Amgen v. Sanofi and Points Beyond

May 30, 2023


By: Jerry Cohen, Esq.


A week ago Thursday, the Supreme Court issued its decision in the Amgen v. Sanofi case, affirming the judgment of the Court of Appeals for the Federal Circuit, that the claims of the two patents Amgen asserted against Sanofi are invalid for lack of enablement. In Amgen, the Supreme Court reiterated an old but simple “statutory command” that has endured “largely intact” since the original Patent Act of 1790 – the broader you claim the more you must disclose to enable the full scope of the claims. While anxiously awaited by the life sciences industry, the result might not have been a surprise, especially considering the direction the case law has been taking us. The Amgen decision joins a confluence of jurisprudence about 25 years in the making (arguably beginning with Regents of the University of California Eli Lilly & Co.) that along with changes in the patent laws brought about by the America Invents Act, are reining in patent rights in life sciences, and particularly those concerning cutting edge, first-in-class breakthroughs, from virtually every statutory angle. Collectively, they point patent applicants in this industry in a single inexorable direction – file for patents limited to the embodiments you have made. 

Summary of Technology and Company’s Marketed Products

The present dispute between the two industrial giants, Amgen and Sanofi, involved biological therapeutics known as monoclonal antibodies. The two patents at issue, both owned by Amgen, relate to therapeutic antibodies that help reduce levels of low-density lipoprotein (LDL) cholesterol. The antibodies target a naturally occurring protein known as PCSK9. This protein degrades LDL receptors that extract LDL cholesterol from the bloodstream. The antibodies are designed to bind PCSK9 and prevent it from binding LDL receptors, thus allowing them to remove LDL cholesterol from the bloodstream. Amgen and Sanofi have their own FDA-approved anti-PCSK9 antibodies, and market them under the tradenames Repatha® and Praluent®, respectively. Both parties have other patents (not involved in the present dispute) that more specifically cover their respective products.      

Procedural Posture

Amgen filed suit against Sanofi in the District Court for the District of Delaware for patent infringement of two of its other patents with broader claims, namely U.S. Patents 8,829,165 and 8,859,741. The asserted claims at issue of the two patents are not directed to any one or more particular anti-PCSK9 antibodies. Rather, the claims are directed to an entire genus of antibodies, described strictly in functional terms, that 1) bind to specific amino acid residue(s) on PCSK9 (which Amgen refers to as the “sweet spot”), and 2) block PCSK9 binding to LDL receptors. The claims broadly cover Sanofi’s Praluent®. Characterizing Amgen’s patent disclosures as providing little more than a trial-and-error process in order to discover the millions of antibodies covered by the claims but not disclosed in the specification, Sanofi argued that the asserted claims were not valid for failing to meet the enablement standard. The district court agreed with Sanofi. The Federal Circuit Court of Appeals affirmed that judgment.         

Amgen filed a petition for a writ of certiorari with two questions presented, the second of which is whether enablement is governed by the statutory requirement that the specification teach those skilled in the art how to “make and use” the claimed invention 35 U.S.C. § 112, or whether it must instead enable those skilled in the art “to reach the full scope of the claimed embodiments” without undue experimentation – i.e., to cumulatively identify and make all or nearly all embodiments of the invention without substantial “time and effort.”  The United States Supreme Court granted certiorari on November 4, 2022, limited to that question.

The Supreme Court Decision

In a unanimous opinion delivered by Justice Gorsuch on May 18, the Supreme Court answered unequivocally as follows: “And we agree with the lower courts that Amgen has failed to enable all that it has claimed, even allowing for a reasonable degree of experimentation.” (emphasis added) Although expressing agreement in principle with Amgen that “there is one enablement standard,” the Court refused to accept Amgen’s argument that the Federal Circuit “raised the bar” for enablement of claims that like Amgen’s encompass an entire genus of embodiments. “Instead, we understand that court to have recognized only that the more a party claims for itself the more it must enable.” This fundamental tenet of the Constitutional quid pro quo reverberates throughout the opinion.   

Consistent with its observation that while Congress has revised the patent laws over time, the “enablement” obligation set forth in the original Patent Act of 1790 has endured “largely intact,” the Court relied heavily on its own precedent, the earliest of which dates back to 1846. Two such cases dealt with the telegraph Samuel Morse’s patent telegraphy method and Thomas Edison’s incandescent lamp that used bamboo as an incandescent material, that allegedly infringed Sawyer and Man’s patent directed to an electric lamp with an incandescing conductor made of carbonized fibrous or textile material. The Supreme Court invalidated both patents based on lack of enablement. These decisions, in the view of the Court, “reinforce the simple statutory command……the specification must enable the full scope of the invention as defined by its claims. The more one claims, the more one must enable.” (emphasis added) The breakthrough nature of the 21st century technology at issue mattered not – “[w]hile the technologies in these older cases may seem a world away from the antibody treatments of today, the decisions are no less instructive for it.  The Court went even further, stating that “while the technology at the heart of this case is thoroughly modern, from the law’s perspective, Amgen’s claims bear more than a passing resemblance to those the Court faced long ago…”

In the view of the Court, Amgen failed to meet its self-created “challenge” by seeking to “claim sovereignty over [an] entire kingdom” of antibodies. In particular, the Court found that the guidance provided in the patents’ common specification, which aside from the amino acid sequences of 26 anti-PCSK9 antibodies, included a “roadmap” and a proposal for “conservative substitution,” amounted to “little more than two research assignments” and did not constitute enablement at all. Rather, the Court likened them to the Sawyer and Man patent that offered a mere “hunting license” while Thomas Edison had to engage in “painstaking experimentation to see what really works.” 

The Aftermath

Now with all that established not only for cutting edge fields of innovation endeavor but all fields, where do we go from here? The Amgen decision reiterates an old but simple message. 

Coupled with other Supreme Court precedents on enablement, it reaffirms the futility of any patent applicant attempting a patent “land grab.” Using claims that define a new composition of matter strictly or substantially in terms of function adds further layers of complication. Most notably, this decision promises far-reaching effects on the antibody field and other analogous fields of limited predictability at any given time. How else could Amgen have cornered the market based on its innovative discoveries?  What else or different could it have done to ensure the enforceability of the claims? All three courts in Amgen were troubled by the “double function” (i.e., binding and blocking) recited in the claims. The Federal Circuit found that the binding limitation alone required undue experimentation. In essence, the pair of recitations raised the enablement bar, and exponentially at that. None of the three courts found that the guidance provided in the specification, especially the “roadmap,” came anywhere close to fulfilling the enablement requirement. Amgen’s roadmap turned out to be a misnomer. The disclosure of 26 anti-PCSK9 antibody sequences paled in comparison to the “millions and millions” of antibodies embraced by the claims at issue, conjuring memories of Carl Sagan’s (mis)quotes and book “Billions and Billions….”  But delaying filing on account of identifying others, especially in the absence of a bright line rule setting forth a threshold number or proportion of embodiments, which there will never be, would have come at significant risk in a post-AIA patent world. Would eliminating the binding limitation (assuming that PCSK9 has one and only one “sweet spot”) and relegating that to the disclosure have eased the burden? Most likely not. On the other hand, recitation of a common structure, e.g., complementarity determining regions (CDRs), or describing them in terms of a consensus sequence showing amino acid residues that are constant in the CDRs, might have made the roadmap actually readable and reliable. Apparently, not even the CDRs of the 26 sequences are amenable to characterization in this manner.

Claiming less broadly and including disclosure regarding “equivalence”, to better position enforcement of the patent from the standpoint of the doctrine of equivalents (DOE) offers another alternative. This approach risks a finding that the disclosure constitutes a dedication to the public, which would foreclose a successful assertion of infringement via the Doctrine of Equivalents. While some large pharma companies employ a “claim the label” or “picture claim” strategy effectively, its universality is questionable. Submitting patent applications with narrow claims is highly problematic for start-ups in life sciences whose sustainability largely depends on attracting investment capital based on broad patent application filings. Broad filings, which are also the norm for not-for-profits, notably research institutes and universities, inevitably result in narrowing claim amendments during prosecution for reasons related to patentability that, absent various exceptions, result in forfeiture of application of DOE and recapture of the surrendered subject matter. For these customers of the Patent Office, DOE is basically a non-consideration.

The 1952 allowability of means-plus-function claims now set forth at 35 U.S.C. § 112(f) narrows claims (rather than broadening them per past aspirations). Their scope is limited to specification-disclosed embodiments and equivalents and can be indefinite if no embodiments are disclosed. 

Yet another possibility entails the use of Jepson claiming, whereby the patent applicant admits that the claim preamble is prior art. Therefore, the addition of a Jepson preamble to the specification might make a difference in establishing that it demonstrates that the inventors were in possession of the full scope of the claimed invention.  Such an approach might have resulted in a different outcome in Juno v Kite (10 F.4d 1330 (Fed. Cir. 2021)), where Kite was successful in invalidating Juno’s U.S. Patent 7,446,190, based on lack of written description. The ‘190 patent broadly covers second-generation CAR Ts that are improved relative to first-gen CAR Ts in terms of including a specific co-stimulatory region. Ironically, the focus of the litigation was not on that “point of novelty,” but rather on the binding element of the CAR, which is broadly recited in claim 1 as “specifically interacts with a selected target,” and further defined in various dependent claims at issue as “a single antibody,” (i.e., a scFv) and “binds to CD19.”  Kite’s CAR T product, YESCARTA, targets CD19 expressed on large B-cell lymphoma cells. By convincing the Federal Circuit that the patent disclosure failed to demonstrate that the inventors were in possession of the full scope of possible CD19 scFvs (about five or so sequences were known at or around the time of filing), Kite got out from under an almost 10-figure judgment and a hefty running royalty rate. Last November, while granting Amgen’s Petition, the Supreme Court denied Juno’s Petition for Writ of Certiorari, and in January, denied its Petition for Rehearing. Given that Juno’s specification describes one-fifth of the population of anti-CD19 scFv’s known at about the time of filing, a relatively high number, and the Federal Circuit’s indication that its prior decision, Capon v. Eshhar, was not controlling authority, the Supreme Court’s denial of Juno’s Petition for Certiorari is disappointing.   

Juno’s CAR Ts included both new and old elements, at least insofar as CAR T is concerned. Given the nature of the “improvement” at hand, which allows the CAR T cells to mount a more potent and prolonged attack on the cancer cells, the use of Jepson claims is intriguing. Unlike Juno, the Amgen case concerned the enablement standard.  Although closely related to written description to the point where the lines of distinction become blurred, we are regularly reminded that they are distinct statutory requirements. Although one could envision resurrecting Jepson claims in the context of inventions directed to “improvements” in existing compositions of matter or in new combinations of old elements, it is hard to point to anything old in Amgen’s anti-PCSK9 antibodies or any new chemical compound or biological that would prompt serious consideration of using this type of claim. 

Yet another strategy involves filing multiple continuation or divisional applications, each with claims of varying scope. Coincidentally, the U.S. Patent Trademark Office is proposing to significantly increase filing fees for continuations filed at certain time points after the parental filing, e.g., three and six years later. Potential cost concerns aside, creation of patent “thickets” may create non-statutory double patenting issues that render one or more patents in a family vulnerable to validity attack, particularly drug patents that benefit from both patent term adjustment (PTA, based on Patent Office delay), and a patent term extension (PTE, based on delays by the FDA in approving a drug product). In this regard, the decision of the Federal Circuit in the pending appeal of In re Cellect, 2021 WL 5755316 (December 1, 2021) might provide greater clarity from the standpoint of whether the terms of patents extended by patent term adjustment are vulnerable to validity attacks on the basis of obviousness-type double patenting, in the absence of a timely filed terminal disclaimer.  

Patent thickets may also be vulnerable to these types of attacks if the progeny patents do not get the benefit of the shield or safe harbor afforded by Section 121. For example, filing a divisional application from a continuation of a parent in which the restriction requirement was first issued, wherein the divisional is filed after the issuance of the parent, has been held as a forfeiture of the safe harbor. See, Ex parte Sauerberg, 2017 WL 150016 (PTAB, Jan. 10, 2017). Although PTAB designated Sauerberg as non-precedential, patent prudence weighs in favor of filing a first divisional application before the issuance of the ultimate parent in which the restriction (or election of species) was first issued.   

Amgen becomes the newest authority in a growing body of patent jurisprudence that has significantly affected patent rights with respect to subject matter eligibility, obviousness, enablement and written description. Through the years the Court and Congress have protected the quid pro quo requirement by such acts as: (a) allowing means-plus-function recitation in a multi-element claim but per 35 U.S.C. § 112(f) discussed above confining its product or method steps disclosed in the specification (or drawings) and equivalents thereof; (b) closely limiting the Doctrine of Equivalents; (c) limited times for reissue revision of claims; (d) new matter preclusion; (e) strict lines of demarcation for claims of divisional applications after a restriction requirement per 35 U.S.C. § 121; (f) a baseline of effective filing date for the beginning of a fixed patent term; and (g) expanding the definition of prior art. All these reflect a continuing recalibration of patent rights not just in life sciences and pharma, but in all fields.    

The great bargain of the patent system admirably limned in the Amgen decision has worked well throughout the history of U.S. patent system and similar patent systems of other countries. It has served well through scope of exclusive right calibration of quid pro quo as to products and processes emerging from physics, chemical, biological and mathematical sciences. Despite the alarmist predictions of some, the enablement jurisprudence has worked well and will continue to do so in a dynamic patent system that balances incentive and reward with the public interest. 

receive news & alerts

Yes, I’d like to receive updates with firm news and insights that are relevant to me.