Recent Case Law Impacting Early-Stage IP Protection

Dustin Lee

Protecting and commercializing inventions from early-stage research faces many challenges. This is particularly true for technologies emerging from universities, which have highly specialized resources yet lack direct paths to markets.[1] Early-stage technologies often show technical potential in idealized conditions, such as in a clinic or on a benchtop.[2] But researchers face lofty financial and technical challenges when translating them to more realistic, complex, and uncontrolled conditions. This manifests as the infamous Valley of Death that early-stage technologies must traverse; the journey from well-defined research proposals to undetermined markets with nebulous commercialization analyses, technical forecasts, and legal landscapes.[3] Success requires funding, resources, and expertise for each challenge.

Intellectual property attorneys and specialists excel at these complex multi-disciplinary challenges. In the legal domain, patent statutes have been nearly static[4] relative to the pace of innovation. Inventions must be useful,[5] novel,[6] and non-obvious.[7] And patent applications must be concise, clear, and enabling descriptions of the invention.[8] But interaction between the commercialization, technical, and case law landscapes is as dynamic as early-stage research. Communication among stakeholders is key to solving multi-disciplinary challenges.[9] Recent federal cases highlight the need for clear and regular communication between commercialization, technical, and IP experts to solve these challenges.

Amgen v. Sanofi—On the Map Toward Enablement

Amgen v. Sanofi[10] tunes the balance between wide claim scope and expensive testing to support a wide scope of biotech claim embodiments. Many early-stage technologies, especially biotechnologies, face the chicken-or-egg problem of lacking funding to support broadly enabling tests when commercial and professional factors necessitate early filing to raise funding. This problem can lead to creative patent applications. The Amgen Court held that a patent application’s description teachings for testing many species within a genus was insufficient to enable claims covering the genus.

Enablement case law[11] balances requiring description of every claimed embodiment against allowing shorthand techniques to fill-in descriptive blanks. Patent specifications must enable a person skilled in the art to make and use the entire claimed class.[12] But even enabled claims can still require reasonable experimentation.[13] They don’t have to describe how to make and use every single embodiment if the specification describes a common class characteristic for the claims’ purpose.[14]

Amgen turned on whether Amgen’s patent specification enabled practice of every claimed embodiment in the genus with only reasonable experimentation. The specification included a “roadmap” for determining how to identify satisfactory embodiments.[15] But the roadmap merely guided substitution and testing of millions of candidate embodiments. Compared to 26 expressly described embodiments, the Court found that the roadmap described trial-and-error experiments.[16] The court essentially ruled that the description enabled research of a genus—not an invention of the genus.

A hazy line separates enabled and unenabled claims in unpredictable technical areas. Tools like roadmaps and substitution methods can be enabling.[17] The line may be relative to the amount of experimental uncertainty. Amgen’s roadmap provided an “uncertain prospect” of reaching a claimed invention “given the state of the art.”[18] This directly relates experimental certainty to the supporting description and to the state of the art. Amgen’s roadmap within the state of the art may have limited the uncertainty for practicing some embodiments, but a person of ordinary skill would still be very uncertain for most others.

Testing that reduces uncertainty for people of ordinary skill is a critical barrier for many early-stage biotechnologies that are already funding-crunched and technically risky. Publishing academic findings improves researchers’ prospects of winning enablement-improving funding and for improving their career prospects. But filing applications too early may fall short of enablement. While delaying filings to raise funds may aid enablement but find an overgrown field of prior art. Both scenarios reduce the likelihood of protecting commercially viable embodiments.

Simple practices can help balance academic realities and enablement requirements. IP professionals should continuously consult with researchers to identify possible enablement gaps and understand how they may be filled. Patent prosecution strategies, like crafting a series of limited-scope applications that allow piecemeal enablement rather than single applications of wide scope, may fill some gaps. And engaging commercial experts early may help identify valuable embodiments to wisely spend limited funds. A policy practice is to advocate for regarding IP generation to be regarded on par with academic publication during career evaluations to reduce early filing pressures, which requires building relationships with administrators. Early and continuous collaboration between university IP stakeholders maximizes possibilities of enabling biotech inventions against academic realities.

Thaler v. Vidal and Thaler v. Perlmutter—AI as the IP Creator

Many researchers use powerful artificial intelligence tools to develop early-stage and application-specific technologies. Two cases narrowed the IP protections—patent and copyright—of technologies created with AI tools. Both cases are reminders that AI can be a great tool for creators and that attorneys should tailor protection to what people created.

Vidal (patent) and Perlmutter (copyright) cases have similar implications. For patent protection, the court restricted AI’s inventive role: AI cannot be listed as an inventor on a patent application.[19] And the copyright-cousin held that human authorship is required for copyright protection based on the plain statutory language.[20] Each case restricts IP protections of AI-created technology.

IP attorneys must understand that AI is a new class of tools that aids inventing and raises new questions. IP attorneys must ask not only “who contributed what,” but also “how do they contribute it?” Arguably, a person could use a prompt that leads AI software to generate an inventive or original work. But who made the contribution—the prompter or the promptee? Was the work an interpolation or an extrapolation of training data? And does an invention with any AI-generated elements negate IP protection of a person’s contributions? Neither court answered whether a patent may be invalid if a claim includes elements invented by an AI tool that cannot be an inventor.[21] And inventors will continue to use more advanced AI tools to generate ideas—which aren’t patentable[22]—hypotheses, and parameters for technical breakthroughs. Patent attorneys must diligently communicate with inventors to carefully answer these new questions.

Attorneys can apply existing tools thanks to newly issued guidance. The Copyright Office issued guidelines for authorship that apply existing case law.[23] Likewise, the USPTO will apply current frameworks to inventorship and subject matter eligibility.[24] This means that varying sets of independent and dependent that contain different elements from AI's contribution could reduce protection risk by providing options and backstops for inventorship and patentability. For example, some patent claims could incorporate AI’s breakthroughs with elements created by inventors. Fortunately, patent attorneys are familiar with similar claim set tactics to mitigate risk of subject matter eligibility rejections under §101.

Sonos v. Google–Strategically Building Patent Portfolios

Sonos v. Google[25] is a chilling reminder that even institutions with long-term horizons need strategies for early-stage technology portfolios. The case between powerhouse technology companies saw an equitable doctrine deny statutory infringement relief—and overturn a jury’s verdict—because of a common patent prosecution strategy.

The case began as a typical patent infringement matter that turned on an unexceptional commercial timeline. Sonos sued Google for infringing multiple patents[26] covering smart speaker technologies and a jury awarded $32 million to Sonos. But the court overturned the verdict and ruled that Sonos’ "targeted continuation" practice rendered its patents unenforceable due to equitable prosecution laches. This doctrine prevents a patent owner from enforcing a patent if their unreasonable and inexcusable patent prosecution delays are prejudicial to a defendant. The decision directly tied Sonos’ inability to justify its statutorily acceptable prosecution strategy with the equitable doctrine.

Sonos followed a typical patent prosecution strategy in its "targeted continuation" practice. Sonos filed several continuation applications claiming priority to a 2006 U.S. provisional patent application that resulted in many patents at issue.[27] The court noted that Sonos filed some applications in response to prior art rejections of other applications and filed others in response to commercial activities involving Google. For example, Sonos filed some continuation applications in 2014 after learning of Google’s commercial development. And Sonos filed some key claims as late as 2019—13 years from the earliest filing date. The court felt that the filing delays were unreasonable and inexcusable since Sonos didn’t point to a strategy that justified them. Thus, the court used equitable prosecution laches to overturn the verdict.

Filing a series of applications to navigate prior art and commercial realities is a typical patent portfolio management strategy. Patent prosecution is a non-linear process spanning years of simultaneous legal, commercial, and technical investments. These are rarely in sync or individually sufficient to eliminate overall uncertainty. Continuation practice reduces uncertainties by encouraging early public disclosure (i.e. filing) in exchange for opportunities to seek different protection scopes that make sense at different times. Each application must find support in the early filing, but applicants can choose to pursue different claims as various factors evolve, including supporting technologies, consumer preferences, and competitive analyses.

Successful patent portfolio management strategies account for evolving technical, commercial, and legal uncertainties. Aside from the Sonos court’s insistence that applicants point to a strategy that justifies an arguably reasonable and excusable delay, a well-developed strategy can reduce prosecution costs and undesirable disclosures.[28] Many early-stage stakeholders embrace less certain early filings and “targeted continuation” practice because their long-term horizon is a competitive advantage (i.e. less need to produce immediate results). These filings are still part of a unified strategy that includes technical investments (e.g., from continuing inventor research), commercial analyses, and legal risks. The resultant patent portfolio often remains commercially valuable as individual factors evolve.

Salix v. Norwich—The IP of Clinical Trials on Trial

Transferring therapeutic technologies from academic research centers to patients requires a careful balance of commercial, scientific, and legal considerations. Difficult scientific hurdles require long-term research programs and costly clinical trials. Intellectual property protection, like patents, allows technologists to recoup and benefit from those expensive investments. But timing the development, commercialization, and protection is important. Salix v Norwich[29] highlights that importance by invalidating a therapeutic’s patent protection in light of a clinical trial summary of that therapeutic.

Salix Pharmaceuticals alleged infringement of several patents covering rifaximin against Norwich Pharmaceuticals. Key patent claims in the case covered administering a high dose of rifaximin three times per day. Norwich’s defense alleged that a Phase II clinical trial summary for rifaximin and a journal article rendered the patents invalid as obvious. The summary described a protocol for administering twice-daily of rifaximin at high doses, and the article described thrice-daily of rifaximin at lower doses.[30] Although the summary was for the clinical trial, arguably an experimental phase of development, the court included it as prior art because filing it indicated a reasonable chance of success in combination with other references.[31] The court found that Salix’s claims were obvious when combining the summary’s twice-daily protocol and the article’s thrice-daily protocol of a higher dose. Thus, Norwich did not infringe upon Salix’s invalid claims.

The court’s decision results from obviousness’s analysis of legal fictions and complex facts. A patent claim is obvious if a person having ordinary skill in the art (“PHOSITA”) would be able to practice the claim based on what was known in the world when the claim was filed.[32] It leaves broad openings for creative answers to the layered questions of what was known at the time of filing, who is the PHOSITA, and what would a PHOSITA do with what was known.[33] A PHOSITA is a fictional person determined for convenience by the court to analyze technical arguments. They are rarely the inventor or an expert, but they are almost exclusively technically-inclined.

Yet the analysis to sponsor a clinical trial is far from fictional. Sponsoring a trial bets a company’s reputation and shareholder value on successfully navigating technical risk, commercial opportunity, and regulatory hurdles. Though courts have taken narrower technical views of “success.”[34] Sponsors also face logistical challenges of organizing researchers, patients, clinics, etc. Actual risks also vary depending on the type of trial and what “success” means to that sponsor at that time.[35] A go-no/go decision often rests on a risk-adjusted business value of the clinical trial. Thus, the presence of a document describing a protocol to study the efficacy of a drug is far from a purely technical factor linked to novelty or inventiveness.

Patent practitioners must translate these real-life complexities into hypothetical questions of obviousness law. Would a PHOSITA (determined by technical factors) look to a regulatory summary of a drug (filed after a commercial decision) to combine with other documents (a legal determination) to practice an invention with a reasonable likelihood of success (a moving multi-factor target)? The question lacks an analytical answer because the law lacks specific inputs for real-life commercial and regulatory factors. Practitioners must instead look to more familiar questions at other layers of obviousness analysis, like what is in the prior art and why would a PHOSITA combine the prior art.

Academic entrepreneurs are also left to familiar solutions. Working closely with technology transfer professionals with patent, commercial, and regulatory expertise can reduce common patent risks. A patent attorney can identify key technical aspects and determine the best time to file a patent application. Licensing and commercialization officers can help identify partners and experts and arrange agreements. And such a team can bridge the technical, legal, and regulatory uncertainties of commercializing academic technologies to further patient care.

New technological breakthroughs have raised new questions for protecting and commercializing inventions from early-stage research. Many new therapeutics, electronics, and AI technologies are ripe for translating from benchtops to markets. IP attorneys and specialists can use existing tools to protect these inventions along the way. But clear and regular communication is key to translating new technologies in a dynamic commercial, technical, and legal environment among multi-disciplinary collaborators.

[1]See Boudou, Justine, and Maria Roche. "Bringing Science to Market: Knowledge Foundations and Performance." Harvard Business School Working Paper, No. 24-021, October 2023. (Revised May 2024.)

[2] Understanding Private-Sector Decision Making for Early-Stage Technology Development, Prepared for Economic Assessment Office Advanced Technology Program National Institute of Standards and Technology, by Auerswald et al., Sept. 2005, at 8, https://www.nist.gov/system/files/documents/2017/05/09/gcr02-841a.pdf

[3] See Butler, D. Translational research: Crossing the valley of death. Nature 453, 840–842 (2008). https://doi.org/10.1038/453840a; Seyhan, Attila. (2019). Lost in translation: the valley of death across preclinical and clinical divide – identification of problems and overcoming obstacles. Translational Medicine Communications. 4 at 3-4. 10.1186/s41231-019-0050-7.

[4]The major post-WWII patent statutes are arguably the Patent Act of 1952 and the Leahy-Smith America Invents Act of 2011. Other statutes, such as the Bayh-Dole Act of 1980 and the Drug Price Competition and Patent Term Restoration Act of 1984, have been critical for early-stage research and litigation but left patentability generally unchanged. See Library guides: Intellectual property research: Patent statutes. Patent Statutes - Intellectual Property Research - Library Guides at Mercer University School of Law. (n.d.). https://guides.law.mercer.edu/c.php?g=1352720&p=9985798. Current copyright law is largely based on the Copyright Acts of 1909 and 1976. Other acts, such as 1998’s Sonny Bono Copyright Term Extension Act and Digital Millenium Rights Act contain provisions, inter alia, for enforcement, international harmonization, and term of copyright life. See https://www.copyright.gov/timeline/.

[5] 35 U.S.C. 101.

[6] 35 U.S.C. 102.

[7] 35 U.S.C. 103.

[8] 35 U.S.C. 112.

[9] E.g., Adam Lindgreen, C. Anthony Di Benedetto, Roderick J. Brodie, Michel van der Borgh, How to undertake great cross-disciplinary research, Industrial Marketing Management, Volume 90, 2020, Pages A1-A5, ISSN 0019-8501, https://www.sciencedirect.com/science/article/pii/S0019850120302583; and LaFrance, D.L., Weiss, M.J., Kazemi, E. et al. Multidisciplinary Teaming: Enhancing Collaboration through Increased Understanding. Behav Analysis Practice 12, 709–726 at 710 (2019). https://doi.org/10.1007/s40617-019-00331-y

[10] Amgen Inc. v. Sanofi, 598 U.S. 594 (2023).

[11] Enablement is a well understood area of patent law. Much of the Amgen Court’s support date to pre-World War I cases (e.g., see id. at 1245-1246)—including inventions involving famous inventors Samuel Morse (O’Reilly v. Morse, 15 How. 62 (1854)) and Thomas Edison (The Incandescent Lamp Patent, 159 U.S. 465 (1895)). The Enablement Clause requires an application’s specification to “enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same.” 35 USC § 112(a).

[12] Id. at 1246.

[13] Id.

[14] Id. quoting The Incandescent Lamp Patent, 159 U. S. at 475.

[15] Id. at 1247.

[16] Id.

[17] Id. at 1256.

[18] Id.

[19] Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022) (cert. denied Apr. 24, 2023).

[20] Thaler v. Perlmutter, No. CV 22-1564 (BAH), 2023 WL 5333236 (D.D.C. Aug. 18, 2023).

[21] See In re Verhoef, 888 F.3d 1362 (Fed. Cir. 2018) (holding that failing to name all inventors invalidated a patent).

[22] Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014).

[23] 88 FR 16190, Copyright Registration Guidance for Works Containing AI-Generated Materials, at 16192, (Mar. 16, 2023).

[24] 89 FR 10043, Inventorship Guidance for AI-Assisted Inventions (Feb. 13, 2024 (implementing the Pannu (Pannu v. Iolab Corp., 155 F.3d 1344, 1351 (Fed. Cir. 1998)); and 89 FR 58128, 2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence, effective on July 17, 2024.

[25] Sonos Inc. v. Google LLC, 20-06754 WHA, 2023 WL 6542320 (N.D. Cal. Oct. 6, 2023).

[26] Including United States Patent Nos. 10,848,885 and 10,469,966 filed in April 2019. Other patents were originally at issue but withdrawn for various reasons.

[27] U.S. provisional patent application No. 60/825,407 filed April 12, 2019.

[28] Notably, there may be some situations that prevent revealing these strategies to justify “delays.”

[29] Salix Pharmaceuticals, Ltd. v. Norwich Pharmaceuticals, Inc., 98 F.4th 1056 (Fed. Cir. 2024).

[30] Id. at 1061-62.

[31] Id. at 1062-63.

[32] 35 U.S.C. 103.

[33] KSR Intern. Co. v. Teleflex Inc., 127 S.Ct. 1727, 1729-30 (2007) (quoting Graham v. John Deere Co. of Kansas City, 383 U.S. 1, 17-18 (1966)).

[34] Bristol-Myers Squibb v. Teva Pharmaceuticals USA, 769 F. 3d 1339, 1355 (Fed. Cir. 2014) (stating that “[t]his is worth noting because it seems to me a serious question whether, in this case and perhaps more generally, the purely in vitro experiments on the lead compound should be deemed to establish a "reasonable" expectation of success. The success that must be reasonably expected in this case would, I think, have to be success in what motivated the investment in the research—an acceptable safety/efficacy profile for human-therapeutic use.”).

[35] See Wu C, Ono S. Exploratory Analysis of the Factors Associated With Success Rates of Confirmatory Randomized Controlled Trials in Cancer Drug Development. Clin Transl Sci. 2021 Jan;14(1):260-267. doi: 10.1111/cts.12852. Epub 2020 Aug 21. PMID: 32702190; PMCID: PMC7877835 (“A study demonstrated that the outcome of the primary end point depended heavily upon the nature of that end point.”).