In 2023, the U.S. Supreme Court will hear oral argument in Amgen Inc. v. Sanofi to decide whether an inventor must describe every example in a category to get an enforceable patent for that category. Must a patent simply give enough detail to allow the reader to make and use one version or a few versions of what is claimed, or should it explain in such detail that the reader can easily achieve all or most versions of what is claimed?
Amgen has patents for drugs that treat high cholesterol. Healthy livers have receptors that bind to and remove bad cholesterol from the blood stream, but sometimes a protein (PCSK9) interferes with this by binding to those same receptors. Amgen developed drugs that preempt this, binding to PCSK9 so it doesn’t disrupt healthy liver function. Here is one of Amgen’s claims to this invention:
Although a jury had found this claim valid, judges disagreed and invalidated this claim as exceeding the scope of the patent description. The Supreme Court will now decide the question.
Should Amgen’s work to discover some liver-protective antibodies allow the company 20-year exclusivity on all antibodies that bind in this way, even if functionally similar antibodies are not described in the patent? Should competitors be allowed to avoid this patent by constructing different proteins that still bind in this way? Is it better for an inventor to quickly describe and race to the patent office with a new breakthrough, even if all relevant permutations have not been described? Commentators think there is legitimate public policy debate on such questions.
For more than a century, U.S. law has wrestled with how much description is required to support broad patent claims. When Thomas Edison was sued on an electric light bulb patent, he showed that a claim to all arched light bulb filaments formed from “carbonized fibrous or textile material” was too broad, partly because the patent description didn’t help him avoid many experiments to finally settle on bamboo. However, in a later mining case, a patent for concentrating ore by adding oil and stirring was upheld, even though a reader may still need to experiment on oil types and best ways of stirring for various ores.
The law has often labeled fields like chemistry and pharmaceuticals “unpredictable,” given the many unknown interactions and permutations of molecules, small and large. As science advances, such fields are becoming more predictable, and the average scientist has more tools to understand implications of patent descriptions. Should the patent rules be different for chemical inventions and mechanical, electronic, or software inventions?
Some decisions focus on claim phrasing (does it state specific physical qualities or aspirational benefits?). Some decisions focus on the full patent description (after reading it, how much more work is necessary to achieve claimed embodiments?). Whatever the focus, legal scholars have found a distinct trend of broad “genus” claims losing in court over the past 30 years.
Tune in to the Supreme Court’s oral argument this term to find out how the court wrestles with these questions, and whether this trend will begin to reverse.