INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

EXPIRATORY FLOW LIMITATION (EFL) occurs in the intrathoracic airways whenever transpulmonary pressure achieves or exceeds the minimum value (critical pressure) necessary to generate maximal flow (9). In healthy subjects, EFL does not occur during tidal breathing at rest, because critical pressure is never achieved and expiratory flow is much less than maximal expiratory flow at the same lung volume. The difference between tidal and maximal flow is the flow reserve available for increasing ventilation. In airway obstruction, EFL may occur during tidal breathing, i.e., tidal expiratory flow becomes equal to maximal expiratory flow and flow reserve is abolished. Under these circumstances, ventilation can be increased or even maintained only by breathing at increased lung volumes, at which higher flows can be generated (17). As a result, functional residual capacity (FRC) increases above the relaxation volume of the respiratory system (dynamic lung hyperinflation). It has been suggested that dynamic hyperinflation may occur because the time constant of the respiratory system increases above the time available for tidal expiration (25). It has also been postulated that neural reflexes originated in the airways compressed downstream from the flow-limiting segments may prematurely activate the inspiratory muscles, thus stopping expiration before the relaxation volume of the respiratory system is attained (14,15). During induced bronchoconstriction, however, an increase in FRC was observed even in the absence of demonstrable EFL (23), thus suggesting that dynamic lung hyperinflation may be independent of EFL. An alternative explanation for this dissociation is that the methods currently used to detect EFL (9, 11, 14, 15, 23), which are based on measurements of mouth flows, are essentially insensitive to regional EFL.

This study was designed to test the hypothesis that EFL during tidal breathing may occur inhomogeneously across the lung, thus remaining largely undetected by mouth flow measurements. For this purpose, we compared the classical physiological methods to detect EFL with radionuclear techniques capable of identifying regional EFL in vivo (20, 21). Specifically, we used Technegas, which is an ultrafine aerosol of carbon molecules labeled with ^99mTc that distributes in the lung like an inert gas (3, 5, 6, 10, 12) and remains trapped in the lung, wherever it is deposited, for a time long enough for a study like this to be performed (3). The study was conducted in asthmatic subjects during pharmacologically induced bronchoconstriction.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Spirometry and Lung Volume Measurements

Screening Bronchial Challenge

Aerosol MCh was administered by a dosimeter (MEFAR, Brescia, Italy) during tidal breathing. Doubling doses of MCh were obtained by changing the number of breaths and/or concentration. After inhalation of saline, the challenge started from 20 µg of MCh and ended when IC decreased beyond its natural variability, which was previously determined to be 9% and 220 ml (16), or when the noncumulative dose of 2,400 µg was achieved. At baseline and after each MCh dose, the subject performed the following set of maneuvers. First, a forced expiratory maneuver initiated with no hesitation from end-tidal inspiration (partial maneuver) was performed in the plethysmograph. Second, at least six regular breaths and a forced expiration from TLC (maximal maneuver) were recorded to compute IC, forced expiratory volume in 1 s (FEV₁), and forced vital capacity. Finally, TLC was measured as above described. Forced expiratory flows from partial (part) and maximal (max) maneuvers were corrected for the amount of thoracic gas compressed. This was the volume difference between mouth flow vs. mouth volume and mouth flow vs. thoracic volume curves. The expiratory flow reserve was calculated as the difference between tidal and compression-free part near FRC (part_FRC).

The dose causing a decrease of FEV₁ by 20% of baseline value was calculated by interpolation of the log dose-response curve. Furthermore, the following three MCh doses were identified: the dose that decreased part_FRC by ~50% of baseline value leaving IC unchanged (dose A), the dose that completely abolished expiratory flow reserve without reducing IC beyond its natural variability (dose B), and the dose that also decreased IC beyond its natural variability (dose C).

Assessment of EFL on the Study Days

The occurrence of EFL was assessed by two methods. The first (partial flow method) was by comparing tidal flow with part corrected for the amount of thoracic gas compressed at that same absolute lung volume as determined in the body plethysmograph during the screening bronchial challenge. After about eight regular breaths, the subject expired forcefully from end-tidal inspiration near RV, then inspired immediately to TLC and expired forcefully to RV. EFL was said to occur whenever tidal expiratory flow was equal to part_FRC. When IC was decreased, tidal flow and part_FRC were compared at baseline FRC (14, 15). In the second method (NEP method) (11), a negative expiratory pressure (NEP) of 4 cmH₂O was applied at the mouth during the expiratory phase of quiet breathing after at least six regular breaths. The expiratory port of the pneumotachograph was connected in series to an interrupter valve (model RV-003, S/N 1003; Aeromech Devices, Almonte, PQ, Canada), a Venturi device (Aeromech Devices), and a tank of compressed air. Pressure at the mouth was measured by a differential pressure transducer (Validyne DP15, ±150 cmH₂O). The negative pressure was applied at regular intervals of ~15 s over 1 min and for 0.2 s after the beginning of expiration and maintained until the next inspiration started. Care was taken that the subject was breathing regularly before the negative pressure was applied and that there were no air leaks around the mouthpiece. All digitized data were stored in a personal computer for subsequent analysis. EFL was said to occur if the flow generated after application of the negative pressure was not increased compared with previous breaths.

Imaging Techniques

Scintigraphic studies were performed by using a large field-of-view gamma camera (General Electric, 400AT) equipped with a low-energy high-resolution collimator. Dynamic radionuclear acquisition was performed in the posterior view every 0.25 s (matrix 64 × 64) during the inhalation of ^99mTc-Technegas. Static acquisitions were performed in the posterior, anterior, and right and left posterior oblique views with a matrix size of 128 × 128; 300 kilocounts were accumulated for each image. Single-photon-emission computed tomography (SPECT) studies were done by rotating the gamma camera around the subject, who was lying supine and with the arms raised above his head. A total of 64 images (matrix 64 × 64) was acquired for 20 s each using a step-shoot method over 360° with 6° increment.

Images acquired in the dynamic phase of the study were added to obtain a single image. A small square region of interest (3 × 3 pixel) was drawn around areas of high uptake of ^99mTc-Technegas ("hot spot") and a time-activity curve was obtained. The same region of interest was drawn on the control study for comparison.

Analysis of the static planar scans was performed by use of a semiquantitative isocount method. Specifically, isocontours of the lungs were drawn according to three cutoff levels (15, 30, and 50% of maximal Technegas radioactivity) with the aid of a special software. The relevant subtended areas were estimated and will be referred to in the present paper as area 15, area 30, and area 50, respectively. The decrease in these areas after exposure to three different doses of MCh reflects the progressive and prevalent distribution of Technegas in the central regions of the lungs. Estimation of the above areas was also extended to the upper and lower zones of each lung after a horizontal line passing through the hylum was arbitrarily drawn.

Transaxial SPECT images (2-pixel slice thickness) were obtained by the filtered backprojection method using a Butterworth filter (cutoff frequency = 0.4 cycles/cm; order no. 15). No attenuation correction was used. Coronal and sagittal images were generated from the transverse slices. SPECT images were used to locate and numerate the area of high activity (hot spots) generated after MCh administration. A hot spot was a well-defined small area with an activity higher than 50% of the activity present in the surrounding field. The numbers of hot spots present in each slice of coronal, sagittal, and transverse planes were identified.

Study Design

Study 1. Four subjects were studied on 2 days in random order. On day 1, lung function was evaluated at baseline by three sets of tidal, partial, and maximal inspiratory and expiratory maneuvers. Then, Technegas was inhaled during 12 tidal breaths with the subject seated in front of a gamma camera to record the posteroanterior scans at 4 Hz for a total of 1 min. Spirogram was synchronized with the scintigraphic scans to identify inspiratory and expiratory phases and to monitor the breathing pattern (Direc 3.1, Raytech Instruments, and ANADAT 5.1, RTH InfoDat, Montreal, Canada). Two minutes later, one more set of tidal, partial, and maximal inspiratory and expiratory maneuvers was recorded. On day 2, the procedures were repeated in the same order except that Technegas inhalation was preceded 3 min earlier by inhalation of the MCh dose C determined on the screening day.

Study 2. Seven subjects were studied on 4 days in random order. On day 1, lung function was evaluated at baseline by three sets of NEP maneuvers and three sets of tidal, partial, and maximal inspiratory and expiratory maneuvers. Then, Technegas was inhaled during 12 tidal breaths with the pneumotachograph placed between the mouthpiece and the inhaler of the Technegas generator. One more set of NEP maneuvers and one of tidal, partial, and maximal inspiratory and expiratory maneuvers were measured 2 min later. On days 2, 3, and 4, the same sets of maneuvers were repeated at baseline, 2 min after one of the predetermined MCh doses (A, B, or C), and 2 min after Technegas inhalation. Soon after lung function measurements were terminated, posteroanterior, anteroposterior, and right and left oblique static scans were taken within 10 min and SPECT over the remaining 25 min of the study.

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study 1

The time-activity curve of total counts on lung windows where the hot spots were maximum on day 2 revealed substantially different deposition pattern between days (Fig. 1). On day 1, the curve was characterized by regular increments of radioactivity synchronous with inspiration and decrements with expiration. During each expiration, about half the counts recorded during the previous inspiratory phase remained in that area so that total radioactivity on the window tended to increase progressively with the number of breaths. In contrast, on day 2 the increase in radioactivity synchronous with inspiration tended to be slightly less in the same region of control, likely due to airway narrowing and consequent decreased ventilation. Nevertheless, little or no decrease was documented during the next expiratory phase, thus suggesting that enhanced deposition of Technegas in the hot spots mostly occurred during expiration.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Time-activity curves of total counts of Technegas on lung windows where the "hot spots" were maximum on day 2 in the 4 cases (A-D) of study 1. On day 1 (control) (thin lines), the curves are characterized by regular increments of radioactivity synchronous with tidal inspiration and decrements with tidal expiration. Total radioactivity of the window progressively increases with the number of breaths. Spirographic signals are not reported in the graph. On the day the airways are exposed to methacholine (MCh; thick lines), the increase in radioactivity synchronous with inspiration is followed by little or no decrease during the next expiratory phase, thus suggesting enhanced deposition of Technegas in the hot spots during expiration.

Study 2

View larger version (79K): [in this window] [in a new window]   Fig. 2.   Typical example of planar posteroanterior scans of the lungs in a subject after inhaling Technegas on day 1 (control; A) and after increasing doses of doses of MCh (B-D). Note the enhanced distribution of the Technegas in the central areas of the lungs even with the smallest dose of constrictor agent.

Data on Technegas lung deposition are presented in Table 6. Total count of the radioisotope was not significantly different within the four sessions, as expected if ventilation remains constant. There was however, a significant and progressive decrease of areas 15, 30, and 50 with airway narrowing in both the upper and lower half of the lung, suggesting enhanced deposition of the isotope in the central regions of the lung and confirming the pattern visually identified on the static and SPECT scans. Typical distribution of the hot spots in the coronal, sagittal, and transverse sections in one subject is shown in Fig. 3.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Distribution of the hot spots in the coronal (A), sagittal (B), and transverse (C) sections in the same subject of Fig. 2 with 80 (dotted lines), 400 (dashed lines), and 800 µg MCh (solid lines). The distribution of the spots is unimodal in the coronal and transverse sections and bimodal in the sagittal section, thus suggesting that the prevalent location of the spots is in the central rather than in the peripheral areas of the lungs. The decrease in number of spots with the highest dose of MCh is due to merging of small and separate spots into larger and bulgy ones. Ant, anteroposterior; Post, posteroanterior.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The main findings of the present study are that 1) EFL during tidal breathing occurred across the lung well before expiratory flow reserve was completely abolished and 2) methods based on mouth flow measurements were fairly insensitive to detect EFL occurring with induced airway narrowing even when dynamic hyperinflation was present.

Comments on Methodology

Breathing pattern during the 12 tidal breaths could have an effect on the deposition of Technegas in the airways due to inertial impaction of the particles when flow velocity increases and eddies form in the large airways (22). However, even though the plastic device connecting the Technegas chamber to the subject offers a perceptible resistance to breathing, tidal volume as well as inspiratory and expiratory flows remained substantially similar in the four sessions.

Comments on Results

As documented by the static and SPECT images and by the progressive decrease in areas 15, 30, and 50, the hot spots always formed in the same regions around the hylum and increased in number or merged with the contiguous ones with the increase of MCh dose. During forced expiration, the flow-limiting segment is known to be located first in the large intrathoracic airways and then to move peripherally although hardly ever beyond subsegmental bronchi (9). Our scintigraphic techniques could not precisely define the anatomical location of the flow-limiting segments during tidal breathing. Yet all the hot spots recorded during tidal breathing were apparently located in the large airways, as reported during cough (20, 21). We speculate that intrinsic anatomical and/or functional differences among parallel airways (13) cause EFL to manifest inhomogeneously across the lung with respect to space and time.

Flow at the mouth during forced expiration is the ultimate result of the emptying of all alveolar units. Its monotonic and fairly linear decrease with lung volume in healthy conditions is in apparent contrast with the above-mentioned structural and functional heterogeneities of the respiratory system, which would tend to desynchronize flows from contiguous sectors. Nevertheless, there are dynamic mechanisms that solidly oppose such regional flow discrepancies (13, 24). Lung emptying during tidal expiration is, on the contrary, a smooth and slow process relatively independent of lung inhomogeneities because it takes place over a time well within the expiratory time imposed by the respiratory drive. Exposing the lung to a bronchoconstrictor agent may cause variable narrowing across the airways, with at the extremes little change in caliber in some airways and severe obstruction associated with occurrence of flow limitation in others. The ensuing decrease in maximum flow is therefore due to any sort of combination of increased frictional losses upstream from the flow-limiting segment in some airways and premature occurrence of EFL in others. Comparing tidal to maximal flow in the tidal breathing range, as done in the present study, documented a progressive decrease in expiratory flow reserve but categorically let regional onset of EFL remain unnoticed until obstruction was so severe that all or most airways were flow limited and the expiratory flow reserve fell to zero. On the basis of these observations, we may depict the following possible scenario.

After exposure of the lung to low doses of a bronchoconstrictor agent, most of the airways slightly decrease caliber. Surprisingly, however, few of them already manifest exaggerated narrowing up to the point that maximal flow was achieved. The effect of the latter is trivial on maximum flow due to compensatory mechanisms opposing and limiting regional inhomogeneities. The expiratory flow reserve under these conditions is still variably present. Further doses of the bronchoactive agent narrow the range of airway response because most or all the central airways manifest dynamic collapse during tidal expiration. Under this condition, increasing expiratory effort as during forced expiration or applying a negative pressure does not produce additional flow because all central airways are flow limited. Thus the complete abolishment of expiratory flow reserve indisputably suggests that EFL has involved all lung regions. In one subject, the NEP method did not identify the presence of EFL on day 4, whereas the partial flow method did. The reason for this is simply that some individuals tend to slightly overdistend the lungs in response to EFL during a bronchial challenge. Therefore, any comparison of tidal to maximal flow that does not consider increase in FRC may miss the presence of EFL.

This study gave us the opportunity to document two other aspects of airflow obstruction with the highest dose of MCh. First, in all the cases we saw on the SPECT scans that Technegas was totally missing in an entire lobe or part of it, mostly in the dependent lung regions and when constriction was severe. In our view, this may be attributed to full airway closure. Under this condition, airway opening pressure is so high that it cannot be achieved during tidal inspiration. The finding is consistent with studies demonstrating the insufficiency of forces of interdependence between airways and surrounding lung parenchyma to prevent full closure in the airways most vulnerable to narrowing (4, 8). Second, it is well known that, when EFL occurs, FRC increases to contrast and limit airway obstruction (17). What one would expect from our study is that the increase in FRC at the transition from the intermediate to the highest dose of MCh in study 2 could counteract EFL and modulate airway narrowing. Surprisingly, we found further decrease in expiratory flow and greater numbers of hot spots, which would apparently speak against a fully protective role of lung hyperinflation during bronchoconstriction. We speculate that airway narrowing and EFL would have certainly been intolerable for the subjects and so shortness of breath would have occurred had FRC remained fixed at baseline values. Thus it appears that increased external elastic load with lung hyperinflation has strong yet limited dilator effects on airways.

A finding of practical importance was that EFL was undetected by either partial flow or NEP method when IC was decreased. In a previous study our laboratory documented that FRC increases mostly when EFL occurs (14). Others (23) found an increased FRC even when EFL could not be detected by the NEP method. They concluded that mechanisms other than EFL cause dynamic hyperinflation during induced bronchoconstriction. The present findings of a significant decrease in IC well before EFL could be detected by methods based on mouth flow measurements suggest that the reported dissociation between dynamic hyperinflation and EFL was due to the low sensitivity of the method used to assess EFL.

To further investigate the relevance of the present findings to overall lung mechanics, dynamic elastance (Edyn) was measured in a single subject during tidal breathing on days 2, 3, and 4 of study 2. As in a previous study (15), Edyn progressively increased after each MCh dose (Fig. 4) but was already increased on days 2 and 3, when IC was still unchanged. This suggests that regional EFL may contribute to increase in Edyn during mild induced bronchoconstriction. We speculate that lung regions served by airways under EFL may become overinflated, thus contributing to the increase in Edyn, as observed with the small doses of constrictor agent in the present study. This mechanism does not exclude the possibility that also undetected airway closure may contribute to increase Edyn under similar circumstances by diverting flow to overdistended lung regions. Clearly, at the highest levels of obstruction an additional increment of Edyn may occur because of airway closure, increase in FRC, and stress relaxation (18).

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Dynamic elastance (Edyn) before (B) and after (A) MCh on days 2, 3, and 4 of study 2 in a subject who accepted to have lung mechanics measured. Values are means ± SD of 8 tidal breaths. Edyn was computed by using the ANADAT 5.1 software (RTH InfoDat, Montreal, PQ) with esophageal and mouth pressures being measured by 2 identical separate differential pressure transducers (DP15, ±150 cmH2O, Validyne). Positioning of an esophageal balloon, filled with 1 ml of air, was made after topical anesthesia and was considered correct if transpulmonary pressure did not change after a gentle expiratory effort against a partially occluded port. Note the progressive increase in Edyn with the different doses of MCh.

In conclusion, the present study supports the notion that EFL during spontaneous breathing in asthma is time and space dependent. Implicit with this assertion is that EFL occurs well before the usual methods based on measurements of mouth flow can detect it.