Emphysema in Silicosis A Comparison of Smokers with Nonsmokers Using Pulmonary Function Testing and Computed Tomography1.2

MARIO KINSELLA,3 NESTOR MULLER, SVERRE VEDAL, CATHERINE STAPLES, RAJA T. ABBOUD, and MOIRA CHAN-YEUNG

Introduction

Silicosis is a fibrotic disease of the lungs caused by inhalation of dust containing crystalline silicon dioxide (l). The diagnosis of silicosisis dependent upon a history of appropriate exposure and confirmed by consistent radiographic and physiologic abnormalities (2). Estimates of disease severityconventionally involve a combination of radiographic and pulmonary function assessment (3). Chest radiography is used to detect lymphadenopathy and the characteristic interstitial changes of silicosis. Pulmonary function tests are used to detect associated functional limitations. It has recently been shown that there is good correlation between the visual computed tomography (CT) assessment of silicosis and the International Labour Organization (ILO) category of profusion (4). It is believed that scarring in the lung can lead to irregular emphysema adjacent to the scar (5). What is not known is whether simple silicosis without progressivemassivefibrosis (PMF) can cause emphysema. The association between emphysema and cigarette smoking is widely recognized (6). Therefore, in patients with silicosis,both cigarette smoking and lung fibrosis are potential causes for emphysema. In order to evaluate the independent effects of silicosis and cigarette smoking on emphysema, we examined 30 men with silicosis by means of CT scanning and pulmonary function testing. Methods Subjects The Workers' Compensation Board of British Columbia routinely refers all pneumoconiosis claimants to the Vancouver General Hospital Pulmonary Function Laboratory for regular follow-up. From this group, 30 subjects with silicosis were recruited to have CT of the chest, with a special effort made to recruit nonsmokers (i.e., smoked less than one cigarette per day for 1 yr). Because of their

SUMMARY The presence of emphysema in silicosis is believed to be secondary to the development of progressive massive fibrosis (PMF). However, it is difficult to separate out other causative factors, particularly cigarette smoking. In order to attempt to distinguish these factors, we examined 30 patients with silicosis by means of pulmonary function testing and computed tomography (CT) scans of the chest. Eighteen of these patients were either exsmokers or current smokers, and 12 of them were nonsmokers. The CTscans were read independently by two observers on two separate occasions. Silicosis was graded on a 5-point scale from 0 to 4; emphysema was graded as a percentage of lung involved. Percent emphysema was associated with level of pulmonary function (FEV1 , FVC, and OLeo) independent of its association with either cigarette smoking or silicosis grade (p < 0.01). Silicosis grade was associated with OLeo(p < 0.05) independent of Its association with either cigarette smoking or percent emphysema, but was not associated with level of FEV1 or FVC. In the group without PMF(silicosis Grade 0, 1, or 2), smokers had worse emphysema than nonsmokers (p < 0.01); there was no such difference among the patients with PMF (silicosis Grade 3 or 4). Only one of the nonsmoking subjects with silicosis but without PMF had any emphysema detected on CT.Our data suggest that silicosis, in the absence of PMF,does not cause significant emphysema, and that it is primarily the degree of emphysema rather than the degree of silicosis that determines the level of pulmonary function. AM REV RESPIR DIS

1990; 141:1497-1500

occupational histories, all of which involved either hard rock mining or sandblasting, the patients wereevaluated annually. Silicosiswas diagnosed in all on the basis of a history of relevant exposure to silica and radiographic changes consistent with silicosis read by a B reader. All of the radiographs had an Il.O classification of 1/0 or greater. Eighteen of the patients were smokers and 12were nonsmokers. Informed consent wasobtained from each patient prior to investigation.

Computed Tomography The CT scans wereperformed on the GE 9800 scanner at l-cm intervals using lO-mm collimation from the apex of the lung to the base of the diaphragm during breathholding after inspiration. Scans were reviewed at window levels appropriate for pulmonary parenchyma (window levels, - 650 to - 700; window width, 1,000 to 2,000 Hounsfield units). The CT scans were scored independently on two occasions each by two observers for the extent of emphysema and silicosis. The CT grading system for silicosis' was similar to that used previously (4): Grade 0, no definite nodules; Grade 1, presence of a small number of nodules; Grade 2, presence of many nodules but with no confluence; Grade 3, confluence of nodules at only one level on the CT scan; Grade 4, confluence of nodules extending over two or more cuts. In

this report, the silicosis subjects weredivided into two groups on the basis of their CT silicosis grade: those without confluence of nodules or PMF (Grade 0,1, or 2) and those with confluence of nodules or PMF (Grade 3 or 4). The CT diagnosis of emphysema was based on the presence of areas of abnormally low attenuation as determined visually (7, 8). All images were assessed and graded based on the percentage of the overallarea that demonstrated emphysematous changes. The mean of the four observations for each subject, for both silicosis and emphysema, was used in the statistical analysis.

Pulmonary Function Jesting All subjects performed tests of spirometry, diffusing capacity, and lung volumes.Spirom(Received in original/arm December 13, 1988 and in revised form December 15, 1989) 1 From the Respiratory Division, Department of Medicine, and the Department of Radiology, University of British Columbia, Vancouver General Hospital, Vancouver, B.C., Canada. 2 Correspondence and requests for reprints should be addressed to N. L. Muller, Department of Radiology, 855 West 12th Avenue, Vancouver, British Columbia, Canada V5Z IM9. 3 Recipient of a Fellowship from the British Columbia Lung Association.

1497

1498

KINSELLA, MULLER, VEDAL, STAPLES, ABBOUD, AND CHAN·YEUNG

etry was done using a rolling seal spirometer. Maximal FVC maneuvers were performed as recommended by the American Thoracic Society (9), with the best FEV I and FVC of three acceptable curves being used for the analysis. Single-breath carbon monoxide diffusing capacity (Dtco) and steady-state helium dilution lung volumes werealso measured using equipment and techniques recommended by the Intermountain Thoracic Society (10). At least two determinations of DLCO were performed so that the ratio of DLeo to singlebreath helium dilution lung volume was reproducible within 10670; the highest DLeo value was used. Reference equations of Crapo and coworkers (11, 12) were used to define percentage of predicted value for the spirometry and lung volumes, and those of Miller and colleagues (13) were used for the diffusing capacity.

Statistical Analysis The interobserver and intraobserver agreement between CT scan readings was assessed using the kappa statistic, which is a measure of agreement accounting for agreement expected from chance alone (14, 15). Percent emphysema and silicosisgrade werecompared using Wilcoxon's rank sum test (16). Differences in level of pulmonary function were tested using a two-sample t test and analysis of variance (17). Multiple linear regression analysis was performed to assess the independent effects of silicosis grade, percent emphysema, and smoking status on pulmonary function (17). Results

Descriptive data, levels of pulmonary function, and CT grades for the 30 subjects with silicosis are summarized in table 1. There were 12 nonsmokers, nine exsmokers, and nine current smokers. All were men ranging in age from 29 to 71 yr. The length of exposure to silica ranged from 36 to 432 months, with a median of 231.5 months. The 18 smokers had a mean cigarette consumption of 35 ± 13.7 (SD) pack-years. There was no significant difference in cigarette consumption between the exsmokers and the current smokers.

CT Grading There was good interobserver and intraobserver agreement for the percent emphysema by CT scanning. Kappa for the interobserver agreement was 0.43 (p < 0.001) and 0.34 (p < 0.001) for the first and second reading, respectively. For intraobserver agreement, kappa was 0.36 (p < 0.001) and 0.31 (p < 0.001) for Observers 1 and 2, respectively. There was also good interobserver and intraobserver agreement for silicosis grade by CT scanning. Kappa for interobserver agreement was 0.56 (p < 0.001)and 0.54 (p < 0.(01)

TABLE 1 CHARACTERISTICS OF 30 SUBJECTS WITH SILICOSIS

Number Age, yr ± SD Exposure to silica, yr Cigarettes, pack-years ± SD Pulmonary function, % pred FEV j * FVC FEV j % FVCt TLC DLeo Silicosis grade, n 0 1 2 3 4 Percent emphysema, n:J: 0 1-5 6-10 11-20 21-30 > 31

Nonsmokers

Smokers

Total

12 53.2 ± 13.3 19.5 ± 11.3

18 60.7 ± 7.9 21.1 ± 9.5 35.0 ± 13.7

30 57.7 ± 10.8 20.5 ± 10.1

90.3 91.8 78.3 91.3 97.5

± ± ± ± ±

23.2 21.2 8.8 16.9 19.0

68.2 84.3 63.8 96.7 94.4

3 0 2

a 7 5 5 0 2

± ± ± ± ±

20.5 18.1 14.3 15.3 18.7

76.9 87.3 69.6 94.7 95.7

1 3 3 1 10

4 3 5 1 17

a

5 13 4

8 4 4 1 1

a a

± ± ± ± ±

23.9 19.4 14.2 15.8 18.5

6 1 1

Smokers significantly different from nonsmokers: • p < 0.05. t p < 0.01. :j: P < 0.001.

for the first and second readings, respectively. For intraobserver agreement kappa was 0.74 (p < 0.(01) and 0.51 (p < 0.(01) for Observer 1 and 2, respectively. Percent emphysema and silicosis grade in nonsmokers and smokers are shown in table 1. Silicosis grades were not different in the smoking and nonsmoking groups. There was a significantly higher percent emphysema in smokers than in nonsmokers in the subjects with silicosis with Grade 0, 1, or 2 (table 2 and figure 1); only one of the five nonsmokers with mild silicosis had any detectable emphysema. All seven smokers with Grade 0, 1, or 2 silicosis had emphysema identifiable on CT scan (figure 1). There was no difference in percent emphysema between

smokers and nonsmokers with Grade 3 or 4 silicosis. Percent emphysema was more severe in subjects with Grade 3 or 4 silicosis than in those with Grade 0, 1, or 2 silicosis (table 2 and figure 1).

Pulmonary Function There were no statistically significant differences in FEV 1 between the patients with Grade 0, 1, or 2 silicosis (no confluence) and Grade 3 or 4 (confluence) in either the smokers or the nonsmokers (table 2), even though the trend was for lower levels in those with higher grades of silicosis. There was a strong association between Dtco and silicosis grade, with the DLCO being significantly lower in the patients with Grade 3 or 4 silicosis

TABLE 2 PERCENT EMPHYSEMA (MEDIAN) AND LEVEL OF PULMONARY FUNCTION (PERCENT OF PREDICTED FEV, AND OLeo) OF THE SMOKERS AND NONSMOKERS CLASSIFIED BY SILICOSIS GROUP (0, 1, 2 AND 3, 4) Nonsmokers

Percent emphysema median * Range FEv,t DLeo:t:

Smokers

Silicosis Grade 0,1,2

Silicosis Grade 3,4

Silicosis Grade 0,1,2

Silicosis Grade 3, 4

0 (D-l) 100.1 ± 14.4 106.8 ± 9.1

1.5 (0-18) 82.3 ± 25.8 89.8 ± 21.3

2.5 (1-12) 72.9 ± 24.3 107.2 ± 38.4

7.3 (1-53) 64.5 ± 14.2 84.3 ± 17.3

• Grade 3, 4 silicosis: no difference in percent emphysema between smokers and nonsmokers (p > 0.05). Grade 0, 1, 2 silicosis: percent emphysema greater in smokers than in nonsmokers (p < 0.02). All subjects: emphysema score higher in silicosis Grade 3 or 4 than in Grade 0, 1, or 2 (p < 0.05). t All subjects: no difference between Grade 0, 1, or 2 silicosis and Grade 3 or 4 silicosis (p = 0.38). :j: Difference between silicosis Grades 0, 1, 2 and 3. 4 significant for both smoking groups (p < 0.005).

1499

EMPHYSEMA IN SILICOSIS USING COMPUTED TOMOGRAPHY

60 o

50

Non-Smokers

• Ever-Smokers

~

- - 40

as

S Fig. 1. Silicosis grade versus percent emphysema in nonsmokers and eversmokers.

~

30

S

20

E

....

~

o

10

o

.

.

....

&.

i

I

o

I

y

o88e o

I

123

I

4

Silicosis Grade than in those with Grade 0, 1, or 2 silicosis (table 2). The Dtoo was also significantly lower in the patients with higher degrees of emphysema than in those with less emphysema (p < 0.001). Levels of FEV 1 and FVC were also associated with percent emphysema (p =;= 0.001). Multiple linear regression analysis was used to determine whether percent emphysema, silicosis grade, or both were independent predictors of level of pulmonary function. Percent emphysema and silicosis grade were both independently associated with DLeo (table 3). In contrast, only percent emphysema was independently associated with level of FEV 1 (table 4) or with FVC (data not shown). After controlling for emphysema, cigarette smoking was associated with a lower FEV 1, but was not associated with lower FVC or DLeo.

Some investigators have argued that emphysema is not a complication of simple silicosis (1, 18-21), whereas others have concluded that emphysemadoes occur in silicosis (22-24). However, only one of the patients with emphysema studied by Kjuus and coworkers (24) did not smoke cigarettes. Chatgidakis (25), who reported data on 800 consecutive autopsies of white South African gold miners, concluded that the incidence of silicosis was higher in patients with emphysema than in those without emphysema, and that the more severe the emphysema, the worse the silicosis. Most investigators be-

TABLE 3 MULTIPLE LINEAR REGRESSION ESTIMATES OF DIFFERENCES IN PERCENT OF PREDICTED DLcO ASSOCIATED WITH PERCENT EMPHYSEMA, HIGH SILICOSIS GRADE, AND SMOKING

Vari'able

Discussion

Thirty subjects with silicosis were examined by means of pulmonary function testing and chest CT scans. Eighteen of them were either current or exsmokers, and 12 were nonsmokers. The degree of emphysema was the predominant independent predictor of pulmonary function impairment; degree of silicosis was also an independent predictor, but only for DLeo. In those subjects who did not have evidence of PMF, smokers had worse emphysema than did nonsmokers. In PMF, the degree of emphysema in smokers and nonsmokers was not statistically different. Our data suggest that silicosis does not cause significant emphysema in the absence of PMF, and that it is primarily the degree of emphysema rather than silicosis that determines the level of pulmonary function in these subjects.

lieve that emphysema does occur in complicated silicosis (PMF), with some (1-3, 22) believing that it occurs only in complicated silicosis. It has also been reported that silica dust exposure predisposes to emphysema independent of silicosis (23, 26). The use of CT in the diagnosis of silicosis has been shown to correlate closely with the ILO category of profusion (4). There were four patients in our study with Grade 0 silicosis on CT scan who had small nodules on their chest radiographs. This relative insensitivity of the CT scan in detecting mild silicosis has been reported previously by Begin and coworkers (27). However, CT is more sensitive than chest X-ray in the diagnosis of emphysema; even mild emphysema can be detected on CT scanning (7), although this has recently been disputed (28). The chest radiograph is insensitive in detecting mild to moderate parenchymal destruction (29, 30). In our study, percent emphysema was the predominant determinant of the level of pulmonary function. A high grade of silicosis was not associated with worsened spirometry, and cigarette smoking was associated only with a reduced FEV 1. For Dr.co, a high grade of silicosis was associated with a 13070 of predicted decrease relative to a low grade of silicosis. Each 10070 increase in emphysema was associated with a 10070 of predicted decrease

High silicosis grade" Percent emphysemat Ever-smoke:t: Constant R2 = 0.56

Percent Differences

Standard Error

-13.2 -0.96 3.4 108.4

4.98 0.24 0.12

p Value

0.02

< 0.001 0.51

• Decrease associated with silicosis Grades 3 and 4 versus Grades 0,1 and 2. Decrease associated with each percent increase in emphysema. :j: Decrease associated with ever-smoking.

t

TABLE 4 MULTIPLE LINEAR REGRESSION ESTIMATES OF DIFFERENCES' IN PERCENT OF PREDICTED FEV1 ASSOCIATED WITH PERCENT EMPHYSEMA, HIGH SILICOSIS GRADE, AND SMOKING

Variable High silicosis grade Percent emphysema Ever-smoke Constant R2 = 0.40 • See table 3.

Percent Differences

Standard Error

p Value

-2.8 -0.94 -15.8 94.9

7.74 0.37 7.61

0.73 0.02 0.05

1500

KINSELLA, MULLER, VEDAL, STAPLES, ABBOUD, AND CHAN-YEUNG

in DLeo, a 9% decrease in FEV 1, and a 70/0 decrease in FVC. In patients with mild silicosis (no PMF), percent emphysema was worse in smokers than in nonsmokers. This is not surprising since cigarette smoking causes emphysema (6). However, in patients with more severe silicosis (PMF), the degree of emphysema was not worse in smokers than in nonsmokers. The severity of emphysema in patients with more severesilicosis seems to be determined by factors other than cigarette smoking. In the nonsmokers, there was little or no emphysema seen with mild silicosis (no PMF); emphysema was largely present only if they had developed PMF. Auerbach and colleagues (6) reported the frequency of emphysema to be approximately 100/0 in nonsmoking subjects older than 60 yr of age but did not report any "severe" emphysema in 166 nonsmoking subjects who had autopsy studies performed. It is interesting to note that six of the seven nonsmoking patients with PMF also had emphysema, although there was one patient with PMF who had no detectable emphysema. Therefore, in most instances, both PMF and emphysema appear to develop together in nonsmokers. However, emphysema is seen not only around the areas of fibrosis but also in nonfibrotic areas. This suggests that there may be some other factor that determines susceptibility to the development of both PMF and emphysema. References 1. Seaton A. Silicosis. In: Morgan WKC, Seaton

A, eds. Occupational lung diseases. 2nd ed. Philadelphia: WB Saunders, 1984; 250-4. 2. Lapp NL. Lung disease secondary to inhalation of non fibrous minerals. Clin Chest Med 1981; 2:219-33. 3. Fraser RG, Pare JAP. Diagnosis of diseases of the chest. Philadelphia: WB Saunders, 1977; 1484502. 4. Bergin CJ, Muller NL, Vedal S, Chan-Yeung M. CT in silicosis: correlation with plain films and pulmonary function tests. AJR 1986; 146:477-83. 5. Thurlbeck WM. Chronic airflow limitation in lung disease.Philadelphia: WB Saunders, 1975; 170. 6. Auerbach 0, Hammond EC, Garfinkel L, Benante C. Relationship of smoking and age to emphysema: whole lung section study. N Engl J Med 1972; 286:853-7. 7. Foster WL Jr, Pratt PC, Roggli VL, Godwin JD, Halvorsen RA, Putman CEoCentrilobular emphysema: CT-pathologic correlation. Radiology 1986; 159:27-32. 8. Muller NL, Staples CA, Miller RR, Abboud RT."Density mask": an objective method to quantitate emphysema using computed tomography. Chest 1988; 94:782-7. 9. Gardner RM, chairman. American Thoracic Society Statement. Snowbird workshop on standardization of spirometry. Am Rev Respir Dis 1979; 119:831-8. 10. Morris AH, Kanner RE, Crapo RO, Gardner RM. Clinical pulmonary function testing: a manual of uniform laboratory procedures. 2nd ed. Salt Lake City: Intermountain Thoracic Society, 1984. 11. Crapo RO, Morris AH, Clayton PO, Nixon CR. Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis 1981; 123:659-64. 12. Crapo RO, Morris AH, Clayton PO, Nixon CR. Lung volumes in healthy nonsmoking adults. Bull Eur Physiopathol Respir 1982; 18:419-25. 13. Miller A, Thornton JC, Warshaw R, Anderson H, Teirstein AS, Selikoff IJ. Single breath diffusing capacity in a representative sample of the population of Michigan, a large industrial state. Am Rev Respir Dis 1983; 127:270-7. 14. Fleiss JL. Statistical methods for rates and proportions. NewYork: John Wileyand Sons, 1973; 143-7.

15. Hunt RJ. Percent agreement, Pearson's correlation, and kappa as measures of inter-examiner reliability. J Dent Res 1986; 65:128-30. 16. Rosner B. Fundamentals of biostatistics. Boston: Duxbury Press, 1982; 277-81. 17. Zar JH. Biostatistical analysis. 2nd ed. Englewood Cliffs: Prentice-Hall, 1984. 18. Parkes RW. Occupational lung diseases. 2nd ed. London: Butterworths, 1982; 134-58. 19. Brinkman GL, Block DL, Cress C. Effects of bronchitis and occupation on pulmonary ventilation over an 11-year period. J Occup Med 1972; 14: 615-20. 20. Elmes PC. Relative importance of cigarette smoking in occupational lung disease.Br J Ind Med 1981; 38:1-13. 21. Morgan WKC. Industrial bronchitis. Br J Ind Med 1978; 35:285-91. 22. Ziskind M, Jones RN, WeillH. Silicosis. Am Rev Respir Dis 1976; 113:643-65. 23. Irwig LM, Rocks P. Lung function and respiratory symptoms in silicotic and nonsilicotic gold miners. Am Rev Respir Dis 1978; 117:429-35. 24. Kjuus H, Istad H, Langard S. Emphysema and occupational exposure to industrial pollutants. Scand J Work Environ Health 1981; 7:290-7. 25. Chatgidakis CB. Silicosis in South African white gold miners: a comparative study of the diseasein its different stages.Med Proc 1963; 9:383-92. 26. Becklake MR, Irwig L, Kielkowski 0, Webster I, de Beer M, Landan S. The predictors of emphysema in South African gold miners. Am Rev Respir Dis 1987; 135:1234-41. 27. Begin R, Bergeron D, Samson L, Boctor M, Cantin A. CT assessment of silicosis in exposed workers. AJR 1987; 148:509-14. 28. Miller RR, Muller NL, VedalS, Morrison NJ, Staples CA. Limitations of computed tomography in the assessment of emphysema. Am Rev Respir Dis 1989; 139:980-3. 29. Pugatch RD. The radiology of emphysema. Clin Chest Med 1983; 4:433-42. 30. Thurlbeck WM, Simon G. Radiographic appearance of the chest in emphysema. AJR 1978; 130:429-40.