key: cord-0824284-2q38mgjx
authors: Yusuf, Erlangga; Vonk, Alieke; van den Akker, Johannes P. C.; Bode, Lonneke; Sips, Gregorius J.; Rijnders, Bart J. A.; de Steenwinkel, Jurriaan; Verkaik, Nelianne J.; Vogel, Marius; van der Eerden, Menno; van Westreenen, Mireille
title: Frequency of Positive Aspergillus Tests in COVID-19 Patients in Comparison to Other Patients with Pulmonary Infections Admitted to the Intensive Care Unit
date: 2021-02-18
journal: J Clin Microbiol
DOI: 10.1128/jcm.02278-20
sha: a37868130585c52c82b4e8cdde8c9d67c94a58aa
doc_id: 824284
cord_uid: 2q38mgjx

The aim of this study was to describe the frequency of positive Aspergillus tests in COVID-19 patients and investigate the association between COVID-19 and a positive Aspergillus test result. We compared the proportion of positive Aspergillus tests in COVID-19 patients admitted to the intensive care unit (ICU) for >24 h with two control groups: patients with community-acquired pneumonia with (i) a PCR-confirmed influenza infection (considered a positive control since the link between influenza and invasive aspergillosis has been established) and (ii) Streptococcus pneumoniae pneumonia (in whom positive Aspergillus tests are mostly considered as colonization). During the study period, 92 COVID-19 patients (mean [standard deviation] age, 62 [14] years; 76.1% males), 48 influenza patients (55 [14]; 56.2% males), and 65 pneumococcal pneumonia patients (58 [15], 63,1% males) were identified. Any positive Aspergillus test from any respiratory sample was found in 10.9% of the COVID-19 patients, 6.2% of the patients with pneumococcal pneumonia, and 22.9% of those infected with influenza. A positive culture or PCR or galactomannan test on bronchoalveolar lavage (BAL) fluid only was found in 5.4% of COVID-19 patients, which was lower than in patients with influenza (18.8%) and comparable to that in the pneumococcal pneumonia group (4.6%). Using logistic regression analysis, the odds ratio (OR) (95% confidence interval) for a positive Aspergillus test on BAL fluid for COVID-19 patients was 1.2 (0.3 to 5.1; P = 0.8) compared to the pneumococcal pneumonia group, while it was 0.2 (0.1 to 0.8; P = 0.02) compared to the influenza group. This difference remained significant when corrected for age and sex. In conclusion, in COVID-19 patients, the prevalence of a positive Aspergillus test was comparable to that in patients admitted for pneumococcal pneumonia but substantially lower than what we observed in patients with influenza.

I nvasive pulmonary aspergillosis (IPA) is a life-threatening disease that typically occurs in severely immunocompromised patients (1) . The mortality of IPA in patients with hematological disease is estimated to be 30% (2) but is substantially higher in critically ill patients (3) . More recently, intensive care unit (ICU) admission for severe influenza has been shown to be a risk factor for IPA, with an incidence varying between 7 and 18% and overall mortality around 50% (4) (5) (6) .

Like influenza, COVID-19 is predominantly a pulmonary disease, and COVID-19 has been linked with Aspergillus detection in respiratory samples by several authors, who have coined the term COVID-19-associated pulmonary aspergillosis (7, 8) . Assessing clinical significance of a positive Aspergillus test is difficult in general (9) and is even more difficult in ICU patients. Clinically, it is impossible to differentiate pulmonary aspergillosis from COVID-19 based on clinical signs and symptoms, especially in patients admitted to the ICU. All of them have pulmonary infiltrates. Moreover, chest computed tomography (CT) scans of COVID-19 patients that show bilateral widespread ill-defined and ground-glass opacification (10) may also obscure any eventual findings of IPA.

Studies showing prevalent positive Aspergillus tests in COVID-19 so far are mostly case reports or observational studies without control groups (8, (11) (12) (13) (14) (15) (16) (17) (18) . While small case series are prone to publication bias, the lack of any control group precludes a valid conclusion (19) . The link between Aspergillus and COVID-19 might, for example, also be explained by the high frequency of bronchoalveolar lavage (BAL) performed in COVID-19 patients admitted to the ICUs to exclude bacterial superinfection. The more diagnostic procedures are performed, the higher the chances of detecting microorganisms that may be colonizers rather than true pathogens, which also applies to the ubiquitous mold Aspergillus. Detection of Aspergillus in respiratory samples in nonimmunocompromised patients is often considered to indicate colonization and does not require antifungal therapy (20) . By considering any positive Aspergillus test in COVID-19 patients clinically significant, antifungal treatment will be initiated in these patients, leading to higher costs, possible adverse events, and toxicity (21) .

Based on these observations, we hypothesized that the number of positive Aspergillus tests in COVID-19 patients would not differ from that in a control group but would be lower than in a group of ICU patients that is now broadly considered to be at increased risk for IPA.

Therefore, the aim of this study was to describe the frequency of positive Aspergillus tests in COVID-19 patients and to investigate the association between COVID-19 status and positive Aspergillus tests in a case-control study using two different control groups, one with an established link with IPA (influenza) and another without (pneumococcal pneumonia).

Study design. This was a case-control study performed in Erasmus University Medical Centre (Erasmus MC), Rotterdam, The Netherlands. Erasmus MC is a large tertiary hospital. It hosted the national coordination center for COVID-19 patient distribution during the first wave of the COVID-19 outbreak in The Netherlands (between March and June 2020). During the first wave, it had 102 ICU beds allocated for COVID-19 (7% of all Dutch ICU beds for COVID -19) .

Included were adult patients (.18 years old) admitted to the ICU for .24 h. COVID-19 cases were those with confirmed COVID-19 based on a positive PCR from respiratory samples between 1 March and 21 April 2020. We used two control groups selected in the period between January 2010 and April 2020 admitted to the same ICU: (i) patients with community-acquired pneumonia due to influenza confirmed by a positive PCR on respiratory samples and (ii) patients with community-acquired pneumococcal pneumonia based on positive respiratory culture with Streptococcus pneumoniae or positive urine antigen test and a negative influenza test. All included patients had infiltrates on the radiograph or CT scan of the chest.

Demographic (age and gender) and microbiology data were obtained from hospital and laboratory information system. This study was a noninterventional observational study using only limited demographic data and was performed under institutional review board approval (METC-2015-306).

Test algorithm. In COVID-19 cases and controls, BAL or other respiratory samples were collected within the first week of ICU admission. BAL fluid was collected when patients showed clinical deterioration leading to differential diagnosis of secondary infection based on clinicians' judgment. At Erasmus MC, it is the standard procedure to do a fungal culture on all BAL samples from ICU patients. Aspergillus antigen test (galactomannan) and Aspergillus DNA detection by PCR are occasionally performed as well. Due to the invasive nature of BAL, and because BAL fluid is used to make a diagnosis and not for followup of the pulmonary infections, BAL was rarely done more than once.

Microbiological tests. Galactomannan tests were performed on the BAL fluid and other types of respiratory samples (when BAL was not performed) or serum of the patients using an immunoenzymatic sandwich microplate assay (Platelia Aspergillus Ag; Bio-Rad Laboratories B.V., Veenendaal, The Netherlands). A galactomannan index of .0.6 was considered positive for serum as well as for BAL fluid. The performance of this test depends on the study population, and studies in nonneutropenic patients reported sensitivities of 22% in serum (22) and 76% in BAL fluid to diagnose invasive pulmonary aspergillosis (IPA) at a cutoff galactomannan index of $0.5 (23) . PCR for Aspergillus was performed using the commercial kit AsperGenius (PathoNostics, Maastricht, The Netherlands), which detects and identifies Aspergillus fumigatus, Aspergillus terreus, Aspergillus spp., and azole resistance markers TR34 and TR46, or in-house real-time PCR assay as described before (24) . The sensitivity of the in-house real-time PCR was 80% in nonneutropenic patients in diagnosing IPA (24) . Aspergillus cultures were performed on BBL Sabouraud dextrose agar with chloramphenicol (BD Diagnostics, Erembodegem, Belgium) and incubated at 26°C and 35°C for 21 days. The sensitivity of culture in detecting invasive pulmonary aspergillosis is between 20% and 50% (25, 26) . These tests were performed in an accredited ISO 15189 microbiology lab. PCR for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) targeting E and RNA-dependent RNA polymerase (RdRp) was performed in the Department of Viroscience of Erasmus MC, which is one of the WHO referral laboratories.

Statistical analysis. Data were analyzed using SPSS Statistics 26 (SPSS Inc., Chicago, IL). Continuous variables are presented as means with standard deviations (SD). Means among group were compared using analysis of variance (ANOVA). Number and proportion of gender and positive tests were calculated and chi-squared test was used to compared proportions between groups.

We determined several possibilities as proxies in defining clinical significance of pulmonary aspergillosis based on microbiology results (culture, PCR, and Aspergillus antigen in respiratory samples), from low probability to higher probability: (i) any positive Aspergillus tests performed in any type of respiratory or serum samples, (ii) any positive Aspergillus tests from BAL samples only, (iii) positive PCR and positive Aspergillus antigen in BAL samples, (iv) positive culture and positive Aspergillus antigen in BAL samples, (v) positive culture and positive PCR from BAL samples, and (vi) positive PCR, positive Aspergillus antigen, and positive culture.

Further, we used logistic regression analysis to calculate the odds ratio (OR) with 95% confidence interval (CI) to determine whether COVID-19 pneumonia was independently associated with positive Aspergillus tests. We performed two analyses: first, by considering positive Aspergillus tests in all type of respiratory samples (BAL fluid, sputum aspirates, and lung tissue), and second, by considering positive Aspergillus tests in BAL fluid only. We adjusted both associations for age and sex.

Demographic characteristics and microbiological tests. During the study period, 92 patients were admitted to the ICU with COVID-19, 48 with influenza, and 65 with pneumococcal pneumonia (Table 1) . COVID-19 patients tended to be older (mean age [SD], 62 [14] years) than patients with influenza or pneumococcal pneumonia (P = 0.1) and were more often male (P = 0.04).

Respiratory samples of approximately one-third of the COVID-19 patients were obtained. BAL sampling was more often performed in COVID-19 patients (29.3%) than in patients with pneumococcal pneumonia (18.5%) but less frequently than in patients with influenza (45.8%). The numbers of performed Aspergillus PCRs and Aspergillus antigen tests were significantly different between the COVID-19 and the control groups, but no statistical difference was found between the groups for numbers of performed fungal cultures.

Positive Aspergillus tests. Any positive Aspergillus test from any respiratory samples was found in 10.9% of the COVID-19 patients, and this proportion was higher than in patients with pneumococcal pneumonia (6.2%) but lower than in the influenza patients (22.9%; P = 0.02) ( Table 2 ). When only the Aspergillus tests performed on BAL fluid were considered, the proportion of positive tests in COVID-19 patients was reduced to 5.4% and comparable to that of the pneumococcal pneumonia group (4.6%) but much lower than what was observed in patients with influenza (18.8%; P = 0.01). Stricter criteria for positivity (e.g., positive PCR and positive Aspergillus antigen in BAL samples or positive culture and positive Aspergillus antigen in BAL samples) delivered a limited number of cases to allow statistical comparisons.

Association between detection of Aspergillus and COVID-19. The OR (95% CI) of having any positive Aspergillus test in any respiratory sample for COVID-19 patients compared with pneumococcal pneumonia patients was 1.9 (0.6 to 6.2; P = 0.3). Taking age and gender into account did not change the estimate. Compared this group with the influenza group, the OR (95% CI) was 0.4 (0.2 to 1.1; P = 0.06) without and 0.4 (0.1 to 0.9, P = 0.04) when age and sex were taken into account.

When only a positive Aspergillus test from BAL fluid was taken into account, the OR was 1.2 (0.3 to 5.1, P = 0.8) for the comparison of COVID-19 patients with pneumococcal pneumonia, while it was 0.2 (0.1 to 0.8; P = 0.02) compared with the influenza group despite the fact that a fungal culture had been performed less often on BAL samples from patients with influenza than from COVID-19 patients. This difference remained significant when corrected for age and sex (OR of 0.2 (0.01 to 0.7; P = 0.01).

In this study, which included 92 patients admitted to the ICU with COVID-19, the proportion of patients in which Aspergillus was detected was comparable to that of patients admitted with pneumococcal pneumonia while significantly lower than in patients with influenza, a patient population in which the link with IPA is well established (4).

In critically ill patients admitted for respiratory failure, such as COVID-19 patients, the suspicion of a hospital-acquired superinfection typically arises when a clinical deterioration is observed. In COVID-19 patients, clinicians are often reluctant to perform BAL (27) and rely on upper respiratory tract specimens to diagnose ventilator-associated pneumonia. Yet the detection of Aspergillus in the upper respiratory tract more often represents colonization than infection (27) . The majority of publications on Aspergillus in COVID-19 are case series (11) (12) (13) (14) (15) (16) 18) and often included the detection of Aspergillus in the upper airway (6) or included the detection of b-D-glucan as a mycological criterion, a test that is not specific for Aspergillus spp. and it has never been shown to be useful to diagnose IPA in ICU patients (9) . Bartoletti and coworkers did perform a study with systematic sampling of BAL fluid in COVID-19 patients admitted to the ICU. In their study, 27.7% of the patients had a serum galactomannan index of .0.5, BAL galactomannan index of .1.0, growth of Aspergillus spp. in BAL fluid, or a cavitating infiltrate on CT scan of the chest (17) . We agree that the 27.7% incidence is surprisingly high, and it is higher than the number in our study. It is very likely that the high incidence is at least partially explained by the very invasive BAL sampling protocol in this study. Indeed, a BAL was performed on admission, on day 7, and at the time of clinical deterioration. Our observations illustrate that a positive test for Aspergillus in COVID-19 should not automatically lead to the initiation of antifungal therapy, a treatment that does not come without risk (21) . Exactly in the ICU patient population, diagnosing IPA is most challenging. Indeed, the typical radiological findings are often absent, and in ICU patients, the specificity of any mycological test is lower than it is in well-defined patient population like those with longstanding neutropenia. In this setting, nonspecific tests such as panfungal b-D-glucan should not be used.

Due to nonspecific clinical symptoms and radiological findings, we considered a comparison of objective microbiological test results in COVID-19 patients with 2 other ICU populations with different respiratory infections a reasonable first comparator in order to estimate the clinical significance of the different Aspergillus tests available. We chose pneumococcal pneumonia as a control group since a comparable group in which invasive respiratory samples were obtained was needed. Patients admitted to the ICU with influenza were chosen as a patient group because ICU admission for respiratory insufficiency due to influenza was repeatedly shown to increase the risk for IPA (4) . Healthy patients as a control group would be ideal, but this was not feasible.

The most reliable data will come from studies in which pre-or postmortem lung biopsies or full autopsies are performed and correlated with diagnostic test results that preceded the findings on biopsy or autopsy. Few published studies of this kind are available. In one such study, postmortem lung biopsies could not conform the premortem IPA diagnosis (28) , and autopsies studies did not report the presence of possible Aspergillus (29) (30) (31) .

In this study, we also observed several clinical practices in diagnosing and treating COVID-19 patients, such as the increased tendency to perform BAL in COVID-19 patients in comparison to pneumococcal pneumonia patients, despite initial fear of SARS-CoV-2 infection of medical personnel during the act of bronchoscopy (27) . For the majority of the BAL fluid samples, an Aspergillus test was requested and performed. Furthermore, we noticed that a positive Aspergillus /	est in our setting was common when samples were obtained from the upper respiratory tract, which suggests isolation of Aspergillus that may represent colonization.

The strength of the present study is the use of a control group and the fact that all patients originated from the same hospital, which should minimize selection bias. It can be assumed, for example, that the reasons to perform BAL are comparable among the groups. Yet we cannot control for all possible confounders and there is still some bias, since cases and controls needed to be selected from different calendar years. Another limitation is that only one-third of the COVID-19 patients' respiratory samples were tested for the presence of Aspergillus.

In conclusion, the frequency of positive Aspergillus tests from BAL fluid of COVID-19 patients was comparable to that of pneumococcal pneumonia patients. Since in this control group positive Aspergillus tests are often considered to indicate colonization, a positive test for Aspergillus in COVID-19 should be interpreted with caution and not automatically lead to the start of an antifungal, especially considering its possible side effects. To get a more reliable estimate of the incidence of IPA in patients with COVID-19, studies are needed in which systematic (minimally invasive) autopsies or directed postmortem lung biopsies are performed. There are no conflicts of interest to report. There are no financial disclosures to report.

Diagnosis and management of Aspergillus diseases: executive summary of the 2017 ESCMID-ECMM-ERS guideline

Aspergillosis case-fatality rate: systematic review of the literature

Epidemiology of invasive aspergillosis in critically ill patients: clinical presentation, underlying conditions, and outcomes

Invasive aspergillosis in patients admitted to the intensive care unit with severe influenza: a retrospective cohort study

High rates of influenza-associated invasive pulmonary aspergillosis may not be universal: a retrospective cohort study from Alberta

Invasive pulmonary aspergillosis complicating severe influenza: epidemiology, diagnosis and treatment

ESCMID Study Group for Infections in Critically Ill Patients. 2020. Diagnosing COVID-19-associated pulmonary aspergillosis

Shimabukuro-Vornhagen A. 2020. COVID-19 associated pulmonary aspergillosis

The impact of culture isolation of Aspergillus species: a hospitalbased survey of aspergillosis

Radiological findings from 81 patients with COVID-19 pneumonia in Wuhan, China: a descriptive study

COVID-19-associated pulmonary aspergillosis

Invasive aspergillosis as an under-recognized superinfection in COVID-19

Fungal co-infections associated with global COVID-19 pandemic: a clinical and diagnostic perspective from China

Invasive pulmonary aspergillosis complicating COVID-19 in the ICU-a case report

COVID-19-associated pulmonary aspergillosis (CAPA) in patients admitted with severe COVID-19 pneumonia: an observational study from Pakistan

Epidemiology of invasive pulmonary aspergillosis among intubated patients with COVID-19: a prospective study

Prevalence of putative invasive pulmonary aspergillosis in critically ill patients with COVID-19

COVID-19, equipoise and observational studies: a reminder of forgotten issues

Repeated Aspergillus isolation in respiratory samples from non-immunocompromised patients not selected based on clinical diagnoses: colonisation or infection?

Adverse effects of voriconazole: over a decade of use

Diagnosis of invasive aspergillosis using a galactomannan assay: a meta-analysis

Diagnostic value of galactomannan antigen test in serum and bronchoalveolar lavage fluid samples from patients with nonneutropenic invasive pulmonary aspergillosis

Validation of a new Aspergillus real-time PCR assay for direct detection of Aspergillus and azole resistance of Aspergillus fumigatus on bronchoalveolar lavage fluid

Combination antifungal therapy for invasive aspergillosis a randomized trial

Isavuconazole versus voriconazole for primary treatment of invasive mould disease caused by Aspergillus and other filamentous fungi (SECURE): a phase 3, randomised-controlled, non-inferiority trial

The challenge of managing COVID-19 associated pulmonary aspergillosis

Late histopathologic characteristics of critically ill COVID-19 patients: different phenotypes without evidence of invasive aspergillosis, a case series

Autopsy fndings in 32 patients with COVID-19: a single-institution experience

Unspecific post-mortem findings despite multiorgan viral spread in COVID-19 patients

Pulmonary post-mortem findings in a series of COVID-19 cases from northern Italy: a two-centre descriptive study