Tobacco Documents Online

The case of chlorine and derivated products Dr Werner Freiesleben

1 THE CASE OF CHLORINE AND DERIVATED PRODUCTS (VCM) Contribution to the international scientific seminar organised by the International Center for a Scientific Ecology in Paris, May 10, 1993 on the question : "Is the concept of linear relationship between dose and effect still a valid model for assessing risk related to low doses of carcinogens ?" by Werner Freiesleben, E.C.V.M., Brussels Introduction The chemical element chlorine has become the centre of a rather controversial dispute in the worldwide search for conditions to keep our global environment inhabitable. While the more extreme environmental activists advocate the attractively simple policy of totally eliminating chlorine as a chemical feedstock and all organo-chlorine chemicals - more moderate environmentalists see only a need for bans on certain chlorine-containing chemicals and the need for a reduction of other organo-chlorines to levels below those known to cause unacceptable risks. The chlor-alkali industry produces presently worldwide close to 40 million tons of chlorine - the majority of which is used in closed-loop operations inside the chemical industry because of its reactivity to bring about a wealth of reactions which - in many cases - would otherwise not be possible. (e.g. There would be no microelectronics without chlorine - no silicon chips, no computers. The disinfection of drinking water with elemental chlorine eliminated practically the waterborne infectious diseases like cholera, typhus, etc. and is considered to have probably been the greatest single advance in promoting human health and well-being, ever.) Elimination of chlorine as chemical feedstock would, moreover, entail irresponsible and severe economic disruptions : About 60% of the activities of chemical industry depend on the availability of chlorine. No other industries would function and be able to operate without the materials supplied by chemical industry which has, thus, a central importance. Among the (roughly) 100.000 chlorine containing substances known there is the full spectrum from thoroughly hazardous candidates for banning to extremely useful, nearly indispensable harmless species. To brand all these

2 compounds dangerous and undesirable would be as simplistic an over-reaction as naive. It's our way of handling materials which determines the danger - just like steel can be an effective tool to save lives or extinguish them. The ordinary chemist is anyway dumb-founded by elemental stereotyping. Why not, then, ban all oxygen-, hydrogen-or carbon-compounds (or other "chemistries") since these elements are present in the most toxic, carcinogenic, teratogenic etc. materials we know of. Wouldn't it be more appropriate - for a fruitful discussion - to classify chemical compounds (whether "natural" or "man-made") according to their environmental or health impact - regardless of the chemical elements they contain and assess/identify problems (and their magnitudes), options (to cope with these problems) and determine priorities ? The Chlorine Institute, Washington D.C. has recently conducted a thorough study with Cantox, Inc., Ontario on the health and environmental aspects of chlorine and its derivatives - especially organo-chlorine compounds. The study - which evaluates and peer-reviews existing scientific literature - will cover as subject areas the environmental fate; b ioconcentrat ion/b ioaccumulation/ food chain interactions; toxicokinetics in various biological systems; aquatic and environmental toxicity; mammalian toxicity as well as epidemiological studies. Toxicology Forum will organize a scientific conference in September 1993 in Washington, D.C. to present and discuss the results among the scientific community.- These results will be published probably this year in a scientific journal. Vinyl chloride monomer (VCM) and PVC use close to 30% of the annual production of chlorine. The health effects of VCM are among the most thoroughly researched risks. It has been, therefore, decided for this seminar to focus on the VCM case. The Vinyl Chloride Case 1. 'sto During the late fifties certain lesions were reported which had been observed with workers - especially autoclave cleaners - in PVC producing plants : acro-osteolysis, Raynaud's syndrome, scierodermat23. Until such findings - in the early period of industrial development and use of VCM and PVC (roughly from 1930) - only symptoms like euphoria, dizziness, headaches and loss of consciousness had been observed at exposure levels of high concentration of VCM in air - along with hazards connected to the lower explosion limit of 38 000 ppm. Trying to reproduce acro-osteolysis in experimental animals P.L. Vio3.a 45 exposed Wistar rats to 30 000 ppm VCM in air

3 and observed that tumours started to appear after the 10th month of treatment. While part of the,industrial/scientific community did not seem to be alarmed about Dr. Viola's results because of the extremely high exposure levels (deemed "unrealistic" at the time) a group of PVC producers (Solvay, ICI, Rhone-Poulenc and Montedison) saw an indication for a risk of cancer from occupational and environmental exposure to VCM and initiated a thorough investigation of its carcinogenicity in laboratory rodents. The study was entrusted to Prof. C. Maltoni who started his work in 1971. (Detailed results of his experiments are tabulated in a book 6.) Late in 1973 Maltoni could not only confirm Viola's results but had convincing evidence of a dose-related incidence of tumours especially of angiosarcomas of the liver 7. A most significant finding was made in January of 1974 when Creech and Johnson a reported that a search of the medical files at the Goodrich plant in Louisville, Kentucky, USA had revealed three cases of death from angiosarcoma of the liver, a very rare form of cancer - among the deceased workers. It was recognised that the cause was likely to be inhalation of VCM at high levels of concentrations over long periods. This finding, in combination with results from experimental studies, initiated an urgent, dramatic and radical worldwide revision of measures for protecting the health of people exposed to VCM and simultaneously led to extensive epidemiological and animal studies. The PVC producing industry changed radically the technology to lower exposure levels - since closing down the production of VCM (about 10 million tons/a at the time worldwide) proved to be a non-viable alternative. The dramatic development might best be illustrated by the weekly-average of VCM concentrations in the atmospheres of 7 polymerisation plants from mid-1974 to spring 1975 : (ppm VCM in air) reported.by A.W. Barnes 9: Plant Mid 1974 Iate 1974 Spring 1975 A 25 15 12 B 10 5 3 C 10 5 4 D 25 13 5 E 2 0 14 4 F 40 12 5 G 30 13 7 To*-days plants reach yearly averages of VCM emissions into the plant atmosphere of 0.2 ppm 10. (See 10 also for detailed description of manufacturing technology for VCM/PVC which I'll not address in this paper.) Costs for the technological improvements as well as some risk/benefit considerations had been reported by W. Freiesleben 11.

4 2. Doses and Effects of VCM 2.1 Popu3.ation groups exposed to VCM Bonnefoy 12 gave an inventory of the population groups liable to be exposed to VCM which shows a steeply- decreasing degree of exposure in the following order : 1. Polymerisation unit workers - the- most heavily exposed because of the many manual operations, frequent opening of equipment and the fact that installations are often situated inside buildings. 2. Workers in monomer production units - only slightly exposed since the continuous process under pressure necessitates sealed systems and the installations are situated in the open air. 3. Workers in plants where PVC is converted into manufactured articles. The exposure levels have always been very low, as the only source of VCM are residual traces released from the resin. 4. People living near plants. 5. Consumers of food and drinks packed in PVC or VC copolymers containing residual monomer. 2.2. Exposure levels. During the early periods of the industrial development and use of VCM and PVC effects of VCM on the central nervous system were recognised in experimental animals and man ~2. The symptoms include euphoria (VC- sniffers), headaches, dizziness and loss of consciousness - manifested in man at concentrations of several thousand ppm. Still in 1970 the official value for maximum allowable concentrations of VCM in air during a lifetime was 500 ppm. Estimated typical exposure levels for people working in PVC polymerisation plants are given in 12 as : ea VCM concentration in air [ppm], S 1945 - 1955 1955 - 1960 1960 - 1970 mid 1973 mid 1974 1000 NJ L:1 ~ 400 - 500 ~ 300 - 400 150 ~ ~ 50 N ~ ~

5 Individual exposure levels might have been much higher. W.J. Nicholsont3 reported that peak exposures may have exceeded 10 000 ppm. Typical procedures for cleaning autoclaves after a polymerisation batch were until the early 70's : 1. Empty the autoclave 2. Purge autoclave with N2/air until residual VCM concentration levels of 300-200 ppm are attained 3. Enter and scratch the crusts of residual PVC from the walls. This scratching liberated occluded quantities of residual VCM and probably gave rise to periodic exposures for autoclave cleaners of several thousand ppm. Such residual quantities depended on the type of resin produced, state and age of the autoclave inner wall etc. Peak exposures and total exposure depended also on the work organisation : whether there was a group of autoclave cleaners proceeding during one shift from one autoclave to the next or whether one group of people stayed with "their" autoclave all the time and did the filling, the reaction, the emptyin4, the purge and the cleaning (before e.g. the next shift did the filling again). In factories with such organisation practices remarkably less cases of angiosarcomas had been observed. Working conditions and general standards of hygiene and housekeeping (I personally saw a polymerisation plant where workers cooled their Coca- Cola bottles with liquid VCM in the 60's) varied from plant to plant. This might be an explanation that in some countries there is a remarkable clustering of cases of angiosarcoma of the liver (ASL) in a small number of factories e.g. cases of ASL in PVC production plants were restricted to 1 out of 16 Japanese factories and 9 out of 28 US factories. (23 out of 33 cases were in just 2 plants.) While average exposure levels of VCM in monomer production plants might have been generally below 10 ppm, individua3l peak exposures during shut-down or maintenance operations might have been very high. Effects of VCM in practice from skin exposure of the liquid phase of VCM are considered negligible 12. Inhalation appears to be the only hazardous intake of VCM. In PVC processing operations - even at the sites of highest exposure levels (hot-mixer) a maximum of 50 ppm had been estimated. . Exposure of the general public must have been always very small. Lahmann 14 measured VCM concentrations at 3 different sites in Berlin and found 0.3 - 0.4 ~,cg/m3. Per smoked cigarette 12.2 ng of VCM were found 12. The average concentration in the Netherlands as a whole (annual average) from emission and dispersion

6 data is 0.00001 ppm (0.2 jLg/m3). Calculated average concentrations close to VCM and PVC plants were found to range 15rom 8g/m3 (1 km) to 1 yv g/m3 (5 km distance) . For the cancer risk in humans, the oral intake of residual VCM in food or drink packed in PVC through migration Til et al. 16 have concluded from animal experiments that a daily intake of 0.1 1,&g of VC/person/day is small enough to be practically neglected. They used a linear model to extrapolate that an oral intake of 0. 4 jAg of VCM/person/day results in a cancer risk of 1 in 106 But they noted that the linear extrapolation model is conservative. The FDA (USA), also using Til's data, said the linear model would exaggerate the risk 12. 2.3. E_pidemiolocry There exist 20 epidemiological studies on the effect of VCM on workers exposed with cohort numbers ranging from 305 to 9677 and follow-up periods (periods over which the expected morbidity or mortality rate is calculated, expressed in person-years) from 8676 to 200 000 person-years, from 10 countries (USA, UK, Germany, Italy, France, Sweden, Norway, Canada, Japan, USSR). The results can be summarized as follows : Occupational exposure to high levels of VCM (hundreds of ppm during several years) causes mortality due to angiosarcoma and possibly other primary liver cancers. There is insufficient evidence to establish any relationship between exposure to VCM and an increased incidence of cancer of the brain, lung, thyroid, lymphatic or haematopoietic tissues, and skin (malignant melarioma). The annual incidence of angiosarcoma of the liver in the general population has been reported to be in the order of 0.25 per million, but may even be lower when stringent pathological criteria are used. In addition to VCM, arsenic and Thorotrast may cause ASL, but for many of the ASL cases studied no plausible cause can be found. Theoretically exposure from the ambient air, in particular near VC or PVC plants, could present a cancer risk but results from the few studies done do not indicate such a possibility or are inconclusive. 2.4. Metabolism of VCM The metabolism and kinetics of VCM have been

7 extensively studied in vitro and in vivo (see literature given in 12). The quantitative studies Zof metabolism in rats by Watanabe 19 and Gehring at 9 different concentrations of VCM, administered orally and by inhalation are considered to be of particular importance since the results help to link between animal and epidemiological studies and provide some understanding of the dose-response relationships in VCM carcinogenicity. The major pathways of VCM metabolism start from the vinyl chloride epoxide which is formed through action of mixed function oxidase. The epoxide can form adducts with macromolecules and finally alkylate the DNA or, through rearrangement to chloro-acetaldehyde or through hydration to the chloro-glycol in both cases end up in the citric acid cycle thereby detoxifying the epoxide. Though Watanabe 19 found a linear relationship between the amount of radioactivity covalently bound to cellular constituents and the level of exposure to radio-labelled VCM even at low levels the mechanisms do not seem to be fully understood by which different metabolic routes are chosen. Also, the oxidation reaction appears to be limited in capacity when the MFO becomes saturated at low dose levels at which point the rate of epoxidation and systemic uptake switch from first-order to pseudo-zero-order kinetics. The different clearance rates of VC from air for different species, reported by Buchter 21 : Species Clearance rate of VC from air (1/h/kg body weight) Man 2.02 Rabbit 2.74 Monkey 3.55 Rat 11.0 Mouse 25.6 indicate that at least different quantities of metabolites are produced in different species - if it is not an indication of even different metabolic pathways. . 2.5. Overall health effects of VCM as assessed by an ECETOC task force in 1988 • VCM has caused primary liver cancer in occupationally exposed human beings, and a

8 A variety of tumours in three species of experimental animals (rat, mouse and hamster). In man, a causal relationship between exposure to VCM in workplace air and the development of angiosarcoma of the liver is well established, but there is insufficient evidence that such exposure is causally-related to the development of tumours at other sites than the liver. There is some evidence that other primary liver tumours were caused by VCM. In animals, the liver is the primary target organ for the toxic and carcinogenic effects of VC, but VC also induces tumours at other sites, for instance the Zymbal gland, the kidney, connective tissue, mammary gland, lung and fore-stomach. ' VCM is mutagenic in vivo and in vitro various assays in the presence of appropriate metabolic activity agents. Its most potent mutagenic metabolite is chloroethylene oxide which has been shown to alkylate DNA and is probably responsible for the mutagenicity and possibly for the carcinogenicity of VCM. From the scant evidence available it seems unlikely that VCM has caused any mutagenic disease in man. ' Exposure to VCM causes chromosomal aberrations in human beings, but only at the levels existing before the marked reduction in occupational exposure in the mid-1970s. It is uncertain whether it alters in the rate of sister chromatid exchange. ' occupational exposure to VCM has led to periportal fibrosis and to changes in the distal vascular tree of the hands followed by the development of Raynaud's phenomenon, scleroderma or acro-osteolysis. There have been no reports of such effects in workers whose exposure to VCM started subsequent to the early 1970s when exposure levels in most countries were lowered to a few ppm. ' Occupational exposure to VCM (by inhalation) has fallen from several hundred ppm, with peaks at several thousand, before the early 1970s, to levels of around 1 ppm subsequently. Although it is not possible to set definitely safe levels of exposure for genotoxic carcinogens, the evidence presented does not suggest that occupational exposure at current levels in compliance with the EEC limit of 3 ppm presents any significant risk to health.

9 At the atmospheric levels to which the general public is exposed (up to 5 ppb, annual average at 1 km, in the vicinity of VC and PVC plants, and less elsewhere), the risk of adverse health effects is even less. ' The content of residual VC in PVC and co-polymers for packaging food and drink has fallen, typically, from about 50 ppm in the early 1970s to below 1 ppm in subsequent years. There has been a corresponding reduction in the VC-content of the food and drink from about 100 ppb, to below 2 ppb from about 1977 onwards. Conservative risk estimates have indicated that the intake of VC from food and drink presents a negligible risk of cancer at current VC levels. 2.6. Predictions of mortality 1976, during-the 4th International Conference of Medichem (Occupational Health in the Chemical Industry) when the knowledge about ASL cases from VCM had been : (I quote) Years of Deaths N' of Deaths Average N' of in Period Deaths/Year Before 1961 0 0 1961 - 1965 2 0.4 1966 - 1970 10 2.0 1971 - 1975 24 4.8 1976 so far 4 > 4.0 i.e. in an atmosphere of dramatic perceptions ("tip-of-iceberg situation") A.W. Barnes gave a paper in which he summarized all knowledge and data available at the time and suggested to make the following corrections (to the otherwise frightening shape of the plot ASL deaths/years) : a) A correction for the probable underrecording in years < 1960 b) A correction for the total numbers .exposed to high concentrations of VCM c) A correction for the decreasing exposures in mid-60's (due to improvements in polymerisation operations) d) A correction for effects of the new standards (e.g. OSHA : 1 ppm) His predictive curve of the "possible true incidence of ASL" (included in Graph) after 17 years proves to have the same characteristics as the curve of the actually recorded cases 18 (see Graph)