Sanofi-Aventis U.S. LLC v. Food and Drug Administration

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA
____________________________________
)
SANOFI-AVENTIS U.S. LLC, )
)
Plaintiff, )
)
v. ) Civil Action No. 10-01255 (ABJ)
)
FOOD AND DRUG ADMINISTRATION, )
et al., )
)
Defendants. )
____________________________________)

MEMORANDUM OPINION

Plaintiff sanofi-aventis U.S. LLC (“Sanofi”) brought this action against the Food and

Drug Administration (“FDA”), its Commissioner of Food and Drugs, Margaret A. Hamburg, and

the Secretary of Health and Human Services, Kathleen Sebelius, alleging that FDA exceeded its

statutory authority under the Federal Food, Drug, and Cosmetic Act (“FDCA”) and unlawfully

departed from agency precedent when it approved a generic version of the Sanofi drug Lovenox.

Sandoz, Inc. (“Sandoz”), the manufacturer of the generic drug, intervened as a defendant, and

now the parties have cross-moved for summary judgment. The Court concludes that 1) the FDA

acted within its statutory authority when it called for Sandoz to file immunogenicity data as part

of its abbreviated new drug application; 2) it did not unlawfully depart from agency precedent by

approving a generic before the listed drug had been fully characterized; and 3) it reasonably

found that the active ingredient in the generic drug was the same as the active ingredient in

Lovenox. Therefore, plaintiff’s motion for summary judgment will be denied, and defendants’

cross-motions will be granted.
 BACKGROUND

I. Statutory Background

The FDCA requires all new drugs to be approved by the FDA before they are introduced

into interstate commerce. 21 U.S.C. § 355(a). It provides two primary pathways for obtaining

approval: (1) the new drug application (“NDA”), described in section 355(b); and (2) the

abbreviated new drug application (“ANDA”) for generic products set forth in section 355(j).

A drug that follows the NDA pathway is referred to as a “pioneer” drug because it is the

first drug of its kind to go through an approval process with the FDA. The NDA procedure

requires the applicant to conduct a spectrum of safety and effectiveness tests and to inform the

FDA of the results. The information that must be provided with an NDA includes in relevant

part: “full reports of investigations which have been made to show whether or not such drug is

safe for use and whether such drug is effective in use,” § 355(b)(1)(A); “a full list of the articles

used as components . . . ” and “a statement of the composition…” of the drug, § 355(b)(1)(B)–

(C); and, “a full description of the methods used in, and the facilities and controls used for, the

manufacture, processing, and packing of such drug,” § 355(b)(1)(D). Once the drug is approved,

it is referred to as a “listed drug.” See 21 C.F.R. § 314.3(b).

In some cases, a new drug applicant may seek to rely on research conducted by a third

party in order to meet the approval requirements. 1 In that instance, the statute sets out a

procedure under section 355(b)(2), which requires the applicant to file additional information

showing that the drug’s approval will not infringe a valid patent. 21 U.S.C.§ 355(b)(2).

1 The statute specifies that this pathway is to be utilized when “[a]n application submitted
under paragraph (1) for a drug for which the investigations [demonstrating safety and
effectiveness] and relied upon by the applicant for approval of the application were not
conducted by or for the applicant and . . . the applicant has not obtained a right of reference or
use from the person by or for whom the investigations were conducted.” 21 U.S.C. § 355(b)(2).
2
 Congress added the truncated ANDA approval process to the FDCA as part of the 1984

Hatch-Waxman amendments, which sought “to make available more low cost generic drugs” by

providing a pathway that was less costly and time consuming than the NDA process. Serono

Labs., Inc. v. Shalala, 158 F.3d 1313, 1316 (D.C. Cir. 1998), citing H.R. REP. NO. 98-857, pt. 1,

at 14 (1984). ANDA applicants must file information showing that the conditions of use, active

ingredient, dosage form, strength, route of administration, and labeling of the generic drug are

“the same as” those of the reference listed drug (“RLD”) 2 that was previously approved. 3 21

U.S.C.. § 355(j)(2)(A)(i)–(iii), (v). They are thereby relieved of the obligation to supply the

extensive testing demonstrating safety and effectiveness that is the hallmark of the NDA process,

see § 355(b)(1)(A), but ANDA applicants are still required to supply the other information

required of a new drug applicant. See § 355(j)(2)(A)(vi)(“An abbreviated application for a new

drug shall contain – . . . (vi) the items specified in clauses (B) through (F) of subsection (b)(1) of

this section[.]”). This means the ANDA applicant must list the components and composition of

the generic drug, and must provide “a full description of the methods used in, and the facilities

and controls used for, the manufacture, processing, and packing of such drug . . . .”

§ 355(b)(1)(a)(D), incorporated into requirements for ANDA applications by § 355(j)(2)(A)(vi).

2 A “reference listed drug,” or RLD, is “the listed drug identified by FDA as the drug
product upon which an applicant relies in seeking approval of its abbreviated application.” 21
C.F.R. § 314.3(b).

3 Part of this opinion concerns the showing to be made under section 355(j)(2)(A)(ii) that
the generic drug’s active ingredient is “the same as” the RDL’s active ingredient, which will be
referred to as the showing of “active ingredient sameness.” The statute also requires a showing
of bioequivalence, see § 355(j)(2)(A)(iv), which Sanofi defines as “the absence of a significant
difference in the rate and extent to which the active ingredient in a pharmaceutical equivalent or
alternative becomes available at the site of drug action, when administered at the same does,
under similar conditions, in an appropriately designed study,” Sanofi’s Mem. at vi.
Bioequivalence is not contested in this case.
3
 But, in accordance with Congress’s goal to keep the ANDA pathway less costly and time

consuming than the NDA pathway, the statute expressly prohibits the FDA from requiring

ANDA applicants to submit any other categories of information. § 355(j)(2)(A), (j)(4). Section

355(j)(2)(A) provides: “The Secretary may not require that an abbreviated application contain

information in addition to that required by clauses (i) through (viii).” In addition, the statute

limits the FDA’s discretion to reject an ANDA. § 355(j)(4). Section 355(j)(4) mandates that

“the Secretary shall approve” an ANDA “unless” he or she makes certain specified findings,

including that the generic drug’s active ingredient is not the same as the listed drug’s active

ingredient, or that “the methods used in, or the facilities and controls used for, the manufacture,

processing, and packing of the drug are inadequate to assure and preserve its identity, strength,

quality and purity[.]” § 355(j)(4).

II. Factual Background 4

A. Sanofi and Lovenox

Sanofi owns the NDA for the injectable anti-coagulation drug Lovenox, which was

approved by FDA in 1993. AR 2881–82. The active ingredient in Lovenox is a compound

called enoxaparin sodium (“enoxaparin”). AR 2882. Enoxaparin is made up of a core protein

from which an assortment of different sugar chains, known as oligosaccharide chains, extend.

AR 5, 12, 2882. To date, no one has fully determined enoxaparin’s complete chemical makeup,

or fully “characterized” it, because the sugar chains are too difficult to identify, and the relative

abundance of the different chains varies from batch to batch of enoxaparin. AR 10–12, 2904.

4 The factual background is also laid out in great detail in the Court’s preliminary
injunction opinion, Sanofi-Aventis U.S. LLC v. FDA, 733 F. Supp. 2d 162, 164–66 (D.D.C.
2010).

4
Apparently, this variation is common among compounds in the class of anticoagulants that

enoxaparin belongs to, called low molecular weight heparins. AR 2884.

On February 19, 2003, Sanofi submitted a Citizen Petition5 urging FDA to withhold

approval of any ANDA for generic enoxaparin “[u]ntil such time as enoxaparin has been fully

characterized . . . unless the manufacturing process used to create the generic product is

determined to be equivalent to [Sanofi’s] manufacturing process for enoxaparin, or the

application is supported by proof of equivalent safety and effectiveness demonstrated through

clinical trials.” 6 AR 1. FDA ultimately rejected this request to forestall the marketing of a

generic. 7 AR 2878–2922.

Instead, FDA found that “enoxaparin has been adequately characterized for the purposes

of approving . . . generic enoxaparin,” and it articulated a five-pronged test to be used to

determine whether the active ingredient in any proposed generic version of Lovenox would be

5 A citizen petition is a document submitted to FDA by a third party under 21 C.F.R.
§ 10.30, which requests that FDA take or refrain from taking a particular action.

6 At the motions hearing, counsel for Sanofi assured the Court, “if a generic applicant were
able – and this is clearly laid out in the Citizens Petition – if a generic applicant were able to
fully characterize themselves enoxaparin and lay out their product and lay out Sanofi’s product
and say, look, it matches up, matches up, matches up, that would be sufficient.” Tr. 42. But
counsel then acknowledged that “it could not be done under current technology.” Tr. 43.
Sanofi’s position has been, then, that a generic version of Lovenox could only be approved at
some unspecified point in the future when the technology necessary for characterization has
evolved. Tr. 43. Absent that, according to Sanofi, a generic should be subject to the full range
of NDA testing for safety and effectiveness, or its manufacturing process should be shown to be
identical to the RLD’s. Tr. 43–44.

7 Sanofi also asked FDA to reject any application for generic enoxaparin that did not show
the drug to contain a certain type of sugar chain (1,6 anhydro ring structure) in similar
concentrations to Lovenox. AR 1. Sanofi identified that chain as important to enoxaparin’s
overall pharmacological effect. AR 1. FDA accepted this part of the Citizen Petition. AR
2879–80.

5
the same as the enoxaparin in Lovenox. 8 AR 2879–80. According to FDA, “each of [the five

prongs] captures different aspects of the active ingredient’s ‘sameness.’” AR 2879–80.

The record indicates that when the five-pronged approach was under consideration, there

was a difference of opinion among two internal FDA units. AR 3836. While the Office of

Generic Drugs (“OGD”) supported the test, the Office of New Drug Quality Assessment

(“ONDQA”) argued that the test was insufficient, and that the only way to show active

ingredient sameness would be to fully characterize enoxaparin. AR 3836. The determination to

adopt the test was made by the Deputy Director of the Office of Pharmaceutical Science, Center

for Drug Evaluation and Research, who, in a memorandum that thoroughly considered both

sides’ arguments, found the five-pronged test to be sufficient. AR 3836–61.

B. Sandoz

On August 26, 2005, while Sanofi’s Citizen Petition was pending, Sandoz filed an

ANDA for generic enoxaparin. See AR 4440. FDA approved the ANDA on July 23, 2010, and

it rejected Sanofi’s Citizen Petition the same day. AR 4440–44. The approval process took just

under five years, and it included lengthy exchanges between Sandoz and FDA as well as multiple

amendments to the ANDA. See AR 4440-44.

At issue here is FDA’s request, two years into the approval process, for information

regarding the potential of Sandoz’s proposed drug to elicit an adverse immune response (its

immunogenicity). AR 4167–73. In making its request, FDA relied on studies showing that

enoxaparin has been known to cause a dangerous immune response in certain patients, called

8 These five prongs address: (1) “the physical and chemical characteristics of enoxaparin”;
(2) “the nature of the source material and the method used to break up the polysaccharide chains
into smaller fragments”; (3) “the nature and arrangement of components that constitute
enoxaparin”; (4) “certain laboratory measurements of anticoagulant activity”; and (5) “certain
aspects of the drug’s effect in humans,” meaning the in vivo pharmacodynamics profile, which is
based upon its effects on two factors, anti-Xa and anti-IIa. AR 2880, 2899.
6
thrombocytopenia, which can be life-threatening. AR 3848–49, 3853. Importantly, the cause of

thrombocytopenia is complex. AR 3854. Although enoxaparin itself can stimulate

thrombocytopenia, it may also be stimulated by impurities in the drug. AR 2918, 3848–49,

3853–54. Furthermore, impurities may affect the strength of the reaction when it occurs. AR

2918, 3854.

In a November 5, 2007 letter to Sandoz, FDA concluded that its ANDA was “not

approvable because the application does not adequately address the potential for immunogenicity

of the drug product.” AR 4167. FDA required Sandoz to either amend the ANDA so that it

addressed that deficiency or to withdraw the application. AR 4167 In a December 4, 2007

follow-up letter, FDA explained its decision, informing Sandoz that its amended ANDA should

address the impurity profile of its generic enoxaparin and suggesting several approaches. AR

4170–74. The letter stated:

FDA is particularly concerned with product and process derived impurities that
may modify the biological activity or enhance the immunogenicity of your
product. Understanding the potential for your product to elicit an immune
response is critical, since low molecular weight heparins are associated with a
serious immune-driven adverse event, heparin induced thrombocytopenia (HIT).
Impurities can interact either with the product or with the host immune system in
ways that alter outcome. Thus, for products that have the potential for
immunologic adverse events and certainly for products with known immunologic
adverse events, the contribution of impurities needs to be carefully considered.

AR 4170. FDA asked Sandoz to address three concerns:

The ability of its generic drug to bind to and form complexes with the compound
PF4, relative to Lovenox. FDA asserted that one known cause of
thrombocytopenia is the presence of certain dangerous complexes that are formed
when enoxaparin binds to PF4. Furthermore, impurities are known to facilitate
the creation of these harmful complexes. Since Sandoz had sufficiently shown
that the enoxaparin in its generic drug was the same as in Lovenox, comparative
information about the ability of its generic drug to bind to and form these
enoxaparin-PF4 complexes relative to Lovenox would shed light on whether the
generic drug contains any harmful impurities.

7
 The amount and nature of potential product contaminants (innate immune
agonists) in its generic drug, relative to those in Lovenox.

The functional immunogenic properties of the generic drug, relative to Lovenox
(i.e., its actual effect on immune response). FDA explained that this could be
tested by in vitro assays or animal models that would show the immune response
elicited by the generic drug as compared to Lovenox.

AR 4170–73.

In response, Sandoz provided FDA with data from laboratory tests that compared the

immunity profile and immunogenicity of its generic enoxaparin to Lovenox. AR 4181–90. The

results submitted compared Sandoz’s generic to Lovenox with regard to: “(a) the ability of

enoxaparin to form complexes with PF4, (b) the presence of impurities that could stimulate the

immune system directly, (c) activation of human PBMC, and (d) the induction of antibodies to

the product in mice.” AR 4433; see also AR 4181–90.

Based on all the information that Sandoz submitted, including the immunogenicity data,

and its application of the five-pronged test described above, FDA found that “Sandoz’s ANDA

for enoxaparin sodium injection [met] the requirements for ANDA approval, including those

regarding active ingredient sameness and purity of the proposed drug.” AR 4437–38.

III. Procedural background

On July 26, 2010, Sanofi filed this action against FDA. Compl. [Dkt. # 1]. Count I

alleges that FDA exceeded its authority under the FDCA, in violation of the APA, by requiring

Sandoz to submit the immunogenicity data as part of its ANDA. Compl. ¶¶ 37–42. Count II

alleges that FDA departed from agency precedent in violation of the APA by approving Sandoz’s

ANDA before enoxaparin had been fully characterized. Compl. ¶¶ 43–46. Count III alleges that

FDA exceeded its authority under the FDA and acted contrary to established agency precedent in

violation of the APA by approving Sandoz’s ANDA without sufficient evidence that the active

8
ingredient in Sandoz’s generic enoxaparin was “the same as” the active ingredient in Lovenox.

Compl. ¶¶ 47–51.

On the same day it filed its complaint, Sanofi filed a motion seeking a temporary

restraining order (“TRO”) and preliminary injunction (“PI”) to compel FDA to withdraw

approval of Sandoz’s ANDA pending a trial on the merits. Pl.’s Mot. for TRO and PI [Dkt. # 3].

After consolidating the TRO and PI, the Court denied them both, relying in part on its finding

that Sanofi was unlikely to succeed on the merits of any of the three claims. Sanofi-Aventis U.S.

LLC, 733 F. Supp. 2d at 162.

On July 28, 2010, the Court granted Sandoz’s motion for leave to intervene as a

defendant. Sandoz’s Mot. to Intervene [Dkt. # 6]. The parties have now cross-moved for

summary judgment on all counts.

STANDARD OF REVIEW

Summary judgment is appropriate “if the movant shows that there is no genuine dispute

as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P.

56(a). The party seeking summary judgment bears the “initial responsibility of informing the

district court of the basis for its motion, and identifying those portions of the pleadings,

depositions, answers to interrogatories, and admissions on file, together with the affidavits, if

any, which it believes demonstrate the absence of a genuine issue of material fact.” Celotex

Corp. v. Catrett, 477 U.S. 317, 323 (1986) (internal quotation marks omitted). To defeat

summary judgment, the non-moving party must “designate specific facts showing that there is a

genuine issue for trial.” Id. at 324 (internal quotation marks omitted). The existence of a factual

dispute is insufficient to preclude summary judgment. Anderson v. Liberty Lobby, Inc., 477 U.S.

242, 247–48 (1986). A dispute is “genuine” only if a reasonable fact-finder could find for the

9
nonmoving party; a fact is only “material” if it is capable of affecting the outcome of the

litigation. Id. at 248. See also Laningham v. U.S. Navy, 813 F.2d 1236, 1241 (D.C. Cir. 1987).

“The rule governing cross-motions for summary judgment . . . is that neither party waives

the right to a full trial on the merits by filing its own motion; each side concedes that no material

facts are at issue only for the purposes of its own motion.” Sherwood v. Wash. Post, 871 F.2d

1144, 1148 n.4 (D.C. Cir. 1989), quoting McKenzie v. Sawyer, 684 F.2d 62, 68 n.3 (D.C. Cir.

1982). In assessing each party’s motion, “[a]ll underlying facts and inferences are analyzed in

the light most favorable to the non-moving party.” N.S. ex rel. Stein v. District of Columbia, 709

F. Supp. 2d 57, 65 (D.D.C. 2010), citing Anderson, 477 U.S. at 247.

ANALYSIS

I. FDA did not exceed its authority under the FDCA by requiring Sandoz to
submit immunogenicity data as part of its ANDA.

The first question at issue here – whether FDA had the authority to require Sandoz to

submit immunogenicity data for generic enoxaparin as part of its ANDA – can be decided on

summary judgment because it is a pure question of statutory interpretation. Plaintiff cites section

355(j)(2)(A) – the provision that prevents FDA from requiring ANDA applicants to submit

information not listed in the statute – and it asks the Court to declare that FDA exceeded its

authority under the FDCA, acted arbitrarily and capriciously, and abused its discretion by calling

for the comparative test results, thereby violating the APA, 5 U.S.C. § 706(2)(A), (C). Compl.

¶¶ 37–42.

A. The Chevron framework for the review of FDA action

The APA establishes the scope of judicial review of agency action, and the standard of

review under the APA is quite narrow. See Vermont Yankee Nuclear Power Corp. v. Natural

Res. Def. Council, Inc., 435 U.S. 519, 545–49 (1978),

10
 The Court is required to analyze an agency’s interpretation of a statute by following the

two-step procedure set forth in Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S.

837 (1984). First, the Court must determine “whether Congress has directly spoken to the

precise question at issue.” Id. at 842. “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.” Id. at 842–43. Courts “use ‘traditional tools of statutory construction’ to

determine whether Congress has unambiguously expressed its intent,” Serono Labs., Inc., v.

Shalala, 158 F.3d 1313, 1319 (D.C. Cir. 1998), including an examination of the statute’s text,

structure, purpose, and legislative history. Bell Atl. Tel. Co. v. FCC, 131 F.3d 1044, 1047 (D.C.

Cir. 1997).

If the Court concludes that the statute is either silent or ambiguous, the second step of the

review process is to determine whether the interpretation proffered by the agency is “based on a

permissible construction of the statute.” Chevron, 467 U.S. at 843. Once a reviewing court

reaches the second step, it must accord “considerable weight” to an executive agency’s

construction of a statutory scheme it has been “entrusted to administer.” Id. at 844. Indeed,

“under Chevron, courts are bound to uphold an agency interpretation as long as it is reasonable –

regardless whether there may be other reasonable, or even more reasonable, views.” Serono, 158

F.3d at 1321. And the court must defer to an agency’s reading of its own regulations unless it is

“plainly erroneous or inconsistent with the regulation.” Id. at 1320 (internal quotation marks

omitted).

Using this framework, the Court reaffirms the determinations that were made when the

motion for preliminary injunction was denied: first, that the FDCA does not speak directly to the

precise question of whether the FDA may require the submission of comparative

11
immunogenicity data as part of an ANDA; and second, that the FDA’s interpretation of the

FDCA to permit it to require such data was reasonable. Sanofi-Aventis, 733 F. Supp. 2d at 168–

71. Since the statute does not plainly prohibit the agency from requesting the data as plaintiff

suggests, plaintiff’s motion for summary judgment on Count I on Chevron I grounds must be

denied. Rather, the statute is sufficiently broad such that the agency is authorized to make its

own judgment about what kinds of data fall within the broad categories of information it is

statutorily permitted to require and what kinds of data it needs to make the expert assessment it is

statutorily entrusted to make. Accordingly, granting the agency the Chevron II deference to

which it is therefore entitled, the Court finds that the request for the test results was reasonable,

and it will enter judgment for the defendants on Count I.

B. Chevron Step I

Sanofi argues that the FDCA expressly provides that the FDA may not require ANDA

applicants to provide any information beyond the eight categories of information listed by

Congress in section 355(j)(2)(A)(i)–(viii), and that immunogenicity testing is simply not

included in those categories. 9 Pl.’s Mem. in Supp. of Mot. for Summ. J. (“Pl.’s Mem”) at 17

[Dkt. # 38]. Therefore, it submits that the agency lacked the authority to require the test results

9 The notion that section (j)(2)(A) should be read strictly to deprive the agency of the
authority to call for the tests is plaintiff’s core contention: counsel directed the Court’s attention
to the provision repeatedly during the hearing and even cited it as the grounds for why the
agency’s interpretation was flawed under Chevron step II. See Tr. at 4, 5, 12, 15, 17–19, 26–27,
30, 57, and 84.

12
as part of its consideration of the application, and it should have denied the application – or

assessed it under section 355(b)(2) – instead. 10

FDA and Sandoz respond by pointing out that one of the categories listed in section

(j)(2)(A) for ANDA applications – section (j)(2)(A)(vi) – specifically incorporates provisions

from the list set forth in section 355(b)(1) for new drug applications, including section (b)(1)(D).

FDA’s Mem. in Supp. of Cross Mot. for Summ. J. (“FDA’s Mem.”) at 5 [Dkt. # 40]; Sandoz’s

Mem. in Supp. of Cross Mot. for Summ. J. (“Sandoz’s Mem.”) at 11 [Dkt. # 43]. They locate

FDA’s authority to seek the comparative testing in that provision, which directs an applicant to

supply “a full description of the methods used in, and the facilities and controls used for, the

manufacture, processing, and packing of [the] drug.” FDA’s Mem. at 5, quoting 21 U.S.C.

§ 355(b)(1)(D).

Sanofi submits, fairly, that those words do not literally appear to encompass test results

comparing the potential adverse effects of a generic drug to the pioneer. Pl.’s Mem. at 19.

Therefore, the Court cannot enter judgment for the defendants without going beyond the

Chevron step I stage.

Defendants note that the statute requires FDA to approve an ANDA unless it determines

that “the methods used in, or the facilities and controls used for, the manufacturing, processing

and packing of the drug are inadequate to assure and preserve its identity, strength and purity.”

21 U.S.C. § 355(j)(4)(A). In other words, Congress required the agency to assess purity, and the

things the agency may demand be fully described under section (b)(1)(D) are the very things the

agency must deem to be adequate to ensure the purity of the drug. So, the defendants maintain

10 Of course, as defendant Sandoz points out, denial of the ANDA would have extended
Sanofi’s seventeen–year monopoly in the market for enoxaparin. Sandoz’s Mem. in Supp. of
Cross Mot. for Summ. J. (“Sandoz’s Mem.”) at 3 [Dkt. # 43]. But it is that economic interest in
the RLD that gives Sanofi standing to complain in this case.
13
that the agency is authorized to interpret the requirement of a “full description” of the methods

and controls called for by section 355(b)(1)(D) to encompass the information it needs to make

the findings required by section 355(j)(4)(A) – and indeed, that the words “full description” must

be read as a means to accomplish that purpose. See, e.g., FDA Cross Mot. for Summ. J. at 5–6.

For the reasons to be set forth in more detail below, the Court agrees. Through the

ANDA pathway’s specific embrace of the NDA requirements, and the imposition of the clear

demands in section 355(j)(4)(A), Congress rendered the ANDA requirements to be ambiguous

and open to agency interpretation, and not as restrictive as the plaintiffs describe them to be. By

specifically incorporating section 355(b)(1)(D) into the ANDA requirements, Congress gave

FDA the authority to utilize its expertise to determine what information it needs to make the

assessment it is required to make under section 355(j)(4)(A).

1. Circuit precedent suggests that the statute is ambiguous, and that the agency
has been entrusted with its interpretation.

The Court’s conclusion is supported by guidance provided in Serono Laboratories, Inc. v.

Shalala, 158 F.3d at 1324–25, where the Court of Appeals indicated that the clauses enumerating

what the FDA may review in an ANDA should be construed broadly. In Serono, a

pharmaceutical company filed an ANDA for a generic version of Serono’s drug Pergonal, and

Serono opposed it with a Citizen Petition. Id. at 1316. The FDA questioned whether the

concentration of a certain inactive ingredient in the generic drug raised safety concerns, and in

making the ultimate decision that it did not, the agency reviewed three animal studies that the

ANDA applicant had submitted as part of its application. Id. at 1323–4. As in this case, the

manufacturer of the pioneer drug objected to the consideration of the test results. Id. at 1324.

Among other questions in the case, then, the court was asked to address whether the FDA had

the statutory authority to consider animal studies submitted as part of an abbreviated application.

14
 The Court of Appeals observed:

The only provision of the Act to which Serono points for support of its no-animal-
studies proposition is one that states the FDA “may not require that an
abbreviated application contain information in addition to that required by clauses
(i) through (viii)” of 21 U.S.C. § 355(j)(2)(A). Because nothing in those clauses
mentions animal studies, Serono contends they are barred. This provision,
however, does not bear the weight Serono applies.

Id. at 1324 (internal citations omitted). The same principle applies here.

Serono argued that section 355(j)(2)(A) of the FDCA barred FDA from considering

animal studies because they did not fall within the of types of information enumerated in sections

355(j)(2)(A)(i)–(viii) that the agency is permitted to require. Id. Ultimately, the court did not

reach the issue because it based its holding on a circumstance not present in this case: it ruled

that even if the provision did prohibit the FDA from requiring an applicant to submit animal

studies, “[i]t does not bar an applicant from voluntarily submitting additional information –

including animal studies – as part of its ANDA.” 11 Id. at 1324. But the Court went on to

observe that the interpretation being advanced by the NDA holder was too restrictive:

[T]he indicated clauses do not suggest that animal studies are in any way
disfavored. The clauses simply describe what the “information” in an application
must “show.” They do not specify the kinds of studies that can or cannot be used
to satisfy the requirement.

Id. The Court then cited one of the categories in section 355(j)(2)(A) – “An abbreviated

application for a new drug shall contain . . . information to show that the active ingredients of the

new drug are the same as those of the listed drug” – as an example of one of the clauses that

11 Sandoz argues that FDA did not actually “require” it to submit the immunogenicity
information. Sandoz Opp. at 14 [Dkt. No. #20]. While it is true that FDA purported to suggest
the kinds of information that would address their three concerns without telling Sandoz what
exactly to file, the Court will proceed on the premise that the information was actually required
because FDA plainly refused to accept Sandoz’s ANDA without the information. AR 4170.
15
identified a broad category of required information but did not specify how it was to be fulfilled.

Id.

While these observations may not have been necessary to the ruling in Serono, they

express a clear view that section 355(j)(2)(A) does not limit the agency’s freedom to determine

what kinds of information will be needed to fulfill the listed ANDA requirements.

Sanofi attempts to distinguish Serono on the grounds that the tests FDA required here

were justified under section 355(b)(1)(D)’s call for a “full description” of manufacturing

processes and controls, incorporated into the ANDA requirements in section 355(j)(2)(A)(vi),

and not under the clauses contained in sections 355(j)(2)(A)(i)–(v), which contain the language

referenced by the Court that authorizes FDA to require “information to show” certain

characteristics of the generic drug. Pl.’s Reply to Opp. to Mot. for TRO and PI (“Pl.’s Reply”) at

8–12 [Dkt. # 21]. But this Court does not find the concept expressed in Serono to be tied to a

parsing of the words “information” or “show” in particular, or to be limited to the first five

categories set out in section 355(j)(2)(A) rather than all eight. The Court of Appeals simply cited

one of the categories listed in section 355(j)(2)(A) as an example of the ANDA statutory

requirements. Id. (“See, e.g., id. § 355(j)(2)(A)(ii)(II)”)(emphasis added).

The Serono court indicated that it was considering the larger question of whether the

(j)(2)(A) prohibition against requiring “information in addition to that required by clauses (i)

through (viii)” should be narrowly construed to bar particular forms of information not

specifically mentioned in the eight categories. See id. at 1324. It answered the question by

drawing a distinction between the broad requirement set forth in each section and “the kinds of

studies that can or cannot be used to satisfy the requirement.” Id. That differentiation applies

equally to the “full description” of processes and controls called for by the incorporation of

16
section 355(b)(1)(D) into section 355(j)(2)(A)(vi). Like the section of the ANDA provisions

analyzed in Serono, the section at issue here, section 355(b)(1)(D), “do[es] not specify the kinds

of studies that can or cannot be used to satisfy the requirement.” Id. And the terms “full

description” are sufficiently broad to warrant the same treatment as “information to show.”

2. An analysis of the text in light of the entire statutory scheme suggests that the
statute is ambiguous.

On that point, it is important to remember that the first step of the Chevron analysis

requires the Court to look not only at the words in question, but at the entire statute. And the text

at issue here – the statutory requirement that an ANDA applicant must supply a “full description

of the methods used in, and the facilities and controls used for, the manufacture, processing, and

packing of [the] drug” – comes straight from the new drug requirements. 21 U.S.C.

§ 355(b)(1)(D). This means there was a deliberate legislative choice to import some of the new

drug pathway requirements into the ANDA pathway verbatim. There is no dispute that in the

context of a new drug application, the statutory requirements are to be construed broadly. See

Tr. at 11. And, as counsel for Sanofi agreed at the hearing, there is no basis to construe the

words differently when they are incorporated into the list of ANDA requirements. Id; 12 see

Ratzlaf v. United States, 510 U.S. 135, 143 (1994) (“A term appearing in several places in a

statutory text is generally read the same way each time it appears.”).

Furthermore, an analysis of the provisions in question in the context of the entire statute

requires that section 355(j)(2)(A) be viewed in light of section 355(j)(4)(A), which specifies that

12 “The Court: My question is simply that the words – you said this is a statutory argument,
they’ve said this is a statutory argument – the words mean the same thing whether we’re talking
about a new drug or an ANDA, right?
A: Yes, they do.
Q: And in the case of a new drug, you have to read them broadly, don’t you?
A: Yes, your Honor.”

17
“the Secretary shall approve an application for a drug unless the Secretary finds the methods

used in, or the facilities and controls used for, the manufacture, processing, and packing of the

drug are inadequate to assure and preserve its identity, strength, quality, and purity.” 13 Through

this section, Congress has directed the FDA to satisfy itself that the processes utilized by the

ANDA applicant will “assure” quality and purity, but none of the categories of information the

agency may require that are listed in section 355(j)(2)(A) expressly provide for the submission of

any information demonstrating quality or purity. This apparent contradiction is another source of

the ambiguity that propels the Court from step one of the Chevron analysis to step two.

Moreover, the fact that section 355(j)(4)(A) parallels section 355(b)(1)(D) suggests that

Congress intended them to be read together. See United Savings Ass’n of Tex. v. Timbers of

Inwood Forest Assocs., 484 U.S. 365, 371 (1988) (“A provision that may seem ambiguous in

isolation is often clarified by the remainder of the statutory scheme – because the same

terminology is used elsewhere in a context that makes its meaning clear, or because only one of

the permissible meanings produces a substantive effect that is compatible with the rest of the

law.”) (citations omitted). While Sanofi is correct that section 355(j)(4)(A) does not expand the

scope of what FDA may require in an ANDA, it does shed light on why Congress wanted FDA

to look at the methods used in, and the facilities and controls used for, the manufacture,

processing, and packing of a generic drug: in part, to ensure the drug’s quality and purity.

13 Sanofi mischaracterizes this language, arguing that “[b]y its plain language, this Section
simply says that FDA must reject an ANDA application if it finds that the limited information
required by section [355](b)(1)(D) is ‘inadequate’ to assure the product’s purity.” Pl.’s Mem. at
28; Pl.’s Reply at 9. The Court notes that this section does not require rejection if the
information submitted falls short; it requires approval unless, given all the information
submitted, FDA makes a determination that the methods, facilities, and controls themselves are
inadequate to assure purity.

18
 Sanofi does not disagree with that much of the analysis. 14 See Tr. at 7, 16. But it insists

that section 355(b)(1)(D) should be interpreted strictly to mean that an applicant can only be

asked to “describe” its manufacturing methods and controls for that purpose, and not to report on

tests that would reveal the effect of those methods and controls on the purity of the product. Pl.’s

Mem. at 26.

But reading the provision in light of the statutory scheme as a whole militates against this

approach. First of all, the words “full description” themselves do not support plaintiff’s rigid

position, and the fact that the ANDA requirement is lifted from the NDA requirements compels a

broad reading. Finally, the fact that Congress described the agency’s task in 355(j)(4)(A)

utilizing the same words found in 355(b)(1)(D) lends meaning to the imprecise words used in

that section and indicates that now that they have been transplanted into the ANDA

requirements, they must be interpreted as a means to facilitate, and not frustrate, the statutorily

mandated evaluation of the purity of a proposed generic drug.

14 The Court: My question to you is, we’ve agreed that the FDA has to think about purity?
Counsel for plaintiff: Yes, your Honor.
Q: So which of the eight does it fall under, which of the things it can require is
going to give it that information?
A: We are not in disagreement over this point. We believe Subsection vi, which
incorporates (b)(1)(D) is the section in which Congress delineated what FDA is
permitted to require and so we think that is the core provision that this Court is
called upon to interpret.

Tr. at 7.
* * *
Q: Well, doesn’t (j)(4)(A) and the fact that it tracks the language from
355(b)(1)(D) suggest that if there’s any purpose to be served by the full
description requirement in (b)(1)(D) at all, its purpose is to illuminate issues such
as quality and purity?
A: Yes. That is the purpose of this information so that FDA can make the
determination . . . .

Tr. at 16.

19
 Sanofi argues that FDA’s request for the immunogenicity information was improper

because it went beyond a request to compare the impurity content of Sandoz’s generic drug with

that of Lovenox, but asked Sandoz to assess whether any difference in impurities would increase

the likelihood of adverse consequences and thus be harmful to consumers. Pl.’s Mem. at 53–54.

It is true that FDA required some of the studies in order to show whether the difference in the

impurity profiles of Sandoz’s generic drug and Lovenox made the generic more likely to cause

immune responses than Lovenox. 15 See AR 4170–73, 4433. But FDA did not call for the sort of

safety and effectiveness tests that are part of an NDA and excluded from the ANDA process; the

tests were expressly requested to answer questions about the purity of the product. See AR 4170.

And, as counsel for Sandoz pointed out at the motions hearing, if one reads section 355(j)(4)(A)

within the context of the ANDA amendments as a whole, it is clear that the (j)(4) assessment

necessarily involves a comparison of the generic to the listed drug.

Counsel for Sandoz: . . . I guess what I would say is, first of all if you look at
(j)(4)(A) and it says that FDA has to make findings with respect to manufacturing,
processing, packaging, methods and controls adequate, among other things, to
assure purity. Now, (j)(4)(A) doesn’t say the words “in comparison to the brand,”
but I think given the whole context of the Hatch-Waxman provisions, it has to be
implicit. And that’s the whole nature of this.

Q: Otherwise, what difference would purity make?

A: That’s right. It’s only in comparison to the brand. If the brand has a certain
level of impurity and you are at or below that level, you’re okay. If you come in
with a product that’s otherwise the same but your impurity levels are ten times
higher than the brand, well that’s an inquiry that the FDA has to make, and you’re
not going to get approved under that scenario as a true generic under (j).

Tr. at 77–78.

15 At the hearing, defendants asserted that there is no assertion by any party that the generic
drug is in fact more harmful or impure than the pioneer. The issue here is simply about process.
Tr. at 79–81, 85–86. Sanofi did not contest that assertion. Id.
20
 A close reading of the statute suggests that the argument can be put more strongly, and

that the comparative nature of the inquiry is not merely implicit, but expressed, in the ANDA

provisions. Congress directed FDA to determine whether the generic manufacturer’s processes

and controls were adequate not only to “assure,” but to “preserve” the drug’s “identity, strength,

quality, and purity.” 21 U.S.C. § 355(j)(4)(A). Since section 355(j)(4) is talking about the

approval of generic copies of listed drugs, the purity to be “preserved” must be the purity of the

original. So a comparison of the adverse effects caused by impurities is warranted. And the

directive that the FDA be confident that the processes are adequate to assure not only the quality,

but the “identity” of the generic also indicates that the (j)(4)(A) assessment – to be based on the

(b)(1)(D) “full description” – is a comparative one. Under those circumstances, the statute did

not prohibit the solicitation of comparative tests.

3. An analysis of the text in light of the statute’s purpose suggests that the statute
is ambiguous.

The Chevron step I exercise also involves a consideration of the provisions at issue in

light of the statute’s purpose. See Mova Pharm. Corp. v. Shalala, 140 F.3d 1060, 1067–68 (D.C.

Cir. 1998), quoting Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 51 (1987) (“[I]n expounding a

statute, we must not be guided by a single sentence or member of a sentence, but look to the

provisions of the whole law, and to its object and policy.”). Sanofi cited Mova Pharmaceutical,

id., and Schering Corp. v. FDA, 51 F.3d 390 (3d Cir. 1995), and maintained that “these statutes

are entry-restricting statutes.” Tr. at 28. But the legislative history and the text of the statute

point to the opposite conclusion, and the precedents Sanofi cites do not address the situation

here.

While certain terms of the Hatch-Waxman amendments may have been the result of a

legislative compromise as plaintiff suggested, see Tr. at 51, and they may reflect the balancing of

21
the interests of manufacturers of listed drugs, would-be marketers of generics, and consumers,

the clear purpose of the amendments is set out in the very first sentence of the House Report:

“The purpose of Title I of the bill is to make available more low cost generic drugs by

establishing a generic drug approval procedure . . . .” H. REP. NO. 98-857(1), pt. 1, at 1 (1984).

The instruction in section 355(j)(2) that FDA may call for no categories of information beyond

those enumerated in the statute must be read as a means to fulfill this purpose – to keep the

agency from delaying or impeding the ANDA approval process by placing additional demands

on the applicants. This reading is borne out by the language of section 355(j)(4), which

embodies a congressional preference in favor of ANDA approval.

Neither Schering Corp. nor Mova Pharmaceutical compels a different conclusion. Both

cases raise the market entry concept only in the context of a ruling on the pioneer manufacturer's

standing, and both found particular provisions – not the statute as a whole or the provisions at

issue here – to be barriers to market entry. In Schering Corp., the court found that the

bioequivalence requirement contained in section 355(j)(7)(B) is meant to restrict market entry,

and therefore, the plaintiff, whose economic position would be injured by the approval of a

generic competitor, had standing to challenge the regulations implementing that provision. 51

F.3d at 396. And in Mova Pharamaceutical, the Court of Appeals was looking only at section

355(j)(5)(B)(iv), the provision that accords priority among successive ANDA applicants. It

simply noted that the statutory provision that regulates the timing of generic drug manufacturers'

entry into the marketplace would also have the effect of freeing the pioneer drug manufacturer

from competition as well, so the pioneer company had grounds to intervene in the action. 140

F.3d at 1076–77.

22
 In sum, considering section 355(j)(2)(A) in the context of the entire statutory scheme and

the statute’s purpose, the Court finds the NDA provision that is included in section (j)(2)(A) –

section (b)(1)(D) – to be ambiguous for Chevron step I purposes. Given the ambiguity in the

statute and this Circuit’s direction that courts should construe the clauses in section 355(j)(2)(A)

broadly, this Court cannot hold that the FRCA unambiguously precludes FDA from requiring

immunogenicity data in an ANDA.

C. Chevron Step II

At Chevron step II, the Court must ask whether the FDA’s interpretation of section

355(b)(1)(d) is “based on a permissible construction of the statute,” Chevron, 467 U.S. at 843,

and consistent with the statute’s text and overall scheme, see Nat’l Ass’n of Home Builders v.

Defenders of Wildlife, 551 U.S. 644, 666 (2007). As noted above, the Court must defer to FDA’s

interpretation of the statute if its interpretation is reasonable.

Here, FDA sought tests that compared the immunogenicity of the generic to the parent,

and it specified that the tests were sought to alleviate its concerns about purity. AR 4170. In

Serono, the D.C. Circuit underscored that deference is particularly appropriate when FDA

approval of drugs is involved. 158 F.3d at 1324. It cited the holding in Schering Corp. that

“judgments as to what is required to ascertain the safety and efficacy of drugs fall squarely

within the ambit of the FDA’s expertise and merit deference from us.” Id., quoting Schering

Corp., 51 F.3d at 399. Since section 355(j)(4) expressly calls upon the agency to assess the

purity of a generic, see § 355(j)(4)(A), as well its safety and efficacy, see § 355(j)(4)(F),(H), the

Schering observation approved by this Circuit is equally applicable to a situation where the

agency made a judgment as to what was required to ascertain the purity of a drug. FDA’s close

and careful review of the scientific information – its approval of Sandoz’s application took

23
nearly five years – is further grounds for deference. See Tr. at 81; Bristol-Myers Squibb Co. v.

Shalala, 923 F. Supp. 212, 218–19 (D.D.C. 1996) (seven-year comprehensive review of

scientific testing merits deference).

According FDA the deference required under Serono to make its own determination

about the information it might need, the Court finds that FDA’s interpretation of the ANDA

approval regime was reasonable, and that it was reasonable for the agency to conclude that

immunogenicity studies are encompassed by the “full description” described in section

355(b)(1)(D). 16 The potential for the generic drug to elicit a different adverse response than the

parent could be the result of impurities, which in turn result from the methods, facilities, and

controls used to manufacture, process, and pack a drug. By revealing what impurities remain at

the end of that process, the studies shed light on, or indirectly “describe,” those methods and

controls.

FDA maintains that its call for test results was also fully consistent with its regulations.

The Court must defer to the agency’s reading of its own regulations unless it is “plainly

erroneous or inconsistent with the regulation,” Serono, 158 F.3d at 1320 (internal quotation

marks omitted), and in this case, the Court cannot make such a finding. The FDA regulations

implementing the ANDA approval provisions in the statute describe the information applicants

must provide to the agency to fulfill each of the requirements listed in section 355(j)(2)(A). See

21 C.F.R. § 314.94. For the “chemistry, manufacturing, and controls” called for by section

(j)(2)(A)’s invocation of section (b)(1)(D), though, the ANDA regulations require applicants to

submit “the information required under section 314.50(d)(1),” which is the regulation governing

16 Indeed, Sanofi said very little about why the agency’s interpretation would be
unreasonable other than to repeat its Chevron step I argument that the immunogenicity tests fell
outside the list of items FDA could require under section 355(j)(2)(A). Tr. at 19–20.

24
new drug applications. 21 C.F.R. § 314.94(a)(9)(i). And that regulation makes it clear that the

“full description” of the chemistry, manufacturing, and controls for a drug should include “for

example, tests . . . .” 21 C.F.R. § 314.50(d)(1)(i)–(ii). 17 So the Court agrees with the

determination made in connection with plaintiff’s motion for a preliminary injunction that FDA’s

interpretation of section 355(b)(1)(D) to include immunogenicity testing was reasonable.

D. The statute did not require FDA to abandon the ANDA pathway and invoke section
355(b)(2).

Finally, contrary to Sanofi’s contentions, the ambiguity in the ANDA provisions is not

clarified by the availability of section 355(b)(2), the quasi-third application pathway. Sanofi

argues that the agency was bound to switch tracks to this “hybrid” drug approval method when

the information submitted with Sandoz’s ANDA was found to be insufficient, but there is

nothing in the statute that compels this approach. Indeed, if one reads the statute strictly, as

Sanofi insists one should, there is nothing that would indicate that the (b)(2) pathway would have

been applicable in this case at all; it appears to be more of a subset of NDA applications than a

hybrid approach for a true generic.

Furthermore, as the Court noted in its opinion denying the motion for preliminary

injunction, FDA’s reading of the statute accords with its own regulations on when (b)(2) is to be

employed. Sanofi-Aventis, 733 F. Supp. 2d at 170–71. FDA’s approach is consistent with 21

C.F.R. § 314.54(a), which applies the 355(b) pathway for “[a]ny person seeking approval of a

drug product that represents a modification of a listed drug (e.g., a new indication or new dosage

form) and for which investigations, other than bioavailability or bioequivalence studies, are

essential to the approval of the changes[.]” 21 C.F.R. § 314.54 (emphasis added). Here, FDA

17 See Tr. at 63, where counsel for FDA noted, “I mean, drug applications, particularly the
CMC section, are practically nothing but test results.”

25
assessed Sandoz’s generic drug as a replication of a listed drug, not a modification. Since 21

C.F.R. § 314.54(a) applies only for drug products that are modifications of listed drugs, it was

not inconsistent with this regulation for FDA to decide to proceed under the (j) pathway rather

than the (b)(2) pathway here.

FDA’s determination of what constitutes a modification versus a replication of a listed

drug is a scientific determination within the agency’s area of expertise, and therefore is entitled

to heightened deference from this Court. A.L. Pharma, Inc. v. Shalala, 62 F.3d 1484, 1490 (D.C.

Cir. 1995) (“[C]ourts give a high level of deference to an agency’s evaluations of scientific data

within its area of expertise.”). While “there may well be more than one reasonable way to read”

the word modification, it is not unreasonable for FDA to determine that a mere variance in the

impurity profile of the drug is not a modification, and therefore, this Court is bound to uphold

that interpretation. See Serono, 158 F.3d at 1321.

Sanofi further claims that allowing FDA to require immunogenicity data as part of an

ANDA would render the (b)(2) pathway to be superfluous. Pl.’s Mem. at 30. But both the

relevant FDA regulation, 21 C.F.R. § 314.54(a), and FDA guidance document show that is to be

unlikely. As mentioned above, the section (b)(2) pathway is followed by new drug applicants

who seek to rely on research conducted by a third party without that party’s permission. FDA’s

regulations specify that this pathway covers “a drug product that represents a modification of a

listed drug (e.g., a new indication or new dosage form) and for which investigations, other than

bioavailability or bioequivalence studies, are essential to the approval of the changes[.]” 21

C.F.R. § 314.54. FDA’s guidance document, “Applications Covered by Sections [355](b)(2),”

clarifies that the kinds of modifications covered include changes in dosage form, strength, route

of administration, or substitution of an active ingredient. Guidance for Industry: Applications

26
Covered by Section 505(b)(2) at 4 (2009),

http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/u

cm079345.pdf. While this guidance document is in no way binding on this Court, it does show

that certain types of drugs undergo approval under section 355(b)(2) and will continue to

undergo approval under that section even if generic drugs like Sandoz’s enoxaparin are routed

through the ANDA pathway in section (j). Therefore, this section is not rendered superfluous by

FDA requiring immunogenicity data as part of an ANDA, and section 355(b)(2) does nothing to

clarify Congress’s intent as to section (j).

Since the statute is ambiguous as to whether FDA may require immunogenicity data and

FDA’s interpretation of the statute is reasonable, the Court finds that FDA did not exceed its

authority by requiring Sandoz to submit comparative immunogenicity data as part of its ANDA.

II. FDA did not depart from agency precedent by approving a generic drug that is
not fully characterized.

The next question at issue here – whether FDA departed from agency precedent by

approving a generic version of enoxaparin without it being fully characterized – can also be

decided on summary judgment because it is a pure question law. Sanofi asks the Court to

withdraw FDA’s approval of Sandoz’s generic enoxaparin because it approved the generic drug

before enoxaparin was fully characterized. Sanofi alleges that this action departed from FDA

precedent without reasoned explanation, in violation of the APA. Pl.’s Mem. at 31. In support

of this claim, Sanofi repeats the argument that it asserted in its motion for preliminary injunction:

that before FDA approved Sandoz’s generic enoxaparin, it had refused to approve three other

drugs – Premarin, Hyaluronidase, and Omnitrope – based on the fact that they had not yet been

fully characterized, and that FDA failed to adequately explain why it departed from that

precedent in approving Sandoz’s drug. Pl.’s Mem. at 31–33.

27
 Since Sanofi spends the majority of its summary judgment briefing arguing its first and

third claims and does not assert any new arguments in support of this second claim, the Court

finds no reason to diverge from the reasoning in the memorandum opinion denying the motion

for a preliminary injunction on this Count, and it specifically incorporates that analysis here.

Sanofi-Aventis, 733 F. Supp. 2d at 171–73.

The Court looks to whether the challenged agency decision was “arbitrary, capricious, an

abuse of discretion, or otherwise not in accordance with the law.” Sanofi-Aventis, 733 F. Supp.

2d at 171, quoting 5 U.S.C. § 706(2)(A), citing Motor Vehicle Mfrs. Ass’n v. State Farm Mut.

Auto Ins. Co., 463 U.S. 29, 43 (1983). This Court’s review is “highly deferential” because the

agency’s decision is based on the evaluation of complex scientific information within the

agency’s technical expertise. Id. at 171–72, quoting Bloch v. Powell, 348 F.3d 1060, 1070 (D.C.

Cir. 2003), citing Troy Corp. v. Browner, 120 F.3d 277, 283 (D.C. Cir. 1997). In making its

decision to approve Sandoz’s drug before it was fully characterized, FDA “provided ‘legitimate

reason[s]’ for deciding that enoxaparin should be treated differently than the drugs cited by

Sanofi” and therefore satisfied the minimal standard of rationality required. Id. at 172–73, citing

Bracco Diagnostics, Inc. v. Shalala, 963 F. Supp. 20, 27 (D.D.C. 2007). 18 As such, the Court

will deny plaintiff’s motion for summary judgment as to Count II and grant defendants’ cross

motion.

18 The Court is not certain that the agency’s handling of only three similar situations gives
rise to the sort of precedent from which a departure needs to be justified, but it does not reach
that question since the decision and the manner in which it diverged from previous decisions
were adequately explained in this instance. Sanofi indicated that it would have objected even if
there had only been one previous situation, so it appears that at bottom, its concern was more
with the merits of the decision than with the agency’s consistency in any event. See Tr. at 41.
28
 III. FDA sufficiently proved that Sandoz’s generic enoxaparin has the same active
ingredient as Lovenox.

The final issue here – whether FDA sufficiently proved that Sandoz’s generic enoxaparin

has the same active ingredient as Lovenox – can also be decided on summary judgment because

it is a pure question law. Sanofi asks the Court to reverse FDA’s approval of Sandoz’s generic

enoxaparin based on the way FDA determined active ingredient sameness.

As with Count II, the Court looks to whether FDA’s determination was “arbitrary,

capricious, an abuse of discretion, or otherwise not in accordance with the law.” 5 U.S.C. §

706(2)(A); see Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto Ins. Co., 463 U.S. 29, 41

(1983). The determination constitutes “an agency’s evaluation[] of scientific data within its area

of expertise” and is therefore entitled to heightened deference by this Court. A.L. Pharma, 62

F.3d at 1490. Furthermore, the D.C. Circuit has already held that the validity of FDA’s

interpretation of what makes a generic drug’s active ingredient “the same as” a listed drug’s

active ingredient is a Chevron step II inquiry that deserves such heightened deference. Serono,

158 F.3d at 1319–20. “[T]he statute does not unambiguously require the term ‘same as’ to be

defined as complete chemical identity.” Id. at 1320.

Sanofi picks at the third and fifth prongs of FDA’s five-pronged sameness test, but, as

FDA argues, this attempt to invalidate an individual criterion for failing to show active

ingredient sameness alone ignores FDA’s overarching five-pronged approach. FDA’s Mem. at

35. “Instead of relying solely on . . . any . . . single criterion, . . . FDA relied upon additional

overlapping evidence derived from all five criteria” to show sameness. Id.

In criterion three, FDA required Sandoz to utilize direct sequencing techniques to

compare the chemical makeup of the enoxaparin in Lovenox with the enoxaparin in Sandoz’s

drug. AR 2897. However, rather than requiring Sandoz to sequence and compare all of the

29
sugar chains (oligosaccharide chains) that make up enoxaparin – which would require

completely characterizing enoxaparin – FDA required it to sequence only a comparable subset of

oligosaccharide chains. Id. The subset FDA chose included short chains, but excluded the

longer chains. Id. FDA explained that these short sugar chains are “the result of the most

cleavage reactions of the heparin oligosaccharide chains” and are therefore the most sensitive to

variation in the process conditions used to make the drug. 19 Id. Therefore, a showing that this

subset of sugar chains from both the generic drug and Lovenox possess the same sequence

“provides further corroborative evidence that the generic drug product’s enoxaparin possesses

the same distribution of oligosaccharide sequences as Lovenox’s enoxaparin” and is therefore the

same. Id.

Clearly this factor alone is insufficient to show active ingredient sameness because it

does not show complete equivalence of the two active ingredients. But, FDA did not rely on

sequencing alone to determine active ingredient sameness; rather, it was one of five factors that

FDA considered together. FDA concluded that “[t]hese five criteria together comprise a robust

test that provides overlapping evidence by which an ANDA applicant for enoxaparin can

demonstrate active ingredient sameness for enoxaparin within the meaning of [21 U.S.C.

§ 355(j)] and FDA regulations.” AR 2880. Thus, by selecting only certain sugar chains for

Sandoz to sequence, FDA did not allow Sandoz to take a “short cut” to sequencing all of the

sugar chains, as Sanofi asserts, but FDA defined what was necessary to satisfy prong three of the

sameness test, given the information required under the other four prongs. See Pl.’s Mem. at 36.

19 Sanofi contests FDA’s assertion that the short sugar chains are the result of the most
cleavage reactions. Pl.’s Mem. at 39–41. However, given that FDA supports its assertion in the
record, see AR 2890, 2894–95, 2897, and given the high level of deference the Court must
accord FDA here, the Court cannot find that FDA’s conclusion is unreasonable, see 5 U.S.C. §
706(2)(A); see also Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto Ins. Co., 463 U.S. 29, 41
(1983).
30
 In addition, Sanofi cannot show that Sandoz failed to satisfy factor five of the sameness

test. In so arguing, Sanofi actually misstates factor five. According to Sanofi, that factor

required Sandoz to show the full equivalence of the in vivo pharmacodynamics profiles (i.e.,

equivalent action or effect in the body) between its generic drug and Lovenox. Id. at 42–43.

Sanofi claims that Sandoz failed to satisfy this because it showed equivalence only of its effect

on two factors (anti-Xa and anti-IIa), but not of its effect on a third (TFPI). Id.

However, nowhere in the administrative record does FDA claim that prong five requires a

generic enoxaparin manufacturer to show full equivalence of in vivo pharmacodynamics profiles

between its drug and Lovenox. Although in FDA’s response letter to Sanofi’s citizen petition,

criterion five is titled, “Equivalence of In Vivo Pharmacodynamics Profile,” the body of the

letter explains that “[t]he comparison of in vivo pharmacodynamics profiles is based upon

measurements of in vivo anti-Xa and anti-IIa profiles.” AR 2899.

Furthermore, given the high level of deference this Court must accord FDA’s

determination, Sanofi does not persuade the Court that it was unreasonable for FDA to focus

only on anti-Xa and anti IIa. Although it may be true that “enoxaparin’s effect on TFPI is a part

of its overall pharmacodynamics profile[,]” Pl.’s Mem. at 42, FDA chose anti-Xa and anti IIa as

the most important factors, AR 2899. FDA therefore, could reasonably have decided that if

Sandoz could show that the effect of its drug on the two most important factors was equivalent to

Lovenox, then it did not need to further show that the effect on a less important factor was

equivalent. This is particularly true given the “overlapping” nature of the five prongs in the

sameness test.

Thus, the Court’s analysis turns on whether the five-pronged approach itself was a

reasonable way for FDA to determine active ingredient sameness. Not only did FDA support its

31
approach in a thorough, well-reasoned response to Sanofi’s citizen petition, but it also carefully

considered both sides of the argument internally – to settle the internal dispute over the validity

of the five-pronged test – before doing so. AR 3836–48, 3853–61. “Of course, differing views

among an agency’s staff may indicate that there is more than one reasonable way to read a

statute. . . . But under Chevron, courts are bound to uphold an agency interpretation as long as it

is reasonable – regardless whether there may be other reasonable, or even more reasonable,

views.” Serono, 158 F.3d at 1321. While fully characterizing enoxaparin would have been

another reasonable, or perhaps even more reasonable, way to determine active ingredient

sameness, the Court is satisfied that the five-pronged approach FDA used was reasonable.

The Court is further convinced that the reason FDA required immunogenicity testing was

to determine whether the drug contained harmful impurities, not to settle a last minute worry that

the five criteria were insufficient to establish sameness. FDA has represented all along that it

sought the immunogenicity data in order to determine whether the generic drug contained

potentially harmful impurities. See AR 3849–50, 4193–94, 4433–34. And that is exactly what

the immunogenicity data Sandoz submitted actually told FDA. AR 4433–35. Perhaps it would

have also been rational for FDA to require immunogenicity data to show whether its generic

drug contained the same active ingredient as Lovenox. However, that is not why FDA required

the data here.

32
 CONCLUSION

Because FDA’s request for immunogenicity data in Sandoz’s ANDA was both lawful and

reasonable, its approval of the drug did not constitute an arbitrary departure from agency

precedent, and its determination of active ingredient sameness was not arbitrary and capricious,

an abuse of discretion, or otherwise not in accordance with the law, the Court will grant

defendants’ motions for summary judgment and deny plaintiff’s cross-motion.

AMY BERMAN JACKSON
United States District Judge

DATE: February 7, 2012

33