On Oct 15, 2021, Delaware court found Dapagliflozin compound patent valid in a Hatch-Waxman suit.

 

AstraZeneca sued Zydus for infringement of U.S. Patent Nos. 6,414,126 and 6,515,117 as Zydus filed ANDA with USFDA to seek approval to market generic version of Farxiga®. By trial, the parties narrowed the dispute to the validity of claims 1-3, 14, and 16 of the ‘117 patent. The ‘117 patent is directed to compounds, composition and methods for treatment of diabetes and related diseases through inhibition of sodium dependent glucose transporters (SGLT2).

 

Zydus argued that the asserted claims are invalid as obvious. Zydus’s proffered prior art references-WO ‘128, Hongu, and Kees. WO ‘128 discloses eighty structurally similar compounds as prospective SGLT2 inhibitors. Twenty-five of these fall within the genus Formula IB, which the patentee designates as the “most preferred” set of embodiments. The Formula IB genus has preferred chemical moieties at certain positions, as shown below:

 

Zydus argued that a POSA would have been motivated to select Example 12 of WO ‘128 as a lead compound. Going through WO ‘ 128, Zydus identified a number of preferred substituents that would have found their way into the various positions required to form Example 12. These include, for example, selecting a lead compound with a 4-methoxy substituent because that moiety was most represented at the R4 position. Zydus also argued that disclosure of biological data is not required to show a POSA’s motivation to select Example 12 as a lead compound because the ‘ 117 patent itself does not disclose biological data to do so. Court, however, said that Zydus’s arguments do not address, the premise that identification of a lead compound requires some indication that the preferred scaffold is likely to yield a biologically active compound upon modification. No such activity data is reported in WO ‘ 128 for a POSA to make a choice of lead compound based on the contents therein.

 

Second, Zydus argued the “one small, conservative change” from a methoxy to an ethoxy at the 4-distal position that is required to go from Example 12 of WO ‘ 128 to dapagliflozin would have been obvious to a POSA. Court said that Zydus’s argument presupposes that a POSA would have been able to identify Example 12 as the appropriate lead compound, which Court have already rejected. Even supposing it were, Zydus has not shown clear and convincing evidence that a POSA would have been motivated to make the required change from a methoxy to an ethoxy at the R4 position. Neither Hongu nor Kees supports the shift from methoxy to ethoxy. Moreover, Court found Dr. Batchelor’s testimony credible mentioning that a POSA would not have automatically considered making the required change because none of the eighty compounds shown in WO ‘128 had an ethoxy at the R4 position.

 

Third, Zydus argued that the prior art does not teach away from substitution of the ethoxy at the R4 position. Pointing to Hongu and Kees, Zydus argued that neither reference clearly teaches that a larger alkoxy group is likely to produce less favorable results. But Court said that both Hongu and Kees teach away from substitution of the ethoxy for the methoxy at the R4 position because Hongu disfavors larger alkoxy groups at that position and Kees presents an ethyl group, which is not an alkoxy, as its most preferred group.

 

Court thus held that, based on the lack of biological activity data in WO ‘128 and the teaching away from the use of larger alkoxy groups at the 4-distal position by Hongu and Kees, a POSA would not have been motivated to swap the distal 4-methoxy in WO ‘128 ‘s Example 12 with an ethoxy. Given the lack of available biological data for candidate SGL T2 inhibitors and the unpredictability of changes in biological activity due to modification of chemical structure, a POSA would not have had a basis to expect dapagliflozin to exhibit better glucose-reducing effects than the closest prior art, such as Example 12. Therefore, Zydus failed to prove that the asserted claims are obvious.