MAIN : Feline Medicine : Feline Cryptococcosis
Welcome, VIN Public ! Logout
<< Previous
Next >>
Feline Cryptococcosis
World Small Animal Veterinary Association World Congress Proceedings, 2003
Richard Malik, BVSc, DipVetAn, MVetClinStud, PhD, FACVSc, MASM
Post Graduate Foundation in Veterinary Science, The University of Sydney
NSW, Australia

Taxonomy, Aetiology and Biology

Cryptococcosis is an important disease of man and animals and the most common systemic mycosis of cats. The infection is thought to be acquired from the environment, with no reported cases of disease transmission from one affected animal to another. Thus, cryptococcosis is not a contagious or anthropozoonotic disease.

Cryptococcosis is most commonly caused by two species of the genus Cryptococcus, C neoformans and C bacillisporus. The ability of C neoformans and C bacillisporus to grow at 37oC may in part explain their pathogenicity, because other members of the genus grow poorly at this temperature. Elaboration of a polysaccharide capsule and the enzymes laccase and phospholipase by C neoformans and C bacillisporus are additional virulence factors that contribute to pathogenicity. C neoformans and C bacillisporus are dimorphic, basidiomycetous fungi. They exist in animal tissues as the yeast form (Cryptococcus spp), but are capable of transforming under special laboratory conditions into a filamentous form (Filobasidiella spp). Thus far, the filamentous phase has been demonstrated only under strictly controlled laboratory conditions, but this perfect state is likely to exist in certain natural environments. The importance of the perfect state is that spores resulting from sexual or asexual filamentous reproduction likely represent the infectious propagules that give rise to mammalian disease.

In animal tissues, Cryptococcus neoformans and C bacillisporus exists as a round, yeast-like organism, with a variably-sized polysaccharide capsule as its distinguishing feature. The capsule provides protection from environmental insults (e.g., desiccation) and the phagocytic response of the host. In tissues, Cryptococcus reproduces by forming one or two daughter cells (buds) that are connected to the parent cell by a narrow isthmus. Buds may break off when small and thus the cell population varies in size.

Unlike other dimorphic fungi, the yeast phase of Cryptococcus is found under routine laboratory conditions and in infected tissues. Cryptococcus has a worldwide distribution and, in addition to people, infects a variety of domestic and native mammals. In contrast to the other systemic mycoses, the prevalence of cryptococcosis in cats exceeds that in dogs by an order of magnitude. Historically, five serotypes (A, B, C, D, AD) have been identified on the basis of antigenic differences in capsular polysaccharide. Recent advances in the taxonomy of the genus Cryptococcus have led to a new nomenclature that was proposed at the 5th International Conference on Cryptococcus and cryptococcosis. C neoformans and C bacillisporus differ biochemically, genetically, ecologically and epidemiologically. C neoformans has a world wide distribution, while C bacillisporus is largely restricted to tropical and subtropical climates. C neoformans can be divided into two varieties based on serotyping, C n var grubii and C n var neoformans. Both varieties are strongly associated with disease in immunocompromised human patients, although the same may not be true for companion animals. C n var grubii is by far the most common isolate from cryptococcosis in people and animals worldwide, although C bacillisporus is important in certain geographical regions such as Australia, Papua New Guinea, South East Asia and Central Africa. There is strong evidence that several of Australian eucalyptus trees provide a natural environmental niche for C bacillisporus. Interestingly, koalas seem capable of amplifying the number of cryptococci in certain environments. The definitive environmental niche for C neoformans has not been determined, although there is a strong 'historic' association with weathered bird (especially pigeon) guano and more recent evidence for growth in decaying plant matter in hollows of certain trees. The organism passes through the gut of pigeons, but systemic infection of pigeons is extremely rare. Perhaps the pigeon's high body temperature protects it from infection. Pigeon guano provide an alkaline, hyperosmolar environment that is rich in many nitrogen-containing compounds including creatinine that favour cryptococcal growth. Cryptococci may remain viable for at least two-years in accumulations of pigeon guano protected from drying or sunlight; pigeon lofts provide such an environment.

Most basidiomycetes reproduce sexually in their natural environment, and the teleomorphs of C neoformans (Filobasidiella neoformans) and C bacillisporus (F bacillisporus) can be induced to undergo sexual reproduction in the laboratory and produce dikaryotic hyphae, blastoconidia, basidia and basidiospores. The recent documentation of both α and a-mating types of C bacillisporus in E camaldulensis trees suggests that this may occur in nature. However, recent work has suggested that C neoformans may be evolving into an asexual fungus and that basidiospores may result from haploid (monkaryotic) fruiting as well as by sexual recombination. In either case, the notion that the basidiospore is the infectious propagule for Cryptococcus is attractive, as this stage is suited to dispersal by air currents and has physical properties that favor penetration into the respiratory system, thereby facilitating primary infection of mammalian hosts.


The exact mode of infection is unproven, but the most likely route is via inhalation of air-borne organisms. These may be basidiospores or yeast cells desiccated by environmental exposure. Shrunken, poorly capsulated cryptococci that are small enough for alveolar deposition have been isolated from pigeon guano and soil. Although human patients with cryptococcosis typically present with neurological signs referable to meningoencephalitis, there is strong circumstantial evidence that the infection starts in the lungs and subsequently spreads to the nervous system haematogenously via macrophages. Respiratory involvement usually does not result in symptoms, although lesions can be detected in chest radiographs, thoracic computed tomography scans or at necropsy. The small particle size of infectious propagules is said to be the reason the lung is primary site of infection, as only very small particles are capable of penetrating deep into the lower respiratory tract.

In cats, dogs, koalas and psittacine birds, the nasal cavity is usually the primary site of infection. The reason(s) for this difference (compared to humans) are a matter of conjecture, but possibly the increased development of the nasal passages in animals, with more efficient filtering of small particles, may provide part of the explanation. It is our belief that most cases of feline cryptococcosis begin as mycotic rhinitis following asymptomatic colonization of the nasal cavity. C n var grubii has been shown to be a transient coloniser of the nasal mucus of cats, dogs and koalas in Australia, while C bacillisporus can be isolated in sufficient numbers and with sufficient frequency to be actually be considered part of sinonasal normal flora of koalas in certain environments. Studies in koalas have shown that self-limiting, subclinical infection is common, with limited invasion, granuloma formation and successful eradication or containment of organisms.

When infection ensues, clinical signs of rostral nasal cavity disease such as sneezing, epistaxis and nasal discharge are conspicuous, and sometimes granulomatous protuberances can be seen at the nares. In some cases, destruction of adjacent facial bones facilitates spread of infection to contiguous regions, such as the bridge and side of the nose, the planum nasale or hard palate. When facial distortion develops, the clinical presentation is strongly suggestive of either fungal rhinosinusitis or nasal neoplasia. On the other hand, when infection begins in the caudal portion of the nasal cavity, signs of mycotic rhinitis may be subtle or absent, although it is possible to confirm the sinonasal region as the primary site of infection using cytology, culture, endoscopy or cross-sectional imaging. In some of cases, infection spreads through the cribriform plate into the olfactory bulbs and olfactory tract, giving rise to meningoencephalitis. In these cases, the anatomical proximity of the optic nerves frequently results in concurrent cryptococcal optic neuritis, and secondary retinitis. Clinically, this is manifest as widely dilated pupils that respond poorly to light, swelling of the optic disc and focal retinal hemorrhage. In other cases, caudal nasal cavity involvement gives rise to a mass lesion which occludes one or both choanae, resulting nasopharyngeal signs viz. stertor, snoring, dyspnea or open mouth breathing. In occasional cases the infection spreads to the middle ear via the auditory tube.

Cutaneous involvement, if multifocal, reflects hematogenous dissemination from the primary site of infection, as do lesions in bone (e.g., digits) or periarticular soft tissues. In ferrets, and the exceptional cat, localized cutaneous cryptococcosis can develop following penetrating injury of the skin. In some cats, infection spreads to the mandibular lymph nodes, presumably via the lymphatics from the nasal cavity. Rarely, mandibular lymphadenomegaly can be massive and require surgical debridement. Occasionally salivary gland infection has been documented, although how organisms reach this site is a mystery.

In eastern Australia, about 20-30% of human cryptococcosis cases are caused by C bacillisporus, and a similar proportion is observed in cats. There is a tendency for animals in rural environments to be infected with C bacillosporus, presumably due to increased exposure to eucalyptus material, and all infections recorded in koalas have been attributable to C bacillisporus. The development of granulomatous intracranial or pulmonary mass lesions (cryptococcomas), are strongly associated with immunocompetence. Thus, in human patients, cryptococcomas are more typical of C bacillisporus infections, whereas infections in immunodeficient patients typically result in meningitis with little involvement of the brain parenchyma.

In cats, infections with feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV) have been thought to predispose to cryptococcosis. In North America, there is evidence that some FeLV-positive cats develop cryptococcosis as a result of immune dysfunction because these cats are slower to respond or fail to respond to treatment, and are much more likely to suffer relapses. In Australia, FeLV-positive cats with cryptococcosis are exceedingly rare, probably because of the very low prevalence of persistent FeLV infection in the cat population. Although there are several reports of one or two cats with cryptococcosis allegedly secondary to feline immunodeficiency virus (FIV) infection, large studies in Australia have failed to produce convincing evidence that cryptococcosis is a feline AIDS-defining infection. Rather, it is considered that co-infection usually reflects the high prevalence of FIV infection in Australia. Leukocyte and lymphocyte subset numbers in FIV-positive and FIV-negative cats with cryptococcosis were not different and a positive FIV status did not impart an unfavourable prognosis. Indeed, many FIV-positive cats with cryptococcosis can be cured and do not relapse despite cessation of therapy. Cryptococcosis has very rarely been reported in cats receiving immunosuppressive therapy or chemotherapy for malignancy. Thus, underlying diseases are typically not detected in cats with cryptococcosis, and factors predisposing to infections remain elusive. The authors have treated two FIV-negative cats with cryptococcosis which subsequently developed malignant lymphoma and although no relationship was established between the two conditions; a similar association has been the subject of a previous report. We have also seen two FIV-positive cats with cryptococcosis who then developed lymphoma and mast cell neoplasia, respectively, subsequent to other opportunistic infections; in these cats, long-standing FIV infection may have predisposed to both cryptococcosis and the terminal malignancies. Genetic factors may be involved in the predisposition towards development of cryptococcosis, as Siamese, Birman and Ragdoll cats are significantly over represented amongst reported cases.

Clinical findings

Cryptococcosis is the most common of the systemic mycoses of cats. There is no gender predisposition and the age range of affected cats is broad, although young adult cats (two to three years old) appear at increased risk. It is likely that exposure and self-limiting infection occurs in the first few years of life, with disease in older cats reflecting reactivation of viable cryptococci in residual granulomatous foci. Upper respiratory tract signs are most common and include sneezing; snuffling and mucopurulent, serous or haemorrhagic, unilateral or bilateral, nasal discharge. Signs are usually chronic and in some cases, a polyp-like mass is evident in the nostril. In others, a firm to fluctuant subcutaneous swelling over the bridge of the nose is present. Cats with nasopharyngeal cryptococcosis develop stertor, inspiratory dyspnea, a tendency to open mouth breathing and sometimes secondary otitis media. Often mandibular lymphadenomegaly is evident, and ulcerated or proliferative lesions in the oral cavity are seen occasionally. Lower respiratory tract signs are rare in cats. Thoracic radiographs are usually normal, although small nodular lesions may be present. Rarely, pulmonary nodules, a mediastinal granuloma or cryptococcal pleurisy will be found.

Neurologic signs associated with cryptococcosis in cats are variable and include depression, changes in temperament, bizarre behaviour, seizures, circling, head pressing, ataxia, paresis, head tilt and other vestibular signs, anosmia and blindness. These signs may occur alone or in association with other physical findings and may result from the presence of mass-like cryptococcoma(s) or from meningoencephomyelitis. In people, cerebral cryptococcosis is often associated with increased intracranial pressure and the same is likely also in cats.

Ocular abnormalities occur in some affected cats and almost always are a marker for central nervous system (CNS) involvement. The most common sign is peripheral blindness with dilated, unresponsive pupils due to optic neuritis, exudative retinal detachment and granulomatous chorioretinitis.

Cutaneous lesions are common in cats with cryptococcosis, principally due to secondary involvement of the planum nasale. Very rarely, localized cutaneous cryptococcosis can develop following inoculation of propagules, e.g., after a cat scratch. Multifocal skin lesions are the result of haematogenous dissemination and consist of papules and nodules that are fluctuant to firm and range from 1 to 25 mm in diameter. Larger lesions tend to ulcerate, leaving a raw surface with a serous exudate. Cats with disseminated disease also may have enlargement of one or more lymph nodes, with cryptococci evident in aspirates from affected nodes. Fever is uncommon in affected cats. Cryptococcosis is typically a chronic infection, causing listlessness and weight loss due to poor appetite. Other reported signs include peripheral lymphadenomegaly (unassociated with skin lesions), bone lysis, swollen digits, chronic cough, and renal failure due to kidney involvement.


Specimen collection

In most of the aforementioned conditions, definitive diagnosis is straightforward, based on obtaining representative tissue specimens for cytology, culture and sometimes histology. Suitable specimens may include deep nasal swabs, nasal washings, needle aspirates from cutaneous nodules or enlarged lymph nodes, bronchoalveolar lavage specimens, pleural fluid and cerebrospinal fluid (CSF). A definitive diagnosis of cryptococcosis is defined as the culture and identification of the organism by a reputable laboratory. However, it is possible to obtain a high index of suspicion of cryptococcosis by demonstrating characteristic capsulate, narrow-necked budding yeasts in cytological smears or histology sections.

Cytological examination

Romanowsky-type stains (DiffQuik, Geimsa and Wright's), new methylene blue and Gram's stain are all satisfactory for making a cytological diagnosis. DiffQuik is simple and reliable. With Gram's stain the organism retains the crystal violet, whereas the capsule stains lightly pink with safranin. The organism is often most easily visualised at low power. India ink has been used historically to examine CSF for cryptococci, which appear unstained and silhouetted against a black background. In our hands, however, Indian ink is not as helpful as other stains because lymphocytes and fat droplets are easily confused with the organism. We have more success demonstrating yeasts in CSF using cytocentrifuged preparations stained with DiffQuik, and culture, than with Indian ink preparations. Cytologic examination can also be performed on crush preparations of biopsy samples.

Cytologic examination of nasal or cutaneous exudates, masses, CSF, or ocular fluid demonstrate organisms in a majority of cases, although culture is sometimes a more sensitive test. Urine sediment should be evaluated because many dogs and the occasional cat have subclinical renal infection. In human patients, blood culture is often useful. Although cytologic examination is a rapid test, negative results do not eliminate the possibility of cryptococcosis from consideration. Unencapsulated organisms may be overlooked unless a large number are present, and where a sufficient index of suspicion exists, additional diagnostic tests are indicated.

Serologic Testing

The detection of cryptococcal capsular antigen by the latex agglutination procedure is the most widely utilized serological test and is very useful in a veterinary setting. A variety of methodologies exist. The test used in our laboratory detects polysaccharide antigen using latex particles coated with anti-cryptococcal antibodies. Current tests detect all known serotypes and can be used for serum or CSF. They provide a rapid diagnostic method in suspected cases where organisms have not been visualized or cultured. In human studies, commercial kits typically demonstrate 90% to 100% sensitivity and 97% to 100% specificity. Results have been similar for animals. The sensitivity and specificity of this test has been improved by pretreatment of test specimens with pronase (a proteinase). This digestion step is included with some test kits, but should be performed routinely even when the enzyme is not supplied by the manufacturer.

Testing for cryptococcal antigen in serum may prove advantageous in a patient for which CSF collection may be an unacceptable risk, or when a diagnosis of cryptococcosis is unlikely, but should be excluded. For example, we recommend all cats with intracranial signs are tested for cryptococcal antigen prior to CSF collection or brain imaging. This is because we have seen a number of canine and feline patients whose neurological status deteriorated markedly following CSF collection, despite the use of appropriate anaesthetic regimens. This is hardly surprising, considering that many cases have intracranial mass lesions or increased CSF pressure, with the attendant risk of brain herniation following a cisternal tap.

Antigen titers can be extremely high (>65,536) in cats with disseminated disease, but even a titer of 1 is considered a positive result. Titer results using different kits can vary considerably. Thus, the methodology should not be altered when monitoring the response to treatment. When monitored in a consistent fashion, titers have proven useful in evaluating the progress of patients during therapy. A good prognosis is indicated by a decrease in antigen titer, whereas a persistent titer after treatment suggests continued infection. Reductions in the titer typically lag behind clinical improvement. There is no correlation between pre-treatment antigen titer and outcome.

In human patients, measurement of the antigen titer in CSF is often more sensitive than cytology or culture and is therefore theoretically preferable to serum in animals with neurologic signs. Our experience, however, is that most cases with CNS cryptococcosis have positive antigen titers in serum, presumable because sufficient antigen is elaborated at the primary site of infection. Cryptococcal antigen can also be detected in other body fluids such as pleural fluid or bronchoalveolar lavage fluid.

Although cats, dogs and koalas all make anti-capsular antibodies in response to infection, subclinical disease with seroconversion is common. As titer cutoffs have not been determined with certainty, antibody determinations are at present a research tool, rather than a clinical tool.

Tissue biopsy

Because of the rapidity of cytologic evaluation, impression smears or KOH preparations should always be made from suspect biopsy samples. If no organisms are seen, part of the sample can be used for culture, and the rest processed for routine histology. In haematoxylin and eosin (H&E) stained sections, the organism stains as a faint, eosinophilic, round-to-oval body surrounded by a clear halo (the unstained capsule). The organism is more easily visualized with periodic acid Schiff, methenamine silver, or Masson-Fontana stain, but the capsule still does not stain. Mayer's mucicarmine is the definitive stain, because the cryptococcal capsule takes on a rose-red colour and the organism appears pink against a blue background. Other fungi with similar morphologic features do not stain with this method. The large capsule and the thin cell wall of Cryptococcus differentiate it from Blastomyces. Its budding and lack of endospores distinguish it from Coccidioides. Some C neoformans strains, however, have poorly developed capsules. When doubt exists, and only formalin-fixed tissues are available, immunohistology can be used to differentiate cryptococci from other fungi and furthermore identify the species and serotype. This technique is also very useful for examination of archived material.


The organism can be easily cultured from aspirates, exudate, CSF, urine and tissue specimens. Although C neoformans will grow on almost all laboratory media, Sabouraud's dextrose agar is preferred when fungi are considered in the differential diagnosis. Standard Sabouraud's agar is optimal when culturing a normally sterile site such as CSF, as antibiotics included in the media may inhibit the growth of some cryptococcal strains. When sampling a site normally contaminated by bacteria, e.g., the nasal cavity, inclusion of antibiotics in the medium improves the chances for isolating cryptococci. Growth of Cryptococcus is inhibited by media containing cycloheximide. Culture should be performed at both 25o and 37oC. Colonies become visible within two-days to six-weeks, typically two to three days. The organism forms white, creamy colonies that yellow with age, are mucoid with heavily capsulated strains and dry when the capsule is poorly developed. C bacillisporus colonies are typically much more mucoid than C neoformans colonies. Characteristics employed to identify the organism include its morphology (presence of a capsule, narrow-necked budding), growth at 37 oC, hydolysis of urea, brown-colour-effect on bird seed agar, growth and colonial appearance on canavanine glycine bromothymol blue agar and response to various assimilation tests (available as commercial kits). Cultures of CSF are usually positive in cases with CNS cryptococcosis. Thus, fungal culture is recommended whenever CSF is collected from patients with inflammatory CNS disease, even when the organism cannot be demonstrated cytologically or serologically.

Bird seed agar containing antibiotics can be useful in the diagnosis of cryptococcosis, especially when sampling sites expected to be heavily contaminated, such as nasal exudate. It is also a wonderful media for environmental studies. Normally, bacteria and other fungi outgrow C neoformans on plates; birdseed agar containing antibiotics overcomes this problem by suppressing growth of contaminants, while identifying colonies of Cryptococcus from other yeasts and filamentous fungi by the brown-colour-effect produced in the agar.

Fungal susceptibility testing using disc diffusion, Etest strips, or broth microdilution provides useful information concerning therapy. Although not as accurate as bacterial in vitro susceptibility testing, it is our impression that strains having a low minimum inhibitory concentration for a given drug in vitro, tend to be susceptible to the drug in vivo, although the converse is not invariably true.


A lesion associated with cryptococcosis can vary from a gelatinous mass, consisting almost exclusively of organisms, to a well-ordered cell-mediated immune response resulting in granuloma formation. The gelatinous appearance is a reflection of the large amount of capsular polysaccharide present in lesions. The primary cellular response is composed of macrophages and giant cells, with a few plasma cells and lymphocytes. Generally, the presence of a granulomatous response reflects an immune-competent host, whereas little or absent response suggests immunodeficiency. An exception is the centre of large lesions, where even in normal hosts there is often limited response to organisms.

In cats that die or are euthanized, granulomatous rhinitis is usually evident, and in rare instances the lungs are affected also. Cerebral cryptococcosis consists of either primary meningoencephalitis or a cerebral granuloma(s). Optic neuritis is a common in cats that die of cerebral cryptococcosis, and in some cats the olfactory bulbs are replaced by a gelatinous mass of yeast cells. Other affected organs are skin and subcutaneous tissues, kidneys and lymph nodes that drain infected areas. Renal granulomas have been found in some cases with disseminated disease, as have lesions in the spleen, adrenal glands, thyroid glands, and liver.


The prognosis for most cats with cryptococcosis is good to excellent, given diligent co-operative owners prepared to dose patients for many months and pay for the costs of medication and monitoring. Animals with long-standing extensive disease have a less favorable prognosis than patients diagnosed early with mild signs of disease, although even long-standing severe cases can be cured. Patients with neurological signs always have a guarded prognosis, although many (perhaps two out of three) can be treated successfully using combination therapy including amphotericin B. Two additional considerations apply to patients with cryptococcal meningoencephalitis. Firstly, neurological status often deteriorates soon after starting therapy, presumably because death of cryptococci and the resulting inflammation give rise to a dangerous increase in intracranial pressure. Secondly, neurological deficits (such as blindness and gait abnormalities may or may not persist following successful therapy.

Cats in which a positive FeLV status is confirmed on two consecutive occasions have a poor long-term outlook and drug therapy for cryptococcosis in these cases should be considered palliative. In contrast, many FIV-positive cats can be cured, although some need a prolonged course of treatment, and recurrence or the development of other clinical problems remain possible.


One of the reasons for renewed veterinary interest in cryptococcosis was the availability of a new generation of triazole antifungal drugs (fluconazole, itraconazole, voriconazole) suitable for long term oral administration and the development of new ways to use the old drug amphotericin B more easily and safely. Indeed, the main problem in treating cryptococcosis cases is currently the high cost of therapy, the requirement for continued hospital visits and medication of pets on a regular basis for a protracted period. Treatment regimens for cats with cryptococcosis will be discussed.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Richard Malik, BVSc, DipVetAn, MVetClinStud, PhD, FACVSc, MASM
Post Graduate Foundation in Veterinary Science
The University of Sydney
NSW, Australia

MAIN : Feline Medicine : Feline Cryptococcosis

Copyright 1991- World Small Animal Veterinary Association World Congress Proceedings, 2003