The US government’s health research moonshot agency has launched a five-year, $160 million bet that the era of bespoke gene-editing cures — currently limited to children whose families can access a vast scientific team willing to absorb the cost — can be converted into an industrial platform available to the 30 million Americans living with rare genetic diseases, 95% of whom have no approved treatment for their underlying condition
The Advanced Research Projects Agency for Health (ARPA-H), part of the Department of Health and Human Services, announced on July 9 that it had awarded contracts to seven research teams across academia and biotechnology for its THRIVE program — Treating Hereditary Rare Diseases with In Vivo Precision Genetic Medicines. The awards are milestone-gated across five years, with an explicit mandate to reach first-in-human clinical trials by year three and, by year five, to demonstrate a regulatory architecture — an umbrella investigational new drug application (IND) covering multiple diseases and multiple therapies in a single framework — that has no direct precedent in gene therapy.
“Our current system is built to develop one drug for one disease tested in clinical trials designed specifically for that drug and that disease,” said ARPA-H Program Manager Daria Fedyukina, Ph.D., who leads THRIVE. “Patients with rare genetic diseases, especially children, cannot wait for this legacy approach, it’s too slow and unsustainable. THRIVE will blaze a new trail and change that — using a modular building-block approach that swaps one key piece instead of rebuilding the whole set for every disease.”
The program was delayed for more than a year. Funding was originally approved in early 2025 under then-ARPA-H director Renee Wegrzyn. “Shortly thereafter, she was asked to leave and agency program launches were largely put on hold while the administration introduced new approval processes and a new director could be installed,” said Mimi Lee, who directed THRIVE before the transition. Thursday’s announcement represents the program’s formal start under ARPA-H Director Alicia Jackson, Ph.D., who called it one that “could change the trajectory of genetic disease, expand access to advanced treatments, and reinforce US leadership in the future of medicine.”
What Made a Single Baby a Proof of Concept — and Why That Wasn’t Enough
The animating case behind THRIVE is KJ Muldoon, a Philadelphia toddler born in August 2024 with carbamoyl phosphate synthetase 1 (CPS1) deficiency — a urea cycle disorder affecting roughly one in 1.3 million births, in which the body cannot clear ammonia from the bloodstream when protein is consumed. Half of all babies born with severe CPS1 deficiency die within their first week of life
KJ survived. In early 2025, a team of researchers at Children’s Hospital of Philadelphia designed and administered kayjayguran abengcemeran — the first personalized CRISPR gene-editing therapy ever developed for a single patient. The drug, a base editor delivered via lipid nanoparticles directly into KJ’s liver cells, was described in the New England Journal of Medicine in June 2025. KJ was discharged from CHOP in June 2025, 307 days after admission, his disease transformed from lethal to manageable.
The achievement required an international research collaboration, months of sprint-level effort, and costs that the collaborating companies agreed to write off as unrecoverable. For every other child born with a rare genetic mutation, none of that infrastructure would exist
“How can we build a paradigm that is also sustainable and can treat thousands of babies?” Fedyukina asked at the program’s launch
That is the question THRIVE is funded to answer
Seven Teams, Seven Disease Frontiers, One Architecture
Each of the seven funded teams is responsible for a different organ system and patient population — but all are accountable to the same structural mandate: build a platform, not a one-off, and make it replicable by the broader rare disease research community
Children’s Hospital of Philadelphia received the program’s largest award — up to $38.9 million over five years — to build directly on the KJ Muldoon lineage. The team, led by Rebecca Ahrens-Nicklas, MD, PhD, and Lindsey A. George, MD, will expand the base-editing approach to four groups of rare liver-related genetic disorders: urea cycle disorders (dangerous ammonia buildup in newborns), organic acidemias (toxic metabolites causing metabolic crises in infants), and rare bleeding and clotting conditions including hemophilia A and protein C deficiency.
Broad Institute / Jackson Laboratory received up to $34.5 million to build the Pediatric Epilepsies and Rare CNS (PERC) Gene Editing Platform, targeting Dravet syndrome and alternating hemiplegia of childhood (AHC) — two devastating neurological diseases caused by specific genetic mutations. “PERC gives us an opportunity to stop thinking about each rare disease program as something that has to start from scratch and instead build a process that is more systematic, repeatable, and adaptable,” said Cathleen Lutz, Ph.D., of The Jackson Laboratory. The team’s additional mandate is to pioneer a novel deployment model to keep the cost of delivery sustainable once therapies reach patients.
UC Berkeley / Innovative Genomics Institute will develop in vivo gene-editing platforms for rare inborn errors of immunity — life-threatening conditions in which the immune system itself carries the genetic defect. The IGI, founded by Nobel laureate Jennifer Doudna, will pair editing precision with a rapid non-invasive patient identification method to accelerate the path from diagnosis to treatment
GEMMABio / Profluent Bio carries the program’s most technologically novel approach. The team, led by gene therapy pioneer Jim Wilson at GEMMABio and partnered with Profluent Bio — a Berkeley-based AI company that in April 2024 released OpenCRISPR-1, the first publicly available AI-designed gene editor — will use machine-learning-generated base editors rather than hand-engineered ones. Profluent’s protein language models, trained on more than 115 billion unique protein sequences, have produced a library of base editors designed to correct any transition mutation. ARPA-H described the approach as promising “ultimate scalability as opposed to current bespoke and artisanal editors.” The initial target diseases include homozygous familial hypercholesterolemia (HoFH), a genetic disorder causing dangerously elevated LDL cholesterol from birth, and maple syrup urine disease (MSUD).
St. Jude Children’s Research Hospital will develop gene-editing medicines for bone marrow failure disorders, using the hospital’s decades of pediatric hematology expertise and what ARPA-H described as the most sensitive and accurate off-target detection methods available — addressing one of the core safety questions in the entire in vivo gene editing field
Stanford University has been tapped for rare skin diseases, including epidermolysis bullosa — a group of conditions in which skin blisters and tears with the slightest contact, sometimes called butterfly skin disease — using a topical gene-editing approach that must reach epithelial cells without systemic delivery
Massachusetts General Hospital will target rare genetic diseases of the blood vessels, with a structural innovation in preclinical testing: alongside advancing non-viral delivery of gene editors to vascular tissue, the MGH team will build 3D in vitro models of human blood vessels that can serve as preclinical test beds, reducing dependence on slow animal studies and reflecting the FDA’s 2022 authorization of non-animal testing data in IND submissions
Read more:Gene Editing Could Rescue 28,000 Discarded Organs: CRISPR Roadmap Targets Perfusion Window
Base Editing vs. CRISPR: Why the Distinction Matters for Safety and Scale
The technology underlying most of THRIVE’s platforms is base editing — a refinement of CRISPR that the KJ Muldoon case introduced to the public but that the gene therapy field has been developing for years
Standard CRISPR-Cas9 operates like molecular scissors: a guide RNA directs the Cas9 protein to a specific location in the genome, where it cuts both strands of the DNA double helix. The cell then repairs the break — but repair processes are imprecise, and the most common repair pathway (non-homologous end joining) introduces random insertions or deletions that can disrupt nearby genes and trigger immune responses or, in rare cases, cell death. CRISPR-Cas9 also requires the presence of a protospacer adjacent motif (PAM) sequence near the target site, limiting the addresses in the genome it can reach.
Base editors work differently. A base editor is a fusion protein: a modified Cas9 that has been deactivated as a cutter and linked instead to a deaminase enzyme. The deaminase directly converts one DNA base to another — adenine (A) to guanine (G), or cytosine (C) to thymine (T) — without breaking either DNA strand. No double-strand break means no error-prone repair, lower off-target risk, and no dependence on the cell’s cycle phase. This makes base editing particularly effective in liver cells, blood vessel cells, and neurons — all post-mitotic or slow-dividing tissues where the break-and-repair approach of classical CRISPR works poorly.
The limitation: conventional hand-designed base editors can only correct specific transition mutations. Each new mutation requires months of protein engineering to design a new editor with the right targeting sequence and activity profile. That is the “artisanal” problem THRIVE’s Profluent partnership is directly attacking. Profluent’s protein language models generate base editors computationally, producing a library covering the full range of correctable mutations rather than designing them one at a time. If the approach validates in the clinic, the design time for a new rare disease therapy could fall from months to days.
Delivery also matters. KJ’s treatment used lipid nanoparticles (LNPs) — the same non-viral delivery vehicle used in mRNA COVID vaccines — to carry the base editor directly into his liver cells after intravenous injection. LNPs break down naturally and do not integrate into the genome, avoiding the immune response problems associated with adeno-associated virus (AAV) vectors. MGH’s mandate to develop non-viral delivery to blood vessel cells signals ARPA-H’s broader preference for non-viral approaches throughout the program.
An Untested Regulatory Template Could Change Everything
The most significant structural innovation in THRIVE is not the AI-designed protein library or the non-viral delivery approach. It is the umbrella IND
In standard drug development, a single investigational new drug application covers a single product for a single disease. That structure is tolerable when a patient population numbers in the tens of thousands and a single trial can recruit enough participants to detect a statistical signal. For diseases affecting fewer than 200 patients — or fewer than 50 — it is not. A trial that requires 30 participants for statistical significance cannot be run when only 50 people in the country have the disease, most of them infants who would need to be identified, enrolled, and treated simultaneously at specialized centers.
THRIVE’s umbrella IND model inverts that logic. One investigational application covers multiple gene-editing therapies targeting multiple diseases, with a shared safety and biodistribution framework established in year one. Year three’s deliverable is a first-in-human trial in which multiple different individualized products for multiple different diseases are tested within a single clinical structure. Year five’s deliverable is proof that the framework can be expanded to additional diseases and drug products.
This has no precedent in gene therapy at FDA. ARPA-H’s program language explicitly names “regulatory innovations are mandatory” as a program requirement — acknowledging that the teams will need to co-develop new guidance with FDA rather than follow existing pathways. If the umbrella IND model succeeds, it would create a regulatory template that every rare disease gene therapy developer could inherit. If it stalls on regulatory uncertainty, the framework itself becomes visible as the structural bottleneck it is.
THRIVE teams are required to publish their methods and share their platforms openly — so that, as ARPA-H put it, “their rare disease colleagues across the country” can replicate the path
What to Watch — and When
The program’s year-by-year milestones give researchers, patient advocates, and investors a concrete benchmark schedule. Within roughly a year from the award date, each team must demonstrate a gene-editing platform capable of generating multiple drug products with common biodistribution and toxicology profiles — the proof that the platform concept works across more than one disease. Watch for IND filings and publications beginning in mid-2027
By approximately July 2029, every team must have initiated first-in-human clinical trials. By July 2031, the expanded umbrella IND and deployment models must be validated.
Funding is milestone-contingent throughout. Teams that do not hit their gates do not receive the next tranche. That structure, common in ARPA programs, keeps pressure on performance rather than on the size of the initial award
For the 30 million Americans with rare diseases — and the parents of the next KJ Muldoon — the next three years are the test of whether the government’s $160 million bet lands on the right side of history
Frequently Asked Questions
What is ARPA-H’s THRIVE program, and who does it help?
THRIVE (Treating Hereditary Rare Diseases with In Vivo Precision Genetic Medicines) is a five-year, up to $160 million federal program that funds seven research teams to build scalable gene-editing platforms for rare pediatric genetic diseases. “In vivo” means the gene-editing therapy is delivered directly into a patient’s body — typically via lipid nanoparticles injected into the bloodstream — rather than requiring cells to be removed, edited in a laboratory, and reinfused. The program targets patients with rare genetic conditions caused by single-gene mutations, a population that includes more than 30 million Americans, the majority of them children, for whom 95% of rare diseases currently have no approved treatment.
How does base editing differ from standard CRISPR, and why does it matter for rare disease treatment?
Standard CRISPR-Cas9 cuts both strands of the DNA double helix, relying on the cell’s own repair machinery to fix the break — a process that is imprecise and works poorly in the non-dividing cells (liver, brain, blood vessels) most commonly affected in rare genetic diseases. Base editing skips the cut entirely: a deactivated Cas9 protein carries a deaminase enzyme to the target site and directly converts one DNA base to another without breaking the DNA strand. This dramatically lowers off-target editing risk and works effectively in the cell types that matter most for pediatric rare diseases. Profluent Bio’s AI-designed base editors go further: instead of engineering each new editor by hand (a process taking months), protein language models generate editors computationally, potentially collapsing that timeline to days and making the platform approach economically viable for diseases affecting only dozens of patients.
What makes the umbrella IND design different, and why does FDA have no playbook for it?
In standard drug development, each investigational new drug application covers one drug for one disease. That model becomes economically and statistically untenable for diseases affecting fewer than a few hundred patients — you cannot run a statistically meaningful clinical trial for a condition affecting 40 people. THRIVE’s umbrella IND is designed to cover multiple gene-editing therapies targeting multiple different diseases under a single regulatory filing, with a shared safety framework established in the program’s first year. No gene therapy program has previously demonstrated this structure at FDA, which is why ARPA-H explicitly lists regulatory innovation as a mandatory deliverable alongside the scientific milestones. If THRIVE’s umbrella IND succeeds, it will create a reusable regulatory template for the broader rare disease field. If it reveals gaps in the existing regulatory framework, those gaps will be the structural obstacle that the field needs to address next.
When might a patient with a rare genetic disease actually benefit from THRIVE’s work?
THRIVE’s milestone structure provides a concrete answer. By approximately mid-2027, teams must demonstrate platforms capable of generating multiple therapies with consistent safety profiles — the first public proof of platform validity. By approximately mid-2029, every funded team must have initiated first-in-human clinical trials. THRIVE does not guarantee approved therapies within that window; clinical trials take additional years to complete and regulatory review follows. But the year-three milestone represents the point at which patients and families will be able to see whether enrollment in a THRIVE trial is a real option. Patients with rare genetic diseases affecting the liver, blood, brain, immune system, skin, or blood vessels — the organ systems covered by the seven funded teams — should watch for IND filings and trial announcements beginning in 2027. Patient advocacy organizations in those disease areas may also receive early engagement invitations, as THRIVE’s program language explicitly requires teams to engage patients and families in shaping trial design.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission
Tags:Gene editingCrisprRare DiseaseGene therapyFDA
