Canada Communicable Disease Report

June 2008 - Volume 34 - Number 06

Monthly Report

Cryptococcus gattii Infections on Vancouver Island, British Columbia, Canada: Emergence of a Tropical Fungus in a Temperate Environment

M Fyfe, MD, MSc (1, 2), L MacDougall (MSc) (3, 4), M Romney, MD (5), M Starr, MBBS (6), M Pearce, MPH (1), S Mak, MSc (4), S Mithani, (7), P Kibsey, MD (1)

1. Vancouver Island Health Authority, Victoria, British Columbia, Canada
2. Department of Health Care and Epidemiology, University of British Columbia, Vancouver, British Columbia, Canada
3. Canadian Field Epidemiology Program, Population and Public Health Branch, Health Canada
4. Epidemiology Services, British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada
5. St. Paul’s Hospital, Vancouver, British Columbia, Canada
6. Infectious Diseases Unit, Royal Children’s Hospital, Melbourne, Victoria, Australia
7. Laboratory Services, British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada

Abstract

Background

Cryptococcus gattii causes disease among immunocompetent individuals in the tropics and subtropics. We document the appearance of C. gattii infections on Vancouver Island (VI), a temperate region, and discuss reasons for this emergence

Methods

Data on Cryptococcus hospitalizations for the calendar years 1995 through 2004 were reviewed. Viable historic isolates stored at the provincial public health laboratory between 1987 and 2000 were serotyped. Human cases were mapped by place of residence.

Results

Cryptococcosis among HIV negative individuals diagnosed on VI increased sharply after 1999. C. gattii was not detected in stored isolates prior to 1999. C. gattii cases lived in a specific biogeoclimatic zone on VI. Higher rates of illness were associated with exposure to the central region of VI.

Conclusions

The emergence of C. gattii in a temperate region is unprecedented. Clinicians should consider C. gattii in the differential diagnosis of individuals who travelled to certain areas in British Columbia.

Introduction

Cryptococcus neoformans and Cryptococcus gattii are encapsulated basidiomycetous yeasts and the causal agents of the majority of human and animal cryptococcosis. Inhalation may result in pulmonary infection⁽¹⁾, sometimes followed by dissemination to other body sites including the central nervous system⁽²⁾, skin⁽³⁾ and bone⁽⁴⁾. No outbreaks of cryptococcal disease in humans have been reported in the medical literature⁽⁵⁾.

C. neoformans typically causes disease in patients with compromised cell-mediated immunity, whereas C. gattii affects individuals with apparently normal immune systems^(6,7). C. neoformans has a worldwide distribution and is frequently recovered from soil enriched with the excreta of pigeons and other birds^(8,9). In contrast, C. gattii occurrence has been primarily limited to the tropics and subtropics⁽⁷⁾ and its natural reservoir appears to be plant debris, especially debris from eucalyptus trees^(10,11).

In June 2001, the Animal Health Centre (AHC) in British Columbia (B.C.) notified the BC Centre for Disease Control (BCCDC) of an increase in cryptococcosis occurring in companion animals (dogs, cats and a ferret) on Vancouver Island over the preceding year⁽¹²⁾. Laboratory physicians on Vancouver Island noticed a similar increase in the human population. Vancouver Island is the largest island on the Pacific coast of North America and has a population of 715,000 people. The climate is temperate, and endemic occurrence of C. gattii infections would therefore not be expected.

Eight initially available human clinical isolates were subtyped using CryptoCheck^® (Iatron Labs, Japan) and found to be C. gattii. Subsequent environmental sampling on Vancouver Island also isolated C. gattii from samples of soil, multiple tree species, and air⁽¹³⁾. Molecular subtyping determined that the majority of both clinical and environmental isolates were of a single, unique C. gattii genotype, and a smaller number were of a C. gattii genotype also found in Australia^(13,14). Fingerprinting of the mating locus found that the unique Vancouver Island genotype was likely produced by an unusual α-α sexual cycle⁽¹⁴⁾. By 2005, there was evidence of fungal spread from Vancouver Island to the B.C. Lower Mainland⁽¹⁵⁾.

Most cases presented with pulmonary infection. Common presenting symptoms included cough, dyspnea, chest pain and weight loss although some cases with pulmonary infection were asymptomatic⁽¹⁶⁾. Lung cryptococcomas, infiltrates and cavitary lesions were visible by x-ray. Cases presenting with CNS infection usually had meningitis with or without brain cryptococcomas. Common symptoms included headache, fever, night sweats and weight loss⁽¹⁶⁾.

This article documents the timing of C. gattii emergence on Vancouver Island through a retrospective review of hospital separations data and serotyping of historic isolates. Spatial characteristics of the emergence are explored through geographic mapping of human cases. Potential hypotheses for emergence are discussed.

Methods

Hospital Separations Review
Anonymized data on Cryptococcus hospitalizations from 1995 to 2004 were requested from the Discharge Abstract Database of the BC Ministry of Health. ICD codes do not distinguish between C. neoformans and C. gattii. Since C. neoformans is a common opportunistic pathogen in HIV+ve patients in BC, the analysis of hospital separations was restricted to HIV-ve individuals to increase the specificity. A hospitalization due to Cryptococcus was defined by the presence of any of the following ICD codes on an individual’s hospitalization record(s): ICD 9 = 117.5; ICD 10 = B45.0, B45.1, B45.2, B45.3, B45.7, B45.8, B45.9. Similarly, HIV infection was defined by the presence of any of the following ICD codes on an individual’s hospitalization record(s): ICD 9 = 042; ICD 10 = B20, B21, B22, B23, B24. Repeat hospitalizations for the same individual were removed. Records without a unique person ID were excluded as repeat hospitalizations could not be determined (n = 18). Year of hospitalization was assigned based on the earliest admission date. As patient address was not available, cases were stratified into Vancouver Island or B.C. Lower Mainland based on the location of the hospital(s) to which they were admitted. In the event that cases were admitted to hospitals on both the Island and the Mainland (n = 4), cases were assigned to Vancouver Island since it is more common for Vancouver Island patients to be transferred to the Mainland for care than the reverse.

Retrospective Laboratory Serotyping
Serotyping was conducted on all viable historic isolates stored at the provincial public health laboratory from individuals diagnosed between 1987 and 2000. Clinical specimens were cultured as previously described⁽¹⁵⁾. Isolates were serotyped using CryptoCheck^® (Iatron Laboratories, Japan). Serotypes A or D indicate C. neoformans infection and serotypes B or C, C. gattii.

Geographic Distribution
Cases for mapping were obtained from the BCCDC enhanced surveillance database on C. gattii, which includes individuals notified either retrospectively through microbiologists and via stored laboratory isolates, or prospectively once C. gattii became reportable in B.C. in 2001. A case was defined as a B.C. resident with either 1) culture-confirmed C. gattii infection or 2) laboratory evidence of infection from antigen detection, histopathology or microscopy in an HIV-negative person who had been to Vancouver Island in the year prior to diagnosis. ArcView 3.2 and ArcGIS 9.1(Environmental Systems Research Institute, Redlands, CA) were used to plot human cases of cryptococcosis reported to BCCDC from 1999-2004. Cases were mapped by address of residence against a reference street network data file⁽¹⁷⁾ or their six-digit postal code⁽¹⁸⁾. Ecosystem data were overlaid⁽¹⁹⁾. Incidence rates were calculated using population data corresponding to B.C. Health Service Delivery Areas obtained from BC Stats⁽²⁰⁾.

Results

Hospital Separations Review
We identified 158 cases of cryptococcal infection in HIV negative persons between 1995 and 2004. Figure 1 demonstrates a marked increase, starting in 1999, of cryptococcal diagnoses in Vancouver Island hospitals. Over the same time period, diagnoses from mainland hospitals remained relatively constant. Only three HIV-ve individuals treated on Vancouver Island were diagnosed with Cryptococcus infection prior to 1999 – two in 1995 and one in 1997.

Figure 1. Hospitalizations for cryptococcal infection in HIV-negative residents of British Columbia as determined by ICD codes on discharge. The stratification of Island or Mainland reflects the location of treatment.

Retrospective Laboratory Serotyping

Serotyping was performed on 36 stored isolates. Twenty-seven (75%) were serotype A, six (17 %) were serotype D, and three (8%) were serotype B (C. gattii). Figure 2 shows the distribution of isolates by year and serotype. C. gattii was not detected in any stored isolates prior to 1999.

Figure 2: Serotyping results for isolates stored at BCCDC (n = 36). A=serotype A (C. neoformans), D = serotype D (C. neoformans), B = serotype B (C. gattii)

Geographic Mapping
The distribution of human cases was clustered along the eastern edge of Vancouver Island in areas within the Coastal Douglas-fir and very dry Coastal Western Hemlock biogeoclimatic zones (Figure 3). The distribution of human cases also corresponded directly with the distribution of animal cases(12). Prior to 2005, all human and animal cases residing in the Lower Mainland had travelled to this eastern edge of Vancouver Island within a year of symptom onset. The rate of infection was highest in the Central Vancouver Island region (annualized rate of 31.6 cases per million population) compared to the South Vancouver Island and North Vancouver Island regions (annualized rates of 19.3 and 24.7 cases per million population respectively). Cases were not new to Vancouver Island; those diagnosed from 1999-2001 had lived an average of 16 years in their municipalities (range 2 to 50 years, SD = 12.92).

Figure 3. Geographic distribution of human cryptococcal infections, 1999-2004. All Vancouver Island cases resided on the east coast of Vancouver Island in the Coastal Douglas-fir (CDF) biogeoclimatic zone. All mainland cases had travelled to the CDF zone on Vancouver Island before their illness. The CDF zone is characterized by drier and warmer summers than other parts of the BC coast.

NORTH VANCOUVER ISLAND
BC MAINLAND
CENTRAL VANCOUVER ISLAND
SOUTH VANCOUVER ISLAND
Study Area
CANADA
USA
MEXICO
Cryptococcus...
Biogeoclimatic Zone
Coastal Western Hemlock - very dry subzones
Coastal Western Hemlock - all other subzones
Mountain Hemlock and Alpine Tundra

Discussion

Vancouver Island has a temperate climate and the emergence of Cryptococcus gattii infections is unexpected. Endemic C. gattii infections have previously been reported mainly in tropical and subtropical regions or areas where host eucalyptus were abundant^(7,21). The first cases that could be laboratory confirmed as C. gattii in B.C. occurred in 1999. An increase in the annual number of cryptococcal infections in HIV-ve persons on Vancouver Island also became noticeable in 1999. Given the incubation period described for Vancouver Island infections⁽²²⁾, this suggests that environmental exposure to C. gattii on Vancouver Island increased in the late 1990s.

Neither the analysis of hospital discharge records nor the retrospective typing of isolates reflects the true incidence of C. gattii infections in B.C. Some stored isolates were no longer viable and protocols for isolate submission and storage may have changed over time. The analysis of discharge records was limited by the inexistence of a specific ICD code for C. gattii; we therefore restricted the analysis to HIV-ve individuals to improve specificity in order to retrospectively demonstrate trends. Reportable disease surveillance data puts the true average annual incidence rate of C. gattii infection in humans on the island at 27.9 cases/million population (2002-2006). This is more than three times the rate of 8.5 cases per million per year found in tropical north Australia(6) .

The use of hospital discharge data for case finding has known limitations although this is less problematic for infectious diseases with discrete, non-overlapping IDC codes (e.g. cryptococcosis than for many chronic conditions (e.g. cardiovascular disease). B.C. hospitals all contribute to a single, centralized database of patient visits, representing the entire population at risk. Provincial identifiers allow repeat hospitalizations for the same individual to be removed, limiting bias. The extent to which misclassifications occurred due to ICD omissions or miscoding is unknown.

Three Cryptococcus cases in HIV-negative residents of Vancouver Island were diagnosed earlier than 1999. These may represent errors or omissions in either Cryptococcus or HIV coding, C. neoformans infections in HIV-negative persons or true C. gattii infections in HIV-negative persons, potentially due to travel. Since data provided were anonymous, a more thorough assessment of these cases was prevented. However, some C. neoformans infections are expected in any HIV-negative population and this represents the most likely explanation for these cases. Retrospective reviews in Hong Kong⁽²³⁾, Brazil⁽²⁴⁾, and Australia⁽⁶⁾ all indicate that an important proportion of C. neoformans incidence occurs in HIV-negative individuals.

An unexplained, transitory drop in incidence in HIV-negative mainland residents coincided with the emergence of Cryptococcus in HIV-negative residents of Vancouver Island. Antiretrovirals for HIV treatment became widely available in B.C. in 1996 and may have prevented cryptococcal infections in subsequent years; if cases were misclassified over this time (i.e. if ICD9 codes for HIV infection were omitted from discharge summaries), this might explain the observed drop. Until 1999, infections in mainland residents likely reflect baseline C. neoformans infections. The increased number of cryptococcal infections among HIV-negative mainland residents beginning in 2002 likely represents C. gattii infections among mainlanders who travelled to Vancouver Island and were exposed to the fungus while there.

The emergence of C. gattii on Vancouver Island resulted in an unprecedented number of cryptococcal infections involving, domestic pets, wild animals⁽¹²⁾ and humans. Past cryptococcal epizootics of bovine mastitis⁽²⁵⁾ and caprine respiratory infections⁽²⁶⁾ have been reported in the United States and Spain, respectively. The occurrence of cryptococcal meningitis in three residents of an Oklahoma community in 1959 is the only report describing a temporal-spatial cluster in humans⁽²⁷⁾. Whether C. neoformans or C. gattii was the infecting organism in the above clusters is not known. These past outbreaks have been smaller in scope and have involved a single mammalian species.

Factors causally associated with the emergence of C. gattii on Vancouver Island remain unproven but may include importation, warming of ambient air temperatures and deforestation. Fungi can be transported from one part of the world to another on commercially traded plants, vehicles, or vectors, and can subsequently colonize native plants⁽²⁸⁾. It has been hypothesized that one of the two parent α mating-types from which the Vancouver Island strain emerged was introduced to Vancouver Island from Australia⁽¹⁴⁾. Nevertheless, a review of import records for established tree hosts for C. gattii does not suggest legal introduction of these species to Vancouver Island in the years preceding 1999. A warming trend on the east coast of Vancouver Island may have supported the emergence of C. gattii. Warmer temperatures can decrease a plant’s resistance to fungal colonization⁽²⁹⁾ and may therefore favor the growth of

C. gattii in host trees. The coastal Douglas-fir biogeoclimatic zone, on the island’s east coast, is characterized by warm, dry summers and mild, wet winters⁽³⁰⁾. Research has shown that aerosolization of the fungus is achieved during felling of colonized trees⁽³¹⁾, therefore recent expansion of neighbourhoods on Vancouver Island and construction of a new Island highway through forested areas in the mid 1990s may also have increased host exposure to the fungus.

The emergence of C. gattii on Vancouver Island demonstrates that exotic pathogens can become established in new ecological niches and have considerable impact on human and animal health. Emergence in other temperate areas should be considered when a cluster of cryptococcal infections occurs in immunocompetent humans or animals; this requires close communication between physicians, microbiologists, veterinarians, and public health officials. Education and early diagnosis are essential since while there are no preventive measures available, antifungal treatment has proven successful. Given its popularity as a tourist destination, physicians should be alert to the possibility of C. gattii in residents and visitors to Vancouver Island and parts of the B.C. lower mainland. Physicians should consider C. gattii in the differential diagnosis for patients who develop cough, dyspnea, fever, night sweats, unexplained weight loss or meningitis within a year of travel to these areas.

Acknowledgements

The authors wish to express thanks to Sally Lester (Central Laboratory for Veterinarians, Ltd.), Ron Lewis and Stephen Raverty (British Columbia Animal Health Centre), Karen Bartlett (University of British Columbia), Craig Stephen (Centre for Coastal Health), Louise Stein (BCCDC laboratories) and to Tania Sorrell and David Ellis of the Australasian Cryptococcal Study Group.

References

1. GD. Primary pulmonary cryptococcosis. Am Rev Respir Dis 1966;94(2):236-43.
2. Driver JA, Saunders CA, Heinze-Lacey B et al. Cryptococcal pneumonia in AIDS: Is cryptococcal meningitis preceded by clinically recognizable pneumonia? J Acquir Immune Defic Syndr Hum Retrovirol 1995;9(2):168-71.
3. Sarosi GA, Silberfarb PM, Tosh FE. Cutaneous cryptococcosis. A sentinel of disseminated disease. Arch Dermatol 1971;104(1):1-3.
4. Behrman RE, Masci JR, Nicholas P. Cryptococcal skeletal infections: Case report and review. Rev Infect Dis 1990;12(2):181-90.
5. Casadevall A, Perfect JR. Cryptococcus neoformans. Washington, DC: ASM Press, 1998.
6. Chen S, Sorrell T, Nimmo G et al. Epidemiology and host- and variety-dependent characteristics of infection due to Cryptococcus neoformans in Australia and New Zealand. Australasian Cryptococcal Study Group. Clin Infect Dis 2000;31(2):499-508.
7. Sorrell TC. Cryptococcus neoformans variety gattii. Med Mycol 2001;39(2):155-68.
8. Hajjeh RA, Brandt ME, Pinner RW. Emergence of cryptococcal disease: Epidemiologic perspectives 100 years after its discovery. Epidemiol Rev 1995; 17(2):303-20.
9. Ehrensing EH, Saag MS. Cryptococcosis. In: Sarosi GA, Davies SF, editors. Fungal infections of the lungs. Philadelphia: Lippincott Williams & Wilkins, 2000:91-103.
10. Ellis DH, Pfeiffer TJ. Natural habitat of Cryptococcus neoformans var.gattii. J Clin Microbiol 1990;28(7):1642-44.
11. Ellis DH, Pfeiffer TJ. Ecology, life cycle, and infectious propagule of Cryptococcus neoformans. Lancet 1990;336(8720):923-25.
12. Stephen C, Lester S, Black W et al. Multispecies outbreak of cryptococcosis on Vancouver Island, British Columbia. Can Vet J 2002;43(10):792-94.
13. Kidd SE, Hagen F, Tscharke RL et al. A rare genotype of Cryptococcus gattii caused the cryptococcosis outbreak on Vancouver Island (British Columbia, Canada). Proc Natl Acad Sci 2005;USA 101:17258-63.
14. Fraser JA, Giles SS, Wenink EC et al. Same sex mating and the origin of the Vancouver Island Cryptococcus gattii outbreak. Nature 2005;437(7063):1360-64.
15. MacDougall L, Kidd S, Galanis E et al. Spread of Cryptococcus gattii in British Columbia, Canada and its detection in the Pacific Northwest, USA. Emerg Infect Dis 2007;13(1):42-50.
16. Galanis E, Watter S, Li, M et al. Cryptococcus gattii in BC – Update on an emerging disease. BCMJ 2007;49(7):374-75.
17. dmti Spatial Inc. CanMap Streetfile V6.0 – British Columbia. Markham, Ontario. 2002.
18. Statistics Canada. Postal Code Conversion File – June 2001 Postal Codes. Ottawa, Ontario. 2001.
19. Provincial digital biogeoclimatic subzone/variant mapping - version 4.0. British Columbia Ministry of Forests, Research Branch, Victoria, BC. 2002.
20. Population Estimates & Projections for P.E.O.P.L.E. 32. BC STATS, Service BC, BC Ministry of Labour and Citizens’ Services, Victoria, BC. April, 2007.
21. Sorrell TC, Ellis DH. Ecology of Cryptococcus neoformans. Rev Iberoam Micol 1997;14:42-43.
22. MacDougall L, Fyfe M. Emergence of Cryptococcus gattii in a novel environment provides clues to its incubation period. J Clin Microbiol 2006;44:1851–52.
23. Lui G, Lee N, IP M et al. Cryptococcosis in apparently immunocompetent patients. Q J Med 2006;99:143-51.
24. Ordonez N, Castaneda E. Varieties and serotypes of Cryptococcus neoformans clinical isolates in Columbia. Rev Iberoam Micol 2001;18:128-30.
25. Pounden WD, Amberson JM, Jaeger RF. Severe mastitis problem associated with Cryptococcus neoformans in large dairy herd. Am J Vet Research 1952;13:121-28.
26. Baro T, Torres-Rodriguez JM, De Mendoza MH et al. First identification of autochthonous Cryptococcus neoformans var. gattii isolated from goats with predominantly severe pulmonary disease in Spain. J Clin Microbiol 1998;36(2):458-61.
27. Muchmore HG, Rhoades ER, Nix GE et al. Occurrence of Cryptococcus neoformans in the environment of three geographically associated cases of cryptococcal meningitis. N Engl J Med 1963;368(20):1112-14.
28. Allen EA. Nonindigenous species introductions: A threat to Canada’s forests and forest economy. Can J Plant Pathol 2002;24(2).
29. Harvell CD, Mitchell CE, Ward JR et al. Climate warming and disease risks for terrestrial and marine biota. Science 2002;296(5576):2158-62.
30. Ecosystems of British Columbia. British Columbia Ministry of Forests. Victoria, BC: 1991.
31. Kidd SE, Bach PJ, Hingston AO et al. Cryptococcus gattii Dispersal Mechanisms, British Columbia, Canada. Emerg Infect Dis 2007,13(1):51-57.