On summary judgment, the United States District Court for the Southern District of New York found in favor of plaintiffs Eisai Co., Ltd. and Eisai, Inc. (collectively Eisai) with respect to the validity and enforceability of U.S. Patent No. 5,045,552 (’552 patent). Eisai Co.
v. Teva Pharms. USA, Inc.,
*1356 I
The '552 patent claims rabeprazole and its salts. Rabeprazole is part of a class of drugs known as proton pump inhibitors, which suppress gastric acid production by inhibiting action of the enzyme H +K+AT-Pase. The distinctions between rabepra-zole and its salts are not relevant for this appeal. Therefore this court refers to ra-beprazole and its salts collectively as “ra-beprazole.” Rabeprazole’s sodium salt is the active ingredient in Aciphex, a pharmaceutical approved in 1991 by the FDA for the treatment of duodenal ulcers, heartburn, and associated disorders. Aciphex has been a commercial success, garnering over $1 billion in worldwide yearly sales.
Dr. Reddy’s and Teva each filed Abbreviated New Drug Applications (ANDAs) under the Hatch-Waxman Act, 21 U.S.C. § 355 and 35 U.S.C. § 271(e), seeking to manufacture a generic version of Aciphex before the expiration of the '552 patent. Because filing an ANDA is an artificial, but legally cognizable, act of patent infringement,
see Glaxo Group Ltd. v. Apotex, Inc.,
II
This court reviews a grant of summary judgment without deference.
Dayco Prods., Inc. v. Total Containment, Inc.,
Where, as here, the patent at issue claims a chemical compound, the analysis of the third
Graham
factor (the differences between the claimed invention and the prior art) often turns on the structural similarities and differences between the claimed compound and the prior art
*1357
compounds.
See Eli Lilly & Co. v. Zenith Goldline Pharms., Inc.,
Teva asserts that a combination of three prior art references renders the '552 patent obvious: 1) European Patent No. 174,726 (owned by Takeda), claiming lansoprazole (EP '726); 2) United States Patent No. 4,255,431 (to Junggren), claiming omeprazole (’431 patent); and 3) an article by Brándstrom, et ah, entitled “Structure Activity Relationships of Substituted Benzimidazoles” (Brándstrom). EP '726 teaches, inter alia, the ulcer treatment compound lansoprazole. Lansoprazole differs structurally from rabeprazole at the 4-position on the pyridine ring, as indicated in the diagram below. Lansoprazole has a trifluoroethoxy (OCH2CF8) substituent, whereas rabeprazole has a methoxypropoxy (OCH2CH2CH2OCH3) substituent.
[[Image here]]
[[Image here]]
Appellant Leva’s Br. at 28. Otherwise, the two compounds are identical. See SJ Validity Order at 7. Both rabeprazole and lansoprazole are “asymmetrically substituted” with respect to the 4-position on the pyridine ring because the substituent at the 3-position (a methyl group in both compounds) is not the same as the substi-tuent at the 5-position (a hydrogen in both compounds).
The '431 patent discloses a broad class of gastric acid inhibiting compounds, in-eluding omeprazole, the first commercial proton pump inhibitor, sold as Prilosec. Although sharing the same basic structure, omeprazole is structurally farther afield from rabeprazole than is lansoprazole. For instance, omeprazole’s pyridine ring is symmetrically substituted and has a me-thoxy (OCH3) group at the 4-position.
Finally, Brándstrom describes a class of anti-ulcerative compounds having a benzi-midazole-sulfinylmethyl-pyridine core (the Brándstrom core structure):
*1358 [[Image here]]
Rabeprazole, lansoprazole, and omeprazole are all Brándstróm core structure compounds. Taking the evidence in the light most favorable to Teva, this court assumes that as per EP '726, lansoprazole is twenty times superior to omeprazole for anti-ulcer action, as measured by an indomethaein-induced gastric lesion assay in rats. This court also assumes that lansoprazole has certain traits, including lipophilicity (the ability of a compound to cross lipid membranes) and low molecular weight, that would have made it desirable to a skilled artisan.
Under these assumptions, one of skill in this art may have considered it a candidate for a lead compound in the search for anti-ulcer compounds. To the contrary, the district court emphasized the differences between anti-ulcer action and gastric acid inhibition. The trial court specifically noted that Teva’s expert testified with respect to the EP '726 data that “[t]he level of acid secretion ... from these [anti-ulcer] data ... cannot be determined.”
SJ Validity Order
at 13. In this context, this court consults the counsel of
KSR
that “any need or problem known in the field of endeavor at the time of invention and addressed by the patent can provide a reason for combining the elements in the manner claimed.”
Nonetheless, as the district court noted, the EP '726 reference teaches at best that the fluorinated substituent of lansoprazole provides “a special path to achieving lipo-philicity.” SJ Validity Order at 10 (emphasis in original). And Teva’s expert identified a separate reference teaching that fluorine-substituted groups increase lipophilicity. Id. The record, however, shows no discernible reason for a skilled artisan to begin with lansoprazole only to drop the very feature, the fluorinated sub-stituent, that gave this advantageous property. Indeed, Teva’s pharmacology expert, Dr. John Forte, declined to opine on lansoprazole’s relevance to an examiner assessing the patentability of rabeprazole. J.A. at 14894. And Dr. Reddy’s pharmacology expert, Dr. Simmy Bank, testified in deposition that “I thought [lansoprazole] had nothing to do with this trial.” J.A. at 14756.
This court notes that the district court did not rigidly limit Teva’s obviousness arguments by forcing Teva to select a single lead compound. Rather Teva alone
*1359
selected lansoprazole as the anchor for its obviousness theory, not the district court. In
KSR,
the Supreme Court noted that an invention may have been obvious “[w]hen there [was] ... a design need or market pressure to solve a problem and there [were] ... a finite number of identified, predictable solutions.”
In other words, post-N&R, a prima facie case of obviousness for a chemical compound still, in general, begins with the reasoned identification of a lead compound. Teva cannot create a genuine issue of material fact on obviousness through the unsupported assertion that compounds other than lansoprazole might have served as lead compounds. Further, the record contains no reasons a skilled artisan would have considered modification of lansopra-zole by removing the lipophilicity-confer-ring fluorinated substituent as an identifiable, predictable solution. In sum, the district court properly concluded that the record did not support a case of obviousness of the '552 patent as a matter of law.
Ill
As with other summary judgment issues, this court reviews a district court’s summary judgment on inequitable conduct without deference.
Innogenetics, N.V. v. Abbott Labs.,
Inequitable conduct in prosecuting a patent application before the United States Patent
&
Trademark Office may take the form of an affirmative misrepresentation of material fact, a failure to disclose material information, or the submis
*1360
sion of false material information, but in every case this false or misleading material communication or failure to communicate must be coupled with an intent to deceive.
Innogenetics,
On appeal, Teva and Dr. Reddy’s allege that Eisai misled the Patent Office in five ways: 1) failing to disclose Eisai’s own co-pending '013 application, which claimed the “ethyl homolog” of rabeprazole (compound SHKA 661); 2) withholding rejections from the '013 application’s prosecution that also would have been applicable to the '552 patent’s prosecution; 3) failing to disclose the prior art “Byk Gulden patent” (WO 8602646); 4) submitting a misleading declaration (the Fujisaki Declaration) to the examiner of the '552 patent; and 5) concealing lansoprazole from the examiner. The district court rejected the fifth assertion on summary judgment, SJ Enforceability Order at 58, and the other four after a bench trial, Trial Order.
Teva and Dr. Reddy’s first and second allegations rely on Eisai’s failure to disclose the fact of, and rejections contained in, Eisai’s patent application claiming the “ethyl homolog” of rabeprazole. Known to Eisai’s scientists as compound SHKA 661, the ethyl homolog differs from rabeprazole as its name suggests. SHKA 661 has one fewer methylene unit at the 4-position of the pyridine ring, giving SHKA 661 an ethoxy group rather than a propoxy group at this position. The district court correctly pointed out that calling SHKA 661 the “ethyl homolog” of rabeprazole in this case could carry a misleading implication with respect to inequitable conduct. The record supplies no evidence to suggest that Eisai’s scientists ever referred to SHKA 661 by this name, or thought of SHKA 661 and rabeprazole “primarily in relation to each other.” Trial Order at 17 n. 7. Rather, the district court found credible the testimony that Eisai scientists considered SHKA 661 separately patentable, even though Eisai ultimately did not pursue that course. Id. at 22-23; 42-43. Furthermore, even if a provisional obviousness-type double-patenting rejection might have issued in the prosecution of the '552 patent due to the co-pending SHKA 661 application, the district court found the materiality of this potential situation low, because applicants routinely overcome this type of rejection, id. at 44, by amending claims or filing a terminal disclaimer. Nonetheless, the district court did not hold that the fact of the copendency of these two applications to be totally immaterial, accurately noting that applicants should be encouraged to disclose closely related applications. Id. at 47.
While disclosure of the co-pending SHKA 661 application to the Patent Office during the prosecution of the '552 patent would have been prudent, Eisai’s failure to do so is by no means fatal, for two reasons. First, the district court had ample evidence from which to conclude that the materiality of the SHKA 611 application *1361 was low, as outlined above. Second, the record is devoid of any real suggestion of intent to deceive the Patent Office, much less the clear and convincing evidence required to support a finding of inequitable conduct.
As for the rejections of the '013 application that would have been relevant to the prosecution of the '552 patent, the district court did not reach materiality because it discerned insufficient proof of intent to deceive. The district court found the documentary evidence (faxed exchange between Eisai employees Mr. Shuhei Miyaza-wa, one of the inventors of the '552 patent, and Mr. Mitsuo Taniguchi, Eisai’s patent agent, regarding Mr. Miyazawa’s presentation to a pharmaceutical trade industry group) to supply no compelling evidence of intent, based on testimony from both parties to the fax. Witness credibility determinations lie squarely within the district court’s discretion.
See Medichem, S.A. v. Rolabo, S.L.,
Finally, the district court found that Teva’s theory that Eisai deliberately hid the ball from the Patent Office by separately fifing the '552 and '013 prosecutions to be “implausibly risky,” given that such similar applications would usually be assigned to the same examiner in the same art unit. Trial Order at 53. The district court thus had ample bases from which to conclude that Eisai’s failure to disclose its co-pending '013 application along with the rejections issued in its prosecution, while not completely forthcoming, did not rise to the level of inequitable conduct.
With respect to the Byk Gulden patent, Teva and Dr. Reddy’s argue that Eisai’s failure to disclose this reference to the Patent Office during prosecution of the '552 patent was material because a reasonable examiner would have used it to issue a new and stronger prima facie obviousness rejection on the basis of Byk Gulden’s disclosure of asymmetrically-substituted compounds having a methoxyethoxy at the 4-position of the pyridine ring. But the district court found Byk Gulden’s teachings cumulative with references already disclosed to the Patent Office (Junggren or Junggren combined with Beecham). As per 37 C.F.R. § 1.56, cumulative evidence is definitionally not material evidence.
See Monsanto Co. v. Bayer Bioscience N.V.,
As for the Fujisaki Declaration, Eisai submitted it during prosecution to overcome an obviousness rejection. Because this reference shows rabeprazole’s pharmacological properties, the trial court found it highly material. Id. at 59. Teva *1362 and Dr. Reddy’s argue that the data presented in the Fujisaki Declaration were misleading. They contend that the comparison with two non-prior art compounds without a comparison of the ethyl homolog of rabeprazole, SHKA 661, sent the examiner on a dead-end side trip. The district court properly characterized this argument as “contorted.” Id. The Fujisaki Declaration indisputably showed a comparison between rabeprazole and the prior art compound called out by the examiner, demonstrating rabeprazole’s superiority. Further, as discussed above, the materiality of SHKA 661 and the patent application claiming it was low. The data from the Fujisaki Declaration were relevant to prosecution, but Eisai had no obligation to include additional, unnecessary data such as a comparison to SHKA 661. Thus the district court did not abuse its discretion in concluding that Eisai did not commit inequitable conduct in failing to include additional data in the Fujisaki Declaration to the examiner. Even here, where the submission to the Patent Office itself was highly material to prosecution, the lack of deceptive intent rendered stillborn yet another allegation of inequitable conduct.
Finally, Teva and Dr. Reddy’s assert that that Eisai deceptively declined to inform the examiner of a patent application for lansoprazole, a prior art proton pump inhibitor (and the active ingredient in Pre-vacid). The district court disposed of this argument on summary judgment. The district court found that Teva and Dr. Reddy’s had presented neither direct evidence of deceptive intent nor any evidence to support an inference of materiality. SJ Enforceability Order at 58. The strongest evidence of some problem was the passing comment of one Eisai “insider” that the similarity of lansoprazole and rabeprazole “bothers me.” Id. at 59. But this vague, subjective statement is not sufficient by any means to establish materiality, let alone intent. Moreover, given lansopra-zole’s fluorinated substituent and its resultant impotence to render the '552 patent invalid, the district court properly rejected this strained theory of inequitable conduct on summary judgment.
IV
In a series of thoughtful, thorough opinions, the district court carefully explained its reasoning with respect to both obviousness and inequitable conduct. Because the district court properly concluded that Teva and Dr. Reddy’s failed to prove that the '552 patent was invalid for obviousness or unenforceable for inequitable conduct, this court affirms the district court’s judgment.
AFFIRMED
COSTS
Each party shall bear its own costs.