While this recent Fed. Cir. decision – Amgen Inc. v. Sanofi, Appeal No. 2020-1074 (Fed. Cir., Feb. 11, 2021) seems predictable, given the fate of antibody claims that recite the target and the function of antibody binding thereto, there are a few interesting wrinkles to comment on. A Fed. Cir. panel of Judges Lourie, Prost and Hughes, Lourie writing, affirmed the district courts finding that Amgen’s claims failed the enablement test of s. 112. Evaluating the Wands factors—there are seven of them—to determine whether or not the claimed antibodies would require undue experimentation to locate ones falling within the scope of the claims, the panel relied on its fairly extensive precedent, most of it unfavorable to Amgen, and concluded that “lack of enablement [was found] due to the undue experimentation required to make and use the full scope of the claimed compounds that require a particular structure and functionality.” As the panel noted, citing McRO 959 F.3d at 100:

“In cases involving claims that state certain structural requirements and also require performance of some function (e.g., efficacy for a certain purpose), we have explained that undue experimentation can include undue experimentation in identifying, from among the many concretely identified compounds that meet the structural requirements, the compounds that satisfy the functional requirements…While functional claim limitations are not necessarily precluded in claims that meet the enablement requirement, such limitations pose high hurdles in fulfilling the enablement requirement for claims with broad functional language.”

Let’s back up a minute. The Patents-in-Suit are U.S. Pat. Nos. 8,829,165 and 8,859,741. They are based on the discovery that the ability of a protein PCSK9 to bind to and disable LDL receptors can be blocked by antibodies that bind to and block PCSK9 and, thus, normalize LDL action. (LDL receptors remove LDL cholesterol from the blood.) The specifications disclose amino acid sequences for 26 antibodies, including one, “21B12,” that is marketed as Repatha^©. The specifications disclose the 3D structures of 21B12 and 31H4 and show how they bind to PCSK9. But the main claims at issue varied in breadth:

1. (of ‘165) An isolated monoclonal antibody (“MCA”), wherein when bound to PCSK9, the MCA binds to at least one of the following residues: S153 [and 15 other residues] of SEQ ID NO. 3, and wherein the MCA blocks binding of PCSK9 to LDLR.

7. (of ‘741) An isolated MCA that binds to PCSK9, wherein the isolated MCA bind an [functional] epitope on PCSK9 comprising at least one of residues 235 or 238 of SEQ ID NO: 3, and where the MCA blocks the binding of PCSK9 to LDLR.

Judge Lourie writes: “The claimed MCA’s are defined by their function: binding to a combination of sites (residues) on the PCSK protein, in a range from one residue to all of them, and blocking the PCSK9/LDL interaction.” While Amgen argued that expert testimony “[showed] that a POSA can make all antibodies within the scope of the claims by following a roadmap using anchor antibodies and well known screening techniques as described in the specification or by making conservative amino acid substitutions in the twenty-six examples.” This evidence was apparently outweighed by Sanofi’s relatively simple argument that the claims cover millions of antibodies. Sanofi argued that Amgen focused on the “number of MCA’s actually known to satisfy the claims, when this court’s precedents require examining the number of candidates that must be made and tested to determine whether they satisfy the claimed function.” This argument is right in Judge Lourie’s wheelhouse, and he quotes his own analogy in AbbVie Deutschland v. Janssen Biotech: “If the genus is analogized to a plot of land, the disclosed species and guidance ‘only abide in a corner of the genus.’”

While Judge Lourie terms the claims “double-function claims” that are not defined by structure, this seems a shaky characterization. There is structure in the claims that defines the binding sites on SEQ ID NO 3. Furthermore, I think that the recitation of the specific binding sites implies the nature of the structures on the MCA’s that would bind to them. While I am sympathetic with Amgen’s arguments that the screening necessary to identify workable antibodies would not be undue these days, the “antibody rule” that biotech practitioners long depended upon—a defined binding site and a method to screen the entities that bind thereto, is not sufficient to satisfy s. 112—has been dead for some time.

However, I don’t think that Claim 7 is “broader than the enabling disclosure”. The functional MCA’s need only to bind to epitopes defined by 3 marker combinations on PCSK9. More than three MCA’s may meet this requirement but I don’t see why it would require undue experimentation to identify MCA’s that both bind to the epitope(s) so as to block PCSK binding to LDLR.  Is claim 7 unreasonably broad in a patent that does have working examples, even if they are “narrow in scope”? While Lourie tries to cabin the undue effort holding by writing “We do not hold that the effort required to exhaust a genus is dispositive”, he contradicts himself in the same paragraph by concluding that “no reasonable jury could conclude under these facts that anything but ‘substantial time and effort’ would be required to reach the full scope of claimed embodiments.” We are left to ponder just how full functional biotech claims must be to survive this test.