The first personalized CRISPR therapy: how CHOP and UPenn pulled it off

In February 2025, a baby named KJ — full name KJ Muldoon — became the first human being ever treated with a fully personalized, bespoke CRISPR gene editing therapy: a therapy designed specifically for his unique genetic mutation, and no one else's.^[1]

This was not a repurposed therapy. It was precision medicine at its most literal: one patient, one disease, one tool built to match.

KJ was born with severe carbamoyl phosphate synthetase 1 (CPS1) deficiency, a rare and usually fatal metabolic disorder that prevents the liver from safely processing ammonia. Without treatment, toxic ammonia accumulates in the blood and causes catastrophic neurological damage or death in infancy.^[1] He spent his first months of life in the hospital on a severely restricted diet while a team at CHOP (Children's Hospital of Philadelphia) and Penn Medicine raced to build something that had never existed before: a therapy custom-designed for his exact mutation.^[1]

The Scientific Challenge

Within six months of KJ's birth, the team designed, manufactured, safety-tested, and administered a custom-built base-editing therapy — delivered via lipid nanoparticles (LNPs) directly to liver cells — that corrected the specific mutation in his genome.^[1] He received three doses, in February, March, and April of 2025. After treatment, he tolerated more dietary protein, needed less ammonia-lowering medication, and by late 2025 was home with his family, walking, and hitting normal developmental milestones.^[1,4]

None of this was built from nothing. Base editing technology itself was pioneered by David Liu at the Broad Institute between roughly 2016 and 2019, and Musunuru's own lab had spent years — from around 2015 to 2024 — conducting preclinical research on CPS1 and related urea cycle disorders before KJ was ever born.^[1] What changed in 2024 and 2025 was the ability to compress everything that normally happens over a decade — mutation characterization, guide RNA design, manufacturing, and safety testing — into a period of months, for a single patient who could not wait for a normal trial timeline.

This case is widely described as a preview of the future of medicine: proof that personalized gene therapy for any of the roughly 7,000 known rare genetic diseases can, in principle, be designed and deployed in months rather than decades.^[1]

Who Made It Possible

Dr. Kiran Musunuru served as co-lead investigator. Born in New York City and raised in Florida, he is of Telugu-American descent; his father immigrated from India in 1976 to practice cardiology.^[1] Musunuru's own educational path is unusually wide-ranging: an AB (Artium Baccalaureus, the Latin form of a Bachelor of Arts) in Biochemical Sciences from Harvard College in 1997, a PhD in Biomedical Sciences from Rockefeller University in 2003, an MD from Weill Cornell in 2004, an MPH in Epidemiology from Johns Hopkins in 2009, a law degree (ML) from Penn Law in 2019, and a Master's in Regulatory Affairs from Penn's Perelman School of Medicine in 2024.^[1] He trained in Internal Medicine at Brigham and Women's Hospital and in Cardiovascular Medicine at Johns Hopkins, with postdoctoral work at Massachusetts General Hospital and the Broad Institute.^[1] He now holds the Barry J. Gertz Professorship for Translational Medicine at Penn and co-directs the Penn Medicine/CHOP Orphan Disease Center. His honors include a Presidential Early Career Award for Scientists and Engineers under the Obama administration, and his broader research focus includes work toward a single-shot gene-editing "vaccine" against heart attacks.^[1]

Dr. Rebecca Ahrens-Nicklas, a metabolic physician at CHOP and director of its Gene Therapy for Inherited Metabolic Disorders (GTIMD) Frontier Program, served as the other co-lead investigator and was the clinical lead for KJ's care from diagnosis all the way through treatment.^[1]

Several other groups were essential to making the timeline possible at all. The Innovative Genomics Institute (IGI) — a UC Berkeley/UCSF institute co-founded by CRISPR pioneer Jennifer Doudna — characterized KJ's specific mutation, designed the guide RNAs, and performed the safety assessments the FDA needed to see; this partnership is the critical reason FDA review took one week rather than years.^[1] Danaher Corporation provided in-kind manufacturing and instrumentation support for clinical-grade LNP production, while Acuitas Therapeutics, Integrated DNA Technologies, and Aldevron contributed in-kind work on LNP formulation and DNA synthesis.^[1]

The work itself took place across CHOP, one of the world's oldest and most respected pediatric hospitals, and Penn's Perelman School of Medicine, with the IGI's mutation analysis and guide RNA design happening at UC Berkeley.^[1]

The Policy Implications — A New Regulatory Pathway

KJ's case forced a genuinely new regulatory moment at the FDA. The normal pathway for a gene therapy runs through preclinical studies, an IND submission, Phase I/II/III clinical trials over 10 to 15 years, a BLA submission, FDA review, and finally approval.^[1] None of that timeline was available to KJ.

Instead, the team filed a single-patient expanded-access IND — also called a compassionate use or emergency IND — under 21 CFR 312.36, a mechanism that allows treatment of a life-threatening condition outside a standard clinical trial when no approved alternative exists.^[1,6] The FDA processed and approved that IND in approximately one week, exceptional even for emergency INDs, which normally take weeks to months.^[1] No standard Phase I, II, or III trial data existed for this therapy; the FDA instead accepted preclinical animal model data along with the IGI's safety assessments, and IRB review was conducted on an expedited basis given the emergency.^[1]

The story didn't end with KJ's treatment. In November 2025, FDA Commissioner Makary and CBER Director Prasad announced an entirely new bespoke gene therapy approval pathway, created directly in response to this case.^[1,7] Its key features: no requirement for standard clinical trial data; the therapy must target a specific, well-characterized molecular abnormality; evidence of successful target editing is required; marketing authorization follows demonstrated success across several consecutive patients; and long-term real-world evidence is collected after approval rather than required beforehand.^[1] The pathway is explicitly designed to let personalized therapies for rare genetic diseases reach patients in months instead of decades. The Accelerating Medicines Partnership (AMP) Bespoke Gene Therapy Consortium is doing parallel work, streamlining requirements across academic institutions so that the next case doesn't have to improvise a regulatory path from scratch.^[1,8]

Every subsequent personalized therapy will be easier because of this case — not because the science gets simpler, but because the regulatory precedent, the manufacturing relationships, and the safety-data expectations are now established rather than invented under emergency pressure.

Funding

The work was primarily funded through federal NIH grants, including the NIH Somatic Cell Genome Editing (SCGE) Program (U01TR005355, U19NS132301) along with several additional grants (R35HL145203, U19NS132303, DP2CA281401, P01HL142494).^[1,9] CHOP's own Research Institute supported the work through its GTIMD Frontier Program, and a set of companies — Acuitas Therapeutics, Integrated DNA Technologies, Aldevron, and Danaher — contributed in-kind technology and manufacturing support rather than direct cash funding.^[1] The Accelerating Medicines Partnership Bespoke Gene Therapy Consortium represents a further public-private layer, working specifically to streamline FDA approval pathways for customized rare-disease gene therapies going forward.^[1]

IP and the Commercial Landscape

The specific therapy built for KJ is, by its very nature, not commercially licensable — it was designed for one unique mutation and has no market of its own. What is commercially relevant is the IP around the underlying platform and delivery mechanism.^[1]

Base editing technology was pioneered by David Liu at the Broad Institute and is licensed through Beam Therapeutics, founded in 2017 and publicly traded on NASDAQ as BEAM since its 2020 IPO; Beam holds broad IP on base editors generally.^[1,10] LNP delivery technology — the same delivery system used in COVID-19 mRNA vaccines — is licensed from Acuitas Therapeutics and Alnylam Pharmaceuticals, among others, and was a critical enabling technology for the entire approach.^[1,11] Musunuru's own lab at Penn holds patents on specific gene editing approaches for cardiovascular and metabolic diseases, though the exact patents tied to KJ's particular therapy haven't been publicly disclosed, and CHOP's Gene Therapy Program has its own commercialization track managed through Penn's technology transfer office.^[1]

The deeper commercial question is one of business model: how do you build a sustainable company around therapies that are each, by design, unique to a single patient? As one outside expert has put it, that's the real challenge facing this entire category.^[1] The AMP Consortium and the FDA's new pathway are both aimed at solving this by enabling platform-level approval across families of related mutations, rather than requiring a bespoke regulatory process for every single patient. Beam Therapeutics, Prime Medicine, Intellia, and Editas Medicine are the public companies best positioned to commercialize the next generation of gene editing platforms that KJ's case has now validated in practice.^[1]

Sources & Further Reading

All claims in this post are traceable to the sources below.

Key Publication (Primary Source)

Institutional Press Releases & Announcements

News & Conference Coverage

Regulatory Sources

IP & Commercial Platform Sources

NIH Grant Numbers (for funding verification)