Document yb9Zg4BZnxXKGa21QmLdanwNE

TOMCk L OG t ASP ATPUM' 4?,4I1-J!J Attachment IV SHORT COMMUNICATIONS Carcinogenic Risk Assessment: Ethylene Dibromide Ca'cmofirnic Rivk Asttviiicm Eihvicnt Dihrorniclc R xxmx C Pink C N Oti, M G *'0 CruKiNf. T J {1970|, Toitrol Jppl Pharinccvl 47. 411 414 t.lhvlrnc iiihri'midc nm hccn 'honri in icm.Ii in jn increased incidence of faxtrie nimi'n in mo foliov*in inmhnnun ji -in mp kf d,i\ 7;icL'S Er\ironnirr.u! I'roiccnor, Afcncx 1tI'A i h-n uu d .1 one nn c.nono feme model m;h p.nameicrx derived f'om ihn how.-iv in ini' in r-urroic thi ink of cancel in human' amine Ito-n mh.ilnnnn exposure lo cihvlenc dihrnmidc The I PA I'li--nicd .in ulmon Ilk) .. lifcimic incidence of cancel in he expected in *nikeis exposed in IU ppm of eihiiene dihinmidc ai enrus fum.if atmn cenie's Thix commurncition etpofi* a icxi of the inhdm nf ih.n p-.-jicnor hai-ed nn n compvivor nf ihc incidence of cancel predicied hi the mcihnd used hi the I.P \ iih thn: oh't'icJ m a fioup nf l>h uorkeis emplnied in iht produenon nf ethilcne 0 "I'midc The one Im model a` u't] l! hi the fcPA picdiciej a tt'ial nr S,` ii.'inoo ahmc ikr ne-mnl taekfour 1 iniidr'ce ir ihr prnup nf en.plnief A im.'.l of rip hi lum n run been OCkcucd Tncicloit. U'C r` thn hir'd I 10 pndict carcmofcnic lesponse in huninr' nppc.i'v io iC'uli in h;t'h|\ cx.ieecfntcv' me cmuo if E;h\ lent- dihromide (CDFu is a chci meal used extenso tlx ax a gasoline audios c and io ; lesser extent as a grain ami citrus fumigant. It has been shovxn to result in an incrc.TuI incidence of castrie lunion in rats ixthen administered ht gaxage (Olson ci ai. |<J' Poxxerx ci i 1y 7 s) Ch> Care nogen Afscssmcn; Group (CAOl of the l S Environmental Protection Agencx has reccnih used data from this carcinogenic bidussax in rats to estimate the risk of increased cancer in humar'. arts,up from exposure to EDR The CAG has estimated an almost lOO'S. lifetime incidence of cancer to oe expected humans expovetl tti a concentration of t> J rpm (3.1 nk nt'lof E.DR !me Jo scars (Pesticide anu Toxic Chemical News. HG7) The reliahihtx oi iht' alarming prediction is depenoent upon the validity of the carcinogenesis model as 'sell as its extrapolation from rats to a human population. To test these hxpotnccs. we raxc compared the' incidence of cancer predicted by the CAG carcinogenesis model with that obserxed in a group of workers employed in the production of EDO. METHODS Tr.e model used bx the CAG to jjbtain the risk estimate quoted aooxc is a one hit model in xx Inch the probabiht) (p) of|t umor formation is po cn by P(x.i) = I - Cxp(-/?.\7'). (II The parameter is the exposure to EDB expressed as milligrams per kilograms per dnx i\craped oxer the entire lifetime. The time span for xxhich the risk is estimated is represented bx j. expressed as a fraction of the total lifetime (I1` xxccks for rats. 70 ill t/ij i i. *\ -v h.vj it ; r*i r. ( op i f : t 1^'* hi Aci.`fn hf f-x All nfhts i' u`pn>i,l,:i''f i* (xrA f ji: SHORT ff' MMCMCXT toss yea-s for nun I The parameters /? and are estimated from the experimental data. /; reflects ihc potency of the carcinogenic agent (with dimensions of reciprocal dosck while ; is a dimensionless parameter which reflects the observed time to tumor formation \\ hen the risk is estimated for an entire lifetime. I = I and Eq. (11 reduces to p[x) I -- exp(--p.\Y (2) The pafametcr estimates used were ihb-sc derived by the CAG from the agcspceific incidence of tumor formation in male-fats presented with EDB by gavage in corn oil at a do>e level of JO mg kg'day. The res e values were/? = 31.75 kg-day'mg and ; 6.V5. The parameter /? includes an idj ustment to human dosage equivalents on a mill cram per surface area basis. Tne epidemiologic studs which provtqed the basis for this comparison was reported in detail by Ot; and Scharnwcber ( 197*1, The study included a cohort of 156 employee' trom two EDB production facilities. on: located in Michigan and another m lexas Fift) seven of the emplovccs' were re ^ictor and Still operators from the Michigan facuuv. which operated continuously from the mid 1920s until 1976 Industrial hygiene surveys conducted in 1 y50 showed ai'borne concentrations ol EDB from I 0 to 10.6 ppm in the breath ng zone of still operators and m IV51 fiom 1 to el ppni between stills. Time-weighted average (TWAI concentrations of EDB in the breathing zone of reactor and still operafors tr. 197 I and 19*2 ranged from 2.9 to J.O ppm. The Vd employees from the Tevas facility were comprised of reactor and Mill operators, foremen, and lead burners. N o industrial hygiene data were available from tnts location. EDB was the only organic bromide product manufactured at the Texas facility. while employees at the Michig an location were exposed to Other products (including carbon tetrachloride and chloir/jforml as well as EDB. Tne one-ht; carcinogenic model wa applied to the two groups ot employees as Kiliows: Tne duration of exposure to E DB at both locations w;o determined ov the work history records of each employee. Based on the tndjst'tal hygiene surveys ai the Michigan location, the Tiva concentrat ion of EDB was conservatively assumed io be o.O ppm. (25 mg m:). To test the effect of a differer.t exposure concentration on the precieted incidence of tumor formation. Another analysts was conducicd at an assumed TW A of 0.9 ppm (6.9 mg nr'). All compijt aborts were core in a manner identical to that used by the CAG. Thus, the daily dose of EDB in mill.grams per kilogram per day was calculated by assuming the individual boJdy weight' were "0 kg. the inspiratory volume for an average work da> v\as 14 m\ and all inhaled EDB was absorbed. All calculations were made or, the basis of 250 working days per year Age was taken as the age either ai me dai; of death or at termination the study ir. 1^76 (with a maximum of 70 yeirsi. L'sing the CAG parameter esti rfiatcs from the carcinogenic bioassay in rats ne'tioned previously, the probability of tumor formation for each individual was then calculated as p s l - exp i--(3 l.75)(daiiy doseii250 365) iy ears exposed age it age 7('T'`I. The individual probabilities were then accumulated to yield the total probabilitv for tumor torm.v.ion within each group of employees. i. ."si-;f aiii .:*.rieai .. . " .C. - I |iR i..w,- - -vn .so. A.' n O" pr.i" n.iiicalu:-* ih.n .if-tuoI .-- .v. .-r i-.-.:-'.P .\'o"i'>,v > - ,i, . -aiivH ' *0:i i.'../. i"*4 > V I CMA 049491 V SHOK1 C OV\'l Mr Vi If IS s j; .* ' KLSLLrs A\D WSCI. SMO's Tjblc 1 shows [lie number of malignant neoplasms observed tup to J.inuarv I'i'M \nrbm e.ich croup of cmplovces and the number predicted based on probabilities Cjleui.iieJ with ihe one hu model As reported b\ On ana Schamuebtr : Iw ). no increase in rhe incidence of malignancies was noied m Marker* a; me fesas Ideation In ihe Nlichican production umi (here have been five dentns due m malignant neoplasms vs expected based on a comparison with ihe United Stairs Mime male moriabiv rais's The pt'M O'Crprcdiction of ihe incidence of maherant ncop'asms bv the one lut mode' is apparent at cither of the assumed TWA concentrations of b.DH. highl malignin' 7 ASLL I ( >I'S| R\ I I > ISO I'KI OK 1 I O M VLIC.S VS 1 S I M 1*1 Kin IS | IIP I'l/t .| ,1 i | IMS 1\||MH| \ I 7 II l i It 11 ! 'II Numi'e: .'i ma.itn a" i-eopi. sim Lee uior Nu-bi- of emp nice' I'Tt dn U ,* th'ie'iee. * a pra* liu;';';" Mwtvijtan lev.is Tom' 1 s V " vsi'rLnrt of 'leepl.c 'v. ,i r e \ i ",e ne\T >, M.-lcv -s,` nn.,a. .i ru,1..;.,,' xn- m. is l.r s.U.r , r. (. \('i \.' u.. -.. r,K- . t v- \ , ' ! iits I *u. .' .'ll i iepi." . J.e , ' pe" . . ir ..' .. . . - lr.. a a. s i *. '1 p .i `.'i,'1! . , 1 *'r .1 . . .* -f ' . - . - ' .1. ,i 1 " .i: u 'i. I-..; . . .. i.. nci'pia-ms have Peen observed thus tar in both e~t'i.ver c'oi.p-. s embmeu. whereas eitne- PP or pJ neoplasms above the norm.-:, baes.'o.aid me.Oenee vi^r: p'edieuv at Tk\ A concentrations of b.DB of 3.0 or O.s) ppm. rcspcc: \eb.. bv [lie ora "a! model The limited number of tmplovccs m the studv popeLioon mehsPt.s the need tor caution in [he interpretation of these results There o also a need to' ;i commune fol!s'vs-up of tins croup of cnioiov ecv Lp to Jniiwi'v ot iron v-.-c ,i >o;ai << '< deaths in the cob or; of I pb men in both croups I sev en due io maiie na,-.t ne op'a* n;s), . r d the average ace ('f the remaining population 'U' ,*p sears (ranee. 3! to 7m I l.e possibilits umsis that a portion of the remaining popaLuiou nav po'-.'s umliiutul neoplasms. Howe'er, the piob.-.biiaic' of n.mor lorr-.a-ion c.'iieulauo b\ ihe on. hit model arc .uitunesl for the aee of each tp.jis icua! T net .-tore ii n cvuenidv unlikc'iv that the dramaticnllv nigh incidence of cancer cstunaicd I" tins techne-a o a u.ilwnc prediction. The apparent failure of me one hi: model as used bv the C-Ui to prcd.st me m.Kleitieol m,i,ij.nant neoplasms niav be tar resu t : e e1I,is tors Smee' the ' nie ir.Osie1 par.iirteters were derived I'otn d.v.i .uisir.s. er'sUiore'iu. biu.-ssa; ;r it... 1 - lIiihu ki -'.I- net *i.,ii'n;.i!i -le-ul:.- ; j, --e u 'b l i -e . 414 SHORT COMMLMC'ATIOSS ihe extrapolation from an inpmd strain of laboratory rats to a heterogeneous human population with diverse environment mat be one of the basic deficiencies in the model. Furthermore, the excessiv elv high dose let el from w htch the parameters were denv ed. as well as dilTerences in the route of exposure 'intubation vs inhalation) may also contribute to failure of the predictive value of the model EDB is metabolized in the rat bv conjugation with hepatic glutathione (Nachtomi. 1970). a process which is known to be saturable at doses sufficiently high to deplete available glutathione concentrations. Thus, nonlinearities in the dose-re'ponsc relationship at low dose levels may obviate the predictive value of data derived from higher dose levels (Cornfield. 1977; Gehrtng and Blau. 19*71. The foregoing reasons, plus the fact that no provision is made for DNA repair, make suspect the biological plausibility of'.his mooel as used b\ the CAG Whatever the reasons for failure of this particular method in predicting carcinogenic response in humans, it is apparent that a great deal cr caution must be used in the application of such methods in predicting the hazard to be expected from exposure to potential carcinogenic agents REFERENCES Cornfield J I19**l Carcinogenic risk assessment Scarce 198 691-699 Gehring. P J wd Blau G F., (Ia'"i, Mechanisms of carcinogenesis. Dose response. J Em nor Paihol Toxicol 1. 167- I W N xchto'-u E ( 19'0i The metapohsm of ethy lene dibrcmioe in me rat--The enzymic reaction with glutathione in 1mo and in t o o Biochcr Phannoco! 19.2853-2860. Olson. 0> A , H aberxi vxx. R T , W eisbl hgs r. E K,, W vxa J M.. and W eisburger. J H 11975) Induction of stomach cancer in rats and mice oy haloeenated aliphatic fumigants J Sen Cancer Inn 31, 1995-1995 Ott. M G, Horns r B B . vxd Gviiddx. H L M 974 Rv-praiory cancer and occupational exposure to anemcals 4/r/t Emooi Health 29. 250-255 Ott. M G and Scharnvv eber H C (I9"*i The MnucJm Experience of 161 Emplo\cc<- Exposed :o Eihilcnc Dibi omidc'm 1 no Pu/ducnon L mu Corporate Medical Department. The Dow Chemical Companv, Midland. Mich - Pesticide and Toxic Chemical Sens October 12. 19*7. pp, 5-4. Powers. M B,. Voelrer. R w , Page, n P,. A eisblrcer. F_. K. . and Kraybill. H F. 11975i Carcinogenicity of ethylene dibromide (EDB) and 1,2 dibromo-3-chloropropnne (DBCP) after oral administration m rats and mice Toxicol, Appl. Pharmacol. 33. 171-172 TomloIozi Repeat ch Lahmaiom Heahh and Em tiomnemal Reseaich Don Chemical Co. 'shchond. Michitan 4864(9 J. C. Ramsey C. N. Park M G. Ott P. J. Gehring Rcceit ed Fehi nary 2d. 19 75. ucccpnd Aupm: .'. 1976