Anti-TNFα Therapy in Inflammatory Lung Diseases bn

Abstract

Increased levels of tumor necrosis factor (TNF) α have been linked to a number of pulmonary inflammatory diseases including asthma, chronic obstructive pulmonary disease (COPD), acute lung injury (ALI)/acute respiratory distress syndrome (ARDS), sarcoidosis, and interstitial pulmonary fibrosis (IPF). TNFα plays multiple roles in disease pathology by inducing an accumulation of inflammatory cells, stimulating the generation of inflammatory mediators, and causing oxidative and nitrosative stress, airway hyperresponsiveness and tissue remodeling. TNF-targeting biologics, therefore, present a potentially highly efficacious treatment option. This review summarizes current knowledge on the role of TNFα in pulmonary disease pathologies, with a focus on the therapeutic potential of TNFα-targeting agents in treating inflammatory lung diseases.

1. Introduction

Proinflammatory cytokines including tumor necrosis factor (TNF) α, interleukin (IL)-1, IL-6 and are key modulators of inflammation that initiate and drive many pulmonary pathologies and diseases. TNFα is especially important as its actions are numerous and quite diverse. These include stimulating leukocyte accumulation, proliferation and differentiation at the sites of injury and infection as well as oxidative stress, necrosis, apoptosis, angiogenesis, and tissue remodeling (Figure1) (Aggarwal, 2003; Mukhopadhyay et al., 2006). TNFα is primarily produced by macrophages and monocytes (Aggarwal et al., 2012; Suzuki et al., 2013). It is synthesized as a cell surface bound precursor, known as transmembrane TNFα (tmTNFα), a homotrimer of 26 KDa subunits. tmTNFα is cleaved by TNFα-converting enzyme (TACE) to a biologically active soluble form (sTNFα), a homotrimer of 17 KDa subunits (Horiuchi et al., 2010). TNFα is a hormone-like peptide that can act locally in an autocrine or paracrine manner or at distant sites by entering the bloodstream. Many inflammatory mediators stimulate the production of TNFα including bacterially-derived lipopolysaccharide (LPS), interleukin (IL)-1, IL-2, interferon-γ, granulocyte macrophage colony stimulating factor and platelet derived growth factor, as well as TNFα itself (Barbara et al., 1996; Semenzato, 1990; Suzuki et al., 2013).

Fig. 1.

Multiple roles of TNFα in the pathophysiology of inflammatory diseases.

The multiple activities of TNFα are mediated via binding to cell surface receptors. Two types of structurally distinct TNF receptors have been identified: Type 1 or TNFRI (a 55-kDa protein) and Type 2 or TNFRII (a 75 kDa protein) (Aggarwal, 2003). TNFRI contains a death domain and is expressed on most cell types. In contrast, TNFRII expression is limited mainly to immune cells, endothelial cells, and nerve cells (Aggarwal et al., 2012). Although the presence of a TNF receptor is a prerequisite for biological responses, there doesn’t appear any relationship between the number of receptors and the magnitude of responses to TNFα (Semenzato, 1990). Both TNFRI and TNFRII bind TNFα, as well as lymphotoxin, a cytotoxic protein secreted by lymphoid cells, with approximately equal affinity. TNFα binding to its receptors initiates a signaling cascade involving mitogen-activated protein kinase and c-Jun Nterminal kinase, culminating in activation of the transcription factors, nuclear factor-kappa B (NF-κB) and activated protein 1 (AP-1) (Garg and Aggarwal, 2002). Activation of these intracellular signals is important in TNFα-mediated apoptosis, differentiation and proliferation, and production of proinflammatory proteins including IL-6, IL-8, IL-18, chemokines, inducible nitric oxide synthase, cyclooxygenase and lipoxygenase enzymes, as well as TNF, itself (Aggarwal et al., 2012). Downregulation of inflammation is associated with shedding of the extracellular domain of TNFRs and decreases in TNFα activity.

2. Biological activities of TNFα

TNFα is identical to cachectin, a peptide recognized for its ability to induce fever and wasting (Aggarwal et al., 2012; Clark, 2007). As a promoter of cachexia, TNFα inhibits the synthesis of lipoprotein lipase, an enzyme that cleaves fatty acids from triglycerides (Clark, 2007). TNFα is also a master regulator of inflammation which is an important component of its pathogenic actions. TNFα is a neutrophil and eosinophil chemoattractant, and it stimulates the production of macrophage chemokines, such as CCL2. It also promotes inflammation by upregulating adhesion molecules important in leukocyte trafficking to inflammatory sites, including intracellular leukocyte adhesion molecule, endothelial leukocyte adhesion molecule-1, and vascular cell adhesion molecule-1(Kelly et al., 2007). In addition, TNFα stimulates the release of eicosanoids and platelet activating factor (Camussi et al., 1991; Kelly et al., 2007), which contribute to inflammation by promoting vasodilatation, and leukocyte adhesion and migration to sites of injury (Camussi et al., 1991; Michel et al., 2014; Semenzato, 1990). Oxidative and nitrosative stress are hallmarks of inflammatory diseases (Laskin et al., 2010; Thomas, 2001). TNFα is thought to be a major inducer of oxidative and nitrosative stress in inflammatory cells (Blaser et al., 2016; Rahman, 2000). This leads to the activation of redox sensitive transcription factors including NF-κB and AP-1 which upregulate inflammatory gene expression. TNFα also depletes intracellular glutathione which contributes to its prooxidant effects (Obrador et al., 1998; Rahman, 2000).

Increases in TNFα are associated with cytotoxicity, a response mediated by its binding to the p55 receptor (TNFRI). This results in recruitment of TNFR-associated death domain and Fas associated death domain and caspases, key pro-apoptotic enzymes (Aggarwal et al., 2012). TNFα can also activate caspases and promote apoptosis by stimulating mitochondria to release reactive oxygen species, cytochrome c and Bax, and by activating sphingomyelinases (Aggarwal et al., 2012; Garcia-Ruiz et al., 2003). Both soluble and transmembrane TNFα are equally effective in inducing apoptosis (Klimp et al., 2002).

TNFα is also a potent mitogen stimulating proliferation of epithelial cells (Lu et al., 1997). This is thought to be due in part to activation of the transcription factor AP-1 and upregulation of cyclin-D1 (Mukhopadhyay et al., 2009; Rahman, 2000). TNFα-mediated proliferation is thought to contribute to epithelial thickening and pulmonary fibrosis (Allen and Spiteri, 2002; Sasaki et al., 2000). TNFα also promotes fibrosis by inducing focal accumulation of fibroblasts and collagen deposition. It upregulates expression of matrix metalloproteinases and transforming growth factor (TGF) β, which are involved in tissue remodeling and fibrogenesis (Oikonomou et al., 2006; Piguet, 1990; Sasaki et al., 2000; Sullivan et al., 2005).

3. TNFα inhibitors

The recognition that TNFα plays a key role in inflammatory diseases and pathologies has led to the development of a number of drugs and biologics that target TNFα. Both chimeric mouse/humanized monoclonal anti-TNFα antibody (infliximab) and fully human monoclonal anti-TNFα antibodies (golimumab and adalimumab) have been developed; these bind to and inactivate membrane bound and soluble TNFα. Etanercept, a soluble fusion protein consisting of two p75 TNF receptors attached to an Fc fragment of human IgG1, is also available which mainly binds and inactivates sTNFα. Etanercept has better avidity and affinity for sTNFα than tmTNFα, and does not induce complement activation. In contrast, monoclonal anti-TNFα antibodies (i.e., infliximab, golimumab and adalimumab) bind to both monomeric and trimeric soluble and transmembrane forms of TNFα, which can activate complement cascade resulting in cytotoxicity (Liang et al., 2013). Despite differences in their mode of administration, efficacy and safety profile, these TNFα targeting agents have effectively been used to treat patients with TNFα-associated diseases such as Crohn’s disease, psoriatic arthritis, rheumatoid arthritis and ankylosing spondylitis with minimal toxicity (Hasegawa et al., 2001; Raychaudhuri and Raychaudhuri, 2009).

4. Role of TNFα in pulmonary diseases

Macrophages are the major source of TNFα in the lung; however, epithelial cells, eosinophils, and mast cells also have the capacity to release TNFα upon activation (Finotto et al., 1994; Gosset et al., 1991; Herfs et al., 2012; Khair et al., 1994). Increased levels of TNFα have been linked to a number of pulmonary inflammatory diseases including asthma, chronic obstructive pulmonary disease (COPD), acute lung injury (ALI)/acute respiratory distress syndrome (ARDS), sarcoidosis, and interstitial pulmonary fibrosis (IPF). Each of these pathologies is characterized by airway injury, inflammation, and bronchial and parenchymal remodeling. TNFα contributes to these inflammatory diseases by recruiting inflammatory cells, stimulating the generation of inflammatory mediators, increasing oxidative and nitrosative stress, and inducing airway hyperresponsiveness (Anticevich et al., 1995; Choi et al., 2005; Herfs et al., 2012; Hughes et al., 1995; Shah et al., 1995). In this review, we describe the role of TNFα in pulmonary disease pathologies, with a focus on the therapeutic potential of TNFα-targeting agents in treating pulmonary diseases (Table 1).

Table 1.

Summary of Studies of TNFα Targeting Agents in Inflammatory Lung Diseases

Disease/ pathology	Treatment	Rodent/ human	References	Outcome
Asthma	Etanercept	human	Berry et al., 2006; Holgate et al., 2011; Howarth et al., 2005; Morjaria et al., 2008	Improvement in asthma symptoms, lung function and quality of life
		mouse	Dejager et al., 2015	Reversal of GC insensitivity
	Anti-TNF antibody	mouse	Busse et al., 2009; Kim et al., 2006	Reduced lung inflammation and mucus cell metaplasia
	Infliximab	rat	Cai et al., 2011	Reduced airway inflammation and hyperreactivity
		mouse	Deveci et al., 2008	Reduced inflammatory cell accumulation and decreased cytokine and chemokine release
		human	Taille et al., 2013	Improved asthma control and reduced frequency of exacerbations
	Adalimumab	mouse	Catal et al., 2014	Attenuation of lung damage
	Golimumab	human	Wenzel et al., 2009	No significant effect on forced expiratory volume in 1 sec. or asthma exacerbations
	TNFα antisense	mouse	Luo et al., 2012	Reduced inflammatory cell infiltration and mucus secretion
COPD	Etanercept and infliximab	human	Suissa et al., 2008	Reduced COPD exacerbations following etanercept treatment
	Etanercept	human	Aaron et al., 2013	Reduced acute COPD exacerbations; equally effective as prednisone
	Infliximab	human	Dentener et al., 2008; Rennard et al., 2007; van der Vaart et al., 2005	No significant effect on quality of life, inflammation or disease exacerbations
ALI/ARDS	Anti-TNFα monoclonal antibody	human	Abraham et al., 1998; Abraham et al., 1995	No significant effect on mortality in septic shock patients
		mouse	Bertok et al., 2012	p55 receptor-specific domain antibody but not anti-TNFα antibody inhibited lung injury, edema and inflammation
		pig	Mullen et al., 1993	Combined ibuprofen and anti-TNFα antibody protect against ALI
	p55 TNFR fusion protein	human	Abraham et al., 1997	Reduction in mortality in patients with severe sepsis
	Etanercept	rat	Guthmann et al., 2009	Hyperoxia-induced lung injury and BAL cell content inhibited
Pulmonary sarcoidosis	Infliximab	human	Baughman et al., 2006; Chebib et al., 2014; Rossman et al., 2006; Russell et al., 2013; Sweiss et al., 2005	Improvement in FVC and pulmonary disease
	Adalimumab	human	Callejas-Rubio et al., 2006; Milman et al., 2012; Sweiss et al., 2014	Decrease in dyspnea, cough and disease pathology. Improvement in FVC and Borg dyspnea score
	Golimumab	human	Judson et al., 2014	No significant effect
	Etanercept	human	Utz et al., 2003	Terminated early due to treatment failure
IPF	Etanercept	human	Raghu et al., 2008	Reduction in disease progression
	Pirfenidone	human	Azuma et al., 2011; King et al., 2014; Maher, 2010; Takeda et al., 2014; Taniguchi et al., 2011	Reduction in vital capacity decline, disease progression and increase in progression-free survival, suppression of cough and dyspnea
Experimental models of lung injury	Anti-TNFα antibody	mouse	Bhalla et al., 2002; Piguet et al., 1990; Shvedova et al., 1996	Inhibition of ozone, silica or cotton dust induced lung injury, decreased inflammatory cytokine release, inflammation and fibrosis
		rat	Malaviya et al., 2015	Marked inhibition in vesicant-induced lung injury, oxidative stress, numbers of cytotoxic, proinflammatory macrophages and fibrosis
	Infliximab	rat	Zhang et al., 2011	Reduced airway inflammation and improvement in cigarette smoke-induced histopathology; protection from emphysema
	Pentoxifylline	rat	Sunil et al., 2014	Inhibition of vesicant-induced lung injury, inflammation and oxidative stress

Asthma

Asthma is a chronic lung inflammatory disease characterized by persistent eosinophilic inflammation, airway hyperreactivity, mucus secretion and reversible airway obstruction. Increased levels of TNFα have been described in the airways of patients with severe asthma (Bradding et al., 1994; Howarth et al., 2005; Noguchi et al., 2002). Alveolar macrophages and peripheral blood monocytes isolated from patients with asthma produce increased amounts of TNFα and TACE, and express higher levels of TNFα receptors (Berry et al., 2006). In patients with allergic asthma, high sputum TNFα levels are observed within 24 h of allergen challenge (Keatings et al., 1997); TNFα has also been reported to increase during asthma exacerbations or after allergen challenge in patients suffering from asthma (Thomas, 2001). In healthy individuals, inhalation of recombinant TNFα increases airway hyperresponsiveness and sputum neutrophils (Thomas, 2001; Thomas and Heywood, 2002). This is thought to be due to a direct effect of TNFα on airway smooth muscle cells and release of leukotrienes (Anticevich et al., 1995; Choi et al., 2005). Blood monocytes and alveolar macrophages from asthmatic subjects have been reported to produce increased amounts of TNFα, as well as IL-8 and granulocyte macrophage colony stimulating factor, following LPS stimulation, when compared to normal subjects, suggesting selective augmentation of cytokine production (Hallsworth et al., 1994).

Animal studies have confirmed that TNFα plays a role in the pathophysiology of asthma and bronchial hyperresponsiveness. Thus, exposure of rats to endotoxin upregulates TNFα production by bronchial epithelial cells and alveolar macrophages, a response associated with bronchial hyperresponsiveness (Ermert et al., 2003; Kips et al., 1992). Ovalbumin-induced asthma in rats is also associated with increases in TNFα in serum and lung (Cai et al., 2011). Additionally, mice lacking TNFα or TNFR are protected from lung inflammation, mucus secretion and late airway hyperresponsiveness in an ovalbumin-induced model of asthma, a response mimicked by administration of anti-TNFα antibody to wild type animals (Busse et al., 2009; Choi et al., 2005).

The efficacy of antagonizing TNFα in asthma has been evaluated both in animal models and in humans. In animal models of asthma, the results of anti-TNFα antibody treatment have been encouraging. In ovalbumin challenged rodents infliximab inhibited airway smooth muscle hyperreactivity and inflammatory damage (Cai et al., 2011; Deveci et al., 2008). Monoclonal anti-TNFα antibody treatment of mice also mitigated house dust induced allergic inflammation, including increases in eosinophils, lymphocytes, macrophages, and neutrophils in bronchoalveolar lavage (BAL), and histopathological changes in the lung (Kim et al., 2006). Methacholine-induced airway hyperreactivity was also inhibited by infliximab. Similarly, in a murine model of acute asthma, adalimumab therapy reduced lung inflammation and inflammatory cell infiltration (Catal et al., 2014). Local administration of TNFα antisense nucleotide has also been reported to suppress allergic inflammation in mice by reducing release of TNFα and other Th2 cytokines, mucus secretion and inflammatory cell influx (Luo et al., 2012). Airway mucus cell metaplasia and hyperresponsiveness were also inhibited by monoclonal anti-TNF antibody in mice (Busse et al., 2009). In another study, etanercept was found to restore the therapeutic efficacy of glucocorticoids (GC) in an ovalbumin induced GC-insensitive mouse models of airway hyperinflammation (Dejager et al., 2015).

In humans with asthma the efficacy of antagonizing TNFα is less clear. Howarth et al. (2005) reported that treatment of patients with severe asthma with etanercept was associated with improvement in asthma symptoms, lung function and bronchial hyperresponsiveness. Additionally, in patients with corticosteroid refractory asthma, both etanercept and infliximab were reported to improve asthma, lung inflammation, lung function and quality of life and to reduce the frequency of asthma exacerbations and hospitalizations (Berry et al., 2006; Morjaria et al., 2008; Taille et al., 2013). Infliximab was also observed to reduce asthma exacerbations in patients with moderate asthma (Erin et al., 2006). Conversely, etanercept had no effect in patients with moderate-to-severe persistent asthma (Holgate et al., 2011). Similarly, in a large multicenter trial of patients with severe asthma, golimumab treatment for 12 months failed to demonstrate a favorable risk-benefit profile (Wenzel et al., 2009). Together these findings suggest that TNFα targeting may particularly be an effective strategy to treat severe or refractory asthma.

COPD

COPD is an inflammatory disease characterized by chronic progressive airway obstruction due to prominent localization of inflammatory cells including neutrophils, macrophages, T cells and mast cells in the airways and thickening of the airway walls (Aoshiba and Nagai, 2004). In transgenic mice overexpressing TNFα, progressive histopathologic changes are observed including chronic inflammation, thickened interstitium, alveolar air space enlargement, septal wall destruction and bronchiolitis, consistent with emphysema, and increases in collagen; these mice also develop fibrosis (Fujita et al., 2001; Lundblad et al., 2005; Miyazaki et al., 1995). Alveolar macrophages from cigarette smokers and patients with COPD release increased quantities of TNFα (Chung, 2005; Lim et al., 2000). In addition, higher levels of TNFα and sTNFRII are observed in BAL from chronic smokers and sputum from patients with COPD. Moreover, levels of TNFα and sTNFRII increase further during COPD exacerbations (Chung, 2005; Keatings et al., 1996; Woodruff et al., 2014). Sputum levels of sTNFRII have been reported to be inversely related to forced expiratory volume 1 in patients with COPD, and serve as a prognostic biomarker for the risk of exacerbation (Takabatake et al., 2000; Vernooy et al., 2002; Woodruff et al., 2014). TNFα is thought to contribute to COPD by upregulating expression of adhesion molecules resulting in inflammatory cell influx, and increased production of matrix metalloproteinases and tenascin, which promote tissue damage and remodeling (Chung, 2001). TNFα may also contribute to cachexia in COPD patients via activation of the transcription factor, NF-κB (Wagner, 2008).

Despite evidence suggesting a role of TNFα in COPD disease pathology, TNFα antagonists have shown only limited clinical efficacy. In patients with RA and COPD, etanercept, but not infliximab, reduced COPD hospitalizations (Suissa et al., 2008). Conversely, in a randomized double blind controlled trial, etanercept had no beneficial effect for treatment of acute exacerbations of COPD (Aaron et al., 2013). Similarly, in controlled studies, infliximab had no beneficial effect in patients with mild, moderate or severe COPD (Rennard et al., 2007; van der Vaart et al., 2005). Infliximab has been reported to have minor effects on systemic inflammation in cachectic patients with COPD, but local inflammation was unaffected (Dentener et al., 2008). The reason for the lack of efficacy of TNF antagonists in COPD is unclear. COPD is a heterogeneous inflammatory disease mediated by multiple cytokines; it is possible that blocking just one cytokine is not sufficient to control the disease pathology (Barnes, 2007; Matera et al., 2010).

ALI/ARDS

ALI and ARDS are life-threatening manifestations of an inflammatory response of the lung to various insults, and are characterized by severe hypoxemia, diffuse infiltration in the chest X-ray, and a substantial reduction in pulmonary compliance (Matuschak and Lechner, 2010). The early inflammatory phase of ALI, characterized by alveolar epithelial and endothelial barrier dysfunction, hemorrhage and protein rich pulmonary edema is followed by a proliferative phase involving alveolar epithelial cell proliferation, interstitial fibrosis and air space obliteration (Tomashefski, 2000; Vasudevan et al., 2004). Later in the pathology, fibrosis and emphysema are observed, along with loss of normal lung structure. ARDS represents a more severe and late phase of ALI.

Increased levels of TNFα have been detected in BAL, serum and epithelial lining fluid from patients with ALI and ARDS (Antonelli et al., 1994; Bauer et al., 2000; Hamacher et al., 2002; Li et al., 2010; Reper and Heijmans, 2015; Roten et al., 1991; Singh et al., 2015; Suter et al., 1992; Vaillant et al., 1996). Higher serum levels of TNFα during early stages of ALI are associated with increased mortality at 2–3 months (Makabe et al., 2012). Higher levels of TNFα and sTNFRII in BAL are observed in patients in early stage ALI, when compared to patients with late phase ARDS (Hamacher et al., 2002). Pulmonary microvascular endothelial cells isolated from ARDS patients also express increased levels of TNFRII relative to microvascular endothelial cells from control patients (Grau et al., 1996). A multicenter study (Parsons et al., 2005) demonstrated that increased baseline plasma sTNFRI and sTNFRII levels are strongly associated with mortality and morbidity in ALI patients. Moreover, genetic variations in TNF receptor-associated factor 6 gene are linked to susceptibility to ALI in patients with sepsis (Song et al., 2012). In vitro, BAL from early stage ALI patients induced endothelial cell cytotoxicity, suggesting a role of TNFα in endothelial-interstitial barrier dysfunction (Hamacher et al., 2002).

In animal models of ALI or ARDS induced by endotoxin, mechanical ventilation, or extracorporeal circulation, TNFα has been implicated in disease pathogenesis (Kao et al., 2006; Li et al., 2013; Matuschak and Lechner, 2010). As observed in patients with ALI, TNFα accumulates rapidly in large quantities in the lung after injury in animals, and is considered an initiating cytokine in early disease pathology (Li et al., 2013). Pretreatment of rats with etanercept protected against hyperoxia-induced ALI, which is characterized by increases in TNFα and TNFRI, and disruption of alveolar-epithelial barrier functions (Guthmann et al., 2009). Monoclonal anti-TNFα antibody alone, or in combination with ibuprofen has also been reported to attenuate ALI in pigs (Mullen et al., 1993). Mice lacking TNFRI were protected from ventilation or acid-induced ALI, whereas TNFRII knockout mice developed pulmonary edema (Maniatis et al., 2012; Wilson et al., 2007). Similarly, inhibition of TNFR p55 using a specific antibody mitigated ventilation, or endotoxin plus ventilation-induced ALI, while anti-TNFα antibody was ineffective (Bertok et al., 2012). Consistent with these findings, clinical trials of monoclonal anti-TNFα antibody in patients with sepsis-induced ALI did not improve survival (Abraham et al., 1998; Abraham et al., 1995), whereas p55 TNFR fusion protein treatment of patients with severe septic shock reduced mortality (Abraham et al., 1997). These studies suggest that divergent effects of TNFα binding to p55 and p75 receptors. Thus, while ligand activation of the p55 receptor promotes lung inflammation, activation of the p75 receptor is protective; this difference may underlie the lack of efficacy of non-selective anti-TNFα treatment in ALI.

Pulmonary sarcoidosis

Pulmonary sarcoidosis is a disease of unknown etiology characterized by the presence of granulomatous inflammation which can progress to fibrosis (Amin et al., 2014; Crommelin et al., 2014). Increased numbers of CD4⁺ T cells and inflammatory macrophages are present in the lungs of sarcoidosis patients which may contribute to disease progression (Allen et al., 1998; Oswald-Richter et al., 2013). Lung fibrosis in sarcoidosis is an irreversible process which, in combination with ongoing inflammation, leads to bronchial distortion, cystic changes and loss of normal lung architecture and function (Mornex et al., 1994; Rozy et al., 2006). TNFα, released by alveolar macrophages, is known to play a role in promoting inflammation and Th1 driven granuloma formation and propagation (Bachwich et al., 1986; Baughman et al., 1990; Nunes et al., 2005). Increased levels of TNFα are observed in exhaled breath condensate and BAL from patients with pulmonary sarcoidosis; moreover, macrophages isolated from patients with sarcoidosis express relatively greater levels of TACE, and release increased amounts of TNFα (Baughman et al., 1990; Rozy et al., 2006). Plasma levels of TNFα are also increased in patients with sarcoidosis (Baydur et al., 2011).

Pentoxifylline, a methylxanthine phosphodiesterase inhibitor known to block TNFα production (Fernandes et al., 2008), has been shown to downregulate spontaneous or LPS-induced TNFα release from alveolar macrophages isolated from sarcoidosis patients (Tong et al., 2003). Inhibition of TNFα using infliximab or adalimumab significantly improved symptoms in patients with chronic sarcoidosis (Baughman et al., 2006; Callejas-Rubio et al., 2008; Chebib et al., 2014; Crommelin et al., 2014; Rossman et al., 2006; Russell et al., 2013; Sweiss et al., 2005); specifically, infliximab improved forced vital capacity (FVC) after 24 weeks of therapy (Baughman et al., 2006). Results with humanized anti-TNF antibody in sarcoidosis have been mixed. In an open label, single center study in patients with refractory sarcoidosis, treatment with adalimumab stabilized or improved FVC, 6-minute walk test distance of 50 m or greater, Borg dyspnea score and Physician’s and Patient’s Global Assessment (Sweiss et al., 2014). Improvement in overall health status compared to baseline was noted after 24 weeks, a response which persisted for at least 52 weeks. A reduction in disease severity following adalimumab treatment has also been described in patients with recalcitrant sarcoidosis or prednisone- and methotrexate-resistant sarcoidosis (Callejas-Rubio et al., 2006; Milman et al., 2012). In contrast, in a randomized, double-blind, placebo-controlled trial of pulmonary sarcoidosis, golimumab failed to improve FVC at week 16 or 28 compared with placebo (Judson et al., 2014). Treatment with etanercept was also ineffective in treating pulmonary sarcoidosis (Utz et al., 2003).

Idiopathic pulmonary fibrosis

Idiopathic pulmonary fibrosis is a chronic, progressive, life threatening lung disease of unknown etiology (Gross and Hunninghake, 2001). It is clinically characterized by a progressive decline in pulmonary function (Raghu et al., 2008). Increased levels of TNFα have been observed experimentally in the lungs of animals with pulmonary fibrosis and patients with IPF (Altintas et al., 2016; Lopez-de la Mora et al., 2015; Losa Garcia et al., 1999; Lozo Vukovac et al., 2014; Pan et al., 1996; Riha et al., 2004; Schupp et al., 2015; Vaillant et al., 1996; Zhang et al., 1993; Ziegenhagen et al., 1998). Lung tissue and alveolar macrophages isolated from patients with IPF release increased amounts of TNFα and sTNF receptors compared to healthy subjects (Cu et al., 2009; Piguet et al., 1993; Zhang et al., 1993). Additionally, in rodents, targeting TNFα attenuates pulmonary fibrosis suggesting role of TNFα in fibrogenesis (Malaviya et al., 2015; Piguet et al., 1993; Piguet and Vesin, 1994; Sunil et al., 2014; Thrall et al., 1997; Zhang et al., 1993).

Raghu et al. (2008) investigated the efficacy of etanercept in a randomized, prospective, double-blind, placebo-controlled, multicenter exploratory trial in patients with clinically progressive IPF. Although etanercept had no significant effect on FVC and lung diffusing capacity, it did decrease the rate of disease progression. Similarly, in clinical trials, pirfenidone, a non-peptide synthetic molecule with TNFα inhibitory activity, slowed disease progression, improved survival, and attenuated decreases in FVC in IPF patients (Azuma et al., 2011; King et al., 2014; Maher, 2010; Takeda et al., 2014; Taniguchi et al., 2011). Partial pressure of arterial oxygen level at rest and exercise capacity, as evaluated by the 6-min walk test were also improved following pirfenidone treatment (Hagmeyer et al., 2016; Miyamoto et al., 2016). In addition to inhibiting TNFα, pirfenidone reduced lipid peroxidation and oxidative stress, and suppressed TGFβ production (Salazar-Montes et al., 2008). The improved efficacy of pirfenidone in IPF patients, when compared to etanercept, may be attributed to these additional actions.

Experimental models of acute lung injury and fibrosis

Exposure of humans and experimental animals to pulmonary irritants including ozone, particulate matter, cigarette smoke, silica, bleomycin, chlorine or mustard vesicants induces an acute inflammatory response in the lung characterized by multifocal inflammatory lesions, macrophage accumulation, perivascular and peribronchial edema, interstitial thickening, cytotoxicity, bronchiectasis and bronchiolization of alveolar walls which in long term may lead to alterations in lung function, tissue remodeling and pulmonary fibrosis (Churg et al., 2009; Fakhrzadeh et al., 2004; Padilla-Carlin et al., 2011; Pendino et al., 1995; Razavi et al., 2013; Weinberger et al., 2011; Wollin et al., 2014; Yadav et al., 2010). Evidence suggests that macrophage-derived TNFα plays role in these pathogenic responses (Fakhrzadeh et al., 2008; Gossart et al., 1996; Lim et al., 2000; Malaviya et al., 2010; Michael et al., 2013; Weinberger et al., 2011). Both macrophages and epithelial cells release TNFα after exposure to ozone, mustards, silica or particulate matter (Barrett et al., 1999; Karacsonyi et al., 2009; Laskin et al., 2003; Michael et al., 2013; Osterlund et al., 2005; Ovrevik et al., 2009; Rusznak et al., 1996). Silica-induced TNFα expression has been shown to persist for more than 70 days in the lungs of mice (Piguet et al., 1990). In vitro TNF induces binding of particulates to rat tracheal explants (Xie et al., 2000). In rodents, this is associated with an increase in TNFRI signaling in the lung; TNFRII expression is also upregulated (Cho et al., 2007; Ortiz et al., 1999). Consistent with a role of TNFα in lung injury and disease pathology are findings that TNFRI^−/−, TNFRII^−/−, or TNFRI/II^−/− mice are protected from ozone, silica or vesicant-induced lung injury and fibrosis (Cho et al., 2001; Laskin et al., 1998; Ortiz et al., 2001; Pryhuber et al., 2003; Sunil et al., 2011). Similarly, treatment of rodents with pentoxifylline or anti-TNFα antibody attenuates histopathological alterations in the lung, inflammatory cytokine release, alveolar cell apoptosis and the development of emphysema following exposure to various pulmonary toxicants (Bhalla et al., 2002; Malaviya et al., 2015; Shvedova et al., 1996; Sunil et al., 2014; Zhang et al., 2011). Damage to the alveolar-epithelial barrier, measured by increases in BAL protein and cell content following mustard exposure, along with expression of the oxidative stress markers, heme oxygenase-1 and lipocalin-2, is also reduced by pharmacologic inhibition of TNFα, (Malaviya et al., 2015; Sunil et al., 2014). Treatment of rats with anti-TNFα antibody also reduces mustard-induced increases in expression of the profibrotic mediator, TGFβ. This is associated with a marked inhibition of mustard-induced collagen deposition in the lung and fibrosis (Malaviya et al., 2015). Similar findings have been described in a silica-induced lung injury model (Piguet et al., 1990). In murine lung epithelial cells, anti-TNFα antibody inhibits silica-induced chemokine release and oxidative stress (Barrett et al., 1999). Taken together, these findings provide strong support for a role of TNFα in pulmonary injury and fibrosis induced by environmental toxicants and chemical threat agents.

5. Summary and conclusions

TNFα is a key mediator of local damage and inflammation in the lung. Several specific TNFα antagonists, including etanercept, infliximab, adalimumab and golimumab are currently used clinically for treatment of immune-inflammatory diseases. Although these agents are potent neutralizers of TNFα bioactivity, their efficacy varies in different pulmonary diseases. Whereas etanercept is more effective in reducing severe asthma or COPD, infliximab and adalimumab are efficacious in treating sarcoidosis. Fundamental differences in the molecular structure, dosing method and schedule, binding characteristics, and mode of action of TNF targeting agents are potentially responsible for the observed variability in clinical responses of these agents. Impairment of mechanical barriers of the lung may also play role. In injured lung, disease pathology is associated with increases in epithelial permeability, squamous metaplasia, increases in goblet cells, inflammatory cells and protein rich pulmonary edema. Together, these changes may hinder overall drug availability. Differences in expression of other inflammatory molecules or heterogeneity of the disease may also play a role. Nonetheless, significant therapeutic responses of TNFα neutralizing agents in lung injury and inflammation are intriguing. Further research on the role of TNFα in acute and chronic pulmonary diseases may help to develop successful treatment strategies using TNF targeting agents.

Abbreviations

TNFα

tumor necrosis factor

tmTNFα

transmembrane TNFα

sTNFα

soluble TNFα

TACE

TNFα-converting enzyme

LPS

lipopolysaccharide

IL

interleukin

NF-κB

nuclear factor-kappa B

AP-1

activated protein 1

TGFβ

transforming growth factor β

COPD

chronic obstructive pulmonary disease

ALI

acute lung injury

ARDS

acute respiratory distress syndrome

IPF

interstitial pulmonary fibrosis

BAL

bronchoalveolar lavage

GC

glucocorticoids

FVC

forced vital capacity.