Cryptococcus gattii: Clinical Importance and Emergence in North America
- First Online:
- Cite this article as:
- Lockhart, S.R. & Harris, J. Curr Fungal Infect Rep (2010) 4: 151. doi:10.1007/s12281-010-0021-y
- 55 Views
Cryptococcus gattii is an emerging fungal pathogen in the Pacific Northwest of North America, where it has caused more than 50 human infections since its emergence in 2004. Among residents of British Columbia, where the disease emerged in 1999 on Vancouver Island, many infections have occurred in immunocompetent persons. The cause for the emergence is currently unknown. The pathogenic profile of Cryptococcus gattii in North American patients appears to be different from that seen previously for C. gattii and from the profile of infection among patients with Cryptococcus neoformans. Treatment duration and the need for patient follow-up may be different between patients infected with C. gattii and C. neoformans. For this reason, physicians treating atypical patients with Cryptococcal spp infection, particularly HIV-uninfected patients, should obtain a travel history and obtain a species identity for Cryptococcus isolates.
KeywordsCryptococcusCryptococcus gattiiCryptococcus neoformansFungal epidemiologyPacific NorthwestFungal infectionCryptococcoma
The genus Cryptococcus contains approximately 37 species of basidiomycetous yeast, of which Cryptococcus neoformans—comprising the two variants Cryptococcus neoformans var. grubii and Cryptococcus neoformans var. neoformans—and Cryptococcus gattii cause the vast majority of human and animal infections. Although the species are closely related genetically, they occupy different environmental niches , affect different subpopulations, and appear to cause distinct (although similar) disease courses [2–4]. Cryptococcus neoformans, primarily an opportunistic infection among immunocompromised persons [3, 4], is distributed globally in association with soil contaminated with pigeon guano and decaying organic matter . C. neoformans is well known as a devastating opportunistic infection among AIDS patients. The incidence of C. neoformans infections surged during the 1980s and early 1990s, but with the introduction of highly active antiretroviral therapy in the 1990s, a decrease in the incidence of disease occurred among patients in industrialized countries [5, 6]. Currently, it is estimated that 600,000 persons with HIV around the world succumb each year to C. neoformans disease [7•].
C. gattii is a much less common cause of cryptococcal infection globally. The clinical presentation of C. gattii in immunocompromised individuals is very similar to that of C. neoformans . However, C. gattii often infects immunocompetent persons and is associated with more pulmonary and central nervous system (CNS) cryptococcomas than C. neoformans [2, 3, 9••]. Historically, C. gattii has been associated with decaying organic matter and eucalyptus trees in tropical and subtropical regions . During the past decade, however, C. gattii has emerged as a primary pathogen of humans and animals in the temperate climatic zones of British Columbia (BC) and the US Pacific Northwest (USPNW) [9••, 10], including Washington, Oregon, California, and Idaho [10, 11•].
Emergence in North America
Emergence on Vancouver Island
Vancouver Island (VI), in southwestern BC (just north of Washington State), has the mildest climate in Canada, with warm, dry summers and mild, wet winters during which temperatures rarely dip below freezing. In 1999, C. gattii, an etiologic agent previously virtually unknown to this region, began appearing in both animals and humans living on the southeast coast of VI and among mainland BC residents who had had contact with this region [12–14]. By 2007, 218 human cases of C. gattii had been documented in BC, mostly among residents of or visitors to VI [9••]. Infections have also been identified in a variety of mammalian species, including companion animals, wildlife, and aquatic mammals [12, 14].
Three molecular subtypes of C. gattii have been identified in BC: VGI, VGIIa, and VGIIb. Between 90% and 95% of clinical and environmental isolates on VI are represented by the subtype VGIIa, often referred to as the “major strain” because of its predominance in the North American outbreak . The VI major strain is genetically indistinguishable from a 1975 sputum sample isolate from a Seattle, Washington patient (travel history for this patient is unknown) and a 1992 isolate from a eucalyptus tree in San Francisco, California [15, 16]. A minority of infections in North America are caused by the related subtype VGIIb (the “minor strain”) or by the subtype VGI [15, 16]. The VI minor strain VGIIb is genetically indistinguishable from several isolates from Australia and may have been imported to North America directly, possibly through international trade in trees or other products . The origin of strain VGIIa is unknown, but there is speculation that it represents meiotic progeny of strain VGIIb . To date, VGIIa has been found only in North America .
In 2003, C. gattii became a reportable disease in BC . Beginning in 2004, the first case-patients without exposure to VI or other C. gattii–endemic areas were reported in BC, suggesting that the pathogen was spreading geographically. Most of these patients were infected with C. gattii subtype VGIIa .
Emergence in the US Pacific Northwest
The first contemporary case of C. gattii infection in the USPNW occurred in 2004, in a 59-year-old Oregon man who had undergone kidney transplantation in 2003 . This patient also had occupation-associated lung tissue scarring, which may have contributed to his infection risk; his isolate was genotyped as VGIIa. The second case involved an 87-year-old man with a history of chronic lymphocytic leukemia and recent oral steroid use . Interestingly, his isolate represented a novel strain, designated VGIIc, not detected anywhere else in the world [19•]. To date, this strain has been found only in Oregon and Washington patients (and in one patient who visited Oregon 11•); however, early genetic analyses identified all VGIIc isolates as VGIIb, and because not all of the VGIIb isolates from BC have undergone testing to distinguish between the two types, some VGIIc isolates may also exist in BC [10, 11•, 19•]. Neither of the first two US patients had any contact with VI or any other known C. gattii–endemic area.
The first instance of human C. gattii infection in Washington State was detected in early 2006 in a 74-year-old man living on Orcas Island in the Puget Sound. This patient had a history of large granular lymphocytic leukemia and oral steroid use, and presented with a cough and a lingular cryptococcal nodule . This patient’s isolate was found to be C. gattii subtype VGIIa.
In the US, C. gattii infections are not formally reportable in any state, although Washington State requests reports of C. gattii as a “rare disease of public health significance” (http://www.doh.wa.gov/notify/nc/cryptococcus.htm). Reports of animal infections have been requested by the state from veterinary laboratories in Washington since 2006, and in Oregon since January 2009. Cats, dogs, ferrets, llama, sheep, camelids, elk, horses, and porpoises have all been found with the infection in Washington, Oregon, and California (C. gattii Public Health Working Group, unpublished). It is thought that animal infections may precede the emergence of human infections in a given geographic region [12, 21, 22] and could act as sentinels for subsequent emergence of human infections.
Both longitudinal and cross-sectional environmental studies have been carried out since 2001 on VI and in mainland BC and the USPNW. Since 2001, samples from air, soil, lake water, river water, seawater, and several types of trees have yielded C. gattii isolates from all regions . The most heavily colonized areas appear to be the Coastal Douglas Fir biogeoclimatic zone in BC and the analogous climatic zones in the USPNW [13, 20, 23, 24]. Some spots appear to be only transiently positive for C. gattii spores, but others have consistently yielded positive cultures over months or years, suggesting permanent colonization in these areas . Wheel wells of vehicles and human shoes have also been shown to be capable of mechanical transport of C. gattii spores, and spores have been found at boat launch sites at public parks, in wood chips, and in the air around tree-felling areas, suggesting multiple mechanisms of dispersal . Most environmental isolates found to date belong to the VGIIa subtype, but isolates of subtypes VGIIb and VGI have also been recovered [10, 23, 24]. Despite being represented in patient samples in the United States [11•], subtype VGIII has not yet been reported from environmental samples [11•, 19•].
Epidemiology of Infection
The epidemiology of the BC emergence has been reported recently, with a cumulative case count of 218 [9••]. The average annual incidence of C. gattii on VI between 1999 and 2007 was 25.1 per million population, much higher than the 8.5 per million annual rate seen in the endemic Northern Territory of Australia or the annual rate of 0.8 to 1.1 per million seen in the Australian states with a temperate climate similar to that of VI . In all of BC, the average annual incidence of C. gattii infection was 5.8 per million population [9••]. The mean age among BC patients was 59 years (range 2–92). Only 3 of the 218 reported cases occurred in persons less than 18 years old, and 2 of these occurred in children taking inhaled corticosteroids. Most patients (77%) sought treatment for respiratory syndrome, with the most commonly reported symptoms being cough, dyspnea, and chest pain. A minority (8%) sought treatment for CNS symptoms, with headache, night sweats, weight loss, anorexia, and neck stiffness as the most common symptoms. The remaining patients had both respiratory and CNS symptoms (10%) or other disease (6%). Approximately 7% were asymptomatic. Chest radiographs were abnormal in 54% of case-patients; 75% of these patients had single or multiple lung nodules. The case-fatality rate was 8.7%. Thirty-eight percent were considered immunocompromised, including being HIV-infected or having a history of invasive cancer, organ transplantation, or corticosteroid use during the past year [9••]. The incubation range (2–13 months; average, 6–7 months) was calculated using patients who had a known single exposure to VI .
At least some clinical aspects of the BC emergence appear to differ from the typical presentation of C. gattii infection. Although other reports have suggested that CNS syndrome is the most common cause for seeking treatment for C. gattii infection [2, 26], only 8% of patients in BC sought treatment for a CNS syndrome [9••]. Epidemiologic studies from Australia suggest that between 66% and 98% of patients with C. gattii infection are immunocompetent [2–4], but in the BC emergence, only 62% of the C. gattii patients could be considered immunocompetent, indicating that differences may exist between the patient populations in Australia and North America, or between the Australian and North American C. gattii subtypes.
Beyond the differing underlying patient populations susceptible to C. gattii and C. neoformans infections, other pathogen-specific differences appear to include a greater propensity among C. gattii–infected patients to develop cryptococcomas and a slower response to antifungal treatment among C. gattii–infected patients than among C. neoformans–infected patients [3, 4].
Laboratory Identification and Typing
Several methods exist for laboratory identification of cryptococcal infection. The cryptococcal antigen test (carried out with serum or CSF), India Ink staining of CSF, and mucicarmine staining of fixed tissue are the most common diagnostic tests for cryptococcal infection. Plating on DOPA-containing agar (Niger, birdseed, Staib, or 3,4-dihydroxyphenylalanine) allows melanin production in both C. neoformans and C. gattii, causing colonies to turn brown and allowing easy differentiation of these two species from other yeast species. None of these tests, however, permits differentiation between C. neoformans and C. gattii. To definitively distinguish C. gattii from C. neoformans, canavanine glycine bromothymol blue (CGB) agar or molecular techniques must be used [27, 28]. On CGB agar, a pH-mediated colorimetric change to the agar occurs in the presence of C. gattii but not C. neoformans. Although some non-gattii Cryptococcus species can also cause this colorimetric change, most will not produce melanin on DOPA agar or grow at 37°C. It must be cautioned that rare C. gattii isolates will not cause a colorimetric change on CGB agar, nor will they produce melanin on DOPA agar; thus, these methods are not foolproof for identifying C. gattii isolates (S. Lockhart, J. Peterson, unpublished).
In the age of molecular diagnostics, genotyping has replaced serotyping (the historical method of definitively distinguishing C. gattii from C. neoformans) as the gold standard for cryptococcal strain identification and differentiation. Two methods have been used extensively for typing: amplified fragment length polymorphism (AFLP)  and multilocus sequence typing (MLST) [16, 30••]. MLST is easy to perform and reproducible between laboratories, and the resulting sequences can be archived [30••]. MLST is currently being used by public health laboratories in BC and the United States to differentiate isolates in both the BC and USPNW emergences of C. gattii [L. Huang and M. Morshed, personal communication; 11•]. Through the work of laboratories in Australia, the Netherlands, and the United States, a universal nomenclature has been established for strain identification [30••, 31].
IDSA Guidelines and Recommendations for C. gattii
The most recent guidelines from the Infectious Disease Society of America (IDSA) for the management of cryptococcal disease provide specific recommendations for the treatment of C. gattii infection [32•]. For patients presenting with pulmonary disease, the treatment recommendations are identical to those for C. neoformans infections. However, for patients presenting with CNS or disseminated C. gattii disease, the guidelines recommend more vigilance for cryptococcomas and hydrocephalus than is normally recommended for patients with C. neoformans infections. Because C. gattii is more likely than C. neoformans to induce hydrocephalus, the placement of a ventriculoperitoneal shunt may be required in addition to antifungal therapy [32•].
Antifungal Resistance in C. gattii
The IDSA guidelines recommend against routine drug susceptibility testing of Cryptococcus isolates because clinical resistance is not typically seen, and interpretive breakpoint levels defining resistance have not been established [32•]. However, during the recent emergence in the USPNW, elevated minimum inhibitory concentration (MIC) values to all of the azole antifungal agents have been specifically linked to C. gattii genotypes VGIIb and VGIIc [11•] when compared with MIC values of VGIIa, VGI, and VGIII isolates. Data from BC indicate that infection with subtype VGIIb is associated with a worse outcome than infection with subtype VGIIa [9••]; to date, there are no published outcome data for subtype VGIIc infections. In light of these data, it may be prudent to perform MIC testing or genetic subtyping of isolates from patients before assigning them to maintenance therapy, which can last as long as 18 months.
Several reports have documented C. gattii cryptococcosis in visitors to BC and the USPNW [33–38]. In February 2007, the first known case of C. gattii infection in North Carolina was reported . The patient was a 46-year-old healthy, HIV-negative man who presented to his physician with a thigh mass. A CT scan revealed an additional mass in his lungs. Upon surgical removal of the thigh lesion, identified as Cryptococcus (unspeciated), he was discharged from the hospital receiving 400 mg per day of fluconazole and appeared to be recovering. However, in May 2007, the patient reported to a local hospital with tonic-clonic seizures. An MRI revealed the presence of a large brain mass. Biopsy, serotyping, and subtyping of cells from the mass revealed it to be C. gattii, subtype VGI. The only plausible exposure to C. gattii for this patient was a trip to San Francisco, California, the previous September and October. The genotype of the patient’s isolate was found to be indistinguishable from a 1992 C. gattii isolate from a liver-transplant patient in San Francisco .
In addition to the North Carolina patient, there have been at least three cases of C. gattii infection in European tourists who visited VI and later presented in their home country with C. gattii cryptococcosis with the VI-specific VGIIa genotype [33, 36, 38]. Two of the European patients had no predisposing risk factors to infection; one had an 8-month history of corticosteroid use . These situations indicate that travel can be an important component of patient history when cryptococcal infection is suspected in an atypical patient. Interestingly, these infections helped define the known extremes of incubation for C. gattii infection, as one patient became ill 6 weeks after exposure  and another became ill 13 months after exposure . It is recommended that clinicians determine a travel history for patients with symptoms of cryptococcal infection.
Why Does C. gattii Infect Immunocompetent Individuals?
One of the more intriguing questions surrounding C. gattii infection is why the immune profile of infected patients differs from that of patients infected with C. neoformans. There are some suggestions that C. gattii does not elicit the same adaptive T helper (Th)1-type immune response required to respond to infection as does C. neoformans. Previous work has shown that culture filtrates of C. gattii antigens inhibited neutrophil migration both in vitro and in vivo [39, 40], and in a 2009 study, Cheng and colleagues [41•] showed that multiple different strains of C. gattii failed to elicit a protective tumor necrosis factor (TNF)-α and interferon (IFN)-γ cytokine response during murine infection. In the same study, C. gattii strains either failed to provoke or prevented migration of neutrophils to the site of infection, thus eliminating one of the primary initial cellular responses. This lack of response was not seen in the C. neoformans strain tested concurrently in the same murine model.
Clinical data from humans also suggest differences in the cytokine response between C. gattii and C. neoformans infection. Two separate reports show that levels of TNF-α, IFN-γ, and interleukin (IL)-6 in HIV-negative patients with C. gattii meningitis were reduced when compared with levels in HIV-positive patients with C. neoformans infection [42, 43]. Another intriguing study showed that the kinetics of complement component C3 binding to the cryptococcal capsule was much slower for C. gattii strains than for C. neoformans strains, which may enhance the pathogen’s evasion of the immune response . Although these findings are intriguing, it is still not completely clear why more immunocompetent individuals get C. gattii infection than get C. neoformans infection.
Why This Strain, and Why Now?
There has been some suggestion that the VI C. gattii major strain, VGIIa, is more virulent than other C. gattii strains. Two independent studies have demonstrated that VGIIa strains from VI are more virulent in a mouse model of infection than other strains of C. gattii [16, 41•]. In addition, a recent study by Ma and colleagues  showed that VGIIa strains proliferated in phagocytes more rapidly than non-VGIIa strains. This increased proliferation rate may be linked to a change in mitochondrial morphology and cellular gene expression of the C. gattii cell, which allows C. gattii to adapt better to the intracellular environment of the phagocyte . Interestingly, Galanis and coworkers [9••] found that patients infected with a VGIIb subtype isolate were more likely to die of their infection than a patient with VGIIa or VGI infection, although they state that this difference may also reflect the increased ages of the patients with VGIIb infections. This result is in direct contrast to the results of Fraser and coworkers , who found that a VGIIb isolate was much less virulent than a VGIIa isolate in a murine model. This difference in outcome highlights the difficulty in comparing a murine model with human infection and may imply that meaningful variation exists within subtypes.
Increased virulence could explain the predominance of the VGIIa strain over other strains in this outbreak, but it does not explain why C. gattii is emerging in the USPNW now. Has a recent climatic change permitted proliferation of already-existing niches of C. gattii in the region? Alternatively, perhaps it has now been “seeded” into the area with sufficient frequency through international travel and traffic that it has been able to establish a niche where one could not be maintained before. Perhaps the answer is a combination of several scenarios. Recent genetic changes in North American strains of C. gattii in combination with climatic changes in the USPNW could also have allowed this dormant pathogen to adapt and proliferate in a new niche and infect a new and naïve population of residents.
Evidence demonstrates that the emergence of C. gattii is not likely to be the result of improved surveillance or patient testing. In 2008, the British Columbia Centre for Disease Control performed a retrospective study of all Cryptococcus spp clinical isolates collected in BC between 1987 and 2000. For isolates collected before 1999, 31 of 36 were not C. gattii, but 3 of 5 isolates collected in 1999 and 2000 were C. gattii . In addition, a review of all Cryptococcus cases identified in BC between 1995 and 2004 showed that, beginning in 1999, the number of infections with Cryptococcus spp identified on VI increased sharply, whereas the number on the BC mainland remained steady . In the United States, retrospective review of 49 Cryptococcus cases identified at two large tertiary care hospitals in Seattle, Washington, between January 1997 and December 2004 did not identify any cases of C. gattii in this period . Taken together, these data suggest that C. gattii is likely to be truly emergent in North America and that the outbreak is not the result of increased surveillance or testing practices.
C. gattii cryptococcosis is an emergent and sometimes fatal disease in North America, most recently in the Pacific Northwest region of the United States. Most of our previous clinical information about C. gattii infection came from experience outside of North America. As noted above, the clinical experience in the North American outbreak is somewhat different from previous experience with C. gattii infections, and there is still much to learn about the clinical management of this disease. Because many clinical labs lack the ability to culture or speciate this emerging pathogen, it is likely that C. gattii infections are sometimes misdiagnosed as C. neoformans. Physicians treating immunocompetent or HIV-negative patients with cryptococcal infections, either in the USPNW or among patients who have traveled to the USPNW or BC, should consider submitting the patient’s isolate to their state public health laboratory for species identification.
The findings and conclusions of this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
The authors wish to thank members of the C. gattii Public Health Working Group (Emilio E. DeBess and Dawn Daly, Oregon Department of Human Services; Ron Wohrle, Nicola Marsden-Haug, Marcia Goldoft, and Cyndi Free, Washington State Department of Health; Ben Sun and Duc Vugia, California Department of Public Health; Tom Chiller, Naureen Iqbal, and Joyce Peterson, Centers for Disease Control and Prevention; Linda Hoang and Muhammad Morshed, BC Centers for Disease Control; and Karen Bartlett, University of British Columbia) for thoughtful discussions, and Mary Brandt and Tom Chiller for critical reading of the manuscript.
No potential conflicts of interest relevant to this article were reported.