INTRODUCTION

TOP ABSTRACT INTRODUCTION SPUTUM INDUCTION AND... EOSINOPHILIC BRONCHITIS WITH... EOSINOPHILIC BRONCHITIS IN COPD TREATMENT OF EOSINOPHILIC... REFERENCES

The term "bronchitis" generally refers to inflammation of the conducting airways of the lungs but is defined clinically by the spontaneous expectoration of sputum. In this article we refer to lower respiratory secretions as sputum and to conditions of the conducting airways associated with an excess proportion of eosinophils ( 3%) in spontaneous or induced sputum as eosinophilic bronchitis. Eosinophilic bronchitis is a common feature of asthma defined by the presence of variable airflow limitation (1), but it can occur in people without asthma, and in smokers with chronic bronchitis with or without chronic airflow limitation.

The recognition, investigation, and treatment of eosinophilic bronchitis, as well as other types of airway inflammation, have been facilitated by the application of refined methods of sputum examination to the analysis of induced sputum. These have enabled study of the characteristics of airway inflammation by a relatively noninvasive technique that can be used on repeated occasions when measurements are needed. This brief article presents our own contributions to these areas of work, some of the hypotheses our findings have led us to propose, and future directions of research.

	SPUTUM INDUCTION AND EXAMINATION

TOP ABSTRACT INTRODUCTION SPUTUM INDUCTION AND... EOSINOPHILIC BRONCHITIS WITH... EOSINOPHILIC BRONCHITIS IN COPD TREATMENT OF EOSINOPHILIC... REFERENCES

Sputum induction with an aerosol of hypertonic saline was introduced by Bickerman and colleagues (2) in 1958 for the diagnosis of lung cancer. It was later used to investigate the diagnoses of tuberculosis and opportunistic lung infections. In the early 1990s, Pin and coworkers (3) and Fahy and colleagues (4) modified the method of induction to make it safer to use in patients with asthma to investigate the characteristics of the airway inflammation. The method we use now is slightly modified from that described in our earlier work (5). The FEV₁ and VC are first measured to identify the presence and severity of any airflow limitation. An inhaled ₂-agonist (salbutamol, 200 µg) is then delivered to inhibit any possible bronchoconstriction from the saline inhalation, which is itself a bronchoconstrictive stimulus, and the FEV₁ is remeasured. Hypertonic saline aerosol is then inhaled from a Medix ultrasonic nebulizer (previously known as Fisoneb) (Clement Clarke International Ltd, Harlow Essex, UK), which has a relatively low output (0.87 mL/min) and a large particle size (5.58-µm aerodynamic mass median diameter) in concentrations of 3% followed by 4% and then 5%, each for an interval of 7 min. At the end of each inhalation, or earlier if symptoms of chest tightness or dyspnea develop, the FEV₁ is measured again. The inhalations are discontinued if there is a fall in FEV₁ of  20%. After each inhalation, the subject is asked to blow the nose, rinse the mouth with water, and swallow to reduce contamination of the expectorate with postnasal drip or saliva. The subject is then asked to cough sputum into a Universal container. This method is successful in raising enough sputum to be examined (70 mg being the smallest amount required), and success is enhanced by the experience of both the subject and of the person doing the induction. Overall, sputum induction is successful in more than 85% of attempts in healthy subjects or patients with airway disease. Attention must be paid to assure the safety of the procedure. We have found it to be safe in milder asthma (3, 5), and the procedure can be modified to make it safer in exacerbated asthma (6) or chronic obstructive pulmonary disease (COPD) (7) by starting with normal saline and by shortening the periods of inhalation.

Methods for examining sputum have been refined to achieve the reproducibility, validity, and responsiveness that are required of any good test (4). The methods we use are as follows (5, 9). The sputum is examined as soon as possible, always within 2 h of its induction. It is poured into a petri dish, and the more opaque and gelatinous portions, which look different from saliva, are all selected with blunt forceps and put into another petri dish. The selected portion is then moved around to remove a bit more of the saliva and then put into a weighed Eppendorf tube. If there is doubt about the selected portions, they can be confirmed as sputum by examining them under an inverted microscope to confirm that they contain macrophages and nonsquamous cells. The Eppendorf tube is weighed again to obtain the weight of the selected sputum. Four volumes of 0.1% dithiothreitol in relation to the weight is added, and the mixture is then gently aspirated in and out of a pipette, vortexed, and rocked on a bench rocker for 15 min. A further 4 volumes of Dulbecco's phosphate-buffered saline is then added, and the mixture is rocked for another 5 min before being filtered through 48-µm nylon gauze to remove debris and any remaining mucus. Total cell count and cell viability (using the trypan blue method) are determined in a hemocytometer. Cytospins are prepared and stained with Wright stain for a 400 differential nonsquamous cell count (and with toluidine blue and a 1,500 cell count if a metachromatic cell count is required). The cytospins can also be prepared for later immunocytochemistry or in situ hybridization (10, 11). The remainder of the suspension is centrifuged and the fluid phase is collected and stored at 70° C for later analyses of molecular components.

The effects of the induction and processing procedures on the content of sputum have been evaluated (12). They do not seem to interfere with the cell or many of the fluid-phase measurements. Induced sputum has characteristics similar to those of spontaneous sputum, except that the viability of the cells and the quality of the cytospins are better (13, 14). The duration of induction can influence the number and proportions of cells recovered (15, 16). For example, the later portions have lower total cell counts and fewer neutrophils and eosinophils, suggesting that cells from the more peripheral airways are being recovered. Because of this, it is appropriate in research to keep the duration of inhalation constant between measurements made on different occasions. The cell counts for neutrophils, eosinophils, macrophages, and metachromatic cells are highly repeatable, but the repeatability of total cell counts and lymphocytes is not as reliable (5). The reason for the poor repeatability of total cell counts has not been studied; it may be methodological or real. For the present it has led us to put more reliance on changes in the proportion of individual cell types rather than on changes in the absolute counts, although we still measure total cell count and take it into account, particularly when there is neutrophilia. The poor repeatability of lymphocytes is methodological; it is due to the difficulty in accurate recognition of the cells and to their small numbers, and can be improved by counting more cells.

At present, normal values are limited to small numbers of subjects; the results from larger samples of healthy subjects need to be reported. At present we regard as within the normal range a total cell count of < 5.5 million per gram of selected sputum, eosinophils < 3%, neutrophils < 36%, and lymphocytes < 2% (5, 17). The proportion of neutrophils in specimens with normal total cell counts is quite variable. The proportion of macrophages obviously depends on the proportion of neutrophils and eosinophils present. Bronchial epithelial cells are not seen frequently; when examined in the hemocytometer, they are nonviable, and they may be lost in the subsequent cytospin preparations.

The measurements obtained will be reliable only if quality control is maintained. To ensure this, we use support staff with permanent appointments. Research nurses and pulmonary function technologists perform sputum induction and hematology-trained registered technologists perform sputum processing and examination. The induction procedure can take up to 30 min; but, in clinical practice, the procedure can be aborted when sufficient sputum is obtained. Processing sputum and performing cell counts takes between 1 and 2 h, but with some automation of the procedures of rocking, cytospins, and staining, the amount of technologist time can be reduced to about 30-45 min.

	EOSINOPHILIC BRONCHITIS WITH AND WITHOUT ASTHMA

TOP ABSTRACT INTRODUCTION SPUTUM INDUCTION AND... EOSINOPHILIC BRONCHITIS WITH... EOSINOPHILIC BRONCHITIS IN COPD TREATMENT OF EOSINOPHILIC... REFERENCES

Eosinophilic bronchitis is a common but not consistent feature of asthma (4, 5, 18). It can occur when symptoms are absent and the FEV₁ is normal, but it is more likely in symptomatic subjects (5). The total cell count is normal or slightly increased, and the proportion of eosinophils above 3%. Free eosinophil granules may be seen. The proportion of neutrophils is also slightly increased compared with healthy subjects. In severe exacerbations, there can be considerable cell degeneration (6). In the fluid phase, measurements that one would expect to be increased in a helper T cell type 2 (Th2) response are elevated, e.g., different eosinophil proteins, tryptase, and interleukin 5 (IL-5) (4, 5). The latter is more difficult to demonstrate since, for reasons presently unknown, some IL-5 is lost during the processing procedures. There are also increases in fluid-phase fibrinogen and albumin, in keeping with an increase in microvascular leakage, and of mucin-like glycoprotein (19), in keeping with an increase in secretion of mucus.

Not all exacerbations of asthma are associated with an eosinophilic bronchitis. Some mild (20), severe (21), or fatal (22) exacerbations have been associated with neutrophilia. One cause of this has been identified to be respiratory viral infections, specifically influenza A and B (23). However, there are also other causes of neutrophilia that could occur in subjects with asthma, such as smoking (5), exposure to endotoxin from bacterial infections or in the environment (24), or exposure to atmospheric pollutants (25). The frequency of noneosinophilic exacerbations of asthma and the mechanisms involved are unknown.

One of the earliest observations to come from the use of reliable sputum examination was the first identification of eosinophilic bronchitis in patients without asthma (26). Seven patients were described who presented with a chronic productive cough. While their sputum demonstrated an increase in the proportion of eosinophils and metachromatic cells, features of variable airflow limitation were absent. Their spirometry, daily peak expiratory flow variability, and methacholine and adenosine monophosphate airway responsiveness were normal (27).

Eosinophilic bronchitis without asthma has been observed in 10% of patients with a chronic dry or productive cough referred to a tertiary care center (28). The patients may be atopic or not, or smokers or not. The condition has been observed to be caused in some patients by inhaled allergens (29) or occupational chemical sensitizers (30). It can be transient (31) or persistent and can require regular treatment with inhaled corticosteroid prednisone (32). If the condition is left untreated, methacholine airway hyperresponsiveness and other features of asthma can develop, and we have observed one patient who had a progressive fall in FEV₁ that could not be reversed by ₂-agonist (Table 1) but was reversed by prednisone, and was not associated with methacholine airway hyperresponsiveness (32).

	EOSINOPHILIC BRONCHITIS IN COPD

TOP ABSTRACT INTRODUCTION SPUTUM INDUCTION AND... EOSINOPHILIC BRONCHITIS WITH... EOSINOPHILIC BRONCHITIS IN COPD TREATMENT OF EOSINOPHILIC... REFERENCES

In smokers with a chronic productive cough consistent with the definition of chronic bronchitis, but who do not have chronic airflow limitation, induced sputum shows inclusions in the macrophages and a mild neutrophilia (Table 2) (5). When COPD develops, sputum neutrophilia becomes more pronounced and can be seen to correlate with the reduction in FEV₁ (33) and with the rate of deterioration in FEV₁ (34). The sputum shows an increase in neutrophil proteins such as myeloperoxidase, elastase, and human neutrophil lipocalin, and IL-8 (7, 35, 36). There is also an elevation of eosinophil proteins and this is thought to be due to a nonselective increase in eosinophils owing to an increase in the total cell count (37).

In some smokers with COPD, an eosinophilic bronchitis can also occur (38). This has implications as to its cause and the effects of treatment. The occurrence of eosinophilia during exacerbations in patients with smokers' COPD has been attributed to the same disease, i.e., to the inflammatory effects of cigarette smoking (39). We hypothesize that it is more likely the result of other causes such as inhaled allergen (29) or chemical sensitizer (25, 30) or whatever other cause is responsible for eosinophilic bronchitis with or without asthma. The fact that there are many different causes of airway inflammation and that these cause inflammation of different types, associated with different clinical effects, make it likely that these different causes of inflammation during life will interact with the effects of cigarette smoking and the resulting clinical presentation. This theory has important implications for treatment with avoidance strategies or corticosteroid.

	TREATMENT OF EOSINOPHILIC BRONCHITIS

TOP ABSTRACT INTRODUCTION SPUTUM INDUCTION AND... EOSINOPHILIC BRONCHITIS WITH... EOSINOPHILIC BRONCHITIS IN COPD TREATMENT OF EOSINOPHILIC... REFERENCES

The response to treatment with corticosteroid in asthma, in chronic cough without asthma, and in COPD, seems to depend on whether sputum eosinophilia is present.

In asthma, eosinophilic bronchitis responds to corticosteroid treatment (6, 40). The evidence that the absence of eosinophilic bronchitis predicts nonresponsiveness is limited to isolated case reports (6, 43). Systematic investigation of the effect of steroid treatment in these patients is needed.

Chronic cough associated with eosinophilic bronchitis always responds to steroid treatment (26). Cough improves, and sputum eosinophilia resolves. While this usually occurs after treatment with inhaled steroid, prednisone is sometimes required (26). Occasionally, the condition requires regular prednisone treatment (32). Chronic cough without sputum eosinophilia does not seem to respond to steroid treatment (44).

In COPD associated with eosinophilic bronchitis, a single blind sequential cross-over treatment trial of eight subjects with eosinophilic bronchitis has shown inhaled corticosteroid treatment to improve quality of life (chiefly the dyspnea domain) and FEV₁ (Figure 1) (7, 45). This trial of sputum examination was done double blind and showed eosinophilia to reverse and ECP and fibrinogen levels to fall; neutrophils were unchanged, unlike the reduction reported by Confalonieri and coworkers after 2 mo of treatment with inhaled steroid (46). Ten subjects without sputum eosinophilia (< 3%) did not benefit. Despite the improvement in sputum eosinophilia in this study, the overall improvement in FEV₁ was only 200 ml. Was this because some subjects had an eosinophilic bronchitis without asthma, or was it because the FEV₁ is an insensitive measure of improvement when chronic airflow limitation is severe? In future studies, more sensitive objective outcome measures of airway function improvement need to be used, such as inspiratory capacity or exercise endurance (47, 48).

View larger version (36K): [in this window] [in a new window]   Figure 1.   Sputum eosinophils (%), postbronchodilator FEV1 (L), and quality of life score in smokers with severe chronic airflow limitation between treatment, 2 wk after placebo treatment, and 2 wk after prednisone treatment (30 mg daily). Those with raised eosinophils at baseline (closed circles) have significant improvement in each outcome after prednisone treatment. *p = 0.05; **p < 0.001. (Reprinted from Reference 45 with permission from the publisher; data published in Reference 7.)