Coronavirus disease 2019 (COVID-19)-associated pulmonary aspergillosis (CAPA) is a life-threatening fungal infection that affects critically ill patients with COVID-19_1-3. Invasive Aspergillus infection is an opportunistic infection that affects immunocompromised hosts_4,5. Aspergillosis cases continue to increase with the growing number of the immunosuppressed population such as patients with hematologic malignancy and organ transplantation recipients₆. Severe influenza and COVID-19, which deteriorates the mechanical and immunologic barrier functions of airways, are novel risk factors for invasive Aspergillus infection_5,7. CAPA has become a cause of excess mortality in critically ill patients with COVID-198, following the announcement of the European Confed- eration of Medical Mycology/International Society for Human and Animal Mycology (ECMM/ISHAM) consensus criteria for CAPA diagnosis₁. Previous multicenter prospective_9-11 and retrospective_10,12 cohort studies have suggested underlying pulmonary disease and treatment with high-dose cortico- steroids as risk factors for CAPA development. Similar results were reproduced in two large, multicenter, retrospective cohort studies in Korea, which showed chronic pulmonary disease and treatment with high-dose steroids as risk factors for CAPA development_13,14. However, CAPA diagnosis remained challenging because of its atypical presentation with manifestations ranging from fungal colonization to angioinvasive disease_7,15. Patients diagnosed with CAPA are usually less immunosuppressed than with invasive pulmonary asper- gillosis (IPA). It contributed to less prominent angioinvasion in patients with CAPA, which is associated with lower serum galactomannan detection prevalence₉, and infrequent typical imaging finding such as halo, reversed halo, and air-meniscus signs₁₆. Additionally, distinguishing between clinical deterioration caused by COVID-19 progression and CAPA development is difficult_3,16. Needs for CAPA diagnosis and treatment remained unmet as the COVID-19 pandemic continues. The present report describes 11 patients diagnosed with CAPA in a tertiary care hospital in the Republic of Korea and summarizes their clinical presentation, management, and outcomes.

This retrospective, observational consecutive, case series evaluated patients who developed CAPA at Seoul National University Hospital (SNUH), which is a large, tertiary care hospital in Seoul, Republic of Korea. Inclusion criteria were patients aged >18 years, had laboratory-confirmed COVID-19 and were admitted to the intensive care unit (ICU) at SNUH from January 1, 2020, to August 31, 2021. Patients positive on galactomannan (GM) assays and those with con- firmed Aspergillus growth on fungus culture were included. Serum GM optical density index of >0.5 or bronchoalveolar lavage (BAL) GM optical density index of ≥1 were regarded as GM positive according to mycological evidence of CAPA in 2020 ECMM/ISHAM consensus criteria. The electronic medical records of patients with definite, probable, or possible CAPA were reviewed according to ECMM/ISHAM consensus criteria₁. The demographic characteristics, underlying medical conditions, total prescribed corticosteroid dosage, and duration before CAPA diagnosis were recorded. Clinical data on patient presentation, management, and outcomes were summarized. This study was approved by the Institutional Review Board of SNUH (H-2108-172-1248). Details of the case series were reported according to PROCESS guidelines₁₇.

This stud included 11 patients, including 6 (54.5%) males and 5 (45.5%) females, with a median age of 72 years (range, 49~85 years) were diagnosed with probable CAPA according to ECMM/ISHAM consensus criteria1. Their under- lying medical conditions are shown in Table 1. One patient (patient 2) had a hematologic malignancy, which is a classical host factor4 for IPA, and she was the only patient who has the classical host factor before admission. This patient had undergone autologous hematopoietic stem cell transplan- tation for diffuse large B-cell lymphoma a month before the COVID-19 diagnosis. Three patients had a history of solid cancer, including two with active malignancy receiving chemotherapy. None had documented evidence of prolonged neutropenia. All patients had received corticosteroid therapy before the CAPA diagnosis. The mean total dose of cortico-steroid therapy from admission to CAPA diagnosis was 220 mg of dexamethasone equivalent dose (range, 80~572 mg), and the mean duration of corticosteroid treatment was 15 days (range, 4~34 days). The mean daily dexamethasone dose before CAPA diagnosis was 14.6 mg (range, 9.2~16.8 mg). The median time from ICU admission to CAPA diagnosis was 12 days (range, 5~36 days). Initial mycological evidence in 10 patients consisted of serum GM positivity (mean optical density index: 1.59), whereas only one patient (patient 1) was diagnosed based on the presence of GM in BAL fluid. The BAL sample from patient 1 was obtained after the end of isolation because an in-house regulation limited the BAL sample analysis from isolated patients with active COVID-19 due to concerns about transmission. The respiratory specimens of two other patients with probable CAPA were culture positive for Aspergillus, although all microbial identification results were obtained after the clinical CAPA diagnosis.

Upon CAPA diagnosis, all 11 patients showed aggravation on chest X-ray images. One patient (patient 7) showed a cavitary change and two patients (patients 3 and 8) had pulmonary nodules, which is a pulmonary aspergillosis indicator_1,4. The other patients showed increased pulmonary consolidation or patchy opacities and findings that are nonspecific and indistinguishable from signs of aggravated COVID-19 pneumonia. Computed tomography (CT) was not considered upon diagnosis because of the risk of transmission and clinical instability although some patients were evaluated by CT. All patients underwent bronchoscopy during their hospital stay, but only four were evaluated before their clinical CAPA diagnosis. Two patients (patients 3 and 4) showed patchy lesions and one (patient 3) had ulcerative lesions on the airways, which are signs considered suspicious of Aspergillus tracheobronchitis1. Nonspecific bronchoscopic findings, such as hyperemic mucosa and secretions, were observed in nine patients.

Voriconazole-based antifungal treatment was administered in eight patients, whereas specific treatment targeting Aspergillus species was not administered in the other three patients (Table 2). One patient (patient 8) started fluconazole 10 days after the CAPA diagnosis because of azole-sensitive Candida albicans from one pair of blood cultures from a central catheter, but it was irrelevant to CAPA. The fluconazole was discontinued after the 7-day treatment course because C. albicans was regarded as a contaminated organism rather than a true pathogen. Other fungal pathogens and any other antifungal regimens for therapy among 11 patients were not reported. The median time from CAPA diagnosis to treatment initiation was 2 days (3 days before diagnosis to 13 days after diagnosis). One patient (patient 4) was started on voricona- zole before meeting the ECMM/ISHAM diagnostic criteria. This patient was suspected of having CAPA based on ongoing respiratory distress accompanied by radiologic deterioration with two consecutive elevated serum 1,3-β-D-glucan levels and unspecified mold growth in respiratory samples despite the lack of definitive mycological evidence. This patient was later found to be serum positive for GM, and a culture of a BAL specimen showed growth of Aspergillus fumigatus.

The median duration of antifungal drug administration was 30.5 days (range, 2~189 days) in patients who received antifungal treatment. The median antifungal treatment dur- ations of survived and deceased patients were 123.5 days and 16.5 days, respectively. Among eight patients who received voriconazole, three completed antifungal therapy (Table 2). Two of them (patients 3 and 4) survived and were transferred to the general ward, and subsequently to a nursing hospital. The other three patients (patients 5, 7, and 10) discontinued the antifungal medication due to the grave clinical course and medical futility. They died within 10 days of treatment discontinuation. One patient (patient 6) died 2 days after the antifungal treatment initiation. Another patient discontinued voriconazole because of an uncertain CAPA diagnosis although the patient met the ECMM/ISHAM criteria. Overall, all patients who did not receive antifungal treatment died, as did 6 (75%) of the 8 patients treated with voriconazole treatment.

This study describes 11 patients who presented with pro- bable CAPA from January 1, 2020, to August 31, 2021, at SNUH, which is a tertiary care hospital in the Republic of Korea. All patients received high-dose corticosteroid therapy before CAPA diagnosis, but most did not have classical host factors for IPA before admission. Classical host factors for IPA include prolonged neutropenia, hematologic malignancy, receipt of an allogeneic stem cell transplant or solid organ transplant, prolonged use of corticosteroids (≥0.3 mg/kg of corticosteroids for ≥3 weeks in the past 60 days), treatment with immunosuppressants, inherited severe immunodeficiency, and grade III-IV acute graft-versus-host disease₄. This study revealed that only one patient had underlying hematologic malignancy as a classical host factor before admission. However, three patients received more than a three-week course of corticosteroids after admission which also met the classical risk factor for IPA. The international guideline on COVID-19 treatment recommended that hospitalized, critically ill patients with COVID-19 should be treated with 6 mg/day of dexamethasone for up to 10 days_18,19. Additionally, a multicenter, randomized controlled trial published before the COVID-19 pandemic revealed the survival benefit of early administration of high-dose dexamethasone (20 mg/day for 5 days, followed by 10 mg/day for the next 5 days) on patients with moderate-to-severe acute respiratory distress syndrome (ARDS)₂₀. All patients in this study who required mechanical ventilation had been already receiving 6 mg of dexamethasone when they were admitted to the ICU. We escalated the dose and duration of dexamethasone because they progressed to moderate-to-severe ARDS despite the usual dexamethasone dose. The mean daily dose of dexamethasone before CAPA diagnosis in the present study (14.6 mg/ day of dexamethasone equivalent) was higher than that of a previous Korean nationwide multicenter cohort study (7.5 mg /day of dexamethasone equivalent for the first 10 days)₁₃, but less than that of another multicenter cohort study in Korea (22.4 mg/day of dexamethasone equivalent)₁₄. Com- paring the use of corticosteroids in different studies is difficult because it depends on the severity and comorbidities of patients. However, it might be considered in clinical practice because many previous studies suggested high-dose cortico- steroids as a risk factor for CAPA development_9-14.

All individuals in this study showed deterioration on simple radiography upon the initial diagnosis. Abnormal radiographic findings alone were insufficient to suspect CAPA because most chest X-ray findings are nonspecific. CT scanning is a more sensitive imaging modality, but the risks of transmission and clinical deterioration during intrahospital transport were considered. Bronchoscopy was more frequently performed in our hospital because it can be performed at the bedside. Additionally, bronchoscopy has been reported safe for medical personnel who use personal protective equipment_21,22. However, invasive procedures, such as biopsy and BAL, were not routinely performed due to the risk of complications such as bleeding and respiratory deterioration. Bronchoscopy findings of tracheobronchial ulceration, nodule, pseudomem- brane, plaque, or eschar are required for a CAPA tracheobronchitis diagnosis₁, with only 2 of the 11 patients in the present study showing significant findings.

Mycologic evidence has a crucial role in CAPA diagnosis in practice because radiologic and bronchoscopic abnormalities are often nonspecific. The presence of GM in serum played a pivotal role in CAPA diagnosis because the ability to obtain BAL samples from isolated COVID-19 patients was limited in our hospital. This might result in CAPA underestimation, as <20% of patients with CAPA were reported to have sera positive for GM_23,24. Moreover, the high serum GM positivity rate in the present study may explain the high mortality rate (81.1%) among these patients because serum GM positivity has been associated with more angioinvasive disease and poor outcomes₂₃.

The treatments of choice for CAPA consist of antifungal agents, including voriconazole and isavuconazole_1-3. The pre- sent study revealed that 8 of 11 patients received voriconazole-based antifungal therapy. Isavuconazole was not available in our center during the study period because it was approved in January 2020 in Korea although isavuconazole has com- parable efficacy but fewer toxicities than voriconazole₂₅. Of the 8 patients who received voriconazole, 2 (25%) survived; these patients have a longer antifungal therapy duration than those who did not survive. A minimum of 6~12 weeks of antifungal treatment has been recommended₁, but the optimal treatment duration has not been determined, and few studies to date have evaluated long-term outcomes.

In conclusion, Aspergillus infection can be complicated in critically ill patients with COVID-19. Clinical suspicion and mycologic evidence from optimal specimens are key to early diagnosis because of its atypical presentation. An antifungal drug that targets Aspergillus should be initiated upon CAPA suspicion, as a lack of treatment is associated with poor outcomes.