Introduction

The outbreak of coronavirus 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has become a major threat to human health. Like SARS-CoV-1 and the Middle East respiratory syndrome coronavirus (MERS-CoV), SARS-CoV-2 is responsible for lower respiratory tract infection and can cause acute respiratory distress syndrome (ARDS) (Song et al. 2020).

Bacterial and fungal infections are common complications of viral pneumonia, especially in critically ill patients. These infections cause an increased need for intensive care and an increased mortality rate (Antinor et al. 2020). In patients with COVID-19, the predisposing features of the patients’ host environment allow coinfections to occur (Hoenigl et al. 2022).

Among the co-infectious agents in COVID-19 patients, Aspergillus species cause invasive pulmonary aspergillosis (IPA). IPA is difficult to diagnose and is associated with high morbidity and mortality (Lai and Yu 2021). IPA has been increasingly detected as an infectious agent in critically ill patients who are hospitalized in the intensive care unit (ICU) in recent years. The incidence of COVID-19-associated pulmonary aspergillosis (CAPA) differs significantly according to the diagnostic criteria and the study design used. While some studies screened report the incidence of CAPA as low as 5%, several studies using a screening protocol have found up to 34% of ICU patients to have CAPA (Er et al. 2022). In this context, further studies are needed to determine the accurate incidence, optimize diagnoses, improve patient management, provide up-to-date epidemiological data, and investigate the risk of IPA in critical patients diagnosed with COVID-19. The aim of this study is to identify Aspergillus spp. isolates from respiratory tract samples of COVID-19 patients hospitalized in ICUs in Niğde Ömer Halisdemir Training and Research Hospital, Niğde, Turkey, by different methods and to determine their antifungal susceptibility profiles.

Methods

Patient group

A total of 50 adult (≥ 18 years old) patients with ARDS and COVID-19 were hospitalized in COVID-19 ICUs between 13 March and 25 December 2020 and were included in this study. All demographic, clinical, and microbiological data were obtained from the clinical records of the patients. The classifications of IPA in the patients included in the study were done by the ECMM/ISHAM consensus criteria (Koehler et al. 2021).

Microbiological analysis.

SARS-CoV-2 RT-PCR

A viral nucleic acid isolation kit (Bioeksen, Turkey) was used for the isolation of SARS-CoV-2 from oro-nasopharyngeal swab and tracheal aspirate (TA) samples. In accordance with the recommendations of the manufacturer, a sample of 10 μL (final volume) was used. RT-PCR kit (Bioeksen, Turkey) targeting N and ORF1ab genes of SARS-CoV-2 was used. Amplification was performed on the Qiagen Rotor-Gene Q 5plex HRM instrument (Qiagen, Hilden, Germany).

Phenotypic identification

Blood agar (5%), eosin-methylene-blue agar (EMB), chocolate agar, and Sabouraud dextrose agar (SDA) media were used for the cultivation of TA and sputum samples of LRT of COVID-19 patients sent to the microbiology laboratory of our hospital. After bacteriological evaluation, the SDA medium was incubated at 25 °C for at least seven days to monitor mold growth. Growing mold colonies were passaged on potato dextrose agar (PDA) medium. For the differentiation of Aspergillus spp. that grow in the PDA medium, macroscopic and microscopic methods were used. Microscopic examination was done by looking at the texture of the colony (colony size and color, surface appearance, and pigment formation) and the color of the mycelia. The growing mold was evaluated microscopically with lactophenol cotton blue. The species-level distinction was made by looking at the number of sterigmata, vesicle structure, arrangement of phialides (single or double row), and location, as well as the structure and color of conidiophores under the microscope (McClenny 2005; Bilgi and Kiraz 2019).

Molecular identification

DNA isolation

Heliosis DNA isolation kit (Metis Biotechnology, Turkey) was used to obtain nucleic acid from the isolates in accordance with the test kit procedure. The total amount of DNA obtained was measured in a DNA measuring device (NanoDrop, USA), and the presence of DNA was checked. DNAs were stored at -20 °C until use.

Amplification of DNA samples by PCR

PCR experiment was performed using internal transcribed spacer 2 (ITS2) ribosomal DNA primers. The amplification reaction was performed using primer set fITS7 5′-GTGARTCATCGAATCTTTG-3′ and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′). The amplification reactions were carried out for 25.0 μL total volume, using 2.0 μL DNA template, 12.5 μL 2 × Master Mix, 0.5 μL of each primer (10 μM), and 9.5 μL ddH2O. The thermo-cycling conditions were adapted as an initial enzyme activation step at 95 °C for 10 min, followed by 35 cycles of denaturation at 95 °C for 30 s, annealing at 53 °C for 30 s (touchdown PCR, starting from 10th cycle with 0.2 °C decreases), and elongation at 72 °C for 30 s.

The amplification products as the single-band were validated by agarose gel electrophoresis run at 50 V for 15 min. PCR products were purified and sequenced by ABI 3730XL sequencer using the BigDye Terminator v3.1 Cycle Sequencing Kit. The species identification was done based on the NCBI n-blast search (https://blast.ncbi.nlm.nih.gov/Blast.cgi) as well as considering the colony morphologies. The phylogenetic tree was constructed using Qiagen CLC workbench program, including outgroup sequences obtained from the NCBI database.

Antifungal susceptibility profiles of isolates

Susceptibility testing of all isolates to antifungal agents was performed by liquid microdilution method according to the Clinical and Laboratory Standards Institute (CLSI) M38-A2 guidelines (CLSI 2008). Amphotericin B (AmB), itraconazole (ITR), voriconazole (VOR), and caspofungin (CSP) were used in the experiment. Serial dilutions were made between 0.03–16 μg/mL for AmB, ITR, and VOR and 0.06–32 μg/mL for CSP. Inoculated microplates were incubated at 35 °C for 48 h. The MIC ranges and MIC90 (MIC value inhibiting 90% of the isolates) of the tested Aspergillus isolates were also determined. The MIC values obtained were compared with the proposed CLSI epidemiological cutoff values (ECVs) to determine the susceptibility profiles of the isolates (Espinel-Ingroff and Turnidge 2016).

Detection of Aspergillus antigen by ELISA

In order to diagnose IPA infection early and accurately in COVID-19 patients, blood samples were collected from 35 patients with Aspergillus overgrowth, and serum galactomannan (GM) was investigated serologically in only one single serum sample that was taken from the patients. Platelia™ Aspergillus antigen ELISA kit (BioRad, France) was used for this. According to the manufacturer’s recommendation, a serum sample was considered positive when the optical density (OD) value was ≥ 0.5.

Statistical analysis

The analysis of the obtained data was performed using the SPSS 20.0 package program (IBM, Armonk, NY, USA), and the chi-square test was used.

Ethical approvement

The study protocol was approved by the Niğde Ömer Halisdemir University Faculty of Medicine Non-Interventional Clinical Research Ethics Committee (protocol number: 2021/32).

Results

Demographic findings of the patient group

Demographic characteristics of the patient groups included in the study are given in Table 1. Of the patients included in our study, 21 (42%) were females and 29 (58%) were males. The median hospital stay of the patients was 16 (IQR = 12–22) day (between 9 and 50 days), and their median age was 72.5 (IQR = 48.25–79) (age range: 37–87 years). Aspergillus spp. growth was detected in 35 (70%) of the clinical samples taken from the patients. Included patients were classified according to ECMM/ISHAM criteria (Koehler et al. 2021). Accordingly, in the patients included in the study, nine patients (18%) were diagnosed with possible IPA, 11 (22%) patients were diagnosed with probable IPA, and 15 (30%) patients were diagnosed with Aspergillus colonization. For the remaining 15 patients (30%), IPA was not diagnosed because they did not meet the criteria of ECMM/ISHAM. There was no patient diagnosed with proven IPA in the study.

Table 1 The demographic, clinical, and mycological characterization of COVID-19 patients
Full size table

Patients with confirmed IPA (20; 40%) had various types of underlying conditions. Among them are corticosteroid therapy, dyspnea, Aspergillus positive culture of LRT specimens, abnormal chest computed tomography (CT) scan (compatible with COVID-19), and worsening respiratory failure despite appropriate antibiotic therapy and respiratory support. Of the patients with Aspergillus positive culture of LRT samples, 29 (82.8%) had hypertension (HT), 20 (57.1%) chronic obstructive pulmonary disease (COPD), 19 (54.3%) diabetes mellitus (DM), two (5.7%) chronic kidney disease (CKD), three (5.6%) heart failure, three (5.6%) asthma, and one (2.8%) atrial fibrillation (Table 1).

A total of 30 (60%) patients included in the study died. Among them, 25 were IPA patients (seven possible, 10 probable, and eight with Aspergillus colonization).

Among the patients with Aspergillus growth, the presence of IPA was found to be significantly higher in patients with COPD (n = 17; 85%; p < 0.0001). No significant difference was found in patients with HT.

In 22 patients, corticosteroid treatment ≥ 40 mg/day prednisone equivalent was administered IV. All patients received broad-spectrum antibiotic therapy (such as teicoplanin, meropenem, piperacillin – tazobactam, and tigecycline). Antifungal therapy (liposomal AmB and VOR) was used as prophylaxis in three IPA patients (two possible and one probable IPA), and 30 (60%) of the patients were intubated and received mechanical ventilation support.

Phenotypic characterization

When TA and sputum samples of a total of 50 patients were evaluated with the traditional method, Aspergillus was recorded in only 35 patients. Among the Aspergillus spp., 20 (57.1%) A. fumigatus species complex, six (17.1%) A. flavus species complex, three (8.6%) A. niger species complex, three (8.6%) A. terreus species complex, and three (8.6%) Aspergillus spp. were identified from the clinical samples of 50 patients (34 (68%) were TA and 16 (32%) were sputum). Of the patient samples that were classified according to IPA identification algorithm, 20 (57.1%) were TA and 15 (42.8%) were sputum. The clinical specimens and Aspergillus species are shown in Table 2.

Table 2 The clinical sample type and isolated Aspergillus spp. in the patients
Full size table

Molecular identification

By sequencing the ribosomal DNA ITS region, which we used as a reference method in our study, 20 of the A. fumigatus samples, six A. flavus, four A. niger, three A. terreus, and two A. welwitschiae species were identified. Two phenotypically unidentified Aspergillus isolates were identified by DNA sequence as A. welwitschiae (Table 3 and Fig. 1).

Table 3 The phenotypic and genotypic characterization of Aspergillus isolates
Full size table
Fig. 1
figure 1

Phylogenetic tree of the Aspergillus isolates using the maximum likelihood method based on the combined sequences of the ITS region

Full size image

Antifungal susceptibility testing

The MIC90 and MIC ranges of AmB, ITR, VOR, and CSP tested against 35 Aspergillus isolates are presented in Table 4. Overall, the lowest MIC values (0.03 μg/mL) for all 35 isolates tested in the study were seen in the ITR, followed by AMB and CSP (0.06 μg/mL) and VOR (0.125 μg/L). As shown in Table 4, the tested antifungal agents showed good activity (MICs ≤ ECV) against A. fumigatus isolate. While two A. flavus isolates were susceptible to all antifungal agents (MIC90 range of 0.125– 4 μg/mL; MIC90 ≤ ECV), four isolates were resistant to AMB with MICs of > 4 μg/mL according to the proposed ECVs. For A. niger isolates, AMB, ITR, VOR, and CSP antifungals showed good activity with MIC values lower than the proposed ECV values. Two A. terreus isolates were susceptible to ITR, VOR, and CSP (MIC90 ≤ ECV). On the other hand, one isolate was resistant to ITR, and two isolates were resistant to AmB (MIK90 16 > ECV). Two A. welwitschiae isolates were found to be sensitive to all tested agents with 1– 0.125 μg/mL MIC90 values.

Table 4 Antifungal susceptibility profiles of Aspergillus species
Full size table

Serum GM test

In the study, serum GM antigen positivity was found in 11 of the patients, and these patients were considered as probable IPA. The group of patients with Aspergillus colonization was negative for serum GM. The non-IPA patient group was not included in the serum GM test because they did not meet the evaluation criteria of ECMM/ISHAM (Koehler et al. 2021).

Discussion

Serious viral pulmonary infections such as COVID-19 are associated with an increased risk of superinfection, including IPA, especially in those with hematological malignancies and in immunocompromised patients. Emerging fungal infection such as aspergillosis is profoundly identified in critically ill patients (Chong and Neu 2021; Al-Tawfiq et al. 2021).

CAPA is defined as IPA in most of the critically ill COVID-19 patients. Consequently, a question arises about the burden of IPA among these patients (Gouzien et al. 2021). Risk factors identified in IPA patients associated with COVID-19 include advanced age, lymphopenia, chronic respiratory diseases, corticosteroid therapy, antimicrobial therapy, mechanical ventilator, or cytokine storm (Lai and Yu 2021; Hoenigl et al. 2022). In a study, secondary infections developed in 65.96% of patients with COVID-19 infection (De Bruyn et al. 2022). In another study, 20 (13.3%) of 150 patients diagnosed with COVID-19 had secondary infections. The rate of secondary infection in ICU patients (72%) was found to be significantly higher than in patients in general service (1.6%) (Arıcı et al. 2022).

In a study conducted at San Salvatore Hospital in Pesaro, Italy, out of a total of 89 patients with COVID-19 in the ICU, 68 (76.4%) developed a secondary infection. Bacteria constituted most of the isolates (94.6%) in the study, followed by fungi with 5.4% (Caiazzo et al. 2022). In a multi-center study conducted in France, IPA was diagnosed in 129 (25.1%) of 366 COVID-19 patients admitted to the ICU (Dellière et al. 2020).

The increased incidence of IPA in people with severe respiratory virus infection has raised concerns that it may also occur in patients with acute respiratory failure due to COVID-19 infection. In particular, this infection causes an inflammatory environment that allows pulmonary damage and fungal infection (White et al. 2021).

In parallel with the increase in the number of immunocompromised patients, it has been reported that the incidence of IA increases in patients hospitalized in intensive care, transplant, and burn units (Weber et al. 2009). Although Aspergillus species are less seen in intensive care patients, they are increasingly detected as infectious agents. The presence of COPD, steroid use, and multiple organ failure in non-neutropenic ICU patients facilitate the development of IPA (Sánchez Martín et al. 2022).

In addition, negative pressure application in the COVID-19 ICU may be the source of air pollution by Aspergillus spp., which increases the risk of opportunistic infection (Ichai et al. 2020). Despite a large number of case reports and extensive studies, Aspergillus spp. can cause devastating inflammatory and invasive pathology in individuals with severe influenza, and culture results are mistakenly reported as respiratory tract colonization by many clinicians. It is always difficult to distinguish between respiratory tract colonization and potential disease caused by Aspergillus spp. The definition of IPA is still difficult, especially in patients with severe COVID-19 infection in the ICU. Therefore, the focus has been on the development of new algorithms based on symptoms compatible with Aspergillus spp. cultures, host factors, and abnormal imaging to improve the process of obtaining information about the disease agent and prognosis (Marr et al. 2021; White et al. 2021; Castro-Fuentes et al. 2022). Current studies particularly highlight the need to monitor COVID-19 patients who develop ARDS and remain in intensive care for IPA (Chiumello et al. 2022).

In this study, we aimed to collect data on the incidence of IPA, risk factors, identification, and susceptibility profile of Aspergillus agents in patients hospitalized in the COVID-19 ICUs of our hospital who developed ARDS. Clinical, radiological, and mycological criteria in ECMM/ISHAM logarithm were used for IPA definitions. EORTC definitions were not included in our study due to the absence of immunocompromised individuals (such as acute myeloid leukemia, solid organ transplantation, and neutropenia) in our hospital among ICU patients.

In our study, 11 patient were considered probable IPA according to the results of Aspergillus culture positivity, serum GM, and abnormal CT scan of the lungs in addition to the lesions attributed to typical COVID-19. The incidence of possible and probable IPA in the whole population of the study was determined to be 18% and 22%, respectively. These results were found to be compatible with similar studies (Salas et al. 2022).

Our results show that in patients having SARS-CoV-2-associated pneumonia and in patients without immunosuppression are more susceptible to increased risk of IPA (40% overall). Due to the difficulty of obtaining samples for histopathological confirmation in these critically ill patients, a proven diagnosis of IPA could not be made.

The serum GM test was negative in 30% of the patients, and mycological evidence was not detected as Aspergillus species did not grow in the respiratory samples. The isolation of Aspergillus species from respiratory specimens in the absence of pneumonia symptoms in 30% of our cases was thought to represent colonization. Based on these results, we cannot rule out that damaged pulmonary epithelium is an indicator of colonization with Aspergillus hyphae prior to the development of active IPA. Some publications indicate that colonization is an important risk factor for the development of IPA (Arastehfar et al. 2020).

Isolation of Aspergillus species from respiratory specimens in critically ill patients is significantly associated with both the diagnosis of the underlying disease (such as COPD) and corticosteroid therapy (Townsend and Martin-Loeches 2022). In our study, results were obtained supporting that the presence of COPD comorbidity in patients with Aspergillus growth in the LRT sample would be a risk factor for the development of IPA.

In the current study, Aspergillus positive cultures were obtained in 35 of 50 respiratory samples. The most frequently isolated species was A. fumigatus, followed by A. flavus, A. niger, A. terreus, and A. welwitschiae (Table 2). In our study, when the results of the conventional and molecular methods were compared, it was found that although 91.4% agreement was observed, the conventional method could not identify three isolates (8.6%). The susceptibility of Aspergillus isolates identified was evaluated in vitro against AmB, ITR, VOR, and CSP. Overall, the tested antifungal agents showed good activity against the isolates. Limited literature available on CAPA, which suggests that serum GM will not be the best marker to distinguish between IA and colonization, and that perhaps serum GM testing from a bronchoalveolar lavage (BAL) sample should be performed (Verweij et al. 2020). Moreover, it is already recommended for diagnosing CAPA (White et al. 2021). However, bronchoscopy for the collection of BAL samples from COVID-19 patients poses a significant risk to healthcare workers as it produces aerosols (Nasir et al. 2020). Therefore, this procedure was avoided in our study, and a serum sample was preferred instead. In our study, serum GM positivity was detected in a total of 11 patients diagnosed with IPA.

The reason for the high mortality associated with IA in critical non-neutropenic COVID-19 patients has also been attributed to difficulties in timely diagnosis due to non-specific clinical manifestations and lack of definitive diagnostic criteria (Palacios and Moffarah 2021). In our study, 25 (71.4%) of 35 patients evaluated according to IPA algorithms died.

Obtaining mycological evidence of airway invasive aspergillosis in patients with COVID-19 is complicated by the reduced use of diagnostic bronchoscopy necessary to protect healthcare workers from aerosol exposure and the low sensitivity of circulating GM detection in serum. Also, the detection of Aspergillus in upper respiratory tract specimens such as sputum or TA generally does not distinguish between aspergillus colonization and invasive disease. TA and sputum samples are usually positive in critically ill COVID-19 patients but may represent upper airway colonization (Caggiano et al. 2022; Rouzé et al. 2022).

Conclusions

The development of new diagnostic tests, assessment of host immune responses, mycological screening (biomarkers and mycological diagnosis) of patients infected with COVID-19, and regular air quality checks of the ICU may perhaps lead to faster diagnosis and immediate initiation of antifungal therapy to manage the poor prognosis of IPA and reduce the risk of mortality. Considering all these may lead to the development of new prevention strategies for secondary infections.