Quiet Breakout: A Liver-Hijacking Pill Just Picked Up Novartis

In April, the US Patent Office issued grant number 12,605,451 to Novartis. The title was deliberately dry: “Targeted plasma protein degradation.” The chemistry is anything but. The molecule has two ends. One end clips onto PCSK9, the cholesterol-raising protein that the injectable drug Repatha was built to inhibit. The other end clips onto a sugar receptor that is expressed almost nowhere in the human body except the surface of liver cells. In between is a linker. The whole assembly does something that would have sounded absurd in a pharmacology seminar five years ago: it tricks the liver into eating proteins that are floating loose in the bloodstream.

Novartis is the fourth identifiable player in a patent race that, until now, has been quiet enough that most pharma-watchers haven’t noticed it is happening. The other three are Lycia Therapeutics, a Stanford-adjacent startup founded by 2022 Nobel laureate Carolyn Bertozzi. Avilar Therapeutics, a Waltham biotech funded by RA Capital. And Yale University, whose chemistry department is currently suing Avilar in federal court for stealing the idea in the first place.

All four are working on the same physical thing. The category has a name: LYTACs, short for lysosome-targeting chimeras. Bertozzi’s lab coined it in a Nature paper in July 2020. The paper now has over 1,000 citations in OpenAlex. It has also produced, as best the patent record can show, exactly zero clinical trials and four lawsuits’ worth of arguments about who gets the spoils.

What the molecule actually does

A traditional small-molecule drug works by sitting in the active site of a protein and blocking it. Statins do this to HMG-CoA reductase. Aspirin does it to COX. The problem is that roughly forty percent of disease-relevant human proteins don’t have a tidy active site to block. They are scaffolds, receptors, secreted hormones, or rogue antibodies whose job is to drift around and bind something. You can’t really turn them off. You have to remove them.

For intracellular proteins, the industry already solved this. The technology is called PROTAC. A PROTAC is a bifunctional small molecule: one end grabs the target protein, the other end grabs an E3 ubiquitin ligase, and the cell’s own disposal machinery does the rest. Arvinas’s vepdegestrant, branded VEPPANU, became the first FDA-approved PROTAC earlier this year. It degrades the estrogen receptor in breast cancer patients with ESR1 mutations. The trial that supported approval was published in The New England Journal of Medicine.

PROTACs only work on proteins inside cells, because the E3 ligase machinery is inside cells. Everything in the bloodstream — antibodies, cytokines, growth factors, lipoprotein cargo — was off-limits. For those, you needed an antibody drug, which has to be infused, manufactured in giant bioreactor farms, and engineered around the half-life of human IgG.

What Bertozzi’s group showed in 2020 was that you could take a small molecule, hang an extracellular protein on one end of it, and hang on the other end a sugar that grabs the asialoglycoprotein receptor — a galactose-recognizing receptor that hepatocytes use to clear glycoproteins from circulation. The receptor swallows whatever you handed it. The whole package gets sent to the lysosome, the cell’s acidic shredder. The target gets digested. The bifunctional molecule walks back out, finds another copy of the target, and does it again. It is, in chemical terms, catalytic.

The liver, in other words, becomes a leasable demolition crew, available for any target you can write a binder for.

Reading the patent claims

The grant filings make the mechanism unambiguous. Avilar’s foundational patent, US 11,819,551, granted in November 2023, claims “compounds and compositions that have an asialoglycoprotein receptor (ASGPR) binding ligand bound to an extracellular protein binding ligand for the selective degradation of the target extracellular protein in vivo.” The company has since stacked six follow-on grants, including specific filings for the degradation of immunoglobulins, the family of proteins that drives most autoimmune disease.

Yale’s recent grant, US 12,364,766, issued July 2025, claims the same architecture: a circulating protein binding moiety, a linker, and a cellular receptor binding moiety that targets a degrading cell. The named inventors are David Spiegel, who runs the Yale chemistry lab where the technology was developed, and two of his postdocs. The disclosed targets include PCSK9, TNF-α, MIF, and VEGF — a list that covers cardiovascular disease, rheumatoid arthritis, lupus, and several oncology indications.

Novartis’s April 2026 grant claims a similar bifunctional architecture, again targeting the asialoglycoprotein receptor, and specifically calls out PCSK9 and FHR3 as proof-of-concept targets. PCSK9 is the protein that Amgen’s Repatha and Regeneron’s Praluent are designed to neutralize, and that Novartis’s own siRNA drug Leqvio silences upstream. Owning a small molecule that removes PCSK9 from the blood would put Novartis in competition with two of its own franchises. They filed anyway.

Lycia, the Bertozzi-founded company, has only one issued US grant so far — number 12,258,332, from March 2025 — and is keeping its pipeline quiet under a 2021 collaboration with Eli Lilly worth up to $1.6 billion in milestones. Lilly’s stated interest was pain and immunology.

The lawsuit

Avilar exists because, according to the complaint filed in the District of Delaware in 2023, RA Capital signed a confidentiality agreement with Yale in April 2019 to evaluate licensing Spiegel’s MoDE-A technology. Talks broke down that August. RA Capital then incorporated a shell company called DegraderCo, renamed it Avilar, and began filing patents that Yale and its commercial partner Biohaven argue read directly on the trade secrets disclosed under the CDA. In May 2025, Judge Jennifer Hall denied Avilar’s motion to dismiss the trade-secrets and breach-of-contract counts. The case is proceeding to discovery.

The legal fight is not incidental to the technology story. It is the technology story. The Bertozzi paper described a class of molecules; whoever owns the freedom to make and sell them gets to compete for a market that, if it works, replaces several lines of injected biologics. Avilar’s published pipeline already includes an oral ApoB degrader for atherosclerosis — a category Amgen serves with monthly injections — and an oral IgG degrader for autoimmune disease, which would put it in collision with argenx’s Vyvgart, the infused antibody fragment that booked $4.2 billion in global sales in 2025. None of these molecules has entered a human trial, according to ClinicalTrials.gov, which lists no studies under the keywords LYTAC, MoDE-A, or ATAC. The whole field is still preclinical, which means the patent estate is the asset.

Who cares and why

The reader who should be paying attention is whoever underwrites antibody-drug capacity. Antibody manufacturing is the most expensive, slowest, and most physically constrained part of modern pharma. Bioreactor capacity is measured in years of waiting list, and the molecules themselves require cold chain and intravenous administration. A small-molecule LYTAC that does the same job — sequester an immunoglobulin, shut down a cytokine, clear a lipoprotein — gets made in the same plants that make ordinary pills. It ships at room temperature. It is taken orally. It costs, in commodity-chemistry terms, roughly nothing.

That is the bet Novartis just placed. It is the bet Lilly placed in 2021 when it signed the Lycia deal. It is the bet the four litigants and Biohaven are paying their lawyers to win. The interesting question isn’t whether the technology works in animals; Bertozzi’s original paper and Avilar’s follow-on disclosures already established that. The interesting question is whether the existing biologics franchises can survive an oral competitor that targets the same proteins through the patient’s own liver.

The first human trial is not yet scheduled. The first injunction motion, on the other hand, is probably six months out.

Method

Patent counts come from a search of issued US utility grants in the USPTO bulk grant record for assignees Avilar Therapeutics, Lycia Therapeutics, Yale University, and Novartis AG whose abstracts describe bifunctional molecules combining an extracellular protein-binding ligand with an ASGPR ligand for degradation. Avilar’s count of seven issued US grants and Lycia’s count of one reflect filings explicitly assigned to those companies; Yale’s filing referenced is US 12,364,766 (inventors Spiegel et al.); Novartis’s is US 12,605,451. Counts include only granted patents, not pending applications, and reflect the public record through May 22, 2026. Literature figures, including the 1,027-citation count for Bertozzi’s 2020 Nature paper, are from OpenAlex. Clinical trial absence is verified against the ClinicalTrials.gov bulk record. The Avilar-Yale-Biohaven litigation history is drawn from public filings in Biohaven Therapeutics Ltd. et al. v. Avilar Therapeutics, Inc., 1:23-cv-00328, D. Del.