Promega’s law team must have been celebrating last month, though no one’s admitting it. After nearly two years of waiting for a decision in its lawsuit with Roche about the pharma’s Taq polymerase patent, the US Court of Appeals for the Federal Circuit announced that it would not overturn the district court’s original decision that the patent was unenforceable, and would send the case back to that court for a final verdict.
Taq DNA polymerase, the reagent used widely for PCR, gene sequencing, and other applications, is owned by Roche through a progressive series of patents. Roche initially sued Promega for patent infringement in 1992, and the small, Madison, Wis.-based company turned around and countersued, claiming that the first patent in the estate was “unenforceable due to inequitable conduct” — in short, that it had been obtained through misrepresentation.
The district court decided in 1999 that the patent was indeed unenforceable, meaning that Roche couldn’t collect further royalties from it. (That would likely affect Applied Biosystems as well; ABI is the exclusive licensee to the R&D side of the PCR and Taq patents from Roche.) At the time, Promega estimated the global market of Taq sales at $200 million per year.
Predictably, Roche appealed the decision, which went to court in May 2001 and was finally released on March 31 of this year. With one judge dissenting, the appeals court upheld two of the three claims for unenforceability, and returned the case to district court Judge Vaughn Walker in San Francisco to write the final verdict, which is expected to adhere closely to the findings of the appeals court.
One of the key components of the court’s finding of inequitable conduct relied on what’s referred to as “Example VI” — a detailed protocol listed in the patent application with certain results that showed the enzyme was different enough from previous enzymes to be patentable. However, Cetus scientist David Gelfand, who was on the original Taq team, admitted that the protocol had never actually been performed that way, nor had those particular results been obtained from the experiments that were done.
Roche argued that the fact that the sentence listing results was written in the present tense (the rest of the protocol was described in the past tense) would lead any reader to understand that the experiment was simply an illustration of something that could have been done. “That argument,” wrote the majority for the appeals court, “is entirely unpersuasive.”
The dissenting judge argued that all three claims for inequitable conduct should be overturned, allowing the patent to stand. Roche lawyers were not available for comment at press time.
Now it’s back to work for both companies. For Promega, the decision means little change, at least immediately: “We had always taken the position that the Taq patent is invalid,” says Brenda Furlow, general counsel for Promega, explaining that the company had never paid royalties to Roche in the first place. “We’re following the law as we believe it to be.”
And while Promega won’t talk about future plans, the fall of the first patent could mean open season on the rest of the estate — for Promega or any other company looking to take on Roche.
— Meredith Salisbury