ArticlePDF AvailableLiterature Review

Abstract and Figures

The incidence of invasive mycoses is increasing, especially among patients who are immunocompromised or hospitalized with serious underlying diseases. Such infections may be broken into two broad categories: opportunistic and endemic. The most important agents of the opportunistic mycoses are Candida spp., Cryptococcus neoformans, Pneumocystis jirovecii, and Aspergillus spp. (although the list of potential pathogens is ever expanding); while the most commonly encountered endemic mycoses are due to Histoplasma capsulatum, Coccidioides immitis/posadasii, and Blastomyces dermatitidis. This review discusses the epidemiologic profiles of these invasive mycoses in North America, as well as risk factors for infection, and the pathogens' antifungal susceptibility.
Content may be subject to copyright.
Fungi have emerged in the last two decades as major
causes of human disease, especially among those who
are immunocompromised or hospitalized with serious
underlying diseases (Horn et al. 2009; Kollef et al. 2008;
Lockhart et al. 2009; Neofytos et al. 2009a; Perlroth et al.
2007; Pfaller et al. 2006c; Richardson et al. 2008). High
risk groups include individuals undergoing hematopoe-
itic stem cell transplantation (HSCT), solid organ trans-
plantation (SOT), major surgery (especially gastrointes-
tinal [GI] surgery), those with AIDS, neoplastic disease,
immunosuppressive therapy, advanced age, and pre-
mature birth (Table 1) (Fishman et al. 2007; Fridkin et al.
2006; Mean et al. 2008; Morris et al. 2008; Neofytos et al.
2009a; Neofytos et al. 2009b; Perlroth et al. 2007; Procop
et al. 2004; Singh. 2003; Singh et al. 2008; Walzer et al.
2008; Zaoutis et al. 2005; Zaoutis et al. 2007).
Serious invasive fungal infections (IFI; those of blood,
normally sterile body uids, deep tissue and organs),
may be broken into two broad categories: (1) oppor-
tunistic and (2) endemic. e opportunistic mycoses
almost always represent health care-associated infec-
tions (HAI) and may be either nosocomial (occur after
~72 h hospitalization) or community onset in nature
(Kollef et al. 2008; Perlroth et al. 2007; Procop et al.
2004). Contributing factors include exposure to broad-
spectrum antibacterial agents, corticosteroids, cyto-
toxic chemotherapeutic agents, and prolonged use of
intravascular catheters (Table 1). e most important
agents of the opportunistic mycoses are Candida spp.,
Cryptococcus neoformans, Pneumocystis jirovecii, and
Aspergillus spp., although the list of potential pathogens
is ever expanding (Table 2).
e endemic mycoses are those in which susceptibil-
ity to the infection is acquired by living in a geographic
area constituting the natural habitat of the particular
fungus. e most commonly encountered endemic
mycoses in North America are due to Histoplasma capsu-
latum, Coccidioides immitis/posadasii, and Blastomyces
dermatitidis (Table 2) (Chiller et al. 2003; Chu et al. 2006;
Kauman 2001; Kauman 2007; Parish et al. 2008). e
organisms in this group are considered primary sys-
temic pathogens, owing to their ability to cause infec-
tion in both “normal” and immunocompromised hosts
and for their propensity to involve the deep viscera after
dissemination of the fungus from the lungs following its
inhalation from natural sources.
e prevention, diagnosis, and therapy of both
opportunistic and endemic mycoses remain extremely
Critical Reviews in Microbiology
Critical Reviews in Microbiology, 2010; 36(1): 1–53
2010
36
1
1
53
Address for Correspondence: Dr. Michael A. Pfaller, University of Iowa, 200 Hawkins Dr., Dept. of Pathology C606B GH, Iowa City, IA 52242. E-mail: michael-
pfaller@uiowa.edu
05 June 2009
00 00 0000
07 August 2009
1040-841X
1549-7828
© 2010 Informa UK Ltd
10.3109/10408410903241444
REVIEW ARTICLE
Epidemiology of Invasive Mycoses in North America
Michael A. Pfaller, and Daniel J. Diekema
Departments of Pathology and Medicine, Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa
City, USA
Abstract
The incidence of invasive mycoses is increasing, especially among patients who are immunocompromised
or hospitalized with serious underlying diseases. Such infections may be broken into two broad categories:
opportunistic and endemic. The most important agents of the opportunistic mycoses are Candida spp.,
Cryptococcus neoformans, Pneumocystis jirovecii, and Aspergillus spp. (although the list of potential patho-
gens is ever expanding); while the most commonly encountered endemic mycoses are due to Histoplasma
capsulatum, Coccidioides immitis/posadasii, and Blastomyces dermatitidis. This review discusses the epide-
miologic proles of these invasive mycoses in North America, as well as risk factors for infection, and the
pathogens’ antifungal susceptibility.
Keywords:
MCB
424318
(Received 05 June 2009; accepted 07 August 2009)
ISSN 1040-841X print/ISSN 1549-7828 online © 2010 Informa UK Ltd
DOI: 10.3109/10408410903241444 http://www.informahealthcare.com/mby
2 M. A. Pfaller, and D. J. Diekema
dicult. Increased recognition of the importance of
these infections has spurred eorts to develop new diag-
nostic and therapeutic approaches, as well as to expand
our knowledge of the epidemiology and pathogenesis of
the mycoses.
is review discusses selected aspects of the epide-
miologic proles of the invasive mycoses that may be
encountered in North America as well as risk factors for
infection with various fungal pathogens. e susceptibil-
ity of pathogens to antifungal agents is also discussed.
Opportunistic mycoses
e frequency of IFI due to opportunistic fungal patho-
gens has clearly increased in recent years (Table 3)
(Pfaller et al. 2007a; Rees et al. 1998; Reingold et al.
1986; Wilson et al. 2002). A study of the epidemiology of
sepsis in the United States (U.S.) found that the annual
number of cases of sepsis caused by fungal organisms
increased by 207% between 1979 and 2000 (Martin et al.
2003). In the Surveillance and Control of Pathogens of
Epidemiological Importance (SCOPE) Study, a 49-center
study of 24,179 nosocomial bloodstream infections (BSI)
recorded between 1995 and 2002, 9.5% of the infections
were fungal in origin (Wisplingho et al. 2004). Candida
spp. were the fourth leading cause of nosocomial
BSI, surpassed only by staphylococci and enterococci
(Table 4) (Wisplingho et al. 2004). Notably, in a recent
population-based study of candidemia, Sofair et al.
(Sofair et al. 2006) determined that 31% of 1,143 cases of
candidemia were community onset infections, leading
participants in a recent HAI summit to conclude that cli-
nicians should be aware of the potential for candidemia
to be a cause of BSI in patients presenting to the emer-
gency department (Kollef et al. 2008). A majority of the
HAI summit panelists and Infectious Diseases Society of
America (IDSA) members responding to a Web-based
survey voiced support for the concept that “patients with
serious HAIs who have risk factors for fungal infections
require early empiric antifungal therapy to reduce mor-
tality,” further underscoring the recognized importance
of opportunistic mycoses occurring both inside and out
of the hospital setting (Kollef et al. 2008).
e most well-known causes of opportunistic
mycoses include Candida albicans, Cryptococcus neofor-
mans, Aspergillus fumigatus, and Pneumocystis jirovecii
(Fishman et al. 2008; Pfaller et al. 2006c; Pfaller et al.
2004a; Procop et al. 2004; Tellez et al. 2008; Zilberberg
et al. 2008). e estimated annual incidence of invasive
mycoses due to these pathogens is 72–290 infections
per million population for Candida, 30–66 infections
per million for C. neoformans, and 12–34 infections
per million population for Aspergillus species (Table 5)
Table 1. Predisposing factors for opportunistic and endemic mycoses.
Factor Possible role in infection Major fungal pathogens
Antimicrobial agents (no. and
duration)
Promote fungal colonization
Provide intravascular access
Candida spp., other yeast-like fungi
Adrenal corticosteroids Immunosuppression Candida spp., Cryptococcus neoformans, Aspergillus spp.,
Zygomycetes, other moulds, Pneumocystis, endemic fungi
Chemotherapy Immunosuppression Candida spp., Aspergillus spp., Pneumocystis
Hematologic/solid organ
malignancy
Immunosuppression Candida spp., Aspergillus spp., Zygomycetes, other moulds,
yeast-like fungi, Pneumocystis, endemic fungi
Previous colonization Translocation across mucosa Candida spp., Trichosporon spp.
Indwelling catheter (vascular) Direct vascular access,
contaminated product
Candida spp., other yeast-like fungi
Total parenteral nutrition Direct vascular access
Contamination of infusate
Candida spp., Malassezia spp., other yeast-like fungi
Neutropenia (<500 cells/μL) Immunosuppression Aspergillus spp., Candida spp., other moulds and yeast-like
fungi, Pneumocystis
Extensive surgery or burns Route of infection
Direct vascular access
Candida spp., Aspergillus spp., Fusarium spp., Zygomycetes
Assisted ventilation Route of infection Candida spp., Aspergillus spp.
Hospitalization or intensive care
unit stay
Exposure to pathogens
Exposure to additional risk factors
Candida spp., other yeast-like fungi, Aspergillus spp.,
Pneumocystis
Hemodialysis, peritoneal
dialysis
Route of infection
Immunosuppression
Candida spp., Rhodotorula spp., other yeast-like fungi
Malnutrition Immunosuppression Pneumocystis, Candida spp., C. neoformans, endemic fungi
HIV infection/AIDSaImmunosuppression (T-cell decit) C. neoformans, Pneumocystis, Candida spp., Histoplasma,
Coccidioides
Extremes of age Immunosuppression
Numerous co-morbidities
Candida spp., Histoplasma
aHIV, human immunodeciency virus; AIDS, acquired immunodeciency syndrome.
North American invasive mycoses 3
(Hajjeh et al. 2004; Pfaller et al. 2007a; Rees et al. 1998;
Reingold et al. 1986; Wilson et al. 2002; Zaoutis et al.
2005; Zilberberg et al. 2008). Although the incidence of
Pneumocystis pneumonia (PCP) has declined from a
high of approximately 90 infections per 1,000 person-
years in 1995, prior to the introduction of highly active
antiretroviral therapy (HAART), to approximately 30 per
1,000 person-years in 2001 (HAART introduced in 1996),
PCP remains the most common AIDS-dening oppor-
tunistic infection in the U.S. (Morris et al. 2004; Tellez
et al. 2008).
In addition to these agents, the growing list of “other”
opportunistic fungi is of increasing importance (Table
2) (Perlroth et al. 2007; Pfaller et al. 2004a; Pfaller et al.
2006c; Pfaller et al. 2009a; Procop et al. 2004; Richardson
et al. 2008). New and emerging fungal pathogens include
species of Candida and Aspergillus other than C. albicans
and A. fumigatus, opportunistic yeast-like fungi such as
Trichosporon spp., Saccharomyces spp., Rhodotorula
spp., and Blastoschizomyces capitatus, the Zygomycetes,
hyaline moulds such as Fusarium, Acremonium,
Scedosporium, Scopulariopsis, Paecilomyces, and
Trichoderma species, and a wide variety of dematia-
ceous fungi (Table 2) (Castelle et al. 2008; Cortez et al.
2008; Nucci et al. 2007; Pfaller et al. 2007d; Pfaller et al.
2009a; Procop et al. 2004; Richardson et al. 2008; Roden
et al. 2005; Sutton. 2008). Infections caused by these
organisms range from localized infections involving
lung, skin, and paranasal sinuses, to catheter-related
fungemia or peritonitis, to widespread hematogenous
Table 2. Agents of opportunistic and endemic mycoses.a
Opportunistic pathogens
Candida spp. Zygomycetes
C. albicans Rhizopus spp.
C. glabrata Mucor spp.
C. parapsilosis Rhizomucor spp.
C. tropicalis Absidia spp.
C. krusei Cunninghamella sp.
C. lusitaniae
C. guilliermondii Other hyaline moulds
C. rugosa Fusarium spp.
Cryptococcus neoformans and other Scedosporium spp.
opportunistic yeast-like fungi Acremonium spp.
C. neoformans Paecilomyces spp.
Cryptococcus gattii Trichoderma spp.
Trichosporon spp. Dematiaceous moulds
Rhodoturula spp. Alternaria spp.
Saccharomyces cerevisiae Bipolaris spp.
Blastoschizomyces capitatus Cladophialophora spp.
Aspergillus spp. Curvularia spp.
A. fumigatus Exophiala spp.
A. avus Exserohilum spp.
A. niger
A. versicolor Wangiella spp.
A. terreus
A. lentulus Pneumocystis
A. ustus P. jirovecii
Endemic pathogens
Blastomyces dermatitidis Coccidioides posadasii
Coccidioides immitis Histoplasma capsulatum
var. capsulatum
aList not all inclusive.
Table 4. Nosocomial bloodstream infections : most frequent
associated pathogens— Surveillance and Control of Pathogens of
Epidemiological Importance (SCOPE) surveillance program.a
Rank Pathogen % of isolatesb
1 Coagulase-negative staphylococci 31.3
2Staphylococcus aureus 20.2
3Enterococcus spp. 9.4
4Candida spp. 9.0
5Escherichia coli 5.6
6Klebsiella spp. 4.8
7Pseudomonas aeruginosa 4.3
8Enterobacter spp. 3.9
9Serratia spp. 1.7
10 Acinetobacter baumannii 1.3
aData compiled from Wisplingho et al. (2004).
bPercentage of a total of 20,978 infections.
Table 5. Incidence and case-fatality ratios for selected invasive
fungal infections.a
Pathogen
Incidence: no. of
cases per million
per year
Case-fatality
ratio (%) for rst
episode
Candida spp. 72.8 33.9
Cryptococcus neoformans 65.5 12.7
Coccidioides immitis 15.3 11.1
Aspergillus spp. 12.4 23.3
Histoplasma capsulatum 7.1 21.4
Zygomycetes 1.7 30.0
Hyalohyphomycosis 1.2 14.3
Phaeohyphomycosis 1.0 0.0
Sporothrix schenckii <1 20.0
Malassezia furfur <1 0.0
Total 178.3 22.4
aData compiled from Rees et al. (1998).
Table 3. Cumulative incidences of selected invasive mycoses.
Incidence per million per year (period)
CPHAaCDCbNHDScCDCdNHDSe
Mycosis (1980–1982) (1992–1993) (1996) (2000) (2003)
Candidiasis 2.6 72.8 228.2 100.0 290.0
Histoplasmosis 13.9 7.1 13.6 NAfNA
Aspergillosis 8.4 12.4 34.3 NA 22.0
Cryptococcosis 4.0 65.5 29.6 13.0 NA
aCPHA, Commission on Hospital and Professional Activities
(Reingold et al. 1986).
bCDC, Centers for Disease Control and Prevention (Rees et al. 1998).
cNHDS, National Hospital Discharge Survey (Wilson et al. 2002).
dCDC (Hajjeh et al. 2004; Mirza et al. 2003)
eNHDS (Pfaller et al. 2007a).
fNA, data not available.
4 M. A. Pfaller, and D. J. Diekema
dissemination (Pfaller et al. 2004a; Procop et al. 2004;
Richardson et al. 2008). Many of these fungi were previ-
ously thought to be nonpathogenic and now are recog-
nized causes of IFI in compromised patients. Estimates
of the annual incidence of the less common mycoses
have been virtually non-existent; however, data from
a population-based survey conducted by the Centers
for Disease Control (CDC) indicate that zygomycosis
occurs at a rate of 1.7 infections per million per year,
hyalohyphomycosis (Fusarium, Acremonium etc) at 1.2
infections per million per year, and phaeohyphomyco-
sis (dematiaceous moulds) at 1.0 infection per million
per year (Table 5) (Rees et al. 1998). Recent data from a
multicenter, prospective fungal registry, the Prospective
Antifungal erapy (PATH) Alliance, conducted in
23 U.S. medical centers between 2004 and 2008, shows
that the distribution of both common and uncommon
opportunistic IFIs may vary greatly according to the
clinical service and underlying condition of the patient
(Table 6) (Fishman et al. 2008; Horn et al. 2007; Horn
et al. 2009; Lockhart et al. 2009; Neofytos et al. 2009a;
Neofytos et al. 2009b). Among patients at highest risk of
fungal infection are solid organ transplant (SOT) recipi-
ents and hematopoietic stem cell transplant (HSCT)
recipients (Tables 6 and 7). For SOT recipients, the type
of organ transplanted may predispose a patient to one
type of fungal infection over another (Table 7) (Fishman
2007; Fishman et al. 2008; Neofytos et al. 2009b; Procop
et al. 2004), while for HSCT recipients, risk for fungal
infection depends on the degree of immunosuppres-
sion (e.g., higher for allogeneic than for autologous
transplants) (Garcia-Vidal et al. 2008; Marr et al. 2000a;
Marr et al. 2002a; Martin et al. 2003; Morgan et al 2005b;
Neofytos et al. 2009a). Risk factors for fungal infections
in transplant recipients include the use of large doses
of corticosteroids, multiple or acute rejection episodes
(SOT), graft-versus-host disease (HSCT), hypergly-
cemia, poor transplant function, leukopenia, and
advanced age (Fishman. 2007; Garcia-Vidal et al. 2008;
Singh. 2003).
Candida Species Infection
Although the array of fungal pathogens known to cause
IFI is very diverse (Table 2), most of these infections are
due to Candida spp. (Perlroth et al. 2007). Candida spp.
accounted for 88% of all nosocomial fungal infections in
the U.S. between 1980 and 1990 (Jarvis. 1995; Jarvis et al.
Table 6. Distribution of invasive fungal pathogens based on the clinical service or underlying condition of the patient.a,b
Pathogen group
% Infections by clinical service (N)
GMED HEME SCT HIV NICU SOT ST SURG Total
(3,640) (1,010) (377) (263) (54) (886) (863) (1,906) (6,031)
Candida spp. 81.7 42.6 31.6 32.7 96.3 57.2 89.2 91.2 75.0
Cryptococcus spp. 4.0 2.1 0.0 48.7 0.0 6.4 1.6 1.0 4.5
Other yeastsc1.2 3.3 2.7 3.4 0.0 1.0 1.2 0.8 1.4
Aspergillus spp. 8.3 33.8 50.7 4.9 1.9 26.0 4.9 3.4 12.3
Zygomycetes 1.1 5.2 6.4 1.1 1.9 1.7 0.0 0.6 1.4
Other mouldd1.6 7.6 6.4 1.5 0.0 4.7 1.3 1.5 2.7
Endemic fungi 1.9 1.2 0.5 7.6 0.0 2.6 0.8 0.7 1.6
aData compiled from Horn et al. (2007); Horn et al. (2009); Neofytos et al. (2009 a,b).
bGMED, general medicine; HEME, hematologic malignancy; SCT, stem cell transplant; HIV, human immunodeciency virus/acquired
immunodeciency syndrome (AIDS); NICU, neonatal intensive care unit; SOT, solid organ transplant; ST, solid tumor; SURG, surgical
(nontransplant).
cOther yeasts include 6 cases of Malassezia spp., 26 Pneumocystis, 12 Rhodotorula, 21 Saccharomyces, and 6 Trichosporon.
dOther moulds include 2 cases of Acremonium, 9 Alternaria, 3 Bipolaris, 53 Fusarium, 10 Paecilomyces, 13 Scedosporium apiospermum, 6 S.
prolicans, and 1 Sporothrix.
Table 7. Compiled incidence of invasive fungal infections (IFIs) in organ transplant recipientsa
Organ transplant Incidence of IFIs
% IFIsb
Aspergillus Candida Cryptococcus Otherc
Renal 0–20% 11.9 60.6 19.3 8.2
Heart 5–21% 25.0 65.0 2.5 7.5
Liver 5–42% 7.9 78.7 7.1 6.3
Lung and heart-lung 15–35% 63.0 23.9 2.2 14.2
Small bowel 40–59% 2.2% 80–100% NAd0–11%
Pancreas and pancreas-kidney 6–38% 10.5 76.3 0.0 13.2
aData compiled from Fishman et al. (2008); Neofytos et al. (2009a); Singh (2003).
bIFIs were not mutually exclusive (patients could have >1).
cOther, infections due to Zygomycetes (10 infections), endemic fungi (10 infections), other yeast (7 infections), and other moulds (20 infections).
dNA, data not available.
North American invasive mycoses 5
1992; Rees et al. 1998). A more recent multicenter survey
found that Candida spp. accounted for 75% of IFI in hos-
pitalized patients, although the frequency of Candida
associated IFIs varied according to the clinical service
and underlying condition of the patient (Table 6) (Horn
et al. 2007; Horn et al. 2009; Neofytos et al. 2009a; Neofytos
et al. 2009b). Between 1995 and 2002, the frequency of
nosocomial candidemia in U.S. hospitals rose signi-
cantly from 8% to 12% of all reported BSI (Wisplingho
et al. 2004). Wenzel and Gennings (Wenzel et al. 2005),
extrapolating from these data, estimate the annual bur-
den of candidemia to be 10,500 to 42,000 infections in
the U.S., associated with between 2,800-11,200 deaths
per year. Zilberberg et al (Zilberberg et al. 2008) used
billing codes to study the secular trends in candidemia-
related hospitalization in the U.S. and found that the
incidence of candidemia rose by 52% between 2000 and
2005. A similar increase in incidence was seen among all
age groups; however, there was approximately a ten fold
dierence in the candidemia-related hospitalization
incidence between the youngest group (1.53–2.26 cases
per 100,000 population among those aged 18–44 years)
and the oldest group (17.32–25.01 cases per 100,000
population among those aged at least 85 years) through-
out the study period (Table 8). ese data lend support
to the earlier ndings of Wilson et al (Wilson et al. 2002)
and of Pfaller and Diekema (Pfaller et al. 2007a), who
used National Hospital Discharge Survey (NHDS) data
to show that estimates of invasive candidiasis incidence
have been steady or increasing between 1996 and 2003 at
22–29 infections per 100,000 population (Table 3). ese
data include not only candidemia but also other forms
of invasive candidiasis that may not be associated with
positive blood cultures, which may explain why the esti-
mates are higher than several population-based studies,
of candidemia incidence (Table 8) (Perlroth et al. 2007).
is increasing incidence of invasive candidiasis over-
all combined with data from the National Nosocomial
Infection System (NNIS) survey, which show a decline in
the frequency of candidemia among intensive care unit
(ICU) patients in the U.S. (Trick et al. 2002), suggest that
the burden of invasive candidiasis is shifting from the
ICU to the general hospital (and even outpatient) setting
(Hajjeh et al. 2004; Kollef et al. 2008; Sofair et al. 2006).
e predominant source of infection due to Candida
spp., from supercial mucosal and cutaneous disease
to hematogenous dissemination, is the patient. at is,
most types of candidiasis represent endogenous infection
in which the normally commensal host ora take advan-
tage of the “opportunity” to cause infection (Pfaller. 1996).
In order to do so, there must be a lowering of the host’s
anti-Candida barrier. In cases of Candida BSI, transfer
of the organism from the GI mucosa to the bloodstream
requires prior overgrowth of the numbers of yeasts in
their commensal habitat coupled with a breach in the
integrity of the GI mucosa (Agvald-Ohman et al. 2008;
Eggimann et al. 2003; Marco et al. 1999; Pappas. 2006).
Exogenous transmission of Candida may also
account for a proportion of certain types of candidia-
sis (Asmundsdottir et al. 2008). Examples of vehicles
which may introduce Candida into the human host
include the use of contaminated irrigation solutions,
parenteral nutrition uids, vascular pressure trans-
ducers, cardiac valves, and corneas (Pappas 2006).
Transmission of Candida spp., from healthcare work-
ers to patients and from patient to patient has been
well documented, especially in the ICU environment
(Asmundsdottir et al. 2008; Bliss et al. 2008; Mean et al.
2008). e hands of healthcare workers serve as poten-
tial reservoirs of nosocomial transmission of Candida
spp. (Pfaller et al. 1998b; Strausbaugh et al. 1994; Van
Asbeck et al. 2007).
Although >100 species of Candida have been
described, only a few species have been implicated
in clinical infections (Hajjeh et al. 2001; Hazen 1995;
Horn et al. 2009; Pfaller et al. 2007a; Pfaller et al. 2007d).
Candida albicans is the species most commonly recov-
ered from clinical material and generally is responsible
for 90–100% of mucosal infections and for 40%-70% of
episodes of candidemia, although this may vary con-
siderably according to the clinical service on which the
patient is hospitalized (Table 9) (Hachem et al. 2008;
Hajjeh et al. 2004; Horn et al. 2009; Pappas et al. 2003;
Perlroth et al. 2007; Pfaller et al. 2007a; Trick et al. 2002;
Wisplingho et al. 2004).
Approximately 95–97% of all Candida-associated
IFIs are caused by ve species: C. albicans, Candida
glabrata, Candida parapsilosis, Candida tropicalis, and
Candida krusei (Table 9) (Hajjeh et al. 2004; Horn et al.
2009; Pappas et al. 2003; Pfaller et al. 2007a). Among
these common species, only C. glabrata and C. krusei
can be said to be truly “emerging” as causes of IFI, due in
part to their intrinsic and acquired resistance to azoles
and other commonly used antifungal agents (Hachem
et al. 2008; Hajjeh et al. 2004; Haulata et al. 2007; Pfaller
et al. 2003; Pfaller et al. 2007a; Pfaller et al. 2007b; Pfaller
et al. 2008b; Sobel. 2007; Vos et al. 2006). Specic aspects
of each of these species will be addressed below.
e remaining 3–5% of Candida-associated IFIs are
caused by 15–18 dierent species including Candida
guilliermondii, Candida lusitaniae, and Candida rugosa
(Hawkins et al. 2003; Pfaller et al. 2004a; Pfaller et al.
2004b; Pfaller et al. 2006d; Pfaller et al. 2006e; Pfaller
et al. 2007d). Although these species must be consid-
ered to be rare causes of candidiasis, several have been
observed to occur in nosocomial clusters or to exhibit
innate or acquired resistance to one or more established
antifungal agents (Colombo et al. 2003; Dick et al. 1985;
Dube et al. 1994; Hawkins et al. 2003; Kabbara et al. 2008;
Nucci et al. 2005; Sobel 2007).
6 M. A. Pfaller, and D. J. Diekema
Candida albicans
Among the various species of Candida capable of caus-
ing human infection, C. albicans predominates (Table
9). Supercial infections of genital, oral, and cutaneous
sites almost always (>90% of cases) involve C. albicans
(Pfaller et al. 2007d). A wider array of Candida species
causes BSI (Table 9), and although C. albicans predomi-
nates, the frequency with which this and other species
of Candida are recovered from blood samples varies
according to the age of the patient and the local, regional,
or global setting (Abi-Said et al. 1997; Antoniadou et al.
2003; Eggimann et al. 2003; Hachem et al. 2008; Horn
et al. 2009; Kao et al. 1999; Pfaller et al. 2002c; Pfaller
et al. 2004b; Pfaller et al. 2007a; Sofair et al. 2006; Trick
et al. 2002). Globally, a decreasing trend in the rate of
C. albicans isolation (7%-10% decrease) was noted over
an 8.5-year period (1997–2005) among 134 sentinel sur-
veillance sites in 40 countries (Pfaller et al. 2007d). C.
albicans accounted for only 45.6% of Candida BSI in a
recent U.S. multicenter survey, ranging from only 22.4%
of infections among stem cell transplant recipients to
69.2% of infections in the neonatal ICU (NICU) (Table
9) (Horn et al. 2009). BSI due to C. albicans have been
shown to occur less frequently with increasing patient
age (Diekema et al. 2002; Horn et al. 2009; Kao et al. 1999;
Pfaller et al. 2002a; Pfaller et al. 2002c; Pfaller et al. 2004a;
Pfaller et al. 2007a), after exposure to azole antifungals
(Abi-Said et al. 1997; Goldman et al. 2000; Hachem et al.
2008; Hope et al. 2002; Laverdiere et al. 2000; Marr et al.
2000a; Pelz et al. 2001), and in the ICU setting (Trick et al.
2002). Recently, Hachem et al (Hachem et al. 2008) found
that in the setting of a cancer center, factors that were
predictive of C. albicans candidemia were an absence of
neutropenia, the presence of an underlying solid tumor,
and no prior use of prophylactic uconazole (Table 10).
Whereas C. albicans accounted for 45% of Candida BSI in
patients with solid tumors, it was observed in only 14% of
patients with hematologic malignancies (Hachem et al.
2008). ese ndings are supported by those of Horn
et al (Horn et al. 2009) who found C. albicans in 47.6%
of Candida BSI in patients with solid tumors versus only
27.4% in those with hematologic malignancies (Table
9). Chow et al. (Chow et al. 2008) also found C. albicans
to be favored in ICU patients without prior uconazole
exposure; however, this was not supported by Schorr
et al (Schorr et al. 2007) who found that among patients
whose fungemia was diagnosed when they were in an
ICU, no variable dierentiated infection with C. albicans
from that with non-albicans species. ese ndings
emphasize the need to understand that experiences in
various hospital settings may dier greatly with regard to
the epidemiology of candidemia (Riddell et al. 2008). It
Table 10. Multiple logistic regression analysis of independent risk
factors predisposing patients to candidemia caused by dierent
species.a
Candida spp. Risk factor Odds ratio
95% Condence
interval
C. albicans No neutropenia 1.53 1.2–2.44
No uconazole
prophylaxis
3.33 2.04–5.56
Solid tumor 2.50 1.54–4.0
C. tropicalis Neutropenia 2.325 1.287–4.202
C. glabrata Fluconazole
prophylaxis
2.041 1.361–3.060
C. krusei Fluconazole
prophylaxis
5.26 2.922–9.468
Neutropenia 5.378 2.696–10.727
C. parapsilosis Catheter-related
candidemia
2.470 1.587–3.845
aData compiled from Hachem et al. (2008).
Table 9. Species distribution of Candida bloodstream infection isolates by clinical service.a
Candida spp.
% Isolates by species and clinical service (N)b
GMED HEME SCT HIV NICU SOT ST SURG Total
(1,339) (197) (58) (41) (26) (166) (351) (662) (2,019)
C. albicans 46.3 27.4 22.4 43.9 69.2 39.2 47.6 47.9 45.6
C. glabrata 26.6 25.9 32.8 29.3 0.0 38.6 26.8 24.0 26.0
C. parapsilosis 15.7 11.7 15.5 9.8 26.9 12.0 12.8 17.7 15.7
C. tropicalis 7.5 17.3 8.6 7.3 0.0 6.0 7.4 7.3 8.1
C. krusei 1.9 13.7 15.5 4.9 0.0 1.8 2.6 1.4 2.5
Otherc2.0 4.0 5.2 4.8 3.9 2.4 2.8 1.7 2.1
aData compiled from Horn et al. (2009).
bGMED, general medicine; HEME, hematologic malignancy; SCT, stem cell transplant; HIV, human immunodeciency virus/acquired
immunodeciency syndrome (AIDS); NICU, neonatal intensive care unit; SOT, solid organ transplant, ST, solid tumor; SURG; surgical
(nontransplant).
cOther: 17 cases of C. lusitaniae, 5 of C. guilliermondii, 7 of C. dubliniensis, 11 other, and 3 unknown Candida spp.
Table 8. Population incidence of candidemia in the United States by
age group: 2001–2005.a
Age group
(y)
Rate per 100,000 population by year
2001 2002 2003 2004 2005
18–44 1.53 1.80 2.20 2.26 2.01
45–64 5.06 5.97 6.16 6.66 6.81
65–84 14.16 15.51 16.84 16.49 18.64
≥85 17.32 19.46 21.48 21.86 25.01
aData compiled from Zilberberg et al. (2008).
North American invasive mycoses 7
is vital to have knowledge of local epidemiologic trends
in order to guide initial therapy for Candida BSI (Kollef
et al. 2008; Pappas et al. 2009; Riddell et al. 2008; Rijnders
et al. 2000). Although C. albicans is usually considered to
be an endogenous pathogen (i.e., infection arises from
the patient’s own ora), exogenous transmission from
patient to patient via the hands of healthcare personnel
is well documented (Asmundsdottir et al. 2008; Bliss
et al. 2008; Marco et al. 1999).
Candida glabrata
C. glabrata has emerged as an important and poten-
tially antifungal-resistant opportunistic fungal pathogen
(Abi-Said et al. 1997; Alexander et al. 2005; Antoniadou
et al. 2003; Diekema et al. 2002; Eggimann et al. 2003;
Hachem et al. 2008; Hajjeh et al. 2004; Horn et al. 2009;
Kao et al. 1999; Klevay et al. 2008; Magill et al. 2006;
Malani et al. 2005; Marr et al. 2000a; Nucci et al. 2005;
Panackal et al. 2006; Pfaller et al. 2004a; Pfaller et al.
2004b; Pfaller et al. 2007a; Trick et al. 2002). Trick et al
(Trick et al. 2002) have demonstrated that, among the
Candida species, C. glabrata alone has increased as a
cause of BSI in U.S. ICUs since 1993. On a global scale,
the frequency of C. glabrata as a cause of BSI varies from
22–26% in North America to 4%-6% in Latin America
(Diekema et al. 2009a; Pfaller et al. 2004b; Pfaller et al.
2004d; Pfaller et al. 2007d; Pfaller et al. 2009b). Within
the U.S., the proportion of fungemia due to C. glabrata
has been shown to vary from 11% to 37% across the 9 US
Bureau of the Census Regions (Pfaller et al. 2003; Pfaller
et al. 2004b) and from <10% to >30% within single insti-
tutions over the course of several years (Baddley et al.
2001a; Malani et al. 2005). e variation in frequency
of C. glabrata as a cause of BSI across clinical services
has clearly been shown by Horn et al (Horn et al. 2009)
(Table 9) and by Hachem et al (Hachem et al. 2008).
Horn et al (Horn et al. 2009) found that patients with C.
glabrata fungemia were more likely than other patients
with candidemia to be older and to receive a solid organ
transplant, whereas Hachem et al (Hachem et al. 2008)
found that antifungal prophylaxis with uconazole was
a predisposing risk factor for C. glabrata BSI among
cancer patients (Table 10). Review of fungal surveillance
programs conducted in North America from 1992 to the
present shows that the proportion of Candida BSIs due
to C. glabrata has increased signicantly from 8–12%
in 1992 to 24–26% in 2004–2008 (Table 11). Numerous
studies have shown that both colonization and infection
with C. glabrata are rare among infants and children
and increase signicantly with patient age (Diekema
et al. 2002; Hajjeh et al. 2004; Horn et al. 2009; Kao et al.
1999; Kauman. 2001; Laupland et al. 2005; Malani et al.
2005; Pfaller et al. 2002a; Pfaller et al. 2002c; Pfaller et al.
2003; Pfaller et al. 2007a; Rangel-Frausto et al. 1999;
Saiman et al. 2000). Importantly, more than one-third
of Candida-associated BSIs among patients >60 years of
age are due to C. glabrata (Diekema et al. 2002; Kao et al.
1999; Malani et al. 2005). is dramatic variation in the
incidence of C. glabrata fungemia appears to be multi-
factorial (Malani et al. 2005; Pfaller et al. 2004d; Riddell
et al. 2008). It has been shown that the prevalence of this
species is potentially related to disparate factors, includ-
ing geographic characteristics (Pfaller et al. 2004b; Pfaller
et al. 2004d; Pfaller et al. 2007c), age (Diekema et al.
2002; Malani et al. 2005; Pfaller et al. 2003; Pfaller et al.
2007a), and characteristics of the patient population
studied (Abi-Said et al. 1997; Hachem et al. 2008; Marr
et al. 2000a; Marr et al. 2000b; Pasqualotto et al. 2008).
Because C. glabrata is relatively resistant to uconazole,
the frequency with which it causes BSI has important
implications for therapy (Collins et al. 2007; Hachem
et al. 2008; Klevay et al. 2008; Pappas et al. 2009; Parkins
et al. 2007; Riddell et al. 2008; Rijnders et al. 2000).
Candida parapsilosis
C. parapsilosis is the third most common species
of Candida recovered from blood cultures in North
America, accounting for 10%-20% of Candida B S I ( Tables
9 and 11) (Clark et al. 2004; Diekema et al. 2009a; Hajjeh
et al. 2004; Horn et al. 2009; Kao et al. 1999; Laupland
Table 11. Temporal variation in species distribution among bloodstream infection (BSI) isolates of Candida in North America.
Location Study period Reference
No. of
isolates
% Total by Candida spp.a
CA CG CP CT CK
United States 1992–1993 Kao et al. (1999) 837 52 12 21 10 4
United States 1993–1995 Pfaller et al. (2002a) 79 56 15 15 10
United States 1995–1997 Pappas et al. (2003) 1,593 46 20 14 12 2
United States 1995–1998 Pfaller et al. (2002a) 934 53 20 10 12 3
United States 1998–2000 Hajjeh et al (2004) 935 45 24 13 12 2
Canada 1999–2004 Laupland et al. (2005) 209 51 22 6 6 5
North America 2001–2004 Pfaller et al (2007a) 2,773 51 22 14 7 2
North America 2001–2006 Pfaller et al (2008a) 1,489 50 24 14 8 2
North America 2001–2007 Pfaller et al. (2009b) 11,682 49 21 14 7 3
North America 2004–2006 Diekema et al (2009a) 1,657 52 23 14 7 1
North America 2004–2008 Horn et al. (2009) 2,019 46 26 16 8 3
aCA, C. albicans; CG, C. glabrata; CP, C. parapsilosis; CT, C. tropicalis; CK, C. krusei.
8 M. A. Pfaller, and D. J. Diekema
et al. 2005; Pappas et al. 2003; Pfaller et al. 2002a; Pfaller
et al. 2006c; Pfaller et al. 2008a; Pfaller et al. 2008d;
Pfaller et al. 2009b; Safdar et al. 2002; Trofa et al. 2008).
C. parapsilosis is an exogenous pathogen that may be
found on skin rather than mucosal surfaces (Almirante
et al. 2006; Bonassoli et al. 2005; Clark et al. 2004; Kuhn
et al. 2004; Strausbaugh et al. 1994; Trofa et al. 2008; Van
Asbeck et al. 2007). C. parapsilosis is known for the abil-
ity to form biolms on catheters and other implanted
devices (Clark et al. 2004; Kuhn et al. 2004; Levy et al.
1998; Trofa et al. 2008), for nosocomial spread by hand
carriage, and for persistence in the hospital environ-
ment (Almirante et al. 2006; Fridkin. 2005; Kwon-Chung
et al. 2002; Perlroth et al. 2007; Reiss et al. 2008; Safdar
et al. 2002; San Miguel et al. 2005; Sarvikivi et al. 2005;
Trofa et al. 2008; Van Asbeck et al. 2007). It is also well
known for causing infections in infants and neonates
(Table 9) (Fridkin et al. 2006; Kuhn et al. 2004; Levy et al.
1998; Lupetti et al. 2002; Pfaller et al. 2008d; Reiss et al.
2008; Saiman et al. 2000; Sarvikivi et al. 2005; Saxen et al.
1995; Trofa et al. 2008; Van Asbeck et al. 2007; Zaoutis
et al. 2005; Zaoutis et al. 2007). C. parapsilosis aects
critically ill neonates and ICU patients likely because
of its association with parenteral nutrition and central
venous catheters (Clark et al. 2004; Hachem et al. 2008;
Horn et al. 2009; Kuhn et al. 2004; Sarvikivi et al. 2005).
Horn et al. (Horn et al. 2009) found that patients with C.
parapsilosis BSI were more likely than other candidemic
patients to have undergone recent surgery and to have a
peripherally inserted central venous catheter, whereas
Hachem et al (Hachem et al. 2008) reported that C. par-
apsilosis was the most common non-albicans species of
Candida among solid tumor patients and that catheter-
related infection was an independent risk factor for C.
parapsilosis BSI (Table 10). Notably, a recent report by
Kabbara et al (Kabbara et al. 2008) describes break-
through C. parapsilosis BSI in HSCT patients receiv-
ing long-term caspofungin therapy and Forrest et al
(Forrest et al. 2008) found a strong correlation between
caspofungin usage and a 400% increase in cases of C.
parapsilosis BSI. e close MIC-to-therapeutic eect of
the echinocandins towards C. parapsilosis may account
for the presumed selection pressure for this organism
(Forrest et al. 2008; Kabbara et al. 2008; Pfaller et al.
2008a; Pfaller et al. 2008e). Fortunately, BSI due to this
species is associated with a signicantly lower mortality
rate than are infections due to other common species
of Candida (Abi-Said et al. 1997; Almirante et al. 2006;
Hajjeh et al. 2004; Horn et al. 2009; Nguyen et al. 1996a;
Pappas et al. 2003; San Miguel et al. 2005; Trofa et al.
2008).
C. parapsilosis isolates can be divided into three
groups on the basis of molecular studies (Lockhart et al.
2008; Tavanti et al. 2005). e C. parapsilosis complex is
now known to be comprised of three separate species,
C. parapsilosis (formerly C. parapsilosis group I), C.
orthopsilosis (formerly C. parapsilosis group II), and C.
metapsilosis (formerly C. parapsilosis group III) (Tavanti
et al. 2005). Lockhart et al (Lockhart et al. 2008) recently
demonstrated that among 1,929 isolates of presumed C.
parapsilosis, 91.3% were C. parapsilosis, 6.1% were C.
orthopsilosis, and 1.8% were C. metapsilosis. Notably,
the percentage of C. parapsilosis isolates that were C.
orthopsilosis increased from 4.5% in the years 2001–2004
to 8.3% in the years 2005–2006 (Lockhart et al. 2008). C.
orthopsilosis accounted for only 5% of C. parapsilosis
complex isolates from North America, 59% of which
were isolated from patients in the ≥60-year-old age
group (versus only 38% of C. parapsilosis isolates). None
of the C. orthopsilosis or C. metapsilosis isolates were
resistant to uconazole, and all were susceptible to the
echinocandins (Lockhart et al. 2008).
Candida tropicalis
C. tropicalis has long been considered to be an impor-
tant cause of IFI in patients with cancer, especially
leukemia, and in HSCT patients (Abi-Said et al. 1997;
Baran et al. 2001; Hachem et al. 2008; Horn et al. 2009;
Kontoyiannis et al. 2001; Marr et al. 2000a; Wingard.
1995). Horn et al. (Horn et al. 2009) found C. tropicalis
to be especially prominent among patients with hema-
tologic malignancies (17.3% of Candida BSI) versus
patients on other clinical services (0–8.6%) (Table 9) and
Hachem et al. (Hachem et al. 2008) reported that the
presence of neutropenia was an independent risk factor
for C. tropicalis candidemia in cancer patients (Table
10). Among patients with neutropenia who are found
to be colonized with C. tropicalis, as many as 60–80%
eventually develop invasive infection with this species
(Pfaller et al. 1987; Sanford et al. 1980, Wingard. 1995).
As such, C. tropicalis has been considered to exhibit
increased virulence, especially in those individuals
with disrupted mucosal integrity (Kontoyiannis et al.
2001; Walsh et al. 1986; Wingard. 1995). Given these
considerations, prophylaxis treatment with uconazole
for patients with neutropenia has been used in an eort
to decrease infections due to C. tropicalis, as well as C.
albicans (Abi-Said et al. 1997; Antoniadou et al. 2003;
Marr et al. 2000a). Indeed, Hachem et al. (Hachem et al.
2008) has shown a signicant decrease in the frequency
of C. tropicalis BSI associated with the widespread use
of uconazole prophylaxis in patients with leukemia,
lymphoma, and HSCT during the 1990s and 2000s. C.
tropicalis accounted for 23% of Candida BSIs (2nd in
rank order) in the pre-uconazole era (1988–1992) ver-
sus only 9.6% (5th in rank order) in the years following
the introduction of uconazole (1993–2003) (Hachem
et al. 2008). Overall, C. tropicalis accounted for 10–12%
of Candida BSI in North America during the 1990s and
7–8% in the 2000s (Table 11).
North American invasive mycoses 9
Candida krusei
C. krusei accounts for 1–5% of all Candida-associated
BSIs (Table 11) and is best known for its propensity
to emerge in settings where uconazole is used for
prophylaxis (Abi-Said et al. 1997; Antoniadou et al. 2003;
Hachem et al. 2008; Hope et al. 2002; Horn et al. 2009;
Marr et al. 2000a; Pfaller et al. 2008b). Similar to C. tropi-
calis infections, C. krusei infections occur most often in
patients with neutropenia, and colonization of patients
is often predictive of BSI with this species (Abi-Said et al.
1997; Antoniadou et al. 2003; Hachem et al. 2008; Hope
et al. 2002; Horn et al. 2009; Pfaller et al. 1987; Sanford
et al. 1980, Wingard. 1995). Both neutropenia and proph-
ylaxis with uconazole were independent risk factors
for C. krusei BSI at M.D. Anderson Cancer Center (Table
10) where C. krusei increased as a cause of Candida BSI
from 7.4% of infections (5th in rank order) in the pre-
uconazole era (1988–1992) to 24.2% of infections (2nd
in rank order) following the introduction of uconazole
(1993–2003) (Hachem et al. 2008). Likewise, Horn et al
(Horn et al. 2009) found C. krusei BSI to be associated
most commonly with prior use of antifungal agents,
hematologic malignancy, neutropenia and receipt of
HSCT (Table 9). Although C. krusei is best known for
resistance to uconazole, it may also exhibit decreased
susceptibility to amphotericin B and ucytosine, further
complicating therapy (Pappas et al. 2009; Pfaller et al.
2007a; Pfaller et al. 2008b; Spellberg et al. 2006). BSI due
to C. krusei is associated with a high mortality rate (80%
crude mortality and 40% attributable mortality), pos-
sibly related to its poor response to antifungal therapy
(Antoniadou et al. 2003; Horn et al. 2009; Viudes et al.
2002). It should be noted that colonization and infection
with C. krusei were apparent in certain medical centers
well in advance of the use of uconazole (Baran et al.
2001; Iwen et al. 1995; Merz et al. 1986; Wingard. 1995).
Other Candida species
Among the remaining 15–18 species of Candida known
to cause invasive candidiasis (Pfaller et al. 2007d), there
are several that merit discussion either because they
have been shown to cause clusters of infection in the
hospital setting, because they appear to be increasing
in frequency, or because they exhibit decreased suscep-
tibility to one or more antifungal agents and therefore
pose a threat of emergence in certain settings (Atkinson
et al. 2008; Barchiesi et al. 2006; Diekema et al. 2009b;
Kabbara et al. 2008; Nucci et al. 2005; Perlroth et al. 2002;
Pfaller et al. 2004a; Pfaller et al. 2006d; Pfaller et al. 2006e;
Pfaller et al. 2007a; Richardson et al. 2008; Spellberg et al.
2006). ose species addressed in this review include C.
lusitaniae, C. guilliermondii, and C. rugosa.
C. lusitaniae most often causes fungemia in patients
with malignancies or other serious comorbid conditions
(Atkinson et al. 2008; Hawkins et al. 2003). Atkinson and
colleagues (Atkinson et al. 2008) recently compared 13
episodes of C. lusitaniae-related fungemia with 41 epi-
sodes of C. albicans fungemia and found that patients
having C. lusitaniae fungemia were more likely to have
neutropenia, stem cell transplantation, and to have
received prior antifungals. C. lusitaniae is often men-
tioned in the literature as being capable of developing
resistance to amphotericin B during the course of therapy
and may present as breakthrough fungemia in immuno-
compromised patients (Blinkhorn et al. 1989; Hawkins
et al. 2003; Holzshu et al. 1979; McClenny et al. 2002;
Merz 1984; Miller et al. 2006; Minari et al. 2001; Nguyen
et al. 1996b; Pappagianis et al. 1979; Sanchez et al. 1992;
Yoon et al. 1999). Indeed, Atkinson et al (Atkinson et al.
2008) found that in contrast with patients with C. albi-
cans BSI, patients with candidemia due to C. lusitaniae
had an increased treatment failure rate when they were
treated with an amphotericin B-based regimen (10%
vs. 38%, respectively; p = 0.028), and a greater need for
subsequent ICU admission (22% vs. 54%, respectively;
p = 0.04). C. lusitaniae appears to be unique among
Candida species due to an acquired or inducible abil-
ity to exhibit high-frequency phenotypic switching from
amphotericin B susceptibility to resistance upon expo-
sure to the drug (Atkinson et al. 2008; McClenny et al.
2002; Miller et al. 2006; Yoon et al. 1999). Atkinson et al
(Atkinson et al. 2008) demonstrated that amphotericin
B resistance may be readily selected out from originally
amphotericin B-susceptible strains in vitro and that
amphotericin B is considerably less fungicidal against
C. lusitaniae when compared with C. albicans (eec-
tive concentration 50% [EC50] of 4 μg/ml vs 0.5 μg/ml,
respectively). ese ndings indicate that C. lusitaniae,
even originally susceptible to amphotericin B, may be
less amenable to therapy with this agent.
C. guilliermondii and C. rugosa are uncommon spe-
cies of Candida that appear to be increasing in frequency
as causes of invasive candidiasis (Pfaller et al. 2006d;
Pfaller et al. 2006e). Both species have been responsible
for clusters of infection in the hospital setting, and both
demonstrate decreased susceptibility to amphotericin
B, uconazole, and the echinocandins (Colombo et al.
1999; Colombo et al. 2003; Dick et al. 1985; Diekema
et al. 2009b; Dube et al. 1994; Masala et al. 2003; Pfaller
et al. 2004a; Pfaller et al. 2006d; Pfaller et al. 2006e).
C. guilliermondii is best known as a cause of ony-
chomycosis and supercial cutaneous infections
(Dignani et al. 2003; Hay 2003); however, Kao et al (Kao
et al. 1999) found candidemia due to this species to be
common among patients with prior cardiovascular or
abdominal surgery. Likewise, Masala et al. (Masala et al.
2003) reported a nosocomial cluster of catheter-related
infections due to C. guilliermondii among surgical
patients. e infections were successfully managed with
catheter removal and administration of uconazole.
10 M. A. Pfaller, and D. J. Diekema
One of the initial descriptions of invasive candidiasis
due to C. guilliermondii was that of a fatal case of dis-
seminated infection in which the patient succumbed
despite amphotericin B therapy (Dick et al. 1985). e
isolate was subsequently shown by in vitro testing to
be resistant to amphotericin B. More recently Kabbara
et al. (Kabbara et al. 2008) reported a breakthrough C.
guilliermondii BSI in a HSCT recipient receiving caspo-
fungin prophylaxis. As with C. parapsilosis (Pfaller et al.
2008d), C. guilliermondii is known to show reduced
susceptibility to the echinocandin class of antifungal
agents (Pfaller et al. 2006d; Pfaller et al. 2008a; Pfaller
et al. 2008e). is reduced susceptibility may come into
play when infections with C. guilliermondii involve
anatomical sites where adequate free drug levels can-
not be readily obtained (Pfaller et al. 2008a; Pfaller
et al. 2008e).
C. rugosa is a rare cause of fungemia (Reinhardt et al.
1985); however, it has been implicated in clusters of
nosocomial fungemia in burn patients in the U.S. (Dube
et al. 1994). C. rugosa is reported to exhibit decreased
susceptibility to polyenes, azoles, and echinocandins
and may cause catheter-related fungemia in seriously ill
patients (Diekema et al. 2009b; Dube et al. 1994; Nucci
et al. 2005; Pfaller et al. 2006e; Reinhardt et al. 1985).
In a multicenter survey of invasive candidiasis (Pfaller
et al. 2006e), we have found that C. rugosa was recovered
most often in cultures of blood and urine obtained from
patients hospitalized on medical and surgical in-patient
services.
Risk factors
e burden of invasive candidiasis is tremendous in
terms of morbidity, mortality, and cost, and it is clear
that we must do more than simply seek better therapeu-
tic agents if we are to impact this burden (Diekema et al.
2004; Fridkin 2005; Garey et al. 2006; Garey et al. 2007;
Morgan et al 2005a; Morrell et al. 2005; Parkins et al. 2007;
Wenzel et al. 2005). Fungal BSIs have been shown to have
some of the highest rates of inappropriate therapy and
hospital mortality among all etiologic agents examined
(Harbarth et al. 2002; Ibrahim et al. 2000; Klevay et al.
2008; Morgan et al 2005a; Morrell et al. 2005; Parkins
et al. 2007). e most common causes of inappropriate
therapy for fungal BSIs are omission of initial empirical
therapy (Garey et al. 2006; Klevay et al. 2008; Morrell
et al. 2005; Parkins et al. 2007) followed by incorrect dos-
ing of uconazole (Armstrong-James. 2007; Garey et al.
2007; Horn et al. 2009; Parkins et al. 2007; Patel et al.
2005). Such inadequate therapy has been linked directly
to mortality (Garey et al. 2006; Klevay et al. 2008; Morrell
et al. 2005; Parkins et al. 2007). us, despite an unprec-
edented array of new, potent, and non-toxic antifungal
agents, we are failing in the management of these infec-
tions (Diekema et al. 2004; Garey et al. 2006; Garey et al.
2007; Morrell et al. 2005; Parkins et al. 2007; Pfaller et al.
2007a; Puzniak et al. 2006).
Lack of specic clinical ndings and slow, insensitive
diagnostic testing complicate the early recognition and
treatment of invasive candidiasis (Alexander et al. 2006;
Morrell et al. 2005; Munoz et al. 2000; Ostrosky-Zeichner
et al. 2006; Parkins et al. 2007; Schorr et al. 2007). Most
authors recommend the use of clinical risk factors to
identify patients who may benet from prophylactic or
empirical antifungal therapy in the proper clinical set-
ting (Charles 2006; Kollef et al. 2008; Munoz et al. 2000;
Ostrosky-Zeichner 2004; Paphitou et al. 2005; Procop
et al. 2004; Rex. 2006; Sobel et al. 2001; Wenzel et al. 2005).
Unfortunately, the predominant risk factors for invasive
candidiasis are common iatrogenic and/or nosocomial
conditions (Table 1) (Perlroth et al. 2007). Additional
meaningful stratication of identied risk factors will be
required to identify those select high-risk patients who
would derive maximal benet from early therapeutic
interventions (Sobel et al. 2001; Wenzel et al. 2005).
It is important to realize that the risk for invasive can-
didiasis is a continuum (Diekema et al. 2004; Lockhart
et al. 2009; Pfaller et al. 2006c; Sobel et al. 2001; Wenzel
et al. 2005). Certain hospitalized individuals are clearly
at increased risk of acquiring candidemia during hos-
pitalization as a result of their underlying medical
condition: patients with hematologic malignancy and/
or neutropenia, those undergoing GI surgery, prema-
ture infants, and patients greater than 70 years of age
(Table 1). Among patients with candidemia in the U.S.,
the mean time to onset of candidemia was 22 days of
hospitalization (Pappas et al. 2003; Wisplingho et al.
2004). us, it must be emphasized rst and foremost
that invasive candidiasis typically aects individuals
with severe illness who have prolonged hospitalizations
(Perlroth et al. 2007).
Within the high-risk groups, specic additional expo-
sures have been recognized to further increase the risk
of invasive candidiasis: the presence of vascular cath-
eters, exposure to broad-spectrum antimicrobial agents,
renal failure, mucosal colonization with Candida spp.,
prolonged ICU stay, and receipt of total parenteral nutri-
tion (TPN) (Table 12) (Blumberg et al. 2001; Ostrosky-
Zeichner et al. 2007; Wenzel et al. 2005; Wey et al. 1989).
Compared to controls without the specic risk factors
or exposures, the likelihood of these already high-risk
patients contracting candidemia in hospital is approxi-
mately 2 times greater for each class of antibiotics they
receive, 7 times greater if they have a central venous
catheter, 10 times greater if Candida has been found to
be colonizing other anatomic sites, and 18 times greater
if the patient has undergone hemodialysis (Table 12)
(Wenzel et al. 2005; Wey et al. 1989). Hospitalization in
the ICU provides the opportunity for transmission of
Candida among patients (Asmundsdottir et al. 2008;
North American invasive mycoses 11
Mean et al. 2008) and has been shown to be an addi-
tional independent risk factor (Blumberg et al. 2001;
Eggimann et al. 2003; Ostrosky-Zeichner. 2004; Saiman
et al. 2000; Wenzel et al. 2005). Notably, the single most
important risk factor for candidemia among patients
hospitalized in the surgical ICU is prolonged (>7 days)
stay in the ICU (Ostrosky-Zeichner et al. 2007; Pelz et al.
2001). Several investigators have now used these risk
factors to develop clinical risk assessment strategies that
could be used in the ICU to (1) predict certain rates of
invasive candidiasis, (2) capture a substantial propor-
tion of patients who actually go on to develop invasive
candidiasis, and (3) be practical for use as selection
tools for risk-targeted prevention (prophylaxis) or treat-
ment (preemptive or empirical) strategies (Leon et al.
2006; Ostrosky-Zeichner et al. 2007; Wenzel et al. 2005).
Preliminary application of these strategies show that risk
stratication is possible and practical; however, their
clinical utility remains to be established in prospective
studies (Mean et al. 2008).
Mortality, length of stay, and cost
e consequences of candidemia in hospitalized
patients are severe. Patients with candidemia have
been shown to be at a 2-fold greater risk of death during
hospitalization than are patients with noncandidal BSI
(Pittet et al. 1997). Among all patients with nosocomial
BSI, candidemia was found to be an independent pre-
dictor of death during hospitalization (Miller et al. 1987;
Pittet et al. 1997). In a multicenter U.S. study of candi-
demia, risk factors for mortality included an APACHE II
score >18 (P < .001), cancer (P = .002), the presence of a
urinary catheter (P = .004), male sex (P = .004), the use of
corticosteroids (P < .001) and the presence of an arterial
catheter (P < .001) (Pappas et al. 2003).
Estimates of the mortality attributable to candidemia
and other forms of invasive candidiasis have been
reported from retrospective matched-cohort studies
conducted in single institutions (Gudlaugsson et al.
2003; Pelz et al. 2000; Puzniak et al. 2006; Wey et al. 1988)
and in the context of population surveillance studies
(Table 13) (Morgan et al. 2005a; Zaoutis et al. 2005). e
weight of the evidence provided by these studies sug-
gests that candidemia or invasive candidiasis is associ-
ated with an important attributable mortality ranging
from 10% to 49% (Pfaller et al. 2007a; Wenzel et al. 2005).
Furthermore, these data demonstrate that candidemia
carries no less risk of death during hospitalization today
than it did 20 years ago (Diekema et al. 2004), despite the
introduction of new antifungal agents with good activ-
ity against most species of Candida (Perlroth et al. 2007;
Spellberg et al. 2006).
Treatment of candidemia is often found to be inad-
equate due to delay in administration of therapy, treat-
ment with an agent to which the organism is resistant,
inadequate dose or duration of treatment, or no treat-
ment at all (Armstrong-James 2007; Atkinson et al. 2008;
Garey et al. 2006; Garey et al. 2007; Klevay et al. 2008;
Morgan et al. 2005a; Morrell et al. 2005; Parkins et al.
2007). Several studies have now shown that delays in
the initiation of adequate antifungal therapy of >12 h
(Morrell et al. 2005), >24 h (Garey et al. 2006; Garrouste-
Orgeas et al. 2006; Parkins et al. 2007), and >48 h (Blot
et al. 2002) were independently associated with mortal-
ity in candidemia patients. A population-based study of
candidemia found that removal of vascular catheters, in
addition to receipt of at least 5 days of antifungal treat-
ment, was independently associated with a decreased
risk for both early and late mortality (Almirante et al.
2006). Likewise, Morgan et al (Morgan et al. 2005a)
demonstrated that the attributable mortality rate was
lower among patients who received adequate (>7 days)
treatment for candidemia (11% in Connecticut and
Table 12. Factors for increased risk of high-risk patients contracting
candidemia in the hospital setting compared with control subjects
without specic risk factors or exposures.a,b
Risk factors Fold increased risk
Each class of antimicrobial received 2
Patient has a central venous catheter 7
Candida colonization 10
Patient has undergone acute hemodialysis 18
aData compiled from Wenzel et al. (2005) and Wey et al. (1989).
bHospitalization in an intensive care unit is an independent risk
factor.
Table 13. Incidence, mortality rates, and costs attributable to candidemia in the United States.a
Period
Location/patient
population
No. cases per
100,000/yr Mortality (%) LOS (days) Cost ($) Reference
1983–1986 Iowa NA 38.0 30.0 NA Wey et al. (1988)
1996 Baltimore, MD/SICU NA 19.0 17.0 21,590 Pelz et al. (2000)
1997–2001 Iowa 6.0 49.0 10.5 NA Gudlaugsson et al. (2003)
1998–1999 Connecticut 7.1 19.0 3.4 6,214 Morgan et al (2005a)
1998–2000 Baltimore, MD 24.0 24.0 12.9 29,094 Morgan et al (2005a)
2000 U.S./pediatrics 43.0 10.0 21.1 92,266 Zaoutis et al. (2005)
2000 U.S./adults 30.0 14.5 10.1 39,331 Zaoutis et al. (2005)
2000 St. Louis, MO NA 35.7 NA 44,051 Puzniak et al. (2004)
aSICU, surgical intensive care unit; NA, data not available; LOS, length of stay in hospital.
12 M. A. Pfaller, and D. J. Diekema
16% in the Baltimore metropolitan area) than among
patients who did not receive adequate treatment (31% in
Connecticut and 41% in Baltimore). Finally, Parkins et al
(Parkins et al. 2007) found that empirical therapy with
an antifungal agent to which the organism was suscep-
tible in vitro was associated with a signicant reduction
in all-cause mortality from 46% to 27% (P = .02). Notably,
empirical uconazole therapy was more likely to be
deemed inadequate (due to both inadequate dosing as
well as in vitro resistance) and inadequate therapy was
an independent predictor of death in hospital (Parkins
et al. 2007). us, reduction of the mortality due to
candidemia and invasive candidiasis is dependent on
administration of appropriate antifungal therapy (right
drug and dose) early in the course of infection and for
an adequate duration.
Several studies have examined the excess length of
stay (LOS) and hospital costs attributable to IFI due to
Candida (Table 13). Candidemia patients have been
shown to have between 3 and 30 more hospital days than
uninfected patients with the same underlying disease
and disease severity (Table 13). Given the prevalence
of Candida infections and their attributable impact
on mortality and LOS, it is not surprising that these
infections are associated with substantial health care
costs (Fridkin 2005; Miller et al. 2001a; Pelz et al. 2000;
Perlroth et al. 2007; Rentz et al. 1998; Wilson et al. 2002).
e excess costs attributable to candidemia range from
$6,214 to $92,266 per episode depending on geographic
location and patient type (Table 13). It is estimated
that 85% of the increase in cost of care for patients with
candidemia is due to the excess LOS (Rentz et al. 1998).
Because each case of candidemia adds tens of thousands
of dollars to hospitalization costs, the estimated health
care cost associated with hematogenously disseminated
candidiasis is $2–4 billion/year in the U.S. alone (Miller
et al. 2001a; Perlroth et al. 2007; Rentz et al. 1998; Wilson
et al. 2002; Zaoutis et al. 2005).
Antifungal susceptibility
Among the eight species of Candida discussed in this
review, C. albicans, C. parapsilosis, C. tropicalis, C.
lusitaniae, and C. guilliermondii remain reliably sus-
ceptible to ucytosine, the azoles (except for itracona-
zole), and the echinocandin antifungal agents (Table
14) (Diekema et al. 2009a; Diekema et al. 2009b; Pfaller
et al. 2002b; Pfaller et al. 2002d; Pfaller et al. 2004e;
Pfaller et al. 2005a; Pfaller et al. 2005c; Pfaller et al.
2007b; Pfaller et al. 2008a; Pfaller et al. 2008e). Both C.
parapsilosis and C. guilliermondii are known to exhibit
higher MICs than other species of Candida for the echi-
nocandins (modal MIC, 0.25–2 μg/ml vs. 0.06-0.12μg/
ml, respectively); however, 92–100% of all clinical BSI
isolates of these species are susceptible to echinocan-
dins at the Clinical and Laboratory Standards Institute
(CLSI) breakpoint of ≤2 μg/ml (Pfaller et al. 2008a;
Pfaller et al. 2008e).
Although 94% of C. guilliermondii isolates are sus-
ceptible to uconazole, the MIC90 value of 8μg/ml is
considerably higher than that determined for other
uconazole-susceptible species (e.g., 0.5–2 μg/ml for
C. albicans, C. parapsilosis, C. tropicalis, C. lusitaniae).
Likewise, C. glabrata, C. krusei, and C. rugosa are all
inherently less susceptible to uconazole (Table 14).
Although both voriconazole and posaconazole are
active against the majority of these isolates, cross-
resistance within the azole class is well documented
for C. glabrata and C. rugosa (Ostrosky-Zeichner et al.
2003; Pfaller et al. 2006a; Pfaller et al. 2006b; Pfaller et al.
2006e; Pfaller et al. 2007a; Pfaller et al. 2007c; Pfaller
et al. 2008c). Both C. glabrata and C. krusei are very
susceptible to the echinocandins (Pfaller et al. 2008a;
Pfaller et al. 2008e), whereas these agents are consider-
able less potent against C. rugosa (Table 14) (Diekema
et al. 2009b). In addition to its intrinsic resistance to u-
conazole, C. krusei also shows decreased susceptibility
to ucytosine (Table 14) (Pfaller et al. 2002b).
Agar-based susceptibility testing methods such as
Etest (AB BIODISK Solna, Sweden) have proven to be
the most sensitive and reliable means by which to detect
resistance to amphotericin B among Candida species
(Clancy et al. 1999; Krogh-Madsen et al. 2006; McClenny
et al. 2002; Park et al. 2006; Pfaller et al. 1998a; Pfaller
et al. 2004c; Wanger et al. 1995). Although interpretive
breakpoints for amphotericin B have not been estab-
lished, isolates of Candida for which MICs are >1μg/ml
are unusual and possibly “resistant” or, at the very least,
may require high doses of amphotericin B for optimal
treatment (Pappas et al. 2009; Rex et al. 2002; Spellberg
et al. 2006). Given these considerations, it is now evi-
dent that C. glabrata, C. krusei, and C. rugosa exhibit
decreased susceptibility to amphotericin B compared
with C. albicans (Table 15) (Diekema et al. 2009b; Pfaller
2005; Pfaller et al. 2004c; Pfaller et al. 2007a). Whereas
C. guilliermondii and C. lusitaniae have been described
as amphotericin B-resistant Candida species (Atkinson
et al. 2008; Dick et al. 1985; Pfaller et al. 2006d), both of
these species appear to be susceptible to amphotericin
B upon initial isolation from blood (Table 15). us,
resistance to amphotericin B may develop secondarily
during treatment and repeat amphotericin B suscepti-
bility testing is recommended for patients with persist-
ent infection with either species while on amphotericin
B therapy (Atkinson et al. 2008; McClenny et al. 2002;
Pfaller et al. 2006d).
Cryptococcus Infection
Cryptococcosis is a systemic mycosis caused by the
encapsulated, basidiomycetous, yeast-like fungi
North American invasive mycoses 13
Cryptococcus neoformans and Cryptococcus gattii.
ese species were previously classied on the basis
of antigenic dierences in the capsular polysaccharide
into ve serotypes of C. neoformans (A, B, C, D, and AD)
which were grouped into two varieties: var neoformans
(serotypes A,D, and AD) and var gattii (serotypes B and
C) (Kwon-Chung 1975; Kwon-Chung 1976). Given their
signicant divergence at the molecular level, the two
Table 14. Comparative in vitro susceptibility of clinical isolates of Candida spp. to ucytosine, azoles, and echinocandins determined by Clinical
and Laboratory Standards Institute broth microdilution methods.a
Candida spp. Antifungal agent No. tested
Minimum inhibitory concentration (MIC) (µg/ml)b%c
Range 50% 90% S R/NS
C. albicans Flucytosine 5,208 0.12–>128 0.25 1 97.0 3.0
Fluconazole 5,827 0.007–>128 0.25 0.5 99.3 0.1
Itraconazole 7,647 0.007–>8 0.06 0.12 96.1 0.7
Posaconazole 5,827 0.007–2 0.015 0.06 99.9 0.0
Voriconazole 5,826 0.007–4 0.007 0.015 99.9 <0.1
Anidulafungin 2,869 0.007–2 0.03 0.06 100.0 0.0
Caspofungin 2,869 0.007–0.5 0.03 0.06 100.0 0.0
Micafungin 2,869 0.007–1 0.015 0.03 100.0 0.0
C. glabrata Flucytosine 1,267 0.12–>128 0.12 0.12 99.0 1.0
Fluconazole 1,517 0.25–>128 8 32 53.1 9.6
Itraconazole 1,929 0.03–>8 1 2 2.3 59.9
Posaconazole 1,517 0.03–>8 1 2 79.6 8.3
Voriconazole 1,516 0.007–8 0.25 1 91.1 5.7
Anidulafungin 747 0.015–4 0.06 0.12 99.9 0.1
Caspofungin 747 0.015–>8 0.03 0.06 99.9 0.1
Micafungin 747 0.007–1 0.015 0.015 100.0 0.0
C. parapsilosis Flucytosine 1,047 0.12–>128 0.12 0.25 99.0 1.0
Fluconazole 1,542 0.12–>128 0.5 2 96.2 0.6
Itraconazole 1,508 0.15–2 0.25 0.5 47.8 3.1
Posaconazole 1,542 0.007–1 0.06 0.12 100.0 0.0
Voriconazole 1,541 0.007–8 0.015 0.06 99.6 0.1
Anidulafungin 759 0.03–4 2 2 92.5 7.5
Caspofungin 759 0.015–4 0.25 1 99.9 0.1
Micafungin 759 0.015–2 1 2 100.0 0.0
C. tropicalis Flucytosine 759 0.12–>128 0.25 1 92.0 7.0
Fluconazole 1,198 0.12–128 0.5 2 99.2 0.2
Itraconazole 1,207 0.015–>8 0.12 0.5 53.0 3.1
Posaconazole 1,198 0.015–2 0.06 0.12 99.9 0.0
Voriconazole 1,197 0.007–8 0.03 0.06 99.8 <0.1
Anidulafungin 625 0.007–2 0.03 0.06 100.0 0.0
Caspofungin 625 0.007–>8 0.03 0.06 99.8 0.2
Micafungin 625 0.007–1 0.03 0.06 100.0 0.0
C. krusei Flucytosine 184 0.12–>128 16 32 5.0 28.0
Fluconazole 305 8–>128 32 64 1.3 34.1
Itraconazole 306 0.12–4 1 1 1.0 56.2
Posaconazole 305 0.03–4 0.25 1 99.0 0.3
Voriconazole 305 0.007–4 0.25 0.5 99.7 0.3
Anidulafungin 136 0.015–0.5 0.06 0.06 100.0 0.0
Caspofungin 136 0.015–1 0.12 0.25 100.0 0.0
Micafungin 136 0.015–0.25 0.06 0.12 100.0 0.0
C. lusitaniae Flucytosine 82 0.12–>128 0.12 0.5 93.0 5.0
Fluconazole 171 0.12–64 0.5 1 97.0 1.0
Itraconazole 129 0.03–2 0.25 0.25 48.8 2.3
Posaconazole 171 0.06–1 0.06 0.12 100.0 0.0
Voriconazole 171 0.06–1 0.06 0.06 100.0 0.0
Anidulafungin 171 0.06–1 0.25 0.5 100.0 0.0
Caspofungin 171 0.06–4 0.25 0.5 99.0 1.0
Micafungin 171 0.06–1 0.12 0.25 100.0 0.0
Table 14. continued on next page
14 M. A. Pfaller, and D. J. Diekema
varieties of C. neoformans have recently been awarded
separate species status, as C. neoformans and C. gattii.
In addition, dierentiation of C. neoformans into two
varieties, var grubii (serotype A) and var neoformans
(serotype D) has been proposed (Boekhout et al. 2001;
Kwon-Chung et al. 2002). For purposes of this review we
will refer to C. neoformans (serotypes A, D, and AD) and
C. gattii (serotypes B and C).
C. neoformans is found worldwide, invariably isolated
from pigeon droppings and soil (Emmons.1955). More
recently, the environmental habitat of C. neoformans
appears to be related to trees and plant material, speci-
cally a specialized niche resulting from the natural bio-
degradation of wood (Lazera et al. 1996). Pigeons only
contribute to the propagation and spread of the fungus
by providing an enriched medium for fungal growth and
by dispersing the fungus via their contaminated beaks
and feet (Littman et al. 1968).
C. gattii is usually found in tropical and subtropical
regions in association with Eucalyptus trees. Recently,
however, C. gattii has been isolated from several envi-
ronmental samples (e.g. r and oak trees) in Vancouver
Island, British Columbia mainland, as well as in the states
of Oregon and Washington associated with human and
animal infections (Bartlett et al. 2007; Kidd et al. 2007;
McDougal et al. 2007).
Mode of acquisition
Cryptococcosis is usually acquired by inhaling aero-
solized cells of C. neoformans or C. gattii from the envi-
ronment. Subsequent dissemination from the lungs,
usually to the central nervous system (CNS), produces
clinical disease in susceptible individuals. Primary
cutaneous cryptococcosis may occur rarely following
transcutaneous inoculation (Naka et al. 1995); however,
the majority of cases of cutaneous cryptococcosis reect
hematogenous dissemination (Hay 1985; Sarosi et al.
1971).
Incidence
Prior to the AIDS epidemic, cryptococcal infection was
diagnosed in less than 1,000 patients per year in the U.S.
(Graybill et al. 2000; Hajjeh et al. 1999). Indeed, among
persons without AIDS the incidence of cryptococcosis
in the U.S. has been shown to be 0.2 to 0.8 per 100,000
population per year and has remained unchanged
for more than a decade (Hajjeh et al. 1999; Mirza
et al. 2003).
Candida spp. Antifungal agent No. tested
Minimum inhibitory concentration (MIC) (µg/ml)b%c
Range 50% 90% S R/NS
C. guilliermondii Flucytosine 100 0.12–4 0.12 0.5 100.0 0.0
Fluconazole 162 0.5–64 4 8 94.0 0.0
Itraconazole 90 0.06–4 0.5 1 4.0 27.0
Posaconazole 162 0.06–2 0.25 0.5 98.0 0.0
Voriconazole 162 0.06–2 0.06 0.12 99.0 0.0
Anidulafungin 162 0.06–4 1 2 93.0 7.0
Caspofungin 162 0.06–8 0.5 1 96.0 4.0
Micafungin 162 0.06–8 0.5 1 99.0 1.0
C. rugosa Fluconazole 16 0.5–16 2 16 69.0 0.0
Itraconazole 15 0.03–0.5 0.12 0.5 60.0 0.0
Posaconazole 16 0.06–0.25 0.06 0.25 100.0 0.0
Voriconazole 16 0.06–0.25 0.06 0.25 100.0 0.0
Anidulafungin 16 0.06–8 1 8 80.0 20.0
Caspofungin 16 0.06–4 0.5 2 91.0 9.0
Micafungin 16 0.06–0.25 0.06 0.25 100.0 0.0
aData compiled from Diekema et al. (2009b); Pfaller et al. (2002b; 2005a; 2006a; 2006b; 2007b; 2008a).
b50% and 90%, MIC encompassing 50% and 90% of all isolates tested, respectively.
cS, susceptible at MIC breakpoint for itraconazole (≤0.12 µg/ml); ucytosine (≤4 µg/ml) uconazole (≤8 µg/ml); posaconazole and voriconazole
(≤1 µg/ml); anidulafungin, caspofungin, micafungin (≤2 µg/ml); R, resistant at MIC breakpoint for itraconazole (≥1 µg/ml); ucytosine (≥32 µg/
ml); uconazole (≥64 µg/ml); posaconazole and voriconazole (≥4 µg/ml); NS, nonsusceptible at MIC breakpoint ≥ 4 µg/ml for anidulafungin,
caspofungin, and micafungin.
Table 14. Continued.
Table 15. Is amphotericin B uniformly active against Candida spp.?a
Candida spp.
No isolates
tested
Minimum inhibitory
concentration (MIC) (µg/ml)b
50% 90%
C. albicans 4,195 0.5 1
C. glabrata 949 2 4
C. krusei 234 4 8
C. lusitaniae 171 0.25 0.5
C. dubliniensis 101 0.25 0.5
C. guilliermondii 162 0.25 0.5
C. rugosa 16 2 4
aData compiled from Diekema et al. (2009b); Pfaller (2005); Pfaller
et al. (2004c; 2004d).
bAmphotericin B MICs were determined by using Etest (Pfaller et al.
[2004c]). 50% and 90%, MICs for 50% and 90% of isolates tested,
respectively.
North American invasive mycoses 15
A dramatic increase in the incidence of cryptococco-
sis was observed with the advent of the AIDS pandemic
(Table 3), and subsequently HIV infection has been
associated with more than 80% of cryptococcosis cases
worldwide (Dromer et al. 1996; Hajjeh et al. 1999; Mirza
et al. 2003). During the 1990s, the prevalence of crypto-
coccosis progressively declined in developed countries,
rst as a result of the widespread use of uconazole and
later due to successful treatment of HIV infection with
the use of highly active antiretroviral therapy (HAART)
(Tables 3 and 16) (Mirza et al. 2003). A population-based
surveillance study conducted by the CDC during 1992
through 2000 in Atlanta, GA and Houston, TX showed
that the overall mean annual incidence of cryptococ-
cosis decreased from 4–5 cases per 100,000 population
in 1992/1993 to 0.4–1.3 cases per 100,000 population in
2000 (Table 16) (Mirza et al. 2003). e majority of cases
(89%) occurred in persons known to be HIV infected. e
annual incidence of cryptococcosis in the Atlanta area
among persons with HIV infection ranged from 66 cases
per 1,000 persons in 1992 to 7 cases per 1,000 in 2000
and in the Houston area the incidence ranged from 23.6
cases per 1,000 persons in 1993 to 1.6 cases per 1,000 per-
sons in 2000 (Table 16) (Mirza et al. 2003). Despite this
progress, cryptococcosis continues to carry a signicant
morbidity and mortality: the annual case fatality ratio in
Atlanta and Houston was 11% among HIV-infected per-
sons and 21% among HIV-uninfected individuals and
did not change signicantly over the 8-year study period
(Mirza et al. 2003). In industrialized countries crypto-
coccosis continues to occur in those with undiagnosed
HIV infection and in socio-economically disadvantaged
HIV-infected people without access to HAART or other
HIV-supportive care (Mirza et al. 2003). In Atlanta and
Houston fewer than one-third of patients infected with
HIV who had cryptococcosis had received HAART
before being diagnosed with cryptococcosis (Mirza et al.
2003). Lack of access to HAART and antifungal therapy
is a major problem in resource-limited regions such as
Africa and Southeast Asia where both AIDS and crypto-
coccosis are rampant (Viviani et al. 2009).
e incidence of cryptococcosis among HIV-
uninfected persons remains low. It varied from 0.4 cases
per million population in 1992 in the Atlanta area to 5
cases per million population in 1994 in the Houston
area, with no signicant changes noted during the study
period (Mirza et al. 2003). Compared with HIV-infected
persons with cryptococcosis, HIV-uninfected persons
with cryptococcosis tended to be older (median age, 36
years vs. 56 years, respectively) and most cases occurred
in white persons (40% vs. 66% of cases, respectively).
At least one underlying condition (diabetes, cancer,
lung disease) was reported in 82% of cases (Mirza et al.
2003).
Cryptococcus neoformans is the most common spe-
cies aecting patients with AIDS and other immuno-
compromising conditions (Chayakulkeeree et al. 2006;
Mirza et al. 2003; Viviani et al. 2009), whereas C. gattii
infections occur mainly in immunocompetent hosts
in endemic regions throughout the world (Speed et al.
1995). e emergence of C. gattii infections in immu-
nocompetent human and animal populations in the
Pacic Northwest region of North America is nothing
short of remarkable (Bartlett et al. 2007). Prior to 1999
no evidence of C. gattii infections existed on Vancouver
Island, but their sudden emergence since this time is
well documented (Bartlett et al. 2007; Kidd et al. 2005).
Study of the environment in this and other areas of the
Pacic Northwest showed a striking change in the envi-
ronmental niches for this species to a variety of trees in
a temperate climate (Bartlett et al. 2007). e incidence
in this new endemic area has reached 35 cases per mil-
lion population per year, markedly higher than rates
reported in other endemic areas such as Australia (1.2
cases per million per year) (McDougal et al. 2007; Sorrell
2001). Molecular studies have shown the potential
direct link between the environmental and clinical iso-
lates from the Vancouver Island outbreak (Bartlett et al.
2007). Furthermore, the major genotype was found to
be more virulent in an animal model than the parental
strain (Fraser et al. 2005). Compared with C. neoformans
infections, cryptococcosis caused by C. gattii is associ-
ated with a lower mortality rate, but is characterized by
more severe neurologic sequelae due to the formation of
granulomas that require surgery and prolonged therapy
(Speed et al. 1995). Although C. gattii clearly is a patho-
gen of immunocompetent hosts, a recent epidemiologi-
cal survey in Southern California showed that C. gattii
produced disease in a substantial number of persons
infected with HIV (Chaturvedi et al. 2005).
Risk factors
Cryptococcosis, particularly meningitis, commonly
occurs in patients with underlying immunodeciency
Table 16. Incidence of cryptococcosis in two U.S. cities before
and after the introduction of highly active antiretroviral therapy
(HAART).a
Location Period (y)
Annual incidence
%
Decrease
Cases/100,000
population
Cases/1,000
HIV-infectedb
Atlanta, GA Pre-HAART
(1992)
5 66
Post-HAART
(2000)
1.3 7 89
Houston, TX Pre-HAART
(1993)
4 23.6
Post-HAART
(2000)
0.4 1.6 92
aData compiled from Mirza et al. (2003).
bHIV, human immunodeciency virus.
16 M. A. Pfaller, and D. J. Diekema
(Table 1) (Chayakulkeeree et al. 2006); however, both
local and disseminated infections are observed in
patients with no known immunologic defect (Mirza
et al. 2003; Mitchell et al. 1995; Pappas et al. 2001).
Characteristically, normal or immunocompetent indi-
viduals may develop cryptococcosis due to C. gattii in
those regions where this species is endemic; however, C.
gattii is now being found in AIDS patients (Chaturvedi
et al. 2005); suggesting that the reduced frequency of
C. gattii infections in immunosuppressed patients may
be partly caused by a limited environmental exposure
as opposed to a specic tropism for those individu-
als who are immunocompetent (Bartlett et al. 2007).
Cryptococcosis is uncommon in children, with a preva-
lence of approximately 1% among children with AIDS.
Immunosuppressed patients at particular risk for
cryptococcal infection include those with malignancies
or sarcoid and those receiving corticosteroid therapy,
organ transplants, or immunosuppressive therapy.
Among neoplastic disorders lymphoproliferative malig-
nancies, mainly Hodgkin’s lymphoma, are known to be
the major predisposing diseases (Viviani et al. 2009).
e AIDS epidemic has clearly demonstrated the
importance of a defect in cell-mediated immunity as the
major immunologic risk factor for cryptococcosis. ose
individuals with CD4+ lymphocyte counts of less than
100 per cubic mm are at high risk for CNS and dissemi-
nated cryptococcosis (Mirza et al. 2003). Despite the
favorable impact of HAART on the incidence of crypto-
coccosis in AIDS patients, the number of patients with
cryptococcosis in medically developed countries has
not approached zero as risk groups broaden in concert
with new developments in transplantation medicine
and the creation of new therapies to manipulate immu-
nity (Fishman 2007; Fishman et al. 2008). e increased
number of patients undergoing SOT and the use of cor-
ticosteroids and other immunosuppressive agents, such
as alemtuzumab and iniximab, can produce a pro-
foundly immunosuppressive state and allow reactivation
of a cryptococcal infection (Hage et al. 2003; Nath et al.
2005). e type of immunosuppression after