The n e w e ng l a n d j o u r na l of m e dic i n e

review article

Medical Progress

Aspergillosis
Brahm H. Segal, M.D.

F
From the Departments of Medicine and ilamentous fungi (molds) are ancient lineages that have existed
Immunology, Roswell Park Cancer Insti- for approximately 1 billion years1 and thrive in soil and decomposing vegeta-
tute, Buffalo, NY. Address reprint re-
quests to Dr. Segal at the Departments of tion independent of an animal host. Thus, the evolution from primitive im-
Medicine and Immunology, Roswell Park mune systems that rely principally on antimicrobial peptides, such as those in insects,2
Cancer Institute, Elm and Carlton Sts., to the complex immune system in mammals occurred with continued exposure
Buffalo, NY 14263, or at brahm.segal@
roswellpark.org. to fungi. The immune system, therefore, must not only recognize inhaled molds
and control their growth but also restrain injurious inflammation and allergy.
N Engl J Med 2009;360:1870-84. We regularly inhale the spores of aspergillus species, yet fungal disease is uncom-
Copyright © 2009 Massachusetts Medical Society.
mon. Aspergillus-related diseases are associated with a spectrum of disorders of im-
munity. Invasive aspergillosis, the focus of this review, is typically a disease of highly
immunocompromised persons and is a leading cause of infection-related death in
patients with acute leukemia and recipients of allogeneic hematopoietic stem-cell
transplants. At the other end of the immunologic spectrum, allergic forms of asper-
gillosis, such as allergic bronchopulmonary aspergillosis, result from a poorly
controlled inflammatory response to hyphae colonizing the sinopulmonary tract.
Important advances have been made in fungal diagnostics and in the antifungal
armamentarium. In addition, new insights have been gained regarding host defense
against aspergillus species and the immunopathogenesis of aspergillus-related dis-
eases that may pave the way to new prevention and therapeutic strategies.

Hos t Defense ag a ins t A spergil lus

Innate Immunity
Respiratory epithelial cells act as an anatomic barrier to invasion by inhaled asper-
gillus species, promote mucociliary clearance, and ingest inhaled conidia (spores).
The ability of aspergillus species to survive within epithelial cells may enable evasion
of host defense by phagocytes.3 Alveolar macrophages constitute the first line of
phagocytic host defense against inhaled conidia.4 Peripheral-blood monocytes and
neutrophils are subsequently recruited to sites of infection. After fungal germination
(transformation from conidia to hyphae), neutrophils are the dominant host de-
fense against hyphae, the tissue-invasive form of molds.4 Natural killer cells are
recruited to the lungs by chemokines early in experimental aspergillosis and play an
important host-defense function.5
The NADPH oxidase in phagocytes is essential in host defense against aspergil-
losis, as demonstrated in patients with chronic granulomatous disease, an inherited
disorder of NADPH oxidase. Chronic granulomatous disease is associated with re-
current bacterial and fungal diseases, and invasive aspergillosis is a major cause of
death in patients with this disease.6 The activation of NADPH oxidase results in the
conversion of oxygen to superoxide anion and the generation of downstream reactive
oxidant metabolites with antimicrobial activity. In neutrophils, the activation of
NADPH oxidase is coupled with activation of antimicrobial proteases sequestered

1870 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

in primary granules.7 The neutrophil NADPH oxi- sure on the surface of germinating aspergillus
dase is activated by constituents of the fungal-cell conidia.19-21 Coordinated ligation of specific patho-
wall and is required to induce hyphal damage. In gen-recognition receptors through the identifica-
contrast, neutrophil-mediated inhibition of conid- tion of stage-dependent fungal motifs probably
ial growth is independent of NADPH oxidase but calibrates the immune response to contain fungal
requires lactoferrin-mediated iron depletion, illus- growth while avoiding excessive inflammation af-
trating the stage-specific antifungal activity of ter inhalation of ubiquitous aspergillus spores.
neutrophil-regulated host defense.8 In addition to Myeloid differentiation factor 88 (MyD88) is a
its host-defense function, NADPH oxidase has a downstream adapter for most TLRs. The work of
counterbalancing role to restrain inflammation von Bernuth et al.22 showed that children with
induced by fungal-cell-wall constituents,9,10 a func- autosomal recessive MyD88 deficiency have life-
tion that is probably mediated in part by the acti- threatening bacterial infections but not other in-
vation of tryptophan catabolism.10 fectious complications. These findings suggest
Pathogen-recognition receptors in the host that a redundancy in antifungal host-defense path-
identify specific microbial motifs, such as cell-wall ways exists and that TLR-independent pathways
products (e.g., endotoxin produced by gram-neg- may be sufficient for protection. However, TLR
ative bacteria and beta-glucans produced by fungi) signaling is probably important in defense against
and microbial DNA and RNA, and trigger innate aspergillosis for patients in immunocompromised
immune responses. Several classes of cell-associ- states, such as occurs after transplantation.23
ated and soluble pathogen-recognition receptors
that identify fungal motifs include toll-like recep- Cellular Immunity
tors (TLRs), dectin-1, surfactant proteins A and The activation of pathogen-recognition receptors
D,11 mannose-binding lectin,12 and pentraxin-313 generally induces maturation of antigen-presenting
(Fig. 1). TLRs are a conserved family of pathogen- cells that prime T-cell immunity. Interferon-γ is the
recognition receptors that are homologous to signature cytokine of type 1 helper T cells (Th1)
interleukin-1 receptor 1 and share a similar sig- and stimulates cellular immunity. CD4+ T-cell dif-
naling cascade, leading to the activation of tran- ferentiation during experimental aspergillosis
scriptional factor nuclear factor κB (NF-κB) and occurs in stages, with TLR-independent signals
mitogen-activated protein kinases. Activation of promoting Th1 differentiation in the lung and
TLRs, in general, induces the expression of proin- priming of TLR-dependent Th1 occurring in lymph
flammatory cytokines. Studies in TLR-deficient nodes.24 In murine aspergillosis, augmentation
mice have shown that specific TLRs recognize of Th1 responses by administration or depletion
different fungal motifs and regulate the inflam- of specific cytokines enhances antifungal host de-
matory response in aspergillosis.14-16 In addition, fense.25 In patients with invasive aspergillosis, in-
the stimulation of specific TLRs can modulate creased Th1 versus type 2 helper T-cell (Th2) cy-
human neutrophil antifungal activity.14 tokine responses were associated with improved
Dectin-1 is a receptor and immunomodulator outcomes.26,27
of beta-glucans, which are ubiquitous cell-wall Th2 cells produce interleukin-4, interleukin-5,
constituents of fungi and plants. Dectin-1 is a and interleukin-13 and are implicated in allergy.
natural-killer-cell–receptor-like C-type lectin that Allergic bronchopulmonary aspergillosis develops
is expressed at high levels within the pulmonary from sensitization to bronchial airway A. fumigatus
and gastrointestinal tract, where exposure to antigens that prime Th2 responses.28 Suppression
pathogens regularly occurs.17 Dectin-1 and TLR2 of Th2 responses attenuates experimental allergic
recognize distinct beta-glucan motifs and stimu- bronchopulmonary aspergillosis.28 Interleukin-17–
late proinflammatory cytokine production.18 Li- producing CD4+ T cells are a distinct subgroup
gation of dectin-1 may stimulate NADPH oxidase of T helper cells that are implicated in autoim-
activation.18 mune diseases.29 Interleukin-17 stimulates the
TLRs and dectin-1 permit host cells to dis- production of specific myelopoietic growth factors
tinguish between resting conidia, germinating and cytokines and chemokines that promote neu-
conidia, and hyphae of Aspergillus fumigatus. Fungal trophil recruitment, but paradoxically, interleu-
beta-glucans trigger inflammatory responses in kin-17 can impair host defense in experimental
macrophages through their time-dependent expo- aspergillosis.10,30 Regulatory T cells have a coun-

n engl j med 360;18  nejm.org  april 30, 2009 1871

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Innate immunity Adaptive immunity

Germlings

Conidia Hyphae Mannose-binding ↓ Fungal

Th1
lectin burden
Pentraxin 3

TLR-4 TLR-9

Th2 ↑ Allergy

TLR-2

Dectin-1
↑ Expression
of mediators
Th17
of neutrophil
Activation of nuclear factor κB recruitment
Up-regulation of specific cytokines
and chemokines
Maturation of dendritic cells that
prime T-cell responses

↓ Cellular
NADPH Reactive oxidant Tregs immunity
oxidase activation generation and allergy

Figure 1. Model of Immune Response to Inhaled Aspergillus Species.

Alveolar macrophages and epithelial cells are the first host cells in the lung to engage inhaled aspergillus conidia. Pathogen-recognition
receptors on host cells, such as dectin-1 and toll-like receptors (TLRs), and soluble pathogen-recognition receptors, such as pentraxin 3
and mannose-binding C O L lectin,
OR FIG recognize
URE specific fungal motifs. For example, fungal-cell-wall–associated beta-glucans ligate TLR-2 and
dectin-1, and aspergillus DNA contains unmethylated CpG sequences that ligate TLR-9. Ligation of pathogen-recognition receptors gen-
Draft 3 04/09/09
erally leads to induction of chemokines and cytokines that activate and recruit neutrophils and other inflammatory cells. NADPH oxi-
Author Segal
dase can be activated by dectin-1 and potentially primed by other pathogen-recognition receptors such as TLR4. NADPH oxidase acti­
Fig # 1
vation leads
Title
to generation of reactive oxidant intermediates and in neutrophils is coupled to the activation of antimicrobial granular
proteases.ME
Dendritic cells also sense aspergillus motifs through pathogen-recognition receptors and stimulate antigen-dependent re-
sponses DE in helper T cells (Th) and regulatory T cells (Tregs). The cross-talk between dendritic cells and T cells in regulating T-cell devel-
opment Artist
is an area SBL
of intensive research interest that is broadly relevant to host defense, allergy, and vaccine development. Most of the
data in this modelAUTHOR
are derived from in vitro studies and animal models of aspergillosis.
PLEASE NOTE:
Figure has been redrawn and type has been reset
Please check carefully

Issue date
terregulatory function by inducing tolerance and turn, mediates control of fungal burden and al-
protecting against allergy in experimental asper- lergy versus tolerance (Fig. 1).
gillosis.31 The balance between the promotion and
the suppression of specific classes of cytokines and Spec t rum of Hum a n Dise a se
chemokines and between the expansion and the in A spergil l osis
contraction of specific T-cell populations modu-
lates the extent and nature of the immune re- Aspergillosis encompasses a spectrum of diseases
sponse to aspergillus species. This response, in related to host factors (Table 1).32-37 Acute invasive

1872 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

Table 1. Predisposing Host Factors and Clinical and Histologic Features Associated with Invasive Pulmonary Aspergillosis.*

Population of Patients Predisposing Host Factors Clinical and Histologic Features

Acute leukemia, myelodysplastic syndrome, Neutropenia from chemotherapy or under­ Hyphal angioinvasion with vascular thrombo-
aplastic anemia, and other causes of lying hematologic disease sis and tissue infarction; scant inflam­
marrow failure matory response; possible evolution to
cavitation
Allogeneic hematopoietic stem-cell trans- Immunosuppression for GVHD (e.g., the use Inflammatory fungal pneumonia and an­gio­
plantation after neutrophil recovery of corticosteroids, T-cell depletion, inhibi- invasion with coagulative necrosis, which
tion of tumor necrosis factor α) are classically associated with neutro­
penia32-34
Solid-organ transplantation Immunosuppression to prevent allograft Acute inflammatory pneumonia, chronic ne-
­rejection crotizing aspergillosis; tracheobronchitis
affecting the anastomotic site and causing
dehiscence in lung-transplant recipients
Advanced AIDS CD4+ T-cell count usually <100 per cubic Acute to slowly progressive necrotizing pneu-
­millimeter; other immu­nocompromising monia; variable histologic findings: neu-
conditions (e.g., neutropenia) trophilic infiltrates, vascular invasion,
walled-off abscesses, and cavitation35
Chronic granulomatous disease Defective NADPH oxidase Acute to slowly progressive pneumonia; pyo-
granulomatous inflammation without
­hyphal vascular invasion or coagulative
necrosis; “mulch pneumonitis,” an acute
hypersensitivity response to a large, aero-
solized exposure36
Preexisting structural lung disease (e.g., Coexisting conditions (e.g., diabetes and Chronic necrotizing pulmonary aspergillosis;
emphysema, previous cavitary tubercu- ­malnutrition) and the use of inhaled slowly progressive invasive fungal pneu-
losis) and systemic corticosteroids monia with inflammatory necrosis37

* AIDS denotes acquired immunodeficiency syndrome, and GVHD graft-versus-host disease.

aspergillosis is a rapidly progressive, frequently fa- In va si v e A spergil l osis

tal disease that occurs in highly immunocompro-
mised persons. In contrast, chronic forms of pul- Epidemiology
monary aspergillosis (e.g., chronic necrotizing Patients who are at risk for invasive aspergillosis
pulmonary aspergillosis or fibrocavitary aspergil- include those with prolonged neutropenia, recipi-
losis) typically occur in patients without severe im- ents of hematopoietic stem-cell transplants or sol-
mune impairment, progress over months to years, id-organ transplants, and patients with advanced
and require prolonged antifungal therapy.38,39 acquired immunodeficiency syndrome or chronic
Aspergilloma is a fungal mass that develops in a granulomatous disease.41 Host factors govern both
preexisting lung cavity. Medical therapy is of un- the risk and clinicopathologic features of invasive
certain value for aspergillomas; in cases of persis- aspergillosis (Table 1).32-34 The number of asper-
tent hemoptysis, surgical resection of the infected gillosis-related deaths increased by a factor of four
lung cavity, if feasible, is the definitive therapy.40 in the 1980s and 1990s in the United States on the
In immunocompetent persons, aspergillus spe- basis of autopsy reports,42 a reflection of an in-
cies can induce allergic responses, manifesting creased number of immunocompromised patients.
as sinusitis and asthma. Allergic bronchopulmo- In patients with neutropenia, the degree and
nary aspergillosis is a severe Th2-mediated aller- duration of neutropenia predict the risk of invasive
gic disease that occurs in 1 to 2% of patients with aspergillosis. Intensive cytotoxic chemotherapy
asthma and in 1 to 15% of patients with cystic fi- (e.g., induction regimens for acute leukemia) causes
brosis.28 Saprophytic hyphal elements that colo- prolonged neutropenia. The incidence of invasive
nize bronchial airways cause recurrent inflamma- aspergillosis in patients with acute leukemia was
tion and airway plugging, transient pulmonary approximately 5% in a multicenter European reg-
infiltrates, and central bronchiectasis as a long- istry.43 Patients with refractory leukemia who are
term consequence. treated with multiple cycles of cytotoxic chemo-

n engl j med 360;18  nejm.org  april 30, 2009 1873

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

therapy (and who may have aplasia from the un- (e.g., vasculitis), can rarely be complicated by
derlying cancer) are at particularly high risk for invasive aspergillosis.54,55
aspergillosis. There is also a growing appreciation of invasive
In recipients of allogeneic hematopoietic stem- pulmonary aspergillosis in patients without clas-
cell transplants, three periods of risk for invasive sic risk factors, such as critically ill patients with-
aspergillosis and other mold diseases occur: first, out documented immunodeficiency.56 Research on
neutropenia after the conditioning regimen; sec- these patients may identify unrecognized host-
ond, acute graft-versus-host disease (GVHD); and defense pathways against aspergillosis.
third, chronic GVHD (occurring at least 100 days
after transplantation).44 Invasive aspergillosis oc- Clinical Manifestations and Diagnosis
curs more often during GVHD than during neu- Invasive aspergillosis principally involves the sino­
tropenia in recipients of allogeneic hematopoietic pulmonary tract, reflecting that inhalation is the
stem-cell transplants, with the severity of GVHD most common route of entry of aspergillus spores;
and the intensity of immunosuppressive therapy other entry sites such as the gastrointestinal tract
being the principal risk factors.45-47 and skin occur on rare occasions. Asymmetric fa-
The degree of disparity between the donor and cial swelling, epistaxis, proptosis and cranial-nerve
the recipient in human leukocyte antigen (HLA) abnormalities (reflecting orbital disease or cavern-
is the major predictor of the incidence and the ous sinus involvement), ischemia of the palate, and
severity of GVHD. In a prospective U.S. surveil- bone erosion are signs suggestive of invasive fun-
lance study, the aggregate cumulative incidence of gal sinusitis. Fever, cough, and dyspnea are fre-
invasive aspergillosis at 12 months after allogeneic quent although nonspecific findings of pulmo-
hematopoietic stem-cell transplantation was 2.3% nary aspergillosis; the lung is the most common
with an HLA-matched related donor (providing site of invasive aspergillosis. Vascular invasion may
the highest level of HLA donor–recipient concor- manifest as pleuritic chest pain from pulmonary
dance), 3.2% with an HLA-mismatched related infarction or as hemoptysis. Poorly controlled in-
donor, and 3.9% with an unrelated donor.48 In fection may lead to extension to mediastinal and
contrast, the incidence of invasive aspergillosis was chest-wall structures and hematogenous dissem-
0.5% after autologous hematopoietic stem-cell ination that can involve virtually any organ. Involve-
transplantation. Similar results were observed in ment of the central nervous system is a devastat-
a multicenter Italian registry, with aspergillosis ing consequence of disseminated aspergillosis and
being more common and more likely to be fatal may be manifested by seizures or focal neurologic
in recipients of allogeneic hematopoietic stem cells signs from mass effect or stroke.
than in recipients of autologous hematopoietic Diagnosis of invasive aspergillosis remains dif-
stem cells.49 T-cell depletion of allografts to pre- ficult in that clinical manifestations are not spe-
vent GVHD impairs reconstitution of cellular im- cific; radiologic findings can be suggestive but
munity and increases the risk of invasive asper- none are pathognomonic, and cultures of respira-
gillosis and cytomegalovirus infection.50 tory samples lack sensitivity. Histologic demon-
Among recipients of solid-organ transplants, stration of invasive hyphae or a positive culture
the incidence of invasive aspergillosis is highest from a normally sterile environment (e.g., pleural
after lung transplantation.48 Aspergillus species fluid) represents proven invasive fungal disease.
that colonize the airways in end-stage lung disease Newer antigen-based assays facilitate the diag-
may be a source of fungal infection after single- nosis of probable invasive aspergillosis and can
lung transplantation.51 In recipients of solid-organ obviate the need for an invasive procedure. The
transplants, late-onset invasive aspergillosis (oc- diagnosis of probable invasive aspergillosis re-
curring 3 months after transplantation) is becom- quires a combination of host factors (e.g., pro-
ing increasingly recognized and is associated with longed neutropenia and organ transplantation),
the intensity of immunosuppression to treat al- compatible radiologic findings, and mycologic
lograft rejection52 and retransplantation.53 In pa- criteria.57 These diagnostic criteria, though de-
tients who have not undergone transplantation, signed for clinical research, can be applied to
intensive immunosuppressive regimens, such as clinical practice. Mold-active treatment is frequent-
those used to treat serious autoimmune disorders ly and appropriately initiated on the basis of a

1874 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

lower level of evidence for invasive aspergillosis with significant heterogeneity in diagnostic accu-
than these intentionally restrictive criteria, given racy among different groups of patients.65 The use
the potential for rapid progression of aspergil- of antifungal agents with activity against molds
losis and death. decreases the sensitivity of the galactomannan as-
In patients with neutropenia, persistent fever say.66 There are causes of false positive results,
may be the only sign of invasive fungal disease. including concomitant use of piperacillin–tazobac-
Computed tomography (CT) of the chest is more tam, other beta-lactam antibiotics, and gluconate-
sensitive than radiography for the detection of containing intravenous fluids. Cross-reactivity with
early pulmonary aspergillosis58 and should be con- other fungi (e.g., Histoplasma capsulatum) that have
sidered in patients with 10 to 14 days of neutro- a cell-wall galactomannan similar to that of asper-
penia (neutrophil count, <500 per cubic millime- gillus species can cause positive results.67
ter) and persistent or recurrent fever of unknown In a patient who is at high risk for invasive
cause that is unresponsive to empirical antibac- aspergillosis and who has a compatible radiologic
terial agents.59 The earliest radiologic sign of in- lesion (e.g., nodule or infiltrate), a positive serum
vasive aspergillosis is a nodule.60 A “halo sign,” galactomannan assay or culture of an aspergillus
defined as a macronodule surrounded by a perim- species from respiratory secretions provides strong
eter of ground-glass opacity corresponding to al- evidence for invasive aspergillosis and can avert
veolar hemorrhage, is suggestive of invasive asper- the need for an invasive procedure. Galactoman-
gillosis in patients with compatible host factors. nan detection in bronchoalveolar lavage fluid ap-
The initiation of treatment on the basis of this pears to be more sensitive than detection in se-
sign has been associated with an improved re- rum56,67 and can be used to support a diagnosis
sponse, as compared with initiation for more ad- of probable aspergillosis.57 Diagnostic testing of
vanced fungal disease.60,61 However, other mold bronchoalveolar lavage fluid should include bac-
pathogens (e.g., zygomycetes) and bacterial patho- terial, fungal, and viral pathogens on the basis of
gens capable of angioinvasion (e.g., Pseudomo- the nature of the immunocompromised state and
nas aeruginosa) can produce a similar appearance. radiologic findings.59 As an alternative to bron-
Other radiographic findings that are associated choscopy, percutaneous lung biopsy may be at-
with invasive aspergillosis are consolidation, tempted for analysis of peripheral nodules. Tho-
wedge-shaped infarcts, and cavitation, with the racoscopic lung biopsy should be considered in
latter typically occurring after neutrophil recov- a patient whose condition is deteriorating when
ery (Fig. 2). less invasive procedures have produced negative
Mycologic criteria require either isolation of results. Thrombocytopenia may limit the ability
aspergillus species from the sinopulmonary tract to perform invasive procedures.
or positive antigen-based laboratory markers. Cul- Diagnosis of invasive fungal diseases with the
tures of bronchoalveolar lavage fluid have at best use of polymerase-chain-reaction assay, although
a sensitivity of 50% in focal pulmonary lesions.62 promising, is currently investigational. Potential
Antigen-based diagnosis relies on serum detection advantages include rapidity, low cost, and the abil-
of either galactomannan or beta-d-glucan, two ity to establish a diagnosis at the species level and
constituents of fungal-cell walls. The galactoman- to detect genes that confer antifungal resistance.
nan assay is relatively specific for invasive aspergil- Limitations include a lack of standardized meth-
losis, whereas the beta-d-glucan assay also detects ods, difficulty in reliably distinguishing fungal
other invasive fungal diseases, including candidi- colonization from disease, and the potential for
asis, other mold pathogens (excluding zygomy- contamination with fungal DNA.68
cetes), and Pneumocystis jiroveci (formerly called
P. carinii).63,64 Therapy for Invasive Aspergillosis
The serum galactomannan assay has been prin- The guidelines of the Infectious Diseases Society
cipally studied in patients with leukemia and re- of America40 recommend the use of voriconazole
cipients of hematopoietic stem-cell transplants, as the primary therapy for invasive aspergillosis.
and its performance varies in different reports. In Voriconazole was more effective than amphoteri-
a meta-analysis, the serum galactomannan assay cin B deoxycholate as initial therapy for invasive
had a sensitivity of 71% and a specificity of 89%, aspergillosis and was associated with significantly

n engl j med 360;18  nejm.org  april 30, 2009 1875

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

improved survival (71% vs. 58%) in a randomized

Figure 2 (facing page). Host Factors in Radiologic
trial.61 The rate of successful outcomes was also and Pathologic Features of Invasive Aspergillosis.
superior among patients who received voriconazole, Panel A shows a computed tomographic chest scan of a
as compared with amphotericin B deoxycholate patient with neutropenia who has invasive aspergillosis
(53% vs. 32%). The poorest responses to antifun- in the left upper lung (arrow). A positive serum galacto-
gal therapy occurred in patients with extrapulmo- mannan test established the diagnosis of probable inva-
sive aspergillosis, which averted the need for an invasive
nary aspergillosis and in recipients of allogeneic
diagnostic procedure. Panel B shows cavitation of the le-
hematopoietic stem-cell transplants. sion after a successful response to therapy and neutrophil
Urgent débridement of localized aspergillosis, recovery (arrow). Panel C shows vascular invasive asper-
such as sinusitis, cutaneous disease, and osteo- gillosis, which is classically associated with neutropenia
myelitis, should be performed. Since zygomycosis but can occur in patients with other conditions, such as
in this case of fatal aspergillosis in a recipient of an al-
is frequently manifested in sino-orbital disease,69
logeneic hematopoietic stem-cell transplant who did not
therapy with an amphotericin B formulation, have neutropenia but did have severe graft-­versus-host
which has activity against aspergillus and zygo- disease. A low-power micrograph shows vascular throm-
mycete species, is advised in cases of suspected bosis (with an arterial vessel outlined by arrows) that is
invasive fungal sinusitis, pending culture and his- surrounded by extensive coagulative necrosis and hemor-
rhage (hematoxylin and eosin). In Panel D, a high-power
tologic results.
micrograph shows hyphae (arrowheads) transversing
Nonlinear elimination of voriconazole in adults the blood vessel wall (outlined by arrows) and intravas-
has important implications for dose selection. cular invasion (Gomori methena­mine-silver, with hyphal
According to the package insert, it is estimated walls staining dark). The septated hyphae, some branched
that increasing the oral dose of voriconazole from at acute angles, are morphologically consistent with
aspergillus species. However, other molds can have a
200 mg every 12 hours to 300 mg every 12 hours
similar appearance, so culture or molecular-based analy-
leads to an increase in exposure in adults by a sis at a reference laboratory is required for a definitive
factor of 2.5 (area under the concentration–time diagnosis. Panel E shows ex­perimental aspergillosis in
curve). In contrast, clearance of voriconazole in a knockout mouse model of chronic granulomatous dis-
children is linear and more rapid than in adults, ease, an inherited disorder of NADPH oxidase. Densely
inflammatory pyogranulomatous pneumonia without
which necessitates a higher dose per kilogram of
vascular invasion or tissue infarction is visible (hematox-
body weight to achieve an exposure similar to that ylin and eosin), with invasive hyphae in the lung as seen
in adults (Table 2).70-76 There can be considerable with silver staining (inset). These histologic features are
variability over time in voriconazole exposure, both similar to those observed in patients with chronic gran-
between patients and in the same patient.77,78 ulomatous disease and aspergillosis and suggest that
NADPH oxidase–independent pathways are able to de-
Small studies have noted a relationship between
fend against hyphal ­invasion of blood-vessel walls.
low plasma voriconazole levels and treatment fail-
ure79,80 and between high voriconazole levels and
toxicity.80,81 Therapeutic monitoring of voricon- lactomannan assays have been used as a thera-
azole should be considered in cases of refractory peutic marker for invasive aspergillosis, in which
fungal disease or suspected drug toxicity. a falling level correlates with a successful response
When a failure in therapy is suspected, several and a rising level with failure.86 Such an approach
things should be evaluated.82 First, the diagnosis is promising, but more research is required to
of invasive aspergillosis may be incorrect (or two evaluate its predictive value in different groups
infections may exist simultaneously). Second, in of patients.
patients with persistent neutropenia, pulmonary There is a paucity of data to guide the admin-
lesions can increase in size, despite eventual re- istration of antifungal therapy in patients with
sponse to antifungal therapy.83 Third, pulmonary invasive aspergillosis resistant to voriconazole; op-
lesions may increase in size after neutrophil re- tions include a lipid amphotericin B formulation,
covery because of immune reconstitution rather an echinocandin, and combination antifungal
than failure of therapy.84 Fourth, subtherapeutic therapy.40 Lipid formulations of amphotericin B
systemic levels of mold-active azoles may occur. are less nephrotoxic than amphotericin B deoxy-
Finally, resistance of aspergillus species isolates cholate (Table 3).87-90 In one study, amphotericin
to mold-active azoles is uncommon,85 but it is also B colloidal dispersion had a similar efficacy but
a potential cause of therapeutic failure. A. terreus less nephrotoxicity than amphotericin B deoxy-
is resistant to amphotericin B. Serial serum ga- cholate as primary therapy for invasive aspergil-

1876 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

A B

C D

losis.90 An analysis of a large data registry on the tolerability in patients with renal impairment.89
use of amphotericin B lipid complex therapy for Liposomal amphotericin B that was admin­istered
in­vasive aspergillosis showed encouraging findings at a daily dose of 3 mg per kilogram per day was
regarding efficacy and safety, including
ICM the drug’s
AUTHOR segal associated with similar
RETAKE 1st efficacy, less nephrotoxic-
REG F FIGURE fig 2 a_e 2nd
CASE 3rd
TITLE Revised
nEMail Line
engl j med 360;18  nejm.org  4-C 30, 2009
april 1877
Enon ARTIST: mleahy SIZE
H/T H/T
FILL Combo 33p9
Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .
AUTHOR, PLEASE
Copyright © 2009 Massachusetts NOTE: Society. All rights reserved.
Medical
Figure has been redrawn and type has been reset.
Please check carefully.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 2. Characteristics of Mold-Active Azoles.*

Antifungal Agent Dose† Pharmacologic Properties‡ Comments

Voriconazole§ Intravenous therapy: 6 mg per kilo- Oral bioavailability, approximately Drug of choice as primary therapy for
gram every 12 hr for 2 doses, then 95%; substantial variability in sys- invasive aspergillosis (with superi-
4 mg per kilogram every 12 hr as temic exposure among patients; ority over amphotericin B in a ran-
61; oral therapy: for
initial therapy genetic polymorphisms for domized trial)61; consideration of
adults, 200 mg twice daily or 4 mg CYP2C19, with 15 to 20% of therapeutic drug monitoring in se-
per kilogram twice daily40; for chil- Asians expected to have a slow lect cases; poor activity against zy-
dren 2 to 11 yr of age, 7 mg per ki- rate of metabolism gomycetes
logram twice daily without a load-
ing dose; for children >11 yr of
age, adult dose
Itraconazole§¶ Oral capsules: for adults, 400 mg dai- Oral bioavailability of the solution is Recommended monitoring of serum
ly (in either one or two doses); better than the capsules, but both trough levels aiming for >0.25 μg
oral solution: 2.5 mg per kilogram formulations have significant vari- per milliliter on high-performance
twice daily; for children >5 yr of ability in systemic exposure; oral liquid chromatography40 because
age: 2.5 mg per kilogram twice bioavailability of capsules (but not of negative inotropic effects; con-
daily; studies in younger children the solution) is increased when traindicated in patients with sub-
are lacking; intravenous therapy: taken with food and is reduced by stantial ventricular dysfunction or
200 mg twice daily for four doses, increased gastric pH a history of congestive heart fail-
then 200 mg daily40 ure; effective as prophylaxis in
chronic granulomatous dis-
ease70,71; effective as therapy for
corticosteroid-dependent allergic
bronchopulmonary aspergillo-
sis72,73
Posaconazole Oral prophylaxis: 200 mg three times Only available as an oral formulation, Effective as prophylaxis in patients
daily in patients ≥13 yr of age at which should be taken with food with neutropenia associated with
high risk for invasive fungal dis- or liquid nutritional supplement; a the myelodysplastic syndrome or
ease74,75; doses of 200 mg four high-fat meal increases bioavail- acute myelogenous leukemia74
times daily and 400 mg twice daily ability; substantial variability in and in recipients of allogeneic he-
have been evaluated as salvage oral bioavailability among patients matopoietic stem-cell transplants
therapy for invasive mycoses76; with severe GVHD75; lack of stud-
drug has not been approved by ies on drug as primary therapy for
the FDA as primary or salvage invasive mycoses
therapy for invasive fungal diseas-
es; for children 8 to 17 yr of age,
a small study showed steady-state
plasma levels were similar to
those in adults receiving the
same dose

* CYP2C19 denotes cytochrome P-450 2C19, FDA Food and Drug Administration, and GVHD graft-versus-host disease.
† Pediatric doses are often derived from the results of small pharmacokinetic studies, some of which are described in the package inserts or
published in abstract form.
‡ The mold-active azoles are potent inhibitors of the hepatic cytochrome P-450 3A4 isoenzyme, which has a major role in the metabolism of
drugs from multiple classes.
§ The intravenous formulations of itraconazole and voriconazole should be used with caution in patients with substantial renal impairment
(creatinine clearance, <50 ml per minute) because of a potentially systemic accumulation of the cyclodextrin vehicle that can, in turn, cause
renal toxicity. This concern does not apply to oral formulations.
¶ Intravenous itraconazole is no longer available in the United States.

ity, and a trend toward improved 12-week survival, Echinocandins are antifungal agents acting on
as compared with a dose of 10 mg per kilogram the cell wall that inhibit beta-glucan synthesis
daily as primary therapy for invasive asper­gillo­ (Table 4).91-96 Caspofungin is approved by the Food
sis; this study showed that increased doses of and Drug Administration as salvage therapy for
amphotericin B should not be equated with invasive aspergillosis. There is substantial interest
greater efficacy.87 Pediatric patients may require in pairing echinocandins, which have cell-wall ac-
a dose per kilogram of body weight that is higher tivity, with either amphotericin B formulations,
than the adult dose to achieve a similar systemic which have cell-membrane activity, or mold-active
exposure.88 azoles as therapy for invasive aspergillosis. Clin-

1878 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

Table 3. Characteristics of Amphotericin B Formulations.*

Antifungal Agents Dose Comments

Amphotericin B de- For invasive mold diseases: 1.0 to 1.5 mg per kilo- Amphotericin B formulations are alternative options for invasive
oxycholate gram daily aspergillosis, such as in patients who cannot tolerate vori-
conazole or with refractory aspergillosis. Saline hydration
(generally 1 liter for an adult of average size with normal
­cardiac and renal function) may prevent or ameliorate
azotemia; aggressive electrolyte replacement is also re-
quired. Because of nephrotoxicity and poor tolerance of
long-term use in patients with invasive aspergillosis,61 a lipid
formulation should be considered. An amphotericin B formu-
lation should be used for suspected or proven zygomycosis.
Liposomal ampho­ 3 to 5 mg per kilogram daily; clearance and vol- One randomized trial showed that 3 mg per kilogram per day
tericin B ume of distribution are influenced by body was as effective as but less toxic than 10 mg per kilogram
weight in children with cancer; higher weight- per day as initial therapy for invasive aspergillosis.88
based doses may be optimal in patients weigh-
ing less than 20 kg87
Amphotericin B lipid 5 mg per kilogram daily This drug was evaluated in a large registry database of invasive
complex aspergillosis.89
Amphotericin B col- 5 to 6 mg per kilogram daily This drug (at a dose of 6 mg per kilogram) had a similar efficacy,
loidal dispersion lower nephrotoxicity, and greater infusional toxicity than daily
amphotericin B deoxycholate (at a dose of 1 to 1.5 mg per ki-
logram) as primary therapy for invasive aspergillosis.90

* The suggested doses of amphotericin B formulations are based on those used in clinical trials and in the guidelines of the Infectious
Diseases Society of America.40 The lipid formulations of amphotericin B have different pharmacokinetic properties, but the clinical signifi-
cance of these differences is unknown.

ical data on combination antifungal therapy for interactions. Indeed, several agents that are used
invasive aspergillosis are limited but encouraging. to treat cancer (e.g., cyclophosphamide and vinca
In one provocative study, Marr et al.97 reported a alkaloids) and as immunosuppressive therapy in
survival advantage of voriconazole plus caspo- transplant recipients (e.g., calcineurin inhibitors
fungin, as compared with voriconazole alone, in and sirolimus) are metabolized through hepatic
a retrospective analysis of salvage therapy for in- CYP3A4. All azoles can cause hepatotoxicity, in-
vasive aspergillosis. However, this database in- cluding hyperbilirubinemia and liver-enzyme ab-
volved a small number of patients, and the two normalities, both of which are usually reversible.
groups were noncontemporaneous; therefore, other Voriconazole commonly causes reversible visual
host and infection-related factors may have in- symptoms that uncommonly require drug cessa-
fluenced the outcome. A noncomparative study of tion. The main toxic effects that are associated
caspofungin in combination with other antifun- with amphotericin B formulations are infusional
gal agents as salvage therapy in patients with in- reactions and nephrotoxicity. Echinocandins are
vasive aspergillosis resulted in a successful out- generally well tolerated.
come in 25 of 51 patients (49%),91 a success rate The optimal duration of therapy is not straight-
similar to that in a previous study of caspofungin forward and depends on the burden of disease and
monotherapy.98 A randomized trial comparing host immunocompetence. The guidelines of the
voriconazole with voriconazole plus anidulafungin Infectious Diseases Society of America recom-
(an echinocandin) has begun. mend a duration of therapy of at least 6 to 12
An important component in therapy for inva- weeks for pulmonary aspergillosis.40 In immu-
sive aspergillosis involves anticipating and man- nocompromised patients, therapy should be con-
aging the toxic effects of various drugs. Azoles tinued throughout the period of immunosup-
cross-react with and can be inhibitors and sub- pression and until the resolution of lesions.40
strates of mammalian cytochrome P-450 isoen- In patients with previously diagnosed invasive
zymes. Inhibition of the CYP3A4 isoenzyme by aspergillosis, antifungal therapy should be con-
azoles (especially mold-active azoles, as compared tinued or reinitiated during subsequent periods
with fluconazole) accounts for the majority of drug of immunosuppression (e.g., additional cytotoxic

n engl j med 360;18  nejm.org  april 30, 2009 1879

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 4. Characteristics of Echinocandins.

Antifungal
Agent Dose Pharmacologic Properties Comments
Caspofungin Adult dose: 70 mg daily for the first Slowly metabolized by hydrolysis Only echinocandin that is FDA-
dose, then 50 mg daily; 70 mg daily and N-acetylation and under- approved as salvage therapy for in­
can be considered for invasive goes spontaneous degradation vasive aspergillosis; encouraging
aspergillosis91; in patients with ­results from a preliminary, nonran-
moderate hepatic insufficiency domized database regarding com­
(Child–Pugh score, 7 to 9), a first bination voriconazole and caspo-
dose of 70 mg, then 35 mg daily for fungin as therapy for invasive
patients with moderate liver disease; ­aspergillosis
pediatric dose: 50 mg per square
meter of body-surface area (35 mg
per square meter in patients with
moderate liver disease)92
Micafungin Prophylaxis in recipients of stem-cell Hepatically metabolized by arylsul- A nonrandomized trial showed safety
transplant recipients during neutro- fatase and catechol-O-methyl- alone and paired with other agents
penia: 50 mg daily93; for candidemia transferase; although a substrate as primary and salvage therapy for
and other forms of invasive candidi- of CYP3A4 in vitro, CYP3A4 does invasive aspergillosis95
asis, 100 mg daily; for esophageal not play a significant role in me-
candidiasis, 150 mg daily; optimal tabolism in vivo
dose for invasive aspergillosis has
not been defined; pediatric dose:
i­ncreased clearance as a function
of decreasing age occurs94
Anidulafungin For candidemia and other forms of in­ Undergoes slow chemical degrada- No clinical trial data are available re-
vasive candidiasis: 200 mg for first tion; not hepatically metabolized garding therapy for invasive asper-
dose, then 100 mg daily; optimal gillosis, although a randomized trial
dose for invasive aspergillosis has is under way comparing voricona­
not been defined; pediatric dose: zole plus anidulafungin with vori-
0.75 mg and 1.5 mg per kilogram conazole alone as primary therapy
daily resulted in drug exposure simi- for invasive aspergillosis
lar to that of adult doses of 50 and
100 mg daily, respectively96

chemotherapy or hematopoietic stem-cell trans- geted to high-risk patients. Mold-active azoles,

plantation) to prevent recrudescence.40,99 echinocandins, and amphotericin B formulations
When feasible, immunosuppressive therapy are options.40,59 Prophylactic inhalation of ampho-
(e.g., corticosteroids) should be reduced or discon-
tericin B formulations, which may prevent system-
tinued. Adjunctive myeloid colony-stimulating fac- ic toxic effects, merits further evaluation. Posa-
tors (granulocyte or granulocyte–macrophage col- conazole was effective as prophylaxis in patients
ony-stimulating factor) should be considered in with neutropenia who had the myelodysplastic syn-
patients with neutropenia who have severe infec- drome and acute myelogenous leukemia74 and in
tions, such as aspergillosis.100 Treatment with patients with severe GVHD.75
granulocyte transfusions in patients with persis- Limitations to mold-active prophylaxis exist.
tent neutropenia and the use of recombinant Mold-active azoles can cause serious drug inter-
interferon-γ (which activates neutrophils and mac- actions. The diagnosis of breakthrough aspergil-
rophages) can be considered in patients who have losis may be hampered by false negative results
refractory or disseminated aspergillosis; howev- on serum galactomannan assays. There is also a
er, the benefit of such therapies versus their toxic
gap in knowledge regarding optimal therapy for
effects has not been established.25 breakthrough invasive aspergillosis in patients re-
ceiving mold-active prophylaxis; changing the an-
Prevention of Invasive Aspergillosis tifungal class is a reasonable, although untested,
Prevention of invasive aspergillosis relies on en- option. There is debate about the overall benefit,
vironmental infection-control guidelines to reduce risks, and cost-effectiveness of mold-active pro-
mold exposure101 and antifungal prophylaxis tar- phylaxis in high-risk patients as compared with a

1880 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

preemptive approach, in which a mold-active agent etic stem-cell transplants. Alterations in plasmin­
is targeted to patients who meet prespecified crite- ogen can affect fungal virulence in a number of
ria on the basis of radiologic findings, laboratory ways, including fungal adherence to extracellular
markers, or both.102,103 matrix and a direct effect on innate immunity.
Knowledge of these host genetic factors may prove
F u t ur e Per spec t i v e s to be important in stratifying the risk of invasive
aspergillosis during periods of immunosuppres-
The development of new diagnostic and therapeu- sion, in guiding donor selection and targeted anti-
tic approaches will be facilitated by knowledge of fungal prophylaxis, and in identifying new thera-
both host and pathogen characteristics that pre- peutic targets.
dispose patients to aspergillus-associated diseases. An increased understanding of fungal genetics
Knowledge of innate and T-cell immunity against and biochemistry has led to therapeutic strategies
aspergillus may pave the way to new immunomod- at the preclinical level, such as inhibition of cell
ulation strategies, including vaccine development.25 stress-response pathways.107,108 The recent discov-
In addition, antifungal agents have immunomod- ery that A. fumigatus has a sexual reproductive cycle
ulatory effects, including the activation of patho- provides insight into its evolution and genomic
gen-recognition receptors and unmasking of proin- variability and offers a valuable tool to analyze the
flammatory constituents of fungal-cell walls that genetic basis of pathogenicity.109 In addition, ge-
may be clinically relevant and therapeutically ex- nomic analysis has shown that aspergillus species
ploitable.104,105 are extremely diverse.110 One potential benefit of
Polymorphisms in host genes that mediate in- comparative genomics among pathogenic and
nate immunity may influence the risk of invasive nonpathogenic aspergillus species is to identify
aspergillosis during periods of immunosuppres- genes associated with virulence that can be tar-
sion. As examples, Bochud et al.23 found that gets for drug development.
specific donor TLR4 haplotypes influenced the risk
of invasive aspergillosis in recipients of alloge- Dr. Segal reports receiving speaking fees from Merck, Pfizer,
and Schering-Plough and consulting fees from Schering-Plough,
neic hematopoietic stem-cell transplants. Zaas et Pfizer, and Astellas. No other potential conflict of interest rele-
al.106 found that polymorphisms in the plasmi- vant to this article was reported.
nogen allele affected the outcome of experimen- I thank Dr. Nikolaos Almyroudis of the Roswell Park Cancer
Institute and Dr. David Andes of the University of Wisconsin
tal aspergillosis in mice and the risk of invasive School of Medicine for their constructive review of an earlier
aspergillosis in recipients of allogeneic hematopoi- version of the manuscript.

References
1. Cornell MJ, Alam I, Soanes DM, et al. human and mouse phagocytes. J Clin In- Doerschuk CM, Dinauer MC. Absence of
Comparative genome analysis across a vest 1982;69:617-31. respiratory burst in X-linked chronic gran-
kingdom of eukaryotic organisms: spe- 5. Morrison BE, Park SJ, Mooney JM, ulomatous disease mice leads to abnor-
cialization and diversification in the fun- Mehrad B. Chemokine-mediated recruit- malities in both host defense and inflam-
gi. Genome Res 2007;17:1809-22. ment of NK cells is a critical host defense matory response to Aspergillus fumigatus.
2. Lemaitre B, Nicolas E, Michaut L, mechanism in invasive aspergillosis. J Clin J Exp Med 1997;185:207-18.
Reich­hart JM, Hoffmann JA. The dors- Invest 2003;112:1862-70. 10. Romani L, Fallarino F, De Luca A, et
oventral regulatory gene cassette spätzle/ 6. Segal BH, DeCarlo ES, Kwon-Chung al. Defective tryptophan catabolism under-
Toll/cactus controls the potent antifungal KJ, Malech HL, Gallin JI, Holland SM. lies inflammation in mouse chronic gran-
response in Drosophila adults. Cell 1996; Aspergillus nidulans infection in chronic ulomatous disease. Nature 2008;451:211-5.
86:973-83. granulomatous disease. Medicine (Balti- 11. Madan T, Kishore U, Singh M, et al.
3. Wasylnka JA, Moore MM. Uptake of more) 1998;77:345-54. Surfactant proteins A and D protect mice
Aspergillus fumigatus conidia by phago- 7. Reeves EP, Lu H, Jacobs HL, et al. Kill- against pulmonary hypersensitivity induced
cytic and nonphagocytic cells in vitro: ing activity of neutrophils is mediated by Aspergillus fumigatus antigens and al-
quantitation using strains expressing green through activation of proteases by K+ flux. lergens. J Clin Invest 2001;107:467-75.
fluorescent protein. Infect Immun 2002; Nature 2002;416:291-7. 12. Hogaboam CM, Takahashi K, Ezekow-
70:3156-63. 8. Zarember KA, Sugui JA, Chang YC, itz RA, Kunkel SL, Schuh JM. Mannose-
4. Schaffner A, Douglas H, Braude A. Kwon-Chung KJ, Gallin JI. Human poly- binding lectin deficiency alters the develop-
Selective protection against conidia by morphonuclear leukocytes inhibit Asper- ment of fungal asthma: effects on airway
mononuclear and against mycelia by poly- gillus fumigatus conidial growth by lacto- response, inflammation, and cytokine pro-
morphonuclear phagocytes in resistance ferrin-mediated iron depletion. J Immunol file. J Leukoc Biol 2004;75:805-14.
to Aspergillus: observations on these two 2007;178:6367-73. 13. Garlanda C, Hirsch E, Bozza S, et al.
lines of defense in vivo and in vitro with 9. Morgenstern DE, Gifford MA, Li LL, Non-redundant role of the long pentraxin

n engl j med 360;18  nejm.org  april 30, 2009 1881

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

PTX3 in anti-fungal innate immune re- covery and biology of IL-23 and IL-27: re- al. Trends in mortality due to invasive my-
sponse. Nature 2002;420:182-6. lated but functionally distinct regulators cotic diseases in the United States, 1980-
14. Bellocchio S, Moretti S, Perruccio K, et of inflammation. Annu Rev Immunol 1997. Clin Infect Dis 2001;33:641-7.
al. TLRs govern neutrophil activity in asper- 2007;25:221-42. 43. Pagano L, Caira M, Picardi M, et al.
gillosis. J Immunol 2004;173:7406-15. 30. Zelante T, De Luca A, Bonifazi P, et al. Invasive Aspergillosis in patients with
15. Ramirez-Ortiz ZG, Specht CA, Wang IL-23 and the Th17 pathway promote in- acute leukemia: update on morbidity and
JP, Lee CK, Bartholomeu DC, Gazzinelli flammation and impair antifungal im- mortality — SEIFEM-C Report. Clin In-
RT, Levitz SM. Toll-like receptor 9-depen- mune resistance. Eur J Immunol 2007;37: fect Dis 2007;44:1524-5.
dent immune activation by unmethylated 2695-706. 44. Marty FM, Rubin RH. The prevention
CpG motifs in Aspergillus fumigatus DNA. 31. Montagnoli C, Fallarino F, Gaziano R, of infection post-transplant: the role of
Infect Immun 2008;76:2123-9. et al. Immunity and tolerance to Aspergil- prophylaxis, preemptive and empiric ther-
16. Netea MG, Warris A, Van der Meer JW, lus involve functionally distinct regulato- apy. Transpl Int 2006;19:2-11.
et al. Aspergillus fumigatus evades immune ry T cells and tryptophan catabolism. 45. Marr KA, Carter RA, Boeckh M, Mar-
recognition during germination through J Immunol 2006;176:1712-23. tin P, Corey L. Invasive aspergillosis in
loss of toll-like receptor-4-mediated signal 32. Chamilos G, Luna M, Lewis RE, et al. allogeneic stem cell transplant recipients:
transduction. J Infect Dis 2003;188:320-6. Invasive fungal infections in patients with changes in epidemiology and risk factors.
17. Brown GD. Dectin-1: a signalling non- hematologic malignancies in a tertiary Blood 2002;100:4358-66.
TLR pattern-recognition receptor. Nat Rev care cancer center: an autopsy study over 46. Jantunen E, Ruutu P, Niskanen L, et
Immunol 2006;6:33-43. a 15-year period (1989-2003). Haemato- al. Incidence and risk factors for invasive
18. Gantner BN, Simmons RM, Canavera logica 2006;91:986-9. fungal infections in allogeneic BMT re-
SJ, Akira S, Underhill DM. Collaborative 33. Shaukat A, Bakri F, Young P, et al. In- cipients. Bone Marrow Transplant 1997;
induction of inflammatory responses by vasive filamentous fungal infections in 19:801-8.
dectin-1 and Toll-like receptor 2. J Exp allogeneic hematopoietic stem cell trans- 47. Wald A, Leisenring W, van Burik JA,
Med 2003;197:1107-17. plant recipients after recovery from neu- Bowden RA. Epidemiology of Aspergillus
19. Hohl TM, Van Epps HL, Rivera A, et al. tropenia: clinical, radiologic, and patho- infections in a large cohort of patients
Aspergillus fumigatus triggers inflamma- logic characteristics. Mycopathologia 2005; undergoing bone marrow transplantation.
tory responses by stage-specific beta-glucan 159:181-8. J Infect Dis 1997;175:1459-66.
display. PLoS Pathog 2005;1(3):e30. 34. Stergiopoulou T, Meletiadis J, Roil- 48. Morgan J, Wannemuehler KA, Marr
20. Gersuk GM, Underhill DM, Zhu L, ides E, et al. Host-dependent patterns of KA, et al. Incidence of invasive aspergil-
Marr KA. Dectin-1 and TLRs permit mac- tissue injury in invasive pulmonary asper- losis following hematopoietic stem cell
rophages to distinguish between different gillosis. Am J Clin Pathol 2007;127:349-55. and solid organ transplantation: interim
Aspergillus fumigatus cellular states. J Im- 35. Nash G, Irvine R, Kerschmann RL, results of a prospective multicenter sur-
munol 2006;176:3717-24. Herndier B. Pulmonary aspergillosis in veillance program. Med Mycol 2005;43:
21. Steele C, Rapaka RR, Metz A, et al. The acquired immune deficiency syndrome: Suppl 1:S49-S58.
beta-glucan receptor dectin-1 recognizes autopsy study of an emerging pulmonary 49. Pagano L, Caira M, Nosari A, et al.
specific morphologies of Aspergillus fu- complication of human immunodeficiency Fungal infections in recipients of he-
migatus. PLoS Pathog 2005;1(4):e42. virus infection. Hum Pathol 1997;28:1268- matopoietic stem cell transplants: results
22. von Bernuth H, Picard C, Jin Z, et al. 75. of the SEIFEM B-2004 study — Sorvegli-
Pyogenic bacterial infections in humans 36. Siddiqui S, Anderson VL, Hilligoss anza Epidemiologica Infezioni Fungine
with MyD88 deficiency. Science 2008;321: DM, et al. Fulminant mulch pneumonitis: Nelle Emopatie Maligne. Clin Infect Dis
691-6. an emergency presentation of chronic 2007;45:1161-70.
23. Bochud PY, Chien JW, Marr KA, et al. granulomatous disease. Clin Infect Dis 50. van Burik JA, Carter SL, Freifeld AG,
Toll-like receptor 4 polymorphisms and 2007;45:673-81. et al. Higher risk of cytomegalovirus and
aspergillosis in stem-cell transplantation. 37. Binder RE, Faling LJ, Pugatch RD, Ma- aspergillus infections in recipients of
N Engl J Med 2008;359:1766-77. hasaen C, Snider GL. Chronic necrotizing T cell-depleted unrelated bone marrow:
24. Rivera A, Ro G, Van Epps HL, et al. pulmonary aspergillosis: a discrete clini- analysis of infectious complications in
Innate immune activation and CD4+ T cell cal entity. Medicine (Baltimore) 1982;61: patients treated with T cell depletion ver-
priming during respiratory fungal infec- 109-24. sus immunosuppressive therapy to pre-
tion. Immunity 2006;25:665-75. 38. Denning DW, Riniotis K, Dobrashian vent graft-versus-host disease. Biol Blood
25. Segal BH, Kwon-Chung J, Walsh TJ, et R, Sambatakou H. Chronic cavitary and Marrow Transplant 2007;13:1487-98.
al. Immunotherapy for fungal infections. fibrosing pulmonary and pleural asper- 51. Hadjiliadis D, Sporn TA, Perfect JR,
Clin Infect Dis 2006;42:507-15. gillosis: case series, proposed nomencla- Tapson VF, Davis RD, Palmer SM. Out-
26. Roilides E, Sein T, Roden M, Schaufele ture change, and review. Clin Infect Dis come of lung transplantation in patients
RL, Walsh TJ. Elevated serum concentra- 2003;37:Suppl 3:S265-S280. with mycetomas. Chest 2002;121:128-34.
tions of interleukin-10 in nonneutropenic 39. Sambatakou H, Dupont B, Lode H, 52. Gavalda J, Len O, San Juan R, et al.
patients with invasive aspergillosis. J In- Denning DW. Voriconazole treatment for Risk factors for invasive aspergillosis in
fect Dis 2001;183:518-20. subacute invasive and chronic pulmonary solid-organ transplant recipients: a case-
27. Hebart H, Bollinger C, Fisch P, et al. aspergillosis. Am J Med 2006;119(6):527. control study. Clin Infect Dis 2005;41:
Analysis of T-cell responses to Aspergil- e17-527.e24. 52-9.
lus fumigatus antigens in healthy individ- 40. Walsh TJ, Anaissie EJ, Denning DW, et 53. Singh N, Pruett TL, Houston S, et al.
uals and patients with hematologic ma- al. Treatment of aspergillosis: clinical Invasive aspergillosis in the recipients of
lignancies. Blood 2002;100:4521-8. practice guidelines of the Infectious Dis- liver retransplantation. Liver Transpl 2006;
28. Stevens DA, Moss RB, Kurup VP, et al. eases Society of America. Clin Infect Dis 12:1205-9.
Allergic bronchopulmonary aspergillosis in 2008;46:327-60. 54. Filler SG, Yeaman MR, Sheppard DC.
cystic fibrosis — state of the art: Cystic Fi- 41. Segal BH, Walsh TJ. Current ap- Tumor necrosis factor inhibition and in-
brosis Foundation Consensus Conference. proaches to diagnosis and treatment of vasive fungal infections. Clin Infect Dis
Clin Infect Dis 2003;37:Suppl 3:S225-S264. invasive aspergillosis. Am J Respir Crit 2005;41:Suppl 3:S208-S212.
[Erratum, Clin Infect Dis 2004;38:158.] Care Med 2006;173:707-17. 55. Segal BH, Sneller MC. Infectious com-
29. Kastelein RA, Hunter CA, Cua DJ. Dis- 42. McNeil MM, Nash SL, Hajjeh RA, et plications of immunosuppressive therapy

1882 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

in patients with rheumatic diseases. Rheum patients. Clin Microbiol Infect 2008;14: 82. Nucci M, Perfect JR. When primary
Dis Clin North Am 1997;23:219-37. Suppl 4:37-45. antifungal therapy fails. Clin Infect Dis
56. Meersseman W, Lagrou K, Maertens J, 69. Kontoyiannis DP, Lionakis MS, Lewis 2008;46:1426-33.
et al. Galactomannan in bronchoalveolar RE, et al. Zygomycosis in a tertiary-care 83. Caillot D, Couaillier JF, Bernard A, et
lavage fluid: a tool for diagnosing asper- cancer center in the era of Aspergillus- al. Increasing volume and changing char-
gillosis in intensive care unit patients. Am active antifungal therapy: a case-control acteristics of invasive pulmonary asper-
J Respir Crit Care Med 2008;177:27-34. observational study of 27 recent cases. gillosis on sequential thoracic computed
57. De Pauw B, Walsh TJ, Donnelly JP, et J Infect Dis 2005;191:1350-60. tomography scans in patients with neu-
al. Revised definitions of invasive fungal 70. Mouy R, Veber F, Blanche S, et al. tropenia. J Clin Oncol 2001;19:253-9.
disease from the European Organization Long-term itraconazole prophylaxis against 84. Miceli MH, Maertens J, Buvé K, et al.
for Research and Treatment of Cancer/In- Aspergillus infections in thirty-two patients Immune reconstitution inflammatory syn-
vasive Fungal Infections Cooperative Group with chronic granulomatous disease. J Pe- drome in cancer patients with pulmonary
and the National Institute of Allergy and diatr 1994;125:998-1003. aspergillosis recovering from neutrope-
Infectious Diseases Mycoses Study Group 71. Gallin JI, Alling DW, Malech HL, et al. nia: proof of principle, description, and
(EORTC/MSG) Consensus Group. Clin In- Itraconazole to prevent fungal infections clinical and research implications. Cancer
fect Dis 2008;46:1813-21. in chronic granulomatous disease. N Engl 2007;110:112-20.
58. Caillot D, Casasnovas O, Bernard A, et J Med 2003;348:2416-22. 85. Verweij PE, Mellado E, Melchers WJ.
al. Improved management of invasive pul- 72. Stevens DA, Schwartz HJ, Lee JY, et al. Multiple-triazole-resistant aspergillosis.
monary aspergillosis in neutropenic pa- A randomized trial of itraconazole in al- N Engl J Med 2007;356:1481-3.
tients using early thoracic computed to- lergic bronchopulmonary aspergillosis. 86. Woods G, Miceli MH, Grazziutti ML,
mographic scan and surgery. J Clin Oncol N Engl J Med 2000;342:756-62. Zhao W, Barlogie B, Anaissie E. Serum
1997;15:139-47. 73. Wark PA, Hensley MJ, Saltos N, et al. Aspergillus galactomannan antigen values
59. Segal BH, Freifeld AG, Baden LR, et Anti-inflammatory effect of itraconazole strongly correlate with outcome of invasive
al. Prevention and treatment of cancer- in stable allergic bronchopulmonary asper- aspergillosis: a study of 56 patients with he-
related infections. J Natl Compr Canc gillosis: a randomized controlled trial. matologic cancer. Cancer 2007;110:830-4.
Netw 2008;6:122-74. J Allergy Clin Immunol 2003;111:952-7. 87. Hong Y, Shaw PJ, Nath CE, et al. Popu-
60. Greene RE, Schlamm HT, Oestmann 74. Cornely OA, Maertens J, Winston DJ, lation pharmacokinetics of liposomal am-
JW, et al. Imaging findings in acute inva- et al. Posaconazole vs. fluconazole or itra- photericin B in pediatric patients with
sive pulmonary aspergillosis: clinical sig- conazole prophylaxis in patients with neu- malignant diseases. Antimicrob Agents
nificance of the halo sign. Clin Infect Dis tropenia. N Engl J Med 2007;356:348-59. Chemother 2006;50:935-42.
2007;44:373-9. 75. Ullmann AJ, Lipton JH, Vesole DH, et 88. Cornely OA, Maertens J, Bresnik M, et
61. Herbrecht R, Denning DW, Patterson al. Posaconazole or fluconazole for pro- al. Liposomal amphotericin B as initial
TF, et al. Voriconazole versus amphoteri- phylaxis in severe graft-versus-host disease. therapy for invasive mold infection: a ran-
cin B for primary therapy of invasive asper- N Engl J Med 2007;356:335-47. [Erratum, domized trial comparing a high-loading
gillosis. N Engl J Med 2002;347:408-15. N Engl J Med 2007;357:428.] dose regimen with standard dosing (Am-
62. Levine SJ. An approach to the diagno- 76. Walsh TJ, Raad I, Patterson TF, et al. BiLoad trial). Clin Infect Dis 2007;44:
sis of pulmonary infections in immuno- Treatment of invasive aspergillosis with 1289-97.
suppressed patients. Semin Respir Infect posaconazole in patients who are refrac- 89. Chandrasekar PH, Ito JI. Amphoteri-
1992;7:81-95. tory to or intolerant of conventional ther- cin B lipid complex in the management of
63. Odabasi Z, Mattiuzzi G, Estey E, et al. apy: an externally controlled trial. Clin invasive aspergillosis in immunocompro-
Beta-D-glucan as a diagnostic adjunct for Infect Dis 2007;44:2-12. mised patients. Clin Infect Dis 2005;40:
invasive fungal infections: validation, cut- 77. Pascual A, Nieth V, Calandra T, et al. Suppl 6:S392-S400.
off development, and performance in pa- Variability of voriconazole plasma levels 90. Bowden R, Chandrasekar P, White
tients with acute myelogenous leukemia measured by new high-performance liquid MH, et al. A double-blind, randomized,
and myelodysplastic syndrome. Clin In- chromatography and bioassay methods. controlled trial of amphotericin B colloi-
fect Dis 2004;39:199-205. Antimicrob Agents Chemother 2007;51: dal dispersion versus amphotericin B for
64. Marty FM, Koo S, Bryar J, Baden LR. 137-43. treatment of invasive aspergillosis in im-
(1→3)β-D-glucan assay positivity in patients 78. Trifilio S, Pennick G, Pi J, et al. Moni- munocompromised patients. Clin Infect
with Pneumocystis (carinii) jiroveci pneu- toring plasma voriconazole levels may be Dis 2002;35:359-66.
monia. Ann Intern Med 2007;147:70-2. necessary to avoid subtherapeutic levels 91. Maertens J, Glasmacher A, Herbrecht R,
65. Pfeiffer CD, Fine JP, Safdar N. Diag- in hematopoietic stem cell transplant re- et al. Multicenter, noncomparative study
nosis of invasive aspergillosis using a ga- cipients. Cancer 2007;109:1532-5. of caspofungin in combination with other
lactomannan assay: a meta-analysis. Clin 79. Smith J, Safdar N, Knasinski V, et al. antifungals as salvage therapy in adults
Infect Dis 2006;42:1417-27. Voriconazole therapeutic drug monitor- with invasive aspergillosis. Cancer 2006;
66. Marr KA, Balajee SA, McLaughlin L, ing. Antimicrob Agents Chemother 2006; 107:2888-97.
Tabouret M, Bentsen C, Walsh TJ. Detec- 50:1570-2. 92. Walsh TJ, Adamson PC, Seibel NL, et
tion of galactomannan antigenemia by 80. Pascual A, Calandra T, Bolay S, Buclin al. Pharmacokinetics, safety, and tolera-
enzyme immunoassay for the diagnosis of T, Bille J, Marchetti O. Voriconazole ther- bility of caspofungin in children and ado-
invasive aspergillosis: variables that affect apeutic drug monitoring in patients with lescents. Antimicrob Agents Chemother
performance. J Infect Dis 2004;190:641-9. invasive mycoses improves efficacy and 2005;49:4536-45.
67. Wheat LJ, Walsh TJ. Diagnosis of inva- safety outcomes. Clin Infect Dis 2008; 93. van Burik JA, Ratanatharathorn V,
sive aspergillosis by galactomannan anti- 46:201-11. Stepan DE, et al. Micafungin versus flu-
genemia detection using an enzyme im- 81. Tan K, Brayshaw N, Tomaszewski K, conazole for prophylaxis against invasive
munoassay. Eur J Clin Microbiol Infect Troke P, Wood N. Investigation of the po- fungal infections during neutropenia in
Dis 2008;27:245-51. tential relationships between plasma vori- patients undergoing hematopoietic stem
68. Einsele H, Loeffler J. Contribution of conazole concentrations and visual adverse cell transplantation. Clin Infect Dis 2004;
new diagnostic approaches to antifungal events or liver function test abnormalities. 39:1407-16.
treatment plans in high-risk haematology J Clin Pharmacol 2006;46:235-43. 94. Seibel NL, Schwartz C, Arrieta A, et

n engl j med 360;18  nejm.org  april 30, 2009 1883

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 medical progress

al. Safety, tolerability, and pharmacoki- the Infectious Diseases Working Party of migatus through stage-specific effects on
netics of Micafungin (FK463) in febrile the European Group for Blood and Marrow fungal beta-glucan exposure. J Infect Dis
neutropenic pediatric patients. Antimi- Transplantation. Blood 2006;108:2928-36. 2008;198:176-85.
crob Agents Chemother 2005;49:3317-24. 100. Smith TJ, Khatcheressian J, Lyman 105. Lamaris GA, Lewis RE, Chamilos G,
95. Denning DW, Marr KA, Lau WM, et GH, et al. 2006 Update of recommenda- et al. Caspofungin-mediated beta-glucan
al. Micafungin (FK463), alone or in com- tions for the use of white blood cell growth unmasking and enhancement of human
bination with other systemic antifungal factors: an evidence-based clinical practice polymorphonuclear neutrophil activity
agents, for the treatment of acute invasive guideline. J Clin Oncol 2006;24:3187-205. against Aspergillus and non-Aspergillus
aspergillosis. J Infect 2006;53:337-49. 101. Sullivan KM, Dykewicz CA, Long- hyphae. J Infect Dis 2008;198:186-92.
96. Benjamin DK Jr, Driscoll T, Seibel NL, worth DL, et al. Preventing opportunistic 106. Zaas AK, Liao G, Chien JW, et al.
et al. Safety and pharmacokinetics of in- infections after hematopoietic stem cell Plasminogen alleles influence susceptibil-
travenous anidulafungin in children with transplantation: the Centers for Disease ity to invasive aspergillosis. PLoS Genet
neutropenia at high risk for invasive fungal Control and Prevention, Infectious Diseases 2008;4(6):e1000101.
infections. Antimicrob Agents Chemother Society of America, and American Society 107. Cowen LE, Lindquist S. Hsp90 poten-
2006;50:632-8. for Blood and Marrow Transplantation tiates the rapid evolution of new traits:
97. Marr KA, Boeckh M, Carter RA, Kim Practice Guidelines and beyond. Hematol- drug resistance in diverse fungi. Science
HW, Corey L. Combination antifungal ogy Am Soc Hematol Educ Program 2001: 2005;309:2185-9.
therapy for invasive aspergillosis. Clin In- 392-421. 108. Steinbach WJ, Reedy JL, Cramer RA Jr,
fect Dis 2004;39:797-802. 102. Gonzalez AV, Ullmann AJ, Almyroudis Perfect JR, Heitman J. Harnessing calcineu-
98. Maertens J, Raad I, Petrikkos G, et al. NG, Segal BH. Broad-spectrum antifungal rin as a novel anti-infective agent against
Efficacy and safety of caspofungin for prophylaxis in patients with cancer at high invasive fungal infections. Nat Rev Micro-
treatment of invasive aspergillosis in pa- risk for invasive mold infections: point. biol 2007;5:418-30.
tients refractory to or intolerant of con- J Natl Compr Canc Netw 2008;6:175-82. 109. O’Gorman CM, Fuller HT, Dyer PS.
ventional antifungal therapy. Clin Infect 103. Maertens J, Buvé K, Anaissie EJ. Discovery of a sexual cycle in the opportu-
Dis 2004;39:1563-71. Broad-spectrum antifungal prophylaxis in nistic fungal pathogen Aspergillus fumig-
99. Martino R, Parody R, Fukuda T, et al. patients with cancer at high risk for inva- atus. Nature 2009;457:471-4.
Impact of the intensity of the pretransplan- sive mould infections: counterpoint. J Natl 110. Fedorova ND, Khaldi N, Joardar VS,
tation conditioning regimen in patients Compr Canc Netw 2008;6:183-9. et al. Genomic islands in the pathogenic
with prior invasive aspergillosis undergo- 104. Hohl TM, Feldmesser M, Perlin DS, filamentous fungus Aspergillus fumiga-
ing allogeneic hematopoietic stem cell Pamer EG. Caspofungin modulates in- tus. PLoS Genet 2008;4(6):e1000046.
transplantation: a retrospective survey of flammatory responses to Aspergillus fu- Copyright © 2009 Massachusetts Medical Society.

images in clinical medicine

The Journal welcomes consideration of new submissions for Images in Clinical
Medicine. Instructions for authors and procedures for submissions can be found
on the Journal’s Web site at NEJM.org. At the discretion of the editor, images that
are accepted for publication may appear in the print version of the Journal,
the electronic version, or both.

1884 n engl j med 360;18  nejm.org  april 30, 2009

Downloaded from www.nejm.org by MR WINSTON FONTES on April 29, 2009 .

Copyright © 2009 Massachusetts Medical Society. All rights reserved.