The clinical spectrum of aspergillosis in chronic obstructive pulmonary disease

Introduction

Chronic obstructive pulmonary disease (COPD) is the third leading cause of death worldwide [1]. It affects more than 500 million people [2] (more than 7% of the global population), posing significant morbidity, mortality, health and economic burden [3–5]. COPD is associated with chronic debilitating respiratory symptoms, including breathlessness, cough and sputum production, as well as systemic symptoms, including fatigue, muscle wasting and deconditioning [6]. COPD exacerbations, punctuating the natural history of the disease, are associated with burdensome symptoms, quality of life decline and a high mortality rate, which in the case of severe exacerbations (those requiring hospital admission) exceeds 15% within 3 months from an event [7]. These events are frequent, since each year, 22–40% of all patients with COPD experience at least one exacerbation, while 9–16% experience more than one exacerbation. Although COPD is both preventable and treatable, the clinical course is usually progressive and irreversible.

The aetiology of exacerbations is very heterogeneous, and includes bacterial [8], viral [9] or possibly fungal infections [10], enhanced eosinophilic inflammation, mechanical and environmental causes. However, in clinical practice COPD exacerbations are still managed as a single disease entity and treated with antibiotics, bronchodilators and systemic corticosteroids, that are not always required, and pose a significant treatment burden. Frequent courses of antibiotics promote antimicrobial resistance and microbial dysbiosis, while systemic corticosteroids are associated with significant side effects, that include a predisposition to infections, such as pneumonia, but also fungal infections. Better characterisation of the aetiology of COPD exacerbations in clinical practice is therefore crucial.

Considering that isolation of Aspergillus fumigatus from respiratory secretions is a common occurrence in COPD patients, surprisingly little is known about its significance and implications. Aspergillus spp. are ubiquitous fungal pathogens causing a wide range of airway disorders with A. fumigatus the most frequently isolated species [11]. Given their small spore size (2–3 µm) and abundance (inhaled air can contain up to 100 conidia/m³) [12] Aspergillus conidia are able to reach the lung alveoli. However, these conidia are normally promptly removed by the innate immune system of the immunocompetent host and are unable to cause disease. Immunocompromised hosts like haematological malignancy or transplant patients are at the highest risk of developing invasive aspergillosis. The airway defences of patients with COPD are less effective at killing Aspergillus spores [13, 14], and they can escape killing by both professional phagocytes and epithelial cells and germinate. Recently however, there has been increased interest in the role of Aspergillus in groups previously considered to be at low risk of aspergillosis, including COPD patients.

Various clinical syndromes related to Aspergillus have been described in COPD [15]. Invasive aspergillosis (IA) is well recognised in COPD patients in intensive care units (ICU) and is associated with high mortality [16, 17], partly due to delays in diagnosis. Aspergillus may also contribute to more indolent pathology in COPD as a subset of patients with COPD will develop chronic pulmonary aspergillosis (CPA) [18]. This entity remains underdiagnosed due to its indolent nature, but may progress to severe end-stage disease, poor quality of life and life-threatening haemoptysis. In addition, A. fumigatus is isolated from the sputum of patients with COPD at steady state and during exacerbations in up to a third of cases, though its contribution to the symptoms and COPD progression is unclear [15]. Aspergillus sensitisation or presence of A. fumigatus in the airways has been linked with bronchiectasis and worse forced expiratory volume in the first second (FEV1) in COPD [19]. Finally, allergic bronchopulmonary aspergillosis (ABPA) has been diagnosed in COPD patients with or without asthma, but this entity is poorly characterised, and may not be a discrete complication of COPD. [20]

In this review, we present the clinical spectrum and pathogenesis of syndromes caused by Aspergillus in COPD namely IA, community-acquired Aspergillus pneumonia, CPA and Aspergillus sensitisation. Some of these entities are clearly linked to COPD, while others may coexist, but are less clearly linked directly to COPD (Table 1). We discuss current uncertainties as these pertain to IA in COPD cohorts and explore areas for future research in this field.

Invasive aspergillosis

IA most commonly affects people with a wide spectrum of immunosuppressive disorders [21]. In these settings, the European Organization for Research and Treatment of Cancer/Mycosis Study Group (EORTC/MSG) criteria define IA as proven, probable, or possible based on level of proof [22]. This ranges from the presence of decisive histopathological evidence of fungal invasion (proven) to a combination of clinical data and host risk factors either associated (probable) or not (possible) with the positivity of mycological criteria [at least 1 of the following: (1) cytology, direct microscopy and/or culture indicating presence of Aspergillus spp. in a lower respiratory tract specimen; (2) galactomannan antigen index > 0.5 in plasma/serum and/or galactomannan antigen > 0.8 in bronchoalveolar lavage (BAL) fluid. [22]

Pulmonary aspergillosis is the most frequent clinical manifestation of IA, but haematogenous spread could occur resulting in the involvement of extra-pulmonary sites such as the brain and skin [23]. Aspergillus becomes invasive in the lungs when epithelial cells and alveolar macrophages fail to clear the conidia which germinate into branching filaments called hyphae that invade the lung tissue [24] (Fig. 1). This gives rise to several invasive entities such as tracheobronchitis, acute and sub-acute disease (Fig. 2). Acute IA occurs rapidly with clinical presentation ranging from days to 3–4 weeks. Sub-acute invasive aspergillosis (SAIA) typically occurs in patients who are not profoundly immunocompromised but may be very debilitated and usually runs a slowly progressive course over 4–12 weeks [25]. IA may be angio-invasive and non-angio-invasive with angio-invasion being especially common in neutropenic patients with invasive pulmonary aspergillosis (IPA). This angio-invasion occurs as either invasion of major proximal pulmonary arteries with resultant thrombosis and distal tissue infarction or infiltration on a vessel in the centre of a well circumscribed nodule by hyphal elements [26]. The frequency of angio-invasion in IPA in COPD is uncommon [27]. Non-angio-invasive IPA is relatively more common in corticosteroid-treated patients, including those with COPD, with micro-abscesses, nodules and acute and chronic inflammation seen more frequently than angio-invasion. In a review by Franquet et al. involving nine patients with COPD and semiinvasive aspergillosis proven at autopsy or by thoracoscopically guided lung biopsy in Spain, the radiologic findings consisted of parenchymal consolidation (n = 6), nodules larger than 1 cm in diameter (n = 3), cavitation (n = 2) and multiple cavitated nodules with a variable degree of central necrosis (n = 3) [28].

Community-acquired Aspergillus pneumonia and/or pneumonitis

Invasive aspergillosis in immunocompromised patients may be acquired in the community [79, 88, 89]. Recent reports of invasive aspergillosis in newly hospitalised COPD patients, many of whom, but not all, were treated with corticosteroids, emphasise the high frequency of Aspergillus acquisition in the community [37, 39, 40]. Virtually all cases of ABPA, Aspergillus bronchitis and CPA are, or presumed to be, acquired out of hospital [90]. This is hardly surprising given the ubiquitous nature of A. fumigatus and other species [91, 92]. It would therefore not be surprising that susceptible patients may develop community-acquired Aspergillus pneumonia.

Reports of community-acquired Aspergillus pneumonia and/or pneumonitis are uncommon. The terminology is also variable. There appear to be 3 common features to these reports: A remarkable exposure to airborne spores, prior influenza or other viral infection and/or underlying pulmonary disease, usually emphysema or chronic obstructive pulmonary disease (COPD). Only a few cases are described (Fig. 3).

Clancy and Nguyen presented one case and summarised the prior literature in 1998 [93]. All 12 cases they identified were infected with A. fumigatus and ranged in age from 14 months to 67 years. Microscopy for hyphae was positive in one patient, and sputum cultures were positive for Aspergillus in 7 of 10 patients (criteria for inclusion in the series). All 12 patients died, despite amphotericin B therapy in six.

Two other series require mention—both reporting a slowly progressive ‘necrotising pneumonia’ [94, 95]. Kennedy et al. report 4 cases of a necrotising pneumonia in 4 patients in Edinburgh, none immunocompromised but 2 with overt prior pulmonary disease and one a heavy smoker [94]. Binder et al. present 4 additional cases of a chronic necrotising pneumonia (chronic necrotising pulmonary aspergillosis) attributable to A. fumigatus or in one case A. flavus [95].

Since that time additional cases attributable to exposure related to damp and decomposing bark have been described [96, 97] and other cases attributable to other ‘vegetal’ [98] or decomposing plant materials [99] and gardening [100–102]. The radiological patterns described are primarily those of a miliary, bilateral diffuse pattern, with high level exposure, but unilateral, usually upper lobe cavitary disease is also well described.

Late diagnosis and lack of antifungal therapy are associated with early death and/or progressive disease. Amphotericin B therapy is probably ineffective for severe cases [93, 98, 103]. Voriconazole is recommended for treatment in the acute phase, with the addition of corticosteroids for those with overwhelming infection and/or pneumonitis, in the current state of our relative ignorance of this condition. More slowly progressive infection may respond to itraconazole.

Chronic pulmonary aspergillosis

Aspergillus nodules

Aspergillus nodules are an unusual form of CPA mimicking carcinoma of the lung, metastases, cryptococcal nodule, coccidioidomycosis or other rare pathogens. In immunocompromised patients, IA may present as nodules. But nodules may also occur in non-immunocompromised patients and are referred to as Aspergillus nodules.

Aspergillus nodules may be single or multiple with or without cavitation, most of which are smaller than 3 cm across [118]. Aspergillus nodules have non-specific clinical and radiological manifestations, may be asymptomatic; the nodules are usually incidental discoveries on chest CT scan (Fig. 6). The Aspergillus nodules may not have distinctive features on chest CT thus making it challenging to distinguish them from malignant or other inflammatory nodules. Commoner symptoms include dyspnoea, cough, haemoptysis and weight loss [119]. This entity can only be definitively diagnosed following removal or biopsy of the nodule(s), but is suggested if some of the multiple nodules are cavitary, other infectious aetiologies are ruled out and Aspergillus IgG (or other microbiology) is positive.

Aspergillus IgG antibody testing appears to have a low sensitivity for diagnosing Aspergillus nodules as this was positive in only a small proportion of patients in a recent study [119]. Further research is required to clearly establish the role of Aspergillus IgG antibody in the diagnosis of Aspergillus nodules and identify other means of making the diagnosis without requiring biopsy or resection. Treatment modalities include antifungal therapy and surgical procedures such as lobectomy or sublobar resection depending on the number of nodules, their location and each patient’s lung function.

Aspergillus sensitisation

Aspergillus sensitisation has been associated with increased severity of asthma. However, the relationship between COPD without asthma and fungal sensitisation remains unclear. COPD patients have been shown to have higher Aspergillus IgG (31 mg/L vs 20.7 mg/L) and more frequent Aspergillus sensitisation than controls (18% vs 4%) [120]. Research has shown that Aspergillus-sensitised individuals with COPD more often have bronchiectasis [120] as well as worse FEV1 compared to non-sensitised COPD patients (39% vs 51% predicted) [121], although this was not found in all studies [2]. The average prevalence of Aspergillus sensitisation in COPD from 5 studies comprising more than 1000 patients was found to be 13.6% [2]. Other studies have reported a prevalence of Aspergillus sensitisation in COPD of between 8.5 and 18% [122].

Allergic bronchopulmonary aspergillosis (ABPA) typically occurs in the setting of asthma or cystic fibrosis. In the recent past, there have been several articles have reported an association between COPD and ABPA though the exact pathophysiologic mechanism underpinning this association is yet to be elucidated. The first report of this association involved a 50‐year‐old smoker with COPD who presented with features of an exacerbation and was subsequently diagnosed with ABPA on the basis of positive Aspergillus skin tests (both Type I and III), elevated total serum IgE, Aspergillus specific IgE and IgG, eosinophilia and a positive sputum culture for A. fumigatus [123]. Both COPD and ABPA are characterised by the occurrence of airway inflammation, mucous hypersecretion, impaired mucociliary clearance and airflow obstruction.

A prospective case–control study involving 200 patients with COPD in India identified Aspergillus hypersensitivity in 17 (8.5%) of the patients with COPD as compared to none in the control group [124]. The serologic criteria for the diagnosis of ABPA were fulfilled in two (1.0%) of the patients. There have since been other case reports of ABPA occurring in patients with COPD who had neither asthma nor CF [125, 126]. This is a dynamic area of study, with uncertainty about the implications for clinical care.

Aspergillus bronchitis

Aspergillus bronchitis is an uncommon manifestation of Aspergillus infection. It is a chronic superficial infection of the lower airways (trachea and bronchi) with Aspergillus, without important tissue invasion, lung parenchymal destruction or overt allergic response [127]. Aspergillus bronchitis is characterised by superficial invasion of airway mucosa by Aspergillus hyphae. This gives rise to mucoid impaction, production of thick tenacious sputum with bronchial plugging. Aspergillus IgG antibody is typically detectable in serum and Aspergillus spp. repeatedly detectable in sputum or bronchoscopy fluid with culture or PCR.

Aspergillus bronchitis is most common in those with bronchiectasis, but has also been reported to occur in patients with COPD. A retrospective analysis of 38 patients presenting with tracheobronchitis among non-neutropenic/non-transplant adult patients with at least two cultures of respiratory samples yielding Aspergillus spp. identified 26 (81.3%) patients with GOLD III to IV COPD [128]. A smaller series involving 17 patients with Aspergillus bronchitis identified 6 (35%) patients with COPD [129].

The various Aspergillosis entities in COPD patients and therapies are summarised in Table 2.

Bronchiectasis

Bronchiectasis is a chronic airways syndrome usually defined in the presence of permanent airway dilatation (often demonstrated on high-resolution CT scanning), variable mucociliary clearance, recurrent airway symptoms of daily sputum production and episodic infective exacerbations [130–132].

The common manifestations of Aspergillus disease in bronchiectasis include CPA, APBA pulmonary aspergilloma and IA [133]. Aspergillus diseases such as ABPA may be complicated by the development of bronchiectasis whereas patients with already established bronchiectasis and lung architectural disruption can develop Aspergillus-related diseases such as aspergillomas on top of post-TB bronchiectasis [134]. Bronchiectasis and COPD are reported to coexist in 20–60% of cases, and coexisting disease is linked to greater symptomatology, reduced therapeutic options and poorer prognosis when compared to bronchiectasis or COPD alone [135–138]. Misdiagnosis of the two conditions frequently occur with computed tomography (CT) without high-resolution algorithms being identified as the main cause of under-diagnosis of bronchiectasis [134].

Bronchiectasis does share some pathogenic similarities with COPD. In both diseases, chronic bronchial infection by pathogenic microorganisms and consequent chronic inflammation with remodelling of the airways leads to impairment of local defence mechanisms. This ultimately results in the persistence of microorganisms such as Aspergillus in the bronchial tree despite treatment [139]. In bronchiectasis, Aspergillus infection of the airways is thought to directly damage the airways and impair mucociliary clearance. In normal (or asthmatic/COPD) airways in susceptible individuals, it has been theorised the Aspergillus proteases upregulate mucous production and/or drive a Th2 phenotype thus drive airway inflammation towards frank bronchiectasis [130].

Role of Aspergillus colonisation in COPD

The isolation of Aspergillus from respiratory secretions of COPD patients is common and usually dismissed as representing mere colonisation or contamination, especially outside the setting of critical illness and in the absence of imaging findings suggestive of IA or CPA. However, several studies attempted to assess its significance [27]. A prospective study conducted across four hospitals in Spain involving 240 hospitalised patients with a severe COPD exacerbation reported a 16.6% prevalence of Aspergillus spp. isolation on admission and 14.1% at 1-year follow-up. No patients were found to have IA. Those who isolated Aspergillus were more likely to have had an exacerbation in the preceding year and to also have other pathogens and especially Pseudomonas aeruginosa in sputum [140]. Aspergillus isolation was associated with an increased length of stay (11.8 vs 7.5 days). In a prospective study from the UK, 37% of patients isolated A. fumigatus at baseline at steady state. Isolation of A. fumigatus was associated with higher inhaled corticosteroid dose and higher sputum total cell and neutrophil count. However, there was no link with exacerbation frequency or FEV1 [119]. The association with a higher dose of inhaled corticosteroids was also shown in a retrospective pair-matched study from China; patients with Aspergillus isolation had more prolonged hospital stay (15 vs 12 days) [141]. Finally, patients with COPD who reported activities related to fungal exposure reported more exacerbations. This may suggest a potential association with Aspergillus exposure [142].

Pathogenesis of aspergillosis in COPD

The pathogenesis of COPD involves infections, chronic inflammation and oxidative stress [143]. A. fumigatus is the most common pathogen isolated from the sputum of patients with COPD at steady state and during exacerbations [14]. The phagocytosis of bacteria and apoptotic cells by the alveolar macrophages in COPD patients is impaired though it is uncertain if this applies to the phagocytosis of Aspergillus conidia [144]. Patients with COPD are thought to have impaired ciliary function resulting from exposure to cigarette smoke and recurrent chest infections. This may promote the binding of the Aspergillus conidia to the airways in a concentration-dependent manner. The epithelial responses to A. fumigatus result in remodelling of the respiratory epithelium with a heightened expression of Th2 cytokines [13]. Cytoskeletal collapse and apoptosis eventually occur. In COPD, it is possible that a combination of several factors limits the ability to eliminate the conidia of Aspergillus. Defects of the innate immune system (inherited or acquired), regular use of high-dose steroids in addition to frequent use of broad-spectrum antibiotics in case of exacerbations, may all play a role in promoting the colonisation of the respiratory epithelium by Aspergillus.

Pentraxin 3 (PTX3) is an acute phase protein released during inflammation or injury by several cells. PTX3 binds to the conidia of A. fumigatus and serves as an opsonising agent [145] while also triggering the classical complement pathway and mediating innate immunity to A. fumigatus [146–148]. A recent study carried out among a Chinese COPD population showed a significant association between PTX3 rs1840680 single nucleotide polymorphisms and the susceptibility to pulmonary aspergillosis [149]. Plasma PTX3 levels have been found to substantially increase in the setting of fungal infection [150] and may have utility in the diagnosis of IA in COPD patients.

There is growing evidence that severe COPD predisposes to the development of IA especially, but not exclusively, in the setting of steroid use [31, 32, 37, 39, 48]. It is still unclear why some COPD patients get colonised by Aspergillus spp., whereas others develop IA. High doses of oral steroids reportedly promote the growth of Aspergillus [151] while enhancing Th2 cytokines and dampening Th1 cytokine response [152]. Steroids also inhibit neutrophil action and decrease alveolar macrophage antifungal activity by limiting the production of reactive oxidant intermediates [153, 154]. In vitro studies suggest that steroids promote the growth of A. fumigatus as 30–40% increase in growth rate of A. fumigatus and A. flavus exposed to pharmacological doses of hydrocortisone was demonstrated [151]. The exact mechanism for this increased growth is not fully understood, but the presence of a ligand/receptor system in the fungi has been proposed as a putative mechanism for this association.

A retrospective study of 239 patients admitted with COPD with Aspergillus isolation identified the accumulated dosage of corticosteroids equivalent to > 700 mg prednisolone in the three months prior to admission to be an independent predictor of IA [37]. In this same study, the receipt of antibiotic treatment in the 3 months prior to admission was also a predictor of IA. Inhaled corticosteroids have also been identified as a risk factor for IA [155]. A study conducted among 11 patients with COPD from which Aspergillus isolates were cultured from lower respiratory tract samples revealed that 8 (72.7%) of the patients had used an inhaled steroid, while 10 (90.9%) had used a systemic steroid within the last 3 weeks [41].

A retrospective case–control study of 30 COPD patients with IA and 60 COPD control patients without IA in China identified treatment with three or more antibiotics during hospitalisation and antibiotic treatment longer than 10 days to be risk factors for IA [40]. The use of broad-spectrum antibiotics in the setting of COPD is thought to disrupt the normal flora of the respiratory epithelium thereby promoting the proliferation of fungal pathogens such as Aspergillus [156]. The ability of Aspergillus to form biofilms on bronchial epithelial cells which are difficult to eradicate with antifungals may also be a contributory factor [157].

Uncertainties, future research

Further research is required in several areas. Optimum means of diagnosing IA in COPD with sputum culture/PCR, Aspergillus antibody and antigen, BDG and identification of new biomarkers are key areas to explore further. Prospective studies are needed to establish the significance of Aspergillus isolation in COPD exacerbations, including intervention studies, and consensus reached on whether this requires treatment. The utility of Aspergillus biomarkers such as BDG and Aspergillus antibody assays in COPD monitoring of therapy for IA in COPD needs to be clearly elucidated.

Likewise, genetic risk factors for IA in need to be investigated. There are many described in leukaemia, haematopoietic stem cell transplantation and solid organ transplantation, but this is largely unexplored in COPD. In addition to PTX3 polymorphisms, other genetic defects may be implicated, such as those affecting the Th1 cytokine pathway. The role of fungal exposure in the development of CPA in patients with COPD also needs to be better defined. The efficacy of IFNγ immunotherapy in the setting of CPA/COPD overlap remains an underexplored but promising area of research.

In conclusion, COPD poses a significant global health concern and urgent measures are required to mitigate the current situation. Healthcare professionals need to be aware of the risk of IA and various forms of aspergillosis in their COPD patients to aid early diagnosis, facilitate prompt management and reduce mortality.

Author contributions

AO, CK and DWD wrote the main manuscript text with contributions from AM and CI. CI prepared Figs. 1 and 2. All authors reviewed the manuscript.

Funding

No funding was received for this article.

Availability of data and materials

Not applicable.

Declarations