The opinion of the court was delivered by: Gerard E. Lynch, District Judge
The Hatch-Waxman Act, 21 U.S.C. § 355 and 35 U.S.C. § 271(e) (1994) (codified as amended), permits would-be manufacturers of generic versions of an already approved, patented drug to seek expedited approval from the Food and Drug Administration ("FDA") before expiration of the patent, by means of an Abbreviated New Drug Application ("ANDA"). See Yamanouchi Pharmaceutical Co., Ltd. v. Danbury Pharmacal, Inc., 231 F.3d 1339, 1342 (Fed. Cir. 2000). Filing an ANDA constitutes an "artificial" but legally cognizable instance of patent infringement, often triggering a lawsuit in which the validity and enforceability of the patent may be tested. Glaxo Group Ltd. v. Apotex, Inc., 376 F.3d 1339, 1344 (2004).
Dr. Reddy's Laboratories, Ltd., and Dr. Reddy's Laboratories, Inc.
(collectively, "Reddy"), and Teva Pharmaceuticals USA, Inc. ("Teva"),
the defendants in these actions, each filed an ANDA, seeking to
manufacture generic versions of a gastric-acid inhibitor that is
marketed under the brand name Aciphex. Plaintiffs Eisai Co., Ltd. and
Eisai Inc. (collectively, "Eisai"), hold the patent on rabeprazole
sodium, the active ingredient of Aciphex. Eisai duly brought these
actions for patent infringement against defendants.*fn1
Defendants argue that Eisai's patent should not be enforced
because Eisai engaged in inequitable conduct before the U.S. Patent
and Trademark Office ("PTO") during prosecution of its patent
application. Eisai now moves for summary judgment on all defendants'
allegations of inequitable conduct. For the
reasons below, the motion is granted in part, but for the most part
To address the parties' contentions requires reciting in some detail the history of the prosecutions of both the rabeprazole patent and another Eisai patent application that defendants argue is closely related. The following facts regarding those prosecutions are undisputed except as otherwise noted. Given the scope and somewhat specialized nature of the record, however, citations are provided to identify the source of certain facts. Where citations point to a statement in a party's brief or other non-evidentiary submission, that statement is generally uncontested and usually itself points to evidentiary material in the record.
I. Prosecution of The Patent-in-Suit
Eisai, a pharmaceutical company, owns U.S. Patent No. 5,045,552 ("'552 patent"), the patent-in-suit. The '552 patent claims, among other things, the chemical compound rabeprazole and its salts.*fn3 Rabeprazole sodium is the active ingredient in a drug that Eisai developed and made commercially available under the brand name Aciphex. The drug was approved by the FDA in 1991 for treatment of certain gastric ailments including duodenal ulcers and heartburn. Rabeprazole functions as a "proton pump inhibitor," because it suppresses gastric-acid production by inhibiting the action of an enzyme, H⯠ጽ, which pumps into the stomach protons (H ions) that combine with chloride ions to form hydrochloric (gastric) acid. (Ds. Joint R. 56.1 Stmt. ¶ 15.) The market for Aciphex is a lucrative one: Eisai's worldwide sales of the drug have been reported at over $1 billion per year. (Ds. Joint R. 56.1 Stmt. ¶¶ 1-5.)
On November 10, 1987, Eisai's attorney Arthur R. Crawford filed the
application that resulted in the '552 patent.*fn4
('552 File History, DRLRAB 51.)*fn5 The application
reported that the compound now referred to as rabeprazole*fn6
belonged to a known class whose chemical structures include a
benzimidazole ring on the left side of the molecule as diagramed in
the standard chemical notation, and a pyridine ring on the right,
joined by a sulfinylmethyl group.
(P. Ex. 1, '552 patent, col. 1, lines 40-44.) Numerous compounds
feature this basic chemical structure, including prominently
omeprazole, the first commercially available proton pump inhibitor;
omeprazole was disclosed in a group of patents (known as "Junggren"
inventor) owned by the Swedish company now known as AstraZeneca AB. It
is the active ingredient in the drug marketed as Prilosec.*fn7
(Ds. Joint R. 56.1 Stmt ¶ 19; '552 Patent File History,
DRLRAB 54, 431; see also Teva Mem. in Opp. to P. Mot. for Summ. J. of
Patent Validity, 1-2.) The compounds are formally distinguishable
primarily by the particular "substitutions" of chemical groups for
hydrogen atoms around their pyridine rings. The structure of
rabeprazole's pyridine ring reflects a pattern of substitution
referred to in this litigation as "asymmetrical," because its
3-position is substituted (with a methyl group) while the 5-position
is unsubstituted (that is, it is bonded to a hydrogen atom); the
4-position is substituted with a methoxypropoxy group
(OCH2CH2CH2OCH3), a type of alkoxy group.*fn8 (P. Mem., glossary at
7.) Other related compounds, some of which are potentially useful in
the inhibition of gastric-acid formation, have different patterns of
pyridine-ring substitution, including the use of different alkoxy
groups at the 4-position. The benzimidazole ring of rabeprazole is
Eisai reported having synthesized rabeprazole by working from omeprazole. (P. R. 56.1 Stmt. ¶¶ 18, 24.) Eisai disclosed summaries of three sets of pharmacological data comparing, relevantly, omeprazole to rabeprazole. (Ds. Joint R. 56.1 Stmt. ¶ 37.) These disclosures will be further discussed in relation to defendants' inequitable conduct claims, but, in brief, they purported to demonstrate rabeprazole's superior potency and its enabling of faster post-dosage recovery of acid secretion as compared to omeprazole.*fn9 (P. R. 56.1 Stmt. ¶¶ 19, 21, 23-27, 36-39 (citing '552 patent)). The submitted data were not the only existing pharmacological comparisons involving rabeprazole, as will further be discussed in relation to the inequitable conduct claims.
A. First Rejection and Eisai's Response Patent examiner Jane Fan rejected the rabeprazole claims three times, prompting various persuasive efforts by Eisai, before ultimately allowing the claims to issue as the '552 patent. On September 21, 1988, Fan rejected the claims as obvious in light of certain prior art: the Junggren patents and Great Britain Patent No. 2,234,523 ("GB '523"). (Ds. Joint R. 56.1 Stmt. ¶¶ 41-42.) "[Junggren] and [GB '523] generically teach R4 [the 4-position on the pyridine ring] being methoxyethoxy or ethoxyethoxy," wrote Fan.*fn10 ('552 File History at DRLRAB 292.) Fan specifically cited and diagramed a prior art-compound, Example 27 of Junggren, which bears a methoxyethoxysubstituent at the 4-position of its pyridine ring. ('552 File History, DRLRAB 292-93.) Fan concluded that the "[prior] art compounds are homologs of the claimed compounds rendering the claimed compounds unpatentable." (Id. at 293.) The rejection of the rabeprazole claims was not based on lack of novelty. (Ds. Joint R. 56.1 Stmt. ¶ 42.)
Eisai reacted to this first rejection in various ways. On March 21, 1989, Eisai attorney Crawford offered a response that addressed both the structural and functional distinctiveness of rabeprazole. With respect to chemical structure, the response distinguished rabeprazole's 4-position substituent, methoxypropoxy, from those of compounds in the prior art. Eisai submitted that Junggren and GB '523 disclosed only compounds bearing a methoxyethoxy group at the 4-position of the pyridine ring and "thus novelty is established" -- presumably, although not explicitly, contrasting rabeprazole's methoxypropoxy substituent. ('552 Patent File History, DRLRAB 429.) Further, Eisai pointed out that the "[s]pecific compounds disclosed in [GB '523] and [Junggren] are substituted at both the 3- and 5-position by methyl groups, as in the GB ['523] patent, or unsubstituted in both the 3- and 5-positions." (Id. at DRLRAB 429, emphasis in original.) The response continued, "Applicants' claims allow for the possibility of unsubstitution or a lower alkyl at the 3-position with no substitution at the 5-position; preferably, the 3-position . . . is methyl" -- describing the asymmetrical pyridine-ring substitution pattern that is found in rabeprazole. (Id.)
In addition to discussing structural traits, Eisai's March 21 response also addressed the pharmacological properties of the claimed compounds. Eisai professed that compounds "in which the substituent at the 4-position is propoxymethoxy"*fn11 exhibited "unexpected anti-ulcer activity." ('552 Patent File History, DRLRAB 429.) It also submitted additional pharmacological comparisons with omeprazole, alleging omeprazole's overall inferiority with respect to acid-inhibition and post-dosage recovery. (Id. at DRLRAB 541.)*fn12 The response
sought to justify Eisai's choice of omeprazole as a comparator:
As . . . the [prior] art shows a preference for a specific compound currently under development and known as Omeprazole, . . . applicants have compared their elected compound with Omeoprazole [sic] and not compounds more nearly structurally related to the claimed compounds . . . . This is appropriate in accordance with In re Fouche, [439 F.2d 1237 (C.C.P.A. 1971),] where the Court accepted a showing of unexpected advantage over a structurally more remote compound because the prior art clearly showed a preference for this structurally more remote compound. ('552 Patent File History, DRLRAB 430.)
In a May 4, 1989, "Supplemental Response," Eisai offered still further omeprazole-related arguments in favor of its claimed compounds' superiority, again explaining its choice of omeprazole as comparator: "The prior publications and patents describe a host of substituted benzimidazole compounds having antisecretory activity . . . . Among these compounds, the single most interesting compound is known as omeprazol[e]." ('552 Patent File History, DRLRAB 437-38.) The response presented certain test comparisons as showing the superior acid-inhibition properties of Eisai's claimed compounds. (Id. at 438-39.)
Also in response to the first rejection, Eisai submitted a copy of a prior-art application to European Patent No. 167,943 (referred to as "Beecham") which it had mentioned in the '552 patent application. ('552 Patent File History, DRLRAB 354-418.) The Beecham application discloses the asymmetric pyridine-ring substitution pattern of rabeprazole (substitutions at the 3-and 4- positions, with none at the 5-position), but does not disclose a methoxypropoxy or methoxyethoxy group at the 4-position. (See id.)
B. Second Rejection and Eisai's Response
On July 14, 1989, unsatisfied by Eisai's further submissions, Fan issued her second rejection of the rabeprazole claims based on the same prior art references. Referring to Eisai's March 21 submission asserting the distinctiveness of rabeprazole's asymmetrical substitution pattern, she expressed incomprehension, writing, "The claimed compounds encompass both 3.5 positions [3- and 5-positions] substituted by lower alkyl, thus the statement at middle part of page 9 of applicants' remark is not understood." ('552 Patent File History, DRLRAB 441.) Additionally, Fan rejected Eisai's reasoning in submitting comparison data for omeprazole rather than for a structurally closer prior-art compound: "The rationale of In re Fouche does not apply herein since isomeric difference [from the prior art] is noted therein whereas in the instant case homologous difference exists. Furthermore, a single compound is claimed in [F]ouche whereas a great many compounds are claimed in the instant case." (Id.) She continued, "In the unpredictable medicinal field, one can not assume a compound's pharmaceutical activity. Actual comparison has to be made in order to establish unexpected property."*fn13 (Id.)
Crawford spoke with Fan on August 3, 1989, after she rejected the rabeprazole claims for the second time. (P. R. 56.1 Stmt. ¶ 56.) Fan's record of the discussion states, in pertinent part, "If applicants are willing to limit the invention to [rabeprazole], then only methoxyethoxy cpd need [sic] to be compared."*fn14 ('552 File History, DRLRAB 449.) On December 28, 1989, Crawford filed a continuation of the rabeprazole application, to avoid expiration of the claims.
C. Third Rejection and Eisai's Response
On July 17, 1990, Fan issued her third rejection of the rabeprazole application. She reiterated reasons from prior rejections, including that the claimed compounds were obvious in light of prior art, that Eisai's argument concerning asymmetric substitution was "not understood," that Eisai's rationale for comparing only omeprazole was not persuasive, and that "[a]ctual comparison has to be made in order to establish unexpected property." ('552 Patent File History, DRLRAB 455-56.) This time, however, she added, "The closest prior art compounds should be compared with the claimed compounds." ('552 File History, DRLRAB 455.) "In the instant case," she continued, certain prior-art compounds "should all be compared," identifying in particular Example 27 of Junggren and "the compounds of EP 74341," a "newly cited" prior art application that Fan referred to in her third rejection as "Carlsson." (Id. at 455, 457.)
Following Fan's third rejection of the '552 patent application, Eisai both narrowed its claims and attempted to strengthen its substantive case. On January 11, 1991, Eisai submitted an amendment, limiting its claims to only those relating to rabeprazole. ('552 Patent File History, DRLRAB 464-66.) On the same date Eisai submitted a Rule 132 declaration*fn15 -- known in this litigation as the "Fujisaki Declaration," as its contents are attested by Eisai inventor Hideaki Fujisaki -- purporting to show rabeprazole's superior performance in an in vitro acid-inhibition test comparing three compounds, Example 27 of Junggren and two others. ('552 Patent File History, DRLRAB 468-471.) "I compared the . . . four compounds for their ability to inhibit H⯠ጽ, an assessment of the compound's ability to inhibit the secretion of gastric acid," Fujisaki reported. (Id. at 469.)
None of the three comparators (compounds 2, 3, and 4) shares rabeprazole's (compound 1) asymmetrical pyridine-ring substitution pattern, instead exhibiting substitution or no substitution at both their 3- and 5-positions. (Id. at DRLRAB 470.) But in an accompanying submission by Crawford, Eisai did not mention the asymmetry feature as a factor distinguishing rabeprazole from the comparators, emphasizing instead rabeprazole's particular 4-position substituent: "[T]he compound of the invention having a methoxy-propoxy at the 4' position of the pyridine ring . . . exhibits surprisingly unexpected [acid-]inhibitory effects . . . in comparison with closely related compounds of the type referred to by the examiner. Here attention is invited to item 5, page 3 of the examiner's letter where applicants have indeed compared the closest sulphonyl compound ethoxy methoxy with the claimed compound." (Id. at DRLRAB 466.) That specified item, Junggren Example 27, like the other two comparators referred to in the Fujisaki Declaration, bears a methoxyethoxy, instead of rabeprazole's methoxypropoxy, at the 4-position of its pyridine ring. (Id. at 470.)
Soon after, on April 3, 1991, Fan allowed the rabeprazole application to issue as patent '552. ('552 Patent File History, DRLRAB 473.) The patent's file history does not reveal why Fan ultimately decided that Eisai should receive the rabeprazole patent.
II. Prosecution of the Co-Pending Prosecution
On June 16, 1988,
approximately seven months after filing the application for the
'552 patent, Eisai via attorney Crawford filed the application for U.S.
Patent No. 5,708,013 ("'013 patent"). ('013 Patent File History,
DRLRAB 2281.) The '013 patent application listed the same 14 inventors
as did the '552 patent application (P. Ex. 1, '552 Patent; P. Ex. 2,
'013 Patent), and claimed inter alia a compound similar to rabeprazole
in structure, prior art, and
asserted properties.*fn16 ('013 Patent File History,
DRLRAB 2282.) Indeed, Eisai had written into its rabeprazole
application a limitation purporting to exclude certain similar
compounds -- including the compound covered by the '013 application --
from being considered as part of the '552 application.*fn17
('552 Patent File History, DRLRAB 230-34.) Eisai prosecuted
the two patent applications separately before different PTO examiners
and, during some three years' overlap in the pendency of the two
prosecutions did not disclose in either one the existence of, or any
occurrences during, the other.*fn18 (See '552 Patent
File History and '013 Patent File History for identities of different
examiners and absence of disclosure.) There is no evidence that Fan
knew of the '013 application, or that the examiners of that
application knew of the '552 application.
The structure of the compound claimed in the '013 application differs from rabeprazole's by a single methylene unit (CH2) at the 4-position of the pyridine ring. It there bears a methoxyethoxy (OCH2CH2OCH3) substituent, where rabeprazole bears a methoxypropoxy (OCH2CH2CH2OCH3) substituent. The two compounds are otherwise structurally identical. (P. Mem., glossary at 7-8.) The similarity but for one methylene unit renders the two compounds homologous. Like defendants, the Court will refer to the compound claimed by the '013 application as the "ethyl homolog" of rabeprazole, after the ethyl segment that differs from rabeprazole's propyl segment.*fn19 (Teva R. 56.1 Stmt. ¶ 24.)
Eisai pursued both rabeprazole and the ethyl homolog for their gastric-acid inhibiting properties. (Teva R. 56.1 Stmt. ¶¶ 32-35, 49, for citations to contemporaneous acid-inhibition tests of both compounds by Eisai and to portions of the patent file histories of both compounds making similar assertions about their inhibitory properties). Eisai also made several identical, noteworthy representations in both prosecutions, for instance identifying Junggren and especially Junggren's omeprazole compound as prior art and reporting that both homologs showed "unexpectedly . . . potent inhibitory activity." (Id. at ¶¶ 48-49, for citations to comparable portions of the patent file histories of the two compounds).
A. First Rejection and Eisai's Response
On April 21, 1989, the PTO rejected the ethyl homolog claim (claim 8) of the '013 application pursuant to 35 U.S.C. § 102(b), that is, for lack of novelty -- specifically, for being "anticipated by" the Juggren prior art reference, which also appeared in the '552 patent prosecution. ('013 Patent File History, DRLRAB 2471.) In the same action, the PTO rejected claims in the '013 application other than the ethyl homolog for being obvious in light of a combination of Junggren and either of two prior art references not mentioned in the '552 patent prosecution. One of these two references, which is significant to the issues in this case, is referred to by the parties as "Byk Gulden." (Id. at DRLRAB 2473.) The PTO rejection reads:
The difference between the Junggren reference [and] applicant's claimed compounds [9, 13, and 17] is the trifluoromethyl group on the benzene ring of the benzimidazole. Both [Byk Gulden] and Rainer teach trifluor[o]methyl substitution of the benzene ring in similar compounds . . . . Since these compounds are all useful as gastric inhibitors and are all closely related in structure, one of ordinary skill would be motivated to combine the reference to produce applicant's compounds. (Id.) This comment seems clearly to refer to claims other than the ethyl homolog, as the benzimidazole ring of the ethyl homolog -- like that of rabeprazole -- does not exhibit a trifluoromethyl substituent. (Compare diagrams of rabeprazole, the ethyl homolog, and Byk Gulden's "WO '646" compounds at P. Mem., glossary at 7-8.)
Eisai responded to the PTO's first rejection of the '013 application in much the same way it did to the first rejection of the rabeprazole application, namely by emphasizing the asymmetrical pyridine-ring substitution pattern of Eisai's claimed compounds and asserting superiority to omeprazole. On October 20, 1989, Crawford on behalf of Eisai narrowed the '013 patent application claims and argued:
The outstanding [rejection] . . . cites and applies primarily Junggren . . . either alone as an anticipation of certain claims or combined with other references . . . . Junggren . . . contains identical substituents in the 3- and 5-positions . . . . By contrast, in the claims of the present application, the 5-position of the pyridine ring is always occupied by hydrogen (that is, it is unsubstituted), while the 3-position contains methyl. ('013 Patent File History, DRLRAB 2507; emphasis in original.) Also as in the rabeprazole prosecution, Eisai urged its claimed compounds' superiority to omeprazole in inhibiting gastric acid secretion. ('013 Patent File History, DRLRAB 2507-08.) Going beyond its rabeprazole argument, however, in prosecuting the '013 application Eisai not only argued omeprazole's inferior efficacy but also its lack of asymmetrical substitution: "Note that omeprazole has a methyl group on both the 3- and 5-positions of the pyridine ring, [and] thus may be considered symmetrical in this regard." (Id. at DRLRAB 2508, emphasis in original.) Crawford further wrote, "Considering these structural differences and the associated differences (improvements) in gastric inhibitory activity as compared with even the preferred compounds of [Junggren] . . . the examiner will quickly appreciate the inventiveness and patentability of the claims now under review." (Id. at DRLRAB 2508-09.)
B. Second Rejection and Eisai's Response
These arguments failed to persuade the '013 application's examiners. On December 6, 1989, the PTO again issued a rejection. This time, the ethyl homolog claim, along with others, was specifically rejected "under 35 U.S.C. [§ 102(b)] as anticipated by or, in the alternative, under 35 U.S.C. [§] 103 as obvious over Junggren . . . in view of [Byk Gulden]." ('013 Patent File History, DRLRAB 2631.) Unlike the previous rejection, which had cited Byk Gulden in denying claims other than the ethyl homolog, this rejection clearly cited Byk Gulden -- a prior art reference never volunteered by Eisai in either prosecution -- directly regarding the ethyl homolog of rabeprazole.
The PTO's December 6, 1989, action also dismissed Eisai's
asymmetrical-substitution argument, finding that the prior art
actually revealed this feature. "Junggren . . . is not committed just
to the symmetrical substituents. The generic formula discloses instant
compounds by teaching asymmetrical substituents at the R3
and R5 positions. [Certain specified
examples of Junggren] . . . all teach asymmetrical substituents at the
R3 and R5 positions."
('013 Patent File History, DRLRAB 2631.) Junggren combined with Byk Gulden
rendered the ethyl homolog claims unpatentable, the PTO determined,
because "[Byk Gulden] discloses compounds which teach methoxyethoxy at
the 4-position and methyl at the 3-position" -- precisely the 3- and
4-position substituents of the ethyl homolog. (Id.) The rejection
pointed to specific compounds in Byk Gulden that served to obviate
Eisai's claims. (Id.) "One of ordinary skill in the art would be
motivated to combine these references to produce the applicants'
claimed compound," the rejection concluded. (Id.)
The PTO also cited yet another combination that rendered the ethyl homolog claims unpatentably obvious. Byk Gulden combined with another prior art application, Carlsson -- which, as described above, the '552 patent examiner Fan had mentioned as a prior art compound of rabeprazole -- taught asymmetrical substitution at the 3- and 5-positions and a methoxyethoxy substituent*fn20 at the 4-position, according to the PTO. ('013 Patent File History at DRLRAB 2632.) Eisai's data purporting to demonstrate the pharmacological superiority of the ethyl homolog claims over omeprazole were dismissed as "not . . . persuasive to overcome prima facie obviousness." (Id.)
Eisai tried again on June 6, 1990, with another submission responding to the second rejection of its '013 patent application. In this submission, Crawford sought to diminish the import of the prior-art teachings the PTO had cited, arguing that they were "broad" and "generic" and should not, in light of legal precedent, wholly preclude the patentability of later, more specific claims. ('013 Patent File History, DRLRAB 2646.) Eisai then asserted "evidence of unexpectedly good inhibition of gastric secretion," as it had in the rabeprazole prosecution, as a basis for patentability. (Id.) Its response emphasized, again, the pyridine-ring substitution pattern: "The specific selection of the two substituents [methyl at the 3-position and methoxyethoxy at the 4-position] and [nonsubstituent] hydrogen on the pyridine ring provides the strong . . . gastric secretion inhibitory activity." (Id. at ...