The Federal Circuit has issued its opinion in Amgen v. Sanofi, a case which addressed the enablement of antibody claims containing very large numbers of embodiments that rely upon functional claiming. See Case No. 2020-1074, 2021 WL 501114 (Fed. Cir. Feb. 11, 2021).
The opinion, issued on February 11, 2021, is likely to raise some questions about the viability of functional claiming. After several recent decisions clarifying and perhaps heightening the requirements for enablement of functional claims, the Federal Circuit is continuing its trend of invalidating broad pharmaceutical claims containing very large numbers of species, drafted with functional elements.
While the Amgen opinion may render functional claiming more difficult, functional claims that follow its guidance may still have an important role to play in pharmaceutical patents.
After outlining the Federal Circuit’s decision in Amgen, this brief article identifies practical implications – as well as pitfalls to avoid – to obtain valid and enabled functional claims.
The underlying decision
In the District of Delaware, Amgen asserted that certain Sanofi antibodies infringed several Amgen patents directed to antibodies used to treat elevated levels of “bad” LDL cholesterol in the bloodstream. Certain receptors – called LDL receptors – “remove LDL cholesterol from the bloodstream.” However, a protein called PCSK9 can interfere with the action of LDL receptors by binding to and blocking their ability to regulate LDL cholesterol. Amgen’s patents claimed any antibodies that bind PCSK9, meeting two functional requirements: (1) binding at least one of fifteen specific amino acid residues of the PCSK9 protein and (2) blocking PCSK9 from binding to the LDL receptors. The patents claim any antibodies that meet these two functional requirements, whether or not the structure of these antibodies is similar to the antibodies disclosed in the patents’ specifications.
After a stipulation of infringement from Sanofi, a jury in the District of Delaware found Amgen’s patents to be valid over obviousness and enablement challenges. Following the trial, the district court granted Sanofi’s motion for judgment as a matter of law (JMOL) that Amgen’s asserted patent claims were not enabled.
On appeal, Amgen contended that its functional claims were enabled for two reasons: (1) the specification discloses “a roadmap using anchor antibodies and well-known screening techniques” that would easily identify working antibodies; and (2) a person of ordinary skill could make “conservative amino acid substitutions in the twenty-six examples” disclosed in the specification to find many other working antibodies. Between the roadmap and the examples, “Amgen contends that the embodiments in the patent are structurally representative for the purpose” of enablement.
Despite a jury verdict in Amgen’s favor, the Federal Circuit rejected Amgen’s position and affirmed the JMOL of non-enablement. The Federal Circuit framed its analysis by noting that “[e]ach appealed claim in this case is a composition claim defined, not by structure, but by meeting functional limitations,” such that it was undisputed that “millions” of antibodies would fall within the claim limitations. Thus, “[i]f the genus is analogized to a plot of land, the disclosed species and guidance ‘only abide in a corner of the genus.’” According to the panel, “the only ways…to discover” the remaining “undisclosed claimed embodiments” “would be through either ‘trial and error, by making changes to the disclosed antibodies and then screening those antibodies for the desired binding and blocking properties,’ or else ‘by discovering the antibodies de novo’ according to a randomization-and-screening ‘roadmap.’” “Either way, we agree with the district court that the required experimentation ‘would take a substantial amount of time and effort.’”
The Federal Circuit rejected Amgen’s enablement arguments for a number of reasons. It first noted that “the claims are far broader in functional diversity than the disclosed examples.” For example, “there are three claimed” amino acid residues “to which not one” of the twenty-six “disclosed examples binds.” Similarly, “although the claims include antibodies that bind up to sixteen residues, none of Amgen’s examples binds more than nine.” The Federal Circuit also cited concessions from Amgen’s experts that “testing would be required to ensure” that a given antibody would function as claimed, including a concession that it was not possible, at the relevant time, to translate the disclosures in the specification (certain amino acid sequences) into three-dimensional structures. The Federal Circuit finally concluded that the relevant art was “unpredictable,” which increased the amount of experimentation via “trial and error” required to practice the claims.
The practical implications
As an initial matter, Amgen shows that the Federal Circuit will continue to place heavy emphasis on the examples disclosed in the specification. In particular, the Federal Circuit appears likely to look favorably upon examples which do not “abide only…in a corner of the genus,” to the extent such identification is possible.
One issue to avoid is the mismatch between claims and examples which the Amgen panel seized upon. As noted above, Amgen characterized the “broad…functional diversity” of the claims by noting that certain limitations were not present in even one of “the disclosed examples.” Ensuring that each claim has at least one corresponding example which fully embodies all limitations is a safeguard and best practice post-Amgen. More broadly, patentees may also wish to identify a representative swathe of examples, not merely those which cover “a corner of the genus.”
Patentees who rely on a “roadmap” may also seek to introduce specific, factual evidence related to the “time and effort” it would take to identify a representative swathe of the genus of claimed embodiments. The Federal Circuit disagreed with Amgen’s “roadmap” for case-specific, factual reasons, including expert concessions related to the adequacy of the roadmap. Amgen also held that “the effort required to exhaust a genius” is not dispositive, so patentees can identify heuristics, methods, or tools that make sifting through a (representative sample of a) genus easier. Expert testimony to that effect may temper any expert concessions that some testing is necessary, by emphasizing that the testing is limited, inexpensive, conventional, or routine. And modern three-dimensional modelling of antibodies and receptors using current computational and crystallographic techniques may be critical to proving enablement.
Relatedly, patentees should consider introducing evidence to preemptively address the concerns underlying the Federal Circuit’s decision. While the Federal Circuit concluded that “the required experimentation ‘would take a substantial amount of time and effort,” it does not appear that either party introduced evidence quantifying what “time and effort” would qualify as substantial in the relevant industry. In future cases, patentees may choose to introduce specific evidence describing the amount of “time and effort” – as a matter of hours and dollars – to conduct the relevant testing. By confining vague prognostications about “time and effort” to a certain number of laboratory hours, or a certain number of graduate students, patentees can deemphasize the otherwise-looming specter of unlimited work. Particularly when measured against broader research investment budgets – or, in context of a jury trial, against the amount of money awarded against the defendant in damages – an efficient estimate of “time and effort” may set an outer, limited boundary for an otherwise-concerned court. Quantification evidence is an underused tool that can be employed while remaining faithful to established precedent (eg, the first In re Wands factor inquires about “the quantity of experimentation”).
Finally, patentees can implement strategies for drafting narrower claims without sacrificing important subject matter. For example, claims could be directed to single-point mutations of antibody, enzyme, or peptide sequences having a specific minimum percentage of amino-acid identity with the example sequences, where these mutations – for the most part – will not affect the claimed function. Where broad functional claiming is important, patentees should consider also drafting dependent claims or separate independent claims with varying levels of structural detail.
While some commentators have suggested that Amgen substantively heightens the bar for functional claiming of antibodies, the Federal Circuit in Amgen denied that it changed the law. The Federal Circuit reaffirmed that In re Wands – the case cited by the Federal Circuit as “go to” precedent – still governs the enablement standard. And in Wands, broad functional claims to antibodies were found enabled. But, as the Amgen court pointed out: “Wands did not proclaim that all broad claims to antibodies are necessarily enabled. Facts control.”
As a result, functional claims are still viable, but claim drafters must maintain the functional breadth of their claims commensurate in scope with the examples in their patents. Pharmaceutical patentees who seek to draft claims with functional elements still have a number of strategies for responding to concerns underlying Amgen, including the evidentiary strategies and claim drafting approaches explained above.
Thus, claims with functional elements are likely to continue to face scrutiny, but will still play an important role in pharmaceutical patenting after Amgen.