Friday July 24, 1981

Part IV

Department of Health and Human Services Public Health Service Food and Drug Administration

[Docket No. 75F-0355] Aspartame; Commissioner's Final Decision 38284 Federal RegIster I Vol. 46, No. 142 I Fnday, July 24,1981 I Notices

Table of Contents

I. Introduction. A. The Product. B. Histoncal Chronology.- II. Statutory ReqUIrements for Approval of a Food Additive Petition. III. Summary of DecIsIon. A. The BraIn Damage Issues. 1. PhenylalanIne. 2. Aspartic ACId. B. The BraIn Tumor Issue. IV EVIdence on the BraIn Damage Issues. A. Projected Consumption Levels. B. PhenylalanIne. C. Aspartic ACId. V EVIdence on the BraIn Tumor Issue. A. Introduction. B. Background Rate for Spontaneous BraIn Tumors In Charles River CD (Sprague-Dawley) Rats. C. Studies on Aspartame and.DKPoi D. ConclusIon on BraIn Tumor Issue. VI. Mr. Turner's Appeal. VII. Conditions of Use. VIII. ConclusIons. IX. Order.

Appendix A-Board's DeCISIon on Potential BraIn Damage from PhenylalanIne.

Appendix B-Board's DeCISion on Potential BraIn Damage from Aspartic ACId. Federal Register I Vol. 46, No. 142 I FrIday. July 24.1981 I Notices 38285

DEPARTMENT OF HEALTH AND foods consumed as part ofany healthful • to restrict carefully theIr phenylalanine HUMAN SERVICES diet When phenylalanme andaspartic mtake (just as diabetics need to restrict aCId are combmed m a certam way to theIr sugarllltake). The second Public Health Service fonn aspartame, they produce an conditionfor approval was thatwhen llltensely sweet tasting substance. aspartame was to be used as a tabletop Food andDiug Administration approxunately 180 times as sweet as sweetener. its label was reqUIred to bear [Docket No. 75F-0355] sucrose. Accordingly, as a sugar Instructions not to use aspartame in substitute, the amount ofaspartame cookmg'or bakmg. 'This 15 because Aspartame: CommISSioner's Final needed to produce the same degree of aspartame breaks downto DKPwhen DecIsion sweetness IS substantially reduced. as exposed to prolonged heat. with a \vill be the resulting calones. consequentloss ofsweetness. F'mally. i£­ AGENCY: Food and Drug Admimstration. Aspartame does breakdown a food contammg aspartame purported ACTION: Notice; final decision followm8 spontaneously to dikeloplperazine to be. orwas represented. for speCIal a hearmg before a public board of (DKP). Ifpresent m large amounts, DKP dietary uses. as mIght be expected ofa mqwry. can make aspartame lose its sweetness. Jow calone product. itwas required to Under the uses approved m thIs SUMMARY: The COllUIllSSloner ofFood be labeledlll compliance with FDA's deCISion, however. DKPnormally and Drugs IS ISSumg hIs Final DeCISion speCIal dietaryfoads regulations (21 comprIses less than 2% ofthe final concernmg the food additive petitionfor CFR Parf105). aspartame product whIch does not the nutritive sweetener aspartame 2. Objections to FDA Approval: As detract from the product's sweet tastes. submitted by G. D. Searle & Co. The permitted bylaw-(21U.s.C. 348[£)(1)). COllUIllSSloner has determmed that B. HistOrIcal Chronology two parties formally objected to the regulation on safety grounds and aspartame has been shown to be safe 1. InitialFDA Aspartame Approval. requested a formal evidentiary hearing for its proposed uses as a food additive was discovered and formulated by D. G. (21 CPR Part12). These parties were and IS approVlllg_the petition. SearTe &Co. (Searle). SkokJe. Ill. As the John W. Olney. and jomtly"James Specifically, the COllUIllssloner finds law reqUIres for all food additives. M.D.. S. Turner. Esq.• andLabel, Inc. (Legal that the available data establish that Searle petitioned the Food and Drug Action for Buyers' Educationand there 15 a reasonable certamty that AdmmlStration (FDA) forappro'lt'al to Labeling). Dr. OIney. then asAssociate human consumption of aspartame: (1) At market aspartame as a sweetenmg agent Professor ofPsychiatry atthe projected consumption levels, \vill not m certam foods (38 FR 5921. March 5. Washmgton Umversity School of pose a nsk ofbram damage resulting m 1973). Searle's petition contaIned Medicme, StLows. Mo. (now mentalretardation, endocnne volummous amounts ofdata purporting disfunction, or both; and (2) will not to establish the safety ofaspartame. Professsor). hadperformedresearchin cause bram tumors. Accordingly. the On July 26.1974. FDA approved animals regarding the tOXlC effects on Initial DeCISIOn ofthe Public Hoard of Searle's petitionand l5sued a regulation thebram ofcerttun. Ammo aCIds., Inqwry IS affinned m part and reversed authonzmg the use ofaspartame m mc1uding aspartic aCId. Mr. Turner;. a m part, as modified and supplemented certam foods and for certam lawyer.representedhimselfandLaheJ. herem. technological purposes (39'FR 27317i Inc.. a consumer-oriented group EFFECTIVE DATE: October 22, 1981. correctionnotice, 39 FR 34520. Sept 26. concerned about the regulation of cheInlca1s mfoods. Bothparties ADDRESS: The transcript ofthe hearmg~ 1974}.That regulation became codified objected pnmarily to the use of eVIdence submitted, andall other m 21CFR 172.804. Aspartame was aspartame by asserting that documentslisted m tb.lS decisIOn may be specificallyapproved for useas a children. the product mIght cause braindamage seen m theDockets Management Branch sweetenerm the followmg-foods: resulting fn mentalretardatioD. (fonnerly theHearmg Clerk's office) a. Dry. free-flOWIng sugar substitutes endocnne dysfunction. or both. (HFA-305), FoodandDrug for table-use (not to lllclude use In Adm1lllstration, Rm. 4-62, 5600 Fishers cookmg) m package units. not to exceed These parties later waived their right Lane, Rockville. MD 20857, from 9 a.m. the sweetenmg equIvalent of 2 to a formal evidentiary hearmg to 4p.m., Monday through Foday. teaspoonfuls ofsugar. conditionedupon the establishmenlo!a Public Board ofInquiry f'Boardj FOR FURTHER INFORMATION CONTACT: b. Sugarsubstitute tablets for sweetenmg hotbeverages. mcluding consisting ofthree qualified SCIentists Ted Hennan, Regulations Policy Staff from outSide the agency (21 CFRPart (HFC-I0), Food and Drug coffee and tea. 13). This wouldbe the first time FDA Adm1lllstration, 5600 Fishers Lane, c.Cold breakfast cereals. d. Chewing gum. bad ever used this alternative Rockville, MD 20857, 301-44~480. e. Dry bases for: (i) Beverages; (Ii) procedure. Searle agreed to delay SUPPLEMENTARY INFORMATION: The mstant coffee and tea; (iii) gelatins. marketing ofaspartame temporarily. purpose of thIs proceeding has been to puddings andfillings; and (iv) drury pending resolution ofthe safety deCIde whetheraspartame has been products and toppmgs.In chewmggum, questions. shown to be safe under section 409 of aspartame was also approved for use as 3. Audit ofSearle Smdies. Hefore a the Federal Food, Drug, and Cosmetic a flavor enhancer m addition to use as a Boardcouldbe convened, prelimmary Act (the act) (21 U.S.C. 348). sweetener. results from anauditofthe records of I. Introduction ThIs approval had three conditions certam arumal studies conductedby or regarding finalproductlabeling. First, for Searle. mcluding studies on A. The Product the label ofanyfood contllUllng aspartame. mdicated,a needfor a Aspartame [L-aspartyl-L­ aspartame was reqUIred to bearthe compreheD.Sl1:e reVIew of the phenylalanme methyl ester] 15 a followmg statement! authenticity of the aspartame research dipeptide composed prunarily of two "PHENYLKETONURICS: CONTAINS data. As a result, pursuant to 21 U.s.c. ammo aCIds. phenylalanme and aspartic PHENYLALANINE." ThIs requirement 348(e). FDA formally stayed the aCId. These, alongwith otherammo was deSigned to alert persons who. regula tion authorizmg the marketingof aCIds. ar~ nonnal constituents of protem because of specific health reasons, need aspartame (40 FR 56907. Dec. 5. 1975). ~ 38286 Federal RegIster I Vol. 46, No. 142 l FrIday, July 24, 1981 I Notices

With the knowledge and approval of 3. Based on answers to the above 29, 1981, thus making the Issue ripe for Searle, the aspartame data iii. 15 pIvotal questions. the final agency deCISIon studies were thoroughly audited to (a) Should aspartame be allowed for use 10 determme theIr authenticity. Three of foods. or. 10stead should approval of n. Statutory Reqmrements for Approval aspartame be withdrawn? of a Food Additive Petition these studies were audited by FDA and (b) Ifaspartame IS allowed for use In foods. 12 by the Umversities AssoCIated for I.e.• if its approval IS not withdrawn. what Section 409 of the act (21 U.S.C.340), Research and Education m Pathology, conditions of use and labeling and label sets forth the statutory reqUIrements for Inc. (UAREP). ThIS was a maSSI're statements should be reqUIred. if any? approval of a food additive 4 polition. undertaking and took over two years to (44 FR at 31717) With the enactment of the Food complete. UAREP concluded that the In the Federal RegIster of January 14, Additives Amendment in 1958, Congross studies were authentic and, on 1980, FDA announced the time and place established a premarket approval becember 13, 1978, submitted its 1,062 of the hearmg (45 FR 2908). The Board system whereby Ule company soeklng to page report to FDA (Vols.ll0, 111 and heard 3 full days of testimony, pnmarily market a food addit~ve must first obtain 112).1 The agency agreed with UAREP from Dr. Olney and representatives from approval from the FDA. GThrough this that those 12 studies, as well as the the Bureau of Foods and Searle. 2 The mechamsm Congress sought to shield three studies whICh it had reVIewed, hearmg dates were January 30 and 31 the public from unsafe or potenlially were mdeed authentic. FDA then turned 'and February 1, 1980. Post-heanog bnefs unsafe products. its attention to arrangIng the public and!or rebuttal statements were filed by hearmg. Section 409(c) (3) of the act, 21 U.S.C. Dr. Olney, the Bureau, and Searle. 348(00)(3), directs FDA not to approvo a 4. Establishment ofPublic Board of 5. The Board's DeCISIOn. On October food a:dditive petition: InqUIry 1, 1980, the Board Issued its declSlon. The Board agreed with the Bureau and • * • If a fou' evaluation of the data boforo Dr. Olney, Searle, and FDA's Bureau Searle on the first Issue, finding that the Secretary !t­ of Foods (the Bureau) all submitted aspartame consumption would not pose (A) fails to establish that the pl'oposed usa nommees for Board membershIp to then an mcreased nsk of bram damage, of the food additive, undor the conditions of Acting CommIssIoner Sherwm Gardner resulting m mental retardation, use to be speclfiod In the regulu lion, w1ll bo who chose the followmg panel: Walle J. endocrme dysfunction, or both. safe: * * * H. Nauta, M.D., Ph. D., Institute However, the Board agreed with Dr. ThIS provlSlon 10 the law. known as tho Professor, Department of Psychology Olney on the second Issue, finding that "general safety clause," is thoroughly and Bram SCIence, Massachusetts the available data on laboratory rats did analyzed m the CommIssioner's Institute of Technology; Peter J. Lampert, not rule out the possibility of DeclSlon on Cyclamate (Cyclamate M.D., Professor and ChaIrman, aspartame's causmg bram tumors, and DeCISIon) (45 FR 61474, 61476-77, Sopt; Department of Pathology, University of that, mdeed, the eVidence suggested that 16,1980). Two pomts of that discussion Califorma (San Diego); and Vernon R. aspartame mIght mduce bram tumors. warrant repeating here. Young, Ph. D., Professor of Nutritional Accordingly, the Board concluded that First, by requmng that tho data BiochemIstry, Department of Nutrition aspartame should not be approved for "establish" safety, Congress clearly and Food SCIence, Massachusetts marketing until further anImal testing placed the burden of proving safety on Institute of Technology. Dr. Nauta was was conducted to resolve the bram the sponsor of a food additive pelition, named chall'man. tumor Issue. Because of the Board's in thIS case Searle. FDA does not have As the Issues for the hearing were finding on the bram tumOr1ssue, the to prove that the product IS unsafe. This bemg framed, Dr. Olney raIsed an Board withdrew approval of Searle's distinction IS very important because it additional concern about aspartame's food additive petition and, after IS possible that the data may fa11m the potential to cause bram tumors. vacating the stay on the aspartame "grey area" where the food additive has Although the Bureau disagreed with Dr. regulation (21 CFR 172.804), revoked that not been shown convincingly eithor to Olney's assessment, then CommISSIoner regulation m its entirety. be safe or unsafe. In such a situation Kennedy agreed to add thIS Issue to the Dr. Olney, Searle and the Bureau all further testing may be necessary to hearmg agenda (see letter to Dr. Olney, filed detailed exceptions to those resolve the Issue. ThIS was the agency's dated November 17,-1978, Vol. 120 portions of the Board's deCISIOn m position on cyclamate (45 FR at 61471. [correspondence filed chronologIcally]). wluch the Board disagreed with theIr col. 3). Similarly, Dr. Olney and Mr. On June 1, 1979, FDA announced the respective positions (21 Cl)R 12.125(a)) 3 Turner contend that the data on establishment of the Public Board of (hereafter "Exceptions"). Mr. Turner aspartame fall mto this "grey area" Inqmry to help resolve the issues also filed exceptions. objecting to the whICh would reqUIre further testing surrounding the proposed marketing of scope of eVIdence conSIdered by the before marketing. aspartame (44 FR 31716). These Issues Board. Searle and the Bureau each filed The second essential pomt in were defined, m relevant part, as replies to both Dr. Olney's and Mr. mterpreting the general safety clause is follows: Turner's exceptions (21 CFR 12.125(c)) the meamng of the term "safe." 1. * * • whether the Ingestion of (hereafter "Replies"). Under the aspartame, either aalone or together with established time fraines, the 4The tenn "food additive" Is defined In 21 U.S.C. "'­ glutamafe, poses a risk of contributing to admmlstrative record closed on January 321(s}. There Is no question that aspart!lme Is a food mental retardation. bram damage, or additive. undeslreable effects on neuroendocrme 2 Mr. Turner made only bnef presentations (see GAny product contaming an unapproved food regulatory systems • * * Tr./ll/pages 187-200 and Tr./II1/pages 237-39). Two addItive IS automatically deemed adulterated 2.' • * whether the mgestion of additional heanng partiCipants. Richard J. Wurtman. therefore unlawful. 21 U.S.C, 340(a). aspartame may 10duce bram neoplasms M.D.• and lloyd J. Filer. Jr•• M.D.. also addressed the oThls decision has beon delegated 10 the FDA Board. as did two consultants to the Board. William Commissioner. 21 CPR 5.10(U)[1) (formerly secllon (tumors) 10 the rat * * * Nyhan. M.D. and Milton Bnghtman. M.D. 5.1(aJ[1)) and Is not subject to the Secretory's 3 Although the Board's heanng procedures are set reservallon of authority under 5 CPR 5.11 or 1 All citations to matenals m the adnllOlstrative out In 21 CPR Part 13. procedures followmg issuance Executive Order 12291 bocauso decisions on food record refer to the filing system m FDA's Dockets of the Board's deCISion are detennmed by Subparts additives are subject to 5 U.S.C. 550 lind 657 (40 FR Management Branch (HFA-305). G and H of 21 CPR Part 12. See 21 CPR 12.32(f)(3). 13193. Feb. 19. 1981. and 46 FR 26052. May 11.1001). Federal Regtster I Vol. 46, No. 142 I FrIday, July 24. 1981 I Notices 38287

Although not defined m the statute possible "bram damage" (tOXIcity). shown to be safe for its proposed uses. itself, FDA regulations clearly reflect the concerns the fonner category whereby My reasons-for thiS conclUSion, detailed legislative hIstory. by stating that safety aspartame may be marketed so long as in Sections IV-VlI below, may be means: the prOjected consumption levels fall summnnzed as follows. suffiCiently below the estimated toXIC * * * there IS areasonable certamty m the A. The Bram Damage Issues mmds of competent sCientists that the threshold. In contrast, with respect to substance IS not harmful under the Intended the second Issue m thIs proceeding, The fIrSt set of objections to the conditions of use. relating to possible "bram tumors" aspartame regIIlation concerned two 21 CFR 170.3(i) (emphasls added). (carcmogemcity), aspartame must be distinct types of bram damage, one CongresslOnal reports show that the shown to a reasonable certamty not to assocJated with each of aspartame's two legislators were particularly IDlpressed cause bram tumors at all, for food ammo aCid components, phenylalamne by expert testimony emphaslZmg the additives producmg carcmogemc effects and aspartic aCid. The Board disagreed ·at any level are deemed to be unsafe per with these objections and found IDlpossibility of provlliing, withm the s bounds of SCientific knowledge, the ·se. aspartame to be safe m terms of absolute harmlessness of any cheffilcal In summary, the general safety clause potential bram damage. places on Searle the burden of provmg Two pomts stand out whIch reqUIre _ substance. H.R. Report 2284, 85th Cong., that the data m the adnurustrative 2d Sess., pp. 4-5, 1958; Senate Report affimung the Board's deCiSion on these record establish that there lS a bram damage Issues. One IS the 2422, repnnted m (1958) U.S. Code Congo reasonable certamty that aspartame will and Admtn. News 5301. 7 Congress enonnously large amounts of aspartame not be harmful under the prescribed that a nonnal person would need to therefore advocated the more realistic, conditions of use. Only ifSearle meets yet still ngorous, standard of reasonable consume before reachIng even a thIs burden can the food additive cautiously estimated toXIC threshold. certamty of no harm, later embodied m petition be approved. FDA's regulation quoted above. The second IS the remarkably low The statute leaves the methods and m. Summary of DeClSloDs amount of ammo aCId mtake whIch critena for mterpreting data up to the The purpose of thIs proceeding has would result, from even the 99th discretion and expertise of the agency. been to deternune whether aspartame percentile of estimated aspartame Congress did, however, direct FDAlo has been shown to be safe under consumption. m relation to the conslder·the followmg three factors: Section 409 of the Federal Food, Drug, prevalence of these same ammo aCids m (A) The probable consumption of the and Cosmetic Act, 21 U.S.C. 348. The common protem foods. additive and of any substance formed m or legal standard for approvmg Searle's 1. PhenyJalamne: The concern that on food because of the use of the additive: Food additive petition is whether there has been raised over aspartame's (B) The cumulative effect of such additive IS a reasonable certamty m the minds of phenylalamne (PH£) mOletyls that In the diet of man or anImals, taking mto component SCientists that aspartame sus tamed plasma-PHE levels above a account any chemIcally or pharmacolOgically certam toXIC threshold may cause related substance or substances m such diet; will not be harmful to the public under and its proposed. conditions of use. What is mental retardation, espeCIally m the (e) Safety factors whIch m the opInIon of reqUIred of the agency, therefore. is the unborn fetus, sIDlilar to that resulting experts qualified by sCientific trammg and consClentious exercise of pnnclpled, from phenylketonurIa (PKU). expenence to evaluate the safety of food SCientific Judgment. As CODIDllssioner of The toXIC threshold for plasma-PHE additives are generally reCOgnIZed as Food and Drugs, my responsibUity IS to levels is 100 mtcromoles per deciliter appropnate for the use of arumal reView the eVldence and evaluate it bunol!dl) for nonnal persons. mcluding expenmentation data. frurly, state my reasons for crediting or infants, and 50 JUIlol/dl for pregnant 21 U.S.C. 348(c)(5). In the case of not crediting certam eVidence, weight all women in order to protect their aspartame, the product's mass the eVldence, apply the correct legal fetuses.1o Nonnal plasma PHE levels marketing potential and expected standard, and deCide. range from 6 to 12 JUI10l/dl. Ingestion m consumption by persons of all ages, ThIs I have done, and I have a smgle sitting by an adult of a loading espeCially children, are aspects that concluded that aspartame has been dose of aspartame, comparable to the have been consldered m the safety 99th percentile of proJected aspartame evaluation. •But see Monsanto v. Kennedy. 613 F. 2d M7, S55 consumption for an entire day, cansed The general safety clause applies to (D.c. Cir. 1979), which Indlcaled that the ogl!llCY has plasma PHE levels to nse from a fasting discretion 10 delermine that the qunnUty of 11 all types of health nsks. For exampl~, polentially carcinogenic substllllcc found In 11 food level of 6 ILmol/dl to a peak of only 11 the proVlslon was recently applied to "may be 50 negligible as to presl!llt no public health ILmol!dl, still withm the nonnal range - both carcmogemcity and mutagemcity or safety concerns •••" after eating and nowhere close to the (45 FR 61474). Cf. Ethyl Corp. V. E.P.A., Itshould also be noted thai another portion of In climcal section 409, dealing specificallY with carcinogenicity toXIC threshold. testing it took 541 F. 2d 1 (D.C. Cir.) (en banc), cerl. (the so-called Dellllley Clouse), Is noloppllcable In approxtnlately SIX times that amount to demed, 426 U.S. 941 (1976) (analogous !Ius proceeding. Thai provision prohibits tho induce plasma-PHE nse to the 50 JUI10I/ statute applied to lead POlSOmng). approval of any food addltiva petition whero the dllevel. For a 60 kilogram adult [132 additive has been shown conclusively 10 be Food additives presenting health risks carcinogenic [see gencrolly, Cyclamate Decl5lon. 45 pounds), thIs corresponds to 600 may be diVlded mto two categones for FR at 61476, coL 3J. As noted obo\'o. however, those aspartame tablets or 24 liters safety evaluation purposes: (1) Those opposing approval of aspartame do so on the [approximately 61h gallons) of whIch are safe at or below certam levels grounds thaf the dolo on aspartame fall Into the aspartame-sweetened beverage "grey area" (Lo., safety hils not been demonstrated), but unsafe at other, hIgher levels; and (2) not that aspartame Is conclusively carcinogenic. consumed m a smgle sitting. Such an those whIch may be unsafe at any level. 'This Sumnl!uy'was made publicly avnlloble on Theanalysls for-tliese two categones lS July 15, 1961 as a vehicle for nnnounclD8 the ~1he dUfcrcnce for fetuses Is that the placenta necessarily different For example, the decision as soon as possible. without awolting maintains 111:2 ratio gradient between the maternal publication of the enliro Fmal Decision In the I1I1d fetal c:lrcuIaUon In the plasma PHE first lssue m thIs proceeding, relating to Federnl Rcglster. Minor changes or deletions ha..-o concentrations {BoOId's Decision at 13}_Thus. a been mode 10 conform to or ovoId unneccssmy plasma PHElevel of 50 l'IIIol/ellln an expectant 'These CongreSSIOnal reports are quoted at redundancy with other portions of this Final mother aeates a plasma PHE level of 100 l'IIIol[ell length m 45 FR at 61477, coL 1. Decision, or to correctlypographlca1 enol'll. for her fetus.. -38288 Federal Register I Vol. 46, No. 142 I Fnday, July 24, 1981 I Notices enormous mtake at one time, let alone levels above the toXiC threshold can the level mduced by the first dose" (id. continuously over a sustamed perIOdi IS cause bram damage, unlike the case at 37). mconcelvable. Thus. it appears that wiili phenylalanme tOXicity where a Finally, as was the case the consumption of aspartame In reasonable sustamed high plasma level IS needed. phenylalanme component. the ratio of amounts, or even m unreasonable but The Board established the plasma prOjected aspartic aCId consumption physIcally possible amounts, will not GLU+ASP level oflOO Jtmol/dl as the resulting from aspartame to that derived cause the type of bram damage of tOXiC. threshold for rIsk assessment fl:om a normal diet is quile small. For concern here (see generally, Board's ·purposes. Tills was an estimate, and an example, m the age group of most DemsIOn at 13-15). extremely conservative one, based on concern, young children, consumpUon of I also agree with the Board's experImental findings m the most aspartame at the 99th percentile level of conclusIon that the marketing of sensitive speCIes at the most sensitive prOjected consumption (34 I mg/kg/day) aspartame will not create any additional age (the mfant mouse). Even with this will only mcrease total aspartic acid rIsk to PKU children not on a restrIcted cautious approach, the data clearly consumption by apprOXImately 4%, diet, mdividuais heterozygous for PKU, establish safety for antiCIpated clearly an mSlgnificant amount. undetected cases ofPKU, or pregnant aspartame consumption. 11 The conclusion compelled by thase women with the spemal condition of Climcal testing m adults usmg rugh fmdings IS that the addition of hyperphenylalanenua (see Board's aspartame doses (eqwvalent to 1 % aspartame to the diet, in any DemsIOn at 15-20). times those at the 99th percentile of conceIvable amount (far beyond the Another way to consIder the prOjected daily consumption), upper prOjected consumption levels), phenylalanme Issue IS to compare the adnumstered at a smgle sitting, showed will not cause focal bram lesions of the prOjected aspartame consumption to the no SIgnifIcant rIse m either the plasma type alleged by the objectors to the ASP or GLU concentrations. Even with amount of phenylalanme present In aspartame regulation. common protem products.l"or example, an enormous aspartame dose consumption of aspartame at the (eqwvalent to six times that at the 99ili B. The Bram Tumor Issue prOjected 99th percentile level (34 mgt percentile of prOjected daily Tills was the Issue on which the Board kg!day) would mcrease the normal consumption, or 600 to 800 aspartame disagreed wiili ilie Bureau of Foods and overall PHE daily mtake by only about 6 tablets) adnumstered at a smgle sitting, concluded that furthel: tesling was percent. Even consumption of aspartame the plasma GLU+ASP level rose from necessary before aspartame could be at the unlikely level of three times that 2.7 Jtmol/dl to only 7 Jtmoljdl, still marketed. With due respect to the prOjected 99th percentile level would withm ilie nonnal range found after Board,'I agree with the Bureau of Foods mcrease the normal overall PHE daily eating. A further study usmg 1 year old that the data presented at the hearIng mtake by only 15 to 20%, still withm mfants showed slIDilar mmor rIses m establish that there is a reasonable expected, normal varIations m protem plasma ASP+GLU levels. Finally, other certamty. that aspartame does not cause consumption (Board's DeCISIon at 20­ studiesm adults showed that mgestion braII1 tumors m laboratory rats. This 21). Thus, from the standpomt of of aspartame (eqwvalent to the 99ili conclUSIon is confinned by addifional phenylalarune mtake, aspartame percentile of prOjected daily eVIdence submitted after the Board appears to present no greater hazard consumption) did not further mcrease Issued its deCISIOn. elevations of plasma GLU+ASPlevels than common protem rIch foods 1. Spontaneous 111cldence rate 0/ brain conSIdered essential for proper nutrition. caused by the adnumstration of very tumors. The most controversIal issue Ilt 2. Aspartic Ai::Id:'I'he concern raIsed hIgh doses ofMSG alone (see generally, the hearmg was whether a significant over mcreased aspartic amd (ASP) Board's DeCISIOn at 32-33).12 disparity eXIsted between the brain consumption stems from anImal studies The Board also addressed the rIskto tumor mmdence rates as reported in tho the PKU heterozygote, the nursmg showmg that extremely hIgh doses of Searle studies m a certmn stram of rat mfant, and the unbomfetus, and ASP, glutanuc aCId (GLUJ, and other and the spontaneous mcidence rate (or concluded that these groups were at "excitatory" ammo amds can cause background rate) for bram tumors in this focal b}.'am leSIOns. prImarily m areas of least as safely protected as normal stram as repol:ted m the scientific adults or. children (id at 33-34). I agree the bram that regulate the endOCrIne literatUre. The Board found that such a system. With two Important differences with these conclUSIons also. An additional pomt worth noting IS disparity eXIsted, and that the disparity described below, the analYSIS parallels that the plasma ASP+GLU levels that was so great as to preclude the key that In the phenylalarune section m were observed were short-lived, Searle rat studies (E-33/34 and E-70) terms of fIrst setting a tOXiC threshold receding to theIr baseline value after from prOVIding adequate evidence qf and then determmmg whether the three hours. Thus, as the Board aspartame's safety (Board's Declmon III prOjected consumption of aspartame eJWlamed, "repeat-doses of the same 43-45). The Bureau of Foods disagreed will keep plasma levels suffiCIently enormous magnitude, when spaced 3 with the Board, believing that I:ellable below that threshold. The fIrst hours apart, are unlikely to escalate the data m the record established a difference IS that it IS the combmed GLU +ASP concentration much beyond spectrum of reported spontaneous brain plasma ASP+GLU levels whICh must be tumor inCidences that encompassed the 'scrutimzed, both because adnumstration rates reported In the Searle studies. IIIn contrast to the analysis m the phenylalanine of either GLU or ASP mcreases plasma section. the above analYSIS does not set the toXiC I agree with the Bureau's assessment levels of both ammo amds, and because threshold for pregnant women [for protection of the. of the background rata for brain tumors the two amino acids are eqwpotent and fetus) as half that for normal mdivlduals because. in the pertinent stram of rats. Although mutually additive m producmg the the placenta forms an effective bamer a8amst the transfer of both ASP and GLU to the fetus [Board's the Board placed considerable weIght on leSion (see Board's DemSIon at 22':"23f, DeCISIOn at 34). .< published studies reporting spontaneous GlutamIC amd is prevalent m the food 12Dr. Olney has asserted that an additional study bram tumor inCIdence rates of less than supply, often as the food additive m children is necessary to measure the effects of 1% (.09%, .6%, and .7%), these studies all monosodium glutamate (MSG). The aspartame admmlStered m conJunction with MS(}' For the reasons discussed 111, Section JV(C)[3)[d) had some flaws and, in addition, must second difference IS the SCIentifIc belief below. I do not believe sucn an additional study IS be supplemented by other data that a SIngle surge of plasma GLU+ASP necessary. presented at the heanng reporting higher Federal Regtster I Vol. 46, No. 142 I FrIday. July 24. 1981 I Notices 38289 spontaneous mCIdences (e.g., 2.2% "and concluded recently 10 Japan and public health ramifications. These 3.2%) wInch are consIstent with those m submitted mto the record after the Board sCIentific Issues were debated the Searle studies. One reason for gIvmg Issued its deCISIon, also appears to be VIgorously at the hearmg, and the Board weIght to these studies reporting Ingher negative m terms of bram ttmlors. performed admIrably m mamtammg a mCIdences IS that the chances are Although tlus study has not been JudiCial decorum and m crystalizmg its consIderably greater that additional critiqued by the hearmg participants. the vIews of the Issues m its InItial DeCISIOn. tumors may have been mIssed m the data on theirIace proVIde additional I would be remISS ifI did not express to 10w-mCIdence studies than that tumors support for roy conclUSIon on tlus Issue. each of the Board members the were mcluded by mIstake or 3. ConcluslOn on broIn tumors. The apprecIation of both the agency and the mIsdiagnosed 10 the hIgh-mcidence available data. VIewed as a whole. public for the mvaluable servtce wInch ones.13 establish that aspartame is safe m terms they performed. ofbram ttmlors for its proposed uses. Of speCIal sIgIlificance IS the IV EVIdence on the Bram Damage reference reporting an mCIdence of 2.2%. C. Conditions ofUse Issues These data were collected by the National Cancer Institute [NC!) from its The labeling conditions set forth m the The fIrst Issue at the hearmg was as carcmogenesis bIoassay program.,The aspartame regulation (21 CFR 172.804) follows: before it was stayed shall still be partiCIpating orgaruzations were NCI The question has been l'3JSed whether the and Hazelton Laboratones, the same reqUIred. These mclude a prommently ingestion of aspartame, either alone or -laboratory used by Searle for its key displayed alert to persons with PKU that together with glutamate. poses a ns.\ of aspartame rat studies. Moreover, the the product contams phenylalamne: contribuUngto mental retardation. bram rats used for the NCI data were all directions not to use aspartame m damage or undesirable eIIects on the control anImals of the same stram and cookmg or bakmg because the neuroendocnne regulatory systems. From commerCIal source used by Searle, and compound loses its sweetness when available evidence. what can be concluded m the SIZe of NCI's sample population was exposed to prolonged heat: and labeling relation to t1us question? The objecting m compliance with FDA's speCIal parties believe that the mgestion of nearly Identical to the control groups aspartame, either alone or together \vith (combmed) m Searle's rat studies. The dietary foods regulations (21 CFR Parl 105) where appropnate. In addition. glutamate, does pose a nsk of contributing to reported spontaneous ·mcidence rates these eIIects. The Bureau of Foods believes were also nearlY-identical: 2.2% (B/368) because the safety assessment on the that the ingestion of aspartame, either alone for the NCI data and apprmamately 2% bram damage Issues IS tied closely to or together \vith glutamate, does not pose a (7/356) for the combmed control groups prOjected aspartame consumption risk of contributing to these effects. levels. as a condition for approval 10 the Searle studies. " Searle IS to monitor the actual use levels (44 FR at 31717). The Board considered 2. Comparlson with concurrent tlus Issue m two parts, one relating to controls. Given the consIstency between of aspartame and to prOVIde such mformation on aspartame's use to the the phenylalanme component of the control mCIdence rates 10 the Searle Bureau of Foods as the Bureau may aspartame. and one mvolvmg the studies and the background rate, I find deem necessary. aspartic aCId component. Aspartame's that Searle's studies should be two ammo aCIds are each aSSOCIated evaluated pnmarily by comparmg the D. Concluding Remarks with a different kmd ofbram damage. aspartame-treated anImals to their The safety evaluation of aspartame Only the aspartic aCId component concurrent controls. Usmg these _ has been a long and arduous process. interrelates with glutamate. The Board's compansons, as analyzed by the Bureau spanmng the tenure of several FDA subdiVISion IS followed m tIns deCISIOn. ofFoods, I find that Searle studies E-70 COmmISSIoners. Although my conclUSIon Before discussmg the specific bram and E-33/34 both are negative studies. IS the same reached by the agency damage Issues. however. itIS necessary It IS undisputed by the heanng nearly seven years ago. the mtervening to address the projected consumption partiCIpants that the E-70 study IS a years have not been without theIr levels of aspartame. negative study when the treatment benefits m terms of the eVidence groups are compared to the concurrent showmg the safety of aspartame. Much A. Projected Consumption Levels controls. The only controversy lies 10 of the data, espeCially climcal data. Because the non-toXIcity ofaspartame the E-33/34 study, where the Board relied upon by Searle at the public is based on safe levels of use, the -found a possible dose response and hearmg, came from studies conducted at projected estimates ofaspartame accelerated tumor onset, both potential the firm's behest durmg the mtenm. consumption are central to the safety mdicators of carcmogemcity. The Also takmg place durmg this penod was evaluation. Three methods have been finding of the dose response, however, IS the detailed mdependent audit of used to amve at these estimates. each of largely dependent on a smgle, very Searle's preclimcal data conducted by whIch attempt to exaggerate projected early-occurnng, unusual tumor (a the Umversities AsSOCIated for Research consumption levels 10 order to account medulloblastoma), wInch was probably and Education in Pathology. Inc. for potential heavy users ofaspartame­ not aspartame related, and the finding of [UAREP) and the agency. Few sweetened products. accelerated tumor onset was based 10 compounds have withstood such -The first method, used by the Bureau part on factual errors. For these reasons, detailed testing and repeated. close ofFoods. IS to assume that aspartame 15 as detailed m Section V below, I agree scrutiny. and the process through whIch substituted for allsucrose 10 the diet of with the Bureau of Foods that E-!Ja/34 IS aspartame has gone should proVIde the an average'60 kg man. In tIns situation. also a negative study. public with additional confidence ofits aspartame consumption would be Finally, a thtrd long-term study safety. approXlDlately B.3 mg/kg/day rrr./I/ assessmg aspartame's carcmogemc The pmnacle of tlus process was the page 60. line It-page 61.line-Z). potential usmg a different stram of rat, heanng befpre the Public Board of Although tlus figIIre IS based on the Inqmry. the fIrSt of its kmd to be needs of an average consumer rather t3Tlus WIder spectrum of reported spontaneous mcidence rates IS further supported by data convened. The SCIentific Issues than a "heavy user," tIns shortcOIDIDg IS submitted mto the record by Searle and tlie Bureau presented to it were mtellectually counterbalanced by the assumption that of Foods after the Board ISsued its deCISIon. complex and carried WIde rangIng aspartame would replace all sucrose m 38290 Federal RegIster I Vol. 46. No. 142 I Fnday. July 24. 1981 I Notices the diet. an unlikely event. Moreover. if Usmg a 70 kg adult. Dr. Stegmk the Board found that the cautionary aspartame IS substituted for all estimated tlus maXlIDum aspartame label proposed by the Bureau of Foods carbohydrates as well as all sucrose. the usage to be 23-Z5 mg/kg/day (Tr./I/ ("Phenylketonuncs: ContaIns average 60 kg man would consume page 94. line 1-page 95. line 3). Dr. Phenylalanme") would suffiCIently approximately' Z5 mg/kg of aspartame StegInk also calculated aspartame protE!ct these mdivlduals who are per day (Tr./I/page 61. lines 7-23). mtake for a 10 kg mfant. assummg accustomed to controlling carefully their Surely thiS would appear to be a lughly aspartame was substituted for all energy dietary mtake of phenylalamne (id nt exaggerated figure. reqUIrements now supplied by sucrose. 21). For the "unfortunate case" of the The second set of consumption and arnved at an estimated mtake of 19 pregnant woman who does not know estima:tes was based on data generated mg/kg/day (Tr./I/page 96. line 6--page she has hyperphenylalanemia. or for by the Market Research Corporation of 97. line 14).15 ff undetected cases of PKU children, the America ("MRCA ). submitted mto the Although these figures are only Board concluded that "the normal food­ admmlstrative record by the General estimates, the consistency of the figures denved PRE poses a much greater risk Foods Corporation (Vol. 103). The across different methods adds to the patient (or the unborn child) than MRCA has a large computer bank whICh significantly to therr credibility. In order tabulates actual dietary records kept by would aspartame, even when consumed to be as cautious as possible. the Board m large amounts" (I'd.). 4.000 households (approXlIDately 12.000 used the 34 mg/kg/day figure as the mdivlduals) over a 2-week penod. benchmark for use m the nsk I agree with the Board's conclusions staggered throughout the year. These assessment analYSIS. This was the and careful diSCUSSIOn of these complex estimates are based on what people m lughest figure obtamed from any of the Issues, and therefore adopt the Board's gIVen age brackets actually eat and are estimates and represented the 99th declSlon as my own. The relevant broken down mto different percentile percentile for all age groups from the portion of the Board's decision (pages levels. to account for both the average MRCA survey. I agree with the Board's 11-22) IS ;reproduced m Appendix A to and heavy users (Vol. 103. pages 1-3: use of tlus 34 mg/kg/day figure. More thiS. decISion. 11 Tr./I/pages 92-93; Tr./ill/page 105). Importantly. as detailed m the followmg Dr. Olney's raised two exceptions of One additional aspect of the MRCA sections. the non-toXicity ofaspartaIpe thiS Issue, The first exception relates to estimates IS that the survey covered two has been clearly demonstrated m all age the percentage of PKU children who aro groups of products: "Group A" products groups atlevels several times tlus 99th not diagnosed at birth (Olney's • were those m Searle's current food percentile figure. Exceptions at 1). The Board used a additive petition; and "Group B" figure of 10%, apparently relying on the B. Phenylalamne products mcluded seven additional testimony of Dr. Koch (Tr./II/page 11). categories for which aspartame has As noted m Section ill(A)(1) above, Dr. Koch testified that 10% of all PKU marketing potential. mcluding the Board concluded that the prOjected cases may be missed "due to the lack of carbonated soft dnnks. probably the level of aspartame consumption by a good quality program" (id.). Dr. Olney largest potential use (Vol. 103. pages Z­ normal humans "cannot be expected to asserts that these 10% are missed due to 3). InclusIOn of the Group B categones mcrease the mCldence of that particular the error mherent m the screening prOVides an extra "cusluon" for form of mental retardation that IS method, and that another group of PKU purposes of Searle's current petition. aSSOCIated with sus tamed elevation of babies (approximately ZO% of those The survey showed young children plasma-PRE levels" (Board's DeCISion at afflicted) are missed because they are (age 2-5) to be the largest consumers m ZO). This conclUSIOn also applies to among groups of mfants that had not proportion to' therr body weights. For fetuses, mfants. andmdiVlduals Groups A and B combmed m tlus age been scree~ed at all. Dr. Olney adds heterozygous for PKU (id. at 14-15). For these two figures and concludes that group. the mean potential exposure was mdiVlduals on a PRE restricted diet (i.e., 11.1 mg/kg. and the 90th percentile value PKU children and pregnant women 14 with this condition may give bIrth to brnln damnged was 25.0 mg/kg. The "under 2" and "6­ known to have hyperphenylalanem18) 16 babies If they do not keep themselves on a 101V lZ" age groups were the next biggest phenylnlamne diet. users. each with mean and 90th "The amount of aspartame that would be used In 17 One mmor modification to the Board's decision percentile aspartame exposure levels of specific products under conSideration. as supplied Is necessary. The change In no way affects the by the General Foods Corporation and Searle. are validity of its conclusIons. In varIous places, the approXimately 6 ancf16 mg/kg/day. as follows: Board uses as a benchmark. for comparison respectively. The other age groups, had a. Table Top Sweetener: 40 mg m a free flOWing purposes, the amount of ph enyin10nIne Intake that decreasmg exposure with age. with the packet; 20 mg m the tablet. These are eqUivalent to would be consumed by II 4 ounco hamburger.ll1to "25 and older" category havmg the l\vo and one teaspoons of sugar. respectively. phenylalanine content of II 4 ounte hamburger used lowest 90th percentile level of only 5.9 b. DryBeverage Mix (e.g., Kool Aidor Tang]: 120 by the Board (4,000 mg] was based on testimony by mg per S oz. glass. one of Searle's wltnosses. Dr. Koch (Tr./II/page 14. mg/kg of aspartame per day. (See lines 21-22). The figure, however. does not appear to c. Gelatin or Pudding: 32 mg per servrng [half detailed chart m Vol. 103. page 6.) For cup). be correct for a 4 ounce cookod hamburger. all age groups, the 99th percentile figure According to Geigy SClOnli/ic Tables (7th Ed., GelOY d. Whipped Topplngs:10 mg per servmg (two Pharmaceuticals (1970) at 516). based on tho was 34 mg/kg of aspartame on a daily heapmg tablespoons). percentage of protein In a cooked hamburger (24.2%) baSIS (Tr./I/page 93. lines 10-15). e. Breakfast Cereals: 90 mg per servmg (one oz. or and the percentage of that protein composed oC Finally, Searle's cluef witness, Dr. one cup). phenylalanine (4.2%J. the PHE content oC a 4 ounce Steglnk, calculated yet a tlurd set of f. CheWing Gum: Smg per stick. cooked hamburger Is approximately 1,l50 mo (not (Searle's Post-Hearmg Bnef. Vol. 155 at 12-13). 4000 018). Although estimates of phenylalanine figures (usmg a method Similar to that - These figures are In straight milligram amounts, content vary dependIng on whether the meat Is employed by the Bureau of Foods) by which need to be diVided by the weight of the considered as cooked. uncooked or dry Wel8ht, Ihe substituting aspartame for all consumer (in kilograms) for companson to the amount of phenylalanine In I\, cooked hamburger Is carbohydrate energy reqUIrements. estimates described above. the mosl appropriate comparison Cor theso purposos ,.Hyperphenylaianemia, as described in more because that Is what people actually eat. A slmllur mcluding those supplied by sucrose. detailed below in relation to Dr. Olney's exceptions, adjustment should be made for a' hot dog. EVen with IS a condition whereby a person'!! plasma PHE these changes. tho upper projected lovel of 1490th percentile means that m 9 out of 10 days an levels are higher than normal but lower than a­ aspartame consumption Is slllllolV when'compared IndiVIdual will have an mtake equal to or less than person with PKU. Although these indiViduals are to the amount of phenylalnnlne which would be the mg/kg figure. . . not themselves brain damaged. pregnant women derived from a proteln·rlch meal. Federal Register I Vol. 46, No. 142 I FrIday, July 24, 1981 I Notices 38291

30% of all PKU children m tlus country only remedy to tIDs problem, however: a context with a bnefbackground remam undiagnosed. problem wblch currently eXlsts wbether diSCUSSIon. I disagree. Although the exact number or not aspartame IS marketed, 18 first to 3. AnalysIs-a. Background. The type or percentage of PKU children who Identify the women wbo have this of bram leSion of concern bere IS one remam undiagnosed at bIrth was subject condition, and then put them on a wruch bas been studied m anImals over to some dispute at the heanog [compare pbenylalanme restrIcted diet just as one the past 12 years. It IS produced by rugh Tr./l/page 39. lines 1-16 with Tr./II/ would a child with PKU [id.). doses of glutamate. aspartate (either page 11. lines 1-8), Dr. William Nyhan, a For the reasons stated above and in glVen as aspartate perse or as consultant to the Board and an the Board's deCISIOn, I find that the data aspartame), and by other "excitatory" acknowledged expert on PKU, establisb that there. is a reasonable ammo aCids and theU' analogs. The emphaSIZed that nearly all PKU children certamty that the proposed use of leSion pnmariiy consists of dead or aspartame will not cause or aggravate who are not diagnosed at bU'th by a dymg nerve cells (neurons. IS The most routine screenmg test are nevertheless the type of diffuse bram damagE! sensitive region of the brain appears to diagnosed by 8-10 months of age by the assoc18ted with sustamed rugh plasma be the arcuate region of the claSSical diagnostic techmques [Le., due levels ofphenylalanme. to abnormal development) [Tr./l/page hypothalamus; other bram regions and 224. line 9-page 225. line 8). Dr. Nyhan C. Aspartic ACId the retina are also affected at rugher also emphaSIZed that an Infant with 1. Issue: The second tOXlcily Issue doses. The affect.ed areas of the PKU, whether diagnosed or not, still before the Board was whether the hypothalamus are mvolved m endocrme needs apprOXImately 90 mg/kg of expected consumption of aspartame, conraI, Vla the pituitary gland. Indeed. phenylalanme per day as an essential either alone or together with glutamate long-lasting endocrme changes have nutritional reqUll'ement rrr./l/page 229, [i.e., as MSG), poses a rISk to humans of been produced by admmlstration oflugh lines 6-10). These nutritional causmg focal bram leSIons [and doses ofMSG to neonatal IDlee and rats l'eqUll'ements. together with the fact that assoc18ted neuroendocrme changes) of (reVlewed m Vol. 126, Tab 67 (Olney. aspartame IS not bemg approved for the type wruch has been demonstrated ''ExcitotoXic AIDIno ACids: Reseal'ch Infant formulas or Infant foods and that m arumals after the admmlstration of Applications and Safety Implications" m PHE levels are not elevated m breast these substances. In addressmg tIDs Filer. Jr. et al., eds., GlutamlcAcJ(i: milk. led Dr. Nyhan to conclude, and I Issue, three questions must be Advances In Biochenustry and agree. that the undiagnosed PKU Infant answered: (1) Based on extrapolation PhYSIology, p. 287+(1979))). will not be at additional nsk by the from anImal data, what IS the tOXlC Significantly, it IS believed that the marketing of aspartame rrr./l/page 232, threshold, m terms of the plasma levels leSion can be produced by a smgle surge line 1-page 233. line 5; cf. Tr./II/page of aspartic aCid [ASP) and glutamlc aCid ofplasma

elevations m plasma GLU/ ASP levels ASP elevations (i.e., Iasting subjects, Finally. Dr. Olneihas cited data resulted. as noted above. with MSG gIven m low-carbohydrate showmg that a gIVen oral dose of GLU The Board also cited data showmg consomme and aspartame given In in young anImals produces higher -that aspartame did not potentiate (i.e., unsweetened Kool-ald) m WhiCh, plasma GLU/ASP levels than does the augment) the effect on plasma GLU followmg the addition of both MSG and same dose on adult anImals. He then levels mduced by MSG. for example, aspartame, plasma GLU/ ASP levels in analogtZes tcihumans and suggests that the Board cited a human study showmg some mdivlduals were observed to be children will have Ingher plasma levels that aspartame at 23 mg/kg had no nearly hVlce as high as those found than adults after mgestion of aspartame effect on the plasma GLUIASP levels follOWIng the addition of MSG alone m conJunction \ ..ith MSG (Olney's resulting from Ingestion .of a hamburger­ (Olney's Exceptions at 3). Exceptions at 3). milkshake meal to whIch 150 mg/kg of The Board's deCISion did not discuss There are two answers to tlus·pomt. MSG had been added: SImilar results the study referred to (Vol. 139, Tab 12 First. the anImal data m the literature were also .found uSing doses of 34 mg/kg and Vol. 152 [Steglnk VI-B at 22-25]).:13 are mconslstent. Although some studies aspartame and 34 mg/kg MSG (Board's However, Dr. Olney's statement that the do show that plasma GLUIASP levels DeCIsion a132, 33, referring to Vol. 152 "glu + aspartame meal caused GLU/ are hIgher m llllJllature anImals as (Stegmk VI-B at 1~22)). Similar results ASP blood levels In some IndiViduals to compared to adult ammals given an were found m an aspartame-MSG rIse nearly hVlce as lugh as those equal oral load of GLU or ASP (Vol. 126. mteraction study performed under Induced by glu alone" (emphaSIS added) Tab 15 [Stegmk et aI., "Comparative "hIgher rIsk" conditions (Vol. 139, Tab presents a one-sided view of the data. In Metabolism ofGlutamate in the Mouse. 12) as discussed below. tIns study, the addition of50 mg/kg Monkey and Man" m Filer. Jr., supra, d. Dr. Olney's exceptioI]s regarding MSG alone resulted In a mean elevation page 85 +]: Vol. 138. Tab 7 (Oppermann the effect ofaspartame on plasma GLU/ In plasma GLU + ASP from a fasting and Ranney,/ournaiofEnvironmenfa/ ASPlevels In humans. Dr. Olney has level of 4.4 ± 1.2 p.mol/ dlto a peak of Pathology andTOXIcology. 2:987.1979)). taken strong exception to the Board's 21.0 ± 7.1Ilmol/dl. The further addition Dr. Olney's contention that tIns deCISIon on tIns focal bram leSiOn Issue· of34 mg/kg aspartame resulted m a difference m plasma levels IS "well (see generally, VoL 158, Tab 261). His mean level of25.7 ± 10.5p.mol/dl, wluch established" IS. In my VIew, not Catrect. pnmary concern IS that the Board did was not different, In terms of statistical In Iact, there are some ammal studies in not address the question of plasma Significance, from the mean level the record, not mentioned by Dr. Olney. GLU/ ASP levels m children folloWIng reached with MSG alone (vol. 139. Tab whIch do not show tIns effect (VoL 123. the Ingestion of aspartame in 12 and Vol. 152 (Stegmk VI-B at 24)). Tab 14 (O'hara. etal.,/ournal of conjunction with MSG under "hIgh-rISk" Singling out those mdiVidual subjects TOXIcology SCIence 2:281. 1977); Vol. conditions,22 and that such a study who did show an mcrease m plasma • 136. Tab 10 (Bizzi. et al. TOXICOlogy should be performed In VIew of the fact levels (over MSG alone) after receiVIng LaUer 1:123, 1977)). that children are already exposed to both compounds results, m my opiruon, More lIDportantly. In human studies food products contammg large amounts m a SCIentifically mcorrect usmg aspartame. it was shown [under of GLU and ASP To· support hIS Interpretation of the data. Plasma level "hIgh nsk" conditions) that doses ofup position, Dr. Olney cites the follOWIng data such as thiS almost always show a to 100 mg/kg were handled aswell by 8­ three lines of eVIdence whIch were not certaIn degree vanation between 12 month old mfants as by adults subjects and even m the same subject: discussed m the Board's deCIsIon~ (regarding mcreases m plasma GLU + First. Dr. Olney states that some thiS IS why hypotheses are accepted or ASP). Tlus study was cited by the commerCial soup products con tam rejected uSing mean values; standard Board, whIch stated: "ThIs finding e.nough added GLU (as MSG) to prOVIde errors, and.statistical methods. Based on appears to refute any suggestion that 100 to 130 mg/kg for a young child, and mean values obtamed m tlus study, aspartic aCId mIght be metabolizedless _ aspartame did not have a Significant that a sunilar dose of GLU, when fed In effiCIently In mfanls than In adults:' noncarbohydrate solution to human effect. Citing results only Ior the (Board's DeCISion at 33. referrIng to Vol: adults, caused a surge of blood GLU/ IndiVIduals who fell on one Side of the 152 (Stegmk VI-B at 31)}. I agree. There mean severely bIases presentation of a ASP to levels substantially exceeding also other human studies In the study's findings. In Iact, there were are the 100 p.mol/ dllevel which the Board In some subjects In thiS study which had record. noL mentioned either the determmed to be the toXiC threshold Board's deCISion or In Dr. Olney's (Olney's Exceptions at 2-3J. lower GLU + ASP levels with the combmation than with MSG alone, but exceptions. showmg that infants This pomt. however, concerns the risk one would not want to conclude from (including tliose whIch were premature assoCiated with mgestion ofMSG itself tIns that aspartame antagoruzes (i.e., or or low bIrth weight) have the whIch IS not at Issue here. What IS at counteracts) the effect ofMSG.::t capability to metabolize dietary GLU Issue IS whether the addition of and ASP as well as adults (cited m VoL aspartame to the food supply will :sAs noted aQo\'c. Ihe Board did discuss a study 152 (StegInk VI-B at 25-31)). mcrease or potentiate any elevations of In which the effect of nspartrunc on MSG-Induccd I therefore find. after a conSideration .plasma GLU/ASP WhICh mIght be nse In plasma GLU/ASPwas examined In human ofDr. Olney's exceptions and the caused by MSG, and, as discussed adults (no effect of aspartame was noted). allhough available data, that the proposed use of t.lus study was not done under "lugh·risk" below and m the Board's deCISIon, the condlIIons, I,e., Ihe vehicle was a hambwger· aspartame, either alone or together with avaihible eVIdence suggests that it will milkshake meal (Board's DecIsion at 32-33). glutamate. does not pose a rIsk offocal not. 2'It can of course be argued Ihnt aspartame might bram damage In humans. The data is Dr. Olney next pomts to a study In have a potentIating effect In some individuals, conVInCIng that that plasma GLU/ASP human adults,.performed undel'''lugh although a more likely explanatIon. Inlight of the facllhat In other sludics nspartame alone at thIs rIsk" conditions regarding plasma GLU/ and higher doscs hed no effect on plnsma GLU/ Ihnt aspartame did not cause II statistically -ASP. Is Ihnt Ihe higher levels seen In80me sublects slgnlficant potentiatIon of the effect ofMSG. "Le.. by usmg noncarbohydrate liquid vehicles,. recclVlng bolh compounds represents variabUily at II might also be noted Ihnt the hIghest plasma whIch would nunuc the animal gavage studies by different times In a single sublect's response to GLU + ASP level reached by an mdividualin!lus proVIding a bolus dose and thus allow for mllXlIllal MSG. Whate\'er the explanatIon. howe\,cr. the stud)' wns 39.6 p.mol/dL which is still weD below Increases In plasma GLU!ASP. conclusion which must be drawn from thIs study is the toxic threshold ns detennlned above. 38294 Federal RegIster I Vol. 46, No. 142 I FrIday, July 24, 1981 I Notices • levels m human adults recelvmg admlmstered m amounts large enough to rodents by subcutaneous mJections of unusually large doses of aspartame cause Identifiablellypothalamlc leSIOns" MSG, the Board suggested that"· • • it remamed withm normal after eating (Board's DeCISion at 31), More seems reasonable to assume that In the limits, far below the lowest level even lffiportantly, no endOCrIne tOXicity due same speCIes * • * admmistration of suspected of bemg neurotoXic. to aspartame at subneurotoXic doses aspartame bygavage * * • (at a doso Moreover, the available data m human has been reported mammals. contalmng an eqUIvalent amount of mfants strongly suggest that this group I therefore conclude that, as with the aspartic aCId) * * * would have similar handles loads of aspartame, GLU and Issue of focal bram leslons,-aspartame endOCrine consequences" (emphasis ASP as well as adults. The lack of would not pose an additional risk to added). The Bureau's concern is that tho significant potentiation of the effects of children of neuroendOCrine changes, and route of admmstration might MSG (on plasma GLU/ASP levels) by that Dr. Olney's proposed study IS Significantly affect the plasma levols, aspartame, as shown m adults under unnecessary. and that dosmg orally by gavage is "high risk" conditions, provides further f. Other exceptions. (1) The Board likely to reqUIre a higher dose to eVidence of aspartame's safety when noted that the record contamed one produce the same effects as would a consumed with MSG. The addition of apparent exception to its general subcutaneous dose (Bureau's Excoptions aspartame to the food supply, therefore, statement that no leSIOns had been at 9-10). should not create any additional risk of observed mammals as a result of the I agree with the Bureau that somo focal bram leSions m children. I find that voluntary consumption of GLU or .A.SP clarification of the Board's statement Dr. Olney's proposed study m children by miXing the test compound with the would be helpful. The critical value is IS unnecessary.25 ammal's regular food. The one exception not the dose used, but the plasma GLU e. Possible adverse effects at subtoxlC occurred m a study'in whICh 10 + ASP level reached after levels. Dr. Olney has also cited ammal weanling mice were offered admImstration of that dose. Because it is studies (Vol. 125, Tab 66 (Olney, et ai., concentrated solutions of either GLU or possible that different routes of Bram Research 112:420, 1976); Vol. 125, Gl.u + ASP + aspartame after havmg admlmstration may affect resulting Tab 55 (Terry, et ai., Federal been deprived of water overmght. All 10 plasma levels, it would be necessary to Proceedings 36:364, 1977)) whICh have mIce developed leSIOns (Board's test the Board's hypotheSIS WhICh, as tho purportedly shown that doses of MSG DeCISion at 27, referrmg to Olney, Vol. Board itself noted. had not yet boon that are sub tOXIC, I.e., below those 157, Tab 205). done (Board's DeCision at 29). I needed to produce focal bram leSIOns, The Bureau takes exception to emphaSize, however, that thiS point is produced acute changes m plasma categorizing trus experiment as a valid raised for clarification purposes and in hormone levels (prestlmably via model for voluntary dietary no way changefl the Board's decision. excitatory effects on hypothalamiC consumption (Bureau's Exceptions at 6­ neurons which control pituitary 9), and I agree with the Bureau. These 4. ConclUSIOn. For the reasons stated hormone secretion) and that a study above and m the Board's decision, I find anlffials were water-deprived, and they - that there IS a reasonable certainty that with aspartame + MSG, mvolvmg the drank a small amount of highly measurement of neuroendOCrine the proposed use of aspartame, eithor concentrated solution over a short alone or together with glutamate, will function, should be performed m period of time at doses known to mduce children for thiS reason also (Olney's not cause focal bram leSions m man or leSIOns m weanling mICe.when other adverse effects on the Exceptions at 4, 6). admmlstered by gavage. Thus" the I disagree. As stated' in the Board's experimental deSign was essentially no neuroendocrme system. deCISIOn (pp. 30-31), the hormone level different from the prevIous gavage V EVidence on the Bram Tumor Issuo changes noted m these studies were studies. 2s withm the range of normal fluctuation, Finally, although it may be argued A. Introduction. and may "have reflected no more than a that under some conditions human 1. Issue Presented. The second major normal CIrcadian or ultradiaI1 periodicity voluntary consumption of aspartame Issue at the hearing was defined as of (hormone) release," and two other may mlffilC gavage dosmg (i.e., follows: research groups were not able to mdiViduals may drink large amounts of 125, 68 The question has been raised whether Ute replicate these findings (Vol. Tab aspartame-sweetened beverage at one mgestion of aspartame may Induce brain (YoneJam and Matsuzawa, TOXIcology sitting), human studies performed under neoplasms m the rat. From available Letter 1:207, 1977); Vol. 137, Tab 25 these conditions showed that plasma eVIdence, what can be concluded In relation (Nemeroff et ai., Bram Research 156:198, GLU/ASP levels were not substantially to thIS question? The objecting parties bellevo 1978)). Moreover, there was no proof in mcreased (Vol. 140, Tab 5 and Vol. 115, that available eVIdence suggests, without the "positive" studies that the doses Section III). adequately ruling out. uposs\ble association , used (1000 mg/kg s.c. or I.p. m adult (2) The Bureau has also taken between aspartame Ingestion and an rats) did not cause hypothalamiC exception to one statement; apparently mcreased mCldence of brain neoplasms In the leSions. The Board thus concluded that made by the Board m passmg, whICh rat. The Bureau of Foods believes that "endocrme disorders are mduced by available eVIdence docs not show that may reqUIre clarification. The statement Ingestion of aspartame results In an MSG only when thiS substance IS appears on page 29 of the Board's mcreased mCldence of brain neoplasms In tho deCISIOn where, after discussmg rat. .. Dr. Olney has requested. In the event aspartame neuroendocrme disorders mduced In is approved for marketing without requmng hIS (44 FR at 31717). In layman's terms, proposed study In children. that the Board and neoplasms are tumors. Thus, stated In myself jointly sign an affidavit stating that we_ • Interestingly. another group of investigators consider such a study to be safe (Olney's have suggested that: the context of the legal standard, the Exceptions at 6). I decline to do so. My sole • • • Apparently water·reSlncted weanling mice Issue IS whether the data establish that responsibility in thIS proceeding IS to render a lose theIr ability to regulate subsequent drlnkmg there IS a reasonable certainty that decision on the Issues rmsed at the heanng. as behaVIOr. and consume hyperosmola solutions aspartame does not cause bram tumors defined in the July'l. 1979 Federal Reglstor notice. whose osmolarity and sweetness would be averSIve This I have done. As for the Board. its to humans. m laboratory rats. responsibilities were fulfilled with the completion of (Takasakl. et 01., Searle's Reply. Vol. 161, 2. Background. The bram tumor issuo the Initial DeCIsion. Appendix at 1). falls under the general category of Federal RegIster I Vol. 46. No. 142 I FrIday, July 24. 1981 I Notices 38295 carcInogemcity In wh1ch the agency has col. 2). In the case of aspartame. the Japanese study) (Searle's-Exceptions. long, exerc1sed its cons1derable expertise parties dispute the propel' category mto Appendix 2). The study tested (see, e.g., CormmsslOner's DeClS1on: Wh1Ch the three rat studies should be aspartame as well as an aspartame-DKP Cyclamate, 45 FR 61474, Sept 16, 1'980; placed. IDlXture. Searle also submitted Comm1ss1oner's Dec1s1on: FD&C Red No. 3. Positions ofthe parties. Dr. Olney additional data on the spontaneous rate 2,45 FR 6253, Jan. 25, 1980; would classify E-33/34 as suggestive 'r Issue (Searle's Exceptions. Appendices 3 COImmsslOner's DeC1S1on: and E-70 as defic1ent. thereby and 4). as did the Bureau ofFoods Diethylstilbestrol (DES), 44 FR 54852. concluding that Searle's petition should (Bureau's Exceptions. Appendix 3]. Sept. 21,1979). To assess the not be approved without further testing. Because th1S proceeding is mtended to carcmogemc potential of food additives. The central thes1s m Dr. Olney's .' be a sClentific mqUIry aImed at the Bureau of Foods reqwres chromc position IS that the spontaneous rate of evaluating the safety ofaspartame uSIng studies m two rodent speCles, usually bram tumors 1Il Sprague-Dawley rats. as all the available eVIdence. I have the rat and the mouse (Tr./I/page·16, reported m the sC1entific literature. 1S conSidered these materIals as eVidence lines 23-25). Proof of safety must entail slgnificantly below the lIlc1dence of m thIS proceeding. acknowledgmg that negative findings from both speCles • bram tumors found m both the control neither the Board nor the partiClpants to because it 1S not known wmch spec1e 1S ammals III E-70 and the treated animals the hearIng have commented on them. In more slmilar to man. As noted m Section m E-33/34. Dr. Olney also cons1ders the so domg, I note that none ofthese N(C) above. it IS the agency's practice, data m E-33/34. on the1r own. to suggest additional matenals have served as a m the absence of more preClse sClentific a dose response and accelerated tumor central bas1s for my deCls1on. but rather knowledge. to adopt findings from the onset. both lIldicators ofpossible only confirIIl the large body ofe\'ldence most sensative speCles m order to carclIlogencity. presented at the hearmg. maXimIZe protection of the public The Bureau of Foods and Searle health. cons1der both aspartame rat studies to 6. Commissioner's deciSion: With due In keepmg "with the Bureau's be negative, thereby Justifymg approval respect to the Board. I disagree with its.­ requirements, Searle submitted chromc of the food additive petition. In response assessment of the background rate of feeding studies on aspartame for both to Dr. Olney's concerns. they maintain spontaneous bram tumors m Charles the rat (E-33/34, Vols. 43-44; E-70, that the inCldence rates at 1ssue m the River CD (Sprague-Dawley) rats. and. Vo1.81) and the mouse (E-75. Vol. 82). aspartame studies represent-normal therefore. I also disagree \\ith th~ Similar studies were performed on the levels ofbackground spontaneous ~ Board's characlenzations ofstudies E­ brea'kdown product diketop1perazme mCldence. and that E-33/34 70 and E-33/34. wblch charactenzations. (DKP), also m both speCles (E-77/78 m demonstrates neither a dose response especHllly regarding E-70, were largely the rat, Vols. 89-90, and E-76 m the nor early tumor onset dependent on the background rate mouse, Vol. 88). Because the parties 4. The Boara's declSlon. The Board assessment. As IS explamed m more have agreed that the mouse data are agreed essentially with Dr. Olany that detail below. I agree with the Bureau of negative, only the three rat studies were the background rate for spontaneouS" Foods that the Incidence rates reported subject to the Board's reView and are bram tumors m thts stram of rat was m the Searle studies fall withm consIdered m thts deClslOn. very low, Ute Board finding the rate to reasonably expected bounds of These three rat studies had the be approXImately 0.7% :5 (Board's spontaneous InCldence for the type of follOWIng des1gns: DeClS10n at 43-45). Given that rat and study SlZe used. and that the prImary evaluation of these studies a. E-33/34 was a 1M-week study on deterIIllIlation. the Board disIDlssed the aspartame, with exposure-begmrung after E-70 study as "b1zarre" because the should be between the treated anImals weamng. Dose levels were 0, 1. 2. 4, and 6-8 control group there showed a 3.5% and their concurrent controls. Usmg tins gldg body weight mCldence of bram tumors (id. at~47). approach. I find that the data m E-33/34 b. E-70 was a two generation study with Also based on its assessment of the do not suggest. 1Il terIIlS of bIOlOgIcal aspartame exposure In utero, dunng background rate, the Board found that, significance. a dose-response lactation, and then for 104 weeks. Dose levels relationslup or early tumor onset. were 0, 2 and 4 g/kg body weight regarding study E-33/34: "By itself, the c_ E-77/78 was a l1S-week study on DKP, 3.5% mC1dence of bram tumors (in the Accordingly. I conclude that the two with exposure begmrung after weanmg. as m treated anImals) glVes cause for aspartame studies reVIewed by the E-33/34_ Dose levels used were 0, 0.75, 1.5 concern" (id. at 46). The Board's concern Board are neither "bIZarre" (E-70) nor and 3.0 g/kg body weight about E-33/34 was augmented by its even of major "concern" (E-33/34). but Inall three studies,-the test anImal used agreement with Dr. Olney that the data rather they are negative studies. was the Charles River CD (Sprague­ suggested a dose response and that 7 Conduct ofthe studies. Dr. Olney Dawley) a1bmo rat. 27 there was a h1gh mC1dence ofgliomas and Mr. Turner have questioned the The agency: has set forth the general (pr1mary bram tumors) at a relatively manner m WhIch the aspartamejDKP prmClples of statistical and blOloglCal early age. Accordingly, the Board. like studies were conducted and their slgnificance wmch gwde the evaluation Dr. Olney. would also classify E-33/34 credibility and usefulness for of carcmogemcity studies (45 FR at as a suggestive study and E-70 as a meanmgful Interpretation. The Board 61477-81). Based on these criterIa. defic1ent one. cons1dered these Issues to be beyond studies may be classified as: (aJ 5. Additional eVldence. After the the scope ofits charge and declined to Positive; [b) mconcluslVe but suggestive Board 1ssued -its dec1s1on. Searley. as rule on them (Board's DeCls10n at 6-8). of a carcmogemc effect ("suggestive"): part of its exceptions. submitted a Mr. Turner has taken exception to this (c) negative: or (d) defic1ent (id. at 61481. recently completed 10ng-terIIl study deCls10n by the Board and has requested conducted on Wistar rats by the that the Board be reconvened to "Sprague-Dawley IS the general stram ofrat Japanese firm. Ajinomoto 90.• Inc. (the cons1der these 1ssues (Turner's used. Different commerctal suppliers have Exceptions m their entirely). The"­ developed their own colomes ofSprague-Dawley UThe Bonni was not as conscrvaU\'o In lis rats. and the Charles River Laboratones from estimate as was Dr. Olney, wbo considered tho conduct of the studies andMr.Turner's Wilmmgton. Massachusetts IS one such supplier. background rate to lie 0.15% [fr./Ill/page139,llnes request for a new bearmg are discussed" "CD" Simply means caesanan-denved. 6-23). m detail m Section VI below. 38296 Federal RegIster I Vol. 46, No. 142. / Friday, July 24, 1981 I Notices • .= B. BackgroundRate for Spontaneous people (Koestner Testimony, Tr./m/ same laboratory as were Senrle's BramTumors in Charles RiverCD page 257, lines 21-25; see also studies (at least ID part) reported a (Sprague-Dawley) Rats MacKenzIe and Garner, Vol. 134.Tab 20 hIgher mCIdence than did the. studies 1. OvervIew. As noted above, the at 1252-53). relied on by the Board (Gart. et al. Vol. cornerstone of the Board's decIsion IS A second factor IS the SIZe of the 154, Tab 7, Table 4, discussed below). that the background rate for study population. The smaller the SIze of ­ Second, the test populations In tho spontaneous tumors 10 Charles River CD the test sample, the largerwill he the Searle studies tended to be smaller than (Sprague-Dawley) rats IS very low, VarIation assocIated willi the estimated 10 the studies relied on by the Board. approxlmately 0.7% (Board's Decislon at of the spontaneousmCldence rate. and thus mcreasmg the vanation observed in 43-45). Specifically, the Board cited on VIce versa. the spontaneous mcidence rates. Third. four studies wmch showed spontaneous The thIrd consIderation IS the m the Searle studies. a very detailed bram tumor mCldence rates of 0.09%, methodology used, espeClally the hIstopatholOgIcal exammntion was 0.6%, 0.7%, and 3.2%; The Board gave meticulousness of the search. For performed (involvmg either 7' or 8 brain 10 extra welgh~. to. the two studies showmg example, studies wmch ammal organs sections per animal) wmch mcreases the 0.6% and 0.7% mCldence rates because are observed only by-the naked eye are chances of detecting tumors. Finally. tho the rats 10 those studies were obtamed likely to turn up fewer tumors than Charles River rat used by Searle was from the same commerCIal source would a study 10 wmch the anImal not uniformly utilized 10 the reported (Charles River Laboratorles) as those organs are routinely exammed under a literature relied upon by the Board (i.e.• used 10 the Searle studies. The Board nucroscope. Similarly~ where more Mawdesley-Thomas and Newman gave less weIght to the .study showIng a sections of the bram are exammed, the study. discussed below). 3.2% mCIdence rate because the number chances are greater that tumors will be· Accordingly. I find that the of rats used 10 that study (125) was found, thereby mcreasmg the tumor spontaneous mCldence rates in tho consIdered to be too'small for a reliable mCIdence reported (Tr./m/page 258, Searle studies are consistent with tho determmation of spontaneous tumor lines 1-9). normal background rate. as determined mCldence (id.). Finally, as the Board noted, the stram from the data 10 the admmIstraUvo Both Searle and the Bureau of Foods and commerCIal source of ratused are record of tills proceeding. have taken strong exception to this Important because ammals denved from 2. Studies Relied Upon by the Board: portion of the Board's oplnton (Searle's different colomes may acqUIre different As noted above. the. Board cited four Exceptions at 16-22 and Bureau's charagtenstics (Board's DeClSlOn at 45, studies 10 malong its determmation of Exceptions at 24-32). Ingeneral, both citing MacKenzIe and Garner, VoL 134. the background rate for spontaneous Searle and the Bureau argue that the Tao 20). bram tumors 10 Charles River CD Board gave too much weIght to the These four factors help explam why (Sprague-Dawley) rats. These studios studies at the lower end of the spectrum there are such. varied results among the may be reVJewed, as follows: and Ignored additional studies 10 the reported studies. As the Board noted. a. Mawdesley-Thomas andNewman admlmstrative record whICh reported emphasIzmg different methodologIes (Vol. 135. Tab 18): ThIs study reported spontaneous mCldence rates as mgh as used: 38 tumors (24 m males) 10 approxlInotely 5%. Searle and the Bureau therefore It IS difficult to conclude from the archIval 41,000 rats for an mCldence of 0.09%. conclude that the mCldence rates literature wluch ofvanous published fi.,oures These rats were fed either a control diet reported 10 the Searle studies fall withm most accurately reflects the "normal" (ie., on one of a varlety of test compounds. the normal spontaneous range. presumably non-toxogemc} mCldence of bram The rats were of the general Sprague­ After a thOl:ough reVIew of the studies tumors m the Sprague-Dawley rat stram. Dawley stram but werenot obtained Several'published reports are basedon 10 the admlmstrative record submitted from Charles River LaboratorIes. on thIs subject, I conclude as follows: (1) findings m rats that had been used m long­ term studies deSIgned to check the potential Searle critiClzes thIS study because. in No smgle study and no group of studies toxicity of a: particular compound, or of a subsequent publication (Vol. 135, Tab submitted 10 thIs proceeding are rrradiated foods. Other reports fail to state 19). the same authors reported that tho suffiClent to stand alone as a defmitive the-protocol followed m exanumng the bram mCldence rate 10 tills study was statement of the background rate for tumors: Gross-anatormcal tumor probably closer to 1% than 0.1% (Searle's spontaneous bram tumors 10 tills stram Identification only, or routine-hIstological Exceptions at16-17). The Bureau of rat; (2) although several studies cited exanunation of each. bram? criticizes tms study for three reasons: (1) .by the Board do report spontaneous (Board's DeClsion at 43).30 The published Because the hIstologIcal examinaUon mCldences 10 the area of1% odower, literature also vanes considerably 10 was more limited than in the Searle these studies are partially flawed and terms ofthe study population SIze, the. studies; (2) because tumors were must be supplemented by other data commerClal source ofrat used, and the elimmated whenevel" they wero presented at the hearmg wluch reported time and place the data were collected. suspected of beIng compound-related; mCldence rates comparable to those 10 These four factors also help explam and (3) because not all tho slides wore the Searle studies; 2!tand, therefore: (3) why the Searle studies reported a Wider reVIewed by the authors themselves the prImary evaluation of the Searle range of mCldence rates than reported 10 (Bureau's Exceptions at 26-27). studies should be on the baSIS of those studies relied on most heavily by I agree with Searle, the Bureau. l,md comparIson with concurrent, not the Board. First, the studies relied on by the authors themselves that the reported hlstorlcal, control data. the Board were conducted at different Four factors seem to playa SIgnificant mCIdence of 0.09% IS probably too low. laboratones than were Searle's studies. The testpopulation mcluded both role 10 creating thlS spectrum offindings. Indeed, the one background rate study The first IS the varlation that would be treated and control animals. The. authortl 10 the record whIch was done by the expected among tests run at different elimmated any tumors from the times and at different places by different mCIdence count that were "suspected" 3.Indeed. one Bureau witness suggested that of bemg compound related, but did not virtually all the reported studies. whetherthey also eUmmate the other "treated" 29ThISJ;oncluslOn IS further supported by the reported high or low spontaneous~rates. have some spontaneous rute materlUls submitted with Searle's methodolOgical defiCienCieS (Tr./Ill/page 195. lines ammals as well (Vol. 135, Tab 18 at 108). and the Bureau's exceptions. 7-11). ThIS approach likely mflated the number Federal RegIster / Vol. 46, No. 142 / FrIday, July 24, 1981/ Notices 38297 of tumor-free anunals and lowered the searches could have the effect of control arumals used m the reported spontaneous mCIdence rate. It lowerIng the reported spontaneous carCInogeneSIS bIoassay program. The therefore appears that the authors' mCIdence rate. Second. the authors participating organrzations were NCI subsequent statement, that the true reported that anunals were sacrificed at and Hazelton Laboratones. the spontaneous mCIdence rate among this unspecified mtervals (Vol, 134. Tab 2.2 at laboratory used by Searle for studies E­ group of anunals was probably lugher 265). Early deaths may also have helped 33/34 and E-10 frr./m/page 214. lines than ongmally reported, IS correct lower the reported mcldence. 14-24 and page 211, lines 11-16). The b. MacKenzie and Carner (Vol. 134, d. Thompson, et al. (VoL 134, Tab 18): data were denved from Charles River Tab 20): TIns study reported three bram TIns study reported four bram tumors m CD (Sprague-Dawley1 rats and showed tumors m a two-year study conducted 125 rats for an mCIdence of 3.2%. The an mCldence rate for bram tumors of , on 535 rats, for.an mCldence rate of0.6%. Board found, however. that "[t]he 2.2% (8/368) frr./m/page 197, lines 13­ The test anunals mcluded both those,fed •number of rats used in this study is too 16). rrradiaten feed and those on a control small for a reliable determination of As Dr. Olney pomts out, it15 true that diet Although a breakdown by sex IS spontaneous-tumor inCIdence" (Board's these data are reported only m tabular not grven for these three tumors, when DeCISIon at 44). Searle believes the form without a detailed descnption of the authors combmed all the strams of Thompson data are nevertheless valid, the methodologres used (see Tr./m/page rats tested, over two-thirds of argumg that smaller study populations 218. lines 1-6). Nevertheless. the fact spontaneous bram tumors were found m tend to produce a WIder vanation m that these data were all denved from males (Vol. 134, Tab 20 at1251, col. 1). reported mCIdence rates (Searle's control anImals IS suffiClent to consIder Searle criticIZes this study because Exceptions at17-18). the mformation m this proceeding, -the authors themselves (at page 1252 Both the Board and Searle make vaUd espeClally because there 15 such an state that their fin~s cannot be pomts that are not mutually exclUSive. overall sparceness oHruly "control" compared with others' because of As noted above, it is true that small test mCldence data available m the record differences m methodology and populations lead to greater ImpreCIsion (see Tr./m/page 217. lines 2-5). diagnostic critena (Searle's Exceptions m estimating spon.taneous mCldence The spontaneous mCldencEr"ofbram at 17). The BlIreau makes only the rates. For this reason, the Thompson tumors reported by NCI (2.2%) IS general criticIsm, as it does with the study would not serve as a reliable approXImately trIple the mCldence ·remamlng two studies relied on by the mdicator of the "true" spontaneous rate. reported by the MacKenzJe, etal. and· Board, that there IS not enough detail However, when a study's test Fitzgerald, et al. studies relied on about the methodology to enable a population IS small, the varJation in the heavily by the Board. Significantly, correct assessment of the thoroughness observed inCidence rate will be large, these data allwere denvedfrom the of the hrstologrcal exammation and the frequency of observing both same commerCIal source (Charles'River) (BlIreau's Exceptions at 27). 1ngh and low mcidence rates will be and were housed, at least m part, m the Although this study IS clearly entitled mcreased. Thus, due to this increased same laboratory (Hazelton) used by to some weight, I believe that the Board vanability, the mCldence m the Searle. As noted above. these factors overemphasIZed its Importance. The relatively small Thompson study (4/125, are known to affect reported mam plIrPose of this publication was to 3.2%) IS acceptable for that SIZe test spontaneous mcidence rates 'compare spontaneous mCldence rates m population. When the Thompson data studies conducted at different (MacKenzie and Garner, Vol. 134. Tab are added to the other three studies 20 at1253. cols.1-2). laboratones and at different times. And. cited by the Board (.09% [38/41,000], .6% mdeed. the authors reported SIgnificant One direct companson between the [3/535], .7% [5/650] and 3.2% [4/125]), NCI data and Searle's control data IS differences. Although it IS true that the the spectrum IS not unlike that in the specific mCldence,rate relied upon by • three control groups in the Searle quite strikmg. Ifthe controls from all three Searle studies are combmed, the the Board was based on rats denved studies, .8% (1/119) in E-33/34, 1.6% (2/ from the same commercIal sOlIrce as 123) m E-77/78, and 3.5% (4/115) in E-70. resulting mCldence rate 15 very were the rats m the Searle studies. InclUSIon of the Thompson study m this comparable to the NCI data for sample commerCIal source m only one of type of comparIson is valid because populations ofnearly Identical SIZe: 2.0% several factors that can affect the Thompson's study SIZe (125) is (7/356) for combmed Searle control data mCIdence rate (see diSCUSSIon m this comparable to those m the Searle and 2.2% (8/368) for NCI control data very study at1252. col. 2). Moreover, the control groups (approXImately 120 each). frr./m/page 195, line 2.2-page 196. line authors state that "many small tumors" Dr. Olney suggests that the 3.2% figIIre 9 and Tr./m/page 197. lines 13-16).31 found m other studies would not be IS too hrgh because three of the four b. AdditionalData. Other relevant called neoplasms by them (id.). a tumors were found m anImals fed a diet non-aspartame studies reported practice whrch could have lowered the ofirradiated feed (Olney's Exceptions at mCldence rates for control arumals of5% reported mCldence rate. 2). Dr. Olney's pomt IS valid to the (2/40),3.3% (2/60) (reported tWIce), 2% L c. Fitzgerald, et al. (Vol. 134. Tab 22): extent that it mdicates a flaw in this (8/400).1.9% (7/368).1.5% (13/876). and Thrs study reported five bram tumors m study (i.e., USIng both "treated" and 0.5% (3/575). Moreover, all ofthese data 650 rats for an mCldence of 0.7%. Once "control" anunals), but it 18 a flaw were based on control ammals and were agam, some of the test anunals were fed common to all of the studies relied upon obtamed from the Charles River test compounds while others were on a by the Board. Such flaws underscore the Laboratones. the Same source used by control diet The authors reported that need to consIder the truly "control data" Searle. The utility ofthese data 15 these tumors predommated m the males. described below. somewhat limited because the data although an exact breakdown by sex 3. Other eVIdence. In addition to these were not available for the Board's was not grven. f fOlIr studies cited by the Board, I also consIderation (Searle's Exceptions. There are two weaknesses m this conSIder the follOWIng data to be reported study. First, the authors did not relevant: ~l A second. more s:onlirmalory type of state how many bram sections were a. Cart, et (Vol. 154, Tab 1, Table compwon Is that NCI and Searle both reported a al. higher sponlancous occunence ofbram tumors m routinely exammed mIcrOSCOPICally. As 4). These data were collected by the males than In females. 'I1us Is COllSlStent with the noted above, less-than-meticulous National Cancer Institute ("NCr') from otherreporled sludics. 38298 Federal RegIster I Vol. 46, No. 142 I FrIday. July 24, 1981 I Notices

,Appendices 3 and 4: Bureau's concurrent controls" (Gart.. et al.• Vol. were verified by UAREP (Vol. 111 at 391, Exceptions, Appendence 3), and 155. Tab 7 at 962). As noted above, the Table IV-201. 34 Most of these tumors because theyreport fmdings only In agency has set forth general prInCiples were gliomas. although one tumor in the summary form without any detailed of statistical and biologIcal significance high dose female group was diagnosed description of the testmethods.32 willch guIde the evaluation of by the Board as a medulloblastoma' Nevertheless. tIns fuformation tends to carcmogenicity studies (45"FR 61477-81). (Board's DeclSlon at 40-41). confirm my conclusIons drawn from the Factors usually conSIdered are: "the Before analyzing the data. it is data presented at the hearIng and methodology of the study Involved. the necessary to resolve two disputes about therefore may be consIdered as eXistence of a dose response exactly how many tumors were found additional support for those conclUSIons. relationship. the rarity of tumors. and among the test animals. The controversy 'As noted above, several of these the presence ofsImilar results mother IS due to variations m tumor count reported bram tumor mCIdences were studies" (id at 61478. col. 2). Other among the several persons or groups SIgnificantly higher than those reported relevant conSIderations mclude the who Viewed the slides: Dr. Innes on by the Board, with. sIZable studies tumor InOldence~In treated anImals behalf of Searle.35 the UAREP reporting mOldences ofl.5%. 1.9% and versus concurrent controls rrr./m/page Committee. and the Board. 2.0%. Moreover. one particular study 190. lines 13-17). any acceleration of (1) Male control group: In this illustrates quite well the pomt that small tumor onset (Tr./m/page 191. lines 1-2; category. Dr. Innes reported no tumors, studies are subject to wIder variation in Tr./m/page 250. lines 14-15). and study while UAREP and the Board each reported spontaneous mCIdences (both SIze (45 FR at 61482). reported one tumor. Dr. Olney took hIgher and lower) than are larger These criteria will be discussed below exception to tIns finding by the Board. studies. These data, which showed an In the context of the Searle studies m asserting that the Board's diagnOSIS,. overall spontaneous InCIdence rate for which they arIse. "most likely a metastatic carCInoma." 2. E-33/34 (Vols. 43-44) 33_ . Study bram tumorsof2.0% (8{400) (Searle's a meant that it was not a prImary brain Exceptions, Appendix 3) was actually DeSIgn. TIns study was conducted on tumor (Olney's Exceptions at 2). The Charles River CD (Sprague-Dawley) rats broken down Into the followmg four Bureau and Searle each counted this separate control groups that were run USing aspartame as the test com'pound. Four treatment groups. conSIsting of 40 tumor. relymg onUAREP's diagnosis conclll1'ently: 4% (4/100): 3% (3/100): 1% that the tumor was an astrocytoma (1/100): and 0% (0/100). ThIs varIation IS rats per sex pel' group. were fed aspartame as part of their regular diets (Bureau's Exceptions at 14; Searle's remarkably sImilar to that seen In the Exceptions at 26). Searle also asserts at dosage levels ofl. 2. 4. and 6-8 g/kg control data from the three Searle. that, In the absence of carcmomas found studies: 3.5% (4/115) In E-70; 1.6% (2/ body welghtl day, respectively. for a period of 2 years. begmmng after In other organs. the tumorcould'not 123) m E-77/78; and 0.8% (1/119) m E­ have been metastatic 30 (Searle's 33/34. weaning at 4 weeks of age. A control group of 60 rats per sex were fed the Exceptions at 26J. 3. ConclusIons on spontaneous rate. I agree with the Board, UAREP, the Based on the above analYSIS. I find that same diet without the aspartame. At the conclUSIOn of the 2-year panod. all the Bureau and Searle that this tumor the Board's conclusIon that the should be counted. for several reasons. background rate for spontaneous bram SurvIVlDg test anImals were sacrificed. and their brams (as well as other First. tumor findings by a group with tumors m Charles River CD (Sprague­ UAREP's expertise are entitled to Dawley) rats was apprOXImately 0.7% organs) were exanImed histolOgically. conSiderable weight. espeCially positive was unduly low. Although it IS true that Eight coronal sections were eventually exammed from each ammal's bram. fmdings: which are much more difficult three studies relied on by the Board to discount than negative ones. Second, showed tumor InOldences ofless than b. StudyResults. Examination ofthe brams revealed a total of13 tumors, one although the Board "tentatively" 1%. these studies each had some flaws, diagnosed the tumor as metastatio (it or discussed above, and other credible In the control group and 12 spread among the four treatment groups (Board used the qualifymgwords ''most data reported spontaneous mOldence likely"), the Board did include this tumor rates for bram tumors m the mId-l% and' at 40-41). The breakdown by sexand dosage level IS as follows~ m its control group count (0.8% VS. O.O~&). 2% range. Also Important are additional something the Board should not have data derived from relatively 'small done had the tumor been clearly studies (comparable to E-70 controls) Group Males Females metastatic. Finally. even by Including reporting spontaneous inCIdences In the COnlrol.__.______._ 1/59 0/59 tIns tumor. the control inCidence is still 3% range or even higher, whIch results 1 g/kg.___ 2130 2140 2 g/kg.• ______1/40 0/40 only 0.8%, which lies at the lower end of may be attributed to the variation that 4 g/kg__ .______4/40 1140 the spontaneous mOldence spectrum may reasonably be expected from 0-8 g/kg 0/39 2/38 (see Subsection B above). Accordingly, studies with small sample populations. I the weight of the evidence strongly therefore find the histOrical control data The numerators shown above to be consistent with the rates reported represent the figIlres found by the Board "Decreasing the denominators Cor this reason in the Searle studies, and therefore the (Id.). The denommators represent the makes the data base moro accurato and reliable. No safety of aspartame should be evaluated appreciable effect Is seen In the staUstlcul on the baSIS of comparison with total number of anImals at Ilsk which evaluations as a result of this chanso. Statlstlcal concurrent controls. results reported in thls dcclalon arc based on the "This study IS reported at several different data base listed above and thereCoro vary sUghtly C. Studies on Aspartame and DKP places In the admmlstrative record. The onglnal from the results reported by the Bureau of Foods. report Is designated as &-33/34 and IS found In 2:1 Dr. Innes' review superseded an carllerrevlew 1. General prmcIples. ItIS generally volumes 43 and ~4. That report 15 supplemented by conducted by Experimental Pathology Laboratorlo8 accepted that "the first and foremost a pathology-report performed fot Searle. deslgIIlIted (EPL). also on behalf of Soarle. Dr. innes' review as E-87 or the "Innes Report," and found In volume was based on a more detalled sectlonlna of the test comparIson of a treated grOUP'IS to its 98. Both reports were subsequently reViewed by animals' brains than was EFL's review (see UJ\REP UAREP in Vo!. 110. pages 5-14. Vol. 111. pages 256­ Report. Vol. 112 at 833-45). 32The one exception was the study reporting a 5% 457. and Vol.112. pages 833-45. as part of the sOTho term "metastatic" means that the tlmlor Incidence (Ulland. et al., Vol. 155. Tab 4). to winch authentication procedure described ill Section I oflgmatcd at another site and then transferred tn neither shortcoming applies. above. the brain. Federal Regtster I Vol. 46, No. 142 I FrIday, July 24, 1981 I Notices 38299

UPI=E" favors treating thIs hlJ:9.or as a pnmary­ (a) Tumor InCIdence. The tumor rIVO bram tumor. mCidences found, stated m percentage ar:-~ (2) 1 g/kg male group: In thIs category, form. are as follows: (percent) Dr. Innes and the Board reported two

1.~3.91 ______tumors while UAREP reported only one 5.0 tumor. Because it IS more likely that a F~2.5 3.8

~n~______-.--______qualified pathologIst IDlght IDlSS one 1______1.7 0.0 tumor than mcorrectly diagnose a non­ 5.6 5.0 24, ______r Both the Bureau and Searle take tumor as a tumor, I agree with Dr. Innes 2.5 0.0 10.0 2.5 exception to the Board's conclUSIon on and the Board that two tumors should 6-8 0.0 5.3 be counted m thIs group. dose-response, asserting: (1) That the c. Analysls- (1) Board's declSlon. dosage levels should not be combmed m The Board consIdered thIs study to be Tumor mCidence has been analyzed thIs fashIon, and (2) the data do not suggestive of causIng bram tumors, for statistically by the Bureau of Foods produce biolOgIcally significant results three reasons: (1) The mcreased usihg the Fishers Exact test. one-tailed usmg appropnate analyses (Seade's mCIdence ofbram tumors m aspartame­ £:fr./III/page 198, lines 21-22). This test Exceptions at 24-26; Bureau's fed rats (reported as 3.75%) when calculates the probability of obtammg Exceptions at 21-22). compared to hIstoncal-controls; (2) a the observed or more extreme results, if I fmd that the statistical trend tests possible dose-response, as seen by there was no difference between the utilized by the Bureau ofFoods (Cox comparmg the mCIdences m the lower treated and control groups. The smaller and Breslow tests), whIch two treatment groups combmed (3.1.%) the calculated probabilities, the greater stmultaneously conSIder all dose levels. with that of the upper two treatment the likelihood that the results are not are more appropnate for analYZIng these groups combmed (4.3~); and (3) the due to cJIance alone. but may be data thanls the method used byDr. prevalence of early-occurnng gliomas, treatment-related (see 45 FR at 61478. Olney and the Board. These trendtests two allegedly m the first year oflife and col. 1). are espeCIally useful for the data m E­ three m the second year (Board's The P values from the Fishers Exact 33/34 because both tests accountfo~ DeCISIOn at 41 and 46-47). test are as follows: diffenng SurvIval times between the (2) Positions ofthe parties. Both treated and control groups ofeach sex Searle and the Bureau filed extenSIve Group (gmms pot ki1ogmm) and make the appropnate adjustments. exceptions to thIs-portion of the Board's ().32 0.16 The Breslow test also glves extra weIght deCISIOn (Searle's Exceptions at 22-29 o.6S 1.00 to tumors whIch are observed early ffr./ and Bureau's Exceptions at 14-22 and o.os o.~ 1.00 0.15 ill/page 200. line 23-page 202, line 3). 32-35). PrInCIpally, they claIm: (1) The Cox and Breslow tests for trends Appropnate statistical analyses show yteld the followmg P values: no SIgnificant mcrease m tumor Although at least one P value (P = .08, mCIde'nce m the treated anImals when 4 g/kg males) may m some cases be compared to concurrent controls; (2) the cause for concern, it IS not a cause for Ua!es Females Board's method for evaluating a dose concern here because the finding IS not 0.04 B~,, 0.44 response was not valid, and more repeated m any other dosage group 0.47 0.02 appropnate tests show no dose because the males did not exhibit a dose response; and (3) the Board made response. As a general rule: The P values raIsmg ObVIOUS concern factual errors m noting the time of death The factors to be considered in determining for certam'rats. bIolOgical SIgnificance [including lack of a are those for the females. The Bureau (3) Study evaluation. In evaluating dose response] may increase or decrease that argues that these values are not thIs study, data for the males and confidence [that may othen'llse be placed in biolOgIcally Significant because they are females have been analyzed separately. low P values]. largely dependent on a smgle ThIs IS because the treated males lived medulloblastoma (found m one ofthe (45 FR at 61481, col. 1 (emphaSIS added): hIgh dose females at12 weeks) which. longer than theIr concurrent control cf. 45 FR at 61478, col. 3). I therefore find counterparts, and the treated females accordingly to the Bureau, was probably that the tumor InCIdence analYSIS does not caused by aspartame (Bureau's died sooner ffr./III/page 323, line 19 to not mdicate biOlogically significant page 233, line 11). Moreover, the males Exceptions at 32-35; see also Searle's findings. ExceptioQs at 27-29). Ifthe produced more tumors than the females, (b) Dose response. The Board whIch IS consIstent ""ith the results m medulloblastoma IS excluded from these concluded that the data suggest a dose­ analyses, the P values become 0.15 for the background studies discussed m response relationshIp (Board's DeCISion Subsection B above, espeCIally, Gart, et the Cox test and 0.13 for the Breslow at 46-47). The Board reached tIns test. al (Vol. 154, Tab 7 at Table 4).37 USIng conclUSIOn by combmmg the data as thIs approach. study ~33/34 may be " follows. as advocated by Dr. Olney (Tr,./ The crux of the Bureau's argIIIDent is evaluated as follows: III/page U6.lines 12-16): that, because the medulloblastoma caused death at age 12 weeks. the tumor :nOne acknowledged mconslStency with !Ius most likely ongmated m embryonic DeCISIOn IS that the background rate Issue has been bram tissue before aspartame was ever analyzed by combmmg'tbe sexes while the compansons to concurrent controls IS bemg adnurustrated. (In &-33/34. mgestion of analyzed with males and females separately. Sexes aspartame began after weanmg, at four were combmed in the background rate analysIS Both sexes COI11lmed 3.1 pc!CCn1l..!____ weeks ofage.) The Bureau believes its because the reported studies relied on by the Board hypotheSIs IS confirmed because of the did not gIve a breakdown by sex. However. the Searle studies do gIve such a breakdown. and for failure to detect any additional the reasons stated m the text. a separate anslysls Although the Board did not separate medulloblastomas either m tIns study or for males and females IS appropnate (cf. 45 FR at the anImals by sex, sImilar results are in E-70. where the animals were 61486, coL 2 and 61489. coL 2]. found if tIns IS done! exposed to aspartame m utero, during I 38300 Federal RegIster I Vol. 46, No. 142 I FrIday, July 24, 1981 I Notices

lactation and then for 104 weeks. Searle (3) ConclUSIOn on E-83/34: For the the aspartame-treated groups combined, agrees with the Bureau on these pomts::~ reasons stated above, I conSIder-E-33/34 2.5% (4/157), was "well above the Whenever a smgle tumor has such a to be a negative study. normative figures" (id,). Finally, the large unpact 'On observed probabilities, 3. Study E-70 (Vol. 80) 38_a•Study Board critiCized the study's size, staUng: caution should be used m evaluating the deSIgn. ThIs study was also conducted "thIs critically important study should results from any statistical test. Under on Charles River CD (Sprague-Dawley) have mcluded a larger number of the uruque facts presented here, I agree albmo rats usmg aspartame as the test experImental arumals" (id.). that the results with the Bureau of Foods compound. The protocol differed from (2) Positions ofthe parties. The from the Cox and Breslow tests should E-33/34 In that.the treated arumals were not be consIdered bIologIcally esposed to aspartame, through theIr Bureau and Searle have taken exception to thIs portion of Board's decision also SIgnificant. ThIs tumor likely ongmated mother's diets, both In utero and durmg ii1 embryomc tissue before aspartame / lactation, and then for 104 weeks as part (Searle's Exceptions at 29-30 and was admInIstered, and the absence of of theIr own diets. The Bureau requested Bureau's Exceptions at 35-36). They additional medulloblastomas In thIs Searle to perfOi'Ill a study with argue thatthe incidence rates are study or In E-70 support thIs conclUSIon aspartame exposure begmnmg at conSIstent with a correct assessment of (Tr./II1/page 204, lines 2-19 and page conception because of the known the histoncal control data, and that this 262, line 23-page 263, line 11). I sensitivity of the fetal or mfant rodent to study, evaluated on its own, showed therefore conclude that E-33/34 does toXiC effects from hIgh doses of glutamIC neither statistically nor biologically not exhibit a dose response. aCId and aspartic aCId [fr./II1/page 205, SIgnificant findings. Dr. Olney agrees (c) Time oftumor onset: The Board lines 18-24) (see generally Section N(G) with the Board that this study is was also concerned about what it above). defiCIent, due to his comparison to the perceIved as a "hIgh mCIdence of E-70 used two dosage levels, 2 and 4 histoncal control data (Tr./III/page 154, gliomas at a relatively early age: 5 rats g/kg body weIght/ day, m groups of 40 lines 1-3). died with glioma before completing the arumals per sex. A control group (3) Evaluation ofstudy. As explained second year of life" (Board's DeCISIon at ongmally conSIsting of 60 anunals per In detail m Subsection B above, I 46). According to the Board, ,these sex was also used. A treatment group disagree with the Board's determination ammals died at weeks 8, 16. 66. 84 and comparable to the hIghest dose In the E­ of the background rate for spontaneous 100 {id.}. 33/34 study (6-8 g/kg) could not be used brain tumors In Charles River CD Both the Bureau and Searle claun that because of exhibited non-specific toXiC (Sprague-Dawley) rats and, accordingly, the Board made a factual error with effects m fetal tissue caused by respect to the two arumals whIch decreased food consumption m the I disagree with the Board's dismissing allegedly died with a glioma at age 8 mother (Tr./II1/page 206, lines 11-24). the results of thIs study. and 16 weeks. Searle and the Bureau Test anunals were necropsIed at the Based on compansons with assert that the'''8 week" anunal really time of death, or at 104 weeks after concurrent controls, it IS clear that this died at 68 or 69 weeks, and the "16 wearung, whIchever occurred first. Eight IS a negative study. As Dr. Olney week" ammal really died at 76 weeks brams sections per anunal were admitted at the heanng: (Searle's Exceptions at 23 and Bureau's exammed hIstologIcally. As one can see, there Is no significant Exceptions at 15). Thus, the dispute here b. Study results. The tumor count 39 difference between the incidence of brain IS whether these anunals died early In and the number of anunals at nsk (as tumors between control and experimental theIr first year or well mto theIr second verified by UAREP (Vol. 111, page 559, ammals In the second aspartame study. year of life. Table V-20]) are as follows: After a reView of the relevant (Tr./II1/page 153, line 24-page 154, line documents, I agree with Searle and the Grams per kilogram Males Females 1);ThIS conclusion was confirmed by Bureau that the Board did mdeed make statistical analyses performed by the two factual errors. In actuality, these Controls2 ____ _ 3/58 1/57 Bureau of Foods (Tr./Ill/page 207,line 4 ___* •___ 2136 1/39­ anunals died at approXImately 69 and 76 1/40 1/40 2-page 208, line 19), Because of the In weeks, respectively (UAREP, VoI.l11 at utero exposure, the results of this study 403 [Animal No. 83-766} and 396 c. AnalysIs-(l) The Board's DecJSJon: alleViate the concern raised by Dr. (Anunal No. 83-837)). As noted above, the Board discounted Olney of mcreased nsk to children in The corrected figures are certamIy completely the results from thIs study, terms of bram tumors (Koestner, Vol. less dramatic; all arumals with gliomas calling them "bIZarre" because the 152, Section XI at 10). died either durmg the second year of life mCIdence of bram tumors m control One additional point which needs to or were sacrificed at the conclUSIon of anunals, 3.5% (4/115) was conSIdered be addressed bnefly IS study size. As the study. Moreover, none of the gliomas completely out ofIine with the were confirmed as bemg the cause of noted above, the Board suggested that background rate m hIstoncal controls thIs study should have included more death. As Dr. Koestner testified at the (Board's DeCISIon at 47). The Board also hearmg: ammals. The protocol used In E-70 found that the bram tumor mCIdence of called for 40 rats/sex In each of the two • * -* these ammals Just happened to die treated groups and 60 rats/sex in the from a non-tumor related cause and 38 Like &-33/34. t1us study IS reported m Its histological exammation of the bram reveals ongmal fonn [Vol. 80). in a pathology report by Dr. _ control group. Searle has demonstrated an unexpected microtumor wmch eventually Innes (E-87, Vol. 98), and m reVIew fonn by UAREP that thIS allocation of treated and would have shown up as a grossly detectable [Vol. 110 at 5-15; Vol. 111 at 458-577; and Vol. 112 at control ammals is comparable to the neoplasm had the ammal been permitted to 833-45). Bureau's current allocation standard (50 live. •• Although the Board reported only two tumors in the 2 g/kg group (both sexes combmed (Board's anunals/sex for both treated and conlrol (Tr./II1/page 255, lines 18-22; see also DeciSIOn at 42]), Dr. Innes and UAREP each groups) in terms of its ability to detcct Tr./II1/page 225, lines 5-11). reported-three tumors, two m the males and one in an mcreased tumor rate (Searlc's Accordingly, I find that there are no the females [Vol. 112 at 838, Table 9-1). The OJllISSIOn of the tIurd tumor may have been an Exceptions at 40. Chart 1). Thus. I do not biologIcally Significant findings of early oversight by the Board. All three have been collntcd share the Board's concern about study tumor onset. here. .slze. Federal RegISter I Vol. 46, No. 142 I Friaay. July 24. 1981 I Notices 38301

[q.) ConclUSIOn on E-70. For the scattered among the 4 test groups. There wns not exclude eVidence relating to the reasons stated above, I consIder E-70 to no sJgIlificant difference in the incidence of quality or appropnateness ofthe be a negative' ·study. bram tumors between control and test expenmental deSign ofthe studies or the groups. It is concluded that neither APM nor 4. E-77/78 (Vols. 89-90) 4O_a. Study DKP caused brain tumors in rats in thIs study. SCientific conclUS1ons that can validly be DesIgn. TIns study differed from &-33/34 drawn from the studies" (Board's and E-70 m that the test compound used Takmg the available mformation at face DeCISion at 7). What the Board did was.cliketopiperazme (DKP), a value. this appears to be a negative decline to do was to "undertake a breakdown product of aspartame [less studym terms ofbram tumors. Without retrospective quality inspection ofall­ than 2%) (VoL 112 at 30-31). Charles a reVIelV of the Bureau offoods. the studies presented to it" which the River CD (Sprague-Dawley) albmo rats however, as well as by other uiterested Board conSidered had already been were also used m this'study. Three parties. I do not believe it proper to base accomplished by UAREP andFDA (id.), treatment groups, consIsting of 36 rats approval of aspartame on this study's Quite clearly. the Board conSIdered its per sexper group, were fed DKP as part results. Nor IS such Ii course necessary charge, as delineated m the June 1. 1979 of their regular diets at dosage levels of m this Instance. The three chromc Federal Register statement, to relate 0.75, 1.5 and 3.0 g/kg body weIght/day studies discussed above (&-33/34. E-70. only to mterpretation ofthe data and for115weeks, begmmng after weanmg. andE-77/78) are suffiClentforme to not conduct ofthe studies. make a final determmation oh the safety A control group of72 anImals per sex Both Searle and the Bureau agreed was fed the same diet without DKP. Test of aspartame in terms ofits potential for bram tumors m rats, However, because with the Board's ruling on Mr. Turner's anImals were sacrificed at the end of the appeal (Searle's Replyto Turner's dosmg penod, and-their brams [seven the Japanese study suggests that Appeal. Vol. 157, Tab 200 and Searle's sections per anImal) were exammed aspartame does not cause bram tumors Reply to Turner's Exceptions; Bureau's hIstologlCally. m a second stram ohat. the SLC Wistar b. Studyresults. The Board reported rat, this studyprOVides additional Reply to Turner's Appeal. Vol. 157. Tab the followmg results and offered the sUEPort for my conclUSion on the bram 208; and Bureau's Reply to Turner's followmg evaluation: tumor ISsue. Exceptions). I believe the problem IS partly one of In the &-77/78 study concemmg the D. ConclusIon on BraIn Tumor Issue diketoplperazme of aspartame 5 tumors were semantics, as the phrase "SCientific recorded: 2 m the control group of123 rats For all the reasons stated above. I ' validity" mayhave several different (1.6%), and the remammg 3 among the 198 conclude that the available data, taken meanmgs. The Board11Ilderstood Mr. ammals of the three experunental groups as a whole, establish that there IS a Turner to mean thatit should redo (1.5%). Two of the 5 gliomas could have been reasonable certamty that aspartame and UAREP's work whIch was to noted on gross mspection of the bram. DKP do not cause bram tumors in authenticate the data (ie., make sure This study shows no difference between laboratory rats. 'flus conclUSion IS based that the studies were actually experunentai and control groups, and the recorded percentages fall within the lugh on studies &-33/34, E-70, and E-77/78: conducted). Clearly, the board was range ofnormal mCidence reported from all ofwhIch were conSIdered at the correct m not attempting to repeat vanous normative studies. hearmg. Additional support for thIs UAREP's work. The Board. inturn. uses (Board's DeCISion at 43) conclUSIon IS found m the Japanese the term "SCientific validity" to mean the study, submitted by Searle after the conclUSIOns that can be drawn from the c. AnalYSIS ofand ConclusIon on E­ Board Issuedlts deCision. Accordingly, data presented. mcluding study deSIgn. 77/78. None ofthe hearmg partiClpants under the act's general safety clause, I These conclUSIOns were clearly within challenge this mterpretation of the data. find that the available data establish the the Board's domam, and itwas based on Accordingly, I agree with the Board that safety of aspartame. m terms ofbrain these considerations that the Board E-77/78 IS a negative study. tumors, for its proposed use. reached its ultimate findings. There IS a 5. Additional eVIdence: The Japanese tlurd area, however, thatlies study [Searle's Exceptions, Appendix 2). VI. Mr. Turner's Appeal somewhere between those two. This This study was conducted only recently Mr. Turner and Dr. Olney have by the Japanese firm Ajinomoto Co .• relates to the mannerm whIch the repeatedly challenged the quality of studies were conducted. Even ifthe Inc., and concluded after the Board data produced m Searle's anImal Issued its declSlon. A prelinImary report studies were not fraudulent. thatdoes studies. Indeed, Mr. Turner has not necessarily mean that they were was submitted by Searle as part ofits petitioned for the Public Board of exceptions. I have consIdered this study well conducted. A non-fraudulent study InqUlrYlo be reconvenedbecause of the Jrught be conducted m such a poor as eVidence m thIs proceeding, Board's refusal to consIder what he acknowledgmg that neither the Board manner that its results would not be called the ·'sclentific validity" of the conSidered meanmgful [cf. 45 FR at nor the hearmg partiClpants have studies ~1 (Vol. 153, Tab 187; see also formally commented on it. , 61478, col. 2). As then FDA Crnef. Turner's Exceptions). Counsel Richard A. Menill wrote to Mr. The prelimInary report at{page 1) The Board disagreed with Mr. contams the followmg summary:. Turner on February 24. 1977. questions Turner's characterIZation that itfailed to regarding the "execution ofthe studies" The bram tumorgemcity of aspartame ·conslder the "sClentific validity'· of the could be l'alsed at the public hearmg (APMJ and ofits diketoplperazme (DKP) was studies, asserting that the Board "did studied m 860 SLC WlStar rats. APM at (Attachment No.1 to Turner's AppeaL dietary levels of1 g/kg, 2 g/kg. 4 g/k.g or Vol. 153, Tab 187). ClMr. Turner's full prayer for rellef Included: AMP+DKP (3:1) 4 g/kg was fed for 104 1. An order directIng the Board to reconvene and I conclude. however. that a nell/' ,_ weeks. One atypical astrocytoma was found consider whether certain studies hnve been hearmg need not be held. With one m a control rat and 2 astrocytomas, 2 validated: • exception discussed below, Mr. Turner oligodendrogliomas and1 ependymoma were 2. An order direcUng an oddlUOIU1! Board or other has not stated with particularity any public InvesUgatory body to vollpote these Gtudles; deficiencies m the conduct ofany of the 40Tlus study was Dot reVlewed byDr. Innes. and Neither was itreVlewed by UAREP. although a 3. Withholding of nsportamll's approval unUl such pertinent studies whIch he believes, smillar authentication was performed by the agency valJdaUon Is complilted. either alone orcollectively. are (voL 151, Tab 167]. (VoL 153, Tab 187 at 24-ZS] sufficiently senous as to warrant a

,~ 38302 Federal RegIster I Vol. 46, No. 142 I FrIday, July 24, 1981 I Notices

study's lOvalidation. 42 Rather, Mr. conducted by FDA. 43 The pertinent (Vol. 110 at 2) (emphasis added). Indeed, Turner's (and Dr. Olney's) malO documents were placed lOtO the. record UAREP addressed such issues as: (1) critiClsms appear to be mere by the Bureau shortly after ilie hearIng, Protocols: (2) climcal observations: (3) speculations WhICh fail to ralse any at the request of Dr. Olney and Mr. body welght, food, and compound genwne Issue of materIal fact. 'Turner (Volume 151, Tab 167). The consumption: (4) survival data: (5) For example, Mr. Turner and Dr. documents lOclude portions of FDA's climcallaboratory studies: (6) Olney rely heavily on the 1976 on/site lOspection report of Searle as ophthalmoscopIC observations: (7) CongressIOnal testimony of then well as a Task Force memorandum necropsy: and (8) histopathology (Vol. COIDmlsslOner Alexander M. Sclmudt lOterpreting and commenting on that 110 at 5-15) as well as (9) personnel,' who characterIzed Searle's anlIDal report. facilities and methods: (10) animals and laboratory practices as "sloppy" lfr./ The agency's mvestigation culmmated ammal care: and (11) data production, III/page 129, lines 1-4). That testimony 10 a Bureau Task Force Report wruch handling and storage (Vol. 110 at 20-22). was based,on findings of an FDA thoroughly discussed the homogeneity The FDA portion of the audit had a lOvestigation of two of Searle's drug Issue. The Task Force concluded that, slIDilar scope. These are very similar studies whlCh only perIpherally although the homogeneity Issue could subject areas to those which Mr. Turner concerned aspartame. The relevance of not be conclUSIvely resolved, no senous raIses in hIS appeal (see Vol. 153, Tab thIS lOvestigation to the aspartame problems were encountered wruch 187 at 14-15). Yet, not once does Mr. proceeding IS that it trIggered the would mvalidate the study. The remedy Turner cite examples from the UARE!> detailed audit conducted by UAREP and advocated by the Bureau, and adopted report as eVIdence of poor conduct of the agency, and therefore, for the by the agency, was to notify Searle by the studies. His request for a new purposes of tills proceeding the drug letter of laboratory practices wruch "validation" reVIew, therefore, appears study investigation was superceded by should be corrected 10 the future (see to be merely a fishing expedition for the UAREP/FDA audit. Nevertheless, Memorandum for the Files, dated eVldence of "sloppy" laboratory based on the "sloppy" laboratory September 26, 1977 prepared by Taylor practices. M. Qumn, and draft letter to Searle from practices theory, Dr. Olney attributed It should be emphaSIzed that UAREP. COIDmlSSIOner Kennedy (undated), both the slightly rugher lOCldence of bram a consortium of nine unlVersities. has 10 Vol, 151, Tab 167). tumors found 10 the E-70 control .Dr. Olney's one pIece of "hard unquestioned expertise in the area of ammals over concurrently treated eVldence" was a photograph of a feed preclimcal ammal testing and that its ammals to a hypothetical nux-up that nuxture shoWIng DKP particles larger reVlew of Searle's studies was may have occurred between the control than that of the feed, so that the anlIDals undertaken with complete neutrality. and treated groups (Olney's Pre-Hearmg Although UAREP, like the agency. noted 10 the treated group IDIght have Position Paper, Vol. 151, Tab 160, Part III some procedures and irregularities that discnmmated 10 favor of the smaller at 15). The speculation mherent 10 tills non-DKP particles (photograph attached warrant Improvement, none were of allegation was eVldenced at the hearmg to Olney letter of February 6, 1980, Vol, such a senous nature as to invalidate an when, as the Issue of the rugher control 151, Tab 165). entire study. Indeed, UAREP noted, and lOCldence 10 the E-70 anlIDals arose, 'the I agree with the Bureau that the I agree, that review of the follOWIng exchange took place: eVldence IS not suffiClent to mvalidate rustopathoiogic slides provides a better Dr. Spitznagel [Consultant to Dr. Olney]: tills study. The photograph 10 question baSIS for validation of the data than Our only comment on that is we have our was taken by a sample prepared -many of the other parameters (Vol. 110 SUspiCions, mainly that some of the controls espeCIally for stability testing purposes, at 23). On tills pomt UAREP noted were actually treated. not feeding purposes. As the '{~sk Force general agreement between its Dr. Bussey [Consultant to Searle]: Do you wrote: "it could not be determmed pathologIsts' reVIews and the original have eVidence to that effect? whether these samples were diagnoses (id. at 24-25). UAREP also Dr. Spitznagel: No, we really don't other representative of the diets fed to the noted that both Searle and Hazelton than the assertion of the Commissioner of rats, SlOce the batches were made up Laboratones were accredited by the FDA. specifically for tills analYSIS and were Amencan ASSOCIation for Accreditation rrr./lli/page 242, lines 20-25). made 10 smaller amounts" (Vol. 151, Tab of Laboratory Ammal Care which, at the 167, Task Force Report, Appendix A at time, carned out the most through and The only specific allegation by either 10-11). Thus, Dr. Olney's allegation here critical nationwide evaluation of animal Mr. Turner or Dr. Olney relates to the E­ also appears to be speculative. care facilities (id. at 20). 77/78 carcmogemcity study conducted Nor IS itnecessary to order a new Therefore, based on the extensive on DKP Dr. Olney cites a Bureau of validation of these studies, as Mr. mformation available 10 the record Foods report that raIses the possibility Turner suggests. Although the UAREP regarding the conduct of Searle's that the DKP-contaIDlDg feed may not audit was undertaken to determme studies, and Mr. Turner's failure to raise have been homogeneous (Report from .whether the aspartame studies were with particularity any specific issues Bureau 'of Foods' Task Force, September authentic or fraudulent, the three other than the one discussed above, Mr. 29, 1979, pages 10-11, Volume 151, Tab volume report coverIng over 1,000 pages Turner's appeal is denied. 167). Dr. Olney's POlOt here IS that the contam detailed observations of how non-homogeneous feed may have these studies were conducted. vn. Conditions of Use resulted 10 the "treated" ammals' UAREP has addressed itself to the question The thud Issue at the hearing was selectively not eating the DKP of whether the expenments were carrIed out deflOed as follows: The Bureau of Foods' documents at according to protocol plans and the accuracy Issue relate to the authentication reVlew and reliability with which the expenments Based on answers to tho abovo questions, were performed and reported to the FDA. (a) Should aspartame be allowed Cor use In foods, or, instead should approval of C2 Mr. Turner has had ample opportunity to do so, either at the he!lnng, as part of his "appeal" 43'&-7'1/'18 was one of the three studies which aspartame be withdrawn? submitted after the heanng, or as part ofhis FDA. rather than UAREP, audited (see Section I (b) Ifaspartame Is allowed for Use In Coodo, exceptions filed after the Board's decision. above). i.e,. ifits approval is not withdrawn. what Federal RegIster I Vol. 46, No. 142 I FrIday, July 24. 1981 I Notices 38303

conditions of use and labeling and label be harmful under its proposed uses. See mformation on aspartame's use to the statements shoUld be reqUIred, ifany? Section II. Bureau of Foods as the Bureau may by (44 FR at 31717). 3. The act places the burden on order deem necessary. prOVing safety on the company seeking The Initial DeCIsion of the Public The conclusIons reached m Sections approval of the food additive petition. IV and V>above compel the conclUSIOn Board of Inqmry IS affirmed m part and See Section II. reversed m part. as modified and ,­ that aspartame should be approved for 4. For Searle to obtam approval of its use m certam foods •. as listed m 21 CFR supplemented herem. food additive petition, it must prove that In accordance with section 409[£)[3) of 172.804. Equally clear IS the fact that the the data m the record establish that post-marketing restrIctions advocated the act (21 U.S.C. 348(£)(3)], the effective there IS a reasonable certamty that the date of thiS order IS October 22, 1981. by Dr. Olney (restrIct aspartame to use proposed uses of aspartame will not be only by obese and diabetic patients) and harmful. See Section II. Daled: July 18, 1981. Mr. Turner (reqUIre a warmng statement 5. The data m the record establish that Arthur Hull Hayes, Jr.. on all labels stating that aspartame the maXImum projected daily Comnusslonero/Food and Drugs. should not be used by children) are not consumption of aspartame IS 34 mg/kg/ supported by the SCIentific eVidence. AppendixA-Board's Deciston on day. See Section !V(A). Potential Bram Damage From The conditions for use stated m the 6. Based on the maxImum projected aspartame regulations (21 CFR 172.804), daily consumption, the data In the Phenylalanme mcluding.labeling reqUIrements, are record establish that there IS a A. Diffuse Bram Damage AssaclOted affirmed m thell' entirety. These labeling reasonable certamty that the ingestion With Abnormally High Plasma­ reqUIrements mclude: (1) A prommently of aspartame. either alone or together Phenylalamne Levels: Phenylketonuria displayed alert to persons with PKU that with glutamate. does not pose a nsk of the product conlams phenylalanme contributing to mental retardation, bram Phenylketonuna (PKU) IS an mherited disorder m the metabolism of ("Phenylketonuncs: Contams damag~, or undesll'able effects on the Phenylalanme"); (2) directions not to use neuroendocrme regulatory systems, See phenylalanme. It IS transmitted by an aspartame m cookmg or bakmg because Sections !V (B) and (C)., autosomal receSSlVe gene, and its the compound loses its sweetness when 7 The data m the record establish that inCIdence m the United States IS about 1 in15,000. The disorder results from the exposed to prolonged heat; and (3) there 15 a reasonable certamty that the labeling m compliance 'with FDA's mgestion of aspartame does not mduce absence of an enzyme (phenylalanme speCIal dietary foods regulations (21 bramneoplasms (tumors) m the rat. See hydroxylase) that converts CFR Part 105) ifthe food contammg Section V. phenylalanme (PHE) to tyrosine; as a aspartame purports. or IS represented, to 8. Searle has met its burden of provmg consequence PHE accumulates in body be for speCIal dietary uses: that aspartame IS safe for its proposed tissues-mcluding blood-m abnormally Ingh concentration: m untreated -­ The safety evaluation m Section IV uses. Aspartame should therefore be allowed for use In foods as set forth m phenylketonuncs plasma-PRE levels above calls for' one additional post­ usually range between 120-600 p.mol/ dl marketing reqUIrement One assumption 21 CFR 172.804. See Sections ill, IV. and V. (20-100 mg %) mstead of the nonna16-12 ,m tlns proceeding IS that extremely hIgh p.mol/dl. Through mechamsms not yet amounts of aspartame's component 9. All the conditions of use contamed m the aspartame regulation (21 CFR fully understood. these grossly elevated anImO aCIns may cause bram damage. PHE concentrations are correlated with Aspartame IS being approved only 172.804). Including labeling reqUlrements. should be required. In 'severely lDlpmred development of the because the available data establish immature bram m general. and of the ·that theJIlaxIIDum projected addition. post-marketing surveillance by Searle of aspartame's actual use levels myelin sheaths of its nerve fibers m consumption of aspartame IS still far, far particular. The clinIcal consequence of below any level even suspected ofbemg 16 necessary to ensure that actual use remams well below suspected tOXiC tlns developmentallDlpaU'IDenhs a tOXiC. Neverthless, prudence dictates profound mental retardation, often that these estimated use levels be levels. See Section VII. The foregOing Final DeciSion m its accompamed"by epileptic selZ\ll'eS and compared to actual use levels to ensure chrome dermatitis. Children born with the validity of the safety assessment. As entirety constitutes my findings of fact and conclUSions oflaw. ' the enzyme defiCIency can develop to a condition for approval, therefore, adults of normal mtelligence, prOVided Searle IS to monitor the actual use levels IX. Order thell' condition IS recogruzed soon after of aspartame and to proVide such In accordance with subsections bll'th. and appropnate dietary treatment mformation on aspartame's use to the mstitutedpromptly thereafter. Ills , Bureau of Foods as the Bureau may (c) (3)(A), [£)[1), and (£)[2) of section 409 of the act [21 U.S.C. 348[b)(3)(A). (£)[1). estimated that the PKU newborn loses deem necessary by an order. m the form one percentage pomt of future of a letter, to Searle. and (£)[2)) and 21 CFR 12.130, and under the authority delegated to the intellectua1.capacity for each postnatal VIII. ConclusIOns COmmIssioner (21 CFR 5.10 (formerly 21 week the condition goes uhrecogruzed CFR 5.1)), it IS hereby ordered that: (et Dr. Richard Koch's testimony at the Based on the foregOing, I conclude 1. Approval of the food additive public hearmg). Treatment IS anned at that: petition for aspartame (FAP 3Ma85) is keepmg plasma-PHE concentrations at 1. Section 409[c)[3)(A) of the act (21 granted. or below 71HlO p.mol/ dl by restrIcting U.S.C. 348[c)(3)(A)) 'permits FDA to 2. The stay of the effectiveness of the the dietary mtake ofPHE. Iftlns approve a food additive petition only if regulation for aspartame (21 CFR preventative regImen IS to successfully a fall' evaluation of the data establishes 172.804lis vacated and the regulation mamtamed. families with a that the food additive will be safe under remstated. phenylketonunc child must lDlpose upon its proposed uses. See Section II. 3. As a further condition for approval the child a strIct dietary disCIpline that 2."Safe" means a reasonable not listed m 21 CFR 172.804. Searle is to cannot be relaxed until the child IS certamty m the mmds of competent monitor the actual use levels of adolescent. ItIS lDlportant to note, sCientists that the food additive will not aspartame and to proVide such however. that phenylketonurIC mental 38304 Federal RegIster I jlo1. 46, No. 142 I FrIday, July 24, 1981 I Notices aLi t - retardation IS conditional upon corresponds to 600 aspartame tablets, or 4. Undetected cases of sustaIned hIgh plasma levels (If PHE, m 24 liters of aspartame-sweetened p/lenylketonuna. The question has been contrast to the more focal bram damage beverage consumed m a smgle sitting by raIsed whether a nsk might occur in that can result-as will be emphaslZed a 50-kg adult, or to 100 tablets of 20 mg umdentified PKU children as a m a subsequent section-from a smgle, aspartame aCCIdentally mgested by a 3­ consequence of the presence of short-lived surga of glutamIC or aspartic year old child. Only m tlus grossly aspartame m the food supply. The aCId concentration m the blood plasma. abUSIve amount could aspartame number of children in thiS category is The essential question with wluch the mgested by a pregnant woman be unknown but thought to be very small. Board found itself confronted m expected to mduce plasma-PRE Sereenmg of newborns for PKU is examInIDg the phenylalamne Issue IS: at concentrations lugh enough to cause, mandatory In 47 states. and it has been what level of mgestion could aspartame through placental transfer, fetal plasma­ estimated that about 10% of the 200 PKU mduce a rIse m plasma-PHE PHE levels approachmg-for a few children born annually m the United concentration to 100 p,mol/ dl or lngher­ hours at least-the lower limit of States nught remain undiagnosed and the levels assocIated with lIDpalred potential tOXIcity. However, it seems hence at great nsk to grow up retarded bram development? It IS clear that tlus mconcelvable that so large a dose (cf. Dr. Richard Koch's testimony at the question IS of particular lIDportance m would be takan m a smgle sitting. Whan public hearmg). An undetected the case of children under 12, whose consumed over a 16-hour penod-as phenylketonunc mfant would be bram IsstilllmIDature, and m the case adversely affected by the phenylalanine of women m the child-bearmg age. The would seem nearly unaVOIdable-it would undoubtedly mduce a more prOVIded m breast milk protem (or Importance of the question for the latter mfant formula) wluch may furrush levels category IS accentuated by the well­ sustamed plasma-PHE elevation remammg well below the 50 p,mol/dl ofphenylalanme mtake in the vicinity of established fact that the placenta 80 mg/kg/day. (Tlus compares with a mamtams between the maternal and peak mduced by the same amount of aspartame taken as a loading dose. projected mean phenylalanme mtake fetal CIrculations a 1:2 gradient m the from aspartame m children under 2 plasma concentrations ofmost amInO 2. In the normal one-year oldmfimt, a years of3 mg/kg/day). The argument aCIds, mcluding phenylalamne. Tlus loading dose of 34 mg/kg body weIght that asartame m the food supply would means that or the fetal plasma-PHE causes the plasma-PHE concentration to sIgnificantly mcrease the risk of mental concentration to reach the 100 JLmol/ill nse from a fasting level of 6p,mol/dl to retardation in the umdenUfied level, the maternal plasma-PHE 10 p,mol/ dl. receding to baseline withm phenylketonunc IS not supported by concentrations needs to nse no lugher 4 hours. It appears from tlus finding that these conSIderations. An undiagnosed than 50 p,mol/ dl. the 1-year old normal child metabolizes PKU child is at nsk fIrst and foremost by Of the eVIdence presented the Board PHE at least as effectively as does the beIng undiagnosed and hence permitted conSIders the followmg data of normal adult. to consume meals that are standard for particular sIgIlificance: 3.·InmdivIduals heterozygous for normal children. This pomt is 1. In normal human adults, the phenylketonuna, a 34 mg/kg loading mgestion ob smgle loading dose of34 emphasized further under the next item dose of aspartame mduces a lugher and \ of conSIderation. mg/kg body weIght aspartame (the 99th longer-lasting plasma-PHE elevation. percentile of projected aspartame 5. PKUchildren who are not on a Instead of the 11p,mol/ dl peakresulting restrIcted diet. As PKU childran get consumption for an entire day) from such a loading dose m the normal dissolved m orange Jwce mduces a nse older they may be allowed larger human. the peak reaches 16 p.mol/ dl m helpmgs of "free" food or they even go m plasma-PHE concentration from a the PKU heterozygote and. m addition, fasting level of 6 JLmol/dl to 11 p.mol/ dl, off theIr earlier retncted diet. This may the plasIl!a-PHE curve declines more a level normally found m adMts and not be harmful provided that the ochild'9 children follOWIng ingestion of a protem­ slowly than it does m normal tolerance to phenylalarune IS carefully rICh meal. Tlus peak value IS reached mdiVlduals. A loading dose of100 mg/kg monitored by blood tests. However, the about one hour after the aspartame aspartame-an abuse load even whan question anses whether the availability mgestion, and recedes to fasting level Ingested over a 16-hour penod-Is of aspartame in the food supply would withm about 8 hours. followed by a plasma-PHEnse reachmg comprOmIse the health and well-being of Ingestion of larger loading doses 42 JLmol/dl. about twIce as !ugh as m the PKU children in tlus category. There mduces proportionately higher plasma­ normal human. Even folloWIng tlus appear to eXIst no explicit data based on PHE elevations. A 50 mg/kg loading enormous smgle load. however, the peak controlled studies to answer this dose (in a 60 kg person 3.000 mg value remams below the level at wluch' question, but it IS possible to seek an aspartame. or 150 aspartame tablets. or m the case of a pregnant woman, a nsk answer by conSIdering the amounts of 6 liters of aspartame-sweetened to her unborn child nnght arlSe. phenylalanme that such children would beverage, but with its 50% content of Moreover, an abuse dose of 100 mg/kg be exposed to through usual food PHE.eqwvalent to less than half the aspartame would m the real-life sources. m companson with the PHE 4,000. mg PHE contamed m one 4-0Z. situation not be mgested m a SIngle prOVIded by aspartame. For example. a hamburger) causes the plasma-PHE sitting, as' itwas m the cited 4-oz. hamburger supplies about 4.000 rog level to rIse from 6 to 16 p,mol/ dl. expenments. but, rather. consumed over phenylalarune, and a normal chUd Followmg a 100 mg/kg loading dose an extended time penod. Under these would consume an average of about 200 (eqUivalent to 12 liters of aspartame­ more natural conditions. the plasma­ mg phenylalanme per kg/day from sweetened beverage consumed m a PHE concentration could be expected to normal food protem sources. ThIs intake smgle sitting) the plasma-PHE level nses remam well below the 42 p,mol/dl level. level compares with a projected dally to 20 p,mol/ dl. Only a 200 mg/kg loading It IS of interest to note that a 100 mg/kg aspartame-based phenylalanine intake dose was found to mduce a nse to 50 mtake of aspartame by a 50-kg woman of 17 mg/kg by those children whose /Lmol/ dl. and only folloWIng tlus very • would add less to her dietary PHE aspartame consumption would reach tho large dose did the plasma-PHE consumption than would be added by an upper 99th percentile of the population. concentration' take more than 8 hours to extra 4-0Z. hamburger: 3.000 mstead of (For a 30-kg child thiS would correspond '.retum to baseline. ThIs 200 mg/kgdose 4.000 mg PEE. to a daily consumption of 2 helpings of Federal RegIster I Vol. 46, No. 142 I FrIday. July 24. 1981 I Notices 38305 , aspartame-coated breakfast cereal plus In evaluating the rIsk Inherent In consumed In very large amounts. The 8 cans of aspartame-sweetened aspartame consumption by hyperphenylalanmennc graVida not on a beverage). Thus. for children on an hyperphenylalanmemlcs. it IS obVIOUS PHE-restrIcted diet would add 5-6% to unrestrIcted diet aspartame Ingestion that aspartame as a source ofPHE can her dietary PHE mtake when consunnng even at thls hIgh level would contribute only contribute further to the already aspartame at the prOjected upper one­ less than 10% of the total daily PHE hIgh plasma-PRE levels. Itshould be percentile level. mtake. For children whose protem consIdered. however. that even the unlikely abuse mtake of100 mg/kg of Appendix B.-Board's Decision on mtake IS restrIcted the relationshIps Potential Bram Damage From Aspartic between food protem-denved and aspartame per day by a 6O-kg woman aspartame-denved phenylalanme would would supply less PHE (3,000 mg) than ACId differ. but agam the total Intake -would be supplied by an extra 4-oz. B. Focal Bram'LeSIons prOVided by aspartame remams small. hamburger (4.000 mg). and that the more In considenng the daily vanation In likely (although still very lngh) intake of Since first demonstrated m 1969 by protem Intake and the concentration of 34 mg/kg/day would be the PHE­ Olney and coworkers m the monse, it phenylalanme prOVided by normal foods eqmvalent oflittle more than two extra has become generally reCOgnIZed that it IS eVident that the Ingestion of glasses of milk. It thus seems fair to the aCIdic, dicarbo~lic ammo aCIds aspartame could not pose a sIgIlificant conclude that the glutannc aCId (GLU) and aspartic aCId extra nsk to PKU children whose dietIs hyperphenylalanmennc woman is at (ASP). when present m the blood plasma eithernot restrIcted or only partially much hIgher rIsk from the consumption in adequately hIgh concentration, can restrIcted. The sIgIlificant nsk to their of natural foods that she would be from cause death ofnerve cells In the central health IS clearly from the phenylalanme the use of aspartame. Itshould be nervous system. As far as IS known at m the protem furnIshed by standard reiterated that the real problem of present, thIs neuronal necrosIS IS focal foods: In a 30 kg youngster one extra hyperphenylalaninenna lies In the rather than diffuse: it IS certam thatit hamburger would add 100-150 mg/kg. usually covert nature of the anomaly. preferentially affects (1) the infundibular one extra hot dog about 50 mg/kg. one ConclusIOns Regarding Aspartame­ regIon of the hypothalamus, (2) the so­ extra glass of milk 15 mg/kg or nearly as Induced Mental Retardation called cIrCumventrIcular organs (the much as the total amount of PHE areapostrema. the subforrucal organ, In the Board's oplIDon, aspartame supplied by a 34 mg/kg Intake of the subcomnnssura1 organ, the vascular cannot aspartame. consumption bynormal humans organ of the lamma termmalis). and (30 be expected to Increase the InCIdence of the retina. 6. HyperphenylalanmemIa. ThIs term that particular form of mental 'refers to a condition In whIch plasma­ retardation that IS assocIated with The eVidence that aCIdic ammo aCIds PHE levels anomalously range between sustamed elevation ofplasma-PHE are potential neurotoXInS naturally has 25 and120 p,mol/ dl. Most of those levels to (or beyond) 120 p.mol/dl durmg rrused questions with respect to the afflicted with thls abnormality .are of munature stages ofbram development safety ofaspartame as a food additive. normal intellect. and smce they are ThIs conclUSIOn IS based on the Roughly one half ofaspartame'e usually asymptomatic also; neither they consIderation that even the hIghly molecular weIght IS contributed by its nor others are likely to be aware of their unlikely daily consumption level of100 aspartic-aCId mOIety, and itIS condition unless it has'been Identified mg/kg of aspartame (3 times the appropnate to ask whether its by a newborn-screenmg test The prOjected upper one-percentile of consumption could entail a nsk offocal InCIdence ofhyperphenlyalanmenna IS aspartame consumption) would add no bram damage. Before considenng the about-lhoooo. and it has been estimated more than 15-20% to the normal dietary eVidence it IS necessary to pomt out that that In the United States the condition PHE Intake, less than would be added m there are at least two reasons why thIS affects-about 1.750 women of a 50-kg IndiVidual by an extra 4-oz. question concemmg aspartic aCId cbildbeanng age. ItIS thls latter hamburger. Consumed at the estimated cannot be exammed separately and category that gIves the most reason for upper one-percentile level ofM mg/kgl must be conSIdered together with a concern. smce the 50% among these day, aspartame would mcrease the similar question concernIng glutannc women who have plasma-PHE levels normal daily Intake of PHE by no more aCId, a food additive already InWIde use rangmg between 60 an120 p,mol/dl are than SIX percent. These figures lie well in the United States and elsewhere: (1] at lngh nsk ofgIVIng birth-to bram­ withm the limits of day-to-day Both of these ammo aCIds appear to be damaged Children destined to grow up variations In dietary protein eqmpotential and mutually additive in mentally retarded. The only effective­ consumption. their neurotoXiC effects, and (2) a prevention of thIS consequence of InmdivIduals on a PHE-reslrIcted SIgnificant proportion of ingested hyperphenylalanmenna would COnsISt In diet deSigned to prevent critically aspartic aCId m the course ofits a systematic reduction of dietary PHE elevated plasma-PHE levels, aspartame metabolismIS transannnated to glutamic Intake through pregnancY-In other IS to be handied as any other source of aCId. For these reasons, itIS the words. In treating the prospective phenylalanme. Since these mdiVIduals combmed GLU-ASP content of blood mother much as a phenylketonunc child (phenylketonuric children and pregnant plasma that ultimately.mnst be would be ,treated. Su~ prophylactic women known to have considered, rather than the plaSl;lla ASP measures. however. are naturally hyperphenylalanmanna) would follow a level alone. Itis also for these reasons contingent upon Identification of the carefully prescribed diet, a cautionary that the Board permitted a volnnnnous anomalous condition before or shortly label explicitly identifymg aspartame as • body ofdata concermng glutannc aCId to after the begInnmg of the pregnancy. It a PHE source should forestall a liberal be presented, even though aspartame follows that until such time as all use of thls sweetener by such patients. itselfis free from thls ammo aCId. hyperphenylalaInnenncs are Identified In the unfortunate case ofUnIdentified Throughout the follOWIng survey ofdata by screenmg tests a complete prevention hyperphenylalanmemIa, the normal itIS assumed that glutannc aCId or of congenital bram damage caused by food-denved PHE poses a much greater monosodium glutamate (MSG) is maternal hyperphenylalanmenna cannot risk to the patient (or the unborn child) exchangeable with aspartic aCId or realistically be hoped for. than would aspartame, even when sodium aspartate m the sense that the 38306 Federal Register / Vol. -46, No. 14Z I FrIday, July 24, 1981 I Notices. neurotoxlC threshold levels of these studied only ill a small number of stands at present unconfirmed) and substances m the blood plasma appear speCIes. Two separate groups of' conSIders that it Imposes a qualification to be approXImately the same. mvestigators have reported that ill the upon those statements according to Focal Bram LeSIOns Induced In pregnant mouse MSG must be mJected whIch no focal brain lesions have been ExperImental Ammals byMonosodium ill very large amounts (5000 mg/kg) to mduced m any speCIes by voluntary Glutamate mclude hypothalamIC lesIons ill her consumption of any amount of GLU or fetuses. ThIS fmding accords well with its monosodium salt. A rough There IS general agreement among the eVidence (consIdered m more detail calculation suggests that the weanling mvestigators that rugh doses ofMSG below) that the placenta ill the monkey rats had mgested a mirumum of 13 mg o( admlmstered either by subcutaneous, mamtaIns a rughly effective bamer GLU with the drmkmg water. Assuming mtraperitoneal or mtravenous mJection, agamst both GLU and ASP' only at that body WeIghts ranged between 10 or by gavage (stomach mtubation), can grossly elevated maternal plasma-GLU and 15 g, ibIS illtake corresponds to a mduce hypothalamIc lesIons m a vanety levels (280 /Lmol/ dl) does GLU ill thIs loading dose of 900 mg/kg to 1300 rng/kg of rodent speCIes. Of all expenmental mammalian speCIes begm to enter the body weIght. ammals used m such experIments the fetal cIrculation. A somewhat sImilar Focal Brrun LeSIons Induced in mfant mouse, 1-10 days old, has been barrier appears to be mailltamed by the experImental AnImals byAspartame found most vulnerable to the neurotoXIc mammary gland: In the lactating human action ofMSG: a SIngle dose of 350 mgt female at least, the mgestion of In the Infant mouse. 2000 mg/kg kg mJected subcutaneously, or of 500' relatively rugh doses ofMSG does not aspartame admmIstered by gavage in mg/kg admInIstered by gavage, IS SIgnificantly affect the GLU content of the form of an aqueous slurry has been enough to cause, withm a few hours her milk (see below). reported to cause hypothalamic lesions time, a mIcrOScopICally Visible lesIon of Dietary Intake ofMSG by m 39% of subjects. No such lesions were the hypothalamus ill about half of the experImental arumals. In all of the found many 9-day old mouse gIven 500 mfant mIce so treated. Correlatild with anImal expenments mentioned m the mg/kg aspartame by gavage. It seems thIS 50%-effectiveness level ofintake IS foregomg account, MSG was either reasonable to assume that m the infant a rise m plasma-GLU concentration from mJected, or admmIstered by stomach mouse the nsk of hypothalamic lesions a baseline value of about 15 /Lmol/ dl to tube ill the form of an aqueous solution. begms to arise at a dqse level o(1000 100 /Lmol/dl. With mcreasIng maturity Markedly different effects upon plasma­ mg/kg aspartame administered by mICe become.more resIstant to MSG: ill GLU concentrations have been reported gavage. ThIs dose approximately weanling mIce a 50% effect reqUIres an from expenments in wruch mIce were corresponds to 500 mg/kg aspartic acId. MSG dose of 1200 mg/kg admmIstered by gavage and resulting m a plasma­ gIven MSG mIXed with food. Mixed with Since neither the same dose nor very GLU concentration of about 380 /Lmol/ "infant formula" or with a "soup diet," much hIgher doses of aspartame dl. In adult mIce the critical plasma-GLU and admuustered by stomach tube, MSG consumed by lIDIDature mIce as part o( concentration lies near 600 /Lmol/dl. m weanling mICe has been reported to the daily diet have been found to induce Other non-pnmate mammalian mduce a rIse of the plasma-GLU endocnne disorders (see below) it seems speCIes seem generally less vulnerable concentration only one-fifth to one-thIrd warranted to conclude that the to the neurotoXIc action of MSG. as large as thatcaused by the same resorption and/or metabolism of Although the Infant rat is nearly as amount ofMSG mIXed with water. aspartic aCId depends upon the route by sensitive to MSG as the Infant mouse, Ingested-by adult mIce as a food wruch thIs ammo aCId IS admintstered. the 50%-effect dose m the adult rat lies additive illthe enormous amount of Much like MSG. aspartic aCId ingested near 4000 mg/kg by gavage. The critical 20,000 mg/kg, MGS has been reported to as a food additive has been reported to dose m the 2-3 day old gurnea PIg IS mduce peak plasma-GLU concentrations mduce elevations of the plasma-ASP about 2000 mg/kg. In dogs 3-35 days old no rugher than 174/Lmol/dl, little more level smaller than those induced by an mtake of1100 mg/kg by gavage fails than one-quarter of the plasma level aspartic aCId admmIstered by gavage or to mduce hypothalamIC leSIons, as do (630 /Lmol/dl) that IS correlated with subcutaneous illJection. Further data doses of up to 4000 mg/kg ill adult dogs. hypothalamIC lesIOns caused by concermng this pomt will be considered Data for the monkey are controversIal. subcutaneous mJection of 1500 mg/kg m a-subsequent review of aspartame The Board IS unable to resolve the MSG. ItIS relevant m thIs context that consumption in the human. the arcruvalliterature mcludes no report conflicts that arose over thIs lssue-at the Neuroendocrme Disorders Induced by public hearmg. However, to remam on of bram leSIons mduced m any speCIes by dietary illtake of any amount of MSG andAspartame In experImental the SIde of safety it accepts the claIms: AnImals (al That a dose of1000 mg/kg ofMSG MSG. admmlstered by gavage or subcutaneous A.postscrIpt to these negative .findings ·In View of the topograpruc Injection can cause mIcroscopIcally must be made. In a post-hearmg characterIstics of its neurotoXIc effects it detectable hypothalamIC lesIOns ill commumcation dated April 3, 1980, to IS not surprIsmg that MSG administered mfant monkeys rangmg between the Board and to lus co-partiCIpants ill ill large amounts by subcutaneous prematurely born and 7 days of age, the heanng, Dr. Olney reported havmg illJection has been found to illduce and, (b) that mtravenous Injection of found clear-cut hypothalamIC leSIOns m endocrme disorders in mtce, rats, and 2000 mg/kg of MSG m the pregnant all of 10 weanling mICe who-after hamsters. In all of the studies from monkey can mduce such lesIOns in her haVIng been depnved ofwater wruch such disorders were reported, fetus. Despite eXIsting controverSIes the overmght-had drunk 0.2-0.35 ml of subjects had received either a Gingle Board also accepts the suggestion that either a 10% aqueous GLU (pt;esumably subcutaneous illjection of 3000 rng/kg the plasma-GLU level critical for the l-glutamIC aCId) solution or a solution MSG on the second postnatal day, or a occurrence of hypothalamIc lesIons m contammg 6.5% GLU, 3.5% ASP, and 1% daily injection of ZZO(}-4000 rng/kg for 10 the Immature monkey lies ill the ViCInity aspartame, while concurrently days starting on day 2. Prominently of120 /Lmol/ dl. consummg anunspecified amount of listed among the consequences of such MSG neurotoXIcity ill pregnant or Purma mouse chow. The Board accepts treatments are: stunting of body growth, -lactating.ammals appears to have been thIs eVidence (acknowledgmg that it obesity, and sterility in the female. Federal Register / Vol. 46. No. 142 I FrIday. July 24. 1981 I Notices 38307

Although apparently not explicitly aspartame rangmg between 1800 and 4. In one-year oldmfants, aloading demonstrated thus far. it seems 3700 mg/kg, begmnmg on postnatal days dose of100 mgfkg aspartame inducesa reasonable to assume thatm the same 10-20 and ending on days90-100, did nse ofthe plasma ASP levelfrom a speCIes subcutaneous mjection of not affect fertility, ge.station, live birth, baseline oU.5 pmol/dlto 2.6 pmoI/dl. smillar amounts of aspartate. or litter SlZe, or nursing m either the. receding to baseline m l-zhours.This admmrstration ofaspartame-by gavage: experxmental subjects or thelr offspnng. finding appears to refute any sugge.stion m twIce these amounts. would have The results of several further studies that aspartic aCId nught De metabolized snnilar endocrme consequences. Itmust presented at the hearmg likeWISe. less efficlentIyinmfants thanm adults. be stressed. however. that no studies mdicate that endocnne disorders are S. In PKUheterozygote adults concermng the endocrme effects of mduced by MSG only when thrs aspartame loading doses of34 mg/kg subcutaneous 01: mtragastric substance IS adminIstered m amounts and 100 mg/kg are metabolizedmuch as admmrstration of either MSG or large enough to cause Identifiable they are m normal mdiVlduals. The aspartate appear to havebeendone in hypothalamIC leslOns. The experrmental resulting rise m plasma GLU level IS specIes other than rodents. Hence. at 'eVIdence thus appears to argue agaxnst VIrtually thesame m bothcategones of presentnothmg can be SaId concermng the notion of sub-neurotoXIc e.ffects subjects. while the nse m plasma ASP the relagye susceptibility ofthe upon the neuroendocrme 8.XlS. level IS slightly, butnot SlgnificantIy. endocrine system ofvarious non-rodent Glutamate one! Aspartame Consumption hIgher: Plasma GLU+ASP level reaches speCIes to parenteralfy admmistered mtheHuman a mean of4.5 pmol/dlm normal adults. MSG or aspartate. a mean of4.8 JUIlol/dl m PKU Among the data presented on thIs heterozygote adults. Neuroendocrme Effocts ofSub­ subject, the. Board conSIders the 6. In the lactating lVoman. a loading ,neurotoXic Doses ofMSG and followmg pragmatic eVIdence of Aspartame dose of50 mg/kg aspartame (about 150 particular relevance. aspartame tablets) mduce.s no adult, a loading dose oL34 One ofthe oblecting;parties has 1.In the Significant elevation ofplasma ASP or mg/kg aspartame (the 99th percentile of stressed the possibility that a routine GLU levels. This dosage raises the ASP mtake ofMSG ot-aspartame several a prOjected mean daily consumption of concentration m hermilkfrom 2.3 to 4.8- times a day bX children throughout theIr 7-9 mg/kg, androughly eqUIvalent to formative years could entail repetitive 100 tablets of20 mg aspartame) JLmol/dl, the GLU conce.ntration from -disturbances m several neuroendocrine dissolved m orange JUIce mduces no 109 to 120 JUIlol/dJ.AttlnshighleveIof axes (e.g.• gonadotropms. growth SIgnificant elevation ofeither plasma maternal aspartame mta.'

ASP levels: The mother by a lughly It IS of some hlstonc mterest that aspartame (100 aspartame tablets) to tho effective barrIer ofASP resorption and! much of the eVIdence reported to the meal. or metabolism, the fetus m addition by Board concernmg the aforementioned 3. The PKU heterozygote adult is no an equally effective placental barrIer. question dates from recent years (1976­ less effectively protected against The mother herself has no comparably 1979), and consequently was not • aspartame-mduced surges of plasma effective defense agamst GLU, but available-at the time the objections to GLU+ASP concentration than the plasma GLU levels high enough to place the approval of aspartame as a food nonnal human. her at rIsk are not reflected m the fetal additive were.ongmally filed. With a 4. In the breast-fed infant, a blood plasma. smgle exception, the followmg consumption of 50 mg/kg aspartame by Risk Evaluation statements can at present be conSIdered the lactating mother results m an Justified by the results of experIments mcrease of no more than 0.77 mg/kg In attempting to assess the rIsk of done directly m the human rather than GLU+ASP over the normal daily intake focal (in particular, hypothalamIC) bram m one or more anlIDal speCIes: of 366 mg/kg GLU+ASP damage connected with human '1. The human orgamsm, Infant as well 5. The speculation that aspartame aspartame consumption, the Board as adult, IS protected agawst hIgh surges consumption by the pr()gnant women deCided to adopt a 100 IlmoI/dl of ASP concentration m either blood. could expose her Ietus to a high rJsk of concentration of GLU+ASP m the blood plasma or erythrocytes by a bIOlOgIcal focal bram damage ,cannot be plasma as the critical level. ThIs barrIer mechamsm presumably located mvesfigated directly m tha human. conservative assumption was made for m the gastromtestinal mucosa and/or However, expenmental findings in the liver. The effectiveness of thIs protective reasons of caution: 100 IlmoI/dlis the monkey mdicate that the primate concentration at whIch a 50% mechamsm IS illustrated by the placenta mamtams a nearly occurrence of focal bram lesions has observation that loading doses of msurmountable barrier agamst any aspartame as hIgh as 200 mg/kg body been reported for the Infant mouse, the transfer of GLU and ASP from the weight (in a 60 kg mdiVldual eqwvalent ammal form generally thought to be maternal to the fetal CIrculation. most sensitive to the neurotoXIc effects to 600 aspartame tablets or 20 liters of of glutamIC and aspartic aCId. the aspartame-sweetened beverage ConclUSIOn Regarding Aspartame.­ problem thus became reduced to the consumed m a smgle sitting) mduce an Induced Focal Brain LeSIons question whether, and at what level of elevation of plasma and erythrocyte consumption by the human aspartame GLU+ ASP concentration of no more In the Board's opImon, the most ,could mduce plasma GLU+ASP than 5 IlmoI/dl above a baseline level of pertinent eVIdence presented at the elevations approachIng the 100 IlmoI/dl 2.5-3 IlmoI! dl. It IS of added SIgnificance public heanng convlDcmgly level when taken alone, or alternatively, that these elevations are short-lived, demonstrates that the rIsk offocal brain whether it could significantly contribute receding to baseline in about 3 hours damage assOCIated with aspartame to such elevations mduced by MSG time. It follows that repeat-doses of the consumption In the human IS negligible, consumption. It should be recalled m same enonnous magnitude, when Elevation of plasma GLU+ASP • thiS connection that-unlike the bram spaced 3 hours apart, are unlikely to concentration even to the lowest level damage aSSOCIated with escalate the GLU+ASP concentration that could be suspected of bemg Ilhenylkalamne-the focal bram leSions much beyond the level mduced by the neurotoXIC (100 IlmoI/dlJ would require assoClated with GLU and ASP first dose. an mconcelVably high oral aspartame neurotoXIcity are not contingent upon a 2:The ASP plasma-entry barrIer IS mtake. Such levels mIght in fact prove long-mamtamed hIgh plasma unaffected by SImultaneously mgested attamable only by parenteral ASP concentration of the causative agent: It MSG: the 25 IlmoI/ dl plasma GLU+ASP admmistration deSIgned to bYl?ass the IS eVident from ammal experIments that concentration achIeved by adding to a highly effective intestinal and/or hepatio focal hypothalamic leSIOns can be protem-rIch meal a very large dose of barrIer mechamsm guarding agamst mduced by a smgle elevation of the MSG (150 mg/kg, or 9000 mg ill the case surges of plasma ASP concentration. plasma GLU and/or ASP concentration of a 60 kg person) IS not augmented by IFR Doc. 81-21696 Filod 7-22-81: 11:25 om) to the level of 100 Ilmol/ dl. the further addition of 34 mg/kg BILLING CODE 4UD-03-M