Structure of bivalirudin, via Wikimedia Commons.

Life is short, the art long, opportunity fleeting, experiment perilous, decision difficult.

The world has changed beyond measure since Hippocrates laid down that maxim 2400 years ago. Yet that fundamental difficulty of knowing what to expect in the context of medicine was at the heart of a cutting-edge pharmaceutical patent decision that Judge Amy J. St. Eve handed down in the Northern District of Illinois on October 27.

The patent in suit was US Patent 7,582,727 , filed in 2008, which covers “pharmaceutical formulations of bivalirudin and processes of making the same.” Specifically, the patent covers a way of producing it with a reduced level of the common “Asp9” impurity, in which the ninth amino acid is degraded from asparagine to aspartic acid. Bivalirudin is used as an anticoagulant for IVs.

The plaintiff and patent owner was The Medicines Company (TMC) of New Jersey, which sells bivalirudin under the brand name Angiomax . TMC is the sole seller of bivalirudin, and naturally would like to stay that way as long as possible.

The defendant was Pennsylvania-based generic drugmaker Mylan Pharmaceuticals , which had filed an Abbreviated New Drug Application (ANDA) for generic bivalirudin with the FDA.

An ANDA is part of the Hatch-Waxman framework for simplifying pharmaceutical patent litigation: when a generic drugmaker files an ANDA, the original and generic drugmakers can fight the patent issues out in court before any actual damages arise.

Several other drugmakers had also filed ANDAs for generic bivalirudin, but most settled quietly in return for a chance to start selling their generic in 2019 rather than 2028. Unlike those cases, and unlike another patent case we covered recently, this one went all the way to trial.

Mylan argued the patent was anticipated because TMC had previously produced at least one batch that was within the claimed purity range. But Judge St. Eve found that this batch was not representative. After all, as Hippocrates could have told us, the outcomes of compounding medicines are variable. One batch may come out with low impurities, and the next one may not.

Mylan next argued the patent was obvious because any person of skill in the art would have known that they needed to use TMC’s technique (lowered pH) in order to slow the reaction that caused the Asp9 impurity. But Hippocrates won the day again. Judge St. Eve pointed to Federal Circuit precedent stating that “in the medical arts ‘potential solutions are less likely to be genuinely predictable,’ as compared with other arts.” TMC’s technique might have been one possibility among dozens for addressing the impurity problem, but there was no reason to think it was especially likely to work—until it did.

Mylan finally argued the patent was not enabled , because even a person of skill in the art would have had to engage in undue experimentation in order to achieve the claimed level of purity. It pointed to erratic production results: a contractor had produced a batch with high impurity levels even though the contractor claimed to have used TMC’s method. But here, Judge St. Eve ruled, a different kind of uncertainty afflicted Mylan’s case: there was no reason to believe that the contractor had really followed the assigned procedures.

Having thus found the patent valid, Judge St. Eve turned to whether Mylan’s proposal would infringe on it. Here the court leaned heavily on a Federal Circuit precedent, Sunovion Pharmaceuticals v. Teva Pharmaceuticals USA . Under Sunovion, an ANDA infringes the original drugmaker’s patent whenever the ANDA would permit the generic manufacturer to make an infringing drug—even if the ANDA also allows a wide range of noninfringing drugs, and even if the generic manufacturer promises not to infringe. In this case, TMC’s patent covered Asp9 impurity levels from 0 to 0.6%, and Mylan’s ANDA covered impurity levels from 0 to 2.0%. Because Mylan’s ANDA did not specifically exclude the 0 to 0.6% range, Judge St. Eve concluded, it ineluctably infringed TMC’s patent.

This outcome goes to show the paradoxical state of pharmaceutical patent law today. The original bivalirudin patent expires next month, but under the Sunovion rule, competing drugmakers seemingly cannot bring generic versions to market unless their ANDA specifically disclaims the purity range staked out by the newer patent.

And it is surely no coincidence that no drugmaker has been able to clear that hurdle to date. After all, the FDA is unlikely to look kindly on a proposal for a minimum level of impurities in an injectable drug.

In any event, although TMC was victorious in Illinois, its trials are not yet over. In a Delaware case earlier this year, drugmaker Hospira eked out a win: the District of Delaware found TMC’s patents valid but not infringed. That case is now on appeal to the Federal Circuit.