COVID-19-Associated Pulmonary Aspergillosis (CAPA)

Abstract

Invasive Pulmonary Aspergillosis (IPA) has been recognized as a possible secondary infection complicating Coronavirus disease 2019 (COVID-19) and increasing mortality. The aim of this review was to report and summarize the available data in the literature concerning the incidence, pathophysiology, diagnosis, and treatment of COVID-19-Associated Pulmonary Aspergillosis (CAPA). Currently, the incidence of CAPA is unclear due to different definitions and diagnostic criteria used among the studies. It was estimated that approximately 8.6% (206/2383) of mechanically ventilated patients were diagnosed with either proven, probable, or putative CAPA. Classical host factors of invasive aspergillosis are rarely recognized in patients with CAPA, who are mainly immuno-competent presenting with comorbidities, while the role of steroids warrants further investigation. Direct epithelial injury and diffuse pulmonary micro thrombi in combination with immune dysregulation, hyper inflammatory response, and immunosuppressive treatment may be implicated. Discrimination between two forms of CAPA (e.g., tracheobronchial and parenchymal) is required, whereas radiological signs of aspergillosis are not typically evident in patients with severe COVID-19 pneumonia. In previous studies, the European Organization for Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG) criteria, a clinical algorithm to diagnose Invasive Pulmonary Aspergillosis in intensive care unit patients (AspICU algorithm), and influenza-associated pulmonary aspergillosis (IAPA) criteria were used for the diagnosis of proven/probable and putative CAPA, as well as the differentiation from colonization, which can be challenging. Aspergillus fumigatus is the most commonly isolated pathogen in respiratory cultures. Bronchoalveolar lavage (BAL) and serum galactomannan (GM), β-d-glucan (with limited specificity), polymerase chain reaction (PCR), and Aspergillus-specific lateral-flow device test can be included in the diagnostic work-up; however, these approaches are characterized by low sensitivity. Early treatment of CAPA is necessary, and 71.4% (135/189) of patients received antifungal therapy, mainly with voriconazole, isavuconazole, and liposomal amphotericin B . Given the high mortality rate among patients with Aspergillus infection, the administration of prophylactic treatment is debated. In conclusion, different diagnostic strategies are necessary to differentiate colonization from bronchial or parenchymal infection in intubated COVID-19 patients with Aspergillus spp. in their respiratory specimens vs. those not infected with severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2). Following confirmation, voriconazole or isavuconazole should be used for the treatment of CAPA.

Introduction

Typically, patients with Coronavirus disease 2019 (COVID-19) are immunocompetent, of advanced age, and with comorbidities (mainly hypertension, diabetes, chronic heart, and renal disease) [1], [2], [3]. Approximately 5% of those will develop a severe form of the disease with respiratory failure, complex immune dysregulation, and cytokine storm, requiring hospitalization in the intensive care unit (ICU) [4]. Different studies showed that these patients are at increased risk of secondary infections [5].

The syndromes of pulmonary aspergillosis complicating severe viral infections are distinct from classic invasive pulmonary aspergillosis (IPA). IPA, particularly that associated with hematologic malignancies and transplantation, is most frequently encountered in patients with neutropenia and other immuno compromised individuals. Numerous studies have recognized influenza-associated pulmonary aspergillosis (IAPA) associated with respiratory epithelium damage. Of note, local anosoparalysis may render patients with influenza more susceptible to IPA [6,7]. Since the first evidence of secondary aspergillosis reported in China, several studies have shown that steroid and other immune-modulatory therapies are linked to an increased risk of a similar syndrome associated with severe COVID-19, termed COVID-19-associated pulmonary aspergillosis (CAPA). Actually, the first reports referred to post-mortem results raising concerns regarding this infection as an additional factor contributing to patient mortality [7], [8], [9], [10], [11], [12]. Three different grades (e.g., possible, probable, and proven CAPA) have been suggested by an international panel of experts [13], enabling investigators to stratify patients in research registries and clinical trials.

This review focused on key controversies in CAPA due to its contribution to mortality among patients with COVID-19. An analysis of the available literature (reported until January 2021; search list presented in Appendix) was performed to identify differences in the incidence, pathophysiology, diagnosis, and treatment of CAPA and IPA.

Incidence and Risk Factors

Owing to differences in diagnostic criteria, methods, definitions, and local practices, the incidence of CAPA varies. Therefore, the estimation of CAPA incidence is challenging due to the lack of a gold standard and limitations in diagnostic tests. For this reason, definitions used for IAPA were applied in most studies; however, this approach generated a wide degree of variability in the incidence of CAPA among ICU patients (range: 3.8–34%) [[7], [8], [9], [10],[14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]]. In general, the diagnosis of CAPA is delayed vs. that of IAPA, and the incidence of the angioinvasive parenchymal form is lower. Table 1 shows all published studies describing the incidence of CAPA and associated comorbidities until January 2021. The majority of those studies suggested that CAPA mostly occurs in severely ill, mechanically ventilated patients with COVID-19. Among them, three prospective studies reported an incidence ranging 14–20%, while more recently published studies indicated a lower incidence ranging 3–15% (a total of 2407 patients were included) [8,11,26,28,[33], [34], [35], [36], [37], [38], [39], [40], [41]]. Excluding seven case reports, the diagnosis of CAPA has been confirmed in 223 of 2400 ICU patients (incidence of 9.2%). Nevertheless, this percentage may be slightly lower since two studies did not report the total number of patients admitted in the ICU during the investigation period. Thus, by deducting 17 patients from those two studies, the total number of ICU patients with COVID-19 and a diagnosis of proven/probable/putative CAPA is 206 (incidence: 8.6%). Interestingly, among eight studies involving >100 patients each, the incidence of CAPA ranged from 3.3% to 27.7%, though six of them reported an incidence <10%. This is in accordance with evidence from Chinese studies [42,43]. Notably, the majority of CAPA cases were documented in European countries (26/32 studies); one and two studies were performed in the USA and China, respectively.

Table 1.

Incidence and risk factors for CAPA (31 studies identified in the literature).

Study	Country	Environment, n (%)	Comments	Comorbitities
Bartoletti et al. [8]	Italy	ICU, 108 (27.27)	Prospective study	Obesity, AH, DM, Coronary disease, COPD, CReF, Hemodialysis, Cerebrovascular disease, Malignancies, Solid-organ transplant, Chronic steroid treatment
Alanio et al. [9]	France	ICU, 27 (33.3)	Putative IPA mainly	AH, Obesity, DM, BA, Cardiac disease
Rutsaert et al. [10]	Belgium	ICUs, 20 (35)	None	AH, DM, Hypercholesterinemia, CReF, Obesity, AML, HIV
Chen et al. [14]	China	ICU, 48 (27.1)	Mixed fungal infections High IL-6 levels	DM, AH, Heart disease
Lescure et al. [15]	France	5 patients	None	Gout, AH, Thyroid cancer
Koehler et al. [16]	Germany	ICU, 19 (26.3)	None	AH, COPD, DM, Obesity OSA
van Arkel et al. [17]	Netherlands	ICU, 31 (19.4)	3 probable/3 possible CAPA	Cardiomyopathy, COPD, BA
Blaize et al. [18]	France	ICU Case report	Putative aspergillosis	MDS, AH, Hashimoto disease
Prattes et al. [19]	Austria	ICU Case report	69-year-old patient	COPD, OSA, Obesity, DM, AH, Cardiac disease
Antinori et al. [20]	Italy	ICU Case report	None	DM, AH, Atrial fibrillation, obesity
Wang et al. [21]	China	Hospital, 104 (8.5)	Mainly elderly patients	DM, AH, Cardiac disease, COPD, CReF
Meijer et al. [22]	Netherland	ICU Case report	None	Reflux, polyarthrosis
Mohamed et al. [23]	Irelend	ICU Case report	Triazole-resistance	DM, AH, Hyperlipidaemia, Obesity
Ghelfenstein-Ferreira et al. [24]	France	ICU Case report	Triazole-resistance	DM, AH, Hyperlipidemia, Obesity
Santana et al. [25]	Brazil	ICU Case report	Autopsy confirmed CAPA	AH, DM, CReF
Nasir et al. [26]	Pakistan	ICU, 23 (21.7)	4 patients with colonization	DM, AH, Atrial myxoma, Recent stroke
Lamoth et al. [27]	Switzerland	ICU, 118 (3.8)	None	AH, DM, Obesity, Pulmonary fibrosis, BA
Van Biesen et al. [28]	Netherlands	ICU, 42 (21.4)	None	COPD, DM, AH, Chronic steroid treatment, Neutropenia, Stem cell transplant, Immunodeficiency
Flikweert et al. [29]	Netherlands	ICU, Unknown number of included patients	7 CAPA cases	AH, DM, CReF
Falces-Romero et al. [30]	Spain	ICU, Unknown number of included patients	10 CAPA cases	MDS, HIV, DM, COPD, Ankylosing spondylitis, Acquired hemophilia A, Hypothyroidism, CLL, Cardiac Disease
Helleberg et al. [31]	Denmark	ICU, 8 (25)	Patients under ECMO	AH, BA
White et al. [32]	UK	ICU, 135 (14)	None	DM, AH, CReF, Obesity, Cancer, CRF malignancy, Hyperlipidaemia, Cardiac and vascular disease, Autoimmune disorders
Gangneux et al. [33]	France	ICU, 45 (15.6)	8 patients with colonization	DM, AH, Cancer, Hemopathy, CRF, Cardiovascular disease
Lahmer et al. [34]	Germany	ICU, 2 cases	None	Pulmonary fibrosis
Borman et al. [35]	UK	ICU, 719 (20)	15 possible/probable CAPA cases
Machado et al. [36]	Spain	ICU, 239 (3.3)	5 patients with colonization	DM, AH, BA, Obesity, COPD, CReF, CLL, Non-alcoholic fatty liver disease, CNS disease
Dellière et al. [37]	France	ICU, 366 (5.7)	None	Kidney transplant recipient, HIV, Cancer, Steroids treatment
Brown et al. [38]	UK	ICU, 62 (20)	(+) GM test: 6 patients (+) PCR 5 patients
Fekkar et al. [39]	France	ICU, 145 (4.8)	6 CAPA cases, 1 fusariosis	DM, AH, Obesity, Tabagism, Kidney and liver transplantation, Steroids treatment, Dyslipidemia
Razazi et al. [40]	France	ICU, 90 (11)	Probable (8%), Putative (2%) Aspergillus tracheobronchitis (1%)	DM, AH, COPD, Dialysis, Stroke, CHF (NYHA classification 3–4), Arrhythmias, CReF
Chauvet et al. [41]	USA	ICU, 46 (13.3)	None	Atrial fibrilation, COPD, AH, OSA, DM, CRF, Coronary Disease, CHD, ESRD, Nephrectomy, Vasculitis, Junctional tachycardia, Bipolar disorder, Hypercholesterolemia, Obesity, Hypothyroidism, Gastric ulcer, Atherosclerosis, Sarcopenia

The incidence includes proven/probable and putative cases of CAPA, while patients with colonization are mentioned in comments.

AH: Arterial hypertension; AML: Acute myeloid leukaemia; BA: Bronchial asthma; CAPA: COVID-19-associated aspergillosis; COPD: Chronic obstructive pulmonary disease; COVID-19: Coronavirus disease 2019; CHD: Congenital heart disease; CHF: Congestive heart failure; CLL: Chronic lymphocytic leukemia; CNS: Central nervous system; CRF: Chronic respiratory failure; CReF: Chronic renal failure; DM: Diabetes mellitus; ECMO: Extracorporeal membrane oxygenation; ESRD: End stage renal disease; GM: Galactomannan; HIV: Human immunodeficiency syndrome; ICU: Intensive care unit; IL-6: Interleukin-6; IPA: Invasive pulmonary aspergillosis; OSA: Obstructive sleep apnea; MDS: Mylodysplastic syndrome; NYHA: New York Heart Association; PCR: Polymerase chain reaction.

The medical history, underlying conditions, comorbidities, and risk factors related to the development of CAPA are shown in Table 1. Two previous studies have identified long-term treatment with corticosteroids as a risk factor for infection with Aspergillus spp. [8,28]. Moreover, the role of dexamethasone (doses) in the early stage of COVID-19 and that of tocilizumab in the development of a cytokine storm warrant further investigation [10,44,45].

Possible Pathophysiologic Mechanisms

Currently, the pathogenesis of CAPA is not well defined. In influenza, disruption of the lung epithelium with dysfunctional mucociliary clearance, local immuno-paralysis due to influenza-induced hypoxia, treatment with corticosteroids, acute respiratory distress syndrome, and suppression of the nicotinamide adenine dinucleotide phosphate oxidase complex favor tissue invasion by Aspergillus spp. In addition, invasive Aspergillus tracheobronchitis has been described in approximately 55% of patients with IAPA. Examination through bronchoscopy has revealed the presence of epithelial plaques, pseudo membranes, and ulceration, as well as sporulating heads of Aspergillus spp. inside the bronchi [46]. Similar to influenza, in severe COVID-19 pneumonia, destruction of the bronchial mucosa and alveolar injury caused by the virus create favorable conditions for fungal growth [47]. The main histo-pathological findings in 14 patients with COVID-19 were diffuse alveolar damage in the acute or organizing phase of the disease, with the virus located in the pneumocytes and tracheal epithelium, and scarce focal pulmonary micro thrombi. This observation indicated that the increased pulmonary epithelial and vascular permeability facilitates the invasion of Aspergillus spp. [2,48].

It is well established that Aspergillus spp. live in the environment, and the inhalation of its spores (conidia) leads to pulmonary disease depending on the host immune status. Severe COVID-19 is frequently associated with immune dysregulation, characterized by a decrease in the number and functionality of CD4+ T and CD8+ T-cells and a hyper inflammatory state. Overexpression of pro- and anti-inflammatory cytokines contributes to a highly permissive inflammatory environment that enhances fungal growth [49,50]. Furthermore, damage-associated molecular patterns, which are implicated in the pathogenesis of aspergillosis, are released during infection with severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2), thereby leading to an excessive inflammatory response and lung injury [49].

COVID-19-associated immune dysregulation and immunosuppressive treatment, rather than the pathophysiology of IPA, are host factors for CAPA. Except for corticosteroids (a known risk factor for IPA), tocilizumab (monoclonal antibody against the interleukin-6 [IL-6] receptor), interferon 1β (IFN1β), and a combination of tocilizumab with corticosteroids were more frequently administered in patients with CAPA than in those without aspergillosis (71.4% vs. 33.3%, P = 0.050; 71.4% vs. 20.9%, P ≤ 0.050; and 57.1% vs. 28.7%, P = 0.180, respectively) [51]. Blockage of IL-6 inhibits the development of protective T-helper cells (Th17 cells), leading to a defective immune response against infection with Aspergillus fumigatus (A. fumigatus). In addition, it has been reported that IL-6 enhances epithelial integrity during injury while the wide use of dexamethasone in critically ill patients with COVID-19 renders these patients more susceptible to IPA [52]. The maturation of phagosomes that degrade A. Fumigatus spores via the process of phagolysosomal fusion is impaired by corticosteroids compromising the host defense against the Aspergillus spp. [53].

Imaging

The use of computed tomography (CT) imaging may be unsuitable in these patients. It may be difficult to document changes indicative of CAPA in the parenchyma through CT imaging. This is because mechanically ventilated COVID-19 patients without invasive aspergillosis often have nodular infiltrates, complicating the identification of surrounding halos in the scans. Radiological characteristics of IPA (e.g., solitary or multiple pulmonary nodules, halo sign, reverse halo sign, ground-glass opacity, air crescent, and cavitation) may not be distinct in severe COVID-19 pneumonia, which presents in CT scans with bilateral, peripheral ground-glass opacities, crazy-paving pattern, consolidation, and broncho vascular thickening. Patti et al. [54] reported a mechanically ventilated patient with COVID-19, whose CT analysis showed bilateral peripheral ground glass infiltrates, with newly formed thin-walled cavitary lesions occupied by fungal ball-like lesions. Aspergillus flavus was isolated in respiratory cultures. Moreover, a CT is not always feasible due to the risk associated with the transportation of critically ill patients. White et al. [32] proposed a diagnostic algorithm for CAPA that, along with clinical and mycological criteria, incorporated typical radiological signs of IPA, the presence of new infiltrates, and evidence of sinusitis [36].

Mycological Criteria

Regarding the mycological criteria for the diagnosis of CAPA, 85 of 208 (42%) patients included in all studies underwent bronchoscopy [Table 2] (one study did not report the number of patients who underwent bronchoscopy) [26]. At the start of the epidemic, the use of bronchoscopy was avoided due to shortages in personal protective equipment. In two studies, non-directed bronchoalveolar lavage (NBL) was performed; although this method is less invasive and safer than bronchoscopy, it is linked to a higher risk of sample contamination by upper respiratory flora [28,55]. Table 2 presents the total number of positive respiratory cultures for Aspergillus spp. (155 of 189 tested patients [82%]), including bronchoalveolar lavage (BAL), NBL, tracheal aspirate (TA), and sputum cultures. In the majority of these cultures (e.g., 105), A. fumigatus was the most commonly isolated pathogen.

Table 2.

Mycological results.

Study	(+) serum GM	(+) serum BDG	Bronchoscopy	GM (BAL/TA)	PCR (BAL, TA, serum)	(+) culture (BAL, NBL, TA, sputum)	Aspergillus spp. (n)
Machado et al. [36]	4/8	2/8	2/8	2/8	NA	8/8	A. fumigatus (6), A. citrinoterreus (1), A. lentulus (1)
Koehler et al. [16]	2/5	NA	3/5	3/3	4/5	3/5	A. fumigatus (3)
Nasir et al. [26]	0/5	1/5	NA	NA	NA	5/5	A. flavus (4), A. fumigatus (1), A. fumigatus (1),
Alanio et al. [9]	1/9	4/9	9/9	2/9	4/9	7/9	A. fumigatus
Dupont et al. [61]	1/12	NA	9/19	5/8	NA	16/19	A. fumigatus (14), A. calidoustus (1), A. niger (1)
Prattes et al. [19]	0/1	0/1	0 /1	NA	NA	1/1	A. fumigatus (1)
Wang et al. [21]	NA	NA	4/8	NA	NA	8/8	A. fumigatus (8)
Bruno et al. [64]	0/1	NA	1/1	1/1	NA	1/1	A. fumigatus (1)
Meijer et al. [22]	0/1	1/1	0/1	1/1	NA	1/1	A. fumigatus (1)
Santana et al. [25]	1/1	NA	0/1	NA	NA	NA	A. penicillioides (1)
Borman et al. [35]	5/15	13/15	5/15	4/5	3/11	5/8	A. fumigatus
Segrelles-Calvo et al. [51]	NA	NA	6/7	NA	NA	7/7	A. fumigatus (3), A. flavus (2), A. niger (2)
Patti et al. [54]	0/1	NA	0/1	NA	NA	1/1	A. flavus (1)
White et al. [32]	2/4	14/18	0/25	17/19	15/15	11/11	A. fumigatus (11), A. versicolor (1)
Helleberg et al. [31]	1/2	NA	1/2	2/2	NA	2/2	A. fumigatus (2)
Trujillo et al. [65]	0/1	1/1	0/1	0/1	NA	1/1	A. fumigatus (1)
Spadea et al. [66]	1/1	1/1	0/1	0/1	NA	NA
Van Biesen et al. [28]	NA	NA	0/9	9/9	NA	7/9	A. fumigatus (5), A. flavus (1), A. terreus (1)
Lamoth et al. [27]	1/3	NA	0/3	NA	NA	3/3	A. fumigatus (3)
Mitaka et al. [67]	1/3	NA	0/4	NA	NA	4/4	A. fumigatus (4)
Nasri et al. [68]	1/1	NA	0/1	NA	NA	NA	NA
Gangneux et al. [33]	2/7	NA	0/7	NA	7/7	6/7	A. fumigatus (6)
Roman-Montes et al. [58]	6/14	NA	0/14	8/14	NA	9/14	A. fumigatus (6), A. flavus (2), A. niger (1), A. versicolor (1), Aspergillus spp. (1)
Blaize et al. [18]	0/1	0/1	0/1	0/1	1/2	1/1	A. fumigatus (1)
Schein et al. [55]	1/1	NA	0/1	1/1	1/1	0/1	NA
Fernandez et al. [69]	1/1	NA	0/1	NA	NA	1/1	A. flavus (1)
Ghelfenstein-Ferreira et al. [24]	0/1	0/1	0/1	NA	1/2	1/1	A. fumigatus (1)
Falces-Romero et al. [30]	1/2	NA	1/8	1/1	NA	8/8	A. nidulans (1), A. fumigatus (7)
Mohamed et al. [23]	1/1	1/1	0/1	1/1	NA	1/1	A. fumigatus (1)
Antinori et al. [20]	1/1	NA	0/1	NA	NA	1/1	A. fumigatus (1)
Lahmer et al. [34]	1/2	NA	2/2	2/2	NA	2/2	A. fumigatus (2)
Chauvet et al. [41]	NA	NA	3/6	2/3	1/1	4/6	A. fumigatus (4)
Bartoletti et al. [8]	1/30	NA	30/30	30/30	20/30	19/30	A. fumigatus (15), A. niger (3), A. flavus (1)
van Arkel et al. [17]	0/3	NA	3/6	3/3	NA	5/6	A. fumigatus (5)
Rutsaert et al. [10]	1/6	NA	6/7	6/6	NA	6/7	A. flavus (1), A. fumigatus (5)

BAL: Bronchoalveolar lavage; BDG: β-d-Glucan; GM: Galactomannan; NA:Not available; NBL: Non directed BAL; PCR: Polymerase chain reaction; TA: Tracheal aspirate.

Other mycological methods, including polymerase chain reaction (PCR) and the lateral flow test, require validation. The Aspergillus-specific lateral-flow device test detects an extracellular glycoprotein antigen secreted by Aspergillus spp. only during active growth. This method has been validated in serum and BAL; it has shown a 79% sensitivity and 85% specificity for probable or proven IPA in non-COVID-19 ICU patients [56]. The lateral-flow device test was used in two cases in the literature and is currently being evaluated in patients with influenza and COVID-19 IPA [19,22] (AspiFlu study ISRCTN51287266; https://doi.org/10.1186/ISRCTN51287266); however, it requires validation. Serial assessment of serum galactomannan (GM) (despite its low sensitivity) and serum β-d-glucan in combination with multiple cultures of TA/bronchial aspirate or NBL and PCR testing of serum and respiratory specimens have been included in the diagnostic work-up [50]. Multiple and repetitive positive mycology tests in combination with typical radiological and clinical criteria contribute to the diagnosis of CAPA. In tracheobronchial cases suspected of CAPA, bronchial biopsies are required. Biomarkers are often negative. The use of different diagnostic criteria may contribute to the varying mortality rates reported in the literature. Indeed, the mortality rate is higher among patients with positive cultures than those with probable CAPA detected using GM. Moreover, the mortality rate is higher in patients with multiple positive Aspergillus spp. results than in those with a single positive biomarker.

Biomarkers

Systemic markers are suboptimal for the diagnosis of CAPA, with sensitivity <50%. Furthermore, some techniques require validation. The GM assay has been validated in BAL and serum, although numerous studies have also used endotracheal aspirates. The specificity of GM in BAL is suboptimal and does not satisfactorily discriminate infection from contamination. In six patients with COVID-19 acute respiratory distress syndrome and a positive value for GM in BAL, the diagnosis of probable CAPA was not confirmed post mortem [29,46]. Thus, the identification of some “probable” cases through this approach remains uncertain.

Serum GM is a sensitive biomarker for IPA in patients with neutropenia; however, in non-neutropenic critically ill patients, serum and BAL GM exhibited a sensitivity of 25% and 88–90%, respectively [47,57]. Similarly, in a prospective multicenter study, 28% of ICU patients who were screened for CAPA and 100% of those classified as probable CAPA had a positive BAL GM-index >1; only 1% and 3% of those had positive serum GM [8]. Overall, the GM index in serum (>0.5) and respiratory samples (BAL, NBL, TA) was positive in 37/144 (26%) and 100/129 (78%) patients [Table 2]. Antifungals and chloroquine/hydroxychloroquine, which exhibit in-vitro activity against A. fumigatus, may interfere with the GM measurement and decrease its sensitivity [52]. Furthermore, the reduced release of serum GM may reflect that CAPA is characterized by more pronounced pulmonary invasion and less fungal angioinvasion. Therefore, a negative serum result cannot exclude the diagnosis of CAPA.

Although β-d-glucan is more sensitive than serum GM, it lacks specificity because it is detected in various invasive fungal infections [58]. Its levels were elevated in 38/62 (61.3%) patients with CAPA [Table 2]. Although PCR is mainly performed for patients with hematologic malignancies and hematopoietic stem cell transplants, it has shown higher sensitivity for the diagnosis of CAPA compared with cultures using respiratory samples [30,33]. PCR using serum and/or respiratory samples was positive in 57/81 (70.4%) patients with CAPA [Table 2]. Based on these findings, the revised European Organization for Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG) mycological criteria include PCR using BAL and serum samples; this method is recommended for screening and confirmation of the diagnosis of probable IPA [59].

Diagnosis

The presentation of CAPA is variable and involves two distinct forms, namely tracheobronchial and parenchymal CAPA. It is important to distinguish between these two forms, as this may impact the diagnostic and therapeutic approaches. Discrepancies in the onset of CAPA could be due to differences between these forms, and examination through bronchoscopy may be required. The currently available literature suggests that some patients with CAPA survive without receiving antifungal therapy. This indicates a potential distinction between angio-invasion and minimal invasive disease/tissue invasion.

The diagnosis of CAPA is challenging. The discrimination between invasive infection and colonization is difficult. This is because the available diagnostic tests, except for histopathological examination, do not yield absolute evidence of infection. There is considerable clinical uncertainty regarding the accurate identification of “probable” cases of CAPA [7,46,59,60]. Thus far, 213 (including 7 case reports) cases of CAPA have been reported in the literature [Table 3]: 6 proven and 31 probable according to the EORTC/MSG criteria [59]; 133 putative according to the AspICU algorithm [7,60]; 38 probable according to the IAPA criteria proposed by Verweij et al. [46]; and 5 according to the criteria proposed by White et al. [32]. Four studies reported 24 cases of Aspergillus spp. colonization that did not meet the clinical, mycological, and radiological criteria for classification as proven/probable or putative infection.

Table 3.

Diagnostic criteria used for CAPA diagnosis and all cause mortality.

Study	Study design	Diagnostic criteria	Proven	Probable	Putative	IAPA criteria	Colonization	Mortality
Machado et al. [36]	Prospective	EORTC/MSG modified AspICU	0	8	0	0	9	8/8
Koehler et al. [16]	Retrospective	Modified AspICU algorithm	0	0	5	0	0	3/5
Nasir et al. [26]	Retrospective	Modified AspICU algorithm	0	0	5	0	4	3/5
Alanio et al. [9]	Prospective	EORTC/MSG modified AspICU	0	1	8	0	0	4/9
Dupont et al. [61]	Prospective	Modified AspICU algorithm	0	0	19	0	0	7/19
Prattes et al. [19]	Case report	AspICU algorithm	0	0	1	0	0	1/1
Wang et al. [21]	Retrospective	EORTC/MSG	0	8	0	0	0	NA
Bruno et al. [64]	Case report	Modified AspICU algorithm	0	0	1	0	0	0/1
Meijer et al. [22]	Case report	Modified AspICU algorithm	0	0	1	0	0	1/1
Santana et al. [25]	Case report	Post-mortem histopathology	1	0	0	0	0	1/1
Borman et al. [35]	Retrospective	Modified AspICU algorithm	0	0	15	0	0	NA
Segrelles-Calvo et al. [51]	Prospective	EORTC/MSG	0	7	0	0	0	5/7
Patti et al. [54]	Case report	IAPA criteria	0	0	0	1	0	0/1
White et al. [32]	Prospective	AspICU algorithm, modified AspICU algorithm, CAPA criteria	0	0	25	0	0	13/25
Helleberg et al. [31]	Case series	AspICU algorithm	0	0	2	0	0	2/2
Trujillo et al. [65]	Case report	EORTC/MSG	0	1	0	0	0	0/1
Spadea et al. [66]	Case report	EORTC/MSG	0	1	0	0	0	0/1
Van Biesen et al. [28]	Cohort study	AspICU algorithm	0	0	9	0	0	2/9
Lamoth et al. [27]	Cohort study	Modified AspICU algorithm, IAPA criteria	0	0	2	1	0	1/3
Mitaka et al. [67]	Retrospective	AspICU algorithm	0	0	4	0	3	4/4
Nasri et al. [68]	Case report	EORTC/MSG	0	1	0	0	0	1/1
Gangneux et al. [33]	Prospective	AspICU algorithm, Modified AspICU algorithm	0	0	7	0	8	2/7
Roman-Montes et al. [58]	Prospective	Modified AspICU algorithm	0	0	14	0	0	8/14
Blaize et al. [18]	Case report	AspICU algorithm	0	0	1	0	0	1/1
Schein et al. [55]	Case report	EORTC/MSG	0	1	0	0	0	1/1
Fernandez et al. [69]	Case report	EORTC/MSG	0	1	0	0	0	1/1
Ghelfenstein-Ferreira et al. [24]	Case report	AspICU algorithm	0	0	1	0	0	1/1
Falces-Romero et al. [30]	Retrospective	EORTC/MSG, AspICU algorithm	0	1	7	0	0	6/8
Mohamed et al. [23]	Case report	AspICU algorithm	0	0	1	0	0	1/1
Antinori et al. [20]	Case report	EORTC/MSG post mortem	1	0	0	0	0	1/1
Lahmer et al. [34]	Case report	AspICU algorithm	0	0	2	0	0	2/2
Chauvet et al. [41]	Retrospective	EORTC/MSG, AspICU algorithm, Modified AspICU algorithm	0	1	5	0	0	4/6
Bartoletti et al. [8]	Prospective	IAPA criteria	0	0	0	30	0	13/30
van Arkel et al. [17]	Cohort study	IAPA criteria	0	0	0	6	0	4/6
Rutsaert et al. [10]	Case series	AspICU algorithm	4	0	3	0	0	4/7
Total			6	31	138	38	24	105/190 (55%)

Cases according to AspICU algorithm is 133, EORTC/MSG criteria is 37, IAPA criteria is 38, CAPA criteria is 5.

AspICU algorithm: A clinical algorithm to diagnose Invasive Pulmonary Aspergillosis in intensive care unit patients; CAPA: COVID-19-associated aspergillosis; EORTC/MSG: European Organization for Research and Treatment of Cancer/Mycoses Study Group; IAPA: Influenza-associated pulmonary aspergillosis; ICU: Intensive care unit.

The EORTC/MSG criteria [Table 4] are targeted toward immuno-compromised patients. Host factors that are a prerequisite for the diagnosis of probable IPA are not typically present in ICU patients [59]. Additionally, histopathological confirmation in critically ill patients is difficult either due to coagulation abnormalities or the risk of complications caused by mechanical ventilation (e.g., pneumothorax). Blot et al. [60] developed an AspICU algorithm concerning ICU patients. In contrast with the definition established by the EORTC/MSG, where the presence of host factors and specific radiological signs (e.g., halo sign, air-crescent sign, or a cavity) is required, the AspICU algorithm includes general radiological abnormalities observed on CT or chest X-ray examination; host factors must be present in case of negative mycological criteria in BAL. Schauwvlieghe et al. [7] proposed a modified AspICU algorithm for patients with influenza and IPA, in which the GM indices for BAL and serum are included in the mycological criteria along with a positive BAL culture and histo pathologic or direct microscopic evidence of Aspergillus spp. [Table 4]. A positive BAL culture was found in 60% of patients with IAPA; in the remaining patients, the diagnosis was based on a positive GM index for BAL [32]. In a prospective cohort study including 135 ICU patients with COVID-19, 8 patients were classified as putative IPA based on the AspICU algorithm. Twelve more patients were identified following the application of the modified version of this algorithm [8]. Limitations of the aforementioned studies include the small number of patients and the different diagnostic algorithms used. Larger prospective studies are needed to elucidate whether infection with SARS-CoV-2 predisposes patients to aspergillosis, as well as the participation of other risk factors in the development of IPA.

Table 4.

Diagnostic criteria and algorithms used for the diagnosis of CAPA.

EORTC/MSG criteria, 2020 revision [59]
Proven Aspergillosis: 6/213 (2.8%) CAPA cases
Histopathologic, cytopathologic, or direct microscopic examination of a specimen obtained by needle aspiration or biopsy in which hyphae are seen accompanied by evidence of associated tissue damage.
Culture on sterile material: recovery of Aspergillus spp. by culture of a specimen obtained by lung biopsy. Amplification of fungal DNA by PCR combined with DNA sequencing when molds are seen in formalin-fixed paraffin-embedded tissue.
Probable Aspergillosis: 31/213 (14.6%) CAPA cases
At least 1 host factor, a clinical feature and mycologic evidence.
Host factors
1. Recent history of neutropenia (<0.5 × 109 neutrophils/L [<500 neutrophils/mm3] for >10 days) temporally related to the onset of invasive fungal disease.
2 .Hematologic malignancy.
3. Receipt of an allogeneic stem cell transplant.
4. Receipt of a solid organ transplant.
5. Prolonged use of corticosteroids (excluding among patients with allergic broncho pulmonary aspergillosis) at a therapeutic dose of ≥0.3 mg/kg corticosteroids for ≥3 weeks in the past 60 days.
6. Treatment with other recognized T-cell immuno suppressants, such as calcineurin inhibitors, tumor necrosis factor-a blockers, lymphocyte-specific monoclonal antibodies, immunosuppressive nucleoside analogues during the past 90 days.
7. Treatment with recognized B-cell immuno suppressants, such as Bruton's tyrosine kinase inhibitors, e.g., ibrutinib.
8. Inherited severe immunodeficiency (such as chronic granulomatous disease, STAT 3 deficiency, or severe combined immunodeficiency).
9. Acute graft-vs.-host disease grade III or IV involving the gut, lungs, or liver that is refractory to first-line treatment with steroids.
Clinical features
The presence of 1 of the following 4 patterns on CT:
1. Dense, well-circumscribed lesions(s) with or without a halo sign.
2. Air crescent sign.
3. Cavity.
4. Wedge-shaped and segmental or lobar consolidation.
Mycological evidence
1. Aspergillus recovered by culture from sputum, BAL, bronchial brush, or aspirate.
2. Micro scopical detection of fungal elements in sputum, BAL, bronchial brush, or aspirate indicating a mold.
3. GM antigen detected in plasma, serum, BAL. Any 1 of the following:
- Single serum or plasma: ≥1.0.
- BAL fluid: ≥1.0.
- Single serum or plasma: ≥0.7 and BAL fluid ≥0.8.
4. Aspergillus PCR. Any 1 of the following:
- Plasma, serum, or whole blood 2 or more consecutive PCR tests positive,
- BAL fluid 2 or more duplicate PCR tests positive,
- At least 1 PCR test positive in plasma, serum, or whole blood and 1 PCR test positive in BAL fluid.
Possible Aspergillosis
A host factor and a clinical feature but not mycological evidence
AspICU algorithm, 2012[60]
Putative (all four criteria must be met): 133/213 (62.4%/) CAPA cases
1. Aspergillus-positive lower respiratory tract specimen culture (entry criterion)
2. Compatible signs and symptoms (one of the following)
- Fever refractory to at least 3 days of appropriate antibiotic therapy.
- Recrudescent fever after a period of defervescence of at least 48 h while still on antibiotics and without other apparent cause.
- Pleuritic chest pain or pleuritic rub.
- Dyspnea.
- Hemoptysis.
- Worsening respiratory insufficiency in spite of appropriate antibiotic therapy and ventilatory support.
3. Abnormal medical imaging by Chest X-ray or CT scan of the lungs
4. Either 4a or 4b
4a. Host risk factors (one of the following conditions)
- Neutropenia (absolute neutrophil count < 500/mm3) preceding or at the time of ICU admission
- Underlying hematological or oncological malignancy treated with cytotoxic agents
- Glucocorticoid treatment (prednisone equivalent, >20 mg/day)
- Congenital or acquired immunodeficiency
4b. Semiquantitative Aspergillus-positive culture of BAL fluid (+ or ++), without bacterial growth together with a positive cytological smear showing branching hyphae
Colonization
When ≥1 criterion necessary for a diagnosis of putative IPA is not met
Modified AspICU algorithm, 2018[7]
AspICU algorithm 1,2,3
Mycological criteria
One or more of the following:
- Histopathology or direct microscopic evidence of dichotomous septate hyphae with positive culture for Aspergillus from tissue
- A positive Aspergillus culture from a BAL.
A GM optical index on BAL of ≥1
A GM optical index on serum of ≥0.5.
IAPA criteria 2020[46]
Probable: 38/213 (17.8%) CAPA cases
A: Pulmonary infiltrate and at least one of the following:
Serum GM index > 0.5 or BAL GM index ≥ 1.0 or
Positive BAL culture
B: Cavitating infiltrate (not attributed to another cause) and at least one of the following:
Positive sputum culture or
Positive TA culture

AspICU: A clinical algorithm to diagnose Invasive Pulmonary Aspergillosis in intensive care unit patients; BAL: Bronchoalveolar lavage; CT: Computed tomography; CAPA: COVID-19-associated aspergillosis; EORTC/MSG: European Organization for Research and Treatment of Cancer/Mycoses Study Group; GM: Galactomannan; IAPA: Influenza-associated pulmonary aspergillosis; ICU: Intensive care unit; IPA: Invasive pulmonary aspergillosis; PCR: Polymerase chain reaction; TA: Tracheal aspirate.

Finally, a group of international experts proposed consensus criteria for a definition of CAPA [13]. The diagnosis includes three different grades (e.g., possible, probable, and proven) based on host factors, clinical factors, and mycological evidence. Proven CAPA is based on direct microscopic detection or histopathological analysis of fungal elements morphologically consistent with Aspergillus spp., showing invasive growth into tissues and associated tissue damage. In non-proven CAPA, classification relies on respiratory cultures or biomarkers detected using BAL or NBL in case of probable and possible CAPA, respectively. PCR and lateral flow assay using BAL or NBL are also considered useful in the diagnosis of CAPA along with GM in BAL, NBL, and serum. The recognition of probable tracheobronchitis relies on characteristic lesions observed through bronchoscopy in conjunction with positive mycological evidence. In this context, distinguishing between invasive infection and colonization is challenging, and probable/possible definitions may lead to overdiagnosis.

A more comprehensive diagnostic work-up should be pursued for mechanically ventilated COVID-19 patients with a positive TA culture, plaques (if bronchoscopies are performed), worsening hypoxemia, and clinical deterioration. Although screening could be useful, surveillance is not possible for some large hospitals or those unable to perform onsite testing. Biopsy should be performed for tracheobronchial CAPA to confirm it is caused by Aspergillus spp. Tracheobronchial airways appear as white plaques; thus, physicians should exercise caution to avoid misdiagnosis of infection with Candida.

Treatment

Voriconazole (VRC) or isavuconazole (ISV) have been used as the first-line treatment options for possible, probable, and proven CAPA; liposomal amphotericin B (L-AMB) has also been administered as an alternative agent. In all reported studies thus far, 135 of 189 (71.4%) patients received treatment [Table 5], whereas data regarding treatment were not reported for 24 patients. The main reason for not administering antifungal treatment was early death. The most commonly antifungal agent was VRC, followed by ISV and L-AMB. Dupont et al. [61] reported a non-statistically significant lower mortality rate among patients with putative aspergillosis who were treated with VRC (three deaths in nine patients [33.3%]) vs. those not treated (five deaths in 10 patients [50%])]. VRC is hepatically metabolized, and patients should be monitored for possible drug interactions with cytochrome P450 family 2 subfamily C member 19 (CYP2C19) and CYP3A. Therapeutic drug monitoring is required as toxic levels may lead to hepatotoxicity and neurotoxicity. ISV exhibits a safer profile with less severe adverse events and fewer drug–drug interactions, while the role of L-AMB is limited by acute kidney injury complicating severe COVID-19 [51]. ISV should be preferred in patients for whom liver toxicity is a concern.

Table 5.

Treatment of CAPA patients.

Study	Patients treated (n)	Type of antifungal drug (n)
Machado et al. [36]	5/8	ISV (4), L-AMB (2), VRC (2)
Koehler et al. [16]	5/5	VRC (4), ISV (1)
Nasir et al. [26]	5/5	L-AMB (2), VRC (3)
Alanio et al. [9]	2/9	VRC (1), Caspofungin (1)
Dupont et al. [61]	9/19	VRC (9)
Prattes et al. [19]	1/1	VRC (1)
Wang et al. [21]	NA	NA
Bruno et al. [64]	1/1	VRC (1)
Meijer et al. [22]	1/1	VRC (1)
Santana et al. [25]	0/1	NA
Borman et al. [35]	NA	NA
Segrelles-Calvo et al. [51]	4/7	L-AMB (1), Itraconazole (3)
Patti et al. [54]	1/1	VRC (1)
White et al. [32]	17/25	VRC (16), L-AMB (6), Caspofungin + VRC (2), L-AMB + Anidulafungin (1)
Helleberg et al. [31]	2/2	VRC (2)
Trujillo et al. [65]	1/1	ISV+ neb L-AMB (1)
Spadea et al. [66]	1/1	L-AMB (1)
Van Biesen et al. [28]	9/9	L-AMB + VRC (9)
Lamoth et al. [27]	3/3	VRC (3)
Mitaka et al. [67]	4/4	VRC (3), Caspofungin (1)
Nasri et al. [68]	1/1	L-AMB (1)
Gangneux et al. [33]	7/7	VRC or ISV
Roman-Montes et al. [58]	12/14	VRC (10), Anidulafungin (2)
Blaize et al. [18]	0/1	NA
Schein et al. [55]	1/1	VRC (1)
Fernandez et al. [69]	1/1	VRC (1)
Ghelfenstein-Ferreira et al. [24]	0/1	NA
Falces-Romero et al. [30]	6/8	VRC (4), VRC + Caspofungin (1), L-AMB (5), ISV (2)
Mohamed et al. [23]	1/1	L-AMB (1)
Antinori et al. [20]	1/1	L-AMB (1)
Lahmer et al. [34]	2/2	L-AMB (2)
Chauvet et al. [41]	4/6	L-AMB (2), VRC (1), VRC + ISV (1), Caspofungin (1)
Bartoletti et al. [8]	16/30	VRC (13)
van Arkel [17]	6/6	VRC + Anidulafungin (5), L-AMB (1)
Rutsaert et al. [10]	6/7	VRC (5), ISV (2)

CAPA: COVID-19 associated aspergillosis, COVID-19: Coronavirus disease 2019; ISV: Isavuconazole; L-AMB: Liposomal amphotericin B; NA: Not availableVRC: Voriconazole.

Patients should also be monitored for the development of resistance by Aspergillus spp. to azoles. Triazole-resistant A. fumigatus was isolated in a patient who was possibly exposed to organic matter. The patient expired due to massive pulmonary embolism shortly after the diagnosis of putative aspergillosis [24]. The second case reported in the literature was also a patient with daily exposure to fungicides [23]. The TR34L98H mutation in the CYP51A gene, which is associated with resistance, was identified in Aspergillus strains isolated from both patients. Susceptibility testing is recommended in regions with a resistance rate >5%. In case of azole failure or in regions with a resistance rate >10%, VRC/ISV in combination with an echinocandin or L-AMB or monotherapy with L-AMB should be administered. New antifungal agents are currently evaluated in clinical trials. Rezafungin, which belongs to echinocandins, and ibrexafungerp (formerly termed SCY-078), which inhibits 1,3-β-d-glucan synthase, have shown in-vitro activity against Aspergillus spp., including azole-resistant A. fumigatus isolates. The antifungal agents olorofim and fosmanogepix, which have also demonstrated activity against Aspergillus spp., are currently under development [62].

The diagnosis of CAPA is challenging, and patients with severe COVID-19 pneumonia complicated by IPA may be associated with a worse prognosis. Therefore, the administration of prophylactic treatment with posaconazole, VRC, itraconazole, or inhaled amphotericin B (recommended for prophylaxis in patients with hematological malignancies and transplants) is currently under debate. Rutsaert et al. [10] after finding an unexpected number of COVID-19 patients with suspected IPA, subsequently used nebulized L-AMB (12.5 mg) for prophylaxis in every mechanically ventilated patient they treated. Nevertheless, this approach was linked to a risk of sudden complications in ventilated patients due to obstruction of expiratory filters. In this study, the rate of all-cause mortality was 55%, while a significantly higher 30-day mortality rate was observed in patients with CAPA vs. those without CAPA (44% vs. 19%, P = 0.002 and 74% vs. 26%, P< 0.001 for probable and putative IPA, respectively) [8]. Indeed, there is a general consensus against the use of prophylactic therapy in patients with COVID-19. This position was supported by a recent randomized clinical trial, which failed to show significant benefit after the administration of prophylactic therapy [63].

Conclusion

The incidence of CAPA appears lower than that of IAPA, but varies regionally. In addition, it is influenced by differences in the standard of care, risk conditions, and poor performance of diagnostic tests. Different diagnostic strategies are necessary to differentiate colonization from bronchial or parenchymal infection in intubated COVID-19 patients with Aspergillus spp. in their respiratory specimens vs. those not infected with SARS-CoV-2. The usefulness of imaging techniques (e.g., CT scans) in the diagnosis of CAPA in patients with COVID-19 is limited, whereas bronchoscopy (under safe conditions) is adding value to the diagnostic process. Following confirmation, VRC or ISV should be used for the treatment of CAPA. The appropriate treatment duration for CAPA is currently unknown.

Conflicts of Interest

Jordi Rello served in the speakers bureau and as a consultant for Pfizer. The other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Managing Editor: Jingling Bao

Appendix

Search strategy

Data for this work were identified by searches of MEDLINE, PubMed using the search string “(Aspergill”) AND (“invasive” OR “infection” OR “case” OR “patient” OR “report”) AND (“COVID*” OR “corona”), AND (“SARS-CoV-2”) AND (“Aspergill*”), AND (“aspergill*”) AND (guideline OR treatment OR therapy OR diagnosis). Only articles published in English until January 31, 2021 were included.