Document nkxv7ZoGwz0qa7mdYG7KpqJN8

LetterSo The Editor In November 1983 the Collegium Ramazz.m. held an lnumational Conference on Benzene, which had as its objective a scientific update on the toxicity associattd with occupational exposure to burzcne [Am J M Med, 19851. A second meeting to discuss benzene risk assesma was held by the Collegium aad hosted by the New York University Medical Center in March 1985. Subsequent to these meetings, in a prtface to the publication of the proceedings of the 1983 Collegium mcczing, Dr. Mehlman focused his commtms on `the apparrm need to develop hforrnation" and suggcsted that despite the considerable research alnady undcmkm more information was needed before the scientific community d d properly urge substantial reduction of worker exposure to benzene [Mthlman, 19851. I suggest that this type of approach to the rrgulatron of toxic su- can k characterized as "Thke no protective Mion until ak$mnve eviaknce becwnes m.1- table." Since "definitive mi-" iuelf is subject to varying interprctauons and dtbate, we face siruatioos in which appeals to scientific maidousncss and extensive rrpeationmay serve todclay and abonthe promulgation of Wth standards n#.rssary to protect Amencan worlccrs and the general public. The preface m e r declares that `to avoid all exposure [to benzene] until all questions relacing LO its toxicity arc dctcrrm.acd is neither liircly nor pracncal." Unformnattly, that is hardly the problem. On the conmy, commonly very tiale or no protection against toxic substances is affordai until exhaustive recarch rcIating to toxicity has k n performed. This prevailing practice d t s in continuing exposure to toxic materials while scieatific inquiry raises more questions than it answers. `Studying a subject to death" unfomrnatcly somctimts includes the deaths of those exposed. This paper wu submmed for publraoon July 22.1985. '& VKWS ex+ do not n r m u n l y represent tbosc of OSHA. Address repnnt rrquesu U) Dr. M e r F. Infanu. Htrlth sunduds Progruas. OSHA. U.S. Depurmenr of Labor. 200 Constmuon Avenue N.W.. Wastungm DC 20210. Acccpcd for pubirPtton Much 11, 1986. - -c HOW MUCH INFORMATlONIS NECESSARY? The cnut of the matter is how much information is ~lcctssaryto trigger public health intervention. This may be quite M m n t from how much we would like to know about the mecfianisms of action for benzcne toxicity because of scientific curiosity or for approaches to medical management. For example, aithough the determination of whether then is a "practical threshold dose" or a 'no effect" level (Mehlman, 19851may be of scientific interest. it would seem to have little bearing on the issue of promulgating a standard to protea workers from uccdicss exposure to benzene in light of cummly available data demonswting toxic effects below any exposure Limit that could bc r a p r e d of industry at this time. Yet, argumcm for the prior establishment of a "w effect" lwei is the most reccm in the series of reasons that have often been the basis for argumcm that oppose OSHA promulgating a new benzenestandard. POINTlCOUNTERPOINR A PROLONGED OVERTURE A brief review of past major arguments shows hterphy between knowledge of benzene's toxicity and a series of argumcnrs raised sequentidly that obspuatd rulemalung.Inthelatc 19609andwellim6thc1wostberrwerts0mewhoargucdthat leukemia observed in workers cxposcd to knzerw could not be due to benzene e x p u r e becaw studies of btnztnc toxicity in expermvaral animalzhadnotdemonstmud the inducnon of cancer (Olson, 1m.Although this argument was never convincing, it has ken answcrai msoudi@y. smni chdicc presented at rhe Collcg~ummetring demonspate tbe inrllrlxinn of multiple site-specitic caclccrs in expenmemalanimais. In 1976, NIOSH updatcd its criteria documcm on btazcne and judged the clinical and epidemiologic evidence that benzene is leukemogenic to be conclusive. Although there is no evidence for safe e x p u r e levels for Qucinogcns, because of feasibility limitations NIOSH rc~mmtndedthat no worker be exposed to benzene concentrations in excess of 1 ppm. Subsequently. OSHA proposcd lowering the permissible exposure limit (PEL)to 1 pprn as an %hourtimc-weighted average. In response to this effort. urdusay wimesscs raised several questions (Tabashaw and Lamm. 1 9 7 about one of the major epidemiologic studies [Infante et al. 1977aI submitted to the OSHA record. These issues were responded to and clarified by NIOSH in 1977 Ffante a al, 1977bI. More recently. in an article entitled "Statistics and Ethics in Medical Research" (Alrman. ISSO], the reanalysis by industry witnesses of data from the MOSH benzene cohort was found to be inapproPriatt. At the OSHA fact-tbding hearing on the proposed benzene standard held in 1977. an additional argument was raised that an increased risk of leukcmia in workers exposed to benzene was in conflict with 'the general experience of physicians pracucing occupational medicine" [Olson. 1 9 7 . This form of negauve clinical impression. of course. is of highly limited utility for many reasons. including. its inabiliry to demonscrate that a substance is not a canhogen. Morwver. clinical observauons among Italian occupational health physicians w e n used more than a decade prior to the OSHA hearing to support evidence of h n c ' s carcinogenicity [vigliani and Saita. 19641. As a result of the 1977 OSHA fact-finding hearing, a final benzene standard allowing a PEL of 1 ppm was promulgated by the Secretary of Labor. This standard was cballcagd in court and inv- because a pluniiry of Supaemc Courtjustices acccped the lndusuy qumcnt that OSHA had mtdaarm*nedw h a k a s i g d i c a u t risk existed at the 10 ppn PEL. kmc incorreniy Coaclutk dmt tbc Court found no haurrdassoc'uted w i t h ~ u t p o w v t b t l o ~ 1 O p p m . T o t b c c a n a a r y , t h c c o U r t did not qutstlon the validity of dre Secraary of Latror's 6ndiags the basis of substantial evidence, that (1) exposure to bcnztnt crcatcs a risk of canccr, c d s o d damage, and a variety of non-malignant but potentially fatal blood disorders, cvSn at the level of 1 ppm or that "(2) no safe lwei of exposure-&been sho~n. [Induspid Union Depanmcnt. AFL-CIO v American parolwmInstitute, 19801. As a result of the Supreme C o w decision. a quano;tatve leukemia risk assess- ment of benzene was undertaken by OSHA staff, submitted for peer review, and published in 1982 Wtc a al, 19821. The rtsults indicate that a "significant" risk of developing leukemb iS associated with Lifetime ocqmional exposure (45years) to 10 ppm benzene avcraged over an 8-hour day and that a significant reduction in that risk could be achieved by lowering the PEL to 1 ppm. In rrsponse. toxicologists empioyed by the peaoleum industry have argucd that the assasmu overtsfimatcd the leukemia risk because the bcnzme exposures cxpaicnccd by the cohorts used for the asu_umcnts w e n undcrcstimated [VanRaalteaal. 1984).Thi~Qesnotappear to be the case basedon the best avaiia, bialaeetdnaftnaII 19841. More recently, Mch&an [19W statcd that the urposun data uscd in the risk a~scssmentarc "generally bctieved to be in m o r . " 'Ibis assmion was unsupported. Although rcprtsmtativcsof a u s a y or thcir collsultamshive chalkngd the cxposurc asstssmc~.relatively few. if any, mdepcmbt s c i e have done so. The assess- ment assumed that &hour time-weighted average exposuns to thc NIOSH cohort ranged between 10 and 150 ppm dependtng on tbe calendar time period that the benzene txposurts occurred. The assumpon s c ~ l l sc o m v e since it, in fan, overcstimatts average exposun based on available data for the NIOSH cohort. Rcasctnabic assumptions a h were made for the Dow Chemical Company &rt. which exptritnced average txposurts of about 5 ppm [Infnnte ami white, 198q. RISK ASSESSMENTS In April 1984 a draft copy of thc DCW OSHA-pmpsal benztne standard with its preamble was submitted i n f ~ d byy OSHA to tk office of Management and Budget ( O m )for review. OMB tequested that OSHA h v c a h n c risk assessment performed by a sour^^ extcnral to the Agency. As a d t , OSHA rrqueJted Dr. Kenny Crump to perform an ladepcndmt exposure assesmmt of the NIOSH cohon and a risk assessment. This risk assessmenf was based on data tapes of the bertlem cohorts studied by NOSH aad the Dow chemicalCompany cohort. and on the rtpon of a Chemical Manufacturers Assocl'an'on(CMA) cohon study of benzmt exposed workers. Using a cumulative dose and relative risk mQdcl (similar u) drat used by White et al[1982) and data from the threesaxiics justmentionai. Crump and AUen I19841 estimated that 40 years of occupaonal cxposurr to 10 ppm btnzcae would result in 88 (95%confidence interval = 34-174) excfss leukemia deaths ptr 1.ooO workers. 'Ibis eStimattis very similarto the risks of 44-152 per 1.ooO projected by White et ai [I9821 or to the risks of9.5-170 per Loo0 estimated by extraplation of the IARC assessment [IARC. 19821 to benzene levels of 10 ppm llnfantc and White. 19851. For excess i~ukcmiarisk a~so~kuwdith lifetime occupationalexposue c to 1 ppm benzene. C m p and Allen [1984]stimakd 9.5 per 1,OOO based on 40 years exposure while White et al had estimated 5-14 per 1,OOO based on 45 years exposure. The results of these risk assessments support the assumptiolls and estimates of leukemia risk made by White et al in 1982.Moreover, the benzene risk assessments conduccui independently met ct al. 1982;M C , 1982;Crump and Men. 19841 lead to estimated risks of death from occupational leukemia that have been chamcurizui by OSHA as si@cant [OSHA,1983)as a resuit of average benzene exposures to the current PEL of 10 ppm or to the OSHA draft proposal of 1 ppm. Subsequent to the review of these benzene risk assessments, OMB asked the Agency to have Crump and A k n perform an additional risk assesmcnt based solely on the CMA study of benzem exposed workers. This was done. The tesults indiGucd an even higher cstimatc of 121 excess leukemia deaths per 1.OOO workers exposed for 40 years to benzcne levels of 10 ppm and 13 per 1.OOO for those exposed to 1 ppm for 40 years [ C m p and Allen. 19841. REPRISE As the above issues seem to have been exhaustively explored. there is now a new demand for scientific certainty mehiman, 19851. It calls for information showing that bone marrow lesions do not occur below a certain lwei of benzene exposurc in ordcr to establish a `no effmlevel" for sucb expornre. In response to rqucsts by OSHA [19851, a study addressing low-level benzmc Cxposurc has rtcwtly becn released by tht Chemical Indusay lnstmnt of Toxkdogy [Neal. 198% Ertxson et ai. 19861.The results demonstrate that a single &hour exposure to 1 ppm benzene in rats nsuited in a significant increase in micronuclei in bone marrow polychromatic erythrocytes and sister chromaad exchsngcs in peripheral blood lymphocytes. Thcse findingsare both qualitatively and quanatatively consistent with other recent ex@mental findings [Baarston a al. 1984: Rosm et al. 1984: Ti= et ai. 1982: GadelKarimet al. 19841 and epidemiologic study results [PiccianO. 1980;Sarto a al. 1985). While one may argue about the existcncc of a `no effcct levei" for benzene-induccd bone marrow lesions or their llcctssary relevance to the long-tcnn carcinogenicity of the chemical. such a level is ci&y lower than the. 1 ppm permissible exposure limit that OSHA has recommended in its wideiy circulated draft proposal. The proposai was cransmitttd f o d y to OMB for review in April 1985. Every day the scientific and political debate over benzene continues to delay necessary mandafed conml of exposure. workers arc subjtcted to the risk of u~tcessary disease and suffering, and-costs to society as a whole arc increasing. For the most pan. these costs. to date. have not been paid by industry. In "An Interim Repon to Congress on Occupauonal Diseases" [U.S.Depamnent of Labor. 19801, for example. the ! k c r e v of Labor concluded that 86% of the income suppon for those severly disabled from an occupational discase was provided by public programs. ie. 53% from social security, 17% from veterans benefits, and 16% from welfm. whereas only 5 1 was paid by workers' compensation. which is paid by industry as pan of its cost of production. The incrcase in succcssfid plaintiff toxic ton Liability suits has raised questions whether the (inexpensivei workers' compensation systcm will suffice to shield industry from these liabilities [Kit&r~U.1985). Through passage of the Occupaaonal Safety and Health Act of 1970.Congress clearly intended to control and eliminate. as far as possible. Occupational disease and death. Congress also intended to place the economic cost of occupational illness !r REFERENCES T r k n h w [R. kmm SH (1977): Benztae.adleukunu. -the editor. LuraU:867-868. Tice RR. Vogt TF.Casta DL (1982): Cytog.cnet~e: ffsoofinbkd bcazm~ID lll~vllybonc m w . In Tie RR. Cosu DL KM (ais): 'Gcnoamc Efieco of Airbornt Agents." New Yorlc: P-l RCSSp.p 257-275. U.S. DepMmem of lrbor (1980): "An Imam Rcpon to Congresr 011 Occuproonrl Diseases." Wuhingtoo. DC:Amsunt Sarruy for Policy. Evaluaaon a d Rcsarch. Depvtment of Labor. Van Rule HGS. Grass P. Imine D (1984): Tackling a very difiicuit problem. Letter to the Editor. Risk -4: 1-2. V i g W EC.Uta G (1964): Benzene .ndleukemrr. New Eng J Med 271:872-876. White MC. lnfnare PF.Chu KC (1982): A qwaopnve of larlraau m o d i t y USOClPICd wth aaupmad uposure to btnzaw. Risk AMI 2:195-20Q. Author's Re8ponse -Benzene: Evaluationof Control Measures not Competitive With Basic Research Requirements Dr. Infimte has charactcrizcd my st~ttlllt~in: the Ramazn* n'tRoceedingson bcnztnc fjvlefilman. 19851 that "more rcscIvch is n d a J kfm the scientific am- munity P m m Y urge . rrduaion of workcrcxposurr to benzene" as 'take no prorecave action untii definitive evidence kcames available." The two quotations arc rn even appioximauiy equivalent. The raiuction of exposures to a safe levd is a wdlesrablishcd principle in the fitlds of occqatiod and-ne health. Many scimhts bctiwe that a level of 10 ppm is adequate to nduce the indalce oflollaria in workers o c x q m i o d y exposed to benzene to a level undetcaable in tht huxnan popllanon. R#xnt proposals to lower the standard further to 1 ppm arc basedon a conap of preventive prudence. I have no quam1wirh Dr.Infante's conciusion that then exists evidmcc in his study for a qualitative miationship bawan benzene exposure and leukemia. How- ever. since Dr. Infante has also stated that the underlying scienfic basis for risk assessment is lcss than certain p t c et al. 19821, further scientific inquiry into the medunism of action of benzene may increase OUT knowledge not only for application to risk assessment methoQlogy but also for udmmdm* g acarmentaadcurcsfor the disease. Scientific inquiry may indeed raise more qutstions than it answers. but that is the nature of Scienct. As Goldstein and Snyder [1982] have noted, 'Considerable funds have been spent on the regulatory process and only a relatively d amount on the scientific efforts which could resolve the problem." Surely regulatory policy should be based on a solid foundation of valid information. Myron A. Mehlman. RD 7 Bouvant Drive Princeton. NJ 08543 REFERENCES GoldWn ED.Snyder CA (1982) BCIIZCIE1U-. h i m a Sci RCS 25:277-289. Mehlman MA (1985) Rchcc. Am J Ind M a l 7:M1-363. White NC. I n f m PF. Chu KC (1982) A quprmutiVe d i n a t e of leukemia monality wciarcd with C X ~ O S Uto~k m c ~R.isk A d 2: 195-203. Acccped for prbliauon March I I . 1986. Address q n m rtquesu to Dr. M.A. Mehlrmn. Director. Dcpumnn of Touaoiogy, Mobd Oil Corpwmon. PO Box 1029. Pnncaon. NJ 08450. i