INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

On the basis of histological analyses in individuals with fatal asthma, airway wall area has been reported to be increased from 50% to 300% compared with nonasthmatic control subjects and in individuals with nonfatal asthma increases from 10% to 100% have been documented (1). These findings suggest that the increases are in part related to disease severity. In nonfatal disease, thickening is most marked in small membranous and mid-sized airways. In contrast, in fatal cases wall thickening has been reported to be more generalized throughout the bronchial tree (3). Although it has been postulated that the airway wall is thickened by the remodeling process associated with chronic inflammation as the duration of asthma increases, there is little direct evidence for this concept. However, physiological data showing decreased distensibility and irreversible airflow obstruction with time support this hypothesis (6).

The present study is based on the hypothesis that if duration of disease contributes to airway wall remodeling, the airways of older individuals with asthma with fatal disease would be thicker than those of younger individuals who died of asthma. A corollary of this hypothesis is that factors in addition to remodeling and thickening, such as lumenal obstruction and airway smooth muscle shortening, contribute to the pathogenesis of a fatal attack. Conversely, if geometric factors are paramount in the causation of a fatal attack, no differences between groups should be evident.

Access to a large repository of lung tissue from cases of fatal asthma collected at one coroner's morgue enabled us to randomly select cases from younger and older age groups. Information concerning the cause of death and background history, including reviews by expert panels and nurse interviews of relatives, was available (10, 11).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All cases documented as fatal asthma in the storage records of the Coroner's unit, Department of Pathology, University of Auckland Medical School, from 1978 to 1986 were evaluated. Two hundred thirty cases of fatal asthma occurred in individuals who were less than 50 yr of age at death. Histology reports, and information from family, physicians, nurse interviewers, and police interviews, were evaluated. Two hundred one individuals were considered by the reviewer (TRB) to represent cases of fatal asthma with no other concomitant disorders likely to contribute to the fatal attack or to affect pulmonary pathology. To exclude ethnic influences, only white subjects (n = 101) were considered and all cases between 19 (males) or 17 (females) and 23 yr (young, n = 15) and between 40 and 49 yr (old, n = 15) were selected. Two of the old group with fatal asthma and one of the young group with fatal asthma had to be omitted because the available sections did not contain suitable airways. The median age of onset of symptomatic asthma in the cohort from which these subjects were drawn was 10.5 yr of age (12). The individual data for each case are incomplete, as some records of interviews with relatives of the deceased from the 1980s are no longer locatable. The available data show an average duration of symptomatic asthma of 7.6 ± 4.8 yr in the young group and 24.0 ± 13.4 yr in the old group. To define normal airway dimensions at these age ranges, tissue blocks from five young and six older subjects who suffered sudden nonpulmonary death (motor vehicle accidents) were obtained from a lung registry.

Airway Wall Morphometry

Wall dimensions. Sections (4 µm) from paraffin-embedded formalin-fixed blocks taken from the hilar region were selected from each case. The lungs had not been inflated prior to fixation. Sections were double stained with Gomori-trichrome and aldehyde fuschin, which allows accurate definition of connective tissue borders and smooth muscle. The morphometric analysis was restricted to airways with a basement membrane perimeter (Pbm) between 2 and 10 mm, as these were the predominant size range of airways in the block selected. Wall areas of 1 to 11 airways per subject were measured by a blinded observer using standard morphometric techniques as previously described (13). Airway wall measurements included the long and short diameters of the airway lumen as well as perimeters at the basement membrane (Pbm), at the outer border of smooth muscle, and at the outer border of adventitia. All dimensions were measured using the highest magnification that allowed each airway component to be visualized in one field, ranging from ×25 to ×250. Airways with a short/long diameter ratio < 0.6 were excluded. The corresponding areas subtended by the perimeters were calculated. The inner wall area (WA_i) was calculated as the difference between the areas defined by the outer border of smooth muscle and the basement membrane, the outer wall area (WA_o) was calculated as the difference between the areas of the outer border of adventitia and the outer border of smooth muscle, and total wall area (WA_t) was calculated as the difference between areas of the outer border of adventitia and the basement membrane.

Subepithelial collagen. The thickness of the subepithelial collagen layer (SEc) was measured in all airways of individuals with asthma with Pbm 2-10 mm (n = 55) and in a random sample of control airways with Pbm 2-10 mm (n = 13). Airways from one subject with asthma were omitted because the subepithelial collagen was too diffuse to accurately identify and quantify. Each airway was sampled at two points 180° apart at ×500 magnification. The total length and area of subepithelial collagen in each field were measured using a Nikon microscope with a camera lucida, a digitizing pad, and the Bioquant BQ system software. SEc thickness per airway was calculated as the mean of the two area/ length quotients (14). The mean standard deviation between the two thickness measurements was 2.2 µm for airways of individuals with asthma. SEc thickness was also evaluated in a sample of 17 airways of individuals using both our technique and a previously described technique (15) that involved calculating the mean of multiple (10) point to point (perpendicular to the basement membrane) thickness measurements using the same system as above. SEc measured with both methods were not found to be significantly different.

Airway smooth muscle area fractions. The proportion of smooth muscle in the airway wall cross sections was quantified by projecting a computer-generated grid of points onto the microscope field. The area fraction of smooth muscle (Asm) was calculated as the number of points falling on muscle/total number of points falling on the total wall area excluding the epithelium and basement membrane ×100. The highest magnification that included both lumen and the area outside the airway wall in a single field was used. Typically this was ×250 or ×500 magnification. In addition, the area fraction of muscle cells per muscle bundle (Asm/bundle) was also measured at ×1,250 magnification. The area fraction of muscle cells was calculated as the number of points falling on muscle cells/total number of points falling on muscles bundles ×100. The latter area included the connective tissue matrix between cells. The absolute amount of smooth muscle cell area in airway wall cross sections was calculated by multiplying the two area fractions thus obtained by the total wall area ([Asm/bundle] × [Asm] × WA_t).

Airways narrowing and lumenal content. The degree of airway narrowing was calculated as the ratio of the measured lumenal area (Abm) and the calculated lumenal area for a relaxed airway (A*bm) (13), expressed as a percentage. The smaller the ratio, the more constricted the airway. Residual lumenal content after fixation was also quantified on a 0-4 scale where 0 = no mucus, exudate, or debris, 1 = < 25% obstruction, 2 = 25-50% obstruction, 3 = 50-75% obstruction, and 4 = > 75% obstruction.

Statistical Analysis

The distribution of airway sizes between groups was examined by the Kolmogorov-Smirnov technique. A random effects regression analysis was used as previously described (13) to compare wall areas of airways as a function of Pbm between subject groups. Parametric or nonparametric (Wilcoxon) t tests were used for other analyses. p Values less than 0.05 were considered significant. All values are expressed as mean plus or minus standard error.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Table 1 indicates demographics of subjects in each group, airway sizes, and numbers of airways examined. The distribution of airway sizes was not different among the four groups. An example of Gomori-trichrome, aldehyde-fuschin staining in a subject with fatal asthma is shown in Figure 1.

View larger version (175K): [in this window] [in a new window]   Figure 1.   Fatal asthmaan example of a small airway stained with Gomori-trichrome, aldehyde-fuschin. Scale bar is 10 µm.

There was no significant difference in the relationship between absolute wall areas and Pbm comparing young and old control subjects (Figure 2 and Tables 2 and 3). In contrast, airways in individuals with fatal asthma showed a significant increase in the slope of this relationship, with approximately a threefold increase in average wall area in the older individuals with fatal asthma compared with the younger individuals (Figure 3 and Tables 2 and 3). This increase was primarily due to differences in adventitial (WA_o) area between the two asthma groups (p = 0.004), but a similar trend was evident for inner wall area (p = 0.06). In an overall analysis of all individuals with asthma compared with all control subjects, absolute wall areas (mm₂) for both inner and outer wall areas were significantly increased in individuals with fatal asthma (WA_i: asthma = 0.56 ± 0.08, control = 0.37 ± 0.04, p = 0.046; WA_o: asthma = 1.91 ± 0.40, control = 1.24 ± 0.24, p = 0.04).

View larger version (20K): [in this window] [in a new window]   Figure 2.   Relationship between the square root of outer (A) and inner (B) wall area (mm2) and internal perimeter (2-10 mm) in airways of young and old normal subjects. Mean regression lines and 95% confidence intervals are shown.

View larger version (21K): [in this window] [in a new window]   Figure 3.   Relationship between the square root of outer (A) and inner (B) wall area (mm2) and internal perimeter (2-10 mm) in airways of young and old individuals with asthma. Mean regression lines and 95% confidence intervals are shown.

Point counting of smooth muscle area fractions showed that the mean area fraction of smooth muscle, uncorrected for changes in connective tissue matrix within muscle bundles, was higher in individuals with asthma (19.3 ± 1.3) than in control subjects (14.5 ± 1.9, p = 0.038) with a progressively increasing trend from normal subjects to young individuals with fatal asthma to old individuals with fatal asthma (Table 4). Calculation of smooth muscle area, which was done by multiplying the area fraction of smooth muscle by the total wall area for each airway, showed that old individuals with fatal asthma had more muscle than young individuals with fatal asthma (p = 0.05), and each asthma group had more muscle than age-matched control subjects (p = 0.04). However, the smooth muscle cell area fraction within smooth muscle bundles was larger in control subjects (90.9 ± 1.4) than in individuals with asthma (80.9 ± 1.7, p < 0.003, Table 4, Figure 4). There was no difference in the proportion of connective tissue matrix between young and old subjects in either the control group or group with asthma. By two-way ANOVA there was also a group effect (p = 0.008) but no age effect. Recalculation of smooth muscle area after correcting for the reduced cellular content of smooth muscle bundles in individuals with asthma (Asm/bundle) still showed a significant increase in smooth muscle area in individuals with asthma compared with age-matched control subjects (young control subjects, 0.09 ± 0.02 mm²; old control subjects, 0.11 ± 0.01 mm²; young individuals with fatal asthma 0.18 ± 0.04 mm²; old individuals with fatal asthma = 0.45 ± 0.17 mm²), with a trend toward a progressive increase in muscle with age in individuals with fatal asthma (p = 0.07, young versus old).

View larger version (10K): [in this window] [in a new window]   Figure 4.   Area fraction of smooth muscle cells within muscle bundles in normal subjects and individuals with asthma. Mean data of all airways averaged per case are shown, along with a horizontal bar indicating mean values per group ± SE. A decrease in area fraction equates to greater extracellular matrix around individual cells. Circles = control subjects, squares = individuals with asthma.

Older individuals with asthma had more airway narrowing (38.5 ± 3.9%) than the younger individuals with asthma (50.7 ± 3.5%, p = 0.03; Table 2, Figure 5). Overall subjects with asthma had more airway narrowing per case than the control subjects (Abm/A*bm%: individuals with asthma 45.3 ± 2.7, control subjects 68.0 ± 4.1, p = 0.0001). There was no relationship between Abm/A*bm% and Pbm. Lumenal mucus-debris was significantly increased in both groups with asthma (49.9 ± 5.0) compared with control subjects (7.3 ± 7.3, p < 0.0003). There were no differences between the young and old subjects within the control groups or groups with asthma (Table 5). Lumenal mucus-debris was present in some but not all airways in each subject with asthma.

View larger version (11K): [in this window] [in a new window]   Figure 5.   Percentage airway narrowing in individuals with fatal asthma in young (circles) and older (squares) subgroups and in control subjects. Mean data of all airways averaged per case are shown, along with a horizontal bar indicating mean values per group ± SE.

The thickness of the subepithelial collagen was 11.3 ± 0.8 µm in the young group with fatal asthma and 13.7 ± 1.3 µm in the old group with fatal asthma (Table 5), a twofold increase over control values (p = 0.0001), with no age effect.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of this study show that among individuals with fatal asthma older people with a longer duration of disease have an increase in airway wall area, predominantly due to an increase in adventitial area, as well as greater postmortem reduction in the airway lumen. Lumenal content, subepithelial collagen deposition, and the content of connective tissue matrix within smooth muscle bundles are similar in both groups with asthma, although greater than control values. Both old and young subjects with fatal asthma have greater amounts of airway smooth muscle compared with normal subjects, with a trend to greater amounts of muscle in older individuals. Surprisingly, the younger individuals with fatal asthma did not have increased wall dimensions compared with age-matched normal subjects. These results suggest that with increased duration of asthma there is ongoing remodeling with an increase in airway tissue. Alternatively the results may reflect age-related changes rather than duration by itself, although there was no age effect in the control subjects. The observation that total wall thickness was not greater in young individuals with fatal asthma than in control subjects suggests that factors other than airway geometry contribute to the pathogenesis of fatal attacks in this age group. The similar morphological changes within the airway lumen in both groups with asthma suggest that lumenal obstruction is probably the most likely factor contributing to fatal attacks in young and old alike. Finally our data suggest it is very uncommon to die without substantial lumenal obstruction.

Previous Reports and Consequences of Structural Changes

There is increasing evidence that the chronic nature of asthma is significantly influenced by changes in the structure of the airways that are a consequence of a continuing chronic inflammatory process that is episodic in nature. Theoretical and morphometric studies have indicated that many of the important functional changes in individuals with asthma such as increased maximal airway narrowing, paradoxical or deficient response to deep inspiration, and progressive loss of airway distensibility may be primarily related to the structural changes collectively termed airway remodeling (16). Previous detailed morphometric studies from a number of laboratories, including our own, provide direct evidence in support of the notion that the inner wall, airway smooth muscle, and adventitia are all thickened in the asthmatic state. One notable exception was a study in which only large airways were examined (17). The results of several studies have suggested that there is a gradient in airway wall dimensions from normal subjects to individuals with moderate and severe fatal asthma (1, 18).

In contrast to our study, however, all the previous studies have analyzed much older individuals with asthma and control subjects, and the individuals with asthma probably had a longer duration of asthma than our young group with fatal asthma. Our results indicate that there may be a tendency for airway wall area to decrease with age in normal subjects, potentially exaggerating morphometric differences between airways in normal subjects and in individuals with asthma in previous studies that have included older control subjects. The view that only severe structural change can cause airway narrowing severe or persistent enough to cause death (2) is not supported by our data in the young individuals with asthma.

It is possible we have found an age effect rather than an asthma duration effect on airway structure in individuals with asthma. There is a range of duration of asthma in the two groups, but the available evidence indicates that the median duration of asthma is much longer in the older group compared with the younger group. Most frequently individuals had onset of asthma in childhood (median age of onset 10.5 yr) (12), that is, the median duration of symptomatic asthma for the young group would be 9 yr, and for the old group 32 yr. More complete data would likely decrease the confidence intervals of the results but not the mean. It is commonly stated that most people who die of asthma have had the disease for many years (19). Review of a series of deaths from asthma from similar white adult populations collected through 1960- 1990, where data are reported, shows similar duration trends, for example, Carroll and coworkers (individuals with fatal asthma aged 18-69, mean 46 yr) show mean duration of 28 yr (20); Tough and coworkers (individuals with fatal asthma aged 15-34 yr) report 85% of subjects diagnosed by age 10 (21). The British Thoracic Society study showed 80% of deaths from asthma occurred in subjects with duration greater than 10 yr, and 59% for > 20 yr (22). However, 57% of the patients in this study were aged 45-64 yr, and as there is a second peak of asthma onset in middle life, excluding these older subjects would likely make the mean asthma duration data even longer. In the control groups in our study no age-related effects were detected, which supports the argument that duration rather than age effects are being observed in our study. We have not performed correlative analyses between reported duration and structural changes because (1) recall of age of onset of asthma by relatives is unreliable because of recall biases, (2) age at diagnosis is a poor indicator of the age of onset of symptoms (23), and (3) in considering morphological changes, the likely variable and unquantifiable presymptomatic period may be important.

Thickened airways can greatly augment lumenal obstruction for a given degree of airway smooth muscle shortening. Adventitial thickening may reduce the stabilizing influence of the lung parenchyma, the phenomenon of parenchymal interdependence. However, thickening of the smooth muscle layer may be the most important of the airway wall changes in individuals with asthma, assuming that smooth muscle contractility is normal or increased (24). Although thickening of the wall internal to the smooth muscle could augment narrowing, it may alternatively provide a load opposing smooth muscle shortening, that is, protect against exaggerated airway narrowing (24, 27). The results of our study, showing thicker adventitial areas in the older group of individuals with fatal asthma, suggest that loss of parenchymal interdependence might be a more important mechanism of severe attacks in this age group. The excess postmortem airway narrowing detected in old individuals with fatal asthma may be secondary to loss of interdependence or increased muscle mass, or may reflect an age-related reduced ability of the airway to relax. Conversely, in the young individuals with fatal asthma, factors other than geometric influences, for example, a relative lack of collateral ventilation or increases in smooth muscle contractility, may be more important (28). Young adults with fatal asthma may have intrinsically less severe disease, but may die because of poor adherence to or underprescription of antiinflammatory regimens (29). The effect of prior drug therapy on the variability in structural changes noted in this study cannot be evaluated because of lack of documentation of prescriptions and compliance. We have previously shown an increased adventitial area following high dose chronic ₂-agonist administered in an animal model (13), therefore we cannot exclude a drug effect in the causation of adventitial changes.

Adventitial thickening may be due to inflammatory cell infiltrate, edema, matrix deposition, or an increased number or dilation of bronchial vessels. We have some uncertainties as to the cause of the adventitial thickening in the formalin-fixed sections used in this study. Edema or matrix deposited may have been washed out during fixation and hence what we sometimes observed as "space" between tissue components may have been matrix or edema-filled. Alternatively, in the absence of special stains to delineate lymphatic and blood vessels walls, the space may represent an increased number of, or dilated vessels. Finally we cannot exclude that there is increased separation of tissue components secondary to increased peribronchial pressure.

Lumenal Content and Reduced Lumenal Area

Lung fixation without inflation leads to preservation of lumenal content. However, no further measures were taken to preserve lumenal content during fixation and processing, and therefore the true degree of lumenal obstruction may have been underestimated. Despite this, approximately 50% of the airway lumen was obstructed in individuals with fatal asthma, underlying the importance of mucus exudate and cell debris in the pathogenesis of fatal asthma. Airways from all blocks showed at least some partial occlusion, suggesting that, contrary to recent reports (30), dry airways are a very uncommon finding in individuals with fatal asthma. We found that older individuals with asthma had greater postmortem airway narrowing than younger individuals, with both groups demonstrating greater narrowing than control subjects. These findings were present despite variable treatment prior to death, postmortem changes in airway tone, variable airway lumenal content, and lack of fixed inflation pressure. If remodeling reduces lumenal area, equal amounts of mucus exudate production in young and old subgroups with asthma may lead to more adverse consequences in the older group. Thus, lumenal content was relatively greater in old individuals with fatal asthma, although confounded by greater airway narrowing in these patients, which will increase apparent obstruction of airways, despite an equal absolute amount of content. Conversely, the postulate that death from asthma at a young age compared with older subjects is primarily secondary to a greater degree of lumenal obstruction by mucus exudate and cell debris is unlikely to be correct, based on our data.

Inhalation of cigarette smoke or other airborne pollutants may cause airway inflammation and result in reduced lumenal area. In regard to smoking history, this was unknown in some subjects although most (> 75%) were not current smokers. Subjects were excluded if there was evidence of chronic obstructive pulmonary disease (COPD) on gross or microscopic examination. However, we cannot exclude a contribution from passive smoke exposure or past smoking to the reduced lumenal area found in older individuals with asthma. The frequency of passive exposure should have been equal by chance in both groups, albeit for a longer period in older subjects, and lumenal area was preserved in older control subjects. Our previous studies suggest that smoking should affect only smooth muscle and airway narrowing/lumenal content, as we showed no differences in wall dimensions in subjects with COPD versus control subjects (1). In addition, Carroll and coworkers showed no differences in airway dimensions in smoking versus nonsmoking individuals with asthma less than 58 yr of age (20).

Subepithelial Collagen

One morphometric alteration shared between the young and old groups with fatal asthma is a similar thickening of the subepithelial collagen. Our discussion of age-related morphometric factors in causation of death and altered airway function excludes the potential importance of increased subepithelial collagen (24, 27, 31, 32), in that it has been hypothesized that overall wall thickness is less important in the pathogenesis of abnormal airway function than subepithelial collagen. We had hypothesized that if airway wall area increased with age, reflecting a longer duration of inflammation/repair, then the thickness of the subepithelial collagen would also be increased. We did not find this, confirming the findings of previous investigators (33, 34). The measurement technique used for subepithelial collagen proved robust, yielding normal values within the range of previous reports. Thickness values for individuals with asthma are at the upper limit of previous estimates based on endobronchial biopsies from subsegmental or larger airways, where the normal subepithelial collagen is thicker.

Smooth Muscle

The proportion of smooth muscle cells per muscle bundle was larger in control subjects than in individuals with asthma, suggesting edema or an increase in proteoglycans or other connective tissue elements surrounding the muscle fibers in asthma, as has been previously described using both light (17, 35) and electron microscopy (unpublished data). However, to our knowledge, this study is the first to accurately quantify this change in small airways in individuals with asthma. An increase in pericellular matrix or edema may alter the mechanics of smooth muscle shortening, by reducing radial forces limiting shortening (36). Our results concerning the increased muscle in the airway wall in older individuals with fatal asthma are comparable with higher estimates in previously reported results, showing a 4.5-fold increase in smooth muscle in airways of 2 to 10 mm internal perimeter compared with age-matched control subjects, even when muscle area is corrected for increased connective tissue matrix around smooth muscle cells.

Conclusions

The results of this study provide new data in support of the hypothesis that duration of asthma or older age is associated with progressive remodeling of the airway wall, including a progressive increase in airway smooth muscle volume and a greater reduction in the airway lumen. The results provide a basis for understanding the more rapid decline in lung function reported in individuals with asthma. The airway modeling studies based on structural changes associated with asthma (24, 27, 31) suggest that a progressive increase in the components of the airway wall should result in greater responsiveness following challenge. The comparison of young and old individuals with fatal asthma reported here shows that there is a similar increase in the lumenal contents and the connective tissue matrix in the subepithelium and around smooth muscle cells in both groups. The other novel aspect of this study is that younger individuals with asthma who die of their disease do not have increased airway wall thickness; only the smooth muscle volume was increased compared with age-matched control subjects. This result suggests that extensive remodeling is not a prerequisite for a fatal asthma attack. The observation that the old and young individuals with asthma shared postmortem airway narrowing and extensive airway lumen obstruction suggests that smooth muscle contraction and production of an inflammatory mucus exudate are important mechanisms for fatal attacks in both age groups.