recorded in early publications. RGM are strongly linked with localized infections of immunocompetent hosts. This is because they are well adapted to a saprophytic existence and inherently have low virulence for mammals. Thus, they do not produce disease unless a breakdown in normal defence barriers provides them with a portal of entry to a favourable tissue environment. Once introduced, RGM are generally constrained by a vigorous immunological response that may or may not eradicate them from the tissues, but is effective enough to prevent haematogenous or lymphatic spread. RGM can produce widely disseminated disease, but only in severely immunocompromised individuals.
Mycobacterial panniculitis refers to a syndrome characterized by chronic infection of the subcutis and skin with RGM. This condition is quite common in cats, especially in Australia, and a series of 49 cases has been reported. RGM replicate in mammalian tissues when introduced through some breach in the skin. This typically follows penetrating injury, especially when the wound is contaminated by dirt or soil. Preference of RGM for fat is a key factor in the pathogenesis of these infections and results in a tendency for disease to occur in obese individuals and in tissues rich in lipid, such as the subcutaneous panniculus and especially the inguinal fat pad. Experimental infections cannot be induced in cats that do not have appreciable subcutaneous fat depots. Adipose tissue offers a favourable environment for survival and proliferation of RGM by providing triglycerides for growth of organisms or protecting them from the phagocytic and immune responses of the host.
Initial reports suggested mycobacterial panniculitis was more common in warm humid climates, however cats from temperate regions, including parts of Australia, Canada, Finland and Germany, have subsequently been reported. In Australia, the M smegmatis group accounts for the majority of feline cases, whereas it is a much less common cause of equivalent infections in human patients.
Clinical signs
Infections tend to start in the inguinal region, usually following environmental contamination of cat fight injuries, e.g., raking wounds inflicted with the hind claws. The infection may spread to contiguous subcutaneous tissues of the ventral and lateral abdominal wall and perineum. Penetrating injury by sticks, metallic objects and vehicular trauma may also give rise to these infections, as can cat and dog bite injuries contaminated with soil or dirt.
Sometimes infections start in the axillae, flanks or dorsum and spread into adjacent tissues.
Early in their course, infections resemble cat fight abscesses, but without the characteristic foetid odour and turbid pus. Instead, a circumscribed plaque or nodule is apparent at the site of injury. Later, there is progressive thickening of the nearby subcutis to which overlying skin becomes adherent. Affected areas become denuded of hair and numerous punctate fistulae appear, discharging watery exudate. Fistulae are intermingled with focal purple depressions (thinning of the epidermis over accumulations of pus). The 'lesion' gradually increases in area and depth, and may eventually involve the entire ventral abdomen, adjacent flanks or limbs. If cats are presented promptly for veterinary attention and the lesion confused with an anaerobic cat-bite abscess, surgical drainage and administration of a µ-lactam is typically followed by wound breakdown and development of a non-healing suppurating tract surrounded by indurated granulation tissue. Some affected cats with infections develop constitutional signs; they become depressed, pyrexic, inappetent, lose weight and are reluctant to move. Occasional cats develop the hypercalcaemia of granulomatous disease, although this is rarely symptomatic. Surprisingly, other cats remain comparatively well despite extensive disease. Usually the problem remains localized to the skin and subcutis. Although adjacent structures such as the abdominal wall can be affected eventually, spread to internal organs or lymph nodes is very unusual.
Diagnosis
Sample collection, Cytology and Histology
A tentative diagnosis of mycobacteriosis can be confirmed by collection of pus or deep tissue specimens. This material is used to confirm the diagnosis using appropriately stained cytology preparations, histological sections and mycobacterial culture. A histological diagnosis is unnecessary if appropriate samples for cytology and culture have been procured. It is vital to give the laboratory warning that a mycobacterial aetiology is suspected so special procedures for processing can be adopted.
In our experience, samples of pus obtained from needle aspirates of affected tissues through intact skin provide the best laboratory specimens. This material can be obtained from a palpably abnormal portion of the subcutis. The overlying skin should be carefully disinfected with 70% ethanol prior to obtaining the specimen to preclude the isolation of saprophytic mycobacteria from the skin surface. It may be necessary to carefully move the needle in the subcutaneous space, while applying constant negative pressure, until a pocket of purulent material is encountered. Aspirated purulent fluid should be submitted for cytology and mycobacterial culture, or inoculated immediately into a commercially-prepared mycobacteria culture bottle which is subsequently submitted to the laboratory. It is only necessary to suck a small amount of liquid material into the hub of the syringe. It is easiest to submit the entire syringe to the laboratory after replacing the needle with a sterile cover. Exudate from draining sinus tracts is heavily contaminated secondary invaders and represents an inferior sample. If deep biopsies are obtained, they should be triturated in brain heart infusion broth using a sterile mortar and pestle to produce a tissue homogenate suitable for cytology and culture.
Smears prepared from aspirates or tissue homogenates should be stained using Diff Quik, Burke's modification of the Gram stain and a modified acid-fast procedure (decolorizing with 5% sulphuric acid for only three to five minutes; RGM are not as acid-fast as other mycobacteria). Cytology invariably demonstrates pyogranulomatous inflammation and it is generally possible to visualize Gram positive and/or acid fast bacilli (AFB) in smears, although an exhaustive search of several smears is sometimes required.
Histologically, there is pyogranulomatous inflammation of subcutaneous adipose tissue, overlying dermis and underlying abdominal fascia and musculature. AFB may be hard or impossible to find in Ziehl-Neelsen stained tissue sections and are often located in lipid vacuoles.
Bacteriology and antimicrobial susceptibility testing
Tissue homogenates and pus should be streaked onto duplicate 5% sheep blood agar plates and a mycobacterial medium such as Lowenstein-Jensen medium or 1% Ogawa egg yolk medium and incubated aerobically at 37oC and 25oC. If available, the BACTEC system can also be utilised. Moderate to heavy growth of pin-point, non-hemolytic colonies is usually detected after two to three days (occasionally longer) on sheep blood agar at 37oC. Where only contaminated specimens are available, tissue homogenates can be treated with 4% sodium hydroxide followed by neutralization with dilute hydrochloric acid prior to inoculation onto media. Another method which can be used to differentiate RGM from contaminant flora is by primary isolation around antibiotic sensitivity discs (first generation cephalosporins or isoxazolyl penicillins) applied to the plate after inoculation.
There is great value in determining species identification and susceptibility data in every case, as this has a big impact on antimicrobial drug strategies. Species identification can be carried out in a well equipped veterinary bacteriology laboratory although it if often more convenient to send the strain to a Mycobacteria Reference Laboratory following primary isolation. Identification takes into account a number of phenotypic and biochemical features. Minimum inhibitory concentrations (MICs) for ciprofloxacin, gentamicin, trimethoprim, clarithromycin and doxycycline can be determined easily using the Etest (AB Biodisk, Solna, Sweden) method. This methodology is less demanding than the 'gold standard' of broth microdilution. Antimicrobial susceptibility of clinical isolates can also be determined using disc diffusion methodology.
Therapy
The management of mycobacterial panniculitis continues to evolve over time according to evolving clinical experience, availability of new antimicrobial agents and development of new surgical techniques. There is great variation in the severity and extent of lesions from patient to patient. Difficulty in making a prompt diagnosis is partly responsible for the chronicity, severity and refractoriness of these infections. Briefly, treatment should commence with oral antimicrobial(s) (doxycycline, a fluoroquinolone and/or clarithromycin), initially chosen empirically, but subsequently on the basis of in vitro susceptibility data. Sometimes long-term administration of such an agent or agents is sufficient to effect a cure, but in more severe cases it is eventually necessary to surgically resect recalcitrant tissues so that oral antimicrobial therapy will be able to cure the infection permanently. Given the extent and severity of the pathology in many of these cases, it is understandable that adequate levels of antimicrobials may not be achieved throughout all affected tissues and that in these cases the best chance for a successful outcome is to remove as much infected tissue as possible following preliminary antimicrobial therapy. Residual foci of infection can then be targeted by the high concentrations of antibiotics achieved during and after surgery. Peri- and post-operative antimicrobial therapy is vital to ensure primary intention healing of the surgical incision.
FELINE LEPROSY SYNDROMES
History
The term feline leprosy is used to refer to a mycobacterial disease in which single or multiple granulomas form in the skin or subcutis in association with large numbers of acid-fast bacilli (AFB) which are nonculturable using standard methods. The condition was first recorded in the literature by Australian and New Zealand researchers in the early 1960s. Since then, the disease has been reported in Western Canada, the Netherlands, France, the UK and USA.
Feline leprosy is more common in certain geographical locations (North Island of New Zealand, the Netherlands and British Columbia). Further, it appears to be more prevalent in temperate coastal areas and port cities, as opposed to inland or tropical habitats. Historically, the causative agent of feline leprosy was purported to be Mycobacterium lepraemurium. This bacterium causes murine leprosy, a systemic mycobacterial infection of rats. Cats are thought to contract M lepraemurium following bite injuries from infected rodents.
M lepraemurium is a fastidious, slow-growing mycobacterial species which, with difficulty, can be cultured from large inoculae on Ogawa's egg yolk medium under strictly controlled conditions, or in enriched liquid medium at a critical pH. Although a few investigators have successfully grown M lepraemurium from infected cats, the basis of ascribing this bacterium as the etiological agent of feline leprosy was dependent on transmission studies and the results of delayed hypersensitivity reactions to intradermally-injected tissue extracts. Several groups were able to show that material obtained from feline lesions can be used to transmit disease to rats, and subsequently back to cats. In such studies, the incubation period varied from two months to one year or more. Interestingly, some cats appeared much more susceptible to infection than others.
According to the literature, cats with feline leprosy are typically young adults (< 5 years-of-age), perhaps with a preponderance of males. Presumably these patient characteristics reflect the need for the cat to interact with a rat to become infected. The initial lesion is a focal granuloma of the subcutis. Owners become aware of solitary, or more commonly multiple, painless, raised, fleshy, tumor-like lesions, from a few millimeters up to 4 cm in diameter. These granulomas are freely movable over underlying tissues. Lesions can develop rapidly and when large, may ulcerate. Infection spreads to adjacent areas and may invade underlying tissues and drain to regional lymph nodes. Lesions can occur anywhere, but tend to be concentrated on the head and limbs. Small lesions are occasionally found on the tongue, lips and nasal plane. Lesions, even if multiple, tend to be initially concentrated in one region and have the propensity to recur following excision.
Pathologically, feline leprosy was subdivided into lepromatous or tuberculoid forms on the basis of the number of AFB present (multibacillary v paucibacillary) and the host immunological response (lepromatous v tuberculoid). Because the causal mycobacteria are slow-growing organisms capable of intracellular survival, the histologic picture depends on the host's immune response. When this response is poor, lepromatous (multibacillary) disease develops with infiltration of the dermis with large sheets of 'incompetent' foamy macrophages containing enormous numbers of organisms. AFB are usually arranged in the cytoplasm of macrophages as dense parallel accumulations which displace the nucleus to an eccentric position. Lymphoid cells and plasma cells are virtually absent from the lesions. If the host's immune response is more effective, histiocytic cells are accompanied by moderate numbers of lymphoid cells and plasma cells and multiplication of the organism is limited--the so-called tuberculoid response. The tuberculoid form accounts for perhaps two-thirds of the cases in Western Canada, a large proportion of cases in New Zealand and the Netherlands, but a minority of the cases encountered in Australia. Invasion of local nerves, a prominent feature of human leprosy, is rarely observed in patients with feline leprosy, although a recent report described a cat without skin lesions which presented for mycobacterial infiltration of one sciatic nerve.
AFB in smears and tissue sections appear as long slender rods. In smears stained with Romanowsky stains such as DiffQuik or Geimsa, organisms appear as negative-staining bacilli. In smears or sections stained with modified acid-fast stains such as Ziehl-Neelsen (ZN) or Fite's stain, organisms take up the carbol fuschin and are acid/alcohol fast.
Molecular insights
Molecular methodologies were recently used to investigate presumptive feline leprosy. Of eight cases of invasive or disseminated cutaneous mycobacterial disease investigated by Hughes and colleagues using material collected largely from New Zealand cats, four were shown to have M lepraemurium infections. Of the remaining cases, one cat had a disseminated M avium infection, the aetiology in one cat was undetermined and in two cases infection was attributable to a novel mycobacterial species. This information encouraged a reappraisal of Australian feline leprosy cases. Interestingly, cats could be divided into two groups on the basis of the patients' age, lesion histology, clinical course and sequence of 16S rRNA PCR amplicons obtained from lesions.
One group consisted of young cats (typically < 4-years) which initially developed localised nodular disease affecting the limbs. Lesions progressed rapidly and sometimes ulcerated. Sparse to moderate numbers of AFB were identified using cytology or histology, typically in areas of caseous necrosis and surrounded by pyogranulomatous tuberculoid inflammation. Organisms did not stain with haematoxylin and ranged from 2 to 6µm (usually 2 to 4 µm). M lepraemurium was diagnosed based on the sequence of a 446 bp fragment encompassing the V2 and V3 hypervariable regions amplified from lesions using PCR and universal primers.
A second group consisted of old cats (>9-years) with generalised nodular skin lesions associated with multibacillary lepromatous histology. Some cats initially had localised disease which subsequently became widespread, while others had generalised disease from the outset. Disease progression was protracted, typically taking months to years, and skin nodules did not ulcerate. Microscopically, lesions consisted of sheets of epithelioid macrophages containing large to enormous numbers of AFB 2 to 8 µm (mostly 4 to 6µm) which stained also with haematoxylin. A single unique sequence spanning a 557 bp fragment of the 16S rRNA gene was identified in lesions from these patients. The sequence was characterised by a long helix 18 in the V3 region, suggesting the new species was likely to be a fastidious, slow-grower. The 16S rRNA sequence had greatest nucleotide identity with M leprae, M haemophilum and M malmoense, and contained an additional 'A' nucleotide at position 105 (the only other mycobacterial database sequence with the same extra nucleotide being M leprae). A very slow, pure growth of a mycobacterium species was observed on Lowenstein-Jensen medium (supplemented with iron) and semisolid agar in one case. The environmental niche of this new mycobacterium species has yet to be determined, although the preponderance of cases from rural or semi-rural areas suggests it is a saprophyte found more commonly in these locations than in metropolitan environments.
The establishment or spread of infection with the novel mycobacterial species suggests a requirement for decreased immunological surveillance to permit the development of disease with an organism of limited virulence. Furthermore, the aetiopathogenesis needs to account for the absence of young cats amongst this cohort of patients. The presence of a foamy histiocytic infiltrate of the dermis and subcutis in human patients with mycobacteriosis is observed almost exclusively in association with profound immunodeficiency such as that seen with terminal HIV infection. Feline leprosy caused by the novel mycobacterial species may likewise represent a manifestation of deteriorating immune competence in elderly cats with longstanding FIV infection. Decreased cellular immunity associated with renal disease may also predispose cats to infection, as renal disease is common amongst infected cats. Alternately, renal disease may occur as a consequence of the mycobacterial infection, as it does in rats with disseminated M lepraemurium infection.
These findings suggest that feline leprosy comprises two clinical syndromes, one tending to occur in young cats caused by M lepraemurium and another in immunosupressed elderly cats caused by a single novel mycobacterium species. To make matters even more complex, recent work by Appleyard & Clark has demonstrated a third mycobacterial syndrome in cats from western Canada and the USA (Idaho and Oregon) called 'feline multisystemic granulomatous mycobacteriosis'. This disease is caused by a slow-growing taxa provisionally called M visibilis that give rise to diffuse (rather than nodular) cutaneous disease and widespread dissemination to multiple internal organs. Sequence analyses demonstrate a number of nucleotide differences between M visibilis and both M lepraemurium and the novel species reported by Hughes and colleagues.
Diagnosis
Diagnosis of the 'feline leprosy' syndromes is usually straightforward, provided that the clinician has a high index of suspicion for the condition. Needle aspirates, crush preparations of biopsy material and histological sections stained with ZN or similar methods contain easily demonstrable AFB surrounded by variable granulomatous to pyogranulomatous inflammation. In DiffQuik stained smears mycobacteria can be recognized by their characteristic negative staining appearance and location within macrophages and giant cells.
Material should be submitted also for culture, because occasionally slowly-growing species such as MAC and M geneavense and the tubercle bacillus (M bovis or M microti) can produce an identical clinical presentation; in such cases optimal antimycobacterial therapy can be selected more readily on the basis of in vitro susceptibility results and information available in the literature. In the majority of cases, however, conventional mycobacterial culture is negative due to the fastidious nature of the causal organisms and a mycobacterial aetiology can only be proven using molecular techniques such as PCR amplification and nucleotide sequence determination of gene fragments. PCR has the additional advantage of providing a rapid diagnosis. Fresh (frozen) tissue delivered to a mycobacterium laboratory with PCR facilities provides the optimal sample, although freeze-dried specimens may be more conveniently sent where tissues need to travel long distances. Sometimes PCR can be performed successfully on formalin-fixed paraffin-embedded material, although fixation conditions invariably cause some DNA degradation which may limit the success of the procedure. Recently, Hughes and colleagues have developed specific PCR assays to diagnose infections due to M lepraemurium and the novel species; furthermore, use of a simple restriction enzyme digest allows these assays to distinguish M visibilis strains also.
M lepraemurium infections have a number of distinguishing features that suggest this aetiology even where molecular testing is not practicable (see above). Although a young age at presentation supports the diagnosis of a M lepraemurium infection, age alone is not a reliable criterion, as we have recently diagnosed M lepraemurium in cats as old as 9 years.
Therapy
Too few cases with a documented aetiology have been reported to provide definitive treatment guidelines. Although M lepraemurium and the novel species can be cultured in vitro, it is currently not routine or reliable to isolate these organisms due to their slow growth and fastidious requirements. Determination of in vitro susceptibility data for individual isolates is therefore not possible.
Only limited experimental studies have been undertaken to determine effective drug therapy for M lepraemurium in vitro or in vivo and as yet we have limited data only for the novel mycobacterial species. Portaels and colleagues found the minimum inhibitory concentration for rifampicin of two strains of M lepraemurium to be 4 and 8 µg/mL, levels that should be just obtainable in vivo based on extrapolation from pharmacokinetic studies in humans and dogs. Other drugs shown to have activity against M lepraemurium in vitro include ansamycin compounds (rifabutin) and sulpha drugs. There is a good deal of evidence that clofazimine has efficacy in vivo, while it is likely that clarithromycin would be also be effective based on its wide spectrum of activity against slow-growing mycobacterial species.
The literature suggests that when M lepraemurium infection is diagnosed early, while disease is localized, wide surgical excision of infected tissues provides the best chance to simply and rapidly effect a cure. Aggressive resection techniques should be adopted, with en bloc resection of all lesions, and reconstruction of resulting tissue deficits using appropriate surgical techniques. Such an approach should be combined with adjunct antimicrobial therapy beginning a few days prior to surgery, so that effective levels of drugs are present in blood and tissues intra- and postoperatively to ensure primary intention healing.
Clofazimine (at a dose of up to 10 mg/kg once daily orally; typically 25 to 50 mg every 24 to 48 hours) has the best reported success rate, although it is likely that combination therapy using two or more drugs will eventually prove superior. Drugs that could be combined with clofazimine include rifampicin and clarithromycin, although sulpha drugs, doxycycline, new fluoroquinolones such as gatifloxacin or amikacin may in time also prove to be useful.
In feline leprosy cases caused by the novel mycobacterium species, we believe combination therapy using two or three of clofazimine (25 to 50 mg per cat orally every day or every other day), clarithromycin (62.5 mg twice daily) or rifampicin (10 to 15 mg/kg per day) represents optimal therapy. However we are currently unsure of which will prove to be the best combination, and side-effects in individual cats may affect which two drugs are used in a given patient. Currently, we recommend a combination of rifampicin and clarithromycin as initial therapy.