Gilead Sciences Inc. dodged a bullet two years ago when U.S. District Judge Beth Labson Freeman wiped out a $200 million patent infringement verdict based on opponent Merck & Co.'s unclean hands.

On Monday, the U.S. Court of Appeals for the Federal Circuit put Gilead back on the firing line.

Judge Richard Taranto grilled Gilead attorney Juanita Brooks of Fish & Richardson over a 95-minute hearing. He repeatedly asked whether the conduct of a former Merck patent prosecutor—whom everyone seemed to agree had breached an ethical fire wall and then lied about it at his deposition—actually had a material impact on Merck's patents or the trial's outcome.

Taranto also accused Brooks of taking the patent prosecutor Philippe Durette's testimony out of context. Brooks protested that her reading was exactly the same as Freeman's. “To that extent, that finding is clearly erroneous,” Taranto replied.

On the other hand, Judge Raymond Clevenger sounded as incensed by Merck's conduct as Freeman. “From the perspective of your adversary and of the court,” Durette “sullied the environment. He fouled the ground of the courthouse.” Why, he asked, isn't the appropriate response, “You're out of here. Come back [in] some other case?”

That appeared to leave Judge Raymond Chen as the deciding vote. His questioning suggested the case may come down to whether an ethically compromised attorney can be “de-tainted” once information he's learned improperly becomes a matter of public record.

Gilead Sciences v. Merck is actually small potatoes in the larger patent war between the two pharma giants over a groundbreaking hepatitis C medicine. Merck obtained a $2.5 billion infringement judgment in Delaware against Gilead based on different patents. That verdict came down in December 2016, but Chief U.S. District Judge Leonard Stark of the District of Delaware is still mulling Gilead's posttrial motion.

Still, the $200 million at stake in this case—plus a $12.5 million fee award that followed but isn't yet on appeal—is a lot of money. Gilead also faces whopping supplemental damages for posttrial infringement if Merck prevails on appeal.

➤➤ Want more IP news and commentary? Geek out with Skilled in the Art, Law.com's new email briefing on everything IP. Learn more and sign up here.

Jurors found that Gilead's sofosbuvir compound, an active ingredient in such meds as Sovaldi and Harvoni, infringed Merck's 7,105,499 and 8,481,712 patents. Gilead had obtained sofosbuvir with its 2011 acquisition of Pharmasset Inc. for $11 billion. By the time of trial, Sovaldi and Harvoni had rung up $20 billion in sales.

But Freeman had thrown out the verdict. She found that Durette had listened in with Merck's knowledge on a 2004 phone conference with Pharmasset to discuss a possible collaboration. In that call, Pharmasset disclosed the structure of the compound that would eventually become sofosbuvir.

The meeting was subject to a confidentiality agreement and intended only for Merck employees outside of the company's HCV program. Merck had been experimenting in a similar but broad area, and Durette later narrowed Merck's existing patent claims to focus more on the subgenus that includes sofosbuvir. He swore repeatedly at his deposition that he wasn't on the phone conference—then said he couldn't remember if he was on or not. Merck and its Williams & Connolly counsel didn't disclose which option Durette would choose until trial. Freeman found that Durette “fabricated testimony in this case and that Merck supported that bad-faith conduct.” All of that together rendered the patents unenforceable against Gilead, Freeman ruled.

Merck sent MoloLamken's Jeffrey Lamken to argue its appeal Monday. He acknowledged that Durette was “not a great witness. In trials you have bad witnesses. This happens.”

But, Lamken argued, untruthfulness or misconduct trigger unclean hands only if it produces an unfair outcome. And Durette didn't use what he'd learned on the call until Pharmasset publicly disclosed the structure of its compound in a 2005 patent application by researcher Jeremy Clark.

Chen seemed to imply some cynicism on Merck's part. He said the company wanted to tell a story that it was untainted by anything that happened with Pharmasset. “But once that didn't work, we'll say, 'We were completely tainted by it, but it was the Clark application.”

Lamken also argued that despite Durette's constant waffling, Merck had adequately disclosed that Durette had been on the phone call. “Sometimes there's trial reality and there's reality reality,” he told he court. “The reality is nobody except Dr. Durette ever had any doubt that he was on that call.”

Brooks had a rougher ride for Gilead. Taranto cited a “mountain of evidence” that companies in this business were looking at each others' publications and patents. Once the Clark application went public, “How in the world would Merck not have claimed that?” he asked.

Brooks argued that Merck had multiple opportunities to avoid the unclean-hands finding. It could have recused Durette from its HCV team. He could have told the truth at his deposition. “Once you're caught lying at your deposition, don't come in to court and lie to the court,” she said. “And once his attorneys know he's gonna recant, tell someone” before opening statements. Because of Durette's ever-shifting stories, cross-examining him “was like drinking from a fire hose,” Brooks said. “I had no idea what was next.”

Taranto didn't sound moved at all. He suggested that Gilead could have cleared up the confusion about Durette by deposing others. Instead, he said, both parties made strategic decisions to remain silent ahead of trial. “You [were] hoping to create this wonderful credibility problem … for a trial that otherwise would be focused elsewhere,” Taranto said. “And they [were] waiting just to spring the cleansed version—or not quite cleansed but much more clean version—at trial.”

Brooks insisted that Durette's only reason for denying being on the call was to hide Merck's improper conduct. “And they did everything in their power, and the court so held, to cover that up,” she told the judges.

But why would the improper conduct matter if everything said in the phone call made it into the Pharmasset patent application a year later, Chen asked. “Is there a way to de-taint a tainted person now, thanks to the fact that the information is free and clear for all to see, including Dr. Durette himself?” he asked.

“The answer is there is not a way to de-taint a tainted person,” Brooks replied.

“Even if Dr. Durette was not de-taintable, why wasn't Merck de-taintable?” Taranto pressed.

“They would have been if he'd recused himself,” Brooks said. “But he didn't. And that's why we're here.”