- Email: [email protected]
Association Between Sputum Atypia and Lung Cancer Risk in an Occupational Cohort in Yunnan, China
Ii CHEST
Original Research LUNG CANCER
Association Between Sputum Atypia and Lung Cancer Risk in an Occupational Cohort in Yunnan, China* Ya-Guang Fan, MD; Ping Hu, SM, SeD; YongJiang, MPH; Run-Sheng Chang, MD; Shu-Xiang Yoo, MD; Wendy Wang, PhD; Jie He, MD; Philip Prorok, PhD; and You-Lin Qiao, MD, PhD Background: Individuals with cytologic atypia in sputum may be at high risk for the development of lung cancer. Method.: A prospective cohort study was conducted among occupational tin minen in Yunnan China based on an annual lung cancer screening program. Sputum samples were collected prospectively at baseline and the following seven annual screenings. The associations between risk factors and sputum cytology were analyzed by univariate and multivariate logistic regression. A proportional hazard model was used to analyze the association between the baseline sputum results and the incidence of lung cancer. The effect of consecutive sputum cytology on the increase of lung cancer risk was analyzed by logistic regression. He8ultB: Sputum cytologic atypia was associated with age, smoking, occupational radon and anenic exposure, and asthma. Sputum cytologic atypia was an independent risk factor for lung cancer with an adjusted hazard ratio (RR) of 3.82 (95% confidence interval [CI], 2.82 to 5.18) in comparing normal to moderate or wone atypia. Compared to the lung cancer risk associated with normal sputum, the risk was significantly higher according to the degree of atypia for squamous carcinomas, small cell lung cancer and central lung cancer, with adjusted IIRs of 5.70 (95% CI, 3.78 to 8.59), 3.32 (95% CI, 1.31 to 8.45), and 4.93 (95% CI, 3.51 to 6.92), respectively. Conclusions: Sputum atypia is associated with an increased risk of lung cancer. Sputum cytologic examination combined with other screening examinations may play an important role in the early detection of lung cancer or in the selection of the optimal target population for more intensive lung cancer screening among this occupational cohort or simi1ar population. Trial registration: CIinica1tria1s.gov Identifier: NCT00340405. (CHEST 2009; 135:118-185) Key words: atypia; lung cancer; sputum; Yunnan, China Abbreviations: CI = confidence interval: HR = hazard ratio; OR cancer; TO to 3 = time 0 to 3; YTC = Yunnan Tin Corporation
Lung cancer is the most common cancer worldwide. In 2005 it was estimated that there were > 490,000 new cases of lung cancer in China.' Lung cancer is also a highly lethal disease; the 5-year survival rate for all stages combined is 15% in the United States.2•3 The reasons for the poor prognosis are late diagnosis and lack of effective treatment for symptomatic disease. In addition, the prognosis for lung cancer is related to the stage at which it is detected.' Sputum cytologic examination as a lung cancer screening tool is not recommended by any organization for the general population or any high-risk
na
= odds ratio; Q = quartile; SCLC = small cell lung
population.S" The reason is that prior screening trials 7- 9 have not shown a reduction oflung cancer mortality. Although low-dose CT is considered the most potentially effective screening tool, it may be insensitive to superficial, preinvasive cancers in the central airways that may be manifested in sputum.'? Therefore, biomarkers that might be identifiable in the sputum may complement radiologic imaging for lung cancer. In addition, lung cancer risk has been demonstrated to increase with increasing levels of sputum cytologic dysplasia, particularly at levels of moderate atypia or worse. 1 1- 1S Original~h
The aim of this study was to investigate whether sputum atypia can be used as a predictor of increased lung cancer risk among participants in an occupational cohort in Yunnan, China. Although annual chest radiography was also conducted, we focus our analyses in this report on the results from sputum cytology examinations.
MATERIALS AND METHODS Study Design and Participants The Yunnan Tin Corporation (YTC) study is a one-armed prospective cohort study initiated in 1992 with annual follow-up through 2001. Participants were tin miners in Yunnan, China and with at least10 years of underground radon and/or arsenic exposure. Detailed inclusion criteria are presented elsewhere." At the time of enrollment, all participants completed standardized baseline questionnaires that included demographic characteristics; residential, occupational, smoking, and medical histories; and a dietary assessment component. From 1992 to 1999, a total of 9,295 eligible tin miners were enrolled in this study. Eight rounds of annual lung cancer screenings with chest radiography and sputum cytology were performed during this period. Among them, 9,084 participants underwent at least one sputum cytologic screening examination with adequate results. The YTC study received approval from the institutional review board, and informed consent was obtained for each participant. Tobacco was consumed mainly in the form of cigarettes, water pipes, and Chinese long-stem pipes. The never-smoker was defined as having a smoking history of < 6 months. Screen"From the Department of Biology (Dr. Fan), Institute of Radiation Medicine, Chinese Academy of Medical Sciences, Tianjin, People's Republic of China; Department of Cancer Epidemiology (Drs. Jiang and Qiao), Cancer Institute, Chinese Academy of Medical Sciences, Beijing, People's Republic of China; Third People's Hospital of Honghe Autonomous District (Dr. Chang), Gejiu, Yunnan Province, People's Republic of China; Department of Epidemiology (Dr. Yao), Kunming Medical College, Kunming, Yunnan Province, People's Republic of China; Cancer Biomarkers Research Group (Dr. Wang), Division of Cancer Prevention, National Cancer InstitutelNational Institute of Health, Bethesda, MD; Cancer Institute (Dr. He), Chinese Academy of Medical Sciences, Beijing, People's Republic of China; and Biometry Research Group (Drs. Hu and Prorok), Divisions of Cancer Prevention, National Cancer InstitutelNational Institute of Health, Bethesda, MD. None of the authors have any financial interests in the subject matter or any actual or potential conflicts of interest. This study was supported by the National Cancer Institute! National Institutes of Health, grant No. 263-MQ-511694. This article describes an ancillary study for the National Cancer Institute project "The Study of Early Markers of Lung Cancer Among Tin Miners in Yunnan China," a federally funded registered clinical trial. Manuscript received June 1B,200B;revision accepted September 16,2008. Reproduction of thisarticle is prohibited without written permission from the American College of chest Physicians (www.cliestjournal. org/mtsc/reprints.shtml). Correspondence to: Ping Hu, SM, SeD, Mathematical Statistician, Biometry Research Group, Divisions of Cancer Prevention, National Cancer InstitutelNational Institute of Health, Executive Plaza North, Suite 3131, 6130 Executive Blvd, MSC 7354, Bethesda, MD 20892-7354; e-mail: [email protected] DOl: lO.13781chest.08·1469 www.chestjoumal.org
detected cancers were defined as those diagnosed within 12 months of a positive screen. All other cancers were classified as interval cancers. Detailed definitions of occupational radon and arsenic exposures are given elsewhere.'? In this study, occupational radon and arsenic exposure were grouped into four quartiles (Ql to Q4) based on each individual's cumulative radon or arsenic levels, respectively. Sputum Collection: Processing and Interpretation Sputum samples were collected either spontaneously or by an induction procedure, and stored at room temperature in Saccomanno preservative. Specimens were smeared on glass slides and stained with Papanicolaou stain for routine cytologiC examination. Slides were independently screened by two cytopathologists from the Labor Protection Institute in the YTC. Sputum cytologic screening results were classified into degrees as follows: (1) 1 = negative; (2) 2-1 = slight atypical metaplasia; (3) 2-2 = moderate atypical metaplasia; (4) 2-3 = grave atypical metaplasia; (5) 3 = suspicious for cancer; (6) 4 = highly suspicious for cancer; (7) 5-1 = epidermoid carcinoma; (B) 5-2 = adenocareinoma; (9) 5--3 = undifferentiated carcinoma (includes both large cell and small cell); and (10) other cancer. The positive cytology slides were defined as the screening results of degrees 2-3 and above, and reread by a confirmatory reader. In this article, we focus on moderate atypical metaplasia (degree 2-2) or worse (degrees > 2-2) to investigate whether sputum atypical metaplasia can be used as a predictor of lung cancer. Statistical Methods The associations between risk factors and sputum cytology results were analyzed by univariate and multivariate logistic regression. IS Person-years of follow-up were calculated from the date of the first sputum cytologic screening to the date of lung cancer diagnosis or lung cancer death or censoring as of December 31, 2001. The proportional hazards model was used to analyze the association between baseline sputum results and lung cancer incidence. The effect of consecutive sputum cytology on the increase of lung cancer risk was also analyzed by logistic regression. IS The Kruskal-Wallis test was used to compare participant age, radon and arsenic exposures. The X2 test of independence was used to evaluate the statistical significance of differences in proportions. All statistical tests were two-sided.
RESULTS
In this study, all analyses were restricted to the 9,084 participants who received at least one sputum cytology screening and had adequate results. Characteristics of the study participants are presented in Table 1. Most participants were men (93.9%), with women accounting for only 6.1% of participants. At the time of enrollment, 66.1% of participants were 40 to 59 years old, Of the participants, 23.9% never attended school, and 84.9% had a history of smoking. Table 1 alsoshowscumulative occupationalarsenic and radon exposures and information on prior pulmonary diseases such as chronic bronchitis, asthma, and silicosis. The association between personal/occupational characteristics and sputum cytology results in the baseline sputum screening is also given in Table 1. CHEST/ 135 / 3 / MARCH, 2009
779
Table l-A.aociation Bet1Deen Participant Characteristics and Sputum Cytology Among the fTC Cohort in BoNfine Sputum Screening OR for Moderate Atypia or Worse, Compared to Normal Cytology
Sputum Cytology, No. (%)
Characteristics
All
Gender Female Male Age group, yr 40-49 50-59 60--69 ~70
Education No 056yr >6yr Smoking status Never Cigarette only Water/stem pipe only Both Cigarette or waterl stem pipe Arsenic level Ql (0-1455) Q2 (1,456-7,102) Q3 (7,102-17,435) Q4 (~ 17,435) Radon level Ql (0-82046) Q2 (82.47-205.80) Q3 (205.81-507.29) Q4 (~ 507.29+ ) Asthma No Yes Chronic bronchitis No
Yes Silicosis No Yes
Total Participants, No.(%)
Normal
Slight Atypia
l
Worse Than Moderate Moderate Atypia Atypia'"
Moderate Atypia or Worse
Crude OR (95%CI)
Adjusted OR (95% CI)t
9,084 (100)
8,069 (88.8) 847 (9.3)
124 (1.4)
44(0.5)
168 (1.8)
553 (6.1) 8,531 (93.9)
542 (98.0) 9 (1.6) 7,527 (88.2) 838(9.8)
2 (004) 122 (104)
o(O.O) 44(0.5)
2 (004) 166 (1.9)
I 5.98 (1.48--24.16)
I 1.23 (0.28-5.37)
3,786 (41.7) 2,219(2404) 2,459 (27.1) 620 (6.8)
3,599 (95.1) 177 (4.7) 1,935 (87,2) 244 (11.0) 2,032 (82.6) 348 (14.2) 503 (8Ll) 78 (12.6)
10 (0.3) 30 (104) 53(2.2) 31 (5.0)
0(0.0) 10 (004) 26 (1.1) 8 (1.3)
10(0.3) 40 (1.8) 79(3.3) 39 (6.3)
1 7.44 (3.71-14.91) 13.99 (7.23-27.08) 27.91 (13.84-56.25)
I 4.36 (1.99-9.54) 5.67 (2.55-12.64) 9.36 (3.99-21.95)
2,174 (23.9) 4,376 (48.2) 2,534 (27.9)
1,794 (82.5) 298 (13.7) 3,872 (88.5) 435 (9.9) 2,403 (94.8) 114 (4.5)
62 (2.9) 49 (Ll) 13 (0.5)
20(0.9) 20 (0.5) 4 (0.2)
82(3.8) 69 (1.6) 17(0.7)
1 0,39 (0.28--0.54) 0.16 (0.09-{).26)
1 0.79 (0.57-1.12) 0.80 (0.44-1.47)
1,372 (15.1) 2,215 (2404) 591 (6.5)
1,294 (94.3) 2,087 (94.2) 505 (8504)
70(5.1) III (5.0) 70 (11.8)
8 (0.6) 12 (0.5) 14 (204)
0(0.0) 5 (0.2) 2(0.3)
8(0.6) 17(0.8) 16 (2.7)
1 1.32 (0.57--3.06) 5.13 (2.18--12.05)
1 1047 (0.63--3.47) 1.21 (0.50--2.94)
4,906 (54.0) 7,712 (84.9)
4,183 (85.3) 596 (12.1) 6,775 (87.9) 777 (10.1)
90 (1.8) 116 (1.5)
37 (0.8) 44 (0.6)
127 (2.6) 160 (2.1)
4.91 (2.40-10.06) 3.82 (1.87-7.79)
1.93 (0.92-4.06) 1.75 (0.84--3.65)
2,364 (26.0) 2,248 (24.7) 2,252 (24.8) 2,220 (24.4)
2,236 (94.6) 1,955 (87.0) 1,856(8204) 2,022 (91.1)
118 (5.0) 263 (11.7) 318 (14.1) 148 (6.7)
9(0.4) 21 (0.9) 55 (2.4) 39(1.8)
1 (0.0) 9(0.4) 23 (1.0) II (0.5)
10 (004) 30 (1.3) 78 (3.5) 50(2.3)
1 3.43 (1.67-7.04) 9.40 (4.85-18.20) 5.53 (2.80-10.93)
1 1.06 (0.46-2.44) 2.02 (0.86-4.75) 2.60 (1.10-6.18)
2,340 (25.8) 2,232 (24.6) 2,251 (24.8) 2,261 (24.9)
2,252 (96.2) 78(3.3) 2,084 (93.4) 128 (5.7) 1,949 (86.6) 255 (11.3) 1,784 (78.9) 386 (17.1)
8 (0.3) 17 (0.8) 29 (1.3) 70 (3.1)
2 (0.1) 3(0.1) 18(0.8) 21 (0.9)
10 (004) 20(0.9) 47 (2.1) 91 (4.0)
1 2.16 (1.01-4.63) 5.43 (2.74-10.78) 11.49 (5.96-22.14)
1 3.96 (1.71-9.17) 4.52 (2.10-9.70) 5.26 (2.52-10.96)
8,417 (92.7) 667 (7.3)
7,531 (89.5) 747 (8.9) 538 (80.7) 100 (15.0)
102 (1.2) 22(3.3)
37(0.4) 7(1.0)
139 (1.7) 29(4.3)
1 2.92 (1.94-4.40)
1 1.80 (1.18--2.74)
6,682 (73.6) 2,402 (26.4)
6,087 (9Ll) 497 (7.4) 1,982 (82.5) 350 (14.6)
76 (Ll) 48(2.0)
22 (0.3) 22 (0.9)
98 (1.5) 70 (2.9)
1 2.19 (1.61-2.99)
1 1.22 (0.86-1.73)
8,633 (95.0) 451 (5.0)
7,699 (89.2) 789 (9.1) 370 (82.0) 58 (12.9)
110 (1.3) 14 (3.1)
35(0.4) 9 (2.0)
145 (1.7) 23 (5.1)
1 3.30 (2.10-5.19)
1 1.29 (0.8D-2.07)
"'Worse than moderate atypia is included as "grave atypical metaplasia," "suspicious for cancer," "highly suspicious for cancer." or "any types of carcinoma. " t Adjusted for gender, age. education, smoking, occupational radon and arsenic exposure, and prior pulmonary diseases.
Moderate atypia or worse was found more often in men than women in univariate analysis (odds ratio [OR], 5.98; 95% confidence interval [CI], 1.48 to 24.16), but there was no significant association with gender in multivariate analysis. Risk of moderate or worse atypia was elevated with the increase of age level. There was an inverse relationship between education level and moderate atypia or worse, but it was not statistically significant. Compared with neversmokers, cigarette and water/stem pipe smokers had 780
a significantly higher risk of moderate atypia or worse, according to univariate analysis, This signiflcant result was not observed in the multivariate analysis. There was a positive relationship between moderate atypia or worse and radon exposure level. According to the multivariate analysis, participants with arsenic exposure levels in Q4, had a significantly higher risk of moderate atypia or worse than those with exposure levels in Q1; however, those with levels in Q2 to Q3 did not have a significantlygreater Original Research
risk than those in Q1. As to prior pulmonary diseases, only asthma was a risk factor of moderate atypia or worse. By 2001, 502 lung cancer cases were confirmed. A total of 464 participants with lung cancer received at
least one sputum cytologic screening and had valid diagnosis dates. Of these participants with lung cancer, 438 died by the end of 2001. The 38 participants whose diagnoses were confirmed only by cause of death, without a date of diagnosis, are not
Table 2-Lung Cancer IncidencelDea'" Among ,he YTC Sputum Screening Co'Iaorl Lung Cancer Incidence", 1/100,000 Person-yr
Characteristics
All
Gender Female Male Age group, yr ~9
50-59 60-69 2= 70 Education, yr No s6 >6 Smoking status Never Cigarette only Water/stem pipe Both Cigarette or water/ stem pipe Arseniclevel Ql (0-1455) Q2 (1456-7102) Q3 (7102--17435) Q4(17435+) Radon level Ql (0-82.46) Q2 (82.47-205.80) Q3 (205.81--507.29) Q4 (507.29+) Asthma No Yes Chronic bronchitis No Yes Silicosis No Yes Baselinesputum cytology Normal Slight atypia Moderate atypia Worse than moderate atypia Moderate atypiaor worse
Lung Cancer Incidence Rate
Lung Cancer Death, 1/100,000 Person-yr
Person-yr
Lung Cancer Deaths
Lung Cancer Death Rate
71,937.4
438
608.9
Person-yr
Lung Cancer Cases, No.
71,279.0
464
651.0
4,159.9 67,119.2
2 462
48.1 688.3
1.00 14.32(4.90-41.81)
4161.0 67,776.6
2 436
48.1 643.3
29,404.1 18,880.6 19,069.3 3,925.2
10 114 268 72
34.0 603.8 1,405.4 1,834.3
1.00 17.76(11.08--28.44) 41.32 (28.38-60.16) 53.94(37.30-78.00)
29,419.5 19,031.1 19,417.1 4,069.7
8 98 253 79
27.2 514.9 1,303.0 1,941.2
16,787.5 35,541.8 18,949.8
206
230
1,370.1 579.6 147.8
1.00 0.42 (0.35-0.51) 0.11 (0.~.15)
17,140.6 35,806.4 18,990.4
218 197 23
1,271.8 550.2 121.1
1.00 0.43 (0.52--0.36) 0.10 (0.13-0.07)
1.00 1.62 (0.92--2.85) 7.67 (4.94--11.91) 4.68 (3.06-7.14) 4.07 (2.65-6.24)
10,578.2 16,540.5 4,582.9 40,235.9 61,359.3
19 45 60 314 419
179.61 272.06 1,309.21 780.40 682.86
1.00 1.51 (0.85--2.71) 7.29 (4.67-11.37) 4.34 (2.83--6.67) 3.80 (2.47--5.86)
28
Incidence Rate Ratio(95% CI)
Death Rate Ratio(95% CI)
1.00 13.37(4.50-39.72) 1.00 18.93(11.27-31.82) 47.90 (32.02--71.71) 71.37 (49.05--103.90)
10,549.2 16,477.2 4,486.6 39,766.1 60,729.9
445
180.11 291.31 1,381.89 842.43 732.75
17,912.6 18,876.2 17,626.4 16,864.0
14 86 247 117
78.2 455.6 1,401.3 693.8
1.00 5.83 (3.52-9.66) 17.93(12.13-26.49) 8.88 (5.59-14.09)
17,922.4 19,060.4 17,941.1 17,013.5
13 75 238 112
72.5 393.5 1,326.6 658.3
1.00 5.43 (3.1S-9.25) 18.30(12.22-27.37) 9.08 (5.63-14.63)
17,588.9 17,597.2 18455.0 17,638.0
25 39 134 266
142.1 221.6 726.1 1,508.1
1.00 1.56 (0.91-2.67) 5.11 (3.47-7.52) 10.61(7.63-14.75)
17,625.9 17,673.1 18,668.0 17,970.4
20 33 129 256
113.5 186.7 691.0 1,424.6
1.00' 1.65(0.91-2.99) 6.09 (4.02-9.24) 12.55(8.81-17.90)
66,222.8 5,056.4
398 66
601.0 1,305.3
1.00 2.17 (1.6S-2.81)
66,797.6 5,139.8
373 65
558.4 1,264.6
1.00 2.26 (1.74-2.94)
52,410.4 18,868.7
258 206
492.3 1,091.8
1.00 2.22 (1.85--2.65)
52,825.0 19,112.4
241 197
456.2 1,030.7
1.00 2.26 (1.88-2.72)
67,964.4 3,314.7
409
55
601.8 1,659.3
1.00 2.76 (2.10-3.63)
68,556.9 3,380.5
384
54
560.1 1,597.4
1.00 2.85 (2.16-3.76)
63,219.8 6,969.0 944.7 145.6
336 79 21 28
531.5 1,133.6 2,222.9 19,230.8
1.00 2.13 (1.67-2.72) 4.18 (2.75--6.34) 36.18(28.44-46.04)
63,682.9 7,076.7 988.0 189,9
320 68 22 28
502.5 960.9 2,226.7 14,744.6
1.00 1.91 (1.47-2.48) 4.43 (2.95--6.66) 29.34 (22.76-37.83)
1,090.4
49
4,493.8
8.46 (6.57-10.88)
1,177.9
50
4,244.8
8.45 (6.57-10.86)
19
48
62 335
"Lung cancer cases without date of diagnosis were not analyzed. www.chestjoumal.org
CHEST / 135/ 3 / MARCH, 2009
781
discussed further in this article. Results presented in Table 2 for lung cancer incidence ratios showed that gender, age, radon and arsenic exposure, smoking status, and prior pulmonary diseases were all potential risk factors for lung cancer. We observed an inverse relationship between education level and the risk of lung cancer. The lung cancer incidence ratio for cigarette and water/stem pipe smokers compared to never-smokers was 4.68 (95% CI, 3.06 to 7.14). As to occupational arsenic exposure, the incidence ratios from low to high (Q1 to Q4) were 1.00,5.83 (95% CI, 3.52 to 9.66), 17.93 (95% CI, 12.13 to 26.49), and 8.88 (95% CI, 5.59 to 14.09), respectively. For radon exposure, the incidence ratios increased from 1.00 to 10.61 (95% CI, 7.63 to 14.75). Lung cancer risk was higher among men than women and among participants with prior chronic bronchitis or silicosis. The same pattern for all of the variables discussed above was also observed in the results of the mortality ratios (Table 2). The association between sputum results and lung cancer incidence is displayedin Table 3. In the baseline sputum screening, moderate, worse than moderate atypia, or both were associated with a higher risk of lung cancer than was normal cytology. A total of 4,269 participants received the baseline (time 0 [TO]), and three subsequent annual (TI to T3) sputum screenings. In the multivariate model, the odds ratio increased with the frequency of consecutive sputum screenings from 2.60 (95% CI, 1.41 to 4.79) for one baseline measurement (TO) to 4.62 (95% CI, 3.14 to 6.81) for four measurement (TO to T3) [Table 3]. The associations between occupational risk factors of lung cancer and baseline results of sputum cytol-
ogy are shown in Table 4. Radon and arsenic exposures were highest among participants with moderate atypia or worse, followed by slight atypia, and finally normal sputum (p < 0.(01). Participant age increased with the degree of atypical metaplasia from normal to moderate atypia or worse (p < 0.(01). Smoking status and the degree of sputum atypia were related, with a higher proportion of normal sputum results among never-smokers than among cigarette and water/stem pipe smokers (94.3% vs 85.3%, respectively; p < 0.001). The reverse distribution was observed for moderate atypia or worse, which was Significantly less common in never-smokers than in cigarette and water/stem pipe smokers (0.6% vs 2.6%, respectively; p < 0.(01). The relationship between sputum atypia and lung cancer risk by histology and location is presented in Table 5. Compared to the lung cancer risk associated with normal sputum, the risk was Significantly higher according to the degree of atypia for squamous carcinomas and small cell lung cancer (SCLC), with an adjusted hazard ratio (HR) of 5.70 (95% CI, 3.78 to 8.59) and 3.32 (95% CI, 1.31 to 8.45), respectively. Similarly, the adjusted HR for lung cancer with a central location was 4.93 (95% CI, 3.51 to 6.92). Of note, the degree of sputum cytology atypia was not related to the incidence of adenocarcinoma or other types of carcinoma.
DISCUSSION
In this study, we mainly investigated the association between sputum atypia and lung cancer risk.
Table 3-As8ociation Between Sputum Results and Lung Cancer Incidence Among the 'fTC Sputum Screening Cohort Compare Moderate Atypia or Worse to Normal Variables Baseline Normal Slight atypia Moderate atypia Worse than moderate atypia Moderate atypia or worse Subsequent sputum screening (TO--T3) Baseline TO TO--Tl
Crude HR (95% CI)
Adjusted HR (95% CI)*
Adjusted HR (95% CI)t
1.00 2.12 (1.66-2.71) 4.16 (2.68--6.47) 37.54 (25.49--55.28)
1.00 1.35 (1.05-1.73) 1.92 (1.23-2.99) 16.04 (10.82-23.77)
1.00 1.28 (1.00--1.64) 1.85 (1.19-2.89) 17.05 (11.49-25.30)
8.46 (6.27-11.4.'3)
3.89 (2.87,5.27)
Crude OR (95% CI)
Adjusted OR (95% CI)
6.05 (3.35--10.92) 8.48 (5.31-13.54) 9.48 (6.20--14.49) 9.94 (6.87-14.37)
2.60 (1.41-4.79) 3.93 (2.41-6.41) 4.62 (2.96-7.22) 4.62 (3.14-6.81)
3.82 (2.82-5.18)
TO--T2
TO--T3
*Adjusted for age. radon, arsenic. and smoking. t Adjustt'd for age, radon, arsenic, smoking, gender, education, and prior pulmonary diseases.
782
OriginalResearch
Table 4-AAociation Bet1Deen Rilk Factors of Lung Cancer and Sputum Reaulu Among the "fTC Cohort in Ba8eline Sputum Screening· Baseline Sputum Group
Normal
Slight Atypia
Moderate Atypia or Worse
Radon (median) Arsenic (median) Age at baseline (median), yr Smoking status Never Cigarette only . Water/stem pipe only Both Cigarette or water/stem pipe
183.1 6,513.0 52.0
455.2 8,189.0 60.0
587.3 11,445.0
1,294 (94.3) 2,087 (94.2) 505 (85.4) 4,183 (85.3) 6,775 (87.8)
70(5.1) III (5.0) 70 (11.8) 596 (12.1) 777(10.1)
8(0.6) 17 (0.8) 16(2.7) 127 (2.6) 160 (2.1)
I
63.0
*Data are presented as No. (%) unless otherwise indicated.
Our results showed that age, asthma, and occupational radon and arsenic exposure were independent risk factors for sputum atypia (moderate atypia or worse). In addition, male gender, older age, less education, smoking, occupational radon and arsenic exposure, and prior pulmonary disease were all potential risk factors of lung cancer (incidence or death). These findings were similar to the earlier studies conducted in the ITC and elsewhere. 19-25 This study also indicated that sputum atypia increased the risk of lung cancer in both univariate and multivariate models. There is little controversy that severe sputum atypia may lead to/or be associated with either invasive cancer or carcinoma in situ.26 - 29 However, our study indicated that moderate atypia was also associated with an increased risk of lung cancer. The same association between moderate atypia and lung
cancer risk was also observed in the Colorado study.12 Since all participants were exposed to radon and/or arsenic and most of them had a history of smoking or other risk factors for lung cancer, and all these risk factors were associated with an increase in moderate atypia, it was hard to distinguish moderate atypia as an independent risk factor of lung cancer from occupational exposure or just part of a multistep process in the development of lung cancer. However, after adjustment for other risk factors, there was still an increased risk for lung cancer from moderate atypia or worse, which showed that other risk factors might not account for the entire relationship between sputum atypia and lung cancer risk. Thus, sputum atypia is an independent risk factor for lung cancer, though it might be partly explained by the process of field cancerization in the respiratory epithelium.P?
Table 5-ABsociation Between Sputum &aulta and Lung Cancer Rilk Among the "fTC Sputum Screening Cohort by Hiltology Type and Location Histology Non-small cell Squamous
Adenocarcinoma or other
Small cell
Location Central
Noncentral
Baseline Sputum Result
Crode HR (95% CI)*
Adjusted HR (95% CI) t
Normal Slight atypia Moderate atypia or worse Normal Slight atypia Moderate atypia or worse Normal Slight atypia Moderate atypia or worse
1 2.45 (1.70-3.54) 12.91(8.64-19.27) 1 1.60 (1.07--3.03) 2.94 (1.08--8.03) 1 2.02 (1.02-4.00) 6.93 (2.75-17.44)
1 1.46 (1.01-2.12) 5.70 (3.78-8.59) 1 1.17 (0.70-1.97) 1.39 (0.51--3.80)
Normal Slight atypia Moderate atypia or worse Normal Slight atypia Moderate atypia or worse
1 2.06 (1.51-2.80) 10.82(7.76--15.10) 1 2.35 (1.56-3.54) 4.47 (2.18-9.17)
1 1.24 (0.91-1.69) 4.93 (3.51--6.92)
1
1.29 (0.65-2.58) 3.32 (1.31-8.45)
1
1.44 (0.95-2.17) 1.91 (0.92--3.94)
*Adjusted for age, radon, arsenic, and smoking. HR = hazard ratio. tAdjusted for age, radon, arsenic, smoking, gender, education, and prior pulmonary diseases. www.chestjournal.org
CHEST/135/3/ MARCH, 2009
783
The ITC cohort had prominent occupational characteristics. First, all participants had serious radon and/or arsenic exposure, and second, squamous carcinomas and SCLC were the most common types of lung cancer among the ITC cohort. In our study, although sputum atypia might be an independent risk factor for lung cancer, age, smoking, occupational radon and arsenic exposure appear to be major confounding factors. In the Colorado study,12 there were very small changes in the HR for lung cancer associated with sputum atypia in the multivariate model as compared with of the HR changes in the univariate model. In our study, the crude HR was higher than the adjusted HR after age, smoking, and occupational radon and arsenic exposure were included. The adjusted HRs differed slightly depending on whether the adjustment included age, smoking, radon and arsenic exposure or included these variables plus gender, education, and prior pulmonary disease (Table 3). In addition, Table 4 shows that the distribution of smoking status and median age at baseline, and median occupational radon and arsenic exposure varied with the results of the baseline sputum cytologic examination (p < 0.001). Another implication of our study is that sputum atypia only increases the risk of squamous carcinoma and SCLC, ie, centrally located lung cancer. This may be because the central tumors arise from the larger bronchi, and adenocarcinomas are peripheral and arise from bronchioloalveolar epithelium. Centrally located cancers are likely induced by an inhaled carcinogen that is deposited in the central airway mucosa, thus implicating something in the workplace of these tin workers and not having the same effect on the periphery where adenocarcinomas are typically found. Univariate analysis results showed that lack of education was a potential risk factor for lung cancer. This result might be confounded by occupational radon and arsenic levels. The tin miners with more education were more likely employed in a job with less occupational exposure and higher education. Although sputum cytologyis an insensitive screening tool,31 our results showed that increased lung cancer risk is associated with the frequency of consecutive annual sputum examinations. Similar results are also observed in other studies.32 This might imply that consecutive annual sputum examination increases the chance of detection of sputum atypia and thus, increases the chance of detection of lung cancer. The strength of this study is the methodology. First, it involved a prospective cohort design, Personal and occupational characteristics and sputum samples were all obtained before the diagnosis of lung cancer. Second, both sputum cytologyand chest 784
radiographs were used as the screening tools for the consecutive annual lung cancer screenings. On one hand, this combination of screening tools was helpful in identifying all lung cancer cases during the study period and reducing the number of missed diagnoses. On the other hand, since the two screening examinations were performed on the same day but by two independent departments, it is unlikely that sputum cytology results were affected by chest radiography results or vice versa. In addition, our analysis showed that the level of chest radiograph compliance is very high and consistent in all the subgroups. A limitation of our studyis that information about some risk factors was obtained by self-reporting. Both radon and arsenic exposure level were estimated based on work histories combined with general environmental monitoring, and monitoring data were for areas and not at the individual level. In conclusion, our study of persons at high risk of lung cancer shows that moderate or worse sputum cytologic atypia predicts an increased risk of lung cancer. Sputum atypia might be helpful to identify high-risk individuals who could benefit from more intensive diagnostic examination andlor be enrolled into lung cancer chemoprevention trials. Further studies in different populations and using more sensitive molecular markers are warranted to investigate the relationship between sputum cytologic atypia and molecular markers in the sputum. ACKNOWLEDGMENT: We thank the staff of The Third People's Hospital of Honghe Prefecture, Gejiu City. Yunnan, Chiiia for their assistance in the collection of the follow-up data on the YTC screening cohort. We thank William Hocking, MD, Marshfield Clinic, and Paul Kvale, MD, Henry Ford Hospital, for their critical comments.
REFERENCES 1 Yang L, Parkin OM, Ferlay OM, et al. Estimates of Cancer incidence in China for 2000 and projections for 2005. Cancer Epidemiol Biomarkers Prev 2005; 14:243-250 2 Brenner H. Long-term survival rates of cancer patients achieved by the end of the 20th century: a period analysis. Lancet 2002; 360:1131-1135 3 Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA Cancer J Clin 2006; 56:106-130 4 Mountain CF. The international system for staging lung cancer. Sernin Surg Oncol 2000; 18:106-115 5 Lung cancer screening: recommendation statement. Ann Intern Med 2004; 140:738-739 6 Smith RA, Cokkinides V, Eyre HJ. American Cancer Society guidelines for the early detection of cancer, 2003. CA Cancer JClin 2003; 53:27-43 7 Fontana RS, Sanderson DR, Woolner LB, et al. Lung cancer screening: the Mayo program. J Occup Med 1986; 28:746-
750
8 Kubik AK, Parkin OM, Zatloukal P. Czech study on lung cancer screening: post-trial follow-up of lung cancer deaths up to year 15 since enrollment. Cancer 2000; 89:2363-2368 9 Marcus PM, Bergstralh EJ, Fagerstrom HM, et al. Lung Original ResearcIl
cancer mortality in the Mayo Lung Project: impact of extended follow-up. J Nat! Cancer lnst 2000; 92:1308-1316 10 McWilliams A. Mayo J, MacDonald S, et al. Lung cancer screening: a different paradigm. Am I Respir Crit Care Med 2003; 168:1167-1173 11 Kennedy TC. Proudfoot SP, Franklin WA, et aI. CytopatholOgical analysis of sputum in patients with airflowobstruction and significant smokinghistories. Cancer Res 1996;56:4673-4678 12 Prindiville SA, ByersT, Hirsch FR, et al. Sputum cytologteal atypia as a predictor of incident lung cancer in a cohort of heavy smokers with airflow obstruction. Cancer Epidemiol Biomarkers Prev 2003; 12:987-993 13 Risse EK, Vooijs GP, van't Hof MA. DiagnosticSignificance of "severe dysplasia" in sputum cytology. Acta Cytol 1988; 32:629-634 14 Fontana RS, Sanderson DR, Taylor WF, et al, Early lung cancer detection: results of the initial (prevalence) radiologic and cytologic screening in the Mayo Clinic study. Am Rev Respir Dis 1984; 130:561--565 15 Kennedy TC, Franklin WA. Prindiville SA, et al. High prevalence of endobronchial malignancy in high-riskpatients with moderate dysplasia in sputum. Chest 2004; 125:109S 16 RatnasingheD, Yao SX, Tangrea JA,et al, Polymorphisms of the DNA repair gene XRCCI and lung cancer risk. Cancer Epidemiol Biomarkers Prey 2001; 10:119-123 17 Qiao YL, Taylor PR, Yoo SX, et al. Risk factors and early detection of lung cancer in a cohort of Chinese tin miners. Ann Epidemiol 1997; 7:533--541 18 Hosmer DW, Lemeshow S. AppliedlOgistiC regression. Wiley series in probabilityand statistics. New York, NY: John Wiley & Sons, 1999 19 QiaoYL, TaylorPR, Yao SX, et al. Relationof radon exposure and tobacco use to lung cancer among tin miners in Yunnan Province, China. Am j Ind Med 1989; 16:511-521 20 Taylor PR, Qiao YL, Schatzkin A, et al. Relation of arsenic exposure to lung cancer among tin miners in Yunnan Province, China. Br J Ind Med 1989;46:881-886
www.chestjoumat.org
21 Lubin JH, Qiao YL, TaylorPR, et al, Quantitative evaluation of the radon and lung cancer association in a case control study of Chinese tin miners. Cancer Res 1990; 50:174-180 22 Yao SX, Lubin JR, Qlao YL, et al. Exposureto radon progeny, tobacco use and lung cancer in a case-control study in southern China. Radiat Res 1994; 138:326-336 23 Xuan XZ, Lubin JH, Li ]y, et al. A cohort study in southern China of tin miners exposed to radon and radon decay products. Health Phys 1993; 64:120-131 24 van Loon AJ, Goldbohm RA, van den Brandt PA. Lung cancer: is there an association with socioeconomic status in The Netherlands? J Epidemiol Community Health 1995; 49:65--69 25 MaoY,Hu I. Ugnat AM,et al. Socioeconomic status and lung cancer risk in Canada. Int J Epidemiol 2001; 30:809-817 26 Saccomanno G, Archer VE, Auerbach 0, et al. Development of carcinoma of the lung as reflected in exfoliated cells. Cancer 1974;33:255-270 27 Frost IK, BallWC Ir, Levin ML, et al. Sputum cytopathology: use and potential in monitoring the workplace environment by screening for biological effects of exposure. J Occup Med 1986; 28:692-703 28 Frost JK, Ball WC Jr, Levin ML, et al. Early lung cancer detection: results of the initial (prevalence) radiologtc and cytologiC screening in the Johns Hopkins study. Am Rev Respir Dis 1984; 130:549-554 29 Tockman MS, Gupta PK, Myers ID, et al. Sensitive and specific monoclonal antibody recognition of human lung cancer antigen on preserved sputum cells: a new approach to early lung cancer detection. J Clin OnooI1988; 6:1685-1693 30 Hirsch FR, FranklinWA. Gazdar AF. et al. Early detection of lung cancer: clinical perspectives of recent advances in biologyand radiology. Clin Cancer Res 2001; 7:5-22 31 Frost JK, Fontana RS, Melamed MR, et al. Early lung cancer detection: summary and conclusions. Am Rev Respir Dis 1984; 130:565-570 32 Thunnissen FB. Sputum examination for early detection of lung cancer. J Clin Pathol 2003; 56:805-810
CHEST/135/3/ MARCH, 2009
785
Original Research LUNG CANCER
Association Between Sputum Atypia and Lung Cancer Risk in an Occupational Cohort in Yunnan, China* Ya-Guang Fan, MD; Ping Hu, SM, SeD; YongJiang, MPH; Run-Sheng Chang, MD; Shu-Xiang Yoo, MD; Wendy Wang, PhD; Jie He, MD; Philip Prorok, PhD; and You-Lin Qiao, MD, PhD Background: Individuals with cytologic atypia in sputum may be at high risk for the development of lung cancer. Method.: A prospective cohort study was conducted among occupational tin minen in Yunnan China based on an annual lung cancer screening program. Sputum samples were collected prospectively at baseline and the following seven annual screenings. The associations between risk factors and sputum cytology were analyzed by univariate and multivariate logistic regression. A proportional hazard model was used to analyze the association between the baseline sputum results and the incidence of lung cancer. The effect of consecutive sputum cytology on the increase of lung cancer risk was analyzed by logistic regression. He8ultB: Sputum cytologic atypia was associated with age, smoking, occupational radon and anenic exposure, and asthma. Sputum cytologic atypia was an independent risk factor for lung cancer with an adjusted hazard ratio (RR) of 3.82 (95% confidence interval [CI], 2.82 to 5.18) in comparing normal to moderate or wone atypia. Compared to the lung cancer risk associated with normal sputum, the risk was significantly higher according to the degree of atypia for squamous carcinomas, small cell lung cancer and central lung cancer, with adjusted IIRs of 5.70 (95% CI, 3.78 to 8.59), 3.32 (95% CI, 1.31 to 8.45), and 4.93 (95% CI, 3.51 to 6.92), respectively. Conclusions: Sputum atypia is associated with an increased risk of lung cancer. Sputum cytologic examination combined with other screening examinations may play an important role in the early detection of lung cancer or in the selection of the optimal target population for more intensive lung cancer screening among this occupational cohort or simi1ar population. Trial registration: CIinica1tria1s.gov Identifier: NCT00340405. (CHEST 2009; 135:118-185) Key words: atypia; lung cancer; sputum; Yunnan, China Abbreviations: CI = confidence interval: HR = hazard ratio; OR cancer; TO to 3 = time 0 to 3; YTC = Yunnan Tin Corporation
Lung cancer is the most common cancer worldwide. In 2005 it was estimated that there were > 490,000 new cases of lung cancer in China.' Lung cancer is also a highly lethal disease; the 5-year survival rate for all stages combined is 15% in the United States.2•3 The reasons for the poor prognosis are late diagnosis and lack of effective treatment for symptomatic disease. In addition, the prognosis for lung cancer is related to the stage at which it is detected.' Sputum cytologic examination as a lung cancer screening tool is not recommended by any organization for the general population or any high-risk
na
= odds ratio; Q = quartile; SCLC = small cell lung
population.S" The reason is that prior screening trials 7- 9 have not shown a reduction oflung cancer mortality. Although low-dose CT is considered the most potentially effective screening tool, it may be insensitive to superficial, preinvasive cancers in the central airways that may be manifested in sputum.'? Therefore, biomarkers that might be identifiable in the sputum may complement radiologic imaging for lung cancer. In addition, lung cancer risk has been demonstrated to increase with increasing levels of sputum cytologic dysplasia, particularly at levels of moderate atypia or worse. 1 1- 1S Original~h
The aim of this study was to investigate whether sputum atypia can be used as a predictor of increased lung cancer risk among participants in an occupational cohort in Yunnan, China. Although annual chest radiography was also conducted, we focus our analyses in this report on the results from sputum cytology examinations.
MATERIALS AND METHODS Study Design and Participants The Yunnan Tin Corporation (YTC) study is a one-armed prospective cohort study initiated in 1992 with annual follow-up through 2001. Participants were tin miners in Yunnan, China and with at least10 years of underground radon and/or arsenic exposure. Detailed inclusion criteria are presented elsewhere." At the time of enrollment, all participants completed standardized baseline questionnaires that included demographic characteristics; residential, occupational, smoking, and medical histories; and a dietary assessment component. From 1992 to 1999, a total of 9,295 eligible tin miners were enrolled in this study. Eight rounds of annual lung cancer screenings with chest radiography and sputum cytology were performed during this period. Among them, 9,084 participants underwent at least one sputum cytologic screening examination with adequate results. The YTC study received approval from the institutional review board, and informed consent was obtained for each participant. Tobacco was consumed mainly in the form of cigarettes, water pipes, and Chinese long-stem pipes. The never-smoker was defined as having a smoking history of < 6 months. Screen"From the Department of Biology (Dr. Fan), Institute of Radiation Medicine, Chinese Academy of Medical Sciences, Tianjin, People's Republic of China; Department of Cancer Epidemiology (Drs. Jiang and Qiao), Cancer Institute, Chinese Academy of Medical Sciences, Beijing, People's Republic of China; Third People's Hospital of Honghe Autonomous District (Dr. Chang), Gejiu, Yunnan Province, People's Republic of China; Department of Epidemiology (Dr. Yao), Kunming Medical College, Kunming, Yunnan Province, People's Republic of China; Cancer Biomarkers Research Group (Dr. Wang), Division of Cancer Prevention, National Cancer InstitutelNational Institute of Health, Bethesda, MD; Cancer Institute (Dr. He), Chinese Academy of Medical Sciences, Beijing, People's Republic of China; and Biometry Research Group (Drs. Hu and Prorok), Divisions of Cancer Prevention, National Cancer InstitutelNational Institute of Health, Bethesda, MD. None of the authors have any financial interests in the subject matter or any actual or potential conflicts of interest. This study was supported by the National Cancer Institute! National Institutes of Health, grant No. 263-MQ-511694. This article describes an ancillary study for the National Cancer Institute project "The Study of Early Markers of Lung Cancer Among Tin Miners in Yunnan China," a federally funded registered clinical trial. Manuscript received June 1B,200B;revision accepted September 16,2008. Reproduction of thisarticle is prohibited without written permission from the American College of chest Physicians (www.cliestjournal. org/mtsc/reprints.shtml). Correspondence to: Ping Hu, SM, SeD, Mathematical Statistician, Biometry Research Group, Divisions of Cancer Prevention, National Cancer InstitutelNational Institute of Health, Executive Plaza North, Suite 3131, 6130 Executive Blvd, MSC 7354, Bethesda, MD 20892-7354; e-mail: [email protected] DOl: lO.13781chest.08·1469 www.chestjoumal.org
detected cancers were defined as those diagnosed within 12 months of a positive screen. All other cancers were classified as interval cancers. Detailed definitions of occupational radon and arsenic exposures are given elsewhere.'? In this study, occupational radon and arsenic exposure were grouped into four quartiles (Ql to Q4) based on each individual's cumulative radon or arsenic levels, respectively. Sputum Collection: Processing and Interpretation Sputum samples were collected either spontaneously or by an induction procedure, and stored at room temperature in Saccomanno preservative. Specimens were smeared on glass slides and stained with Papanicolaou stain for routine cytologiC examination. Slides were independently screened by two cytopathologists from the Labor Protection Institute in the YTC. Sputum cytologic screening results were classified into degrees as follows: (1) 1 = negative; (2) 2-1 = slight atypical metaplasia; (3) 2-2 = moderate atypical metaplasia; (4) 2-3 = grave atypical metaplasia; (5) 3 = suspicious for cancer; (6) 4 = highly suspicious for cancer; (7) 5-1 = epidermoid carcinoma; (B) 5-2 = adenocareinoma; (9) 5--3 = undifferentiated carcinoma (includes both large cell and small cell); and (10) other cancer. The positive cytology slides were defined as the screening results of degrees 2-3 and above, and reread by a confirmatory reader. In this article, we focus on moderate atypical metaplasia (degree 2-2) or worse (degrees > 2-2) to investigate whether sputum atypical metaplasia can be used as a predictor of lung cancer. Statistical Methods The associations between risk factors and sputum cytology results were analyzed by univariate and multivariate logistic regression. IS Person-years of follow-up were calculated from the date of the first sputum cytologic screening to the date of lung cancer diagnosis or lung cancer death or censoring as of December 31, 2001. The proportional hazards model was used to analyze the association between baseline sputum results and lung cancer incidence. The effect of consecutive sputum cytology on the increase of lung cancer risk was also analyzed by logistic regression. IS The Kruskal-Wallis test was used to compare participant age, radon and arsenic exposures. The X2 test of independence was used to evaluate the statistical significance of differences in proportions. All statistical tests were two-sided.
RESULTS
In this study, all analyses were restricted to the 9,084 participants who received at least one sputum cytology screening and had adequate results. Characteristics of the study participants are presented in Table 1. Most participants were men (93.9%), with women accounting for only 6.1% of participants. At the time of enrollment, 66.1% of participants were 40 to 59 years old, Of the participants, 23.9% never attended school, and 84.9% had a history of smoking. Table 1 alsoshowscumulative occupationalarsenic and radon exposures and information on prior pulmonary diseases such as chronic bronchitis, asthma, and silicosis. The association between personal/occupational characteristics and sputum cytology results in the baseline sputum screening is also given in Table 1. CHEST/ 135 / 3 / MARCH, 2009
779
Table l-A.aociation Bet1Deen Participant Characteristics and Sputum Cytology Among the fTC Cohort in BoNfine Sputum Screening OR for Moderate Atypia or Worse, Compared to Normal Cytology
Sputum Cytology, No. (%)
Characteristics
All
Gender Female Male Age group, yr 40-49 50-59 60--69 ~70
Education No 056yr >6yr Smoking status Never Cigarette only Water/stem pipe only Both Cigarette or waterl stem pipe Arsenic level Ql (0-1455) Q2 (1,456-7,102) Q3 (7,102-17,435) Q4 (~ 17,435) Radon level Ql (0-82046) Q2 (82.47-205.80) Q3 (205.81-507.29) Q4 (~ 507.29+ ) Asthma No Yes Chronic bronchitis No
Yes Silicosis No Yes
Total Participants, No.(%)
Normal
Slight Atypia
l
Worse Than Moderate Moderate Atypia Atypia'"
Moderate Atypia or Worse
Crude OR (95%CI)
Adjusted OR (95% CI)t
9,084 (100)
8,069 (88.8) 847 (9.3)
124 (1.4)
44(0.5)
168 (1.8)
553 (6.1) 8,531 (93.9)
542 (98.0) 9 (1.6) 7,527 (88.2) 838(9.8)
2 (004) 122 (104)
o(O.O) 44(0.5)
2 (004) 166 (1.9)
I 5.98 (1.48--24.16)
I 1.23 (0.28-5.37)
3,786 (41.7) 2,219(2404) 2,459 (27.1) 620 (6.8)
3,599 (95.1) 177 (4.7) 1,935 (87,2) 244 (11.0) 2,032 (82.6) 348 (14.2) 503 (8Ll) 78 (12.6)
10 (0.3) 30 (104) 53(2.2) 31 (5.0)
0(0.0) 10 (004) 26 (1.1) 8 (1.3)
10(0.3) 40 (1.8) 79(3.3) 39 (6.3)
1 7.44 (3.71-14.91) 13.99 (7.23-27.08) 27.91 (13.84-56.25)
I 4.36 (1.99-9.54) 5.67 (2.55-12.64) 9.36 (3.99-21.95)
2,174 (23.9) 4,376 (48.2) 2,534 (27.9)
1,794 (82.5) 298 (13.7) 3,872 (88.5) 435 (9.9) 2,403 (94.8) 114 (4.5)
62 (2.9) 49 (Ll) 13 (0.5)
20(0.9) 20 (0.5) 4 (0.2)
82(3.8) 69 (1.6) 17(0.7)
1 0,39 (0.28--0.54) 0.16 (0.09-{).26)
1 0.79 (0.57-1.12) 0.80 (0.44-1.47)
1,372 (15.1) 2,215 (2404) 591 (6.5)
1,294 (94.3) 2,087 (94.2) 505 (8504)
70(5.1) III (5.0) 70 (11.8)
8 (0.6) 12 (0.5) 14 (204)
0(0.0) 5 (0.2) 2(0.3)
8(0.6) 17(0.8) 16 (2.7)
1 1.32 (0.57--3.06) 5.13 (2.18--12.05)
1 1047 (0.63--3.47) 1.21 (0.50--2.94)
4,906 (54.0) 7,712 (84.9)
4,183 (85.3) 596 (12.1) 6,775 (87.9) 777 (10.1)
90 (1.8) 116 (1.5)
37 (0.8) 44 (0.6)
127 (2.6) 160 (2.1)
4.91 (2.40-10.06) 3.82 (1.87-7.79)
1.93 (0.92-4.06) 1.75 (0.84--3.65)
2,364 (26.0) 2,248 (24.7) 2,252 (24.8) 2,220 (24.4)
2,236 (94.6) 1,955 (87.0) 1,856(8204) 2,022 (91.1)
118 (5.0) 263 (11.7) 318 (14.1) 148 (6.7)
9(0.4) 21 (0.9) 55 (2.4) 39(1.8)
1 (0.0) 9(0.4) 23 (1.0) II (0.5)
10 (004) 30 (1.3) 78 (3.5) 50(2.3)
1 3.43 (1.67-7.04) 9.40 (4.85-18.20) 5.53 (2.80-10.93)
1 1.06 (0.46-2.44) 2.02 (0.86-4.75) 2.60 (1.10-6.18)
2,340 (25.8) 2,232 (24.6) 2,251 (24.8) 2,261 (24.9)
2,252 (96.2) 78(3.3) 2,084 (93.4) 128 (5.7) 1,949 (86.6) 255 (11.3) 1,784 (78.9) 386 (17.1)
8 (0.3) 17 (0.8) 29 (1.3) 70 (3.1)
2 (0.1) 3(0.1) 18(0.8) 21 (0.9)
10 (004) 20(0.9) 47 (2.1) 91 (4.0)
1 2.16 (1.01-4.63) 5.43 (2.74-10.78) 11.49 (5.96-22.14)
1 3.96 (1.71-9.17) 4.52 (2.10-9.70) 5.26 (2.52-10.96)
8,417 (92.7) 667 (7.3)
7,531 (89.5) 747 (8.9) 538 (80.7) 100 (15.0)
102 (1.2) 22(3.3)
37(0.4) 7(1.0)
139 (1.7) 29(4.3)
1 2.92 (1.94-4.40)
1 1.80 (1.18--2.74)
6,682 (73.6) 2,402 (26.4)
6,087 (9Ll) 497 (7.4) 1,982 (82.5) 350 (14.6)
76 (Ll) 48(2.0)
22 (0.3) 22 (0.9)
98 (1.5) 70 (2.9)
1 2.19 (1.61-2.99)
1 1.22 (0.86-1.73)
8,633 (95.0) 451 (5.0)
7,699 (89.2) 789 (9.1) 370 (82.0) 58 (12.9)
110 (1.3) 14 (3.1)
35(0.4) 9 (2.0)
145 (1.7) 23 (5.1)
1 3.30 (2.10-5.19)
1 1.29 (0.8D-2.07)
"'Worse than moderate atypia is included as "grave atypical metaplasia," "suspicious for cancer," "highly suspicious for cancer." or "any types of carcinoma. " t Adjusted for gender, age. education, smoking, occupational radon and arsenic exposure, and prior pulmonary diseases.
Moderate atypia or worse was found more often in men than women in univariate analysis (odds ratio [OR], 5.98; 95% confidence interval [CI], 1.48 to 24.16), but there was no significant association with gender in multivariate analysis. Risk of moderate or worse atypia was elevated with the increase of age level. There was an inverse relationship between education level and moderate atypia or worse, but it was not statistically significant. Compared with neversmokers, cigarette and water/stem pipe smokers had 780
a significantly higher risk of moderate atypia or worse, according to univariate analysis, This signiflcant result was not observed in the multivariate analysis. There was a positive relationship between moderate atypia or worse and radon exposure level. According to the multivariate analysis, participants with arsenic exposure levels in Q4, had a significantly higher risk of moderate atypia or worse than those with exposure levels in Q1; however, those with levels in Q2 to Q3 did not have a significantlygreater Original Research
risk than those in Q1. As to prior pulmonary diseases, only asthma was a risk factor of moderate atypia or worse. By 2001, 502 lung cancer cases were confirmed. A total of 464 participants with lung cancer received at
least one sputum cytologic screening and had valid diagnosis dates. Of these participants with lung cancer, 438 died by the end of 2001. The 38 participants whose diagnoses were confirmed only by cause of death, without a date of diagnosis, are not
Table 2-Lung Cancer IncidencelDea'" Among ,he YTC Sputum Screening Co'Iaorl Lung Cancer Incidence", 1/100,000 Person-yr
Characteristics
All
Gender Female Male Age group, yr ~9
50-59 60-69 2= 70 Education, yr No s6 >6 Smoking status Never Cigarette only Water/stem pipe Both Cigarette or water/ stem pipe Arseniclevel Ql (0-1455) Q2 (1456-7102) Q3 (7102--17435) Q4(17435+) Radon level Ql (0-82.46) Q2 (82.47-205.80) Q3 (205.81--507.29) Q4 (507.29+) Asthma No Yes Chronic bronchitis No Yes Silicosis No Yes Baselinesputum cytology Normal Slight atypia Moderate atypia Worse than moderate atypia Moderate atypiaor worse
Lung Cancer Incidence Rate
Lung Cancer Death, 1/100,000 Person-yr
Person-yr
Lung Cancer Deaths
Lung Cancer Death Rate
71,937.4
438
608.9
Person-yr
Lung Cancer Cases, No.
71,279.0
464
651.0
4,159.9 67,119.2
2 462
48.1 688.3
1.00 14.32(4.90-41.81)
4161.0 67,776.6
2 436
48.1 643.3
29,404.1 18,880.6 19,069.3 3,925.2
10 114 268 72
34.0 603.8 1,405.4 1,834.3
1.00 17.76(11.08--28.44) 41.32 (28.38-60.16) 53.94(37.30-78.00)
29,419.5 19,031.1 19,417.1 4,069.7
8 98 253 79
27.2 514.9 1,303.0 1,941.2
16,787.5 35,541.8 18,949.8
206
230
1,370.1 579.6 147.8
1.00 0.42 (0.35-0.51) 0.11 (0.~.15)
17,140.6 35,806.4 18,990.4
218 197 23
1,271.8 550.2 121.1
1.00 0.43 (0.52--0.36) 0.10 (0.13-0.07)
1.00 1.62 (0.92--2.85) 7.67 (4.94--11.91) 4.68 (3.06-7.14) 4.07 (2.65-6.24)
10,578.2 16,540.5 4,582.9 40,235.9 61,359.3
19 45 60 314 419
179.61 272.06 1,309.21 780.40 682.86
1.00 1.51 (0.85--2.71) 7.29 (4.67-11.37) 4.34 (2.83--6.67) 3.80 (2.47--5.86)
28
Incidence Rate Ratio(95% CI)
Death Rate Ratio(95% CI)
1.00 13.37(4.50-39.72) 1.00 18.93(11.27-31.82) 47.90 (32.02--71.71) 71.37 (49.05--103.90)
10,549.2 16,477.2 4,486.6 39,766.1 60,729.9
445
180.11 291.31 1,381.89 842.43 732.75
17,912.6 18,876.2 17,626.4 16,864.0
14 86 247 117
78.2 455.6 1,401.3 693.8
1.00 5.83 (3.52-9.66) 17.93(12.13-26.49) 8.88 (5.59-14.09)
17,922.4 19,060.4 17,941.1 17,013.5
13 75 238 112
72.5 393.5 1,326.6 658.3
1.00 5.43 (3.1S-9.25) 18.30(12.22-27.37) 9.08 (5.63-14.63)
17,588.9 17,597.2 18455.0 17,638.0
25 39 134 266
142.1 221.6 726.1 1,508.1
1.00 1.56 (0.91-2.67) 5.11 (3.47-7.52) 10.61(7.63-14.75)
17,625.9 17,673.1 18,668.0 17,970.4
20 33 129 256
113.5 186.7 691.0 1,424.6
1.00' 1.65(0.91-2.99) 6.09 (4.02-9.24) 12.55(8.81-17.90)
66,222.8 5,056.4
398 66
601.0 1,305.3
1.00 2.17 (1.6S-2.81)
66,797.6 5,139.8
373 65
558.4 1,264.6
1.00 2.26 (1.74-2.94)
52,410.4 18,868.7
258 206
492.3 1,091.8
1.00 2.22 (1.85--2.65)
52,825.0 19,112.4
241 197
456.2 1,030.7
1.00 2.26 (1.88-2.72)
67,964.4 3,314.7
409
55
601.8 1,659.3
1.00 2.76 (2.10-3.63)
68,556.9 3,380.5
384
54
560.1 1,597.4
1.00 2.85 (2.16-3.76)
63,219.8 6,969.0 944.7 145.6
336 79 21 28
531.5 1,133.6 2,222.9 19,230.8
1.00 2.13 (1.67-2.72) 4.18 (2.75--6.34) 36.18(28.44-46.04)
63,682.9 7,076.7 988.0 189,9
320 68 22 28
502.5 960.9 2,226.7 14,744.6
1.00 1.91 (1.47-2.48) 4.43 (2.95--6.66) 29.34 (22.76-37.83)
1,090.4
49
4,493.8
8.46 (6.57-10.88)
1,177.9
50
4,244.8
8.45 (6.57-10.86)
19
48
62 335
"Lung cancer cases without date of diagnosis were not analyzed. www.chestjoumal.org
CHEST / 135/ 3 / MARCH, 2009
781
discussed further in this article. Results presented in Table 2 for lung cancer incidence ratios showed that gender, age, radon and arsenic exposure, smoking status, and prior pulmonary diseases were all potential risk factors for lung cancer. We observed an inverse relationship between education level and the risk of lung cancer. The lung cancer incidence ratio for cigarette and water/stem pipe smokers compared to never-smokers was 4.68 (95% CI, 3.06 to 7.14). As to occupational arsenic exposure, the incidence ratios from low to high (Q1 to Q4) were 1.00,5.83 (95% CI, 3.52 to 9.66), 17.93 (95% CI, 12.13 to 26.49), and 8.88 (95% CI, 5.59 to 14.09), respectively. For radon exposure, the incidence ratios increased from 1.00 to 10.61 (95% CI, 7.63 to 14.75). Lung cancer risk was higher among men than women and among participants with prior chronic bronchitis or silicosis. The same pattern for all of the variables discussed above was also observed in the results of the mortality ratios (Table 2). The association between sputum results and lung cancer incidence is displayedin Table 3. In the baseline sputum screening, moderate, worse than moderate atypia, or both were associated with a higher risk of lung cancer than was normal cytology. A total of 4,269 participants received the baseline (time 0 [TO]), and three subsequent annual (TI to T3) sputum screenings. In the multivariate model, the odds ratio increased with the frequency of consecutive sputum screenings from 2.60 (95% CI, 1.41 to 4.79) for one baseline measurement (TO) to 4.62 (95% CI, 3.14 to 6.81) for four measurement (TO to T3) [Table 3]. The associations between occupational risk factors of lung cancer and baseline results of sputum cytol-
ogy are shown in Table 4. Radon and arsenic exposures were highest among participants with moderate atypia or worse, followed by slight atypia, and finally normal sputum (p < 0.(01). Participant age increased with the degree of atypical metaplasia from normal to moderate atypia or worse (p < 0.(01). Smoking status and the degree of sputum atypia were related, with a higher proportion of normal sputum results among never-smokers than among cigarette and water/stem pipe smokers (94.3% vs 85.3%, respectively; p < 0.001). The reverse distribution was observed for moderate atypia or worse, which was Significantly less common in never-smokers than in cigarette and water/stem pipe smokers (0.6% vs 2.6%, respectively; p < 0.(01). The relationship between sputum atypia and lung cancer risk by histology and location is presented in Table 5. Compared to the lung cancer risk associated with normal sputum, the risk was Significantly higher according to the degree of atypia for squamous carcinomas and small cell lung cancer (SCLC), with an adjusted hazard ratio (HR) of 5.70 (95% CI, 3.78 to 8.59) and 3.32 (95% CI, 1.31 to 8.45), respectively. Similarly, the adjusted HR for lung cancer with a central location was 4.93 (95% CI, 3.51 to 6.92). Of note, the degree of sputum cytology atypia was not related to the incidence of adenocarcinoma or other types of carcinoma.
DISCUSSION
In this study, we mainly investigated the association between sputum atypia and lung cancer risk.
Table 3-As8ociation Between Sputum Results and Lung Cancer Incidence Among the 'fTC Sputum Screening Cohort Compare Moderate Atypia or Worse to Normal Variables Baseline Normal Slight atypia Moderate atypia Worse than moderate atypia Moderate atypia or worse Subsequent sputum screening (TO--T3) Baseline TO TO--Tl
Crude HR (95% CI)
Adjusted HR (95% CI)*
Adjusted HR (95% CI)t
1.00 2.12 (1.66-2.71) 4.16 (2.68--6.47) 37.54 (25.49--55.28)
1.00 1.35 (1.05-1.73) 1.92 (1.23-2.99) 16.04 (10.82-23.77)
1.00 1.28 (1.00--1.64) 1.85 (1.19-2.89) 17.05 (11.49-25.30)
8.46 (6.27-11.4.'3)
3.89 (2.87,5.27)
Crude OR (95% CI)
Adjusted OR (95% CI)
6.05 (3.35--10.92) 8.48 (5.31-13.54) 9.48 (6.20--14.49) 9.94 (6.87-14.37)
2.60 (1.41-4.79) 3.93 (2.41-6.41) 4.62 (2.96-7.22) 4.62 (3.14-6.81)
3.82 (2.82-5.18)
TO--T2
TO--T3
*Adjusted for age. radon, arsenic. and smoking. t Adjustt'd for age, radon, arsenic, smoking, gender, education, and prior pulmonary diseases.
782
OriginalResearch
Table 4-AAociation Bet1Deen Rilk Factors of Lung Cancer and Sputum Reaulu Among the "fTC Cohort in Ba8eline Sputum Screening· Baseline Sputum Group
Normal
Slight Atypia
Moderate Atypia or Worse
Radon (median) Arsenic (median) Age at baseline (median), yr Smoking status Never Cigarette only . Water/stem pipe only Both Cigarette or water/stem pipe
183.1 6,513.0 52.0
455.2 8,189.0 60.0
587.3 11,445.0
1,294 (94.3) 2,087 (94.2) 505 (85.4) 4,183 (85.3) 6,775 (87.8)
70(5.1) III (5.0) 70 (11.8) 596 (12.1) 777(10.1)
8(0.6) 17 (0.8) 16(2.7) 127 (2.6) 160 (2.1)
I
63.0
*Data are presented as No. (%) unless otherwise indicated.
Our results showed that age, asthma, and occupational radon and arsenic exposure were independent risk factors for sputum atypia (moderate atypia or worse). In addition, male gender, older age, less education, smoking, occupational radon and arsenic exposure, and prior pulmonary disease were all potential risk factors of lung cancer (incidence or death). These findings were similar to the earlier studies conducted in the ITC and elsewhere. 19-25 This study also indicated that sputum atypia increased the risk of lung cancer in both univariate and multivariate models. There is little controversy that severe sputum atypia may lead to/or be associated with either invasive cancer or carcinoma in situ.26 - 29 However, our study indicated that moderate atypia was also associated with an increased risk of lung cancer. The same association between moderate atypia and lung
cancer risk was also observed in the Colorado study.12 Since all participants were exposed to radon and/or arsenic and most of them had a history of smoking or other risk factors for lung cancer, and all these risk factors were associated with an increase in moderate atypia, it was hard to distinguish moderate atypia as an independent risk factor of lung cancer from occupational exposure or just part of a multistep process in the development of lung cancer. However, after adjustment for other risk factors, there was still an increased risk for lung cancer from moderate atypia or worse, which showed that other risk factors might not account for the entire relationship between sputum atypia and lung cancer risk. Thus, sputum atypia is an independent risk factor for lung cancer, though it might be partly explained by the process of field cancerization in the respiratory epithelium.P?
Table 5-ABsociation Between Sputum &aulta and Lung Cancer Rilk Among the "fTC Sputum Screening Cohort by Hiltology Type and Location Histology Non-small cell Squamous
Adenocarcinoma or other
Small cell
Location Central
Noncentral
Baseline Sputum Result
Crode HR (95% CI)*
Adjusted HR (95% CI) t
Normal Slight atypia Moderate atypia or worse Normal Slight atypia Moderate atypia or worse Normal Slight atypia Moderate atypia or worse
1 2.45 (1.70-3.54) 12.91(8.64-19.27) 1 1.60 (1.07--3.03) 2.94 (1.08--8.03) 1 2.02 (1.02-4.00) 6.93 (2.75-17.44)
1 1.46 (1.01-2.12) 5.70 (3.78-8.59) 1 1.17 (0.70-1.97) 1.39 (0.51--3.80)
Normal Slight atypia Moderate atypia or worse Normal Slight atypia Moderate atypia or worse
1 2.06 (1.51-2.80) 10.82(7.76--15.10) 1 2.35 (1.56-3.54) 4.47 (2.18-9.17)
1 1.24 (0.91-1.69) 4.93 (3.51--6.92)
1
1.29 (0.65-2.58) 3.32 (1.31-8.45)
1
1.44 (0.95-2.17) 1.91 (0.92--3.94)
*Adjusted for age, radon, arsenic, and smoking. HR = hazard ratio. tAdjusted for age, radon, arsenic, smoking, gender, education, and prior pulmonary diseases. www.chestjournal.org
CHEST/135/3/ MARCH, 2009
783
The ITC cohort had prominent occupational characteristics. First, all participants had serious radon and/or arsenic exposure, and second, squamous carcinomas and SCLC were the most common types of lung cancer among the ITC cohort. In our study, although sputum atypia might be an independent risk factor for lung cancer, age, smoking, occupational radon and arsenic exposure appear to be major confounding factors. In the Colorado study,12 there were very small changes in the HR for lung cancer associated with sputum atypia in the multivariate model as compared with of the HR changes in the univariate model. In our study, the crude HR was higher than the adjusted HR after age, smoking, and occupational radon and arsenic exposure were included. The adjusted HRs differed slightly depending on whether the adjustment included age, smoking, radon and arsenic exposure or included these variables plus gender, education, and prior pulmonary disease (Table 3). In addition, Table 4 shows that the distribution of smoking status and median age at baseline, and median occupational radon and arsenic exposure varied with the results of the baseline sputum cytologic examination (p < 0.001). Another implication of our study is that sputum atypia only increases the risk of squamous carcinoma and SCLC, ie, centrally located lung cancer. This may be because the central tumors arise from the larger bronchi, and adenocarcinomas are peripheral and arise from bronchioloalveolar epithelium. Centrally located cancers are likely induced by an inhaled carcinogen that is deposited in the central airway mucosa, thus implicating something in the workplace of these tin workers and not having the same effect on the periphery where adenocarcinomas are typically found. Univariate analysis results showed that lack of education was a potential risk factor for lung cancer. This result might be confounded by occupational radon and arsenic levels. The tin miners with more education were more likely employed in a job with less occupational exposure and higher education. Although sputum cytologyis an insensitive screening tool,31 our results showed that increased lung cancer risk is associated with the frequency of consecutive annual sputum examinations. Similar results are also observed in other studies.32 This might imply that consecutive annual sputum examination increases the chance of detection of sputum atypia and thus, increases the chance of detection of lung cancer. The strength of this study is the methodology. First, it involved a prospective cohort design, Personal and occupational characteristics and sputum samples were all obtained before the diagnosis of lung cancer. Second, both sputum cytologyand chest 784
radiographs were used as the screening tools for the consecutive annual lung cancer screenings. On one hand, this combination of screening tools was helpful in identifying all lung cancer cases during the study period and reducing the number of missed diagnoses. On the other hand, since the two screening examinations were performed on the same day but by two independent departments, it is unlikely that sputum cytology results were affected by chest radiography results or vice versa. In addition, our analysis showed that the level of chest radiograph compliance is very high and consistent in all the subgroups. A limitation of our studyis that information about some risk factors was obtained by self-reporting. Both radon and arsenic exposure level were estimated based on work histories combined with general environmental monitoring, and monitoring data were for areas and not at the individual level. In conclusion, our study of persons at high risk of lung cancer shows that moderate or worse sputum cytologic atypia predicts an increased risk of lung cancer. Sputum atypia might be helpful to identify high-risk individuals who could benefit from more intensive diagnostic examination andlor be enrolled into lung cancer chemoprevention trials. Further studies in different populations and using more sensitive molecular markers are warranted to investigate the relationship between sputum cytologic atypia and molecular markers in the sputum. ACKNOWLEDGMENT: We thank the staff of The Third People's Hospital of Honghe Prefecture, Gejiu City. Yunnan, Chiiia for their assistance in the collection of the follow-up data on the YTC screening cohort. We thank William Hocking, MD, Marshfield Clinic, and Paul Kvale, MD, Henry Ford Hospital, for their critical comments.
REFERENCES 1 Yang L, Parkin OM, Ferlay OM, et al. Estimates of Cancer incidence in China for 2000 and projections for 2005. Cancer Epidemiol Biomarkers Prev 2005; 14:243-250 2 Brenner H. Long-term survival rates of cancer patients achieved by the end of the 20th century: a period analysis. Lancet 2002; 360:1131-1135 3 Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA Cancer J Clin 2006; 56:106-130 4 Mountain CF. The international system for staging lung cancer. Sernin Surg Oncol 2000; 18:106-115 5 Lung cancer screening: recommendation statement. Ann Intern Med 2004; 140:738-739 6 Smith RA, Cokkinides V, Eyre HJ. American Cancer Society guidelines for the early detection of cancer, 2003. CA Cancer JClin 2003; 53:27-43 7 Fontana RS, Sanderson DR, Woolner LB, et al. Lung cancer screening: the Mayo program. J Occup Med 1986; 28:746-
750
8 Kubik AK, Parkin OM, Zatloukal P. Czech study on lung cancer screening: post-trial follow-up of lung cancer deaths up to year 15 since enrollment. Cancer 2000; 89:2363-2368 9 Marcus PM, Bergstralh EJ, Fagerstrom HM, et al. Lung Original ResearcIl
cancer mortality in the Mayo Lung Project: impact of extended follow-up. J Nat! Cancer lnst 2000; 92:1308-1316 10 McWilliams A. Mayo J, MacDonald S, et al. Lung cancer screening: a different paradigm. Am I Respir Crit Care Med 2003; 168:1167-1173 11 Kennedy TC. Proudfoot SP, Franklin WA, et aI. CytopatholOgical analysis of sputum in patients with airflowobstruction and significant smokinghistories. Cancer Res 1996;56:4673-4678 12 Prindiville SA, ByersT, Hirsch FR, et al. Sputum cytologteal atypia as a predictor of incident lung cancer in a cohort of heavy smokers with airflow obstruction. Cancer Epidemiol Biomarkers Prev 2003; 12:987-993 13 Risse EK, Vooijs GP, van't Hof MA. DiagnosticSignificance of "severe dysplasia" in sputum cytology. Acta Cytol 1988; 32:629-634 14 Fontana RS, Sanderson DR, Taylor WF, et al, Early lung cancer detection: results of the initial (prevalence) radiologic and cytologic screening in the Mayo Clinic study. Am Rev Respir Dis 1984; 130:561--565 15 Kennedy TC, Franklin WA. Prindiville SA, et al. High prevalence of endobronchial malignancy in high-riskpatients with moderate dysplasia in sputum. Chest 2004; 125:109S 16 RatnasingheD, Yao SX, Tangrea JA,et al, Polymorphisms of the DNA repair gene XRCCI and lung cancer risk. Cancer Epidemiol Biomarkers Prey 2001; 10:119-123 17 Qiao YL, Taylor PR, Yoo SX, et al. Risk factors and early detection of lung cancer in a cohort of Chinese tin miners. Ann Epidemiol 1997; 7:533--541 18 Hosmer DW, Lemeshow S. AppliedlOgistiC regression. Wiley series in probabilityand statistics. New York, NY: John Wiley & Sons, 1999 19 QiaoYL, TaylorPR, Yao SX, et al. Relationof radon exposure and tobacco use to lung cancer among tin miners in Yunnan Province, China. Am j Ind Med 1989; 16:511-521 20 Taylor PR, Qiao YL, Schatzkin A, et al. Relation of arsenic exposure to lung cancer among tin miners in Yunnan Province, China. Br J Ind Med 1989;46:881-886
www.chestjoumat.org
21 Lubin JH, Qiao YL, TaylorPR, et al, Quantitative evaluation of the radon and lung cancer association in a case control study of Chinese tin miners. Cancer Res 1990; 50:174-180 22 Yao SX, Lubin JR, Qlao YL, et al. Exposureto radon progeny, tobacco use and lung cancer in a case-control study in southern China. Radiat Res 1994; 138:326-336 23 Xuan XZ, Lubin JH, Li ]y, et al. A cohort study in southern China of tin miners exposed to radon and radon decay products. Health Phys 1993; 64:120-131 24 van Loon AJ, Goldbohm RA, van den Brandt PA. Lung cancer: is there an association with socioeconomic status in The Netherlands? J Epidemiol Community Health 1995; 49:65--69 25 MaoY,Hu I. Ugnat AM,et al. Socioeconomic status and lung cancer risk in Canada. Int J Epidemiol 2001; 30:809-817 26 Saccomanno G, Archer VE, Auerbach 0, et al. Development of carcinoma of the lung as reflected in exfoliated cells. Cancer 1974;33:255-270 27 Frost IK, BallWC Ir, Levin ML, et al. Sputum cytopathology: use and potential in monitoring the workplace environment by screening for biological effects of exposure. J Occup Med 1986; 28:692-703 28 Frost JK, Ball WC Jr, Levin ML, et al. Early lung cancer detection: results of the initial (prevalence) radiologtc and cytologiC screening in the Johns Hopkins study. Am Rev Respir Dis 1984; 130:549-554 29 Tockman MS, Gupta PK, Myers ID, et al. Sensitive and specific monoclonal antibody recognition of human lung cancer antigen on preserved sputum cells: a new approach to early lung cancer detection. J Clin OnooI1988; 6:1685-1693 30 Hirsch FR, FranklinWA. Gazdar AF. et al. Early detection of lung cancer: clinical perspectives of recent advances in biologyand radiology. Clin Cancer Res 2001; 7:5-22 31 Frost JK, Fontana RS, Melamed MR, et al. Early lung cancer detection: summary and conclusions. Am Rev Respir Dis 1984; 130:565-570 32 Thunnissen FB. Sputum examination for early detection of lung cancer. J Clin Pathol 2003; 56:805-810
CHEST/135/3/ MARCH, 2009
785