Federal Circuit Delivers Billion-Dollar Lesson on Written Description

By Robert Asher. Co-Chair of the Patent Practice Group

March 2011 IP Update

For patent owners, patiently biding time in the PTO with repeated continuation filings, waiting for a hook on which to catch infringers, is a common practice. Here is a story in which patience pays off, but the BIG one gets away. The Federal Circuit demonstrates its greater willingness to use the written description requirement to place limits on the freewheeling practice of filing continuations.

Centocor Ortho Biotech, Inc. and New York University (collectively, “Centocor”) file a patent application in March 1991. Centocor then proceeds to file continuation-in-part after continuation-in-part (“CIP”). Finally, a couple of divisional applications are filed. A total of twelve patent applications are filed, at least six of which become patents.

The patent applications track Centocor’s development of a mouse antibody for tumor necrosis factor α (“TNF-α”). Overproduction of TNF-α can cause autoimmune conditions such as arthritis.

Centocor proceeds to develop and patent a chimeric (part mouse and part human) antibody.  The chimeric antibody has a mouse variable region and a human constant region. It solves the problem of achieving an antibody that binds with and neutralizes human TNF-α with an acceptable reduced level of immunogenic reaction.

Centocor jumps on a new opportunity to expand its chimeric antibody patent family in 2001. Competitor Abbott Laboratories receives regulatory approval and begins marketing a different technology to likewise neutralize human TNF-α. Whereas Centocor made use of mouse regions in its antibody, Abbott developed a fully human antibody, which it sold under the name Humira.

Soon after Abbott comes out with its product, Centocor files a continuation application, its thirteenth patent application in the family. Centocor’s patent applications had mentioned the possibility of using a human variable region in the antibody but it did not disclose one that would work. Indeed, Centocor only made the single chimeric antibody.

Nevertheless, in the new continuation application, Centocor for the first time drafts claims to cover a fully human antibody so as to cover Abbott’s Humira. To Abbott’s chagrin, the PTO allows the claims and a patent issues in 2006.

A few months later, Centocor files a lawsuit in the reputedly patent-friendly venue of the Eastern District of Texas. Centocor survives claim construction and motion practice to win the opportunity to stand before the Texas jury. The jury rules that Abbott infringes the patent. In view of Abbott’s highly profitable sales, the damages award is $1.67 billion, the largest patent jury verdict in history.

To hold onto its big catch, Centocor needed only to successfully defend its Texas verdict in the Federal Circuit Court of Appeals. The claims in the thirteenth patent application had to find adequate support in Centocor’s 1994 CIP patent application. The claims would otherwise be invalidated by Abbott’s activities, which included a patent application on the fully human antibody filed back in 1996, which resulted in a patent to Abbott.

In order for Centocor’s after-the-fact claims to benefit from the filing date of its earlier-filed CIP application, that application must provide a written description demonstrating that Centocor was in possession of the claimed invention back in 1994.

The Federal Circuit ruled that Centocor could not show in its earlier CIP application that it had invented the claimed invention prior to the filing of Abbott’s patent application. The 2006 patent claims covered new subject matter that had not been discussed in the 1994 patent application. The court found those claims invalid, overturning the billion-dollar verdict and depriving Centocor of the blocking power that its patent filing strategy had been designed to achieve.

In the end, the Federal Circuit concluded, “The scope of Centocor’s right to exclude cannot overreach the scope of [its] contribution to the field of art as described in the patent specification.” The invention of a chimeric antibody for binding to the known human TNF-α did not support patent claims, added to a chain of patent applications years later, to cover a fully human antibody.