On December 9, 2020, the Federal Circuit heard oral arguments on the validity of Amgen’s patents (U.S. Patent Nos. 8,829,165 and 8,859,741) on cholesterol-lowering drug Repatha. Specifically, the question came down to whether the patents, claiming a genus of antibodies by their functional properties, are enabled under 35 U.S.C § 112. The panel consisted of Judges Sharon Prost, Alan D. Lourie, and Todd M. Hughes. While we are still awaiting the Court’s decision, the oral argument strongly suggested that the Federal Circuit is likely to continue its trend of holding functional genus claims invalid under Section 112.
Claim 1 of Amgen’s ’165 patent, which is representative, recites “an isolated monoclonal antibody, wherein, when bound to PCSK9, the monoclonal antibody binds to at least one of [certain listed amino acid residues], and wherein the monoclonal antibody blocks binding of PCSK9 to LDLR.” Amgen explains that its claims cover antibodies that bind a certain epitope on the PSK9 protein, and that its inventors discovered that antibodies that bind this “sweet spot” will block PSK9 from binding to LDL receptors, ultimately lowering cholesterol.
Within seconds after counsel for Amgen began his argument, Judge Lourie stepped in and pointed out that the patent claims are composition of matter claims defined by function (in fact, two separate functions – “binding” and “blocking”) rather than structure. Given the breadth of this functional language, and the district court’s finding that “a substantial amount of time and effort” would be required to determine the scope of the claims, Judge Lourie stated that “that sounds like undue experimentation,” making it a “pretty uphill battle” for Amgen to overcome.
Amgen’s counsel responded that the claim is directed to a specific structure, namely the specific arrangement of “sweet spot” binding sites in the form of a 3D structure, which results in the functional “blocking.” However, Judge Lourie seemed unconvinced that the claims were not purely based on function.
Judge Prost then addressed the issue of “roadmap predictability,” referring to Amgen’s argument that the specification provides a roadmap enabling those skilled in the art to make and use the full scope of antibodies within the claim. Judge Prost asserted that there was no direction or guidance as to whether a specific antibody will bind, or exactly how to make the 400 distinct antibodies that Amgen argues fall within the claimed genus. She stated that only 26 of the antibodies were described in the specification, with only nine of those antibodies exhibiting successful binding to the “sweet spot.” As to the other 384 antibodies, she noted that there was simply no evidence of what those antibodies may be, and that they may very likely cover competitor antibodies whose structures are quite different from those described in the specification.
Counsel for Sanofi continued with this “numbers” argument, questioning whether 400 antibodies was even an accurate scope of the claimed genus. Citing the testimony of a lead inventor and an expert who stated that they didn’t know how many antibodies the genus would cover, and noting the many different permutations based on the dual functionality and the effects of a change of just a single amino acid, he argued that the number is more likely in the millions, or “astronomical.” Using an analogy to the California gold rush (“the fact that you knew that there were gold in the hills and that you knew you had a pan to find it, doesn’t mean that you’re entitled to gold on every square mile of the California countryside”), he argued that determining the scope of the claims in this case is a classic example of undue experimentation.
Judge Hughes poised an interesting question about why the number of experiments matters if the test being conducted is the same across all samples and is not qualitatively difficult. He also stated that he was having difficulty distinguishing the Wands case from the facts at issue here. Counsel for Sanofi responded that repeated testing, regardless of whether that test is the same each time, is just labor intensive, to such a degree that no scientist would consider doing it. He also argued that the “hit rate” (or amount of antibodies found to be within the genus) was of particular importance – in Wands, the hit rate was nearly 50%, whereas here, he argued, it’s searching for a needle in a haystack.
Overall, the oral arguments provide insight into the Judges’ views and shed some light on the future of antibody genus claims. The Court seems likely to follow its decisions in Idenix and Wyeth, where functionally-defined claims to a genus of small molecules were held invalid under Section 112, extending this precedent into the antibody field.[1] The USPTO has already made it difficult for patent owners to obtain claims to a functionally defined genus of antibodies. We expect that patent owners claiming a novel genus of antibodies will need to recite structural features such as specific complementary determining region (CDR) sequences, or perhaps a percentage of identity to a given amino acid sequence. Hopefully, the Federal Circuit’s opinion will offer some guidance to this effect, but biotech companies should not expect new support from the Federal Circuit for functional genus claims. Given that the breadth of acceptable claims for antibodies is narrowing, patent attorneys will need to explore other options for drafting enforceable claims to continue to motivate innovation in the field.
[1] Idenix Pharm. LLC v. Gilead Scis. Inc., 941 F.3d 1149 (Fed. Cir. 2019), cert denied 2021 WL 161021 (Jan. 19, 2021); Wyeth & Cordis Corp. v. Abbott Labs., 720 F.3d 1380 (Fed. Cir. 2013).