Fungal Infection during COVID-19: Does Aspergillus Mean Secondary Invasive Aspergillosis?

To the Editor:

We read with great interest the letter “COVID-19–associated Pulmonary Aspergillosis” by van Arkel and colleagues (1). The authors report a high incidence of presumed invasive pulmonary aspergillosis (6 of 31; 19.4%) among patients with coronavirus disease (COVID-19) admitted to intensive care.

In light of the recent studies revealing the high incidence of influenza-associated pulmonary aspergillosis (2), it seems natural to expect similar complications in severe forms of COVID-19 pneumonia. However, we would like to discuss some particular points.

Among the six patients presented by the authors, two had chronic obstructive pulmonary disease (COPD), another had asthma with inhaled corticoid therapy, and a fourth received oral prednisone. The association between aspergillosis and COPD is well known; in a recent prospective study, 14% of patients with COPD exacerbations had respiratory samples with Aspergillus spp. of unclear clinical significance (3). Corticosteroid therapy is also a known risk factor for Aspergillus colonization (4). Furthermore, these four patients had an A. fumigatus–positive culture on a single respiratory sample, and aspergillosis was diagnosed within 3–5 days after ICU admission. Because no previous negative respiratory sample was available, the preexisting presence of Aspergillus cannot be ruled out. Finally, the fifth patient had only a single positive galactomannan on a BAL.

Eight years ago, Blot and colleagues proposed criteria for defining invasive aspergillosis in critically ill patients using histopathology-controlled cases as references (5). For immunocompetent patients, the direct examination of hyphae in respiratory samples was mandatory to classify the case as putative, which would have excluded all the presented cases. Criteria have evolved since then and are now similar to those suggested by the authors for the diagnosis of COVID-19–related aspergillosis, that is, Aspergillus spp. cultured from BAL (without direct examination) or a galactomannan index of 1 or greater on BAL or 0.5 or greater on serum. However, diagnosing an invasive aspergillosis in an immunocompetent individual solely on a single positive respiratory specimen culture or a single galactomannan index might be adventuresome. Indeed, this generates a risk of artificially increasing the incidence of aspergillosis and the overuse of antifungal treatments.

To illustrate this point, we present two cases of patients hospitalized for severe COVID-19 in our institution. Following the same criteria, secondary invasive aspergillosis would have been diagnosed in them. Nevertheless, they presented favorable outcomes without any antifungal treatment, which from our point of view makes the diagnosis refutable.

The first patient was a 66-year-old immunocompetent man with type 2 diabetes who was hospitalized for 21 days in our ICU. A BAL performed on Day 7 was positive for galactomannan (index = 3.2). Antifungal treatment was not administered because the direct examination, the A. fumigatus PCR, and the culture were negative, as were the serum galactomannan and β-d-glucan. The patient presented a favorable outcome and was later discharged from the hospital.

The second patient was a 38-year-old woman with obesity, hypertension, type 2 diabetes, and rheumatoid arthritis treated with methotrexate. She required venovenous extracorporeal membrane oxygenation and was therefore hospitalized in our ICU for 11 days. Several colonies of A. niger were found on a protected distal respiratory sample performed on Day 4 but not on later samples. This patient did not receive any antifungal treatment, but her respiratory state improved nonetheless.

The diagnosis of invasive aspergillosis is difficult and based on a body of arguments. This is especially true for critically ill patients in whom clinical arguments are not discriminant and computed tomographic scan is either lacking or difficult to analyze. This is why mycological arguments play a crucial role in the diagnostic approach. Consequently, we ask ourselves whether the use of more stringent criteria that are not limited to a single mycological argument would be preferable.

Also, careful attention should be paid to clearly differentiate aspergillosis as a subsequent complication of severe COVID-19 pneumonia from aspergillosis in patients with underlying chronic respiratory diseases (that may be occult noninvasive forms preexisting COVID-19).

We acknowledge that COVID-19 might be an independent risk factor for subsequent aspergillosis. It is also possible that underlying pulmonary conditions may favor COVID-19–associated aspergillosis. We also fully agree with the authors that classifying aspergillosis cases is very challenging, especially in the ICU setting. Common efforts should therefore be made to further assess the pathogenic nature of the presence of Aspergillus in respiratory samples in ICU patients with severe COVID-19.