Study population

This was a prospective observational study conducted in a major teaching hospital. Critically ill patients with SARS-CoV-2 pneumonia admitted to the ICU since the onset of the COVID-19 pandemic were included. The study period ran from March 2020 to June 2022. Patients were eligible if they were: (1) 18 years or older; (2) receiving invasive ventilation for COVID–19 SARS–CoV-2 infection confirmed by RT–PCR. Exclusion criteria were patients not requiring invasive mechanical ventilation, or in whom bronchoscopy was deemed unsafe. Diagnostic fibrobronchoscopy was performed in all mechanically ventilated patients on admission to the ICU or with suspected secondary infection. No patients received antifungal therapy prior first bronchoscopy bronchoalveolar lavage.

Samples processed

Bronchoscopy and bronchoalveolar lavage (BAL) was performed with 150 ml of physiological saline solution, divided into three aliquots. The first 20 ml of the BAL was discarded, and a sample of the remaining fluid was collected for microbiological analysis. Fibrobronchoscopy was performed with disposable AMBU aScope™ 4 Broncho Regular OD 5.8 / 2.8 mm bronchoscopes (Copenhagen, Denmark). The bronchoscope was inserted using a special adapter valve to avoid generating aerosols during the procedure. A specialist pulmonologist with 30 years of experience not related to the research study randomly analyzed the BAL samples to perform a quality control; those with few alveolar macrophages and/or excessive numbers of airway-derived cells were considered unsuitable due to the low quality of the sample. No local anesthetics were administered, and aspiration was not attempted through the bronchoscope channel until it was in position at the site to be studied. BAL samples were sent to microbiology the service to be processed.

Microbiological studies

Qualitative galactomannan (GM) detection, culture on fungal media and calcofluor white staining were performed within the first hours, all tests were performed from the same BAL fluid. All BAL fluids were stored at -80ºC for quantitative GM detection and quantitative PCR (qPCR) assay for detection of the Aspergillus genus.

Qualitative GM detection

The Aspergillus-specific lateral flow device (AspLFD, OLM Diagnostics, Newcastle upon Tyne, United Kingdom) is a rapid immunochromatography test for the qualitative detection of Aspergillus diagnostic GM antigen. The results were reported as weak positive or strong positive in less than one hour.

BAL samples were processed according to the manufacturer’s recommendations [15]. Nevertheless, all BAL samples, both blood-free and bloody, were pretreated.

Calcofluor white staining

Direct examination of BAL fluids by fluorescence microscopy used calcofluor white stain (Becton Dickinson) for microscopic observation of filamentous fungi. Staining was positive when we observed septate hyaline hyphae branching at 45 degrees, suggesting Aspergillus species. Smear microscopy was performed using potassium hydroxide (KOH) and calcofluor white.

Culture on fungal media

BAL fluid was cultured on Sabouraud dextrose agar (SDA) with chloramphenicol (Becton Dickinson) prior to sample processing. For this, 10–20 ml of BAL was centrifuged for 10 min at 3000 rpm. 100 µl of sediment was inoculated onto fungal media. Fungal cultures were incubated at 30ºC for seven days. Aspergillus species were identified by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry and/or using colony characteristics and conidiophore and conidia morphology.

Quantitative GM detection

The Platelia Aspergillus galactomannan assay (Platelia, Bio Rad) is an enzyme immunoassay (EIA) that detects the GM antigen produced by Aspergillus species during active growth. This GM-EIA was performed twice a week on BAL fluid from these patients. Two different cut-off points were considered to determine positive results: GM index ≥ 0.5 ng/ml and ≥ 1.0 ng/ml.

Aspergillus-qPCR

In cases where there were discrepancies because only one of the previously described tests was positive, a qPCR assay (Aspergillus species ELITe MGB® Kit) was performed for the detection and quantification of the DNA of the Aspergillus genus: Aspergillus niger, Aspergillus nidulans, Aspergillus terreus, Aspergillus flavus, Aspergillus versicolor and Aspergillus glaucus. Only copy numbers > 150 of the genus Aspergillus (Ct < 36) were considered positive.

DefinitionsCAPA

The 2020 ECMM/ISHAM consensus criteria proposed three different grades: possible, probable, and proven [16]. Lung biopsies were not performed in this patient population, so there were no cases of proven CAPA in our cohort. All samples analyzed involved BAL fluid, as classifying patients as possible CAPA was not appropriate. Probable CAPA, according to the consensus criteria, depends on a single piece of mycological evidence. In this study, probable CAPA was defined on the basis of at least two positive diagnostic tests:

1) Detection of GM by AspLFD and/or EIA and BAL culture-positive for Aspergillus species or with calcofluor white stain suggestive of Aspergillus species.

2) Detection of GM by AspLFD and/or EIA and/or BAL culture-positive for Aspergillus species and positive Aspergillus-qPCR;

Negative case

Patients were defined as negative for CAPA when all the diagnostic tests returned negative results or only one test was positive.

Statistical analysis

For the baseline characteristics of the study population, data were reported as absolute counts (%) or as medians [25th-75th percentile] and p values were assessed by Mann Whitney U test, Pearson’s χ2 test or Fisher’s exact test.

The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR(+)) and negative likelihood ratio (LR(-)) of AspLFD, Platelia Aspergillus galactomannan assay, culture on SDA with chloramphenicol and calcofluor white staining were analyzed.

In a receiver operating characteristic (ROC) curve analysis for the Platelia Aspergillus galactomannan assay, the area under the curve (AUC), including 95% confidence intervals (CI), was examined to assess the ability of the diagnostic test to discriminate between patients with CAPA and critically ill patients without CAPA. The Youden index was applied to determine the optimal ROC curve cut-off value.

Observed percentage agreement and the kappa statistic were used to measure agreement between AspLFD and the Platelia Aspergillus galactomannan assay at a cut-off index of ≥ 0.5 and ≥ 1.0. The strength of agreement was defined according to Altman’s scale as: < 0.20 (poor); 0.20 to 0.40 (fair); 0.40 to 0.60 (moderate); 0.60 to 0.80 (good); and 0.80 to 1.00 (very good agreement).

Any P value less than 0.05 was interpreted as statistically significant.

The Statistical Package for Social Sciences Software (SPSS version 25.0; IBM SPSS statistics Inc., Chicago, IL, USA) was used for statistical analysis. The local research ethics committees approved this study.