Invasive Aspergillosis in the ICU

Abstract & Commentary

By David J. Pierson, MD, Editor, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, is Editor for Critical Care Alert.

Synopsis: With better immunosuppressive therapy and ICU care, invasive aspergillosis is being encountered more often. Making the diagnosis is challenging, especially in lower-risk patients such as those with COPD and cirrhosis. Despite availability of effective new antifungal agents with less toxicity than amphotericin B, the effectiveness of these drugs in critically ill patients is uncertain, and the prognosis remains poor.

Source: Meersseman W, et al. Clin Infect Dis. 2007;45:205-216.

Meersseman and colleagues at Gasthuisberg University Hospital in Leuven, Belgium, have extensive experience in studies of invasive aspergillosis (IA) in critically ill patients, including those without the traditional risk factors.1 In this comprehensive review, these investigators summarize current knowledge of the risk factors, clinical manifestations, available diagnostic techniques, and treatment of IA, focusing on patients in the ICU.

In the general inpatient hospital population, invasive fungal infections have become more prevalent during the last 10-15 years. Much of this increase is accounted for by IA, although few data on IA specifically in critically ill patients have been published. Patients with neutropenia (less than 500 neutrophils/mL) are at the highest risk for IA, as are those with hematologic malignancies and allogenic bone marrow transplants. However, categories of patients who are at lower but still-substantial risk have increasingly been identified. These include those with prolonged treatment with corticosteroids prior to ICU admission, and patients with solid-organ cancer, HIV infection, lung and autologous bone marrow transplantation, and systemic diseases requiring immunosuppressive therapy. Also recognized as being at increased risk for IA are patients with COPD and cirrhosis, particularly in the latter case if they have been in the ICU for more than 1 week. Patients at relatively lower risk for IA include those with burns, other solid-organ transplants (ie, kidney, heart, or liver), prolonged ICU stays, malnutrition, and treatment with corticosteroids for 1 week or less.

Because of increased awareness of the possibility of invasive fungal infections in ICU patients who are being evaluated for prolonged fever or pulmonary infiltrates, it has become commonplace to send sputum, blood, and other microbiology specimens for fungal culture as well as for routine bacteriology. In an attempt to improve the efficiency of the microbiology services in their hospital, Bouza and colleagues in Madrid evaluated 404 isolates of Aspergillus fumigatus in 260 patients, 37% of whom were in ICUs or special hematology wards.2 In their study, A. fumigatus was isolated 2.1 times per 1,000 admissions, and 1 time per 1,000 microbiology samples, representing 5.6% of the total fungal isolations during the 3-year study period. These investigators used clinical information about the patients with positive cultures to derive a predictive model for the probability of their having IA (see Table 1).

In their study of 102 patients with positive cultures for aspergillus in their medical ICU,1 Meersseman et al found that almost all of them had required mechanical ventilation. Of the 56 patients with IA (26 with underlying hematological malignancy and 30 without malignancy), more than half had evidence of IA at the time of ICU admission. Individual case reports suggest that some patients acquire the infection while in the ICU, although available evidence indicates that most cases involve activation or progression of previously acquired infection in the context of critical illness.

Clinical presentations of IA most often encountered in the ICU include (a) the aggressive, angio-invasive form typically seen in neutropenic patients, (b) cavitating pulmonary infiltrates most often observed in patients on corticosteroids, those with COPD or cirrhosis, or in solid-organ transplant recipients, (c) anastomotic infections in lung transplant recipients, and, rarely, (d) miscellaneous presentations such as wound infections, mediastinitis (in cardiac surgery patients), and endocarditis.

The diagnosis of IA in ICU patients who are not in the classic high-risk category is challenging, because the presentation tends to be clinically nonspecific and the sensitivity and specificity of most commonly used tests vary. Angio-invasive IA typically produces multiple small nodules with characteristic halos on chest CT, but these signs are seldom present in patients in the lower-risk categories. Most patients in these categories are in the ICU because of processes—such as pneumonia or acute lung injury—whose clinical and radiographic signs obscure or mimic those of IA. Cultures of respiratory specimens in such patients are both insensitive and nonspecific, and fungal stains of such specimens are negative in at least half of patients subsequently proven to have IA. Serologic tests such as galactomannan and β-1,3D-glucan, and PCR techniques for detection of fungal DNA, are increasingly available, although very few published data on the effectiveness of these tests are from ICU patients. In one study of IA in a general medical ICU population, serum galactomannan was positive in only 53% of patients with documented IA.1 Although there has been the suggestion that this test is more sensitive in bronchoalveolar lavage fluid (BALF) than in serum, this is yet to be confirmed, and making the diagnosis continues to be challenging.

For ICU patients in all risk categories, IA carries a very unfavorable prognosis and responds poorly to available antifungal therapies. Voriconazole has recently become the standard of care for treating IA, replacing the more toxic amphotericin B.3 Other antifungal agents of potential future value in this condition include posaconazole and the echinocandins caspofungin and anidulafungin. Lipid-based formulations of amphotericin B, which are touted as being less toxic than the traditional version, have also seen increasing use. However, for all these antifungals, data on treating IA in ICU patients are exceedingly sparse, and for multiple reasons the response rates in such patients would be expected to be less favorable than in most of those included in clinical trials to date.


With ever-increasing availability of new immunosuppressive drugs, along with improvements in life support and other aspects of ICU care, clinicians can expect to encounter IA more and more frequently in critically ill patients. As this thorough review by Meersseman et al demonstrates, diagnosis and treatment of IA in patients who are not in the classic high-risk groups are especially challenging. Table 2 summarizes current information about approaching the possible diagnosis of IA in critically ill patients.

The relative likelihood of this infection is much greater in certain patient categories than others, and fortunately the diagnosis is often easier to confirm in them than in the larger numbers of patients at lower yet still important risk. Neither positive nor negative findings on chest imaging, stains and cultures of respiratory tract specimens, and serum tests can be considered definitive, and the clinician is faced with synthesizing complex and sometimes contradictory results in attempting to diagnose IA. In most instances, definitive diagnosis continues to require histologic demonstration of tissue invasion.


  1. Meersseman W, et al. Invasive aspergillosis in critically ill patients without malignancy. Am J Respir Crit Care Med. 2004;170:621-625.
  2. Bouza E, et al. Workload due to Aspergillus fumigatus and significance of the organism in the microbiology laboratory of a general hospital. J Clin Microbiol. 2005;43:2075-2079.
  3. Herbrecht R, et al. Voriconazole versus amphotericin B for primary therapy of invasive aspergillosis. N Engl J Med. 2002;347:408-415.