Drug patent survives invalidity challenge based on lack of teaching or suggestion in the art, teaching away, unexpected results and long-felt need
| August 15, 2017
Millennium Pharmaceuticals Inc. v. Sandoz Inc., et al.
July 17, 2017
Before: Newman, Mayer and O’Malley. Opinion: Newman
Summary
The inventors discovered that freeze-drying mannitol with an existing, but unstable, drug resulted in an ester which solved stability and solubility problems of the drug alone. The patent survived an invalidity challenge since there was no teaching or suggestion of freeze-drying this drug in the presence of mannitol. Non-obviousness was further bolstered by a teaching away, unexpected results and evidence of a long-felt need.
Details
Background
Millennium is the exclusive licensee of U.S Patent No. 6,713,446, which discloses D-mannitol N-(2-pyrazine)carbonyl-L-phenylalanine-leucine boronate. This compound is a boronate ester of bortezomib and D-mannitol, with the circled portion below forming the bonds between the bortezomib moiety and the D-mannitol moiety.
One of the claims in question recites a lyophilized (freeze-dried) version of this compound.
Prior to the ‘446 patent, bortezomib was known as an effective drug against some cancers, but it never obtained FDA approval due to its instability, rapid degradation in liquid formulations and insolubility. Despite evaluating about 20 formulations, a formulation sufficiently stable for storage, transportation and administration could not be obtained. Bortezomib was disclosed in U.S. Patent No. 5,780,454, referred to as the “Adams Patent.” Later, the inventor of the ‘446 patent consulted with the inventors of the Adams patent. About 25 more formulations were evaluated, but the same problems persisted.
Eventually, the inventors experimented with lyophilized formulations, and produced a formulation which is a lyophilized version of bortezomib and mannitol, a known bulking agent. This formulation did not have the problems described above. The improvement was due to the formulation of an ester of bortezomib and mannitol during the lyophilization. This ester acted as a “prodrug” which releases the active ingredient upon administration.
The resulting product is claimed in the ‘446 patent and is marketed commercially as Velcade. The FDA quickly approved the drug, and it is used in treatment of multiple myeloma.
Sandoz and other defendants filed an ANDA application, and Millennium filed suit. The defendant’s defense is based purely on alleged invalidity of several claims of the patent.
District Court
At the district court, the claims were held to be obvious as an inherent result of an obvious process: lyophilizing bortezomib in the presence of mannitol. Millennium argued that lyophilization would not have been obvious because bortezomib was known to be unstable “even if the dry state as a freestanding compound.” However, the district court sided with Sandoz’s expert, who stated that lyophilization was a well-known technique used when a liquid formulation is not successful.
Although the prior art did not disclose reacting bortezomib with mannitol or that an esterification reaction occurs, and also did not disclose that lyophilization would result in a new chemical compound, the district court considered the claimed subject matter obvious. In particular, the district court considered that the lyophilizing bortezomib with mannitol was an option “from which the prior art did not teach away.” Although the Adams Patent did not mention mannitol (it mentioned boronate esters and other sugar alcohols such as glycerol), the district court stated that it “pointed directly to” mannitol.
Despite hearing unrebutted testimony that the formulation was not expected upon lyophilization, the court held that the ester is the “natural result” of freeze-drying bortezomib with mannitol. The court held that the product of a chemical reaction is inherent in the reaction and would have been obvious.
As to objective indicia of nonobviousness, the district court criticized Millennium for not comparing the claimed compound to a lyophilized ester of bortezomib and glycerol, and thus discounted their argument regarding unexpected results. As to long-felt need, the court stated that the claimed formulation did not “solve any problem having persisted over a long period of time without resolution by the prior art.”
CAFC
At the CAFC, Sandoz argued that lyophilization and the use of bulking compounds were well known, which the panel did not dispute. However, the panel stated that there was nothing in the art which would lead one to use mannitol to make an ester via lyophilization, or that such an ester would solve the problems in the art.
Although Sandoz cited art which allegedly suggested the claimed invention, the CAFC did not agree. The references cited by Sandoz either (1) disclose lyophilization, but without reference to bortezomib, (2) disclose mannitol as a bulking agent, but without reference to bortezomib, or (3) disclose generally that boronic acids can form esters, but without reference to mannitol. As an example of a reference which disclosed that boronic acids can form esters generally, the Adams Patent cited 10 alcohols, none of which was mannitol. None of the expert witness testimony sufficiently filled in the gaps in the cited art.
Next, the CAFC discussed the teaching away argument. The district court stated that lyophilizing bortezomib with mannitol was an option “from which the prior art did not teach away.” However, Millennium presented sufficient evidence that one skilled in the art would not have modified bortezomib because they would be afraid to upset the anti-cancer properties of the compound by blocking a large portion of bortezomib with mannitol.
The CAFC then discussed the inherency position of the district court. Sandoz argued that even if the lyophilization produced an unexpected result, this was “inevitable” and therefore inherent and not inventive. However, the CAFC returned to the point that the reaction between bortezomib and mannitol was unexpected, as was the fact that the resulting ester avoids the problems in the prior art.
Finally, the CAFC discussed the objective indicia of nonobviousness. Although evidence was presented comparing the lyophilized mannitol-bortezomib ester to bortezomib itself, the district court instead argued that the closest prior art was not bortezomib alone, but rather a glycerol-bortezomib ester. The Adams Patent mentioned glycerol, it did not actually disclose or test a glycerol-bortezomib ester or that any ester which solved the problems in the art. The CAFC concluded that bortezomib alone should have been considered the closest prior art, since unexpected results must be compared to that which is known, not what is unknown.
The CAFC also discussed the evidence of a long-felt need as further illustrating non-obviousness. The district court wrongly considered the commercial success of Velcade to be due to bortezomib alone, since this is not a commercially available product. Rather, the CAFC held that the commercial success of Velcade is due to the structure of the claimed ester, and the resulting improved stability and solubility.
For all of the above reasons, the CAFC concluded that the finding of obviousness was incorrect.
Takeaways
Here, the district court seemed to be making several leaps from the Adams Patent. Although the Adams Patent generally mentioned bortezomib esters and glycerol, it did not explicitly disclose a glycerol-bortezomib ester. The district court seemed to have assumed that the Adams Patent made such a disclosure, and made a “jump” from glycerol to mannitol. This case serves as a reminder that obviousness needs to flow from the disclosures themselves and the skill of those in the art, without skipping any steps or making assumptions.
When an inherency argument is presented in an obviousness rejection, it appears that the best course of action is to focus on the non-obviousness of the product or process rather than the allegedly inherent effect, function or properties of the claimed product or process.
Unexpected results must be compared to that which is known, not what is hypothetical. It was incorrect for the district court to require a comparison to a compound never disclosed in the art, rather than to compare the inventive compound to the existing lead compound.
Although rare, evidence of a long-felt need can be helpful to further demonstrate non-obviousness in some situations.
