Early Thickening of the Reticular Basement Membrane in Children with Difficult Asthma

Departments of Paediatrics and Lung Pathology, Imperial College at the Royal Brompton Hospital, London, United Kingdom; Department of Pulmonary Medicine, University Hospital, Umeå, Sweden; and Departments of Pulmonary/Critical Care, Baylor College of Medicine, Houston, Texas

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Remodeling of the airway wall occurs in adults with asthma, and reticular basement membrane (RBM) thickening is pathognomonic of the asthma process. To investigate whether RBM thickening is present in children with difficult asthma and comparable to that seen in adults with asthma, we used light microscopy to measure RBM thickness in plastic-embedded endobronchial biopsy sections from 19 children with difficult asthma who were prescribed 1,600 µg/day or more of inhaled steroids (age range, 6–16 years), 10 children without asthma (7–16 years), and three adult groups: 8 healthy control subjects (21–42 years), 10 mild steroid-naive subjects with asthma (18–41 years), and 6 adults (3 steroid naive and 3 on inhaled steroids) intubated after a life-threatening attack of asthma (20–64 years). RBM thickness in the children with asthma was similar to that in adults with either mild or life-threatening asthma (median 8.2 [range 5.4–11.1] versus 8.1 [5.8–10.0] and 7.2 [2.8–10.0] µm, respectively) and greater than either adult or pediatric control subjects (8.2 [5.4–11.1] versus 4.4 [3.2–6.3] µm, p < 0.01, and 4.9 [3.7–8.3] µm, p < 0.01). We conclude that RBM thickening is already present in children with difficult asthma and to a similar extent to that seen in adults with asthma. In addition, we find no association with age, symptom duration, lung function, or concurrent eosinophilic airway inflammation.

　

Key Words: airway remodeling • endobronchial biopsy • pediatrics

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Changes in airway wall structure, referred to as airway wall remodeling, are a characteristic feature of asthma in adults (1). Thickening of the epithelial reticular basement membrane (RBM) is pathognomonic of the adult asthma phenotype. In hematoxylin and eosin–stained tissue sections, the thickened layer has a hyaline appearance on light microscopy and, ultrastructurally, appears to consist of a plexiform arrangement of reticulin seen as fibrils in an amorphous matrix. The thickened reticular layer is associated with deposition of immunoglobulins and/or collagens I and III, tenascin, and fibronectin (2–4). In contrast, the basal lamina, occasionally referred to as the "true" basement membrane, is of normal thickness in the adult with asthma.

RBM thickening has been described in children with asthma. Cutz and colleagues demonstrated RBM thickening in four children using tissue obtained postmortem (n = 2) or from open-lung biopsies (n = 2) performed for other reasons (5). More recently Cokugras and colleagues described qualitatively the presence of RBM thickening in children with moderately severe asthma (6). However, there are no objective measurements of RBM thickness in children published and no data on the relationship of RBM thickness with age or duration of symptoms in children. It is not known whether the RBM is thickened to the same degree in both children and adults with asthma.

It is also unclear whether the RBM gradually thickens with time as a consequence of repeated episodes of allergen exposure and bouts of acute inflammation. An alternative is that the RBM thickens maximally early in the natural course of the disease or after the establishment of chronic inflammation. Until recently, chronic airway inflammation has been considered to be the primary abnormality in asthma and remodeling a secondary consequence (1, 7). As a result, antiinflammatory treatment with inhaled steroids is recommended to control the inflammation. It is generally thought that such treatment may also prevent the development or persistence of airway wall remodeling. At present, it is not known when RBM thickening begins, and to our knowledge, there is no evidence that treatment of any sort can prevent the development of RBM thickening. However, there is evidence that long-term treatment with inhaled steroids may, at least partially, reduce the thickness of the RBM in adults with asthma (8, 9). The suppressive effect of inhaled steroids on eosinophilic airway inflammation can be demonstrated much earlier than any change in RBM thickness (9). Further long-term studies are indicated to investigate accurately the timing of RBM thickening and its relationship to airway inflammation. It is possible that undetermined factors drive inflammation and remodeling as separate processes (10–12) and that the processes are not equally steroid sensitive.

The aims of this cross-sectional study were to investigate whether RBM thickening is already present in children with difficult asthma; to examine the relationship with age, duration of symptoms, FEV₁, and markers of airway inflammation; and to investigate whether the RBM is thickened to a similar degree in both children and adults with asthma. We also aimed to obtain objective data on RBM thickness in children with asthma to help design future longitudinal studies. We used light microscopy and plastic-embedded sections to examine endobronchial biopsies from a group of children with difficult asthma who were undergoing bronchoscopy and biopsy as part of their clinical assessment (13). The data were compared with those from biopsies obtained from a group of children without asthma undergoing flexible bronchoscopy for other clinical indications, and with biopsies from three groups of adults: normal healthy nonsmoking control subjects, mild nonsmoking patients with asthma, and adults intubated for a life-threatening attack of asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Subjects
Subject demographics are summarized in  . There were 19 children (aged 6–16 years) with difficult asthma, defined as previously described (13, 14), who underwent flexible bronchoscopy with endobronchial biopsy after a 2-week course of oral prednisolone (40 mg/day). Spirometry was performed before and after short-acting bronchodilator (salbutamol 1 mg via a spacer) the day before bronchoscopy (15). Exhaled nitric oxide (FE_NO) was measured before and after the course of prednisolone (16). Endobronchial biopsy was also performed in 10 children without asthma undergoing bronchoscopy for recent or recurrent hemoptysis (n = 4), chronic cough that was unresponsive to bronchodilators, and prolonged treatment with inhaled corticosteroids (n = 3), tracheomalacia (n = 2), and persistent left upper lobe collapse (n = 1).

fig.ommitted TABLE 1. Clinical characteristics

 
The adult subjects included 10 steroid-naive patients with asthma and 8 healthy nonsmoking control subjects. Their details have been described previously (17). Endobronchial biopsy was also performed in six adults, with a diagnosis of asthma and less than 15 pack-years of tobacco use, intubated after a life-threatening attack of asthma.

Ethics Approval
Performance of endobronchial biopsy for studying airway inflammation and remodeling was approved by the local ethics committees. Informed consent was obtained from subjects (adults) or their parents (children)/next of kin (intubated patients). The study was performed according to the Declaration of Helsinki.

Bronchoscopy
Bronchoscopy with endobronchial biopsy was performed in children under general anesthesia and in adults with mild asthma and adult control subjects under sedation, as previously described (13, 17). Adults with life-threatening ("severe") asthma underwent fibreoptic bronchoscopy (Olympus type BF P10 or BF P20D; Olympus Corporation, Melville, NY) under deep sedation with continuous intravenous propofol with or without neuromuscular blockade (vecuronium 8–10 mg). The bronchoscope was passed through an endotracheal tube (size 7 mm or larger), and the airway was inspected and suctioned clear of any secretions. During subsequent passes, the bronchoscope was preloaded with single-patient use alligator biopsy forceps (Microvasive Radial Jaw, 1.8-mm external diameter; Boston Scientific Corporation, Natick, MA) and endobronchial biopsies taken from third-, fourth-, and fifth-order bronchial divisions from either the left side or the right side.

Processing of Biopsies
For measuring RBM thickness, biopsies were fixed immediately in 2.5 or 3% glutaraldehyde in 0.05-M cacodylate buffer (pH 7.4), postfixed in 1% osmium tetroxide (in the same buffer), and dehydrated and embedded in epoxy resin (Araldite). Sections (1 µm thick) were cut and stained with alkaline toluidine blue. For assessment of mucosal eosinophilic inflammation in children with difficult asthma, biopsies were processed and stained with an antibody to major basic protein as previously described (14).

Quantification
Toluidine blue sections were selected for measurement of RBM thickness providing that the epithelium, RBM, and submucosa could be easily identified, and the length of RBM was adequate to allow multiple measurements (i.e., approximately 1 mm). Sections were coded and examined by light microscopy in random order by the same observer, who was unaware of the origin of the sections. Using previously validated criteria, RBM thickness was assessed using computer-aided image analysis (NIH Image 1.33; National Institutes of Health, Bethesda, MD) by making 40 measurements at 20-µm intervals (18), and expressed as the geometric mean of the 40 measurements (18). Intraobserver repeatability was assessed by measuring the same section four times, expressed as the percentage of coefficient of variation of the four measurements. The within-biopsy variability (the percentage of coefficient of variation) was assessed by measuring four different sections of a single biopsy. The between-biopsy variability (the percentage of coefficient of variation) for a single subject was assessed by measuring a single section from each of three biopsies from the same subject. The within-biopsy and between-biopsy variabilities were compared using the F test (19). Major basic protein immunostaining was assessed by point counting as previously described (14).

Statistical Analysis
Nonparametric tests were applied to test for intergroup differences. A comparison between all groups was made using the Kruskal-Wallis test, followed by a Mann-Whitney U-test, if a significant difference (p < 0.05) was found (19). The Bonferroni correction was made for multiple comparisons (19). Associations were looked for by Spearman rank correlation. Power calculations were performed to determine the number of subjects needed for future pediatric studies designed to detect differences in RBM thickness (20).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Bronchoscopy
The safety and tolerability of bronchoscopy for the adults with mild asthma and control subjects have been described elsewhere (17). For the intubated adult subjects, complications of the procedure were a mild increase in bronchospasm, as evidenced by increase in peak inspiratory pressure, airway resistance, and auto positive end-expiratory pressure in all subjects. An average of 1.5 doses of nebulized albuterol was administered after the procedure. Bronchoscopy and endobronchial biopsy did not result in any other changes in management and did not result in a prolonged duration of intubation and ventilation compared with other patients ventilated for life-threatening asthma who did not undergo bronchoscopy and endobronchial biopsy. There were no instances of pneumothorax, hypotension, arrhythmia, or fatality. The mean time between intubation and bronchoscopy was 15.6 hours. The number of intubated patients reported here is small. The safety of this procedure in a larger group of intubated subjects will be the subject of a separate article that is currently in preparation. We have previously described the safety of bronchoscopy in an overlapping group of children, which included some of the subjects in this study (13). Of the pediatric subjects reported here, both subjects with and without asthma, bronchoscopy was tolerated well. There were no significant adverse events, in particular instances of pneumothorax or serious bleeding. However, two children with asthma experienced an increase in symptoms for several days after the bronchoscopy.

Microscopy
The biopsies processed for measurement of RBM thickness were generally of good quality in all groups and sufficient to allow multiple measurements of RBM thickness in all subjects ( and ) . Of the sections originally selected for analysis, only three (one section from each of three different subjects) were rejected and replaced by suitable sections from another biopsy. The size of the pediatric biopsies varied between 0.8–1.6 mm in maximum diameter. A variety of mucosal structures could be identified, including surface epithelium, RBM, submucosal glands and vessels, and bronchial smooth muscle. Areas of epithelial loss were seen in subjects with and without asthma. In two pediatric subjects with asthma, areas of squamous metaplasia were present. Areas that included smooth muscle were present in biopsies from 15 out of 19 children with asthma and 9 out of 10 children without asthma. Of the biopsies assessed for mucosal eosinophilic inflammation, biopsies from 13 out of 19 subjects were suitable for analysis (14).

fig.ommitted Figure 1. Endobronchial biopsy section from a child with difficult asthma stained with toluidine blue (low power).

 

fig.ommitted Figure 2. Higher power view of an endobronchial biopsy section from a child with difficult asthma demonstrating darker staining of the RBM.

 
Variability
The percentage of coefficient of variation of RBM thickness for four repeat measurements on one and the same section was 2.4%. Within a single biopsy, the between-section percentage of coefficient of variation for four different sections was 1.6%, and the between-biopsy variability for each of three subjects was 5.2, 8.5, and 9.6%, represented by an average of 7.8%. The between-biopsy variability was significantly greater than that within (p < 0.05).

RBM Thickness
The results of measurements of RBM thickness in the five groups are shown in  . RBM thickness in the children with asthma was similar to that seen in the adults with either mild or severe asthma (median 8.2 [range 5.4–11.1] versus 8.1 [5.8–10.0] and 7.2 [2.8–10.0] µm, respectively). The values in the pediatric patients with asthma were significantly greater than both the adult and pediatric control subjects, which were similar to each other (8.2 [5.4–11.1] versus 4.4 [3.2–6.3] µm, p < 0.01, and 4.9 [3.7–8.3] µm, p < 0.01, respectively). In the children with difficult asthma, there was no significant correlation between RBM thickness and age (r = -0.004, p = 0.98), duration of asthma (r = -0.04, p = 0.87), or both prebronchodilator and postbronchodilator FEV₁ after a 2-week course of prednisolone (r = 0.18, p = 0.47, and r = 0.2, p = 0.4, respectively). Analyzing all of the subjects with asthma as one group (n = 35), there was no significant correlation between RBM thickness and either age (r = 0.06, p = 0.68) or duration of asthma (r = 0.07, p = 0.68).

fig.ommitted Figure 3. RBM thickness in adult and pediatric subgroups. *p < 0.01; **p < 0.05.

 
Airway Inflammation
The median (range) values for FE_NO in the pediatric group with asthma before and after prednisolone were 16.9 (1.2–33.4) and 8.1 (1.3–24.5) parts per billion, respectively. The upper limit of normal in our department is 12.5 parts per billion (16). For mucosal eosinophilic inflammation, the median (range) eosinophil score in the group with asthma was 6.4% (0.6–51.5%). We have previously described a median (range) for children without asthma of 0.8% (0.1–5.4%) (14). There was no significant correlation between RBM thickness and FE_NO before or after prednisolone (r = 0.39, p = 0.11 and r = 0.43, p = 0.08, respectively)  or between RBM thickness and mucosal eosinophilic inflammation after prednisolone (r = 0.44, p = 0.13).

fig.ommitted Figure 4. Relationship between FENO after prednisolone and RBM thickness in children with difficult asthma (ppb = parts per billion).

 
Power Calculations
The data for RBM thickness in the group of children with difficult asthma approximated to a normal distribution, with a standard deviation of 1.5 µm. The sample size required per group to detect a difference in RBM thickness of 2 µm at the two-sided 5% significance level with 90% power would be 14 (20). A sample size of 11 per group would detect a difference of 2.25 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
This study has demonstrated that RBM thickening is present in a group of school-age children with difficult asthma and that the thickness is similar to that seen in adults with either mild intermittent or acute life-threatening asthma. Within the group of children with difficult asthma, there is no correlation between RBM thickness and age, duration of symptoms, FEV₁, FE_NO, or mucosal eosinophilic inflammation. Analyzing all of the subjects with asthma as one group, there is no association between RBM thickness and either age or duration of asthma.

The results of this study confirm the findings of Cutz and colleagues and Cokugras and colleagues that RBM thickening is a feature of asthma in children (5, 6). However, this is the first study to provide objective measurements of RBM thickness in children with asthma and to compare these with data from adults. It is also the first study to provide reference data for children without asthma. In the group of children with difficult asthma, the lack of an association between RBM thickness and asthma duration is in agreement with previous studies involving adults with asthma that were unable to demonstrate any such relationship (21, 22). With regard to asthma severity, Chu and colleagues did not find any association between RBM thickness and the clinical severity of asthma in adults (23). This study involved three groups of subjects with asthma of differing severity (steroid-naive adults, adults with life-threatening asthma, and children with persistent symptoms despite high-dose inhaled steroids). The comparable degree of RBM thickening among the three groups is in keeping with the findings of Chu and colleagues (23). No significant correlation was demonstrated in this study between RBM thickness and FEV₁ in the pediatric subjects. An association between RBM thickness and FEV₁ in adult asthma has been described by some investigators (24) but not by others (23). This suggests that RBM thickness is only one of several factors that contribute to airways obstruction in asthma.

The lack of an association between RBM thickness and asthma duration or severity and the similar degree of RBM thickening in all three groups with asthma, raises the possibility that RBM thickening, once established, varies very little within an individual. However, without follow-up biopsies, it cannot be determined whether RBM thickness is maximal for any individual. No significant correlation was found between RBM thickness and two different markers of airway inflammation in the children. However, all of the children had been prescribed long-term, high-dose inhaled steroids, as well as a 2-week course of prednisolone just before the bronchoscopy. Ward and colleagues have demonstrated that the effects of treatment with inhaled steroids have a different time course on airway eosinophils (in bronchoalveolar lavage) and on RBM thickness (9). The cross-sectional design of this study therefore does not allow accurate conclusions to be drawn about the precise relationship between RBM thickness and airway inflammation. Further longitudinal studies involving infants and young children, with biopsies performed at a number of different time points, are needed. However, the design of such studies is extremely challenging and requires careful consideration of the practical and ethical implications of performing repeat endobronchial biopsies in young children. This study provides preliminary data that should prove useful for the design of future studies.

We acknowledge that this study has a number of potential weaknesses. First, the children with asthma studied were not typical of the majority of children with asthma but represent those with the most severe asthma, with persistent symptoms despite maximal conventional therapy. Unlike adult studies, bronchoscopy cannot be performed in children purely for research purposes. The opportunity for obtaining endobronchial biopsies from children with asthma is currently limited to those subjects undergoing bronchoscopy for clinical indications (13, 25). The children in this study underwent bronchoscopy and endobronchial biopsy as part of the clinical evaluation of their disease (13), and the opportunity was taken to obtain tissue for the purposes of research, after ethics committee approval and written informed consent from the parents (26, 27). A second point is that the control group without asthma was made up of children having a bronchoscopy for other clinical indications. They cannot therefore be regarded as healthy control subjects. Although most of the control subjects clearly did not have asthma, three were being investigated for persistent cough without wheeze. These had no evidence of reversibility to bronchodilator assessed by spirometry, and there was no improvement in symptoms after previous treatment with either short-acting bronchodilator or inhaled corticosteroids. Cough-variant asthma has been described, with evidence of RBM thickening (28), but it is unlikely that this diagnosis was present in any of the "control" subjects reported here. With appropriate ethical approval, future studies to obtain samples from children with less severe asthma and healthy control subjects are likely to involve subjects who are intubated and undergoing a general anesthetic for another elective procedure. Krawiec and colleagues have used this approach to obtain reference data for bronchoalveolar lavage in young children (29).

RBM thickening has been described in atopic adults without asthma (30). Only one of the children with asthma in this study was not atopic, and interestingly, this subject had the thinnest RBM of all the pediatric patients with asthma (5.41 µm). The most appropriate pediatric control group would therefore consist of atopic children without asthma, although it might be difficult to recruit a sufficient number given the difficulty in obtaining pediatric control data of any sort. In this study, the atopic status of most (9 out of 10) of the pediatric control subjects (and all the severe adults with asthma) was assessed clinically (a personal history of eczema or allergic rhinitis). Radioallergosorbent tests to house dust mite, cat, dog, and grasses were performed in one pediatric subject without asthma with chronic cough, and these were all negative. We acknowledge that the lack of any objective tests of atopic status in the remaining nine control subjects is a weakness of the study, as we cannot exclude the possibility that RBM thickening in the children with asthma is a feature of atopy rather than asthma. RBM thickening has also been described in a subgroup of adults with chronic obstructive pulmonary disease (31). However, the thickening appears to be associated with features of asthma (airway eosinophils, response to corticosteroids) in this subgroup rather than the effects of smoking.

To examine the relationship between RBM thickness and both age and asthma duration across a wide time span, data from all of the subjects with asthma (n = 35) were pooled. However, the subjects in the three groups differed considerably in the level of treatment prescribed. They were also recruited at different times and at different centers. They cannot therefore be regarded as a single group, and no firm conclusions can justifiably be drawn from analysis of the pooled data. However, analysis of the data has the potential to raise questions and generate hypotheses. Another difference between the groups studied is the site from which the biopsies were taken. However, we have previously demonstrated that RBM thickness does not vary between different airway generations (17), and the different biopsy sites used in this study are therefore unlikely to have a significant effect on the interpretation of the data.

In conclusion, these findings demonstrate that RBM thickening is present in children with difficult asthma, similar to that seen in adult asthma. There is no association with age, duration of symptoms, FEV₁, FE_NO, or concurrent eosinophilic inflammation. Longitudinal studies in younger children or infants are needed to determine when the RBM begins to thicken and whether this is dependent on the prior establishment of chronic inflammation (32). If such studies were to demonstrate the presence of early and maximal RBM thickening in preschool children with a clinical diagnosis of asthma, this would lend considerable support to the hypothesis that RBM thickening is a fundamental component of the airway pathology of asthma: it may even precede the development of chronic inflammation.

	Acknowledgments

 
The authors are grateful to Tim Oates for assistance with processing and staining the pediatric biopsies for major basic protein. This article was presented at the 2001 meeting of the American Thoracic Society, San Francisco, CA.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 

1. Bousquet J, Jeffery PK, Busse WW, Johnson M, Vignola AM. Asthma: from bronchoconstriction to airways inflammation and remodelling. Am J Respir Crit Care Med 2000;161:1720–1745.

2. Laitinen A, Altraja A, Kampe M, Linden M, Virtanen I, Laitinen LA. Tenascin is increased in airway basement membrane of asthmatics and decreased by an inhaled steroid. Am J Respir Crit Care Med 1997;156:951–958.

3. Roche WR, Beasley R, Williams JH, Holgate ST. Subepithelial fibrosis in the bronchi of asthmatics. Lancet 1989;1:520–524.

4. Jeffery PK, Wardlaw AJ, Nelson FC, Collins JV, Kay AB. Bronchial biopsies in asthma: an ultrastructural, quantitative study and correlation with hyperreactivity. Am Rev Respir Dis 1989;140:1745–1753.

5. Cutz E, Levison H, Cooper DM. Ultrastructure of airways in children with asthma. Histopathology 1978;2:407–421.

6. Cokugras H, Akcakaya N, Seckin, Camcioglu Y, Sarimurat N, Aksoy F. Ultrastructural examination of bronchial biopsy specimens from children with moderate asthma. Thorax 2001;56:25–29.

7. Kips JC, Pauwels RA. Airway wall remodelling: does it occur and what does it mean? Clin Exp Allergy 1999;29:1457–1466.

8. Sont JK, Willems LN, Bel EH, van Krieken JH, Vandenbroucke JP, Sterk PJ. Clinical control and histopathologic outcome of asthma when using airway hyperresponsiveness as an additional guide to long-term treatment: the AMPUL Study Group. Am J Respir Crit Care Med 1999;159:1043–1051.

9. Ward C, Pais M, Bish R, Reid D, Feltis B, Johns D, Walters EH. Airway inflammation, basement membrane thickening and bronchial hyperresponsiveness in asthma. Thorax 2002;57:309–316.

10. Jeffery PK. Remodeling in asthma and chronic obstructive lung disease. Am J Respir Crit Care Med 2001;164:S28–S38.

11. Warner JO. Unmet needs in the treatment of asthmatic children and adolescents: 1. Clin Exp Allergy 2000;30:70–72.

12. Holgate ST. Inflammation and remodelling in asthma and allied disorders: summary and conclusions. Clin Exp Allergy Rev 2001;1:9–13.

13. Payne DNR, McKenzie SA, Stacey S, Misra D, Haxby E, Bush A. Safety and ethics of bronchoscopy and endobronchial biopsy in difficult asthma. Arch Dis Child 2001;84:422–425.

14. Payne DNR, Adcock IM, Wilson NM, Oates T, Scallan M, Bush A. Relationship between exhaled nitric oxide and mucosal eosinophilic inflammation in children with difficult asthma, after treatment with oral prednisolone. Am J Respir Crit Care Med 2001;64:1376–1381.

15. American Thoracic Society. Standardization of spirometry, 1994 update. Am J Respir Crit Care Med 1995;152:1107–1136.

16. Payne DNR, Wilson NM, James A, Hablas H, Agrafioti C, Bush A. Evidence for different subgroups of difficult asthma in children. Thorax 2001;56:345–350.

17. Jeffery PK, Godfrey RW, Ädelroth E, Nelson F, Rogers A, Johansson S-A. Effects of treatment on airway inflammation and thickening of basement membrane reticular collagen in asthma. Am Rev Respir Dis 1992;145:890–899.

18. Sullivan P, Stephens D, Ansari T, Costello J, Jeffery PK. Variation in the measurements of basement membrane thickness and inflammatory cell number in bronchial biopsies. Eur Respir J 1998;12:811–815.

19. Altman DG. Practical statistics for medical research. London: Chapman & Hall; 1991.

20. Campbell MJ, Julious SA, Altman DG. Estimating sample sizes for binary, ordered categorical, and continuous outcomes in two group comparisons. BMJ 1995;311:1145–1148.

21. Hoshino M, Nakamura Y, Sim JJ. Expression of growth factors and remodelling of the airway wall in bronchial asthma. Thorax 1998;53:21–27.

22. Bai TR, Cooper J, Koelmeyer T, Pare PD, Weir TD. The effect of age and duration of disease on airway structure in fatal asthma. Am J Respir Crit Care Med 2000;162:663–669.

23. Chu HW, Halliday JL, Martin RJ, Leung DY, Szefler SJ, Wenzel SE. Collagen deposition in large airways may not differentiate severe asthma from milder forms of the disease. Am J Respir Crit Care Med 1998;158:1936–1944.

24. Kasahara K, Shiba K, Ozawa T, Okuda K, Adachi M. Correlation between the bronchial subepithelial layer and whole airway wall thickness in patients with asthma. Thorax 2002;57:242–246.

25. Fregonese L, Sacco O, Fregonese B, Verna A, Mereu C, Gambini C, Rossi GA. Difficult-to-treat asthma-like symptoms in a 12-yr-old atopic female. Eur Respir J 2000;15:1128–1131.

26. McIntosh N, Bates P, Brykczynska G, Dunstan G, Goldman A, Harvey D, Larcher V, McCrae D, McKinnon A, Patton M, et al. Guidelines for the ethical conduct of medical research involving children: Royal College of Paediatrics: Child Health: Ethics Advisory Committee. Arch Dis Child 2000;82:177–182.

27. Bush A. Guidelines for the ethical conduct of medical research involving children . Arch Dis Child 2000;83:370.

28. Niimi A, Matsumoto H, Minakuchi M, Kitaichi M, Amitani R. Airway remodelling in cough-variant asthma. Lancet 2000;356:564–565.

29. Krawiec ME, Westcott JY, Chu HW, Balzar S, Trudeau JB, Schwartz LB, Wenzel SE. Persistent wheezing in very young children is associated with lower respiratory inflammation. Am J Respir Crit Care Med 2001;163:1338–1343.

30. Milanese M, Crimi E, Scordamaglia A, Riccio A, Pellegrino R, Canonica GW, Brusasco V. On the functional consequences of bronchial basement membrane thickening. J Appl Physiol 2001;91:1035–1040.

31. Chanez P, Vignola AM, O'Shaugnessy T, Enander I, Li D, Jeffery PK, Bousquet J. Corticosteroid reversibility in COPD is related to features of asthma. Am J Respir Crit Care Med 1997;155:1529–1534.

32. Pohunek P, Roche WR, Turzikora J, Kudrmann J, Warner JO. Eosinophilic inflammation in the bronchial mucosa of children with asthma . Eur Respir J 1997;10:160s.