As I wrotelast week, the Federal Circuit held in Centocor Ortho Biotech, Inc. v. Abbott Laboratories, that the human antibody claims that Centocor had asserted against Abbott's Humira® product are invalid for failing to satisfy the written description requirement of 35 USC § 112. The holding is getting lots of attention because it will negate a $1.67 billion damages award, but it also may be important as a cautionary tale for those who would "ensnare" competitors with later-filed claims alleged to be supported by an early-filed disclosure.
The Antibody Dilemma
As described in the Federal Circuit decision, both Abbott and Centocor developed therapeutic antibodies that could be used to treat autoimmune conditions associated with the overproduction of TNF-α, such as arthritis. To be suitable for therapeutic use, the antibodies have to:
- bind to TNF-α with high affinity,
- neutralize the biological activity of TNF-α, and
- exhibit low immunogenicity when administered to humans.
As reflected in the asserted patent (U.S. 7,070,775), Centocor's approach was to engineer antibodies with a murine variable region that exhibited a high affinity for human TNF-α and had neutralizing activity, and a human constant region that made the antibody less immunogenic than a fully murine antibody would be. The chimeric antibody described in the '775 patent is referred to as the cA2 antibody.
As reflected in the accused product (Humira®), Abbott's approach was to engineer a fully human antibody, with both a human constant region and a human variable region. According to the Humira® prescribing information, the antibody is a recombinant human IgG1 monoclonal antibody with human derived heavy and light chain variable regions and human IgG1:k constant regions.
The Patent At Issue
The asserted patent (U.S. 7,070,775) was filed in 2002, with a lengthy priority claim that includes numerous "continuation-in-part" relationships reaching back to the original application filed in 1991. Abbott has its own patent (U.S. 6,090,382) that was filed in 1996, so Centocor had to assert that the relevant claims of the '775 patent were entitled to claim priority to the 1994 series of applications. As the Federal Circuit put it:
To ensnare Abbott with later-filed claims, Centocor must use a priority date from an earlier application.
The Written Description Analysis
Claim 2 is representative of the asserted claims, and recites:
2. The antibody or antigen-binding fragment of claim 1, wherein the antibody or antigen binding fragment comprises a human constant region and a human variable region.
For reference, claim 1 recites:
1. An isolated recombinant anti-TNF-α antibody or antigen-binding fragment thereof, said antibody or antigen-binding fragment comprising a human constant region, wherein said antibody or antigen binding fragment (i) competitively inhibits binding of A2 (ATCC Accession No. PTA-7045) to human TNF-α, and (ii) binds to a neutralizing epitope of human TNF-α in vivo with an affinity of at least 1x108 liter/mole, measured as an association constant (Ka), as determined by Scatchard analysis.
The Federal Circuit framed the analysis as whether the 1994 applications provide adequate written description for an antibody to human TNF-α with
- a human constant region,
- a human variable region,
- high affinity for human TNF-α,
- neutralizing activity, and
- the ability to bind to TNF-α in the same place as Centocor’s A2 mouse antibody (“A2 specificity”).
As noted by the court, the "overwhelming majority" of the disclosures of both the ’775 patent itself and the 1994 applications "describe the A2 mouse antibody and the single chimeric antibody that Centocor made based on A2’s mouse variable region."
Although the disclosures broadly discuss antibodies with human variable regions, "the specification does not describe a single antibody that satisfies the claim limitations." As found by the Federal Circuit,
[Centocor's application] does not disclose any relevant identifying characteristics for such fully-human antibodies or even a single human variable region. . . . Nor does it disclose any relationship between the human TNF-α protein, the known mouse variable region that satisfies the critical claim limitations, and potential human variable regions that will satisfy the claim limitations. . . . There is nothing in the specification that conveys to one of skill in the art that Centocor possessed fully-human antibodies or human variable regions that fall within the boundaries of the asserted claims.
The court cited "undisputed" testimony from Abbott's expert witness to the effect that the disclosed murine variable region sequence "does not serve as a stepping stone to identifying a human variable region within the scope of the claims." Thus, the court found:
the asserted claims constitute a wish list of properties that a fully-human, therapeutic TNF-α antibody should have . . . . The specification at best describes a plan for making fully-human antibodies and then identifying those that satisfy the claim limitations.
The court stated:
At the time the 1994 CIP applications were filed, it was entirely possible that that no fully-human antibody existed that satisfied the claims.
Although the court noted that "the written description requirement does not demand either examples or an actual reduction to practice," it explained that the specification must at least "demonstrate constructive possession." Because Centocor's 1994 applications at best set forth "a description of the problem to be solved" and "a wish or a plan" for doing so, the jury verdict finding adequate written description support was not supported by substantial evidence.
A Cautionary Tale
Although language in the Federal Circuit decision suggests that Centocor's problem stemmed from its priority claim to an earlier application with a different disclosure, I found little relevant differences between the disclosure of the '775 patent and the disclosure of the 1994 CIP applications (such a U.S. 5,698,195). Some may think that, in the wake of Ariad, the court applied a more strict interpretation of § 112 than was in effect when the patent was granted. But, this theory is undermined by the § 112 rejections that were raised in the CIP applications, where the claims were limited to antibodies with non-human variable regions.
I think the real moral to this story is that the Federal Circuit is taking the written description requirement seriously, and will evaluate compliance based on evidence within "the four corners of the specification." If the disclosure does not amount to a "constructive possession" of the subject matter at issue, patent holders may find their patents vulnerable to invalidity challenges, regardless of how much guidance is provided as to how someone could carry out the steps that should lead to the subject matter at issue.