Abstract
Background
Infection by Cryptococcus gattii can lead to pulmonary or central nervous system (CNS) disease, or both. Whether the sites of infection and disease severity are associated with C. gattii species and lineages or with certain underlying medical conditions, or both is unclear. We conducted a retrospective cohort study to identify factors associated with site of infection and mortality among C. gattii cases.
Methods
We extracted data on 258 C. gattii cases from Australia, Canada, and the United States reported from 1999 to 2011. We conducted unadjusted and multivariable logistic regression analyses to evaluate factors associated with site of infection and C. gattii mortality among hospitalized cases (N = 218).
Results
Hospitalized C. gattii cases with CNS and other extrapulmonary disease were younger, more likely to reside in Australia, and be infected with variety gattii I (VGI) lineage but less likely to have comorbidities and die as compared to cases with pulmonary disease. The odds of having CNS and/or other extrapulmonary disease were 9 times higher in cases with VGI infection (adjusted odds ratio [aOR] = 9.21, 95% confidence interval [CI] = 3.28–25.89). Age ≥70 years (aOR = 6.69, 95% CI = 2.44–18.30), chronic lung disease (aOR = 2.62, 95% CI = 1.05–6.51) and an immunocompromised status (aOR = 2.08, 95% CI = 1.05–6.51) were associated with higher odds of C. gattii mortality.
Conclusions
Among hospitalized cases, C. gattii species and lineage are associated with site of infection but not with the risk of death, whereas older age and comorbidities increase the risk of death.
Keywords: Cryptococcus gattii, mortality, infection, predictive factors
Hospitalized C. gattii patients infected with variety gattii I (VGI) lineage had higher odds of central nervous system and/or other extrapulmonary disease, and those aged ≥70 years, with chronic lung disease or immunocompromise had higher odds of death.
Graphical Abstract
Graphical Abstract.
This graphical abstract is also available at Tidbit: https://tidbitapp.io/institutional-portal/clinical-infectious-diseases/tidbits/cid-124054/update
Cryptococcus species are ubiquitous environmental fungi that, when inhaled, can lead to subclinical infection, or symptomatic pulmonary or extrapulmonary (mainly central nervous system [CNS]) disease [1]. C. gattii species complex (hereafter referred to as C. gattii) is divided into species and lineages, which are proposed to have different virulence and health outcomes [2]. However, there is ongoing debate about the relative contribution of C. gattii species and lineages and underlying medical or immunocompromising conditions [2, 3] to the clinical manifestations and outcomes of infection.
In 2008, British Columbia, Canada, Australia, and the Pacific Northwestern states of Washington and Oregon in the United States embarked on separate but similar cohort studies of C. gattii patients to document clinical progression and to identify factors associated with C. gattii-related mortality in an attempt to clarify this issue. The study findings varied, partly due to differences in settings and methods; differences in species and lineage distribution between sites may also have contributed [4–7].
In British Columbia, lineage variety gattii IIa (VGIIa) predominated, and most cases (69%) presented with pulmonary disease [7]. In Australia, variety gattii I (VGI) caused all infections, and most (85%) presented with CNS disease [5]. In the US Pacific Northwest, VGIIa and VGIIc predominated, and most cases had pulmonary disease [4]. In British Columbia and Australia, an immunocompromised state increased the risk of death but in the United States, it did not [4, 5, 7]. In British Columbia, CNS disease increased the risk of death, whereas in the United States, it decreased the risk of death [4, 7].
Given the different findings in the 3 endemic study areas, the roles that C. gattii species and lineage, site of infection, and underlying conditions play in the severity and outcome of C. gattii infection remain unresolved. The findings of these individual studies were affected by small sample size and different case definitions, inclusion criteria, and case ascertainment methods. The different populations at risk, C. gattii lineages, and clinical management in these settings, however, provided the opportunity to assess their impact on clinical presentation and outcome. In the present study, we combined all 3 patient cohorts to increase sample size and standardized analyses in order to identify factors associated with clinical site of infection and mortality among C. gattii patients.
METHODS
Study Population
We included cases of C. gattii disease reported from Australia, British Columbia, and the US Pacific Northwest (Washington and Oregon States) reported from 1999 to 2011. Cases were defined as confirmed by culture using CGB agar and genotyped as VGI, VGII, or VGIII using multiple locus variable number of tandem repeats analysis (MLVA) in British Columbia, multilocus sequence typing (MLST) in the United States, and both rapid amplified polymorphic DNA (RAPD) and MLST in Australia [4, 7–9]. Australian isolates that were not sequenced were classified as VGI due to the near dominance of this lineage in Australia at that time [9]. Cases infected with less common lineages VGIIb and VGIII along with any British Columbia and US Pacific Northwest missing lineages were combined into an “other lineage” category in the models.
The chart abstraction tool was developed jointly by Australian and Canadian researchers, modified slightly for each study site and used as a model by American researchers. Data were collected from medical charts using these tools, as described elsewhere [4, 5, 7]. Data from all 3 study sites were combined, and a standard set of variables developed by Baddley et al for a prior study were used [10].
The studies were approved by the institutional review boards of each site. Activity related to this study was reviewed by the Centers for Disease Control and Prevention (CDC) and was conducted consistent with applicable federal law and CDC policy 45 C.F.R. part 46.102(l)(2), 21 C.F.R. part 56; 42 U.S.C. Sect. 241(d); 5 U.S.C. Sect. 552a; 44 U.S.C. Sect. 3501 et seq.
Definitions
We divided site of infection into pulmonary or extrapulmonary. We defined pulmonary disease as a positive culture from a respiratory sample (eg, sputum, bronchoalveolar lavage fluid, bronchial aspirate); we defined extrapulmonary disease as a positive culture from an extra-pulmonary site (eg, brain parenchyma, cerebrospinal fluid, blood) OR with compatible imaging findings from brain MRI or CT scan, as defined by Baddley et al [10]. Patients with extrapulmonary disease were further described.
We defined chronic lung disease as per Baddley et al [10]. We defined immunocompromising conditions as at least 1 of the following: prior solid organ transplant, prior stem cell transplant, human immunodeficiency virus (HIV) positivity, current connective tissue disorder, current rheumatic disease, past or current hematological malignancy. We defined immunocompromising therapy as any immunosuppressant drug use (eg, cyclosporine, tacrolimus, methotrexate, chemotherapy) in the 12 months prior to C. gattii diagnosis. We defined systemic steroid use as steroid use at any time prior to diagnosis, excluding patients with asthma or chronic obstructive pulmonary disease (COPD) who are more likely to have used inhaled steroids. The latter were combined with cases not reporting steroid use. We defined immunocompromised status as a case with an immunocompromising condition, on immunosuppressive therapy, or on systemic steroids.
We defined recommended induction therapy as at least 14 days on induction therapy (amphotericin B with or without 5-flucytosine). We defined C. gattii deaths as deaths due to C. gattii infection as per the medical chart or death certificate and any death as death due to any cause in the 12 months following C. gattii diagnosis.
Statistical Analyses
We described patient characteristics by region and site of infection using frequencies and proportions. Differences between groups were compared using the χ2 test.
We conducted unadjusted and multivariable logistic regression analyses to evaluate factors associated with site of infection and C. gattii mortality. We assessed the following factors: age, sex, lineage, comorbidities (diabetes, HIV infection, asthma, COPD, chronic lung disease, renal insufficiency, cirrhosis, liver diseases, hematological malignancy, all malignancy, rheumatological disease, immunosuppressant therapy, systemic steroids, immunocompromising conditions, immunocompromised status) and site of infection (mortality model only). Cases who died of causes other than C. gattii infection were excluded from the mortality model. Region (British Columbia, US Pacific Northwest, Australia) was added as a random effect in the mortality model.
Analyses were conducted using SAS 9.4 (SAS Institute Carey, North Carolina, USA).
RESULTS
Two hundred and fifty-eight C. gattii cases from Australia, British Columbia, and the US Pacific Northwest (Washington and Oregon States) from 1999 to 2011 were included (Table 1). Case ascertainment differed by region; only British Columbia systematically included non-hospitalized cases. Geographic location, age, lineage, site of infection, comorbidities, and outcomes differed by hospitalization status (Supplementary Table 1). Therefore, 40 non-hospitalized cases (37 from British Columbia and 3 from the US Pacific Northwest) were excluded from further analyses.
Table 1.
Methods for Identification and Confirmation of C. gattii Cases Included in the Study by Contributing Region (N = 258)
| British Columbia (N = 111) | US Pacific Northwest (N = 79) | Australia (N = 68) | ||||
|---|---|---|---|---|---|---|
| Hospitalized | Non-hospitalized | Hospitalized | Non-hospitalized | Hospitalized | Non-hospitalized | |
| Number of cases | 74 | 37 | 76 | 3 | 68 | 0 |
| Study period | 1999–2007 | 2004–2011 | 1999–2010 | |||
| Case ascertainment | Laboratory-based surveillance (reportable since 2003) | Hospital-based diagnosis and laboratory-based surveillance (reportable in WA since 2006 and in OR since 2011) | Hospital-based diagnosis | |||
| Lab confirmation | Culture confirmed as C. gattii and typed using MLVA | Culture confirmed as C. gattii and typed using MLST | Culture confirmed as C. gattii and typed using MLST and RAPD or not typed | |||
Abbreviations: C. gattii, Cryptococcus gattii; MLST, multilocus sequence typing; MLVA, multiple locus variable number of tandem repeats analysis; OR, Oregon State; RAPD, rapid amplified polymorphic DNA; WA, Washington State.
Of 218 hospitalized cases, 74 (33.9%) had pulmonary disease only. One hundred and forty-four (66.1%) had extrapulmonary disease, including 132 (91.7%) with CNS disease, 7 with positive blood culture (of which 2 also had pulmonary disease), 2 with positive urine culture, 2 with positive bone or joint culture (of which 1 also had pulmonary disease), and 1 with positive soft tissue culture.
In univariate analyses, hospitalized C. gattii cases (N = 218) differed by region (Table 2). Those in Australia were younger (age ≤49 years: 61.8% in Australia, 36.5% in British Columbia, 43.4% in US Pacific Northwest), all were infected or presumed infected with VGI, and almost all had CNS or other extrapulmonary disease (92.6%). Those in British Columbia and the US Pacific Northwest were mostly infected with VGIIa or VGIIc (in US Pacific Northwest only), and approximately half presented with isolated pulmonary disease. All British Columbia and US Pacific Northwest cases infected with VGI had CNS disease.
Table 2.
Demographic and Clinical Features of C. gattii Hospitalized Cases by Region (N = 218)
| BC (N = 74) N (%) |
US PNW (N = 76) N (%) |
Australia (N = 68) N (%) |
P Value | |
|---|---|---|---|---|
| Sex | ||||
| Male | 41 (55.4%) | 36 (47.4%) | 41 (60.3%) | .288 |
| Age, y | ||||
| 0–49 | 27 (36.5%) | 33 (43.4%) | 42 (61.8%) | .004 |
| 50–69 | 26 (35.1%) | 33 (43.4%) | 19 (27.9%) | |
| ≥70 | 21 (28.4%) | 10 (13.2%) | 7 (10.3%) | |
| Lineage | ||||
| I | 5 (6.8%) | 4 (5.3%) | 68 (100.0%)a | <.001 |
| III | 0 (0.0%) | 1 (1.3%) | 0 (0.0%) | |
| IIa | 60 (81.1%) | 46 (60.5%) | 0 (0.0%) | |
| IIb | 6 (8.1%) | 3 (3.9%) | 0 (0.0%) | |
| IIc | 0 (0.0%) | 21 (27.6%) | 0 (0.0%) | |
| Unknown | 3 (4.1%) | 1 (1.3%) | 0 (0.0%) | |
| Site of infection | ||||
| Extrapulmonary | <.001 | |||
| CNS (± pulmonary) | 40 (54.1%) | 31 (40.8%) | 61 (89.7%) | |
| Other non-CNS sitesb | 3 (4.1%) | 7 (9.2%) | 2 (2.9%) | |
| Pulmonary only | 31 (41.9%) | 38 (50.0%) | 5 (7.4%) | |
| Comorbidities | ||||
| Diabetes | 11 (14.9%) | 19 (25.0%) | 2 (2.9%) | .001 |
| HIV | 4 (5.4%) | 3 (3.9%) | 1 (1.5%) | .454 |
| Any transplantc | 2 (2.7%) | 14 (18.4%) | 1 (1.5%) | <.001 |
| Asthma | 5 (6.8%) | 4 (5.3%) | 1 (1.5%) | .304 |
| COPD | 9 (12.2%) | 4 (5.3%) | 0 (0.0%) | .009 |
| Chronic lung disease | 23 (31.1%) | 17 (22.4%) | 3 (4.4%) | <.001 |
| Renal insufficiency | 2 (2.7%) | 15 (19.7%) | 1 (1.5%) | <.001 |
| Cirrhosis | 3 (4.1%) | 5 (6.6%) | 1 (1.5%) | .306 |
| Any malignancy | 19 (25.7%) | 14 (18.4%) | 4 (5.9%) | .007 |
| Hematological malignancy | 7 (9.5%) | 5 (6.6%) | 1 (1.5%) | .128 |
| Immune compromise | ||||
| Immunocompromising conditionsd | 15 (20.3%) | 27 (35.5%) | 5 (7.4%) | <.001 |
| Immunosuppressive therapye | 4 (5.4%) | 27 (35.5%) | 4 (5.9%) | <.001 |
| Systemic steroid usef | 10 (13.5%) | 31 (40.8%) | 10 (14.7%) | <.001 |
| Immunocompromised statusg | 19 (25.7%) | 39 (51.3%) | 14 (20.6%) | <.001 |
| Therapy | ||||
| Flucytosine | 26 (35.1%) | 32 (42.1%) | 54 (79.4%) | <.001 |
| Induction ≥ 14 d | 19 (25.7%) | 17 (22.4%) | 51 (75.0%) | <.001 |
| Deaths | ||||
| C. gattii deaths | 23 (31.1%) | 11 (14.5%) | 8 (11.8%) | .006 |
| Any death by 12 mo | 25 (34.7%) | 13 (17.1%) | 9 (13.6%) | .005 |
Abbreviations: BC, British Columbia; C. gattii, Cryptococcus gattii; CNS, central nervous system; COPD, chronic obstructive pulmonary disease; HIV, human immunodeficiency virus; US PNW, US Pacific Northwest; VGI, variety gattii I.
aIn sum, 40 Australian cases with unknown lineage were classified as VGI as this was the strain circulating in the region at the time.
bIncludes 12 cases with other invasive disease, including 7 with positive blood samples (of which 2 also had pulmonary disease on imaging), 2 with positive urine samples, 2 with positive bone or joint sample (of which 1 also had pulmonary infection), and 1 with positive soft tissue sample.
cIncludes solid organ and stem cell transplant.
dIncludes at least one of the following: prior solid organ transplant, prior stem cell transplant, HIV positivity, current connective tissue disorder, current rheumatic disease, past or current hematological malignancy.
eIncludes any immunosuppressant drug use (e.g., cyclosporine, tacrolimus, methotrexate, chemotherapy) in the 12 months prior to C. gattii diagnosis.
fIncludes steroid use at any time prior to diagnosis, excluding patients with asthma or COPD who were more likely to have used inhaled steroids.
gIncludes cases with immunocompromising conditions and/or on immunosuppressive therapy and/or on systemic steroids.
Hospitalized C. gattii cases in British Columbia and the US Pacific Northwest had higher rates of comorbidities than those in Australia, particularly: diabetes, transplant, COPD, renal insufficiency, and malignancy (Table 2). The US Pacific Northwest had the highest proportion of cases with an immunocompromised status (51.3%). Australia had the highest proportion receiving induction therapy for at least 14 days (75.0%). British Columbia hospitalized cases were the most likely to die of C. gattii infection (31.1%).
Hospitalized cases differed by site of infection (Table 3). Those with extrapulmonary disease were more likely to be ≤49 years old (extrapulmonary 53.5%, pulmonary 33.8%). Cases with isolated pulmonary disease were more likely to reside in British Columbia or the US Pacific Northwest (extrapulmonary 56.3%, pulmonary 93.3%) and more likely to be infected with VGIIa (extrapulmonary 37.5%, pulmonary 70.3%). Cases with pulmonary-only infection had more comorbidities including diabetes, COPD, any chronic lung disease, renal insufficiency, history of malignancy, and immunocompromised status. Half (55.6%) of the hospitalized cases with CNS or other extrapulmonary disease were on induction therapy for at least 14 days compared with 9.5% of those with isolated pulmonary disease (P < .001). Hospitalized cases with isolated pulmonary disease were more likely to die of any cause by 12 months (extrapulmonary 16.0%, pulmonary 25.7%, P = .038).
Table 3.
Demographic and Clinical Features of Hospitalized C. gattii Cases by Infection Site (N = 218)
| Extrapulmonarya (N = 144) N (%) |
Pulmonary Only (N = 74) N (%) |
P Value | |
|---|---|---|---|
| Geographic region | |||
| Australia | 63 (43.8%) | 5 (6.8%) | <.001 |
| BC | 43 (29.9%) | 31 (41.9%) | |
| US PNW | 38 (26.4%) | 38 (51.4%) | |
| Sex | |||
| Male | 83 (57.6%) | 35 (47.3%) | .147 |
| Age, y | |||
| 0–49 | 77 (53.5%) | 25 (33.8%) | .006 |
| 50–69 | 49 (34.0%) | 29 (39.2%) | |
| ≥70 | 18 (12.5%) | 20 (27.0%) | |
| Lineageb | |||
| I | 72 (50.0%) | 5 (6.8%) | <.001 |
| III | 0 (0.0%) | 1 (1.4%) | |
| IIa | 54 (37.5%) | 52 (70.3%) | |
| IIb | 2 (1.4%) | 7 (9.5%) | |
| IIc | 13 (9.0%) | 8 (10.8%) | |
| Unknown | 3 (2.1%) | 1 (1.4%) | |
| Comorbidities | |||
| Diabetes | 15 (10.4%) | 17 (23.0%) | .013 |
| HIV | 5 (3.5%) | 3 (4.1%) | .829 |
| Any transplantc | 8 (5.6%) | 9 (12.2%) | .085 |
| Asthma | 4 (2.8%) | 6 (8.1%) | .075 |
| COPD | 3 (2.1%) | 10 (13.5%) | .001 |
| Chronic lung disease | 14 (9.7%) | 29 (39.2%) | <.001 |
| Renal insufficiency | 8 (5.6%) | 10 (13.5%) | .043 |
| Cirrhosis | 4 (2.8%) | 5 (6.8%) | .162 |
| Any malignancy | 16 (11.1%) | 21 (28.4%) | .001 |
| Hematological malignancy | 8 (5.6%) | 5 (6.8%) | .723 |
| Immune compromise | |||
| Immunocompromising conditions | 25 (17.4%) | 22 (29.7%) | .036 |
| Immunosuppressive therapy | 18 (12.5%) | 17 (23.0%) | .046 |
| Systemic steroid use | 25 (17.4%) | 26 (35.1%) | .003 |
| Immunocompromised statusd | 37 (25.7%) | 35 (47.3%) | .001 |
| Induction therapy | |||
| ≥ 14 d | 80 (55.6%) | 7 (9.5%) | <.001 |
| Death | |||
| C. gattii deaths | 23 (16.0%) | 19 (25.7%) | .085 |
| Any death by 12 mo | 25 (17.7%) | 22 (30.1%) | .038 |
Abbreviations: BC, British Columbia; C. gattii, Cryptococcus gattii; COPD, chronic obstructive pulmonary disease; HIV, human immunodeficiency virus; US PNW, US Pacific Northwest.
aIncludes 132 cases with central nervous system (CNS) disease and 12 cases with other disease, including 7 with positive blood samples (of which 2 also had pulmonary disease on imaging), 2 with positive urine samples, 2 with positive bone or joint sample, and 1 with positive soft tissue sample.
bIn sum, 40 Australian cases with unknown lineage were classified as variety gattii I (VGI) due to circulating strain in the region at the time.
cIncludes solid organ and stem cell transplant.
dIncludes cases with immune compromising conditions or on immunosuppressive therapy or on systemic steroids.
In the multivariable analysis of factors associated with site of infection, VGI was associated with CNS and other extrapulmonary disease (adjusted odds ratio [aOR] = 9.21, 95% confidence interval [CI] = 3.28–25.89) (Table 4). Chronic lung disease and immunocompromised status were associated with a decreased odds of CNS and other extrapulmonary disease. Excluding the 12 cases with non-CNS extrapulmonary disease did not change the model results.
Table 4.
Multivariable Logistic Regression of Factors Associated With CNS and Other Extrapulmonary Disease Among Hospitalized C. gattii Cases (N = 218)
| Factors | Unadjusted OR (95% CI) | Adjusted OR (95% CI) |
|---|---|---|
| Age, y | ||
| 0–49 (N = 102) | Ref | Ref |
| 50–69 (N = 78) | 0.55 (.29, 1.04) | 0.89 (.42, 1.89) |
| ≥70 (N = 38) | 0.29 (.13, .64) | 0.57 (.22, 1.46) |
| Sex | ||
| Female (N = 100) | Ref | |
| Male (N = 118) | 1.52 (.86, 2.66) | 1.48 (.76, 2.89) |
| Lineage | ||
| VGIIa (N = 106) | Ref | Ref |
| VGI (N = 77) | 13.87 (5.19, 37.06) | 9.21 (3.28, 25.89) |
| Others (N = 35) | 0.53 (.17, 1.7) | 0.99 (.43, 2.27) |
| Comorbidities | ||
| No chronic lung disease (N = 175) | Ref | Ref |
| Chronic lung disease (N = 43) | 0.17 (.08, .34) | 0.25 (.11, .56) |
| No immunocompromised status (N = 146) | Ref | Ref |
| Immunocompromised status (N = 72) | 0.39 (.21, .70) | 0.45 (.23, .91) |
Abbreviations: C. gattii, Cryptococcus gattii; CI, confidence interval; CNS, central nervous system; OR, odds ratio; Ref, reference group; VGI, variety gattii I; VGIIa, variety gattii IIa.
The majority (n = 33, 78.6%) of C. gattii deaths were among individuals who were ≥70 years of age with chronic lung disease or an immunocompromised status (data not shown). An additional 6 were aged ≥70 years and had another comorbidity. Three fatal cases were younger than 70 years with no reported comorbidities; all had CNS infection.
In the multivariable analysis of factors associated with C. gattii mortality, age ≥70 years (aOR = 6.69, 95% CI = 2.44–18.30), chronic lung disease (aOR = 2.62, 95% CI = 1.05–6.51) and an immunocompromised status (aOR = 2.08, 95% CI = 1.05–6.51) were associated with greater odds of death (Table 5). C. gattii lineage and site of infection were not independently associated with mortality. Including the 9 cases who died of other causes did not change the model results.
Table 5.
Multivariable Logistic Regression With Random Effects of Factors Associated With C. gattii Death Among Hospitalized C. gattii Casesa (N = 209)
| Factors | Unadjusted OR (95% CI) | Adjusted OR (95% CI) |
|---|---|---|
| Age, y | ||
| 0–49 (N = 100) | Ref | Ref |
| 50–69 (N = 73) | 1.75 (.74, 4.17) | 1.58 (.63, 3.98) |
| ≥70 (N = 36) | 8.09 (3.27, 20.0) | 6.69 (2.44, 18.30) |
| Sex | ||
| Female (N = 96) | Ref | Ref |
| Male (N = 113) | 1.49 (.75, 2.99) | 1.71 (.77, 3.79) |
| Lineage | ||
| VGIIa (N = 101) | Ref | Ref |
| VGI (N = 74) | 0.36 (.16, .82) | 0.54 (.19, 1.58) |
| Others (N = 34) | 0.45 (.16, 1.28) | 0.34 (.1, 1.17) |
| Infection site | ||
| Pulmonary only (N = 70) | Ref | Ref |
| CNS and other extrapulmonary (N = 139) | 0.53 (.27, 1.06) | 1.33 (.53, 3.31) |
| Comorbidities | ||
| No chronic lung disease (N = 170) | Ref | Ref |
| Chronic lung disease (N = 39) | 3.31 (1.54, 7.11) | 2.62 (1.05, 6.51) |
| No immunocompromised status (N = 142) | Ref | Ref |
| Immunocompromised status (N = 67) | 2.05 (1.02, 4.10) | 2.08 (1.05, 6.51) |
Abbreviations: C. gattii, Cryptococcus gattii; CI, confidence interval; CNS, central nervous system; OR, odds ratio; Ref, reference group; VGI, variety gattii I.
aNine individuals who died due to other reasons were excluded.
DISCUSSION
We conducted an international study by combining C. gattii infected case cohorts from endemic areas in Australia, the United States, and Canada to clarify the role that species and lineage, site of infection and underlying conditions play in the site of infection and outcome of C. gattii infection.
Cases differed substantially by hospitalization status. All non-hospitalized cases had pulmonary disease only, and none died; they were more likely to be older, infected with VGIIa, have chronic lung disease, and be from British Columbia. Given these differences, we limited analyses to hospitalized cases. We found that C. gattii VGI infection was associated with CNS and other extrapulmonary disease, whereas comorbidities, particularly chronic lung disease and immunocompromised status, were associated with isolated pulmonary disease. We also found that neither lineage nor site of cryptococcal infection were associated with C. gattii mortality, in contrast to older age, chronic lung disease and an immunocompromised status.
When we included non-hospitalized cases in our multivariable regression models (data not shown), we found that cases with pulmonary disease had a lower risk of C. gattii death, a trend that was also noted by Phillips et al in British Columbia [7]. Because British Columbian researchers included these milder pulmonary disease cases, they found that CNS disease was associated with C. gattii death. We limited our analysis to hospitalized C. gattii infected cases (a more severely ill and homogeneous patient group) and found that site of infection was not significantly associated with C. gattii death. In sum, C. gattii pulmonary disease has a wider spectrum of severity, and patients generally have a better prognosis. However, patients hospitalized with pulmonary disease are just as likely to die as patients hospitalized with CNS or other extrapulmonary disease.
We found that among hospitalized cases, VGI was strongly associated with CNS and other extrapulmonary disease, as had Australian and British Columbian researchers [5, 7]. This finding remained true whether we included or excluded Australian cases. This may be due to a predilection of this lineage to infect the CNS or to unexplained host, clinical, or environmental factors. We found that chronic lung disease and an immunocompromised status were both associated with isolated pulmonary infection, as had Phillips et al [7]. Others have also found underlying lung disease to be associated with pulmonary cryptococcosis [11, 12]. Chronic lung disease may cause anatomic abnormalities, defective mucus secretion, decreased ciliary clearance or other changes enabling C. gattii to establish infection. Other studies have found that cryptococcal lung infection is associated with known immunocompromising conditions, immunosuppressive therapy or steroid use [13]. There is growing evidence for the role of unrecognized immunological risks in C. gattii infection, such as autoantibodies against the granulocyte-macrophage colony-stimulating factor [14]. However, it is unclear why an immunocompromised state is more strongly associated with pulmonary than CNS and other extrapulmonary infection. Our interpretation is limited by the lack of detail on the type of immune deficiency in each case.
Where past studies disagreed on the role of infection site in C. gattii mortality, our combined analysis found that site of infection and lineage were not associated with risk of death in hospitalized patients, once adjusted for confounders. We found that older age was strongly associated with C. gattii mortality, as did Harris et al and Phillips et al [4, 7]. We also found chronic lung disease and an immunocompromised state were associated with C. gattii mortality. Phillips et al and Chen et al both found an association between an immunocompromised status and death, and Harris found an association between oral steroid use and death [4, 5, 7]. In our combined analysis, the vast majority (92.9%) of all fatal cases were either aged 70 years and above, or had comorbidities, or both.
Treatment guidelines recommend induction therapy for cryptococcal disease affecting the CNS, for severe pulmonary disease and for other severe or disseminated disease [15–17]. The most recent guidelines recommend similar management for C. gattii as for C. neoformans in most cases [17]. However, it has been reported that many C. gattii-infected patients have not been treated in line with these guidelines [6, 7, 18], which may have influenced their outcome. All hospitalized cases in our cohort had CNS infection, other disseminated disease, or were assumed to have severe pulmonary disease given their admission to hospital and should have received induction therapy. However, only 40.0% received at least 14 days of induction therapy.
Our comparable treatment data across the 3 regions was limited to duration of induction. We found that Australian cases and those with CNS or other extrapulmonary disease were more likely to receive induction therapy for ≥14 days than other cases (Tables 2 and 3). However, after adjusting for other factors, we found that induction therapy was not associated with site of infection or mortality risk. This is more likely to reflect our inability to accurately measure the impact of therapy on outcome due to lack of detailed treatment courses in our cohorts rather than reflect a true lack of impact. We recommend that the impact of treatment on the outcome of C. gattii cases, particularly in those with isolated pulmonary disease, be studied further.
Our study was limited by incomplete data-set comparability. Although the chart abstraction tools were very similar between study sites, there were some differences in the variables collected and in the amount of detail included, such as treatment type and duration; and some inconsistency in how some variables, such as comorbidities, were recorded. The main limitation was that only 1 site consistently identified and included outpatients. Given the important differences between hospitalized and non-hospitalized cases, the latter were excluded from most analyses to ensure comparable data sets. Our findings are therefore not generalizable to non-hospitalized C. gattii infected cases who tended to be younger, infected with VGIIa, and with milder, isolated pulmonary disease (Table 3). Also, only 1 VGIII infected case was included, and our findings cannot be extrapolated to this species, which tends to cause severe illness.
The lack of detailed, standardized data greatly limited our ability to combine and contrast treatment. Our study cannot confidently comment on the impact of treatment on site of infection or mortality risk. This should be assessed using a prospective cohort or randomized trial. We were also unable to measure the time between symptom onset and diagnosis consistently. This factor may have affected access to treatment and the risk of mortality. Time-to-event models were assessed but were not feasible due to differences in data capture and lack of dates in some study regions; logistic regression models were used instead. Finally, not all Australian cases were genotyped; these cases were assumed to all be infected with VGI, given the overwhelming dominance of this lineage in Australia during the study period. Even if a few were misclassified, this would not have influenced the overall findings.
In conclusion, by combining data from 3 endemic areas and addressing differences in case ascertainment, we found that C. gattii species and lineage are associated with site of infection but not with the risk of death; rather, older age and certain underlying conditions increase the odds of C. gattii death among hospitalized cases. We recommend aggressive clinical management of patients presenting to hospital with severe C. gattii infection of the lungs, CNS, or other sites, including induction therapy as per the most recent clinical guidelines [17].
Supplementary Material
Contributor Information
Eleni Galanis, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada.
Laura MacDougall, British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada.
Caren Rose, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada.
Sharon C A Chen, Centre for Infectious Diseases and Microbiology Laboratory Services, Institute of Clinical Pathology and Medical Research, New South Wales Health Pathology, Westmead Hospital, Westmead, New South Wales, Australia; Sydney Infectious Diseases Institute, Faculty of Medicine and Health, University of Sydney, Sydney, New South Wales, Australia.
Hanna N Oltean, Washington Department of Health, Seattle, Washington, USA.
Paul R Cieslak, Oregon Department of Health, Portland, Oregon, USA.
Emilio DeBess, Oregon Department of Health, Portland, Oregon, USA.
Mei Chong, British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada.
Tania C Sorrell, Sydney Infectious Diseases Institute, Faculty of Medicine and Health, University of Sydney, Sydney, New South Wales, Australia.
John W Baddley, School of Medicine, Johns Hopkins University, Baltimore, Maryland, USA.
Linda M N Hoang, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada.
Shawn R Lockhart, Mycotic Diseases Branch, Centers for Disease Control and Prevention, Atlanta, Georgia, USA.
Peter G Pappas, Heersink School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA.
Peter Phillips, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; Division of Infectious Diseases, St. Paul's Hospital, Vancouver, British Columbia, Canada.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Acknowledgments . The authors thank Dr. Catriona Halliday for identifying C. gattii in Australia.
They thank C. gattii database managers (Sophie Li, Min Li) and research assistants (Max Xie) in British Columbia. They thank chart abstractors (Lana MacDonald) and research assistant (Albert Isaacs) in British Columbia.
They also thank Dr Muhammad Morshed and the British Columbia Centre for Disease Control (BCCDC) Public Health Laboratory for molecular typing.
Disclaimer . The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention and other institutions.
Financial support . J. W. B. received a Merck Inc. grant for analysis of data published in 10 and which were used for this study. P. P. reports Astellas Pharma Canada and Pfizer Canada.
Potential conflicts of interest . The authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.
References
- 1. McMullan BJ, Sorrell TC, Chen SCA. Cryptococcus gattii infections: contemporary aspects of epidemiology, clinical manifestations and management of infection. Fut Microbiol 2013; 8:1613–31. [DOI] [PubMed] [Google Scholar]
- 2. Montoya MC, Magwene PM, Perfect JR. Associations between Cryptococcus genotypes, phenotypes, and clinical parameters of human disease. A review. J Fungi 2021; 7:260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Chen SC, Meyer W, Sorrell TC. Cryptococcus gattii infections. Clin Microbiol Rev 2014;27:980–1024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Harris JR, Lockhart SR, Debess E, et al. Cryptococcus gattii in the United States: clinical aspects of infection with an emerging pathogen. Clin Infect Dis 2011; 53:1188–95. [DOI] [PubMed] [Google Scholar]
- 5. Chen SCA, Slavin MA, Heath CH, et al. Clinical manifestations of Cryptococcus gattii infection: determinants of neurological sequelae and death. Clin Infect Dis 2012; 55:789–98. [DOI] [PubMed] [Google Scholar]
- 6. Smith RM, Mba-Jonas A, Tourdjman M, et al. Treatment and outcomes among patients with Cryptococcus gattii infections in the United States Pacific Northwest. PLoS One 2014; 9:e88875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Phillips P, Galanis E, MacDougall L, et al. Longitudinal clinical findings and outcome among patients with Cryptococcus gattii infection in British Columbia. Clin Infect Dis 2015; 60:1368–76. [DOI] [PubMed] [Google Scholar]
- 8. Carriconde F, Gilgado F, Arthur I, et al. Clonality and α-a recombination in the Australian Cryptococcus gattii VGII population—an emerging outbreak in Australia. PLoS One 2011; 6:e16936. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sorrell TC, Chen SC, Ruma P, et al. Concordance of clinical and environmental isolates of Cryptococcus neoformans var. gattii by random amplification of polymorphic DNA analysis and PCR fingerprinting. J Clin Microbiol 1996; 34:1253–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Baddley JW, Chen SCA, Huisingh C, et al. MSG07: an international cohort study comparing epidemiology and outcomes of patients with Cryptococcus neoformans or Cryptococcus gattii infections. Clin Infect Dis 2021; 73:1133–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Baddley JW, Perfect JR, Oster RA, et al. Pulmonary cryptococcosis in patients without HIV infection: factors associated with disseminated disease. Eur J Clin Microbiol Infect Dis 2008; 27:937–43. [DOI] [PubMed] [Google Scholar]
- 12. MacDougall L, Fyfe M, Romney M, Starr M, Galanis E. Risk factors for Cryptococcus gattii infection, British Columbia, Canada. Emerg Infect Dis 2011; 17:193–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Setianingrum F, Rautemaa-Richardson R, Denning DW. Pulmonary cryptococcosis: a review of pathobiology and clinical aspects. Med Mycol 2019; 57:133–50. [DOI] [PubMed] [Google Scholar]
- 14. Wang SY, Lo YF, Shih HP, et al. Cryptococcus gattii infection as the major clinical manifestation in patients with autoantibodies against granulocyte–macrophage colony-stimulating factor. J Clin Immunol 2022; 42:1730–41. [DOI] [PubMed] [Google Scholar]
- 15. Saag MS, Graybill RJ, Larsen RA, et al. Practice guidelines for the management of cryptococcal disease. Clin Infect Dis 2000; 30:710–8. [DOI] [PubMed] [Google Scholar]
- 16. Perfect JR, Dismukes WE, Dromer F, et al. Clinical practice guidelines for the management of cryptococcal disease: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis 2010; 50:291–322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Chang CC, Harrison TS, Bicanic TA, et al. Global guideline for the diagnosis and management of cryptococcosis: an initiative of the ECMM and ISHAM in cooperation with the ASM. Lancet Infect Dis 2024; 24(8):e495–512 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Chen SCA, Korman TM, Slavin MA, et al. Antifungal therapy and management of complications of cryptococcosis due to Cryptococcus gattii. Clin Infect Dis 2013; 57:543–51. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.