Document GKpB71Lgr6XO9Z9a3Q61r3Owx

1. BENZENE: AN HISTORICAL PERSPECTIVE OF PUBLIC HEALTH RESPONSE IN RELATION TO KNOWLEDGE OF TOXICITY IN THE AMERICAN AND EUROPEAN OCCUPATIONAL SETTING 1.1. EARLY WARNING OF THE DANGERS 1.1.1. Early reports demonstrate benzene is apowerful bone marrowpoison Since the 1897report of Santessen, who observed aplastic anemia (AA) among young women engaged in the manufacture of bicycle tires in Sweden and the report in th.e same year by Lenoir and Claude, who observed hemorrhaging in a young man engaged in a dry cleaning operation in France, benzene has been known to be a powerful bone marrow poison. Similar reports of workers developing benzene related diseases of the bone marrow increased dramatically through the first half of the 20th century. Between 1910 and 1914,the first major use of benzene as a solvent in the rubber industry took place. Production of benzene also was stimulated greatly by the demand for toluene in the manufacturing of explosives during World War I. Expanded use of benzene in industry after the War led to an increased use of benzene as a solvent in the artificial leather industry, rubber goods, glue manufacturing, hat manufacturing, rotogravure printing, paint, adhesives, coatings, dry cleaning, automobile manufacturing, tin can assembly, as a starting material in organic synthesis, in petroleum products and in the blending of motor fhels. This expansion in the industrial uses of benzene was accompanied by a vast increase in the numbers of reported cases of AA, generally referred to as "benzene poisoning." Some 1 individuals were diagnosed with benzene poisoning within a few weeks of initial employment and some died within a few months of beginning their jobs (Hogan and Schrader, 1923). These poisonings were associated with benzene levels ranging mostly between 200 ppm and 1,000 ppm. Greenburg et al. (1926) made a survey of 12 plants in the United States (USA) that used benzene and observed that 32% of the workers had abnormally low white blood cell counts (WBC), e.g., below 5,500 WBCs per cc. Twelve percent had WBCs below 4,000 per cc. The benzene exposures levels associated with this extremely high prevalence of abnormal WBC counts were 90 ppm and higher. Greenburg et al. (1926) recommended medical removal if an employee developed clinical symptoms of poisoning, or if blood values for individual workers dropped 25% or more. 1.1.2. First report of benzene induced leukemia In 1928, Dolore and Borgomano published the first case of benzene induced leukemia. This case of acute lymphatic leukemia (ALL) was identified in a pharmaceutical worker, who had a job considered dangerous because of high benzene exposure levels. Another worker at the same plant died of AA and the authors were of the opinion that some previous cases of AA reported in association with benzene exposure may also have been leukemia. The company's means of dealing with the high exposure levels that caused these blood diseases was to rotate the workers out of the specificjob every month. 1.2. ACTlONSllNACTlONS RELATED TO KNOWLEDGE OF BENZENE TOXICITY 1.2.1. Steps toward control using exposure recommendations By 1939, the vast number of benzene poisonings among workers in all parts of the world led to the recommendation for the substitution of benzene with other solvents by a number of investigators (Greenberg 1926; Erf and Rhodes 1939;Mallory et al. 1939). In 1939, Hunter and Mallory et al. reported 89 cases of "poisoning" and three cases of leukemia among workers exposed to benzene in a variety of occupations. Two of the "poisonings" were associated with benzene levels of less than 25 ppm and 10 ppm, respectively. Yet, in 1946, the American Conference of Governmental Industrial Hygienists (ACGIH) recommended a limit of 100 ppm for benzene exposure in the workplace (ACGIH 1946). Subsequently,the recommended value was reduced to 50 ppm in 1947 and to 35 ppm in 1948 (ACGIH 1948). Because of evidence of "hypersusceptibility" to the bone marrow suppressing effects of benzene, a document published in 1948 by the American Petroleum Institute (API) concluded that the only absolutely safe level from exposure to benzene was zero ( API 1948). 2 This opinion was based on observations of some workers developing various blood diseases indicative bone marrow depression who had worked alongside other workers whose blood counts were in the normal range. The API then proceeded to recommend a limit of "50 ppm or less." In 1957, the ACGIH lowered its recommended 8-hour time-weighted average exposure limit to 25 ppm for benzene (ACGIH 1957). 1.2.2. Disregardfor exposure recommendations In spite of the recommendations mentioned above, cases of AA, and central nervous system (CNS) toxicity manifested by headache, nausea, giddiness, staggering gate, paralysis and unconsciousness (leading to death in 13 cases in the UK) continued to be reported in the 1940s and 50s (Browning 1965). These CNS symptoms are thought to be associated with benzene exposures ranging from 3,000 ppm to 20,000 ppm (Flury 1928 )--levels 200 to 800 times higher than the limits recommended in the 1940s and 1950s and 2000 times higher than the 10 ppm level already associated with AA by Hunter and Mallory et al. (1939). In the 1950s and 1960s, an obvious lack of precaution for workers exposed to benzene was taking place in many parts of the world, including Italy, Turkey, Britain, France, USA, Russia and other countries as documented by the publication of case reports of blood diseases as a result of benzene exposures well above levels known to be toxic to the bone marrow and to the central nervous system of workers. For example, Vigliani and co-workers, in 1964, reported that the risk of acute leukemia among workers "heavily exposed to benzene" in the rotogravure and shoe industries in the Italian Provinces of Milano and Pavia was at least 20 times higher than for the general population. Vigliani (1976) reported that over 200 cases of benzene hemopathy, including 34 cases of acute leukemia had been treated at the Institutes of Milano and Pavia between 1942-1965. These workers were exposed to benzene levels ranging mostly from 200500 ppm, with occasional peaks above these levels. In 1967, Goguel reported 44 cases of benzene-induced leukemia, mostly chronic forms, from the Paris region of France that occurred between 1950-1965. Blood poisonings among workers exposed to benzene were being documented in other parts of Europe as well. As stated by Aksoy (1977), because "benzene containing glue adhesives were extremely practical and cheaper in the market, they replaced their customary petroleum containing adhesives with the new product" in the shoe and slipper industry in Turkey in 1961. Benzene exposures experienced by these workers were reported to have been between 150-650 ppm (Aksoy 1978). By the mid-1970s, an epidemic of AA and leukemia among Turkish shoe workers began to unfold as reported by Aksoy (1971; 1977; 1978). The majority of the individuals who were identified in the reports from these European countries as dying from leukemia as well as other benzene related blood diseases were exposed to levels shown decades earlier to cause benzene poisoning. For a review of case reports of benzene related blood diseases associated with the countries mentioned above, see IARC (1974). 3 1.2.3. Epidemiological evidencefor leukemia Beginning in the early 1970s, the University of North Carolina in the USA published a series of epidemiologic studies demonstrating significant excesses of leukemia, mostly chronic forms, among workers exposed to presumably low atmospheric levels of benzene (McMichael et al. 1975). The benzene exposure resulted primarily from use of rubber solvents such as petroleum naphtha, toluene, mineral spirits, etc. that were contaminated with benzene ranging in volume from about 1%-5% in the 1940s to about 0.5% for petroleum naphtha in the 1970s. In 1977, Infante et al. published the results of the first cohort study of workers exposed specifically to benzene. The workers were engaged in the manufacture of a rubberized food wrap called Pliofilm. The study demonstrated a 5- to 10-fold risk of leukemia among workers exposed to technical grade benzene at levels that were generally considered within the various limits recommended over the time period 1940-1971,e.g., a 10 ppm TWA to a maximum limit of 100 ppm (Infante et al. 1977a). Up until this time, benzene was considered a cause of leukemia based not upon epidemiological studies, but rather upon case reports of leukemia and the clinical observation that individuals with benzene induced aplastic anemia and other blood diseases transformed into acute leukemia. 1.2.4. Attempt to control occupational exposure through regulation in the USA In 1977,on the basis of the Infante et al. (1977a) study results supplemented by the world literature on benzene and leukemia, the US Department of Labor (OSHA 1977a) issued an Emergency Temporary Standard (ETS) that would have reduced the occupational benzene exposure limit in workplace air to 1 ppm as an 8-hour time-weighted average (TWA). The 1 ppm exposure limit was based on OSHA's policy at the time that exposure to carcinogens in the workplace should be lowered to the lowest feasible limit. mote: feasibility analyses take into consideration both technological and economic feasibility.] The new OSHA ETS was stayed in 1977, however, in response to a challenge in the US Court of Appeals by the American Petroleum Institute, who in essence argued that there was no increased risk of developing leukemia as a result of benzene exposures below the old limit of 10 ppm. Then, OSHA proposed a permanent standard, requested comments and held a public hearing (OSHA 1977b). In 1978, OSHA issued a Final Standard (OSHA 1978) that included a 1 PPM atmospheric exposure limit. This standard also was challenged by the API on the same grounds. The US Court of Appeals again vacated the Final Standard, and that decision was appealed to the Supreme Court. [Note: Benzene was voluntarily withdrawn from consumer products in the USA after it was shown that use of a paint stripper in the home could generate atmospheric levels up to 200 ppm in a short period of time (Young et al. 1978).] 4 1.2.5. The USA Benzene Decision and dose response analyses In July 1980, the U.S. Supreme Court (IUD, AFL-CIO vs API, 1980) issued what has become known as the Benzene Decision. This Decision has had a major impact on OSHA's ability to control exposures to benzene and other toxic substances in the workplace. The Court stated that before OSHA can promulgate any permanent health standard, the Secretary of Labor is required to make a threshold finding that a place of employment is unsafe in the sense that significant risks are present and can be eliminated, or lessened by a change in practices. Although the Benzene Decision recognized the uncertainties involved, it indicated that the determination of "significant risk" should, if at all possible, be established on the basis of an analysis of the best available evidence through such means as quantitative risk assessments. The Supreme Court in its general guidance for fhture OSHA rulemaking noted that the requirement that a significant risk be identified is not a mathematical straitjacket and that it is the Agency's responsibility to determine what it considers to be a significant risk based largely on policy considerations. In the only concrete example of significant risk, the Court stated that if, for example, the odds are one in a million that a person will die from cancer by taking a drink of chlorinated water, the risk clearly could not be considered significant. On the other hand, if the odds are one in a thousand that regular inhalation of gasoline vapors that are 2% benzene will be fatal, a reasonable person might well consider the risk significant and take appropriate steps to decrease, or eliminate it. [Note: Since the Benzene Decision, OSHA has considered an occupational lifetime risk (over a 45 year period) of one extra case of cancer, or other material impairment of health consequence, per 1,000 workers to be "significant." It has not yet considered the other end of the range, e.g., what it considers to be a non-significant risk, because all of the health standards that OSHA has promulgated since the Benzene Decision, with the exception of perhaps formaldehyde, have resulted in estimates of excess risk that are greater than one per 1,000 over an occupational lifetime.] For quantitative estimates of risk for health standards promulgated by OSHA since the Benzene Decision, see Infante (1995b). A straightjacket, however, appropriately describes the risk analyses that OSHA currently engages in prior to proposing any regulatory action. The long delay recognized in OSHA promulgating standards before 1980 has now been superseded by additional detailed analyses of risk related to exposure. While on the surface such analyses may seem appropriate, they have become encumbered by additional analyses that take into account the mechanism by which the substances being regulated may cause cancer. Since the exact mechanism by which any substance causes cancer (including benzene which has been studied for decades) has not been identified, speculation and argument about various unproven hypotheses for cancer causation are time consuming. Many other issues also have been added to the debate on risk assessment procedures such as whether the mouse, the hamster, or the rat is the most appropriate species to use when human data are not available. Since the Agency is required to review and comment on all possible cancer mechanisms, appropriateness of species, etc., the entire "risk assessment'' process has created additional years of delay in standard setting. Instead of reasonable precautions being promulgated by government and employers, years pass as analyses are performed to determine the dose response between exposure and risk of disease. These analyses can also include incorporation of speculative mechanistic data that have not been validated. 1.2.6. Cost in lives of the prolonged regulatoryprocess in the USA Eleven years after the OSHA promulgation of an ETS for benzene, a new benzene standard that included a 1 ppm 8- hour TWA exposure limit was finally issued (OSHA 1987). The new limit was based on "economic feasibility," not elimination of "significant risk" as an occupational lifetime risk of 10 extra leukemia deaths per 1,000workers was associated with the 1 ppm limit. Other estimates of leukemia risk only and limited to the NIOSH cohort data (as updated by Paxton et al. 1994)and to deaths from acute myelogenous and monocytic leukemias only, indicate a range of 0.02 to 5.1 per 1,000excess deaths depending upon the estimates of benzene exposure and the model chosen (Crump 1994). [Note: these analyses are based on a follow-up period of the NIOSH cohort that results in selection bias for the reasons stated below.] Based on OSHA's final quantitative risk assessment for benzene and estimates of extra benzene exposure to the USA workforce during the 11 years that it took to complete the benzene standard, it has been estimated that an extra 198 deaths from leukemia and 77 extra deaths from multiple myeloma will eventually develop among US workers as a result of the 11 year delay--deaths that could have been prevented (Infante and DiStasio, 1988). This estimate of preventable deaths from benzene exposure did not include blood diseases other than leukemia, which were known at the time to be caused by benzene exposure--the reason being that there were no dose response data available for the other blood diseases. 1.2.7. Expansior: of lymphohematopoietic diseases related to benzene exposare The quantitative estimate of extra deaths from benzene exposure indicated above did not include non-Hodgkin's lymphoma, which has been shown more recently to be associated with occupational exposure to very low levels of benzene (Hayes et al. 1997). Quantitative risk assessment (Infante 1997) based on the Hayes et al. (1996) study results suggests an extra risk of 54 deaths from leukernidlymphoma per 1,000 workers exposed over an occupational lifetime (45 years). This risk is 54 times greater than a level considered significant by OSHA. The Hayes et al. (1997) study results and dose response analyses based upon those data clearly demonstrate the inadequacy of the 1 ppm exposure limit for benzene based on cancer risk alone. Fortunately, most occupational settings in the USA are able to achieve benzene exposure levels in the range of 0.2- 0.3 ppm, or lower. Additionally, the USA standard includes provisions ancillary to the PEL such as exposure monitoring, medical surveillance and hazard training--all of which should presumably hrther reduce the risk of benzene exposure and related diseases. In addition, to quantitative risk assessment, direct observation of data from the Hayes et al. (1997) study 6 demonstrates significantly elevated relative risks for all lymphohematopoietic cancers combined, and for acute non-lymphocytic leukemia and myelodysplastic syndrome combined (ANLLMDS) among the group of workers exposed to a constant average benzene level of only 1.2 ppm for 5.5 years for a total cumulative dose of 6.7 ppm-years (Hayes, personal communication, 1999), e.g., a much lower cumulative dose than that allowed by a limit of 1 ppm over an occupational lifetime of 45 years (45 ppm-years of cumulative dose). Some individuals within the benzene cohorts who have died from leukemia or lymphoma experienced estimated benzene exposures of 0.5-2 ppm for only 1-2 years, or less (Infante, 1992). By the 1 9 9 0 t~ox~icologic research demonstrated the multiple-site carcinogenicity of benzene in experimental animals and additional epidemiological studies and case reports of exposed workers expanded the carcinogenicity of benzene to all major forms of leukemia in the aggregate (Savitz and Andrews 1997; Infante 1995a; Wong 1987b) and specifically acute myelogenous leukemia (AML) and its variants (Hayes et al. 1997; Browning 1965; Rinsky et al. 1987;Bond et al. 1986; Decoufle et al. 1983), ALL (Hernberg et al. 1966; Shu et al. 1988), chronic lymphatic leukemia (McMichael et al. 1975;), chronic myelogenous leukemia (Browning 1965; Goguel et al. 1967; Tareeff et al. 1963; Infante 1995a;Wong 1987b;) and some minor forms such as hairy cell leukemia (Aksoy 1987; Flandrin and Collado 1987), myelodysplastic syndrome (Hayes et al. 1997),myeloproliferative disorders (Rawson et al. 1941;Tondel et al. 1995)as well as nonHodgkin's lymphoma (Hayes et al. 1997), including multiple myeloma (DeCoufle et al. 1983; Rinsky et al. 1987; Ireland et al. 1997; Goldstein 1990). [Note: An "updated" analysis of the NIOSH benzene cohort (Infante et al. 1977; Rinsky et al. 1987) by Wong (1995) concluded that the excess of multiple myeloma in the NIOSH cohort was no longer statistically significant. Wong (1995), however, changed the beginning date of follow-up from 1950 to 1940. In doing so, he introduced selection bias into his analysis because the company removed records for individuals who died at one of the study locations for several of the years prior to 1950. Thus, anaIyses of the NIOSH cohort by Wong (1995) that begin follow-up prior to 1950 cannot be relied upon for estimating risk of death from multiple myeloma, or any other cause of death.] 1.3. DISCUSSION The response to information on the toxicity of benzene has at times demonstrated concern by some in the field of occupational health, particularly in the earlier years when large numbers of workers in various sectors and jobs were being surveyed to determine the extent of their blood diseases. Nevertheless, even during this period, benzene exposure levels were not reduced to levels commensurate with the toxicity data available at the time, and the epidemic of poisonings and leukemia among benzene exposed workers continued through the first six decades of the 20th century. Various reasons for the limited public health response and subsequent overexposure and disease in light of the knowledge of benzene toxicity are apparent. 7 1.3.1. Some reasonsfor lack of precaution in benzene exposure to workers Lack of knowledge: The lack of precaution about exposure to benzene has been attributed in part to lack of knowledge of its toxicity during the first four decades of the 20th Century. For example, even though Santessen (1897) reported four cases of AA among women manufacturing raincoats in Sweden in 1897, exposures were not lowered enough and Helmer (1944) reported 60 cases of benzene poisoning (58 were women) at a single raincoat manufacturing facility in the same country in 1940 and 1941. The epidemic of benzene poisoning in Sweden was ascribed in part to a lack of knowledge of the toxicity of benzene by plant management and workers (Helmer 1944). Cost of solvents: Several investigators, who surveyed workforces and identified workers with various benzeneinduced blood diseases, namely Greenberg and colleagues in 1926, Erf and Rhodes in 1939 and Mallory et al. in 1939 recommended the substitution of benzene with other solvents. Yet, worldwide consumption of benzene in the marketplace continued to expand after World War 11. One of the reasons for expanded use of benzene in the synthetic rubber industry is that it was such a good rubber solvent. Another reason, as expressed by Dr. Aksoy (1977), is that benzene was cheaper than other solvents used in the shoe and slipper industry in Turkey. Thus, economic consequences led to other solvents being replaced with benzene in the Turkish shoe manufacturing industry as late as 1961. This substitution to the use of benzene led to high level workplace atmospheric benzene concentrations and to the epidemic of leukemia, preleukemia, pancytopenia and other blood diseases as reported by Aksoy (1971;1977; 1978). Thus, economic considerations in the 1960s (Aksoy 1977) fixther contributed to the overexposure and subsequent benzene related diseases. Consensus recommendations and potential influence by corporate representatives: The 1939 reports by Hunter and by Mallory et al. indicated that some cases of benzene poisoning were associated with levels of 25 ppm and 10 ppm. Yet, in 1946,the ACGIH recommended a limit of 100 ppm. Although the ACGIH recommendation for benzene was lowered to 35 ppm in 1948, this level was still higher than the levels reported in association with benzene poisonings. [Note: From my personal experience over the years, it is apparent that one of the problems being faced in the occupational health community for benzene (as well as for other toxic substances) is that consensus organizations usually base their recommend exposure levels on what is easily achievable in the workplace. Data related to level of exposure and toxicity are reviewed, but are not translated into health based exposure limit recommendations.] Castleman and Ziem (1988) investigated this behavior by the ACGIH and concluded that the ACGIH threshold limit values (TLVs) were based heavily on corporate influence. Thus, consensus recommendations were inadequate and corporate influence may have played a role in these recommendations and in the 8 resultant proliferation of benzene induced diseases in the workplace. Castleman and Ziem (1988) were of the opinion that an international effort was needed "to develop scientifically based guidelines to replace the TLVs in a climate of openness and without manipulation by invested interests.'' To date, this goal has not been achieved (Castleman and Ziem 1994). Anti-public health attitude (call for scientific certainty): In the 1970s, benzene manufacturers and users began a new approach for conveying knowledge about the toxicity of benzene to the public in general, and to workers and plant managers specifically, which contributed to a continuation of overexposure to benzene. This is the period when manufacturers began to hire consultants to downplay the importance of the scientific observations related to the toxicity of benzene and to introduce unresolvable arguments about dose response analyses, which has had an impact in delaying much needed government regulations that sought to reduce benzene exposure in the workplace. Economic considerations were again being given a higher priority than concerns for public health, but this time the economic concerns appeared to be based on the cost of lowering exposures (OSHA 1987), and perhaps on the increasing cost of litigation and liability related to workers contracting benzene related diseases on the job. This era has fostered the development of arguments that seek to minimize or misrepresent study results. In my opinion, it is part of a new anti-public health approach that calls for scientific certainty in terms of causality for every specific lymphohematopoieticdisease related to benzene by exposure level. As a result, workers employed world-wide in benzene exposure operations today may not be afforded the appropriate protection to the extent feasible in order to reduce their risk of hematopoietic diseases and many of those who develop benzene related diseases may receive little or no compensation. For example, during the OSHA rulemaking hearings held in 1977, it was argued that the benzene cohort study conducted by the National Institute for Occupational Safety and Health (NIOSH), which demonstrated a 540 10-fold elevated risk of leukemia (Infante et al. 1977a) was meaningless in terms of public health intervention to reduce exposures in the workplace. Although not persuasive, one of the arguments raised by consultants to the industry was that the study had simply identified a random leukemia cluster, and since clusters of leukemia are known to occur in time and space, this cluster of leukemiajust happened to be identified among a cohort of benzene exposed workers (Tabershaw and Lamm, 1977). See Tabershaw and Lamm (1977) and Infante et al. (1977b) for a series of arguments and rebuttal about the study findings. It was further argued that benzene could not cause leukemia in workers because there was no evidence that benzene caused cancer in experimental animals (Olson, 1977). This argument was clearly fictitious given the overwhelming evidence of carcinogenicity provided by the study of humans. In any case, shortly thereafter the evidence of carcinogenicity of benzene in experimental animals became available (Maltoni and Scarnato 1979;NTP 1986). In the 1980s, when OSHA again published a new benzene proposal taking into consideration the guidance offered by the Supreme Court's Benzene Decision on determination of the significance of health risk, attention during the rulemaking focused on the dose response analyses prepared by 9 OSHA and its consultants in support of its new standard (OSHA 1987). Most of the argument addressed the benzene exposure assessment portion of the risk assessment, with new estimates of exposure to the benzene cohort members being provided for time periods in which there were no exposure data available. These "educated" guesses by the various parties involved in the rulemaking could not be confirmed. Proposed government regulation of benzene, however, resulted in a number of new dose response analyses and a protracted debate about which exposure assessment was the most appropriate for use in the quantitative risk assessment--the resolution of which can never be determined with scientific certainty. Argument about the type of dose response model that was most appropriate for benzene exposure and risk of leukemia also was raised. While nobody would object to debate on these issues, the continuation of the arguments for protracted periods of time resulted in workers and the public in general being unnecessarily exposed to benzene levels that could have otherwise been reduced through a shorter regulatory process. Studying a subject to death, often results in the death of those we are trying to protect (Infante, 1987). In the 1990s, the USA National Cancer Institute (NCI) in collaboration with the Chinese Academy of Preventive Medicine (CAPM) published a series of ongoing studies of Chinese workers exposed to benzene. These studies (Hayes et al. 1997; Hayes et al. 1996; Dosemeci et al. 1996)demonstrate a dose response for exposure to benzene and leukemia, lymphoma, myelodysplastic syndrome and aplastic anemia. They also demonstrate through direct observation, high relative risks for leukemia, myelodysplastic syndrome and non-Hodgkin's lymphoma as a result of very low average benzene exposures, e.g., around 1 ppm. Since the results of these well conducted studies may be used in the future by governments in Europe, the USA and other countries for estimating benzene related diseases from low level exposure to the general population, the findings have broad implications for public health intervention. In response to the publication of these studies, consultants to the chemical industry have published critiques of both the health findings and the benzene exposure estimates related to those findings (Wong 1998; Wong 1999; Budinsky et al. 1999), which in the opicion of some misrepresent the data on health effects as well as the benzene exposure estimates made by the NCI/CAPM investigators (Hayes, et al. 1998; Hayes personal communication). Some of these same consultants have also expressed surprising views about the more general findings related to benzene exposure and disease (Wong 1995; Wong 1996; Bergsagel et al. 1999). For example, based on his analysis of data from the NIOSH benzene cohort study, one author concluded that benzene can only cause acute mylogenous leukemia (AML), in contrast to other types of leukemia, and that the threshold for benzene to cause AML is between 370 and 530 ppm-years of exposure (Wong 1995). In this publication, he failed to include data from his own benzene study whereby he reported a statistically significant dose response for benzene exposure and leukemia among workers whose cumulative benzene exposures ranged from less than 15 ppm-years to more than 60 ppm-years (Wong 1987a; Wong 1987b).Furthermore, in the latter study, none of the leukemia deaths were from AML. Thus, the findings and conclusions he drew from his own study contradict his opinion that benzene causes only AML and that the cumulative exposure threshold for benzene to cause leukemia is between 370 and 530 ppm-years. 10 Wong (1995) further concluded that there is no evidence that benzene exposure was associated with multiple myeloma in the NIOSH study because he could not identifL a dose response based on four cases of multiple myeloma in the study population. Lack of ability to observe a dose response relationship based on four deaths from multiple myeloma is essentially meaningless because four is too few cases to allow for enough statistical power to observe a dose response if in fact one were present. [Note: Rinsky et al. (1987) of NIOSH had previously demonstrated a significant excess of death from multiple myeloma among the benzene cohort members and concluded that low level benzene exposure was related to multiple myeloma.] Wong (1995) also has argued that benzene in general is not associated with multiple myeloma, and contrary to the findings in the NCVCAPM study, he has argued that there is no evidence that benzene is associated with non-Hodgkin's lymphoma (Wong 1998)--conclusions that appear to be at odds with the views of others on benzene toxicity (Goldstein and Shalat 2000; Goldstein 1990; Rinsky et al. 1987; Decoufle et al. 1983; Infante 1995a; Savitz and Andrews 1996; Savitz and Andrews 1997;Hayes et a1 1998; Hayes et al. 2000). The arguments about the NCIKAPM study seem reminiscent of the protracted debate and delay in necessary government action that took place after release of the NIOSH benzene study in 1977 (Infante et al. 1977a). It will be unfortunate if more precaution is taken with the use of data from the NCIKAPM study than with the protection of populations exposed to levels of benzene that can be reduced with technology currently available. Scientific certainty is difficult to achieve, but stressing the uncertainty does not do justice to the data on benzene exposure and related diseases. In the case of benzene historically, taking precaution to maintain exposure levels in the workplace in accordance with the scientific data available at the time would have eliminated much needless suffering and death among the workers. In my opinion, the protracted argument about the dose response for benzene related diseases among exposed workers and denial about the lymphohematopoieticdiseases most likely associated with benzene exposure is counterproductiveto efforts to provide a workplace relatively free of harm. While the continuation of this debate may be interesting from an academic viewpoint, it also raises the question of whether it may be more of a reflection of economic concerns and potential liability on the part of companies rather than a concern for the public's health based on a reasonable interpretation of available scientific data. 1.3.2. Benzene in gasoline as a continuing hazard Most consumers and many medical personnel are not aware of the fact that gasoline contains benzene. In the USA, gasoline has contained an average of about 1.5% benzene for the past two decades, but excursion concentrations may reach 5% by volume (Infante et al. 1990). Historically, petrol in most European countries has contained more benzene than USA varities and the trend apparently still existed through 1994 (Deschamps 1999, but the benzene content supposedly has been reduced more recently. Not surprisingly, epidemiologic studies, analyses 11 and case reports have demonstrated an association between gasoline exposure and leukemia (Schwartz 1987; Jakobsson et al. 1993; Infante et al. 1990),other blood diseases (Infante et al.1990; Lumley et al. 1990; Naizi and Fleming 1989), chromosomal defects (Lumley et al. 1990; Hogstedt et al. 1991) and other manifestations of genetic damage (Nilsson et al. 1996). Yet, gasoline station pumps do not provide adequate information on the cancers known to be associated with benzene exposure. Nor do the material safety data sheets (MSDSs) for gasoline provide the available evidence on chromosomal, or genetic damage. Because of this lack of candor about the hazards of benzene in gasoline, garage mechanics and highway maintenance workers take unnecessary risks by using gasoline as a solvent in cleaning auto parts (Infante 1993) and consumers take unnecessary risks by using gasoline as a solvent and don't take the necessary precautions when using gasoline in various home appliances such as lawn mowers, weed trimming devices, power saws, etc. In addition, a study of roadside vendors who sold repackaged gasoline in Nigeria has demonstrated that 26% had neutropenia as compared to 2-10% in controls--a significant difference (Naizi and Fleming 1989). The hazards of benzene in gasoline have been recognized since at least 1928, when Askey (1928) reported a case of aplastic anemia in a U.S. worker exposed to gasoline, and more recently by the report of a case of myelofibrosis in a Swedish petrol station attendant (Tondel et al. 1995) and a case of aplastic anemia in a U.S. roofer who used petrol to clean seams before fitting rubberized roofing material (Infante et al. 1990). Despite the overwhelming literature on the hazards of benzene in petrol, the public health community and safety officials, including those employed by industry have yet to come to grips with task of adequately informing workers and consumers about this hazard. 1.4. CONCLUSIONS AND LESSONS FOR THE FUTURE The available knowledge on the toxicity of benzene and the failure to take precautions to protect workers (and the public in general) in light of this knowledge over the past century is cause for concern. The inaction, or inadequate actions by consensus organizations and governments alike throw into question the ability of these organizations to protect the health of the public. In the case of benzene exposure in the workplace, the Precautionary Principle is not relevant. Recommendations made in the USA and in the UK in the 1920s for the substitution of benzene with other solvents known to be less toxic to the bone marrow went unheeded for decades even though high percentages of workers being surveyed demonstrated blood disorders. Furthermore, benzene was not withdrawn from consumer products in the USA until 1978 and this was done by manufacturers on a voluntary basis, and it has never been adequately validated. It is also difficult to accept the claimed ignorance of the toxicity of benzene in the raincoat manufacturing industry in Sweden in the 1940swhen 60 cases of blood poisoning in a single factory were reported. Forty years earlier (in 1897)published case reports of aplastic anemia among women employed in this same industry in Sweden appeared in the literature. The claim that management was unaware of the hazards of benzene exposure in such a small industry in a country known for its humanitarian concerns is incomprehensible. In the 1940s and 1950s, 13 12 deaths from the neurotoxic effects of benzene were reported to have occurred in the UK. The benzene exposure levels associated with these acute deaths were most likely more than 200 to 800 times the occupational exposure levels recommended at the time. Clearly, this situation could, and should, have been avoided had there been any serious concern for worker health. With the knowledge available at the time, it is also difficult to understand how benzene was substituted for other petroleum solvents in the shoe industry in Turkey in the 1961. Aksoy states that the substitution was made because ber- , ?ne was cheaper than the other solvents. Data on the costs of benzene and the other solvents in `iurkey in the 1960s are not available, but it is unlikely that the difference could have been more than a few cents a gallon. Yet, the epidemic of leukemia and other fatal blood diseases that followed this substitution had to have been very costly in terms of the worker's diseases, the associated expenses for health care and the loss of wages, etc. This is simply a case, as in other instances, of the cost of production being more important to the manufacturers than the cost of human life. Even though numerous case reports (in the thousands) of benzene related blood diseases, including leukemia, were reported in the literature, precautionary measures to reduce exposure levels below those known, or reasonably anticipated, to cause blood diseases were not taken, and recommended exposure limits by consensus organizations like the ACGIH were based on those that were easily achievable in the workplace. According to Castleman and Zeim, such recommendations stemmed from the participation of scientists employed by various corporations on the TLV Committee that made the exposure recommendations. Thus, one of the lessons to be learned, if not already obvious, is that consensus organizations in the process of developing exposure limits for chemicals should maintain distance from the producers of the chemicals and their "consultants" when evaluating evidence for the diseases of concern. Finally, affixing a warning label on gasoline pumps that includes the cancers and other diseases known, or likely to be caused by benzene exposure, may serve to reduce unnecessary benzene exposure to garage mechanics, petrol service station attendants, highway maintenance workers and consumers, who fill their own gas tanks, but who more unknowingly use the gasoline in consumer products at home, and not infrequently use gasoline as a solvent at home without full knowledge of its cancer and non-cancer disease risks. To not adequately inform the public of the cancers, bone marrow proliferative diseases and genetic hazards associated with benzene in gasoline, is to repeat our failures of the 20th Century into the 21st Century and to make a mockery of public health education. References American Conference of Governmental Industrial Hygienists (1946) Transactions of the Eighth Annual Meeting. Cincinnati, Ohio ACGIH, p. 40. 13 American Conference of Governmental Industrial Hygienists (1948) Transactions of the Tenth Annual Meeting. Cincinnati, Ohio ACGIH, p. 31. American Conference of Governmental Industrial Hygienists (1957) Transactions ofthe Nineteenth Annual Meeting. Cincinnati, Ohio ACGIH, p. 47. American Petroleum Institute Toxicological Review: Benzene. (1948) Department of Safety, American Petroleum Institute, New York, September. Aksoy M. (1987) Chronic lymphoid leukaemia and hairy cell leukaemia due to chronic exposure to benzene: report of three cases. Br J Haematol66,209-211. Aksoy M. (1977) Leukemia in workers due to occupational exposure to benzene. New Istanbul Contrib Clin Sci 12,3-14. Aksoy M. (1978) Benzene and leukaemia. Lancet i, 441 Aksoy M., Dincol K., Akgun T., Erdem S. and Dincol G. (1971) Haematological effects of chronic benzene poisoning in 217 workers. Br J Industrl Med 28,296-302. Askey J.M. (1928) Aplastic anemia due to benzol poisoning. Calif West Med 29,262-263. Bergsagle D.E., Wong O., Bergsagel P.L., Alexanian R., Anderson K., Kyle R.A. and Raabe G.K. (1999) Benzene and multiple myeloma: Appraisal of the scientific evidence. Blood 94, 1174-1182. Bond G.G., McLaren E.A., Baldwin C.L. and Cook R.R. (1986) An update of mortality among workers exposed to benzene. Br J Indust Med 43,685-691. Browning E. (1965) Toxicity and Metabolism of Industrial Solvents. Elsevier Publishing Co., New York, pp 3-65. Budinsky R.A., DeMott R.P., Wernke M.J. and Schell J.D. (1999) An evaluation ofmodeled benzene exposure and dose estimates published in the Chinese-National Cancer Institute collaborative epidemiology studies. Reg Toxicol Pharm 30,244-258. Castleman B.I. and Ziem G.E. (1988) Corporate influence on Threshold Limit Values. Am J Ind Med 13,531-559. Castleman B.I. and Ziem G.E. (1994) American Conference of Governmental Industrial Hygienists: Low threshold of credibility. Am J Ind Med 26, 133-143. 14 Crump K.S. (1994) Risk of benzene-induced leukemia: A sensitivity analysis of the Pliofilm cohort with additional follow-up and new exposure estimates. J Tox Environ Health 42,219-242. DeCoufle P., Blattner W.A., and Blair A. (1983) Mortality among chemical workers exposed to benzene and other agents. Environ Res 30, 16-25. Delore P. And Borgomano C. (1928) Acute leukemia in the course of benzene poisoning: the toxic origin of certain acute leukemias and their relationship to severe anemia. J Med Lyon 9, 227-233. Deschamps F. (1995) Unleaded gasoline: a continuing hazard. Occup Med 45,55. Dosemeci M., Yin S-N., Linet M., et al. (1996) Indirect validation of benzene exposure assessment by association with benzene poisoning. Environ Health Perspect 104 (Suppl6), 1343-1347. Erf L.A. and Rhoads C.P. (1939) The hematological effects of benzene (benzol) poisoning. J Industrl Hyg Toxicol2 1,421-435. Flandrin G. And Collado S. (1987) Is male predominance (4/1) in hairy cell leukaemia related to occupational exposure to ionizing radiation, benzene and other solvents? Br J Haematol 67, 119120. Flury F. (1928) 11. Modem occupational intoxications. In. Modern occupational intoxications from the aspect of pharmacology and toxicology. Arch Exp Path Pharmakol 138,65-82 (Ger) Cited in: Occupational Exposure to Benzene, Criteria for a Recommended Standard. (1974) US Dept Health Educ Welfare, PHS, CDC, NIOSH, Pub No. (NIOSH) 74-137. Goguel A., Cavigneaux A. And Bernard J. (1967) Benzene leukemia in the Paris region between 1950 and 1965 (A study of 50 cases). Nouv Rev Fr Hemato 7,465-480. Goldstein B.D. (1990) Is exposure to benzene a cause of human multiple myeloma? Ann NY Acad Sci 609,225-230. Goldstein B.D. and Shalat S.L. (2000) The causal relation between benzene exposure and multiple myeloma, Letter-to-the-Editor. Blood 95, 1512-1513. Greenburg L. (1926) Benzol poisoning as an industrial hazard. Pub Health Rep 41, 1516-1539. Hayes R.B., et al. (1997) Benzene and the dose-related incidence of hematologic neoplasms in China. J Natl Cancer Inst 89, 1065-1071. 15 Hayes R.B. et al. (1996) Mortality among benzene-exposed workers in China. Env Health Perspect 104 (Suppl6), 1349-1352. Hayes R.B. et al. (1997) Benzene and the dose-related incidence of hematologic neoplasms in China. J Natl Cancer Inst 89, 1065-1071. Hayes R.B. et al. (1998) Benzene and the dose-related incidence of hematologic neoplasms in China--Response, Letter-to-the-Editor. J Natl Cancer Inst 90,470-471. Hayes R.B., et al. (2000) In press Helmer K.J. (1944) Accumulated cases of chronic benzene poisoning in the rubber industry. Acta Medica Scand 118,254-275. Hernberg S., Savilahti M., Ahlman K. And Asp S. (1966) Prognostic aspects of benzene poisoning. Br J Indust Med 23,204-209. Hogan J.F. and Shrader J.H. (1923) Benzol poisoning. Am J Pub Health 13,279-282. Hunter F.T. (1939) Chronic exposure to benzene (benzol)--11. The clinical effects. J Industrl Hyg Toxicol21, 331-354. Hogstedt B., Holmen A., Karlsson A., Raihle G., Nillius K.and Vestlund K. (1991) Gasoline pump mechanics had increased frequencies and sizes of micronuclei in lymphocytes stimulated by pokeweed mitogen. Mutat Res 263, 51-55. Hunter F.T. (1939) Chronic exposure to benzene (benzol). 11. The clinical effects. J Industrl Hyg Toxic0121, 331-354. Industrial Union Department, AFL-CIO v. American Petroleum Institute. (1980) 448 U.S. 601. Infante P.F. (1992) Benzene and leukemia: The 0.1 ppm ACGIH proposed TLV for benzene. Appl Occup Environ Hyg 7,253-262. Infante P.F. (1997) Quantitative risk from leukemidlymphoma from occupational exposure to benzene. The Toxicology Forum 22nd Annual Winter Meeting, pp. 485-488, c Toxicology Forum, Inc, Washington, D.C. Infante, P.F. (1995a) Benzene and leukemia: Cell types, latency and amount of exposure associated with leukemia. (In, Update on Benzene, Advances in Occupational Medicine and Rehabilitation. Imbriani, M., Ghittori, S., Pezzagno, G., Capodaglio, E., Eds. Fondazione Salvatore Maugeri Edizioni, Pavia, Italy) pp. 107-120. Infante P.F. (1995b) Cancer in blue-collar workers: Who cares? New Solutions 5, 52-57. 16 Infante P.F. (1987) Benzene toxicity: Studying a subject to death. Am J Industrl Med 11,599604. Infante P.F. (1993) State of the science on the carcinogenicity of gasoline with particular reference to cohort mortality study results. Environ Health Perspect 101 (Suppl6), 105-105. Infante P.F. and DiStasio M.V. (1988) Occupational benzene exposure: Preventable deaths. Lancet i, 1399-1400. Infante, P.F., Rinsky, R.A., Wagoner, J.K. and Young, R.J. (1977a) Leukaemia in benzene workers. Lancet ii, 76-78. Infante, P.F., Rinsky, R.A., Wagoner, J.K. and Young, R.J. (1977b) Benzene and leukaemia. Lancet ii, 867-868. Infante P.F., Schwartz E. and Cahill R. (1990) Benzene in petrol: a continuing hazard. Lancet 335,814-815. International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Man. (1974) Some anti-thyroid and related substances, nitrofurans and industrial chemicals. IARC, Lyon, France, 7,203-221. Ireland B., Collins J.J., Buckley C.F. and Riordan S.G. (1997) Cancer mortality among workers with benzene exposure. Epid 8 , 318-320. Jakobsson R., Ahlbom A., Bellander T and Lundberg I. (1993) Acute myeloid leukemia among petrol station attendants. Arch Environ Health 48,255-259. LeNoir and Claude (1897) On a case of purpura attributed to benzene intoxication. Bul Mem SOCMed Hop 3,1251-1261. Lumley M., Barker H. and Murray J.A. (1990) Benzene in petrol. Lancet 336, 1318-1319. Mallory T.B., Gall E.A. and Brickley W.J. (1939) Chronic exposure to benzene (benzol)--111. The pathologic results. J Industrl Hyg Toxicol21,355-377. Maltoni C and Scarnato C. (1979) First experimental demonstration of the carcinogenic effects of benzene. Long-term bioassays on Sprague-Dawley rats by oral administration. Med Lav 70, 352-35 7. McMichael A.J., Spirtas R., Kupper L.L. and Gamble J.F. (1975) Solvent exposure and leukemia among rubber workers: An epidemiologic study. J Occup Med 17,234-239. 17 National Toxicology Program (NTP) (1986) Toxicology and Carcinogenisis Studies of Benzene (CAS No. 71-43-2) in F344/N rats and B6C3Fi Mice (Gavage Studies) (NTP Technical Report 289;NIH Publ. No. 86-2545), Research Triangle Park, NC. Naizi G.A. and Fleming A.F. (1989) Blood dyscrasia in unofficial vendors of petrol and heavy oil and motor mechanics in Nigeria. Tropical Doctor 19,55-58. Nilsson R.I., Nordlinder R.G., Tagesson C., Walles S., Jarvholm B.G. (1996) Genotoxic effects in workers exposed to low levels of benzene from gasoline. Am J Ind Med 30, Occupational Safety and Health Administration. (1977a) Occupational Exposure to Benzene: Emergency Temporary Standards-Hearing,May 3, Part IV. Fed Register 42,225 16-22529. Occupational Safety and Health Administration. (1977b) Occupational Exposure to Benzene: Emergency Temporary Standards-Hearing,May 27, Part VI. Fed Register 42,27452-27478. Occupational Safety and Health Administration. (1978) Occupational Exposure to Benzene: Occupational Safety and Health Standards, February 10, Part 11. Fed Register 43, 5918-5970. Occupational Safety and Health Administration. (1987) Occupational Exposure to Benzene: Final Rule, September 11, Part 11. Fed Register 52,34460-34578. Olson, R.E. (1977) Testimony before the US Department of Labor, Occupational Safety and Health Administration, Washington, D.C., OSHA Docket No. H-059, Exhibit No. 149E. Paxton M.B., Chinchilli V.M., Brett S.M. and Rodricks J.V. (1994) Leukemia risk associated with benzene exposure in the Pliofilm cohort: Mortality update and exposure distribution. Risk Analysis 14, 147-154. Rawson R., Parker Jr. F. and Jackson Jr. H. (1941) Industrial solvents as possible etiologic agents in myeloid metaplasia. Science 93, 541-542. Rinsky R.A., Smith A.B., Hornung R., Filloon T.G., Young R.J., Okun A.H. and Landrigan P.J. (1987) Benzene and leukemia. An epidemiologic risk assessment. N Eng J Med 316, 1044-1050. Santesson C.G. (1897) Chronische vergifiungen mit steinkohlentheerbenzin:Vier Todesfalle. Arch Hyg Bakteriol31,336-376. Savitz D.A. and Andrews K.W. (1997) Review of epidemiologic evidence on benzene and lymphatic and hematopoietic cancers. Am J Industrl Med 31,287-95. 18 Savitz D.A. and Andrews K.W. (1996) Risk of myelogenous leukaemia and multiple myeloma in workers exposed to benzene. Letter-to-the-Editor,Occup Environ Med 53,357. Schwartz E. (1987) Proportionate mortality ratio analysis of automobile mechanics and gasoline service station workers in New Hampshire. Am J Ind Med 12,91-99. Shu X.O., Brinton L.A., Linet M.S., Tu J.T., Zheng W. And Fraumeni J.F. (1988) A populationbased case-control study of childhood leukemia in Shanghai. Cancer 62,635-643. Tabershaw I.R. and Lamm S.H. (1977) Benzene and Leukaemia, Letter-to-the-Editor.Lancet ii, 867-868. Tareeff E.M., KontchalovskayaN.M. and Zorina L.A. (1963) Benzene leukemias. Acta Unio Internat Contra Cancrum 19,751-755. Tondel M., Persson B. and Carstensen J. (1995) Myelofibrosis and benzene exposure. Occup Med 45,51-52. Vigliani E.C. (1976) Leukemia associated with benzene exposure. Ann NY Acad Sci 271, 143151. Vigliani E.C. and Saita G. (1964) Benzene and leukemia. N Engl J Med 271,872-876. Wong 0.(1998) Re: Benzene and the dose-related incidence of hematologic neoplasms in China. Letter-to-the-Editor. J Natl Cancer Inst 90,469-470 Wong 0.(1999) A critique of the exposure assessment in the epidemiologic study of benzeneexposed workers in China conducted by the Chinese Academy of Preventive Medicine and the US National Cancer Institute. Reg Toxicol Pharm 30,259-267. Wong 0.(1995) Risk of acute myeloid leukaemia and multiple myeloma in workers exposed to benzene. Occup Environ Med 52,380-384. Wong 0.(1987a) An industry wide mortality study of chemical workers occupationally exposed to benzene. I. General results. Br J Ind Med 44,365-381. Wong 0. (1987b) An industry wide mortality study of chemical workers occupationally exposed to benzene. 11. Dose response analyses. Br J Ind Med 44,382-395. Wong 0.(1996) Risk of myeloid leukaemia and multiple myeloma in workers exposed to benzene--Response. Letter-to-the-Editor. Occup Environ Med 53,357-358. Young R.J., Rinsky R.A., Infante P.F. and Wagoner J.K. (1978) Benzene in consumer products. Science 199,248. 19