Opinion
Civil Action No. 03-1405 (RMC).
January 19, 2005
MEMORANDUM OPINION
The question raised by the instant complaint is whether the Food and Drug Administration ("FDA") has properly classified Periostat®, a drug product developed by CollaGenex Pharmaceuticals, Inc., as an "antibiotic," thereby exposing it to competition from generic drug manufacturers. CollaGenex sues Tommy Thompson, Secretary of Health and Human Services, Lester Crawford, Acting Commissioner, and the FDA, to obtain a change in classification that would protect Periostat from generic versions of its product. Finding that the FDA classification decision is supported by the Food and Drug Modernization Act of 1997 ("FDAMA"), Pub.L. No. 105-115, 111 Stat. 2296, and by the administrative record, the Court will dismiss the CollaGenex complaint and dissolve the preliminary injunction issued on July 22, 2003, effective immediately. See CollaGenex Pharm., Inc. v. Thompson, No. 03-1405, 2003 U.S. Dist. LEXIS 12523 (D.D.C. July 22, 2003).
I. BACKGROUND FACTS
The basic facts are not in dispute. They are taken from the Declaration of Murray M. Lumpkin, M.D. ("Lumpkin Dec."), A.R. Part A, and the Memorandum prepared by David Roeder ("Roeder Mem."), A.R. Part B. Part A reconstructs the FDA decision in 1998 that Periostat is an antibiotic drug ("1998 Decision") while Part B reflects a review conducted in 2003 ("2003 Decision"). CollaGenex appealed the 1998 Decision; FDA asserts that both the 1998 and 2003 Decisions are final and subject to review.
CollaGenex submitted a New Drug Application ("NDA") to the FDA on August 30, 1996, for the drug product Periostat (doxycycline hyclate) 20 mg, as a non-antibiotic drug under Section 505 of the Federal Food, Drug, and Cosmetic Act ("FFDCA"). Periostat was the first — and remains the only — systemic drug that contributes to fighting periodontitis, a severe gum disease that can lead to tooth loss. Periostat helps keep the tooth attached to the gum, over and above the normal treatment procedures.
21 U.S.C. § 355 ("Section 505").
Shortly after receiving CollaGenex's NDA, FDA informed CollaGenex that the NDA should be amended to reflect an application as an "antibiotic" drug under Section 507 of the FFDCA. CollaGenex resubmitted its application on September 18, 1996, reserving the right to argue about its classification. FDA gave the application a "50 series" antibiotic NDA number, 50-744.
21 U.S.C. § 357. ("Section 507").
After further discussions with FDA reviewers were unsuccessful in achieving a reclassification, CollaGenex submitted a request to the FDA Ombudsman in the Office of the Commissioner on September 11, 1997, asking that Periostat be designated as a non-antibiotic drug under Section 505 and not an antibiotic drug under Section 507. The importance of the distinction is that Section 505 drugs were protected from competition from generic manufacturers while most Section 507 antibiotic drugs were not. The Ombudsman referred the request to the Center for Drug Evaluation and Research ("CDER"), as raising issues within its policy role.
As explained in greater detail below, drugs approved under Section 505 were entitled to new drug exclusivity, patent listing, patent certification, and, if an application for a generic equivalent were submitted, a 30-month stay on FDA approval of the generic so that the patent holder could defend its patent.
Congress passed FDAMA on November 20, 1997, while the CollaGenex application for Periostat was still pending before FDA. FDAMA repealed Section 507 and provided that all applications for new drugs would be submitted under Section 505; it also included a "Transition" provision that declared that an application approved under Section 507 before FDAMA would be considered to be an application submitted, filed, and approved under Section 505, with one critical exception. Section 125(d)(2) of Title I of FDAMA exempted antibiotic drugs from the protections of Section 505 if the antibiotic drug had been the subject of any application received under Section 507 before enactment of FDAMA. On July 8, 1998, CollaGenex sent a letter to Dr. Murray Lumpkin, then Deputy Director for Review Management in CDER, renewing its arguments that Periostat is not an "antibiotic" drug and briefly discussing the effect of FDAMA on the Periostat application.
Pub.L. No. 105-115, 111 Stat. 2327 (1997) § 125(d) ( reprinted in 21 U.S.C.A. § 355 Historical and Statutory Notes, "Transition").
Pub.L. No. 105-115, 111 Stat. 2327 (1997) § 125(d)(2) ( reprinted in 21 U.S.C.A. § 355 Historical and Statutory Notes, "Transition").
FDA approved Periostat (doxycycline hyclate capsules) 20 mg. on September 30, 1998. The entirety of that letter's discussion of Periostat's status as an antibiotic drug was contained in two sentences:
Please refer to your new drug application (NDA) dated August 30, 1996, received August 30, 1996, submitted under section 505(b) of the Federal Food Drug, and Cosmetic Act for Periostat™ (doxcycline hyclate USP) Capsules, 20 mg. We note that this application is subject to the exemption provisions contained in section 125(d)(2) of Title I of the FDA Modernization Act of 1997 [FDAMA].
A.R. B, Ex. 6. By its reference to Section 125(d)(2), the FDA proclaimed that Periostat is an antibiotic drug and not subject to the protections of exclusivity, patent protection, etc., that are extended to non-antibiotic drugs.
CollaGenex did not immediately appeal this FDA determination and, instead, jumped into the marketplace with Periostat. By 2002, after CollaGenex had expended the monies necessary to develop Periostat and then to develop a market among dentists for the use of Periostat, it learned that one or more manufacturers of generic drugs was preparing an application for approval of a generic version of Periostat. CollaGenex wrote to FDA's Chief Counsel advancing its arguments that Periostat is not an "antibiotic" drug and thereafter met with staff from the Office of Chief Counsel and CDER. In response, CDER began a review of the 1998 decision that Periostat is an "antibiotic" drug.
Concerned that this review would merely favor re-affirming the 1998 decision and not be unbiased, CollaGenex withdrew its request for reconsideration and, on June 26, 2003, sued FDA, seeking to bar it from approving a generic application referencing Periostat, on the grounds that Periostat is not an "antibiotic" drug and should receive all the protections from generic competition allowed under Section 505 for non-antibiotic drugs. On July 22, 2003, this Court issued a preliminary injunction restraining and enjoining FDA from approving any generic application referencing Periostat until the Court could rule on the challenge to FDA's classification of Periostat.
When CollaGenex filed this suit, it was relying on three U.S. patents: Patent Nos. 5,223,248 (the '248 Patent), 4,666,897 (the '897 Patent), and Re. 34,656 (the '656 Patent). CollaGenex has dropped its claims under the '248 Patent and the '897 Patent expired on May 7, 2004. Therefore, the only patent under which CollaGenex might be able to claim marketing exclusivity for Periostat is the '656 Patent.
FDA submitted the administrative record on September 23, 2003. That record comes in two parts. Part A is the record of the 1998 determination that Periostat is an "antibiotic" drug ("1998 Decision"). However, FDA has not located a single contemporaneous document explaining the basis for the 1998 Decision. Instead, it submits the Lumpkin Declaration to explain its reasoning. Dr. Lumpkin is now the FDA's Principal Associate Commissioner but was, at the time, the Deputy Director for Review Management in CDR. Part B of the administrative record reflects FDA's reconsideration of Periostat's classification during 2003 ("2003 Decision"). While this review started at the request of CollaGenex, it turned into a document to support the agency's position in this lawsuit.
See CollaGenex Reply at 17 ("The 2003 memorandum is plainly a post hoc rationalization of the agency's original decision intended for litigation purposes.").
CollaGenex thereafter challenged the adequacy of the administrative record submitted by the FDA. It also filed a request for documents from FDA under the Freedom of Information Act ("FOIA"), 5 U.S.C. § 552, and, later, a FOIA lawsuit, CollaGenex Pharm., Inc. v. FDA, No. 03-1554 (D.D.C.). CollaGenex seeks to supplement the administrative record with some of this additional documentation, which the FDA opposes. The Court will grant the CollaGenex motion to supplement the record, Dkt. No. 93, for two reasons.: First, the 1998 Decision is singularly lacking in contemporaneous documentation reflecting FDA's decision process. Second, the Roeder Memorandum makes clear that FDA had not developed a studied policy for defining an "antibiotic" drug prior to this litigation and did so only in response thereto. In certain respects, this may call into question the bona fides of the 2003 Decision, a point determined contrary to CollaGenex by this Court, but one that has enough basis in the record that it should be available for review should CollaGenex appeal this decision.
On August 19, 2004 and September 10, 2004 respectively, IVAX Pharmaceuticals, Inc. and CorePharma, LLC filed motions to intervene as co-defendants with the FDA based upon their pending applications for approval of generic versions of Periostat, which, they state, would have been approved by FDA but for the Court's preliminary injunction. The motions were granted and these parties have participated fully in the briefing of this matter.
II. STATUTORY SCHEME
Prior to 1997 and the passage of FDAMA, "antibiotic" drugs were approved under Section 507 of the FFDCA and non-antibiotic drugs were approved under Section 505. This difference had a long history, dating back to the development of penicillin, the first drug to have the capacity to kill microbes, i.e., be "anti-biotic." Because penicillin was manufactured in batches through fermentation, its strength and efficacy could vary depending on the rigor of that process. Congress required that FDA test all batches of penicillin to ensure that appropriate doses were administered to the military during World War II. Initially, Section 507 applied only to penicillin or any derivative of penicillin; other named antibiotic drugs were added to the statute as they were developed. When the FFDCA was amended in 1962, a more generalized definition was added so that the law would not need amending with each new discovery of an antibiotic drug.
See H.R. REP. NO. 79-702 at 2-3 (1945) ("A primary reason for the type of control proposed by this bill is the fact that penicillin is produced by a biological process and is subject to the vagaries inherent in all such processes."). FDA stopped requiring batch certifications for many antibiotic drugs in 1982. 47 Fed. Reg. 39155 (Sept. 7, 1982); see also 21 C.F.R. § 433.1 (1983).
Streptomycin was added in 1957; aureomycin, chloramphenicol, and bacitracin were added in 1949; chlortetracycline was substituted for aureomycin (a trade name for chlortetracycline) in 1953. A.R. Part B, Roeder Mem. at 5 n. 6.
See Fed. Defs.' Motion for SJ at 8 ("Congress added a general definition of antibiotic drug to § 357 to eliminate the listing of each future new antibiotic as it was developed.").
Two key consequences arose from these different treatments. Applicants for generic versions of antibiotic drugs were only requested to show conformance with statutorily-mandated, published standards of identity, strength, quality, and purity for the antibiotic substance, as reflected in antibiotic "monographs" published by FDA. Pharmaceutical companies did not have to submit the safety and efficacy data that was required for pioneer and generic non-antibiotic drugs. Therefore, generic antibiotics were developed and marketed fairly readily. See Glaxo, Inc. v. Heckler, 623 F. Supp. 69, 72 (E.D.N.C. 1985) ("[S]ince the FDA had available on file all safety and efficacy data for a pioneer antibiotic drug, a manufacturer seeking to produce a generic copy of the drug was not required to replicate the scientific data accompanying the pioneer drug's application."); Abbreviated New Drug Applications, Proposed Rule, 54 Fed. Reg. 28872, 28878 (July 10, 1989). However, antibiotic drugs did not receive the patent listing, patent certification, and marketing exclusivity benefits available to pioneer and non-antibiotic drugs after enactment of the Drug Price Competition and Patent Term Restoration Act ("Hatch-Waxman"), Pub.L. No. 98-417, 98 Stat. 1585 (1984).
A. Hatch-Waxman Amendments
The significance of the Hatch-Waxman Amendments to FFDCA ("Hatch-Waxman") cannot be understated. Prior to 1984, all applicants seeking to market pioneer drugs or generic nonantibiotic drugs had to file an NDA containing, inter alia, extensive scientific data demonstrating the safety and effectiveness of the drug. See 21 U.S.C. § 355(a)-(b); 21 C.F.R. § 314.50. As a result, few generic non-antibiotic drugs were approved by FDA. Glaxo, 623 F. Supp. at 72. Hatch-Waxman created an abbreviated approval process for generic non-antibiotic drugs, while retaining incentives for pioneer drugs, such as marketing exclusivity and patent protections. See 21 U.S.C. § 355(jj). The abbreviated new drug application ("ANDA") process shortens the time and effort needed for approval of a generic drug by allowing the applicant to merely demonstrate its product's bioequivalence to the NDA drug, without reproducing the entirety of the NDA's extensive scientific research. See Eli Lilly and Co. v. Medtronic, Inc., 496 U.S. 661, 676 (1990) (describing the ANDA process).
Because Congress still wanted to provide incentives for new drug development, alongside the ANDA process that eased the marketing of generic drugs, Hatch-Waxman entitles an NDA applicant to a period of market exclusivity (3 or 5 years, depending on the degree of innovation reflected in the NDA) which bars FDA approval of a generic ANDA for the NDA product. See 21 U.S.C. § 355 (c)(3)(D)(ii)-(iv), (j)(5)(D)(ii)-(iv). In addition, an NDA applicant must inform FDA about any patent that the NDA applicant claims will protect its exclusivity to market its drug. 21 U.S.C. § 355(b)(1), (c)(2). FDA then publishes patent information for approved drugs in the "Approved Drug Products With Therapeutic Equivalence Evaluations" (the "Orange Book"). See 21 U.S.C. § 355 (b)(1), (c)(2), (j)(7); 21 C.F.R. § 314.53(e). Generic drug manufacturers check the Orange Book to determine if a drug product is patent-protected or if it is available for the development of a generic bioequivalent drug.
An ANDA applicant must certify to FDA that (I) patent information has not been filed; (II) the patent has expired; (III) the patent will expire shortly on a date certain; or (IV) the patent is invalid or will not be infringed by the manufacture, use, or sale of the new drug covered by the application. 21 C.F.R. § 355(j)(2)(A)(vii). Under a paragraph IV certification, the applicant must also notify the NDA holder and patent owner concerning its application and its reasoning for applicability of paragraph IV. 21 U.S.C. § 355(b)(2)(B), (j)(2)(B). The filing of a paragraph IV certification "for a drug claimed in a patent or the use of which is claimed in a patent" is an act of infringement. 35 U.S.C. § 271(e)(2)(A). The holder of the patent for the drug may therefore initiate a patent infringement suit upon the ANDA applicant; if it does so, FDA will stay approval of the ANDA application for 30 months, unless a final court opinion is reached earlier, or for the term ordered by the patent court. 21 U.S.C. § 355(c)(3)(C), (j)(5)(B)(iii). It is these protections from early competition for which CollaGenex sues.
B. FDAMA
When Congress adopted FDAMA in November 1997, it repealed Section 507 of the FFDCA and required that all applications for antibiotic drugs be submitted under Section 505. FDAMA § 125(d)(1) (Transition). In subsection (d)(1), the Transition provided that applications for antibiotic drugs approved under Section 507 before FDAMA would be considered approved under Section 505. Id. However, subsection (d)(2) added the provision that when "the drug that is the subject of the application contains an antibiotic drug and the antibiotic drug was the subject of any application" received by FDA before the enactment of FDAMA, it is exempt from Hatch-Waxman benefits. FDAMA § 125(d)(2); Proposed Rule: Marketing Exclusivity and Patent Provisions for Certain Antibiotic Drugs, 65 Fed. Reg. 3623, 3624-25 (Jan. 24, 2000). Specifically, § 125(d)(2) exempts from Hatch-Waxman:
any application for marketing in which the drug that is the subject of the application contains an antibiotic drug and the antibiotic drug was the subject of any application for marketing received by the Secretary of Health and Human Services under section 507 of such Act ( 21 U.S.C. § 357 [Section 507]) before the date of enactment of this Act.
Pub.L. No. 105-115, 111 Stat. 2327 (1997), § 125(d)(2) ( reprinted in 21 U.S.C.A. § 355 Historical and Statutory Notes, "Transition"). Antibiotic drugs that were the subject of pre-FDAMA applications are known as "old antibiotics" and will be so referenced here.
With the enactment of FDAMA in 1997, Congress moved the definition of an "antibiotic" drug to 21 U.S.C. § 321(jj). It now states that an antibiotic drug is:
any drug (except drugs for use in animals other than humans) composed wholly or partly of any kind of penicillin, streptomycin, chlortetracycline, chloramphenicol, bacitracin, or any other drug intended for human use containing any quantity of any chemical substance which is produced by a micro-organism and which has the capacity to inhibit or destroy microorganisms in dilute solution (including a chemically synthesized equivalent of any such substance) or any derivative thereof.21 U.S.C. § 321(jj).
III. THE ADMINISTRATIVE RECORD
A. The 1998 Decision
One capsule of Periostat contains doxycycline hyclate 20 mg. In 1998, FDA agreed that the amount of doxycycline in Periostat was insufficient to have antimicrobial effect. While its NDA application was pending, CollaGenex argued to FDA:
The 2003 Decision concludes otherwise.
Given the fact that Periostat® does not kill or inhibit microorganisms, it seems both counterintuitive and potentially confusing to treat it as an antibiotic. Further, there is no legal reason to do so: Periostat® does not fit the legal definition of an antibiotic because, among other reasons, it does not have the capacity to inhibit or destroy microorganisms.
. . . .
Periostat® is not intended to nor does it destroy or inhibit microorganisms. To be sure, in dosages substantially higher than those in Periostat®, doxycycline has an antimicrobial effect and doxycycline is approved for that use at dosages of 50 mg. twice daily and above. At the 20 mg. dosage in Periostat®, however, doxycycline does not destroy or inhibit microorganisms. . . .
. . . .
[The FFDCA definition of an antibiotic] clearly contemplates that quantity matters. To be an antibiotic, a drug must contain a "quantity of a chemical substance . . . which has the capacity to inhibit or destroy microorganisms in dilute solution." A quantity of drug that does not have the capacity to inhibit or destroy microorganisms would not fit the definition.
A.R. Part A, Letter from Nancy Luque to Dr. Lumpkin, July 8, 1998, at 2-3.
The FDA "was unable to locate contemporaneous documentation for the 1998 [D]ecision" that Periostat is an "antibiotic." Fed. Defs.' Motion for SJ at 11. It submits the Lumpkin Declaration to explain the 1998 Decision. Dr. Lumpkin acknowledges the multiple pieces of correspondence from CollaGenex, arguing that Periostat should not be classified as an "antibiotic" because "it is not intended to kill or inhibit microrganisms and . . . it does not kill or inhibit microorganisms" due to the low level of doxycycline in Periostat. A.R. Part A, Lumpkin Dec. ¶ 6. Nonetheless, the decision that Periostat should be treated as an antibiotic drug "was not a particularly difficult decision" because the FDA had previously regulated products containing the substance "doxycycline" under Section 507 of the FFDCA. Id. ¶ 8 ("There was already a long history of FDA regulating products containing the substance `doxycycline' under the previous Section 507 antibiotic provisions of the FFDCA.").
Further, Dr. Lumpkin read the definition of "antibiotic" — both in Section 507 of FFDCA and Section 321(jj) of FDAMA — as covering
"any drug . . . containing any quantity of a chemical substance" which meets the two hurdles of (a) production by a microorganism and (b) having the capacity to inhibit or destroy microorganisms in dilute solution. . . . In FDA's opinion and practice[,] the phrase ". . . which has the capacity to inhibit or destroy microorganisms" refers in the definition sentence to the "chemical substance" and not to "any quantity." This interpretation, FDA believes, is consistent with the fundamental historical purpose and intent of the original 507 section of the FFDCA.Id. Considering that it was potential flaws in the manufacturing processes that caused Congress to regulate antibiotics starting with penicillin, Dr. Lumpkin concluded that the two characteristics identified in Section 507 apply to the substance produced by the manufacturing process and not to the individual drug products later made with that substance. Id. ¶ 9. It seemed illogical to conclude otherwise. Id. ("Again, given that these provisions were enacted to address safety concerns during manufacturing, it could not follow that a substance manufactured by this methodology would, under certain circumstances, fall under this Section [507] and in certain circumstances would not fall under this Section."). Doxycycline "clearly is produced by a microorganism and in dilute solution it inhibits microorganisms." Id. ¶ 11. Therefore,
[a]s the product in application 50-774 contains doxycycline ("at any quantity") and is intended for use in man, it was the FDA's decision that the product must be regulated as an antibiotic drug, as defined in the FFDCA, consistent with our regulation of all other doxycycline-containing products. . . . In this case, the status of the product as an antibiotic drug made it subject to the exemption provisions of FDAMA.Id. ¶ 12. Other than correspondence from CollaGenex and the final letter telling CollaGenex of the approval of its NDA, these statements constitute the substance of the administrative record for the 1998 Decision.
B. The 2003 Decision
Initially because CollaGenex requested reconsideration, and then because FDA faced this lawsuit, in 2003 FDA performed a review of the 1998 Decision concerning Periostat. The 2003 Decision concurred that Periostat is properly classified as an antibiotic drug for two reasons: first, that the amount of doxycycline hyclate in a daily dose of Periostat of 40 mg (two 20-mg capsules) "inhibit some micro-organisms as determined by in vitro susceptibility testing;" and second, that the most reasonable interpretation of the statutory language is that "any drug intended for human use containing any quantity of an antibiotic substance is considered to be an antibiotic drug." A.R. Part B, Roeder Mem. at 2. FDA's 2003 Decision recognized that "CDER does not have a formal, written statement of its policy regarding the classification of antibiotic drugs under the second, general, definition of antibiotic drug [in Section 507 and Section 321(jj)]." Id. at 6. Therefore, the Roeder Memorandum was in actuality a research project to determine if FDA has been consistent in its interpretation of the statutory language and, if so, how that interpretation applied to Periostat. Dr. Roeder properly focused on (1) FDA treatment of drugs that meet the FDA definition of "antibiotic" drug but are not indicated or intended for antimicrobial use; and (2) drugs using a drug substance that is classified as an "antibiotic" but in an application or dose that is notanti-microbial.
It would not surprise the Court if one of the reasons for the review in 2003 was the absence of an administrative record to support the 1998 Decision.
The Court declines to rely on the in vitro tests as these results directly contradict testing performed by CollaGenex and accepted in 1998 by the FDA, see Pltf's Motion for SJ at 4 n. 2, 34-36; and because the record is unclear as to the reliability of in vitro tests. See, e.g., A.R. Part B, Cyclosporine 1994 Mem. at 2 ("The assumption that clinically relevant antibiotic activity will correlate with in vitro MIC values determined for all human pathogens is unwarranted."). FDA has interpreted dilute solution to "correspond to concentrations that are found in human tissue at proposed or approved human dosing levels," A.R. Part B, Roeder Mem. at 7, and not concentrations found in a laboratory dish.
Discussing the FDA's treatment of cyclosporine, a drug to fight organ rejection (treated as an antibiotic), lovastatin, a cholesterol lowering agent (not treated as an antibiotic), and simvastatin, another cholesterol lowering agent (not treated as an antibiotic), Dr. Roeder concluded that FDA had consistently determined that, although all three substances are produced by microorganisms, only cyclosporine has an antimicrobial effect in dilute solution when used at the directed dosages. Id. at 8. "FDA therefore concluded that cyclosporine was appropriately classified as an antibiotic, whereas lovastatin was not classified as an antibiotic drug because it did not have adequate antimicrobial activity at concentrations corresponding to the recommended human dose and therefore did not meet the `dilute solution' criterion of the definition." Id. From these examples, Dr. Roeder drew the conclusion that when faced with a drug that meets FDA's definition of an antibiotic drug but that is not intended to be used to inhibit or kill microorganisms, FDA "has looked at whether there is a proposed or approved human dose that corresponds to a concentration at which the substance would have the capacity to inhibit or destroy micro-organisms." Id. at 9.
As to the second question, the Roeder Memorandum acknowledged,
We have been unable to locate documents in which the agency has discussed the question of whether a drug that has previously been correctly classified as an antibiotic drug can, or must, subsequently be classified as a non-antibiotic drug if it is later used for a non-antimicrobial use at doses that correspond to a dilute solution concentration that lacks the capacity to inhibit or destroy micro-organisms.Id. at 10. In other words, there is no written record of FDA ever before facing the issue presented by the Periostat NDA: the use of an "antibiotic" drug at a dose that cannot affect microbes and for a use that is not intended to be anti-microbial. There is only the fact that Dr. Lumpkin concluded in the 1998 Decision that Periostat should be treated as an antibiotic.
The Roeder Memorandum concluded that Periostat is properly classified as an "antibiotic" because "[t]he statute clearly links the antibiotic definition to the specific properties of the chemical substance, not to the particular indication or dose." Id. FDA reads the general definition in Section 321(jj) as:
any other drug intended for human use containing any quantity of any chemical substance which is [1] produced by a micro-organism and [2] which has the capacity to inhibit or destroy micro-organisms in dilute solution. . . .21 U.S.C. § 321(jj) (emphasis added). CollaGenex reads the same language with a different emphasis: any quantity of any chemical substance which is [1] produced by a micro-organism and [2] which has the capacity to inhibit or destroy micro-organisms in dilute solution. Pltf's Motion for SJ at 13-14. FDA concludes that any quantity of a substance meeting the criteria of [1] and [2] "is considered to be an antibiotic. Thus, . . . Periostat contains a quantity of doxycycline, and doxycycline is known to have the capacity to inhibit or destroy micro-organisms in a dilute solution corresponding to approved human doses (ranging from 100 to 200 mg per day taken orally) at which it is indicated for use for the treatment of a variety of infections." A.R. Part B, Roeder Mem. at 10-11. Although all parties prior to the 2003 Decision agreed that the doxycycline hyclate in Periostat is insufficient to have an antibiotic, or antimicrobial, effect, the Roeder Memorandum determined that "[t]he classification of a drug as an antibiotic is based on the specific characteristics of the chemical drug substance or moiety. Therefore, an antibiotic drug cannot be reclassified if it is used, in a particular drug product, at a sub-antimicrobial dose." Id. at 11.
"[A]ctive moiety means the molecule or ion . . . responsible for the physiological or pharmacological action of the drug substance." 21 C.F.R. § 314.108.
IV. LEGAL STANDARDS
FDA's decision that Periostat is an "antibiotic" drug is subject to review under the Administrative Procedure Act ("APA") and will be reversed only if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A). This standard is commonly deferential to agency action. Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416 (1971). When an agency is construing its own organic statute, this deference is at its highest. Chevron U.S.A., Inc. v. Natural Res. Def. Council, 467 U.S. 837 (1984); see also United States v. Mead Corp, 533 U.S. 218 (2001). "If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress." Chevron, 467 U.S. at 842-43. If the language of the statute is not clear and unambiguous, a court may not "simply impose its own construction," but must determine whether the agency's construction is a permissible interpretation. Id. at 843; see also id. at 843 n. 11 (agency's interpretation need not be the only one it could have adopted or even the one the court would have adopted); Barnhart v. Walton, 535 U.S. 212, 218 (2002) (court must decide first, whether statute unambiguouslyforbids agency interpretation, and second, whether that interpretation exceeds the bounds of permissible); Mead, 533 U.S. at 229 (a court cannot reject an agency interpretation that is merely "unwise," if Congress has not spoken clearly and the agency's interpretation is "reasonable.").
The FDA particularly is often accorded special deference when its decisions are based on an evaluation of the scientific record before it. "There is no denying the complexity of the statutory regime under which FDA operates, the FDA's expertise or the careful craft of the scheme it devised to reconcile the various statutory provisions. . . . We therefore accord Chevron deference to FDA's letter decision here. . . ." Mylan Labs., Inc. v. Thompson, No. 04-5296, 2004 WL 2710043, at *6 (D.C. Cir. Nov. 30, 2004). Courts "review scientific judgments of the agency `not as the chemist, biologist, or statistician that we are qualified neither by training nor experience to be, but as a reviewing court exercising our narrowly defined duty of holding agencies to certain minimal standards of rationality.'" Troy Corp. v. Browner, 120 F.3d 277, 283 (D.C. Cir. 1997) (citation omitted). When FDA's interpretation of the FFDCA "rests on `the agency's evaluations of scientific data within its area of expertise'" it is "entitled to a `high level of deference' from this court." Serono Labs., Inc. v. Shalala, 158 F.3d 1313, 1320 (D.C. Cir. 1998) (citation omitted).
Policy judgments made by an agency within its area of expertise are also entitled to deference from the courts:
Such deference, the Supreme Court recently explained, is justified because the responsibilities for assessing the wisdom of policy choices and resolving the struggle between competing views of the public interest are not judicial ones, and because of the agency's greater familiarity with the ever-changing facts and circumstances surrounding the subjects regulated.Nat'l Rifle Ass'n v. Reno, 216 F.3d 122, 132 (D.C. Cir. 2000) (quoting in part FDA v. Brown Williamson Tobacco Corp., 529 U.S. 120, 131 (2000) (citations and internal punctuation omitted)).
Even when Chevron deference does not apply, the courts will give "considerable and in some cases decisive weight" to an agency interpretation of a statute that is "made in pursuance of official duty, [and is] based upon more specialized experience and broader investigations and information" than a court might have, as long as the decision is carefully and thoughtfully made. Skidmore v. Swift Co., 323 U.S. 134, 139-40 (1944).
Upon review under the APA, an agency must demonstrate that it "examine[d] the relevant data and [can] articulate a satisfactory explanation for its action including a `rational connection between the facts found and the choice made.'" Motor Vehicle Mfrs. Ass'n of the United States, Inc., et al. v. State Farm Mutual Auto. Ins. Co., et al., 463 U.S. 29, 43 (1983) (citing Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168 (1962)). "[T]he court must consider whether the decision was based on a consideration of the relevant factors." Overton Park, 401 U.S. at 416. To be sustained, an agency decision must be one that "consider[ed] the relevant factors" and "is within the bounds of reasoned decisionmaking." Baltimore Gas Electric Co. v. Natural Res. Def. Council, Inc., 462 U.S. 87, 105 (1983). In addition, a "fundamental rule of administrative law" is that a court reviewing an agency decision "must judge the propriety of [agency] action solely by the grounds invoked by the agency." SEC v. Chenery, 332 U.S. 194, 196 (1947). Because the court's review "is confined to the administrative record at the time of the agency's decision, it may not include `some new record made initially in the reviewing court.'" Fund for Animals v. Williams, 245 F. Supp. 2d 49, 54 (D.D.C. 2003) (citations omitted). The courts "do not rely on counsel's post hoc rationalization for upholding an agency's action." McDonnell Douglas Corp. v. Air Force, 375 F.3d 1182, 1188 (D.C. Cir. 2004); see also Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 212 (1988) (courts may not accept counsel's post hoc rationalizations for agency orders); Williams Gas Processing-Gulf Coast Co., L.P. v. FERC, 373 F.3d 1335, 1345 (D.C. Cir. 2004) ("[ P] ost hoc rationalizations by agency counsel will not suffice.") (quoting Western Union Corp. v. FCC, 856 F.2d 315, 318 (D.C. Cir. 1988)).
V. ANALYSIS
FDA defends the 1998 Decision and the 2003 Decision as compelled by FDAMA and a reasonable interpretation of the statutory definition of "antibiotic" to which the Court must give deference. Its positions are strongly supported by Intervenors IVAX and CorePharma. CollaGenex just as strongly disagrees and also attacks FDA's reliance on FDAMA as the construct of counsel after the fact and not part of the agency's decision(s).
The Court thanks all parties for their excellent briefs which were submitted on an expedited schedule without loss of any quality.
A. FDAMA
The Court finds that the statutory command in FDAMA is clear and binding on the FDA and the courts. When "the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress." Chevron, 467 U.S. at 842-43. FDAMA explicitly exempts from the benefits of Hatch-Waxman a drug that "contains an antibiotic drug," when that antibiotic drug was the "subject of any application for marketing" received by HHS prior to the enactment of FDAMA. FDAMA § 125(d)(2). The "antibiotic drug" in question is doxycycline, a substance that fully meets the definition of "antibiotic" drug in Section 321(jj). FDA properly applied the exemption to Periostat because Periostat contains doxycycline and doxycycline was the subject of applications submitted to FDA under 21 U.S.C. § 357 prior to the effective date of FDAMA.
This conclusion does not rely on deference to FDA. "Deference to an agency's statutory interpretation `is only appropriate when the agency has exercised its own judgment,' not when it believes that interpretation is compelled by Congress." Arizona v. Thompson, 281 F.3d 248, 254 (D.C. Cir. 2002) (citations omitted) (emphasis in original).
FDA did not argue that FDAMA controls the result when before the Court in 2003 on CollaGenex's application for a preliminary injunction. For that matter, FDA also refused to engage in any discussion or argument concerning the definition of an "antibiotic" drug because the administrative record of the 1998 Decision (which, it turns out, did not exist) had not yet been presented to the Court and, as it turns out, the agency had not developed a studied interpretation of the statutory text.
CollaGenex resists this conclusion on two grounds. First, it argues that the doxycycline hyclate 20 mg. dose in Periostat is not an "antibiotic" drug because it has insufficient strength to be anti-microbial. Second, it argues that doxycycline was not the subject of previous marketing applications to FDA because applications are submitted for drug products, not drug substances. This reading of the statute is not only strained but would also undercut the purposes of Hatch-Waxman and FDAMA to encourage innovate new antibiotics — not new antibiotic uses of old antibiotics. See, e.g., H.R. REP. NO. 105-310, at 77 (1997) ("the granting of market exclusivity be limited to products that achieve the policy objective of increasing research toward the development of new antibiotics.").
A "drug substance" is "an active ingredient that is intended to furnish pharmacological activity or other direct effect in the diagnosis, cure, mitigation, treatment, or prevention of disease. . . ." 21 C.F.R. § 314.3(b). A "drug product" is "a finished dosage form, for example, tablet, capsule, or solution, that contains a drug substance, generally, but not necessarily, in association with one or more ingredients." Id. With court approval, FDA reads the term "drug" differently in different portions of the FFDCA according to its context. See Baker Norton Pharm., Inc. v. FDA, 132 F. Supp. 2d 30 (D.D.C. 2001).
There is really no dispute that doxycycline is an antibiotic drug that was subject to pre-FDAMA approval by FDA. See 21 C.F.R. §§ 446.20 (pre-FDAMA monograph for doxycycline hyclate); 445.120a (pre-FDAMA monography for doxycycline hyclate capsules); 446.121a (pre-FDAMA monograph for doxycycline monohydrate) (1998) (all repealed after passage of FDAMA). Essentially, CollaGenex argues that each of these monographs reflected approval for an antibiotic drug product — using doxycycline at higher doses than in Periostat — so that they are irrelevant to the Periostat classification as an antibiotic. The Court disagrees. There is no indication in the statutory text or legislative history that Congress intended a mere change in the quantity of the drug substance in a drug product to enable an old antibiotic to be treated as a new antibiotic under Hatch-Waxman. Indeed, the legislative history is directly to the contrary. See 143 CONG. REC. H8479 (Oct. 7, 1997) (Rep. Deutch: "I think it is important that any additional exclusivity that we grant in terms of antibiotics, which would be the first time that there would be exclusivity for antibiotic drugs, that it be limited in scope very narrowly to the challenge we face in terms of resistant strains."); 143 CONG. REC. S12243 (Nov. 9, 1997) (Sen. Kennedy: The legislation "provides incentives . . . for development of new antibiotics to deal with emerging, drug-resistant strains of disease.").
Whether Periostat is itself an "antibiotic" drug or not, it clearly "contains" the drug doxycycline which was classified as antibiotic as long ago as 1967. Thus, its NDA applied for marketing of a drug (Periostat) that contains an antibiotic drug (doxycycline) and doxycycline was the subject of marketing applications received by FDA before 1997. The FDAMA exemption in Section 125(d)(2) clearly applies and bars Periostat from the benefits of Hatch-Waxman.
CollaGenex would have the Court ignore the applicability of FDAMA because it does not see any reliance on that statute in either the 1998 Decision or the 2003 Decision. It asserts, "The argument is nothing more than a post hoc rationalization of counsel" and should therefore be rejected. Pltf.'s Opp. to Defs.' Motions for SJ at 3. "An agency decision must be upheld, if at all, on the same basis articulated in the decision by the agency itself." Id.
The Court agrees with CollaGenex on the law but disagrees on the facts. The very reason Dr. Lumpkin found the 1998 Decision "not particularly difficult" was that FDA had long regulated doxycycline as an antibiotic and Periostat contains doxycycline. A.R. Part A, Lumpkin Dec. ¶ 8 ("There was already a long history of FDA regulating products containing the substance `doxycycline' under the previous Section 507 antibiotic provisions of the FFDCA."). This logic exactly follows the contours of the FDAMA exemption — which had only recently been adopted at the time of the 1998 Decision — and presumably explains the lack of a paper record behind the 1998 Decision. In addition, FDAMA was explicitly cited in the letter to CollaGenex approving Periostat as an antibiotic drug.
Because the Court finds that the FDAMA exemption applies to Periostat, it sustains the 1998 Decision by FDA without regard to whether Periostat itself is an "antibiotic" drug. The 2003 Decision is irrelevant to this conclusion.
B. The Statutory Definition of "Antibiotic" Drug
The Court reaches this issue so that this memorandum opinion addresses both sets of arguments, in case on appeal the Court of Appeals disagrees with the FDAMA analysis.
While CollaGenex presents a perfectly plausible interpretation of the definition of "antibiotic" drug at 21 U.S.C. § 321, the contrary interpretation presented by FDA is at least as plausible. Under such circumstances, the Court concludes that the statute is ambiguous and the agency's interpretation is entitled to Chevron step 2 deference.
It is to be remembered that the difference between FDA's interpretation and CollaGenex's interpretation of the statutory language rests on a difference of emphasis: is an antibiotic "any quantity of any chemical substance which is [1] produced by a micro-organism and [2] which has the capacity to inhibit or destroy micro-organisms in dilute solution" or " any quantity of any chemical substance" which has the attributes of [1] and [2]. CollaGenex argues that FDA's interpretation would read out of the statute the references to "any quantity" and "dilute solution" because it "ignores the fact that the drug must contain some amount of an antibiotic, that is, of a chemical substance with the capacity to inhibit or destroy micro-organisms." Pltf.'s Opp. to Defs.' Motions for SJ at 12. Without the benefit of any input from FDA, which stood mute on the question of statutory interpretation at the preliminary injunction stage, this is how the Court first read the law as a "colorable" interpretation. CollaGenex, 2003 U.S. Dist. LEXIS 12523, at *30.
This is not the fault of counsel. As the 2003 Decision makes clear, FDA had never before this litigation articulated a studied interpretation of this section of the law.
It is not illogical to believe that the "quantity of the substance in the drug must be factored into the definition of an antibiotic drug," since only a sufficient quantity can have antimicrobial effect. Pltf.'s Opp. to Defs.' Motions for SJ at 12. However, it is also not illogical to take the approach adopted by the FDA in the 2003 Decision: if the substance in the drug has antibiotic properties and is intended for human use at any such dose, then it is an antibiotic drug at all doses. 2003 Decision at 10. In this approach, FDA does not omit reliance on "any quantity" at all; the words "any quantity" mean that if a chemical substance has the attributes of [1] and [2], "any quantity" of that chemical substance, no matter how small, is an antibiotic drug under the law even if not in effect. "[I]n dilute solution" has meaning under FDA's interpretation because it contributes to defining the nature of the "chemical substance" at issue. While many substances may have an antimicrobial effect at high concentrations, FDA reasonably limits its definition to those concentrations proposed or approved for use in humans — those that have an antimicrobial effect in human tissue when taken as directed. As CollaGenex argues, Congress could have written the statute to achieve this same result in a more direct fashion but the fact that the legislative process ended with some ambiguity does not overcome the legitimate deference due to the agency for any "reasonable" interpretation. Mead, 533 U.S. at 229 (When Chevron deference is applicable, a "reviewing court has no business rejecting an agency's exercise of its generally conferred authority to resolve a particular statutory ambiguity simply because the agency's chosen resolution seems unwise, but is obliged to accept the agency's position if Congress has not previously spoken to the point at issue and the agency's interpretation is reasonable.") (citations omitted).
FDA's treatment of lovastatin, provastatin, and simvastatin as non-antibiotics — even though all three drugs contain a substance made by a micro-organism and that has antimicrobial effects — is not inconsistent with the approach adopted by the 2003 Decision and does not require that Periostat receive the same treatment. There are no dosages of these drugs at which they have antimicrobial effect when administered to humans. Thus, they do not have an antimicrobial effect "in dilute solution," i.e., in human tissue. The same cannot be said of doxycycline hyclate, the active ingredient in Periostat. Doxycycline is prescribed for various human infections at dosages of 50 mg or more, as CollaGenex noted in its letter arguments to FDA. It, therefore, does have an antimicrobial effect "in dilute solution" and FDA can properly find that a drug product with "any quantity" of doxycycline is also an antibiotic.
CollaGenex argues that it is illogical and contrary to its own precedents for FDA to define Periostat as an antibiotic because of the characteristics of a different drug, Vibramycin, which also contains doxycycline but at a much higher dose. Pltf.'s Opp. to Defs.' Motions for SJ at 21. According to the 2003 Decision, FDA has never before articulated a rationale for classifying drugs as antibiotics, making these decisions on a case-by-case basis, because it never before had an application for use of a drug with a non-antimicrobial quantity of a substance that had been used in much higher doses as an antibiotic. 2003 Decision at 10. The 2003 Decision thus represents the very current and new articulation of the agency's interpretation of the statute. Presuming for the sake of argument that there are earlier FDA drug classifications that do not meet the new interpretation, that alone would not prevent FDA from developing its current position and earning the deference of the Court to it.
The Court defers to FDA in its interpretation of the statutory definition of an "antibiotic." The 2003 Decision provides a reasonable interpretation of ambiguous language which is certainly permissible. The 2003 Decision "claim[s] the merit of its writer's thoroughness, logic and expertness," and, as such, is entitled to deference. Mead, 533 U.S. at 235.
VI. CONCLUSION
The complaint will be dismissed and the preliminary injunction entered on July 23, 2003, will be dissolved. FDAMA exempts Periostat from the benefits of Hatch-Waxman and supports the 1998 Decision of FDA to that effect. As articulated in the 2003 Decision (consistent with the 1998 Decision), FDA has adopted a reasonable interpretation of the statutory definition of "antibiotic" drug in the FFDCA to which Chevron deference is due. A separate order accompanies this memorandum opinion.