Structural Changes in Airway Diseases^*: Characteristics, Mechanisms, Consequences, and Pharmacologic Modulation FREE TO VIEW

Asthma and COPD are the most prevalent airway obstructive conditions. Both are complex diseases in which inflammatory and remodeling processes have been depicted. Less is known about remodeling features of other airway diseases, such as reactive airways dysfunction syndrome (RADS), cystic fibrosis (CF), and bronchiectasis. In this review, we will describe briefly the main airway remodeling features observed in obstructive diseases—tentatively defined as a change in the nature, content, and distribution of structural airway elements—their potential functional consequences and clinical relevance, as well as current evidence of pharmacologic modulation of these features in airway diseases.

From a histologic perspective, the human airway can be divided into three layers: the inner wall, the outer wall, and the smooth-muscle layers.¹ The inner wall layer refers to the epithelium, the basement membrane, and the submucosa. The outer wall layer consists of loose connective tissue between the muscle layer and surrounding parenchyma. In comparison to healthy airways, all airway layers have been shown to present some alterations in airway obstructive diseases. Epithelial alterations, goblet-cell and submucosal gland hyperplasia, smooth-muscle cell hyperplasia and hypertrophy, subepithelial fibrosis, microvascular proliferation, cartilage changes, airway wall edema, and inflammatory cell infiltration are the main histologic features of airway diseases. The relative magnitude of the structural changes among the airway diseases is summarized in Table 1 .

The ideal approach to assess airway remodeling is histopathologic analysis of lung tissue derived from autopsy or lung resection. However, these specimens are rarely available. Most studies use airway tissues obtained by endobronchial biopsies performed under flexible bronchoscopy. Morphometric measurements of airways are commonly used to describe structural alterations. These measurements are done after standard or specific staining.²³ Transbronchial biopsies have also allowed to evaluate the remodeling process in the small airways.⁴ Other tools such as induced sputum or BAL have been used to assess indirectly some aspects of airway remodeling and surrogate markers evaluated.⁵⁶⁷⁸ However, analysis of human tissues is still the best way to assess structural changes.

The main characteristics of asthma are the presence of respiratory symptoms associated with a variable airflow limitation, airway hyperresponsiveness (AHR), bronchial inflammation, and structural changes. Structural changes of asthmatic airways are represented in Figure 1 . Typical airway remodeling and inflammatory features of asthma may be found in large and small airways.⁹ Loss of epithelial integrity is a common observation in asthmatic airways¹⁰; most studies, using endobronchial biopsies, have observed epithelial shedding. The procedure of tissue sampling has been held responsible, at least in part, for such epithelial detachment, as it has been reported on biopsies in healthy subjects as well.¹¹¹² However, epithelial shedding has also been observed on postmortem studies, BAL, and sputum analysis, suggesting that it is not exclusively an artifact of biopsy sampling and processing but that a weakened attachment of epithelial cells can also be a feature of asthma.¹⁰¹³¹⁴ Using BAL samples from asthmatic patients and normal control subjects, Montefort and colleagues¹⁵ showed that the epithelial cleavage occurs between the suprabasal and basal cell layers and that epithelium integrity is dependent on various adhesion molecules.¹⁶ Integrin α6β4 is predominant on basal cell layer, while desmosomal protein 1 and 2 are prevalent between columnar and basal cell layers in bronchial biopsy specimens from healthy subjects. However, integrin distribution in asthmatic airways is still unknown. Epithelial cells of asthmatic patients show an increased expression of epidermal growth-factor receptor and CD44 proportional to disease severity, potentially playing a role in bronchial epithelial repair.¹⁷¹⁸ These observations therefore suggest that epithelial damage is not only an artifact of sampling.

The layer of epithelial cells lies on a basement membrane that appears thickened in histologic studies. This apparent thickening of the basement membrane results from subepithelial fibrosis at the level of the lamina reticularis, the lower part of the basement membrane area, and this fibrosis may extend into the submucosal layer. Since it was first observed in 1922 (reviewed by Redington and Howarth¹⁹) in fatal asthma, subepithelial fibrosis has been found in asthma of all severities,²⁰²¹ in atopic rhinitic patients,²²²³ and even in young children with difficult-to-treat asthma.²⁴²⁵ This fibrosis results from an increased deposition of collagens I, III, and V, fibronectin, tenascin, lumican, and biglycan.^2627282930 This increase in macromolecules might not only lead to fibrosis but may also be a compartment where adhesion molecules, cytokines, and other inflammatory mediators are stored, perpetuating the inflammation. Goblet cells and submucosal gland hyperplasia are also observed in asthma and are related to mucus overproduction both in adult and child airways, a feature particularly evident in fatal asthma.¹³²⁵³¹

An increase in smooth-muscle mass from both hypertrophy and hyperplasia of smooth-muscle cells is a common observation in asthmatic airways,¹³ and has been described²⁵ in children with difficult-to-treat asthma. Airway smooth-muscle (ASM) cells were initially regarded as only airway bronchoconstrictor cells,³² but there is now strong evidence of additional immunomodulatory functions. ASM cells express cytokines, chemokines, and cellular adhesion molecules that might play a role in submucosal inflammation and airway hyperresponsiveness.³³³⁴ An in vitro study³⁵ revealed a higher basal proliferative rates in asthmatic ASM cells compared with control subjects. The underlying mechanisms leading to such proliferation remain to be elucidated.

Furthermore, airway cartilage is an important determinant of wall stiffness and integrity. Decreased cartilage volume and increased cartilage proteoglycan degradation are seen in asthmatic airways.³⁶ Reduced cartilage integrity may result in a more powerful bronchoconstriction from ASM load reduction.

Changes in the airway wall microvasculature can contribute to airway wall edema and result from angiogenesis. Increased airway vascularity is seen in asthma³⁷³⁸ in association with a greater expression of the vascular endothelial growth factor.³⁹

All previously described structural changes are common to the asthma phenotype and are believed to be the result of the inflammatory process, although other mechanisms may prevail.⁴⁰ Airway inflammation in asthma involves CD4+ T-cells, eosinophils, neutrophils, and mast cells,^4142434445 as well as mediators and T-helper type 2 (Th2) cytokines such as interleukin (IL)-4, IL-5, IL-9, and IL-13, along with transforming growth factor (TGF)-β, granulocyte macrophage-colony stimulating factor (GM-CSF), lipid mediators, and histamine. Many of these mediators or cytokines have potent fibrogenic properties.

Asthma is a highly heterogeneous condition including nonallergic and allergic asthma and presenting in various states of severity. Nonatopic or intrinsic asthma have a similar inflammation pattern as atopic asthmatics, although this group has been less extensively studied. Increased levels of IL-3, IL-4, IL-5, GM-CSF, and eosinophils are found in endobronchial biopsy specimens of nonatopic asthmatics.^46474849 The degree of structural changes seems to be correlated with asthma severity.²⁰²⁶⁵⁰ Airway neutrophilia and an irreversible component of airway obstruction are commonly seen in severe asthmatics,⁵¹ and are typical of other airway diseases such as COPD and bronchiectasis.

Smoking asthmatic patients constitute a specific asthma subgroup in which little is known about airway remodeling, as these patients have usually been excluded from previous studies. They do, however, show various features of both asthma and COPD.⁵²⁵³ A significant increase in elastic fibers was reported in airways of asthmatic smokers, but we need to know more about other changes in this population.⁵⁴

Structural airway changes such as epithelial damage and subepithelial fibrosis seem to be prominent in asthma from occupational origin.²⁰⁵⁵ Initial report⁵⁶ suggested that changes in airway structure were more prominent in occupational asthma than in other types of asthma such as seasonal asthma. This was particularly true in the condition initially called reactive airways dysfunction syndrome. Inhalation of toxic products or high concentrations of a variety of chemical substances with irritant properties can induce airway obstruction and hyperresponsiveness.⁵⁷ RADS was originally described as the onset of respiratory symptoms in the minutes or hours attributable to a single accidental inhalation of high concentrations of irritant gas, aerosol, or particles, followed by asthma-like symptoms and persistent AHR.⁵⁸ It is also called irritant-induced asthma without a latency period,⁵⁹ which includes asthma from repeated exposure to toxic substances. This condition is associated with respiratory mucosal damage. Desquamation of the airway epithelium, squamous cell metaplasia, and a variable subepithelial fibrosis with airway inflammation, particularly made up of lymphocytes, have been described.⁵⁵⁶⁰ Chan-Yeung and coworkers⁶¹ showed airway infiltration by eosinophils and CD8+ T-cells in two thirds of workers affected by irritant-induced asthma, while the other third had airway infiltrated by CD4+ T-cells. Lemière and coworkers⁶² analyzed serial biopsy specimens from a patient acutely exposed to chlorine, and showed evidences of acute inflammation, evolving toward a more chronic process over time, including airway inflammatory and structural changes, although the subepithelial fibrosis seemed more severe than in other forms of asthma.⁵⁶⁶³⁶⁴

COPD is characterized by fixed or partly reversible airflow limitation and clinical symptoms such as increased sputum production, cough, wheezing, and dyspnea. Cigarette smoking is its main etiologic factor. COPD encompasses two main phenotypes: chronic bronchitis and emphysema. Chronic bronchitis features are increased sputum production and cough, usually associated with airflow limitation. Emphysema is characterized by a loss of lung parenchyma, with an enlargement of peripheral air spaces.

In large and smaller airways, bronchial mucosa of COPD is characterized by squamous-cell metaplasia, loss of epithelial cilia, goblet-cell hyperplasia, mucus gland enlargement, smooth-muscle hypertrophy, and inflammatory cell infiltration (Fig 2 ).^65666768 Airway wall fibrosis and stenotic lesions are also observed in small airways and, in addition to a change in the protease-antiprotease balance, are implicated in the development of emphysema.⁶⁹ These pathologic modifications translate into an increase in bronchial secretions and airway obstruction.

Basement membrane thickening is not a characteristic of COPD, although it has been found to a lesser extent than in asthma, particularly in a subgroup of COPD patients having a predominant eosinophilic inflammatory profile.⁷⁰ The COPD inflammatory pattern is characterized by an increase in neutrophils, macrophages, CD8+ T-cells, eosinophils with IL-8, tumor necrosis factor (TNF)-α, leukotriene B₄, and TGF-β being the main inflammatory mediators.^{6671727374757677} In COPD, all these structural changes are mostly attributed to direct injury and inflammation from cigarette smoke components.

CF is a multisystemic disorder affecting children and young adults. Chronic airway obstruction, infection, and exocrine pancreatic insufficiency are the features of CF. Numerous mutations for the CF transmembrane regulator gene are responsible for CF syndrome. CF is a major medical problem leading to severe chronic lung disease in children, with significant morbidity and mortality rates. Many structural alterations characterize CF airways. Goblet-cell and submucosal gland hyperplasia with extension to bronchioles are seen in CF airways. This leads to mucous overproduction and obstruction of small airways leading to bronchiectasis and bronchiolectasis. Peripheral airways of CF patients show increased thickness of inner-wall and smooth-muscle areas when compared with COPD airways.⁷⁸ A greater height of the epithelium is also observed. Under the basal lamina, a dense fibrous deposit of collagens I and III, tenascin, and elastin is seen in bronchial wall of CF.⁷⁹ In the submucosa, a degradation process occurs in regard to elastic and collagen fibers. Loss of cartilage is also found in CF airways and is related to the severity of bronchiectasis.⁸⁰ CF airways are infiltrated with neutrophils and B- and T-cells.⁸¹ IL-8 and leukotriene B₄ are the main inflammatory mediators in CF.⁸² These alterations contribute to airflow obstruction and possibly airway hyperresponsiveness.

Bronchiectasis refers to an abnormal dilatation of bronchi > 2 mm in diameter and is caused by airway wall destruction strongly related to a postinfectious process. They are frequently associated with chronic bacterial infection. An airway inflammatory process mostly implicating neutrophils, CD4+ T-cells, macrophages, and increased IL-8 and TNF-α levels is detected in this disease.⁸³⁸⁴ In bronchiectasis, destruction of bronchial cartilage and smooth-muscle layer, bronchial ulceration, and obstruction of airways by granulomas are common.⁸⁵ Airway remodeling is also present in other airway diseases, but their extent, origin, and clinical consequences have been poorly depicted.

Asthma is considered to be an inflammatory disease of the airways that is mostly triggered by allergen exposure in sensitized individuals, viral infections, and toxic exposures. The Th2-mediated inflammatory response, a complex interaction between a wide network of inflammatory and structural cells, is typical of asthma.⁸⁶ Many asthma-related mediators and cytokines, such as TGF-β, IL-11, IL-17, and leukotriene D₄ have potent remodeling properties. Th2 lymphocytes, mast cells, and eosinophils are important profibrotic inflammatory cells due to their capacity to express potent profibrotic cytokines.⁸⁷⁸⁸⁸⁹ Resident cells such as smooth-muscle cells, epithelial cells, and fibroblasts also have fibrotic properties through their production of extracellular matrix (ECM) proteins and proteases/antiproteases, expression of cytokines, and stimulation of inflammation.^349091929394 The airway fibrotic process has been linked to many cytokines produced by inflammatory cells as well as structural cells. Among them, TGF-β is the most potent and widely studied profibrotic cytokine and is mainly produced by eosinophils.⁸⁸ TGF-β increases fibroblast production of ECM proteins including collagen I, collagen III, and fibronectin,⁹⁵⁹⁶ and decreases collagenase levels in in vitro models.⁹⁷ In in vitro or animal models,^{9899100101102103} IL-4 and IL-13 have shown similar fibrotic properties. IL-11 is found in increased expression in the airways of patients with severe asthma but not in those of mild asthmatic patients or healthy control subjects, and IL-11 is believed to contribute to fibrosis.¹⁰⁴ IL-17 is produced by T-cells and eosinophils and induces bronchial fibroblasts to increase their expression of IL-11 and IL-6 (T-cell and B-cell growth factors).⁸⁹ Lipid mediators have also been linked to asthma remodeling in in vitro models in which leukotriene C₄ increased collagenase expression by lung fibroblasts.¹⁰⁵ Endothelin is another mediator potentially involved in airway remodeling, although its role remains to be determined.¹⁰⁶¹⁰⁷

Besides the increase in synthesis of ECM proteins, reduced collagen degradation may also result in fibrosis. In asthma, it is believed that the protease/antiprotease balance is on the profibrogenic side. Interstitial cells, macrophages, and neutrophils are the major sources of proteases and antiproteases. The activity of proteases is mainly regulated by antiproteases. Matrix metalloproteinases (MMPs) are a family of proteases implicated in collagen degradation. MMP-2, MMP-3, MMP-8, and MMP-9 are the MMPs related to asthma.^5108109110 Among these, MMP-9 has been intensively linked to asthma.^5111112 MMP-9 levels are significantly higher in the sputum of asthmatic patients compared with control subjects, but levels of MMP-9 inhibitor tissue inhibitor metalloproteinase (TIMP)-1 are similar to those found in healthy control subjects.⁵ This imbalance between MMP-9 and TIMP-1 leads to a profibrotic MMP-9/TIMP-1 ratio. Moreover, MMP-9 level is increased in asthma exacerbation,^108113114 and after allergenic challenge,^115116117 and is decreased by corticosteroids.¹¹³¹¹⁴ MMPs are implicated in airway inflammation through their influence on eosinophil trafficking¹¹¹ and in airway remodeling not only by matrix reorganization but also by its alteration of the angiogenesis¹¹⁸ and smooth-muscle hyperplasia processes.¹¹⁹¹²⁰

Th2 cytokines, including IL-4, IL-5, IL-9, and IL-13, lead to goblet-cell metaplasia in animal models.¹²¹ IL-9 appears to have a major role and has been shown to induce mucus gene expression in airway epithelial cells.¹²² A recently discovered cytokine, IL-25, triggers epithelial cell hyperplasia and increases mucus secretion in experimental animal models of asthma following intranasal administration.¹²³ Various mediators such as TGF-β, IL-13, platelet-derived growth factor, and basal fibroblast-derived growth factor (bFGF), among others, have been implicated in smooth-muscle proliferation.^124125126Moreover, human ASM cells from asthmatic patients have a high rate of basal proliferation compared with normal control subjects.³⁵ A greater vascular endothelial growth factor, angiogenin, and bFGF expression leading to angiogenesis³⁹ may explain the increased airway vascularity observed in asthma.³⁷³⁸ Cytokine production by inflammatory cells as well as structural cells may be the main factor responsible for airway remodeling changes in asthma.

The airway damage in RADS is the result of either an acute inflammatory response or a direct toxic effect.¹²⁷ The induced damage to the airway epithelium may also trigger an acute airway repair process, which can lead to permanent airway remodeling if the inflammatory process is severe or if the host response is abnormal.

In COPD, inflammatory and structural cells also produce remodeling-associated cytokines. A few inflammatory mediators seem to play an important role in triggering the cascade of events leading to COPD. IL-8, TNF-α, TGF-β, and leukotriene B₄ are mostly reported in COPD inflammation.^{128129130131132} Understanding the exact implication of mediators and finding new pertinent mediators will help to improve our knowledge of COPD pathogenesis. Many of these cytokines have chemotactic properties to inflammatory cells, but little information is available on the mechanisms leading to irreversible bronchial structural alterations in COPD. Pathogenesis of COPD can be summarized as being an induced oxidative stress for epithelial cells and macrophages caused by cigarette smoke. Then, the oxygen-free reactive radicals trigger an inflammatory pattern through chemokine production by epithelial cells leading to neutrophil, CD8+ T-cell, eosinophil, and macrophage recruitment. The inflammatory process is believed to induce structural cell apoptosis and tissue degradation leading to chronic bronchitis and emphysema. Epithelial cells release other mediators (insulin-like growth factor-1, prostaglandin, fibronectin), thereby activating fibroblast recruitment and proliferation, matrix synthesis and, therefore, remodeling. In a murine model of emphysema, interferon-γ produced by CD8+ T-cells triggers COPD remodeling by modulation of protease and antiprotease synthesis.¹³³ Proteases are produced by inflammatory cells as well as resident cells. The reported proteases and antiproteases implicated in COPD include neutrophil elastase, proteinase 3, cathepsins, MMP-1, MMP-9, MMP-2, α₁-antitrypsin, secretory leukocyte proteinase inhibitor (SLPI), elafin, and TIMP.¹³⁴ MMP-2 and MMP-9 are increased in COPD BAL fluid, with a greater increase being observed in the levels of MMP-9 as demonstrated by zymography studies.¹³⁵ Furthermore, an increased level of MMP-1, MMP-2, MMP-9, and MMP-8 is found by immunocytochemistry in epithelial cells, fibroblasts, endothelial cells, macrophages, and neutrophils.¹³⁵ MMP-9 and MMP-8 are mainly expressed by neutrophils, while MMP-1 and MMP-2 are restricted to macrophages and epithelial cells. All these various mechanisms associated with cigarette smoke may lead to lung remodeling in COPD.

However, it is not only the inflammatory or toxic airway insult that could per se trigger an abnormal airway repair process. It has been proposed that an alteration of the relationships between the epithelium and airway myofibroblasts (epithelial/mesenchymal trophic unit) could lead to the structural changes in asthma.¹³⁶ According to this hypothesis, epithelial cells present an abnormal response to various stimuli by altering the repair process through epidermal growth factor, TGF-β, and Th2 cytokine production.¹³⁷ If the epithelium/myofibroblast communication is perturbed, it may induce a phenotypic change of myofibroblasts and lead to smooth-muscle proliferation.¹³⁸ This perturbed communication in combination with wound healing-associated mechanical stress stimulates the development of differentiated myofibroblasts, which then show a de novo expression of α-smooth-muscle actin and increased expression of ED-A fibronectin. Furthermore, the epithelial/mesenchymal signaling to the airway environment might lead to the proliferation of fibroblasts, microvessels, and nerves and may play a role in the persistence of inflammation.¹³⁶

Otherwise, we have evidence that mechanical stimuli also lead to airway remodeling. Bronchoconstriction produces a folding of the airway wall, inducing a stress on the epithelial layer.¹³⁹ Such stress stimulates epithelial cells to produce factors influencing fibroblasts and smooth-muscle cells toward a profibrotic profile. Pressure stresses on cultured rat tracheal epithelial cells increase the expression of profibrotic mediators such as TGF-β₁.¹⁴⁰ Mechanical stimuli have a direct effect on other resident cells. Fibroblasts increase the production of fibronectin, collagens III and V, and MMP-9 following mechanical stress.¹⁴¹ Strained asthmatic bronchial fibroblasts have an increased expression of decorin and versican, while nonasthmatic cells have only an up-regulation of versican.¹⁴²

Besides the pathologic and physiologic aspects of asthma, the remodeling process may potentially be influenced by genetic determinants. The asthma phenotype will develop in genetically susceptible individuals exposed to environmental triggers. Genome screening has led to identification of genes or cluster of genes relevant to asthma and atopy.¹⁴³¹⁴⁴ Among them, a disintegrin and metalloproteinase (ADAM-33) has been a focus of interest in the last few years. ADAM-33 has been linked to asthma in a study of 460 white families.¹⁴⁵ ADAM-33 is a membrane-anchored metalloprotease, and abnormal activity of such enzymes can lead to altered airway function, inflammation, and remodeling. In an expression microarray study using tissue obtained from bronchial biopsies of healthy control subjects and of subjects with allergic asthma, Laprise et al¹⁴⁶ found 79 genes showing significant differences in expression in asthmatic compared to control subjects. This type of analysis can potentially guide future research on physiopathology of asthma.

We previously showed that subjects with allergic rhinitis without asthma or AHR display evidence of airway remodeling, although to a lesser extent than in asthma.²² Furthermore, subjects with asymptomatic AHR show focal bronchial subepithelial fibrosis on bronchial biopsies, and when respiratory symptoms develop, the subjects show evidence of an increase in subepithelial fibrosis and in the number of CD25/CD4-positive T-cells.¹⁴⁷ These observations suggest that the structural airway changes occur even before the development of symptomatic asthma and may be involved at an early stage of the development of the disease when the repair process is activated. An abnormal repair process may result in permanent changes leading to or promoting the development of chronic airway diseases. With regard to the influence of early events in childhood on future airway structural changes, Rasmussen et al¹⁴⁸ reported a greater decline in postbronchodilator FEV₁/FVC ratio from age 9 to age 26 in those with asthma and an initial low pulmonary function, and suggested that airway remodeling in asthma, as shown by a reduced pulmonary function, begins in childhood and persists into adulthood.

It has also been proposed that the remodeling process might be beneficial in airway diseases. Lambert and colleagues¹⁴⁹ suggested that airway wall thickening protects against bronchoconstriction. Peribronchiolar fibrosis correlated with milder COPD stage, suggesting that fibrosis can be protective, preventing narrowing.⁶⁷ Collagen deposition in subepithelial matrix may inhibit narrowing by making the airway wall stiffer, representing an additional load on ASM.²³ Hyaluronan and versican deposition in and around the smooth muscle also counteracts airway narrowing and smooth-muscle shortening. Experiments conducted on rat models of allergen-induced asthma have suggested that airway responsiveness may increase following airway inflammation but may decrease with the development of airway remodeling.¹⁵⁰

Duration of asthma has been associated with reduced lung function, increased hyperresponsiveness and asthma symptoms, and greater use of medication.^2021151 Airway remodeling may contribute to these features. The extent of epithelial injury has been statistically correlated to AHR.²⁰¹⁵² Functional consequences of epithelial alterations mostly refer to increased sputum production and airway narrowing.¹³

Some variable correlations have been found between the severity of asthma, AHR or attack score, and subepithelial collagen types I and III deposition in the airways.^104153154 Proteoglycan deposition in the ECM and bronchial fibroblast production of proteoglycans also correlate with airway responsiveness in asthmatic subjects.²⁶⁹⁰ Niimi and colleagues¹⁵⁵ showed than increased airway thickness measured by helical CT correlated inversely with AHR, suggesting that it is not only airway dimension that may affect airway function but probably other factors.

Smooth-muscle cell hyperplasia and hypertrophy have been linked to asthma severity.¹⁵⁶ Functional consequences of this increase in airway smooth-muscle mass have been proposed, such as airflow obstruction through airway wall thickening and increased airway responsiveness in both asthma and COPD.⁶⁶¹⁵⁷

Cartilage degradation can contribute to chronic airway obstruction and allow more powerful bronchoconstriction for a given degree of ASM contraction.¹⁵⁸ Peribronchiolar fibrosis occurring in COPD small airways is believed to lead to centrilobular emphysema, resulting in loss of alveolar attachments and thereby contributing to a loss of elastic recoil and early closure of bronchioles during expiration.⁶⁹ It has been proposed that small airway alterations might be the earlier stage of COPD, a theory supported by the clinical observation of isolated abnormalities of forced expiratory flow, midexpiratory phase in the pulmonary function test of smokers.¹⁵⁹ However, this has never been proven.

The number of infiltrating leukocytes, such as mast cells, eosinophils, and CD8+ and CD45+ T-cells, correlates with AHR in patients treated with inhaled corticosteroids (ICS).¹⁶⁰ Sputum MMP-9 level is positively correlated with the FEV₁ fall after allergen challenge and asthma severity.¹¹¹¹¹⁷ The BAL level of MMP-8 inversely correlates with FEV₁ in asthmatic patients.¹⁰⁹

Other authors have also linked inflammation and structural alterations with the clinical stage of COPD. Sputum neutrophilia and CD8+ T-cells, macrophages, goblet-cell numbers, CD45+ cells, natural killer lymphocytes, and macrophage inflammatory protein-1α+ epithelial cells in bronchial biopsy specimens have all shown a negative relationship with FEV₁.^161162163 An increase in CD8+ T-cells in subepithelium of COPD-derived bronchial tissues is associated with the total pack-years of smoking.¹⁶² Sputum MMP-9 level is positively linked to neutrophil numbers and is negatively correlated with airflow obstruction in COPD.¹⁶⁴

According to previous observations, structural alterations, combined with the inflammatory process, appear to be related to the magnitude of functional abnormalities in COPD and asthma (summarized in Table 2 ). How the structural changes could lead to changes in airway function, however, remains uncertain. It is possible that the changes observed in the airway wall structure in asthma influence the above mechanisms and that the loss of distensibility of the airways or other factors cause a persistent impairment of the smooth-muscle stretch-relaxation protective mechanism.

The Childhood Asthma Management Program study¹⁵¹ has demonstrated an association between asthma duration and reduced lung function, higher AHR, greater asthma symptomatology, and higher use of β₂-agonists. Furthermore, Lange and colleagues¹⁶⁵ showed an accelerated decline in lung function in asthmatic compared to nonasthmatic subjects, with the decline being more severe in smokers. These observations could relate to airway remodeling. Two studies²⁴²⁵ have compared bronchial biopsy specimens of children with difficult-to-treat asthma with those of adult asthmatic patients and found no difference regarding the extent of reticular basement membrane thickness. Furthermore, no correlations were found in terms of age, symptom duration, lung function, or eosinophilic inflammation. Further studies looking at long-term effects of airway remodeling are needed to determine its role in altering lung function in asthma. As lung function is irreversibly changed in COPD, identifying correlations between remodeling features and lung function will be of great interest. However, the exact functional consequences of airway remodeling are largely unknown.

Structural airway changes likely explain the development of a fixed component of airway obstruction, mostly in COPD but also in some asthmatic subjects.¹⁶⁶ Furthermore, the changes in ECM observed in asthma and other conditions may also contribute to the persistence of airway inflammation through an increased storage of cytokines or other mechanisms.¹⁶⁷

Many of the changes described above are part of an abnormal or exaggerated response to an airway insult, and the type and time-course of these alterations will vary according to the type of offending agent, the duration of the process, and the genetic predisposition of the individual for such a response. Airway remodeling, as bronchial inflammation, may be observed in the absence of clinical manifestations, and there may be a threshold after which their combined influence on airway function will be sufficient to induce respiratory symptoms; in asthma, this may occur after a period of subclinical AHR.¹⁴⁷ The clinical relevance of the various airway structural changes remain to be determined, and the correlations with epidemiologic data remain unclear. Although some changes may be protective, they most probably contribute to the persistence of altered airway physiology, particularly AHR, an accelerated decline in lung function, and the development of an irreversible component of airway obstruction. For a certain degree of remodeling, however, changes in the contractile properties of the airways and inflammation may modulate the expression of the disease. It remains possible that if we could block the remodeling response to inflammatory processes, the development of diseases such as asthma or COPD could be prevented. Nevertheless, these hypotheses remain to be explored. Apart from inhibiting inflammation, it is possible that in the future, the transition toward remodeling will be preventable by pharmacologic interventions or gene therapy.

Control of persistent asthma can be usually achieved by early treatment and maintenance therapy with ICS.¹⁶⁸ Some evidence supports the role of ICS in preventing or reducing the decline of airway function in asthmatic patients,¹⁶⁹¹⁷⁰ with a greater benefit if therapy with ICS is started early after asthma diagnosis.¹⁷¹ However, the Childhood Asthma Management Program study¹⁷² reported no significant long-term prevention in the decline of lung function with ICS in children presenting mild-to-moderate asthma. Whether the decline in lung function is preventable in asthma needs to be confirmed. ICS has no effect on FEV₁ decline, respiratory symptoms and exacerbations in mild to moderate COPD,¹⁷³ while in moderate to severe COPD, a positive effect is seen.¹⁷⁴ This reflects the facts that inflammation in COPD is different from asthma-associated inflammation and that damages are mostly irreversible. In CF, oral corticosteroids appear to slow the progression of the disease.¹⁷⁵

ICS are the mainstay of asthma therapy and are currently the most effective way to control inflammation, particularly of the eosinophilic type.¹⁷⁶¹⁷⁷ Reduced number of airway mast cells, eosinophils, CD4+ T-cells, and CD8+ T cells by either ICS or oral corticosteroids is well documented in asthma disease.^{178179180181182} Oral corticosteroids or ICS also decrease the messenger RNA expression of GM-CSF, IL-4, IL-5, and increase interferon-γ.¹⁷⁸¹⁷⁹ Therefore, ICS could possibly influence remodeling through their antiinflammatory effects. However, ICS are much less effective in reversing airway remodeling than inflammation. Oral corticosteroids significantly reduced the expression of two profibrotic cytokines, IL-17 and IL-11, in patients with moderate-to-severe asthma but interestingly had no effect on TGF-β expression.¹⁸³ In RADS, despite corticosteroid treatment, recovery is often incomplete with persistence of inflammation and airway fibrosis.⁶²

The effects of ICS on airway inflammation have been studied on COPD and are disappointing. No reduction in CD8+ T-cells, CD68+ cells, or neutrophils is seen after ICS treatment.¹⁸⁴ Decreased CD8:CD4 ratio in the epithelium, increased neutrophils, and reduced number of mast cells in the subepithelial area following three months of ICS treatment were found in COPD endobronchial biopsies.¹⁸⁴¹⁸⁵ In COPD, it was reported that high doses of ICS had no effect on sputum inflammatory cell number, IL-8 levels, elastase activity, MMP-1, MMP-9, SLPI, or TIMP-1.¹⁸⁶ Systemic corticosteroids are used in a short-course pattern in exacerbated COPD, improving lung function and clinical outcomes.¹⁸⁷¹⁸⁸ However, the antiinflammatory effect of such a short course of systemic corticosteroid treatment has not been addressed in COPD patients. As CD4+ T-cells and mast cells have not been linked to COPD inflammation, there are no convincing data on the potential therapeutic role of corticosteroids in COPD with the exception of exacerbated COPD and COPD with increased eosinophils.^70189190 In CF, ICS failed to reduce airway inflammation.¹⁹¹ In bronchiectasis, corticosteroids decreased T-cell infiltration and IL-8 levels,⁸⁴ but clinical benefits are still not clear.

Corticosteroids have been reported to induce apoptosis in airway epithelial cells.¹⁹² Corticosteroids also reduced IL-6 cytokine production but not TNF-α–mediated increases in IL-8 or GM-CSF secretion in cultured epithelial cells.¹⁹³ This might explain why corticosteroids fail to control neutrophilic inflammation in airway diseases. In another study,¹⁹⁴ corticosteroids decreased chemokine production including growth-related oncogene-α and IL-8 by bronchial epithelial cells. From a clinical perspective, ICS were found to decrease mucus production in asthmatic patients.¹⁹⁵ From the various models studied, the effect of corticosteroids on epithelial cells is still controversial, but they may be beneficial through a decrease in mucus production.

The effect of corticosteroids on fibroblasts has been poorly studied. Cultured nasal fibroblasts stimulated with bFGF or TNF-α have reduced proliferation rate and intracellular adhesion molecule-1 expression when treated with fluticasone.¹⁹⁶ In the same way, bFGF-induced proliferation and DNA synthesis or TNF-α–induced intracellular adhesion molecule-1, monocyte chemotactic protein-1, eotaxin, and IL-6 release in human fetal lung fibroblasts can be reduced by corticosteroids.¹⁹⁷ In an in vitro study,¹⁹⁸ dexamethasone increased 3H-thymidine incorporation (promoting their proliferation) in asthmatic bronchial fibroblasts without having any significant effect on normal fibroblast proliferation. Future studies will help us to understand the precise effect of corticosteroids on fibroblasts.

In vitro studies^199200201 of the use of corticosteroids show a decrease in smooth-muscle cell proliferation. The antiproliferative effect of corticosteroids can therefore benefit asthmatic patients through smooth-muscle mass reduction if it is also effective in vivo. As smooth-muscle cells produce inflammatory mediators, corticosteroids can also reduce cytokine and chemokine production at this level, but this has to be confirmed. Some studies²⁰²²⁰³ have shown that corticosteroids have no effect on the activation of the nuclear factor-κB transcription factor, suggesting a possible immunosuppressive effect by corticosteroids on ASM. Corticosteroids were also ineffective in reducing ECM protein production by human ASM.²⁰⁴ Thus, corticosteroids may improve asthma control through decreased smooth-muscle cell proliferation but are less effective in modulating the synthesis of ECM proteins and cytokines.

Tenascin, an ECM glycoprotein, appears to be responsive to corticosteroid treatment as a short-course of ICS (4 to 6 weeks) significantly decreased tenascin immunoreactivity in the airways of chronic asthmatic patients.²⁷

As subepithelial fibrosis is a major histologic feature of asthma with significant clinical consequences, the effect of corticosteroid treatment on this feature has been assessed. Jeffery,²⁰⁵ Lundgren et al²⁰⁶ and Boulet et al¹⁸⁰ reported no change in basement membrane thickness following long (3 year, 7 year, and 10 year) or short-term (8 weeks) use of ICS. In other studies, treatment with ICS lasting 6 weeks,²⁰⁷ 4 months,¹⁷⁸ or 6 months²⁰⁸ resulted in a modest decrease in basement membrane thickness. With ICS treatment of 1-year²⁰⁹ and 2-year²¹⁰ durations, a modest decrease in basement membrane thickness was observed in asthmatic patients. Sont et al²¹⁰ described a decrease in reticular basement membrane (RBM) thickness in a group of patients receiving a dose of ICS equivalent to 800 μg/d of budesonide. This finding has led to the hypothesis that ICS might be effective in reducing RBM thickness when used for a long period of time and at a higher dose. In a group of patients with moderate-to-severe asthma, no significant differences were seen in types I and III collagen and TGF-β immunoreactivity after a 2-week course of oral corticosteroids.¹⁸³ The inability of ICS to inhibit TGF-β expression may be responsible for the persistent fibrosis seen in this group of patients with severe asthma. The fact that doses of ICS could have been too low and the duration of treatment could have been too short has been suggested as an explanation for the poor response.

Finally, Ward and coworkers²⁰⁹ reported that the variability in AHR can be partly explained by the RBM thickness, the number of BAL epithelial cells, and BAL eosinophils. Part of the improvement in AHR provided by ICS was related to early changes in inflammation, but a progressive and larger improvement was associated with subsequent changes in airway remodeling.

As an imbalance in MMP/TIMP can lead to fibrosis, some studies addressed the effects of corticosteroids on MMP/TIMP. Decreased MMP-9 expression and increased TIMP-1 expression were observed on bronchial biopsy specimens after a 6-month ICS treatment in patients with mild asthma,²⁰⁸ suggesting that ICS can effect the MMP/TIMP ratio leading to a profibrotic effect. However, ICS decreased BAL fluid TIMP-1 in a subgroup of patients with moderate-to-severe asthma treated with corticosteroids when compared with noncorticosteroid-treated patients with mild-to-moderate asthma.¹¹² In another study,²¹¹ 4 weeks of ICS treatment did not affect MMP-9 or TIMP-1 levels in sputum and BAL fluid of patients with mild asthma patients. A correlation has been made between oral corticosteroid responsiveness and blood MMP-9/TIMP-1 ratio in patients with moderate-to-severe asthma.²¹² A low MMP-9/TIMP-1 ratio suggested a predominant fibrogenic process over the inflammatory process, explaining the poor efficiency of corticosteroids. Duration of treatment and asthma severity seemed to modulate the effect of corticosteroids on MMP/TIMP. Finally, in COPD, high doses of ICS had no effect on sputum elastase, MMP-1, MMP-9, SLPI, or TIMP-1 levels.¹⁸⁶

Using high-magnification bronchovideoscopy, control subjects and patients with COPD had similar vessel densities, whereas these densities were significantly increased in asthmatic patients.³⁸ A 5-year course of ICS did not decrease the density of subepithelial vessels. However, these observations were limited to the tracheal wall. In another study, a 6-week course of high dose of ICS was associated with a decrease in the number of vessels and in the total vascular area in bronchial biopsy specimens in patients with mild-to-moderate asthma.²¹³ Overall, the influence of corticosteroids seems modest in remodeling (Table 3 ), although the effect seemed more marked with longer durations of administration.

Montelukast, a cysteinyl leukotriene-1 receptor antagonist, decreases sputum eosinophils after allergen challenge in asthmatic patients.²¹⁴ The cysteinyl leukotriene antagonist may play an important role in the pathogenesis of airway remodeling in addition to its antiinflammatory effects. Montelukast has been shown to significantly inhibit ovalbumin-induced airway smooth-muscle hyperplasia and subepithelial fibrosis in sensitized mice²¹⁵(Table 3). Studies on asthmatic subjects are needed, however, to confirm the antiremodeling effect of cysteinyl leukotriene-1 receptor antagonists.

Antihistamines are used mostly to treat atopic diseases other than asthma. It is believed that antihistamines might be beneficial in asthma. An in vitro study²¹⁶ shows an antiinflammatory role of antihistamines through a decrease in the migration and proliferation of eosinophils and an inhibition of mast-cell and basophil mediator release. In vivo, epithelial cells express fewer adhesion molecules after second-generation antihistamine treatment.²¹⁷ In a murine model of asthma, antihistamines inhibited the Th2 response, lung inflammation, and AHR.²¹⁸ Antihistamines further afforded bronchoprotection against methacholine challenge.²¹⁹ Antihistamines could possibly be useful in preventing the development of asthma in certain patients as reported in cohorts of children with atopic dermatitis.²²⁰ Other antiallergic molecules such as cromolyn and nedocromil stabilize mast-cell membrane and therefore decrease mediator release.²²¹²²² Although it might be of interest to determine whether they could play a role in the prevention of remodeling, up to now we have no data indicating that they may affect bronchial structural changes (Table 3).

β₂-Agonists reduce airway muscular tone and improve expiratory flow. Combinations of long-acting β₂-agonists (LABAs) and low-dose ICS result in a similar control of the disease and reduction of induced-sputum inflammatory features compared with moderate doses of ICS after 1 year of treatment.²²³ LABAs alone modestly decrease airway eosinophilia.²²⁴ The number of airway total leukocytes and mast cells increases after allergen challenge in patients receiving a LABA only.²²⁵ β₂-Agonists induce a change in response toward a Th2 profile in human peripheral blood mononuclear cells by increasing Th2 cytokine expression.²²⁶ Overall, β₂-agonists are not considered to have significant antiinflammatory properties. In regard to remodeling, however, until recently we had little evidence that they may affect airway structure. However, one study²²⁷ suggested that salmeterol could lead to a reduction in airway vasculature. As for theophylline, although an immunomodulatory effect has been suggested,²²⁸ we have no evidence regarding its influence on airway remodeling, although long-term studies have yet to be conducted.

Methotrexate is a folic acid antagonist that inhibits thymidine synthesis and thereby interferes with DNA synthesis. There is no consensus on the use of methotrexate in asthma, and nothing is known about its effects on remodeling.²²⁹ Azathioprine has no proven benefit in asthma. Cyclosporin A has inhibitory effects on mast cells, monocytes, neutrophils, basophils, and T-cells, and might have some benefit in severe corticosteroid-dependent asthma but no effect on remodeling has been described in humans.²³⁰ However, in a cat model of asthma, cyclosporine A decreased inflammation and remodeling processes such as ASM hypertrophia and goblet-cell and submucosal gland hyperplasia. IV Igs have antiinflammatory properties as well; they may decrease the use of corticosteroids and reduce respiratory symptoms.²³¹²³² Potential antiremodeling effects have not been assessed.

The main target of new drugs for the prevention and treatment of asthma or COPD is the inflammatory process. Some agents, such as the CpG oligonucleotides or bacille Calmette-Guérin vaccine, are thought to be able to switch immune responses from a Th2 to a T-helper type 1 (Th1) profile. Anti–IL-5 antibodies target eosinophil-mediated inflammation. Rapamycin, a macrolide analog, has immunosuppressive effects and could influence inflammation and remodeling in experimental asthma mouse models. Selective phosphodiesterase inhibitors, through the breakdown of intracellular cyclic adenosine monophosphate, have bronchodilatory, antiinflammatory, and potential antiremodeling properties.

Soluble recombinant human IL-4 receptor has been used in humans and seems to have the same effectiveness on clinical outcomes as ICS.²³³ Unfortunately, no observations to date in either humans or animals have addressed an antiremodeling effect. In animal models of asthma, IL-4 receptor antagonists inhibit airway inflammation and AHR.²³⁴²³⁵ Monoclonal antibodies against IL-5 were shown to be effective in reducing the deposition of ECM proteins tenascin, lumican, and procollagen in the basement membrane of patients with mild asthma.²³⁶ In addition, anti–IL-5 reduces blood and sputum eosinophilia.²³⁷ Subepithelial fibrosis prevention has been observed with anti–IL-5 in a mouse model of asthma.²³⁸ These data suggest that these agents may influence remodeling, which might be partly preventable or reversible.

As airway inflammation in asthma is considered to be under control of Th2 lymphocytes, new therapeutic interventions that may reverse the Th2 pattern have been conceived. Bacterial DNA, precisely CpG motifs, promotes Th1 immune response in animal models. Synthesized CpG-oligonucleotides, known as immunostimulatory DNA sequence, are currently under study in allergic diseases. Studies²³⁹²⁴⁰ in murine models showed that CpG treatment reduced airway eosinophilia, IL-5, IL-4, IL-13, and GM-CSF production, and increased interferon-γ release. CpG oligonucleotide treatment induces a Th1 response through activation of macrophages and dendritic cells, resulting in increased IL-12 production. This leads to a higher interferon-γ/IL-4 ratio, resulting in a Th1 profile. Animal studies²⁴¹²⁴² on mouse and monkey models report a thinner subepithelial fibrosis and weaker goblet-cell hyperplasia following immunostimulatory DNA sequence treatment. Modulating Th2 toward Th1 pattern by CpG oligonucleotides therapy may possibly reduce the fibrotic process.

Treatment with anti-IgE reduces blood IgE, slightly decreases asthma symptoms and corticosteroid use, but has no effect on lung function.²⁴³ Treatment with anti-IgE (omalizumab) significantly reduces the airway eosinophilia in both the induced sputum, and epithelium and submucosal layers.²⁴⁴ A decrease in circulating cytokines IL-13, and IL-5 is observed with anti-IgE, while no modulation of IL-6, IL-10, and serum intracellular adhesion molecule can be detected.²⁴⁵ Anti-IgE presents potential antiinflammatory effects, but no antiremodeling property has been reported yet.

Selective phosphodiesterase inhibitors have been designed and studied, mainly in COPD. The breakdown of intracellular cyclic adenosine monophosphate following phosphodiesterase inhibition has been suggested to explain bronchodilatory and antiinflammatory effects and could lead to potential antiremodeling properties. The phosphodiesterase-3 inhibitor siguazodan has been shown to reduce in vitro proliferation of human ASM.²⁴⁶ In addition, the phosphodiesterase-4 inhibitor roflumilast reduced inflammation, subepithelial collagen deposition, and thickening of airway epithelium in an asthma mouse model.²⁴⁷

Rapamycin, SAR 943, a macrolid analog, decreases epithelial growth factor-induced proliferation in cultured human ASM cells.²⁴⁸ It has no effect on epithelial cells. In an asthma mouse model, rapamycin decreased fibronectin, mucus-containing cells, IL-4 and IL-5, the number of airway eosinophils, neutrophils, and lymphocytes in BAL fluid, and decreased airway response to methacholine.²⁴⁸

Bronchothermoplasty is an original mode of intervention that consists of applying an electric current to segmental and subsegmental bronchi, with the goal of destroying the smooth muscle and therefore reducing its capacity to contract. Although studies should determine the usefulness of this treatment, it has been shown to alter airway structure in a possibly beneficial way. In fact, Cox and colleagues²⁴⁹ report a persistent improvement of 2.9 doubling dose of methacholine bronchoprovocation test after 1 year of treatment.

Airway remodeling is a complex phenomenon that includes a variety of changes whose specific contributions to the change in airway function need further analyses. The time course and mechanisms by which such changes translate into modifications of airway responsiveness and symptomatic airway obstruction are to be evaluated more extensively in term of effects of the different treatments used for these conditions. Various methodologic aspects should, however, be taken into account when studying these effects.

For example, some agents could possibly be more effective at preventing than reversing airway remodeling, and the duration of treatment and doses of the agents should be considered, as well as the type and duration of the disease. Research on gene therapy in lung diseases is currently under scrutiny. Other than CF or α₁-antitrypsin deficiency, it is highly unlikely that a single vector will be therapeutic, as multiple genes are implicated in asthma and COPD diseases. Gene therapy has been studied mostly in relation to CF.²⁵⁰ Vectors and incorporating techniques need to be optimized to prove helpful in treatment of CF patients. Overall, primary prevention of airway disease is likely to be the most effective tool and can be achieved by allergen avoidance,²⁵¹ smoking avoidance, control of infections, or prenatal diagnosis. Treatment of other allergic diseases might possibly prevent the development of asthma, especially in allergic rhinitic subjects²⁵²²⁵³ and in subjects with asymptomatic AHR and has to be further studied.²⁵⁴

Although this is still controversial, we have evidence that the airway remodeling process may have untoward consequences in obstructive diseases, in the clinical expression of the disease, in its development, and in the decline in pulmonary function. The common factor underlying all the structural changes in airway diseases is an injury/repair process. In asthma, the damage follows an allergenic or nonallergenic Th2 inflammation and mechanical stress. In COPD, the initial trigger is cigarette smoke inducing direct cell toxicity and inflammatory response. An infectious process is both at the origin of the inflammation observed in CF and bronchiectasis patients. More research should be done to identify key changes, effective treatments, and proper interventional timing to counteract these changes. The prevention of the development of asthma, COPD, or other airway diseases through early interventions on structural changes is an exciting avenue. The potential of novel therapeutic agents to reverse or prevent airway remodeling warrants further evaluation.