In Appeal no. 21-257 (May 18, 2023), the Supreme Court affirmed the judgement of the Fed. Cir. by ruling that all of the claims of Amgen’s U.S. Patent nos. 8,829,165 and 8,859, 741 were invalid due to their failure to meet the enablement requirement of s. 112(a). For an introduction of the landmark decision, please re-read my post of Feb. 13, 2021 which summarizes the issues addressed by the Fed. Cir. Judge Lourie summarized the claims as “claiming a monopoly over all antibodies that (1) bind to specific proteins on a naturally occurring protein known as PCSK9, and (2) block PCSK9 from impairing the body’s mechanism for impairing the body’s mechanism to remove LDL cholesterol from the blood stream.”

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

The “sweet spot”, a term the Court adapted, is no more than a binding site of 15 residues on PCSK for the MCA that would permit the MCA to block PCSK9’s interaction with LDLR binding. Both Amgen and Sanofi had MCA products that competed for a share of the cholesterol reduction market. Sanofi also pursued method of treatment claims, but their fate would appear to be tied to Amgen’s. While the Court described Amgen’s general techniques as (1) generate MCA’s [the Court did not discuss MCA’s generated in, e.g., mice, using PSK9 as the antigen], (2) test the antibodies to see if they would bind to PCSK9, (3) see if any of the MCA’s would bind to one or more of the ‘‘sweet spots” and (4) test the antibody/PCSK9 complexes would block PCSK9 from binding to LDL receptors. The Court also noted Amgen’s conservative substitute test (not exactly an untried technique comprising replacing single amino acid residues on known functional MCA’s with residues that might increase or decrease binding).

While Sanofi argued that millions of MCA’s might function as claimed, they also stated that “those methods require scientists to engage in little more than a trial-and-error process of discovery.” While millions of functional antibodies may exist—no one knows—trial and error isolation and evaluation has always been the basis for drug discovery. When I was a pharmaceutical chemist, our group made 100 analogs of PGE1 and PGE2 that were screened for useful bioactivities. Since then, prostaglandin F-alpha was modified and yielded at least a family of 8 drugs that are used to treat ocular hypertension. I feel that the Court just collapsed in the face of all this science, but the Fed. Cir. also bought the millions and millions argument as if was the only flaw in Amgen’s case. Oddly the Court endorsed Tilghman and Proctor as not involving undue experimentation, even though the specification “suggests a trial … with different degrees of heat so as to ascertain that which is best for each particular kind of fat.”

Although there is no statutory requirement that working examples must be described in the patent, Amgen has identified 26 MCA’s that would work and even studied their 3D structures. The Court held that Amgen had not provided “adequate guidance” to make and use the claimed antibodies “beyond the narrow scope of the working examples. A lengthy discussion of the Morse and the light bulb litigation follows, but I won’t discuss them, since you all know how they turned out (or off). Things get darker when the Court uses the Fed. Cir.’s language that “the specification must enable the full scope of the invention as defined by its claims. The more one claims the more one must enable.” This is the test that Amgen and many amici argued amounted to raising the bar of the enablement requirement.

The Court added insult to injury by endorsing the “undue experimentation” test for the enablement requirement: “Amgen has failed to enable all that it has claimed, even allowing for a reasonable degree of experimentation.” No guidance comes along to help anyone determine what a “reasonable degree of experimentation” entails. The opinion only endorses the meaningless requirement that “the specification [need not] always describe with particularity how to make and use every single embodiment within a claimed class.” However, in the next paragraph the Court states that “it may suffice to give an example … if the specification also discloses ‘some general quality … that gives it ‘a peculiar fitness for the particular purpose.’”

The Court’s emphasis on “undue experimentation” leans heavily on Wands factors 1 and 8; so heavily I would argue that the factors for meeting the enablement requirement posited in Ariad look rather paltry to the extent that meeting the enablement requirement may subsume the written description requirement. However the Court keeps the WDR at a distance and quotes it only when it quotes s. 112(a). For example, at page 15 of the slip. op., the Court cites Mowry v. Whitney for the proposition that “[w]hat is reasonable in any case will depend on the nature of the invention and the underlying art….(“'(The definiteness of a specification must vary with the nature of its subject. Addressed as it is to [POSA], it may leave something to their skill in applying the invention.)’ … [C]ourts  cannot detract from the basic statutory requirement that a patent’s specification describe the invention ‘in such full, clear, concise, and exact terms as to enable any [POSA]’ to ‘make and use’ the invention [citing s. 112(a).” Although the Court begins with factor 6 of the Wands factors, the quote goes on to suggest that it is the specification that must describe how to make and use the invention. Page 16 of the slip opinion states that the more a party claims, the more it must enable. In other words the WDR do no more than meet the standards for making and using. Then its work is done.

Clearly the Court had to learn something about biotech, not its favorite subject area, and it leaned heavily on expert testimony. Perhaps it will use this knowledge to take another look at Ariad.