MAIN	cases	docket	decisions	orders	briefs	rules	guides	calendar
No. 98-1152: FDA v. Brown and Williamson Tobacco Corp.

No. 98-1152

In the Supreme Court of the United States

FOOD AND DRUG ADMINISTRATION, ET AL., PETITIONERS

v.

BROWN AND WILLIAMSON TOBACCO CORP., ET AL.

ON WRIT OF CERTIORARI
TO THE UNITED STATES COURT OF APPEALS
FOR THE FOURTH CIRCUIT

REPLY BRIEF FOR THE PETITIONERS


SETH P. WAXMAN
Solicitor General
Counsel of Record
Department of Justice
Washington, D.C. 20530-0001
(202) 514-2217



REPLY BRIEF FOR THE PETITIONERS


Respondents do not challenge here the findings that led FDA to concludethat the nicotine in tobacco products is intended to affect the structureor function of the body. In particular, respondents do not dispute that(1) the nicotine in tobacco products is highly addictive and acts as a sedative,stimulant, and appetite suppressant, (2) consumers use tobacco productspredominantly for those purposes, (3) manufacturers have known for yearsthat consumers use their products predominantly to obtain nicotine's pharmacologicaleffects, (4) manufacturers have privately referred to nicotine as a drugand cigarettes as devices for delivering that drug, and (5) they have longengineered their products to deliver to consumers the precise doses of nicotinethey need to obtain its powerful effects. Gov't Br. 3-8. The sole questionpresented therefore is whether, given those unchallenged findings, tobaccoproducts are drug-delivery devices within the meaning of the FDCA. As weshow in our opening brief, FDA reasonably concluded that they are.

A. Structure/Function Definitions

1. a. Respondents contend (RJR Br. 11-13) that a product falls within thestructure/function definitions only if a manufacturer makes a structure/function"claim." The term "claim," however, does not appearin the definitions. Instead, those definitions encompass as "drugs"and "devices" products that are "intended" to affectthe structure or function of the body, 21 U.S.C. 321(g)(1)(C) and (h)(3),and "intended" simply does not mean the same thing as "claimed."The dictionary definition of "intend" is "to have in mindas a design or purpose." Webster's Third New International Dictionary1175 (1986). The Court long ago stated that "[t]he law presumes thatevery man intends the legitimate consequence[s] of his own acts," Agnewv. United States, 165 U.S. 36, 53 (1897), and more recently, it interpreted"primarily intended for use" in an analogous statutory contextas "the item's likely use." Posters 'N' Things Ltd. v. UnitedStates, 511 U.S. 513, 521 (1994). In contrast, the definition of "claim"is "an assertion, statement, or implication (as of value, effectiveness,qualification, eligibility)." Webster's Third at 414. Where a provisionof the FDCA is meant to turn on such representations, it specifically soprovides. Gov't Br. 26.

Because manufacturers ordinarily have a financial incentive to make claimsabout a product so that customers will be be induced to buy it, the claimsthat are made in connection with a sale usually reflect a product's "intended"effects. In some cases, however, manufacturers can count on consumers tounderstand the uses and pharmacological effects of a product, and to buyit for those reasons, even in the absence of any claims by the manufacturers.In such cases, FDA is not powerless to protect the public health. It maytreat those products as drugs or devices when it finds, based on all theobjective evidence, that the pharmacological effects of the product are"intended." That is the situation here: The evidence convincinglyshows that the nicotine in tobacco products is intended to be used by consumersto sustain addiction and for sedation, stimulation, and weight control.It would be contrary to the fundamental public health purposes of the Actto conclude that a product is altogether excluded from regulation (evento prevent its adulteration or to improve its safety) precisely becauseits drug-like attributes are so widely known and thoroughly embedded inthe behavior of consumers and manufacturers as to render claims to thateffect superfluous. Gov't Br. 25.

Even respondents shrink from that consequence of their position, for theyconcede that FDA could regulate a product such as Prozac if it were soldonly by its name, without any representations about its uses or pharmacologicaleffects. Respondents would reach that result, however, on the theory thatthe product name has taken on a "secondary meaning" that constitutesan "implied claim" about the product's uses and effects, attributesthey say tobacco products do not have. RJR Br. 15; B&W Br. 24-25. Wedisagree that tobacco products do not have similar attributes. But thereis no need to resort to concepts such as "secondary meaning" or"implied claim" (which do not appear in the Act) to ensure thata product marketed simply as Prozac is covered by the FDCA. The reason theAct applies in that example is that consumers would be aware of the usesand effects of the product based on its name alone, they would buy and usethe product accordingly, and manufacturers could count on them to do so.The same is true for tobacco products.

Respondents' claims-only theory threatens to open a gaping hole in the Act'sprotection of the public health. Under respondents' theory, manufacturersof potent drugs could escape regulation by marketing their products withthe same chemical name as a brand name product, but without accompanyingdrug claims. A manufacturer could freely market in that manner such drugsas "fluoxetine" (the chemical name for the compound in Prozac)and "sildenafil citrate" (the chemical name for the compound inViagra), and FDA would be unable to assure their safety or effectiveness.1

b. Respondents' claims-only interpretation also conflicts with FDA's "intendeduse" regulations, which have been in effect since 1952. See Gov't Br.26-27 & n.5. Those regulations, which are entitled to deference underChevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S.837 (1984), and are not challenged here, provide that "intended use"refers to the "objective intent" of the person legally responsiblefor the labeling, not to that person's "claims." 21 C.F.R. 201.128(drug), 801.4 (device). The regulations do provide that a manufacturer's"labeling claims" and "advertising matter" can constituteevidence of "objective intent." Ibid. The regulations make clear,however, that such evidence is not the exclusive basis for determining thatintent. Also relevant are: (1) all of the manufacturer's "oral or writtenstatements," (2) "the circumstances surrounding" a product'sdistribution, (3) the manufacturer's "knowledge" that a productis "offered and used for a purpose for which it is neither labelednor advertised," and (4) the manufacturer's "knowledge of factsthat would give him notice" that a product "is to be used"for purposes other than those for which the manufacturer offered it. Ibid.As FDA has explained, the "intended use" regulations contemplatethat FDA will consider "all of the relevant evidence" and decide,"from the perspective of a reasonable fact-finder," whether theproduct is intended to affect the structure or function of the body. 61Fed. Reg. 45,153 (1996).

Respondents' "claim" requirement also conflicts with FDA's regulatorypractice. Products regulated without market claims include "caine,"a street drug marketed as incense; "khat," a stimulant; cosmeticscontaining hormones; toothpaste containing fluoride; interferon; a foodsupplement containing thyroid; and novelty condoms. Gov't Br. 29-30. Respondents'effort (B&W Br. 26-27) to distinguish those regulatory measures doesnot accord with FDA's authoritative explanations for its actions. See 61Fed. Reg. at 45,186-45,191; 60 Fed. Reg. 41,527-41,531 (1995).

c. There is no merit to respondents' contention (RJR Br. 15-17; B&WBr. 28-32) that FDA's longstanding interpretation will interfere with physicians'ability to prescribe approved drugs and devices for uses other than thoseon the labeling and inhibit the development of new uses for approved products.FDA does not prohibit physicians from prescribing approved products foroff-label uses. Insofar as the manufacturer is concerned, however, FDA regulationshave provided ever since 1938 that, unless FDA grants an exception, labelingshould contain adequate directions for those purposes for which a drug ordevice is "commonly used," even if the manufacturer has not chosento promote those purposes in its labeling or advertising.2 When a particularoff-label use becomes widespread, FDA may fairly "impute the requisiteintended use[] to [the] manufacturers" (J.A. 62). As FDA explainedin 1972, when that occurs, "FDA will investigate it thoroughly"and "take whatever action is warranted to protect the public,"including, if appropriate, "[r]equiring a change in the labeling towarn against or approve the unapproved use, seeking substantial evidenceto substantiate its use [as safe and effective], restricting the channelof distribution, and even withdrawing approval of the drug and removingit from the market in extreme cases." 37 Fed. Reg. 16,504 (1972) (quotedat 61 Fed. Reg. at 45,182); see, e.g., 64 Fed. Reg. 10,994 (1999) (addressingoff-label use of diet drug in "Phen Fen," which resulted in seriouscardiac events); B&W Br. 31 (labeling requirement for baby aspirin usedby adults to reduce heart attack risk). These steps do not interfere withthe practice of medicine or the proper development of new uses for approveddrugs or devices.

d. Respondents derive (RJR Br. 12; B&W Br. 11-12) their non-textual"claim" requirement from a single sentence in a 1934 Senate Reportstating that "[t]he manufacturer of the article through his representationsin connection with its sale can determine the use to which the article isto be put." S. Rep. No. 493, 73d Cong., 2d Sess. 3 (1934). That sentence,however, says only that representations "can determine" an article'sintended use, not that the presence or absence of a representation is alwaysdispositive.3

By contrast, FDA's longstanding position that "intended" effectsare not limited to manufacturer claims was confirmed by the House Reporton the 1976 device amendments. That report specifically rejected the propositionthat a claim is dispositive and explained that the Secretary "may consideractual use of a product in determining whether or not it is a device."H.R. Rep. No. 853, 94th Cong., 2d Sess. 14 (1976). Contrary to respondents'assertion (B&W Br. 17 n.14), the 1976 House Report cannot be dismissedon the ground that it "interprets language enacted 38 years earlier,"because the structure/function "device" definition was reenactedin the 1976 amendments. Gov't Br. 27.4

e. In sum, the text, legislative history, and administrative interpretationof the Act all make clear that intended effects under the drug and devicedefinitions are not limited to those claimed by manufacturers. And numerousjudicial decisions, including many cited by respondents (RJR Br. 12, 14,B&W Br. 11-12, 21-24), confirm that intent may be determined from "anyrelevant source," including consumer use. Gov't Br. 28.

2. Respondents err in contending (UST Br. 9) that the structure/functiondefinitions are limited to products with a medical purpose. The term "medicalpurpose" does not appear in the Act's definitions, and there is noterm in the definitions that could serve as the basis for such a limitation.

Moreover, the drug and device definitions include as separate categories(1) products "intended for use in the diagnosis, cure, mitigation,treatment, or prevention of disease," and (2) products "intendedto affect the structure or any function of the body." 21 U.S.C. 321(g)(1)and (h). Since the dictionary defines "medical" as "of, relatingto, or concerned with physicians or with the practice of medicine,"Webster's Third at 1402, and "medicine" as "a substance orpreparation used in treating a disease," ibid., respondents' medical-purposegloss on the latter definition would eliminate any meaningful distinctionbetween the two. Congress, however, added the latter definition becausecertain dangerous and ineffective products that were intended to affectthe structure or function of the body, such as weight-loss products, didnot fit within the treatment-of-disease category. Gov't Br. 20. If the structure/functiondefinition required proof of a medical purpose, it would resurrect the formerregime and reopen the loophole that Congress sought to close.
Respondents' medical-purpose test also conflicts with FDA's practice ofregulating products that do not have a medical purpose-tanning booths, wrinklecreams, hair-growing products, stimulants (such as NoDoz), aphrodisiacs,and athletic performance enhancers. 61 Fed. Reg. at 44,677-44,678. Respondentsseek to reconcile their medical-purpose test with that FDA practice by equatinga medical purpose with a purpose to produce a beneficial effect on the body(UST Br. 28-30). But if that is the relevant inquiry, tobacco products qualify."Tobacco industry scientists have themselves argued that tobacco productsprovide 'needed physiological benefits (increased mental alertness; anxietyreduction, coping with stress)," and that "nicotine is a veryremarkable beneficent drug." 61 Fed. Reg. at 44,680.5

3. Respondents contend (RJR Br. 21) that, because Congress has exemptedtobacco products from other health and safety statutes, those products shouldbe excluded from the FDCA as well. Respondents have it backwards. The specificexemptions in other statutes demonstrate that Congress knows how to exempttobacco products when it wants to. Since Congress did not exempt tobaccoproducts from the "drug" and "device" definitions inthe FDCA (even though it did exempt tobacco from the definition of "dietarysupplement," 21 U.S.C. 321(ff)(1)), those products are covered by theFDCA-if, as FDA has found, they are "intended to affect the structureor any function of the body."6

4. Contrary to respondents' assertion (B&W Br. 9-10; PM Br. 7), FDA'spredecessor agency did not announce in 1914 that it could not regulate tobaccoproducts unless they were marketed with medical claims. The agency statedthat "tobacco and its preparations, when labeled in such a manner asto indicate their use for the cure, mitigation, or prevention of disease,are drugs within the meaning of the act," and that "tobacco andits preparations which are not so labeled and are used for smoking or chewingor as snuff and not for medicinal purposes are not subject to the provisionsof the act." See Gov't Br. 41 n.10. That statement, concerning thetreatment-of-diseases definition now in 21 U.S.C. 321(g)(1)(B), indicatesthat, while health claims are sufficient to subject tobacco products tocoverage, they are not necessary. Indeed, the italicized portion-which wouldbe superfluous under respondents' reading of the Act-reinforces FDA's positionthat consumer use of a product can support a finding of "intended"effects even in the absence of claims.

Respondents also rely (RJR Br. 13 n.12; B&W Br. 19-20) on statementsby FDA officials during the 1960s and 1970s that tobacco products were notcovered by the FDCA. Those statements were made in the context of growingawareness, sparked by the Surgeon General's 1964 Report, that tobacco productscause cancer and other serious health conditions. The statements reflectedFDA's view that it did not have jurisdiction to regulate tobacco productsbased on those adverse health effects alone: because consumers used tobaccoproducts in spite of, not because of, their cancer-causing and other harmfulproperties, those effects on the structure or function of the body werenot "intended" within the meaning of the FDCA. Cf. Personnel Administratorv. Feeney, 442 U.S. 256, 279 (1979). The statements also reflected FDA'sbelief that, absent manufacturer claims, there was then insufficient evidenceto warrant the conclusion that tobacco products were "intended"to affect the structure or function of the body in some other way.
FDA did not have such evidence until recently. In 1980, when FDA deniedthe petition filed by Action on Smoking and Health (ASH) due to the absenceof evidence of intended effects (J.A. 50-68), no major health organizationhad yet determined that nicotine was addictive. By 1994, every leading healthorganization had concluded that it was. In 1980, evidence did not show thatmost consumers use tobacco products to sustain addiction and for stimulationand sedation. Evidence developed since 1980 shows that the overwhelmingpercentage of consumers do so. Most dramatic, recently released internalindustry documents show that tobacco manufacturers have long known thatconsumers use their products for their pharmacological effects and havedeliberately engineered them to deliver active doses of nicotine. In 1980,that evidence was not available. Gov't Br. 38-39.7

While respondents now launch an "everyone has always known" attackon FDA's decision, respondents' chief executives represented to Congressin 1994, under oath, that the nicotine in tobacco products is not addictive,Regulation of Tobacco Products: Hearing Before the Subcomm. on Health andthe Environment of the House Comm. on Energy and Commerce, 103d Cong., 2dSess. Pt. 1, at 628 (1994), and that respondents do not engineer their productsto deliver active doses of nicotine, id. at 542, 544, 558, 598. Respondentscontinued to make those same assertions in this rulemaking proceeding.8And, for more than 30 years, respondents withheld from the public criticalinformation about the intended effects of the nicotine in their products.Gov't Br. 5-7.

B. Structure Of The Act

1. Respondents erroneously contend (RJR Br. 24-25) that FDA cannot regulatetobacco products as drugs or devices because the FDCA provides that drugsand devices must be proven safe, and FDA has not found that tobacco productsare safe. At the outset, we note the following: First, FDA has chosen toregulate tobacco products as devices, a choice within its authority when,as here, a product is a combination of a drug and a device. 61 Fed. Reg.at 44,400-44,403.9 Second, the relevant standard for permitting the saleof a device is that the regulatory controls to be applied must provide a"reasonable assurance of * * * safety" for the device. 21 U.S.C.360c(a). Third, that determination is made at the end of the classificationprocess, after an expert panel studies the product and makes a recommendation,after FDA issues a proposed rule concerning the appropriate regulatory classand controls for the product, and after the public has an opportunity tocomment. 21 U.S.C. 360c(b), (c) and (d). That process is often time-consuming,compare Medtronic, Inc. v. Lohr, 518 U.S. 470, 478 n.3, 479-480 (1996),and as respondents concede (RJR Br. 29), the Act does not impose a deadline.Fourth, consistent with its usual practice, FDA decided to apply generalcontrols to tobacco products first, rather than delaying regulation untilcompletion of the lengthy classification process. 61 Fed. Reg. at 44,412.

Respondents contend, however, that, regardless of what additional controlsmay be imposed in the classification process, FDA will not be able to findthat a reasonable assurance of safety exists, and that tobacco productstherefore will have to be banned. That result is so unthinkable, respondentsargue, that, despite the overwhelming evidence that tobacco products aredevices for delivering nicotine to the body, they cannot be drug-deliverydevices within the meaning of the FDCA. FDA has reasonably determined, however,that the FDCA does not require a ban on the sale of tobacco products toadults.
In classifying a device, the Act requires FDA to weigh "any probablebenefit to health from the use of the device against any probable risk ofinjury or illness from such use." 21 U.S.C. 360c(a)(2)(C). Under thatstandard, devices that are dangerous in the ordinary sense of that wordmay be permitted to be sold if FDA finds (as it has, e.g., for certain cancer-treatmentproducts) that the health benefits to those who use them outweigh the risks.Gov't Br. 32. FDA has made such a judgment here. Although FDA decided toprohibit the sale of tobacco products to children, it found that, becauseso many adults are addicted to the products, it would be more dangerousto the health of those adults and to the public health overall to removetobacco products from the market completely than to leave them on the marketfor adults, subject to the Act's general controls and whatever additionalcontrols may be imposed as a result of the classification process. In particular,FDA found that leaving tobacco products on the market provides health benefitsbecause many addicted adults would suffer from nicotine withdrawal and therewould not be adequate pharmaceuticals available for treatment, and becausethe black-market products that would predictably replace existing tobaccoproducts would be even more dangerous for those users. 61 Fed. Reg. at 44,413.Respondents' assertion (RJR Br. 26) that FDA's judgment rests on factorsoutside FDA's mission, such as adverse effects on law enforcement, the economy,and society at-large, ignores FDA's rationale.

Even if the Act were to require tobacco products to be banned, however,that would not invalidate FDA's threshold determination that tobacco productsare "devices" for delivering the "drug" nicotine. Itwould mean that Congress might then have to consider whether to amend theAct to permit the continued sale of those drug-delivery devices, just asit permitted the continued sale of products containing saccharin after FDAconcluded that the Act required them to be banned. Gov't Br. 36-37. Respondentsassert (RJR Br. 32) that a ban of tobacco products "was not reasonablyin the contemplation of the enacting Congress." But if, as we haveshown, compelling new evidence establishes that the nicotine in tobaccoproducts is intended to sustain addiction and for sedation, stimulation,and weight control-and if, as respondents assert, tobacco products cannotbe marketed with a reasonable assurance of safety-the Act would requirea ban. The fact that the legislators who voted for the 1938 Act did notanticipate that such evidence would come to light and that tobacco productswould be covered by the Act as a result-or that some might have regardeda ban as undesirable even in those circumstances-is simply not relevantto the statutory inquiry. Congress deliberately crafted broad definitionsof "drug" and "device" in 1938, and "it is ultimatelythe provisions of our laws rather than the principal concerns of our legislatorsby which we are governed." Oncale v. Sundowner Offshore Servs., Inc.,523 U.S. 75, 79 (1998).

2. In addition to disagreeing with FDA's judgment concerning whether tobaccoproducts must be banned, respondents argue that FDA has misinterpreted severalother provisions of the FDCA. Those criticisms are misguided.

a. Noting that a device is "misbranded" if "it is dangerousto health when used in the * * * manner * * * suggested in the labeling,"21 U.S.C. 352(j), respondents object (RJR Br. 27) that FDA did not explainwhy that provision is inapplicable to tobacco products. The reason is that,just as the benefits of some cancer treatments outweigh their health risks,the benefits of allowing tobacco products to remain on the market, subjectto regulatory controls, outweigh the health risks of removing them fromthe market. They are therefore not "dangerous to health" withinthe meaning of that provision. Even if that provision were applicable, however,FDA would have discretion to decide that, given the danger to the healthof addicted adults of removing tobacco products from the market, it shouldnot be enforced against those products. Chaney, 470 U.S. at 835.10

b. Respondents contend (RJR Br. 28) that, because a device is misbrandedif it fails to bear "adequate directions for use," and FDA hasnot required such directions for tobacco products, "it must be FDA'sview that adequate directions for use of tobacco products cannot be written."FDA may grant an exemption from the adequate directions requirement, however,when it determines they are "not necessary for the protection of thepublic health." 21 U.S.C. 352(f). One such circumstance is when "adequatedirections for common uses [of the device] are known to the ordinary individual."21 C.F.R. 801.116. Because it is "common knowledge" how tobaccoproducts are used, FDA reasonably decided an exemption was warranted. 61Fed. Reg. at 44,465.

c. Finally, respondents argue (RJR Br. 28-29) that FDA failed to apply amisbranding provision that requires "adequate warnings against use* * * by children." 21 U.S.C. 352(f)(2). FDA concluded, however, thatthe familiar "Surgeon General's warnings" required by other federalstatutes satisfy Section 352(f)(2). 61 Fed. Reg. at 44,465. That rationaleis not "disingenuous," as respondents suggest. It reflects FDA'sreasonable judgment that no warnings are likely to be effective for children,id. at 44,468, 44,511, and that the Surgeon General's warnings serve thepurposes of Section 352(f)(2) as well as any that FDA could devise.

C. Tobacco-Specific Statues

Respondents concede (RJR Br. 36) that "the tobacco-specific statutesdo not repeal any part of the FDCA or 'preempt' any action by FDA."They nevertheless submit (ibid.) that those very same statutes should precludeFDA from regulating tobacco products. Respondents are wrong.

1. a. Respondents contend (PM Br. 37-41) that FCLAA precludes FDA regulationbecause FCLAA forecloses a ban on tobacco products, while FDA's determinationthat tobacco products are drug-delivery devices would necessarily lead toa ban. For three reasons, that contention is without merit. First, as wehave shown, FDA's determination that tobacco products are covered by theAct as drug-delivery devices does not mean that they must be banned.

Second, FCLAA does not foreclose a ban on tobacco products. By its terms,FCLAA only prevents FDA from requiring any "statement relating to smokingand health, other than the statement required by" FCLAA itself. 15U.S.C. 1334(a). Respondents' contention (PM Br. 38) that the "policy"statement in 15 U.S.C. 1331 precludes a ban finds no support in that provision'stext, and it ignores the holding in Cipollone v. Liggett Group, Inc., 505U.S. 504 (1992), that FCLAA "merely prohibit[s] state and federal rulemakingbodies from mandating particular cautionary statements on cigarette labels,"id. at 518, and that Section 1331 states not a broad policy of protectingthe continued marketing of cigarettes, but a far more limited policy of"protecting the national economy from the burden imposed by diverse,nonuniform, and confusing cigarette labeling and advertising regulations,"id. at 514. See also id. at 534 (Blackmun J., concurring and dissenting);Medtronic, 518 U.S. at 488 (FCLAA preempts only a "limited set"of requirements).
Third, even if FDA's coverage determination were to result in a ban on thesale of tobacco products under the FDCA and Section 1331 were to precludeone, that still would not undermine FDA's coverage determination. It wouldsimply mean that FCLAA (as the more specific statute) would preclude a ban,and FDA would therefore be required to adopt measures short of a completeban to regulate tobacco products. Nothing in FCLAA, for example, would precludeFDA from continuing with its current regulatory program of preventing salesof tobacco products to minors or from requiring safer ingredients or a saferfilter.

b. Respondents further argue (PM Br. 41-42) that FDA's coverage determinationis precluded by FCLAA because the FDCA requires FDA to impose labeling requirements,such as adequate directions for use and warnings for children, while FCLAAprevents FDA from imposing those requirements. As we have explained, however,FDA exempted tobacco products from the FDCA's adequate directions requirement,and it reasonably determined that the Surgeon General's warnings are sufficientwarnings for children. See p. 16, supra.11 In any event, to the extent thatFCLAA precludes FDA from imposing particular restrictions on tobacco productsthat the FDCA otherwise would require, the more specific statute would governand FDA would be limited to regulating tobacco products in other ways.12

2. Respondents contend (NACS Br. 14, 18-19) that FDA's coverage determinationis precluded by ADAMHA because ADAMHA generally permits States to decidewhat measures to adopt to curb youth tobacco use, while FDA's tobacco regulationspreempt state laws that are "different from, or in addition to,"FDA's requirements. 21 U.S.C. 360k(a)(1). ADAMHA, however, simply conditionscertain federal funding on the States' enactment of their own laws againsttobacco use. Gov't Br. 47. It does not address whether FDA may concludethat tobacco products are drug-delivery devices and subject to federal regulationas well. Moreover, contrary to respondents' assertion (RJR Br. 46-47), FDA'sregulations do not divest States of authority to regulate tobacco products.States are free to impose whatever requirements they choose when there isno parallel FDA requirement; and where there are federal requirements, theStates may impose substantially similar ones. Medtronic, 518 U.S. at 496-497;21 C.F.R. 808.1(d)(2). In addition, the FDCA authorizes FDA to exempt frompreemption state laws that impose more stringent requirements than FDA's,21 U.S.C. 360k(b), and FDA has done so on many occasions. And while respondentsobject to that regime as insufficiently sensitive to state interests (NACSBr. 18-19), an amici brief joined by 40 States concludes (Br. 19) that "FDA'sauthority to regulate tobacco products is authorized by law, and is a criticallyimportant part of the effort to limit the use of tobacco products by minors."

D. Chevron Deference

Respondents err in contending (RJR Br. 47-50) that this case should be resolvedentirely outside the Chevron framework. As respondents note (RJR Br. 48),this case involves the construction of both the FDCA, which FDA enforces,and tobacco-specific statutes, which it does not. But that does not meanthat the Chevron framework should be discarded. Instead, the Court shouldfirst decide under Chevron whether FDA's interpretation of the Act it administersis permissible. If the Court concludes that it is, the Court should thendecide independently whether FDA's authority under the FDCA has been divestedby the tobacco-specific statutes. NASA v. FLRA, 119 S. Ct. 1979, 1984-1985(1999). As we have shown, FDA's interpretation is based on a permissiblereading of the FDCA, and the tobacco-specific statutes do not withdraw FDA'sauthority.

Respondents similarly err in contending (RJR Br. 49) that the Chevron frameworkdoes not apply because FDA has changed its position on whether tobacco productsare covered by the Act. Under Chevron, a change in agency position is entitledto full deference, as long as the agency offers a reasoned analysis forthe change. See 467 U.S. at 863-864; Rust v. Sullivan, 500 U.S. 173, 186-187(1991); Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463U.S. 29, 42 (1983). FDA supplied such an analysis here. FDA adhered to itslongstanding legal position that a finding of "intended" effectsmay be based on evidence other than manufacturer claims, and it found compellingnew evidence that tobacco products are intended to be used to sustain addictionand for stimulation and sedation.

Finally, respondents argue (RJR Br. 47-48) that this case should be resolvedagainst FDA at the first step of Chevron. The relevant indicia of congressionalintent, however, do not come close to establishing that Congress "directlyaddressed the precise question at issue" and "unambiguously expressed[its] intent" that tobacco products fall outside the reach of the FDCA.467 U.S. at 843. To the contrary, the text, legislative history, and administrativeinterpretation of the Act strongly support FDA's conclusion that, giventhe overwhelming evidence that the nicotine in tobacco products is intendedto be used to sustain addiction and as a sedative, stimulant, and appetitesuppressant, tobacco products are drug-delivery devices within the meaningof the FDCA. At the very least, FDA's conclusion is based on "a permissibleconstruction" of the Act. Ibid.

* * * * *

For the foregoing reasons and those in our opening brief, the judgment ofthe court of appeals should be reversed.

SETH P. WAXMAN
Solicitor General


OCTOBER 1999

1 Respondents concede (RJR Br. 15) that undertheir theory, nicotine inhalers would escape FDA review as long as the manufacturerpromoted them for "breathing pleasure." In an attempt to avoidthat anomaly, respondents suggest that the Consumer Product Safety Commissioncould regulate that product. Yet respondents offer no reason why Congresswould have wanted a product that manufacturers intend and consumers useas a drug-delivery device, and that poses the health risks of such a device,to be regulated by an agency with no expertise in that area.

2 Those regulations specify what constitutes "adequate directions foruse" in the labeling of a drug or device under 21 U.S.C. 352(f)(1).They provide that, absent an exception, such directions must be adequatefor the purposes for which the drug or device "is intended" (cross-referencingthe "intended use" regulation), and then state that directionsmay be inadequate if they omit statements of "all conditions, purposesor uses for which the drug [or device] is intended"-"including"not only those for which it is "prescribed, recommended, or suggestedin its oral, written, printed, or graphic advertising," but also thosefor which it "is commonly used." 21 C.F.R. 201.5 (drugs), 801.5(devices). That requirement, adopted in its current form in 1952 (17 Fed.Reg. 6818 (1952) (21 C.F.R 1.106(1)), derives from a regulation adoptedin 1938. See 3 Fed. Reg. 3167 (1938).

3 Respondents also take the sentence entirely out of context. It is partof a paragraph discussing when a product that is concededly subject to theAct will be regulated as either a "drug" or a "food"(or both). The sentences preceding the one respondents quote explain that"if [the product] is to be used only as a food it will come withinthe definition of food and none other," while "[i]f it containsnutritive ingredients but is sold for drug use only, as clearly shown bythe labeling and advertising, it will come within the definition of drug,but not that of food"; the sentence immediately following the one respondentsquote then states that a manufacturer of a laxative that is a medicatedcandy or chewing gum could bring its product within the definition of drug"and escape that of food" by "representing the article fairlyand unequivocally as a drug product." S. Rep. No. 493, at 3 (emphasisadded). The paragraph as a whole thus suggests no more than that a manufacturer,by "clearly," "unequivocally," and "fairly"representing a product as "only" one thing (e.g., a drug), cannegate the conclusion that it is another (e.g., a food). Here, respondentshave carefully avoided making claims that reflect the intended uses of theirproducts. Nor have they made claims that their products are intended onlyfor some other use that negates their status as drugs or devices, and suchclaims could not "fairly" be made, in light of the overwhelmingevidence of intended pharmacological effects. Compare Gov't Br. 27-28.

4 Respondents' effort (B&W Br. 12-15) to find support for their positionin the Drug Amendments of 1962 is unavailing. The 1962 amendments preventa drug manufacturer from making claims about a new product, or new claimsabout an existing product, without first establishing that the product iseffective or generally recognized as such. See Weinberger v. Hynson, Westcott& Dunning, 412 U.S. 609, 622 (1972). Specifically, they require manufacturersto establish that a new drug will have the effect it purports or is representedto have under the conditions "prescribed, recommended, or suggested"in its labeling. 21 U.S.C. 355(d)(5), (e)(3). In justifying that requirement,several committee reports, Members of Congress, and administration witnessesstated that drugs should be shown to be effective for their "intended"uses or purposes before they are marketed. See B&W Br. 13-15.
The quoted passages did not advert to the wholly different situation, presentedhere, of a product that has been marketed for many years and whose "intended"effects are shown by pervasive practices of consumers and manufacturers,without any need for claims. None of those passages purported to interpretthe structure/function definition or to equate "intended" effectsin that definition with uses "prescribed, recommended, or suggested"on the labeling in 21 U.S.C. 355(d). Indeed, the 1962 amendments establishthat Congress did not equate those two concepts. Section 107(c)(4) of theamendments afforded grandfather protection for a drug "when intendedsolely for use under conditions prescribed, recommended, or suggested inlabeling with respect to that drug" prior to the amendments' effectivedate. 76 Stat. 789; see USV Pharmaceutical Corp. v. Weinberger, 412 U.S.655 (1972). That clause plainly contemplates that there can be "intendeduses" other than those identified in the labeling. See H.R. Rep. No.2464, 87th Cong., 2d Sess. 12 (1962). After the 1962 amendments, as before,if an approved drug develops such a use-i.e., if the drug becomes commonlyused for an off-label purpose-FDA can respond in various ways, includingby requiring that the labeling contain adequate directions for that additionaluse. See pp. 4-5, supra. If FDA imposes such a requirement, then conditionsfor that additional use will be "prescribed, recommended, or suggested"in the labeling, and that use will therefore have to satisfy the new drugstandards of safety and effectiveness. See 21 U.S.C. 321(p)(1), 355(d)(1)and (5), 355(e)(1) and (3).

5 Respondents contend (UST Br. 15; RJR Br. 18) that without a "medicalclaims" limitation, products such as guns, thermal clothing, air conditioners,exercise equipment, scuba-diving gear, mattresses, and even roller coastersand horror movies could be considered "devices" under the Act.FDA has never interpreted the FDCA to reach any of those products, cf. Churchof the Holy Trinity v. United States, 143 U.S. 457, 459 (1892), and it plainlywould have discretion not to take enforcement action even if they were thoughtby some to be covered. Heckler v. Chaney, 470 U.S. 821 (1985). If FDA nonethelessattempted to regulate such products, the question would arise whether itwould be reasonable to press the words of the structure/function definitionto the point of treating as "devices" products that do not deliverdrugs to the body, that do not have intended effects similar to any otherproduct regulated by FDA, and that implicate more directly the consumer-safetypurposes of the Consumer Product Safety Act (CPSA), 15 U.S.C. 2051 et seq.,than the health concerns of the FDCA (see RJR Br. 19; B&W Br. 20-21n.17). This case does not remotely raise that question. Tobacco manufacturershave themselves characterized nicotine as a powerful drug and cigarettesas devices for delivering nicotine to the body; the intended effects oftobacco products are the same as many other products regulated by FDA (Gov'tBr. 23-24), and tobacco products directly implicate the health concernsof the FDCA.

6 Respondents assert (RJR Br. 19-20) that the exemption in the CPSA fortobacco products is superfluous if they are drugs or devices under the FDCA,because there is a separate exemption in the CPSA for FDCA drugs and devices.When Congress enacted the CPSA in 1972, however, FDA had not yet found sufficientevidence that tobacco products are intended to affect the structure or functionof the body. An express exemption was therefore necessary to exclude tobaccoprodcuts from the CPSA.
Respondents also argue (RJR Br. 20) that, because the Controlled SubstancesAct (CSA), 21 U.S.C. 801 et seq., includes FDCA "drugs" and excludes"tobacco," 21 U.S.C. 802(6), (12), if the nicotine in tobaccoproducts is a "drug," then it would simultaneously be includedand excluded from the definition of "controlled substance." Thereis no such contradiction. Under familiar principles of statutory construction,the CSA's specific exclusion of "tobacco" prevails over its generalinclusion of "drugs." Nor does the CSA's definition of "controlledsubstance" to mean both "drugs" and "any other substance"included in one of the CSA schedules (see 21 U.S.C. 802(6)) contradict ourcontention that products taken into the body for pharmacological effectshave the classic characteristics of products subject to FDA regulation.The reference to "any other substance" simply permits certainsubstances to be controlled even without a showing that they are intendedto be used as drugs.

7 Respondents err in asserting (PM Br. 25) that FDA's 1980 decision wasbased on a supposed recognition "that its lack of jurisdiction wasinherent in the FDCA and not due to lack of evidence." FDA twice statedthat consumer-use of a product as a device could be a basis for finding"intended" effects or use, but that ASH had failed to producesufficient consumer use evidence. J.A. 56-57, 61-62; see also id. at 54,58. The decision also stated that FDA's statement to Congress in 1965 concerningits lack of jurisdiction was likewise based on an absence of evidence ofthe requisite intended use. J.A. 57. Respondents similarly quote (PM Br.23) an ambiguous statement from FDA's 1977 decision (J.A. 44-49) withoutnoting that the D.C. Circuit authoritatively interpreted the 1977 decisionas likewise resting on lack of evidence. ASH v. Harris, 655 F.2d 236, 239(1980); accord Gov't Br. 38 (quoting FDA brief in D.C. Circuit).

8 E.g., 61 Fed. Reg. at 44,617, 44,670-44,671, 44,706-44,707, 44,776-44,777,44,783, 44,789, 44,800, 44,958-44,959, 44,965 44,983, 44,986-44,987, 45,065,45,067, 45,115, 45,141.

9 Because FDA has chosen to regulate tobacco products as devices ratherthan drugs, there is no merit to respondents' contention (RJR Br. 27-28)that, under FDA's theory, tobacco products are "new drugs" thatFDA may not permit to be marketed unless approved as safe and effective.

10 For the same reason, there is no merit to respondents' argument (RJRBr. 30) that FDA has failed to comply with a section of the Act providingthat, "[i]f [FDA] finds that there is a reasonable probability thata device * * * would cause serious, adverse health consequences or death,[FDA] shall issue an order requiring the [manufacturer] * * * to immediatelycease distribution of such device." 21 U.S.C. 360h(e)(1)(A). FDA'sfinding that banning tobacco products would create greater dangers thanleaving them on the market subject to regulatory controls makes that provisioninapplicable. But even if it were applicable, FDA would have discretionnot to issue a cease-distribution order. Chaney, 470 U.S. at 835. Althoughthe provision uses the word "shall," FDA has interpreted it aspermissive rather than mandatory. See 21 C.F.R. 810.10(a) (FDA "mayissue a cease distribution and notification order"). FDA's interpretationis reasonable, because "shall" sometimes means "'should,''will,' or even 'may,'" Gutierrez de Martinez v. Lamagno, 515 U.S.417, 432 n.9 (1995); because enforcement authority is usually discretionary;and because a cease-distribution order is only an interim step in a processthat leads to a "recall order," which is itself discretionary,21 U.S.C. 360h(e)(2)(A).

11 Respondents erroneously contend (PM Br. 42) that FCLAA precludes FDAfrom requiring tobacco-product labeling to bear the statement "Nicotine-DeliveryDevice for Persons 18 or Older." Because that statement simply informsconsumers about the products' intended and lawful use, and does not containany warning about the health dangers of tobacco use, it is not a statement"relat[ing] to smoking and health" within the meaning of 15 U.S.C.1334(a). Even if it were, however, that would lead only to invalidationof that requirement. It would not affect the conclusion that tobacco productsare drug-delivery devices under the FDCA.

12 There likewise is no inconsistency between FDA's actions and the ComprehensiveSmokeless Tobacco Health Education Act of 1986, 15 U.S.C. 4401 et seq.,since as respondents concede (UST Br. 31), it was modeled on FCLAA and containsthe same basic requirements. Gov't Br. 46.

APPENDIX


Section 801.4 of 21 C.F.R. states as follows:

§ 801.4 Meaning of "intended uses."
The words intended uses or words of similar import in §§801.5,801.119, and 801.122 refer to the objective intent of the persons legallyresponsible for the labeling of devices. The intent is determined by suchpersons' expressions or may be shown by the circumstances surrounding thedistribution of the article. This objective intent may, for example, beshown by labeling claims, advertising matter, or oral or written statementsby such persons or their representatives. It may be shown by the circumstancesthat the article is, with the knowledge of such persons or their representatives,offered and used for a purpose for which it is neither labeled nor advertised.The intended uses of an article may change after it has been introducedinto interstate commerce by its manufacturer. If, for example, a packer,distributor, or seller intends an article for different uses than thoseintended by the person from whom he received the devices, such packer, distributor,or seller is required to supply adequate labeling in accordance with thenew intended uses. But if a manufacturer knows, or has knowledge of factsthat would give him notice that a device introduced into interstate commerceby him is to be used for conditions, purposes, or uses other than the onesfor which he offers it, he is required to provide adequate labeling forsuch a device which accords with such other uses to which the article isto be put.

Section 801.5 of 21 C.F.R. states as follows:

§ 801.5 Medical devices; adequate directions for use.
Adequate directions for use means directions under which the layman canuse a device safely and for the purposes for which it is intended. Section801.4 defines intended use. Directions for use may be inadequate because,among other reasons, of omission, in whole or in part, or incorrect specificationof:

(a) Statements of all conditions, purposes, or uses for which such deviceis intended, including conditions, purposes, or uses for which it is prescribed,recommended, or suggested in its oral, written, printed, or graphic advertising,and conditions, purposes, or uses for which the device is commonly used;except that such statements shall not refer to conditions, uses, or purposesfor which the device can be safely used only under the supervision of apractitioner licensed by law and for which it is advertised solely to suchpractitioner.

(b) Quantity of dose, including usual quantities for each of the uses forwhich it is intended and usual quantities for persons of different agesand different physical conditions.

(c) Frequency of administration or application.

(d) Duration of administration or application.

(e) Time of administration or application, in relation to time of meals,time of onset of symptoms, or other time factors.

(f) Route or method of administration or application.

(g) Preparation for use, i.e., adjustment of temperature, or other manipulationor process.