METHODS

Section:

ChooseTop of pageAbstractMETHODS <<RESULTSDISCUSSIONReferencesCITING ARTICLES

Spirometry was obtained by means of standardized equipment (Vmax 20C; SensorMedics, Bilthoven, The Netherlands). From the forced expiratory curve, we used FEV₁, FVC, and forced expiratory flow after exhalation of 75% FVC (FEF₇₅); predicted values for FEV₁ and FEF₇₅ were based on those provided by the European Community for Steel and Coal. We also measured single breath carbon monoxide transfer factor (dl_CO) and specific airway conductance (sGaw; Models 2200 and 6200; SensorMedics). A Micro Smokerlyzer (Bedfont Scientific, Rochester, UK) was used to monitor exhaled CO for a double-check of self-reported smoking cessation.

The N₂ MBW test—that is, 20 to 25 inspirations of 1 L pure O₂, starting from functional residual capacity with continuous monitoring of N₂ concentration and volume—was performed according to standardized methods extensively reported elsewhere (8, 10, 11). For the theory of MBW N₂ phase III slope analysis leading to indices of conductive (Scond) and acinar (Sacin) ventilation heterogeneity, see the online supplement. It implies that Scond and Sacin are intrinsically independent and not comparable (11). Because Scond and Sacin are derived from phase III slopes, their value increases when ventilation heterogeneity increases. In particular, Sacin will increase if an alteration of the intraacinar asymmetry occurs—for example, by differential airway wall thickening of any two daughter branches. Scond will increase if the conductive airways and their subtended units show an increase in heterogeneity of specific ventilation or flow asynchrony (such that the best-ventilated units empty preferentially early in expiration).

RESULTS

Section:

ChooseTop of pageAbstractMETHODSRESULTS <<DISCUSSIONReferencesCITING ARTICLES

Data reported here are lung function and ventilation heterogeneity parameters obtained after dilation on each visit in four successively smaller subgroups of subjects who remained abstinent from smoking for at least 1 wk (1-wk subgroup, n = 57), 3 mo (3-mo subgroup, n = 28), 6 mo (6-mo subgroup, n = 24), and 12 mo (12-mo subgroup, n = 18). The proportion of subjects receiving nicotine replacement therapy or bupropion was not significantly different among the four subgroups (p > 0.1). The characteristics of all subject subgroups are provided in Table 1, showing no statistical differences in any of the displayed parameters; note the substantial intersubject variability in most of these parameters. Although intersubject variability does not matter for this longitudinal study, we did include a control group of persistent smokers to double-check the absence of longitudinal intrasubject variability for these measurements in this particular type of study population. The 16 persistent smokers (6 male, 10 female) who participated in all visits (4 of 20 did not return for the 12-mo visit) were generally younger (32 ± 9 [SD] yr) and had a shorter smoking history (13 ± 8 [SD] pack-years) than the smoking cessation participants.

Table 2 shows all lung function and ventilation heterogeneity parameters that have been shown to be meaningful for the characterization of smokers (8). The persistent smoker group (n = 16) did not show significant changes in any of the parameters in Table 2. For smoking cessation participants in the 1-wk subgroup (n = 57), significant but marginal improvements were observed in FEV₁ (+1% predicted) and dl_CO (+2% predicted) after 1 wk. A marked decrease (i.e., improvement) in Sacin was also seen after 1 wk, but this improvement no longer reached statistical significance in the 3-mo, 6-mo, or 12-mo subgroups. On the other hand, Scond showed a marked decrease (i.e., improvement) after only 1 wk, and this improvement persisted across all subgroups, that is, for sustained smoking cessation up to 1 yr of study.

Figures 1A and 1B graphically represent the time evolution (in log scale) of the indices of small airway ventilation heterogeneity in the conductive and acinar lung zone. Using the mean values for Scond and Sacin recently obtained from a group of 63 normal never-smokers without bronchial hyperresponsiveness to histamine (Scond = 0.028 ± 0.007 [SD] · L⁻¹; Sacin = 0.072 ± 0.025 [SD] · L⁻¹) (8), the cut-offs for abnormality (i.e., means plus 1.96 times SD) were established at Scond = 0.042 · L⁻¹ and Sacin = 0.12 · L⁻¹ (shaded areas in Figures 1A and 1B). As can be appreciated from Figure 1, the smokers from the 12-mo subgroup (n = 18; closed circles) showed a similar evolution pattern of ventilation heterogeneity with smoking cessation time compared with the larger subgroups that had resumed smoking before the end of the study (open circles). For instance, the change in Scond seen after only 1 wk in the 1-wk subgroup (n = 57) is nearly indistinguishable from that observed in the 3-mo, 6-mo, and 12-mo subgroups (n = 28, 24, and 18, respectively).

Figure 1. (A) Time evolution of conductive ventilation heterogeneity (Scond; mean ± SEM) with smoking cessation in the 1-wk (n = 57), 3-mo (n = 28), and 6-mo (n = 24) subgroups (open circles), and in the 12-mo subgroups (n = 18; closed circles). The x axis is in log scale and the y axis starts at the average normal Scond value (0.028 · L⁻¹); the shaded zone up to 1.96 SD above the average delimits an area of Scond normality. Statistical significance of Scond changes at the different time points can be found in Table 2. (B) Time evolution of acinar ventilation heterogeneity (Sacin; mean ± SEM) with smoking cessation in the 1-wk (n = 57), 3-mo (n = 28), and 6-mo (n = 24) subgroups (open circles), and in the 12-mo subgroups (n = 18; closed circles). The x axis is in log scale and the y axis starts at the average normal Sacin value (0.072 · L⁻¹); the shaded zone up to 1.96 SD above the average delimits an area of Sacin normality. Statistical significance of Sacin changes at the different time points can be found in Table 2.

Download Figure (Part 1) | Download Figure (Part 2) | Download Powerpoint

[More] [Minimize]

Figure 1A shows that the average Scond was above the line of Scond abnormality and fell below it after smoking cessation. Considering that normal Scond is 0.028 · L⁻¹, the Scond decrease from an average of 0.046 · L⁻¹ (baseline) to 0.040 · L⁻¹ (1-wk cessation) in the 1-wk subgroup (n = 57) represents a 33% reduction (= [0.046 – 0.040]/[0.046 – 0.028]) of baseline conductive airway abnormality. The corresponding changes in the 12-mo subgroup (n = 18) were 30 and 42% reduction in Scond abnormality after 1 wk and 12 mo, respectively. Considering a normal Sacin of 0.072 · L⁻¹ (Figure 1B), similar computation yielded a 40% reduction in the acinar airway abnormality in the 1-wk subgroup after only 1 wk (n = 57), but this improvement was less consistent across all subgroups, and did not persist for longer smoking cessation periods (Table 2).

DISCUSSION

Section:

ChooseTop of pageAbstractMETHODSRESULTSDISCUSSION <<ReferencesCITING ARTICLES

Conductive airway malfunction is one of the earliest signs of smoking-induced small airway abnormality, typically observed in smokers with smoking histories greater than 10 pack-years and preceding the appearance of spirometric airflow obstruction (8). The present study shows that conductive airway abnormality could be improved by approximately one-third after only 1 wk of smoking cessation. In addition, this improvement was sustained, and after 1 yr, a 42% reduction of conductive airway abnormality was obtained in the smokers under study. Given that neither FEV₁ nor FEF₇₅ showed any change, we conclude that the sustained Scond decreases seen here reflect a functional recovery of the smallest airways that are still positioned in the conductive lung zone. Of all parameters in Table 2, Scond, Sacin, and dl_CO were the ones that had been previously shown to pick up early signs of smoking-induced airway changes from 10 pack-years onwards (8). Nevertheless, the two indices related to the alveolar compartment, dl_CO and Sacin, either showed a minor recovery (dl_CO) or a transient one (Sacin) with smoking cessation up to 1 yr. This indicates a more obstinate smoking-induced damage of small airways in the acinar lung zone.

Few published reports exist about the fate of small airway function or inflammation after smoking cessation in smokers without airway obstruction. Even in smokers with FEV₁/FVC around the 70% cut-off for obstruction, FEV₁/FVC was not necessarily related to inflammatory markers (4). Nevertheless, Swan and colleagues (4) showed that macrophages recovered from induced sputum markedly decreased after only 1 mo and the decrease was maintained after 12 mo of smoking cessation. These authors pointed out that this prompt cellular response contrasted with the slow changes in conventional lung function parameters in their study. The quick response of small airway ventilation heterogeneity seen in the present study may well be the functional equivalent of inflammatory changes Swan and colleagues observed after smoking cessation. Three decades ago, Buist and colleagues (7) had already shown the potential of ventilation heterogeneity to monitor the smoker's lung, observing marked decreases in the single breath phase III slope after 1 mo of smoking cessation. However, from these phase III slopes, it is impossible to derive in which lung zones these improvements in ventilation heterogeneity take place.

The respective patterns of change in the conductive and acinar compartments observed here with Scond and Sacin parallel those found in the airways and alveolar NO production derived from exhaled NO measurements reported by Hogman and colleagues (13). In that study, 4 wk of smoking cessation in smokers with FEV₁ and dl_CO values comparable to ours resulted in an increase of the abnormally low airway NO to normal values and no change in the alveolar NO. Regardless of the exact reason for the abnormally low NO in the airway compartment of the smokers versus never-smokers, it is the airway and not the alveolar space that appears to most consistently respond to smoking cessation. On the border between conductive and acinar lung compartments, respiratory bronchiolitis has been observed in all current smokers (ranging from 6 to 200 pack-years) and in approximately 50% of ex-smokers (ranging from 5 to 180 pack-years) without any correlation between occurrence of respiratory bronchiolitis and number of pack-years (14). Although obtained in a cross-sectional study with a very wide range of smoking histories, these pathology data indicate that there is a potential for repair but they also point to an irreversible component in the damage of these small airways situated around the entrance to the gas-exchanging lung zone.

In the acinar space, inflammatory processes with different time constants and functionalities appear to be at work (15, 16). For instance, Skold and colleagues (15) found that cells recovered from bronchoalveolar lavage fluid of smokers (with ∼ 23 pack-years smoking history) decreased consistently after 1 mo of smoking cessation, but that considerable fluorescent tobacco material trapped inside alveolar macrophages persisted in the alveolar space for up to 15 mo. In another study, Skold and colleagues (16) showed transient changes in accumulated extracellular matrix compounds such as albumin and hyaluronan in bronchoalveolar lavage fluid within the first few months of smoking cessation, with a return to baseline for longer smoking cessation periods up to 15 mo. A direct relationship between any of these separate inflammatory processes and a cumulative parameter of acinar airway function such as Sacin is unlikely. Rather, the transient pattern of Sacin recovery after 1 wk and subsequent return to baseline abnormality may reflect the presence of different underlying inflammatory damage and repair processes with different time constants.

We conclude that small airway function as measured by MBW-derived parameters is abnormal in smokers without spirometric evidence of airway obstruction, and that these abnormalities in small airway ventilation heterogeneity can be partly reversed after smoking cessation. This recovery already occurs after 1 wk and persists in the conductive lung zone, more specifically in the airways peripheral to the small airways sampled by spirometric endexpiratory flows. Such findings of small airway improvement could be used in the encouragement of the so-called asymptomatic smoker to start or maintain smoking cessation. A recent Cochrane database systematic review (17) judged existing studies of insufficient quality to be conclusive about the effectiveness of biomedical risk assessment such as genetic susceptibility to positively influence smoking cessation. Although measurement of abnormal small airway function at baseline could provide another index of biomedical risk, its improvement as smoking cessation follow-up proceeds could provide a positive stimulus to maintain adherence to smoking cessation, as was shown when ex-smokers were confronted with their improving inflammatory markers during smoking cessation follow-up (4). Finally, the noninvasive measures of peripheral lung function presented here provide clues as to how deep in the lungs and in what time frames, other, possibly more invasive techniques, could monitor in situ pathology during smoking cessation protocols.