In a development having significant ramifications for many biopharma companies, the Federal Circuit has denied Amgen’s petition for rehearing en banc a decision invalidating a genus patent covering its antibody cholesterol treatment, Repatha®. Amgen’s petition, joined by several amici, argued that the Federal Circuit’s heightened enablement standard under section 112 of the Patent Act improperly invalidates virtually all genus claims having functional limitations, and that enablement should be a question of fact, not law.1 In a non-precedential per curiam order issued last week, the court declined to rehear the case without stating its reasons. The members of the original panel (Judges Lourie, Prost, and Hughes), however, provided a seven-page opinion concurring in the denial of rehearing. As well as providing a vigorous defense of its original decision, the panel’s opinion supporting the denial of rehearing provides important guidance on the claiming of a chemical or biological genus.
Amgen’s patent claims at issue cover a genus of monoclonal antibodies that purportedly bind to the protein PCSK9 and lower low-density lipoprotein (“LDL”) cholesterol levels by blocking PCSK9 from binding to LDL receptors. Like many other antibody patents, Amgen’s claims were written in functional terms, requiring that, when bound to PCSK9, the antibodies (1) bind to at least one of a list of residues of a particular sequence, and (2) block binding of PCSK9 to the LDL receptors. After weighing the so-called Wands factors, the Federal Circuit panel held Amgen’s patent specification “did not enable preparation of the full scope of these double-function claims without undue experimentation.”2 In its rehearing petition, Amgen argued that the panel had applied a new enablement test that “does not examine the effort required to find any embodiment, but the effort to find every embodiment.”3 According to Amgen, this new test is legally unsupported because section 112 does not provide a “separate or heightened ‘full scope of claimed embodiments’ test for claims with functional limitations.”4
In their opinion concurring in the denial of Amgen’s rehearing petition, the original Amgen panel members stated that they had not created a “new” test for enablement, commenting that a genus claim has always had to be “enabled by a disclosure commensurate with the scope of the genus.”5 Where such genus claims cover compositions not actually reduced to practice, the panel advised, they must be “described constructively . . . with procedures and names of resultant compositions, as with chemical compositions.”6 The panel stated that enablement requires invention and that “[d]rawing a broad fence around subject matter, without filling in the holes, is not inventing the genus.”7 The panel rejected the argument forwarded by various amici that the panel’s decision introduced a “numbers-based standard,” stating that the “problem was not simply that the claimed genus was numerous—it was that it was so broad, extending far beyond the examples and guidance provided.”8
The panel pushed back on the “sky is falling” view of Amgen and amici that the panel’s decision would “devastate” incentives to invest in drug discovery.9 Rather, the panel stated, Amgen was the one suppressing innovation by “asserting such broad, unsupported claims . . . by trying to control what it has not invented.”10 The court noted that Amgen has another patent that covers Amgen’s Repatha® antibody, and that the doctrine of equivalents already protects against compositions that are “so close” to the species disclosed and claimed by a first entrant.11
With respect to functional claiming, the panel cautioned: “[i]t is not the law that one can put forth an idea, or a result or function, and claim all methods of achieving it; one cannot claim everything that works.”12 With respect to functional means claim limitations, the court advised:
This court has already considered the impact of functional means claim limitations on whether a disclosure is commensurate in scope with the claim. The answer is that single means claims claim too much. See In re Hyatt, 708 F.2d 712, 714 (Fed. Cir. 1983) (Rich, J.) (“The proper statutory basis for the rejection of a single means claim is the requirement of the first paragraph of § 112 that the enabling disclosure of the specification be commensurate in scope with the claim under consideration. The long-recognized problem with a single means claim is that it covers every conceivable means for achieving the stated result, while the specification discloses at most only those means known to the inventor.”). Multiple means claims simply compound the problem.13
Regarding Amgen’s second argument for rehearing—that enablement should be a question of fact—the panel members stated that it was bound by precedent to treat enablement as a question of law, and Amgen offered no compelling reason to change that standard. In doing so, the panel members stated that enablement, like claim construction and obviousness (both questions of law), “involves interpreting the specification and the scope of the claims.”14 The panel did not specifically address why enablement should be treated differently than written description, which also involves interpreting the specification and the scope of the claims.
The practice of protecting biopharma discoveries, such as antibodies, with broad functional genus claims appears to be largely dead, at least for the foreseeable future. The original Amgen panel’s strong defense of its heightened enablement standard for such claims, and the fact the denial of Amgen’s rehearing request by the Federal Circuit as a whole was “non-precedential” signals the court is not inclined to lower the enablement bar for such claims anytime soon. Further, it appears unlikely that the Supreme Court will grant cert. considering its recent denial of cert. in Idenix v. Gilead, a case that invalidated Idenix’s functional genus claims under section 112.15 Patentees should, therefore, consider alternative ways to protect their biopharma inventions, including where possible filing genus claims that describe compounds in structural terms rather than purely in functional terms. Where genus claims do include functional language, patentees must take special care to provide sufficient actual and constructive examples in the patent specification to support the breadth of the claim.