The immune system is said to be a double-edged sword: On one hand, it protects us from foreign invaders such as bacteria and viruses. On the other, it can recognize its host as the enemy, causing autoimmune disorders like rheumatoid arthritis. A similar double-edged sword principle holds true for prior art references. In a recent decision in Momenta Pharmaceuticals, Inc. v. Bristol-Myers Squibb Company (IPR2015-01537), prior art cited by Momenta in a § 103 challenge to Bristol-Myers Squibb’s CTLA4Ig formulation patent turned out to be the very instrument that killed its case.

The patent at issue in Momenta claims stable liquid formulations of the therapeutic molecule CTLA4Ig, which is a protein molecule used to treat immune system diseases and disorders such as rheumatoid arthritis and adverse transplant reactions. In a § 103 challenge, the petitioner relied on the teachings of the “Carpenter” reference, arguing that the claimed formulation could be achieved using the approach and limited set of possible ingredients taught therein. The petitioner had the burden of establishing that a person of ordinary skill in the art would have had a reasonable expectation of success in making the formulation.

In order to establish this, the petitioner pointed to a statement in Carpenter that the authors were “quite confident” that the generalized formulation procedure they described could be coupled with screening strategies to facilitate drug development. However, the Board was not convinced that the term “quite confident” spoke to the ability to formulate specific proteins. Seizing a golden opportunity, the patent owner pointed to disparaging statements made in Carpenter that effectively turned the reference against the petitioner. Namely, Carpenter expressly stated “[i]t [could] be assumed that most proteins will not exhibit sufficient stability in aqueous solution,” and “for most proteins maintaining physical and chemical stabilities in aqueous solution for an extended period of time is extremely difficult.”

It followed that the challenged claims were not obvious in view of the petitioner’s cited art. In particular, the statements in Carpenter that the guidance for formulating proteins may not lead to a successful result were found to be persuasive to the Board. Unfortunately for the petitioner, its choice of art killed its chances of establishing a reasonable expectation of success to prove obviousness.


For petitioners, this case exemplifies how prior art references cited against a patent owner have the potential to be self-defeating rather than self-serving. Therefore, carefully evaluate a reference from the perspective of the patent owner—especially longer references, such as books, that often have multiple teachings. For patent owners seeking protection for specific formulations of known drugs, this case increases the confidence that claims to drug formulations, even if developed via well-known or routine methods, can survive obviousness challenges by casting doubt on their anticipated success.