Stanford v. Roche (Fed. Cir. 2009)

Stanford v. Roche (Fed. Cir. 2009) 08-1509r.pdf (PDF via www.patentlyo.com).

Using federal NIH funding, Stanford scientists developed and patented a method of using PCR to measure HIV virus concentration in blood plasma. The development was done in partnership with the PCR maker Cetus. The Cetus PCR business was later purchased by Roche, and Stanford subsequently offered to license the patent rights to Roche (for a substantial royalty). After negotiations stalled, Stanford sued Roche for patent infringement. Roche claims that it has ownership rights or - at least - shop rights to the patents based on its acquisition of Cetus's PCR assets. The district court (N.D. Cal.) rejected Roche's claim of rights, but did find the patents invalid as obvious (on summary judgment).

The Fed. Cir. made a number of findings about the nature of Roche's defense as well as the validity of the patent itself. What I found interesting was the discussion on assignment of rights.

Roche provides some interesting practice notes for both the small litigator as well as the patent prosecutor. In fact, small law firms which combine both (such as my own) would do well to study Roche for all the practical morsels of legal knowledge that the Fed. Cir. is serving up. Roche is a good example of the conventional wisdom that the "little things" end up being a big litigated issue. In Roche that "little thing" was the often over-looked portion of client / new matter application intake, the assignment. Assignment intake and filing are generally a routine matter, something that takes a back seat to IDSs and application drafting. However, Roche shows the importance of not only a) making sure that the client provides the full scope of information for the assignment in take form, but b) that the practitioner ask some follow up questions to make sure that the client is not "on notice" about rights transfers, and if they are "on notice", that suitable steps be taken to resolve the assignment issues.

As a first issue the Fed. Cir. reaffirmed that contractual issues surrounding assignments of patents are intimately tied into issues of standing, and hence, are subject matter ripe for federal review. As a more practical matter, Roche brings to light is something that every employment / contract agreement drafter should be aware. The assignment between the inventor and Stanford and the Inventor and Roche (by way Cetus) determined the order of the transfer of ownership rights, even though the temporal order and the transfer order of relationships was reversed.

The Fed. Cir. held:


Recordation of Rights: Patent ownership regularly transfers without the new owner recording the assignment with the USPTO. Recordation is not required, but it does offer some benefits. Notably, where two entities both claim ownership (as there are in this case), a second assignee can hold title if it records first.

A [prior] assignment . . . shall be void as against any subsequent purchaser or mortgagee for a valuable consideration, without notice, unless it is recorded in the Patent and Trademark Office within three months from its date or prior to the date of such subsequent purchase or mortgage. 35 U.S.C. 261.

One caveat in the statute is that the subsequent assignee must be a bona fide purchaser - i.e., take "without notice" and pay "valuable consideration." Here, the Federal Circuit found that Stanford was at least on inquiry notice of the relationship between the inventor and Cetus and the potential for a rights transfer.

Here, the inventor "agreed to assigned" to assign his patent rights to Stanford. Roche at 11. The Court interpreted this as requiring a subsequent written instrument in order to effect the transfer.

This is in contrast to the Cetus transfer which stated "do hereby assign", which effected a present assignment of the future invention without requirement of a subsequent additional written instrument. Roche at 12,

The lesson learned by Stanford and its attorneys is to make sure that if a recordation of the assignment is taking place at the time desired by the parties. To do this, you must insure that the proper terms of art are being used to effect immediate transfer.

Conversely, if immediate transfer is not desired, we now have several Fed Cir decisions that will allow for a proper time-line of assignments to be drafted.

I think a lot of licensees and contractual drafters are going to be copying and pasting "do here by assign" on to all of their new drafts.

(GA)

Martek Biosciences V. Lexicographer

Martek v. Nutrinova presents an interesting case for the burgeoning lexicographer. The term at issue is "animal" and the question is whether or not Humans are covered by said term. The majority held that an applicant, when they have not expressly defined limits to a term (i.e. animals = all animals -humans and kittens), there is no inherent exclusion to members of the term. (If you claim "Rock Bands" then "Creed" is coverd by that disclosure. Arguments before the CAFC that "Creed is christian Rock, Not rock rock" will get you nowhere, since you never limited Rock Bands to "good rock bands").

The majority goes on to point out that regardless of if the preferred embodiments in the specification appear to limit the definition, preferred embodiments are just subsections that have been highlighted for the overall disclosure, and not limiting on the overall specification.

Judge Rader dissents, finding that the totality of the specification should be used to determine if there is a inherent limitation in the lexicographer's definition. Specifically, the usage of the term "raised" with respect to animal made it so the term animal could never apply to people.

Brief aside: Rader here is making a moral / western civilization post modern argument that would not exist 200 years ago. While it is true that Humans are not raised (in the same manner as livestock) at the moment, it was not always the case. The Chattel Slavery system was designed to treat people as live stock, and it is possible that this specific invention would have been useful to someone who wanted to raise Omega-3 intensive individuals. While I agree with Rader's assertions that the term raised makes the definition of Human suspect, I am not a huge fan of using moral arguments to craft borders on the term.


Regardless of the whether Rader and the dissent is correct, the case provides a helpful bit of practical drafting guidelines. It makes sense, if you are going to be you own lexicographer, to put the broadest definition possible in your specification. However, if the broad reading needs some specific tweaking, that tweaking should be done in the preferred embodiments, and not in the general specification. In Martek's case, the reference to raised animals could have been placed in a preferred embodiment clause that spoke directly to the raising of animals. In this way the argument would not have rested on if "raised" modified "animal", but if perferred embodiments can limit general disclosurs. (one is settled law, and one was up for interpertation).

I have often remarked that I draft applications starting with the specification, but always draft the preferred embodiments last. Doing it this way, it is possible to pick out some of the areas that are going to give you problems down with the Examiner, and hopefully shore them up.

Willian H. Conner Inn of Court

On April 15, 2008, I had the great honor of participating in a Moot Appellate Court presentation for William H. Conner Inn of Court. The Conner Inn of Court presented the continuing saga of the Wright v. Curtis patent dispute. The program called for an Appellate level oral argument by both sides, including standard rebuttals and even a Sur Rebuttal. The participants, themes, and analysis will be discussed in depth in a future post. Stay tuned.