Document vBLj6ejnX8LRma74xDX7EV8qZ

- 3M ECPoLmOp0aOn)y Page 10f25 FINAL REPORT Epidemiology Medical Department 3M Company 220-3W-05 St. Paul, MN 5144 Date: September 4, 1998 PTietrlfel:uAornooEcptiadneomiicolAocgiidcPIrnovdeuscttigiaotnioWnoorfkePrlsasma Cholecystokinin and Hepatic Function in Study Start Date: September 3, 1997 IRB Approval Date: September 3, 1997 Protocol Number: EPI-0003 IRB Approval Exempt Expedited x `Principal Investigator: Co-investigators: ` Geary W. Olsen, DVM, PhD" Jean Burris, RN, MPH" Michele M. Burlew, MS" JeffreHy.Mandel, MD, MPH" Study Director Jeffrey H. Mandel, MD, MPH" 1. Occupational Medicine, 3M Company, 220-3W-05, St. Paul, MN 55115 03510 ABSTRACT 3M ECPoLm0p0a0n3y Page 20125 Perfluorooctanoic acid (PFOA) is a peroxisome proliferator which increased the incidenceof pancreas acinar cell adenomas in rats. Recent research suggested that these tumors may be the consequence oaf mild but sustained increase in cholecystokinin (CCK) as 2 consequence of hepatic cholestasis. In addition, an epidemiologic investigation had suggested that PFOA may modulate hepatic responses to obesity and alcohol consumption in these production workers. To further assess these hypotheses, we conducted three cross-sectional analyses of the employees" serum PFOA levels and medical surveillance data collected in 1993 (n = 111), 1995 (n= 80) and 1997 (n= 74). Plasma CCK was only measured in 1997. Serum PFOA was measured by mass spectrophotometry methods and plasma CCK was assayed by radioimmunoassay. Mean serum PFOA levels, by year, were: 1993, mean = 5.0 ppm (range 0.0 - 80.0 ppm); 1995, mean 6.8 ppm (range 0.0- 114.1 ppm); and 1997, mean = 6.4 ppm (range 0.1 - 81.3 ppm). CCK values (mean = 28.5 pg/ml, range 8.8-86.7 pg/ml) approximated the assay's reference range (up to 80 pg/ml) for a 12 hour fast. Employees' serum PFOA levels were not positively associated with either clinical hepatic toxicity as measured by various serum liver enzyme ests, chlestass or sevated plasma CCK levels Nor did serum PFOA levels modulate hepatic responses (e.g.,liver enzymes and high density lipoprotein) to obesity and alcohol, respectively. 03511 INTRODUCTION Eines Page 3of25 Perfluorocarbons are structurally analogous to hydrocarbons, except the hydrogens are replaced by fluorine (Bryce, 1964] and may contain other elements such as oxygen, nitrogen and sulfur. Ammonium perfluorooctanoate is a potent synthetic surfactant used in industrial applications which rapidly dissociates in aqueous solution to perfluorooctanoic acid (PFOA, C:F15CO;H). ' In laboratory animals, PFOA and is salts are: 1) absorbed by ingestion, inhalation or dermal; 2) not metabolized; 3) distributed primarily in the plasma and liver of male rats and the liver, plasma and kidneyinfemale rats; and 4) eliminated in the male rat via feces and urine whereas in the female rat there is a greater rate in renal excretion [Griffith and Long, 1980; Ophaug and Singer, 1980; Hanhijarvi et al, 1982; 1987; Just et al, 1985; Kennedy 1985; Kennedy et al, 1986; Yiinen et al, 1990; Vanden Heuvel etal, 1991], In rats, PFOA results in peroxisome proliferation, uncouplingofmitochondrial oxidative `phosphorylation, altered lipid metabolism, hypolipidemia and an increased incidence of fiver, Leydig cell and pancreas acinar cell adenomas [Griffith and Long 1980; Kennedy, 1985; Kennedy et al., 1986; Sibinski 1987; Haughom and Spydevold, 1992; Keller et al, 1992; Cook et al., 1992; 1994]. The inductionofthese tumors most likely occurs via nongenotoxic mechanisms because PFOA is not mutagenic [Griffith and Long, 1980; Biegel etal. 1995]. Causal mechanisms may include the roleof oxidative stress in the liver tumors and increased estradiol levels, via induction ofhepatic aromatase activity, in the developmentof Leydig cell tumors [Cook et al, 1992; 1994; Rao and Reddy, 1996], The pancreas acinar adenomas were hypothesized to be a result ofa mild but sustained 03512 : 3PMagECePodLmof Op20a05n)y increase in cholecystokinin (CCK) levels secondary to hepatic cholestasis [Oboumn et al, 1997). CCK is released from the "M" cells in the duodenal `mucosa in response to the presence of food, binds to receptors on the pancreas acinar cells and subsequently stimulates the releaseof pancreatic enzymes into the duodenum [Pandol, 1998]. Itis controled by a negative feedback cycle involving monitor peptide and trypsin. CCK has been shown, in some animal models, to produce pancreatic hypertrophy, hyperplasia and neoplasia [Longnecker, 1986; 1990;1991 Pour et al, 1981; 1988). ! Hepatic toxicity, hypolipidemia and abnormal hormone levels have not been observed in PFOA production workers [Ube et al, 1980; Gilliland and Mandel, 1996; Olsen et al,, 1998]. Gilliland and Mandel [1996] did report that PFOA. may negatively `modulate the effect alcohol has on `high density lipoprotein (HDL) levels and exacerbate the effect that obesity has on liver enzyme tests. However, this workforce was not found to be at an increased mortality risk for liver cancer or liver disease [Gilliland and Mandel, 1993). There were 4 pancreatic cancer deaths compared to 2 expected (Standardized Mortality Ratio 1.96, 95% Confidence Interval 0. 53-5.01). Oneofthese four pancreatic cancer deaths had worked in th building where PFOA is produced at this chemical plant "The purpose ofthis epidemiologic investigation was to re-examine the workforce in this PFOA production plant in order to determine: 1) whether CCK levels are positively associated with serum PFOA levels among production employees; and 2) whether PFOA `may modulate hepatic responses to obesity and alcohol. 03513 METHODS 3M Company PagEePSLaf O20053 PFOA Production PFOA production at this 3M plant began in 1947. PFOA, a white powder, is produced by an electrochemical process [Bryce, 1954]. Production involves a four-stage process: isolating and converting the chemical to a salt slurry, converting the slurry to a , salt cake, dryingthecake, and packaging. The greatest likelihood for exposure to PFOA. occurred in the drying area althoughjob history was not predictive of total serum fraorine levels (a surrogate for serum PFOA) [Gilland and Mandel, 1996). Subject Selection and Data Collection Voluntary medical surveillance examinations were offered biennially (1993, 1995 and 1997) to the fluorochemical production workers. The total number ofsubjects, by year, who participated in these three cross-sectional investigations were: 1993 (n = 111); 1995 (n = 80); and 1997 (n = 74). Eligible voluntary participation rates among these production workers approximated 70 percent. There were 68 subjects in common for 1993 and 1995, 20subjectsin common between 1993 and 1997 (lower number due to employee tumover and re-assignments); and 17 subjects in common for all three years. Surveillance activites included a self-administered questionnaire, measurementofheight, weight and pulmonary function, standard biochemical and urinalysis tests, PFOA determination and several male reproductive hormone assays. The hormone data were collected only in 1993 and 1995 and results have been reported elsewhere [Olsen et al, 1998]. Serum biochemical tests included: alkaline phosphatase, gamma glutamyl 03514 Crea transferase (GGT), serum glutamy] oxaloacetic transaminase (SGOT), serum Page 6of glutamyl 25 pyruvic transaminase (SGPT), total bilirubin, direct bilirubin, cholesterol, low-density lipoproteins (LDL), high-density lipoproteins (HDL), triglycerides, blood urea nitrogen (BUN), creatinine and glucose. Hematology tests included: hematocrit, hemoglobin, red blood cells (RBC), platelets and white blood cells (WBC). In 1997, employees' plasma CCK-33 levels were determined. CCK exists in various forms and lengths although sulfated CCK-3 (ie, a 33 amino acid arrangement) appears to be the predominant form Employees were required to have fasted for 12 hours prior to their venipuncture. One employee serlf-reported that he did not fast and thus he was excluded from the study. His CCK level was 123 pg/dl. This exclusion left 74 employees available for analysis in 1997. Serum chemistries and hematology were evaluated at United Hospitals (St. Paul, Minnesota). Plasma CCK-33 was measured by direct radioimmunoassay by Inter Science Institute (Inglewood, California). Serum PFOA was determined by thermospray (1993 and 1995) and electrospray (1997) high performance liquid chromatography mass spectrometry methods [Johnson et al, 1996; Advanced Bioanalytical Services Inc. 1997]. Data Analysis Simple and stratified analysis, Pearson correlation coefficients, analysisofvariance (ANOVA), and ordinary multivariate regression were used to evaluate linear and nonlinear associations between PFOA and the biochemical parameters with adjustment for potential confounding variables (SAS, 1990). For stratified analyses, employees were divided into four PFOA categories: 0 - <1 ppm, 1 - <10 ppm, 10 - <30 ppm, and 230 ppm in order to determineifan effect existed at the highest serum levels. These categories had been 03515 - 3M Company previously used to examine associations between male reproductive hormones PagEeP7LoOf02O5Y and PFOA among these workers in 1993 and 1995 [Olsen et a, 1998]. For multivariable regression evaluation, PFOA, age, body mass index (BMI), alcohol use, and cigarette use were `examined as both categorical and continuous variables. Alcohol use was analyzed as less than 1 drink per day, > drink per day (with almost all subjects between 1-3 drinks/day), and non-response to the questionnaire item. Linearand nonlinear transformations of PFOA were used to test for associations. In particular, the multivariable models employed by Gilliland and Mandel [1996] were re-examined to determine whether PFOA has a modulating effect on obesity and alcohol consumption in regards to hepatic serum chemistries (SGOT and SGPT) and HDL, respectively. RESULTS Mean serum PFOA levels,by year, were: 1993, mean = 5.0 ppm (SD = 123, range 0.0 - 80.0 ppm); 1995, mean 6.8 ppm (SD = 16.0, range 0.0 114.1 ppm); and 1997, mean = 6.4 ppm (SD = 14.3, range 0.1 - 81.3 ppm). In 1997, the mean CCK value was 28.5 pg/ml (SD = 17.1 pg/ml, range 8.8-86.7 pg/ml). All but two CCK values were within the assay's reference range (up to 80 pg/ml). Thesetwo CCK values (80.5 pg/ml and 86.7 pg/ml) were from employees with 0.6 ppm and 5.6 ppm serum PFOA levels, | respectively. Serum PFOA levels were not consistently correlated with anyof the potential confounding variables, serum chemistries or hematological parameters. The Pearson correlation coefficients (in parentheses) between PFOA and the variables for 1993, 1995 and 1997 respectively, were: age (-22, -.14, 02); alcohol (10, 18, 01), BMI (10, .10, 03516 . 3M Company PagEePgLoOr02G53 ~01), cigarettes (.07, 11, -02), alkaline phosphatase (11, .14, - 07), SGOT (12, -01, 02), SGPT (10, .04, .14), GGT(07, - 01, -.05,), total bilirubin (~02, ~ 14, -.08), direct bilirubin (01, -32, -.04), cholesterol (15, .14, .18), LDL (-.01, -.07, .11), HDL (-11, -19, 03) triglycerides (17, 37, .11), glucose (~08, .04,-.04 ), BUN (~12, -.11, .0)5, creatinine (07, .17, -.01), hematocrit (22, .08, -.10), hemoglobin (22, 1, - 11), REC (09, -01, -.19), platelets (-.10, .04, .11) and WBC (-.01, .06, -01). In 1997, the Pearson correlation coefficient for PFOA and CCKwas -.20 (p = 09). ! `Table I provides the mean, standard deviation and rangeofthe potential confounders, serum chemistries and hematologies by four levels ofPFOA categorization (0-<1, 1-<10, 10-<30, and 230 ppm) for the three years (1993, 1995 and 1997) ofmedical surveillance examinations. The meanof the PFOA ppm categories differed significantly with each other and there were two orders of magnitude difference between the lowest and highest PFOA categories in each year. There were no statistically significant (p <.05) F values for any clinical chemistry test or hematological parameter examined for any of the three surveillance years. It should be noted that the mean CCK values were 50 percent lower among employees with serum PFOA values > 10 ppm. ` Figure 1 is a scatterplotofthe relation between CCK (transformed via natural log) and PFOA. The linear regression equation was: In CCK =3.3 - 0.008 PFOA (p value of PFOA coefficient = 07, r* of model = 04). We did not observe any significant differences in mean serum chemistry values for those employees with high CCK values (e.g. > 40) compared to those with lower CCK values or for those subjects with high PFOA values (e810 ppm) compared to those with lower values. Useof multivariable regression 03517 - IM Company models (data not shown) continued to indicate a weak negative association PagEeP9LoOf02O53 between CCK. and PFOA adjusting for potential confounding variables (e.g., age, body mass index, alcohol, cigarettes and clinical chemistry measures of hepatic function) : Based on the multivariable model used by Gilliland and Mandel [1996], Table I provides the change in HDL levels associated with a 10 ppm increase in serum PFOA levels among moderate drinkers (2 1 drink/day) compared to light drinkers (< 1 ) drink/day). Included in Table IL is the change originally reported by Gilliland and Mandel [1998] in this workforce with their 1990 surveillance data. It should be noted, however, that their 1990 model was based on total serum organic fluorine measurements rather than serum PFOA levels. Unlike 1990, there was not a substantial modulation in HDL levels with increased PFOA serum levels among moderate drinkers. Likewise, Table: III presents the resultsofmultivariable analyses, including those originally reported using the 1990 surveillance data [Gilliland and Mandel, 1996), regarding the potential modulating effect of PFOA on hepatic responses to obesityinthe three subsequent surveillance years. Whereas SGPT levels increased considerably with a 10 ppm change in total serum organic fluorine whenthe BMI wa>s 30, this association `was not observed in 1993, 1995 or 1997. DISCUSSION `We observed a weak negative association between serum PFOA and plasma CCK. among 74 workers engaged in the production ofammonium perfluorooctanoate. This finding was opposite that hypothesized based on the toxicological findings ofObourn et al [1997] who fed diets to rats containing either 0 or 100 ppmof Wyeth-14,643, a potent 03518 " 3M Company EPLO00) peroxisome proliferator, which causes the same triadoftumors, including pancrPeaagse a1c0i.n0a1r25 cell adenomas, as PFOA. After six months, the mean pancreatic weights ofthe treated rats were 17 percent above control animals (p < 05), mean plasma CCK levels were 44 percent higher (p <.05) and markersof cholestasis (total bile acids, alkaline phosphatase and bilirubin) were also significantly elevated. The clinical pathology data indicative of cholestasis were associated with alterations in bile flow and bile acid output. Obourn et al [1997] had also conducted in viro experimentsof both Wyeth-14,643 and PFOA which arguedagainst other biological pathways known to elevate plasma CCK levels including CCK receptor agonism, trypsin inhibition and increased dietary fat content. Obourn et al (1997) concluded that chronic exposure to Wyeth-14,643 may induce pancreatic adenomas via amild but sustained increase in CCK levels secondary to hepatic cholestasis. `We offer several explanations for the lack ofa positive association between PFOA and CCK in our study. First, the primary setof biochemical and cellular events identified in rodents susceptible to the hepatocarcinogenic effectsofperoxsisome proliferators have not been identified in either liver biopsies from humans exposed to peroxisome proliferators or in in vitro studies with human hepatocytes; however, the peroxisome proliferator-activated receptor (PPAR-<) is expressed at very low levels in the human liver [Oboum etal, 1997; Cattley etal, 1998). Consequently, an expert panel has recently opined that it is unlikely that peroxisome proliferators are carcinogenic to humans under anticipated conditions and levels of exposure, however, their carcinogenic potential cannot be ruled out under extreme conditionsofexposure [Cattleyet al., 1998]. Second, evenif the mechanism existed in humans, the serum measurements in these production `workers may have been t00 low to cause an effect. Third, CCK receptors appear to be 03519 :" 3M Company PageEP1110o0f0235 different between the rat and human. Recent studies indicate, that unlike the pancreas of the rat and dog, the human pancreas has no detectable CCK, receptors and litle to no mRNA for the receptor [Wank et al, 1994]. Human, cynomologus and hesus monkeys lack specific binding sites for the selective CCK ligand *[H]L-364,718 [Gavin et al, 1996, 1997). Because the CCK receptor activity of the rat may be quite dissimilar to the human, 2 cynomologus monkey may be a more appropriate animal model to study the pancreatic pathophysiology consequenceofexposure to PFOA in the human. Fourth, `whether CCK initiates or promotes pancreatic cancer is not 2 new question and the research published, to date, remains controversial [Axelson et al, 1992). Data from more than seventy laboratory animal studies have variably suggested that CCK has positive. trophic effects, inhibitory effects, or no involvement in pancreatic tumor growth [Herrington and Adrian, 1995). CCKhas promoted growth ofhuman pancreatic cancers in cell cultures(Palmer-Smith et al, 1991]. Onthe other hand, fasting plasma concentrations of CCK in unresected pancreatic cancer patients did not differ from healthy controls [Rehfeld et al, 1994]. ~ Fifth, the rat maybe an inappropriate model in the study ofhuman pancreas carcinogenesis. Carcinogens in rats induce acinar cell malignancies whichare rare in the human [Anderson et al, 1996]. Hamster pancreas cancer models are ofductal cell origin which resemble human pancreatic cancer (adenocarcinomasof the ductules) [Pour et al, 1981]. Activation of the c-K-ras gene is frequent in both human and hamster pancreatic cancer but is not found in azaserine-induced pancreatic cancer models in the rat [van Kranen et al, 1991; Caldas and Kem, 1995). Finally, we must ask whether the weak negative association observed in our study represents an entirely different biological relationship than what was originally postulated based on the findings 03520 - by Obourn et al (1997). 3M ECoPmLpOaKn)y Page 120725 We do not believe so because: 1) all CCK values observed in this study were within the assay's reference except for two values (which were not associated with high serum PFOA values); and 2) there was no suggestion of cholestasis `which was considered the underlying reason for the elevated CCK levels in the rat. We were unable to replicate in three separate years the orginal suggestion that PFOA may modulate hepatic responses to obesity and alcohol. Several explanations for the disparate findings exist. First, there may be an association that was not observed by us. In the original report [Gilliland and Mandel, 1990], total serum. organic fluorine was used as a surrogate variable for PFOA exposure because the assay was less expensive and technically easier to perform at the time. The use ofa total serum `organic fluorine may represent other perfluorocarbons, which could be peroxisome proliferators, however, data suggest that PFOA would represent the greatest fractionoftotal serum organic fluorine levels in this employee population [Ubel, 1980]. Another `explanation for the disparate BMI findings is that there may have been measurement error regarding body mass index in the original studyorin our study. We have previously noted the lackofan expected positiveassociationbetween BMI and estradiol in the 1990 data [Olsen etal, 1998). In ronly oneoftheyearswas there the expected [Bums et al, 1997] strong positive correlation between BMI and SGPT values (1990, r= 20, p = 02,1993, r = .16, p=.09; 1995, r=.13, p= 27; 1997, r= 43, p= .0001). Self-reported alcohol data collected in the occupational setting should also be questioned for its reliability as well as validity. To partially address the issueofreliability, we examined the analysesofthe 68 employees who participated both in 1993 and 1995. The data showed good correlation for the confounding factors ofBMI (r = 94, p = 0001), alcohol consumption (c = 67, p = .0001) 03521 |- 3M Company PageEP13Io0f02053 and cigarette smoking (r = 84, 0001). The few employees in common for all three years (n= 17) prevent any conclusions regarding the reliability of self-reported data across all three years. The trend in the correlations was comparable for the 1993 and 1995 analyses (e8., 1995/1997 correlations were BMI: r = 91, p < 0001; alcohol 37, r= 37,p <.15; cigarettes, r= 99,p< 0001). A few additional issues need to be considered in evaluating the results from this study. The cross-sectional design does not allow for a direct analysisof the temporality of an association. Given that the half.fe ofPFOA is estimated to be 18 to 24 months [Ubel etal, 1980}, it is conceivable that there may be some biological accommodation to the effects ofPFOA as suggestedbyBiegel et al [1995]. Also, there were fewer employees analyzed in 1995 and 1997 reducing the statistical powerof the study although the number ofsubjects with serum PFOA measurements > 10 ppm remained comparable. Finally, the issue remains that the lack ofa clinical hepatotoxic effect observed by Gilliland and Mandel [1996] and ourselves does not negate the possibility that PFOA may have a subclinical effect in this production population that has yet to be observed. Results from additional laboratory animal studies may provide further insight. ' In conclusion, our data do not suggest that, at the serum levels measured, PFOA is associated with a mild increase in plasma CCK levels. Neither the original epidemiological findings [Gilliland and Mandel, 1996] or our findings suggest clinical hepatic toxicity at the PFOA levels observed. The fact that we were unable to demonstrate, on three separate occasions, PFOA modulationofhepatic responses to `obesity and alcohol, leads us to believe that this, too, is unlikely at the serum PFOA levels measured in this study. 03522 . REFERENCES IM CEPoLmOp0aOn}y Page 14of25 Anderson KE, Potter JD, Mack TM (1996). Pancreatic cancer. (In) Schottenfeld D, OFxrfaoumrednUinJivFe(resdist)y "PrCeasnsc,eprpE7p2i5d-e7m7i1ology and Prevention (2* edition)" New York. k Axelson J, Thse I, Hakanson R (1992). Scand J Gastroenterol 27:993-998. Pancreatic cancer: the roleof cholecystokinin? pBeirefgleuloLrBo,ocLtiaunoRatCeMo,nHLuerytdtiMgEc,ellCfouoncktiJonC:(1i9n9v5i)t.ro, Eifnfveicvtos,oafnadmemxovniivuomstudies. Toxicol Appl Pharmacol 1995;134:18.25. J Bryce H (1964). Industrial and Utilitarian AspectsofFluorine Chemistry. In: Simons J, ed. Fluorine Chemistry. New York:Academic Press, pages 297-492 `Burns CJ, Boswell IM, Olsen GW (1996). Liver enzyme activity and body mass index. J Occ Env Med 38:1248-1252 Caldas C, Kem SE (1995). Pancreatol 18:1-6. K-ras mutation and pancreatic adenocarcinoma. Int J PCaotptpleJyAR,CR,obDienLsuocnaDJE,,ElSccohmwbetezC,B,FeTnungerw-oCordisJp, PW,ahLialkeWB(G1,99M8)a.rsDmoanpDeSr,oxPiassotmoeor TA, proliferating compounds Pharmacol 27:47-60. pose a hepatocarcinogeric hazard to humans? Reg Toxicol bCyooakmmJoC,niMuurmrpaeyrfSlMuo,roForcatamneoaStRe,: HaurptotssMibEle(e1n9d9o2c).rineI-nrdeulcattieodnomfeLcheayndiisgm.cellTaodxiecnoolmAapspl Pharmacol 1992;113:209-217. iCndouocktiJoCn, bHuyrptetrMoExi,soFmreampreolSiRf,erBaiteogrselinLCBrl(:19C9D4)B.RM(eCcDh)anraitss.msToofxiecxotlroaghiespta1t4i:c30t1umor (abstract) Gavin human CE, Martin pancreatic NP, Schlosser MJ (1996). membranes. Toxicologist Absenceof 30:334. specific CCK-A binding sites on oGfapvainncrCeEa,tiMcalchnloosekceysJtAo,kiWnhiint-eA J, Schlosser MJ (1997). receptors. Toxicologist Species 36:1180 differences (abstract). in expression Gilliland FD, Mandel JS production plant. JOM 3(51:999530).-95M4o.rtality among employees ofa perfluorooctanoic acid G03523 IM Company PageEP1L5-o0f00235 Gilliland FD, Mandel JS (1996). Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins and cholesterol: a studyofoccupationally exposed men. Am J Ind Med 129:560-568. Griffith FD, Long JE (1980). Animal toxicity studies with ammonium perfluorooctanoate. Am Ind Hyg Assoc J 41:576-583. Hanhijarvi H, Phaug R, Singer L (1982). The sex-related difference in perfluorooctanoate excretion in the rat. Proc Soc Exp Biol Med;171:51-55. Hanhijarvi H, Ylinen M, Kojo A, Kosma VM (1987). Elimination and toxicity of perfluorooctanoic acid during subchronic administration in the Wistar rat. Pharmacol , `Toxicol 61:66-68. Herrington MK, Adrian TE (1995). On the roleofcholecystokinin in pancreatic cancer. Int J Pancreatol 17:121-138. HpearufgluhooomcaBtn,oiScpyadceivdo(lPdFOOA)(1,99p2e)r.fluTohreoomcetcahnaensiuslmphuonndiecralcyiidng(tPhFeOhSyAp)olainpdemcilcofeifbfreicctacoifd. `Biochim Biophys Acta 1128:65-72. Johnson JD, Wolter JT, Colaizy GE, Rethwill PA, Nelson RM (1996). Quantification of paenrdflhuigohropoecrtfaonoramtaencaendlipqeuridflcuhorroomoacttaongersauplhfyo-ntahteerimnohsupmraaynmsaesrsumspuescitnrgoimoent-rpyaiwritehxtraction authomated sample preparation. Company. 3M Environmental Laboartory Report. St. Paul:3M JBuisotchWeWmi,caGlorefgfaesctsK,anHdarzto!nFaUl,heHteeirongeemnaenitnyPo,fSpaelrzoexriMs,omSecphriolmiafsersaetHikon(1i9n8d9u)c.ed by perfluorocarboxylic acids in rat liver. Hepatology 9:570-581. Keller B, Marsman D, Popp J, ThurmanR (1992). uncouple mitochnodrial oxidative phosphorylation. Several nongenotoxic carcinogens Biochim Biophys Acta 1102:237- 244 Kennedy G. Dermal toxicityofammonium perfluorooctanoate. 1985:81:348-355. Toxicol Appl Pharmacol Kennedy G, Hall G, perfluorooctanoate. Brittelli J, Fd Chem Chen H. Toxicol (1986). Inhalation 24:1325-1329. toxicity ofammonium Longnecker DS (1986). VLW (ed) The Exocrine Experimental models of exocrine pancreatic Pancreas: Biology, Pathology and Diseases, tumors. (In) Go New York: Raven, pp 443-458. 03524 . 3M Company PageEP16L.0f-200503 Longnecker DS (1990). Experimental pancreatic cancer: role of species, sex and diet Bull Cancer 77 27-37. Longnecker DS (1991). Hormones and pancreatic cancer. InJt Pancreatol 981-86, Oboum ID, Frame SR, Bell RH, Longnecker DS, Eliott GS, Cook JC (1997). Mechanisms for the pancreatic oncogenic effectsofthe peroxisome proliferator Wyeth 14,643. Toxicol Appl Pharmacol 145:425-436. Olsen GW, Gilliland FD, Burlew MM, Burris JM, Mandel JS, Mandel JH (1998). An epidemiologic investigationofreproductive to perfluorooctanoic acid. J Occ Env Med hormones in men 1998,40:614-622 with occupational exposure , SOopchaEuxgpRB,ioSl MiendLg(1169e38:01r)9.-23.Metabolic handlingofperfluorooctancic acid in rats. Proc `Palmer-Smith J, Krame ST, SolomToEn (1991). CCK stimulates growth ofsix human pancreatic cancer cell lines in serum-free medium. Reg Peptides 22:341-349. Pandol SJ (1998). Fordtran JS (eds) Pancreatic physiology and secretory testing. Gastrointestinal and Liver Diseases, Volume (In) Sleisenger M, 1, Philadelphia:W.B. Saunders Co, pp 771-782. oPfasatmomoor nTiP,umLepeerKfPl,uoPreororcitManAo,ateGi-lilnedsucPeJd (h1e9p8a7)t.omeBgiaolcyheamnidcpaelraonxdismoomrephporlooligfiecraatliosnt.udies Exp Mol Pathol 47:98-109. (P1o9u8r1)P.MC,urRruenngtekRnGow,lBeidrgteDo,fGpianngcerlelatRi,cLcaawrscionnogTe,neNsaigseilnDt,heWhaalmlsctaevre aLn,dSaitlsmraeslievance to thehumandisease Cancer 47:1573-1587. fPoonupranPcMr,eatLiacwcsaornciTn,ogHeenlegsiessoinn tSh,eDhoanmnsetlelrymTo,deSlt.epaCnarKci(n1o9g8e8n)e.siEsff9e:c5t9o7f-6c0h1olecystokinin Rao MS, Reddy JK (1996). Acad Sci 804:573-587. Hepatocarcinogenesis ofperoxisome proliferators. Ann NY Rehfeld JF, van Solinge WW (1994). Adv Cancer Res 63:295-347. The tumor biology ofgastrin and cholecystokinin. SAS Institute, Institute, Inc. Inc. (1990). SAS Users Guide: Statistics. Version 6. Cary, NC: SAS Sibinski LJ (1987). FC-143in rats. St. TPawuol-,yMeaNr oRriakler(diLeatb)ortaotxoirciietsy./carcinogenicity studyoffluorochemical 03525 . 3M Company PageEP1I700000235 Ubel F, Sorenson S, Roach D (1980). Health statusofplant workers exposed to fuorochemicals: A preliminary report. Am Ind Hyg Assoc 41:584-589. Vanden Heuvel J, Kuslikis B, Van Refelghem M, Peterson R (1991). Tissue distribution, metabolism and elimoifpnerfaluotroioctoanonic acid. J Biochem Toxicol 6:83-92. vanKranen HJ, Vermeulen E, SchorenL, Bas J, Woutersen RA, van lersel P, van Kreijl CF, Scherer E (1991). Activationofc-K-ras is frequent in pancreatic carcinomas of Syrian hamsters, but is absent in pancreatic tumorsofrats. Carcinogenesis 12:147-1482. Wank SA,PisegnaJR, deWeerth A (1994). Cholecystokinin receptor family. AnnNY J `Acad Sci 713:49-66. `Yiinen M, Koho A, HanhijarHv,i PeurPa (1990). in the rat after single and subchronic administration. Dispositionofperfluorooctanoic acid Bull Environ Contam Toxicol 44:46- 53 ' C03526 . 3PC1o-n0p0a0n3y Page 16 of 25 "HTeabmlaetLoMleoagbnySSetarnudmarPdFDOeAviaLteivoenlso,fManedaYnea(rSDo)fDaantdaRaCnlgleeotfioPnerfluorocanoic Acid (PFOA), Demographic,; Clinical Chemises oPrmon Men S593DDitRa ane 1995 Daa 1997 Da. Men SD Fae Mem SD Fame PrOA 0-<1 ppm 1-<10 ppm 1>030-<p3p0mppm 048 027 000099 338 217 103892 F6106u.21a63l'' i=e326340981.5.93111=.6.9000-802101.000 031 032 0000.90 303 184 110-820 51F75.u91a16)lie63=39720796,p=31.40023000-0-1121842100 047 026 005092 313 212 1057.66 FS137a12l87e =25411294155 p31=70.510.1-0.02803]11335 01--<t0pppmm 2130-00<p0m ppm 6 982 w2a6 a53s7os a25e4 Fralue =33, p=.02 Axe eao 8ss umsa 4357452 s2sss Fualue =09,p = 45 / a 8os7 eas sonosn mmsst Fua=0t3.i7e8 01--<<110ppppmm 21300-p3p0mppm 2W9o 4235 2W693250 8#53 2186 m26e9x3n002 Fualie=11.p=.33 BM 22:66 4324 2W11943532 9288 4180 2282233236 Fualue = 08.p=.52 22975 3593 2294S68 20569 2300 0280232320 Fua= 1t4,pi=.e26 01--<<110ppppmm 1203-0p3p0mpom ` 01--<<110ppppmm 21300-<p3p0mppm 01--<<10ppopmm 10-30 ppm 230 ppm 007s o0s7 000031ss 0089 0068 00042210 Fualue =29,p =.04 26 71021 ooad s 11050 0o%x Fuatue =13.p = 26 Not done Alcohol 00ss 00s7 00002199 005s 0076 0000211s Fualue=09,p=43 Cimarenes s 5 s60 oowa 56 81552 ooswo Fualue =13.p= 29 cx Notdone 00s7 0160 000032s6 0074 006s ooi0lLss Fuatue =07,p=.58 37 1710s 003a0 F50 a=li1ss7u.p=oe1s8 3B0s 11972 8134886075 157 41 1420 F2a06lue =71225.p=.102.69299 03527 cToanbtilnLeued PFOA em 01--<<1i0ppppmm 21300-3o0m ppm 01--<<iip0ppmpm 1203-00mppm 1100---<<1310p0ppppmpmm 230 pom 01--<<it0ppppmm 1203-0350mpom ` 01--<<t10ppppmm 10-<50ppm 230ppm 01 -<<it0pppmpm 1203-0<p3o0mppm --1993Daa Mean SD Range s 2 o5 3aaa 3 0 83 s7aaom Fal=1u.6,e =.19 03010m ensme 6w 142 wm Fra=l10u.pe= 41 52%3 41 n1e16em0 Fualue =61.1,p =2336s s 8u nasm 655 nsisa: Fua=l0.u4,ep=.75 006%8 00228 00200113300 055 017 000% F06u5al=0u0173e, p=0..5550080 00178 000077 001100003#0 001188 000170 001100003300 Fualue =0.1, p=. --1995Daa Mean SD Range AlkalinePhoshassse B0 as2s ssaees B50523 ssase Fualie =10.p=.39 ger esoa7 e1wa 2 11s n26s Fuahie =06.p= 61 sqot 2u n1 an6 13sn3aa6s F0uali4e =05.p=.s6a6s srr 5 " sBn 7 mars s sonn mnawl Fralue =09,p =.46 "ToulBlinbia 008785 002362 004400123200 06 021 040100 0F64ua=0l11.7u5.ep=0.52020.90 Direst Bilinkin 0022 00005s 001200003300 00119s 0000s6 00110000300 Fale =22.p=.10 31CPo1m-p0a0n0y3 Page 19 of 25 --1997Daa___ Mean SD Range n non 2easm B81%6 eseeo Fualue =09,p =.45 ! 0m002 Ilseaeo 2 Ba u eisdso Fualue =05.p=91 2386 772 nlBe4md1s 7Fualu5e =05,p =m84a Bao0 as lnesm 06 n 0 s iss Fualue =10.p=.42 00789 004&82 00300223400 0 03 030-120 F0.73ua=020l24.up=0.e47060.90 0O0l5 000150 000000002600 000058 00005i 00000000..1100 Fualue =02.p = 89 03528 Tcoanbtlienued PoFeOmA 0-<tppm 11-0<-1300pppmpm 230 ppm 01--<<110ppppmm 21300-p3p0mppm 01--<<1l0ppppmm 21300-p3p0mppm 01--<<110ppppmm 21300-p3p0mppm ` 0-<1 ppm 1102-<<1300ppppmm 230 ppm 0-1<-<t10ppppmm 1203-05p0pmpom 3ECPoIm-p0a0n0y3 Page 20 of 25 Man 1S9D0mFang ass 25 39 1n55e.3w09 3206 a328 u1s2e6s Fralue =0.6, p =.65 aa u1nB nBww 2v00B au4e Fualue =20,p=.11 wBss 3o8w m2aamm BLS37 6e1s8i8s Fale =02,p=.91 0msoon40s8 13268%4s M68120140 740.35664 Falue =0.5,p=.67 154 Bos seans nos FrBoalue = 09,p=.n44w 01090 0012 00761l34 01s00011 oossll Fual=u05e, p=.68 M_e_a_n19S5DDwaRage Cholesterol 022 336 3ns2asm 2245 43s6 1821a2sss Fl=u06.ep=.63 HDL 2an8 uanwe a 3% 89 2a6s4s6 Fralue =09,p = 46 LoL 1B33IR40 2382.10 2Be1046 e10a01n57 Fale =02,p =.92 Triglycerides 7m5o 91 sseasn 2269 114870 714553.5963 Fualue =20,p =.13 BN 1155s4 6a26 3 62 4 6nis Fralue =0.4,p=.75 Creatinine 01s0 0011 00861132 11000 0011 00991112 Fralue =1.4.p=.25 M--a_n19S9D70Fmamse W199oe30 s129s28 2 20 s2 2l0e2nss Fua=1l8,pu=e16 a "4 9on 2 sa S51100 206s0 Fra= 0l6,pu=.e62 Bmees28 2s6o2u8s 1BS322446 7191200666 Fua=1l.5,up=.e22 2196 81s6 4s4a36s0 241% 270 6187813860 Fra=l1.u4,ep=.25 16534 e7xs 6vos6 enxa Fal= 0i9,ep=.44 010 0011 00781132 1100 0012 o008s114l Fua=1l1.pu=.e37 03529 | Table L continued PsoFmOA 01--<<110ppppmm 210.3- 0 p<3p0mppm 0-<lppm 110--<<1300ppppmm. 30 ppm. 01--<<1l0ppppmm. 2130-0p<p30mppm 0-<1 ppm 1-<10ppm 10-<30 ppm. 230 ppm , 01--<<110ppppmm 10-<30 ppm. 230 ppm. 34 Company. PagEPe1-2010o03f 25 Man 99S3DDataRange 88 382 essuass 88 910 n69w7 Fvalue = 02, p =87 4s 2 4081 44s 522 aadse 485 ase Fvalue = 1.6,p =.20 115566 0088 11326117742 116564 0189 115402-116993 Fva=l19u,pe= 14 5103 4663 50 03 4589 4903 444 5405 4860 Fvalue =23,p =.08 20 51 len 241 48 156370 29 85 lesdes F2v3alue =300.5,p =1.8792-285 Mean 99SDData Range Glucose 99 114 T7e5ss 897 1102 s7el6l Fyalue =07,p=.59 `Hematocrit "2 as " "2 2 a04e8 455 ase Falue = 06,p=.63 `Hemoglobin 11419 0088 1137217617 118572 0280 114402-119617 Fvalue =1.0,p = 0.40 RBC 4499 0033 445556s0 48 03 4452 50 05 4556 Fvalue =0.4,p =.78 Platelets me 1603 2260 4468 118603333397 F2v7alue =400.5,p=.17729267 1 Mean 7 SD D Range 2100 3186 e712s55 80 512 77967% Fra= 0l.9.up =e#7 4s 3 es 44s 522 a04a7 "43 ade Fva=0.l5 pu=.e66 11584 30098 11372117800 115520 0069 113492115680 Fva=l0.u4,ep=.79 5511 0054 44137601 48 03 432 48 03 45s1 Fale =1.4,p=.25 24 32 16288 2270 4545 112192337088 F260va=361l.0,pu=2e4152300 03530 . Table continued PFOA oem 0-<lppm 110--<3100ppppmm 230ppm 3Company PagEePI-22000o3f 25 --1Me9a9n 3SDDaaange 65 65 21 19 33150 3817 66 63 20 19 461l4 4487 Fyalue=01p=99 M--e_a1n99S5DDmaRage WBC 60 63 13 18 3eus 41108 63 12 4986 6F5value1=203p=55188)3 Me--a_n19S97DDmRaa_n_ge 63 Ll 7123 4389 42168 68 61 22 09 50109 5072 Evalu=e 1.=0.3p9 12.. MMeeaannssiiggnniiffiiccaannttllyyddiiffffeerreenntt((BBonofenrrfoenirtt-ertsetsot,,npp<i<.0055))tthhaannetahceh1o0-f3t0hpepomthaenrPdF>O3A0ppppmmcPatFeOgAorciaetse.gories. *Study Population byPFOA (ppm)CategoryandYear PFOA Category 01--<<L1p0pppmm 10- <30 ppm 230ppm 993 5 1995 39 1997 29 3 9 2% 10 3 7 4 54 ' 03531 rSere s = LN(CCK) vs. PFOA " p 2 g< . 3= : i" g ' rr ee we 888 . PFOA(opm) : 3M Company EP1-0003 Page 24of25 `Table IL. Change in HDL" from Light Alcohol Drinker (< 1 drink/day) to Moderate Alcohol Drinker (> 1 drink/day) Associated with a 10 ppm Change in Serum PFOA Level Moderate Moderate Drinker with Year Drinker 10 ppm increase in PFOA T9507 355 52 1993 +48 "a4 / 1995 +51 +41 997 +57 +38 "Determined from multivariable model adjusted for age, body mass index and smoking (al four years) and testosterone (1990, 1993 and 1995 only). M* aDnadteal,in 11999960) analyzed toal serum organic fluorine (see Gilliland and ` 03533 3M Company EPL-0003 Page 25of25 `Table IIL. Change in Serum Glutamic Oxaloacetic Transaminase and Serum Glutamic in Serum PFOA Pyruvic Transaminase Associated with a 10 ppm Change BMI (kg/m) 25 30 35 SGOT 1990 1993 24 13 37 [] 9.7 2s ' 1995 03 ol os 1997 21 o1 "19 SGPT 1990" 1993 1995 1997 3.0 25 280 59.0 15 os 08 10 21 53 01 38 "mDaestserinmdienxedanfdrosmmomkuilntgi.variable regression model adjusted for age, body * Datain Mandel, 1990 1996). analyzed total serum organic fluorine level (see Gilliland and ` OC524 3M CEoPmLpOa0nGy Page lof42 PROTOCOL Epidemiology Medical Department 32M20C-o3mWp-a0n5y St. Paul, MN 55144 Date: September 3, 1997 Tide: An Epidemiologic Investigation of Plasma Cholecystokinin and Hepatic Function in , Perfluoroocatanoic Acid Production Workers Study Start Date: September 3, 1997 Estimated Date of Final Report: April 15, 1998 IRB Approval Date: Protocol Number: EPI-0003 IRB Approval Exempt Expedited x Principal Investigator: Co-investigators: ' Geary W. Olsen, DVM, PhD Jean Burris, RN, MPH' Michele M. Burlew, MS* Jeffrey H. Mandel, MD, MPH' Study Director: Jeffrey H. Mandel, MD, MPH' 1. Occupational Medicine, 3M Company, 220-3W-05, St. Paul, MN 55115 03535 3M CEoPmLp0a0n3y Page 20f42 ABSTRACT `Two-year feeding studies in Crl:CD BR (CD) rats at a maximum amount of 300 `ppm perfluorooctanoic acid (PFOA) showed, in addition to liver adenomas and Leydig cell adenomas, an increased incidence of pancreasacinarcell adenomas. However, PFOA was not found to be mutagenic; thus the induction of these tumors likely occurs via nongenotoxic mechanisms. Recent research has suggested tha the pancreas adenomas are a secondary effect from elevated cholecystokinin (CCK) levels due to hepatic cholestasis. Increased plasma CCK levels has produced pancreatic hypertrophy, hyperplasia and neoplasia in some, but not al, animal models. A cynomologus monkey is currently being used in a PFOA feeding study to further assess the relation with CCK because the rat `pancreas CCK receptor may be quite dissimilar to that ofthe human and monkey. `Because the role of CCK in the promotion of pancreatic cancer remains quite: `controversial in laboratory animals, additional insight is warranted into any possible association between occupational exposure to PFOA and plasma CCK levels. Therefore, the purposeofthis cross sectional epidemiological study design is to determine whether there is an association between plasma CCK-33 levels, as measured by tadioimmunoassay, and serum PFOA levels among 3M Cottage Grove fluorochemical production workers. We will also examine hepatic enzymes, bilirubin and lipoproteins in relation to the employees" serum PFOA levels due to the toxicological issue regarding hepatic cholestasis. 03536 INTRODUCTION 3M CEoPmpOan}y Page3 042 Fluorocarbons are compounds of fluorine, carbon and other elements such as oxygen, nitrogen and sulfur. Perfluorocarbons are structurally analogous to hydrocarbons, except the hydrogens are replaced by fluorine (Bryce, 1964]. In general, perfluorocarbons are inert and heat stable; thus they are often used in high temperature applications and make excellent insulators and surfactants. Synthesis has been , accomplished by electrochemical fluorination, direct fluorination, teleomerization, and ' catalytic methods using high valence metals. Although fluoride (inorganic ionic fluoride) was identified in human blood 140 years ago [Nickles, 1856], the presence of fluorine ina free ionic siate as well as a covalently bound organic state was first reported in 1968 [Taves, 1968; 1968b]. Guy [1972] subsequently identified perfluorooctanoic acid (PFOA, CyF;sCO3H) as a major component of the serum organic fluorine fraction. Ammonium perfluorooctanoate, a `potent synthetic surfactant used in industrial applications, rapidly dissociates in aqueous Solution to PFOA. Since Tave's and Guy's observations [Taves, 19682;1968b:Guy 1972], PFOA has been the subject of several toxicologic studies. ` In laboratory animals, PFOA acid or is salt is absorbed by ingestion, inhalation or dermal exposure [Griffith and Long, 1980; Kennedy 1985; Kennedy et al, 1986]. PFOA is not metabolized (Ophaug and Singer, 1980; Ylinen et al., 1990; Vanden Heuvel etal., 1991; Kuslikis etal, 1992). PFOA is distributed primarily in the plasma and liver of male rats and the liver, plasma and kidney in female [Vanden Heuvel et al. 1991]. The major routeofelimination in themalerat is via urine and feces whereas in the female rat there is G03537 3M CEPoLmOp0a0nYy a 10-fold greaterratein renal excretion [Vanden Heuvel et al, 1991; HanhijarviPetaagl.e,of2 1982; 1987). Caswated male rats treated with eswadiol have PFOA urinary excretion ratessimilarto female rats [Ylinen et al, 1990; Vanden Heuvel etal, 1991) Peroxisome proliferators, like PFOA, are a diverse class of chemicals that cause hepatic peroxisome proliferation and enzyme induction, liver hyperplasia and, in some instances, hepatocarcinogenesis in rats and mice [Ikeda et al, 1985; Pastoor et al., 1987; Sibinski 1987; Cook etal., 1992; Biegel etal., 1995; Liu et al., 1995; Lemberger etal, / 1996]. Peroxisome proliferators bind to and activate peroxisome proliferator-activated receptors (PPAR) belonging to the superfamily of nuclear hormone receptors. Upon bindingofa peroxisomal proliferator or other ligands, such as fatty acids, PPAR interacts with RXR, another nuclear hormone receptor activated by 9-CIS retinoic acid. This heterodimer binds to specific hormone recognition elements called peroxisome proliferator response elements (PPRE) in the promoter of target genes, resulting in the coordinated transactivation ofa set of genes in peroxisomal, mitochondrial, microsomal and cytosolic cell compartments involved in lipid homeostasis. Two-year feeding studies in Crl:CD BR (CD) rats at a maximum amountof300 ppm PFOA showed, in addition to liver adenomas and Leydig cell adenomas [Sibinski 1987: Conk. 1994], an increased incidence of pancreas acinar cell adenomas [Cook etal. 1994]. PFOA was not found to be mutagenic [Griffith and Long, 1980]; thus the induction of these tumors most likely occurs via nongenotoxic mechanisms [Biegel et al, 1995). Evidence swongly implicates the role of oxidative stress in liver tumor development for peroxisome proliferators [Rao and Reddy, 1996]. Cook et al [1994] 03538 . 3M CEoPmLpOa0nGy3 showed that the Leydig cell tumorsare likely the result of increased estradiol levPealgsedSuoef2to induction of hepatic aromatase activity. `The pancreas tumors have been hypothesized to be a secondary consequence of PFOA' effect on the rat liver via cholestasis-induced increased plasma cholecystokinin (CCK) concentrations (Obournetal, 1997; 1997b). Specifically, Oboum et al [199721997] examined the possible mechanisms for the pancreatic oncogenetic effects oftwo potent peroxisome proliferators: Wyeth 14,643 and ammonium perflurooctanoate / (C8). Both compounds in virro failed to: 1) bind to the CCK-A receptor in a competition binding assay: and 2) inhibit trypsin in a continuous spectrophometric assay. Rats fed 100 `ppm Wyeth 14,643 for 60 days were found to have no pancreatic weight effects, increases in plasma CCK, acinar cell proliferation or increased fecal fat. However, rats fed 100 ppm Wyeth 14,643 for 3 and 6 months had increased pancreatic weight (6% and 17% above control, respectively), increased CCK plasma levels and acinar cell proliferation. Increased serum concentrations of serum bile acids, alkaline phosphatase and bilirubin suggested choleostasis and this was confirmed by measuring bile flow. Relative bile flow (relative to liver weight) was decreased at6months to 72% of the control group. Oboun etal concluded that Wyeth 14,643 and also infeed for PFOA) causes liver effects, including choleostasis, thatresultin increased plasma CCK levels. Increased CCK levels have been shown in other animal models to produce pancreatic hypertrophy, hyperplasia and neoplasia [Longnecker, 1986; 1990; Pour etal, 1981] Appendix A provides a literature review which offers a more detailed appreciation of the role CCK plays in pancreatic physiology. hypertrophy and neoplasia. Included in Appendix A are reviews of the following topics: 1) an overview of the anatomy and 03539 3M CEoPmL0p0a0n3y Page Gof 2 histology of the pancreas; 2) pancreatic exocrine physiology as it relates primarily to CCK: 3) the epidemiology of pancreatic cancer; 4) animal models of pancreatic cancer; 5) CCK receptors and the pancreas; and 6) the role of CCK and pancreatic cancer. ~ A very brief reviewofthe information provided in Appendix A is summarized below. `The principal inorganic components of exocrine pancreatic secretions are water, Sodium, potassium, chloride and bicarbonate [Pandol, 1993). The principal organic constituents are proteins, primarily digestive enzymes, produced from the acinar cells. The enzymes are secreted into the pancreatic ductules in an inactive form with activation occurring in the intestinal lumen primarily by cleavage with trypsin. Regulationof enzyme secretions is controlled by both neuro- and humoral stimulation of the acinar cells Receptors located on the basolateral surface of the acinar cells have been reported for CCK, acetylcholine, bombesin, substance Pand vasoactive intestinal peptide (VIP) in the guinea pig, rat and mouse. However, CCK-A receptors were not found in human, cynomologus and rhesus monkeys [Gavin et al, 19972; 1997b). The intestinal phase: represents the most important aspect of pancreatic enzyme secretion which is mediated by both enteropancreatic vagovagal reflexes and hormones. Secretin is the major humoral medistr of ductal bicasborate and water secretion whereas CCK is the major humeral mediator of meal-stimulated enzyme secretion. CCK is a potent regulator peptide that also stimulates gall bladder contraction, potentiates secretin-induced pancreatic bicarbonate secretion and slows gastric emptying time. CCK is also a major neurotransmitter in the brain. CCK exists in various forms and lengths although sulfated CCK-33 appear to be the predominant form. The octa peptide retains full activity of the 33 peptide molecule. CCK in the duodenum is inhibited by a negative feedback mechanism 603540 3M CEoPIm-p0a0n0y3 Page7of 42 involving primarily trypsin. Whether CCK initiates or promotes pancreatic cancer remains highly controversial. Iti clear that administration of exogenous CCK or its analog cerulein does induce pancreatic hypertrophy and hyperplasia in several species as measured by increased DDN synthesis, DNA content, RNA content and glandular weight Various methodologic approaches have examined the issue of whether CCK causes `pancreatic neoplasia. These endeavors have included the administration of exogenous CCK, manipulation of endogenous CCK, administration of CCK receptor antagonists, measurement of CCK receptor binding activity and detection of CCK receptor mRNA. Data from more than seventy laboratory animal studies have provided mixed resulis [Herrington and Adrian, 1995). CCK-8, CCK-9 and CCK-39 have promoted growth of human pancreatic cancers in cell culture in serum-free media (Palmer-Smith et al., 1991] However, fasting plasma concentrations of CCK in unresected pancreatic cancer patients did not differ from healthy controls [Rehfeld et al., 1994;Adrian etal., 1994]. Because the CCK receptor activity of the rat maybe quite dissimilar to the human [Gavin etal, 19972:1997b], a cynomologus monkey may be a more appropriate animal model to use for apancreatic pathophysiology consequence of exposure to PFOA in the: human. In this regard, 3M, in conjunctionwiththe APME, has inated a minimum 3 month feeding studyofPFOA to cynomologus monkeys. A primary study hypothesis is `whether increased CCK levels and pancreatic hypertrophy are associated with serum PFOA levels in these cynomologus monkeys. Because the pancreatic effects (e.g. increased CCK levels, increased pancreas weights) in rats were not observed by Obourn et al [19972:1997] until the sixth study month, it is anticipated that at least the same 03541 3M CEoPmLp0a0n3y Page of42 `comparabletime period, if not much more, will be neededto determine whether such an effect exists, if atall, in cynomologus monkeys. Because the role of CCK with pancreatic cancer remains quite controversial in laboratory animals, further insight into any role exposure to PFOA may have in relation to CCK in humans should be considered. Inthis regard, we propose to examine, via a cross Sectional epidemiological study design, the plasma CCK levels of 3M Coniage Grove fluorochemical productionworkersin relation to their serum PFOA levels. Due tothe # hypothesis that elevated levels of CCK in the ratare a result of choleostasis, we will also `examine hepatic enzyme function, bilirubin and lipoproteins. METHODS Study Design and Population `The study is a cross-sectional design. Cottage Grove production have biennial medical surveillance examinations as specified in the 3M Medical Surveillance Protocol for fluorochemicals. Eligible employeces are all building 15 employees (department numbers 3020 and 3060), building 6employees (department number 3036) and building 3 employees (hey will automaticaly be included 1 theyar staffed fom buildings 6 and 15). In addition, al plant engineering employees and craftworkers who sevice buildings 6 and 15 are eligible for the study (departments 0688and 0690). I is anticipatedthatthere will be a maximum of 100 employees eligible for the medical surveillance examination. Data Collection 03542 3M CEoPmLpO0a0n3y As specified by the 3M Medical Surveillance Protocol for fluorochemicalPsa,geth9e.of42 `medical history, tests and physical examination requirements include: a medical questionnaire along with a self-administered special questionnaire designed for the fluorochemical production area (Appendix B), spirometry, serum chemistries, hematology, urinalysis, serum PFOA measurements, and the measurement of the employee's height, weight, blood pressure and pulse. These medical surveillance examinations are scheduled to take place between September 15, 1997 and October 31, 1997. Clinical laboratory ~~ / evaluations will occur at United Hospitals (St. Paul, Minnesota). Serum PFOA levels `will be evaluated by Dr. Jack Henion at the Advanced Bioanalytical Services Inc. (Ithaca, New York). In addition to the above medical surveillance parameters, we will also draw 10 ml of blood for plasma CCK-33 measurements. CCK-33 will be measured by direct radioimmunoassay by Inter Science Institute (Inglewood, California). The Inter Science Institute' technical coordinator for this project will be Alan Kacena. The requirements for plasma CCK determination are a 10-12 hour fast prior to collection of the specimen. Antacid medication or medications affecting intestinal motility should alsobe discontinued, if possible, for aleast48 hous prior to collection. The Inter Science Institute will supply SMa 10 md EDTA coletion tube which contains the special G.I. preservative. Trasylol. Plasma should be separated from the cells immediately after collection and then frozen in a plastic vial. Specimens should be shipped on dry icet the laboratory. Specimens can be shipped in a batch sample. 03543 3M CEPaLmOpOaGnSy Page 100142 `The expected upper reference level of CCK at this laboratory is 80 pe/mi. The laboratory has provided the following data regarding quality assurance aspects of its CCK33 radioimmunoassay. `Specificit-y Cross reactivity was determined a the 50% inhibition of binding level. CC cocmr Kpoo uns d s-reacti1.v 00 ity Gastrin 0.04 Secretin 002 Glucagon <0.01 / Insulin <0.01 Vasoactive Intestinal Polypeptide <0.01 Gastric Inhibitory Polypeptide ~~ <0.01 Pancreatic Polypeptide. <001 Modlin <001 Other compounds tested <001 Recovery - Specimens were spiked with a known quantity of CCK and measured to determine the amount of recovery. Specimen (oo/m)____ 17 18 2 Amount added 20 50 50 Amount measured 34 66 131 Amount expected 37 68 42 Recovery (5% 9917.09 922 7 Ifnotrrat-harseseacyVoanrwioalbsilaistsyay-eTdhien moenaenr,unstaarnedalridstdedevbiealtoiwo.n and coefficient of variation Mean Co16mb 94 50 Standard Deviation Ge2m3bh 74 130 Coefficient of Variation (1%4).4 7892 03544 3M CEoPm1p0a0n0y3 Inter-assavVariation - The mean, standard deviation, and coefficient ofPvaagreia1t1ioofn42 forthree controls assayed in different runs are listed below. I Mean Standard Deviation 91 8.1 163 s2 Coefficient of Variation 89 93 A consent form will be signed by all luorochemical production employees who participate in the determinationoftheir plasma CCK levels (Appendix C). In essence, the only additional item that willbe collected beyond that already specified in the Medical Surveillance Protocol Manual for fluorochemical workers is the 10 mi of blood necessary for plasma CCK radioimmunoassay analysis. Data Analysis After consideration of normality of the data, stratified analyses, ANOVA, Pearson correlation coefficients, and linear multivariate regressions willbeused to evaluate for positive or negative associations between PFOA and CCK-33, hepatic enzymes and lipoproteins. It is anticipated that the following PFOA levels will be used for the fategorical analyses: 0 - < 1 ppm, 1 - < 10 ppm, 10 - < 30 ppm an>3d0 ppm. These levels were used in aprevious study of reproductive hormone levels and PFOA among the same Cottage Grove fluorochemical production workforce [Olsen et al., 1997). Four `potential confounders will be considered in the analyses: alcohol consumption, cigarette smoking, body mass index and the employee's age. Swdy results will be tabulated and analyzed by packaged procedures in the SAS System SAS, 1990). 03545 3M CEoPmLpO0aGn3y Page 120042 DISCUSSION `The purpose of this epidemiologic investigation is o determine whether plasma CCK-33 levels ar associated with serum PFOA levels among fluorochemical production employees at the 3M Cottage Grove plant. There are several strengths to the proposed study. First, although it is hypothesized that PFOA may cause CCK-induced pancreatic acinar adenomas in the rat, this model may be irrelevant as the histology and CCK-A ~~ # receptor activity of the rat appears to be quite different than that observed for humans. Second, this study will measure plasma CCK and serum PFOA levels. This level of specificity of exposure is often missing in occupational epidemiology investigations. Qualitative/subjective data will only be used in the analysis of alcohol consumption and cigarette smoking. Third, the results of this study may have considerable importance in the interpretation of CCK-related data obtained from the cynomologus monkey study. Although an infrequent occurrence, results from this epidemiologic investigation may direct further toxicological studies. Fourth, in addition to plasma CCK-33 levels, we will also examine hepatic enzymes, bilirubin and lipoprotein levels in relation to serum PFOA levels because choleostassi the proposed mechanism for enhanced CCK levels in the rat. Several methodological issues will need to be considered in evaluating the results from this study. First, the cross-sectional design does not allow for adirect analysis of the temporality of an association. Given the long-half life of PFOA, itis conceivable that there may be some biological accommodation to the effects of PFOA, as suggested by Biegel etal (1995). Second, although there may be upwards of 80 employees eligible for CCK-33 determination,notall subjects may participate. Any refusals wil decrease the 03546 3M CEPoIm.p0a0n0y3 Page 13 (42 statistical power of the study and result in nonresponse bias. Third, there exists potential `measurement error for the confounders of interest. Currently, alcohol consumption and cigarette usage are politically sensitive issues and thus the accuracy of these self-reported data must be questioned. Fourth, it is recommended that CCK measurements be done after a 10-12 hour fast. Employees will be remindedofthis fact; however, whether they adhere to this remains-to-be-seen. `The protocol, any addenda to the protocol, data analyses, and a copy of the final / reportwill undergo a Quality Assurance audit. Permanent recordsofall other data `generated during the course of this study are subjectto privacy and confidentiality considerations. All data gathered or generated including protocol addendum and the final report will be archived by the Medical Department, 3M Company, St. Paul, Minnesota. Costs for routine medical surveillance and the epidemiologic analyses will be bome by the Medical Department. Costs for CCK-33 analysiswillbe covered by the Specialty Chemicals Division. Inter Science Laboratory has quoted $120 per sample for CCK-33 radioimmunoassay based on an 100person batch sample. Thus, costs should not exceed $12,000 for the CCK-33 analyses. | Tis esimated that daa collection will be completed by November 15, 1997. Sample analyses should be completed by January 15, 1998. Data analysis and report writing will take approximately 3 months. Thus, adraft manuscript ready for review and approval should be available by April 1, 1998. Upon completion of this study, results will be communicated to management, fluorochemical production employees and the technical community. A manuscript for 03547 3M CEoPmLp0aOn3y Page 14 of 2 publication consideration in a scientific journal will be prepared and submited for peerreview. ` ' 03548 APPENDIX A 3M C`EoPm1p0a0n0y3 Page 15 of 42 u ' 03549 `Pancreas Anatomy and Histology. 3M CEoPmLpOa0n}y Page 16.0f42 `The pancreas first appears embryologically at the fourth week of gestation [Ermak and Grendell, 1993]. The pancreas arises from a dorsal and ventral outpouching. The tail, body and partofthe headofthe pancreas are formed by the dorsal component. The remainder of the headofthe pancreas develops from the ventral outpouching. In the adult, the pancreas measures 12 to 15 cm in length and weighs between 70 and 100 gm. The. circulationofthepancreas is derived from branchesoftheciliac and superior mesenteric / arteries. The venous drainage flows into the portal system. The visceral efferent innervation of the pancreas is through the vagi and splanchnicnervesvia the hepatic and celiac plexuses, respectively. `The pancreas functions as both an exocrine and endocrine organ. The endocrine pancreas is located in the isletsof Langerhans. Glucagon, insulin, somatostatin and pancreatic polypeptide are produced in the A, B, D and PP cellsofthe islets of Langerhans, respectively. B cells are the most abundant representing 50 to 80 percent of islet volume. The exocrine pancreas consistsof countless acini with their accompanying ductules. The islets are intermixed with the acini. Acinar cells surrounding the islets of Langerhans appear tobe morphologically and biochemically different from acini further moved rom he ets. An snus is a network of. adjoining acinar cells which can take ona variety of shapes (e.g., spherical, tubular or imegular). The intercellular connection between acinar cells is the gap junction which functions as a pore to allow small molecules (500 to 1000 daltons) to pass between the cells. The acinar cells synthesize, store and secrete digestive enzymes into the lumen of the acinus. On an acinar cell's basolateral `membrane are receptors for hormones and neurotransmitter that stimulate enzyme 03550 3M CEoPmLpOaODn3y Page 17 of 42 secretion. The basal region (furthest away from the lumen of the acinus) of the acinar cell contains rough endoplasmic reticulum for protein synthesis and comprises about 20 percentofthe cell volume. The apical region (nearest the lumen of the acinus) of the acinar cell contains zymogen granules which hold the digestive enzymes. The Golgi complex is located between the nucleus and zymogen granules. The lumenofthe acinus is the origin of the secretory duct and contains centroacinar cells. The centroacinar cells function similarly to the ductule epithelial cells by secreting ions and water. The lumen of! the acinus leads to intralobular ducts which eventually anastomose to create interlobular ducts. The interlobular ducts anastomose to form the main pancreatic duct. The primary function of the ductal system is to secrete an alkaline solution which aids in the transport of digestive enzymes to the lumen of the intestine. This alkalinity also raises the. intralumen pH of the intestine necessary for digestive enzyme action. Pancreas Physiology `The principal inorganic components of exocrine pancreatic secretions are water, sodium, potassium, chloride and bicarbonate [Pandol, 1993]. Smaller quantities of calcium, magnesium, zinc, phosphate and sulfate are also secreted. The purposeofthese secretions is to aid in the deliveryofdigestive enzymetso the intestinal lumen and neutralize gastric acid emptied into the duodenum. These inorganic pancreatic secretions, produced primarily from the ductal system as the result of secretin stimulation in the upper intestinal mucosa, vary from 0.2 ml/min in the resting state to 4.0 ml/min during stimulation with the total daily volume equal to 2.5 liters. The bicarbonate concentrate varies between 25 mEQ/L (low flow) to 120 mEqL (high flow). Chloride concentrations 03551 3M CEoPmL0p0a0n3y vary inversely. Secretin stimulates secretion by activating adenylate cyclase Page and 18.042 increasing cyclic adenosine monophosphate (cAMP) which ultimately results in an `exchange of chloride for bicarbonate ions in the lumenofthe pancreatic ductule. `The principal organic constituents of the pancreas are proteins, primarily digestive enzymes, produced from the acinar cells [Pandol, 1993]. The major enzymes function in the following manner: digest starch and glycogen (amylase): hydrolyze wiglycerides (lipase and phospholipase A2) as well as cholesterol esters, lipid soluble vitamin esters, " diglyceride and monoglyceride (carboxylesterase); and cleave peptide bonds (trypsin, chymotripsin and elastase). `The concentration of protein in the pancreatic secretions depends on the output rate from the acinar cells. Theenzymesare secreted into the pancreatic duct in an inactive `precursor form because they could potentially digest the pancreas. Digestive enzymes in the acinar cell include: proteolytic enzymes (trypsinogen, chymotypsinogen, proelastase, procarboxypeptidase A, procarboxypeptidase B); amyolytic enzyme (alpha amylase); lipolytic enzymes (lipase, pro-phospholipase As, carboxylesterase lipase), nucleases (deoxyribonuclease, ribonuclease) and other enzymes (pro-colipase and trypsin inhibitor). Activation ofthe recur form to the fina enzyme occurs in the intestinal lumen. Trypsinogen is activated to trypsin via the action of enterokinase. The trypsin then, in tum, activates other types of proenzymes to their final enzyme form (c.g.. chymorrypsinogen, proelastase, procarboxypeptidases A and B, prophospholipase Asto chymotrypsin, elastase, carboxypeptidase A and B, and phospholipase A;respectively). In addition to the proenzyme activation of these other enzymes, trypsin also activates the conversion of wrypsinogen to trypsin. 03552 3M CEPoLmOp0aGn3y Page 19.0 2 `The synthesis of these digestive enzymes occurs in the rough endoplasmic reticulum of the acinar cells. Once synthesized, these proteins are transported to the acinar cell's Golgi complex where further glycosylation and concentration occur. Finally the enzymes are transportedtothe zymogen granules via vesicles that cycle back and forth between the Golgi complex and the granules. It appears that each zymogen granule contains the entire complement of secretory enzymes although their concentrations may differ between granules. The percent cell volume that the zymogen granules represents / can vary from 1510 20 percent (fasted adult) to less than one percent after stimulation `witha cholinergic drug or the gastrointestinal hormone, cholecystokinin. The enzymes are. believed to be released from the zymogen granules by exocytosis (the fusion of the granule `membrane with the apical cell membrane and subsequent release of the granule content into the ductule lumen). It is thought that different stimuli result in the selective secretion of specific enzymes from the acinar cells. Furthermore, the regulationof protein (enzyme) synthesis can be altered by dietary intake. Increased concentrations of amylase and decreased levels of chymotrypsinogen have been reported with carbohydrate-rich diets. | Studies onthe regulon of enzyme seceon in humans ar limited [Lu eal. 1989; Chey, 1991; Pandol, 1993). The majorityofinformation known, to date, is from animal models. Regulationof enzyme secretion is controlled by neuro- and humoral stimulationofthe acinar cells. ~ Receptors, located on the basolateral surface of the acinar cell, have been reported for cholecystokinin (CCK), acetylcholine, bombesin, substance P. vasoactive intestinal peptide (VIP) in the guinea pig, rat, and mouse. Depending on their mode of stimulus-secretion coupling, the receptors have been divided into two categories. 03553 2 3M CEoPm0p0a0n3y VIP and secretin increase cellular cAMP via activation of adenylate cyclase. AnPiancgree2as0eoif2n CAMP mediates the secretory response. The acinar cell also contains receptors for CCK. acetylcholine, bombesin and substance P. Rather than stimulating cAMP, these compounds increase cellular metabolism of membrane phospinositides and calcium. A major effofeCcCKt, acetylcholine, bombesin and substance P is the mobilization and release of intracellular storesofcalcium. These agonists also increase cellular cyclic guanosine monophosphate, arachidonate release from membrane phospholipids, and ~~ * electrical membrane potential changes. The exact mechanisms by which these steps lead 10 enzyme secretion remain to be fully elucidated. There does appear to be a greater than additive response to the joint effects of agonists which alter cAMP (e.g., VIP and secretin) and calcium (e.8., CCK and acetylcholine) Exocrine pancreatic secretion occurs during fasting (interdigestive) as wel as after ingestion of a meal (digestive). The digestive state is divided into three phases: cephalic, gastric and intestinal. The cephalic and gastric phases are neuro-regulated whereas the intestinal phase is regulated by neurological and hormonal factors. In the cephalic phase, increased pancreatic secretions of bicarbonate and digestive: enzymes will occur afer smelling sting, chewing and swallowing food with or without duodenal chyme or acidification. This is due to increased vagal tone which increases inrapancreatic postganglionic cholinergic input which subsequently stimulates the release of the bicarbonate and enzymes. These neurons in the pancreas are activated by central input during the cephalic phase and by vagovagal reflexes initiated by the stimulation during gastric and intestinal phases. Besides acetylcholine, there are other neurotransmitters in the pancreas that contain peptides, VIP, gastrin-releasing peptide, 03554 . 3M CEoPmLpOa0n0y3 Page 21 of 42 CCK, neuropeptide Y, neuortensin, substance P, enkephalins, calcitonin gene related peptide and galanin. These peptides may coexist with nonpepide transmiters in autonomic nerves and thus appear to play significant roles in the regulation of the exocrine pancreas. The gastric phaseof pancreatic secretion results from food stimuli in the stomach, The gastric stimuli cause primarily enzymatic release with minimum secretion of water and bicarbonate. ' `The intestinal phase represents the most important aspect of pancreatic enzyme secretion. The intestinal phase begins upon enotfcrhymye into the duodenum. The intestinal phase is mediated by both enteropancreatic vagovagal reflexes and hormones. There are two major mediators of these pancreatic enzymatic secretions: secretin and CCK. Thetypeof pancreatic secretion depends on whetheritis released by secretin or CCK. Secretin is the major mediator of ductal bicarbonate and water secretion. CCK is the major mediator of pancreatic enzyme secretion. The release of secretin from the duodenal mucosa depends upon the acid load (minimum pH of 4.5) that is delivered to the duodenum. Fatty acids greater than eight cabons in length and ile acids are likely secondary simulansforsecesn release. Secretion of bicarbonates also occurs via cholinergic input. Although CCK by itself does not invoke bicarbonate release, it can augment secretin-induced bicarbonate secretion. CCK is the major humoral medofimeaal-sttimoularted enzyme secretion. CCK is released from the `I' (or sometimes called *M") cells of the upper small intestinal mucosa by products from fat and protein digestion. Phenylalanine, valine, methionine and tryptophan are the most potent amino acids for CCK release (and subsequent pancreas C03555 . 3M CEoPm1p00a0n3y Page 20(42 enzyme secretion). A low level of intestinal contents mediate pancreatic enzyme secretion by an enteropancreatic neural reflex whereas high loads of intestinal contents mediate enzyme secretion predominantly by the hormonal effects of CCK. The release of secretin and CCK is mediated by releasing peptides secreted from the intestinal mucosa. These releasing peptides are, in tum, inactivated in the upper small intestinal lumen by pancreatic proteases. Specifically, the regulation of CCK occurs in the following manner: 1) chyme enters the duodenum which results in the releasing peptides secreted from the intestinal `mucosa which subsequently release CCK from the I" cells 2) intestinal CCK is absorbed and travels, via the circulatory system, to the pancreas where it binds to receptors; 3) the activated receptors causes the release of pancreatic precursor enzymes from the acinar cells' zymogen granules into the pancreatic ductules which subsequently drain into pancreatic duct and then empties into the intestinal lumen; 4) the precursor enzymes are activated in the intestinal lumen where oneofthese enzymes, rypsin, assists in the digestion of protein: 5) upon cessation of eating, the concentration of free trypsin increases in the intestinal lumenwiththedecreased lumen contents; 6) this free trypsin inhibits two substances that acta simul for CCK release (monitor peptide and CCKreleasing factor); 7) this inhibition of monitor peptide and CCK-releasing factor subsequently results in the cessation of CCK release from the duodenal `I' cells; and finally 8)a decreased CCK circulatory concentration results in less CCK binding activity at the acinar membrane which ultimately decreases pancreatic pro-enzyme secretions from the. zymogen granules. Of course, this negative feedback cycle begins again with the next stimuli entering the duodenum (i.e., the next meal). It appears, albeit to a much lesser 03556 3M CEPoLmOp0a0n3y Page 23 of 2 extent, that circulating insulin, starch, glucose and calcium may also produce an pancreatic enzyme secretory response. The evaluation of exocrine pancreatic function is accomplished by both direct and indirect tests. Because of its large functional reserve, malabsorption does not occur until CCK-stimulated digestive enzyme secretion is reduced to 10 percent of normal; thus most tests have low sensitivity to detect mild to moderate degrees of pancreatic insufficiency. The direct function tests are based on the measurement ofenzyme and bicarbonate + secretion upon duodenal intubation and stimulationofsecretin and/or CCK. A variety of indirect tests are available which examine surrogatesofeither stimuli and/or the actual : digestive enzyme secretions [Pandol, 1993]. Epidemiology of Pancreatic Cancer Briefly, cancer of the pancreaiss the ninth leading cause of cancer and the fourth leading cause of cancer death for men and women in the United States [Anderson et al., 1996). Five-year survivorship is less than 5 percent. Pancreatic cancer occurs fifty percent more frequentlyinmales thanfemalesand blacks than whites. Adenocarcinoma of the pancreatic duculs isth primary hisologi type. The proportion of pancreatic cancer from acinar origin is estimated at 1 to 15 percent. Numerous pancreatic cancer case-control studies published over the last 15 years have failed to show any strong associations. The most consistent association is, on average,a two-fold increased risk with cigarette smoking. Several linesofepidemiologic evidence suggest diets high in : animal fat and low in vegetable andfruits are weakly associated with pancreatic cancer. C03557 3M CEoPmLpOa0nGy3 Page 24 of 42 No consistent associations have been reported for industry, occupation, or specific chemical exposures. `The most important discovery during the past 15 years has been the accumulation of data which shows that mutations in cellular proto-oncogenes and tumor suppressor `genes are important events in pancreatic carcinogenesis [Caldas and Kem, 1995). K-ras mutations area frequent finding in adenocarcinomas of the pancreas in humans with the great majority of these mutations foundin codon 12 of c-Kirsten ras. In fact, pancreatic / cancer is the human mor withthehighest incidence of ras mutations. The ras gene family encodes proteins involved in cell growth and differentiation. Thus, point mutations in ras genes may play an important role in early pancreatic carcinogenesis. Other evidence suggest roles for mutations in other oncogenes (myc, erbB-2) and tumor suppressor genes (apc. pS3). These findings supporta genetic model of pancreatic tumorigenesis: ductal cells driven by almost universal mutationof a dominant oncogene, Keras and deregulation of cell-cycle control. Whether multiple carcinogens may cause `pancreatic cancer remains a plausible hypothesis However, the degreeof consistency in the mutations, transversion and transitions of nucleotides from G to Tor G to A in codon 12 0f K-ras may eventually pinto only afew specific causes. Animal Models of Pancreatic Cancer `There are two animal models of pancreatic cancer (Longnecker, 1990]. A number of nitroso compounds have been used to produce ductal pancreatic cancer in hamsters including N-nitrosobis(2-oxopropyl)amine (BOP), N-nitrosobis(2-hydroxypropyl)amine (BHP), and N-nitroso(2-hydroxypropyl)(2-oxopropylamine (HPOP) [Pour et al., C03558 3M CEoPmLp0a0n0y3 19751981). In the rat, azaserine (o-diazoacetyl-L-serine) is the Page 25 of42 most commonly used carcinogen [Longnecker, 1986]. Howeverthe hamster and rat models are of different cellular origin. Carcinogensin rats induceacinarcellmalignancieswhich are rareinthe human. Hamster models areof ductal cell origin and thus resemble human pancreatic cancer. Furthermore, the patiem of genetic mutations found in tumors from BOP-treated. `hamsters parallels thatofhuman cancers. Activation of the -K-ras gene is frequent in both human and hamster pancreatic cancer but is not foundinazaserine-induced / pancreatic cancer in the rat [van Kranen et al, 1991]. ` 03559 CRCKecepatndotherPanscre,as 3M CEPoIm0p0a0n3y Page26 of42 As discussed previously, CCK is a potent regulatory peptide that is the major stimulus for pancreatic enzyme secretion. CCK-33 refers to its 33 amino acid arrangement: MeLLAyrytsgs----AATAsllspaen----PPSLhreeeoru---SAGeslrpn---GAAlrsygn---AALrsegpu---MTAeystrp---SMPeerrto-.-iGSele-yrV--aTHlri-ps-- , where Ala = alanine Ag = arginine Asn = asparagine Asp = aspartic acid Gln= glutamine Gly = glycine His = histidine: Tle = isoleucine: Leu= leucine LMyest==lymestihnieonine. PPhroe == pprhoelniynlealanine Ser = serine Trp = uyptophan TVaylr== vtaylrionseine ' CCK stimulates gallbladder contraction, potentiates secretin-induced pancreatic bicarbonate secretion, and slows gastric emptying. CCK also exists in various forms and lengths from metabolites CCK-4 to apro-CCK peptide, CCK-58. The octapeptide retains full activity of the 33 peptide molecule. Sulphated CCK-8 and CCK-33 appear to be the predominant forms. CCK acts through more than one type of receptor. In animals, CCK-A receptors are found on pancreatic acinar cells, smooth muscles, vagal afferent fibers and some central neurons. CCK-A receptors have a high affinity for sulphated 03560 3M CEPoIm0p0a0n3y `CCK-8 and recognize gastrin poorly. On the other hand, CCK-B receptors arePaabguen2d7anotfl2y found in the central nervous system and on gastric glands and recognizes both sulfated and nonsulfated CCK-8 and gastrin. Because circulating levels of gastrin are appreciably higherthan those of CCK, the CCK-B receptors are usually activated by gastrin leaving the CCK-A receptors as the predominant form for CCK binding. CCK-A and CCK-B receptors are not consistently found in all species. Recent studies indicate, that unlike ratand dog pancreas, the human pancreas has no detectable # CCK-A receptors and lite to no mRNA for the receptor [Wank et al., 1994], In particular, human, cynomolgus and rhesus monkeys lack specific binding sites for the selective CCK-A ligand *{H]L-364,718 (Gavin et al., 1997). In contrast, the baboon pancreas exhibited a significant number of specific CCK-A bindings sites although not as much as the rat or guinea pig pancreas. As also stated previously, CCK is inhibited by a negative feedback mechanism involving primarily trypsin: lack of trypsin in the duodenum and jejunum results in CCK release, whereas the presence of free trypsin inhibits CCK release. Thus, any interference with this feedback mechanism by orally administered trypsin inhibitors will result in continued pancreatic secretion via the release of CCK. These trophic effects of CCK have been demonstrated in both rats and hamsters (Pour et al. 1981; Longnecker, 1990]. 03561 `TheRoleofCCKandPancreaticCancer 3M CEoPmLpOa0Gn3y Page28of 2 Whether CCK initiates or promotes pancreatic cancer is not a new question and the research published,to date, remains highly controversial with elusive answers [Axelson etal., 1992; Hemington and Adrian, 1995]. What is clear is that administration of exogenous CCK or its analog cerulein does induce pancreatic hypertrophy and hyperplasia in several speciesasmeasured by increased DNA synthesis, DNA content, RNA content, total protein contentand glandular weight [Mainz et al., 1973; Petersen et al., 1978; 4 Zucker etal., 1989). Increased endogenous plasma CCK levels, via pancreaticobiliary diversion (thought to inhibit the negative feedback regulation of CCK release), have also resulted in pancreatic hypertrophy and hyperplasia [Gasslander et al, 1990). Furthermore, administration of a CCK-A receptor antagonist will abolish or markedly reduce the pancreatic. trophic response of CCK. Feeding ratsor hamsters high-fat or high-protein diet induces pancreatic hypertrophy and hyperplasia (Longnecker et l., 19%0].. Various methodologic approaches have examined the sue of whether CCK causes pancreatic neoplasia, not just hypertrophy and hyperplasia These research endeavors have included the administration of exogenous CCK, manipulation of endogenous CCK, administration of CCK receptor antagonists, measurement of CCK j-- binding activity, and detection of CCK receptor mRNA. Herrington and Adrian [1995] have summarized this body of research by stating, "in spite of extensive research, the role of CCKin pancreatic cancer is still very unclear, and the subjecti highly controversial. Data from more than seventy studies have variably suggested that CCK has no positive wophic effects, inhibitory effects, or no involvement in pancreatic tumor growth." 03562 . 3M CEoPLm.p0a0n0y3 Page 29 of42 Provided below for illustrative purposes are five examples offered by Herrington and Adrian [1995] regarding the controversial issues surrounding CCK and pancreatic cancer. f1.r)omSefvreorgalskiinn)vievnohaenxcpeedritmheentdsevweiltohpemxeongteonfoupsanCcrCeaKtoirc ccearnucleeriinn(haaCmsCtKersaneaaltogeddewriitvhed carcinogens [Howatson and Carter, 1985; Satake et al., 1986; Pour etal., 1988] yet other hamster studies showed CCK reduced the incidence of pancreatic carcinomas in hamsters or produced no effect [Pour etal., 1988, Johnson et al., 1988). ' 2.) Endogenous CCK levels have been associated with pancreatic tumors. For example, rpaatnscr[eSatteiwcaorbtileitarayl,di1v9e9r1s)i.onLeonnhga-ntceerdm afzeaesdeirnigneo-ftrreaawtesdopyaancfrleoautrictocraantcs,erwhdiecvehlcoopnmteanintsin p[rMoctGeuasienninehsisbietotrasl,.p1r9o8d0]u.cedHpoawnecvreera,s faceiendairngadreanwosmoaysaafnldouardoernsoycnatrhceitincomparsotease inhibitors to hamsters did not result in increased pancreatic tumors even though plasma. CCK level were increased [Herrington et al., 1994). 3.) CCK receptor antagonists have been used to investigate the role of CCK, exogenously or endogenously, in pancreatic cancer. These antagonists include glutaramic acid derivatives (proglumide, lorglumide, loxiglumide) and nonpeptide compounds (asperlicin, devazepide, 365,260). The results of these studies varied considerably depending upon animal model, the timing of the administration of the antagonist in relation to. administration of the carcinogen, dosage administered and cultured cels. Studies of the CCK receptor antagonist devazepide in patients with pancreatic cancer have not shown significant effects on survival [Abbruzzese et al., 1992]. 4) CCK binding was consistently demonstrated in whole cells and membranes from acinar `c`edlelmtounmstorrasteidn otnheirnattacat,zacsuelrtiunreedmohduelm.anHtouwmeovrerce,llsspeocfifdiucctCalCcKelrleocreipgtionrs[Ahdarvieannoettbael.e,n s1e9v94e)r.al dNiefvfeerrtehnetlepsasn,crCeaCtiKc bcianndcienrgcehlalslbieneesn[dSeimnognhstetraalte1d98i6n;mSemmibtrhaenteaflr,ac1t9io9n4s]fbruotmthis significance [Herrington is uncertain and Adrian, when receptors 1995). cannot be demonstrated on the cell surface: c5.u)ltuCrCe Kin (sCeCruKm--8f,reCeCmKe-d9iaan(dPaClCmeKr--3S9r)itphroetmoatl,ed19g9r1o)w.thOonf tphaencortehaetrichacnadn,cefrastinincgell lpelvaeslmsaincohnecaelntthryatcionotnrsoolsf oCrCiKn piantiuennrtessewcittehdcopalnocnreoartliucncgancacnecreprat[iReenhtfsedliddetnoatl.,differ from 03563 3M CEoPmLOp0aOn3y 1994; Adrian etal, 1994]. CCK levels diddeclinein pancreatic cancer Page 30f 2 patients after dpaunocdreenautomd[uAoddreinaencettomaly.,d1u994t)o.the loss of CCK-producing Icells located in the In conclusion, CCK likely promotes pancreatic tumors in some animals but there is no evidence that it is an inducer of pancreatic tumors. Although CCK receptors have not been extensively investigated in the normal human pancreas, there may well beadifferent pate of receptor expression in humans than other species.Iftue, then resis from laboratoryanimals,in particular the azaserine-rat and nitrosamine-hamster models, may be of minimum applicability to understanding CCK and its relation to abnormal pancreatic function in humans. ' 03564 APPENDIX B 3M Company PageEP31Lo0f00432 ' 03565 3 [eT Jr Je -[ge 3M Health Questionnaire [reo T5o m | FIA 3M Company ay 2. Do you smoke cigarettes now? ( as of one month ago) r + insemap--m--r--e--em--------------"------"" Stoppedsmoking) 8. On average, how many cigarettes do you smoke per day now? oy m m--e--e --r ---- or Atonol (1 alcohol beverage = 1 beer, 1 glass of wine; or 1 shot of hard liquor) 11. Do you currently drink alcoholic beverages? C03566 : 0 34 Company pects [eecrcaneer vesTho 7 Tf Yes, year you were re TT1 Lymphoma (Non- `Name and address of Doctor = Fr E rr r r Tr r r c i eHr E E IEr E por 03567 [omnes --L_z-- TV [NO] TI, Yes, year you were | Name and address of Doctor who `Hepatitiscausedby [|] Pi e e H T E Identify other liver R eCi ee T ---- (Rheumatoid arthritis, (overactive thyroid) (underactive thyroid) FET ---- ann i C03568 APPENDICX 3M CEoPmLpOaOn3y Page 350 42 J ` G03569 `CONSENT FORM 3M CEoPmLpOa0n}y Page 360142 An Epidemiologic InvestTiigalteioonfoSftuChdoylPercyssotcookilnin snd Hepa Funciion among Perluorooctanoic Acid Production Workers YInoturoadnudcytoiuorncolleagues involved in luorochemical production st Cotage Grove ar being invited qtouepsatriioncsipaarte ainnsawreerseedarbcehfostruedyy.ouPmleaaksee areevciieswiotnistocopnasretnitcifpoar.m Tcahreefnuulrlsyeaannddboerpshuyrsiycoiuarn can provide you additonal informationa the ime of your medical surveillance examination. "PTuhreppousrepoosfeSotftuhdiys study is (0 conduct the routine medical surveillance exam tht is offered you} eiavnnefdorrFymCta-wt1io4oy3negalearvi.enlesId.nfCardCoimKttoihsinsa1sh0totuhrdiymsowenixlelatmfhuawrtehisewrnilehlceel1spssuasrcyuhniedncekprasyntocaurnerdabtalinocyoedhnuczhmyoamlneecepyrxsoptdoouksciutnriieonn.(FCTCChK.e)1.43, 1Styuodu cPhrooocseedutorepsaniipate we willbeasking that you fl out the Health Questionnaire and provide cpbrolemosposlduerstea,emhbpellieogsohdtf.ocrwoeulinagtbhoatrnaadtnodriyflfeuesnrteginntfgiua.nlc,tpi~loTinheetleeesttxiancgmo.uwniL;lalbaconordnastbiolsrtoyoodtfecashtemimneigsdtwiricilealsi:snuctvlaeuaidlle,lhaHenDcmeaLftooarlnomdg.yLb:DlLo.od cholestrol. rigyceride, plucose (blood sugar), liver and Kidney function tests and pancreas fu`nwcitthioonnetensete.dlIen saidcdkiatndiwoweinlwlielqlimreaesaurpepfroroFxCi-1m4a3&iineubyloeyu.rPblleoaosde.reTmheemb berl focfaao nsbtfeo odrraad twlneast 4182 hhoouurrss pprriioorr coordinator for e1o0xcthhheedietvessi.ts.ioASnlcshoe,dupllienagsewidlol bnoetcuosoe,ridfiantaalbl yptoshseibMleed,iacnayl aDnetpaacritdmmenetdticoatuigofnhosr wInethneeeqduefsotbioenanbasiorleutyeoluywsuirleblethaatskyeoduirf lyisotuedhcaovnedistpieocnifcicumledlicaexlisdtiss.easFeosr. thFaotrrceearsaoninwperombalyembse, OgeNttLinYgaibnotuotutchhewmietdhiycoaul tco aosknfordoyfoiiunrttpreeirsmtitosusisno.nto contact your physician. This would be. "`rTPehocteeeonsntsliyaalndRdiilsoscrkosms/flDiogirhtsteysoowmueflomlriatnysg fienelhiissafrroema tahfetneetehdeblleosoidck.olYloeucimoany also have some temporary g"Beahnieenrfeeidtifslrobmethniosdsitruedcyt wbielnleffiutrtfhreormheylopurvspaurntdiecripsattainodnainntyihsusmnadny.expHooswuerveer1,FtCh-e1i4n3f,orYmaotuiron individual resulsaswelals the overal results of this sudy will be commicaed you. 03570 3M CEoPLmLp0a0n0y3 Compensation/Medical Treatment From Related liness or Injury Page 37of2 s1tuydoyu, syuofufewrililnjbuerydoirreactmeeddtiocathleco3nMditMiedointchalatDaepppaeraurmsetnotbeforthoebsreersvulattioofnpaanrdtidciiapgantoisnigisn.thAitsyour hreeaqltuhecsoatreatptrhofeesrseicoonmalmeantdnaoticoosnto1f0tyhoeu.3MInMtehdeiecvaelntDoefpaartemseenatr.cyhoruelwaitleldbienjrueryf.ecrormetpodenasnaotthieorn Mwialnldbeel,deMt.eDr.m,in7e3d3-o8n6a70coafsethbey c3aMseMbeadsiiscablyD3eMpa.rtTmheentc.ontact for medical compensation is eff Confidentiality sTthriectilnyfcoornmfaitdieonntiyaoluapnrdowviidleboenutsheedqfuoerstgiroonunpaiarnea,lyosriasnoynliyn.foTrhmeatdiaotnapcroolvleicdteeddtionutsh,iswisltlubdey mkeapyt obreoutsheedridnopcuubmlecnatisoinnstoenrdpedubflorippurbelsiecnactaiiononesx.amYinoautinoanm.e will notberevealed in any publication, p`eYroiuodwiilclhbeealgtihveexnafmeiendabtaicoknaibsonuottyosuurbsitnidtiuvtiedufaolr arnesuelxtasmaisnhaatsiboneebnyydoounredoicnttohre.paWste. eTnchoeurage 30ut0 share the resuls of your tests with your private doctor. ISfubyjoeucthaRvieghatns/yAqvuaeilsatbiiointsayboouftItnhfeosrtmuadtyinoonw. or later,or in the eventof 2 research related injury 1o0r qeumeesrtgieonncsya,bocounttyaocturDrr.igJhetfsfiMn arnedgaerld (073t3h-i8s6r7e0s)eoarrcJhe,aynouBumrraiys,cRonNt.act(7D3r7.-7L8a67r)r.Ry.FoZrobaenls,wers Chair, 3M Institutional Review Board at 733-5181 P`aVrotliucnitpaartyioPnairntitchiispastuidonyaisndvolWuintthadrry.awRaelfusal to participate will involve no penalty or oss of `bYeonuerfidtectioswiohnictohpyaonuiacriepoatteheorrwwiistehdenrtaiwtlefdr.omYtohuiasrsetfurdeyewitlolwniotthadfrfaewcattyaounrywtoirmke sftoartausnyorreason. piserdfetoerrmmainnceedatphpartaiysoaulr.coTnhteiniunevdesptairgtaitcoirpamtaioyn Swoouplydobuerdpeatrtriicmiepnattailontoinyotuhirshsetauldtyhoatiafntyhetismteudifyit objectives are changed. B`SyusbijgecntinCgotnhsiescnonsentform, Icerfythat Iam atl18eyeaarssoltd. Iconthfat ihavreremad ahbiosuctontshiesnctonfsoernm,t afnodm tohratabohauvtetbheeeSntugdiyv.enIaaldseoqucaotnefiorpmporthtautni1tuy ntodasek ranstyhtqeusaecsnotpdieoonfsImmyay have a`pmartsiicginpiantigotnhiins cthoinssSetnutdfyo,ramndvotlhuanttaarlilloyf,mayndquedsetsiiorens(0hapavretibceiepantaenisnwtehriesdSttuodmy.y Isuantidsefracsttiaonn.d t|hat `1aantmicnioptatwiaoinviinntghiosrSrteuldeya.sinTguanndyerosftamnydltehgaatlIrwiiglhltsrebcyeisviegnaincgoptyhisofcotnhissenstigfnoerdm,coornsbenyt form. Signature and Date 3M Emplovee Number 03571 REFERENCES 3M CEoPmLpOa0nGy3 Page 38of2 Abbruzzese JL, Gholson CF, Daugherty K, Larson E, DuBrow R, Belin R, Levin B. A pilot clinical trial of the cholecystokinin receptor antagonist MK-329 in patients with advanced pancreatic cancer. Pancreas 7:165-171. Adian TE, Permet J, Wang Q, Herrington MK, Takahashi T, Larsson J, Pour PM (1994). Is CCK involved in pancreatic ductal cell cancer? Tenth International Symposium on Gaswointestinal Hormones, Santa Barbara, CA. Anderson KE, Porter JD, Mack TM (1996). Pancreatic cancer. (In) Schottenfeld D, Fraumen JF (eds) "Cancer Epidemiology and Prevention (2* edition)." New York: Oxford University Press, pp 725-771. / Axelson J, Ise 1, Hakanson R (1992). Pancreatic cancer: the role of cholecystokinin? ScandJ Gastroenterol 27:993-998 BLieeygdeilgLcBe,l LfuinuctRioCnM:,inHuvirttro,MiEn,viCvooo,kanJdC.ex Evfifveoctsstuodfieasm. moTonxiiucomlpAeprpflluPohroaorcmtaacnoolate on 1995:134:18-25, Bryce H. Industrial and Utilitarian Aspects of Fluorine Chemistry. Fluorine Chemistry. New York:Academic Press, 1964:297-492. In: Simons J, ed. Caldas C, Kem SE (1995). K-ras mutation and pancreatic adenocarcinoma. IntJ Pancreatol 18:16. Chey WY (1991). Regulation of pancreatic exocrine secretion. InJt Pancreatol 9:7-20. C`aomomkoJnCi,uMmuprerraflyuoSrMo,ocFtarnaomaeteS:Ra,pHousrstitblMeEe.ndoIcnrdiuncet-iroenlaotfedLemyedcihganceilslma.deTnooxmiacsolbyAppl Pharmacol 1992;113:209-217. CboyopkerJoCx,iHsuormtetpMrEol,ifFerraatmoresSiRn,CrBile:gCeDlBLBR.(CMDe)chraatnsi. sTmoxsiocfoleoxgtirsathe1p9at9i4c;1t4u:m3o0r1 i(nadbsutcrtaicotn). `Ermak TH, Grendell JH (1972). developmental anomalies. (In) The pancreas: anatomy, histology, embryology, Sleisenger M, Fordtran JS (eds) Gastrointestinal and Diseases, Philadelphia:W.B. Saunders Co., pp 1573-1584. rGeasspsolnasnidbelreTf,orAgxreolwstohn Jo,fHtahkeapnasnocrneRa,s aIfhtseerIp,aLniclrieaatI,icRoebhifleiladryJFd.iveCrhsoiloencyisntroaktsi.niSnciasnd J Gastroenterol 25:1060-1065. 03572 P3aMgeCEPo3Lm9Op00a1Gn43y2 `Gavin CE, Martin NP, Schlosser MJ (1996). Absence of specific CCK-A binding sites on human pancreatic membranes. Toxicologist 30:34. `Gavin CE, Malnoske JA, White J, Schlosser MJ (1997). Species differences in expression of pancreatic chloecystokinin-A receptors. Toxicologist 36:1180 (abstract). Griffith FD,LongJE. Animal toxicity Ind Hyg Assoc J 1980;41:576-583. studies with ammonium perfluorooctanoate. Am cGhuayraWct.eriFzaltuioorn.ocoRmopcoheusntdesr,ofNhYu:maUnnipvlerassimtay:ofanRaloycshiess,tperre,va(lPhe.nDc.e,dpiusrsiefriactaitoino)n,, 1a9n7d2. eHxacnrheitjiaornviinH,thPehraatu.g PRr,ocSiSngoecrELx.pTBhieolsMexe-drel1a9t8e2d;d1i7f1f:e5r1e-n5ce5.in perfluorooctanoate 4 Hanhijarvi H, Ylinen M, Kojo A, Kosma VM. Elimination and toxicity of perflucrooctanoic acid during subchronic administration in the Wistar rat. Pharmacol Toxicol 1987:61:66-63. Howatson AG, Carter DC. Pancreatic carcinogenesis - enhancement the hamster-nitrosamine model. BrJ Cancer 51:1-7-114. of cholecystokinin in Herrington MK, Permert J, Kazakoff KR, Zucker KA, Bilchik AJ, Pour PM, Adrian TE. tEfufmeocrtsinoiftiraatwionsoiynathdeiehtaamnsdtecrh.olCeaccysnteorkiLnetitn 8r2e:c7e-p1t6o.r blockade on pancreatic growth and Herrington MK, Adrian TE (1995). IntJ Pancreatol 17:121-138. On the role of cholecystokinin in pancreatic cancer. Ikeda T, Aibe K, Fukuda K, TanakMa . The inductionofperoxisome proliferation in rat liver by perfluorinated fatty acids, metabolically inert derivatives of fary acids. J Biochem 1985:98:475-482. flohnson FE,LaRegina `pancreatic and hepatic MC, Martin SA, carcinogenesis. Bshiti HM (1983). Cancer Detect Prev 6C:h3o8l9e-c4y0s2t.okinin inhibits Kennedy G. Dermal 198581:348-355. toxicity of ammonium perfluorooctanoate. Toxicol Appl Pharmacol Kennedy G, Hall G, Britelli J, Chen H.. Inhalation toxicity of ammonium perfluoroocianoate. Fd Chem Toxicol 1986;24:1325-1329. pKuesrlfilkuiosroBoIc,tVaannoydle-ncoHeenuzvyemleJAP,fPoertmeartsioonnRiEn.maLlaecakndoffeevmiadleenrcaets.forJpeBrifolcuhoreomdeTcoaxniocyoll- or 1992:7:25-29. 03573 . 3M CEoPmLpOa0n3y Paged00r 42 LnuecmlbeearrgreerceTp.toDressivgenraglinnegBp,aWthawhalyi iWn.lipPiedrpohxyissioolmoegyp.roliAfnernatRoer-vacCteilvlatDeedvreBcieopltors: a 1996:12:335-363. `LiaummRoCnMi,umHpuerrtftlMuoEr,ooCcotaonkoaJtCe,B(iCe8)geolnLhBe.patEifcfeacrtomoafttahseepaecrtoivxiitsyoimneapdruolltifmearlateorC,RI:CD BR (CD) rats. Fund Appl Toxicol 1996:30:220-228. LVoLngWne(cekd)erThDeSE(x1o9c8r6i)n.e EPxapncerreiamse:ntBailolmoogdye,lPsatohfoelxoogcyriannedpDainsceraesaetsi,c NtuemworYs.ork(:In)RavGeon, Pp 443-458. , Longnecker DS (1990). Bull Cancer 77:27:37. Experimental pancreatic cancer: role of species, sex and diet. Longnecker DS (1991). Hormones and pancreatic cancer. Int J Pancreatol 9:81-86. LuL, Louie D, Owyang C (1989). regulation of pancreatic secretion. AAmchoJlPehcyyssitoolki2n5i6n:rGe4l3e0as-iGn4g3p5e.ptide mediates feedback Mainz DL, Black O, Webster PD (1973). Invest 52:2300-2306. Hormonal control of pancreatic growth. J Clin (M1c9G8u0i).nnTehses eEfEf,ecMtsorogfalnonGg-Ht,erLmevfieseodinnDgAo,f sForyaapeflDoLur,oHnopthweoroatdpDan,crWeoars.mslSecyanKdGJ Gastroenterol 15:407-502. Nickles M. Presence du fluor dans la sang. Compt Rend 1856;43:885. WOyboeutrhn1I4D,,643F.raTmoexiScRo,loEglilsitot3G6S:,11C8o1ok(abJsCtr(a1ct9)9.74). Pancreatic oncogeniceffectsof OWbyoeutrhn1J4D,,643F.raHmaeskSeRl,lELlalibootrGatSo,ryCofoorkTJoCxi(c1o9l9o7g)y.anPdanIcnrdeuasttriicalonMceodgiecniince,efEf.eLctsduofPont de Nemours & Company, (unpublished manuscript). eOplisdeenmGioWl,ogGiiclliinlvaensdtiFgDa,tiBonurolferwepMroMd,uctBuirvreihsoJrMm,onMeasndienlmeJSn, wMiatnhdoeclcuJpHa.tiAonnal exposure 10 perfluorooctanoic acid. (Submitted for publication.) EOxpphaBuioglR,MeSding1e9r8L0.;16M3e:t1a9b-o2l3ic handling of perfluorooctanoic acid in rats. Proc Soc 03574 . 3M CEoPmLOp0a0n3y Page 10f 42 Palmer-Smith J, Kramer ST, Solomon TE (1991). CCK stimulates growth of six human pancreatic cancer cell lines in serum-free medium. Regulatory Peptides 32:341-349. Pandol SJ. Diseases, Pancreatic physiology. (In) Sleisenger M, Fordiran PhiladelphWi.aB:. Saunders Co., pp 1585-1600. JS (eds) Gastrointestinal aPamsmtoonriTuPm, LpeereflKuPo,roPoecrtrainMoaAt,e-GiinldluiceesdPJh.epaBtioomcehgeamliycaalndanpdermoorxpihsoolmoegpircoallifsetruadtiioens.ofExp Mol Pathol 1987:47:98-109. aPentdersseecrnetHi,nSoonlpoamnocnreTaEs,oGfrroatss.smAanmMJI.PhyEsfifoelct2o3f4:cEh2r8o3n-i2c8p6e.magastrin, cholecystokinin , Poston GJ, Gillspie J, Guillou PJ (1991). Biology of pancreatic cancer. Gut 32:800-812. Pour P, Kruger FW, Althoff J, Cardesa A, MohUr (1975). A new approach for induction of pancreatic neoplasms. Cancer Res 35:2259-2268. Pour PM, Runge RG, Birt D, Gingell R, Lawson T, Nagel D, Wallcave L, Salmasi (1981). Curent knowledge of pancreatic carcinogenesis in the hamsterand is relevance 10 the human disease Cancer 47:1573-1587. oPnoupranPcMr,eatLiacwcsaornciTn,ogHeenlegseissoinn St,heDohnanmsetlelry mTo,dSetle.paCnarKci(n1o9g8e8n)e.siEsff9e:c5t9o7f-6c0h1o.lecystokinin Rao Sci 1M9S9,6;R8e0d4:d5y73J-K5.87H.epatocarcinogenesis of peroxisome proliferators. Ann NY Acad Rehfeld JF, van Solinge WW Adv Cancer Res 63:295-347. (1994). The tumor biology of gastrin and cholecystokinin. SAS Institue, Inc. Inc., 1990. SAS Users Guide: Statistics. Version 6. Cary, NC: SAS Institute, oSanteaxkpeerKi,mMeuntkaalipRa,ncKraetatoicY,caUrmceiynoammaainK.theEfSfyercitasnogfocledreunlehianmsotnert.he Pnaonrcmraelaspa1n:c2r4e6a-s25a3n.d Sibinski inrats. LJ. St. PTauwlo,-yMeNa:rRiokraelr(dLiaebto)rtaotxoirciietsy,/c1a9r8c7i.nogenicity studyoffluorochemical FC-143 cShionlgehcyPs,tTokoiwnnisnernedceCptMo,rsU(pCpCKJ,-rL)ariindjnaonrimAa,l TahndomcpasncoenroJuCs. hCuhamraanctpearniczraetaiso.n oFfed Proc 45:291. 03575 PRE 3M CEPoLm0p0a0n3y Page 42042 `cShmairtahctJePr,izLaituioGn,oSfoCuCnKd-aBr/agraajsatnriVn,rMecceLpatuogrhsliinnhPuJm,aZnagpaonncrIeS.atiIcdecnatnicfiecraicetlolnlainneds. Am J Physiol 266:R277-R283. Stewart ID, Flaks B, Watanapa P, Davies PW, Williamson RC. Pancreatobiliary diversion `enhances experimental pancreatic carcinogenesis. BrJ Cancer 63:63-66. TdaekmojnasdtVraDt,ioFnoorfttuhnreeKePd,ifPfoelrleontDAst,atSeshoafhtGheN,paWncarneaktiWcAc,hoGleacrysdtnokJeiDnrin(1r9e9c4e)p.torD.irePcrtoc Natl Acad Sci 91:1868-1872. Taves D. Evidencethatthere are two forms of fluoride in human serum. Nature 1968:217:1050-1051. / Taves D. Electrophoretic mobility of serum fluoride. Nature 1968;220:582-583. Vanden HeuveJl, Kuslikis B, Van Refelghem M, Peterson R. Tissue distribution, `metabolism and elimination of perfluorooctanoic acid. J Biochem Toxicol 1991:6:83-92. van Kranen Hl, Vermeulen E, Schoren L, Bas J, Woutersen RA, van lersel P, van Kreijl CSFyr,iaSnchhearmesrteErs(,19b9u1t).is Aabcsteinvtaitinopnaonfcrce-aKt-ircams oisrfrseoqfureanttsi,n pCaanrccrienaotgiecnceasricsi1n2o:m1a4s7-o1f482. `Wank SA, Pisegna JR, deWeerth A (1994). Cholecystokinin receptor family. Ann NY Acad Sci 713:49-66. Yiinen M, Koho A, Hanhijarvi H, Peura P. Disposition of perfluorooctancic acid in the rat afer single and subchronic administration. Bull Environ Contam Toxicol 1990;44:46EX Zucker KA, Adrian TE, Bilchik AJ, Modlin IM. Effects ofthe CCK receptor antagonist 364,718on pancreatic growth in adult and developing animals. Am J Physiol 257:GS11-GS16. 03576