In Teva Pharms. USA, Inc. v. Corcept Therapeutics, Inc., No 21-1360 (Fed. Cir. Dec. 7, 2021), the Federal Circuit affirmed the Final Written Decision of the Patent Trial and Appeal Board (“the Board”) holding Teva failed to show all the challenged claims in PGR2019-00048 would have been obvious.

Background

Corcept markets a 300 mg mifepristone tablet under the name Korlym^®. In approving Corcept’s New Drug Application (“NDA”) for Korlym^®, the U.S. Food and Drug Administration (“FDA”) required Corcept to conduct a drug-drug interaction clinical trial of mifepristone and ketoconazole (a strong CYP3A4 inhibitor). Id. at *3. The FDA also provided an Office of Clinical Pharmacology memorandum (“Lee”) explaining that the drug-drug interaction study was necessary to determine whether there was a safety risk in co-administration of CYP3A inhibitors and mifepristone. Id.

Corcept’s original Korlym^®label recommended a starting dose of 300 mg once daily up to a maximum of 1200 mg once daily, but with a warning to limit the mifepristone dose to 300 mg once daily when used with strong CYP3A inhibitors. Id. at *3-4.

Based on the subsequent drug-drug interaction study, Corcept applied for received U.S. Pat. No. 10,195,214 (“the ’214 patent”). Claim 1 of the ’214 patent read:

1. A method of treating Cushing’s syndrome in a patient who is taking an original once-daily dose of 1200 mg or 900 mg per day of mifepristone, comprising the steps of:

reducing the original once-daily dose to an adjusted once-daily dose of 600 mg mifepristone,
administering the adjusted once-daily dose of 600 mg mifepristone and a strong CYP3A inhibitor to the patient,
wherein said strong CYP3A inhibitor is selected from the group consisting of [ ].

The ’214 patent reflects the drug-drug interaction study by allowing up to 600 mg of mifepristone in combination with the strong inhibitor. The original label clearly contradicts this.

Corcept sued Teva for infringement of the ’214 patent, and Teva then sought post-grant review of claims 1-13. Teva argued the ’214 claims were obvious based on the Korlym^®label and Lee, optionally in combination with FDA guidance on drug-drug interaction studies. Id. at *4-5.

The Board rejected Teva’s obviousness argument, finding that Teva failed to show that a person of ordinary skill in the art (“POSITA”) “would have had a reasonable expectation of success for safe co-administration of more than 300 mg of mifepristone with a strong CYP3A inhibitor.” Id. at *5. This conclusion followed from the Board’s construction of the claims to require safe administration of mifepristone (even though that was not recited in the claims) and the Board’s rejection of Teva’s expert’s testimony that “based on the Korlym^®label and Lee, ‘it was reasonably likely that 600 mg [per day of mifepristone] would be well tolerated and therapeutically effective when co-administered with a strong CYP3A inhibitor.’” Id. The Board found that testimony inconsistent with Teva’s expert’s later testimony that a POSITA “would have no expectation as to whether the co-administration of 600 mg of mifepristone with ketoconazole would be safe.” Id. at *7 (emphasis in original).

Federal Circuit Decision

The Federal Circuit affirmed the Board’s decision. Although Teva argued that the Board erred in its reasonable expectation of success analysis by requiring precise predictability of a specific dosage, the Federal Circuit disagreed. “The Board applied the correct standard, requiring only a reasonable expectation of success and tying its analysis to the scope of the claimed invention.” Id. at *8. Acknowledging that “[a]bsolute predictability is not required[,]” it was Teva’s burden to prove a reasonable expectation of success for a 600 mg dosage. Id. at *6. But Teva failed to show that a POSITA would have had “no expectation as to whether co-administering dosages of mifepristone above the 300 mg/day threshold set forth in the Korlym label would be successful.’” Id. at *7 (emphasis in opinion). “Because there was no expectation of success for any dosages over 300 mg per day, there was no expectation of success for the specific 600 mg per day dosage. … Nothing about this analysis required precise predictability, only a reasonable expectation of success tied to the claimed invention.” Id. Monotherapy doses above 300 mg per day do not change this conclusion since the claim is limited to co-administration doses. The Federal Circuit upheld the Board’s finding that a POSITA would not have expected monotherapy and co-administration dosages to behave similarly. Id. at *10.

The Court went on to hold that substantial evidence supported the Board’s finding that the condition ranges disclosed in the prior art did not overlap with the claimed invention. Id. at *9. As noted above, the prior art warned against co-administration with doses of more than 300 mg per day when used with strong CYP3A inhibitors.

Take-Aways 

Showing obviousness requires both a motivation to modify/combine the prior art AND a reasonable expectation of success in arriving at the claimed invention. Beyond looking to the prior art to determine if it suggests doing what the inventor has done, this case reinforces that one must also consider if the art provides the required expectation of succeeding in that endeavor. “Obviousness does not require absolute predictability, but a reasonable expectation of success is necessary.” In re Clinton, 527 F.2d 1226, 1228 (CCPA 1976).

This case also highlights the importance of claim construction. Corcept’s claims were construed to require safe administration, even though that was not recited in the claims.  But the Board and the Federal Circuit focused on precisely what the clinical trials showed and what Corcept wisely claimed. Claim construction was an important underpinning of the conclusion that the success of the claimed invention could not have been reasonably expected. 

And, as always, cautiousness is obligatory when selecting an expert. Here, the expert’s inconsistent testimony contributed to an unfavorable outcome for Teva.

Jordan Cowger†, Sumaia Tabassum‡, and Stacy Lewis also contributed to this article.