Executive Summary
- Patents on cyclic peptides passed journal articles starting in 2023, and the FDA approved two oral cyclic peptides in 2026.
- A CAS analysis of 46,574 therapeutic cyclic peptides shows a field settling on large, head-to-tail cyclized rings.
- Most of these molecules break classic drug-likeness rules, and approved drugs do too.
- Making enlicitide at commercial scale took solution-phase chemistry and three biocatalytic cascades.
- The bottleneck has moved from binding to delivery and manufacturing.
What 46,574 cyclic peptides reveal about the next wave of oral drugs
In 2026, the FDA approved two oral cyclic peptides: icotrokinra from Johnson & Johnson and enlicitide from Merck, the first oral PCSK9 inhibitor. A recent CAS Insights™ webinar looked at the field from two sides: what the literature and patents say, and what it takes to make one of these molecules at scale.
- Trupti Thite (Information Scientist, CAS) shared an analysis of 46,574 cyclic peptides from the CAS Content Collection™.
- L.-C. Campeau (Associate Vice President and Head of Small Molecule Process Research and Development, Merck) walked through the synthesis of enlicitide.
- Kavita Iyer (CAS Lead Scientist, Life Sciences) moderated.
What CAS data shows about therapeutic cyclic peptides
The CAS Content Collection™ holds more than 23,000 journal articles and patents on cyclic peptides published from 2006 to 2025. Patents passed journal articles in 2023 and stayed ahead.
The scientists who build the CAS Content Collection are united by a shared purpose: accelerating discoveries that improve lives. Every day, they curate, connect, and analyze global literature, knowing their work impacts countless research projects worldwide.
From those documents, the team identified 46,574 therapeutic cyclic peptides and classified each by peptide type, cyclization type, and chemical modification. The method appears in the team’s recent paper in the Journal of Medicinal Chemistry.
Bicycle Therapeutics leads the patent assignees, followed by Bristol Myers Squibb and Merck. Cancer dominates, with more than 25,000 cyclic peptides, and Trupti flagged metabolic and neurodegenerative disorders as areas to watch.
The structures show clear preferences:
- Peptide type. Macrocycles make up about 54% of the set. Cyclic dipeptides and tripeptides come to less than 2%.
- Cyclization. Head-to-tail cyclization accounts for nearly two-thirds, since closing both ends protects against enzymes and often helps permeability.
- Modifications. N-methylation leads at 45% because it improves membrane permeability.
More than two-thirds of the molecules break five of six Lipinski and Veber criteria, and approved drugs sit in the same space. Oral delivery is the most studied route of administration and has risen over the past decade.
How was enlicitide manufactured?
L.-C. Campeau explained that peptide synthesis has changed little since Bruce Merrifield introduced solid-phase peptide synthesis (SPPS). That approach struggles with engineered cyclic peptides in three ways.
Non-canonical amino acids. Oral peptides rely on modified building blocks that can take many steps to make. One hydroxyproline building block for enlicitide once needed an 11-step route. An evolved enzyme now makes it in one step.
Waste. SPPS generates huge amounts of solvent waste at scale. Merck moved to a solution-phase, fragment-based synthesis, published in the Journal of the American Chemical Society, and made more than 750 kg of material with it.
Biocatalytic cascades. That still fell short of a commercial process. The final route uses three enzyme cascades, which halved the steps and cut process mass intensity (PMI) by another order of magnitude. A running joke on the team was that they would know they had succeeded when the whole synthesis fit on a single slide. It now does.
Why molecular design comes first for oral cyclic peptides
Kavita opened the panel with the 2023 crossover, which L.-C. read as a sign that science is moving toward patients. On why head-to-tail rings dominate, he pointed to display technologies, which usually rely on head-to-tail cyclization, so other ring types rarely show up as hits.
Whether you’re aligning modality choices to business strategy, prioritizing drug repurposing opportunities, or staying ahead of your competitors, CAS BioFinder® gives your team the data and analysis you need to make confident decisions.
Both panelists named molecular design as the biggest driver of oral delivery. “It turns out permeability and solubility still matters,” L.-C. said. Trupti put formulation second.
Final thoughts
Trupti described cyclic peptides as a balance between the selectivity of biologics and the stability of small molecules. L.-C. noted that macrocyclic peptides, two to three orders of magnitude smaller than an antibody, now deliver similar potency and selectivity.
“I think 15 years from now we’re going to look back at this era of medicinal and process chemistry and realize it was a golden era for chemists and scientists.”
- L.-C. Campeau, Merck
Watch the full webinar
For the full data set, the manufacturing story, and the complete Q&A, watch the recording on demand.
For molecular targets, the clinical pipeline, and future directions, read the CAS Insights Report, Trends and therapeutic directions for cyclic peptides in drug discovery.
Questions from the audience
Highlights from the audience Q&A:
Question: After enlicitide, what is the biggest unsolved green chemistry problem for macrocyclic peptides?
Answer: Selectivity. These molecules have many reactive groups, so L.-C. wants selective macrocyclization at scale without heavy use of protecting groups, plus better ways to make non-canonical amino acids.
Question: Is cyclization better in solid or solution phase?
Answer: Merck does nearly all of its cyclization in solution. Solid phase can help at small scale, but L.-C. has seen little of it in manufacturing.
Question: How well do in silico methods such as docking work for cyclic peptides?
Answer: L.-C. said docking works well when crystal structures are available. Trupti added that the many rotatable bonds make conformational sampling hard, so it is less effective than for small molecules.
Question: Which AI tools for peptide design and permeability prediction are changing day-to-day decisions?
Answer: Trupti said permeability and structure prediction models already help chemists filter candidates before synthesis. L.-C. noted that many models are trained on Protein Data Bank data, so they handle non-canonical amino acids poorly.
Question: What makes cyclic peptides hard to manufacture?
Answer: Size and complexity, according to L.-C. A typical small-molecule process runs 4 to 15 steps, and enlicitide runs into the twenties. Many peptides also lack crystalline intermediates.
Question: How does macrocyclic peptide manufacturing differ from traditional peptide manufacturing?
Answer: In L.-C.’s experience, it looks more like small-molecule manufacturing. Long peptides such as GLP-1 receptor agonists are made by SPPS and usually injected, while enlicitide is a daily oral pill.
Question: Beyond the top three assignees, where are other companies focusing their work?
Answer: Novartis and Roche focus on smaller cyclic peptides. Roche is pursuing antibacterial peptides, and Entrada Therapeutics antiviral ones.
Question: Which molecular targets appeared most often?
Answer: Macrocyclic peptides showed up most with PD-1, PD-L1, and RAS. MDM2 and p53 showed up with mixed-cyclization peptides.
Question: Which part of the analysis would be most worth repeating in two years?
Answer: Trupti would track whether oral bioavailability keeps turning into approvals, and whether rare cyclization types stay rare.
Question: What innovations could take biocatalysis further?
Answer: Some enzymes for enlicitide took three to four years and 20 to 25 rounds of evolution. L.-C. sees the biggest opportunity in enzymes that make bonds that nature does not.




