INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Healthy humans do not develop airway narrowing in response to bronchoconstrictors under conventional bronchoprovocation tests; however, severe bronchoconstriction to methacholine, a direct spasmogen, may take place in healthy individuals if the challenge is carried out in the absence of deep inspirations (1). In asthmatics, on the other hand, deep inspiration has little influence on the outcome of a stepwise, increasing dose, methacholine challenge. Thus, we and others have suggested that lung inflation is responsible for the maintenance of airway patency in healthy humans and that the loss of this function in asthma plays a major role as an underlying abnormality leading to the phenomenon of bronchial hyperresponsiveness (BHR).

We have recently determined that, in healthy subjects, the effect of deep inspiration is dual (2, 3): it acts as a bronchodilator (that is, once bronchoconstriction is established, deep breaths can partially reverse it) and as a bronchoprotector. In the latter case, deep inspiration maneuvers performed within a few minutes before the administration of methacholine protect the airways from the subsequent exposure to the spasmogen. Using single-dose methacholine bronchoprovocations, we developed the methodology to measure and compare the bronchodilatory with the bronchoprotective effects of deep inspiration (2, 3). In healthy subjects, we have found that bronchoprotection is more potent than bronchodilation (3). In patients with asthma, on the other hand, we failed to detect any evidence of bronchoprotection by deep inspirations (2). The asthmatic population used in that study was selected on the basis of moderate to severe airways hyperresponsiveness to a conventional methacholine provocation (provocative concentration causing a 20% reduction in FEV₁ [PC₂₀] < 1 mg/ml). Hence, the need to examine patients with milder degree of hyperresponsiveness was raised. As for the bronchodilatory ability of deep inspiration in patients with asthma, our previous work suggested that, although reduced, a certain degree of this function was still present (1). However, bronchodilation and bronchoprotection have not been compared in asthmatics.

The present study was designed with two goals: first, to assess whether the lack of bronchoprotection is related to the outcome of a conventional methacholine bronchoprovocation (PC₂₀), and second, to evaluate the extent to which the bronchodilatory effect is also lost in asthmatics. For reference purposes, we included a group of healthy subjects in whom both the bronchodilatory and the bronchoprotective effects of deep inspiration were determined. Finally, to examine whether the defect in the function of lung inflation is related to the phenomenon of airways hyperresponsiveness as opposed to the clinical phenotype of asthma, we also investigated a group of patients with clinical characteristics of allergic rhinitis but no history of lower airways symptoms, who were selected based on the fact that they had mild to borderline airways hyperresponsiveness to methacholine.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We studied a total of 46 subjects: 10 healthy individuals, 10 subjects with allergic rhinitis but no asthma, and 26 asthmatics. Healthy, nonsmoking subjects were recruited among the employees of Johns Hopkins University, whereas rhinitics and asthmatics were selected from a large database of allergic individuals recruited from the community by radio and newspaper advertising.

Healthy subjects had never experienced upper or lower respiratory symptoms consistent with the diagnosis of rhinitis or asthma; all but one, who had mildly positive skin-prick tests to Dermatophagoides pteronyssinus and D. farinae, were skin test negative. Upon routine methacholine bronchoprovocation, all these subjects received the highest dose of spasmogen (75 mg/ml) with less than 15% reduction in FEV₁.

Rhinitic subjects were recruited for reporting a history of nasal symptoms for at least 10 mo every year and being skin test positive to at least one of the perennial aeroallergens tested (perennial allergic rhinitis, n = 3), or a history of symptoms for at least 1 mo but less than 10 mo every year, and one or more positive skin tests to a relevant seasonal aeroallergen (seasonal allergic rhinitis, n = 7). Subjects allergic to pollen were studied out of season. None of the rhinitic subjects reported a history of, or current symptoms consistent with the diagnosis of asthma; however, they were selected for having demonstrated a 20% reduction in FEV₁ during the screening routine methacholine challenge, with PC₂₀ between 1 and 25 mg/ml (mild to borderline hyperresponsiveness) (4). None of them was receiving any nasal or other medications at the time of the study.

The criteria for a subject in our database to be termed asthmatic are either (1) a PC₂₀ < 10 mg/ml in the screening methacholine provocation and the report of at least one seasonal or chronic lower airway symptom consistent with asthma (breathlessness, cough, phlegm, wheezing), or (2) 10 < PC₂₀  25 mg/ml and at least two seasonal or chronic asthma symptoms. These definitions are compatible to those employed and validated by the Collaborative Study on the Genetics of Asthma in which our group participates (5). All asthmatics participating in this study were skin test positive to at least one aeroallergen. They had intermittent or mild persistent disease (6) and they were using only short-acting -agonist on demand, which they withheld for at least 12 h before each visit to the laboratory. Coffee or tea was not allowed in the morning before a test. Subjects were tested at least 4 wk after their most recent upper respiratory infection. We entered two groups of asthmatics in this study: 12 subjects with moderate to severe hyperresponsiveness (PC₂₀ < 1 mg/ml) (4), and 14 with mild to borderline hyperresponsiveness (1 < PC₂₀ < 25 mg/ml) (4). The study was approved by the Johns Hopkins Bayview Medical Center institutional review board and all subjects gave written, informed consent prior to participation.

Study Design

The entire study consisted of three phases. Phase 1 was the screening evaluation. Phase 2 aimed at establishing a single dose of methacholine inducing at least 20% reduction in lung function and at determining the bronchodilatory effect of lung inflation. During Phase 3, the bronchoprotective effect of lung inflation was examined.

Phase 1. The screening evaluation included a respiratory questionnaire for both upper and lower airways, allergy skin testing to a panel of 10 common aeroallergens using the epicutaneous method, and a routine methacholine challenge (7). The routine methacholine challenge was performed as follows. First, reproducible baseline lung function measurements were obtained; sterile diluent (phosphate-buffered saline solution) and then increasing concentrations of methacholine (0.025, 0.075, 0.25, 0.75, 2.5, 7.5, 25, 75 mg/ml) were delivered through a DeVilbiss 646 nebulizer (DeVilbiss Co., Somerset, PA) attached to a Rosenthal dosimeter (Laboratory for Applied Immunology, Inc., Fairfax, VA), which is triggered by the beginning of each inhalation (dose delay: 0.03 s; dose duration: 0.6 s). Saline solution and methacholine were administered by inhalation with five slow breaths, from functional residual capacity to total lung capacity. Spirometric measurements were repeated 3 min after saline inhalation and after each dose of methacholine. After each step, the best FEV₁ among three acceptable maximal spirometric maneuvers was recorded. The test was stopped when either FEV₁ dropped by 20% from the postdiluent value or the highest methacholine concentration (75 mg/ml) was delivered. The provocative concentration of methacholine causing this reduction (PC₂₀) was calculated by interpolation of the dose- response curve.

Phase 2. This part of the study was designed (1) to determine a single dose of methacholine inducing at least a 20% reduction in FEV₁ from baseline value under prohibition of deep breaths, and (2) to evaluate the ability of deep inspiration to reverse the induced bronchoconstriction. The protocol is depicted in Figures 1a and 1b. The bronchoprovocation performed is a single-dose, methacholine challenge, originally employed by Malmberg and colleagues (8) and modified by our group (2, 3). After three baseline spirometric maneuvers, subjects were asked to refrain from taking deep breaths for a 20-min period. As we have shown, this period allows for the bronchoprotective effects of the baseline, deep breath-involving, spirometric maneuvers to dissipate (2, 9). Thereafter, a single concentration of methacholine was administered as five separate tidal inhalations corresponding to 5 independent activations of the dosimeter, and, after 3 min during which deep breaths were still avoided, lung function was again measured with a single spirometric maneuver (Figure 1a). The difference between the postmethacholine FEV₁ and the baseline value was expressed as percent change from baseline. The challenge was performed on several occasions, at least 24 h apart, with increasing single doses of methacholine, until the targeted reduction in lung function was attained. Based on our previous experience with this model, the concentration of methacholine delivered during the first provocation of Phase 2 was 10 mg/ml in healthy subjects, 5 mg/ml in rhinitics, and 0.025 mg/ml in asthmatics. One to three visits per subject were required to complete this phase. To measure the bronchodilatory effect of deep inspiration, the protocol of the single-dose methacholine challenge, which resulted in the targeted > 20% reduction in FEV₁ from baseline values was extended (Figure 1b). Each subject was asked to perform four fast deep inspiratory maneuvers immediately after the postmethacholine spirometry, and lung function measurement was then repeated. The changes from baseline values in the post-deep inspiration spirometry were compared with those recorded by the postmethacholine spirometry.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Schematic of the protocol for methacholine bronchoprovocations employed to establish the single dose of spasmogen inducing the targeted reductions in FEV1 (a) and to assess the bronchodilatory (b) and the bronchoprotective (a and c) effects of deep inspiration. See also Phases 2 and 3 in STUDY DESIGN. DI = deep inspiration; bsl = baseline.

Phase 3. In this phase, the bronchoprotective effect of deep inspiration was assessed. Each subject underwent another single-dose methacholine challenge (Figure 1c), which employed the Phase 2-determined concentration of spasmogen capable of inducing at least 20% reduction in FEV₁ from baseline, in the absence of deep breaths. At the end of the 20-min deep inspiration-free period, immediately before methacholine administration, five fast deep inspiratory maneuvers were performed. The resulting changes from baseline values in postmethacholine spirometric outcomes were compared with those of the same single-dose challenge in which no deep inspirations were included (Phase 2, Figure 1a).

Data Analysis

The outcomes of this study were obtained from spirometry. Because spirometric values follow normal distribution, parametric statistics were applied. One-way analysis of variance (ANOVA) was employed to compare baseline FEV₁, FVC, and FEV₁/FVC among the four groups. ANOVA was also used to compare (1) the mean provocative concentrations of methacholine causing the 20% reduction in FEV₁ in the conventional bronchoprovocation protocol, (2) the mean single doses of methacholine required to induce the targeted reductions in FEV₁ in the absence of deep inspirations, and (3) the percent reductions in FEV₁ attained by the single-dose challenges in the four groups. For bronchoprotection, the percent reductions in the spirometric outcomes from baseline in the protocol devoid of deep inspiration (Figure 1a) were compared, using paired t tests, with those in the protocol in which 5 deep inspirations preceded the methacholine (Figure 1c). Similarly, for bronchodilation, the percent reductions in the spirometric outcomes from baseline after the methacholine administration (Figure 1a) were compared with those obtained after deep inspirations were performed (Figure 1b). Also in primary analysis, to evaluate the magnitude of the bronchoprotective and the bronchodilatory effects of deep inspirations, we calculated the percent bronchoprotection and the percent bronchodilation by using the ratio between the percent induced reduction in FEV₁ in the absence of deep breaths and the percent remaining bronchoconstriction at the end of the bronchoprotective and the bronchodilatory protocols, respectively.

One-sample t tests were applied to evaluate whether the percent bronchoprotection and the percent bronchodilation were different from zero. To compare the percent bronchoprotection and bronchodilation among the four groups, one-way ANOVA was again employed. Post hoc analyses were performed with the Fisher's least significant difference (PLSD) test. In all analyses, two-tailed p values  0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Demographics and baseline lung function characteristics from the screening evaluation for all subjects participating in the study are depicted in Table 1. Both FEV₁ percent predicted and FVC percent predicted did not differ among groups (ANOVA, p = 0.20 and p = 0.35, respectively), but the FEV₁/ FVC did (ANOVA, p = 0.02). Mean baseline FEV₁/FVC value in asthmatics with mild to borderline hyperresponsiveness was significantly lower than that in healthy subjects and in rhinitics (p = 0.04 and p = 0.004, respectively). Also, the FEV₁/FVC in asthmatics with moderate to severe hyperresponsiveness was significantly lower than that recorded in rhinitics (p = 0.03). By design, the mean PC₂₀ clearly differentiated the study groups with the exception of asthmatics and rhinitics with mild to borderline hyperresponsiveness who, again by design, ended up being identical in terms of their methacholine PC₂₀ (p = 0.75). The single dose of methacholine determined from Phase 2 to induce more than 20% reduction in FEV₁ in the absence of prechallenge deep inspirations varied considerably among groups: healthy subjects required a methacholine dose of 53 ± 8.1 mg/ml (mean ± SEM), whereas asthmatics with moderate to severe hyperresponsiveness received 1.2 ± 0.33 mg/ml, and asthmatics and rhinitics with mild to borderline hyperresponsiveness 5.1 ± 1.5 and 11.5 ± 3.3 mg/ml, respectively (ANOVA, p < 0.0001). On the other hand, the reduction in FEV₁ attained by this single-dose challenge was identical among the four groups (mean ± SEM percent induced bronchoconstriction in FEV₁, for healthy subjects: 43 ± 3.9%; for asthmatics with moderate to severe BHR: 41 ± 4.5%; for asthmatics with mild to borderline BHR: 37 ± 4.5%; for rhinitics with mild to borderline BHR: 46 ± 3.9%, ANOVA, p = 0.49).

Bronchoprotective Effect of Deep Inspiration

The ability of deep inspiration to protect the airways against the effect of methacholine (Phase 3) is depicted in Figure 2. The spirometric outcomes are expressed as percent reductions from baseline values. The striking bronchoprotection conferred by deep inspiration in healthy subjects (mean ± SEM of the difference between the two provocations: 34 ± 3.7%, p < 0.0001) was completely lost in both groups of asthmatics and in the rhinitics with mild to borderline BHR: in these three groups, no significant difference was found between the methacholine-induced percent reduction in FEV₁ from baseline in the absence of deep breaths, and the percent reduction when 5 deep breaths were taken before the spasmogen (mean ± SEM percent difference between the two challenges; for asthmatics with moderate to severe BHR: 5.3 ± 5.5%, p = 0.36; for asthmatics with mild to borderline BHR: 1.0 ± 5.7%, p = 0.87; for rhinitics with mild to borderline BHR: 3.0 ± 5.4%, p = 0.58). Similar results were obtained for FVC. These findings were confirmed by calculating the percent bronchoprotection for each group as described under DATA ANALYSIS (Table 2); the four groups were significantly different in terms of the percent bronchoprotection (ANOVA, p = 0.002); however, whereas the percent bronchoprotection was significantly higher in healthy subjects, compared with the other three groups, no differences among the latter were detected. Furthermore, in the three groups with hyperresponsiveness, the percent bronchoprotection was not significantly different from zero, indicating complete loss of bronchoprotection.

View larger version (50K): [in this window] [in a new window]   Figure 2.   Reductions in FEV1 and FVC from baseline values in the course of the protocol aimed at evaluating the bronchoprotective effect of deep inspiration. Bars represent mean ± SEM. Hatched bars represent percent reductions in spirometric outcomes in the protocol in which no deep inspiration preceded the single-dose methacholine challenge. Open bars represent percent reductions in the protocol in which 5 deep inspirations were performed before the methacholine administration. * p < 0.0001, comparing the post-methacholine FEV1 and FVC recorded in the protocol devoid of deep inspiration with the post-methacholine FEV1 and FVC recorded in the protocol in which deep inspirations preceded Methacholine administration. DI = deep inspiration; MCh = methacholine.

Bronchodilatory Effect of Deep Inspiration

The reversal of airway narrowing by deep inspiration is depicted in Figure 3, in which the spirometric outcomes are, again, expressed as percent reductions from baseline values. In contrast to our findings with bronchoprotection, deep inspiration was able to partially reverse bronchial obstruction in healthy individuals, as well as in asthmatics and rhinitics. In all groups, the differences between the methacholine-induced reductions in FEV₁ before deep inspirations and the remaining bronchoconstriction after the deep inspirations were statistically significant (mean ± SEM percent difference; for healthy subjects: 28.6 ± 3.6%, p < 0.0001; for asthmatics with moderate to severe BHR: 15.9 ± 5.7, p = 0.02; for asthmatics with mild to borderline BHR: 17.5 ± 3.4, p = 0.0002; for rhinitics with mild to borderline BHR: 17.4 ± 3.7, p = 0.001). Similar results were obtained when FVC was considered. Having established that deep inspiration induces bronchodilation in all these groups, we further investigated whether the magnitude of this effect was different between groups by calculating the percent bronchodilation as described under DATA ANALYSIS (Table 2): overall, ANOVA demonstrated significant differences among the four groups (p = 0.03). Post hoc analysis showed that the bronchodilatory effect of deep inspiration in healthy subjects was significantly stronger than that in the asthmatics with moderate to severe BHR and in the rhinitics (p = 0.005 and p = 0.02, respectively), but not stronger than that in the asthmatics with mild to borderline BHR (p = 0.12). No statistically significant difference in percent bronchodilation was found among the three groups with BHR.

View larger version (47K): [in this window] [in a new window]   Figure 3.   Reductions in FEV1 and FVC from baseline values in the course of the protocol aimed at evaluating the bronchodilatory effect of deep inspiration. Bars represent mean ± SEM. Hatched bars represent percent reductions in spirometric outcomes in the protocol in which no deep inspiration preceded the single-dose methacholine challenge. Open bars represent percent reductions in spirometric outcomes obtained after the deep inspirations that followed methacholine administration. * p < 0.05, comparing the post-methacholine FEV1 and FVC recorded before and after deep inspirations. For detailed p values refer to RESULTS.

In a previous study (3), we demonstrated that, in healthy subjects, the bronchodilatory effect of deep inspiration is limited by increasing bronchoconstriction, as opposed to the bronchoprotective effect, which is equally strong against concentrations of methacholine inducing high or low levels of airflow limitation. Herein, we evaluated the extent to which deep inspiration-induced bronchodilation is also limited in asthmatics and rhinitics. Two regression lines were constructed to examine the extent to which the dilatory ability depends on the induced bronchoconstriction in the absence of deep breaths. One regression line was constructed with the data from the healthy subjects whereas the other with pooled data from all subjects with airways hyperresponsiveness. The two slopes were compared with t test. These regressions, which are depicted in Figure 4, show that the remaining bronchoconstriction after the bronchodilatory maneuvers tends to be higher when the bronchoconstriction before the bronchodilatory maneuvers is higher. This indicates limitation of the bronchodilatory ability in all subject groups. It is noteworthy that, for the same degree of methacholine-induced obstruction, the group of subjects with BHR appears to have higher remaining bronchoconstriction after deep inspirations, compared with the healthy subjects. However, comparison of the slopes of the two regressions did not yield a statistically significant difference (p > 0.10). No differences in this slope could be found between the three hyperresponsive groups when their data were analyzed separately (data not shown).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Correlations between the degree of bronchial constriction induced by the single dose of methacholine in the absence of deep inspirations and the remaining bronchoconstriction after the bronchodilatory maneuvers. Open circles and the interrupted regression line represent data from subjects with BHR. Closed circles and the continuous regression line represent data from healthy subjects.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have recently established that asthmatics have lost the ability of deep inspiration to prevent airways from constricting, under experimental conditions (2). This protective effect is quite powerful in healthy subjects (2, 3). Here, we have demonstrated that the loss of bronchoprotection is associated with the phenomenon of airways hyperresponsiveness, regardless of the clinical diagnosis of asthma. In addition, we have confirmed that, in individuals who manifest airways hyperresponsiveness in the conventional way, deep inspiration can act as a bronchodilator mechanism, almost as well as in healthy, normoresponsive subjects. Thus, we can propose that (1) lung inflation-induced bronchoprotection operates through a different mechanism than that of lung inflation-induced bronchodilation; and (2) bronchoprotection, much more than bronchodilation, is the property that differentiates airways hyperresponsiveness from the healthy state.

To obtain these results, we have used a model involving single-dose methacholine bronchoprovocations. This was originally developed by Malmberg and colleagues (8) and further modified and characterized by our group (2, 3). The single-dose challenge offers simplicity in the evaluation of lung inflation-induced bronchodilation and bronchoprotection, as a series of deep inspirations can be performed immediately before or after the methacholine administration. The model is quite reproducible (2) and the vast majority of healthy individuals will develop substantial bronchoconstriction, because any protective effects of potential spontaneous deep inspirations taken before the beginning of the experiment will dissipate during the 20-min quiet breathing period that precedes the administration of methacholine (2). One disadvantage of this model is that 1 to 3 single-dose challenges have to be initially performed, in order to identify the dose that is capable of inducing the targeted reduction in lung function. The second disadvantage lies in the fact that, in the protocol where bronchoprotection is evaluated, the maximal breath spirometry that is performed after the provocation inherently produces bronchodilation. We could have used partial expiratory maneuvers or other measurements of airway function, which do not involve deep inspirations, at baseline and after the methacholine provocation to evaluate the effect of deep breaths; yet, the superior reproducibility and reliability of full spirometry led us to adopt the current approach. The risk of overestimating the bronchoprotective effect of deep inspiration is reduced by the fact that the same spirometric maneuver with the same potential bronchodilatory properties is performed at the end of the first component of the bronchoprotection protocol, in which no deep inspirations precede the administration of methacholine. On the other hand, in healthy subjects, the maximal spirometry performed at the end of the single-dose methacholine provocation may result in requirement for higher single-dose methacholine, compared with what would have been adequate to reduce lung function in the absence of the maximal spirometry. This is because the bronchodilatory ability of deep inspiration in healthy subjects is somewhat stronger that that of individuals with airways hyperresponsiveness (Table 2). Another design element that probably also results in the need for higher single-dose methacholine to induce bronchoconstriction in healthy subjects is the single-dose challenge itself, with a provocation in which consecutive increasing doses of methacholine are administered in fixed intervals. The aforementioned aspects of the experimental model that we have used in this study may explain why, in our original publication in which the importance of deep inspiration in protecting healthy individuals from bronchoconstriction was described, the average top concentration of methacholine resulting in significant reduction in lung function was 5 to 6 times less than the one used in the current study (1).

Our results clearly show that the bronchoprotective effect of deep inspiration is not only absent in asthmatics with moderate to severe hyperresponsiveness (2), but is also lost in asthmatics who have a milder degree of hyperresponsiveness. Most importantly, lung inflation did not exert any bronchoprotection in the rhinitics who were selected by virtue of having airways hyperresponsiveness, but who had never experienced symptoms consistent with the diagnosis of asthma (Figure 2). These findings indicate that a complete loss of the bronchoprotective effect of deep inspiration is closely associated with the development of airways hyperresponsiveness, even in its mildest form: once the bronchoprotective ability is lost, increased responses to bronchoconstrictors occur.

The fact that patients with allergic rhinitis may show increased response to methacholine is well established. The prevalence of BHR in the rhinitic population has been reported to range from 22 (10) to 40 (11, 12) or even to 55% (13). Several studies have demonstrated that, among rhinitic patients, those who have airways hyperresponsiveness are more likely to develop asthma (12, 13). In a study by Braman and colleagues (12) carried out on 40 patients with rhinitis, none of the 24 with a negative methacholine test during baseline observation developed asthma in the subsequent 5 yr. Of the 16 patients with hyperresponsiveness, three developed asthma. Because we have now shown that rhinitics with hyperresponsiveness lack the bronchoprotective property of deep inspiration as much as asthmatics, it is not unreasonable to propose that this defect may constitute a very early alteration in lung function in the process of developing asthma. Parenthetically, in preliminary work from another study (14), we have obtained clear evidence that rhinitics without conventional airways hyperresponsiveness have a fully functional bronchoprotective effect of deep inspiration.

The mechanism through which deep inspiration exerts its bronchoprotective role has not been addressed. Studies on animal models and airway smooth muscle in vitro work point out different possible explanations for this phenomenon, such as activation of neural or hormonal pathways, or even stretch-induced changes in the contractile apparatus of the airway smooth muscle. Neural pathways may include inhibition of cholinergic tone (15) or activation of the nonadrenergic, noncholinergic (NANC) bronchodilator system which uses nitric oxide (NO) as its major neurotransmitter (16). These two mechanisms may, actually, interact. It has become evident, at least in the vasculature, that acetylcholine, by stimulating M₂ muscarinic receptors, suppresses the release of NO by NANC nerves (17). If acetylcholine release is inhibited by lung inflation, the M₂ receptor-mediated NO suppression may dissipate, allowing for NO to be released at higher concentrations. Fredberg and colleagues, as well as Gunst and colleagues, have proposed two hypotheses concentrating on the airway smooth muscle, which may offer some other explanation to the bronchoprotective effect of lung inflation (18, 19). Fredberg's work suggests that the stretch imposed by deep inspiration before the administration of methacholine may set the airway smooth muscle in a condition of disequilibrium that makes the development of smooth muscle contraction more difficult to occur (18). According to the hypothesis put forward by Gunst and colleagues (19), the response to methacholine is markedly reduced if volume oscillations are applied on bronchi; this could result from changes in the organization of the contractile elements in relation to smooth muscle cell length (20).

One could question whether the lack of the bronchoprotective ability of deep inspiration, which, in the present study, appears to behave as an "all or none" phenomenon, is the consequence of an inherited abnormality, or whether it is caused by changes in lung physiology induced by airway inflammation. Several studies have shown that airways hyperresponsiveness is more prevalent in relatives of asthmatic subjects compared with those of nonasthmatic control subjects, suggesting a genetic predisposition (21). In 1995, airways hyperresponsiveness was mapped to a region of chromosome 5q31-q33 (25), but it is still not clear how independent this linkage is from the "high IgE" phenotype, which also maps in the same region. In any case, it is tempting to speculate that a strong element associated with airways hyperresponsiveness, such as the lack of deep inspiration-induced bronchoprotection, can be present as a genetic trait in the general population and may constitute a predisposing factor for the development of asthma. It is interesting to note, in this respect, that, over the past few years of working with this model, we have encountered a small number of healthy subjects who lack bronchoprotection. Against the genetic trait notion, however, argues the transient development of airways hyperresponsiveness in rhinitic, normoresponsive individuals, after inhaling repetitive doses of allergen (our unpublished observations). Also, Braman and colleagues (12), during the course of a 5-yr prospective study on rhinitics, reported the development of transient airways hyperresponsiveness in two patients. The observation that airways hyperresponsiveness may diminish or even disappear over the years in some asthmatics (26) further supports the hypothesis that acquired, rather than inherited, factors lie behind this phenomenon.

If the lack of bronchoprotection is an acquired phenotype, the most plausible hypothesis, based on our current understanding of asthma and its related abnormalities, is that bronchoprotection is lost as a result of airway inflammatory changes in the broadest context. In this respect, it is interesting that the presence of an inflammatory process, qualitatively similar to that observed in the asthmatic airways, has been reported in the lower airways of rhinitic patients, even when they are not exposed to natural allergen (e.g., outside the pollen season) (27). It has been recently demonstrated that subjects with allergic rhinitis and airways hyperresponsiveness show significantly higher eosinophil counts in induced sputum compared with rhinitics with normal methacholine responsiveness (28). Others and we have previously discussed the possible mechanisms through which an inflammatory process of the airways may counteract the beneficial effects of deep inspiration (2, 29, 30). Chronic inflammation may confer functional changes to the airway smooth muscle or may inhibit a humoral or neural mechanism that is activated by lung inflation-induced airway stretch, leading to bronchoprotection. For example, in the context of our earlier discussion on the possible role of NO as a mediator of the bronchoprotective effect of lung inflation, it is worth raising the hypothesis that dysfunction of M₂ muscarinic receptors, which has been proposed to occur as a result of asthmatic inflammation (31), may lie behind the lack of bronchoprotection in individuals with airways hyperresponsiveness.

The effectiveness of deep inspiration in reversing bronchoconstriction in all subjects with airways hyperresponsiveness, as well as in healthy individuals, is another interesting and potentially revealing observation. Despite the quantitative (but not very consistent) differences between the healthy subjects and the other groups involved in our study, the bronchodilatory effect of deep inspiration alone does not allow us to differentiate the asthmatic from the nonasthmatic (or the hyperresponsive from the normoresponsive) state, as the bronchoprotective effect does. This raises the possibility that the two phenomena involve two different mechanisms. For example, whereas deep inspirations taken before the administration of the spasmogen, could function by altering the state of the contractile apparatus of the airway smooth muscle (18, 19), so that the smooth muscle is more resistant to a subsequent bronchoconstrictor stimulus after bronchoconstriction takes place, the effect of these maneuvers may merely represent mechanical breaking of latch bridges (32). Of course, other mechanisms, such as the release of relaxant factors from airway tissue or from nerves in response to deep inspiratory maneuvers may account for the bronchodilatory effect.

If lung inflation-induced bronchoprotection and bronchodilation represent two distinct functions of the human lung, one could raise the possibility that abnormalities of both functions interact to produce the asthmatic phenotype. If all groups with airways hyperresponsiveness lack bronchoprotection, the small differences in bronchodilatory ability between these groups could explain the occurrence of asthmatic symptoms. However, in our hands, the bronchodilatory effect did not differentiate asthmatics from rhinitics. It is possible, nevertheless, that the degree to which bronchodilation is suppressed determines the magnitude of airways hyperresponsiveness, measured in the conventional bronchoprovocation way. Because bronchoprotection is completely absent in asthmatics, changes in airways responsiveness observed during the course of the disease, or improvement with anti-inflammatory treatment, may reflect changes in the bronchodilatory ability of lung inflation. Alternatively, such changes may reflect true changes in the contractility of airways smooth muscle.

In conclusion, we confirmed the dual role of deep inspiration as a bronchoprotector and a bronchodilator. Whereas the bronchodilatory ability is present in healthy and asthmatic subjects, as well as in individuals with allergic rhinitis and mild to borderline hyperresponsiveness, the bronchoprotective effect of deep inspiration is only found in healthy subjects. Our main conclusion is that further work aiming at our understanding of airways hyperresponsiveness and its role in asthma should focus on the bronchoprotective effect of lung inflation.