EOSINOPHILIC BRONCHITIS IS AN IMPORTANT CAUSE OF CHRONIC COUGH

To the Editor :

Brightling and colleagues have modified a well validated protocol for the investigation of chronic cough with the analysis of induced sputum (1). They report a different spectrum and distribution of diagnoses compared with a number of published studies of chronic cough, in which postnasal drip syndrome, cough-variant asthma (CVA) and gastroesophageal reflux (GOR), either alone or in combination account for almost all causes of cough (2). They describe a group of patients with eosinophillic bronchitis (EB), which is defined as the presence of sputum eosinophilia (> 3%) and absence of bronchial hyperresponsiveness (PC₂₀ methacholine > 8 mg/ml) accompanied by steroid-responsive cough. This observation occurs with an unknown prevalence of CVA (i.e., hyperresponsiveness) in the absence of variable airflow obstruction and a lower prevalence of individuals with GOR-associated cough compared with the above series. This difference may partly reflect differences in baseline study populations, although we have a number of concerns regarding their study.

Our main concern centers on the definition and existence of a separate entity of EB. The only factor that appears to separate CVA (which is itself associated with an eosinophilic bronchitis) from EB is hyperresponsiveness to methacholine at an arbitrarily defined cut-off point (PC₂₀ < 8 mg/ml). Using this point, hyperresponsiveness to methacholine is seen in virtually all patients with symptomatic asthma (defined as having cough, wheeze and variable airflow obstruction) but it is also present in normal subjects with no symptoms and absent in some subjects with definite clinical asthma (5). Thus, separating CVA from the new entity of EB on this problematic, albeit objective, measurement seems inappropriate. It may be more appropriate to revise the diagnostic threshold for hyperresponsiveness in asthmatic cough.

Irrespective of nomenclature, methacholine challenge is useful in the assessment of chronic cough. We found a negative methacholine challenge (PC₂₀ > 8 mg/ml) had a 100% negative predictive value for CVA and subsequent response to steroids (4). Brightling and colleagues suggest that we may have overlooked subjects with EB and steroid-responsive cough, but in all our subjects with PC₂₀ > 8 mg/ml (n = 26), we found the cough was either due to an alternative diagnosis (n = 18) or failed to respond to sequential trials of all therapeutic modalities, including high dose oral steroids (n = 8). Thus, in our study population, steroid-responsive cough was not overlooked.

We would further suggest that not all cough patients with eosinophils in the airways require steroid therapy. We have recently reported that eosinophil numbers are increased in bronchoalveolar lavage fluid from patients with GOR-related cough (6). These patients responded to antireflux therapy with improvement in capsaicin cough threshold. Micheletto and colleagues have measured elevated levels of ECP in induced sputum from nonasthmatic coughers with GOR (7). If EB does exist as a separate clinical entity, our findings and those of others suggest it would be important to exclude GOR as an underlying cause. Brightling and coworkers infrequently diagnose GOR (7%), and their protocol inadequately assesses GOR, becaue it relies on symptoms to precipitate confirmatory investigations. Previous studies, including our own, have demonstrated that symptoms are insensitive in detecting GOR (up to 50% are missed) and that some of the group with EB may have had silent GOR (8).

Eosinophilic bronchitis is a descriptive pathological term that is likely to encompass a number of different etiologies that cause chronic cough. Introduction of this entity may add confusion to a difficult area where the use of established diagnostic protocols has consistently yielded excellent response rates.

Department of Respiratory Medicine,Belfast City Hospital,Belfast, Northern Ireland

Madeleine Ennis

Respiratory Research Group,Clinical Biochemistry,The Queen's University of Belfast,Belfast, Northern Ireland

1. Brightling, C. E., R. Ward, K. L. Goh, A. J. Wardlaw, and I. D. Pavord. 1999. Eosinophilic bronchitis is an important cause of chronic cough. Am. J. Respir. Crit. Care Med. 160: 406-410 [Abstract/Free Full Text].

2. Irwin, R. S., F. J. Curley, and C. L. French. 1990. Chronic cough: the spectrum and frequency of causes, key components of the diagnostic evaluation and outcome of specific therapy. Am. Rev. Respir. Dis. 141: 640-647 [Medline].

3. O'Connell, F., V. E. Thomas, N. B. Pride, and R. W. Fuller. 1994. Cough sensitivity to inhaled capsaicin decreases with successful treatment of chronic cough. Am. J. Respir. Crit. Care Med. 150: 374-380 [Abstract].

4. McGarvey, L. P. A., L. G. Heaney, J. T. Lawson, B. T. Johnston, C. M. Scally, M. Ennis, D. R. T. Shepherd, and J. MacMahon. 1998. Evaluation and outcome of patients with chronic nonproductive cough using a comprehensive diagnostic protocol. Thorax 53: 738-743 [Abstract/Free Full Text].

5. Casale, T. B., B. J. Rhodes, A. L. Donnelly, and J. M. Weller. 1988. Airway reactivity to methacholine in nonatopic asymptomatic adults. J. Appl. Physiol. 64: 2558-2561 [Abstract/Free Full Text].

6. McGarvey, L. P. A., P. Forsythe, L. G. Heaney, J. MacMahon, and M. Ennis. 1999. Bronchoalveolar lavage findings in patients with chronic non-productive cough. Eur. Respir. J. 13: 59-65 [Abstract].

7. Micheletto, C., E. Burti, L. Mauroner, C. Pomari, P. Turco, and R. Dal Negro. 1997. Induced sputum and serum inflammatory markers in subjects with cough due to gastro-oesophageal reflux (abstract). Eur. Respir. J. 10-317s.

8. Irwin, R. S., J. K. Zawacki, F. J. Curley, C. L. French, and P. J. Hoffmann. 1989. Chronic cough as the sole presenting manifestation of gastro- oesophageal reflux. Am. Rev. Respir. Dis. 140: 1294-1300 [Medline].

To the Editor :

Brightling and colleagues (1) emphasize the importance of nonasthmatic eosinophilic bronchitis in treating patients with chronic cough. Eosinophilic airway disease as a cause of cough has been proposed as "eosinophilic bronchitis" by Gibson and colleagues (2, 3). We have been also studying chronic bronchodilator-resistant nonproductive cough in atopic patients in comparison with cough variant asthma from an independent standpoint and have proposed "eosinophilic tracheobronchitis with airway cough hypersensitivity," in which cough response to inhaled capsaicin is increased (4), eosinophils are documented in the induced sputum (5) and the submucosa of biopsied trachea and bronchi but not the bronchoalveolar lavage fluid (6), cough is nonproductive, corticosteroids but not bronchodilators are effective, and bronchial responsiveness to methacholine is not heightened (4). In our practice, eosinophilic airway disease is diagnosed in half of the patients presenting with chronic nonproductive cough, in agreement with Brightling and colleagues (1). Although it has been shown that the character of cough is not useful in making diagnoses, Brightling and coworkers (1) state that patients with eosinophilic bronchitis present with a dry cough or a cough producing small amounts of viscid sputum in the morning, whereas Gibson and coworkers (2, 3) state that the cough is productive, and doubt whether the airway disease described by Brightling and colleagues is the same as Gibson's eosinophilic bronchitis. Furthermore, the cough is essentially nonproductive in the eosinophilic tracheobronchitis with cough hypersensitivity that we propose.

Thus, I question whether eosinophilic bronchitis and eosinophilic tracheobronchitis with cough hypersensitivity are identical because the following have not been demonstrated in eosinophilic bronchitis: (1) effect of bronchodilator therapy, and (2) cough sensitivity to inhaled capsaicin or citric acid. The only finding suggestive of difference between the two eosinophilic airway diseases is presence or absence of eosinophils in the bronchoalveolar lavage fluid. Gibson and colleagues (3) have clearly shown that the eosinophilic bronchitis and asthma have a common pattern of inflammation in terms of both cellular infiltration and inflammatory cytokine gene expression, whereas no increase in BAL eosinophils is characteristic of the eosinophilic tracheobronchitis with cough hypersensitivity (6). Thus, I recommend that cough sensitivity and effect of bronchodilator therapy (and BAL cell findings as far as possible) should be included for the diagnosis of the eosinophilic bronchitis.

MASAKI FUJIMURA

The Third Department of Internal Medicine

Kanazawa University School of Medicine

Kanazawa, Japan

1. Brightling, C. E., R. Ward, K. L. Goh, A. J. Wardlaw, and I. D. Pavord. 1999. Eosinophilic bronchitis is an important cause of chronic cough. Am. J. Respir. Crit. Care Med. 160: 406-410 .

2. Gibson, P. G., J. Dolovich, J. Denburg, E. H. Ramsdale, and E. F. Hargreave. 1989. Chronic cough: eosinophilic bronchitis without asthma. Lancet 17: 1346-1348 .

3. Gibson, P. G., K. Zlatic, J. Scott, W. Sewell, K. Woolley, and N. Saltos. 1998. Chronic cough resembles asthma with IL-5 and granulocyte-macrophage colony-stimulating factor gene expression in bronchoalveolar cells. J. Allergy Clin. Immunol. 101: 320-326 [Medline].

4. Fujimura, M., Y. Kamio, T. Hashimoto, and T. Matsuda. 1994. Cough receptor sensitivity and bronchial responsiveness in patients with only chronic nonproductive cough: in view of effect of bronchodilator therapy. J. Asthma 31: 463-472 [Medline].

5. Fujimura, M., N. Songur, Y. Kamio, and T. Matsuda. 1997. Detection of eosinophils in hypertonic saline-induced sputum in patients with chronic nonproductive cough. J. Asthma 34: 119-126 [Medline].

6. Fujimura, M., H. Ogawa, M. Yasui, and T. Matsuda. 2000. Eosinophilic tracheobronchitis and airway cough hypersensitivity in chronic non-productive cough. Clin. Exp. Allergy 30: 41-47 [Medline].

From the Authors:

We welcome Dr. Fujimura's comments confirming our finding that nonasthmatic eosinophilic airway inflammation is a common cause of chronic cough (1). There are a number of similarities between the condition described by Fujimura and colleagues (2) known as atopic cough and eosinophilic bronchitis. Both present with a corticosteroid-responsive cough, a sputum eosinophilia, normal spirometric values, normal airway responsiveness and increased capsaicin cough sensitivity (3). Atopy appears to be more common in atopic cough and the cough is usually dry, whereas we and others have noted that the cough in eosinophilic bronchitis is occasionally productive of tenacious sputum. Fujimura proposes that the main difference between the two conditions is the site of eosinophilic airway inflammation. In atopic cough the inflammation is confined to the upper airway with biopsy evidence of a tracheobronchitis without an eosinophilia present in the bronchoalveolar lavage (BAL) (1). In contrast Gibson and colleagues (4) have shown a similar degree of BAL eosinophilia and GM-CSF and IL-5 gene expression in patients with eosinophilic bronchitis compared with subjects with asthma. We have shown that eosinophilic bronchitis is not typically associated with a nasal wash eosinophilia or upper airway hyperresponsiveness (3). These observations would not support a predominant upper airway inflammation and suggest that atopic cough and eosinophilic bronchitis are distinct conditions.

Dr. Fujimura recommends that an assessment of BAL cell findings may be necessary to differentiate the two conditions. This would be particularly important if the natural history and long-term management differs. The natural history of unrecognized or undertreated eosinophilic bronchitis is unknown. However, up to 40% of patients with chronic obstructive pulmonary disease without a history of asthma have sputum evidence of an eosinophilic airway inflammation (5), so one intriguing possibility is that such patients develop irreversible structural changes to their airways and fixed airflow obstruction. In support of this we have reported a patient with eosinophilic bronchitis who had a good symptomatic response to inhaled corticosteroids but had sputum evidence of persistent eosinophilic airway inflammation and developed fixed airflow obstruction over 3 yr (6). If eosinophilic bronchitis is associated with a lower airway inflammation and airway remodelling, in contrast to a possibly more benign natural history with atopic cough, there would be a need for establishing a clear diagnosis, as eosinophilic bronchitis may require more intensive anti-inflammatory treatment aimed at normalizing the sputum eosinophilia as well as the cough.

McGarvey and colleagues (7) question whether some of our patients had cough-variant asthma or eosinophilic airway inflammation secondary to silent gastroesophageal reflux. Methacholine airway responsiveness was well inside the normal range (PC₂₀ > 16 mg/ml) in all our subjects and all had normal spirometric values and was within day peak expiratory flow variability, so it would be difficult to substantiate a diagnosis of cough-variant asthma. The practical importance of our findings is that a significant proportion of patients with corticosteroid-responsive cough have normal airway responsiveness. Therefore, we can not agree with the suggestion of McGarvey and colleagues (3) that a corticosteroid trial is not necessary in patients with normal airway responsiveness.

The incidence of gastroesophageal reflux as a primary cause of chronic cough was low in our series although it was a common codiagnosis in patients with more than one explanation for their cough. Our protocol ensured that 24-h esophageal pH monitoring was performed on all patients whose cough remained unexplained after a full clinical assessment, measurement of airway responsiveness and induced-sputum analysis, irrespective of the presence or absence of symptoms of gastroesophageal reflux; so we are confident that cases were not overlooked. The bronchoalveolar lavage and sputum eosinophilia found in patients with cough secondary to gastroesophageal reflux (8, 9) is interesting. However, it is substantially less than that seen in our patients with eosinophilic bronchitis and below the level where a response to corticosteroids would be expected (10), so we doubt that gastroesophageal reflux is an important explanation for the eosinophilic airway inflammation in our patients.

C. E. BRIGHTLING

J. D. PAVORD

Department of Respiratory Medicine

  and Thoracic Surgery

Glenfield Hospital

Leicester, United Kingdom

1. Brightling, C. E., R. Ward, K. L. Goh, A. J. Wardlaw, and I. D. Pavord. 1999. Eosinophilic bronchitis is an important cause of cough. Am. J. Respir. Crit. Care 160: 406-410 .

2. Fujimura, M., H. Ogawa, M. Yasui, and T. Matsuda. 2000. Eosinophilic tracheobronchitis and airway cough hypersensitivity in chronic non-productive cough. Clin. Exp. Allergy 30: 41-47 .

3. Brightling, C. E., R. Ward, A. J. Wardlaw, and I. D. Pavord. Airway inflammation, airway responsiveness and cough before and after inhaled budesonide in patients with eosinophilic bronchitis. Eur. Respir. J. (In press)

4. Gibson, P. G., K. Zlatic, J. Scott, W. Sewell, K. Woolley, and N. Saltos. 1998. Chronic cough resembles asthma with IL-5 and granulocyte macrophage colony stimulating factor gene expression in bronchoalveolar cells. J. Allergy Clin. Imunol. 101: 320-326 .

5. Pizzichini, E., M. M. M. Pizzichini, P. G. Gibson, K. Paramaswaran, G. J. Gleich, L. Bernan, J. Dolovich, and F. E. Hargreave. 1998. Sputum eosinophilia predicts benefit fom prednisone in smokers with chronic obstructive bronchitis. Am. J. Respir. Crit. Care Med. 158: 1511-1517 [Abstract/Free Full Text].

6. Brightling, C. E., G. Woltmann, A. J. Wardlaw, and I. D. Pavord. 1999. Development of irreversible airflow obstruction in a patient with eosinophilic bronchitis without asthma. Eur. Respir. J. 14: 1228-1230 [Abstract].

7. McGarvey, L. P. A., L. G. Heaney, J. T. Lawson, B. T. Johnston, C. M. Scally, M. Ennis, D. R. T. Shepherd, and J. MacMahon. 1998. Evaluation and outcome of patients with chronic non-productive cough using a comprehensive diagnostic protocol. Thorax 53: 738-743 .

8. McGarvey, L. P. A., P. Forsythe, L. G. Heaney, J. MacMahon, and M. Ennis. 1999. Broncholaveolar lavage findings in patients with chronic non productive cough. Eur. Respir. J. 13: 59-65 .

9. Micheletto, C., E. Burti, L. Mauroner, C. Pomari, P. Turco, R. Dal, and Negro. 1997. Induced sputum and serum inflammatory lmarkers in subjects with cough due to gastroesophageal reflux (abstract). Eur. Respir. J. 10: 317 .

10. Pavord, I. D., C. E. Brightling, G. Woltmann, and A. J. Wardlaw. 1999. Non-eosinophilic corticosteroid unresponsive asthma. Lancet 353: 2213-2214 [Medline].