Enabled to Claim the Unknown? Federal Circuit applies Amgen v. Sanofi to invalidate broad antibody claims

by Dennis Crouch
Baxalta Inc. v. Genentech, Inc. , No. 22-1461, — F.4th — (Fed. Cir. Sept. 20, 2023).
The Federal Circuit recently affirmed a finding of invalidity for lack of enablement in Baxalta v. Genentech , holding that functional antibody claims were insufficiently enabled under 35 U.S.C. §112. Relying heavily on the Supreme Court’s Amgen v. Sanofi decision, the appellate panel found Baxalta’s disclosure to be analogous to Amgen’s “roadmap for further research” rather than a complete teaching of the invention. This result again calls into question the viability of broad functional claims post- Amgen and highlights the importance of structurally defining inventions, especially in unpredictable arts like biotechnology.
As mentioned, the case has close parallels and to the Supreme Court’s recent decision in Amgen Inc. v. Sanofi , 598 U.S. 594 (2023).  Amgen involved a functional-claimed antibody designed to inhibit PCSK9 activity.  The Supreme Court found that the broad claims encompassed a large number of antibody configurations while only describing a limited number of those, along with a trial-and-error roadmap for finding the additional embodiments.  Although some experimentation is permissible, a patentee cannot satisfy the enablement requirement by assigning a research project.  Ultimately, the Court concluded that Amgen had failed to enable the “full scope” of the claimed invention and that the claims were therefore properly held invalid as a matter of law.
As in Amgen , Baxalta’s patent also includes a functionally claimed isolated antibody:
1. An isolated antibody or antibody fragment thereof that binds Factor IX or Factor IXa and increases the procoagulant activity of Factor IXa.
‘590 Patent, Claim 1.

The invention involves a treatment for Hemophilia A to promote blood coagulation.  As you’ll note by looking at Claim 1, it does not claim any particular amino acid sequence or antibody structure, but rather claims all antibodies capable of binding Factor IX or IXa in a way that increases  its procoagulant activity.  The inventors went through a process of injecting Factor IX into mice, and collecting resulting antibody producing white blood cells (B-cells). Those B-cells were then fused with myloma (cancer) cells to promote rapid growth even in vitro . As those fused B cells continued their work, the resulting antibodies were screened to determine which ones increased the procoagulant activity.  Although this process is tedious, it was a familiar and known process at the time of the invention. The result was several hundred antibodies generally identified through the screening, although only 11 were sequenced and disclosed in the patent application.
Genentech’s accused product (Emicizumab) was not one of the 11 disclosed.  The court also noted that, unlike the disclosed antibodies, Emicizumab is a humanized bispecific antibody and would not have been identified through the patentee’s research approach. One note here, unlike in Amgen , the Baxalta patent is not limited to “monoclonal” antibodies even though the disclosed research protocol would only result that type of antibody.
The Patent Act requires that the patentee describe how to make and use the invention:
the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains . . . to make and use the same.
35 U.S.C. § 112(a). Whether or not a particular invention has been enabled is seen as a question of law based upon underlying questions of fact.  Although courts would typically call this a “mixed question” of law and fact, the conceptual separation allows courts more leeway in preemptively deciding the legal question without having to submit the inquiry to the finder of fact (typically a jury).
In this case, Federal Circuit Judge Timothy Dyk served as the district court judge — sitting by designation in D.Del.  In a 2022 decision, Judge Dyk granted summary judgment of invalidity for lack of enablement. On appeal, the Federal Circuit affirmed — finding the facts of the case “materially indistinguishable from those in Amgen .”
There are millions of potential candidate antibodies, but the written description discloses the amino acid sequences for only eleven antibodies with the two claimed functions. . . Just like the roadmap rejected by the Supreme
Court in Amgen, the ’590 patent’s roadmap simply directs skilled artisans to engage in the same iterative, trial-and-error process the inventors followed to discover the eleven antibodies they elected to disclose.
Slip Op. Enablement of a broad genus can also be accomplished by disclosing common qualities of the genus. These are typically structural qualities, but the court here notes that the might include “other” features that help delineate the class.  An example might involve the mechanism of binding, for instance. The court found none of those in the disclosure.
The only guidance the patent provides is “to create a wide range of candidate antibodies and then screen each to see which happen to bind” to Factor IX/IXa and increase procoagulant activity. Amgen makes clear that such an instruction, without more, is not enough to enable the broad functional genus claims at issue here.
Id . (Quoting Amgen ).  One question following Amgen is whether the case disturbed the Wands Factors associated with identifying “undue experimentation.” Here, the court wrote that there is “no meaningful difference between Wands ’ ‘undue experimentation’ and Amgen ’s ‘[un]reasonable experimentation’ standards.”  One difficulty with that result though is that the Wands Factors are generally questions of fact that, if in dispute, preclude summary judgment.  Here though, the court concluded that the case was so analogous to Amgen that summary judgment was appropriate — the effective implied conclusion is that there were not factual allegations presented by the patentee material enough to shift the outcome.
Baxalta reinforces that claimed inventions should be defined structurally insofar as possible, rather than solely by function. While claim breadth is undoubtedly desirable, applicants must now enable the full scope or risk invalidation post- Amgen . Even fields accustomed to broad functional claiming may require narrowing to structurally-defined elements, especially where some aspect of the design remains unpredictable.  A functional limit is potentially problematic any time a claim element has “millions” of possible embodiment designs.  On major caveat here of course is that these two cases are biotech focus and the USPTO in particular has largely resisted strongly expanding enablement doctrine outside of the chemical and biotech arts.  Another caveat of these pair of cases is that neither are combination claims. Rather, each simply claim a single component – the antibody.  I can imagine a situation where more leeway is given to functional components in a combination patent where substantial structure is provided by the combination, including other components in the combination.  Finally, also note that 112(f) is available for combination inventions to claim individual elements in means-plus-function form. That approach should typically save the claim validity at the cost of limited scope.