Lung Inflammation and Fibrosis

PETER A. WARD and GARY W. HUNNINGHAKE

Department of Pathology, University of Michigan, Ann Arbor, Michigan and Department of Medicine, University of Iowa, Iowa City, Iowa

	INTRODUCTION

TOP INTRODUCTION CONCLUSION REFERENCES

Lung research supported by the National Heart, Lung, and Blood Institute (NHLBI) has had important outcomes that have resulted in an infinitely better understanding of how inflammatory responses in lung are triggered, the mediators that are involved, and why under some circumstances the outcome is resolution and return to normal function, while in other situations the result is progression to a chronic, persistent inflammatory response that results in extensive interstitial pulmonary fibrosis and parenchymal collapse. In this article, we have attempted to review some, but not all, of the important highlights of lung research that have led to important conceptual advances and, in some situations, to important clinical studies and trials.

	LUNG MEDIATOR PATHWAYS AND THEIR REGULATION

In the lung, neuropeptides associated with both adrenergic and cholinergic products as well as noncholinergic nerve fibers regulate the inflammatory response (1). Associated with branches of the vagus nerve and cervicothoracic sympathetic ganglia that form the anterior and posterior pulmonary plexuses, these neuropeptides affect a variety of lung responses, including airway and vascular smooth muscle relaxation or contraction and glandular and surface epithelial cellular responses. Many neuropeptides in lung have been identified, such as tachykinins, vasoactive intestinal peptide, neuropeptide Y, galanin, enkephalins, etc. Obviously, the network of neuropeptides is not critical for lung function, as demonstrated in patients receiving lung allografts, but this neuro-regulatory system appears to be essential for optimal homeostatic inflammatory responses of the lung. In a very general way, non-neural lung inflammatory mediators can be classified into vascular-associated and non-vascular-associated. The early work of Ryan and colleagues (see below) demonstrated that the pulmonary endothelium plays an important role in regulation of inflammatory mediators that circulate through the pulmonary vasculature. For instance, in one pass through this vasculature, most kinins are converted into inactive peptides by contact with enzymes on the endothelial surface. Perturbation in the regulation of kinins may be present in individuals with active asthma in whom plasma levels of kinins have been detected.

	THE LUNG IMMUNE SYSTEM

Due to the prominence of secretory immunoglobin A (IgA) in the airway system and because during bacterial infections lung airway secretion of IgM and IgG occurs, the lung has been a focus for investigation of immune responses (2). It is now well established that the unperturbed lung immune response is largely confined to generation of secretory IgA. Using particulate antigens such as sheep red blood cells (RBC), airway instillation into rodents induces an immune response that is confined to regions of lung into which RBC have been instilled. The immune responses to sheep RBC involve bronchoalveolar lymphoid tissue (BALT) and regional lymph nodes. Under such circumstances, the immune response is confined to lung, with no measurable systemic immune response. It is also possible for the lung to acquire immune reactivity through the trafficking of blood lumphocytes into lung lymphoid tissue. The immune response in lung is fundamentally protective, but it is now well established that this response can, under certain conditions, be harmful. For instance, the human lung inflammatory response to respiratory syncytial virus (RSV) occurs as a result of naturally acquired immunity to RSV or as a result of active immunization (6). Indeed, immunization of humans with inactivated RSV led to florid lung inflammation after contact with the wild virus. It is not clear to what extent lung inflammation as seen in cystic fibrosis is chiefly a protective immune response to the presence of bacteria and to what extent the response is excessive and results in harmful outcomes to the lung. The possible relationship between the immune response and pulmonary fibrosis is discussed below.

In the past decade, a major conceptual advance has occurred in the area of immune responses (and subsequent inflammatory reactions), namely, the role of T helper type 1 (Th1) and T helper type 2 (Th2) lymphocytes in the immune response (7). Th1 lymphocytes seem to facilitate generation of interleukin (IL)-2 and interferon gamma (IFN-), resulting in enhanced cellular immune responses, such as delayed-type hypersensitivity (DTH) reactions and allograft rejection. Th2 lymphocytes are associated with production of IL-4 and IL-10, which facilitate Ig production, including IgE. Thus, T-cell phenotypes can be correlated with functional immune responses. This hypothesis has received support from studies of immune and inflammatory responses in mouse lung after immunization protocols that induce either DTH reactions or those that are associated with Ig-dependent outcomes. In these models, Th1 responses appear to result in progressive lung inflammatory responses with development of pulmonary fibrosis, while Th2 responses appear to resolve without a progressive and functionally disabling outcome. This body of knowledge may provide new insights into how selective cytokine blockade may favorably influence a lung inflammatory immune response that results in an adverse inflammatory outcome.

	THE ACTIVATED PULMONARY VASCULAR ENDOTHELIUM

It is apparent that the lung vascular endothelium plays a very dynamic role in the lung inflammatory response (8, 9). For instance, activation of the endothelium can induce the presence of both procoagulant factors (e.g., tissue factor) and anticoagulant factors (e.g., tissue plasminogen activator, thrombomodulin, etc.). In addition, stimulated endothelial cells can express a variety of adhesion molecules (e.g., intracellular adhesion molecule-1 [ICAM-1], vascular cell adhesion molecule-1 [VCAM-1], E-selectin, P-selectin, etc.) as well as inflammatory mediators (e.g., IL-8, MCP-1, platelet-activating factor, etc.). Endothelial adhesion molecules are known to be reactive with counter-receptors contained on leukocytes (10). For instance, the ₂ integrin family (CD11/CD18) of molecules on leukocytes is reactive with ICAM-1 in a manner that promotes leukocyte adhesion to the endothelium. E- and P-selectins of endothelial cells are reactive with sialyl Lewis ^x-containing molecules present on surfaces of leukocytes. This knowledge has been the basis for clinical trials involving blockade of adhesion-promoting molecules such as ICAM-1. Another major advance in the understanding of the biology of adhesion molecules has come from studies in young patients with leukocyte adhesion deficiency syndrome (type I) in which genetic defects in CD18 have been found. These defects preempt the function of ₂ integrin-dependent pathways, including chemotaxis and phagocytosis. Afflicted patients with functionally inactive CD18 have a profoundly depressed ability to recruit leukocytes, resulting in frequent and often fatal bacterial infections. Understanding the key role of CD18 in neutrophil recruitment has led to plans for the use of blocking antibodies to CD18 in clinical trials in which acute inflammatory conditions such as that found in adult respiratory distress syndrome (ARDS) are the target of intervention.

Molecules that promote adhesion of leukocytes to the endothelium (as described above) have only been recognized in the past two decades. These molecules permit leukocyte adhesion only to the activated endothelium. Adherent leukocytes also undergo an activation process not only due to engagement of leukocytes ₂ integrins with endothelial "counter receptors" (ICAM-1) but also because activated endothelial cells produce inflammatory mediators (e.g., IL-8, MCP-1, platelet-activating factor) that have the ability to activate adherent leukocytes. The most significant conceptual advance in this area has been the realization that only the activated endothelium participates in the inflammatory response. For instance, neutrophils will not adhere to the unperturbed endothelium but will adhere to areas of the endothelium that are expressing relevant adhesion molecules. Accordingly, it should be possible to develop therapeutic interventions that are specifically targeted only for factors expressed by the activated endothelium.

	CYTOKINES AND CHEMOKINES

Inflammatory mediators deriving from extravascular sources in lung include a vast array of products from cellular and noncellular sources. Perhaps the most intensively studied mediator system involves the interleukins, which originate not only from lung macrophages but from a variety of nonmacrophage cellular sources (e.g., alveolar epithelial cells, fibroblasts, mast cells, etc.). Work on tumor necrosis factor alpha (TNF-) and IL-1, the early response cytokines, has established the primacy of these mediators in upregulation of lung vascular adhesion molecules. The IL-8 family of cytokines, collectively termed "chemokines," is now known to represent at least a dozen and a half highly reactive peptides, whose biologic activities include chemotactic activity for leukocytes (neutrophils, eosinophils, monocytes, lymphocytes), angiogenic activity, induction of collagen synthesis, and cell proliferation, etc. Excessive intrapulmonary production of IL-8 is associated with ARDS (11). The recent observations that some chemokine receptors facilitate entry of human immunodeficiency virus (HIV) into CD4⁺ lymphocytes and that cells with mutant receptors (CCR-5) for these chemokines are resistant to entry of HIV add a new dimension to chemokine biology (12). Development of chemokine receptor blocking reagents can be expected to provide important new approaches for clinical interventions. Having said this, definition of the in vivo biologic roles of chemokines in a variety of experimental and clinical conditions is critical to determining the most appropriate chemokines for blockade.

Information pertaining to cytokines and chemokines has come at a very rapid pace, far outstripping our ability to understand this vast array of mediators and determine their clinical relevance. While most cytokines are thought to be pro- inflammatory, it has become obvious that there also exists a family of anti-inflammatory interleukins (15). For instance, in the context of lung inflammation, IL-4, IL-6, IL-10, and IL-13 have powerful anti-inflammatory effects, apparently related to their ability to suppress TNF- production (and, correspondingly, to inhibit upregulation of endothelial adhesion molecules, such as ICAM-1). Thus, there is a built-in counterbalance which regulates the pro-inflammatory aspects of cytokines. Indeed, in vivo blockade of IL-10 significantly enhances lung inflammatory responses after intrapulmonary deposition of IgG immune complexes or after ischemia-reperfusion injury in rat lung. These data underscore the checks and balances built into the inflammatory response of the lung and raise the questions as to whether in persistent human lung inflammatory conditions the balance between pro-inflammatory and anti-inflammatory cytokines has been lost.

The chemokine family poses a daunting challenge because of the number of members and the variety of biologic activities. For instance, in the CXC family of chemokines, those molecules with the ERL chemical motif have very strong angiogenic activity, promoting capillary formation, whereas those lacking the ERL motif have angiostatic activity. It would appear that the ratio of ELR and non-ELR CXC chemokines may determine if a given lung inflammatory response will be associated with prominent vascular formation and fibrosis (16).

	OXIDANTS

The role of nitric oxide (·NO) in lung inflammatory responses is an enigma (17). Nitric oxide (whether produced by endothelial cells or by lung macrophages) has in vitro anti-adhesive properties, attenuating adhesion of leukocytes to endothelial cells. If a superoxide (·O₂) generating source is present, ·NO reacts with ·O₂ to form the peroxynitrite anion (·ONOO), reducing ·NO concentrations and diminishing its anti-adhesive effects. In this respect, ·NO downregulates the inflammatory response and can be considered anti-inflammatory. On the other hand, ·ONOO is highly reactive with thiols and other chemical groups and can be converted to the most reactive of all oxygen-centered radicals, the hydroxyl radical (HO·). Thus, ·NO can be construed as being anti-inflammatory or pro-inflammatory, depending on the particular experimental circumstances. In models of experimental lung inflammatory injury, blockade of ·NO production is protective against inflammatory injury without affecting neutrophil recruitment. In this instance, ·NO (or its derivatives) is clearly pro-inflammatory. One of the problems in trying to define the role of ·NO in vivo is that the use of competitive antagonists of L-arginine results in hypertension and vasoconstriction due to loss of endothelial ·NO (which has vasodilatory effects). Until a specific inhibitor of inducible ·NO synthase is available, it will be difficult to define precisely the roles of ·NO and its derivatives. Clinical trials using inhalation of ·NO (at low concentrations) in patients with diminished pulmonary artery blood flow have suggested that blood flow in lung may be improved, perhaps due to vasodilation of the pulmonary arterial system following airway contact with ·NO, but whether this type of therapy will become a useful clinical modality remains to be determined.

Another important oxidant-generating pathway is NADPH oxidase, which appears to play an important role in outcomes of the inflammatory response. Phagocytic cells contain abundant amounts of this enzyme, which is assembled on the cell surface by interaction of plasma membrane components with translocated cytosolic subunits following cell activation (18). The oxidant products include ·O₂, HO·, H₂O₂ and hypochlorous acid (HOCl), all of which are designed for killing of phagocytized bacteria. The fact that the enzyme is assembled on the surface of cell membranes implies that these oxidants are destined for extracellular release or release into phagocytic vacuoles formed after cell activation. When phagocytic cells are activated, tissue matrix damage may occur. Oxidant-induced changes in stromal and cellular targets can lead to a variety of outcomes. There is also evidence that oxidants such as HOCl can directly activate matrix metalloproteinases (e.g., collagenases, gelatinases, stromelysins, etc.). The combined actions of oxidants and proteinases released from activated phagocytic cells seems to be chiefly responsible for the acute damaging effects of the inflammatory response. How to design molecules to block these damaging products or to inhibit their formation remains a challenge. Understanding the multicomponent nature of NADPH oxidase has led to a recognition of mutant forms of components of this enzyme in humans with chronic granulomatous disease (CGD) of childhood. This information has provided explanations for the varying clinical courses (and survival times) of patients with various forms of CGD and may point the way for consideration of intrauterine diagnosis and corrective gene therapy.

	PROTEASES AND ANTIPROTEASES

Proteases and antiproteases play a key role in the outcome of the lung inflammatory response. The well-established relationship between functionally defective forms or absolute deficiency of 1 protease inhibitor (1PI) and development of pulmonary emphysema (reviewed in this supplement by Senior and Anthonisen) has focused on the role of leukocytic elastase in the breakdown of lung elastin, resulting in emphysema. The recognition that oxidants in cigarette smoke can cause chemical changes in 1PI leading to its functional inactivation has provided important clues as to why cigarette smokers have a higher incidence of pulmonary emphysema than do nonsmokers. Defining more precisely the pathways that lead to development of pulmonary emphysema remains an important priority. Clinical trials with recombinant 1PI in humans with 1PI deficiency have been undertaken, but it does not seem likely that this will be an effective, efficient and cost- effective way to treat these patients prophylactically.

Cloning technology has identified important tissue-associated protease inhibitors, such as secreted leukocyte protease inhibitor-1 (SLPI-1) and tissue inhibitors of metalloproteinases (e.g., TIMP-2). Both these enzymes and their endogenous enzyme inhibitors appear to derive from a variety of cells, including macrophages. If left unchecked, these proteases have the potential for directly causing cytotoxicity as well as extensive breakdown of connective tissue matrix. SLPI and TIMPs are secreted constitutively at low concentrations; during the inflammatory response these inhibitors may be induced, resulting in higher tissue levels of antiproteases. Whether deficiencies of these inhibitors exist in humans and to what extent clinical use of these inhibitors will be effective in preventing lung inflammatory damage remains to be determined. Understanding the balance between these enzymes and their inhibitors may yield useful information regarding regulation of inflammatory injury.

	ENDOTHELIAL CELL BIOLOGY AND ITS CLINICAL APPLICATIONS

As indicated above, endothelial cells, in addition to their vital barrier function (which retards outflow of cellular and plasma components), are in a dynamic state of change depending on environmental conditions (e.g., hypoxia, hyperoxia, presence of complement activation or coagulation products, etc.). One of the most important advances in the past two decades has been the development of culture techniques that have permitted the systematic analysis of endothelial cell function (8). Although most endothelial cell studies have involved the use of human umbilical vein endothelial cells, more recent approaches employing endothelial cells from other tissue or organ sources have suggested the presence of significant functional and biochemical differences, depending on the source of the endothelial cells. Such studies are in very preliminary stages. The ability to isolate and grow endothelial cells has allowed the cloning of genes relevant to the inflammatory response. As indicated above, these efforts have resulted in expression (and definition) of adhesion molecules such as ICAM-1, VCAM-1, E-selectin, P-selectin, etc. In turn, the role of these molecules in facilitating leukocyte adhesion and "rolling" (repetitive attachment to the endothelial surface followed by detachment) have been assessed. Access to recombinant adhesion proteins has permitted development of antibodies that have been used to search for expression of vascular adhesion molecules in human tissues from a variety of pathological conditions (e.g., allograft rejection, ARDS, DTH, etc.), providing some hints as to the role of these molecules. By such studies, it has been possible to estimate which adhesion molecules may be candidates for targeted clinical interventions. For instance, blocking of ICAM-1 has been done in patients with rheumatoid arthritis, although the results have been disappointing. Currently, anti-ICAM-1 is being used in clinical trials involving renal allografts to see if greater retention of early renal function following engraftment can be achieved. Identification of human adhesion molecules has facilitated the cloning of animal homologues and development of blocking antibodies. The results have permitted a definition of which of these molecules play a role in controlled experimental conditions, such as ischemia-reperfusion, allograft rejection, acute inflammatory reactions mediated by neutrophils, DTH reactions, etc. The information derived from this work may apply to related human conditions.

A compelling example of how endothelial cell cultures have resulted in vital new knowledge for clinical application comes from the identification of endothelial cell tissue plasminogen activator, a powerful thrombolytic agent. Tissue plasminogen activator is now commonly used for treatment of early stages of acute myocardial ischemia and/or impending myocardial infarction and more recently for treatment of nonhemorrhagic brain ischemia (strokes). These examples show the power of the use of endothelial cell cultures for providing information that bears on a new understanding of the inflammatory response and the application of this knowledge to human diseases.

The use of monolayer cultures of endothelial cells has resulted not only in a better understanding of mechanisms leading to adhesion and ultimate transmigration of blood leukocytes (neutrophils, monocytes, lymphocytes, etc.), but also to ways in which adhesion of activated leukocytes may lead to endothelial cell damage and the manner in which malignant cells co-opt adhesion mechanisms involving endothelial cells to ultimately move beyond the endothelial barrier and form metastatic foci. In this respect, the metastatic process is but a version of the inflammatory response. Studies of the interactions between endothelial cells and leukocytes resulting in endothelial cell damage have provided important and unexpected insights. Whereas leukocytes such as neutrophils produce oxidants and employ their own adhesion molecules, such as the ₂ integrins (CD11/CD18), endothelial cells employ their own adhesion molecules (e.g., selectins, ICAM-1) to facilitate these adhesive interactions. Furthermore, in the case of adherent, activated neutrophils, endothelial cells contribute critical components such as iron and ·O₂ to pathways that cause damage or destruction of endothelial cells. This type of information may be useful in devising in vivo strategies to block leukocyte-mediated damage of the endothelium during the inflammatory response.

	USE OF TRANSGENIC AND KNOCKOUT MICE FOR INFLAMMATORY STUDIES

Development of mice that express proteins not normally produced (transgenic mice) or mice that have been genetically manipulated, resulting in mutant nonfunctional proteins or absence of proteins (knockout mice), has provided important new approaches to an understanding of the biology of the inflammatory response. With respect to how this technology has influenced our understanding of the lung inflammatory system, several examples can be cited. Intrapulmonary expression of proteins has been accomplished by airway exposure to adenovirus constructs containing oligonucleotide coding for products such as chemokines or other interleukins. Transgenic expression of chemokines such as RANTES results in a mononuclear cell infiltrate, confirming the predicted in vivo biologic activity of this chemokine (19). The intrapulmonary transgenic expression of granulocyte/macrophage colony-stimulating factor driven by a surfactant promoter results in a phenotype that is indistinguishable from human alveolar proteinosis (20), providing a model for understanding and treating the human disorder. These experimental observations not only extend our knowledge regarding the biology of these factors but also suggest potential therapeutic applications for the treatment of detrimental inflammatory responses in the lung (e.g., as in sarcoidosis, idiopathic pulmonary fibrosis, etc.). To what extent this technology will be applicable to and effective in the treatment of human lung inflammatory diseases remains to be determined. Conditional knockout mice are being developed, allowing the shut-off in expression of a protein in adult mice. Although this technology is not yet fully developed, it may provide a valuable new approach to defining the role of the involved proteins.

Knockout mice have provided data that are often difficult to interpret. For instance, while P-selectin or ICAM-1 knockout mice show a depressed ability to recruit both neutrophils into the peritoneal cavity and mononuclear cells into DTH reactions in skin, induction of neutrophil-dependent lung inflammatory injury following systemic activation of complement is fully expressed, even though blocking antibodies to these adhesion molecules are highly effective in wild-type mice and ineffective in the relevant knockout mice (21). Thus, some knockout mice appear to have adapted to the loss of an important adhesion-promoting protein, although how this adaptation process occurs is not known.

Following systemic activation of complement in rats or mice, the lung inflammatory response and ensuing capillary injury develop in a neutrophil-dependent manner. This model has been evaluated in knockout mice that lack phagocyte NADPH oxidase. Injury in wild-type mice is oxidant-dependent and can be suppressed by catalase. Surprisingly, in the NADPH oxidase knockout mice injury is fully expressed but, as expected, it is now catalase-insensitive. On the basis of additional studies, it appears that in mutant mice there has been a switch from the NADPH oxidase pathway to the ·NO pathway for oxidant production (22). These data underscore caution that must be used in interpretation of data when knockout mice are employed.

Another example of the complexity in understanding the usefulness of knockout mice is found in attempts to simulate human cystic fibrosis in mice. Loss of expression of the cystic fibrosis transmembrane regulator (CFTR) in knockout mice has failed to result in a distinctive lung phenotype (23). These animals have lungs that morphologically appear normal. Why mice should not develop progressive infection and the related series of morphologic changes found in human lungs of patients with cystic fibrosis remains to be explained and, once again, underscores the caution that is needed in drawing conclusions from the use of knockout mice. Whether the problems with gene targeting, duration of gene expression, and ability to measure correction of CFTR loss in patients with cystic fibrosis can be overcome in human clinical trials remains to be seen.

	USE OF BRONCHOALVEOLAR LAVAGE

One of the most informative strategies for sampling the external environment of the lung during the course of an inflammatory response has been the development and application of bronchoalveolar lavage (BAL), from which can be captured both cellular and soluble products from the inflamed lung. This technique was first described for use in human subjects in the mid-1970s (24, 25). A number of studies have used BAL to evaluate alveolar macrophages and lymphocytes from normal lung. These studies showed that a major function of alveolar macrophages is the clearance of micro-organisms and various other particulates from the alveolar surface of the lung. These cells respond to a variety of stimuli and produce a number of products, including pro- and anti-inflammatory cytokines, chemokines, reactive oxygen and nitrogen species, proteases and antiproteases, complement proteins, and products of the arachidonate cascade. Interestingly, the capacity of alveolar macrophages to produce these factors often differs substantially from blood monocytes. Alveolar lymphocytes from normal human lung have not been studied in as much detail. It is known, however, that their functional repertoire is relatively similar to that of lymphocytes in blood. Analysis of T-cell receptor genes that are expressed on blood lymphocytes suggests that some of these cells are attracted to the lung and/or expanded in the lung in response to exposure to environmental antigens. These studies, in aggregate, have demonstrated the importance of defining inflammatory and immune responses in a site-specific manner. Many of the inflammatory and immune functions of normal lung, however, require further definition.

Brochoalveolar lavage has been very useful in pointing to mechanisms of lung inflammatory conditions. For instance, a precise analysis of leukocytes present in BAL fluids provides important clues as to what type of inflammatory response is occurring in the lung. Using this technique, early studies of patients with ARDS resulted in a realization that neutrophils were an important cellular component in this syndrome. Further, the data suggested that the presence of neutrophils might be contributory to the defects in gas exchange occurring in ARDS. At the same time, oxidant products such as oxidatively inactivated 1PI were found in BAL fluids from patients with ARDS, indicating that oxidant-generating pathways were participating in the chain of events. In addition, the finding in the same BAL fluids of high levels of IL-8 (11) and the lack of parallel production of the regulatory cytokine, IL-10, have suggested a possible role for IL-8 in neutrophil recruitment and the potential therapeutic application of exogenously administered IL-10 or anti-IL-8 in ARDS to suppress neutrophil recruitment. The application of BAL technology to animal models of lung inflammation has demonstrated the close association in BAL fluids between the early response cytokines (TNF- and IL-1) and the corresponding upregulation of lung vascular adhesion molecules such as ICAM-1 and E-selectin. Thus, it may be possible to monitor lung vascular endothelial activation by monitoring BAL levels of these cytokines, and to evaluate the effectiveness of regulatory cytokines (e.g., IL-4, IL-10) that tend to suppress production of TNF- and IL-1, thereby reducing upregulation of lung vascular adhesion molecules (16). Studies using BAL have been especially useful in defining the pathogenesis of various interstitial lung diseases.

	PULMONARY FIBROSIS

Histologic evaluation of lung tissue from patients with pulmonary fibrosis shows evidence of inflammation, a disordering of lung parenchymal cells, and fibrosis. These observations suggested that the inflammatory process results in lung injury and pulmonary fibrosis. While early studies suggested that the fibrosis occurred in the interstitium of the lung, more recent studies have suggested that the initial injury causes intra-alveolar fibrosis and collapse of alveolar capillary units (26, 27). These latter observations have suggested that pulmonary fibrosis may be an ongoing form of acute lung injury occurring sequentially at discrete sites in the lung. Even in those diseases where the cause is known (i.e., asbestosis), the factors that cause progression of the lung disease in specific areas of the lung are not known. The observations emphasize that effective therapies for these disorders must be given early in the natural history of the disease, prior to the development of extensive lung destruction and fibrosis.

An early observation from BAL in pulmonary fibrotic disorders suggested increased numbers of neutrophils and eosinophils in BAL fluid (25). Subsequent studies showed that these cells were activated, releasing a variety of products (like proteases and oxidants) that injure lung parenchymal cells. Other studies suggested that these leukocytes were attracted to the lung in two ways: (1) by chemoattractants directly released from alveolar macrophages; and (2) by chemoattractants released from lung parenchymal cells. The stimulus for the release of chemoattractants by lung parenchymal cells was, at least in part, due to cytokines (IL-1 and TNF-) generated from alveolar macrophages. The specific agents that stimulate alveolar macrophages to release these cytokines and chemoattractants depend on the etiology of the pulmonary fibrotic conditions (e.g., asbestos fibers in asbestosis). In idiopathic forms of the disease, the stimulus may be immune complexes (28). However, it is likely that these inflammatory processes are much more complex and involve multiple interactions among inflammatory cells, lung parenchymal cells, oxidant and antioxidants, proteases and antiproteases, and coagulation proteins that are described above.

The fibrotic process in the lung appears to result from a complex interaction between fibroblasts, other lung parenchymal cells, and macrophages (29). Injury to the epithelium and basement membranes appears to be necessary for the fibrotic process to occur (33). Fibroblasts migrate into areas of acute lung injury and are stimulated to secrete collagen and other matrix proteins. These cells also release various proteases that have the capacity to degrade and remodel these matrix proteins. The stimuli that activate fibroblasts to remodel the lung are not well defined but likely include components of blood (like fibrin), matrix degradation products, and mediators (like transforming growth factor beta) that are released from macrophages and lung parenchymal cells. Factors and circumstances that determine whether areas of the lung heal with minimal injury or progress to irreversible injury need to be defined.

Several studies have correlated lung histology with changes in pulmonary function and/or outcome of the process. In general, correlations have been found between measures of lung distensibility (compliance) and pulmonary fibrosis; however, this has not been confirmed in all studies. Many patients with pulmonary fibrosis are current or ex-cigarette smokers. All studies have shown that cigarette smoking alters pulmonary function in patients with pulmonary fibrosis, suggesting that compromised lung function may result from the combined effects of pulmonary fibrosis and emphysema (34). Patients with significant pulmonary fibrosis (honeycombing) on biopsy, clearly, have more advanced disease and a worse short-term prognosis.

Recent studies using high-resolution computerized tomographic (CT) scans have suggested that overall lung pathology can be estimated from appearance of the lung on scans. "Honeycombing" on lung CT scans appears to correlate with histologic evidence of honeycombing. "Ground-glass" changes on lung CT scans may be due to inflammation and/or early fibrosis. It is of interest that honeycombing usually develops in areas of lung that have previously demonstrated ground-glass changes. Normal-appearing lung on CT scans can be histologically normal, or it can represent early inflammation and fibrosis not detectable by CT technique. In a general way, these studies suggest that the lung disease progresses, from areas of normal-appearing lung to a ground-glass appearance, to honeycombed lung. A recent study combining evaluations of lung scans and whole lung morphometry showed that aggregate lung tissue in patients with pulmonary fibrosis is not increased but is reorganized and composed of increased amounts of connective tissue proteins (35). In pulmonary fibrosis there also is a marked decrease in the alveolar surface areas. This report is consistent with pathologic studies suggesting that the lung disease is associated with intra-alveolar fibrosis and collapse of alveolar capillary units followed by reorganization of the lung parenchyma.

Early studies suggested that the prognosis of pulmonary fibrosis could be determined by use of BAL, (i.e., patients with increased numbers of neutrophils and eosinophils and decreased numbers of lymphocytes had a worse prognosis than patients without these findings) (36). Although these findings have been confirmed, it is now clear that this approach is not necessary since similar data can be obtained according to duration of symptoms, severity of changes on chest X-rays and lung CT scans, and severity of changes on pulmonary function (31, 37). Male gender and cigarette smoking also appear to affect the prognosis of the disease. Although patients with this disease are often treated with corticosteroids and/or cytotoxic agents, optimal therapy of this disease has not clearly been defined, nor is it known if current treatment alters the natural history of the disease. Carefully controlled studies are required to define the natural progression of this disease in subsets of patients and to determine if various forms of therapy alter this natural progression.

Although the etiology of some forms of pulmonary fibrosis is known (i.e., exposure to asbestos or bleomycin), the cause of pulmonary fibrosis in most patients is unknown and is an important area for future study. For some patients, there appears to be a genetic component since multiple members of their families develop pulmonary fibrosis. For most patients, however, there is no family history of fibrotic lung disease. Several studies have suggested that cigarette smoking increases the risk of developing pulmonary fibrosis (38). It is unlikely, however, that cigarette smoking is the sole or primary cause of this disease, because it also develops in lifetime nonsmokers. Important future studies will need to address the genetic and environmental factors that underlie the development of idiopathic forms of this disease.

	GRANULOMATOUS LUNG DISEASE

Based on BAL studies, sarcoidosis was considered a disease characterized by depressed cellular immunity. This was due to observations that patients with sarcoidosis exhibited anergy to skin tests and studies of peripheral blood cells, suggesting a decreased cellular immune response. The first study of BAL in this disease, however, showed that pulmonary sarcoidosis was associated with an intense cellular immune response (39). A number of subsequent studies showed that the immune response in the lung is a classic TH1 type response (40). Lung lymphocytes in the lung spontaneously proliferate and release IL-2 and interferon gamma. The results of several studies that have evaluated expression of lung T-cell receptors have suggested that lung T cells exhibit a profile consistent with a response to an antigen (41). In this disease alveolar macrophages also have enhanced antigen-presenting capacity (42) and release cytokines like TNF- and IL-1, which are considered to be involved in granuloma formation; these cells also release the Th1 cytokine IL-12. In many sarcoid patients, the intrapulmonary granulomatous inflammation resolves without significant lung injury; however, in other patients this inflammatory response results in the development of pulmonary fibrosis and significant lung injury. Less is known about the immune response in chronic sarcoidosis compared with early sarcoidosis. Further, it is not known if the development of chronic sarcoidosis is due to failure to clear an unknown antigen (i.e., infectious agents) or if this is due to a genetically-determined aberrant immunologic response to common environmental exposures. Important areas for future study will be to identify the cause(s) of sarcoidosis and to determine how the genetic background of the patient influences the outcome of the disease. This is currently being evaluated in a multicenter study funded by the NHLBI.

	SUMMARY

TOP INTRODUCTION CONCLUSION REFERENCES

In summary, studies supported by the NHLBI have markedly increased our knowledge of basic mechanisms of inflammation and fibrosis in the lung. A remarkable feature of these studies is the translation of this knowledge into studies of patients with inflammatory and fibrotic lung disease and the prospects of therapeutic advances on the horizon.

	Footnotes

Correspondence and requests for reprints should be addressed to Peter A. Ward, Department of Pathology, University of Michigan, M5240 Medical Science I, 1301 Catherine Rd., Ann Arbor, MI 48109-0602.

	References

TOP INTRODUCTION CONCLUSION REFERENCES

1. Maggi, C. A., A. Giachetti, R. D. Dey, and S. I. Said. 1995. Neuropeptides as regulators of airway function: vasoactive intestinal peptide and the tachykinins. Physiol. Rev. 75: 277-322 [Free Full Text].

2. McDermott, M. R., A. D. Befus, and J. Bienenstock. 1982. The structural basis for immunity in the respiratory tract. Int. Rev. Exp. Pathol. 23: 47-112 [Medline].

3. Kaltreider, H. B.. 1976. Expression of immune mechanisms in the lung. Am. Rev. Respir. Dis. 113: 347-379 [Medline].

4. Mulligan, M. S., J. S. Warren, C. W. Smith, D. C. Anderson, C. G. Yeh, A. R. Rudolph, and P. A. Ward. 1992. Lung injury after deposition of IgA immune complexes: requirements for CD18 and L-arginine. J. Immunol. 148: 3086-3092 [Abstract].

5. Lukacs, N. W., and P. A. Ward. 1996. Inflammatory mediators, cytokines and adhesion molecules in pulmonary inflammation and injury. Adv. Immunol. 82: 257-304 .

6. Graham, B. S.. 1995. Pathogenesis of respiratory syncytial virus vaccine-augmented pathology. Am. J. Respir. Crit. Care Med. 152: S63-S66 .

7. Mosmann, T. R., H. Cherwinski, M. W. Bond, M. A. Giedlin, and R. L. Coffman. 1986. Two types of murine helper T cell clone: I. Definition according to profiles of lymphokine activities and secreted proteins. J. Immunol. 136: 2348-2359 [Abstract].

8. Ryan, U. S.. 1986. Pulmonary endothelium: a dynamic interface. Clin. Invest. Med. 9: 124-132 [Medline].

9. Zimmerman, G. A., S. M. Prescott, and M. McIntyre. 1992. Endothelial cell interactions with granulocytes: tethering and signaling molecules. Immunol. Today 13: 93 [Medline].

10. Imhof, B. A., and D. Dunon. 1995. Leukocyte migration and adhesion. Adv. Immunol. 58: 345-416 [Medline].

11. Donnelly, S. C., R. M. Strieter, S. L. Kunkel, A. Walz, C. R. Robertson, D. C. Carter, I. S. Grant, A. J. Pollork, and C. Haslett. 1993. Interleukin-8 and development of adult respiratory distress syndrome in at-risk patient groups. Lancet 341: 643-647 [Medline].

12. Dragic, T., V. Limin, G. P. Allaway, S. R. Martin, Y. Huang, K. A. Nagashima, C. Cayanan, P. J. Maddon, R. A. Koup, J. P. Moore, and W. A. Paxton. 1996. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature 381: 667-673 [Medline].

13. Deng, H., R. Liu, W. Ellmeier, S. Choe, D. Unutmaz, M. Burkhard, P. DiMarzio, S. Marmon, R. E. Sutton, C. M. Hill, C. B. Davis, S. C. Peiper, T. J. Schall, D. R. Littman, and N. R. Landau. 1996. Identification of a major co-receptor for primary isolates of HIV-1. Nature 382: 661-666 .

14. Alkhatib, G., C. Combadierre, C. C. Broder, Y. Feng, P. E. Kennedy, P. M. Murphy, and E. A. Berger. 1996. CC CKR5: A RANTES, MIP-1alpha, MIP-1beta receptor as a fusion cofactor for macrophage-tropic HIV-1. Science 272: 1955-1958 [Abstract].

15. Moore, K. W., A. O'Garra, R. deWall, Malefyt, P. Vieira, and T. R. Mosmann. 1993. Interleukin-10. Annu. Rev. Immunol. 11: 165-190 [Medline].

16. Kunkel, S. L., R. M. Strieter, N. W. Lukacs, and S. W. Chensue. 1994. Molecular aspects of granulomatous inflammation. EOS. J. Immunol. Immunophar. 14: 71-77 .

17. Gaston, B., J. M. Drazen, J. Loscalzo, and J. S. Stamler. 1994. The biology of nitrogen oxides in the airways. Am. J. Respir. Crit. Care Med. 149: 538-551 [Abstract].

18. Barnes, P. J.. 1990. Reactive oxygen species and airway inflammation. Free Radic. Biol. Med. 9: 235-243 [Medline].

19. Braciak, T. A., K. Bacon, Z. Xing, D. J. Torry, F. L. Graham, T. J. Schall, C. D. Richards, K. Croitoru, and J. Gauldie. 1996. Overexpression of RANTES using a recombinant adenovirus vector induces the tissue-directed recruitment of monocytes to the lung. J. Immunol. 157: 5076-5084 [Abstract].

20. Huffman, J. A., W. M. Hull, G. Dranoff, R. C. Mulligan, and J. A. Whitsett. 1996. Pulmonary epithelial cell expression of GM-CSF corrects the alveolar proteinosis in GM-CSF-deficient mice. J. Clin. Invest. 97: 649-655 [Medline].

21. Doerschuk, C. M., W. M. Quinlan, N. A. Doyle, D. C. Bullard, D. Vestweber, M. L. Jones, F. Takei, P. A. Ward, and A. L. Beaudet. 1996. The role of P-selectin and ICAM-1 in acute lung injury as determined using blocking antibodies and mutant mice. J. Immunol. 157: 4609-4614 [Abstract].

22. Kubo, H., D. Morgenstern, W. M. Quinlan, P. A. Ward, M. C. Dinauer, and C. M. Doerschuk. 1996. Preservation of complement-induced lung injury in mice with deficiency of NADPH oxidase. J. Clin. Invest. 92: 2680-2684 .

23. Clarke, L. L., B. R. Grubb, J. R. Yankaskas, C. U. Cotton, A. McKenzie, and R. C. Boucher. 1994. Relationship of a non-cystic fibrosis transmembrane conductance regulator-mediated chloride conductance to organ-level disease in Cftr(-/-) mice. Proc. Natl. Acad. Sci. U.S.A. 91: 479-483 [Abstract/Free Full Text].

24. Reynolds, H. Y., and H. H. Newball. 1974. Analysis of proteins and respiratory cells obtained from human lungs by bronchial lavage. J. Lab. Clin. Med. 84: 559-573 [Medline].

25. Reynolds, H. Y., J. D. Fulmer, J. A. Kazmierowski, W. C. Roberts, M. M. Frank, and R. G. Crystal. 1997. Analysis of cellular and protein content of bronchoalveolar lavage fluid from patients with idiopathic pulmonary fibrosis and chronic hypersensitivity pneumonitis. J. Clin. Invest. 59: 165-175 .

26. Katzenstein, A. L.. 1985. Pathogenesis of `fibrosis' in interstitial pneumonia: an electron microscopic study. Hum. Pathol. 16: 1015-1024 [Medline].

27. Kuhn, C. III, J. Boldt, T. E. King Jr., E. Crouch, T. Vartio, and J. A. McDonald. 1989. An immunohistochemical study of architectural remodeling and connective tissue synthesis in pulmonary fibrosis. Am. Rev. Respir. Dis. 140: 1693-1703 [Medline].

28. Hunninghake, G. W., J. E. Gadek, T. J. Lawley, and R. G. Crystal. 1981. Mechanisms of neutrophil accumulation in the lungs of patients with idiopathic pulmonary fibrosis. J. Clin. Invest. 68: 259-269 .

29. Antoniades, H. N., M. A. Bravo, R. E. Avila, T. Galanopoulos, J. Neville-Golden, M. Maxwell, and M. Selman. 1990. Platelet-derived growth factor in idiopathic pulmonary fibrosis. J. Clin. Invest. 86: 1055-1064 .

30. Chapman, H. A., C. L. Allen, and O. L. Stone. 1986. Abnormalities in pathways of alveolar fibrin turnover among patients with interstitial lung disease. Am. Rev. Respir. Dis. 133: 437-443 [Medline].

31. Aston, C., J. Jagirdar, T. C. Lee, T. Hur, R. L. Hintz, and W. N. Rom. 1995. Enhanced insulin-like growth factor molecules in idiopathic pulmonary fibrosis. Am. J. Respir. Crit. Care Med. 151: 1597-1603 [Abstract].

32. Broekelmann, T. J., A. H. Limper, T. V. Colby, and J. A. McDonald. 1991. Transforming growth factor ₁ is present at sites of extracellular matrix gene expression in human pulmonary fibrosis. Proc. Natl. Acad. Sci. U.S.A. 88: 6642-6646 [Abstract/Free Full Text].

33. Brody, A. R., P. Soler, F. Basset, W. M. Haschek, and H. Witschi. 1981. Epithelial-mesenchymal associations of cells in human pulmonary fibrosis and in BHT-oxygen-induced fibrosis in mice. Exp. Lung Res. 2: 207-220 [Medline].

34. Cherniack, R. M., T. V. Colbe, A. Flint, W. M. Thurlbeck, J. A. Waldron Jr., L. Ackerson, M. I. Schwarz, and T. E. King Jr.. 1995. Correlation of structure and function in idiopathic pulmonary fibrosis. Am. J. Respir. Crit. Care Med. 151: 1180-1188 [Abstract].

35. Coxson, H. O., J. C. Hogg, J. R. Mayo, H. Behzad, K. P. Whittall, D. A. Schwartz, P. G. Hartley, J. R. Galvin, J. S. Wilson, and G. W. Hunninghake. 1997. Quantification of idiopathic pulmonary fibrosis using computed tomography and histology. Am. J. Respir. Crit. Care Med. 155: 1649-1656 [Abstract].

36. Schwartz, D. A., R. A. Helmers, J. R. Galvin, D. S. Van Fossen, K. L. Frees, C. S. Dayton, L. F. Burmeister, and G. W. Hunninghake. 1994. Determinants of survival in idiopathic pulmonary fibrosis. Am. J. Respir. Crit. Care Med. 149: 450-454 [Abstract].

37. Schwartz, D. A., D. S. Van Fossen, C. S. Davis, R. A. Helmers, C. S. Dayton, L. F. Burmeister, and G. W. Hunninghake. 1994. Determinants of progression in idiopathic pulmonary fibrosis. Am. J. Respir. Crit. Care Med. 149: 444-449 [Abstract].

38. Baumgartner, K. B., J. M. Samet, C. A. Stidley, T. V. Colby, J. A. Waldron, and Collaborating Centers. 1997. Cigarette smoking: a risk factor for idiopathic pulmonary fibrosis. Am. J. Respir. Crit. Care Med. 155: 242-248 [Abstract].

39. Hunninghake, G. W., J. D. Fulmer, R. C. Young Jr., J. E. Gadek, and R. G. Crystal. 1979. Localization of the immune response in sarcoidosis. Am. Rev. Respir. Dis. 120: 49-57 [Medline].

40. Thomas, P. D., and G. W. Hunninghake. 1987. State of the art: current concepts of the pathogenesis of sarcoidosis. Am. Rev. Respir. Dis. 135: 747-760 [Medline].

41. Klein, J. T., T. D. Horn, J. D. Forman, R. F. Silver, A. S. Teirstein, and D. R. Moller. 1995. Selection of oligoclonal V-specific T cells in the intradermal response to Kveim-Siltzbach reagent in individuals with sarcoidosis. J. Immunol. 154: 1450-1460 [Abstract].

42. Lem, V. M., M. F. Lipscomb, J. C. Weissler, G. Nunez, E. J. Ball, P. Stastny, and G. B. Toews. 1985. Bronchoalveolar cells from sarcoid patients demonstrate enhanced antigen presentation. J. Immunol. 135: 1766-1771 [Abstract].