Convergence in the Epidemiology and Pathogenesis of COPD and Pneumonia

ABSTRACT

Chronic obstructive pulmonary disease (COPD) is one of the main causes of human mortalities globally after heart disease and stroke. There is increasing evidence of an aetiological association between COPD and pneumonia, the leading infectious cause of death globally in children under 5 years. In this review, we discuss the known risk factors of COPD that are also shared with pneumonia including smoking, air pollution, age and immune suppression. We review how lung pathology linked to a previous history of pneumonia may heighten susceptibility to the development of COPD in later life. Furthermore, we examine how specific aspects of COPD immunology could contribute to the manifestation of pneumonia. Based on the available evidence, a convergent relationship is becoming apparent with respect to the pathogenesis of COPD and pneumonia. This has implications for the management of both diseases, and the development of new interventions.

KEYWORDS:

• Chronic airway obstruction
• corticosteroid
• lung inflammation
• respiratory infection

Introduction

The broad classification of human illnesses as being either communicable or non-communicable has existed for decades and is imbedded in our approach to disease prevention and management. While it holds true for many conditions, there are instances where diseases, previously regarded as non-communicable, were later found to have an infectious component. A renowned example of this was the discovery of the aetiological role of Helicobacter pylori in gastritis, peptic and duodenal ulceration, and gastric cancer (1, 2). It is, therefore, plausible that a crossover between communicable and non-communicable disease may exist with regard to other body systems, including the respiratory system.

Defined as chronic airflow obstruction that is progressive and only partly reversible, chronic obstructive pulmonary disease (COPD) includes chronic bronchitis, emphysema and chronic asthmatic bronchitis (3). It commences with structural damage to the small airways and surrounding alveoli, and develops over several years leading to a loss in elastic recoil and an increase in airway resistance (4). This results in defects in lung function, detectable using spirometry, which become more severe with age (4). COPD is emerging as the third largest cause of human mortality worldwide after heart disease and stroke, killing over 3 million people worldwide each year (5). The World Health Organization (WHO) estimates that 65 million people have moderate-to-severe COPD (6) and that the disease is increasing in prevalence (7). Although COPD is a non-communicable chronic illness, its epidemiological convergence with a number of major respiratory infectious diseases including tuberculosis (8) and pneumonia is of concern.

Pneumonia is an acute infection of the parenchyma of the lung caused by viruses, bacteria, fungi or other pathogenic microorganisms, which is primarily contracted via aspiration or inhalation of infectious particles (9). Inflammation results in migration of neutrophils from blood capillaries into the alveolar spaces (10). Exudate pools in the alveoli leading to cough, painful respiration and shortness of breath. In terms of burden, pneumonia is the single largest infectious cause of death in young children worldwide (11) and a major cause of mortality in the elderly (12, 13). According to Chatila and colleagues, “COPD is more frequently associated with pneumonia compared with other chronic diseases” (14). A recent paper by Hayden and co-workers identified that episodes of pneumonia during childhood significantly increased the likelihood of COPD in later life (15) and raised the question as to whether COPD could “begin in childhood” (16). Other work has established that the use of immunosuppressant corticosteroid-based therapy in COPD patients is linked to pneumonia development (17). In this review, we begin by examining risk factors that are common to both COPD and pneumonia, and then assess the evidence that underlies an overlap in the aetiology of the two diseases.

Shared risk factors of COPD and pneumonia

Tobacco smoking is well established as the most important causative factor for the development of COPD (18, 19) and is related to an estimated 73% of mortality in COPD patients (20). Smoking is also a known risk factor for community-acquired pneumonia (CAP) with reported crude odds ratios (ORs) for current smokers in the range of 1.37 (95% CI: 1.14–1.64) to 1.81 (95% CI: 1.53–2.15) with respect to non-smokers (21). Older adults who are passive smokers are also more susceptible to pneumonia (22). The risk posed by smoking extends to invasive pneumococcal disease (IPD) whereby passive and active smokers have a greater than twofold (OR 2.5, 95% CI: 1.2–5.1) and fourfold (OR 4.1, 95% CI: 2.4–7.3) increased risk, respectively, of developing IPD (23).

It is estimated that 35% of people in low–middle income countries developed COPD due to exposure to indoor smoke from biomass fuels (20). Biomass fuel smoke also increases the incidence of respiratory infectious diseases, including pneumonia (24). Replacement of wood and charcoal-based fuels with cleaner fuels was recommended following a study conducted in Sierra Leone to reduce the risk of acute respiratory infections among children and adult women (25).

Associations have also been identified between workplace air pollutants and lung diseases including COPD (26). Titanium dioxide nanoparticles (<10 nm), widely used in both industry and daily life, have been associated with both damage to lung cell structure and pulmonary alveolar macrophage dysfunction, leading to a reduction in non-specific and specific immune responses (27). In regard to pneumonia in older adults, a significantly higher increased mortality rate has been associated with occupational exposure to metal fumes [risk ratio (RR) of 2.31, 95% CI: 1.35–3.95] and inorganic dust (RR of 1.87, 95% CI: 1.22–2.87) during their working lives (28).

In terms of social determinants of health, COPD has been associated with poor nutrition, household crowding and inadequate access to health care (20, 29). Low birth weight has also been linked to an increased risk of obstructive lung disease in adulthood, as measured by a reduction in expected FEV₁ (forced expired volume of air in the first second of expiration) (30). Similarly, the risk of developing pneumonia in infants has been linked to social deprivation (31) including low household income and maternal education attainment (32).

Alcohol abuse results in higher susceptibility to several types of respiratory illnesses (33). The volatile nature of ethanol enables its movement across the epithelium of the lung airways (34). While mild ethanol exposure leads to enhanced mucociliary clearance, stimulation of bronchodilation, and decreased airway inflammation in asthma and COPD, in contrast, prolonged and heavy exposure disrupts mucociliary clearance, and is believed to worsen lung function and mortality in COPD patients (34). Similarly, Happel and colleagues have related an increase in pneumonia morbidity and mortality to excess alcohol consumption (35).

The risk of pneumonia has been determined from a study of 77,671 age-matched subjects to be higher in both inpatients and outpatients with underlying chronic kidney disease (CKD) (adjusted hazard ratio of 1.97, 95% CI: 1.89–2.05, p  <  0.001) (36). In a smaller study of 181 subjects, the prevalence of CKD was reported to be 31% in the COPD group and 8% in the non-COPD group (OR of 4.91, 95% CI: 1.94–12.46, p = −0.0008) based on estimated glomerular filtration rate with creatinine (37).

COPD and pneumonia share a number of other risk factors including age. The incidence of CAP is strongly age-related with a higher incidence occurring in adults aged 60 years and over (38). In Australia, the median age at death due to pneumonia was 86.2 years for males and 89.9 years for females in 2012 (39). Likewise, a fivefold increase in COPD is reported to prevail among people aged over 65 years compared to those aged 40 years or less (40). An age-related increase in inflammation and reduction in mucociliary clearance during COPD imparts damage to the lung matrix and reduces its ability to repair itself (41,42). Increased tracheobronchial microbial colonisation among people with disrupted mucociliary activity and prolonged inflammation contribute to the occurrence of COPD and pneumonia in the elderly (43).

Immune suppression also plays a role in the manifestation of COPD and pneumonia. The incidence rate of CAP is higher in patients with HIV/AIDS compared to HIV-negative subjects [adjusted ORs: 2.48 (95% CI: 1.34–4.58)] (44). Highly active antiretroviral therapy substantially decreases the risk of pneumonia among HIV-positive individuals, but HIV infection is still a strong risk factor for hospitalisation to treat pneumonia, even in persons with high CD4⁺ T cell counts (45). In particular, incidence rate ratios for HIV-infected individuals with a CD4⁺ T cell count >500 cells/µL were 5.9 (95% CI: 4.2–7.6) during 2005–2007 compared with control individuals (45). According to Morris and co-workers, “HIV-infected persons were noted to have an accelerated form of COPD, with significant emphysematous disease seen in individuals less than 40 years old” (42). In a later section, we will examine the role of iatrogenic immune suppression, in particular due to corticosteroid treatment of COPD, in pneumonia development.

It should be noted that there are a number of risk factors which have been identified in either COPD or pneumonia, but have not been reported for both diseases to date. A population-based cohort study conducted in the Netherlands identified an association between gastric acid suppressive therapy and risk of pneumonia. Among current proton pump inhibitor (PPI) users, the relative risk for pneumonia was 1.89 (95% CI: 1.36–2.62) with respect to those who ceased its use (46). In current H₂ receptor antagonist users, there was a 1.63-fold increased risk of pneumonia (95% CI: 1.07–2.48) (46). Besides increasing the risk of aspiration pneumonia (47), acid-suppressive drugs are believed to modulate the immune system (48, 49) leading to sustained microbial colonisation and potential pneumonia (46, 50). In contrast, PPI use was independently and significantly associated with a lower risk of COPD exacerbations (adjusted OR of 0.23, 95% CI: 0.08–0.62, p = 0.004) (51). Splenectomy has been associated with an increased risk of pneumococcal pneumonia (RR of 2.06, 95% CI: 1.85–2.30) and death due to pneumonia (RR of 1.58, 95% CI: 1.20–2.08) (52) and septicaemia (RR of 3.02, 95% CI: 1.80–5.06) (51), but its role in COPD remains to be established. Conversely, α₁-antitrypsin deficiency has been identified as one of the risk factors for COPD (53), but it is uncertain whether this deficiency poses a risk with respect to the development of pneumonia.

Previous pneumonia history as a risk factor for COPD in later life

In a recent study of 10,192 adult current and former smokers, Hayden and colleagues determined that childhood pneumonia was significantly associated with COPD in smokers (OR of 1.40, 95% CI: 1.17–1.66) (54). This association was supported by airway wall thickness differences as measured by chest computerised tomography, and in significantly lower lung function in terms of post-bronchodilator FEV¹ (69.1% vs. 77.1% predicted), forced vital capacity (FVC) (82.7% vs. 87.4% predicted), and FEV₁/FVC ratio (0.63 vs. 0.67). In terms of acute exacerbations of COPD, a recent study from Korea involving 1,114 subjects over 40 years of age found that the exacerbation rate was many times higher (18-fold, p < 0.001) in COPD patients with a history of pneumonia compared to patients who have not had pneumonia (55).

Other investigators have reported links between childhood pneumonia and lung function. Shaheen and colleagues observed a lower mean FEV₁ in adults who had acquired pneumonia before 2 years of age, even after controlling for smoking (56, 57). More studies are needed on the role of earlier pneumonia in the development of COPD in non-smokers. A mean reduction of FEV₁ in adults by 0.17 L was observed among individuals who had bronchitis or pneumonia in infancy (30). Lange and co-workers have determined that persons with a low FEV₁ (less than 80% of predicted) in early adulthood (before 40 years of age) had a risk of COPD in midlife that was approximately threefold higher than for persons with a higher baseline FEV₁ (p < 0.001) (58). While a low FEV₁ in early adulthood was linked to COPD, the authors also reported that an accelerated decline in FEV₁ from a normal level was not an obligate feature of COPD (58).

Childhood pneumonia is particularly prevalent in low-income countries and is regarded as one of the so-called “childhood disadvantage factors” that leads to ongoing impaired lung function (59, 60). In one study, five childhood disadvantage factors, identified as maternal asthma, paternal asthma, childhood asthma, maternal smoking and childhood respiratory infections were prevalent in the general population and had a significant impact on lung function and development of COPD (59). An increase in COPD was consistent with increasing childhood disadvantage factors, i.e., OR of 1.7 (95% CI: 1.1–2.6) in men and OR of 1.6 (95% CI: 1.01–2.6) in women for one disadvantage factor; and OR of 6.3 (95% CI: 2.4–17) in men and OR of 7.2 (95% CI: 2.8–19) in women for more than three disadvantage factors (60). Reduced weight gain in infants with bronchitis, pneumonia or whooping cough has also been associated with mortality due to COPD in adult life (30).

From the above lines of evidence, it is apparent that childhood pneumonia plays a role in the early origins of COPD (54). Furthermore, inadequate treatment of bacterial lower respiratory tract infection in childhood has been linked to a higher prevalence of COPD among adults in developing countries (43) and to severe COPD exacerbations (54). This highlights the need for the prevention of pneumonia in childhood via measures to reduce social deprivation and through medical interventions such as vaccination, early diagnosis and effective treatment.

COPD and increased susceptibility to pneumonia

It has previously been reported that the risk of developing CAP is higher in COPD patients than in the general population, in particular, in patients aged over 65 years with severe COPD (61). The use of a combination of an inhaled corticosteroid (ICS) and a long-acting beta2-adrenoceptor agonist (LABA) is recommended in the treatment of patients with severe COPD with FEV₁ <50% predicted and ≥2 exacerbations in 12 months (62). ICS therapy has been associated with an increased risk of serious pneumonia in COPD patients (63, 64). Newly diagnosed COPD patients on ICS treatment were 1.38 (95% CI: 1.31–1.45) times more likely to have a hospitalisation for pneumonia than those without current use of ICS (65). A Cochrane systematic review of seven randomised controlled trials highlighted that treatment with a corticosteroid/LABA combination of fluticasone and salmeterol increased the risk of pneumonia in COPD patients by 1.8-fold as compared with placebo (66).

The rate of pneumonia and hospital admission were reportedly higher in patients treated with fluticasone/salmeterol [rate ratio, RR of 1.73 (95% CI: 1.57–1.90) and 1.74 (95% CI: 1.56–1.94), respectively] compared to a budesonide/formoterol combination (67). But pneumonia adverse events have also been associated with the use of 160/9 and 320/9 μg/day budesonide/formoterol in 4.7% and 6.4% of COPD patients, respectively (68). As recently stated by Iannella and colleagues, “Newer studies were not able to rule out budesonide as responsible for pneumonia, as previous evidence suggested, and there is still need for evidence from head-to-head comparisons in order to better assess possible intra-class differences” (69).

What is clearer is that the dosage of ICS relates to pneumonia risk. While lower doses of fluticasone/salmeterol (500 μg/day) have been reported to pose a risk of pneumonia (70), Ernst and co-workers reported an increased RR of hospitalisation for pneumonia with higher ICS doses, i.e., fluticasone at 1,000 μg/day or more (RR of 2.25, 95% CI: 2.07–2.44) (71). As concluded by Eapen et al., “Since then all major studies clinical studies have found a direct correlation between ICS dosage and pneumonia rates in COPD” (72).

Not all of the increased risk of pneumonia in COPD patients is attributable to ICS use, and other factors likely play a role. In a recent Danish study of 179,759 hospitalisations for acute exacerbations of COPD (AECOPD), pneumonia was common in patients hospitalised with AECOPD and was associated with increased health resource utilisation in terms of length of stay (median 7 vs. 4 days) and intensive care unit admission (7.7% vs. 12.5%) (73). However, prior use of ICS differed little between non-pneumonic and pneumonic first-time AECOPD cases (73).

In addition to commonalities in risk factors for COPD and pneumonia, there is also an overlap in a number of the bacterial species that are capable of causing both conditions, in particular Streptococcus pneumoniae and Haemophilus influenzae. Corticosteroid therapy in COPD patients is believed to facilitate infection of the lower respiratory tract infection by these pathogens by suppressing the cellular and humoral arms of immunity (74, 75). However, the precise mechanisms behind the pneumonia risk in COPD patients due to ICS use have not been investigated in great detail to date. Alveolar macrophages play a central role in the host pulmonary defence against bacterial infection. Whereas phagocytosis and intracellular anti-microbiocidals, including reactive oxygen and nitrogen species, fail to kill ingested bacteria, alveolar macrophages can engage apoptosis-mediated killing to clear the bacteria (76). However, the addition of fluticasone to apoptotic cells has recently been shown to reduce the ability of murine alveolar macrophages to kill S. pneumoniae in vitro and to result in a higher pneumococcal lung burden in mice (77). For another pathogen associated with both COPD and pneumonia, Klebsiella pneumoniae, fluticasone impaired nitric oxide production by alveolar macrophages in parallel with reduced bacterial clearance and survival of mice (78). Nevertheless, our understanding of the effect of corticosteroid-based COPD therapies on the pulmonary immune defence against bacterial pneumonia is derived from a relatively small number of experimental studies and, therefore, further investigations are needed.

Immunopathology of COPD and pneumonia

The lung is the largest epithelial surface area of the body in contact with the external environment (79). Airway irritants interfere with the innate defence system by increasing mucus production, reducing mucociliary clearance, disrupting the epithelial barrier, and inhibiting the migration of immune cells (80). The release of various chemical messengers by airway epithelial cells and macrophages, triggered by airborne irritants, participates in the pathological events of COPD (81). Chemotactic factor CC-chemokine ligand 2 (CCL2) attracts monocytes, while CXC-chemokine ligand 1 (CXCL1) and CXCL8 attract neutrophils and monocytes. CXCL9, CXCL10 and CXCL11 attract T-helper 1 (TH₁) cells and type 1 cytotoxic T cells (81). Along with these inflammatory cells, matrix metalloproteinase (MMP)-9 released by macrophages and epithelial cells causes degradation of elastin (82). Small airway fibrosis results from the release of transforming growth factor-β (TGFβ) from epithelial cells and macrophages (83). Additionally, mucus hypersecretion mediated by neutrophil elastase causes airway blockage in COPD (84).

The so-called “persistently activated” CD8⁺ cytotoxic T cells may constitute a large population of lymphocytes in the airways of COPD patients (85). However, cytotoxic T cells are a major contributor in COPD progression with increased abundance in lung compartments, which correlates with the degree of airflow limitation, particularly in smokers (86). In regard to pneumonia, cytotoxic T cells primarily feature in viral forms of the disease (87).

The classical complement pathway is thought to be a major immune mechanism protecting the host against pneumococcal infections (88). Teichoic acid and peptidoglycan of pneumococci activate the alternative complement pathway (89), whereas anti-capsular antibody activates the classical pathway (90), thereby generating various mediators of inflammation. Peptidoglycan and lipoteichoic acid stimulate Toll-like receptor 2, and pneumolysin interacts with Toll-like receptor 4 to induce nuclear factor kappa B (NF-κB) (91). Rather than local proliferation, the peripheral immune response in pneumonia typically requires recruitment of multiple types of haematopoietic cells such as macrophages (92).

Inflammation is a double-edged sword whereby events that facilitate clearance of respiratory pathogens also induce a potentially injurious inflammatory response in the lung (93, 94). Lower respiratory tract infection in childhood has been linked to adverse lung activity, believed to be due to impaired lung growth following injury, direct damage to lung parenchyma, reduction in alveolar development, and disruption of bronchial epithelium and connective tissue (95). It is understood that the lung epithelium, after injury, may either activate the repair and regeneration pathways, or undergo an aberrant remodelling (96). However, a detailed examination of the specific components of long-term lung damage due to pneumonia is needed to define the pathological features that specifically increase later susceptibility to the development of COPD.

Common markers of COPD and pneumonia

Higher levels of interleukin (IL)-8, a potent neutrophil chemoattractant produced by macrophages, have been associated with elevated bacterial load as well as with CAP and hospital-acquired pneumonia (HAP) (97), and a faster decline in FEV₁ (98). Similarly, an increase in IL-8 and another neutrophil chemoattractant, CXCL5, has been observed in bronchial biopsies from severe COPD exacerbations (99). The level of IL-8, and fellow pro-inflammatory cytokine IL-6, has been observed to increase in the sputum of COPD patients with increasing number of exacerbations (100). It has been reported that three or more exacerbation episodes per year are associated with a median IL-6 level of 110 pg/mL (95% CI: 11–215) and an IL-8 level of 6,694 pg/mL of sputum (95% CI: 3120–1195) (100). However, median levels of both IL-6 and IL-8 were reported to decrease to 22 pg/mL (95% CI: 12–93) and 1,628 pg/mL (95% CI: 607–4812), respectively, among those experiencing two or fewer COPD exacerbations per year (100), although wide data confidence intervals indicate a notable level of variance. An increased level of IL-6 has been related to the need for mechanical ventilation and with early death in CAP patients (101). Likewise, Bohnet and colleagues observed a significantly higher IL-8 level in patients with pneumonia as compared to age-matched and healthy controls (p < 0.05) (102). Ferrer and co-workers reported an increased rate of pneumonia among ICS-treated COPD patients. The authors did not find significant differences in the serum level of the majority of systemic inflammatory biomarkers tested in CAP patients with or without ICS treatment, but did detect a decreased level of IL-6 (adjusted OR for each log unit 0.888, 95% CI: 0.775–1.018, p = 0.089) and TNF-α (adjusted OR for each log unit 0.506, 95% CI: 0.361–0.708, p < 0.001) among CAP subjects receiving ICS (103). In contrast, the level of biomarkers IL-6 and TNF-α were similar with or without ICS treatment during the early inflammatory response in COPD exacerbation subjects in work performed by Crisafulli and colleagues (104).

IL-8 induces the secretion of MMP-9, by neutrophils (105). Neutrophils also secrete serine proteases including neutrophil elastase, cathepsin G and proteinase-3, which are powerful stimulants of mucus production (106). MMPs might have an important protective effect following inflammation by inhibiting pulmonary fibrosis (107). MMPs are capable of degrading connective tissue matrices, and tissue inhibitors of metalloproteinase (TIMPs) are believed to function in remodelling and repair after parenchymal damage (108). A number of studies have examined a possible relationship between MMP-9 and COPD. Plasma levels of MMP-9 have been reported to be increased in α₁-antitrypsin deficiency-associated emphysema and in COPD, and to correlate negatively with FEV₁ (109, 110). Increased plasma levels of MMP-9 have been shown to negatively correlate with FEV₁ and indicate a decline in pulmonary function (111). Similarly, an increase in MMP-8 has been identified during exacerbations and airflow obstruction (112). HAP and CAP have also been associated with increased MMP activity that correlates with an increase in other inflammatory markers (109, 110, 113). MMP-9 mRNA expression in peripheral blood mononuclear cells was higher in the acute phase of CAP compared to the control group (110). Hartog and co-workers reported a 10-fold increase of MMP-8 and MMP-9 expression in mini-bronchoalveolar lavage fluid of HAP patients (114). CAP has also been associated with higher MMP-9 expression which correlated with an increase in the neutrophil count (109). This has led to the suggestion that plasma MMP-9 levels, and the molar ratio of MMP-9 to its cognate TIMP-1, could be used in the clinical evaluation of pneumonia diagnosis (115).

In terms of immunoglobulins (Igs), it has been reported by de la Torre et al. that levels of IgA, IgG and subclasses IgG1, IgG2 were significantly lower in CAP patients as compared to healthy controls (116). O'Keeffe and co-workers previously reported that total IgG, as well as IgG1 and IgG2, were lower in COPD patients, both in the presence and absence of ICS therapy, with respect to controls (117).

C-reactive protein (CRP) is often used as a clinical marker of acute systemic inflammation (118). CRP is an independent predictor of future COPD outcomes in individuals with airway obstruction (119, 120). Correlation has been observed between human serum CRP and variables including FEV₁ (r = −0.813; p < 0.01), GOLD (Global Initiative for Chronic Obstructive Lung Disease) stage of COPD (r = 0.797, p < 0.01), and smoking status (r = 0.796; p < 0.01) (121). Furthermore, CRP levels were reported to be higher in COPD patients than in controls (4.82 vs. 0.88 mg/L, p < 0.01) (121). Similarly, CRP has also been observed to increase during COPD acute exacerbations (118). A significant correlation has also been observed in CAP patients between human serum CRP level and the need for intensive care unit admission, oxygen therapy and mechanical ventilation (122).

Earlier work has determined from immunofluorescence microscopy and qPCR that cigarette smoke, the primary risk factor for COPD, induces Platelet Activating Factor receptor (PAFr) expression in lung epithelial cells in both humans and mice (123). PAFr is expressed on the surface of the majority of cells of the innate immune system (e.g., macrophages, polymorphonuclear leukocytes, mast cells) and is an important mediator of the inflammatory response (123). But of particular interest is its elevated expression in bronchial epithelial cells of COPD-smokers (p < 0.005) and COPD-ex-smokers (p < 0.002) compared to smokers with normal lung function (124). Upregulated PAFr correlates with higher levels of adhesion of both S. pneumoniae and non-typeable H. influenzae (NTHi) (125), the main bacterial drivers of acute exacerbations of COPD (126, 128). PAFr is used as a receptor for bacterial adhesion to the epithelium (123) and is, therefore, believed to facilitate colonisation of the airways in COPD patients by pneumococci and NTHi. PAFr may also play a role in pneumonia with PAFr^−/− knock-out mice exhibiting significantly lower susceptibility to pneumococcal pneumonia as measured by decreased bacterial outgrowth in the lungs, diminished dissemination and prolonged survival (129, 130). Higher levels of PAFr expression have been observed in mouse lungs co-infected with S. pneumoniae and influenza virus (131). Therefore, heightened expression of PAFr due to smoking could in turn increase the risk of S. pneumoniae infection and pneumonia development in COPD patients. Immune-modulatory effects of PAFr in host defence include stimulation of migration and degranulation of granulocytes, monocytes and macrophages, and the release of cytokines and toxic oxygen metabolites (129). A significant increase in PAFr expression observed in lung biopsies of aged mice is believed to support a link between age-associated inflammation and severe pneumonia (132).

Table 1 illustrates mediators of the immune response for bacterial pneumonia, COPD and AECOPD. It is apparent that several effectors of lung inflammation have been found to be shared between bacterial pneumonia, COPD and AECOPD, which suggests the involvement of common inflammatory pathways. However, differences also exist whereby some immune mediators have to date not been reported to participate in all three diseases. For example, CXCL-8 has been reported at higher levels in COPD and AECOPD (81) but not in bacterial pneumonia patients. The potential of immune mediators as markers to ascertain disease severity, and to distinguish between illnesses, warrants investigation.

Table 1. Inflammatory mediators in bacterial pneumonia, chronic obstructive pulmonary disease and acute exacerbations of chronic obstructive pulmonary disease.

	Bacterial pneumonia	COPD	AECOPD
Immune cells	APC (94)	Macrophages (133)	Macrophages (133)
	Neutrophils (94)	Neutrophils (134)	Neutrophils (134)
	CD4^+ T cells (94)	CD8^+ T cells (136)	Eosinophils (136)
Cytokines and chemokines	IL-1 (137)	IL-1 (139)
	IL-6 (138)	IL-6 (139)	IL-6 (100)
	IL-8 (138)	IL-8 (139)	IL-8 (99)
	TNF-α (137)	TGF-β (140)	TNF-α (141)
		CXCL-1 (81)	CXCL-8 (81)
		CXCL-8 (81)
		CCL-2 (81)
Matrix metalloproteinases	MMP-2 (142)	MMP-8 (112)	MMP-8 (143)
	MMP-8 (114)	MMP-9 (112)
	MMP-9 (142)	MMP-12 (105)

AE, acute exacerbations; COPD, chronic obstructive pulmonary disease; APC, antigen presenting cells; CCL, CC-chemokine ligand; CXCL, CXC-chemokine ligand; IL, interleukin; MMP, matrix metalloproteinase; TNF, tumour necrosis factor. References are given in parentheses. Immune mediators with a known involvement in one of the three diseases are listed. The non-listing of specific immune mediators for one or more of the diseases is based on absence of published data acquired from the literature.

Conclusions and future perspectives

It has become increasingly apparent that an overlap exists in the epidemiology between non-communicable COPD and infectious respiratory disease including pneumonia. A growing aged population and other common risk factors, such as smoking, air pollution and poor social determinants of health, contribute to the co-morbidities seen with respect to the two diseases. Previous episodes of pneumonia significantly increase the likelihood of COPD in later life such that earlier spirometric screening may be warranted in individuals with a known history of pneumonia. Further research is needed to define the specific immunopathological events that occur during pneumonia and lead to long-term lung damage and increased COPD susceptibility. A number of inflammatory markers such as matrix metalloproteinases are common to both COPD and pneumonia such that they may offer potential new ways to gauge the severity and prognoses of these diseases. In addition, markers that can more accurately distinguish between the two conditions are likely to aid in improving their respective diagnoses. Regarding treatment, investigations are needed into COPD therapeutics, which avert a heightened susceptibility to pneumonia and other infectious diseases due to immunosuppression. While COPD by definition is a chronic lung disease, the realisation of its interaction with a number of acute infectious diseases, including pneumonia and tuberculosis, warrant quantitation of the contribution that acute communicable illnesses make to the overall burden of COPD and predictive modelling of this impact into the future. New research is now needed across a number of scientific disciplines to raise our level of understanding of the associated epidemiologies of COPD and pneumonia, and underpin the development of new interventions for their prevention or management.

Funding

This work was funded by a Research Development Grant from the School of Medicine, University of Tasmania awarded to R.F.O., and a Faculty of Health/School of Medicine Ph.D. Scholarship from the University of Tasmania awarded to S.S.G.