Structurally Similar Chemical Compounds Alone Do Not Render a Compound Obvious
December 26, 2006
Last Month at the Federal Circuit - January 2007
Judges: Rader (author), Schall, Gajarsa
[Appealed from: S.D. Ind., Judge Young]
In Eli Lilly & Co. v. Zenith Goldline Pharmaceuticals, Inc., Nos. 05-1396, -1429, -1430 (Fed. Cir. Dec. 26, 2006), the Federal Circuit affirmed the district court’s holding that the asserted claims of U.S. Patent No. 5,229,382 (“the ’382 patent”) were valid, enforceable, and infringed.
Eli Lilly and Company (“Lilly”) owns the ’382 patent, which discloses olanzapine and its use to treat schizophrenia. Lilly had previously discovered other drugs in the same family of thienobenzodiazepines, including clozapine, flumezapine, ethyl flumezapine, and ethyl olanzapine (“Compound ’222”). In fact, Lilly marketed clozapine as the first “atypical” antipsychotic drug in the late 1960s; however, it was withdrawn in 1975 following the discovery that it caused a potentially fatal blood disorder in one percent of patients. Fourteen years later, no better drug had been developed, and the FDA reapproved the use of clozapine in combination with careful blood monitoring.
In 1996, Lilly began marketing olanzapine as Zyprexa®. Olanzapine differs from other members of the thienobenzodiazepine family in two critical respects. First, olanzapine has a hydrogen atom substituted on the benzene ring, rather than a chlorine atom (as found in clozapine) or a fluorine atom (as found in flumezapine). This halogen substitution, sometimes referred to as the “neuroleptic substituent,” was an electron-withdrawing group widely believed to be responsible for the antipsychotic activity of clozapine, flumezapine, and other antipsychotics before the discovery of olanzapine. Second, the thiophene ring of olanzapine is substituted with a methyl group, rather than an ethyl group, as found on ethyl flumezapine and ethyl olanzapine (both of which did not reach the market because they cause significant side effects).
Zenith Goldline Pharmaceuticals, Inc. (now IVAX Pharmaceuticals, Inc., “IVAX”), Dr. Reddy’s Laboratories, Ltd. (“DRL”), and Teva Pharmaceuticals USA, Inc. (“Teva”) each filed an ANDA for a generic version of Zyprexa®, thereby conceding infringement. Lilly sued. The district court held that the defendants did not prove by clear and convincing evidence that the claims of the ’382 patent were invalid as anticipated or obvious.
On appeal, the Federal Circuit first agreed with the district court that the claims of the ’382 patent were not anticipated by a cited reference (“Chakrabarti”) disclosing millions of compounds in the same general family of thienobenzodiazepines. The Court rejected IVAX’s argument that Chakrabarti’s disclosure of compounds in the thienobenzodiazepine family anticipated claim 1, to olanzapine, because it did not spell out “a definite and limited class of compounds” that enabled one of ordinary skill in the art to “at once envisage” each member of the limited class. In distinguishing the present case from the cited case law, the Court noted that Chakrabarti disclosed millions of compounds, with sixty compounds specifically examined. None of the preferred compounds resembled olanzapine, and the preferred compounds all had a fluorine or chlorine substituent on the benzene ring, rather than a hydrogen, as found in olanzapine. After describing other differences between olanzapine and the prior art compounds, the Court concluded that there was no anticipation because (1) the cited reference preferred complete compounds, not individual substituents; (2) there was no generic disclosure encompassing olanzapine; and (3) there was no suggestion to modify the closest described compound into a preferred compound.
Turning to obviousness, the Federal Circuit first agreed with the district court that claims to olanzapine are not obvious because the prior art taught away from antipsychotics that lack a halogen substituent on the benzene ring. The Court explained that, for a chemical compound, a prima facie case of obviousness requires the prior art to have “structural similarity” to the claimed compound and provide a reason or motivation to make the inventive compound. Lilly’s own prior art patent, U.S. Patent No. 4,115,574 (“the ’574 patent”) disclosed Compound ’222, which, like olanzapine, has a hydrogen atom rather than a halogen substituent. However, the Federal Circuit noted that the ’574 patent expressed a preference for a halogen-containing compound, and that the “prior art references at the time of this invention taught away from using a non-halogenated compound as a substituent in the benzene ring, exactly where olanzapine has a hydrogen atom.” Slip op. at 11.
The Federal Circuit also agreed with the district court’s determination that a person of ordinary skill in the art would not have chosen Compound ’222 as a starting compound to further modify, because it did not contain the neuroleptic halogen substituent. In addition, Compound ’222 has an ethyl group substitution where olanzapine has a methyl group; the Court found no motivation to modify this substituent in the prior art. Going further, the Court found that the art taught away from selection of Compound ’222 as a lead compound. The ’574 patent did not provide any biological data for Compound ’222, but instead indicated that halogensubstituted compounds were preferred, and described the fluorine-substituted ethyl flumezapine as “particularly active.” Id. at 12. Other art taught that substitution with fluorine or chlorine increased antipsychotic activity and reported that Compound ’222 was less active than clozapine, the benchmark for this class of compounds.
The Federal Circuit also rejected IVAX’s argument that olanzapine is rendered obvious by Compound ’222, because structurally, olanzapine is the adjacent homolog of Compound ’222. The Court emphasized that the “patentability for a chemical compound does not depend only on structural similarity.” Id. at 11. If a “relevant property” of a compound is “unexpected and significant,” that property cannot be overlooked, regardless of how structurally similar the compounds may be, but can render the inventive compound nonobvious. Id. Here, although there is some structural similarity of olanzapine to the prior art, olanzapine exhibits “unexpected beneficial properties,” which must be accounted for in the analysis and lead to nonobviousness.
The Federal Circuit also explained that the prior art did not provide motivation to make the modifications required to reach olanzapine. The Federal Circuit dismissed the argument that olanzapine was “bracketed” by two compounds in the prior art with similar structures: combining Compound ’222’s hydrogen-substituted benzene ring and flumezapine’s methyl substitution on the thiophene ring generates the structure of olanzapine, thereby making olanzapine prima facie obvious. Structural similarity is not controlling in this case, and the prior art did not contain any suggestion to make these modifications. Mere identification in the prior art of each component of a composition does not render the combination obvious; the law requires some motivation to select and combine the references to reach the claimed invention.
Even if a prima facie case of obviousness could be established, the Federal Circuit held it would be overcome by Lilly’s extensive secondary considerations. “The record shows a long-felt need for a safer, less toxic, and more effective clozapine-like drug; a decade (or more) of failure to find a replacement for clozapine; a reasonable amount of commercial success for olanzapine; and a number of awards for olanzapine as indicators of industry acclaim.” Id. at 14.
The Federal Circuit next upheld the district court’s finding that Lilly’s clinical trials of olanzapine were an experimental, rather than public use, and therefore, negated any statutory bar under § 102(b). The Court emphasized that a use which occurs in the open will not trigger a statutory bar when undertaken to experiment on or with the claimed invention. Here, phase I clinical trials were performed to test the safety and efficacy of olanzapine. The trials were conducted in the Lilly clinic, with restricted access, security, and confining the volunteers’ movements. The Court concluded that because Lilly had “tailored its tests to their experimental drug safety and efficacy purpose, adequately monitored for results, and maintained confidentiality,” the trial court did not err in finding no public use. Id. at 16.
Finally, the Federal Circuit considered and rejected DRL’s assertion that the ’382 patent should be declared unenforceable due to inequitable conduct. First, when the PTO questioned Lilly about blood cholesterol levels in dog studies, Lilly did not disclose to the PTO its statements to the Swedish Board about the hematotoxic effects of olanzapine in these studies. Because the hematotoxicity findings were “believed not to have clinical relevance to humans,” the Court concluded that Lilly did not fail to disclose information to the PTO. Second, the Federal Circuit did not find that the declaration of a Lilly physician was false, or that certain information was withheld from the PTO with an intent to deceive. On a third inequitable conduct charge, the Court concluded that Lilly’s nondisclosure of Chakrabarti and the ’574 patent was neither a material omission nor done with the intent to deceive, where another Lilly patent with an identical specification was disclosed and Chakrabarti was cited by the examiner during prosecution.