INTRODUCTION

TOP ABSTRACT INTRODUCTION PHENOTYPING THE DISEASE THE VARIOUS LUNG COMPARTMENTS THE INFLAMMATORY PROCESS REFERENCES

Developments in molecular and cell biology have provided investigators of chronic obstructive pulmonary disease (COPD) with novel concepts and tools for the investigation of inflammatory processes in the lung. Techniques allowing precise cell phenotyping and determination of cytokine production and activation markers, among others, have allowed for a much better description of the inflammatory profiles in the lungs of smokers. Interest in the area has generated a welcome explosion of information about lung inflammation in COPD. A review of the literature would show that the neutrophil, the eosinophil, the lymphocyte, and the macrophage are involved in the inflammatory process of COPD. However, the predominant cell varies from publication to publication, often depending on the method used for the assessment of the inflammation or the selection of patients. This disparity of results, although potentially enriching, tends to confuse the issue of inflammation, in part owing to the absence of a unifying approach to interpreting these findings. In response to the obvious need for some standardization in the area, a European Respiratory Society task force has published a supplement on methods for assessment of airway inflammation, in which the various methodologies necessary for the study of airway inflammation in asthma and COPD are reviewed and standards recommended (1).

The main issues that need to be considered in order to understand better the inflammatory process are disease phenotyping, compartmentalization of the lung, and the inflammatory process itself.

	PHENOTYPING THE DISEASE

TOP ABSTRACT INTRODUCTION PHENOTYPING THE DISEASE THE VARIOUS LUNG COMPARTMENTS THE INFLAMMATORY PROCESS REFERENCES

Smoking-induced lung disease is a complex process producing various abnormalities that can be defined pathologically (emphysema), clinically (chronic bronchitis), and functionally (COPD: airflow limitation and other abnormalities in lung function). The complexity is compounded by how the disease is recognized or referred to by various authors when describing the inflammatory changes in the lungs of smokers, as illustrated by the following excerpts: "Leukocyte numbers are increased in the bronchial epithelium and lamina propria of smokers with chronic bronchitis as compared to healthy non-smokers" (2); "However no differences were observed in the number of neutrophils between bronchial biopsies of smokers with chronic airflow limitation and control subjects" (3); "Increased infiltration of central and peripheral airways with CD8+ lymphocytes in subjects with chronic obstructive bronchitis compared to healthy smokers was reported recently" (3, 4).

To understand better what these findings mean, we need first to be certain how best to define chronic bronchitis, chronic obstructive bronchitis, healthy smokers, and chronic airflow limitation. Chronic bronchitis is defined clinically as chronic productive cough. Because the main pathological counterpart of chronic bronchitis is the enlargement and probably inflammation of the bronchial glands in the cartilaginous airways (larger than 2 mm in diameter) (5), the biopsy findings in the large airways of smokers with chronic bronchitis (where glands are present) might be different from the findings in the small airways (i.e., those less than 2 mm and with no cartilage and no glands).

Findings in exacerbations are of interest. Large numbers of neutrophils and eosinophils are found in both sputum and bronchial biopsies (6, 7). Since exacerbations are most often secondary to an infection (viral or bacterial or both) this is not surprising. It would be interesting to compare induced sputum and bronchial biopsy during an acute viral or bacterial bronchitis between smokers and nonsmokers. Unfortunately, smokers are typically compared with healthy nonsmokers.

In emphysema, airflow limitation, the hallmark of this condition, is determined by the decrease in elastic recoil and the increased resistance in the small airways. In an autopsy series Nagai and coworkers (8) found that for the same degree of airflow limitation, smokers with lesser degrees of emphysema had more diseased small airways. Furthermore, Kim and colleagues (9) reported that the inflammation and remodeling of airways in smokers could be different in lungs with centrilobular and panlobular emphysema, with centrilobular emphysema showing a more severe airway inflammation and remodeling. Finkelstein and coworkers (10) showed that in centrilobular emphysema the airways were narrower and thicker than in panlobular emphysema. Large numbers of inflammatory cells, mainly lymphocytes but also polymorphonuclear leukocytes, were present in both. The number of lymphocytes was statistically similar in nonsmokers and in smokers with centrilobular and panlobular emphysema. However, the number of CD3⁺ lymphocytes per cubic millimeter correlated significantly with the thickness of the internal wall and airway reactivity in centrilobular emphysema but not in panlobular emphysema and nonsmokers (10). Further characterization of lymphocytes in the airways of smokers is necessary.

Lams and coworkers (11) reported the immunopathology of the small airway submucosa in smokers with and without COPD. One of their main findings was the presence of eosinophils in the submucosa of the airways of smokers, but not in those of nonsmokers. Furthermore, the number of eosinophils was related to an increased CD8⁺/CD3⁺ cell ratio in all subjects. A wide variability in the number of eosinophils in smokers (0 to 217/mm³) and in the number of eosinophils at each level of CD8⁺/CD3⁺ cell ratio was seen. The variability may be related to the fact that not all smokers react similarly to cigarettes and that the inherent reactivity of their airways is probably a key factor in the inflammatory profile of the airways. Nagai and colleagues (8), in a study of 41 lungs obtained from patients who died of COPD, reported that lack of airway responsiveness (measured by bronchodilator response in FEV₁) was associated with lesions such as emphysema and goblet cell metaplasia. Patients with increased airway responsiveness had less emphysema and responsiveness correlated with the number of eosinophils. These findings complement the findings by Finkelstein and coworkers (10), in which the inflammatory pattern in the airways differed in the various smoker groups. It is quite apparent that not all smokers develop the same type of disease and it is likely that the pattern of inflammation might vary with the different disease phenotypes. Suggested terms to describe the phenotype of cigarette smoke-related lung disease are listed in Table 1.

	THE VARIOUS LUNG COMPARTMENTS

TOP ABSTRACT INTRODUCTION PHENOTYPING THE DISEASE THE VARIOUS LUNG COMPARTMENTS THE INFLAMMATORY PROCESS REFERENCES

Inflammation of the lung in COPD is now being studied by three different methods: sputum, either spontaneous or induced; bronchial and bronchoalveolar lavage; and bronchial biopsies, taken either through a bronchoscope or from surgical specimens. These three methods identify in COPD different inflammatory cell populations; these differences are expected, given that the techniques sample different parts of the lung.

Sputum is a sample of normal or altered airway liquid that consists of secretions from bronchial glands, goblet cells, ciliated cells, and material exudated from the bronchial circulation. Normally the airway liquid, along with the ciliary system, keeps the tracheobronchial tree free of offending organisms or other particles. However, certain inflammatory cells participate in airway defense as well. CD8⁺ and T lymphocytes are more abundant in the epithelial layers and can be found in most sputum samples. Circulating leukocytespolymorphonuclear leukocytes (PMNs) and eosinophilswould be the obvious inflammatory cells able to respond and migrate in response to an insult at the epithelial surface, with the help of NO and other vasodilating substances. Probably under most circumstances, the CD4⁺ T cell placed in the airway wall tissue itself might direct the inflammatory response by the regulation of the endothelial cells to produce vasodilator substances and adhesion molecules, but would not be detected in the sputum itself.

Induced sputum seems to be a combination of "resident" mucus, freshly produced bronchial secretions, and probably vascular transudate. As such, induced sputum seems to have different fluid-phase measurements and higher cell viability compared with spontaneously produced sputum. This suggests that blood-borne cells and other substances might be recruited by the sputum induction maneuver (12).

Bacterial colonization and infection can increase the numbers of neutrophils and eosinophils in sputum, probably by mechanisms totally different from those responsible for the ongoing inflammation of COPD (13). This may create problems in interpretation. Bronchoalveolar lavage (BAL) findings differ from sputum findings because the former technique samples mainly the alveolar compartment whereas the latter samples primarily the bronchial compartment. In general, BAL tends to have more macrophages while sputum is richer in neutrophils. However, there seems to be good agreement in the number of eosinophils between sputum, BAL, and biopsies (14). The results of both sputum induction and BAL seem to be affected by the procedure itself. One report has suggested that when repeating sputum induction at daily intervals, the proportion of neutrophils in induced sputum increases (15). Similarly, the number of neutrophils in BAL is increased 7 and 24 h after performing a BAL (16). Although these findings do not detract from the value of these procedures, they do highlight the fact that neutrophil migration to the airway lumen can be induced by the sampling procedure.

Bronchial biopsies should not be considered the gold standard for comparison with sputum and BAL, since bronchial biopsies represent yet another compartment. Direct comparisons of inflammatory markers in sputum, BAL, and biopsies show that there is a good correlation in the numbers of eosinophils, but no correlation in the numbers of other inflammatory cells, and the main difference in cell counts involves the T lymphocyte (14). Few lymphocytes are seen in BAL and even less in sputum when compared with bronchial biopsies, suggesting that neutrophils and eosinophils migrate into the airway lumen more readily than do lymphocytes.

The site chosen for bronchial biopsies is another important consideration. Biopsies are obtained from fairly proximal cartilaginous airways. However, the site of airflow obstruction in COPD is mainly the membranous bronchiole, which is devoid of cartilage and bronchial glands. Biopsies of the large airways in smokers with chronic bronchitis would show an inflammatory cell infiltrate around the bronchial glands. This inflammation not only could be different from the inflammation occurring in the remainder of the airway wall, but it might not be present in the small airways. However, since sputum probably originates in the large airways, bronchial biopsies (from large airways) might be better than small airway biopsies when comparisons with sputum are made.

The data available suggest that different techniques sample different compartments of the lung, and that each instrument has properties that might differ in usefulness depending on the objective of the examination. More studies correlating the findings in the three compartments are needed in order to understand how the patterns correlate under a wide variety of inflammatory conditions.

	THE INFLAMMATORY PROCESS

TOP ABSTRACT INTRODUCTION PHENOTYPING THE DISEASE THE VARIOUS LUNG COMPARTMENTS THE INFLAMMATORY PROCESS REFERENCES

The neutrophil was the first cell implicated in the pathological abnormalities and pathogenesis of emphysema (17, 18). It was followed shortly by the appearance of the alveolar macrophage (AM) (19, 20). Both cells have the required proteolytic enzymes necessary to degrade the lung parenchyma and produce emphysema. Later, when more precise quantitative methods to assess lung pathology and inflammation were used, the T lymphocyte appeared as a potential protagonist of the abnormalities in airways and lung parenchyma in COPD (7, 21). The eosinophil has also appeared now and then as a contender for lung injury (6, 11).

At first, a review of the extensive literature now available on lung inflammation seems overwhelming and confusing. Each work heralds one protagonist cell and, in general, we tend to defend "our cell" as the cell responsible for the disease process. All of these cells must be important and interrelated as part of the general inflammatory profile seen in smokers. An attempt should be made to synthesize all available data to understand how these inflammatory cells are interacting in smokers.

It is possible that an important consideration in understanding the inflammatory process in COPD is the dose effect, or time at which the study is performed in the life of the subject. One study describing the inflammation of the airways in smokers emphasizes that the increase in CD8⁺ T cells is related to the pack-years of smoking; furthermore, the decrease in these T cells was related to the months of smoking cessation (11). Similarly, our own data on inflammation of the lung parenchyma show that the numbers of CD3⁺, CD4⁺, and CD8⁺ T cells are similar in nonsmokers and smokers with a dose of less than 25 pack-years. Only with greater doses do T cells increase and neutrophils decrease in the lungs, and does emphysema appear (22).

The first inflammatory cell responding to smoking is most likely the neutrophil. In animal studies, exposure to cigarette smoke and other irritants promotes neutrophils to appear promptly in the airways. Hulbert and colleagues (23) showed that when guinea pig airways were exposed to cigarette smoke, the number of neutrophils in the airway epithelium increased fivefold from control values 6 h after injury. In human smokers, Bosken and colleagues (24) demonstrated no differences between smokers with and without COPD in terms of the number of immunohistochemically stained neutrophils in the airways. However, the number of submucosal neutrophils correlated significantly with the number of cigarettes smoked. The mechanism by which cigarette smoke promotes neutrophil accumulation in the lung is not yet clear, but several possibilities can be postulated. Components of cigarette smoke such as nicotine have been found to be chemotactic for human neutrophils (25). More likely, the irritation and injury of the epithelium by smoke could be the result of neurogenic inflammation mediated by substance P. Substance P stimulates airway mucus secretion, increases microvascular permeability and leakage of neutrophils, and promotes alveolar macrophage stimulation. Ablation of substance P-producing sensory neurons by capsaicin markedly inhibits the noxious effects on the lung of cigarette smoke and other noxious stimuli (26).

It is likely, then, that cigarette smoking through epithelial irritation produces an inflammatory response in the lung from the early stages of exposure to cigarettes, and that this stimulus may last throughout the life of the active smoker. Both neutrophils and alveolar macrophages could be recruited in this way. After years of exposure and years of potential injury and degradation of the lung parenchyma mediated by neutrophils and macrophages, T cells become prominent at the site of injury. Available data in lung parenchyma show a significant relation between the number of CD3⁺ T cells and the number of AMs, and both cell types increase in number as the extent of emphysema increases, suggesting a "partnership" in the production of lung injury (21).

Naive, non-antigen-activated T cells do not stay long in the lung. They return to the circulation or die. A role for T cells in causing a particular inflammatory disease is suspected largely because of the demonstration of T cells in the affected organ. For T cells to "home" to the lungs (or any organ) and become more numerous, they need first to be activated by the recognition of an antigen and then to "home" to the organ in which antigen is being produced. Once in the lung, activated T cells can exert their effector functions: the CD4⁺ (or CD8⁺) T cell would do so by the production of cytokines in either a helper T cell type 1 (Th1) (interleukin 2 [IL-2], interferon  [IFN-]) or Th2 (IL-4, IL-5) pattern. The presence of large numbers of AMs and the significant correlation between the numbers of CD3⁺ T cells and AMs in the lungs of smokers (27) suggest that a Th1 CD4⁺ T cell is involved in this inflammatory process. CD4⁺ Th1 T cells produce IFN-, the most potent macrophage-activating cytokine. Cytokines produced by CD8⁺ T cells can initiate the same reaction. On activation AMs express the following functions: (1) they generate reactive oxygen species and high-output nitric oxide synthase that catalyzes the production of NO; (2) they stimulate acute inflammation through secretion of short-lived inflammatory mediators (platelet-activating factor [PAF], prostaglandins, leukotrienes, etc.); (3) they become more efficient as antigen-presenting cells; a significant part of the enhanced antigen-presenting capacity may be attributed to increased surface expression of class II MHC molecules owing to the activation of the transcription of MHC II genes by IFN- (28); and (4) CD4⁺ and CD8⁺ lymphocyte-derived IFN- stimulates AMs to secrete cytokines, including IL-12, which feed back to the T cell line promoting T lymphocytes to differentiate into Th1 subsets promoting CD4⁺ (and CD8⁺) Th1 cytokine production and cytotoxic T lymphocyte (CTL) differentiation (27).

Another extremely important effector function of the T lymphocyte is the alteration of its microvascular environment. Under the influence of cytokines secreted by antigen-activated T cells or through contact-dependent signals, microvascular endothelial cells perform four functions that contribute to inflammation (27): (1) vasodilatation increases local blood flow and delivery of leukocytes to sites of inflammation via prostaglandin I₂ (PGI₂) and NO; (2) by expression of new or increased levels of certain surface proteins, postcapillary venular endothelial cells become adhesive to leukocytes; (3) antigen- activated T cells cause endothelial cells to secrete chemokines such as IL-8 and monocyte chemotactic protein 1 (MCP-1), which act on the leukocytes to promote extravasation; and (4) cytokines or contact-dependent signals from activated T cells cause endothelial cells to undergo shape changes and basement membrane remodeling that favor leakage of macromolecules and extravasation of cells. Thus, the interplay of adhesion molecules and chemokines leads to the multistep model of leukocyte recruitment.

There is now ample evidence that some smokers develop eosinophilic inflammation (11, 14, 28, 29). Furthermore, these subjects seem to have special characteristics resembling those of patients with asthma: they respond to steroids and they have increased airway reactivity. Also, the eosinophils seem to be activated. If this is the case, it would be logical to assume that in some smokers, some of the T cells present in the airways belong to the Th2 subset secreting IL-4 and IL-5. This possibility might be worth investigating, since it could add greatly to our understanding of this disease.

The various inflammatory cells encountered in COPD could be understood better if we try to link them together. Possibly the earliest inflammatory reaction involves the neutrophil, followed by the AM in all the epithelial surfaces of the lung. These cells would in time damage epithelial cells and the interstitial protein structures (elastin, collagenase, proteoglycans, etc.); these proteins could be processed by dendritic cells (and eventually AMs) into peptides with antigenic potential that could eventually be recognized by T cells initiating T cell activation and proliferation. As we have already seen, these activated T cells, as part of their effector function, could recruit other cells such as neutrophils, AMs, and even eosinophils to the site of inflammation. Thus, all of the different inflammatory cells described here and presumed to be working on their own can now be seen as working together toward the production of airway abnormalities, lung destruction, and eventually COPD.