« Under Pressure | Main | Patent Retriever »

June 1, 2008


Thursday, Stanford University had three HIV-monitoring patents invalidated in summary judgment, letting Roche off the hook. Northern California Judge Marilyn Hall Patel ruled that the Stanford patents were obvious in light of a 1991 prior art article.

The patents: 5,968,730; 6,503,705; and 7,129,041, relied upon using HIV RNA as a surrogate marker for monitoring the effectiveness of treatment against HIV. The 1991 Journal of Infectious Diseases article suggested using HIV RNA as a marker.

There had been an ownership dispute. Stanford claimed that the three named inventors were employed by the university at the time of invention, thus Stanford owned the rights. But Roche had asserted the work was a joint collaboration between Stanford and Cetus. Roche claimed it had acquired the rights to the patents through purchase of a Cetus subsidiary. Moot point now.

The vast majority of patents granted prior to the landmark 2007 Supreme Court ruling known as Obzilla (KSR v. Teleflex) are invalid as obvious.

Posted by Patent Hawk at June 1, 2008 1:32 AM | Prior Art



I just found this web site that lets you download patents as PDF files for free. Its http://www.patentretriever.com/

Thought I'd share this little gem with those that are interested.


Posted by: John Segal at June 1, 2008 5:20 AM

Hawk, your last sentence is ironic, right? From where I come from, the vast majority of patents disclose a finite (but maybe only just)novel and inventive enabled contribution to the art, but also an over-broad independent claim. Are such patents "valid" or "invalid" (and if "invalid" is that because they lack a dependent claim to what is truly patentable)?

Posted by: MaxDrei at June 1, 2008 10:20 PM

Hi Max,

There was no intended irony to my last sentence. It was a reiteration of what I've written numerous times. The U.S. Obzilla standard for obviousness makes invalidating most patents via prior art like the proverbial shooting fish in a barrel.

The Patent Hawk kill rate post-KSR is around 90%.

Your statement didn't quite make sense to me. Few dependent claims are so outstanding as to escape the scythe of Obzilla.

Posted by: Patent Hawk at June 2, 2008 12:49 AM

Thanks Hawk. I agree that few dep claims survive the scythe. But that's because most of them are in there to broaden the independent claim (on a claim differentiation theory). My thought was that IF drafters were to include dependent claims directed to credible "fall back positions", THEN many more patents would escape the scythe. In a properly functioning patent system, the only cases that go to trial are those where the patent owner thinks he has a claim that is both valid and infringed, but at the same time the defendant thinks there is no room for any claim construction that can navigate between the Scylla of "infringed but also embracing something old or obvious" and the Charybdis of "Not old, not obvious, not infringed". You offer a 90% failure rate as a disaster. One in ten patent owners winning, post-KSR, strikes me as about right. When everything is perfectly in balance (not now, right?), perfectly balanced cases go to a win for the patent owner one time in four (in a 2 x 2 matrix of Validity/Infringement there are 4 possible outcomes, and only one of them is "infringed, and not invalid").

Posted by: MaxDrei at June 2, 2008 7:01 AM