The Many Faces of Airway Inflammation
Asthma and Chronic Obstructive Pulmonary Disease

P. M. O'BYRNE and D. S. POSTMA

Asthma Research Group and the Department of Medicine, McMaster University, Hamilton, Ontario, Canada; and Department of Pulmonary Diseases, University of Groningen, Groningen, The Netherlands

	INTRODUCTION

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

Airway diseases, predominantly asthma and chronic obstructive pulmonary disease (COPD), are among the world's most prevalent diseases. The prevalence of asthma has been increasing over the past 20 yr in most countries where this has been studied, and it affects up to 10% of the populations of most developed countries. COPD is the sixth cause of death in the world and affects 4-6% of people more than 45 yr of age. In 1985, more than 5.4 million Americans were estimated to have COPD and COPD is the most rapidly rising cause of death among individuals over the age of 65 yr in the United States (1). These diseases constitute a major financial burden to society, with both direct and indirect costs.

Both asthma and COPD are identified by the presence of characteristic symptoms and functional abnormalities, with airway obstruction being the sine qua non of both diseases. The airway obstruction in asthma must be reversible to establish a diagnosis, whereas COPD is defined as a syndrome characterized by abnormal tests of expiratory flow that do not change markedly over periods of several months of observation (2). Both diseases are now known to be caused by lung inflammation induced by different initiating factors, most likely environmental allergens, occupational sensitizing agents, or viral respiratory infections in asthma and cigarette smoking in COPD. However, some patients with asthma who do not smoke also develop irreversible airway obstruction similar to COPD. Also, the treatments used to manage both diseases resemble each other.

This review compares and contrasts asthma and COPD with regard to the following aspects: the risks for development, the functional and structural abnormalities, the inflammatory activity and the site of the inflammation, animal models of both diseases, and current and potential new treatments for the diseases. The review was developed as a result of a workshop held in Whistler, Canada, involving clinical and basic scientists actively working in these two areas. It is hoped that by comparing the two diseases, new insights can be attained about each.

	RISK FACTORS FOR THE DEVELOPMENT OF ASTHMA

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

Asthma usually commences in childhood, but with the exception of occupational sensitizers, risk factors for adult and childhood onset asthma are broadly similar. The strongest determinant of childhood asthma is a parental (especially maternal) history of asthma and atopy (3), but whether asthma reflects polygenic inheritance or a genetic heterogeneity is unclear (4). A child whose identical twin has asthma has a sevenfold higher relative risk for asthma than does a child whose fraternal twin has asthma, despite the shared environment (5). Children with a maternal family history of asthma were more likely to have persistent rather than transient childhood wheezing, when followed for 6 yr (6). Among young adults 20-44 yr old, parental asthma is associated more strongly with onset of asthma before age 15 yr (odds ratio [OR] 5.16) than with onset after age 15 yr (OR 3.95), but for all ages parental asthma is associated with current asthma symptoms (OR 4.53) and airway responsiveness (OR 3.13) (7). Factors operating in utero and during the neonatal period also influence the likelihood of development of asthma, including younger maternal age (8), less weight gain during pregnancy (9) and (in some studies) low birth weight. Most childhood studies have found more asthma among males, probably related to an unexplained greater prevalence of atopy in boys (10).

Early childhood infections are associated with both transient (6) and more persistent (6, 11, 12) childhood wheezing. Children with lower respiratory infections without wheezing have lower levels of IgE than before the episode, and higher levels of interferon  (IFN-), suggesting that lack of IFN- may predispose to IgE production (11). Among adults, lower respiratory tract infection before the age of 5 yr is associated with onset of asthma before age 15 yr (OR 9.05), but is nonsignificantly associated with onset after age 15 yr (OR 2.16) (7). Work suggests that under some circumstances, respiratory tract infections in early childhood may confer protection against allergic sensitization, involving stimulation of helper T cell type 1 (Th1) immune responses (13).

Ethnic factors have variable influence, there being a 50% higher incidence of asthma among African-Americans (14, 15) and a higher likelihood of persistent childhood wheezing among Hispanic children (6), but in the United Kingdom no substantive difference in prevalence was found among three ethnic groups (16). Environmental rather than genetic differences may at least in part explain these findings.

Atopy is the second most obvious childhood risk factor for asthma and clearly is related to genetic risk factors and family history. Atopic children, as demonstrated by positive allergy skin tests (17), or serum IgE levels (6, 20, 21), had a three- to fivefold higher risk for developing asthma (10) and for persistence of early childhood asthma (6). These effects are also seen in adult life, especially in those with an age of onset less than 15 yr (7). Serum IgE is related to airway hyperresponsiveness even in asymptomatic subjects (21). Among children at risk of asthma by reason of a positive family history, all but one of those who did develop asthma had been exposed to substantial levels of house dust mite allergen (Der p 1 > 10 µg/g of dust) in early childhood (22) (Figure 1). Children exposed to furry pets have twice the risk of developing asthma (23) (Figure 1). Among 1,314 newborn children, those who developed allergy to house dust mite and cat by age 3 yr had a three- to fourfold higher exposure to these allergens in the home (22). Moving from a rural to an urban environment, shown in several studies to be a risk factor for asthma (24, 25), may increase allergen exposure and hence the likelihood of developing asthma.

View larger version (32K): [in this window] [in a new window]   Figure 1.   Allergen-specific relative risk, adjusted for the effects of sensitivity to 10 other allergens, for the development of asthma symptoms, airway hyperresponsiveness, and concurrent symptoms and hyperresponsiveness, at any age up to 13 yr, compared with nonsensitivity to that allergen in a longitudinal cohort of New Zealand children. *p < 0.05. [Adapted from Sears, M. R. 1989. Clin. Exp. Allergy 19:419-424.]

Exposure to environmental tobacco smoke during childhood has a dose-related effect on the risk of developing asthma (6, 9, 26, 27). Children of mothers smoking four cigarettes daily had a 14% increased prevalence of asthma, but children of mothers smoking 15 or more cigarettes daily had a 49% increased prevalence (28). Among Boston children 5-9 yr of age, only 1.9% of those with no environmental tobacco smoke exposure developed wheezing, compared with 6.8% of those with one parent smoking in the home, and 11.8% of those with two parents smoking in the home (29). Among children of atopic parents, 62% of those with environmental tobacco smoke exposure developed wheezing by age 5 yr, compared with 37% of those not exposed (30). In a longitudinal study of a New Zealand birth cohort, lung function of wheezing children not exposed to environmental tobacco smoke trended back toward normal values by age 15 yr, whereas wheezy children exposed to environmental tobacco smoke showed progressively worsening airway obstruction through adolescence (31). The risk of development of asthma associated with environmental tobacco smoke exposure is most obvious in those less than 2 yr of age (32), but continuing exposure is associated with a poorer long-term outcome (31). Moreover, the odds ratio on smoke exposure for the development of asthma increases further when associated with home dampness (OR 1.3) or cat or dog exposure (OR 8.0).

An effect of family size, number of older and younger siblings, and birth order on development of asthma and atopy has been reported in some studies of childhood asthma (7) but not others, and may depend on interaction with other factors including risk of infections, allergen exposure, and socioeconomic and income status.

Airway hyperresponsiveness to nonspecific agonists often precedes the development of symptomatic asthma (33). This has been observed not only in children or young adults, but also in middle-aged men, not selected to have an allergic history. More severe airway hyperresponsiveness is associated with the development of more severe symptoms and a steeper fall in FEV₁ (34). Furthermore, those individuals with less severe airway hyperresponsiveness are more likely to lose symptoms and become asymptomatic later in life (35). Although the presence of atopy is a risk factor for the development of airway hyperresponsiveness, it does not fully explain the presence of airway hyperresponsiveness.

Dietary factors may be important in both children and adults. Asian children in the United Kingdom, eating a fully Asian diet, had a lower risk of asthma (OR 0.31) compared with those eating a mixed Asian and English diet (OR 0.99) (36). Australian children eating oily fish regularly had a fourfold reduction in risk of developing asthma (OR 0.26) (37). Among adults, the role of sodium and magnesium intake, consumption of antioxidants, and other dietary factors remains uncertain despite several studies, although a role particularly for dietary antioxidants seems possible (38, 39).

Work-related asthma may be directly related to sensitizing agents in the workplace (40), indirect exposure to industrial or occupational airborne contaminants such as occurred in the Barcelona soya bean epidemic (41), or may reflect nonspecific factors such as exercise-induced bronchoconstriction because of work-related activities. Approximately 15% of adult onset asthma in Japan has been attributed to occupational sensitization (42). The range of sensitizing substances is ever increasing, with particular risks associated mainly with low molecular weight substances.

Smoking by adults is generally identified as a risk factor for adult onset asthma, especially increasing the likelihood of wheeze with cough and sputum (43). The number of anecdotal reports of asthma beginning after smoking cessation in adult life is now sufficiently frequent to suggest there may be other dynamics in this relationship. Immunization of infants has also been suggested to increase risk of developing asthma and allergy (44). There are likely other risk factors remaining to be identified, or suspicions proved. Nevertheless, among some children, and a larger number of adults, no risk factors can be clearly identified.

	RISK FACTORS FOR THE DEVELOPMENT OF COPD

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

To consider appropriately the factors influencing the development of fixed airway obstruction in COPD, there must be a clear model of growth and decline in lung function to evaluate suspected risk factors.

In theory, there are three separate mechanisms that can reduce an individual's FEV₁ and increase the risk of development of COPD (Figure 2). If normal growth and decline are represented by curve a in Figure 2, an individual can simply have reduced childhood growth and lung function leading to a reduced level of FEV₁ in adolescents, with a normal plateau phase in early adult life and a normal decline (Figure 2, curve b). An alternative possibility is normal growth of FEV₁ but a premature decline with attenuation of a normal plateau phase, the stable phase of lung function that usually occurs between 15 and 35 yr of age (Figure 2, curve c). A third alternative is normal growth, a normal plateau phase but accelerated decline in FEV₁ in adult life after the age of 35 yr (Figure 2, curve d). Obviously, an individual can combine elements of curves b, c, and d to achieve a more complex history of growth and decline in FEV₁. The importance of the theoretical model in the various growth and decline curves is that its components are amplified by knowing that pulmonary function has an exceedingly high tracking correlation. In clinical terms, this means that an individual's present pulmonary function predicts the level of his or her pulmonary function far into the future with a great deal of certainty. Tracking of pulmonary function has been demonstrated in children as young as 5 yr of age, when FEV₁ can first be measured. The estimated tracking correlation for pulmonary function is 0.8 to 0.9. The magnitude of this correlation is not constant but decreases with an increase in time interval between the measurements (45). The high tracking correlation does not imply that environmental events are not important in producing disease. It does imply that reduced levels of lung function identify the physiologically susceptible individual. Viewed in this manner, one can examine the factors that influence the three important phases of the pulmonary growth and decline curve: attainment of maximal lung function (growth phase), maintenance of maximal lung function (plateau phase), and the third, or decline, phase.

View larger version (14K): [in this window] [in a new window]   Figure 2.   Hypothetical tracking curves of FEV1 for an individual through life. The normal pattern of growth and decline with age is shown by curve a. A significantly reduced FEV1 can develop by a normal rate of decline after a reduced growth phase (curve b), by an early initiation of decline after normal growth (curve c), or by an accelerated decline after normal growth (curve d ). [Reproduced with permission from Rijcken, B. 1991. Bronchial Responsiveness and COPD Risk: An Epidemiological Study. Doctoral dissertation. University of Groningen, Groningen, The Netherlands.]

The three most important factors influencing growth in pulmonary function are asthma, passive and active cigarette smoking, and sex. Symptomatic childhood asthma is associated with a 0-15% reduction in percent predicted FEV₁ by age 15 yr (46, 47). The severity of the reduction is directly proportional to the severity of asthma symptoms. If symptoms are present on a daily basis, the reduced growth is on the order of 10-15%. In mild asthma, there is some evidence to suggest the effects are greater in females than in males (48). The relative paucity of information does not allow separation of the independent effects of allergy and airway responsiveness on the estimates of asthma effect on lung growth. Passive cigarette smoking is associated with a 1 to 5% reduction in FEV₁ by age 14 yr in children (49). It appears, at the present time, that the bulk of this effect is due to in utero exposure. Active cigarette smoking between the ages of 5 and 20 yr is associated with a 5- 10% decrease in maximal attained FEV₁. It has a relatively small effect on maximally attained lung function because the bulk of cigarette smokers in the United States begin smoking between the ages of 15 and 25 yr and in females FEV₁ is maximized by age 15 yr and in males by age 25 yr. Green and co-workers (50) originally coined the term "dysanapsis" to signify nonisotropic growth of lung airways and lung parenchyma. The female lung is smaller than the male lung, controlling for height, and female airway size is also smaller than male airway size at maximal growth. However, relative to lung size, female airways are larger than male airways. Thus, dysanapsis is more common in males than in females. Sex differences are present at birth but become maximal at some point during the plateau phase, during which continued male growth and vital capacity maximize the sex difference in lung size versus airway size, probably at about age 25 yr. The implications of these anatomic sex effects on long-term risk of COPD are unclear.

The single most important predictor of lung function during the plateau phase is the presence of chronic respiratory symptoms or asthma. The presence of wheezing at age 14 yr and its persistence at 28 yr of age are associated with decrements in pulmonary function of approximately 20% of predicted by the end of the plateau phase in asthmatic subjects (51). Similar results have been obtained when examining airway responsiveness alone (52). Third, as an important influence on FEV₁, is the effect of active smoking during the plateau phase. Beginning cigarette smoking at age 15 yr is associated with a premature decline in lung function, such that by the end of the plateau phase, the loss in lung function may be as much as 5-10% (53).

While it is clear that cigarette smoking is the most important predictor of decline in lung function and the development of COPD after age of 35 yr, it is important to recognize that the attributable risk estimates of active cigarette smoking overestimate its effect. Only 10-15% of active cigarette smokers actually develop COPD (45). In addition, a study by Burrows and colleagues (54) demonstrated that regression models predicting FEV₁, including age, duration of smoking, current number of cigarettes smoked per day, average number of cigarettes smoked per day, and total pack-years, produced a correlation coefficient of 0.38, thus explaining only 15% of the variability in FEV₁. Independent of cigarette smoking, airway responsiveness is an important predictor of accelerated decline in lung function and, hence, COPD risk (55, 56). Thus, these data would tend to support as well the importance of asthma and airway responsiveness as independent predictors of the development of COPD.

At the present time, it appears clear that asthma and airway responsiveness are important predictors of COPD throughout the life cycle. What remains to be established is the magnitude and precise contribution of these factors relative to cigarette smoking in each stage of the life cycle.

	ANIMAL MODELS OF ASTHMA

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

In asthma, the use of animal models has greatly increased the understanding of the mechanisms involved in the pathogenesis and pathophysiology of the disease, and has been critical for drug discovery, characterization, and development. Requirements for an appropriate animal model of disease include a close resemblance to the pathology of the disease in humans, objective measurement of the physiologic parameters (preferably in the absence of anesthetic agents), and reliability and reproducibility. Moreover, the responses of the animal model should reflect the activity of pharmacological agents known to effectively modify the disease in humans. Features of the ideal animal model of asthma include the presence of paroxysmal bronchoconstriction, early and late allergen-induced airway responses, airway inflammation including bronchial eosinophilia, variable but persistent airway hyperresponsiveness, and chronic lung remodeling with deterioration in lung function. Unfortunately, the investigation of the pathophysiological mechanisms of chronic asthma is hampered by the lack of such an ideal animal model. Animals do not spontaneously develop asthma, although some symptoms comparable to human asthma have been seen in horses (57) and in cats (58). A variety of animals have been used as models for asthma, including primates, sheep, mice, rabbits, guinea pigs, rats, and dogs (Table 1).

Guinea pigs have been probably the most widely used animal model for asthma, likely because of an easily demonstrable early bronchoconstriction after ovalbumin sensitization and inhalation. Research based on this animal model has helped elucidate the role of different mediators of asthma and inflammation in acute bronchoconstriction and induced airway hyperresponsiveness, as well as the relationship between airway hyperresponsiveness and airway inflammation. Muscarinic M₂ autoreceptor (59) dysfunction or nitric oxide deficiency (60) has been suggested to contribute to airway hyperresponsiveness. Moreover, studies with antibodies against interleukin 5 (IL-5) (61) or the adhesion molecule VLA-4 (very late activation antigen 4) (62) point to an important role for the eosinophil in this respect. Also, chemotactic factors for eosinophils, such as leukotriene B₄ (LTB₄), IL-8, IL-5, and platelet-activating factor (PAF) have been investigated in this species (63). The disadvantage of guinea pigs is the considerable variation in frequency, severity, and duration of late asthmatic reactions, and the lack of persistent airway hyperresponsiveness.

In the last 10 years, murine models of asthma have been developed, in particular since new methods for lung function measurement in vivo without anesthesia became available. Advantages of the mouse model are the well-characterized immune system and the availability of many different types of immunological reagents and genetically manipulated species, such as genetic knockouts, transgenic mice, severe combined immunodeficient (SCID) mice, and others. These tools can contribute to analyze the role (presence/activation status) of immune cells and cytokines in the pathophysiology and pathogenesis of asthma.

The different ways in which mice may be sensitized to allergens have demonstrated that under different conditions airway hyperresponsiveness can be associated with airway eosinophilia, but can also be clearly dissociated from eosinophilia (64). Concepts about the role of Th1 and Th2 cells are derived from murine research. Also, new ideas with respect to a possible role for INF- and IL-4 in the induction of airway hyperresponsiveness are derived from murine models (64). Also, the mouse is an appropriate model for studying the role of costimulatory molecules in the activation of T lymphocytes and the possibilities for in vivo pharmacological modulation (65).

Mouse models of occupational asthma have recently been developed by using toluene diisocyanate, dinitrofluorobenzene, and picrylchloride (66). These new models have led to new ideas on induction of airway hyperresponsiveness, particularly involving the interaction between a certain type of lymphocyte factor derived from Thy-1⁺B220⁺ lymphocytes and mast cells which seems to be of crucial importance, whereas inflammation in the lung may not be essential for the induction for airway hyperresponsiveness in these models (70, 71). Airway hyperresponsiveness is not associated with increases in IgE. These models may also represent models of nonatopic immunological asthma in mice that would be related to forms of intrinsic asthma in humans. However, this model also lacks persistent airway hyperresponsiveness.

Some animal species have a naturally occurring IgE-mediated respiratory hypersensitivity to Ascaris suum (monkey, sheep, and dogs) (72). In both sheep and primates, a well-defined late asthmatic reaction is found after inhalation of A. suum. Research in primates has shed light on the role of adhesion molecules in airway hyperresponsiveness. Wegner and co-workers (75) showed that anti-ICAM1 (intercellular adhesion molecule 1) significantly inhibits eosinophil influx and activation as well as airway hyperresponsiveness in this species. Of relevance in allergic sheep is the great similarity with humans with respect to the role of different mediators in induction of airway constriction and airway hyperresponsiveness as well of the effects of different pharmacological antagonists, such as antihistaminics, steroids, and leukotriene antagonists, in preventing early and late asthmatic reactions and airway hyperresponsiveness (73).

The canine model has an advantage in that dogs can be spontaneously atopic (basenji-greyhound) or can be sensitized to ragweed or A. suum (74). Interestingly, dogs do not have a nonadrenergic, noncholinergic (NANC) bronchodilator system. In addition, in allergic dogs airway hyperresponsiveness is associated with an impaired response to -adrenergic agonists, associated with decreased cyclic AMP production probably caused by impaired coupling between  receptor and G protein Gs- (74).

Rabbits neonatally immunized to antigen have been shown to mimic key aspects of the allergic asthmatic response in humans, including allergen-induced early and late bronchoconstriction, airway hyperresponsiveness, and airway inflammation (76). Both rabbits and humans are similar, in that they both receive a relatively sparse innervation by sensory nerves and are modestly responsive to contractile actions for capsaicin in vitro. Both late reactions and airway hyperresponsiveness appear to be dependent on the presence of granulocytes in the circulation at the time of antigen exposure.

Rat with high IgE responses (Brown Norway) and low IgE responses (Lewis and Sprague-Dawley) are available. Studies in the Brown Norway rat have shown that antibodies to ICAM-1 prevent airway hyperresponsiveness without suppression of the influx of eosinophils or lymphocytes in the bronchoalveolar lavage (BAL) (77, 78). The pulmonary response to antigen in the rat differs from that in humans in that it is primarily mediated by serotonin. Histamine plays only a small role in this acute response and does not contract rat airway smooth muscle. Interestingly, depletion of CD8⁺ lymphocytes with monoclonal antibody led to a significant increase in the late asthmatic response (79). Late asthmatic response and airway hyperresponsiveness are inhibited by leukotriene antagonists (80).

Comparison of these animal models demonstrates that each can contribute to the knowledge of the pathogenesis and pathophysiology of asthma in humans. In particular, the murine model may contribute in future to the understanding of the immunopathology of allergic diseases in the lung. Research in animal models should lead to innovative concepts that can be tested in humans.

	ANIMAL MODELS OF COPD

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

When evaluating animal models of COPD, it is important to realize that COPD is not one disease and is generally thought to be an umbrella term encompassing chronic (obstructive) bronchitis, emphysema, and small airway disease, although the latter is no longer mentioned in the most recent American Thoracic Society (ATS) definition (81). The diagnosis of chronic (obstructive) bronchitis in humans is based on objective measurements of airway obstruction and on symptoms of chronic cough and/or sputum production. This definition based on symptoms is less than useful when defining the characteristics of an animal model of chronic bronchitis. By contrast, emphysema is an anatomically defined disease with permanent destructive enlargement of air spaces distal to the terminal bronchioles and without obvious fibrosis, and therefore much more amenable to animal models. Because of this, animal models of emphysema and of chronic (obstructive) bronchitis are discussed separately.

Spontaneous emphysema in animals has been reported in several inbred mice strains, including pallid, beige, and tight skin mouse strains (82). In these mice the extent of emphysema differs, as does the age at which it develops. Emphysema has also been reported occasionally in horses that develop so-called COPD, but it has many of the clinical characteristics of human asthma accompanied by marked mucous production (83).

Proteolytic enzyme models are the oldest experimental animal models of emphysema (84). It was demonstrated in 1964 that papain was able to elicit morphological changes in rats similar to human panacinar emphysema (85). Similar effects were produced by several other proteolytic enzymes, most notably elastase. Within minutes of intratracheal administration of a single dose of elastase, hemorrhage occurs with edema and influx of neutrophils and macrophages, and within hours, air space enlargement occurs. Physiologic studies revealed increased lung compliance and total lung capacity and decreased expiratory flows and diffusion capacity. These models have also been established in dogs, guinea pigs, hamsters, rats, and mice. The models have been used extensively to study diaphragm function, structure-function relationships of the airways, and the role of elastin and collagen in emphysema development. Disadvantages of the proteolytic enzyme models are that they are hyperacute, which is in marked contrast to the human development of emphysema; they produce hemorrhages, and they are focused on the protease/anti-protease imbalance theory of emphysema development.

Irritant gas models have also been developed. Inhaled nitrogen dioxide (NO₂) has been shown to elicit centriacinar air space enlargement accompanied by neutrophil infiltration (86). Short-term exposure to ozone (O₃) elicits similar effects, but longer exposure causes more lung fibrosis. In experimentally produced diesel exhaust, soot particles are important components in causing emphysema, together with NO₂ and SO₂ (87).

Because cigarette smoke is the major risk factor for the development of emphysema in humans, many small animal models based on the use of cigarette smoke or its derivatives have been tried. One well-documented model is a guinea pig model (88), which has documented increases in mean linear intercept over a 12-mo period with accompanying physiologic changes. Although some groups have claimed success in inducing emphysema in rats with cigarette smoke, others have been unable to document these changes (89). In each of these research protocols, there are large differences in smoke delivery (whole body versus nose only), in main stream versus side stream smoke, smoke duration, number of days per week of exposure, animal strains, and aerobiological environment. It is difficult to define the common denominator of successful models. A common drawback of all small animal models of cigarette-induced emphysema is the fact that the animals are nose breathers, which likely has a major impact on the smoke composition and concentration reaching the lung. Occasionally, breathing through a tracheostomy has been tried in shorter protocols. Cigarette smoke-induced emphysema has been reported in C57BL/6J mice (90). Interestingly, this could not be produced in macrophage metalloelastase-deficient mice. Less frequently used models of emphysema include hyperoxia, starvation, -aminopropionitril feeding, and cadmium salts. In addition, several combinations of elastase and cigarette smoke have been described.

Several research groups have published reports of cigarette smoke-induced changes of the tracheal and larger central airways. These changes usually occur within 2-4 wk and consist of increases in epithelial thickness, secretory cell hyperplasia, and increases in epithelial mucous cells in rats (91). In rats and mice, increases in BAL total cell count and percentage neutrophils have been documented after smoke exposure, accompanied by increases in superoxide production (92). Documentation of changes in small airways after cigarette smoke inhala tion is much more difficult. In guinea pigs, even after 12 mo of exposure, the small airways had unaltered structure and number of alveolar attachments (93). Nevertheless, there was an increase in secretory cell number and a decrease in expiratory flows. The latter could be due to changes in the parenchyma and associated collapse.

Other research groups have documented the short- and long-term effects of NO₂ and ozone inhalation (94). There are rapid increases in airway resistance and airway responsiveness, with influx of neutrophils first and mononuclear cells later. After chronic exposure, goblet cell hyperplasia has been demonstrated accompanied by increases in mucus production, and an increase in smooth muscle hypertrophy. More recently, SO₂-induced changes have been described, consisting of a marked neutrophil infiltration of the trachea, an increase in mucous glycoproteins, and an increase in resistance (95). Finally, elastase inhalation also elicits increases in BAL total cell counts, and is accompanied by epithelial damage, mucous plugging, and neutrophil and macrophage infiltration in the bronchial mucosa (96).

In summary, the proteolytic enzyme models of emphysema have, in the past, been the major research focus. These models, however, have severe limitations. Newer models focusing more on the oxidant/antioxidant balance have been slow in development. Models of cigarette smoke-induced emphysema have not been easy to establish, but mouse models seem to be promising. The possibilities of knockout and transgenic mice should permit advances in unraveling the pathogenesis of emphysema. Descriptions of small airway lesions are scarce. The role of viral or bacterial infections in induction of COPD in animal models has not received much attention thus far.

	AIRWAY INFLAMMATION IN ASTHMA

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

Asthma is traditionally considered a large airway disease. Research based on bronchoscopy with biopsies to explore the inflammatory response in asthma has usually relied on information about inflammation in the large airways. This has shown that an increase in eosinophils, lymphocytes (predominantly of the CD4 type), and mast cells occurs in all patients with asthma. Neutrophil numbers can increase during an exacerbation. The activation of these cells in the airway wall appears to be of clinical relevance. Hamid and coworkers (97) have shown that a larger increase in eosinophils is present in the inner airway wall when compared with the outer airway wall, but the outer wall eosinophils were more activated. This suggests that the deleterious effects of eosinophils may also occur in the peribronchial area.

The majority of information about small airway inflammation has come from autopsy studies. Carroll and colleagues (98, 99) have evaluated small airways to determine the distribution of inflammation in smooth muscle and mucous glands. They used a stratified sampling method to examine the distribution of eosinophils and lymphocytes in the large and peripheral airways in fatal and nonfatal asthma and compared the results with the distribution in the absence of asthma (98), showing that in the large airways the number of lymphocytes was significantly increased in patients with asthma as compared with control patients. However, the number of lymphocytes was greatest in the membranous bronchioles, the smallest airways sampled, in patients with nonfatal asthma as compared with control patients. Eosinophils were present in all of the airways sampled, regardless of whether subjects were asthmatic or nonasthmatic, but the numbers were greatest in the fatal asthma group. Their findings suggest that asthma is best characterized by increased numbers of lymphocytes, which are evenly distributed throughout the bronchial tree. The eosinophil counts were more variable, with a 23-fold variation in the range of mean eosinophil counts in cases of fatal asthma, 9-fold variation in nonfatal cases, and an 11-fold variation in control cases.

Two studies have demonstrated significant parenchymal inflammation in chronic, stable, and severe asthma. Kraft and colleagues evaluated subjects with nocturnal asthma and nonnocturnal asthma (100). They performed endobronchial and transbronchial biopsies at 4:00 P.M. and 4:00 A.M. The proximal endobronchial biopsy inflammation did not change significantly from 4:00 P.M. to 4:00 A.M., and this observation has been shown by other investigators (101). The distal alveolar tissue inflammation, particularly eosinophils, was increased at 4:00 A.M. and this correlated with the overnight decrement in FEV₁. Wenzel and colleagues evaluated patients with moderate asthma and severe asthma, and control patients without asthma (102). They noted an increased number of neutrophils in the proximal endobronchial and distal transbronchial tissue in patients with severe asthma, as compared with patients with moderate asthma and control patients. Further, Hamid and colleagues (97) have shown eosinophils to be present in both peripheral and central airways in patients with moderate asthma (Figure 3), whereas CD4 cells were predominantly found in the central airways. These findings suggest that parenchymal inflammation is present in asthma, but may be characterized by different cell types depending on the severity of the disease.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Immunocytochemical cell markers in airways less than 2 mm and greater than 2 mm in diameter from patients with asthma. T cells (CD3), total eosinophils (MBP), activated eosinophils (EG2), and mast cells (tryptase) were examined. Eosinophils were significantly higher in airways less than 2 mm in diameter (*p < 0.05). [Reproduced with permission from Hamid, Q., et al. 1997. J. Allergy Clin. Immunol. 100:44-51.]

Other important aspects of the large and small airway structure have been examined in asthma. Carroll and colleagues evaluated the inner and outer airway wall diameter in fatal and nonfatal asthma (99). The inner wall area, defined as the area between the epithelium and smooth muscle, was significantly thicker in patients with fatal and nonfatal asthma as compared with nonasthmatic control patients. Interestingly, the difference in wall thickness between patients with fatal and nonfatal asthma was significant only in the larger airways (> 18 mm, or lobar bronchi), suggesting that small airway changes were present but similar in both asthmatic groups. The outer airway wall thickness, defined as the area between smooth muscle and adventitia, was similar in both fatal and nonfatal asthma groups in terms of the smallest airways evaluated (2-4 mm, small and large membranous bronchioles). The fatal asthma groups exhibited the most significant increases in thickness as compared with the nonfatal asthma group, in terms of the large airways only. When smooth muscle and mucous gland areas were evaluated, the two asthmatic groups showed similar smooth muscle thickness in the small airways; smooth muscle area was greater in the fatal asthma group, but only in the larger airways. Mucous gland area was evaluated only in the areas greater than 4 mm, and was significantly greater in the fatal asthma group as compared with the nonfatal asthma group and controls. Finally, epithelium desquamation was not significantly different between the two asthmatic groups and the nonasthmatic group.

The debate continues as to how inflammation in asthma is related to lung function. Although some studies have illustrated such a relationship (100), others have not (103). This issue brings to light whether small airway inflammation or structural changes exert any effect on the lung function changes appreciated in asthma. Woolcock and colleagues (104) have shown a frequency dependence of dynamic compliance in patients with asthma despite normal resting pressure-volume relationships consistent with nonuniform obstruction of the peripheral airways. In addition, peripheral airway resistance measured directly via bronchoscopy is increased in patients with mild asthma as compared with normal subjects, which correlated with airway hyperresponsiveness (105). Others have confirmed that peripheral resistance is increased in patients with asthma, and further increases in response to cool, dry air (106). Finally, the peripheral airway resistance measured via the bronchoscope increases at night, suggesting that the changes occurring in the lung parenchyma may be affecting nighttime lung function (107).

In conclusion, inflammation and structural changes related to smooth muscle, mucous glands, and basement membrane thickening occur in asthma in both the large and small airways. It is still not clear if the large airway changes appreciated by biopsy mirror the changes in the small airways. A challenge remains in how best to measure the small airway changes occurring in asthma, which at present cannot be done noninvasively.

	AIRWAY INFLAMMATION IN COPD

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

Pathological changes in COPD are found in the central airways, the small airways (bronchioles), and in the lung parenchyma. When the disease progresses, changes are found in the pulmonary circulation, the heart and the respiratory muscles.

The airway narrowing in the large and small airways is caused by changes in the different constituents normally present in the airway wall, and by the consequences of the inflammatory events that are present in this area. Epithelial changes include squamous metaplasia, and, related to this, qualitative and quantitative abnormalities of ciliated cells, and atrophy. In early stages epithelial hyperplasia is observed also with hyperplasia of mucous glands. Hypersecretion of mucus contributes to the airway obstruction, in particular in combination with defective mucociliary clearance. Epithelial changes are considered to be caused by a combination of direct effects of constituents of cigarette smoke and indirect damage caused by effects of inflammatory cells. Epithelial cells influence the inflammatory process by production of a great variety of mediators. This is also true for fibroblasts, which may be involved initially in fibrotic changes in and around the airway wall as a consequence of chronic inflammation, but can be expected to mediate effects on the inflammatory process directly.

Information is now available, from several biopsy studies in smokers with or without COPD, about the inflammatory cells present in the subepithelial area of the mucosa in the central airways (108). The findings are different from those in asthma. The infiltrate shows a predominance of lymphocytes, in particular CD8⁺ cells, in contrast to CD4⁺ T cells in asthma. Although eosinophils were found in exacerbations of COPD, they are not activated and do not degranulate (114, 115). These eosinophils can be considered relatively inert "bystanders" recruited by upregulation of vascular adhesion molecules, caused by mediators from local extensive inflammation. Neutrophils are scarce in the subepithelial area (116, 117). In contrast, neutrophils were found to be increased in the epithelium and in bronchial glands (118, 119), corresponding with findings of prominent numbers of neutrophils in the airway lumen as demonstrated in lavage fluid and induced sputum.

The increased airway resistance in COPD has long since been shown to be due to disease of the small bronchi and bronchioles (120, 121). This pathology has been described as small or peripheral airway disease as well as chronic obstructive bronchiolitis. Although some characteristic pathological changes can be seen, these are not as obvious as observed in several other diseases involving more severe damage to peripheral airways, such as bronchiolitis obliterans, and following inhalation of toxic fumes, extrinsic allergic alveolitis, or infection. Small airways may show a variation in size due to differences in size of individual patients and differences in sample site within the lung.

The main histopathological changes in small airway disease involve an appearance and increase in number of goblet cells, an increase in the amount of mucus in the lumen, the presence of inflammation, an increase in muscle mass in the walls of the bronchioles, and, ultimately, fibrosis and obliteration, causing airway narrowing (122). Data suggest that airway narrowing is also due to loss of alveolar attachments to bronchioles.

The composition of the inflammatory infiltrate in the peripheral airways has not yet been well studied. B cells were described to reside in the adventitia (126) and one study showed infiltration of CD8⁺ T cells in the airway wall (127). As already described, most information about chronic bronchitis has been gained from studies of bronchial biopsies taken from central airway walls. Whether the inflammatory events in central airways during exacerbations of chronic bronchitis are also reflected in the periphery is at present uncertain.

Emphysema is the parenchymal component in the manifestation of COPD. In contrast to chronic bronchitis, which is defined functionally, emphysema is defined by an anatomical substrate as a permanent destructive enlargement of air spaces distal to the terminal bronchioles, without obvious fibrosis (128). The last characteristic is subject to discussion, as in many emphysematous lungs focal areas with fibrosis are observed, often owing to reactive changes associated with larger bullae.

Emphysema is observed in two major patterns of panacinar and centriacinar emphysema (120). In panacinar emphysema the entire acinus is destroyed. In this type of emphysema, adjacent acinar units are involved to a similar degree. In the second major type of emphysema (centriacinar), the abnormal air spaces are initially found associated with respiratory bronchioles, in more severe cases progressing to involvement of the whole acinar unit. In contrast to panacinar emphysema, often quite small lesions can be identified. Both major types of emphysema may coexist. Bullae are subpleural areas of emphysema that are locally overdistended, and are more than 1 cm in diameter.

A diagnosis of emphysema is made by macroscopic observation of the gross lung specimen (120). The mean linear intercept (MLI) technique or the destructive index (DI) are at present used mostly to obtain an estimate of the diameter of the air spaces in performing microscopic assessment of severity (129, 130). However, these techniques are time consuming and require fixation under standardized pressure, making them difficult to use for estimation of severity of emphysema in partial lung resections or lobectomies for bronchial carcinoma (in which also diagnostic procedures regarding the carcinoma must be performed).

The finding of isolated or "free lying" segments of alveolar septal tissue or isolated cross sections of pulmonary vessels is considered characteristic histologic evidence of mild emphysema (131). These free lying segments are viable and often contain blood cells in the capillary lumen. However, using this diagnostic method, it is impossible to estimate microscopically the extent or severity of the disease from histologic sections. When it is not possible to perform standardized fixation and slicing of lung specimen for determination of mean linear intercept as a measure of severity of emphysema, the best alternative may be to perform preoperative high-resolution computerized tomography (CT) scans.

	FUNCTIONAL ABNORMALITIES IN RELATION TO AIRWAY INFLAMMATION IN ASTHMA

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

Airway wall inflammation is thought to play a central role in the development and progression of asthma. Even though the degree and variability of airway obstruction in asthma appear to be associated with many markers of airway inflammation, such as those in induced sputum (132) or in exhaled air (133), the physiological factors determining the severity of airway narrowing are far from being clarified. The currently available knowledge on associations between inflammation and airway function is based on both acute inflammatory events, such as airway smooth muscle contraction and/or (sub)mucosal swelling, secondary to activation of inflammatory and resident cells within the airway wall as induced by allergen challenge (134), as well as chronic airway inflammation, associated with subepithelial collagen deposition, smooth muscle growth, and increased mucosal vascularity (135).

It has been postulated that peripheral airway inflammation will have the greatest physiological impact in asthma, particularly when occurring in the peribronchial area (136). Indeed, inflammation is prominent within the airway adventitia, as appears from the few studies using transbronchial biopsies (100) or surgically resected lungs (97) in patients with asthma.

When considering airway inflammation as the fundamental disease characteristic of asthma, it is obvious that monitoring of inflammation might benefit the clinical management of the disease, particularly because antiinflammatory therapy is presently the first choice. Current attempts to monitor the degree of airway inflammation include determining eosinophilic presence and activity in induced sputum (132), and biochemical markers in urine (137) and in exhaled air (133). However, the presently available cellular or molecular markers predominantly reflect acute inflammatory events only, which change rapidly after exposure to inflammatory stimuli or after antiinflammatory treatment (138).

It can be postulated that the integrative physiological markers of inflammation are better suited for monitoring purposes (139). Interestingly, physiological markers of inflammation are more closely related (140) to the number of eosinophils in the bronchial (sub)mucosa than to the number of eosinophils in induced sputum (141). Moreover, physiological markers are also associated with features of chronic airway inflammation (142) for which few validated cellular or molecular markers exist.

The association between airway inflammation and FEV₁ and peak expiratory flow (PEF) variability is at best weak (140). The postbronchodilator FEV₁ is probably best associated. The significance of small airway inflammation correlating with FEV₁, which is believed to be primarily a large airway measurement, merits discussion. Macklem and Meade (143) have reported that maximal expiratory flows such as the FEV₁ are dependent on three factors: the elastic recoil of the lung, the resistance in the small airways, and the cross-sectional area of the large airways When the FEV₁ decreases it may be the result of changes in one or all of these entities. Thus, one may hypothesize that alveolar tissue inflammation not only increases the small airway resistance, but decreases elastic recoil, the latter potentially owing to an uncoupling of the airways and parenchyma caused by the alveolar tissue inflammation. These changes may result in a fall in the FEV₁ regardless of large airway changes. However, better physiological tests might include the following:

1. Impaired deep-breath bronchodilation: during bronchoconstriction normal subjects respond to a deep breath by transient bronchodilation. Presumably, this hysteresis is due to changes in plasticity of the cellular organization of the contractile filaments of smooth muscle (144). There is overwhelming evidence now that this physiological protection is impaired in asthma, which can be measured rather easily (145). This impairment is likely to be due to mechanical discoupling between airways and parenchyma, secondary to peribronchial inflammation (136), which might induce a force-maintenance "latch" state of smooth muscle (146). Indeed, avoidance of deep breaths in healthy subjects induces hyperresponsiveness of the airways to bronchoconstrictors in vivo, as observed in asthma (147, 148). Because the deep breath-induced bronchodilation decreases after proinflammatory stimuli (149) and improves only after inhalation of steroids (150), this particular measure might be well suited for physiological monitoring of inflammation in asthma management.
2. Excessive airway narrowing in response to bronchoconstrictor stimuli is another major consequence of airway inflammation. This is measured by the presence, height, absence, or resolution of the maximal response plateau in patients with asthma (151). However, for monitoring purposes, this measure might be less suited because of the implicit requirement to reach severe degrees of bronchoconstriction in the laboratory. That is why alternative approaches, such as the recording of FVC at the PC₂₀ (provocative concentration of an agonist causing a 20% fall in FEV₁) level, have been proposed (152).
3. Airway hyperresponsiveness, in terms of the PC₂₀, is the most commonly used measure of the response to bronchoconstrictor challenge tests, and is currently the best validated physiological marker of acute (153) and chronic (137) features of airway inflammation in asthma. It responds rather slowly to antiinflammatory treatment with inhaled corticosteroids, particularly when measured with methacholine (154). This may be an advantage, because it suggests that the methacholine PC₂₀ reflects the slowly improving anatomic remodeling during inhaled steroid therapy (155).

An important issue concerns whether measuring these physiological markers will provide information complementary to the standard information collected in the long-term management of asthma. This issue has been addressed in a randomized, parallel group, follow-up study over 2 yr, which examined whether inclusion of airway hyperresponsiveness (PC₂₀) among the current guides for asthma treatment (symptoms and lung function) improves the clinical and histological outcome of the disease (156). The treatment strategy that included PC₂₀ as a guide for adjustment of therapy resulted in a twofold reduction in cumulative incidence of exacerbations, when compared with the control strategy. Interestingly, this was accompanied by a decrease in subepithelial collagen thickness in the bronchial biopsies of the group with treatment aimed at improving hyperresponsiveness, but not in the control group (157). The reduction in airway hyperresponsiveness in this study was also correlated with a decrease in activated eosinophil counts in the bronchial lamina propria.

The treatment of asthma on the basis of physiological principles leads to a better clinical as well as pathological long-term outcome. It is questionable as to whether this can be accomplished by including cellular markers of inflammation in asthma management, because the integrative physiological measures rather than specific cells or biochemicals are likely to reflect the complex and chronic inflammatory changes within the airways in this disease.

	FUNCTIONAL ABNORMALITIES IN RELATION TO AIRWAY INFLAMMATION IN COPD

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

Less is known about structure-function relationships in COPD than in asthma. Ventilation-perfusion (A/) imbalance is the main determinant of low Pa _O2 in patients with COPD in both acute (158) and chronic (159) conditions; in contrast, the role of increased intrapulmonary shunt is marginal and that of alveolar end-capillary diffusion limitation for O₂, irrelevant (160). The influence of pulmonary emphysema and small airway abnormalities on A/ mismatching in both early mild and advanced COPD has been evaluated (161). This study demonstrated that a macroscopic index of emphysema severity was significantly related to the dispersion of pulmonary blood flow (a functional descriptor sensitive to alveolar units with both normal and low A/ ratios) and also to the variable that characterizes the dispersion of alveolar ventilation (a functional marker sensitive to units with normal and high A/ ratios). The more severe the degree of emphysema, the more abnormal the A/ imbalance, as reported in previous studies (162). The loss of alveolar attachments of bronchiolar walls observed in emphysema may result in both distortion and narrowing of the lumen of bronchioles, thereby reducing alveolar ventilation in the dependent alveolar units, and hence resulting in areas with low A/ ratios. This relationship becomes evident in the dispersion of blood flow distribution. Also, the relationship between emphysema and abnormalities in the ventilation distribution could be, at least in part, related to the loss of pulmonary capillary network in emphysematous spaces. This would lead in turn to the development of lung units with high A/ ratios, and hence to an increase in the dispersion of alveolar ventilation. Alternatively, bronchiolar lesions are also associated with A/ heterogeneity. The airway narrowing results in a nonhomogeneous distribution of inspired air, and hence to the development of areas with an increased dispersion of alveolar ventilation. Thus, the functional and structural narrowing of the conducting airways caused by bronchiolar impairment, due to acute or chronic airway changes such as bronchial wall edema and/or increased lumenal secretions more evident in more advanced stages of COPD (163), would be at the origin of the increased dispersion of alveolar ventilation.

Although A/ inequality is not a barrier to O ₂ exchange when 100% O₂ is breathed, 100% O₂ always worsens A/ mismatch (as assessed by a significant increase in the dispersion of blood flow), in patients with COPD irrespective of the severity of airway obstruction and the underlying clinical condition. This implies release or abolition of hypoxic pulmonary vasoconstriction. Patients with COPD who do not release hypoxic vasoconstriction while breathing 100% O ₂ have more intimal thickness in the small pulmonary arteries, particularly in those with a smaller caliber (< 500 µm), where it is supposed that hypoxic vasoconstriction takes place, as compared with patients who do respond and controls. Furthermore, the thickness of the intimal layer was correlated with the degree of bronchiolar inflammation, possibly suggesting that an underlying inflammatory process might be involved in the pulmonary arterial wall abnormalities (164). Accordingly, thickening in small pulmonary arteries may interfere with the adaptability of these vessels to changes in O₂ concentration and the maintenance of adequate A/ matching. This could explain the presence of arterial oxygenation abnormalities out of proportion to the degree of airway obstruction.

Endothelial dysfunction of pulmonary arteries has been demonstrated in patients with end-stage chronic obstructive lung disease submitted to lung transplantation (165). Furthermore, maximal relaxation of pulmonary artery rings was related directly to Pa _O2 and negatively to Pa_CO2 before transplantation, but not to the FEV₁, thus suggesting that chronic hypoxemia may impair endothelial cell metabolism and synthesis of endothelium-derived relaxing factors. Equally important, a similar degree of endothelial dysfunction along with thickened intimas in small pulmonary arteries has been shown in normoxemic patients with early COPD (166); interestingly, in smokers with normal lung function, pulmonary vascular reactivity was preserved but the intimal layer was thickened (166).

Because the severity of airway inflammation correlates with abnormalities of pulmonary arteries, it is conceivable that smoking-induced inflammation may have an important pathogenic role. Cigarette smoking rapidly induces proliferation of endothelial and smooth muscle cells and fibroblasts in small vessels (167). Likewise, smoking can modulate the continuous release of endogenous nitric oxide (NO) by endothelial cells, hence not only regulating the pulmonary vascular tone but also decreasing NO-induced inhibition of both platelet and neutrophil aggregation (168), reduction of endothelial cell adhesion, and suppression of superoxide anion production by activated neutrophils. This suggests that smoking could play a pathogenic role in the development of the pulmonary vascular alterations shown in COPD, even preceding the functional disturbances of the pulmonary vasculature.

	INFLAMMATORY MARKERS IN ASTHMA

TOP INTRODUCTION RISK FACTORS FOR THE... RISK FACTORS FOR THE... ANIMAL MODELS OF ASTHMA ANIMAL MODELS OF COPD AIRWAY INFLAMMATION IN ASTHMA AIRWAY INFLAMMATION IN COPD FUNCTIONAL ABNORMALITIES IN... FUNCTIONAL ABNORMALITIES IN... INFLAMMATORY MARKERS IN ASTHMA INFLAMMATORY MARKERS IN COPD RISK FACTORS FOR IRREVERSIBLE... RISK FACTORS FOR IRREVERSIBLE... EFFECTS OF INHALED... EFFECTS OF INHALED... NEW TREATMENT APPROACHES FOR... NEW TREATMENT APPROACHES FOR... CONCLUSIONS REFERENCES

The invasive methods to measure markers of airway inflammation include bronchial biopsy, bronchoalveolar lavage (BAL), and bronchial washings. Each of these measures different compartments in the airways. Induced sputum will sample the most proximal airways, including the trachea and major bronchi, possibly down to the sixth airway division. Bronchial wash at bronchoscopy involves the wedging of the bronchoscope in a subsegmental bronchus and so reflects endobronchial lining fluid sampling at that site, with no contribution from the trachea or large proximal airways. Bronchoalveolar lavage will sample the proximal airways and the distal small airways as well as the alveolar spaces. Endobronchial biopsy, performed during fiber-optic bronchoscopy, is from a large airway carina and so will reflect large airway mucosal events. Rationally, biopsies should be the gold standard if the markers measured are repeatable and give an accurate reflection of the abnormalities throughout the airways. Biopsy inflammatory markers are considered to be present when asthma is controlled. BAL predominantly samples the peripheral airways but is confounded by variable dilution with lavage fluid. Lymphocytes and macrophages predominate in BAL. Neutrophils and eosinophils (when present) are more prominent in the bronchial wash, which samples more proximal airways.

The noninvasive methods to measure markers of airway inflammation include the examination of sputum (169), blood (169), and urine (170) and measurement of exhaled NO (171). Of these, only sputum directly examines cell and molecular markers. The methods of sputum examination have been more extensively evaluated than other methods and can be highly repeatable and responsive to change. Two methods have been used to examine the expectorate of sputum plus saliva; one has successfully selected the sputum from saliva, to remove the confounding influence of the latter (172, 173). The other has examined the whole expectorate (174), although various maneuvers have been applied to reduce the amount of saliva (132, 175, 176). The best repeatability of cell and fluid-phase markers has been observed with selected sputum (173); however, the repeatability can also be good with examination of the whole expectorate (132, 174). Provided the sputum is selected, from the expectorate, the same results of cell and fluid-phase markers are obtained whether the sputum has been produced spontaneously or induced with hypertonic saline aerosol (177).

In comparison with blood eosinophil counts, sputum cell counts tak