Document 3JQyy5aE7N0NYR2dMOwYmRBJD

'~ J 08705797 11:41 -> 216 58~ 8314 D. STNDTHEN VIA FAX TO: STROTHER, DALE E. BP AMERICA FROM: ATTN: DALE STROTHER TO: FAX PHONE~: 216-596-8314 JOb Number: 66491746-009-27-0004 TIME: Wed Aug 06 12:24:53 1997 6 pages including coYer sheet BP-00017785 11:41 -> 216 SBo 8314 D. STROTHER !?age 002 August 6, 1-997 MazyPaxt~~ fl[f[J@flili]@ To: Benzene Task force (BTF) Draft comments on lEI?A's revision on benzene's cancer potency Attached is a draft outline of commengs to EPA that can provide a st."liwm<m for our cm.ference can tomorrow at 3:00 EDST. I regret that I will be out at an ali-day meeting on Friday, so iliat is not M alternative for rescheduling, but the .hour on Thursday can be adjusted. [fyou have not already do:oe so. fax to Mary Paxton at 202-682~8270: (fulloame) (company) can participate in the ti'ill"$lt BIT--conference (;a]J on Thl!lt$dlmy, Augma_L mft 3\: !ElD>J"; A hetter time this week would-be: can participate in the seoolllldl BTF conference call on 1l'llllunrntll2JS1, J1~1l11St ~ ma J:OO JED'!r; A better time that week would be: Yeu-wilJ be notified if a change from these two times is necessary. To participate, call 1-800~282-=-2343 at the appointed time. would likc a copy of the Caoadian exposure report. BP-00017786 - 11:42 -> Z16 SB& 831~ D. STROY~Em C<r>mmllll1l~IID1l ~f Q.llllte Amm~rrtt&lffi 1P\eftii'~ll~1lllm llimmn~ 1U. .IEirn~IroliDIITID~JIUmll. IP'zmtl~cti:iillll. A~llll~J9 '6Cmil'~nll!l@~llll.k lEffiYe~efcs ((J)fr" l.$emenne: AIDl UJjp>cdlmft<e'9 EPA's goal: Review the adequacy of tlle current benzene cancer potency factor (cpf) at i:he request of the Oftice of Mobil Sources under the 1991 Clean Air Act Amendments. Is this a worthy goal? Has EPA achieved its objective in a satisfactory fashion? AlPI' s goal: Be constructive. !Encourage any movement away from low-dose linearity default Point out what not considered. A. Issues posed by EPA to Ex~rt .!Review Panel: [Rob Schnatter's submission was focused on addressing these questions.] 1. Is this documem successful in "selecting a quantitative approach to estimating benzeJ!le risks especially at environmental exposure levels"? 2. -Does the document adequately justify use of the Plioflhn cohort C'Rims1cy, 19g7 cohort") as the ba.ctis of its dose-respcnse analysis and risk derivation? Tiris raises the issue of deriving a cpf for a pure substance vs. deriving risk estimares for what (e.g., ruixtwes like gasoline) that most wortcers and the general population are now eJqX>sed to. If ilie latter is the objective. it would argue for also (or instead) looking at smdlies of petroleum worker cohorts. 3.1 lDoes-the discussion of the various mode-of-action hypotheses appropriately descrire the current tllinking and capture consensus where such exists? Most. but not all (studies missed pointed out below), of current research into benzene!s mechanism of leukemogenesis are alluded to. but are not then really factored into JEPA!s modeling and quantitative choices. This leads AP! to pose the long-standing indulstry qoeey [m.~also in Section B], How much data does it take to move from EPA's default position'! Studies missed: DRAFT August 6, 1997, 11:40 AM BP-00017787 -> 2i6 58& 8314 D. SfROjMR 3.2 Ks the ntionale for specific model selection (e.g. li~ar at low dose) adequately eJtp!ained and supported? The tally of points for and against non-linearity does not really establish the relative merits of the argwnents and the overall we-ight of the evidence. Super-linearity in relative proportion of toxic to non-to:xic metabolires observed at levels where metabolism is saturated does not balie the possibility (in faci, li1relihood) dna& the production of toxic metabolites is sub-linear at the much lower. ambient levels-of regulatory concern. Even if Mlichalis-Menton kinetics are lfnea.rfor metabolites all the way to zero from below-the high-dose plateau, the relationship for the target event would not be expected to be linear. EPA equares any type of genotoxicity witlllow-dose linearity, whereas tllose specific types of genotoxicity associ':ted with benzene are now generally regarded as having non-linem doseresponses (Preston attachment). gene conversion in glycophorin assay clastogenicity aneuploidy 33 Is there another modeling approach that could be supponed by consensus infonnation? July 16 ERP meeting demonstrated consensus that there is adequate evidence to justify dle presentation of a range of non-linear estimaies to supplement the range of linear eslim.ates as presented in the current EPA draft. 3.4 Is-there a scientific mtienale for selecting a specific risk estimate from among the range of presented (i.e., low-dose linear risks of 4.7xl03 to 2.5xl0"2 per ppm. whose uppell' limit is justless than the curtent benzene cpf of 2.6xl02 per ppm)? Without explicit guidance, risk managers and regulators will tend to--select me probably overly protective upper lifuit as the-basis of the quantitative judgments. However. EPA's conservative interpretation of tlle information gathered on benzene over the last decade indicates that, even considering only linear estimates, the rislc falls below the existing cpf value. lif one were to put aside the strong arguments in favor of non-linearity. the weight of the evidence would support a point within EPA's stated xange, thereby effectively red111Cing the estimawd benzene cpf that would be applied in risk aJSSeSSments by about five-fold. .AJPlf is well aware mat, given -tfle extend of uncertainty surrounding these est:irnates, a five-fold difference may not represent-a "statistically significant difference," but in the implementation of environmental and occupational regulations it represents a very real and practical difference from the engineering point-of view. !DRAFT -------- 2 August 6, 1997, 11:40 AM BP-00017788 11:43 89/36/97 12:29:39 Via Fdx -> 216 SBG 8314 D. StROtmE~ 4. Have major uncertainties and research needs for the nex-t ntecation of risk assessment beern identified? The document concludes with recommendations for further investigation that are WldUillY focused on age subgroups [as addressed in section D.4]. Current understanding ofbernz..ene's metabolism and preliminary findings of genetic heterogeneity su,ggest that future research related to subpopulations with particular sensitivity to benzene toJcicity would be more productively dirtxted at genetic polyruorphisms in the enzymes involved in benzene's metabolism. lB. Degree of Success as Implementation-of EIPA's 1996 Prooosed Cancer Risk As..c;essment Guidelines (r.a. g.l.) [The substantive issue of this section will be touched upon in A.3.1 and ID.2 or could ~ tucked in elsewhere. Do we want to keep this issue as a separate section?] 1. ln spite of the emphasis in the new r.a. g.L on integrating multiple rou~ of-exposure, no attention is paid to dermal absorption as a significant source of benzene exposure 2. Both extrapolating from dose-response model (although only linear) and margilll of exposure (MOE) approaches were taken, but no resolution firmlY stated. C. Other Teclmical Issues L Question assertion of low-dose linearity for iormation of DNA adducts (Reddy a~hment). prorein contamination in Turteltaub's work? surprising uniformity of levels in various tissues in Ma.ttullo's work relative sensitivity of lllethods - suggestion-at inconsistency in range of overlap 2. The continuing credence EPA gives to Cronkite's animal model for the human endpoint of concem is outmoded. 3. Exposure estimates applied in MOE approach could be replaced with more cu.r:rent and up-rodare estimates. [I!nfonnation available from Will Ollison.] D. Process lissues 1. EPA docwnent citessmdies that are not publicly available in sufficiently detailed form to eennit independent review. DJRAFT 3 August 6. 1997, 11:40 AM BP-00017789 11:44 SB/5~/97 12:29:24 Via Fax -> 21b 58~ 831~ D. S~RUTffiiR Turteliaub abstract Zhang submitted article 2. fu many places, EPA's conclusions are not adequately developed. Allusions to ..other possible pathways.. must be specified and accompanied by supporting citations. permitting evslUJation of the merits of the assertion. Adhering to a default option on the basis of the existence of some remotely possible alternative puts the impossible onus upon the regulated community to prove the negative that nothing else is possible e;,.cept. in the case of benzene, the non-linear mechanisms that have been demonstrated. [ties back to section Bon r.a. g.L] 3. EPA's currem draft represents a good first step toward a revise<i cpffor benzene, but (as noted in these comments) it could be substantially improved. UnfortuEJately, the ''f"mes" will need to go considerably beyond simple editorial changes 8llld insertions. and potentially could have a non-trivial impact-on the benzene cpf ultimately settled upon. Therefore. AP][ would encourage another iteration of revision with the EJRP. The external review p.rtx>eSS fo.r1l:hi.s document should be in aC(;ord with the FACA. process. It would be highly desir.!Lble for this mERP to follow the example of CASAC on NAAQS documents, where the experts the oversight group stay actively involved in the review and revision process untillhey grant closure. 4. The appended assertion that children and tiDe elderly are J.i.keJy sensitive sub--population for benzene toxicity is a scientifically unsupported political st&nre. JE. Editorial Points (Attachment C) Straighten out reference& to paniculmr update of the Pliofilm cohort being referred ~ the basis of a given analysis instead of using the sloppy nomenclature ''Rinsky 1987," which actually only refers to the Pliof"tlm cohort as updated through 1981, which w.as not the basis for EPA's 19S5 deriation -Of its current cpf and which is not the most complete data set available on tllis cohort now available. F. Conclusion Attachments A Preston on low-dose nonlinearity for various genotoxicity endpoint B. Reddy on specificity and-relative sensitivity of techniques of detectin,g DNA addnets C. Editor-ial Points - DRAFT 4 August 6, 1997, ll:)AM TOTAL P.06 BP-00017790