Document QJ4bNj9bKMg47K47e80NpEdx6

Twenty Lessons from Asbe ' A Bitter Harvest of Scientific Information - -. By Dr. Irving J. Selikoff, M.D. it seems that we sometimes learn most from our worst mistakes. This certainly was the case in one of the greatest public sought to evaluate those factors) both prevention of disease in the future and provide help to those for whom health disasters in modern times -- cigarette smoking. When the marked increase in cigarette use began after World War II, there were few predictions prevention is now too late. There has also been the hope that what we have learned from the asbestos tragedy will provide principles that may help to of what was to occur in the 1960s, 1970s prevent similar disasters in the future. and 1980s. More recently, nature has been similarly unforgiving with regard to asbestos, perhaps because we were TWENTY LESSONS reluctant to heed the warnings that we were given. It was found in 1924, for. example, that exposure to asbestos could result in fatal disease. In that year, the British Medical Journal published a report by W. E. Cooke of a young woman who had worked with asbestos and who had died with extensively scarred lungs. In 1927, again in the British Medical Journal, he gave the disease the name it still bears. Pulmonary Asbestosis. By 1930, additional British studies demonstrated that such scarring was very common among workers exposed to asbestos and these observations were soon confirmed in our country by Fulton, "i ; k- ' .v'.v.-.-. ' v ; irr;V -V ' Dr. Irving J. Selikoff far, W. J. Nicholson has calculated that We have been taught much by the asbestos experience. This could be analyzed differently by the industrial hygienist the regulator, corporate risk manager, clinician, industry executive, union official, pathologist insurance company executive, lawyer, physiologi economist, molecular biq'logist and others. But perhaps the most pertinent lesson's of all have been those gleaned from a public health point of view, fror the perspective of how to prevent preventable disease. Twenty have beei selected as being central to EPA responsibilities and concerns. Dreessen, Lanza and their colleagues as there have been more than 100,000 1. Latency: Although tissues and celts well as by other scientists. By the deaths of asbestos-associated disease begin to react to the presence of inhali mid-1930s it was well established that and that we may look forward to more asbestos fibers on a microscopic level asbestos inhalation could frequently than 350,000 additional such deaths within hours and days, clinical effects? cause disease and that such disease before the effects of past exposures run not seen for years or decades. Even WS might be fatal. Scientific research since their course. These projections are the extensive exposure that was freque then has added much information but in concerned with cancer deaths from found in asbestos factories in the past, a 'sense, this largely defined the different occupational sources. There will be was commonplace to find no X-ray or ways that asbestos could kill. Thus, in additional excess cancer deaths from pulmonary function change until five, 1935, Lynch and Smith in the United non-occupational exposures, as well as ten. or more years had passed. These States and Gloyne in Great Britain, noted deaths from asbestosis, but it has not yet . clinical probes are insensitive for the association of lung cancer and been possible to make appropriate demonstrating early changes. In one asbestos work, and during the 1940s and quantitative predictions. Further, the study of 1,117 asbestos insulation 1950s cases of pleura? and peritoneal predictions are predicated on the workers, regularly employed in the mesothelioma were seen in assumption that, after 1980, ssbestos construction industry under asbestos-exposed workers. This exposure will have ceased. Initial circumstances in which significant association was clarified and firmly - experiences suggest that this was a exposure was the rule, more than half established in the first half of the 1960s by dubious assumption, and that the tragic those with less than 20 years from onj Wagner, Selikoff, Churg, Newhouse and toll of death and disease will extend of exposure still had normal X-rays. A- others. Additional neoplasms (malignant longer than we thought. Moreover, the that point, most X-rays were abnorma growths! -- again, further ways of dying 9,000 or so excess cancer deaths from We should not expect to see early - -- were subsequently found related. occupational sources now seen each year evidence of asbestotic change. We are now in the midst of widespread are accompanied by many times that The same constraint is the rule for asbestos disease resulting from number of workers with a'sbestosis of asbestos-associated cancer and for fat exposures during the past 60 years. So greater or lesser severity, with greater or asbestosis, as well. In a prospective lesser disability, but insufficient to study of 17,800 asbestos insulation . directly cause death............... workers, 1967-1976, relatively few - (Dr. Selikoff is Director, Environmental Sciences Laboratory, Mount Sinai School of Medicine of the City University of New York.) Inevitably, the observation of so much serious disease has ted to increased understanding of the circumstances in which it has occurred, (as scientists asbestos- associated deaths were seer less than 20 years from onset of their work exposure. Indeed, most deaths MAY 1334 CAP CO JEN 0010337 's'Fli-'-' occurred 30.40 or more years after ' ' -j; exposure had occurred." The disease and deaths now being'r v experienced are the results of exposures in the 1940s and 1950s. with,the 1960s . ' beginning to make their contribution, the legacy of our mistakes of the past Current exposures will not show their effects until the year 2010 and ubsequently. 2. irreversible errors: Once exposure has occurred (with one exception so far, see below) the die seems cast We know of no way to remove or neutralize fibers in the lung or in other tissues (to which some migrate). Whether this is because of the residual fiber tissue burden or because of cellular and molecular changes is not known. From the point of view of prevention of future disease, control of human exposure, wherever and whenever it is occurring, is an emergency. Sometimes this is not appreciated. Somehow when the disease effect is 30 years off, there is little sense of urgency. This is wrong. There might be less complacency about friable asbestos in schools and public buildings if this were better appreciated. 3. Dose-disease response: Less asbestos, less disease; more asbestos, more disease. This central fact provides guidance for what is to be done. We may not be able to control every last fiber in the environment, but we can take some comfort in knowing that as our ' engineering and regulatory measures become more and more effective, there will be less and less disease. However, the 'dose* of asbestos is cumulative, with newly inhaled fibers added to the burden already present. Therefore, each opportunity for asbestos exposure should be controlled not only because of its own hazard, but because it would be adding to the, risk from other sources. This is a good? example of the correctness of the definition of dose as "intensity x time.' With many agents, it is very difficult to ascertain "dose* associated with disease being seen, since the exposures responsible for such disease occurred decades before, when measurements were: cot made. Seidman and his coiftagues have recently reviewed a unique set of circumstances demonstrating the dose-disease response nature of asbestos disease. They traced the long-term mortality experience of a large group of asbestos factory workers employed during World War II. They were all exposed to the same fiber, making the same products, using the same machinery, in the same plant. They differed, however, in one respect. Because of wartime conditions, some worked for a day, a week, a month, several months. Others worked from the Two workers removing asbestos from a ceiBni time the plant opened in 1941 to when it 5. Disease with (ow-Ievel exposure: the closed in 1954. Since the intensity, for dose-response relationship for asbestos the groups involved, was the same, dose appears to be linear. This predicts was proportional to duration of exposure. disease with low exposures. The model Lung cancer Incidence for the various has been shown to be correct In 1965, groups increased with increasing dose. Newhouse reported mesothelioma 4. Disease with brief exposure: There , have been numerous reports of relatively ! brief exposure and the subsequent occurrence of disease. However, many reflected individual experiences and for . diseases such as lung cancer, they did' not "prove* an association with short exposure. The risk of brief exposure became better established with the study of mesothelioma, a neoplasm which has few known causes in humans other than asbestos. When mesothelioma is found, prior asbestos exposure is looked for and usually found. When asbestos exposure occurs, there is significant risk of subsequent mesothelioma. The extraordinary relationship between asbestos exposure and mesothelioma was perhaps best considered by Cochrane and Webster. They interviewed 107 patients in whom the diagnosis of mesothelioma had recently been established by biopsy. In 106, potential prior exposure to asbestos was elicited. The experiences of Seidman et al (see above) have provided the necessary population-based data to confirm the keen clinical observations previously ' made. The mechanism by which brief exposure subsequently results in disease among individuals whose only known exposure had occurred as.* result of residence in households oflasbestos workers, or by virtue of living within a half-mile of an asbestos pldnt in London. Such family contact and neighborhood exposure mesothelioma has been widely confirmed and its importance documented. Of course, it can be argued that such exposure is not "low," particularly since it results in a significant amount of disease (in one current study, lung cancer risk appears to be about doubled and mesothelioma to be responsible for approximately 1% of deaths occurring 20 orjoore years following the initiatfonof household contact exposure). What will happen at the lowest levels of exposure is still not known. There are other uncertainties. Brief exposure, if fairly intense, produces disease. Long-term exposure, at relatively low levels (household) produces disease. It is not known whether brief exposure to low levels will produce detectable disease. Complicating such analyses is the cumulative nature'of even low-level exposure. The problem is not unique to asbestos; it is also the case with PCBs, dioxins, etc. This again points to the . necessity for control of all sources. ' is not known. It may be related to the 6. Multiple factor interaction: It has long retention of fibers in tissues but it may been suspected that much human not. The same phenomenonTs'seen in disease from exogenous sources is bladder cancer following exposure to multifactorial in nature. Asbestos taught beta-naphthylamine or benzidine or in us that this is indeed so. When the angiosarcoma of the liver after vinyl - experiences of the 17,800 asbestos chloride exposure where there is no insulation workers, with smoking habits . evidence for retention of the chemical known and observed prospectively, were carcinogens. compared with those of 73,736 like men 22 ..EPA JOURNAL CAP CO JEN 0010338 ^ 1 ? .;*rlnthia American Cancer Society'*-\V/>^-f - 8. Industrial origin of environmental T. asbestos-laden schools. Their risk j "J** prospective study of cigarette smoking, a~["disease: -The factory gate and.the factory.-.s .depends upon their age as well as thtik " ~ "remarkable multiplicative effect was . ;' fence are porous. Almost all asbestos-?i 1;. prior asbestos exposure. A 55-year-cfd ' seen. Men'who (fid not smoke and did~'r exposure Is industrial in origin, although'^^-.teacher with only 10 years in such a . ' network with asbestos suffered 11/' . some fibers derive from erosion of *? school nevertheless has important rislc deaths per 100,000 man-years. For" ' ` natural outcroppings, and water may be'- Op. the other hand, since there is littl asbestos workers who did not smoke, it . contaminated as it filters through.v-> - .background risk of mesothelioma, was five times as much, 58. On the other.. asbestos rock formations'. Such .. v asbestos acts, as an initiator with risk hand, individuals who smoked but did environmental contamination is very* .; -L increasing with age by approximately < 'not work with asbestos had a death rate . limited, however, particularly in terms of .* power of fourl Again in school' of 122 per 100,000 man-years, and men disease. ......... **iV .. t. . circumstances, this points to the * who had both exposures, asbestos and agareae smoking, had 601. There is evidence that the same cigarette smoking-asbestos interaction may explain the increased risk of cancer of the esophagus, oropharynx and buccal cavity, and iarynx. There is no such interaction, however, for mesothelioma, cancer of the stomach, coton-rectum or kidney--both smokers and non-smokers suffer equally. Conclusions important for prevention may be drawn. First, all individuals known to have been exposed to asbestos should never start smoking or, if they are smoking, should stop immediately. This is particularly important since data indicate that there can be reversal of risk once smoking ceases. Asbestos insuiation workers who stop smoking, after 5-10 years, have about one-third to one-half the risk of lung cancer of their mates who continue to smoke. While cancer, once it occurs, is not reversible, cancer risk may be. A corollary conclusion, however is inherent in the above observations. Since smoking cessation will not affect risk of mesothelioma or the other neoplasms not associated with smoking, it will be equally necessary to control asbestos exposures. Both measures are needed. 7. Product use: For every worker employed in the manufacture of asbestos products, there may be 500 who would ' use them or be exposed indirectly during such use. It is therefore unfortunate that at the outset of our asbestos experience, we thought of "asbestos workers" -- men and women employed in mining, milling or factory work. The first phase of asbestos exposure and accompanying disease was associated with product manufacture. Later, during the last 40 years or so, there was increasing attention to disease associated with product use in the construction industry, shipyards, powerhouses, chemical plants and refineries, brake maintenance and brake repair, etc. We are now entering a third phase -- in which asbestos exposure will be associated with envmmmemtal exposures, during repair, renovatfiant removal, and maintenance of the asbestos put in place during Phase Two. We have learned the difficult lesson of not thinking of asbestos workers, but ' . 9. Multiple effects/muftiple agents: Asbestos can produce a variety of - ?- importance of prevention of exposure -. children, with long lives ahead of their illnesses, ranging from pulmonary and 12. Complexity of societal consequenc pleural fibrosis to lung cancer, pleural .. ^ It has long been a truism that, from an and peritoneal mesothefibma, * ecological and environmental point of gastrointestinal cancer, cancer of the ' oropharynx and buccal cavity7 laryngeal ' view, everything is related to.everythin else. With asbestos, this dictum applia cancer,.andikidney,,cancer. Other effects],''* to other circumstances, as well. Curren too, are now being seen, includihg.^;.... ; - ' litigation has been marked by bapkrup- immunomodification and serological of major industrial firms, thousands of changes. The other side of the coin.. - ' lawyers face each other in courts ciogq important from a diagnostic point of ' by suits seeking help and redress, view, is that virtually all of these diseases insurance companies are concerned wi and modifications can be caused by., potentially monumental costs. It has other agents, as well. Even - - been variously estimated that asbestos mesothelioma, so highly attributable to disease payments to victims will range asbestos, can be found to have other between 40 and 150 billion dollars, in causes. Already, erionite has been seen .-. . addition. Professor WilUam G. Johnsor to produce pleural and peritoneal of Syracuse has calculated that social mesothelioma among residents of' : - costs of asbestos disease due to previc Cappadocia, Turkey, and there is - .1- exposure will total more.than three considerable concern that other - hundred billion dollars..Industrial materials, particularly man-made fibers, practices are changing, with the adven may eventually be associated with . of substitute materials, many of untest mesothelioma risk. ; toxicity. Doubt has even been cast on t 10. Environmental persistence: It has been said that asbestos has "a half-life of infinity." This is remembered ruefully as ' one considers the 30,000,000 tons of asbestos put in place from 1900 to 1380, in our ships, buildings, schools, chemical plants, refineries, powerhouses, factories, etc. Approximately 700,000 tons of. . effectiveness and applicability of the workers compensation system. We are also beginning to see anothe legal tangle, perhaps of equal or greatc complexity, with legal battles shaping i over who is to pay for the expense associated with abatement of asbestos schools and public buildings. insulation materials were installed in the 113. Early utilization of industrial hygiei same period; much remains., 71; engineering: Failure to respond early t< 11. Complexity of initiation and promotion: There has been much - - scientific interest in recent years - :`: concerning the concept that carcinogenic agents may either initiate the cancer process or, once initiated by other - ; agents, promote its development. Asbestos seems to do both, according to circumstances. Thus, for lung cancer, the data suggest that it acts as a promoter, multiplying the background risk at each attained age. A 50-year- old individual ' has a much greater background risk of . lung cancer than, let us say, one who is 20. Asbestos, in each, multiplies that risk.. It therefore does not achieve very much to restrict hiring to older workers, in the hope that latency would give them a very long life before lung cancer might strike. Two latencies have to be considered -- background exposure and asbestos. This information concerning the disease potential of asbestos carried with it the omission of measures needed to contr< exposure. Asbestos became entwined i industrial procedures with hazards inta When, decades later, there was increasing concern with disease potential, it was doubly difficult to change uses and procedures integral.. with the entire fabric of industrial . . production. Moreover, since the industrial engineering measures that' were neede'd were being telescoped in' a relatively short period of time rather than having been accomplished over many years, attendant costs were correspondingly high. To further complicate matters, these costs had to borne at a time when the product itself was being questioned and sales were decreasing. asbestos-exposed workers. would apply, for example, to teachers in . MAY 5934 CAP CO JEN 0010339 " 14. Disadvantages of fragmentary firms producfng'salmonellarho public - &W.- disease are expected at such levels, it is regulatory approaches: There has been -f relations groups operating on behalf ofV;sj, unlikely that tits vast resources recess.'? less than complete interaction and T~~ the pneumococcus, the diphtheria or this i^..for these studies will ever be made : interdigitation of knowledge, experience, staphylococcus.' ^ ' ' . . t^'available.' Animal experiments at very ' research, regulatory actions. Oreessen of It has become dear thatjustas in the' low levels will always have the the U.S. Public Health Service undertook 1890s, scientific research is necessary for disadvantage of insecurity with regard t? a rather elegant study of asbestos the identification of causes of disease.' 'extrapolation to humans." ' disease potential in the early 30s But the simple gathering of data is only ; The discussions, while interesting and (published in 1338). I expect that It was - one part of the process. Utilization of the . -important from s regulatory point of hardly known to the National Cancer ' information is also required. r 'gulatory :_'"* view, nevertheless have an air of Institute's Advisory Council when, in " measures are needed, often of .. : ! unreality at this moment, with workers 1951, it rejected a proposal by Leroy " considerable complexity. still being exposed to permissible levels U.Gardner, then dean of experimental dust disease pathologists, to study cancer potential of asbestos in animals (he had early hints of such findings in his pneumoconiosis experiments)./ There has been less than complete integration of the interests and studies of the EPA. NIOSH, NIEHS, CPSC, NCI. Fortunately, mechanisms exist for such interdigitation. 14 Science is necessary but not sufficient: When, in the latter half of the : twbapNnbarl1hieesturug6eeetosbc.relmvlepebleoIlneanoabalsmtcertdniusetteymomhesocrtsewairheosoib.alnceearlfTbtcbecssispchleeresaefiegepirf,ortnoaoeetoIprgcarsnplieuoonniiildfnnveccmlgoilegditetuueainptlontbyonnowitlvooeeicdfifocrmcnixeocrloaami:evfatpunmiyestIatrnohstipdekoopditreo,deontfoecanetovtlrlshblernsseukuiyaitistrhbttacsyisrPohisnholewicle'nadmtalyifhrsaer.hwe,eg;ibncaaeilthtpeishss.-t's7=`.;..'. ''nosaCssrdltoaheaufoocooftreppnecmmgnmsanoeoneceeeiomrogrervwutrsrerwnipeannsthcsauaodmihatotaknimhreomtnksebyaroob.taonaaolituesnthunknetr2tetedhsoo0oamnvawfrsfaevmewbtscebhrofehyiceiuilutllslooeerhitlcotourvtooahtmonwneseruwetlarfsltmotilioebihnpntonanvheerotgfeeteertesver1nclsscere0raaipowaa0rufsnueiyllehbtscitnitreooeiemycitdmsnerhwau1sasds.nch0y.aoh0di;tnC?r.- 19th Century, it began to be found that of the potential importance, help to - 19. Limitations of epidemiology: These serious human disease could be caused explain the paucity of exposure data now. are widely acknowledged -- evidence is by exogenous agents (infectious) a at hand. ' ' . ^ ; . _ based upon human disease that has revolution in scientific thinking began; there was now not only description, but causation. (It is instructive to appreciate how recent this has been; 1982 was only the one hundredth anniversary of the discovery of the tubercle bacillus by Koch.) It was soon found that the identification of causes could be followed by tfcecr control. Pasteurization of milk, sewer systems, and clean water supplies were put in place. In the first half of the 20th Century, we again applauded those who discovered still other causes of disease, often metabolic, endocrine, or nutritional. The same approbation has not inevitably met those studies which have identified some of the newer exogenous causes of disease. The tobacco industry has given no testimonial dinners to the researchers who have shown that this year we might expect more than 100,000 deaths of lung cancer due to cigarette smoking (plus additional excess deaths of pancreas, bladder, oropharyngeal, esophageal and larynx cancers, plus deaths of cardiovascular disease and emphysema). As we consider . S-naphthalymine and benzidine, 4*aminobiphenyl, nickel smelting, arsenic, vinyl chloride, lead, cadmium, chromium, etc., we are reminded that, in the 1890s, there were no trade associations for the protection of the cholera vibrio or the tubercle bacillus, no 17. Recruitment of constituencies: An . important asbestos lesson, perhaps related to what has been said before about science being necessary but not sufficient, has been the increasing understanding that application of knowledge can be speeded when those who are directly affected have the information that intimately concerns, them. OSHA operates best, perhaps, when both labor and industry are aware of the facts that form the background for OSHA regulations. EPA's requirements that parents and teachers be told of asbestos findings in schools, is of this genre. Control of asbestos exposure depends at least as much upon understanding at the shop floor, as upon intricate regulations ensconced in the Federal Register. If we don't have '- already occurred, available methods are insensitive in detecting otirthan very gross and marked effects/studies are no - suitable for smaller populations, there is frequent lack of concomitant exposure data, etc. Further, with the inevitable biases and variability inherent in human population studies, residual uncertainties persist and sometimes the best that can be achieved is the acknowledgment of . `'associations' rather than definitive causation. Yet for asbestos disease, epidemiology has served us well and we have had only limited assistance so far from animal studies. It is to be hoped,that in coming years, with other agsrifsTwe will no ` longer have to depend so heavily on epidemiological studies of human , experience. understanding of what has to be done on 20. The concept of 'industry* identity: the part of supervisory personnel and There is probably no such thing as a workers, there will never be enough monolithic industry, each sector being inspectors to insure safety. With identical with all others. Some industry understanding, we will need few. units are knowledgeable, others not All this translates into an important Some are concerned and truly educational function for EPA1 responsible, others couldn't care less. Who, then, speaks for 'industry'? My HOW MANY ANGELS ON own experience with asbestos problems' THE HEAD OF A THRESHOLD? . .. indicates that trade associations do not 18. Disease: There are learned and often esoteric discussions of how much ; disease might be expected at very low levels of exposure. Calculations are made . always speak/or the most knowledgeable and the most involved industry units. This can be an important disadvantage. and projections offered. It will'be very difficult to verify or contradict these. Epidemiologically, very large populations will be required, carefully defined as to . biases and variables. Since few cases of ' 24 EPA JOURNAt CAP CO JEN 0010340