Airway Inflammation as an Assessment of Chronic Nonproductive Cough* Sang Yeub Lee, MD ; Jae Youn Cho, MD; Jae Jeong Shim, MD; Han Kyeom Kim, MD; Kyung Ho Kang, MD, FCCP; Se Hwa Yoo, MD; and Kwang Ho In, MD Objectives: The possibility that a bronchial inflammatory process could be involved with a chronic nonproductive cough without other potential causes such as postnasal drip syndrome, bronchial asthma, gastroesophageal reflux, chronic bronchitis, bronchiectasis, or the use of angiotensinconverting enzyme inhibitors has not been clearly described. We investigated the possibility that a chronic nonproductive cough without other potential causes is associated with airway inflammation, and if this is so, what the relationship might be between this inflammation and the possible etiology of the cough. Subjects: Twenty-five adults with chronic nonproductive cough as an isolated symptom over a 3-week period, and 5 healthy control subjects were studied. Measurements and results: Clinical assessments, cough scores, methacholine challenges, allergy skin prick tests, and bronchoscopies for bronchial biopsies were performed. In the bronchial biopsies, the patients were divided into the following two subgroups: 21 patients who were infiltrated with eosinophils vs the healthy control group (median, 12.0 vs 0.0 cells/mm 2, respectively; p < 0.01); and 4 patients who were infiltrated with lymphocytes vs the healthy control group (median, 84.5 vs 22.0 cells/mm 2, respectively; p < 0.01). With the methacholine challenge test, 5 of the 21 eosinophil-infiltrated patients received diagnoses of cough-variant asthma, and the other 16 patients received diagnoses of eosinophilic bronchitis. In the lymphocyte-infiltrated group, all four patients received diagnoses of lymphocytic bronchitis. Conclusions: These results suggest that a chronic nonproductive cough as an isolated symptom is associated with airway inflammation due to eosinophil and lymphocyte infiltration. The causes of the chronic nonproductive cough were eosinophilic bronchitis, cough-variant asthma, and lymphocytic bronchitis. (CHEST 2001; 120:1114 1120) Key words: airway inflammation; cough; cough-variant asthma; eosinophilic bronchitis; lymphocytic bronchitis Abbreviations: PC 20 provocative concentration of methacholine causing a 20% fall in FEV 1 ; PEF peak expiratory flow Chronic cough is defined as a cough persisting for at least 3 weeks. 1 Its pathophysiology is still poorly understood. Systemic evaluation of the potential causes of chronic cough based on the locations of afferent limbs of the cough reflex has shown that *From the Pulmonary Division (Dr. Lee), the Department of Internal Medicine, College of Medicine, Konkuk University, Chungju, Korea; Korea University (Drs. Cho, Shim, Kang, Yoo, and In), Seoul, Korea; and the Department of Pathology (Dr. Kim), College of Medicine, Korea University, Seoul, Korea. Dr. Lee is currently at the Department of Internal Medicine, Pulmonary Division, Pundang CHA General Hospital, College of Medicine, Pochon CHA University, Kyonggi-do, Korea. This study was supported by Brain Korea 21. Manuscript received October 17, 2000; revision accepted April 6, 2001. Correspondence to: Kwang Ho In, MD, The Pulmonary Division, The Department of Internal Medicine, Korea University Anam Hospital 126 1, 5Ka, Anam-Dong, Sungbuk-Ku, Seoul, Korea, 136 075; e-mail: khin@ns.kumc.or.kr postnasal drip syndrome, gastroesophageal reflux, and asthma are the main etiologies of persistent cough. The so-called cough-variant asthma is also a well-known cause of cough, 2 and nonasthmatic eosinophilic bronchitis, a less common etiology, also has been described. 3,4 Asthma is associated with airway inflammation. Boulet and colleagues 5 reported that the chronic cough observed in nonasthmatic subjects, which was caused by postnasal drip syndrome and gastroesophageal reflux, is also associated with airway inflammation. However, the possibility that a bronchial inflammatory process may be involved in a chronic nonproductive cough without the other potential causes of postnasal drip syndrome, bronchial asthma, gastroesophageal reflux, chronic bronchitis, bronchiectasis, or the use of angiotensin-converting enzyme inhibitors has not been clearly described. 1114 Clinical Investigations
We investigated the possibility that a chronic nonproductive cough without other potential causes is associated with airway inflammation, and if this is so, what the relationship might be between this inflammation and the possible etiology of the cough. Patients Materials and Methods Twenty-five patients, 10 men and 15 women who were 20 to 64 years of age (mean, 39.8 year), who were referred to the Korea University Hospital for persistent dry cough of more than 3 weeks duration were studied. The inclusion criteria included a normal pulmonary function test result, a normal chest examination and radiograph, and nonsmoking status. Exclusion criteria included postnasal drip syndrome, gastroesophageal reflux, the use of a medication known to induce chronic cough (eg, angiotensin-converting inhibitor or -blockers), evidence of a respiratory tract infection in the previous 8 weeks, or a history of asthma, chronic bronchitis, or any other chest or systemic diseases. Postnasal drip syndrome was considered as an etiology if the patients experienced the sensation of having something drip down into their throat and/or had to clear their throat frequently, or if a physical examination of the pharynx revealed secretions or a cobblestone mucosa, with or without signs of chronic sinusitis on the sinus radiograph, as suggested by a thickening of the sinus mucosa. Gastroesophageal reflux was considered if the 24-h measurement of the esophageal ph was positive. Coughers were compared with a group of healthy control subjects (three men and two women; mean age, 57.2 years). The control subjects were nonsmokers and had normal chest radiograph results, normal airway reactivity, and normal results of pulmonary function testing. They had no current respiratory symptoms, were nonatopic, and had had no respiratory tract infection within the previous month. No control subjects were receiving any medications. The study protocol was approved by the ethics committee of Korea University College of Medicine, and written informed consent was obtained from all patients and control subjects. Procedures All patients and control subjects were given a thorough evaluation, which included the subjective measurement of cough severity, methacholine challenge, allergy skin prick tests, and fiberoptic bronchoscopy for bronchial biopsies, in order to evaluate airway inflammation. The severity of the cough was rated using the cough score (Table 1). 6 The methacholine challenge was performed using the following method. Acetyl methacholine chloride (Sigma; St. Louis, Table 1 Cough Verbal Category Descriptive Scores Score Symptom Description 0 No cough 1 Cough for one or two short periods only 2 Cough for 2 short periods 3 Frequent coughing but does not interfere with activities 4 Frequent coughing that interferes with activities 5 Cannot perform most usual activities due to severe coughing MO) was diluted with normal saline solution and was divided into nine concentrated solutions ranging from 0.075 to 25 mg/ml. Each concentrated methacholine solution was inhaled five times through a nebulizer (DeVilbiss Pulmo-aide Compressor/Nebulizer; SM Instruments; Doylestown, PA) and a dosimeter (Micro- Dosimeter; Sunrise Medical; Carlsbad, CA). FEV 1 was measured after each inhalation, and the provocative concentration of methacholine resulting in a 20% fall in FEV 1 from the baseline measurement (PC 20 ) was subsequently calculated from a logresponse curve using linear interpolation. A positive airway hyperresponsiveness was defined when the PC 20 was 8 mg/ ml. Allergy skin prick tests were performed with a solution (Bencard; Brentford, UK) of 55 common allergen extracts. The sizes of the wheal and the area of erythema were measured with a skin test reaction gauge (Bencard) 15 min after the skin prick. A negative result on the test indicated that there was no reaction induced by an allergen but that there was a normal reaction to the controlled histamine solution. A score of 1 was given to areas of erythema of 21 mm, and a score of 2 was given for areas of erythema of 21 mm, with the ratio of the wheal size of 1 compared to that induced by the histamine-controlled solution. The ratios between 1 and 2 were given a score of 3. Ratios between 2 and 3 were given a 4. Scores of 2 were designated as atopic. A bronchoscopic biopsy was performed under local anesthesia with 2% lidocaine spray using flexible bronchoscopy (model BF 1T30; Olympus; Tokyo, Japan). The oxygen saturation and pulse rate were checked by pulse oximetry during the procedure. From the carina of the second and third segment of the right bronchus, three or four pieces of tissue were obtained, were fixed with 10% formalin, were embedded on paraffin, were sliced at a 6- m thickness, were stained with hematoxylin-eosin, and were examined under a light microscope. Assessment of Airway Inflammation Inflammatory cell counts and basement membrane thickness were measured from the histologic sections. To avoid observer bias, all microscopic slides were coded prior to analysis by one pathologist and were analyzed blindly, in terms of the patients clinical data. Inflammatory cell counts were calculated manually from the light microscopy using a graticule (eyepiece micrometer; Olympus) at 200 magnification. The cell counts were measured as the mean number of cells per square millimeter (ie, 1 mm 2 ) measured at three different sections. Differential counts of lymphocytes, basophils, eosinophils, plasma cells, and so forth were performed. Only the surface of the connective tissue was considered, excluding both smooth muscle cells and mucus glands. Cell differentiation was based on the type of nuclei and granules inside the cells. The cells were counted only if the nuclei were observed. The thickness of the basement membrane was assessed on three sections stained with hematoxylin-eosin and was measured from the base of the bronchial epithelium to the outer limit of the reticular lamina of the basement membrane using a micrometer at 400 magnification. The final result was the mean of all measurements obtained for each biopsy specimen. Statistical Analysis All statistical analyses were performed using statistical software (SAS, version 6.12; SAS Institute; Cary, NC). The results were expressed as the mean values of FEV 1, FVC, peak expiratory flow (PEF), and the thickness of the basement membranes, and as the median and range for cell counts from biopsy specimens. The Wilcoxon two-sample test was used to compare the mean values CHEST / 120 / 4/ OCTOBER, 2001 1115
Table 2 Clinical Characteristics of Patients* Characteristics Patients (n 25) Mean age, yr 39.8 12.0 Sex Male 10 Female 15 Cough score 3 12 4 9 5 4 Symptom duration, mo 2 6 FEV 1, % predicted 104 15.8 FVC, % predicted 88.2 12.0 PEF, % predicted 102.3 17.7 Provocation With 5 Without 20 Atopy With 5 Without 20 *Values given as mean SD, unless otherwise indicated. Most frequent duration. Methacholine challenge. between groups of patients. Cell differential counts in bronchial biopsy specimens were compared using the Kruskal-Wallis test. Results Patient Characteristics The patient pool consisted of 10 men and 15 women, ranging in age from 20 to 64 years (Table 2). The mean ( SD) baseline daily cough scores were 3.68 0.75. The duration of the cough symptoms of the patients ranged from 3 weeks to 10 years, with 2 to 6 months being considered the most prevalent duration (10 patients; 40%) (Table 3). The mean values for FEV 1, FVC, and PEF (% predicted) of the patients were 104 15.8, 88.2 12.0, and 102.3 17.7, respectively. Among the 25 patients, 5 had positive results for the methacholine challenge (PC 20, 8 mg/ml). The PC 20 values for the five patients were 0.94, 1.53, Figure 1. The eosinophil-infiltrated group. The bronchial mucosa shows eosinophil infiltration predominantly in the subepithelial connective tissue. Epithelial denudation and basement membrane thickening are also noted (hematoxylin-eosin, original 400). 1.77, 3.67, and 5.88 mg/ml. Five patients had positive results to the allergy skin prick tests. Assessment of Airway Inflammation The patients were divided into an eosinophilinfiltrated group and a lymphocyte-infiltrated group (Fig 1, 2). The eosinophil-infiltrated group was considered to have 3% occupancy of all cells, 4 irrespective of other inflammatory cell counts. The lymphocyte-infiltrated group was considered to have more than twice the number of lymphocytes of the control group (Fig 3) or to have 60% occupancy without increasing the numbers of other inflammatory cells. The eosinophil-infiltrated group consisted of 21 patients with a median eosinophil count of 12.0 cells/mm 2, while the median eosinophil count of the Table 3 Number of Patients According to Duration of Cough Duration No. % 1mo 2 8 1 2 mo 2 8 2 3 mo 4 16 3 6 mo 6 24 6mo 1 yr 3 12 1 5 yr 2 8 5 10 yr 4 16 10 yr 2 8 Figure 2. The lymphocyte-infiltrated group. The bronchial mucosa shows lymphocyte infiltration predominantly in the subepithelial connective tissue and basement membrane thickening (hematoxylin-eosin, original 400). 1116 Clinical Investigations
patients and 2 of 5 cough-variant asthma patients had atopy. The lymphocytic bronchitis patients did not have atopy. Discussion Figure 3. The healthy control group. The bronchial mucosa shows no definite inflammatory cell infiltration and basement membrane thickening (hematoxylin-eosin, original 400). control group was 0.0 cells/mm 2 (p 0.01). The lymphocyte-infiltrated group consisted of four patients whose median eosinophil count was 84.5 cells/ mm 2, in contrast with the 22.0 cells/mm 2 of the control group (p 0.01) (Table 4). Significantly, the mean thickness of the basement membrane of 25 patients was 14.20 5.19 m, which is thicker than that of the control group (mean, 3.50 1.37 m) (p 0.01). Etiologic Diagnosis The causes of chronic nonproductive cough in the patients were inferred from the methacholine challenge results and the pathologic findings. Five of 21 patients of the eosinophil-infiltrated group could receive diagnoses of cough-variant asthma due to positive results to the methacholine challenge. However, 16 patients with negative results to the methacholine challenge could receive diagnoses of eosinophilic bronchitis. All four patients in the lymphocyte-infiltrated group were shown to have negative results for the methacholine challenge and manifested lymphocytic bronchitis (Fig 4). Three of 16 eosinophilic bronchitis Two studies 4,5 indicate that chronic cough is associated with airway inflammation. It is not clear how the airway inflammation affects the cough reflex and airway responsiveness, yet the hypothesis that epithelial cell damage could lead to an increased cough reflex is highly supported. 5 Denudation of the sensory nerve endings by epithelial damage due to the inflammation may result in an increased cough reflex to exogenous or endogenous factors. Epithelial damage along with inflammatory cell infiltration have been observed in patients with chronic cough. This is supported by studies 7 9 showing that steroids assist in healing the damaged epithelial cells by suppressing the infiltration of inflammatory cells, and that they also participate in regenerating the destroyed cilia, which results in a reduction of coughing. Therefore, the pathogenesis of chronic cough can be summarized as the damage of epithelial cells due to airway inflammation. In this study, an increase in inflammatory cell infiltration and a thickening of the basement membrane in the patient group were found, in contrast with the control group. Mostly, the inflammatory cells were either eosinophils or lymphocytes. Typically, no significant differences were observed between the patient and control groups in the number of inflammatory cells, with the exceptions of the number of eosinophils and lymphocytes. Therefore, we evaluated the patient group according to the number of eosinophils and lymphocytes. First, we defined the patients with increased eosinophils, as compared with the number of eosinophils in the healthy control group, to be an eosinophil-infiltrated group. Then, we defined the patients with increased lymphocytes, as compared with the number of lymphocytes in the healthy control group, to be a Table 4 Cell Counts per Square Millimeter of Surface in Bronchial Biopsy Specimens in Each Group According to Dominant Cell Infiltration* Cells Eosinophil-Infiltrated (n 21) Lymphocyte-Infiltrated (n 4) Control Subjects (n 5) Monocytes/macrophages 13.0 (8.0 25.5) 5.0 (4.0 8.25) 8.0 (3.0 8.5) Lymphocytes 242.0 (165.5 810.5) 84.5 (55.0 113.2) 22.0 (11.5 38.5) Eosinophils 12.0 (5.5 174.5) 1.0 (0.25 1.0) 0.0 (0.0 1.0) Neutrophils 39.0 (19.5 52.0) 17.5 (11.0 42.0) 27.0 (21.5 29.0) Plasmacytes 6.0 (0.0 14.0) 1.0 (0.0 2.0) 0.0 (0.0 1.0) *Values given as median (range). Statistically significant (p 0.01). CHEST / 120 / 4/ OCTOBER, 2001 1117
Figure 4. Flow chart of the study analysis. lymphocyte-infiltrated group. On the basis of inflammatory cell infiltration of the bronchial mucosa and the results of the methacholine challenge, we were able to consider the causes of chronic nonproductive cough in the patients as cough-variant asthma, eosinophilic bronchitis, or bronchitis with lymphocyte infiltration. For the investigation of bronchitis with lymphocyte infiltration in this study, we think that a term such as lymphocytic bronchitis perhaps would be a useful label during further investigation of this type of airway inflammation. Our data suggest that the assessment of airway inflammation is an important addition for the investigation of chronic nonproductive cough. 4 We chose to assess airway inflammation with bronchoscopic biopsies instead of sputum induction, since it is difficult to obtain successful results with the sputum induction method in patients with dry cough. Although bronchoscopic biopsies are invasive, we believe that this method is more precise for the assessment of airway inflammation. Cough-variant asthma was first named by Glauser 10 in 1972 and is defined as a condition of nonsmoking patients who have coughed for at least 3 weeks without other symptoms but who have normal results of physical examinations, normal or almost normal spirometry values, and positive results of bronchial provocation testing. 2,10 13 It was reported that cough-variant asthma could reach up to 29% of all patients with chronic cough and up to 7 to 11% of all asthma patients. 13,14 However, many cases were overlooked without being diagnosed. Therefore, patients with a nonspecific chronic cough, such as a nocturnal dry cough or aggravated cough after exercise and exposure to cold air, and with an atopic allergy history, either family or individual, must be suspected. 2,11,13,14 In this study, 40% (two of five patients) of cough-variant asthma patients showed atopy, which was not significantly different from the 25 to 35% seen in cases of classic asthma. Eosinophilic bronchitis was suggested by Gibson et al 3 in 1989 and was defined as an eosinophilic airway inflammation with chronic cough as an isolated symptom with normal spirometry values and normal methacholine airway responsiveness. It is a new disease that does not have the diurnal variability of PEF or airway responsiveness to histamine, adenosine 5 -monophosphate, and methacholine. No specific association between airway inflammation and airway hyperresponsiveness was inferred in cases of eosinophilic bronchitis. 15 Its pathogenesis is not yet clear. However, it was suggested that airway inflammation by eosinophils and mast cells might be associated with the cough and heightened airway responsiveness but were not sufficient for the development of the wheeze, airway hyperresponsiveness, or airway obstruction that is characteristic of asthma. 15 We measured eosinophilic airway inflammation in patients with eosinophilic bronchitis and coughvariant asthma. There tend to be more eosinophils in 1118 Clinical Investigations
patients with cough-variant asthma than in those with eosinophilic bronchitis. However, due to the small number of patients in the study, any statistically significant conclusions could not be made. Further investigations on the relationship between the eosinophilic bronchitis and cough-variant asthma are needed. To our knowledge, there have not been any large studies performed regarding the association with atopy, and nothing has yet been proven. This study demonstrated the relevance of atopy to be 18.8% (ie, 3 of 16 patients). Compared with previous studies, the results of this study suggested that eosinophilic bronchitis may be the causative disease of a chronic nonproductive cough 15 and for the first time provided histopathologic evidence to that end. Furthermore, in this study, 16 patients demonstrated signs of eosinophilic bronchitis, which is three times the rate in coughvariant asthma patients (5 patients). These results indicate that there are more cases of undiagnosed eosinophilic bronchitis because cough-variant asthma has been linked to chronic cough in only 29% of the reported cases. Therefore, more interest and effort are crucial in diagnosing unrecognized eosinophilic bronchitis. Patients with lymphocytic bronchitis were identified by their morphologic features. The patients with this disease may have other specific diseases that the authors could not diagnose, or it may truly be a whole new disease. However, there was no evidence of other possible diseases to account for the finding. Also, Boulet and colleagues 5 reported an increase in lymphocytic infiltration in the bronchial tissues of patients with chronic nonproductive cough without any definitive causes. This issue should be investigated in future studies. Until then, we think that a term such as lymphocytic bronchitis would be a useful label. The specific cause of lymphocytic bronchitis was not identified but was considered to be a nonspecific inflammatory reaction. In the lungs, lymphocytes, along with macrophages, play an important role in the inflammatory and immune reactions. The macrophage, as a phagocytic cell, is not antigen-specific. The lymphocyte represents a significant refinement in the inflammatory response (ie, antigen specificity and discrimination of the self and other). In other words, lymphocytes produce a defense mechanism against the invasion of antigens and microorganisms by antibody production, cytotoxic activity, or cytokine production. 16,17 These lymphocytes are activated in particular when foreign antigens overwhelm mucociliary escalators, macrophages, and neutrophil defenses. 18 The patients with lymphocytic bronchitis did not have any specific history of exposure or any occupational history. But since the lung is stimulated by a large number of foreign infectious and noninfectious antigens on the tide of airflow, it would be very difficult to actually find specific causative antigens. The clinical features of lymphocytic bronchitis such as severity of cough and duration of symptoms did not differ from those of eosinophilic bronchitis. However, the lymphocytic bronchitis patients did not have atopy. Oral prednisolone, 30 mg once a day for 2 weeks, was prescribed to each patient with lymphocytic bronchitis. The cough score was measured before and after the treatment. Three of four patients had improved symptoms (change in cough scores: 3 to 0; 4 to 1; and 3 to 1), but one patient had no improvement (change in cough score, 4 to 3). The effects of steroid therapy need further study such as with a placebo-controlled study. As mentioned above, chronic cough is the result of epithelial damage due to inflammatory cell infiltration. However, the role of lymphocytes in epithelial damage is controversial 19 and requires further investigation. Conclusion Chronic nonproductive cough as an isolated symptom without other potential causes, such as postnasal drip syndrome, bronchial asthma, gastroesophageal reflux, chronic bronchitis, bronchiectasis, or the use of angiotensin-converting enzyme inhibitors, is associated with airway inflammation by the infiltration of inflammatory cells that are composed mainly of eosinophils and lymphocytes. The causes are eosinophilic bronchitis, cough-variant asthma, and lymphocytic bronchitis (in order of frequency). Airway inflammation has a useful role in the assessment of chronic nonproductive cough. References 1 Irwin RS, Curley FJ, French CL. Chronic cough: the spectrum and frequency of causes, key components of the diagnostic evaluation and outcome of specific therapy. Am Rev Respir Dis 1990; 141:640 647 2 Corrao WM, Braman SS, Irwin RS. Chronic cough as the sole presenting manifestation of bronchial asthma. N Engl J Med 1979; 300:633 637 3 Gibson PG, Dolovich J, Denburg J, et al. Chronic cough: eosinophilic bronchitis without asthma. Lancet 1989; 1:1346 1347 4 Brightling CE, Ward R, Goh KL, et al. Eosinophilic bronchitis is an important cause of chronic cough. Am J Respir Crit Care Med 1999; 160:406 410 5 Boulet LP, Milot J, Boutet M, et al. Airway inflammation in non-asthmatic subjects with chronic cough. Am J Respir Crit Care Med 1994; 149:482 489 CHEST / 120 / 4/ OCTOBER, 2001 1119
6 Chang AB, Newman RG, Carlin JB, et al. Subjective scoring of cough in children: parent-completed vs child-completed diary card vs an objective method. Eur Respir J 1998; 11:462 466 7 Lundgren R, Soderberg M, Horstedt P, et al. Morphological studies of bronchial mucosal biopsies from asthmatics before and after ten years of treatment with inhaled steroids. Eur Respir J 1988; 1:883 889 8 Jeffery PK, Godfrey RW, Adelroth E, et al. Effects of treatment on airway inflammation and thickening of basement membrane reticular collagen in asthma. Am Rev Respir Dis 1992; 145:890 899 9 Heino M, Karjalainen J, Ylikoski J, et al. Bronchial ciliogenesis and oral steroid treatment in patients with asthma. Br J Dis Chest 1988; 82:175 178 10 Glauser FL. Variant asthma. Ann Allergy 1972; 30:457 459 11 Hanaway PJ, Hopper DK. Cough variant asthma in children. JAMA 1982; 247:206 208 12 Johnson D, Osborn LM. Cough variant asthma: a review of clinical literature. J Asthma 1991; 28:85 90 13 O connell EJ, Rojas AR, Sachs MI. Cough-type asthma: a review. Ann Allergy 1991; 66:278 285 14 Konig P. Cough variant asthma. J Asthma 1991; 28:83 84 15 Gibson PG, Hargreave FE, Girgis-Gabardo A, et al. Chronic cough with eosinophilic bronchitis: examination for variable airway obstruction and response to corticosteroid. Clin Exp Allergy 1994; 25:127 132 16 BAL Cooperative Steering Committee. Bronchoalveolar lavage constituents in healthy individuals, idiopathic fibrosis, and selected comparison groups. Am Rev Respir Dis 1990; 141:S169 S201 17 Pabst R, Binns RM. Lymphocytes migrate from the bronchoalveolar space to regional bronchial lymph nodes. Am J Respir Crit Care Med 1995; 151:495 499 18 Berman JS, Center DM. Lymphocyte-and macrophage-mediated inflammation in the lung. In: Fishman AP, ed. Fishman s pulmonary diseases and disorders. 3rd ed. New York, NY: McGraw-Hill, 1998; 275 287 19 Ohashi Y, Motojima S, Fukuda T, et al. Airway hyperresponsiveness, increased intracellular spaces of bronchial epithelium, and increased infiltration of eosinophils and lymphocytes in bronchial mucosa in asthma. Am Rev Respir Dis 1992; 145:1469 1476 1120 Clinical Investigations