on 04-Mar-2023 (Sat)

The term "aspergillosis" refers to illness due to allergy, airway or lung invasion, cutaneous infection, or extrapulmonary dissemination caused by species of Aspergillus, most commonly A. fumigatus, A. flavus, A. niger, and A. terreus

Aspergillus species are ubiquitous in nature, and inhalation of infectious conidia is a frequent event. Tissue invasion is uncommon and occurs most frequently in the setting of immunosuppression associated with therapy for hematologic malignancies, hematopoietic cell transplantation, or solid organ transplantation.

Inhaled conidia are met by the innate defenses provided by resident phagocytes, specifically airway epithelial cells and alveolar macrophages [1,2]. Little is known about the contribution of epithelial cells in clearing conidia. Relatively more is known about macrophages, which contribute to both conidial clearance and the production of secondary inflammation. These cells secrete inflammatory mediators after recognition of key cell wall components (eg, beta-D-glucan) exposed after conidial germination into hyphal forms. These mediators result in neutrophil recruitment and the activation of cellular immunity, which are important in killing potentially invasive microbial forms (hyphae) and determining the extent and nature of the immune response. Hence, risks for disease and the type of disease that occurs are the combined result of multiple cellular functions that impact proximal events in conidial clearance, production of inflammation, and killing of invasive forms [1].

Microbial factors that impact disease potential include toxins, proteases, and secondary metabolites that exert multiple effects on the host's local pulmonary and systemic defense. These include cellular products that:

● Inhibit phagocyte nicotinamide adenine dinucleotide phosphate (NADPH) oxidase activation, a key component in host defense against filamentous fungi

● Inhibit macrophage phagocytosis and killing

● Suppress functional T cell responses

Histopathologically, invasive aspergillosis is characterized by progression of the infection across tissue planes. One hallmark of infection is vascular invasion with subsequent infarction and tissue necrosis. Presumably, fungal cell surface components bind to vessel wall components, including basement membrane, extracellular matrix, and cellular constituents, and can cause ischemia and infarction of structures distal to invaded arteries.

Species prevalence — Most invasive infections are caused by members of the A. fumigatus species complex. In a report of 218 infections in 24 transplant centers in the United States, 67 percent were caused by members of the A. fumigatus complex, followed by A. flavus (13 percent), A. niger (9 percent), and A. terreus (7 percent). These data contrast with epidemiologic data from a decade earlier when the vast majority of cases (90 percent) were secondary to A. fumigatus species [4], likely reflecting changes in microbial epidemiology, center-based differences, and/or changes in typing methods.

Classic risk factors include:

● Severe and prolonged neutropenia

● Receipt of high doses of glucocorticoids

● Other drugs or conditions that lead to chronically impaired cellular immune responses (eg, immunosuppressive regimens administered to treat autoimmune diseases and to prevent organ rejection, AIDS)

Use of certain biologic agents (eg, Bruton’s tyrosine kinase inhibitors [ibrutinib], small molecule kinase inhibitors) also raise the invasive aspergillosis [5,6]. Other agents such as venetoclax, a B-cell leukemia/lymphoma-2 inhibitor used in hematologic malignancies, may increase risk for invasive aspergillosis although venetoclax is often used in combination with other chemotherapeutic agents like rituximab so that individual risk is unclear [7].

Specific risk factors in certain immunocompromised patients include:

● In hematopoietic cell transplant (HCT) recipients, risks are different early versus late after receipt of stem cells. Underlying disease, persistent neutropenia, and types of stem cells and conditioning impact risks early after transplant. Additional risk factors during the latter period include graft-versus-host disease (and accompanying therapies) and cytomegalovirus (CMV) infection and disease [9]. Immunologic variables that correlate with risk include impaired production of reactive oxygen species and qualitative and quantitative defects in NK and T cells [10-12].

● Various gene polymorphisms (in HCT donors, HCT and solid organ transplant recipients, or patients with hematologic malignancies) have been associated with an increased risk of invasive aspergillosis. Examples include polymorphisms in genes involved in the innate immune response, such as the toll-like receptor 4 gene [13], interleukin 1 beta and beta-defensin 1 [14], the soluble pattern recognition receptor, long pentraxin 3 [15,16], dectin-1 [17,18], and others [11].

● Airway colonization in lung transplant recipients with cystic fibrosis is a risk factor for disease following transplantation [19,20]. (See "Fungal infections following lung transplantation", section on 'Risk factors'.)

● In liver transplant recipients, increasing immunosuppression, hemodialysis, and CMV infection are risks factors [21,22].

● In renal transplant recipients, a longer duration of hemodialysis before transplantation and leukopenia are risk factors for early-onset infection (≤3 months following transplantation) [23]. Donor CMV seropositivity is a risk factor for late-onset infection (>3 months following transplantation).

Emerging risk factors described patients with no or milder immunosuppression include:

● Chronic obstructive pulmonary disease, with receipt of glucocorticoid therapy [24]

● Intensive care unit (ICU) admission [25-27]

● Certain viral infections (eg, influenza, severe acute respiratory syndrome coronavirus 2 [SARS-CoV-2], respiratory syncytial virus) [26,28-30]

Invasive aspergillosis associated with viral infections typically occurs in patient with respiratory failure in ICUs. In these settings, invasive aspergillosis is thought to occur secondary to epithelial airway injury facilitating Aspergillus colonization and invasion. Influenza-associated aspergillosis (IAA) was evaluated over a seven-year period in Belgian and Dutch ICUs; among patients who had influenza pneumonia admitted to ICUs, 19 percent developed invasive aspergillosis. Forty-four percent lacked classic risk factors [31]. Mortality of IAA was 51 percent compared with 28 percent in patients with influenza alone.

SARS-CoV-2 (COVID-19) has also been associated with pulmonary aspergillosis in multiple case series [28,32-38]. The reported incidence varies widely across studies (0 to 33 percent) [34]. This wide range may in part be due to the difficulty of distinguishing airway colonization from invasive infection and varied diagnostic criteria. When relying on strict criteria for diagnosis (including autopsy data), the incidence appears to be approximately 2 to 6 percent [35,39]. Among reported cases, age >65 years, chronic structural lung disease, receipt of immunomodulatory agents, immunocompromising conditions, and invasive ventilatory support appear to be the most significant risk factors for its acquisition [32-35,40-42]. Extensive use of corticosteroids or receipt of tocilizumab or other interleukin-6 inhibitors or Janus kinase/signal transducer and activator of transcription inhibitors, such as baricitinib, may also increase the risk [28,42].

The nature of the immune suppression, including degree, duration, and type of immunocompromise and concomitant therapy influences the pathogenesis of disease. Thus, invasive aspergillosis can be manifested as a spectrum of diseases involving predominantly the airways (eg, tracheobronchitis), the lung, or both and also as disseminated disease, involving the central nervous system and other organs.

Invasive disease is most frequent when there is a high amount of airway exposure, for instance, in the setting of construction [43] or when the host has a condition in which conidia are not effectively cleared. Historically, the highest risks for invasive infections have been seen in severely immunosuppressed patients, particularly allogeneic HCT recipients, solid organ transplant (especially lung, heart-lung, and liver) recipients, and patients who experience prolonged neutropenia. There are some inherited immunodeficiencies, such as chronic granulomatous disease, that lead to high risk for pulmonary aspergillosis [1].

Invasive aspergillosis most frequently occurs in the lungs or sinuses after inhalation of conidia, although, less commonly, disease can spread from the gastrointestinal tract or result from direct inoculation into the skin.

Pulmonary aspergillosis — Invasive aspergillosis most commonly involves the lungs (picture 1). Patients can present with a number of signs and symptoms: fever, chest pain, shortness of breath, cough, and/or hemoptysis. The classic triad that has been described in neutropenic patients with pulmonary aspergillosis is fever, pleuritic chest pain, and hemoptysis. However, absence of this triad should not discourage consideration of this diagnosis in the patient with risk factors for disease since neutropenic patients frequently present with fever in the absence of localizing pulmonary symptoms. In such patients, lung imaging often reveals pulmonary nodules and/or infiltrates.

Imaging — Chest radiograph is insensitive for detecting the earliest stages of pulmonary disease, but computed tomography (CT) scans typically demonstrate focal lesions (image 1). The type of radiographic abnormality is variable, depending upon the host and the type of disease: bronchopneumonia, angioinvasive aspergillosis, tracheobronchitis, or chronic necrotizing aspergillosis.

Pulmonary aspergillosis typically manifests as single or multiple nodules (with or without cavitation), patchy or segmental consolidation, or peribronchial infiltrates with or without tree-in-bud patterns. One relatively large review noted that small nodules (<1 cm) were most common (20 of 46; 43 percent), followed by consolidation (12 of 46; 26 percent), large nodules (masses, 10 of 46; 21 percent), and peribronchial infiltrates (4 of 46; 9 percent) [44]. Another large study that included neutropenic and nonneutropenic patients showed that the more specific signs (nodules and cavitation) were infrequent, and radiographic findings of consolidation, ground-glass infiltrates, and pleural effusions were more commonly seen [45].

Radiographic findings, as with disease manifestations, can be variable and largely depend upon the host. The radiographic progression of disease has been studied best in neutropenic patients in whom the initial findings typically include nodules that have surrounding ground-glass infiltrates (the halo sign), reflecting hemorrhage into the area surrounding the fungus (image 1) [46]. These nodules typically enlarge, even during appropriate therapy, and may eventually cavitate, producing the air-crescent sign [47,48].

However, in one study, it was noted that neutropenic patients more frequently presented with segmental areas of consolidation rather than demonstrating the classic nodule progression to cavitation (image 2) [45]. Similar findings were reported in a cohort of solid organ transplant recipients, with the majority of patients presenting with "pulmonary pneumonia" (focal infiltrates, 39 percent) or "pulmonary nodular" (21 percent) findings on CT scan [49]. In a cohort of 139 children, the most common radiographic presentation was nodules (35 percent); few children developed cavitary disease (2 percent) [50].

There may be a role for CT pulmonary angiography for distinguishing between invasive mold infections and other causes of pulmonary infiltrates, by virtue of detection of angioinvasion [51,52]. However, the studies that evaluated this modality were small and did not evaluate patients with a heterogeneous group of underlying diseases. In addition, the risk of nephrotoxicity associated with the use of intravenous (IV) contrast has not been adequately evaluated. Larger studies will be needed to define the role of CT pulmonary angiography for diagnosing invasive aspergillosis.

COVID-19-associated infection — Clinical features associated with pulmonary aspergillosis in patients with coronavirus disease 2019 (COVID-19) vary considerably. Most reported cases have occurred in patients who are critically ill and/or mechanically ventilated [38,53,54]. In one case series of 20 patients, pulmonary aspergillosis was diagnosed a median of 11 days after symptom onset of COVID-19 and 9 days after intensive care unit admission [55]. Radiographic findings included ground glass opacities, consolidation, bronchiectasis, and in some cases, nodules and cavities. These findings likely reflect both viral pneumonia, and the array of findings due to aspergillosis, ranging from early airway inflammation, to invasive pneumonia with necrotic cavitary disease.

Because clinical features of COVID-19 and pulmonary aspergillosis vary and overlap, diagnosis is challenging [34,37]. In general, suspicion for invasive aspergillosis should be raised in any patient with COVID-19 with radiographic features suggestive of invasive aspergillosis (ie, multiple nodules or cavitary disease). Because patients with severe COVID-19 often have marked bilateral ground glass opacities, typical findings can be obscured; thus, suspicion should also be raised in patients with prolonged or relapsing respiratory failure, particularly in those with structural lung disease, age >65 years, and/or immunocompromising conditions [28].

Tracheobronchitis — Aspergillus tracheobronchitis has been reported most commonly in lung transplant recipients but has also been described in other types of hosts (eg, recipients of other solid organ transplants, patients with hematologic malignancies, patients with HIV infection) [56]. (See "Fungal infections following lung transplantation", section on 'Tracheobronchial aspergillosis'.)

Affected patients typically present with prominent dyspnea, cough, and wheezing; they occasionally expectorate intraluminal mucus plugs. Chest imaging may be normal or reveal areas of airway thickening, patchy infiltrates, consolidation, or centrilobular nodules.

Different patterns of Aspergillus tracheobronchitis have been described:

● Obstructive bronchial aspergillosis, a condition in which thick mucus plugs filled with Aspergillus hyphae are found in the airways, with little mucosal inflammation or invasion

● Ulcerative tracheobronchitis, in which there is focal invasion of the tracheobronchial mucosa and/or cartilage by fungal hyphae

● Pseudomembranous tracheobronchitis, which is characterized by extensive inflammation and invasion of the tracheobronchial tree with a pseudomembrane composed of necrotic debris and Aspergillus hyphae overlying the mucosa

Tracheobronchial disease can also occur in other hosts, including those who have hematologic malignancies, hematopoietic cell transplantation [57], and chronic obstructive pulmonary disease.

Chronic necrotizing and chronic cavitary pulmonary aspergillosis — Patients who have underlying chronic lung disease are at risk for indolent forms of pulmonary aspergillosis, characterized by cavities or infiltrates that may or may not demonstrate hyphal invasion of tissue on histopathology. Presumably, the slowly progressive nature of this infection is a function of the host immune response, which is enough to hold the organism in check but not to eliminate it. Cough, weight loss, fatigue, and chest pain are common, and the chest radiograph shows a slowly progressive lesion, which can be better defined by CT scanning. However, interpretation of radiologic studies may be complicated by the presence of concomitant lung disease, since this is the setting in which chronic aspergillosis usually occurs.

Disseminated infection — In the presence of angioinvasive disease, Aspergillus spp can disseminate beyond the respiratory tract to multiple different organs, including the skin, brain, eyes, liver, and kidneys. Disseminated infection is associated with a very poor prognosis.

Rhinosinusitis — In the paranasal sinuses, aspergillosis can present in an identical fashion to mucormycosis. However, rhinocerebral aspergillosis is usually seen in neutropenic patients with hematologic malignancy, whereas mucormycosis occurs most commonly in those with diabetes mellitus or hematologic malignancy. (See "Fungal rhinosinusitis".)

The inflammatory response to the infection may be blunted because of neutropenia, and findings may be subtle. Nevertheless, nasal congestion, fever, and pain in the face and around the eye are common presenting features. If the orbit becomes involved, additional symptoms may include blurred vision, proptosis, and chemosis. The infection can also extend locally into the vasculature and the brain, leading to cavernous sinus thrombosis and a variety of central nervous system (CNS) manifestations.

CT findings may be subtle and can include focal soft tissue lesions, subtle focal bony erosions, and focal hypodense areas; magnetic resonance imaging (MRI) may also show soft tissue lesions as well as focal enhancement of the sinus lining [58].

Biopsy is necessary to establish the diagnosis; multiple biopsies are sometimes necessary [58].

Central nervous system infection — CNS aspergillosis may occur in the setting of disseminated infection as well as from local extension from the paranasal sinuses. Patients with CNS involvement with Aspergillus spp may present with seizures or focal neurologic signs [1]. Mycotic aneurysms develop in some cases and can rupture, resulting in a hemorrhagic cerebrovascular accident, subarachnoid hemorrhage, and/or empyema formation [59].

In a study of the CT and/or MRI findings associated with CNS aspergillosis, three patterns were observed [60]:

● Ring-enhancing lesions consistent with brain abscesses (image 3)

● Cerebral cortical and subcortical infarction with or without superimposed hematomas (image 4)

● Mucosal thickening of a paranasal sinus with secondary intracranial dural enhancement consistent with direct extension from the sinuses (image 5)

CNS infection is associated with a very poor prognosis.

Endophthalmitis — Aspergillus endophthalmitis may be a presenting feature of disseminated aspergillosis, in which involvement of the deep structures of the eye results not only from hematogenous spread. In other patients, corneal infection or direct inoculation following trauma is the genesis of infection [61]. Patients present with eye pain and visual changes [62].

Progressive infection is characterized by destruction of multiple components of the eye. The outcome is usually poor with loss of useful vision in the affected eye and a requirement for enucleation in some cases.

Endocarditis — Aspergillus spp are second only to Candida spp as a cause of fungal endocarditis [63]. This infection occurs primarily in patients with prosthetic heart valves. In many patients, infection occurs at the time of surgery, with the fungus contaminating the surgical site. Patients may present at any time postoperatively.

Other patients at risk for the development of Aspergillus endocarditis include patients with indwelling central venous catheters and people who inject intravenous drugs.

Patients typically present with fever and embolic phenomena. Blood cultures are rarely positive, even when a fungal isolator system is used. Microscopic examination of an embolus will reveal the typical hyphae suggestive of aspergillosis, but definitive microbiologic diagnosis depends upon culture of the organism.

The prognosis of Aspergillus endocarditis is poor. Even with combined medical and surgical therapy, the mortality rate has approached 100 percent [64].

Cutaneous aspergillosis — Cutaneous aspergillosis can be either primary, occurring from direct inoculation, usually in the presence of trauma, or secondary, occurring from contiguous extension or bloodborne spread (picture 2 and picture 3) [65].

A review noted that otherwise healthy hosts can develop cutaneous disease in surgical wounds or by traumatic inoculation, especially when this occurs in the presence of high spore counts, such as might occur during a farming accident [65]. While burn victims, neonates, and solid organ transplant recipients tend to develop primary cutaneous disease in the presence of prolonged local skin injury, patients with hematologic malignancies and hematopoietic cell transplant recipients more frequently develop disease due to contiguous or hematogenous spread to the skin.

Diagnosis can only be verified by skin biopsy, which should be taken from the center of the lesion and reach the subcutaneous fat in order to visualize hyphae invading blood vessels of the dermis and subcutis [65].

Gastrointestinal aspergillosis — Aspergillosis can involve the gastrointestinal (GI) tract, causing focal invasion as a primary site of inoculation and presenting as neutropenic enterocolitis (typhlitis), appendicitis, colonic ulcers, abdominal pain, and/or GI bleeding [66-68]. Direct inoculation from the GI tract is likely, with risk factors including neutropenia, receipt of glucocorticoids, and mucosal breakdown (mucositis).

Several fungi can cause invasive infections that are indistinguishable from invasive aspergillosis on clinical or radiographic grounds alone. Furthermore, the risk factors for invasive aspergillosis are similar to those for other invasive fungal infections. It is important to establish a microbiologic diagnosis since the treatment regimen depends upon the species isolated.

Fungi that can cause invasive pulmonary infections with clinical and radiographic features (eg, halo sign) that are similar to those caused by Aspergillus spp include:

● Mucorales (see "Mucormycosis (zygomycosis)")

● Fusarium spp (see "Mycology, pathogenesis, and epidemiology of Fusarium infection")

● Scedosporium apiospermum complex (see "Epidemiology, clinical manifestations, and diagnosis of Scedosporium and Lomentospora infections")

● Lomentospora prolificans (formerly Scedosporium prolificans) (see "Epidemiology, clinical manifestations, and diagnosis of Scedosporium and Lomentospora infections")

The halo sign can also be seen with other fungal infections and Pseudomonas aeruginosa infections [1].

Imaging findings of invasive pulmonary aspergillosis can also overlap with those of tuberculosis [69].

General approach — Culture of Aspergillus spp in combination with the histopathologic demonstration of tissue invasion by hyphae provides definitive evidence of invasive aspergillosis [1,2]. However, biopsy is frequently not feasible due to the risks of complications (eg, bleeding risk in patients with thrombocytopenia).

A rational first step to establishing the diagnosis of invasive aspergillosis involves the use of less invasive modalities, such as serum biomarkers (galactomannan, beta-D-glucan, and polymerase chain reaction [PCR] assays), and obtaining sputum and/or bronchoalveolar lavage (BAL) specimens for fungal staining, fungal culture, BAL galactomannan, and/or PCR.

A positive sputum fungal stain and/or culture should prompt therapy of hosts who are at risk for invasive aspergillosis. In the right clinical context, a positive BAL galactomannan antigen, BAL PCR for aspergillus species, or a positive serum galactomannan antigen can also provide a presumptive diagnosis [2-4]. In contrast, a positive serum beta-D-glucan assay can occur in the setting of various invasive fungal infections, including candidiasis.

Patients with clinical and radiographic findings that are suggestive of an invasive fungal infection, but in whom noninvasive testing and/or bronchoscopy is unrevealing, should ideally undergo lung biopsy.

Options for biopsy include bronchoscopy with transbronchial biopsy, computed tomography-guided transthoracic needle biopsy, and video-assisted thorascopic surgery. The most appropriate technique depends upon the location of the lesion(s), the individual patient's risk of complications from each procedure, and the necessity to establish the diagnosis. The decision regarding the need for a histopathologic diagnosis must be made on a patient-by-patient basis.

Suspicion for invasive aspergillosis should be raised in any patient with COVID-19 who has radiographic findings that suggest aspergillosis (ie, nodular opacities with surrounding ground glass infiltrates or cavitary disease). Because typical radiographic findings may be absent or obscured by those due to COVID-19, suspicion should also be raised in any patient with COVID-19 who has prolonged or relapsing respiratory failure without alternate explanation.

When invasive aspergillosis is suspected, we typically perform a bronchoscopy to directly visualize the airways and perform BAL to obtain samples for microbiologic diagnosis. In general, we perform fungal staining, fungal culture, galactomannan antigen, and/or PCR for Aspergillus on the BAL fluid. While the diagnostic utility of serum biomarkers is less clear in patients with COVID-19, we usually obtain both a serum galactomannan test and a beta-D-glucan test because positive results can help corroborate the diagnosis of invasive aspergillosis. Serum PCR testing should be performed, if available. These tests are typically performed as part of a more comprehensive evaluation for respiratory failure. A blind nonbronchoscopic lavage can be obtained if bronchoscopy is not feasible due to infection control concerns or other factors [9]. Tracheal aspirates can be used in conjunction with other assessments, but positive culture results and/or galactomannan results may reflect colonization rather than invasive disease [7,9].

For patients who have radiographic findings consistent with aspergillosis (ie, nodules or cavities) and lack of concern for other causes, detection of aspergillosis by culture, BAL or serum galactomannan, and/or PCR is typically sufficient for diagnosis. When consistent radiographic findings are absent, it is difficult to determine whether the detection of Aspergillus in the airways represents colonization or active infection; in such cases, we typically make the decision to treat on a case-by-case basis.

Aspergillus species are frequently inhaled into the airways, but because of effective conidial clearance in the majority of individuals, disease usually does not result. Because we inhale conidia constantly, culture isolation of Aspergillus species from the airway does not necessarily indicate disease. Thus, the diagnosis of invasive aspergillosis is based upon both isolating the organism (or markers of the organism) and the probability that it is the cause of disease. The latter is a function of the host's risk factors for disease (eg, immune status) and the clinical presentation. Demonstration of hyphal elements invading tissues from biopsy of any affected site, such as the lung or skin, represents a proven diagnosis [10].

Given the above issues, the diagnosis of invasive aspergillosis is often referred to within a scale of certainty: possible, probable, or proven [2,10,11]. These definitions have been developed in order to maintain consistency in clinical and epidemiologic studies, not to drive therapeutic decision making.

Direct examination of respiratory specimens — Respiratory specimens are usually stained with calcofluor white with 10% potassium hydroxide to detect the presence of fungal elements [12]. Gomori methenamine silver can be used to stain cytology preparations. Organisms can be observed as narrow (3 to 6 microns wide), septated hyaline hyphae with dichotomous acute angle (45°) branching [13]. However, several filamentous fungi, including Scedosporium spp and Fusarium spp, have similar appearances to Aspergillus spp on direct microscopy.

Culture — Culture of the organism, in combination with evidence of tissue invasion on histopathology or culture from a normally sterile site, provides the most certain evidence of invasive aspergillosis [10]. However, both microscopic examination and culture are insensitive [14], and therapy should not be withheld in the absence of such confirmation. Furthermore, performing a biopsy is not feasible in some patients due to bleeding risk or risk of other complications. In patients with risk factors and clinical and/or radiographic features that are suggestive of invasive aspergillosis, culture of Aspergillus spp from respiratory secretions provides adequate evidence of invasive disease.

Aspergillus is a rapidly growing fungus in the laboratory and is often visible in culture within one to three days of incubation. Sporulation is required for examination of spore-bearing structures, and diagnosis at the species complex level. Molecular testing is required for definitive diagnosis of the specific species within the complex.

Occasionally, organisms can be difficult to identify due to slow sporulation. Slow sporulating organisms may have been disregarded as non-pathogens in the past. However, more recently, slow-sporulating species (eg, Aspergillus lentulus, Neosartorya udagawae, and others) with variable susceptibility profiles have been identified and implicated in invasive infections [15]. These organisms should not be disregarded as contaminants based on the slow-sporulating phenotype.

Many patients with documented invasive aspergillosis have negative cultures. This has been observed in surveillance studies, and the prevalence of negative cultures varies depending upon the population being evaluated. As an example, in multicenter surveillance studies, only 25 to 50 percent of hematopoietic cell transplant (HCT) recipients who met criteria for invasive aspergillosis based upon galactomannan antigen results had positive cultures [16,17].

The predictive value of positive sputum or bronchoalveolar lavage (BAL) cultures is dependent upon both the host and the clinical presentation. In a study that evaluated the predictive value of lower respiratory tract cultures in different patient populations with probable or proven invasive pulmonary aspergillosis, the positive predictive value was highest in HCT recipients, patients with hematological malignancies, and granulocytopenic patients (72 percent), compared with solid organ transplant recipients and patients receiving glucocorticoids (58 percent) and patients with HIV (14 percent) [18]. The positive predictive value was highest in BAL cultures, in part because the prevalence of invasive aspergillosis is higher among patients who have radiographic abnormalities warranting bronchoscopy. Clinical and radiographic findings suggestive of invasive aspergillosis were present significantly more often among infected patients compared with uninfected patients.

Histopathology — In the setting of invasive disease, organisms can be observed in biopsy specimens as narrow (3 to 6 microns wide), septated hyaline hyphae with dichotomous acute angle (45°) branching (picture 1 and picture 2 and picture 3). The organism can be seen using Gomori methenamine silver or periodic acid-Schiff staining. However, several hyaline molds, including Scedosporium spp and Fusarium spp, have similar appearances to Aspergillus spp in histopathologic sections. Since the treatment of infections caused by these fungi may differ, it is important to confirm genus and species by culture.

Histopathologic examination can sometimes distinguish the above organisms from the order Mucorales(the agents of mucormycosis), which appear as broad, nonseptate hyphae that exhibit right-angle branching (picture 4 and picture 5). Occasionally, this distinction can be difficult, especially when small hyphal fragments are present or when the organism folds back on itself to create "pseudo-septations." Determining whether the fungus is one of the Mucorales or another fungus is important because the Mucorales are not susceptible to voriconazole, which is the treatment of choice for invasive aspergillosis.

Galactomannan antigen detection — Galactomannan is a polysaccharide that is a major constituent of Aspergillus cell walls. A double sandwich enzyme immunoassay (EIA) that utilizes the monoclonal antibody EB-A2 forms the current commercial (Platelia) assay, which has been approved by the US Food and Drug Administration (FDA) for testing serum and BAL fluid. This antibody detects multiple epitopes on galactofuranose side chains that are linked to the large polymer mannan backbone [19]. Galactofuranose is the six-member ring form of galactose, which is made by non-mammalian eukaryotes and some prokaryotic pathogens, and is a common antigenic moiety of different glycoconjugates [20]. Although once described as being Aspergillus specific, we now know that galactofuranose is present in other fungi and certain other substances. (See 'Caveats' below.)

Interpretation — The galactomannan EIA is performed with an optical read-out that is interpreted as a ratio relative to the optical density (OD) of a threshold control provided by the manufacturer; this ratio is called the OD index. Initial use of the test, largely in Europe, utilized a relatively high OD index (1 to 1.5) as a cut-off for positivity in order to ensure few false-positive results. Subsequent studies demonstrated that lower thresholds (0.5 to 0.7) provide relatively better performance [21,22]. The assay cleared by the FDA has a suggested threshold OD index of 0.5; thus, an OD index ≥0.5 is considered to be a positive result. For clinical trial purposes, in order to ensure a higher likelihood of diagnostic certainty, a threshold OD index of ≥1.0 in serum or plasma has been proposed and included in the European Organisation for Research and Treatment of Cancer (EORTC)/Mycoses Study Group Education and Research Consortium (MSGERC) definitions [2,23].

Serum — In some patients, galactomannan antigen can be detected in the serum before the presence of clinical signs or symptoms of invasive aspergillosis. Several studies have assessed its test characteristics, and the performance of the test has been addressed in a review and meta-analysis [24,25]. Some inconsistency in the reported performance of the test has been introduced by studies that used different cut-offs for positivity as well as other variables, such as different populations evaluated and testing on patients receiving concurrent antifungal drugs. One review noted that the sensitivity of the test varied from 30 to 100 percent, whereas specificity remained relatively high and constant (>75 percent) [26].

A meta-analysis that included 50 studies with a total of 5660 immunocompromised patients, including 586 with proven or probable invasive aspergillosis, reported that using an OD index cutoff of 0.5, sensitivity of the galactomannan EIA was 82 percent (95% CI 73-90 percent) and specificity was 81 percent (95% CI 72-90 percent) [25].

In another meta-analysis, subgroup analyses suggested that the assay performs better in patients who have a hematologic malignancy or who have received HCT; in comparison, the test performance may be limited in solid organ transplant recipients [27]. Whether this is related to biologic differences in the type of invasive disease, such as degree of fungal burden, is unclear. Few cases were included in the studies that have been performed; more patients who have different underlying diseases will need to be studied to establish the usefulness of this assay in disparate populations.

Caveats — The following caveats should be considered when using the galactomannan EIA:

● The sensitivity of detecting galactomannan in serum is decreased by concurrent administration of mold-active antifungal therapy [28-30].

● In the past, false-positive serum results were demonstrated in patients who were receiving intravenous piperacillin-tazobactam due to the presence of galactomannan (or a cross-reactive antigen) in the antibiotic formulation [31,32]. False-positive results persisted for as long as five days after discontinuation of piperacillin-tazobactam [32,33]. Results of more recent studies suggest that current preparations of piperacillin-tazobactam rarely react with the assay [34,35]. False-positive galactomannan results have also been reported with intravenous amoxicillin-clavulanate, a formulation that is not available in the United States [36].

● Many fungi demonstrate galactomannans (or polysaccharides containing galactofuranose residues) on their cell walls. False-positive results of the Aspergillus galactomannan EIA may be seen with infections caused by organisms that share cross-reacting antigens. These include filamentous Ascomycetes that cause similar disease presentations (eg, Fusarium species [37]) and others (Penicillium species [38], Histoplasma capsulatum [39]).

● False-positive results are more likely to occur during the first 100 days following HCT and in patients with gastrointestinal tract mucositis caused by chemotherapy or graft-versus-host disease (GVHD) [40]. The proposed mechanism is that galactomannan in foods or bacteria having cross-reactive epitopes may translocate across the intestinal mucosa during periods of impaired mucosal integrity [31].

● Another possible cause of false-positive galactomannan EIA results is contamination of foods with Aspergillus or closely related fungi, such as Penicillium spp. One report noted false-positive results associated

...

Role of testing — Studies evaluating the serum galactomannan EIA report relatively low (<50 percent) and high (>90 percent) positive and negative predictive values, respectively. These values are largely a function of the low prevalence of disease, even when measured in high-risk populations (usually <20 percent). One should thus consider the clinical context to determine the probability of infection [26]. (See "Glossary of common biostatistical and epidemiological terms", section on 'Predictive values'.)

Some investigators have suggested using the serum galactomannan EIA for screening (weekly or twice weekly) in order to detect invasive aspergillosis prior to the development of clinical signs or symptoms. Biomarker screening as a means to guide pre-emptive therapy was evaluated in a randomized trial; screening with galactomannan antigen and PCR led to reduced empiric antifungal treatment compared with a culture-based strategy [45]. The use of biomarker-driven strategies allows less empiric use of antifungal therapy for fever and more pre-emptive therapy based on positive biomarker results

The utility of the serum galactomannan EIA has been best established in the setting of suspected disease in high-risk patients with hematologic malignancies. In the setting of clinical disease suspicion, prevalence is higher, assuring better predictive performance. The assay may also be useful in other immunocompromised patients at risk for invasive pulmonary aspergillosis, but caution is necessary in interpreting results as sensitivity is not as high in patients who have disease that may be limited to the airways (eg, lung transplant recipients).

Some investigators have suggested a role for serial galactomannan testing in patients with documented invasive aspergillosis, both as a prognostic measure of effectiveness of antifungal therapy and to differentiate between worsening fungal infection and worsening radiographic findings developing as a result of the host immune response [46-50]. The prognostic value of the serum galactomannan EIA is discussed separately.

Bronchoalveolar lavage fluid — The galactomannan EIA detects fungal antigens even when the organism does not grow in the laboratory, providing an indication of potentially invasive disease. The galactomannan EIA performed on BAL fluid provides additional sensitivity compared with culture, estimated in most studies to exceed 70 percent [51-55].

The optimal threshold OD index for positivity continues to be debated; a higher threshold OD index results in lower sensitivity but higher specificity for invasive aspergillosis. In a meta-analysis of studies of BAL galactomannan in immunocompromised patients, an OD cutoff of 0.5 resulted in a sensitivity of 0.88 and a specificity of 0.81, whereas a cutoff of 1.0 resulted in a sensitivity of 0.78 and a specificity of 0.93 [55]. As noted above, the FDA considers an OD index of ≥0.5 to be positive for galactomannan EIA in both serum and BAL fluid, although a revised threshold of 1.0 for BAL fluid is now included in the EORTC/MSGERC definitions [2,23]. Higher OD levels are associated with greater diagnostic certainty.

It is important to note that false-positive results occur, and these are especially common in the setting in which the detection of fungi in the airways represents colonization as occurs in lung transplant recipients or when the fluid that is used for BAL washes is contaminated with galactomannan. However, one study performed in lung transplant recipients suggested that specificity is high (95 percent) [51]. Many centers have begun to utilize this assay on BAL fluid as an adjunct to other diagnostic tests when invasive aspergillosis is suspected. Because performance may be dependent on technical variables, such as type and amount of BAL fluid, institutions should establish specific protocols for testing.

The use of the galactomannan assay on BAL fluid from lung transplant recipients is discussed in greater detail separately.

Other specimen types — Although the galactomannan assay has been approved by the FDA for use only on serum and bronchoalveolar lavage fluid, galactomannan can also be detected in other samples, including cerebrospinal fluid and pleural fluid, depending on the clinical context [56].

Beta-D-glucan assay — 1,3-Beta-D-glucan, a cell wall component of many fungi, is detected by the beta-D-glucan assay. There are several different commercial assays available in different countries. The Fungitell assay has been cleared by the FDA as an aid to diagnose invasive fungal infections. These assays utilize the same principle as serum endotoxin assays, measuring activation of Factor G. Since the different marketed tests utilize different horseshoe crab substrates and different methods to measure output, performance may be variable. The output of the serum assay currently available in the United States is based on spectrophotometer readings, in which optical density is converted to beta-D-glucan concentrations; the results are interpreted as negative (range <60 pg/mL), indeterminate (60 to 79 pg/mL), or positive (>80 pg/mL) [57]. Importantly, these cut-offs were defined in the clinical context of identifying breakthrough invasive fungal infections (primarily, invasive candidiasis) in people who were undergoing treatment for hematologic malignancies. Precise cut-offs to optimize performance of the assay as an aid to diagnose invasive aspergillosis have not been defined and may be different.

In a 2011 meta-analysis that included 16 studies evaluating beta-D-glucan assays for the diagnosis of invasive fungal infections, the pooled sensitivity was 77 percent (95% CI 67-84 percent) and the pooled specificity was 85 percent (95% CI 80-90 percent) [ 58]. A 2012 meta-analysis that included six cohort studies of patients with hematologic malignancies noted a lower sensitivity (50 percent, 95% CI 34-65 percent) and a higher specificity (99 percent, 95% CI 97-100 percent) than previously reported; however, there were few cases of invasive aspergillosis tested to provide definitive conclusions [59]. In individual studies, the sensitivity has ranged from 55 to 95 percent and the specificity has ranged from 77 to 96 percent [57,60-64]. As in studies evaluating other fungal diagnostics, likely reasons for the differences in sensitivity and specificity between studies are that different thresholds were considered positive, different assays were used, patient populations varied, and study design was not uniform. Despite the substantial heterogeneity among different studies, the beta-D-glucan assay has good accuracy for distinguishing patients with proven or probable invasive fungal infections from patients without invasive fungal infection [58].

One study compared performance of the galactomannan EIA with the beta-D-glucan assay in sera from 105 patients with invasive aspergillosis and 50 healthy blood donors [65]. Results demonstrated a higher specificity with the galactomannan test (97 versus 82 percent) but a lower sensitivity (81 versus 49 percent).

Caveats — The following caveats should be considered when using the beta-D-glucan assay:

● The beta-D-glucan assay is not specific for Aspergillus species and can be positive in patients with a variety of invasive fungal infections, including candidiasis and Pneumocystis jirovecii (formerly P. carinii); the latter can be associated with beta-D-glucan levels greater than the upper limits of the assay. Although the beta-D-glucan assay may be positive in patients with a variety of invasive fungal infections, it is typically negative in patients with mucormycosis or cryptococcosis [62,63,66,67]. (See "Epidemiology, clinical manifestations, and diagnosis of Pneumocystis pneumonia in patients without HIV", section on 'Beta-D-glucan assay'.)

● The specificity of the assay can be decreased by multiple other clinical variables; the following factors have been reported to cause false-positive results [66]:

• Hemodialysis with cellulose membranes

• Intravenous immunoglobulin

• Albumin

• Use of cellulose filters for intravenous administration

• Gauze packing of serosal surfaces [68]

• Intravenous amoxicillin-clavulanic acid (a formulation that is not available in the United States) [69]

• Infections with certain bacteria that contain cellular beta-glucans, such as Pseudomonas aeruginosa [70,71]

Role of testing — The beta-D-glucan assay may be used for detecting invasive fungal infections early in the course of infection, prior to the onset of overt clinical findings. One study evaluated a screening strategy in which 95 patients receiving chemotherapy for acute leukemia underwent beta-D-glucan testing twice weekly in the absence of fever and daily in the presence of fever [72]. Screening appeared to shorten the time interval between suspected infection and established diagnosis, when compared with waiting for clinical signs and symptoms of disease. However, this strategy has not been validated in randomized trials. Because of its limited sensitivity, beta-D-glucan testing should not be used to rule out invasive fungal infections, particularly in patients with hematologic malignancies due to the high prevalence of invasive fungal infections in this population [73].

Polymerase chain reaction — When invasive aspergillosis is suspected, we typically include PCR, in conjunction with other biomarkers, in our diagnostic evaluation. PCR can be performed on serum, plasma, whole blood, or BAL fluid. As with other biomarkers, diagnostic certainty rises with two consecutive positive results. In addition to diagnosis, some PCR-based assay can detect mutations associated with antifungal resistance [3].

Molecular assays (eg, by PCR) have undergone extensive evaluation [4,28,39,57,60,74-80]. In a meta-analysis of 25 studies, sensitivity and specificity of PCR to detect invasive aspergillosis were 84 and 76 percent, respectively [79]. When at least two PCR results were positive, the sensitivity was 64 percent and the specificity was 95 percent. Another meta-analysis had similar findings [80]. These results suggest that two positive PCR results are highly suggestive of invasive aspergillosis. Standardized protocols for nucleic acid extraction, sample types, volumes, and processing have been developed by the European Aspergillus PCR Initiative/Fungal PCR Initiative [3].

Combining assays — In a meta-analysis of studies that assessed the diagnostic performance of serum galactomannan and serum or whole-blood PCR as weekly screening in high-risk patients, single positive test results had good sensitivity (92 percent for galactomannan; 84 percent for PCR) and specificity (90 percent for galactomannan; 76 percent for PCR) for detecting proven or probable invasive aspergillosis [88]. However, when positivity was defined as at least one positive result (either galactomannan or PCR), the sensitivity was 99 percent. When both the galactomannan and the PCR results were positive for the same patient, the specificity was 98 percent; specificity was also high with two positive galactomannan results (95 percent) or two positive PCR results (93 percent). In an open-label randomized trial of high-risk patients with hematologic malignancies who were not receiving antifungal prophylaxis, treating physicians were informed of either serum galactomannan and PCR results or of serum galactomannan results only; testing was performed twice weekly [89]. Positivity of either assay prompted chest computed tomography (CT) scan and initiation of antifungal therapy. The use of the serum galactomannan assay and PCR for screening was associated with earlier diagnosis and a lower incidence of invasive aspergillosis than the use of the galactomannan assay alone.

Most studies comparing the performance parameters of different assays (the galactomannan EIA, beta-D-glucan assay, LFD, and PCR) for use on BAL fluid emphasize the insensitivity of culture for isolation of Aspergillus species and suggest that the best approach may be to combine assays [83]. In a total of 78 BAL fluid samples, the sensitivity of all four methods ranged from 70 to 88 percent. The combination of the galactomannan EIA and either PCR or LFD increased sensitivity to 94 to 100 percent without compromising specificity. The beta-D-glucan assay applied on BAL fluid resulted in high false-positive rates.

Precipitin antibodies — Detection of precipitin antibodies may be useful for the diagnosis of allergy to various molds, including Aspergillus, but has no role in the diagnosis of invasive aspergillosis.

Aspergillus is ubiquitous worldwide—found in soil, water, food, and air—and is particularly common in decaying vegetation. The inoculum for infection is not known, but hosts with normal pulmonary host defenses rarely develop disease despite exposure to the organism with normal daily living through airborne conidia, including foodstuffs like pepper, herbs, and so forth.51

who use corticosteroids) that inhibit the activity of pulmonary macro- phages or those who are neutropenic, have increased susceptibility to these organisms.52

Patients with prolonged and profound neutropenia are at high risk for IA, but changing treatment patterns in chemotherapy and transplanta- tion, along with the use of growth factors, have limited the numbers of persistently neutropenic patients.53,54

It is important to recognize that, even among high-risk patients with hematologic malignancy, there is substantial heterogeneity of risk.55,56

Most cases of IA occur in the period of neutropenia after primary induction chemotherapy or in patients with refractory or recurrence of malignancy, with the risk of aspergillosis during consolidation chemotherapy very low.55,56

In patients undergoing HSCT or bone marrow transplantation, a recent increase in the incidence of IA has been reported, and the epi- demiology of infection has changed (Table 257.2).54,60

In patients undergoing HSCT the major periods of risk are bimodal, with peak incidence occurring fewer than 40 days but more than 100 days after transplantation, with mean time to diagnosis more than 180 days in some series.53,54,61–63 In the Transplant-Associated Infection Surveillance Network (TRANSNET) study the median time to onset of IA was 99 days, with 22% and 53% of 335 IA cases occurring 1 and 4 months after transplantation, respectively.54 One of the factors in the changing epidemiology in this patient population is the use of nonmyeloablative transplant procedures, which has shifted the major risk factor in these patients from neutropenia to that of the use of high doses of corticosteroids or other immunosuppressive therapy for the treatment of acute or chronic graft-versus-host disease (GVHD).53

Other immunosuppressed patients are also at risk for IA, including patients undergoing organ transplantation, particularly lung transplanta- tion (see Table 257.2).64,65 The increased risk in lung transplantation is because the transplanted organ is constantly exposed to the environment, ciliary clearance is reduced, and many of these patients are colonized with Aspergillus in either the native or transplanted lung.66

Other patients at higher risk for IA include patients receiving corticosteroids or biologic agents, including the tumor necrosis factor-α (TNF-α) antagonists, such as infliximab, etanercept, adalimumab, and others, including ibrutinib.68

IA occurs in other immunosuppressed patients, including patients with pulmonary diseases, acquired immunodeficiency syndrome (AIDS), chronic granulomatous disease, and other hereditary immunodeficiency syndromes.3,52

IA can occur in critically ill patients in intensive care units (ICUs) independent of traditional risk factors, suggesting additional cumulative risk factors for disease in this setting, although the overall impact of IA in that population is unclear and varies in different series.69,70

Outbreaks of IA have occurred in patients exposed to aerosols of Aspergillus conidia in association with construction and other environ- mental risks.71–74

It should also be noted that, with the prolonged period of risk, these infections become very difficult to prevent with protective environments because much of their health care will occur in a non– hospital-based setting.76

Invasive Syndromes Caused by Aspergillus IA most frequently begins in the lung after inhalation of Aspergillus conidia. Invasion of hyphae into pulmonary vasculature is common, occurring in as many as a third of patients with invasive pulmonary aspergillosis. Disseminated disease occurs either by hematogenous spread to distant sites or by contiguous extension from the lung. Hematogenous dissemination to the CNS or other organs, including the thyroid, liver, spleen, kidney, bone, heart, and skin, is common in patients with severe immunosuppression, such as those undergoing HSCT, and heralds a poor prognosis.3

Invasive Pulmonary Aspergillosis The most common manifestation of IA is invasive pulmonary aspergillosis (IPA). IPA rarely manifests before 10 to 12 days of profound neutropenia, which is a major risk factor for infection.125

The incubation period for developing IPA after inhalation of conidia is not known but in patients with neutropenia has been estimated at approximately 15 days.126 A significant number of patients have manifestations of IPA on admission or within the first 2 weeks of hospital admission, suggesting that community-acquired exposure is common.76,127

Symptoms of IPA include progressive cough, dyspnea, pleuritic chest pain, fever despite coverage with broad-spectrum antibiotics, and pulmonary infiltrates. These symptoms may be reduced in patients who are unable to mount an inflammatory response owing to profound neutropenia. In addition, although fever is common, it may be absent in those receiving high doses of corticosteroids.

Other clinical features of IPA include hemoptysis, pleural effusion, and pneumothorax. However, all the physical findings are nonspecific and may lag significantly behind the disease process.

Clinical characteristics may resemble a pulmonary embolism with pleuritic chest pain, hemoptysis, and dyspnea, which reflect the angioinvasive nature of the organism.

Laboratory studies are also nonspecific but may include elevation in bilirubin, lactate dehy- drogenase, and C-reactive protein, as well as coagulation abnormalities. Life-threatening hypoxia may occur in patients with extensive or progres- sive infection.

In extensive infection multiple diffuse nodular pulmonary infiltrates are readily seen on chest radiographs or chest CT (Fig. 257.7). Although these are not diagnostic, dense nodular lesions are associated with a poorer prognosis.128,129 Other pulmonary radiographic findings of IPA include the classic pleural-based, wedge-shaped densities or cavitary lesions, although the former are not commonly detected, and the latter are present late in the course of infection.128,130 Pleural effusions are more common than previously considered, but whether they are a specific manifestation of IPA is not established.128 The presence of a “halo” of low attenuation surrounding a nodular lesion on CT is an early finding in IA (Fig. 257.8A).128,130 Later in the course of infection, these nodular lesions may cavitate, usually in temporal association with recovery of neutrophils, forming an “air-crescent” sign (see Fig. 257.8B). These radiographic features are characteristic of IPA, but similar findings can also occur with other angioinvasive organisms, including Mucorales, Fusarium, and Scedosporium, as well as bacterial pathogens

Tracheobronchitis Aspergillus in the airways can range in significance from colonization, which is common in lung transplantation, to ulcerative tracheobron- chitis.66 The syndrome of Aspergillus tracheobronchitis typically occurs in patients undergoing lung transplantation and in patients with AIDS and is characterized by extensive pseudomembranous or ulcerative lesions due to Aspergillus. In patients undergoing lung transplantation the infection often occurs at the suture line of the lung transplant and can lead to dehiscence of the anastomotic site.

Symptoms of tracheo- bronchitis are nonspecific and include dyspnea with associated pulmonary function abnormalities, cough, chest pain, fever, or hemoptysis. Symptoms may be mild and can be confused with other causes, including rejection. Results of plain radiographs may be normal so that clinical suspicion is needed to establish the diagnosis, which is accomplished by bronchoscopy with biopsy to document tissue invasion. A prolonged course of a systemic antifungal agent therapy is usually required for treatment, although aerosols of liposomal formulations of amphotericin B have been used for localized disease.117

A proven diagnosis of IA requires a tissue biopsy showing invasion with hyphae and a positive culture for Aspergillus.132 The diagnosis can also be established with positive cultures from a normally sterile site, such as a needle biopsy or cerebrospinal fluid (CSF), although blood cultures are rarely positive.148 Tissue biopsies may not be possible because of potential risks, although Aspergillus hyphae are easily seen with common fungal stains, such as Gomori methenamine silver or periodic acid–Schiff.

Aspergillus hyphae are hyaline, septate, acute-angle branched, even diameter, and 3 to 6 μm in width. Although these features usually distinguish Aspergillus from agents of mucormycosis, they are not distinguishable from a number of other opportunistic molds, including Fusarium, Lomentospora (formerly Scedosporium), Paecilomyces, and others so that a positive culture is needed to confirm the diagnosis.149

Cultures for Aspergillus in respiratory samples in high-risk patients, particularly if obtained via bronchoalveolar lavage (BAL), can support the diagnosis of probable IA.132 Aspergillus can also be cultured from patients in whom no clinical illness is apparent, and positive cultures in patients with a low risk for IA should be interpreted with caution.29

Radiographic findings can also be used in the diagnosis and manage- ment of invasive pulmonary aspergillosis. Plain chest radiographs are of limited diagnostic utility because they are insensitive and findings are nonspecific.128

A “halo” of low attenuation surrounding a nodular lesion on CT imaging is an early finding in invasive pulmonary aspergil- losis and has been used as a marker for initiating early antifungal therapy.4,128,150 The volume of lesions may increase over the first 7 days of infection, even when therapy is successful, so that early radiologic progression should be interpreted cautiously.130 The CT findings of IPA have been validated in high-risk, neutropenic, and bone marrow transplant recipients, but in other patients, including solid-organ transplant recipients and those with nontraditional risk factors in the ICU, the CT findings are not as well defined.

Non–culture-based methods have been used to establish a rapid diagnosis of IA.151

Antibody detection is of limited utility because immunosuppressed hosts fail to mount an antibody response even with invasive infection.86

Detection of galactomannan (GM) by enzyme immunoassay has contributed substantially to the diagnosis of probable IA.152 This assay has been validated in a variety of patient groups, particularly with serial measurements. Studies have suggested a sensitivity as high as 89% and a specificity of 92% in high-risk HSCT patients not receiving antimold prophylaxis.153 Other studies have found the assay to be less sensitive (40%–50%), reflecting the impact of prior antifungal therapy in reducing the level of circulating GM, a limited number of samples per patient, extent of infection, and other variables.152–154 This has resulted in recommendations for a lower cutoff for a positive test result, especially in high-risk patients with greater probability of infec- tion.153,155 A meta-analysis using an optical density index cutoff of ≥0.5 showed a sensitivity of 78% and a specificity of 81%.156 False-positive results have been reported in some children, including in some neonates, which may be due to dietary intake or the presence of cross-reacting antigens with bacteria such as Bifidobacterium, and in patients receiving therapy with piperacillin-tazobactam—which now appears to be uncommon—and other antibiotics.152,157

GM detection has also has been used for other body fluids, such as CSF, and in BAL fluid.152,158 GM results from BAL fluid testing demonstrate sensitivity that is greater than that seen in serum testing and may be less affected by antifungal therapy.159–161 A negative test result may be useful in ruling out a diagnosis of IPA.160 Some colonized patients, such as those with chronic obstructive pulmonary disease or those undergoing lung transplantation, may have a positive GM result and not have evidence of invasive infection.162 Serial assessments of GM may offer prognostic value in outcomes of infection. Other potential markers also include the nonspecific fungal marker (1→3)-β-d-glucan.163,164

Several reports demonstrate the potential for using polymerase chain reaction (PCR) as an early diagnostic marker for IA, which may be more sensitive that other methods.165 These assays have not been tradition- ally standardized, and it was thus problematic to determine their diagnostic utility.166,167 More recently, PCR has become commercially available, allowing standardization and quality control schemes to be implemented. The specificity of BAL PCR appears to be greater than GM testing, providing better utility for a diagnostic test.168

Susceptibility testing for Aspergillus has been standardized, but correlation with clinical responses has not been well established.169

Antifungal resistance to itraconazole has been reported developing after exposure to antifungal therapy and is associated with point mutations in the cyp51A gene target of triazoles antifungals.170 That resistance mechanism, which correlated with lack of efficacy in an animal model, has not been a common finding. However, resistant isolates have also been detected in azole-naïve patients.171,172 In the initial report all cases were observed since 1999, and the prevalence of resistant A. fumigatus ranged from 1.7% to 6%. The authors speculated that resistance in A. fumigatus could be related to the widespread use of agricultural triazole fungicides.172

It is not clear how frequently these mutations occur clinically, but the lack of azole responsiveness in patients with these isolates suggests it may become prudent to consider that possibility. Different mutations induce differential triazole susceptibil- ity, with some mutations causing resistance to voriconazole and isavu- conazole, whereas others cause resistance to posaconazole and itraconazole, and others cause panazole resistance.177 Other species, such as Aspergillus terreus and cryptic Aspergillus spp., may be resistant to amphotericin B or other antifungals, so testing of those species may be warranted.11,30,178