An Oakland company is betting millions of years of worm evolution can solve what blockbuster biologics haven't—without actually infecting anyone
The pitch sounds counterintuitive, maybe even reckless: extract molecular wisdom from parasites that have perfected immune evasion over geological time, then engineer those insights into drugs. No live worms. No intentional infections. Just the proteins.
Ditto Biosciences, a freshly minted Y Combinator alum founded in early 2025, believes the answer to autoimmune disease lies not in more potent immune suppression but in something closer to evolutionary sleight of hand. The three-person Oakland team—credentialed in bacteriophage warfare, genomic evolution, and molecular biology—is mining helminth genomes for immunomodulatory sequences, using AI to predict how those proteins fold and function, then designing therapeutic variants for conditions ranging from rheumatoid arthritis to inflammatory bowel disease.
Timing, as they say, matters. This October, the Nobel Prize in Medicine went to researchers who cracked peripheral immune tolerance and regulatory T cells, validating what a growing cohort of biotech founders already suspected: brute-force immune suppression is running out of road. Around the same time, the last major trial of whole-organism helminth therapy—actual pig whipworm eggs given to ulcerative colitis patients—failed spectacularly, posting placebo-level results after two decades of cautious optimism.
The juxtaposition is instructive. Living parasites are messy, uncontrollable, and now demonstrably ineffective in rigorous trials. But their molecular toolkit, refined across millions of years of host-parasite coevolution, remains as elegant as ever. Ditto's wager is that you can have one without the other.
The Numbers Behind the Need
Start with the patient pool. A 2024 analysis in the Journal of Clinical Investigation pegged diagnosed autoimmune prevalence at 4.6% of Americans—approximately 15 million people—over the 2011-2022 period. Notably, 34% of those patients carried multiple autoimmune diagnoses, layering Hashimoto's atop rheumatoid arthritis or lupus alongside Sjögren's syndrome. UK data from a staggering 22 million electronic health records suggest even higher incidence, hovering around 10% across 19 distinct conditions.
The National Institutes of Health, launching a coordinated research agenda for fiscal 2026-2030, cites prevalence estimates ranging from 23.5 million to 50 million Americans depending on how aggressively you count borderline cases and subclinical presentations. That range tells you something about diagnostic ambiguity—and about how large the addressable market might actually be.
Market forecasts paint aggressive growth curves. Future Market Insights projects the autoimmune therapeutics sector climbing from $168.6 billion to $226.2 billion between 2025 and 2035. Fortune Business Insights tracks the broader immunology segment from $112.3 billion this year to $228.2 billion by 2034, an 8% compound annual growth rate that assumes steady innovation and pricing power.
Except pricing power is eroding faster than analysts expected just two years ago. Adalimumab biosimilars—generic versions of AbbVie's Humira, once the world's top-selling drug—grabbed 22-23% market share by late 2024. CVS Health didn't just preference biosimilars; it dropped branded Humira entirely from its 2024 formulary, steering patients toward Hyrimoz, Cyltezo, and Simlandi. IQVIA data shows newer biosimilars reaching roughly 60% volume share within three years of launch, compressing margins for branded biologics across the immunology franchise.
Safety concerns compound the financial squeeze. JAK inhibitors like Pfizer's tofacitinib carry FDA-mandated boxed warnings for blood clots and elevated mortality at higher doses, with regulators restricting their use to patients who've already failed TNF blockers. The tightening reflects an uncomfortable truth: systemic immunosuppression trades one set of problems—joint inflammation, skin lesions, intestinal damage—for another: opportunistic infections, lymphoma risk, cardiovascular events. Some patients weather the tradeoff fine. Others cycle through therapies, chasing relief while dodging side effects.
What Parasites Know That We Don't
Parasitic worms and ticks have spent evolutionary epochs perfecting a single survival trick: live inside a host without getting killed. Hookworms colonize the human gut for years, suppressing inflammation that would otherwise expel them. Schistosomes manipulate basophils to secrete IL-4 and IL-13, cytokines that skew the host immune response away from parasite-clearing Th1 pathways. Ticks inject complement inhibitors while feeding, preventing the blood from coagulating beneath their mouthparts.
The mechanistic literature is dense with examples. Schistosome egg proteins like IPSE/alpha-1 and omega-1 hijack pattern-recognition receptor signaling. Evasins, isolated from tick saliva, potently sequester CC and CXC chemokines, dampening the inflammatory cell recruitment that would otherwise trigger itch and rejection. A 2019 Nature study documented engineered variants of Evasin-3 with enhanced sulfation and binding activity. These aren't crude toxins; they're exquisitely tuned proteins shaped by millions of years of host-parasite coevolution.
Ditto's founding team brings relevant, if unconventional, credentials. Adair Borges co-authored a 2018 Cell paper on bacteriophage cooperation and anti-CRISPR mechanisms while at UCSF's Bondy-Denomy lab, exploring molecular arms races between viruses and their bacterial hosts. Dennis Sun holds a PhD in Genetics, Genomics, Evolution, and Development from UC Berkeley. Emily C.P. Weiss rounds out the trio on the company's F6S profile, though public details about her background remain thin.
The pitch, at least on Ditto's website, is direct: "combine evolutionary genomics with AI to transform parasite proteins into safe and effective drugs." In practice, that means identifying immunomodulatory sequences across helminth genomes, predicting three-dimensional protein structures with tools like AlphaFold—which now covers over 214 million protein sequences—and engineering optimized variants for human use. Resources like WormBase ParaSite, cataloging 181 helminth species and 240 genomes, provide raw material. The recently released Evo 2 biological language model from Arc Institute and UC Berkeley, trained on 9.3 trillion nucleotides, promises to map sequence-to-function relationships across the tree of life at scales previously unimaginable.
Whether the company can execute on that vision is an open question. But the underlying biology? That's been validated, repeatedly, in animal models. The trick is translating it into something regulators will approve and patients will tolerate.
When Living Worms Fail, Extract the Code
The whole-organism approach to helminth therapy is dying a slow, expensive death—and has been for over a decade. A 2005 randomized controlled trial of Trichuris suis ova, or TSO (essentially pig whipworm eggs), in ulcerative colitis showed tantalizing early promise: 43% of treated patients improved versus 17% on placebo at 12 weeks. Coronado Biosciences raised venture capital and advanced two Phase 2 trials in Crohn's disease, TRUST-I and TRUST-II. Both failed to meet primary endpoints in 2013.
The latest blow landed last year. A double-blind Phase 2b trial called PROCTO randomized 119 ulcerative colitis patients to TSO or placebo. At 24 weeks, clinical remission rates were 30% for TSO and 34% for placebo—p=0.80, statistically indistinguishable from flipping a coin. An earlier 2017 multicenter trial in 252 Crohn's patients found no dose-response relationship for remission. A 2023 pilot study with hookworms demonstrated feasibility and safety but lacked the statistical power to make efficacy claims. Systematic reviews now cite insufficient evidence to rule out placebo effects, even as preclinical work in lupus and arthritis mouse models shows intriguing microbiome modulation and reduced autoantibody titers.
The regulatory environment has hardened in parallel. The European Food Safety Authority issued a 2019 opinion stating it "could not establish the safety" of daily ingestion of 250 viable T. suis ova as a novel food. The FDA has sent warning letters to companies marketing unapproved parasite products. The CDC's clinical guidance on helminths focuses squarely on treatment and eradication, not therapeutic infection.
One parasite-inspired protein did make it to Phase 3, offering a proof-of-concept that purified molecules can advance where live organisms cannot. Nomacopan, derived from the tick Ornithodoros moubata, inhibits both complement C5 and leukotriene B4. Akari Therapeutics secured FDA Orphan Drug, Fast Track, and Rare Pediatric Disease designations for hematopoietic stem cell transplant-associated thrombotic microangiopathy, a devastating complication in bone marrow recipients. Then in 2024, Akari merged with Peak Bio and suspended the HSCT-TMA program during portfolio reprioritization, pivoting instead to antibody-drug conjugates.
The program's trajectory—advanced clinical validation followed by strategic abandonment—illustrates both the translational feasibility of parasite-derived proteins and the financing volatility that plagues specialty indications. It also underscores a harsh reality: getting a drug to Phase 3 doesn't guarantee it will ever treat patients if the business case collapses.
The Tolerance Revolution Arrives
This year's Nobel Prize in Medicine, awarded to Brunkow, Ramsdell, and Sakaguchi for foundational work on regulatory T cells and peripheral immune tolerance, ratifies a paradigm shift already underway in clinical development. CAR-T therapies, engineered Tregs, and antigen-specific tolerance platforms are converging on a deceptively simple idea: reset the immune system rather than suppress it indefinitely.
Consider the clinical momentum. Kyverna Therapeutics' KYV-101, a CD19-targeting CAR-T borrowed conceptually from cancer immunotherapy, holds FDA Fast Track designations for myasthenia gravis, multiple sclerosis, and lupus nephritis. Interim positive data in generalized myasthenia gravis emerged earlier this year, with registrational trial design now underway. Cabaletta Bio reported early clinical signals for CABA-201 in systemic lupus erythematosus and myositis at the EULAR 2024 conference, drawing investor attention to single-dose interventions that might obviate chronic therapy.
Engineered regulatory T cell programs are proliferating across the sector. Sonoma Biotherapeutics inked an undisclosed-milestone collaboration with Regeneron targeting autoimmune diseases with proprietary Treg engineering. The company reported interim Phase 1 signals in rheumatoid arthritis this year. Quell Therapeutics advanced CAR-Treg programs for transplant tolerance and inflammatory bowel disease—AstraZeneca exercised an option on the IBD asset while preclinical work continues in RA and systemic sclerosis.
Antigen-specific tolerance platforms take a different tack, aiming to retrain the immune system's recognition apparatus. Anokion's KAN-101 delivers liver-targeted antigens to induce peripheral tolerance without systemic immunosuppression. In January 2025, the company announced positive symptom data from a Phase 2 celiac disease trial, building on encouraging Phase 1b/2 updates presented at Digestive Disease Week 2024. Selecta Biosciences and Swedish Orphan Biovitrum's SEL-212, which pairs ImmTOR nanoparticles with uricase, met Phase 3 endpoints in chronic refractory gout in 2023, demonstrating that immune tolerance strategies can enable otherwise immunogenic biologics.
These modalities share a compelling thread: durability. CAR-T infusions can produce sustained remissions lasting years. Engineered Tregs persist in circulation, theoretically self-renewing. Antigen-specific tolerance, if truly achieved, could eliminate the need for chronic dosing altogether. Payers battered by the cost of lifetime immunosuppression—Humira alone generated $200 billion in cumulative sales before biosimilar entry—are paying close attention.
Navigating the Wreckage
Ditto enters a market sorting winners from cautionary tales with brutal efficiency. Evelo Biosciences, once a darling of microbiome immunotherapy, shut down in 2023-2025 after Phase 2 failures in atopic dermatitis and psoriasis, unable to secure additional financing despite earlier enthusiasm. The microbiome thesis that attracted hundreds of millions in venture capital during the 2010s collided hard with elevated placebo response rates and persistent mechanism-of-action questions.
The parasite-inspired approach offers distinct advantages—and distinct risks. Defined biologics, whether recombinant proteins or computationally engineered analogs, sidestep the regulatory morass of live organisms entirely. EFSA's refusal to endorse TSO safety and the complete absence of FDA approvals for any helminth therapy signal that purified, characterized molecules will face smoother regulatory paths. AI Protein's $400 million collaboration with Bristol Myers Squibb and $41.5 million Series A in 2024-2025 demonstrate Big Pharma's appetite for computationally designed protein modalities, assuming target validation holds.
Yet the bar for differentiation keeps rising. Biosimilar-driven pricing compression means novel therapies must justify premium pricing with demonstrably superior efficacy, safety profiles, or treatment durability. BCG's 2025 biopharma trend analysis highlights manufacturing complexity, Inflation Reduction Act pricing pressures, and concentrated therapeutic area leadership as structural headwinds. IL-23 inhibitors like AbbVie's Skyrizi, approved for ulcerative colitis in 2024, expand the biologics arms race in core autoimmune indications, raising the efficacy bar for new entrants.
The regulatory calculus arguably favors Ditto's strategy. Rather than persuading agencies to approve controlled parasitic infections—a Sisyphean task given current safety standards—the company can develop engineered proteins with defined pharmacology, predictable toxicology, and manufacturing reproducibility. AlphaFold's 214-million-sequence coverage and WormBase ParaSite's 181-species genomic database provide target-rich hunting grounds. If Evo 2's sequence-to-function predictions hold up under experimental validation, the bottleneck shifts from identifying candidate proteins to selecting the right mechanisms for specific disease pathways—a more tractable problem.
Immune tolerance as a therapeutic goal has officially arrived, accelerated by this year's Nobel recognition. Capital is flowing toward companies promising to reset immunity rather than dampen it indefinitely. Whether Ditto's evolutionary mining thesis delivers a first-in-class tolerance inducer or joins Evelo in the cautionary tale column depends entirely on execution: target selection, preclinical validation, and successfully navigating the perilous valley between promising mouse data and human proof-of-concept.
The worms themselves won't make it to market. Too messy, too unpredictable, too laden with regulatory baggage. But the molecular lessons from their survival strategies—decoded by AI, refined by medicinal chemistry, and packaged into defined biologics—might yet rewire how we treat the 15 to 50 million Americans whose immune systems have turned against them.
That's the bet, anyway. And in a sector desperate for new mechanisms that don't just suppress but reprogram, it's a bet that's starting to look less audacious and more inevitable.
