IDIOPATHIC PULMONARY FIBROSIS: NEW CONCEPTS IN PATHOGENESIS AND IMPLICATIONS FOR DRUG THERAPY

Abstract

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, and usually fatal pulmonary disease for which there are no proven or approved drug therapies. Anti-inflammatory and immunosuppressive agents have been largely ineffective. The precise relationship of IPF to other idiopathic interstitial pneumonias (IIPs) is not known, despite the observation that different histopathological patterns of IIP may co-exist in the same patient. We propose that these different histopathological “reaction” patterns may be determined by complex interactions between host and environmental factors that alter the local alveolar milieu. Recent paradigms in IPF pathogenesis have focused on dysregulated epithelial-mesenchymal interactions, an imbalance in TH₁/TH₂ cytokines and potential roles for aberrant angiogenesis. In this review, we discuss these evolving concepts in disease pathogenesis and emerging therapies designed to target pro-fibrogenic pathways in IPF.

CLINICAL EVALUATION AND DIAGNOSTIC APPROACH TO IPF

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive parenchymal lung disease with a median survival of less than three years following diagnosis, although the clinical course can be highly variable (1, 2). No pharmacologic therapies have proven effective for this disorder (3). IPF is the most common of the idiopathic interstitial pneumonias (IIPs, Figure 1) with a prevalence of 13−20 per 100,000 people in the general population (3, 4). It is more common in men than women and its prevalence increases with age (3, 4). Predictors of a worse outcome include progressive dyspnea, oxygen desaturation during the 6-minute walk (5), worsening pulmonary function and gas-exchange (6, 7), the presence and extent of honeycombing on high-resolution computed tomography (HRCT) (8), and the presence of pulmonary hypertension (2, 9).

Figure 1.

Idiopathic interstitial pneumonias (IIPs) represent an overlapping spectrum of inflammatory and fibrotic tissue reactions or histopathologic patterns in response to an “unknown” injury. At one end of the spectrum are IIPs marked predominantly by diffuse inflammatory cell infiltration; these IIPs tend to be more responsive to anti-inflammatory and immunosuppressive drug therapy. At the other end of the spectrum are IIPs characterized by fibrosis with minimal inflammation; these disease processes tend to have poor responses to currently available pharmacologic agents. Host factors (such as age and genetic polymorphisms) in combination with environmental factors (such as exposure to infectious agents and cigarette smoke) likely determine the resulting histopathological reaction patterns. DIP, desquamative interstitial pneumonia; RB-ILD, respiratory bronchiolitis associated-interstitial lung disease; COP, cryptogenic organizing pneumonia; NSIP, non-specific interstitial pneumonia; AIP, acute interstitial pneumonia; UIP, usual interstitial pneumonia.

The diagnosis of IPF is based on clinical, radiographic and histopathologic evaluations (3). Common clinical features consist of progressive dyspnea, dry cough and the presence of basilar “velcro-like” rales on examination. Digital clubbing and clinical signs of cor-pulmonale may be present. Extrapulmonary signs/symptoms are usually absent, while constitutional symptoms such as fatigue and malaise may be noted. Secondary causes of pulmonary fibrosis such as collagen-vascular disease, chronic hypersensitivity pneumonitis, adverse drug reactions, granulomatous diseases and pneumoconiosis must be excluded. More recently, HRCT has taken a more prominent role in the diagnosis of IPF and can help distinguish IPF from other IIPs (10). A patchy pattern of peripheral, subpleural and predominantly lower lobe reticular opacities combined with honeycombing, traction bronchiectasis and the absence of significant ground glass opacities together constitute the classic radiographic features of IPF. The presence of these features on HRCT, when reported by an experienced chest radiologist, correlates well with the histologic pattern of usual interstitial pneumonia (UIP) on surgical lung biopsy (11, 12). Thus, classic radiographic findings in the context of an appropriate clinical presentation may abrogate the need for a surgical lung biopsy; however, a bronchoscopy with transbronchial biopsy may be advisable in this setting, primarily to exclude infection and malignancy.

In the absence of typical clinical and radiographic features, a surgical lung biopsy is recommended for the definitive diagnosis of IPF. Diagnostic accuracy may be improved if biopsies are obtained from multiple lobes, as recent studies have shown that several distinct histopathologic patterns may co-exist in the same patient and the presence of UIP on any biopsy confer a worse prognosis (13). Histopathologic features of UIP include patchy areas of fibrosis in association with areas of normal lung architecture, the so-called “temporal” heterogeneity of UIP. Mild inflammatory cell infiltration may be present in UIP, but is not a prominent feature. Fibroblastic foci, consisting of aggregates of myofibroblasts underlying “injured,” reparative epithelium, are key histologic features of IPF (14). The presence and extent of fibroblastic foci, while not pathognomonic, are of prognostic value in IPF as the profusion of these lesions correlates with a worse prognosis (15).

PATHOGENESIS OF IPF

The etiopathogenesis of IPF remains enigmatic. Phenotypic changes in alveolar epithelial cells are an early and consistent features of IPF, suggesting that alveolar epithelial cell injury and apoptosis are key to the pathogenesis of IPF (14, 16-18). The cause(s) of alveolar epithelial cell injury associated with IPF is unknown, and host responses to tissue injury are likely to involve a combination of host-specific, genetic and environmental factors (Figure 1). Certain latent viral infections have been associated with IPF suggesting a possible infectious etiology (19-23). Other environmental factors, including cigarette smoke, have also been associated with the development of IPF (24, 25). Genetic factors seen with increased frequency in patients with IPF include mutations in surfactant protein C and polymorphisms of tumor necrosis factor-alpha (TNF-α) (26-29). Additionally, polymorphisms in transforming growth factor-β1 (TGF-β1) are associated with more rapid progression of IPF (30).

The relationship between IPF and other IIPs, such as nonspecific interstitial pneumonia (NSIP) and desquamative interstitial pneumonia (DIP) are not well defined. Recent studies have identified UIP, NSIP and even DIP in biopsy specimens from the same patient, suggesting that host responses may produce different histopathologic tissue reactions to the same stimulus/injury or that these histopathologic patterns represent a continuum that varies temporally and spatially within the same lung (13, 31, 32). Current prevailing hypotheses in the pathogenesis of IPF have focused on dysregulated interactions between epithelial and mesenchymal cells, a shift from TH₁ to TH₂-polarized host immune response, and aberrant angiogenesis. Additionally, recent data have demonstrated that acute exacerbations, marked by rapid clinical deterioration in previously stable patients with IPF, contribute to the increased mortality (33). Interventions to reduce the frequency and severity of these acute exacerbations may improve outcomes in IPF. These evolving concepts of disease pathogenesis and the natural history of IPF are the bases for emerging pharmacotherapies for the treatment of this complex disease.

Epithelial-Mesenchymal Interactions in IPF

Dysregulated interactions between epithelial and mesenchymal cells within the context of a damaged basement membrane and the surrounding extracellular matrix are increasingly recognized as a key control point in the pathogenesis of fibrotic disorders (Figure 2). Morphologic changes suggestive of epithelial cell injury and apoptosis are early features in the histopathology of IPF (14, 16-18). Several studies have described phenotypic alterations in type II alveolar epithelial cells (AECs) including proliferation, (16), bronchiolarization (34, 35), apoptosis and regenerative hyperplasia (36).

Figure 2.

The pathogenesis of IPF is complex and three major hypotheses or paradigms for the pathogenesis of IPF have emerged: dysregulated epithelial-mesenchymal interactions, aberrant angiogenesis and the TH₁/TH₂ cytokine imbalance. Epithelial-mesenchymal interactions between altered epithelial and mesenchymal phenotypes results in dysregulated interactions between these cellular compartments. In response to an unknown stimulus, alveolar epithelial cells in IPF develop a phenotype characterized by increased apoptosis, dysregulated proliferation, impaired regeneration/differentiation and perhaps impaired migration. Alveolar epithelial cells elaborate increased TGF-β1 and endothelin-1 along with decreased levels of PGE₂. Mesenchymal cells in this alveolar microenvironment acquire a contractile, synthetically active myofibroblast phenotype. Myofibroblasts retain the capacity to proliferate and are resistant to apoptosis while secreting large amounts of extracellular matrix proteins, soluble growth factors/cytokines and extracellular oxidants. Soluble mediators and the insoluble matrix elaborated by these myofibroblasts may, in turn, lead to aberrant reepithelialization that perpetuates a feed-forward cycle. IPF fibroblasts/myofibroblasts and alveolar epithelial cells contribute to an imbalance in angiogenic chemokines that may promote neovascularization and aberrant angiogenesis in IPF. A “shift” in the host immune response, favoring a TH₂ cytokine profile, has been postulated to contribute to initiation and/or progression of IPF.

Underlying areas of atypical and apoptotic epithelial cells are fibroblastic foci containing activated contractile myofibroblasts that secrete abundant extracellular matrix (ECM) proteins (14). These myofibroblasts are closely associated with areas of epithelial cell death and regeneration (37). In IPF, dysregulated interactions between epithelial and mesenchymal cells may promote a “vicious cycle” of epithelial injury/apoptosis and mesenchymal cell activation leading to progressive fibrosis.

Epithelial Cell Phenotypes

Reestablishing an intact epithelium following injury is a key component of normal wound healing. This may require orchestrated epithelial cell responses that include proliferation and expansion of progenitor/stem cells, migration and differentiation. The homeostatic function of an intact epithelium may be necessary to suppress mesenchymal cell activation. Altered epithelial cell phenotypes may contribute to fibrosis through excess elaboration of soluble factors, such as TGF-β1, that activate the underlying mesenchyme (38). In addition to secretion of fibrogenic cytokines, altered epithelial cell phenotypes contributing to the pathogenesis of IPF may include exaggerated or inappropriate epithelial cell apoptosis, dysregulated proliferation/differentiation and impaired migration (Figure 2).

There is abundant histopathologic and ultrastructural evidence of ongoing epithelial cell apoptosis in patients with IPF (39-43). Moreover, several studies in animal models have shown that inhibition of apoptosis attenuates pulmonary fibrosis (44) (45). The specific mediator(s) responsible for epithelial cell apoptosis in pulmonary fibrosis remain unclear and several different mechanisms, including TGF-β1, oxidative stress, Fas activation, and angiotensin II have been proposed. In epithelial cells, TGF-β1 typically functions as a tumor suppressor through growth-inhibition and induction of apoptosis (46). In an animal model, conditional activation of TGF-β1 in murine lung epithelium caused marked epithelial cell apoptosis and pulmonary fibrosis (47). Moreover, inhibition of epithelial cell apoptosis, both by pharmacologic caspase inhibition and genetic inactivation of the early growth response gene, blocked pulmonary fibrosis induced by TGF-β1 overexpression (47). Oxidative stress resulting from the release of reactive oxygen species (ROS) by activated phagocytic cells has also been proposed as a mechanism for epithelial cell injury in IPF (48). Although ROS are well recognized to induce cell injury, they are also important regulators of intra- and inter-cellular signaling (49). TGF-β1 is a strong inducer of the sustained production of extracellular oxidants in human lung fibroblasts (50, 51) and extracellular hydrogen peroxide (H₂O₂) generated by IPF myofibroblasts in response to TGF-β1 has been shown to mediate paracrine induction of epithelial cell apoptosis/death (52). Upregulation of the Fas signaling pathway has been demonstrated in human IPF (42) and a number of studies in animal models have explored the role of the Fas-Fas ligand cascade in epithelial cell apoptosis, sometimes with conflicting results (53-55). Angiotensin II has also been proposed as a soluble mediator of epithelial cell apoptosis in the pathogenesis of IPF. Inhibition of angiotensin II or the angiotensin receptor has been shown to block Fas-mediated epithelial cell apoptosis (56-58). Moreover, IPF-derived fibroblasts can induce paracrine epithelial cell apoptosis through generation of angiotensin II (56). In animal models, inhibition of angiotensin II activation and blockade of its receptor abrogates bleomycin-induced fibrosis (44, 59, 60).

Epithelial cell regeneration/proliferation and differentiation are critical components of wound healing that may be impaired in fibrotic tissue responses. Exogenous administration of epithelial cell mitogens/motogens appears to be able to attenuate fibrosis in animal models. One of these growth factors, hepatocyte growth factor (HGF), is secreted by myofibroblasts and functions as an epithelial cell mitogen and motogen with actions that tend to oppose those of TGF-β1 (61, 62). Thus, while HGF stimulates epithelial cell proliferation/migration (63), it also promotes myofibroblast apoptosis (64). Fibroblasts/myofibroblasts isolated from patients with IPF secrete lower levels of HGF than normal lung fibroblasts (65). Moreover, HGF has demonstrated anti-fibrotic activity in animal models of renal, hepatic, pulmonary and myocardial fibrosis (64, 66-68). Interestingly, interferon gamma (IFN-γ) has been shown to upregulate the HGF receptor in alveolar epithelial cells (69).

Keratinocyte growth factor (KGF) is another epithelial cell mitogen/motogen that is secreted by mesenchymal cells (70-72). TGF-β1 has been shown to block the mitogenic effects of KGF on type II alveolar epithelial cells (73). In lung injury models, intratacheal instillation of KGF prior to bleomycin injury has been shown to attenuate fibrotic responses (74). In a recent study, IPF-derived fibroblasts generated equivalent amounts of KGF under basal conditions when compared to control fibroblasts; however, IPF-fibroblasts generated significantly lower levels of KGF following IL-1β stimulation (75). Collectively, these studies suggest that impaired regulation of HGF and/or KGF, may impair reepithelialization and contribute to the pathogenesis of pulmonary fibrosis.

Impaired epithelial cell migration may also contribute to the development of fibrosis by delaying or preventing reconstitution of the alveolar epithelium. In animal models of lung injury, growth factors that induce epithelial cell migration, including HGF and GM-CSF, have been shown to attenuate pulmonary fibrosis (67, 76). The plasminogen activation system has been shown to play an important role in the pathogenesis of bleomycin-induced murine pulmonary fibrosis. Overexpression of PAI-1, which inhibits plasminogen activation, promotes fibrosis while suppression of PAI-1 or overexpression of uPA (a plasminogen activator) attenuates fibrosis (77-80). Although the precise mechanism of protection remains unclear, modulation of this proteolytic system can regulate wound healing and reepithlialization (63, 81, 82).

Once reestablished, an intact epithelium functions to suppress fibroblast/myofibroblast activities; the loss of this suppressive effect may contribute to the pathogenesis of pulmonary fibrosis. Intact epithelial cells secrete PGE_2, a cyclooxygenase (COX)-dependent arachadonic acid metabolite which suppresses fibroblast proliferation (83, 84), migration (85), differentiation (86), and collagen synthesis (87). Animal models support a role for PGE₂, as COX-2 deficient mice develop increased fibrosis and leukotriene-deficient mice are protected from fibrosis following intratracheal bleomycin-induced lung injury (88, 89). Importantly, PGE₂ levels in the epithelial lining fluid are significantly reduced while levels of leukotriene B₄ and C₄ are increased in lung homogenates of patients with IPF, (90, 91).

Recent studies have reported the potential derivation of lung epithelial cells from the bone marrow or systemic circulation. Krause and colleagues showed that transplanted bone marrow-derived stem cells may engraft as type II pneumocytes in the lung (92). Kotton et al. demonstrated increased numbers of bone marrow-derived stem cells in the lung following bleomycin-induced injury (93). More recent studies, however, suggest that some of these earlier findings may have been confounded by technical limitations of in-situ labeling and that true engraftment of the lung by bone marrow-derived stem cells may not occur, even in response to an injury (94). Bone marrow-derived stem cells have been used as a vehicle for transgene expression in the lung epithelium (95). The precise role(s) of circulating bone marrow-derived cells in repair and regeneration of the alveolar epithelium in response to injury is presently unclear.

Finally, there has been recent interest in alternative fates of epithelial cells, particularly with regard to epithelial-mesenchymal transition (EMT). Wnt/beta-catenin, a signaling pathway that is crucial in embryonic development and is implicated in EMT, is activated in epithelial cells at bronchiolo-alveolar junctions in IPF lung (18). Recent evidence suggests that EMT may contribute to the fibrotic responses in lungs of IPF patients (96). Further studies are required to determine the precise role and significance of EMT in the pathogenesis of IPF.

Mesenchymal Cell Phenotypes

Underlying areas of epithelial injury in lungs of patients with IPF are fibroblastic foci composed of aggregates of activated fibroblasts/myofibroblasts, which are the key effector cells in fibrogenesis (14, 97, 98). Fibroblasts have the ability to undergo distinct phenotypic transitions and participate in the stereotypic repair processes of diverse human tissues and organs (99, 100). Myofibroblasts are contractile and highly activated, secretory cells with characteristics intermediate between fibroblasts and smooth muscle cells (100, 101). Following injury, mesenchymal cells infiltrate wounds, differentiate into myofibroblasts and secrete ECM proteins that help constitute a provisional matrix which serves as the “scaffold” for normal tissue repair (101). Contraction of the provisional matrix approximates epithelial cell margins to facilitate reepithelialization while collagen secretion stabilizes the contracted matrix (100). Interestingly, recent studies indicate that lung mesenchymal cells following injury may derive from several sources including resident interstitial fibroblasts (102), circulating fibrocytes (103, 104), bone marrow-derived fibroblast precursors (105) and lung epithelial cells via epithelial-mesenchymal transition (96, 106, 107).

For normal healing to occur, wound myofibroblasts must undergo apoptosis (108); failure of apoptosis leads to myofibroblast accumulation, exuberant ECM production, persistent tissue contraction, and pathologic scar formation (100). TGF-β1 is a critical mediator of the myofibroblast phenotype and has been implicated in the pathogenesis of fibrotic disease in virtually every organ system studied (32, 97, 109, 110). In rodent models, over-expression of TGF-β1 leads to pulmonary fibrosis (111, 112) and blockade of TGF-β signaling abrogates fibrosis (113). In contrast to its effects on epithelial cells, TGF-β1 promotes an anti-apoptotic phenotype in fibroblasts (114, 115).

Multiple other soluble and insoluble substances in the cellular microenvironment have been shown to promote a “pro-fibrotic” mesenchymal cell phenotype. Soluble factors that induce myofibroblast differentiation include endothelin-1 (116), thrombin (117), and thrombospondin-1 (118). Consistent with a pro-fibrotic role, angiotensin II acts as a fibroblast mitogen (119) and the ACE inhibitor, captopril, inhibits fibroblast proliferation (120). Angiotensin II also induces the synthesis of TGF-β1 in several cell systems (121, 122). The ECM itself plays a key role in regulating mesenchymal cell phenotypes, as the composition and biomechanical properties of the ECM may also regulate myofibroblast differentiation, contractile capacity and apoptosis (123-126). Moreover, adhesion-dependent activation of focal adhesion kinase is required for stable induction of myofibroblast differentiation by TGF-β1 (127).

Soluble and insoluble factors secreted by myofibroblasts can contribute to the development of fibrosis through several mechanisms. First, the activated myofibroblast may be a source of soluble mediators that induce epithelial cell injury/apoptosis through paracrine signaling (52, 128). Autocrine secretion of growth factors promotes activation of pro-survival signaling pathways in myofibroblasts (114), and the autocrine induction of TGF-β1 itself may amplify these signaling pathways in fibroblasts while inhibiting epithelial cell growth and/or inducing epithelial cell apoptosis. Additionally, decreased production of “epithelial-protective” factors by myofibroblasts may feed into a state of epithelial-mesechymal dysrepair. For example, fibroblasts/myofibroblasts from IPF patients have a reduced capacity to elaborate the epithelial cell mitogens, HGF and KGF (65, 75). Myofibroblasts are likely the most important source of ECM proteins in the chronically injured lung. TGF-β1-stmulated generation of extracellular hydrogen peroxide has been shown, in certain contexts, to induce oxidative cross-linking reactions in the matrix overlying myofibroblasts (129). Such alterations in the biochemical and biophysical properties of the ECM may perpetuate dysregulated epithelial and mesenchymal phenotypes.

Finally, myofibroblasts may participate in the initiation or propagation of inflammatory responses (130). The myofibroblast has been described as an “inflammatory cell” due to its capacity to secrete multiple cytokines, chemokines and growth factors (130-132). Additionally, IPF-derived myofibroblasts and fibroblasts induced to undergo myofibroblast differentiation in response to TGF-β1 have been shown to generate extracellular oxidants (50, 52).

The Role of Inflammation and the TH₁/TH₂ Hypothesis

The role of inflammation in the pathogenesis of IPF has been the subject of considerable debate (133). For several decades, chronic inflammation was the prevailing and dominant paradigm in IPF pathogenesis (134, 135). However, the failure of anti-inflammatory strategies to improve outcomes in patients with IPF and the lack of prominent inflammation in IPF lung biopsy specimens have prompted a reevaluation of the role of inflammation in pathogenesis (136, 137). A recent study using microarray analyses comparing lung tissues from patients with IPF and hypersensitivity pneumonitis showed that the IPF gene signature was distinguished by the relative lack of typical inflammation-associated genes (138). While a comprehensive review of the inflammatory mechanisms in fibrogenesis is beyond the scope of this review, the reader is referred to existing reviews on this subject (139, 140). The precise role of inflammation and classical inflammatory cells in the pathogenesis of IPF remains uncertain. Evolving hypotheses regarding the role of inflammation now focus more on the balance between a “pro-inflammatory” TH₁ cytokine profile and a “pro-fibrotic” TH₂ cytokine profile.

The TH₁/TH₂ hypothesis proposes that progressive fibrosis results from a maladaptive immune/inflammatory response to a chronic or persistent pathogen/antigen (141, 142). In the typical host, a variety of cells may respond to pathogens/antigens by producing either TH₁ cytokines [such as interferon gamma (IFN-γ) and interleukin (IL)-12] which tend to suppress fibrotic responses or TH₂ cytokines (such as IL-4, IL-10, IL-13) that promote fibrotic responses. This hypothesis proposes that an inappropriate shift in the TH₁/TH₂ cytokine balance, favoring the TH₂ profile, results in the development of fibrotic disease (recently reviewed in reference #142). Much of our understanding of TH₁/TH₂ polarization in the pathogenesis pulmonary fibrosis has come from rodent models of fibrosis provoked by intratracheal bleomycin to induce epithelial injury (143, 144). Since the bleomycin animal model is an imperfect model of IPF (145), the significance of these findings to human IPF is unclear. Interestingly, even in the bleomycin animal model, the precise roles of host immune T-cell responses are not entirely clear. One study showed that T-cell depleted mice are capable of developing bleomycin-induced pulmonary fibrosis (146), while another study showed that CD-28, necessary for co-stimulatory T-cell activation, was required for the fibrotic response to injury (147). Despite the imperfections of animal modeling and the controversial role of inflammation, recent studies in human IPF tissues and IPF-derived cells seem to support the TH₁/TH₂ polarization hypothesis. Increased expression of the receptors for IL-4 and IL-13 was shown in lung tissues from IPF patients compared with those from patients with non-IPF idiopathic interstitial pneumonias (148). Fibroblasts derived from IPF lung similarly demonstrated increased expression of these TH₂ cytokine receptors (149). The use of a chimeric protein comprised of human IL-13 and a truncated version of Pseudomonas exotoxin appears to attenuate the proliferation of IPF-derived fibroblasts to a greater extent than non-IPF derived fibroblasts (149).

The Role of Angiogenesis in IPF

Angiogenesis is an important component of wound healing and aberrant angiogenesis has been associated with fibrosis in cutaneous wound models (99, 150). In rodent models, lung injury is followed by marked pulmonary vascular changes including neovascularization (150). In patients with IPF, early studies demonstrated the presence of pulmonary-systemic vascular anastomoses (151). A more recent study showed increased levels of pro-angiogenic chemokines and cytokines in the plasma of patients with IPF compared to normal volunteers (152).

In a series of studies using both in-vivo murine models and lung tissue from IPF patients, Keane and colleagues have shown differences in the expression of “angiogenic” ELR⁺ CXC chemokines (IL-8/CXCL8, ENA-78/CXCL5, MIP-2/CXCL2) and the “angiostatic” interferon (IFN)-inducible ELR⁻ CXC chemokines (IP10/CXCL10). Patients with IPF have increased expression of angiogenic ELR⁺ CXC chemokines and decreased expression of the angiostatic chemokines (153, 154). Moreover, both inhibition of the angiogenic chemokines and administration of angiostatic chemokines have been shown to attenuate murine pulmonary fibrosis in the bleomycin model (155-157). Mice deficient in the angiostatic chemokine IP10/CXCL10 develop increased pulmonary fibrosis following bleomycin-induced lung injury (158). Collectively, these studies suggest that an imbalance of angiogenic and angiostatic chemokines contributes to the pathogenesis of pulmonary fibrosis and that intervention targeted at this imbalance may be beneficial in the treatment of IPF.

Two recent studies, however, have reported decreased neovascularization in fibroblastic foci of patients with IPF (159, 160). Another recent study reported elevated levels of endostatin, an inhibitor of angiogenesis, in the serum of IPF patients (161). Interestingly, a post-mortem study of lungs from patients with scleroderma-associated lung disease showed early increases in microvascularity with subsequent decreases in later stages of disease (162). Thus, continued investigation into the spatial and temporal regulation of angiogenesis may be necessary to determine the precise role(s) of angiogenic and angiostatic factors in the pathogenesis of human IPF.

Pulmonary Hypertension in IPF

Pulmonary hypertension may be an important risk factor for increased mortality in IPF patients (2). Early studies showed that patients with IPF had elevated mean pulmonary artery pressures compared to patients with non-IPF interstitial lung diseases; pulmonary pressures increased with exercise to a greater extent in IPF patients than in non-IPF patients (163). Another study reported that impaired gas exchange in IPF patients was associated with increases in pulmonary vascular resistance (164). More recent retrospective studies have shed further light on the association between pulmonary hypertension and IPF. In one study, pulmonary hypertension inversely correlated with gas exchange; moreover, IPF patients with pulmonary hypertension had increased mortality (9). In another study of IPF patients during an actue exacerbation, pulmonary hypertension was present in all six patients in whom an echocardiogram was obtained (165).

These potential links between pulmonary hypertension and increased IPF mortality have spurred an interest in vasoactive pharmacologic agents in the treatment of lung fibrosis-associated pulmonary hypertension. A preliminary study examined the effects of various vasodilators - inhaled prostacyclin, intravenous prostacyclin and nitric oxide – administered sequentially to eight patients with pulmonary hypertension associated with fibrotic lung disease; in this study, inhaled prostacyclin decreased pulmonary hypertension without affecting systemic blood pressure, shunt fraction, or ventilation/perfusion matching (166). Similarly, sildenafil, a phosphodiesterase-5 inhibitor approved for the treatment of pulmonary hypertension, was compared with epoprostenol in a randomized trial of 16 patients with pulmonary hypertension associated with fibrotic lung disease. Sildenafil induced pulmonary vasodilatation to the same degree as epoprostenol, but did not adversely affect ventilation/perfusion matching or oxygenation (167). A phase II multicenter trial of the safety and efficacy of inhaled prostacyclin in IPF patients with established pulmonary hypertension is currently underway (www.clinicaltrials.gov/ct/gui/show/NCT00109681).

NOVEL THERAPEUTIC STRATEGIES FOR IPF

Therapeutic strategies targeting inflammation with corticosteroids, azathioprine, cyclophosphamide, cyclosporine A and mycophenalate mofetil have shown no definitive benefit in the treatment of IPF (3, 168, 169). This contrasts with other fibrotic lung diseases in which immunosuppressive therapies appear to have more favorable results (170-174). While one cannot rule out a role for inflammation or the possibility of a dysregulated, ongoing, low-grade TH₂-mediated host response to a persistent or recurrent antigen in the pathogenesis of IPF, inflammation is not a prominent feature when patients come to clinical attention (14, 136). Here, we will highlight some of the emerging therapies that are currently in, or nearing, clinical trials for IPF (summarized in Table 1).

TABLE 1.

Emerging pharmacotherapy targeting fibrogenic pathways in IPF

Name/Class	Mechanism of Action	Clinical trials	Other
Interferon-γ1b	Modulates TH1/TH2 balance; inhibits fibroblast activation	Phase 3 (completed) Phase 3 (ongoing)
Pirfenidone	Suppresses TNF-α; inhibits TGF-β; free radical scavenger	Phase 2 (completed) Phase 3 (planned)
Zileuton	Inhibits 5-lipoxygenase; Modulates eicosanoid imbalance	Phase 2 (ongoing)	Approved for asthma
Etanercept	Inhibits TNF-α receptor binding	Phase 2 (completed)
N-acetyl cysteine	Anti-oxidant; glutathione precursor	Phase 3 (completed)	Approved for acetaminophen overdose and contrast nephropathy
Tetrathiomolybdate	Inhibits angiogenesis; Inhibits TGF-β, TNF-α production; anti-oxidant	Phase 1/2 (ongoing)
Bosentan	Endothelin-1 receptor antagonist	Phase 2/3 (ongoing)	Approved for pulmonary hypertension
Monoclonal Anti-TGF-β Antibodies	Inhibits binding of TGF-β to its receptor	Phase 1/2 in scleroderma and post-trabulectomy scar prevention; initiated in IPF
TGF-β receptor kinase inhibitors	Protein kinase inhibitor; Blocks TGF-β type 1 receptor kinase activity	Planning stages for IPF
Imatinib mesylate	Protein kinase inhibitor; Inhibits c-Abl and PDGF tyrosine kinase	Phase 2 (ongoing)	Approved for CML and GIST
AG1879	Protein kinase inhibitor; Modulates myofibroblast differentiation and survival by inhibiting FAK/Src kinases and Akt.	Pre-clinical
Anti-CTGF	Monoclonal antibody against CTGF	Phase 2 (ongoing)

Interferon Gamma-1b (IFN-γ1b)

IFN-γ appears to target several pathways in the pathogenesis of fibrosis. First, as a TH₁ cytokine, it may modulate the TH₂ polarization reported in such chronic disorders (142). IFN-γ may also mediate anti-fibrogenic effects by reducing fibroblast proliferation, chemotaxis and collagen production (175-178). Additionally, IFN-γ induces the expression of angiostatic ELR⁻ CXC chemokines, thereby blocking potential angiogenic signals (179). Finally, IFN-γ may aid in re-epithelialization through upregulation of the HGF receptor in alveolar epithelial cells (69). In contrast to these effects, early studies suggested that IFN-γ can function as a mitogen for fibroblasts (180-182). Furthermore, a recent study reported that IFN-γ failed to inhibit TGF-β1 stimulated collagen synthesis, contraction or myofibroblast differentiation in keloid-derived fibroblasts, suggesting that mesenchymal cells in fibrotic lesions may be resistant to the suppressive effects of IFN-γ (183).

Treatment with IFN-γ inhibits renal, pulmonary and hepatic fibrosis in animal models (184-186). In humans, an early clinical trial showed promising results with stabilization/improvement in the pulmonary functions in patients administered IFN-γ (187). These favorable results, however, were not confirmed in a follow-up multicenter, randomized, double blind, placebo-controlled trial (188). In this large Phase III trial, no significant benefit was seen in the primary endpoint of progression-free survival or in secondary endpoints of pulmonary function or quality of life. A trend towards decreased mortality was seen in the IFN-γ-treated group in retrospective subgroup analysis of patients with less severe disease (FVC > 55% predicted and DLCO > 35%). There is an ongoing prospective clinical trial designed to further evaluate the efficacy of IFN-γ_1b in this patient subgroup [International Study of Outcomes in IPF with IFN-γ_1b (INSPIRE)].

Pirfenidone

Pirfenidone is a pyridone compound [5-methyl-1-phenyl-2(1H)-pyridone] that inhibits fibroblast proliferation, differentiation and ECM synthesis (189). Several potential mechanisms have been proposed for the anti-fibrotic effects of pirfenidone; one possible mechanism relates to suppression of tumor necrosis factor-alpha (TNF-α), which is thought to be a key cytokine in pathogenesis of IPF. Other potential mechanisms include suppression of TGF-β1 transcription (190) and downregulation of heat shock protein 47 (HSP47), a protein involved in procollagen secretion in rodents treated with bleomycin (191, 192). Additionally, pirfenidone may function as a free radical scavenger to reduce oxidative stress (193, 194). In animal models, pirfenidone attenuates liver (195), peritoneal (196), renal (197), biliary (198) and pulmonary (199, 200) fibrosis.

Several studies have investigated the effects of pirfenidone in humans with pulmonary fibrosis. In a compassionate-use protocol enrolling 54 patients with IPF, pirfenidone treatment was well tolerated, allowed most patients to discontinue conventional therapy and some patients appeared to stabilize lung function (201). In a randomized placebo-controlled trial of 21 patients with pulmonary fibrosis due to Hermansky-Pudlak syndrome, pirfenidone-treated patients had a reduced rate of decline in lung function (202). Recently, a prospective, randomized, placebo-controlled trial of 107 patients with IPF reported no significant change in the primary endpoint [the lowest oxygen saturation (SaO₂) during a 6-minute walk at 6 months] (203). Significant improvements were seen in the subgroup of patients with less severe disease (those who maintained SaO₂ greater than 80% during their baseline 6-minute walk). Moreover, the treatment group had a significant reduction in the number of acute exacerbations of IPF (203), a relevant clinical endpoint as these episodes are increasingly recognized as a major cause of mortality in patients with IPF (33). Given the lack of benefit in the primary endpoint, further study is required before pirfenidone can be recommended for general use in the treatment of IPF.

Zileuton

An imbalance in eicosanoids, arachadonic acid metabolites that include leukotrienes (e.g. LTB₄/LTC₄) and prostaglandins (e.g. PGE₂), has been implicated in the pathogenesis of IPF; in general, leukotrienes mediate pro-fibrotic effects while cyclooxygenase-2 (COX-2)-dependent PGE₂ mediates anti-fibrotic effects (recently reviewed in ref #(204)). Fibroblasts from IPF patients synthesize decreased levels of PGE₂ at baseline (65, 91) and following TGF-β1 treatment (88). Moreover, IPF fibroblasts may be resistant to the suppressive effects of PGE₂, as exogenous PGE₂ has an impaired ability to decrease their proliferative capacity (205).

In animal models, COX-2 deficiency is associated with increased bleomycin-induced fibrosis (88) while leukotriene-deficient mice are protected (89). The chemokine, monocyte chemoattractant protein-1 (MCP-1) inhibits PGE₂ synthesis by alveolar epithelial cells resulting in enhanced fibroblast proliferation (206); mice deficient in the MCP-1 receptor, CCR2, are protected from pulmonary fibrosis (207), supporting a role for MCP-1/CCR2 mediated suppression of PGE₂ in the pathogenesis of fibrosis. Finally, PGE₂ mediates the anti-fibrotic actions of GM-CSF in murine models (208).

Collectively, these data support a role for eicosanoid imbalance favoring increased leukotrienes and decreased prostaglandins in the pathobiology of IPF. Although there are no published data on the use of prostaglandin agonists or leukotriene-inhibitors in humans with IPF, there is an ongoing Phase II clinical trial investigating the safety and efficacy of the 5-lipoxygenase inhibitor, Zileuton, at the University of Michigan.

Etanercept

TNF-α is a pro-inflammatory cytokine that may also have roles in the regulation of the ECM and has been implicated in IPF pathogenesis. Alveolar epithelial cells and macrophages from patients with IPF express elevated TNF-α levels (209, 210). Moreover, several studies have identified associations between TNF-α gene polymorphisms and pulmonary fibrosis (28, 29, 211, 212). Murine models support a role for TNF-α in the pathogenesis of pulmonary fibrosis. Transgenic overexpression of TNF-α induces a form of lymphocytic fibrosing alveolitis (213) and adenoviral-mediated overexpression of TNF-α in rat lungs induces inflammation, patchy fibrosis and myofibroblast accumulation (214). Mice deficient in TNF receptors or treated with a soluble TNF receptor develop decreased bleomycin-induced pulmonary fibrosis(215, 216). Paradoxically, several studies demonstrate that TNF-α may suppress collagen synthesis by fibroblasts (217-220) and activate certain matrix mettaloproteinases (221, 222). TNF-α deficient mice develop an intense and persistent inflammatory response without a decrease in lung hydroxyproline content following intratracheal bleomycin, suggesting that the absence of TNF-α may not, in itself, attenuate the fibrotic response to injury (223). Moreover, TNF-α overexpression actually reduces fibrosis induced by TGF-β1 overexpression (224). A Phase II clinical trial of etanercept (a soluble TNF-R-Fc fusion protein) in IPF has been completed and the results were recently presented in abstract form (225). Eighty-seven patients with mild-moderate IPF were randomized to receive twice-weekly subcutaneous etanercept or placebo. After 48 weeks of therapy, etanercept treatment resulted in no significant differences in the primary endpoints (FVC, DLCO, or the differences in alveolar-arterial oxygen gradient) or in the secondary endopoints (TLC, resting oxygen saturation, and 6-minute walk distance). However, some trends for improvement in physiologic parameters were noted which may yet lead to another trial of this agent.

N-acetyl Cysteine

Oxidative stress is thought to play a critical role in IPF pathogenesis (reviewed in ref #(226)). Localized generation of ROS may induce epithelial cell injury/apoptosis, a key ultrastructural feature of UIP/IPF. Inflammatory cells from patients with IPF generate increased levels of oxidants compared to normal volunteers (48) and the epithelial lining fluid of IPF patients contains both increased concentrations of myeloperoxidase (48) and decreased levels of the anti-oxidant, glutathione (GSH) (227). N-acetyl cysteine (NAC) may mediate anti-oxidant effects through augmentation of GSH synthesis. It is safe, efficacious and approved for the treatment of acetaminophen overdose (228) and in the prevention of contrast nephropathy (229). Both GSH and NAC decrease fibroblast proliferation in-vitro (230). In patients with IPF, aerosolized GSH decreases ROS production by alveolar macrophages in IPF (231). Both oral and intravenous NAC increase GSH levels in bronchoalveolar lavage fluid (232, 233). In a small prospective trial of oral NAC in combination with immunosuppressive therapy in patients with IPF, the NAC-treated group had increased levels of GSH in epithelial lining fluid, decreased markers of oxidative stress and improvements in pulmonary function (234). A preliminary study randomizing 30 IPF patients to inhaled NAC or a control for 12 months reported that inhaled NAC had no effect on quality of life or pulmonary function, but did lead to improvements in the lowest oxygen-saturation during a 6-minute walk (235). A large, prospective Phase III clinical trial of oral NAC combined with prednisone/azathioprine compared to prednisone/azathioprine alone has recently been reported(236). In this study, 182 patients were randomized to receive oral NAC (300 mg, three times daily) or placebo in addition on prednisone and azathioprine. After 12 months of therapy, the addition of NAC resulted in a significant decrease in the primary endpoints of decline in vital capacity and D_LCO. No significant difference was observed in the secondary endpoint of mortality. It is not known if the positive effects of NAC are related to modulation of prednisone and azathioprine treatment since NAC monotherapy and placebo arms were not included in this trial (237).

Tetrathiomolybdate

Tetrathiomolybdate (TM) is a copper-chelating compound that has been shown to protect from fibrogenic effects of bleomycin in animal models of pulmonary fibrosis (238). The mechanism(s) for the anti-fibrotic effects of TM may be mediated, in part, by its anti-angiogenic effects (239). TM may also suppress the in-vivo expression of pro-fibrotic mediators, SPARC and TGF-β1, in response to lung injury (240, 241). The protective effect of oral TM is observed even when the drug was administered several days after bleomycin injury, suggesting that anti-fibrotic effects cannot be explained by the blockade of the early inflammatory response in this model (238). A Phase I/II clinical trial of TM in IPF patients refractory to previous immunosuppressive therapies is nearing completion at the University of Michigan.

Bosentan

Endothelin-1 (ET-1) is smooth muscle cell and fibroblast mitogen that is secreted by alveolar epithelial cells, vascular endothelial cells and macrophages (242). ET-1 has been linked to activated fibroblastic phenotypes (243-246). ET-1 overexpressing transgenic mice develop chronic inflammation and pulmonary fibrosis (247). Bleomcyin-induced pulmonary fibrosis is attenuated by treatment with an endothelin receptor antagonist (248); although, another study showed no effect on collagen deposition in this model (249). Patients with IPF demonstrate increased expression of ET-1 in proliferating type II epithelial cells, endothelial cells, and inflammatory cells (250, 251).

Bosentan is an oral, non-selective endothelin receptor antagonist that is currently approved for the treatment of pulmonary hypertension (252). A Phase II multi-center, double-blind trial randomizing 158 patients with IPF to treatment with Bosentan or placebo has been completed (BUILD-1). The results have not been presented in a peer-review format; however, a press-release from Actelion reported that there was no significant difference seen in the primary endpoint of improvement in 6-minute walk. They did report a trend for improvement in the secondary endpoint of combined death or treatment failure (www.actelion.com).

Inhibitors of TGF-β1 Signaling

A large body of evidence implicates TGF-β1 as a key mediator of human fibrotic disorders. Animal models have demonstrated that blocking TGF-β1 prevents injury-provoked pulmonary fibrosis (111-113). TGF-β1 is upregulated in fibrotic foci of UIP/IPF (253, 254). Anti-TGF-β1 therapies involving monoclonal antibodies, small molecule inhibitors, and other approaches (anti-sense oligonucleotides, receptor kinase inhibitors) are currently being investigated. Recently, Bonniaud and colleagues used an animal model of pulmonary fibrosis induced by the adenoviral-mediated overexpression of TGF-β1 to show that oral treatment with a small molecule inhibitor of type-1 TGF-β receptor (ALK-5) kinase activity could attenuate the development and progression of pulmonary fibrosis (255).

Despite its pro-fibrotic properties, TGF-β1 also has important homeostatic functions as an immune- and tumor-suppressive cytokine (46). Although inhibition of TGF-β1 signaling protects mice from fibrosis, recent animal studies have shown that an extended loss of TGF-β1 signaling can induce emphysematous changes in the lung (256, 257). These concerns temper enthusiasm for the long-term inhibition of TGF-β1 as a therapeutic strategy for the treatment of IPF. Continued investigation and better understanding of the divergent post-receptor signaling mechanisms of TGF-β1 may allow for more targeted blockade of its pro-fibrotic actions in the future (258).

Protein Kinase Inhibitors

Protein kinases are critical intracellular intermediates in the regulation of cell signaling and phenotype. The successful use of the c-Abl kinase inhibitor, imatinib mesylate (Gleevec®) in oncologic disease (259) provided proof-of-principle that protein kinase inhibitors could be used to target altered (activated) signaling pathways in certain human diseases. Recent investigations suggest that several protein kinases are potential targets of inhibition for the treatment of IPF.

c-Abl and Platelet-Derived Growth Factor Receptor Kinase

Recently, TGF-β1 has been shown to activate the Smad-independent c-Abl tyrosine kinase that appeared to mediate ECM synthesis in fibroblasts (260). Treatment with imatinib mesylate decreased bleomycin-induced pulmonary fibrosis in mice (260). Another group reported that this drug decreases fibroblast proliferation and bleomycin-induced murine pulmonary fibrosis through inhibition of platelet derived growth factor (PDGF)-receptor activation (261). Similarly, a recent study with three different inhibitors of PDGF signaling (including imatinib mesylate) demonstrated a reduction in radiation-induced pulmonary fibrosis (262). A multi-center Phase II clinical trail of imatinib mesylate (Gleevec) in IPF is nearing completion.

Focal Adhesion Kinase

Focal adhesion kinase (FAK) is a tyrosine kinase that regulates cell proliferation, migration and survival (263). This adhesion-dependent FAK pathway is required for differentiation of fibroblasts to the myofibroblast phenotype (127). Inhibition of FAK signaling may also induce fibroblast apoptosis (264), while forced activation of FAK signaling can protect fibroblasts from apoptosis through activation of Akt (265). Recent studies demonstrate that FAK and Akt are expressed in areas of fibrosis following intratracheal bleomycin in mice and that treatment with a protein kinase inhibitor (AG1879) that blocks TGF-β1-induced FAK and Akt in myofibroblasts attenuates fibrogenic responses to bleomycin (266).

p38 MAP Kinase

p38 MAP kinase is activated by a variety of stimuli and participates in multiple cellular processes through the activation of downstream transcription factors (267). Early and delayed activation of p38 MAPK by TGF-β1 have been described (268, 269). Furthermore, p38 MAPK activation is necessary for the activation of Akt by TGF-β1 in human lung fibroblasts, supporting a role for p38 MAPK in the anti-apoptotic phenotype of human lung fibroblasts (114). Finally, p38 MAPK activation has been implicated in proliferative responses of fibroblasts in response to TGF-β1 treatment (270). Inhibition of p38 MAPK has been shown to attenuate bleomycin-induced pulmonary fibrosis in murine models (271, 272).

Other Potential Pharmacologic Agents

We have already discussed the potential roles of angiotensin-II in the pathogenesis of IPF. Inhibitors of both angiotensin-converting enzyme and the angiotensin receptor are widely used for the treatment of hypertension, ischemic cardiomyopathy and diabetic nephropathy. Clinical experience with these drugs in IPF, however, is limited. A recent retrospective analysis from the Mayo clinic found no significant differences in the survival of IPF patients receiving an ACE-I compared to IPF patients who were not receiving an ACE-I (273).

Connective Tissue Growth Factor (CTGF) is a growth factor that is produced by fibroblasts in response to TGF-β1 stimulation and is thought to mediate some of the pro-fibrotic effects of TGF-β1 (274). In murine models, CTGF expression is upregulated by bleomycin (275). Adenoviral-mediated overexpression of CTGF in lungs of mice, however, induces only transient fibrosis (276). Increased expression of CTGF has been demonstrated in alveolar epithelial cells and interstitial fibroblasts in IPF (277). Monoclonal antibodies to CTGF are currently in Phase I/II clinical trials in IPF.

CONCLUSION

Idiopathic pulmonary fibrosis is a progressive, debilitating lung disease for which the etiology remains unclear and for which no proven effective therapies exist. Clinical and laboratory investigations have led to changing paradigms in disease pathogenesis and evolving new targets for therapeutic intervention. Therapeutic strategies targeting fibroblast phenotypes (myofibroblast differentiation, proliferation and survival), epithelial regeneration and apoptosis, angiogenesis, ECM regulation and pulmonary hypertension are supplanting traditional anti-inflammatory/immunosuppressive strategies. Continued investigations into the signaling mechanisms supporting these dysregulated cellular phenotypes are likely to identify future targets for intervention.