Patenting New Uses for Old Inventions

A bedrock principle of patent law is that old inventions cannot be patented. And a new use for an old invention does not render the old invention patentable. This is because patent law requires novelty-an invention must be new. But while a new use for an old invention does not make the old invention patentable, the new use itself might be patentable. In fact, new-use patents comprise a significant part of the patent landscape-particularly in pharmaceuticals, when drug companies obtain new-use patents to repurpose old drugs. This trend has fueled debates over follow-on innovation and patent quality. But there is a problem with new-use patents that has escaped the attention of legal scholars and commentators. The problem is when an inventor seeks a new-use patent for an old product that is, on close inspection, not new because the old product is really doing the same thing that it did before. This is a technical question that requires some understanding of the underlying science-how and why a result is achieved. But various evidentiary rules, biases, and perfunctory views of novelty preclude a true and accurate patentability assessment. Sometimes this leads to unwarranted patents; other times it derails meritorious inventions. This Article corrects this problem by offering a new framework for evaluating novelty in new-use patent claims. It proposes a probing novelty inquiry that would require inventors to elucidate and disclose mechanistic information to prove that a claimed new use is truly novel. Providing mechanistic information would promote patent law's disclosure function and improve patent (examination) quality. At a broader level, this Article raises the normative and theoretical question of what it means to be identical-which is what novelty is all about. It also raises policy questions about novelty's gatekeeping function and its role in promoting broader goals of the patent system.