Chronic Cough Due to Nonbronchiectatic Suppurative Airway Disease (Bronchiolitis) ACCP Evidence-Based Clinical Practice Guidelines

Chronic Cough Due to Nonbronchiectatic Suppurative Airway Disease (Bronchiolitis) ACCP Evidence-Based Clinical Practice Guidelines Kevin K. Brown, MD, FCCP Objectives: To review the role of nonbronchiectatic suppurative airway disease (bronchiolitis) in the spectrum of causes of cough and its management. Design/methodology: A MEDLINE search (through May 2004) for studies published in the English language since 1980 on human subjects using the medical subject heading terms cough, causes of cough, etiology of cough, interstitial lung disease, bronchiolitis, bronchiolitis obliterans, diffuse panbronchiolitis, and inflammatory bowel disease was performed. Case series and prospective descriptive clinical trials were selected for review. Any references from these studies that were pertinent to the topic were also obtained. Results/conclusions: In patients with cough in whom other more common causes of cough have been excluded, incomplete or irreversible airflow limitation, small airways disease seen on high-resolution CT scan, or purulent secretions seen on bronchoscopy, should suggest nonbronchiectatic suppurative airways disease (bronchiolitis) as a potential primary cause. Successful management depends on the identification of the specific underlying disorder. (CHEST 2006; 129:132S 137S) Key words: bronchiolitis; diffuse panbronchiolitis; inflammatory bowel disease; small airways disease Abbreviations: CXR chest radiograph; DPB diffuse panbronchiolitis; HRCT high-resolution CT; IBD inflammatory bowel disease; ILD interstitial lung disease; UC ulcerative colitis Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Kevin K. Brown, MD, FCCP, National Jewish Medical and Research Center, 1400 Jackson St, Denver, CO 80206; e-mail: brownk@njc.org 132S Cough may be the initial manifestation, or may develop during the clinical course, of nonbronchiectatic suppurative airway disease (bronchiolitis). While infrequently encountered, these small airways disorders are seen commonly enough by the practicing pulmonologist to warrant serious consideration in the correct clinical setting. A MEDLINE search (through May 2004) was conducted for studies published in the English language since 1980 on human subjects using the medical subject heading terms cough, causes of cough, etiology of cough, interstitial lung disease (ILD), bronchiolitis, bronchiolitis obliterans, diffuse panbronchiolitis (DPB), and inflammatory bowel disease (IBD) was performed. Case series and prospective descriptive clinical trials were selected for review. Any references from these studies that were pertinent to the topic were also obtained. Definition Abnormalities of the small airways and bronchiolar disorders are diseases that affect airways with an internal diameter of 2 mm and do not contain cartilage in their walls. 1 Like other ILDs, they may be affected by variable amounts of cellular inflammation (eg, lymphocytic, neutrophilic, eosinophilic, or granulomatous), fibrosis, and architectural distortion. If the bronchiolar abnormalities associated with asthma, COPD/emphysema, and bronchiectasis are Diagnosis and Management of Cough: ACCP Guidelines

removed, bronchiolitides can be grouped in a clinical classification scheme (Table 1). Patients with all of these conditions can present with or eventually develop cough as part of the clinical syndrome. 2 Epidemiology While few data are available on the frequency of the occurrence of these diseases, it is clear that they are not common. Generally grouped with the ILDs, they make up only a small fraction of the patients seen by a practicing pulmonologist. The prevalence of cough at presentation is unknown. Pathogenesis The pathogenesis of cough in patients with the bronchiolitides is unknown. However, the inflammation, fibrosis, and architectural distortion 3 of the small airways that occur both with and without mucous hypersecretion and bronchial hyperresponsiveness is almost assuredly part, if not the entire cause. Diagnosis As these diseases can be difficult to identify and may result from a variety of causes, and appropriate therapy often requires a specific diagnosis to be successful, a high index of suspicion and a complete evaluation are required. 2 Bronchiolitis should be considered in patients with cough and incompletely or nonreversible airflow limitation associated with a clinical syndrome suggestive of infection, an underlying systemic disorder known to be associated with small airways disease, or concerning exposures. Both the general and specific diagnosis of bronchiolitis requires a comprehensive medical history, physical examination, physiologic testing (spirometry with and without bronchodilator, lung volumes, and gas exchange), and radiographic studies (ie, chest radiograph [CXR] and high-resolution CT [HRCT] scans with expiratory cuts). Because of the anatomic location of disease, CXR findings can often be normal in the presence of bronchiolar disease. 4 The resolution of HRCT scanning is limited to airways that are 2 mm in diameter; therefore, normal bronchioles cannot be visualized. 5 Because of this technical limitation, clinically significant disease may be present in the absence of direct HRCT scan bronchiolar findings, and the absence of these features cannot rule out bronchiolar disease. Direct signs of abnormal bronchioles include dilation or airway wall thickening, which often appears with nodular branching of 2 to 4 mm and linear branching tree-in-bud abnormalities. 4 However, disorders in which the small airways are neither dilated nor inflamed, such as cryptogenic bronchiolitis obliterans, may show no direct radiographic abnormalities and may only be inferred by indirect features such as air-trapping (mosaic attenuation on expiration scans) or subsegmental atelectasis. 4 In the correct clinical setting (such as suggestive physiologic findings and an appropriate exposure or underlying systemic disease), HRCT findings of direct features (ie, small nodules and tree-in-bud pattern) and indirect features (ie, mosaic attenuation on expiratory scanning) of bronchiolar disease, particularly when larger airway abnormalities are also present, may obviate the need for further invasive testing. However, there remains an important role for surgical lung biopsy (to provide pathologic classification) and bronchoscopy (to rule out infection) in the absence of a characteristic clinical scenario. Specific Treatment Therapy for patients with the bronchiolitis should be tailored to the specific cause or diagnosis. Infectious bronchiolitis, while uncommon in adults and usually secondary to a viral etiology (eg, respiratory syncytial virus), can result from bacterial infection Table 1 Clinical Classification of Bronchiolitis Clinical Class Infections Inhalational mechanism Toxins Antigens Systemic diseases Drug reactions Allograft recipients Idiopathic disorders Specific Examples M pneumoniae, respiratory syncytial virus Respiratory bronchiolitis (tobacco smoke), sulfur dioxide Hypersensitivity pneumonitis Collagen vascular disease, inflammatory bowel disease, immunodeficiency Penicillamine, amiodarone Lung and bone marrow transplant Bronchiolitis obliterans (cryptogenic constrictive bronchiolitis), follicular bronchiolitis, DPB www.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 SUPPLEMENT 133S

(eg, Mycoplasma pneumoniae) and if identified may require prolonged antibiotic therapy with or without the addition of corticosteroids. 6 In a retrospective review 7 of patients who had been referred for the evaluation of chronic cough over 7 years, 15 patients were identified with unsuspected polymicrobial bacterial infection ( -streptococcus was common) of the large and small airways as the sole cause of cough. The disease was clinically unsuspected (some had normal pulmonary function test and CT scan findings), but all had purulent secretions seen on bronchoscopy. These patients failed to respond to oral antibiotics, and all required a minimum of 21 days of directed IV therapy. 7 Repeated infections associated with immunodeficiency states such as common variable immunodeficiency disease require immune reconstitution therapy (eg, IV Ig in patients with common variable immunodeficiency disease) as well as prolonged antibiotic therapy. 8 In those disorders in which ongoing exposure to a toxin (eg, cigarettes or sulfur dioxide), an antigen, or a medication/drug is contributing, cessation of the causative exposure is required with or without the use of corticosteroids. When the airways disease is secondary to a systemic inflammatory disease (eg, rheumatoid arthritis), immunomodulatory therapy for the primary disease may be helpful. Recommendations 1. In patients with cough and incomplete or irreversible airflow limitation, direct or indirect signs of small airways disease seen on HRCT scan, or purulent secretions seen on bronchoscopy, nonbronchiectatic suppurative airways disease (bronchiolitis) should be suspected as the primary cause. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 2. In patients with cough in whom more common causes have been excluded, because bacterial suppurative airways disease may be present and clinically unsuspected, bronchoscopy is required before excluding it as a cause. Level of evidence, low; benefit, substantial; grade of recommendation, B 3. In patients in whom bronchiolitis is suspected, a surgical lung biopsy should be performed when the combination of the clinical syndrome, physiology, and HRCT findings do not provide a confident diagnosis. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 4. In patients with infectious bacterial bronchiolitis, prolonged antibiotic therapy improves cough and is recommended. Level of evidence, low; benefit, substantial; grade of recommendation, B 5. In patients with toxic/antigenic exposure or drug-related bronchiolitis, cessation of the exposure or medication plus corticosteroid therapy for those with physiologic impairment is appropriate. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A Definition IBD The lung may be affected in the patient with an IBD (ie, ulcerative colitis [UC] or Crohn disease). Direct involvement by the underlying disease, pulmonary toxicity secondary to a medication, or infection may occur. The underlying disease may involve the entire airway, from larynx to the alveolus repeating the abnormalities found in the bowel including inflammatory (lymphocytic, neutrophilic, and granulomatous), fibrotic, and destructive changes. 9 Epidemiology The prevalence of UC and Crohn disease may each be upward of 100 cases per 100,000 population, and the incidence of extraintestinal manifestations of disease may range up to 40%. The lung is the least commonly affected extraintestinal system, and cough with variable amounts of sputum production has been suggested to be the most common clinical scenario seen by pulmonologists. 9 Pathogenesis The pathogenesis of cough in IBD is unknown. However, it can be assumed that the prominent airway inflammation that is present in patients with these diseases plays a pivotal role. Airways disease with necrosis, bronchiolitis obliterans, granulomatous and lymphocytic inflammation, and pathologic DPB have all been described. 10 Considerable amounts of mucopurulent sputum can be generated, although the cough may occasionally be dry. Both physiologic airflow limitation and bronchial hyperresponsiveness appear to be common 11 and may worsen with the activity of the underlying bowel disease. 12 Diagnosis The diagnosis of lung disease in the patient with IBD requires a complete evaluation including a comprehensive medical history, physical examination, physiologic studies (ie, spirometry, lung volumes, and gas exchange), and radiographic studies 134S Diagnosis and Management of Cough: ACCP Guidelines

(ie, CXR and HRCT scan). 13 As lung involvement secondary to IBD, infection, and drug reaction is always in the differential diagnosis, there remains an important role for bronchoscopy with lavage and biopsy as well as surgical lung biopsy in selected cases. A confident diagnosis of the cause of the respiratory symptoms is necessary before specific therapy can be instituted. Specific Treatment No randomized trial of therapy has been performed in the airways disease associated with IBD. Removal of the affected bowel has not been shown to be beneficial and may lead to an increase in symptoms. 13 In observational studies, therapy with corticosteroids, both oral and inhaled, has been associated with improvement in cough, particularly when the large airways (bronchitis) are involved. 9 With involvement of the smaller airways, the responses to corticosteroids appear to depend on the underlying pathology, with granulomatous and lymphocytic disease being more responsive than bronchiolitis obliterans or necrotizing bronchiolitis. 10 Recommendations 6. In the IBD patient with cough, bronchiolitis should be suspected as a potential cause. Level of evidence, low; benefit, substantial; grade of recommendation, B 7. In patients in whom IBD-related bronchiolitis is suspected, both adverse drug reaction and infection should be specifically considered. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 8. In patients with IBD, therapy with both oral and inhaled corticosteroids may improve cough, and a trial of therapy is suggested. Level of evidence, low; benefit, substantial; grade of recommendation, B Definition DPB DPB is a distinct form of small airways disease found primarily, but not solely, in Japan, Korea, and China. 14 It is associated with the presence of chronic sinus disease, cough with copious purulent sputum, wheezing, and dyspnea. Obstructive physiology, suggestive radiographic features, and characteristic histopathologic findings 15 on surgical lung biopsy specimens are all seen. Epidemiology The disease was initially described and appears to be most common in Japan, Korea, and China, and while genetics appear to be an important factor, 16 environmental factors likely play some role, as the disease appears to be uncommon in persons outside of this geographic region. Pathogenesis The pathogenesis of cough in patients with DPB is unknown. However, it can be assumed that both the prominent large and small airway inflammation and the frequent presence of associated bacterial infection (particularly Pseudomonas) play a critical role. 14 Diagnosis The diagnosis requires a complete evaluation including comprehensive medical history, physical examination, physiologic studies (ie, spirometry, lung volumes, and gas exchange), and radiographic studies (ie, CXR and HRCT scan). Physiology reveals a marked obstructive defect with or without superimposed restriction. Findings on the HRCT scan include symmetric bilateral basalpredominant ill-defined centrilobular nodules ( 5 mm), some of which are connected to distal branching structures (tree-in-bud pattern), thickwalled and ectatic bronchioles (bronchiolectasis), and mosaic air-trapping. 17 When the expected physiologic and radiographic pattern is seen in the appropriate clinical setting, a surgical lung biopsy may not be necessary, and a trial of therapy may be considered. While the findings of the radiographic studies can be suggestive, there remains an important role for bronchoscopy with lavage (for infection) as well as surgical lung biopsy to confirm the diagnosis in selected cases. The pathologic pattern is characteristic and consists of a respiratory bronchiolocentric cellular infiltrate of lymphocytes, plasma cells, and a striking number of foamy macrophages that may expand the contiguous peribronchiolar alveolar septae. 15 Organizing plugs of exudate similar to that seen in organizing pneumonia as well as intraluminal neutrophils are regularly seen. Essentially identical pathologic features have been seen in patients with rheumatoid arthritis, UC, and adult T-cell lymphoma/leukemia. As bacterial infection is often a complicating feature of the disease, it should be specifically sought. Specific Treatment The untreated natural history of DPB includes the development of bronchiectasis, progressive respirawww.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 SUPPLEMENT 135S

tory failure, and death. Up to 50% of untreated patients will die within 5 years, and only 25% will live for 10 years. 18 Therapy with macrolide antibiotics, particularly low-dose erythromycin (eg, erythromycin, 200 to 600 mg/d for 2 to 6 months), has been shown to improve outcome, symptoms, and biomarkers of disease activity. A retrospective study found a significant survival advantage, 19 and another study 20 showed improvements in productive cough, dyspnea, and CXR abnormalities with 2 months of therapy. The mechanism of action of erythromycin (and other macrolides [specifically 14 member-ringed compounds such as clarithromycin and roxithromycin]) appears to be antiinflammatory 21,22 rather than antiinfective, as serum and sputum drug levels achieved with doses of erythromycin up to 600 mg daily are below the minimum inhibitory concentrations of the common superinfecting organisms, Haemophilus influenzae and Pseudomonas aeruginosa. 20 And, while control of infections is important (the presence of Pseudomonas is associated with a worse prognosis), therapy for the infections alone does not improve outcome. 23 Nonsteroidal antiinflammatory drugs may assist in the control of excessive mucus production, 24 while inhaled 2 -agonists or ipratropium bromide may lead to bronchodilation and improved mucociliary clearance. Importantly, while frequently used, there are no data to support the use of corticosteroids. Recommendations 9. In patients with chronic cough who have recently lived in Japan, Korea, or China, DPB should be considered as a potential cause. Level of evidence, low; benefit, substantial; grade of recommendation, B 10. In patients with suspected DPB, an appropriate clinical setting and characteristic HRCT scan findings may obviate the need for invasive testing and a trial of macrolide therapy (erythromycin or other 14-member ring macrolides such as clarithromycin and roxithromycin) is appropriate. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 11. In patients with DPB, prolonged treatment (> 2 to 6 months) with erythromycin (or other 14-member ring macrolides such as clarithromycin and roxithromycin) is recommended. Level of evidence, low; benefit, substantial; grade of recommendation, B Summary of Recommendations 1. In patients with cough and incomplete or irreversible airflow limitation, direct or indirect signs of small airways disease seen on HRCT scan, or purulent secretions seen on bronchoscopy, nonbronchiectatic suppurative airways disease (bronchiolitis) should be suspected as the primary cause. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 2. In patients with cough in whom more common causes have been excluded, because bacterial suppurative airways disease may be present and clinically unsuspected, bronchoscopy is required before excluding it as a cause. Level of evidence, low; benefit, substantial; grade of recommendation, B 3. In patients in whom bronchiolitis is suspected, a surgical lung biopsy should be performed when the combination of the clinical syndrome, physiology, and HRCT findings do not provide a confident diagnosis. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 4. In patients with infectious bacterial bronchiolitis, prolonged antibiotic therapy improves cough and is recommended. Level of evidence, low; benefit, substantial; grade of recommendation, B 5. In patients with toxic/antigenic exposure or drug-related bronchiolitis, cessation of the exposure or medication plus corticosteroid therapy for those with physiologic impairment is appropriate. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 6. In the IBD patient with cough, bronchiolitis should be suspected as a potential cause. Level of evidence, low; benefit, substantial; grade of recommendation, B 7. In patients in whom IBD-related bronchiolitis is suspected, both adverse drug reaction and infection should be specifically considered. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 8. In patients with IBD, therapy with both oral and inhaled corticosteroids may improve cough, and a trial of therapy is suggested. Level of evidence, low; benefit, substantial; grade of recommendation, B 9. In patients with chronic cough who have recently lived in Japan, Korea, or China, DPB should be considered as a po- 136S Diagnosis and Management of Cough: ACCP Guidelines

tential cause. Level of evidence, low; benefit, substantial; grade of recommendation, B 10. In patients with suspected DPB, an appropriate clinical setting and characteristic HRCT scan findings may obviate the need for invasive testing and a trial of macrolide therapy (erythromycin or other 14-member ring macrolides such as clarithromycin and roxithromycin) is appropriate. Level of evidence, expert opinion; benefit, substantial; grade of recommendation, E/A 11. In patients with DPB, prolonged treatment (> 2 to 6 months) with erythromycin (or other 14-member ring macrolides such as clarithromycin and roxithromycin) is recommended. Level of evidence, low; benefit, substantial; grade of recommendation, B References 1 Colby TV. Bronchiolitis. Pathologic considerations. Am J Clin Pathol 1998; 109:101 109 2 Ryu JH, Myers JL, Swensen SJ. Bronchiolar disorders. Am J Respir Crit Care Med 2003; 168:1277 1292 3 Myers JL, Colby TV. Pathologic manifestations of bronchiolitis, constrictive bronchiolitis, cryptogenic organizing pneumonia, and diffuse panbronchiolitis. Clin Chest Med 1993; 14:611 622 4 Lynch D, Brown KK, Lee JS, et al. Imaging of diffuse infiltrative lung disease. Hamilton, ON, Canada: BC Decker, 2000 5 Franquet T, Stern EJ. Bronchiolar inflammatory diseases: high-resolution CT findings with histologic correlation. Eur Radiol 1999; 9:1290 1303 6 Penn CC, Liu C. Bronchiolitis following infection in adults and children. Clin Chest Med 1993; 14:645 654 7 Schaefer OP, Irwin RS. Unsuspected bacterial suppurative disease of the airways presenting as chronic cough. Am J Med 2003; 114:602 606 8 Cunningham-Rundles C, Bodian C. Common variable immunodeficiency: clinical and immunological features of 248 patients. Clin Immunol 1999; 92:34 48 9 Camus P, Colby TV. The lung in inflammatory bowel disease. Eur Respir J 2000; 15:5 10 10 Camus P, Piard F, Ashcroft T, et al. The lung in inflammatory bowel disease. Medicine (Baltimore) 1993; 72:151 183 11 Louis E, Louis R, Drion V, et al. Increased frequency of bronchial hyperresponsiveness in patients with inflammatory bowel disease. Allergy 1995; 50:729 733 12 Herrlinger KR, Noftz MK, Dalhoff K, et al. Alterations in pulmonary function in inflammatory bowel disease are frequent and persist during remission. Am J Gastroenterol 2002; 97:377 381 13 Storch I, Sachar D, Katz S. Pulmonary manifestations of inflammatory bowel disease. Inflamm Bowel Dis 2003; 9:104 115 14 Sugiyama Y. Diffuse panbronchiolitis. Clin Chest Med 1993; 14:765 772 15 Iwata M, Colby TV, Kitaichi M. Diffuse panbronchiolitis: diagnosis and distinction from various pulmonary diseases with centrilobular interstitial foam cell accumulations. Hum Pathol 1994; 25:357 363 16 Park MH, Kim YW, Yoon HI, et al. Association of HLA class I antigens with diffuse panbronchiolitis in Korean patients. Am J Respir Crit Care Med 1999; 159:526 529 17 Nishimura K, Kitaichi M, Izumi T, et al. Diffuse panbronchiolitis: correlation of high-resolution CT and pathologic findings. Radiology 1992; 184:779 785 18 Izumi T. A nation-wide survey of diffuse panbronchiolitis and the high incidence of diffuse panbronchiolitis seen in Japanese respiratory clinics. In: Sarcoidosis and other granulomatous disorders. New York, NY: Elsevier Science Publishers, 1988 19 Kudoh S, Azuma A, Yamamoto M, et al. Improvement of survival in patients with diffuse panbronchiolitis treated with low-dose erythromycin. Am J Respir Crit Care Med 1998; 157:1829 1832 20 Nagai H, Shishido H, Yoneda R, et al. Long-term low-dose administration of erythromycin to patients with diffuse panbronchiolitis. Respiration 1991; 58:145 149 21 Takizawa H, Desaki M, Ohtoshi T, et al. Erythromycin modulates IL-8 expression in normal and inflamed human bronchial epithelial cells. Am J Respir Crit Care Med 1997; 156:266 271 22 Shimizu T, Shimizu S, Hattori R, et al. In vivo and in vitro effects of macrolide antibiotics on mucus secretion in airway epithelial cells. Am J Respir Crit Care Med 2003; 168:581 587 23 Izumi T. Diffuse panbronchiolitis. Chest 1991; 100:596 597 24 Tamaoki J, Chiyotani A, Kobayashi K, et al. Effect of indomethacin on bronchorrhea in patients with chronic bronchitis, diffuse panbronchiolitis, or bronchiectasis. Am Rev Respir Dis 1992; 145:548 552 www.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 SUPPLEMENT 137S