Longitudinal Follow-up Study of Smoker’s Lung with Thin-Section CT in Correlation with Pulmonary Function Tests1

¹ From the Department of Radiology, Calmette University Center Hospital, Blvd Jules Leclerc, 59037 Lille, France (M.R.J., J.R., I.M.); the Medical Research Group, Equipe d’Accueil no. 2682, Lille, France (M.R.J., J.R.); and the Environmental and Occupational Health and Ergonomics Research Center, University of Lille, France (J.L.E., C.B., A.S.). From the 2000 RSNA scientific assembly. Received June 30, 2000; revision requested August 28; revision received April 23, 2001; accepted May 22. 

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To evaluate thin-section computed tomography (CT) in depicting longitudinal changes in the lung parenchyma.

MATERIALS AND METHODS: One hundred eleven volunteers underwent sequential examination with thin-section CT and pulmonary function tests over a mean period of 5.5 years. According to their smoking habits between initial evaluation (T0) and follow-up (T1), the subjects were classified as persistent current smokers (n = 57), persistent nonsmokers (n = 31), persistent ex-smokers (n = 13), or quitters (n = 10).

RESULTS: Significant differences in CT findings between T0 and T1 were seen in only the group of persistent current smokers, who showed a higher frequency of emphysema (40% vs 26%; P = .005) and ground-glass attenuation (42% vs 28%; P = .02). Individual analysis of follow-up CT scans in the 19 persistent current smokers with micronodules at T0 demonstrated (a) no changes in seven cases, (b) a higher profusion of micronodules in seven cases, and (c) replacement of micronodules with emphysema in five cases. Subjects with emphysema and/or areas of ground-glass attenuation at T0 had a significantly more rapid decline in lung function than did those with a normal CT scan.

CONCLUSION: Emphysema and/or ground-glass attenuation are linked with impairment of ventilatory lung function over time in persistent current smokers.

　

Index terms: Bronchiolitis, 60.2191 • Computed tomography (CT), high resolution, 60.12118 • Emphysema, 60.75 • Lung, diseases, 60.75, 60.2191 • Lung, function, 60.75, 60.2191

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Smoking-induced lung disease is a complex group of disorders, varying from the well-known entity of chronic obstructive pulmonary disease to more recently described interstitial lung diseases (1,2). The obstructive ventilatory defect that defines chronic obstructive pulmonary disease reflects the presence of varying degrees of lung parenchymal destructive changes and airway inflammation. Different degrees of severity of small-airway and parenchymal reaction to cigarette smoke characterize several diseases strongly linked to tobacco consumption, that is, respiratory bronchiolitis-associated interstitial lung disease and desquamative interstitial pneumonia.

Whereas the computed tomographic (CT) features of these disorders have been extensively described in the literature, little attention has been paid to the recognition of tobacco-induced lesions in asymptomatic subjects. The authors of two preliminary studies (3,4) have demonstrated that parenchymal micronodules, indicating the presence of respiratory bronchiolitis, and areas of ground-glass attenuation, suggesting an accumulation of inflammatory cells and fibrosis in respiratory bronchioles and alveolar spaces, can be depicted on thin-section CT scans. These findings help support the concept that parenchymal abnormalities can be detected in healthy smokers with normal findings at chest radiography and pulmonary function testing. Recognition of these abnormalities by using a noninvasive morphologic tool opens the field of correlative studies between morphology and function in smokers, the latter having been extensively investigated over the past 30 years. Therefore, the purpose of this study was to evaluate thin-section CT in depicting morphologic and functional changes over time induced by smoking habits in a population of healthy volunteers.

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Subjects
Candidates for this study were drawn from a population of 175 healthy adult volunteers who initially participated in a prospective evaluation of lung changes detectable on CT scans in smokers. These volunteers, all urban dwellers, were recruited from hospital workers in our institution. The details of subject selection for this study have been described in a previous study (3). All subjects from the baseline study were asked to participate in this follow-up study at the time of their annual health-care evaluation after a minimum follow-up period of 4 years.

From the initial population, 62 subjects (47 women and 15 men) were lost to follow-up because they (a) did not want to be included in the present study (n = 30), (b) could not undergo the follow-up investigation because they had moved to a different work site (n = 25), or (c) had developed a chronic (nonrespiratory) disease and could not be considered healthy subjects (n = 7). The subjects lost to follow-up included 35 smokers, 11 ex-smokers and 16 nonsmokers with a mean age (±SD) of 31 years ± 6.9. This subgroup of subjects did not differ from the followed-up population in sex ratios; proportions of smokers, nonsmokers, and ex-smokers; or in their clinical, functional, and CT characteristics at the time of initial evaluation.

The follow-up study population consisted of 113 subjects. Of these subjects, two were not included in the follow-up study, since they had modified their smoking habits several times between the time of initial evaluation (T0) and the time of follow-up evaluation (T1). Consequently, the final study group comprised 111 subjects (69 women and 42 men) who had not modified their smoking habits during the follow-up period and ranged in age from 27 to 59 years (mean, 35 years ± 6.5), with a mean follow-up period of 5.5 years (range, 4.5–7.5 years). At T0, the group comprised 67 current smokers (subjects who had smoked regularly for more than 5 years), 13 ex-smokers (subjects who had not smoked for more than 2 years), and 31 nonsmokers.

According to their smoking habits, the 111 members of the final study group were classified into one of four categories: (a) current smokers at T0 who did not stop smoking between T0 and T1 (n = 57), further referred to as the "persistent current smokers"; (b) nonsmokers at T0 who remained nonsmokers at T1 (n = 31), further referred to as the "persistent nonsmokers"; (c) ex-smokers at T0 who did not smoke between T0 and T1 (n = 13), further referred to as the "persistent ex-smokers"; or (d) current smokers at T0 who gave up smoking between T0 and T1 (n = 10), further referred to as the "quitters."

Clinical evaluation at T0 and T1 included answering a standardized questionnaire, modified from one used by the British Occupational Hygiene Society Committee on Hygiene Standards (5), and a physical examination. Pulmonary function testing and thin-section CT were performed within 3 weeks after the clinical examination (range, 1 day–3 weeks; mean interval, 10 days). Because the purpose of this study was to evaluate tobacco-related lung changes by using thin-collimated scans, care was taken to screen the entire lung for any associated anomalies on chest radiographs, for example, posteroanterior and lateral views, systematically obtained at the time of CT (screen-film examinations). Radiographs were read during daily work-up by a senior chest radiologist not involved in the present investigation and revealed no abnormalities in any of the subjects.

Pulmonary Function Tests
At T0 and T1, pulmonary function tests were performed to obtain flow-volume curves. Maximum expiratory flow-volume curves were obtained by measuring flow with a pneumotachygraph (Spiroanalyser ST-300; Fukuda, Nagareyama, Japan). The calibration of the pneumotachygraph, the maneuvers performed, and the selection of curves met the American Thoracic Society’s guidelines (6,7).

The following parameters were evaluated: forced vital capacity (FVC); forced expiratory volume in 1 second (FEV₁); ratio of FEV₁ to FVC, or FEV₁/FVC; maximal expiratory flow (MEF); MEF at 75%, 50%, and 25% of FVC (MEF 75, MEF 50, and MEF 25); maximal midexpiratory flow (MMEF); MEF between 75% and 25% of FVC (MEF 75-25), and MEF at the end of expiration (MEF 25–15). All spirometric values were expressed as a ratio of measured to predicted values. Prediction equations have previously been given by Knudson et al (8,9) for all parameters except MEF 25-15, which were obtained from Morris (10). Because of the technical limitations of the apparatus used in this protocol, diffusing capacity was not measured. The variation of ventilatory function between T0 and T1 was calculated from ratios of difference between T0 and T1 to predicted values, as recommended by Dales et al (11), divided by the follow-up duration (eg, 100 x FEV₁ [T1] - FEV₁ [T0]/FEV₁ predicted, divided by the follow-up duration). The results of changes in lung function were expressed as percentages.

CT Studies
Technique.—Thin-section CT scans were obtained at T0 and T1, with a 1-mm section thickness, 15-mm intervals, a 512 x 512 reconstruction matrix, and a high-spatial-frequency algorithm. The intersection spacing was that selected in the preliminary investigation (3), then adapted to limit the overall number of sections obtained in each examination combining thin- and thick-collimated scans. At T0, CT was performed either with a model 2400 (Elscint, Hackensack, NJ), with 130 kV, 420 mA, and a 2-second scanning time; or with a Somatom Plus (Siemens, Erlangen, Germany), with 137 kV, 255 mA, and a 1-second scanning time). CT examinations at T1 were performed with a Somatom Plus 4 A (Siemens), with 137 kV, 206 mA, and a 0.75-second scanning time. At T0, thin-section CT was targeted to the right lung by using a 240-mm field of view; follow-up CT scans were systematically reconstructed on a 350-mm field of view. All subjects underwent scanning in the supine position, at deep inspiration. All images were obtained at window levels appropriate for lung parenchyma at thin-section CT (window width, 1,600–1,800 HU; window center, -600 HU).

Analysis.— Thin-section CT scans were interpreted independently and in random order by two observers (M.R.J., J.R.) without knowledge of the subject’s smoking habits; in case of discordant reading, a consensus analysis was systematically obtained at second intention. For each subject, thin-section CT scans obtained at T0 were read first, followed by a simultaneous reading of thin-section CT scans obtained at T0 and T1 to analyze changes over time. Because thin-section CT scans obtained at T0 consisted of targeted images of the right lung, the analysis of lung changes over time was limited to the interpretation of right-sided images. Specific abnormalities evaluated included the following: (a) ill-defined micronodules (<7 mm in diameter); (b) ground-glass opacity, defined by the presence of slightly hyperattenuating areas in which underlying vessels and bronchial walls remained visible, suggestive of smoker’s alveolitis when observed in the nondependent lung; (c) emphysema, including centrilobular emphysema, defined as areas of hypoattenuation, surrounded by discernable interlobular septa; bullous emphysema, defined as regions of decreased attenuation and vascular disruption surrounded by a wall less than 1–2 mm thick; and nonbullous emphysema, defined as regions of decreased attenuation without definable walls. Because of the subjective analysis of bronchial wall thickness and its variability on images obtained with different scanners, this parameter was not included in the present study. In addition, we did not include the evaluation of dependent lung attenuation on the two series of CT scans owing to the lack of spirometrically gated CT examinations at T0 and T1.

The percentage of lung involved with micronodules, ground-glass opacity, and/or emphysema at T0 and T1 was determined by using the following grading system, enabling the definition of an extent score for each anomaly. For that purpose, the right lung was divided into three areas: the upper lung zone, above the level of the carina; the middle lung zone, between the level of the carina and the level of the right inferior pulmonary vein; and the lower lung zone, below the level of the right inferior pulmonary vein. Each zone was assigned a score for each anomaly (a score of 0 indicated no anomaly; a score of 1 indicated an anomaly involving less than 25% of the lung surface; 2, 25%–50% of the lung surface; 3, 50%–75% of the lung surface; or 4, more than 75% of the lung surface). The scores of the three zones were added for a total score of 0–12. A consensus score was obtained for each patient. In addition to determining extent scores, simultaneous reading of CT scans obtained at T0 and T1 led the readers to analyze changes in micronodular profusion; on the CT scans obtained at T1, the profusion was coded as similar, greater, or lower compared with that observed on the images obtained at T0.

Statistical Analysis
Statistical analysis was performed with SAS software (SAS, Cary, NC) and a microcomputer (XPS P 166; DELL, Austin, Tex). Interobserver agreement for CT findings was determined by using the statistic. The findings were described, and the means, SDs, and frequencies were calculated. Comparisons of proportions between current smokers, ex-smokers, and nonsmokers were performed with the ² test or Fisher exact test when tables had expected values of less than five. Changes in proportions between T0 and T1 were evaluated with the McNemar test. Differences between initial and follow-up values of functional parameters were evaluated with the Wilcoxon signed rank test.

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Clinical, Functional, and CT Findings at T0
The characteristics of the study population are reported in Table 1. Apart from a significantly higher mean age among ex-smokers, the three groups did not significantly differ in height or weight (P > .05). Among the largest subgroups, namely, the subgroups of smokers (n = 67) and nonsmokers (n = 33), there was a higher percentage of female than of male subjects. These data reflect the sex ratio in the working population asked to participate in the study.

fig.ommitted TABLE 1. Subject Characteristics at T0

 
Table 2 summarizes the clinical data obtained from the questionnaire for the three groups studied. There was a significant correlation between current cigarette consumption and morning cough (P < .01) and wheezing (P < .001). There was a significantly higher percentage of dyspnea in smokers and ex-smokers than in nonsmokers (P = .02). It is noticeable that a few of the nonsmokers had clinical symptoms. Results of pulmonary function tests are summarized in Table 3, showing overall normal pulmonary function in the three groups studied.

fig.ommitted TABLE 2. Clinical Findings at T0

 

fig.ommitted TABLE 3. Results of Pulmonary Function Tests at T0

 
Thin-section CT scans were normal in 73 subjects (66% of the study group, including 31 current smokers, 11 ex-smokers, and 31 nonsmokers) and showed abnormalities in 38 subjects (34% of the study group, including 36 smokers and two ex-smokers). Detailed results of thin-section CT according to smoking habits are summarized in Table 4, showing a significantly higher frequency of lung abnormalities among current smokers and ex-smokers compared with nonsmokers. The mean cigarette consumption of the 36 current smokers with abnormal thin-section CT scans was significantly higher than that of the 31 current smokers with normal thin-section CT scans, namely 16.0 pack-years ± 7.9 versus 11.0 pack-years ± 7.8, respectively (P < .001). The interobserver agreement was 0.86 for micronodules, 0.95 for emphysema, and 0.87 for ground-glass attenuation.

fig.ommitted TABLE 4. Thin-Section CT Findings at T0

 
Evolution between T0 and T1
The mean interval between T0 and T1 was 5.5 years (range, 4.5–7.5 years). In the group of persistent current smokers, the mean cigarette consumption was 14.0 pack-years ± 8.2 at T0 and 19.0 pack-years ± 5.4 at T1. In the group of quitters, the mean cigarette consumption at T0 was 11.0 pack-years ± 8.1, and the mean duration of tobacco cessation at T1 was 2.4 years ± 1.6. In the group of persistent ex-smokers, the mean duration of smoking cessation was 4.0 years ± 2.5 at T0 and 10 years ± 2.5 at T1. As detailed in Table 5, the clinical presentation at T1 did not significantly differ from that observed at T0 in the four groups. Analysis of the mean annual evolution of lung function showed significant annual decline for all parameters except FVC, without significant differences between persistent current smokers, quitters, persistent ex-smokers, and persistent nonsmokers. However, a trend toward a more rapid decline in lung function was observed in the persistent current smokers and quitters.

fig.ommitted TABLE 5. Evolution of Clinical Findings between T0 and T1

 
Significant differences in thin-section CT findings between T0 and T1 were found only in the persistent current smokers (Table 6). The frequency of areas of ground-glass attenuation was significantly higher at T1 (n = 24 [42%]) compared with that at T0 (n = 16 [28%]; P = .021) without any significant difference between the mean extent scores (mean score at T0, 2.5 [range, 0–12] vs 2.9 [range, 0–12] at T1; P = .337). The areas of ground-glass attenuation were observed in the upper lung zones (at T0 in 13 [81%] of 16 subjects; at T1 in 19 [79%] of 24 subjects) and in the upper, middle, and lower lung zones (at T0 in three [19%] of 16 subjects; at T1 in five [21%] of 24 subjects).

fig.ommitted TABLE 6. Thin-Section CT Changes in the Group of 57 Persistent Current Smokers

 
The frequency and mean extent scores of emphysema were significantly higher at T1 compared with T0 (mean extent score at T0, 1.6 [range, 1–4]; at T1, 2.6 [range, 1–8]; P < .005). At T1, emphysematous changes were detected in eight subjects devoid of emphysema at T0, including bullous (five of eight subjects) and/or nonbullous (five of eight subjects) emphysema, in the upper lung zones (five of eight subjects) or in the upper and middle lung zones (three of eight subjects).

Among the 15 subjects in whom emphysema was present at T0, CT findings at T1 were as follows: No changes in the morphologic categories of emphysema (bullous [six of eight subjects]; nonbullous [six of eight subjects] and/or centrilobular [two of eight subjects]) nor in the extent scores (mean scores at T0 and T1, 1.5 [range, 1–3]) were observed in eight of the 15 subjects; these subjects showed emphysematous alterations exclusively in the upper lung zones (five of eight subjects) or in the upper, middle, and lower lung zones (three of eight subjects). In seven of the 15 subjects in whom emphysema was present at T0, changes in the extent score, location, and/or morphologic features of emphysema were observed: (a) the mean extent score was 5.1 [range, 4–8]) at T1 versus 1.8 [range, 1–4] at T0; (b) at T0, emphysema was located in the upper lung zones (five of seven subjects) or in the upper, middle, and lower lung zones (two of seven subjects); at T1, emphysema was observed in the upper, middle, and lower lung zones in every subject; (c) the morphologic CT features at T0 included nonbullous (six of seven subjects) and/or bullous (four of seven subjects) emphysema; at T1, CT findings included nonbullous (six of seven subjects), bullous (five of seven subjects), and/or centrilobular (three of seven subjects) emphysema (Fig 1).

fig.ommitted   Figure 1a. CT scans from a 6-year longitudinal study in a 29-year-old persistent current smoker (23.0 pack-years at T0; 29.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows areas of hypoattenuation (arrowheads) disseminated throughout the upper lung zone, suggestive of emphysematous lesions. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows extensive emphysematous changes (arrowheads) in both lungs and more prominent septal lines. Note the long posterior junction line (arrow), suggestive of lung distention.

 

fig.ommitted Figure 1b. CT scans from a 6-year longitudinal study in a 29-year-old persistent current smoker (23.0 pack-years at T0; 29.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows areas of hypoattenuation (arrowheads) disseminated throughout the upper lung zone, suggestive of emphysematous lesions. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows extensive emphysematous changes (arrowheads) in both lungs and more prominent septal lines. Note the long posterior junction line (arrow), suggestive of lung distention.

 
No significant difference in the frequency and extent score of ill-defined micronodules was found between T0 and T1 (mean score at T0, 1.45 [range, 0–10]; mean extent score at T1, 1.35 [range, 0–9]) (P = .506), located in the upper lung zones (T0, 15 [79%] of 19; T1, 16 [80%] of 20), or in the upper, middle, and lower lung zones (T0, four [21%] of 19; T1, four [20%] of 20). Individual analysis of the follow-up CT scans of the 19 subjects with micronodules at T0 demonstrated: (a) no changes in seven subjects; (b) a higher profusion of ill-defined micronodules in seven subjects (Figs 2, 3), and (c) a lower extent score of micronodules in five subjects (mean extent score at T0, 5.52; at T1, 2.45), with the subsequent replacement of the micronodular pattern by emphysematous changes (Figs 4, 5).

fig.ommitted Figure 2a. CT scans from a 6-year longitudinal study in a 27-year-old persistent current smoker (9.0 pack-years at T0; 17.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) and areas of mild ground-glass attenuation in the central and peripheral parts of the upper lung zone. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows numerous ill-defined micronodules (arrowheads) diffusely distributed throughout both lungs.

 

fig.ommitted Figure 2b. CT scans from a 6-year longitudinal study in a 27-year-old persistent current smoker (9.0 pack-years at T0; 17.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) and areas of mild ground-glass attenuation in the central and peripheral parts of the upper lung zone. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows numerous ill-defined micronodules (arrowheads) diffusely distributed throughout both lungs.

 

fig.ommitted Figure 3a. CT scans from a 7-year longitudinal study in a 33-year-old persistent current smoker (7.5 pack-years at T0; 16.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) in the upper lung zone. (b) Transverse thin-section CT scan obtained at the same level as in a at T1 shows numerous nodular areas of increased attenuation (arrowheads), some depicted with a patent centrilobular distribution and more prominent septal lines.

 

fig.ommitted   Figure 3b. CT scans from a 7-year longitudinal study in a 33-year-old persistent current smoker (7.5 pack-years at T0; 16.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) in the upper lung zone. (b) Transverse thin-section CT scan obtained at the same level as in a at T1 shows numerous nodular areas of increased attenuation (arrowheads), some depicted with a patent centrilobular distribution and more prominent septal lines.

 

fig.ommitted   Figure 4a. CT scans from a 5-year longitudinal study in a 31-year-old persistent current smoker (15.0 pack-years at T0; 21.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) in the upper lung zone. Arrows points to subpleural emphysematous area. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows replacement of the micronodular pattern (arrowheads) observed in a in the peripheral part of the anterior segment by small areas of hypoattenuation indicative of emphysematous changes, also observed in the central part of the lung. No change in the subpleural emphysematous area (arrow) is seen.

 

fig.ommitted Figure 4b. CT scans from a 5-year longitudinal study in a 31-year-old persistent current smoker (15.0 pack-years at T0; 21.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) in the upper lung zone. Arrows points to subpleural emphysematous area. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows replacement of the micronodular pattern (arrowheads) observed in a in the peripheral part of the anterior segment by small areas of hypoattenuation indicative of emphysematous changes, also observed in the central part of the lung. No change in the subpleural emphysematous area (arrow) is seen.

 

fig.ommitted   Figure 5a. CT scans from a 6-year longitudinal study in a 34-year-old persistent current smoker (30.0 pack-years at T0; 38.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules in the upper lung zones. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows replacement of the micronodular pattern by numerous small areas of hypoattenuation, indicative of emphysematous changes.

 

fig.ommitted Figure 5b. CT scans from a 6-year longitudinal study in a 34-year-old persistent current smoker (30.0 pack-years at T0; 38.0 pack-years at T1). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules in the upper lung zones. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows replacement of the micronodular pattern by numerous small areas of hypoattenuation, indicative of emphysematous changes.

 
Owing to the small number of subjects in the group of quitters, CT changes between T0 and T1 were analyzed individually. At T0, emphysema was depicted in one subject with a similar pattern and extent at T1 (mean extent score at T0 and T1, 1). Among the nine quitters in whom no features of emphysema were depicted at T0, none showed newly developed emphysematous changes at T1. At T0, areas of ground-glass attenuation were identified in four subjects, with a mean extent score of 9.25 (range, 3–12). In one subject, a lower extent score of ground-glass attenuation was observed at T1 (extent score at T0, 10; extent score at T1, 3), whereas ground-glass attenuation was not depicted at T1 in the remaining three subjects. At T0, ill-defined micronodules were found in two subjects. At T1, a similar micronodular pattern was observed in the first subject (extent score at T0 and T1, 3), whereas a lower extent score of micronodules was found in the second subject (extent score at T0, 5; extent score at T1, 2) (Fig 6).

fig.ommitted Figure 6a. CT images from a 7-year longitudinal study in a 41-year-old quitter (20.0 pack-years at T0). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) in the upper lung zones. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows considerable reduction in the profusion of ill-defined micronodules (arrowheads).

 

fig.ommitted Figure 6b. CT images from a 7-year longitudinal study in a 41-year-old quitter (20.0 pack-years at T0). (a) Transverse thin-section CT scan targeted to the right lung at T0 shows faint parenchymal micronodules (arrowheads) in the upper lung zones. (b) Transverse thin-section CT scan obtained at the same level as a at T1 shows considerable reduction in the profusion of ill-defined micronodules (arrowheads).

 
Functional and CT Correlations
No significant difference was found at T0 between the mean values of functional parameters of the subjects with normal CT scans compared with those with CT abnormalities. Significant relationships were found between CT abnormalities at T0 and an abnormally rapid decline in functional parameters between T0 and T1: (a) Subjects with emphysema at T0 (n = 17) had a more rapid decline in FEV₁ (P = .012), MEF 75–25 (P = .004), MEF 25–15 (P = .035), and MEF 50 (P = .003) between T0 and T1 than did those without emphysema (n = 94). (b) Subjects with areas of ground-glass attenuation at T0 (n = 22) had a more rapid decline in FEV₁ (P = .001), MEF 75–25 (P = .001), MEF 25–15 (P = .008), MEF 50 (P = .001), and MEF 25 (P = .006) between T0 and T1 than did subjects without areas of ground-glass attenuation (n = 89). When ill-defined micronodules were depicted at T0, a trend toward a more rapid decline in functional parameters also was observed, without statistically significant differences (FEV₁, P = .99; MEF 75–25, P = .73; MEF 25–15, P = .80; MEF 50, P = .41).

Between T0 and T1, no changes were depicted at CT in 82 of the 111 subjects, including 65 subjects with a normal CT scan and 17 subjects with lung abnormalities. In the 29 remaining subjects, lung changes were depicted with a greater frequency and/or extent. Subjects with a worsening of CT abnormalities between T0 and T1 had a greater cigarette consumption and lower functional parameters at T0 compared with those with stable CT findings (mean cigarette consumption, 15.53 pack-years ± 9.73 vs 7.90 pack-years ± 8.26, P = .003) (FEV₁, 0.98 ± 0.13 vs 1.06 ± 0.13; P = .017; MEF 75-25, 0.89 ± 0.30 vs 1.04 ± 0.25; P = .021; MEF 25-15, 0.93 ± 0.47 vs 1.11 ± 0.36; P = .008; MEF 25, 0.83 ± 0.35 vs 0.99 ± 0.31; P = .018).

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Despite the small size of the several subgroups of subjects followed up in the present investigation, this study provides morphologic information on longitudinal changes in CT findings caused by smoking. The study group was composed of 111 healthy adult volunteers who consented to examination with thin-section CT, as well as pulmonary function testing on two occasions over a mean follow-up period of 4.5 years. The largest group consisted of 57 subjects who did not give up smoking between T0 and T1. In this subgroup, the most notable finding was a worsening of lung parenchymal abnormalities related to cigarette smoking, including emphysema, areas of ground-glass attenuation, and ill-defined micronodules.

With regard to longitudinal changes in destructive lung lesions, follow-up CT depicted more extensive emphysematous alterations in seven of the 15 subjects in whom emphysema was depicted at T0. In addition, CT features of emphysema were identified at T1 in eight subjects devoid of emphysema at T0. Whatever was depicted at T0 and T1, emphysematous changes were seen to predominantly affect the upper lung zones. These morphologic findings confirm results recently reported by Soejima et al (12) regarding longitudinal changes in quantitative CT parameters in current smokers. Measuring annually over a 5-year period the relative area of low attenuation with CT values less than -912 HU, these authors observed that the annual increase in this parameter in the upper lung field was greater for the current-smoking group than for the nonsmoking group. We did not observe major morphologic changes in emphysematous alterations among subjects who remained smokers during the follow-up period. It is noticeable that nonbullous emphysema was the most frequently depicted CT feature, identified in 73% of smokers at T0 and in 78% at T1.

The second notable finding among persistent current smokers was a significantly higher frequency of areas of ground-glass attenuation at T1 compared with T0. Moreover, a higher profusion of ill-defined micronodules was observed at T1 in seven of the 19 subjects in whom a micronodular pattern was depicted at T0. These findings suggest that continuous smoking is responsible for more prominent inflammatory changes at the level of respiratory bronchioles and alveolar spaces, subsequently becoming detectable or depicted with a higher profusion on thin-section CT scans. As previously observed for emphysematous lesions, we observed a predominant or exclusive distribution of smoking-induced inflammatory lung changes in the upper lung zones on the initial and follow-up CT scans. These results are concordant with those of Soejima et al (12), who observed more positive CT values in the lung parenchyma of smokers. However, because their study was based on quantitative CT parameters, the shift toward more positive values was detectable in only the middle and/or lower lung fields.

It is assumed but not proved that smoker’s bronchiolitis is a precursor of the destructive changes that occur in the bronchioles and result in centriacinar emphysema (13–17). Because thin-section CT depicts early lesions within the lung parenchyma of smokers, we paid particular attention to lung changes over time in the 19 subjects with CT features of smoker’s bronchiolitis at T0. In five of the subjects showing ill-defined micronodules at T0, we observed a replacement of the micronodular pattern with emphysematous alterations, providing chronologic CT findings to support the hypothesis according to which the bronchiolar inflammatory process might induce emphysema. However, care should be taken to avoid overinterpretation of CT findings. First, pathologic-CT correlations in smoker’s lung have already pointed out the lack of sensitivity of thin-section CT in identifying smoker’s bronchiolitis (4). Moreover, despite the high sensitivity of thin-section CT in identifying emphysema, a normal CT scan does not exclude the presence of mild pathologic lesions of emphysema (4,18–22). Consequently, both kinds of limitations must be kept in mind when attempting to address the issue of the responsibility of inflammatory bronchiolar lesions in the development of emphysema on thin-section CT scans.

In the followed-up population, 10 current smokers at T0 had quit smoking between T0 and T1, leading us to categorize them as quitters. Their smoking history was characterized by a mean cigarette consumption of 11.0 pack-years at T0 and a mean duration of tobacco cessation of 2.4 years at T1. No statistically significant differences were found between the CT findings at T0 and T1, which may be due to the small number of subjects in this subgroup. However, the individual analysis of these cases led us to observe several morphologic particularities. Whereas no changes in emphysematous alterations were observed during the follow-up period, regression of the CT features of bronchioloalveolar inflammation was found in four of the six subjects in whom areas of ground-glass attenuation or ill-defined micronodules had been depicted at T0. In three of the four subjects with areas of ground-glass attenuation at T0, this CT feature was not depicted at T1. Similarly, a marked regression of the micronodular pattern was demonstrated in one of the two subjects showing ill-defined micronodules at T0. Despite obvious limitations of thin-section CT in detecting mild parenchymal changes, these findings could be interpreted as a CT demonstration of the beneficial effects of smoking cessation on lung parenchyma. Because of the limited number of subjects in this group, it was not possible to identify variable degrees of improvement known to be influenced by the age and quality of pulmonary function at the time of quitting (23). One quitter showed a similar micronodular pattern at T0 and T1. This CT finding may be explained by the persistence of the inflammatory process in the airway mucosa of smokers after smoking cessation in subjects who continue to have symptoms of chronic bronchitis (24).

With regard to functional follow-up, we observed that cigarette smoking was responsible for an accelerated annual loss of pulmonary function, which confirms previous findings (23,25–28). Whereas those who give up smoking usually cease to lose pulmonary function at an accelerated rate (28), we failed to demonstrate this beneficial effect in the current study population. This may be due to the small size of the subgroup of quitters, limited to 10 subjects. Moreover, the mean duration of smoking cessation was 2.4 years, which may not be sufficient to analyze the effects of smoking cessation in a population of healthy smokers. Cigarette smoking has already been questioned as having more detrimental effects on lung function in women than in men (29). However, owing to the limited number of followed-up subjects, we could not evaluate these sex-related changes over time. At the time of the initial evaluation of the study population, we observed that CT could depict smoking-related anomalies in the absence of functional alterations. As previously underlined by Park et al (30), this apparent discrepancy may be due to the possibility of identifying focal abnormalities on CT scans, whereas the spirometric measurements provide a more "global" measure of lung function. These findings suggest that CT allows earlier detection of adverse effects of tobacco smoke on lung parenchyma compared with pulmonary function tests.

Because detection of early changes may influence further smoking habits by means of smoking cessation counseling, we attempted to determine whether CT could help predict functional impairment over time. We observed significant relationships between emphysema and areas of ground-glass attenuation at T0 and a more rapid decline in functional parameters between T0 and T1. Despite the lack of statistically significant relationships, a similar trend was observed in the presence of ill-defined micronodules on CT scans at T0. In addition, we observed that the subjects in whom a worsening of lung changes was depicted on CT scans between T0 and T1 had greater cigarette consumption and lower values of functional parameters at T0 compared with the subjects with stable CT findings between T0 and T1. Because the mean values of functional parameters in these subjects were within the normal range, CT abnormalities at T0 could be considered as objective predictors of changes in lung function.

With regard to the evaluation of CT abnormalities in the present study, one should emphasize several aspects of the study design that may have hampered precise evaluation of subtle lung changes. First, different scanners were used at baseline and follow-up examinations. This led to the analysis of images generated with different reconstruction algorithms, which may have affected the detection of mild ground-glass opacity. In addition, CT scans obtained over time were read side by side. Although this study design was necessary to describe how subtle findings could change over time, this approach introduced a bias with regard to the demonstration that lung changes actually progress over time in smokers. Second, all subjects underwent scanning at deep inspiration, but spirometric gating was used in none of these examinations because of its unavailability at T0. Variable degrees of inspiration may have influenced the detection of nondependent ground-glass attenuation. Third, window width ranged from 1,600 to 1,800 HU according to the CT scanner used, thus introducing another variable with potential consequences for lesion conspicuity. These limitations are inherent to the improvement in CT technology during the interval between T0 and T1. However, their consequences for the CT scan reading in the present study were limited by the readers’ experience of several years of daily activity on each of the three CT scanners involved in the present investigation, as well as by the consensus reading in every case of discordant interpretation. The fourth aspect deals with the fields of view used for image reconstruction, targeted to the right lung at T0 and including both lungs at T1. Thin-section CT was targeted to one lung to optimize the detection of fine parenchymal details, which, to our knowledge, had not been reported in the literature prior to this investigation. At the time of follow-up CT, the description of smoking-related lung changes was more familiar to chest radiologists, thus leading to the imaging of both lungs on each CT scan.

The proportion of the initial population that is lost to follow-up is an important problem in longitudinal studies. In the present study, 62 of the 175 subjects initially evaluated did not undergo sequential evaluation, leading to a follow-up percentage of 63 over 8 years. In studies similar to ours (25,26,31–36), the follow-up rates have varied from 30% to 75%. The highest follow-up rates have been achieved in studies with a shorter length of follow-up—for example, 75% in the study by Beaty et al (26), whereas longer follow-up times have usually been accompanied by a lower response rate—for example, 30% of the original population in the study by Kauffmann et al (32), with a follow-up of 12 years. With regard to the characteristics of the subjects lost to follow-up, it should be emphasized that this subgroup did not differ from the followed-up population in sex ratios; proportions of smokers, nonsmokers, and ex-smokers; or in their clinical, functional, and CT characteristics at T0.

In conclusion, this study provides morphologic information concerning smoking-related lung changes in healthy volunteers, emphasizing the responsibility of continuous smoking for the worsening of emphysematous lesions and bronchioloalveolar abnormalities. Thin-section CT is superior to pulmonary function measurements in detecting subtle abnormalities and may help predict functional deterioration over time in healthy smokers.

　

	REFERENCES

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Baumgarten KB, Samet JM, Sitdley CA, Colby TV, Waldron JA. Cigarette smoking: a risk factor for idiopathic pulmonary fibrosis. Am J Resp Crit Care Med 1997; 155:242-248.

2. Heynemann LE, Ward S, Lynch DA, Remy-Jardin M, Johkoh T, Müller NL. Respiratory bronchiolitis, respiratory bronchiolitis-associated interstitial lung disease, and desquamative interstitial pneumonia: different entities or part of the spectrum of the same disease process?. AJR Am J Roentgenol 1999; 173:1617-1622.

3. Remy-Jardin M, Remy J, Boulenguez C, Sobaszek A, Edme JL, Furon D. Morphologic effects of cigarette smoking on airways and pulmonary parenchyma in healthy adult volunteers: CT evaluation and correlation with pulmonary function tests. Radiology 1993; 186:107-115.

4. Remy-Jardin M, Remy J, Edme JL, Gosselin B. Lung parenchymal changes secondary to cigarette smoking: pathologic-CT correlations. Radiology 1993; 186:643-651.

5. British Occupational Hygiene Society Committee on Hygiene Standards, Sub-Committee on Vegetable Textile Dusts. A basis for hygiene standard for flax dust. Ann Occup Hyg 1980; 23:1-17.

6. American Thoracic Society. Snowbird workshop on standardization of spirometry. Am Rev Respir Dis 1979; 119:831-838.

7. American Thoracic Society. Standardization of spirometry: 1987 update. Am Rev Respir Dis 1987; 136:1285-1298.

8. Knudson RJ, Lebowitz MD, Holberg CI, Burrows B. Changes in normal expiratory flow volume curve with growth and aging. Am Rev Respir Dis 1983; 127:725-734.

9. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow volume curve: normal standards, variability and effects of age. Am Rev Respir Dis 1976; 113:587-600.

10. Morris JF. Spirometry in the evaluation of pulmonary function: medical progress. West J Med 1976; 125:109-118.

11. Dales RE, Spizer WO, Tousignant P, Schechter M, Suissa S. Clinical interpretation of airway response to a bronchodilator. Am Rev Respir Dis 1988; 138:317-320.

12. Soejima K, Yamaguchi K, Khoda E, et al. Longitudinal follow-up study of smoking-induced lung density changes by high-resolution computed tomography. Am J Respir Crit Care Med 2000; 161:1264-1273.

13. Adesina AM, Vallyathan V, McQuillen EN, Weaver SO, Craighead JE. Bronchiolar inflammation and fibrosis associated with smoking: a morphologic cross-sectional population analysis. Am Rev Respir Dis 1991; 143:144-149.

14. Wallaert B, Gressier B, Marquette CH, et al. Inactivation of alpha1-proteinase inhibitor by alveolar inflammatory cells from smoking patients with or without emphysema. Am Rev Respir Dis 1993; 147:1537-1543.

15. Wagner EM, Bleecker ER, Permutt S, Liu MC. Peripheral airways resistance in smokers. Am Rev Respir Dis 1992; 146:92-95.

16. Willems LN, Kramps JA, Stijnen T, Sterk PJ, Weening JJ, Dijkman JH. Relation between small airways disease and parenchymal destruction in surgical lung specimens. Thorax 1990; 45:89-94.

17. Saetta M, Kim WD, Izquierdo JL, Ghezzo H, Cosio MG. Extent of centrilobular and panacinar emphysema in smokers’ lungs: pathological and mechanical implications;. Eur Respir J 1994; 7:664-671.

18. Bergin C, Müller NL, Nichols DM, et al. The diagnosis of emphysema: a computed tomographic-pathologic correlation. Am Rev Respir Dis 1986; 133:541-546.

19. Hruban RH, Meziane MA, Zerhouni EA, et al. High-resolution computed tomography of inflation-fixed lungs: pathologic-radiologic correlation of centrilobular emphysema. Am Rev Respir Dis 1987; 136:935-940.

20. Sanders C, Nath H, Bailey WC. Detection of emphysema with computed tomography: correlation with pulmonary function tests and chest radiography. Invest Radiol 1988; 23:262-266.

21. Klein JS, Gamsu G, Webb WR, Golden JA, Müller NL. High-resolution CT diagnosis of emphysema in symptomatic patients with normal chest radiographs and isolated low diffusing capacity. Radiology 1992; 182:817-821.

22. Kuwano K, Matsuba K, Ikeda T, et al. The diagnosis of mild emphysema: correlations of computed tomography and pathologic scores. Am Rev Respir Dis 1990; 141:169-178.

23. Sherrill DL, Holberg CJ, Enright PL, Lebowitz MD, Burrows B. Longitudinal analysis of the effects of smoking onset and cessation on pulmonary function. Am J Respir Crit Care Med 1994; 149:591-597.

24. Turato G, DiStefano A, Maestrelli P. Effect of smoking cessation on airway inflammation in chronic bronchitis. Am J Respir Crit Care Med 1995; 152:1262-1267.

25. Bosse R, Sparrow D, Rose CL, Weiss ST. Longitudinal effect of age and smoking cessation on pulmonary function. Am Rev Respir Dis 1981; 123:378-381.

26. Beaty TH, Menkes HA, Cohen PH, Newill CA. Risk factors associated with longitudinal change in pulmonary function. Am Rev Respir Dis 1984; 129:660-667.

27. Xu X, Dockery DW, Ware JH, Speizer FE, Ferris BG. Effects of cigarette smoking on rate of loss of pulmonary function in adults: a longitudinal assessment. Am Rev Respir Dis 1992; 146:1345-1348.

28. Scanlon PD, Connett JE, Waller LA, et al. Smoking cessation and lung function in mild-to-moderate chronic obstructive pulmonary disease: the Lung Health Study. Am J Respir Crit Care Med 2000; 161:381-390.

29. Chen Y, Horne SL, Dosman JA. Increased susceptibility to lung dysfunction in female smokers. Am Rev Respir Dis 1991; 143:1224-1230.

30. Park CS, Müller NL, Worthy SA, Kim JS, Awadh N, Fitzerald M. Airway obstruction in asthmatic and healthy individuals: inspiratory and expiratory thin-section CT findings. Radiology 1997; 203:361-367.

31. Fletcher C, Peto R. The natural history of chronic airflow obstruction. Br Med J 1977; 1:1645-1648.

32. Kauffman F, Querleux E, Drouet D, Brille D. Twelve year FEV₁ changes and smoking habits among 556 workers in the Paris area. Bull Eur Physiopathol Respir 1979; 15:723-737.

33. Camilli AE, Burrows B, Knudson RJ, Lyle SK, Lebowitz MD. Longitudinal changes in forced expiratory volume in one second in adults. Am Rev Respir Dis 1987; 135:794-799.

34. Van der Lende R, Kok TJ, Peset RR, Quanjer PH, Schouten JP, Orie NGM. Decreases in VC and FEV₁ with time: indicators for effects of smoking and air pollution. Bull Eur Physiopathol Respir 1981; 123:378-381.

35. Beck GJ, Doyle CA, Schachter EN. A longitudinal study of respiratory health in a rural community. Am Rev Respir Dis 1982; 125:375-381.

36. Kryzyzanowski M, Jedrychowski W, Wysocki M. Factors associated with the change in ventilatory function and the development of chronic obstructive pulmonary disease in a 13-year follow-up of the Cracow study: risk of chronic obstructive pulmonary disease. Am Rev Respir Dis 1986; 134:1011-1019.