Document K6R5Eqvym6pgBJ4BXGM33LrK0

AR226~ 179 - FINAL REPORT Epidemiology, 220-3W-05 Medical Department 3M Company St. Paul, MN 55144 Date: August 14, 2003 Title: Descriptive Analysis of Serum Perfluorooctanoate (PFOA) and Perfluorooctanesulfonate (PFOS) Concentrations of Antwerp Employee Participants from the 2003 Fluorochemical Medical Surveillance Program Protocol Number: N/A IRB Approval: N/A Principal Investigator: Co-investigators: Study Director: Geary W. Olsen' `Marie-Nolle Schmickler' Sabine Pays JKeifmfbreeyrlHy.AM.aYnoduenlg' Jeffrey H. Mandel! 8 =# ox 3 ~ooa Loos & 1. Corporate Occupational Medicine, Medical Department, 3M Company, Mail Stop 220-3W-05St,. Paul, MN 55144. 2. Medical Department, 3M Antwerp manufacturing site, Antwerp, Belgium. Page2 ABSTRACT Thirty Antwerp employees participated in the 2003 fluorochemical medical surveillance program. The serum PFOA concentrations ranged from 0.92 to 5.69 ppm. with amean (arithmetic) of 2.63 ppm (95% C1 2.17 - 3.09) and a geometric mean of 2.35 ppm (95% CI 1.96 - 2.82). Two participants' serum PFOA concentrations (5.40 and 5.69 ppm) exceeded the 3M Biologic Limit Value (BLV) of ppm for PFOA. All thirty employees participated in the BLV program in 2001. These employees' 2001 PFOA values rangedfrom 0.88to 4.44ppmwith a mean of2.31 ppm (95% C1 1.94- 2.69) anad `geometric mean of 2.09 ppm (95% CI 1.76 - 2.49). Themean paired difference in PFOA values (2003 minus 2001 concentration) was +0.32 ppm (95% C1 0.11 - 0.52). The 30 employees' 2001 PFOS values ranged from 0.20 to 2.22 ppm with a mean of 0.87 ppm. (95% C10.64 - 1.09) and a geometric mean of 0.70 ppm (95% C1 0.54 - 0.90). The mean paired difference in PFOS values between 2003 and 2001 was -0.37 ppm [95% CI (0.51) - (-0.23)). Twenty of the 30 employees also participated in the 2000 fluorochemical medical surveillance program. Their mean PFOA difference (2003 minus 2000 concentration) was +0.65 ppm (95% C1 0.16 - 1.13). Their mean PFOS difference (2003 minus 2000 concentration) was -0.31 ppm [95% C1 (-0.48-) (-0.14)]. A cautious interpretation is warranted of any trend in serum concentrations because potential occupational exposure information was not considered in these analyses. Pages INTRODUCTION `The 3M Company offers periodic fluorochemical medical surveillance examinations for its production employees. At 3M's Antwerp (Belgium) manufacturing plant, serum perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS) have been measured in employees' serum since 1995 (Olsen etal. 1999; 2003a). As aconsequence of3M's phase-outofthe manufactureofperfluorooctany-related materials, the number ofemployees engaged in the production of perfluorooctanoic acid (salts) and perfluorooctanesulfony! fluoride (POSF)-related materialshasdeclined since 2000. Beginning in 2003 3M Corporate Occupational Medicine and Industrial Hygiene adopted exposure triggers for inclusion in the fluorochemical medical surveillance program. `These exposure triggers were 0.05 mg/ir? (8 hour-Time Weighted Average (TWA) for PFOS and 0.005 mg/n? 8-hr Time Weighted Average (TWA) for PFOA. The latter is one-halfthe Threshold Limit Value (TLV) for PFOA established by the American Conference of Government Industrial Hygienists (ACGIH). In addition, 3M has a Biologic Limit Value (BLV)of 5 partspermillion (ppm, pg/ml)of PFOA in serum (3M `Company 2001). The purposeofthis report istodescribe the serum PFOA and PFOS. concentrations obtained in the 2003 fluorochemical medical surveillance program and `compare these data to serum concentrations from the BLY program offered in 2001 and the previous fluorochemical medical surveillance program conducted in 2000. METHODS Antwerp employees who were considered to have occupational exposures that likely met the exposure trigger for PFOA (described above) were encouragetdo Pages participate in the 2003 fluorochemical medical surveillance program. Other fluorochemical production-related employees,ifso desired, could also participate. As part of this program, a blood sample was collected from each participating employee for the determinationofPFOA and PFOS. The two fluorochemicals were determined by high performance liquid chromatography electrospray tandem mass spectrometry methods (Hansen ct al. 2001). Analyses were conducted at Tandem Labs (Salt Lake City, UT) using validated methods that have been detailed elsewhere (Tandem Labs 1999; 20013; 2001b). Fluorochemical concentrations are reported as partspermillion (ppm = pg/mL), RESULTS Presented in Figure 1 is the distributionofthe 30 employees' serum PFOA concentrations. The serum PFOA concentrations ranged from 0.92 to 5.69 ppm with a `mean (arithmetic) of 2.63 ppm (95% CI 2.17 - 3.09)and ageometric mean of 2.35 ppm (95% CI 1.96 - 2.82). Two participants' serum PFOA concentrations (5.40 and 5.69 ppm) exceeded the 3M Biologic Limit Valueof 5 ppm for PFOA. Also presented in Figure 1 is the serum PFOS concentration distribution for the 30 employee participants. Serum PFOS concentrations ranged from 0.19 to 1.35 ppm with a mean of 0.50 ppm (95% CI 0.39 0.60) and a geometric meanof 0.43 ppm (95% C1 0.35 - 0.52). All thirty employee participants in 2003 also participated in the BL program in 2001. These employees' 2001 PFOA values ranged from 0.88 0 4.44 ppm with a mean 02.31ppm (95% CI 1.94 -2.69)and ageometric meanof2.09ppm (95% CI 1.76 2.49). Twenty-three (77%)ofthese 30 employees had 2003 serum PFOA concentrations Pages that were higher than those measured in 2001 (Figure 2). The mean paired difference in PFOA values (2003 minus 2001 concentration) was +0.32 ppm (95% CI 0.11 - 0.52). `The 30 employees' 2001 PFOS values ranged from 0.20 to 2.22 ppm with ameanof 0.87 ppm (95% C1 0.64 - 1.09) anad geometric meanof0.70 ppm (95% C1 0.54 - 0.90). All but oneofthese 30 employees' 2003 serum PFOS concentrations were lower than those: `measured in 2001 (Figure 3). The mean paired difference in PFOS values between 2003 and 2001 was -0.37 ppm [95% CI (-0.51) - (-0.23)] Twenty (67%)ofthe 30 employee participants in 2003 also participated in the 2000 Antwerp fluorochemical medical surveillance program. These 20 employees had a `mean serum PFOA concentration in 2000 of 2.12 ppm (95% CI 1.50 - 2.74). The range was from 0.10 - 5.31 ppm. Sixteen of these 20 employees (80%) had higher serum PFOA concentrations in 2003 than 2000 (Figure 4). The rangeofdifferences (2003 minus 2000 concentration) was -0.99 ppm to +3.26 ppm. The mean PFOA difference (2003 minus 2000 concentration) was +0.65 ppm (95% CI 0.16 - 1.13). The mean PFOS serum concentration in 2000 was 0.79 ppm (95% C1 0.52 - 1.05) with a range from 0.21 102.28 ppm. Of these 20 employees who participated in both the 2003 and 2000 fluorochemical medical surveillance programs, 19 (95%) had a decline in their serum PFOS concentrations in 2003 compared to their 2000 assessment (Figure 5). The range of differences in PFOS (2003 minus 2000 concentration) was 1.25 ppm to +0.06 ppm. `The mean PFOS difference (2003 minus 2000 concentration) was -0.31 ppm [95% CI (0.48) - (0.14). Pages DISCUSSION `The 30 employee participants' mean serum PFOA concentration in 2003 (2.63 ppm) was slightly higher than that measured in 2001 (2.31 ppm) and. higher yet than their mean PFOA value measured in 2000 (2.12 ppm). The 2000mean PFOA serum concentration was 1.06 ppm for the 206 male employees who participated in the fluorochemical medical surveillance program at that time (Olsen ct al. 2003). This increase in PFOA concentration measured in 2003 is likely due to several factors including: 1) ongoing potentialforoccupational exposure from production activities of PFOA (salts) between 2000 and mid 2002; 2) potential nom-production related exposures in the workplace into 2003; 3) the long serum elimination halflife of PFOA; and 4) restrictionofthe 2003 fluorochemical medical surveillance program to those employees with the highest potential for occupational exposure to PFOA, similar to the 2001 BLV program but unlike the 2000 fluorochemical medical surveillance program which included many Antwerp employees with non-production exposure to PFOA, (.g., administrative and research workers). Also, it is important to acknowledge that the validation methods for the analysis ofPFOA and PFOS were developed with acceptable criteria of 26 and +20 percent, respectively (Tandem Labs 1999; 2001a; 2001b). Therefore, some upward, or downward, individual employee values over time may be comparable as they are within the errorofthe validated analytical methods. In 2003, two participants had serum PFOA concentrations that were just above the BLV of5 ppm (5.40 and 5.69 ppm). The purpose for the PFOA BLV program is explained, in detail, elsewhere (3M Company 2001). Basically, 3M established the BLV (5 ppm serum PFOA) to represent the best estimate ofa levelofa chemical substance or Page its metabolite in a biological fluid thatifpresent, even on a chronic basis, would not be expected to pose, or correlate with, a significant riskofadverse health effectstothe. worker(s). By establishing this BLV at 5 ppm for PFOA, the 3M exposure guideline committee did not intend to imply that serum levels of PFOA greater than the BLY immediately impose a significant risk of adverse health effects. However, ifemployees serum PFOA level exceed the BLY, it should be understood that corrective actions (which may include temporary removal from the immediate work area) may need to be `applied, on a case-by-case basis, at the directionof 3M Corporate Occupational Medicine. Implementationofthe BLV program by 3M Corporate Occupational Medicine did not allow for employees with serum PFOA concentrations > 10 ppm to be directly engaged in the production of PFOA (Olsen et al. 2003c). 3M Antwerp, Cottage Grove (Minnesota) and Decatur (Alabama) employees' serum PFOA concentrations have not been associated with adverse measurements in hepatic clinical chemistry enzymes, cholesterol, lipoproteins, plasma cholecystokinin, reproductive hormones or thyroid hormones (Gilliland and Mandel 1996; Olsenetal. 1998; 2000; 2003; 2003b). `The 30 participants' average serum PFOS concentration (0.49 ppm) in 2003 was lower than the mean previously reported in 1995 (1.93 ppm), 1997 (1.48 ppm) and 2000 (0.96 ppm) at the Antwerp manufacturing sit for the fluorochemical medical surveillance program (Olsenetal. 1999; 2003). These previous medical surveillance years involved considerably more male employee participants (88, 65 and 206, respectively) than in 2003. We suspect that the primary reason for the reduction in mean serum PFOS concentrations is the resultofthe selection process that occurred in the 2003 fluorochemical medical surveillance program. The 2003 fluorochemical medical Pages surveillance program concentrated on employees with potential exposures to PFOA. By March 2003, perfluorooctanesulfony! fluoride-related production had essentially ceased `and thus the greatest proportion of employee participants would have been those employees engaged in PFOA-related production activities. As observed at the Cottage Grove (MN) plant, the workplace production area is a primary factor in whether employees' PFOA or PFOS serum concentrations may predominate (Olsenet a. 2003b). Antwerp and Decatur employees' serum PFOS concentrations have notbeenassociated with adverse measurements in hepatic clinical chemistry enzymes, cholesterol, lipoproteins or thyroid hormones (Olsen etal. 1999; 2003a). There are several limitations to the analyses presented in this report. The results arise from a small subset of employees who voluntarily participated in the 2003 Antwerp fluorochemical medical surveillance program. The number of employees in the 2003 fluorochemical medical surveillance program is almost 90 percent less than those who participated (maleand female) in the 2000 fluorochemical medical surveillance program (n=255). Atitionof employees due to production cessation ofPFOA and POSFrelated materials occurred between 2000 and 2003. Unlike the 2000 Antwerp. fluorochemical medical surveillance program in which employee interview workplace data were collected and analyzed in relation to serum PFOA and PFOS concentrations (Olsen etal. 2001), such data were not collected in 2003. Therefore, it was not possible. 0 associate serum fluorochemical concentrations to potential workplace exposures in the 2003 time period. Knowledge regarding the serum elimination half-life remains sparse but is suggestive ofa long serum elimination rate (i.., years) (Burris et al. 2002). Asa consequence of these limitations, there needs to be cautious interpretationofany trend Pages analysis of the serum fluorochemical concentrationsofthe 30 Antwerp employees who participated in the 2003 fluorochemical medical surveillance program. Page 10 REFERENCES hBaulrfr-ilsivJeMso, fLsuenvdebrearlgfJlKuo,rOolchseemnicGa,lsS.imSpt.soPnauCl,(MMaNn)d:el3MJ. C2o0m0p2a.nyD.eteUr.mSi.nEantviiornoonfmseenrtuaml Protection Agency docket AR-226-1086. GliilplopirloatnedinFsD,,aMnadncdheollesJSt.ero1l9:a96.stuSdeyruofm opcecrufplautoiroonoacltlaynoeixcpaocsieddamnedn.hepAatmicJeInnzdymMeesd, 29:560-568, HquaannsteitnatKiJv,e CchlaermacetnerLiAz,atiEolnleoffsoonrgMaEn,icJfolhunosroonchHeOm.ica2l0s0i1n.biCoolmopgoiucanld-msapterciciefsi.c,Environ Sci Technol 35:766-770. eOplisdeenmGioWl,ogGiiclliinlvaensdtiFgDat,ioBnurolferwepMroMd,ucBtuirvreihsoJrMm,onMeasndienlmeJSn, wMiatnhdoeclcuJpHa.ti1o9n9a8l. An exposure to perfluorooctanoic acid. J Occup Environ Med 40:614-621. Olsen GW, and hepatic aBunrdrilsipJidM,clMinaincadleclhJeHm,isZtorbyetlesLtRs.in1f9l9u9o.roScehremuimcaplerpfrloudourcotoicotnaneempslulofyoeneast.e J Occup Environ Med 41:799-806. hOelpsaetnicGeWn,zyBmuersr,iscJhoMl,esBtuerrollewanMdMl,ipoMparnodteeilnsJHin. a2m0m00o.niPulmaspmearfclhuoolreocoycsttaonkoiantien and production workers. Drug Chem Toxicol 23:603-620. OJKl,seMnaGndWe,l SJcHh.mic2k00l1e.r MDNes,crTiipetrievnessJuMm,mLaorgyaonfPsWe,ruBmurfrliusorJoMc,heBmuirclalewleMveMl,s aLmuonndbgerg eprmopglroayme.e pSta.rtPiacuilpa(ntMsoNf)t:h3eM yCeoamrp2a0n0y0.AnUt.Sw.erEpnvfilruoonromcehnetmailcaPlromteedciticoanlAsgurevnecilyladnocceket AR-226-1030a020b. OwlosreknerGsWe,ruBmurpreirsflJuMo,roBoucrtlanecwsuMlMfo,naMtean(dPeFlOSJH).an2d00pe3r.fluEopriodoecmtiaonlooagtiec(aPsFsOeAs)sment of concentrations 270. and medical surveillance examinations. J Occup Environ Med 45:260- Olsen GW, function in Butenhoff relation to JL, Mandel JH. an occupational 2003b. biologic Assessment of limit value for lipid, hepatic and thyroid perfluorooctanoate. St. Paul (MN):3M Company. U.S. Environmental Protection Agency docket AR-226. PTaFnOdSeAm,LPaFbsO.SA1A9,99N.-AMscsFaOySvEal-iOdaHt,ioNn-rEetpoFrOt.SEQ,uaFnOtiAtAa,tiavneddePtFeHrmSiniantihounmoafnPsFeOrSu,m by CLiCt/y,MSUT/.M,SU..SS.tEundvyiNroo.nmNeWntBaSl98Pr-o0t8ec2t,ioRneApogretncNoy.dNocWkBeRt9A9R--020265-.12N08W.T Inc., Salt Lake Pagel `PTFaHnSd,emM5La5b6s.an2d00M15a7. 0QuinanhtuitmaatnivseedreutmerbmyinLaCti/oMnSo/fMPSF.OSA,ssPaFyOrSeAva,liPdFatOiSonAAre,poPrtO.AA, SU.tSu.dyEnNvoi.rNonWmBeSn0t0al-0P4ro0t,ecRteiponorAtgNeon.cyNWdBoRck0e0t-A1R0-82.26-N1W20T9.Inc, Salt Lake City, UT., `PTFaHnSd,emM5La5b6s.an2d00M15b.70QuinanhtuitmaatnivseedreutmerbmyinLaCt/ioMnS/ofMSP.FOSA,ssPaFyOrSeAva,liPdFatOiSonAAa,ddPeOnAduAm, rUeTpo,rt.USStuEdnyviNroo.nmNeWnBtSal00P-ro0t4e0ct,ioRnepAogretnNcoy.dNocWkBeRt0A0R--122262-.12N1W0,T Inc. Salt Lake City, f3oMr PCeormflpuaonryo.oct2a0n0o1a.teD[oCcFu3m(eCFn2t)aitCiOozn)o.faSnt. OPcacuulpa(tMiNo)na:l3BMioCloomgpicaanly.LimUi.tS.Value (BLV) Environmental Protection Agency docket AR-226. Pager2 BE -- eyRe on " weron _ i apros rime Page13 Fonz ras tionof3SroPnOBNOCASre ai35n0 Comme 0 3Eg s lrosronsoon Ew eroimen Fiotne.rCoemppans0P3a0rsa San o505rCoSenBcor20r030s2o0t1s Pageld pros 2003) ie HES opros oon) w T2945 8 8 00 RIS Ersrie te da2222 228 27 29% gets E ot T, " Year 2003 ; OYear 2000 i. Fuge 16 eA I A----S -- , jo= OYveeaarr 2200003 ers