- Email: [email protected]
A retrospective lung cancer mortality study of people exposed to insoluble arsenic and radon
LUNQ CANCER ELSEVIER
Lung Cancer 14 Suppl. 1 (1996)Sl37-S148
A retrospective lung cancer mortality study of people exposed to insoluble arsenic and radon Liu Yu-tang*, Chen Zhen Institute
of Occupational
Medicine,
Chinese
Academy
of Preventive
Medicine,
Beijing,
Chinu
Abstract
The incidence of lung cancer for workers who had been exposed to insoluble arsenic in four mines was found to be 290/105. A dose-dependent decrease in the incidence was associated with a reduction in the concentration of insoluble~arsenic in the air. The content of arsenic (expressed as the geometric mean) in the lung of subjects exposed to insoluble arsenic was 51.4 pg/g per dry lung tissue which was 17 times higher than the value of 3.0 pg/g per dry lung tissue found in control groups. Moreover, the content of arsenic was found to correlate with the number of years working in the mine and with the incidence of lung cancer. Metabolic studies of arsenopyrite showed that it is converted to products such as arsenous acid, arsenic acid, methyl arsenate and dimethyl arsenate, which are identical to those generated from As,O,. Although these metabolic products are formed at a lower rate, they nonetheless show that arsenopyrite should be considered a carcinogen. Potential carcinogens such as As, Cr, Ni, Be, and Cd were evaluated in lung specimens of miners with lung cancer and compared with values obtained in controls using logistic regression analysis. Only As was found to be significantly associated with lung cancer. The concentration of As in lung tissues correlated well with the amount found in the air of the mining environment. A retrospective/prospective interference epidemiological investigation performed over a 40-year period showed that the risk of radon had been overestimated. After regulatory measures were implemented in the mines to control for exposure to Rn, the risk of exposure to Rn was found to be RR/WLM (relative risk per working level month) = 0.17%, which was 9 times lower than the values previously estimated. Lung cancer; Insoluble arsenic; Radon; Rn daughters: Arsenic in mines; Tin miners; Metals in mine air
Keywords:
*Corresponding
author.
0169-5002/96/$15.00 0 1996 Elsevier Science Ireland Ltd. All rights reserved SSDI 0169-5002(96)00532-3
S138
L. Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
1. Introduction Arsenic and radon are carcinogens commonly existing in nature in extremely low concentrations. In metal mines, As and Rn usually coexist. The arsenate in the mine deposit is mainly composed of arsenopyrite (FeAsS), which is not considered to be a carcinogen because of its low solubility. Consequently, the etiologic agent for cancer in mines is often attributed to Rn or its daughters. Indeed, the concentration of Rn in many uranium and non-uranium mines is very high, and may be elevated to 4-5 times the recommended working levels (WL). Despite the high levels of radon, the incidence of lung cancer is not increased in these environments. China is a country rich in nonferrous metal mines. High concentrations of insoluble arsenic and radon and its daughters are often detected in the air of many mines. A high incidence of lung cancer is seen among many miners. We have been interested in determining the relationship between insoluble arsenic and lung cancer. The possible confounding by Rn, which coexists with As in ore deposits, was another objective of our research. There were at least two important considerations in our studies: first, if results of our research show Rn to be significantly correlated with the incidence of lung cancer, large amounts of money must be invested to modify mining conditions in order to minimize exposure to Rn. However, if Rn turned out to be not significantly correlated with the risk of lung cancer, then only efforts to protect miners from exposure to As are needed, which would represent savings of 90% monetary investments for remediation. Second, an insight to the relationships between exposure to insoluble As and lung cancer may provide useful leads for the prevention of lung cancer. Our studies show that for effective protection from Rn, positive pressure ventilation should be adopted, whereas wet operation with adequate ventilation affords good protection from As.
2. Results 2. I. Epidemiological investigations of the role of insoluble As in inducing lung cancer 2.1.1. Research on lung cancer in miners in realgar mines Realgar (As&) and orpiment As,& are the main components of the ore. The solubility of orpiment in water is 0.00005% at 18°C. The concentration of As in the mine air (as As) was measured on three separate occasions and the following results were obtained: 1973: n = 6, 0.004-0.577 mg/m3, average 0.23 mg/m3 1981: n = 14, 0.003-0.166 mg/m3, average 0.06 mg/m3; 1988: n = 8, 0.028-1.442 mg/m3, average 0.32 mg/m” A retrospective-prospective cohort study was conducted with follow-up from January 1972 to 1989 [1,2]. The total person years in the prospective study was 6942, which was reduced to 6566 in the follow-up, representing a 5.5% loss. In addition,
L
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
1 (1996)
Sl37-S148
s139
61 persons died and 289 retired or moved during this period. The results showed that: (1) Cause of death: of the 27 cancer deaths, 16 were lung cancers, accounting for 59.2% of the total cancer deaths and raising the possibility that As is a risk factor for lung cancer. (2) Incidence of lung cancer: of the 6566 person years, 17 were diagnosed with lung cancer. Sixteen deaths eventually resulted from these cases and only 1 survived. The incidence of lung cancer was calculated to be 258.8/10’. (3) Relative risk: RR = 20.41, x2 = 15.49, P < 0.01. (4) Standardized mortality rate (SMR): the expected value was 0.3848, the observed value was 16. The SMR was 41.58, P -K 0.01. The mortality in workers with tumor(s) in sites other than the lung was not significantly associated with As. It is well known that the incidence of lung cancer in smelters is associated with the presence of soluble As. Because of the view that the arsenates (As& in realgar mines are not soluble, it is assumed that they do not accumulate in the body and, accordingly, presumed to have no association with lung cancer. Instead, Rn daughters have always been assumed to be the recognized agent for lung cancer. Measurement of Rn daughters in realgar mines, however, showed a value which was within the background level, making it unlikely that there existed a link between Rn daughters and the incidence of lung cancer among realgar miners. Accordingly, the concentrations of various carcinogenic agents in the mining air were measured; inorganic carcinogens were quantified by ICP spectroscopy. In addition to showing the concentration of Rn to be within the background levels, other suspected agents for lung cancer in humans, such as Cr, Ni, and possible carcinogens such as Be, Cd, were also found to be lower than the threshold values (Table 1). An attempt to examine the relationship between smoking and lung cancer yielded inconclusive results because of the small number of cases. Based on the number of cases that were actually collected, no significant association was found, with RR = 3.007, x2, = 1.54, P > 0.05.
Table I Concentration of carcinogens in mine air (mg/m3) conducted in 1988 Carcinogens”
Minimum
Maximum
Average
AS Cr Ni Cd’ Be’
0.0284 n.d.” n.d. o.c002 n.d.
1.4422 0.0004 n.d. 0.0249 0.00001
0.3201 0.0003
“Number of measurements n = 8. %.d., not determined. ‘Possible human lung carcinogens.
0.0047 o.oooo 1
Sl40
L. Yu-tang, C. Zhrn I Lung Cancer I4 Suppl. I (1996) SI37-SI48
2.1.2. Study of lung cancer in tin miners The main arsenate in the three tin mines (Lao (L) mine, Ma (M) mine and Song (S) mine) is arsenopyrite, its solubility in water is 0.0005% at 18°C. The average concentration of As in the air of mines (as As) was measured. A total of 543 air samples were used to measure As concentration. The average concentration of As in the mine air was higher than 0.29 mg/m3 before 1950; 0.29 mg/m3 in the 1950s; 0.022 mg/m3 in the 1960s; 0.015 mg/m3 in the 1970s; and 0.010 mg/m” in the 1980s. These results showed that the concentration of As in the mine air had gradually decreased. An epidemiological study was conducted. Over one thousand cases of male subjects with lung cancer were analyzed in the tin mine study; 90% of these were exposed to insoluble As before 1950. Because of the limited database, a considerable amount of effort went into collecting cases and population controls as far back as possible in order to obtain the crude mortality rate (CMR). In addition, a cohort of 751 persons who started working in the mines between 1960-1969 was established. Follow-up of the cohort continued until 1992. There was a 8.6% loss to follow-up. The CMR was 290/105 in lung cancer cases in which exposure to As occurred before the 195Os, as opposed to a CMR of 150/10s found in lung cancer workers who were exposed to As after 1950. In cases where exposure to As took place only in the beginning of the 196Os,the CMR was only 20/10’. The decrease of lung cancer mortality showed a dose-response relationship to the decrease of insoluble arsenic in the mine environment. Studies were conducted to investigate the possible etiology of lung cancer in tin mine workers [3-53. The concentration of various known carcinogens such as Cr, Ni, polycyclic aromatic hydrocarbons (PAH), or possible carcinogens such as Be and Cd, in the mine air were analyzed. All values were below the threshold levels, regardless of whether the measurements were taken in the early or late stage of the study. The only exception was the concentration of As (Table 2).
Table 2 Concentration investigation Carcinogen
Cr Ni Bea Cd” PAH
of carcinogens/possible carcinogens in mine air measured during different years of the Unit
m8/m3 mkdm3 w/m3 m8/m3 4m3
“Possible human lung carcinogens.
Year 1950
1960
1970
1980
0.010 0.008 0.003 0.005
0.008 0.006
0.003
0.001 to.001 to.001 <0.001
L. Yu-tang.
h’* .J As OHS :
C. Zhen
I Lung
OH I OH-h-OH 11
Cancer
CH,.X v-)
14 Suppl.
I (1996)
CHs I OH-h-OH
,-j
CHs. X
!
Sl41
S137-SI48
CH, OH-A: -CH, II
Fig. 1. Metabolism of As,O,.
The inorganic As in ore mines is mainly composed of FeAsS, known as arsenopyrite, which is formed from FeAsO, .2H,O by oxidation [6]. This compound, as mentioned above, has extremely low solubility and was not considered a carcinogen for a long time. The biological properties of arsenopyrite had not been investigated before 1981. Our work in 1981 demonstrated for the first time that arsenopyrite can be solubilized and metabolized by rats (Fig. 1 and 2). The metabolic products, namely, arsenous acid, arsenic acid, methyl arsenate and dimethyl arsenate are the same as those formed from As203, which is carcinogenic. Compared to the soluble As203, however, arsenopyrite is metabolized at a much slower rate. Nonetheless, the results show that, despite its low solubility, inorganic As may act as a carcinogenic agent for lung cancer in miners. Our work in 1975 [7] found, by chemical analysis, that the content of insoluble As in lungs of patients was 10 times higher than that of the controls. In addition, using electron-probe microanalysis, a large amount of arsenic deposited as insoluble particles was found in the lungs of cancer patients (Fig. 3 and 4). The diameter of
CH.
k-CH. II
Fig. 2. Metabolism of insoluble arsenic.
S142
L. Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
Fig. 3. Demonstration of insoluble arsenic deposits in the lung of a cancer patient.
particles was l-2 p. In some localized areas, the density may be as high as lO/lOO. (Relative density of As/microscopic field.) Our work in 1978 [7], using X-ray spectrometry analysis, found that the insoluble arsenic present in the lung was not in an inert state but was oxidatively metabolized to arsenite and arsenate (Fig. 5 and 6). To further elucidate the involvement of arsenopyrite in the development of lung cancer, the “dose” of As, which referred to the concentration of the putative causative agent actually found in the affected organ, was determined. The following results were obtained. The content of As in the lungs of 42 miners who were exposed to insoluble arsenic and developed lung cancer was 51.4 pg/g per dry lung tissue. The content of As in the lungs of three miners exposed to insoluble arsenic without developing lung cancer was 6.2 pg/g per dry lung tissue. The content of As in the lungs of 38 subjects with no lung disease was 3.0 ,ug/g per dry lung tissue. These results showed that the content of As in the lungs of As-exposed miners was 17 times higher than that of the control group. The amount of As accumulated corresponded with the number of years working in the mine and showed a dose-response
L.
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
SI37-S148
s143
Fig. 4. Characteristic spectrum of arsenic in the lung of a cancer patient.
relationship with the morbidity rate. In summary, although the insoluble As is slowly solubilized and metabolized in the body, it can accumulate over an extended period of time and therefore may exert a pathological effect, e.g., lung cancer which has a characteristic long latency period. In order to ascertain whether other carcinogens or possible carcinogens coexisting with As in the target organ increase the risk of lung cancer due to multi-factor interactions, a logistic regression analysis was performed. Ar.0~ l&20-(4.8702)
21.22’(4.1834)Ar.O.
~6)A*,o‘
3S.SS(2.5231)As.O. 34.65’ (2.SfJ66)AssO.
Fig. 5. X-Ray diffraction analysis of arsenic in a miner’s lung.
S144
L.
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
~.4s63)***o. 34.50’ (2.5875) ;
33.80’(2.6496)Ae,O,
As *O.
37.45’ (2.399OA.,O.
6 4Cb7SS
Fig. 6. X-Ray diffraction analysis of arsenic in a miner’s lung.
A total of 21 male miners with lung cancer, who had worked in the mines for 6-42 years, were matched with 21 controls, i.e., subjects who had cancers in locations other than the lung. The variables considered were the generally recognized carcinogens such as As, Cr, Ni and possible carcinogens such as Be and Cd. They were referred to as Xl, X2, up to X5. A value of 0 was assigned to controls, while a value of 1 was assigned to the cases. EPIPAC software was used in the calculations. The multifactor logistic regression analysis showed that, of the three carcinogens and the two possible carcinogens considered, only As fitted the model and reached statistical significance (Table 3).
Of the 751 subjects that constituted the cohort, 85% were smokers. Through a 29-year follow-up study, the incidence of lung cancer was only 20/105 and was within the expected range. It is noteworthy to mention that miners usually smoked through a bamboo pipe and, thus, only inhaled smoke that had been previously water-filtered. 2.2. Epidemiological investigations of Rn daughters Rn daughters have long been considered to be lung cancer inducing agents. A cumulative dose of 100 WLM (working level month) of Rn daughters is often taken Table 3 Analysis of variables using the logistic model Variable Xl As x2 Cr X3 Ni X4 Be X5 Cd
BETA” 0.8 1678 0.11251 - 0.29078
0.15943 -0.66154
“BETA, regression analysis. bEXP, exponential. ‘S.E., standard error.
EXPb (BETA)
SE. (BETA)
BETA/SE.”
P
0.01851 0.10113
0.21725 0.87614 0.10933
3.7595
0.wO170
0.12841
0.897820 0.790263 0.555202 0.260357
0.97 134
0.11728 0.93558
0.27023 0.58775
0.26597 0.58998 1.1255
L. Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
Sl45
as a level which correlates with the abnormally high incidence of lung cancer in China. The risk of Rn has also been estimated by using an ERR (excess relative risk) lS%/WLM, as recommended by BEIR IV (1988) (risk model developed by the National Research Council, known as the BEIR IV model). Different concentrations of Rn daughters have been found in many mines, including uranium and non-uranium mines. Although many other carcinogens such as As, Cr and Ni are also found in the same locations as Rn, emphasis has always been placed on Rn daughters. In our tin mine study, high concentrations of Rn daughters and insoluble arsenate were present. Thus, to reduce the risk of exposure to these agents, reasonable, economic and effective preventive measures must be taken. Because As and Rn exert their harmful effects in different ways, different preventive measures are needed. To effectively decrease Rn in the mines, positive pressure high ventilation speed is needed, whereas protection from As is adequately achieved with wet operation and negative pressure low speed ventilation. From the economic point of view, energy consumption and thus cost associated with decreasing Rn is 90% higher compared to methods designed to minimize exposure to As. 2.2.1. Retrospective, prospective interference epidemiological study The study was designed to assess the relationship that may exist between changes in concentration of As and Rn in mine air and the incidence of lung disease. Another objective was to study the separate role of these two agents in the etiology of lung cancer. The approach involved first the evaluation of the carcinogenic potential of Rn using epidemiological methods, followed by an estimate of the associated risks using a cumulative “dose” method. The results of the 40-year retrospective/prospective interference epidemiological study (retrospective for 25 years and prospective for 15 years) showed that the concentration of carcinogens and dust in the mine air began to decrease in the mid-1950s after wet operations began to be adopted to prevent pneumoconiosis. In the 1960s the wet operation became fully adopted resulting in As reduction from 0.29 mg/m” in the 50s to 0.015 mg/m3 in the 1970s (Fig. 7). The Rn daughters, naturally decaying products, were not expected to be affected by changes in the production mode of dry versus wet operations. Their concentration remained high (3.1 WLM in “L” mine) from the 1950s to the 1970s the incidence of lung cancer decreased from 150/105 in the 1950s to 20/105 in the late 1980s. Hence, there is no epidemiological evidence to support the association between the incidence of lung cancer and the concentration of Rn daughters. 2.2.2. Radiation cumulative “dose” of Rn daughters and lung cancer The relationship between Rn daughters as a causal agent of lung cancer is based on the cumulative dose of radiation received by the workers. The method for calculating the cumulative dose of Rn daughters is amply illustrated in the literature. As mentioned above, the “dose” of Rn necessary to induce lung cancer is suggested to be 100 WLM in China. Whether such a dose is supported by epidemiological findings is unclear. Thus, in addition to examining the relationship between Rn
L. Yu-tung,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-SI48
0.33 0.29 0.25 0.21
“E 2
0.17
2
0.13 0.09 0.05
Fig. 7. Dose-response relationship between insoluble arsenic and crude death rate (CDR).
daughters and lung cancer, another objective of our study was to find out whether the alleged harmful cumulative dose of Rn is consistent with the data provided by epidemiological studies. Accordingly, mines where there was high exposure to Rn daughters were selected for our study. Exposure to different concentrations of Rn and to other coexisting environmental agents in the mines were evaluated in control groups and subjects with lung cancer. Individuals who began to work in the mine in the 1960s were included in the cohort. Observations continued until 1992. 2.2.3. Radiation cumulative “dose” of Rn daughters Since the installation of a ventilation system in the mines in 1976, the concentration of Rn daughters has been steadily declining: the concentration before 1976 averaged 3.1 WLM; the cumulative total averaged 5.4 WL from 1977 to 1985 (9-year period); the cumulative total averaged 2.7 WL from 1986 to 1991 (6-year period). The year 1965 was chosen as the “median” of years of exposure to Rn. With that as a reference, the cumulative dose was calculated. After 27 years of exposure, the radiation cumulative dose of Rn daughters received by each member in the cohort averaged 619.6 WLM, while the incidence of lung cancer was only 20/10’ (Fig. 8). Thus, even with a cumulative exposure dose of 619 WLM for a total of 27 years, the incidence of lung cancer was still within the expected range. Another group of workers in the same mine whose exposure to Rn began in the 50s and had a greater total exposure of 42 years, had a mean cumulative dose of Rn of 1120 WLM. Taking this set of data as a whole, and comparing it to information provided by the National Institute of Health (NIH) [8], in which the incidence of lung cancer due to radon exposure was reported to be 43.4/105, with an adjusted SMR of 1.72, P < 0.01, it can
L. Yu-rung,
C. Zhen
I Lung
Cancer
14 Suppl.
1 (1996)
s147
S137-S148
3.5
2.5
2 -2 B 0
2.0 s k 1.5
1.0
0.5
Fig. 8. Dose-response relationship between radon and crude death rate (CDR).
be seen that even if one assumes that the cumulative exposure dose was 900 WLM, the adjusted SMR of 1.36 would not be statistically significant (P > 0.05). Because the calculated cumulative dose of 619.6 WLM obtained in our cohort study is below the 900 WLM used in our assumption, the previously reported RR of l.S%/WLM calculated according to BEIR IV was adjusted by a factor of 9, giving a new value of O.l7%/WLM. It is noteworthy that in the new data released by the NIH [S] the RR calculated by BEIR IV (1988) has also been adjusted. These results are very similar to our data in a study conducted in 1990 [S].
3. Discussion
This paper investigates the incidence of lung cancer among miners in four mines who were exposed to insoluble As. Long term exposure to high concentrations of As was associated with a high incidence of lung cancer. Using electron-probe microanalysis we found that the content of insoluble As in the lungs of patients and by chemical analysis found its content was 17 times higher than that of the non-exposed, disease-free persons. Using X-ray spectrometry analysis we found that the insoluble As present in the lung was not in an inert state but it was oxidatively metabolized to arsenite and arsenate. A 40-year interference epidemiological investigation shows that the incidence of
S148
L.
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-5248
lung cancer in tin miners decreased in parallel with a reduction in the concentration of insoluble As in the mine environment, suggesting causality. The so-called insoluble arsenate “arsenopyrite” is solubilized and metabolized in the body. Its metabolic products, i.e., arsenous acid, arsenic acid, methyl arsenate and dimethyl arsenate, are identical to those derived from the generally recognized carcinogen As203, which is consistent with the notion that arsenopyrite may be a carcinogen. The incidence of lung cancer in miners showed a progressive yearly decrease, against a background in which the concentration of Rn daughters remained essentially unchanged. Even when the radiation cumulative exposure dose of Rn daughters averaged 619.6 WLM, the incidence of lung cancer was only 20/105. The data from a 40-year follow-up study of a cohort suggest that the carcinogenic role of Rn was over-estimated in the past. In determining the cause of lung cancer in miners who were simultaneously exposed to multiple factors including Rn, the RR of Rn daughters was calculated to be 0.17% WLM, which was 9 times lower than the past estimate. References [l] [Z] [3] [4] [S]
[6] [7] [S]
Lu YT et al. An epidemiological study on occupational cancer. J Hyg Res 1980; 9(4): lo- 15. Lu YT et al. An epidemiological investigation on occupational cancer in workers exposed to arsenic. Chinese J Ind Hyg and Occup Dis 1986; 4(4): 200-203. Lu YT et al. Chemical etiology research on lung cancer in Yunxi miners. J Hyg Res 1980; 9(4): 15-20. Lu YT et al. On a etiology of lung cancer in Yunxi miners. Chinese J Ind Hyg and Occup Dis 1987; 5(l): 20-21. Lu YT et al. Etiological research on lung cancer excess occurrence in Yunxi miners. The selected papers of the Symposium on Occupational Safety and Health in Asia-Pacific Region: 101. October 7, 1991, Beijing. Lu YT, Chen Z, Wang AD. Metabolic study of insoluble arsenic. J. Hyg Res 1981; lO(4): 50-53. Liu Yu-tang. The chemical etiology study of lung cancer of miners in YTC. University of Hong Kong, 2TH. Discussion on preventive medicine in China-cancer, early diagnosis and control, April 1986. Lubin JH et al. Radon and lung cancer risk. US Department of Health and Human Services. Public Health Service National Institute of Health NIH Publication No. 94-3644. 1994.
Lung Cancer 14 Suppl. 1 (1996)Sl37-S148
A retrospective lung cancer mortality study of people exposed to insoluble arsenic and radon Liu Yu-tang*, Chen Zhen Institute
of Occupational
Medicine,
Chinese
Academy
of Preventive
Medicine,
Beijing,
Chinu
Abstract
The incidence of lung cancer for workers who had been exposed to insoluble arsenic in four mines was found to be 290/105. A dose-dependent decrease in the incidence was associated with a reduction in the concentration of insoluble~arsenic in the air. The content of arsenic (expressed as the geometric mean) in the lung of subjects exposed to insoluble arsenic was 51.4 pg/g per dry lung tissue which was 17 times higher than the value of 3.0 pg/g per dry lung tissue found in control groups. Moreover, the content of arsenic was found to correlate with the number of years working in the mine and with the incidence of lung cancer. Metabolic studies of arsenopyrite showed that it is converted to products such as arsenous acid, arsenic acid, methyl arsenate and dimethyl arsenate, which are identical to those generated from As,O,. Although these metabolic products are formed at a lower rate, they nonetheless show that arsenopyrite should be considered a carcinogen. Potential carcinogens such as As, Cr, Ni, Be, and Cd were evaluated in lung specimens of miners with lung cancer and compared with values obtained in controls using logistic regression analysis. Only As was found to be significantly associated with lung cancer. The concentration of As in lung tissues correlated well with the amount found in the air of the mining environment. A retrospective/prospective interference epidemiological investigation performed over a 40-year period showed that the risk of radon had been overestimated. After regulatory measures were implemented in the mines to control for exposure to Rn, the risk of exposure to Rn was found to be RR/WLM (relative risk per working level month) = 0.17%, which was 9 times lower than the values previously estimated. Lung cancer; Insoluble arsenic; Radon; Rn daughters: Arsenic in mines; Tin miners; Metals in mine air
Keywords:
*Corresponding
author.
0169-5002/96/$15.00 0 1996 Elsevier Science Ireland Ltd. All rights reserved SSDI 0169-5002(96)00532-3
S138
L. Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
1. Introduction Arsenic and radon are carcinogens commonly existing in nature in extremely low concentrations. In metal mines, As and Rn usually coexist. The arsenate in the mine deposit is mainly composed of arsenopyrite (FeAsS), which is not considered to be a carcinogen because of its low solubility. Consequently, the etiologic agent for cancer in mines is often attributed to Rn or its daughters. Indeed, the concentration of Rn in many uranium and non-uranium mines is very high, and may be elevated to 4-5 times the recommended working levels (WL). Despite the high levels of radon, the incidence of lung cancer is not increased in these environments. China is a country rich in nonferrous metal mines. High concentrations of insoluble arsenic and radon and its daughters are often detected in the air of many mines. A high incidence of lung cancer is seen among many miners. We have been interested in determining the relationship between insoluble arsenic and lung cancer. The possible confounding by Rn, which coexists with As in ore deposits, was another objective of our research. There were at least two important considerations in our studies: first, if results of our research show Rn to be significantly correlated with the incidence of lung cancer, large amounts of money must be invested to modify mining conditions in order to minimize exposure to Rn. However, if Rn turned out to be not significantly correlated with the risk of lung cancer, then only efforts to protect miners from exposure to As are needed, which would represent savings of 90% monetary investments for remediation. Second, an insight to the relationships between exposure to insoluble As and lung cancer may provide useful leads for the prevention of lung cancer. Our studies show that for effective protection from Rn, positive pressure ventilation should be adopted, whereas wet operation with adequate ventilation affords good protection from As.
2. Results 2. I. Epidemiological investigations of the role of insoluble As in inducing lung cancer 2.1.1. Research on lung cancer in miners in realgar mines Realgar (As&) and orpiment As,& are the main components of the ore. The solubility of orpiment in water is 0.00005% at 18°C. The concentration of As in the mine air (as As) was measured on three separate occasions and the following results were obtained: 1973: n = 6, 0.004-0.577 mg/m3, average 0.23 mg/m3 1981: n = 14, 0.003-0.166 mg/m3, average 0.06 mg/m3; 1988: n = 8, 0.028-1.442 mg/m3, average 0.32 mg/m” A retrospective-prospective cohort study was conducted with follow-up from January 1972 to 1989 [1,2]. The total person years in the prospective study was 6942, which was reduced to 6566 in the follow-up, representing a 5.5% loss. In addition,
L
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
1 (1996)
Sl37-S148
s139
61 persons died and 289 retired or moved during this period. The results showed that: (1) Cause of death: of the 27 cancer deaths, 16 were lung cancers, accounting for 59.2% of the total cancer deaths and raising the possibility that As is a risk factor for lung cancer. (2) Incidence of lung cancer: of the 6566 person years, 17 were diagnosed with lung cancer. Sixteen deaths eventually resulted from these cases and only 1 survived. The incidence of lung cancer was calculated to be 258.8/10’. (3) Relative risk: RR = 20.41, x2 = 15.49, P < 0.01. (4) Standardized mortality rate (SMR): the expected value was 0.3848, the observed value was 16. The SMR was 41.58, P -K 0.01. The mortality in workers with tumor(s) in sites other than the lung was not significantly associated with As. It is well known that the incidence of lung cancer in smelters is associated with the presence of soluble As. Because of the view that the arsenates (As& in realgar mines are not soluble, it is assumed that they do not accumulate in the body and, accordingly, presumed to have no association with lung cancer. Instead, Rn daughters have always been assumed to be the recognized agent for lung cancer. Measurement of Rn daughters in realgar mines, however, showed a value which was within the background level, making it unlikely that there existed a link between Rn daughters and the incidence of lung cancer among realgar miners. Accordingly, the concentrations of various carcinogenic agents in the mining air were measured; inorganic carcinogens were quantified by ICP spectroscopy. In addition to showing the concentration of Rn to be within the background levels, other suspected agents for lung cancer in humans, such as Cr, Ni, and possible carcinogens such as Be, Cd, were also found to be lower than the threshold values (Table 1). An attempt to examine the relationship between smoking and lung cancer yielded inconclusive results because of the small number of cases. Based on the number of cases that were actually collected, no significant association was found, with RR = 3.007, x2, = 1.54, P > 0.05.
Table I Concentration of carcinogens in mine air (mg/m3) conducted in 1988 Carcinogens”
Minimum
Maximum
Average
AS Cr Ni Cd’ Be’
0.0284 n.d.” n.d. o.c002 n.d.
1.4422 0.0004 n.d. 0.0249 0.00001
0.3201 0.0003
“Number of measurements n = 8. %.d., not determined. ‘Possible human lung carcinogens.
0.0047 o.oooo 1
Sl40
L. Yu-tang, C. Zhrn I Lung Cancer I4 Suppl. I (1996) SI37-SI48
2.1.2. Study of lung cancer in tin miners The main arsenate in the three tin mines (Lao (L) mine, Ma (M) mine and Song (S) mine) is arsenopyrite, its solubility in water is 0.0005% at 18°C. The average concentration of As in the air of mines (as As) was measured. A total of 543 air samples were used to measure As concentration. The average concentration of As in the mine air was higher than 0.29 mg/m3 before 1950; 0.29 mg/m3 in the 1950s; 0.022 mg/m3 in the 1960s; 0.015 mg/m3 in the 1970s; and 0.010 mg/m” in the 1980s. These results showed that the concentration of As in the mine air had gradually decreased. An epidemiological study was conducted. Over one thousand cases of male subjects with lung cancer were analyzed in the tin mine study; 90% of these were exposed to insoluble As before 1950. Because of the limited database, a considerable amount of effort went into collecting cases and population controls as far back as possible in order to obtain the crude mortality rate (CMR). In addition, a cohort of 751 persons who started working in the mines between 1960-1969 was established. Follow-up of the cohort continued until 1992. There was a 8.6% loss to follow-up. The CMR was 290/105 in lung cancer cases in which exposure to As occurred before the 195Os, as opposed to a CMR of 150/10s found in lung cancer workers who were exposed to As after 1950. In cases where exposure to As took place only in the beginning of the 196Os,the CMR was only 20/10’. The decrease of lung cancer mortality showed a dose-response relationship to the decrease of insoluble arsenic in the mine environment. Studies were conducted to investigate the possible etiology of lung cancer in tin mine workers [3-53. The concentration of various known carcinogens such as Cr, Ni, polycyclic aromatic hydrocarbons (PAH), or possible carcinogens such as Be and Cd, in the mine air were analyzed. All values were below the threshold levels, regardless of whether the measurements were taken in the early or late stage of the study. The only exception was the concentration of As (Table 2).
Table 2 Concentration investigation Carcinogen
Cr Ni Bea Cd” PAH
of carcinogens/possible carcinogens in mine air measured during different years of the Unit
m8/m3 mkdm3 w/m3 m8/m3 4m3
“Possible human lung carcinogens.
Year 1950
1960
1970
1980
0.010 0.008 0.003 0.005
0.008 0.006
0.003
0.001 to.001 to.001 <0.001
L. Yu-tang.
h’* .J As OHS :
C. Zhen
I Lung
OH I OH-h-OH 11
Cancer
CH,.X v-)
14 Suppl.
I (1996)
CHs I OH-h-OH
,-j
CHs. X
!
Sl41
S137-SI48
CH, OH-A: -CH, II
Fig. 1. Metabolism of As,O,.
The inorganic As in ore mines is mainly composed of FeAsS, known as arsenopyrite, which is formed from FeAsO, .2H,O by oxidation [6]. This compound, as mentioned above, has extremely low solubility and was not considered a carcinogen for a long time. The biological properties of arsenopyrite had not been investigated before 1981. Our work in 1981 demonstrated for the first time that arsenopyrite can be solubilized and metabolized by rats (Fig. 1 and 2). The metabolic products, namely, arsenous acid, arsenic acid, methyl arsenate and dimethyl arsenate are the same as those formed from As203, which is carcinogenic. Compared to the soluble As203, however, arsenopyrite is metabolized at a much slower rate. Nonetheless, the results show that, despite its low solubility, inorganic As may act as a carcinogenic agent for lung cancer in miners. Our work in 1975 [7] found, by chemical analysis, that the content of insoluble As in lungs of patients was 10 times higher than that of the controls. In addition, using electron-probe microanalysis, a large amount of arsenic deposited as insoluble particles was found in the lungs of cancer patients (Fig. 3 and 4). The diameter of
CH.
k-CH. II
Fig. 2. Metabolism of insoluble arsenic.
S142
L. Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
Fig. 3. Demonstration of insoluble arsenic deposits in the lung of a cancer patient.
particles was l-2 p. In some localized areas, the density may be as high as lO/lOO. (Relative density of As/microscopic field.) Our work in 1978 [7], using X-ray spectrometry analysis, found that the insoluble arsenic present in the lung was not in an inert state but was oxidatively metabolized to arsenite and arsenate (Fig. 5 and 6). To further elucidate the involvement of arsenopyrite in the development of lung cancer, the “dose” of As, which referred to the concentration of the putative causative agent actually found in the affected organ, was determined. The following results were obtained. The content of As in the lungs of 42 miners who were exposed to insoluble arsenic and developed lung cancer was 51.4 pg/g per dry lung tissue. The content of As in the lungs of three miners exposed to insoluble arsenic without developing lung cancer was 6.2 pg/g per dry lung tissue. The content of As in the lungs of 38 subjects with no lung disease was 3.0 ,ug/g per dry lung tissue. These results showed that the content of As in the lungs of As-exposed miners was 17 times higher than that of the control group. The amount of As accumulated corresponded with the number of years working in the mine and showed a dose-response
L.
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
SI37-S148
s143
Fig. 4. Characteristic spectrum of arsenic in the lung of a cancer patient.
relationship with the morbidity rate. In summary, although the insoluble As is slowly solubilized and metabolized in the body, it can accumulate over an extended period of time and therefore may exert a pathological effect, e.g., lung cancer which has a characteristic long latency period. In order to ascertain whether other carcinogens or possible carcinogens coexisting with As in the target organ increase the risk of lung cancer due to multi-factor interactions, a logistic regression analysis was performed. Ar.0~ l&20-(4.8702)
21.22’(4.1834)Ar.O.
~6)A*,o‘
3S.SS(2.5231)As.O. 34.65’ (2.SfJ66)AssO.
Fig. 5. X-Ray diffraction analysis of arsenic in a miner’s lung.
S144
L.
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
~.4s63)***o. 34.50’ (2.5875) ;
33.80’(2.6496)Ae,O,
As *O.
37.45’ (2.399OA.,O.
6 4Cb7SS
Fig. 6. X-Ray diffraction analysis of arsenic in a miner’s lung.
A total of 21 male miners with lung cancer, who had worked in the mines for 6-42 years, were matched with 21 controls, i.e., subjects who had cancers in locations other than the lung. The variables considered were the generally recognized carcinogens such as As, Cr, Ni and possible carcinogens such as Be and Cd. They were referred to as Xl, X2, up to X5. A value of 0 was assigned to controls, while a value of 1 was assigned to the cases. EPIPAC software was used in the calculations. The multifactor logistic regression analysis showed that, of the three carcinogens and the two possible carcinogens considered, only As fitted the model and reached statistical significance (Table 3).
Of the 751 subjects that constituted the cohort, 85% were smokers. Through a 29-year follow-up study, the incidence of lung cancer was only 20/105 and was within the expected range. It is noteworthy to mention that miners usually smoked through a bamboo pipe and, thus, only inhaled smoke that had been previously water-filtered. 2.2. Epidemiological investigations of Rn daughters Rn daughters have long been considered to be lung cancer inducing agents. A cumulative dose of 100 WLM (working level month) of Rn daughters is often taken Table 3 Analysis of variables using the logistic model Variable Xl As x2 Cr X3 Ni X4 Be X5 Cd
BETA” 0.8 1678 0.11251 - 0.29078
0.15943 -0.66154
“BETA, regression analysis. bEXP, exponential. ‘S.E., standard error.
EXPb (BETA)
SE. (BETA)
BETA/SE.”
P
0.01851 0.10113
0.21725 0.87614 0.10933
3.7595
0.wO170
0.12841
0.897820 0.790263 0.555202 0.260357
0.97 134
0.11728 0.93558
0.27023 0.58775
0.26597 0.58998 1.1255
L. Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-S148
Sl45
as a level which correlates with the abnormally high incidence of lung cancer in China. The risk of Rn has also been estimated by using an ERR (excess relative risk) lS%/WLM, as recommended by BEIR IV (1988) (risk model developed by the National Research Council, known as the BEIR IV model). Different concentrations of Rn daughters have been found in many mines, including uranium and non-uranium mines. Although many other carcinogens such as As, Cr and Ni are also found in the same locations as Rn, emphasis has always been placed on Rn daughters. In our tin mine study, high concentrations of Rn daughters and insoluble arsenate were present. Thus, to reduce the risk of exposure to these agents, reasonable, economic and effective preventive measures must be taken. Because As and Rn exert their harmful effects in different ways, different preventive measures are needed. To effectively decrease Rn in the mines, positive pressure high ventilation speed is needed, whereas protection from As is adequately achieved with wet operation and negative pressure low speed ventilation. From the economic point of view, energy consumption and thus cost associated with decreasing Rn is 90% higher compared to methods designed to minimize exposure to As. 2.2.1. Retrospective, prospective interference epidemiological study The study was designed to assess the relationship that may exist between changes in concentration of As and Rn in mine air and the incidence of lung disease. Another objective was to study the separate role of these two agents in the etiology of lung cancer. The approach involved first the evaluation of the carcinogenic potential of Rn using epidemiological methods, followed by an estimate of the associated risks using a cumulative “dose” method. The results of the 40-year retrospective/prospective interference epidemiological study (retrospective for 25 years and prospective for 15 years) showed that the concentration of carcinogens and dust in the mine air began to decrease in the mid-1950s after wet operations began to be adopted to prevent pneumoconiosis. In the 1960s the wet operation became fully adopted resulting in As reduction from 0.29 mg/m” in the 50s to 0.015 mg/m3 in the 1970s (Fig. 7). The Rn daughters, naturally decaying products, were not expected to be affected by changes in the production mode of dry versus wet operations. Their concentration remained high (3.1 WLM in “L” mine) from the 1950s to the 1970s the incidence of lung cancer decreased from 150/105 in the 1950s to 20/105 in the late 1980s. Hence, there is no epidemiological evidence to support the association between the incidence of lung cancer and the concentration of Rn daughters. 2.2.2. Radiation cumulative “dose” of Rn daughters and lung cancer The relationship between Rn daughters as a causal agent of lung cancer is based on the cumulative dose of radiation received by the workers. The method for calculating the cumulative dose of Rn daughters is amply illustrated in the literature. As mentioned above, the “dose” of Rn necessary to induce lung cancer is suggested to be 100 WLM in China. Whether such a dose is supported by epidemiological findings is unclear. Thus, in addition to examining the relationship between Rn
L. Yu-tung,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-SI48
0.33 0.29 0.25 0.21
“E 2
0.17
2
0.13 0.09 0.05
Fig. 7. Dose-response relationship between insoluble arsenic and crude death rate (CDR).
daughters and lung cancer, another objective of our study was to find out whether the alleged harmful cumulative dose of Rn is consistent with the data provided by epidemiological studies. Accordingly, mines where there was high exposure to Rn daughters were selected for our study. Exposure to different concentrations of Rn and to other coexisting environmental agents in the mines were evaluated in control groups and subjects with lung cancer. Individuals who began to work in the mine in the 1960s were included in the cohort. Observations continued until 1992. 2.2.3. Radiation cumulative “dose” of Rn daughters Since the installation of a ventilation system in the mines in 1976, the concentration of Rn daughters has been steadily declining: the concentration before 1976 averaged 3.1 WLM; the cumulative total averaged 5.4 WL from 1977 to 1985 (9-year period); the cumulative total averaged 2.7 WL from 1986 to 1991 (6-year period). The year 1965 was chosen as the “median” of years of exposure to Rn. With that as a reference, the cumulative dose was calculated. After 27 years of exposure, the radiation cumulative dose of Rn daughters received by each member in the cohort averaged 619.6 WLM, while the incidence of lung cancer was only 20/10’ (Fig. 8). Thus, even with a cumulative exposure dose of 619 WLM for a total of 27 years, the incidence of lung cancer was still within the expected range. Another group of workers in the same mine whose exposure to Rn began in the 50s and had a greater total exposure of 42 years, had a mean cumulative dose of Rn of 1120 WLM. Taking this set of data as a whole, and comparing it to information provided by the National Institute of Health (NIH) [8], in which the incidence of lung cancer due to radon exposure was reported to be 43.4/105, with an adjusted SMR of 1.72, P < 0.01, it can
L. Yu-rung,
C. Zhen
I Lung
Cancer
14 Suppl.
1 (1996)
s147
S137-S148
3.5
2.5
2 -2 B 0
2.0 s k 1.5
1.0
0.5
Fig. 8. Dose-response relationship between radon and crude death rate (CDR).
be seen that even if one assumes that the cumulative exposure dose was 900 WLM, the adjusted SMR of 1.36 would not be statistically significant (P > 0.05). Because the calculated cumulative dose of 619.6 WLM obtained in our cohort study is below the 900 WLM used in our assumption, the previously reported RR of l.S%/WLM calculated according to BEIR IV was adjusted by a factor of 9, giving a new value of O.l7%/WLM. It is noteworthy that in the new data released by the NIH [S] the RR calculated by BEIR IV (1988) has also been adjusted. These results are very similar to our data in a study conducted in 1990 [S].
3. Discussion
This paper investigates the incidence of lung cancer among miners in four mines who were exposed to insoluble As. Long term exposure to high concentrations of As was associated with a high incidence of lung cancer. Using electron-probe microanalysis we found that the content of insoluble As in the lungs of patients and by chemical analysis found its content was 17 times higher than that of the non-exposed, disease-free persons. Using X-ray spectrometry analysis we found that the insoluble As present in the lung was not in an inert state but it was oxidatively metabolized to arsenite and arsenate. A 40-year interference epidemiological investigation shows that the incidence of
S148
L.
Yu-tang,
C. Zhen
I Lung
Cancer
14 Suppl.
I (1996)
S137-5248
lung cancer in tin miners decreased in parallel with a reduction in the concentration of insoluble As in the mine environment, suggesting causality. The so-called insoluble arsenate “arsenopyrite” is solubilized and metabolized in the body. Its metabolic products, i.e., arsenous acid, arsenic acid, methyl arsenate and dimethyl arsenate, are identical to those derived from the generally recognized carcinogen As203, which is consistent with the notion that arsenopyrite may be a carcinogen. The incidence of lung cancer in miners showed a progressive yearly decrease, against a background in which the concentration of Rn daughters remained essentially unchanged. Even when the radiation cumulative exposure dose of Rn daughters averaged 619.6 WLM, the incidence of lung cancer was only 20/105. The data from a 40-year follow-up study of a cohort suggest that the carcinogenic role of Rn was over-estimated in the past. In determining the cause of lung cancer in miners who were simultaneously exposed to multiple factors including Rn, the RR of Rn daughters was calculated to be 0.17% WLM, which was 9 times lower than the past estimate. References [l] [Z] [3] [4] [S]
[6] [7] [S]
Lu YT et al. An epidemiological study on occupational cancer. J Hyg Res 1980; 9(4): lo- 15. Lu YT et al. An epidemiological investigation on occupational cancer in workers exposed to arsenic. Chinese J Ind Hyg and Occup Dis 1986; 4(4): 200-203. Lu YT et al. Chemical etiology research on lung cancer in Yunxi miners. J Hyg Res 1980; 9(4): 15-20. Lu YT et al. On a etiology of lung cancer in Yunxi miners. Chinese J Ind Hyg and Occup Dis 1987; 5(l): 20-21. Lu YT et al. Etiological research on lung cancer excess occurrence in Yunxi miners. The selected papers of the Symposium on Occupational Safety and Health in Asia-Pacific Region: 101. October 7, 1991, Beijing. Lu YT, Chen Z, Wang AD. Metabolic study of insoluble arsenic. J. Hyg Res 1981; lO(4): 50-53. Liu Yu-tang. The chemical etiology study of lung cancer of miners in YTC. University of Hong Kong, 2TH. Discussion on preventive medicine in China-cancer, early diagnosis and control, April 1986. Lubin JH et al. Radon and lung cancer risk. US Department of Health and Human Services. Public Health Service National Institute of Health NIH Publication No. 94-3644. 1994.