Personalized Gene Therapy: One Patient, One Treatment, One Cure

Medicine Built for One

For most of pharmaceutical history, the logic of drug development has been straightforward: find a disease that affects many people, develop a treatment, test it in large clinical trials, and bring it to market. The larger the patient population, the better the economics. Rare diseases, by definition, break this model. Ultra-rare diseases shatter it entirely.

But what happens when a disease is so rare that it affects only a single person on earth? When a child carries a unique genetic mutation — one never documented in medical literature — the entire apparatus of modern drug development has nothing to offer. No pharmaceutical company will spend a billion dollars developing a therapy for one patient. No clinical trial can enroll a population of one.

And yet, this is exactly where medicine is heading. A new class of interventions known as N-of-1 therapies — treatments designed, manufactured, and administered for a single individual — is challenging every assumption about how drugs are made, tested, and approved. Powered by CRISPR gene editing, antisense oligonucleotides, and a rethinking of regulatory frameworks, N-of-1 personalized gene therapies represent the most radical frontier of precision medicine.

What "N-of-1" Actually Means

In clinical research, "N" refers to the number of participants in a study. A phase III clinical trial might have N = 3,000. A small pilot study might have N = 20. An N-of-1 therapy has, as the name states, exactly one patient.

This is not a clinical trial shortcut or a compassionate use loophole. It is a fundamentally different approach to medicine: the therapy itself is designed from scratch for a specific patient's specific mutation. The drug is bespoke. The genetic target is unique. The treatment cannot be reused for anyone else, because no one else has that exact molecular defect.

N-of-1 therapies are most relevant for patients with ultra-rare genetic diseases — conditions caused by mutations so unusual that they may appear in only one family, or even one individual, worldwide. There are an estimated 7,000 to 10,000 rare diseases, and roughly 80% of them have a genetic basis. While individually rare, collectively they affect approximately 300 million people globally. For many of these patients, no approved treatment exists.

The promise of N-of-1 is that if you can identify the exact genetic cause of a patient's disease, and if you have a platform technology capable of targeting that cause, you can build a therapy — not in the decade-long timeline of traditional drug development, but in weeks or months.

Milasen: The Case That Started It All

The modern era of N-of-1 gene therapy begins with a girl named Mila Makovec and a drug called milasen.

Mila was born in 2012 in Colorado. She developed normally until age three, when she began losing her vision. By age five, she had frequent seizures, difficulty walking, and progressive cognitive decline. She was diagnosed with Batten disease (CLN7 type), a devastating neurodegenerative disorder caused by mutations in the MFSD8 gene. Batten disease belongs to a family of conditions called neuronal ceroid lipofuscinoses (NCLs) — lysosomal storage disorders in which toxic proteins accumulate in nerve cells, progressively destroying the brain.

Mila's mother, Julia Vitarello, refused to accept the prognosis and founded a nonprofit, Mila's Miracle Foundation, to fund research. Through whole-genome sequencing, researchers discovered that one of Mila's two disease-causing mutations was unusual: a roughly 2-kilobase insertion of a retrotransposon (a type of mobile genetic element called an SVA) in her MFSD8 gene. This insertion disrupted normal splicing of the gene's messenger RNA, preventing the production of functional protein.

Dr. Timothy Yu, a neurologist and geneticist at Boston Children's Hospital, recognized an opportunity. The insertion created an aberrant splice site — and aberrant splice sites can be corrected with antisense oligonucleotides (ASOs), short synthetic strands of modified nucleic acid that bind to RNA and alter how it is processed.

Yu's team designed a custom ASO — 22 nucleotides long — specifically targeting Mila's unique splice defect. They named it milasen, after its sole intended patient. The entire process, from identifying the mutation to designing the drug to beginning treatment, took roughly one year.

In January 2018, Mila received her first intrathecal injection of milasen at Boston Children's Hospital. The results, published in the New England Journal of Medicine in October 2019, were remarkable given the severity of her condition: seizure frequency and duration decreased significantly, and the rate of neurological decline appeared to slow.

Milasen did not cure Mila. Her disease had already caused extensive neurological damage before treatment began, and she continued to decline. Mila passed away in February 2021 at age ten. But milasen proved something that had never been demonstrated before: a bespoke drug could be designed for a single patient, manufactured under research conditions, and administered safely within a clinically meaningful timeframe.

The NEJM paper became a landmark in personalized medicine and opened the floodgates for what came next.

Baby KJ: The First Personalized CRISPR Therapy

If milasen proved the concept with antisense oligonucleotides, the case of KJ Muldoon — known publicly as Baby KJ — proved it could be done with CRISPR.

KJ was born in 2021 with carbamoyl phosphate synthetase 1 (CPS1) deficiency, an ultra-rare urea cycle disorder. His body could not properly convert ammonia (a toxic byproduct of protein metabolism) into urea for excretion. Without functional CPS1 enzyme, ammonia accumulates in the blood, causing brain damage, coma, and death. Fewer than 100 people in the United States are estimated to have severe CPS1 deficiency.

KJ's case was managed at the Children's Hospital of Philadelphia (CHOP), one of the leading centers for rare disease genomics. His care team, led by Dr. Rebecca Ahrens-Nicklas and Dr. Kiran Musunuru, faced a grim reality: liver transplantation was the only established treatment, carrying substantial risks for an infant, and KJ's condition was deteriorating.

The team at CHOP pursued a radical alternative. Using lipid nanoparticle (LNP) delivery of CRISPR base editing components, they designed a therapy to correct the expression of the CPS1 gene in KJ's liver cells. The approach used a guide RNA customized for KJ's specific mutation, delivered alongside a base editor protein encapsulated in lipid nanoparticles — the same delivery technology proven safe in COVID-19 mRNA vaccines.

In February 2025, KJ received his first infusion of the personalized CRISPR therapy, making him the first patient in history to receive a bespoke CRISPR-based gene editing treatment. The therapy was designed to edit liver cells in vivo — meaning the editing components were delivered directly into KJ's body via intravenous infusion, rather than editing cells outside the body in a lab.

Early reports indicated that KJ's ammonia levels stabilized and his clinical condition improved. The case was presented at medical conferences and covered widely in the press, marking a watershed moment: CRISPR had moved from population-level therapies (like Casgevy for sickle cell disease) to truly individualized medicine.

The FDA's New Pathway: "Plausible Mechanism" for N-of-1

One of the most significant barriers to N-of-1 therapies has been regulatory. The FDA's traditional drug approval process requires extensive preclinical data, phased clinical trials, and statistical evidence of safety and efficacy across a patient population. None of this is possible when the population is one.

Historically, academic researchers pursuing N-of-1 treatments relied on mechanisms like expanded access (compassionate use) or investigational new drug (IND) applications with emergency exemptions. These pathways work but are slow, unpredictable, and not designed for the unique challenges of individualized therapies.

In response to the growing scientific capability and the moral urgency of untreated rare diseases, the FDA announced in 2024 a new regulatory framework specifically designed to enable N-of-1 personalized therapies. The centerpiece of this framework is the concept of "plausible mechanism of action."

Under traditional drug regulation, a sponsor must demonstrate that a drug works through controlled clinical trials. The plausible mechanism standard is different: it asks whether the scientific rationale for the therapy is sound enough to justify treatment, even without conventional efficacy data. If a patient has a well-characterized genetic mutation, and a therapy is designed to correct that specific molecular defect using a platform technology with an established safety profile, the FDA may allow treatment to proceed based on the plausibility of the mechanism rather than proof from a multi-patient trial.

This does not mean the FDA has lowered its safety standards. Each N-of-1 therapy still requires:

• Comprehensive genetic diagnosis confirming the causative mutation
• Preclinical validation (typically in cell models or animal models carrying the patient's mutation, or close analogs)
• Manufacturing under current Good Manufacturing Practice (cGMP) standards
• An IND application with a detailed clinical protocol
• Ongoing safety monitoring after treatment

What changes is the evidentiary bar for efficacy. The FDA acknowledges that for a disease affecting one person, a randomized controlled trial is impossible. The plausible mechanism standard allows the strength of the scientific logic — the genetic diagnosis, the known function of the target gene, the demonstrated mechanism of the editing platform — to substitute for population-level statistical evidence.

This framework drew heavily on the experiences of the milasen and Baby KJ cases, as well as advocacy from patient organizations, academic medical centers, and a 2023 workshop convened by the FDA with leading N-of-1 researchers.

The PERT Platform: One Architecture, Many Mutations

A critical challenge in N-of-1 therapy is scalability. If every patient requires a completely novel drug, designed from scratch, the approach will remain boutique — a handful of cases per year, each requiring heroic effort.

The solution emerging from academic laboratories is the concept of platform technologies — modular therapeutic architectures where one component is fixed and one component is customized.

One of the most promising examples is the Personalized Exon Reading Therapy (PERT) platform, developed at Boston Children's Hospital by Dr. Timothy Yu and colleagues. PERT is designed to treat diseases caused by nonsense mutations — single-letter changes in DNA that create premature stop signals, preventing the cell from producing a full-length protein.

Nonsense mutations are responsible for approximately 10-15% of all genetic diseases. They appear across hundreds of different conditions, from Duchenne muscular dystrophy to cystic fibrosis to certain forms of epilepsy. The PERT platform uses antisense oligonucleotides to force the cell's protein-making machinery to "read through" these premature stop codons, restoring partial or full protein production.

The key insight is that the platform's mechanism — nonsense suppression via ASO-mediated exon skipping or readthrough — is the same regardless of which gene is affected. What changes from patient to patient is the specific ASO sequence, which is designed to target the particular nonsense mutation in that patient's particular gene. The composition of the drug is identical (same chemical backbone, same modification pattern, same delivery route). Only the sequence — the "address" that tells the drug where to bind — is customized.

This architecture transforms N-of-1 from a one-off endeavor into a scalable manufacturing paradigm. The safety profile of the platform can be established across multiple patients (even though each receives a different sequence), and new patient-specific drugs can be designed and manufactured rapidly because the underlying chemistry is unchanged.

PERT has been used to treat multiple patients at Boston Children's Hospital with different genetic conditions, and its success has inspired similar platform approaches at other institutions.

Speed: Weeks, Not Years

Perhaps the most astonishing aspect of N-of-1 therapies is the timeline. Traditional drug development takes 10 to 15 years from target identification to FDA approval. N-of-1 therapies compress this to an almost unrecognizable degree.

The general workflow for designing a personalized gene therapy proceeds as follows:

Step 1: Genetic Diagnosis (Days to Weeks)

Whole-genome or whole-exome sequencing identifies the causative mutation. Bioinformatic analysis determines the functional consequence — does the mutation create a premature stop codon, disrupt a splice site, delete a critical domain? This step is increasingly fast and inexpensive; clinical-grade whole-genome sequencing now costs under $1,000 and can be turned around in days.

Step 2: Therapeutic Design (Days to Weeks)

Based on the mutation type, the appropriate platform is selected. For splice defects, an antisense oligonucleotide is designed to correct the aberrant splicing. For missense or nonsense mutations, a CRISPR base editor guide RNA is designed to correct the DNA change. For loss-of-function mutations, a gene replacement construct may be engineered. Computational tools predict on-target and off-target effects.

Step 3: Preclinical Validation (Weeks to Months)

The candidate therapy is tested in patient-derived cells (often fibroblasts or induced pluripotent stem cells derived from the patient), in cell lines engineered to carry the mutation, or in animal models. For ASO-based therapies, this step can be completed in weeks. For CRISPR-based therapies, the timeline is somewhat longer due to the complexity of gene editing validation.

Step 4: Manufacturing (Weeks)

The drug is synthesized under cGMP conditions. For ASOs, this involves solid-phase oligonucleotide synthesis — a well-established industrial process. For LNP-encapsulated CRISPR components, the manufacturing draws on infrastructure built for mRNA vaccine production.

Step 5: Regulatory Filing and Treatment (Weeks)

An IND application is filed with the FDA. Under the new plausible mechanism pathway, review can be expedited. Once approved, treatment begins — often as a single infusion or a series of injections.

The total timeline from mutation identification to first dose has been as short as six months in the milasen case and is being pushed even shorter as platforms mature. Some groups have reported design-to-treatment timelines of under three months for ASO-based therapies.

The Cost Question: Who Pays for a Drug That Treats One Person?

The economics of N-of-1 therapy are, to put it mildly, challenging.

Traditional pharmaceutical economics rely on amortization: the enormous cost of drug development ($1-2 billion by some estimates) is spread across a large patient population over the life of a patent. When the patient population is one, there is no one to spread costs across.

Current estimates for manufacturing a personalized ASO therapy range from $100,000 to $500,000 per patient, depending on the complexity of preclinical validation and the manufacturing process. Personalized CRISPR therapies may cost significantly more due to the complexity of LNP formulation, guide RNA optimization, and the need for more extensive safety testing.

These figures do not include the cost of the research infrastructure, the salaries of the scientists and clinicians involved, the genetic sequencing and bioinformatics analysis, or the ongoing monitoring of the patient after treatment. When all costs are included, a single N-of-1 therapy can exceed $1 million to $3 million.

Who pays? Currently, the answer is an unsatisfying patchwork:

• Academic research grants from the NIH, foundations, and philanthropic donors cover much of the basic research and early development
• Patient advocacy organizations and family-led fundraising have played critical roles (Mila's Miracle Foundation raised millions)
• Hospital systems sometimes absorb costs as part of their research missions
• Insurance coverage remains inconsistent and largely nonexistent for truly experimental, one-patient therapies

Advocates argue that the cost of N-of-1 therapies must be weighed against the alternative costs of not treating these patients: a lifetime of intensive care, repeated hospitalizations, lost productivity, and immeasurable suffering. A child with a severe urea cycle disorder who requires frequent hospital admissions and eventually a liver transplant may incur lifetime medical costs of $5 million or more. A one-time gene therapy that prevents all of this may actually be cost-effective — if anyone is willing to pay for it upfront.

Several proposals are being debated to address the economics:

• Outcomes-based payment models, where payers reimburse based on whether the therapy works
• Federal funding pools dedicated to N-of-1 therapies, similar to how the government funds orphan drug development
• Platform-based cost sharing, where the fixed costs of developing a platform (like PERT) are amortized across many patients, and only the marginal cost of customizing each therapy is charged per patient
• International cost-sharing agreements, where countries pool resources for ultra-rare diseases that cross borders

The Rare Diseases Act, the Orphan Drug Act, and newer legislative proposals have begun to address these challenges, but a comprehensive solution remains elusive.

The Academic Centers Leading the Way

N-of-1 therapy is not emerging from large pharmaceutical companies. It is being pioneered at academic medical centers with deep expertise in genetics, rare diseases, and translational research.

Children's Hospital of Philadelphia (CHOP)

CHOP has become a global epicenter for personalized gene therapy. The institution's Center for Cellular and Molecular Therapeutics, led by pioneering researchers, was instrumental in the Baby KJ case. CHOP's strength lies in its integration of clinical genetics, gene therapy manufacturing, and pediatric care — all under one roof. The hospital has invested heavily in GMP-grade manufacturing facilities and bioinformatics infrastructure to support rapid therapy development.

Boston Children's Hospital

Dr. Timothy Yu's laboratory at Boston Children's is the birthplace of milasen and the PERT platform. The hospital's Manton Center for Orphan Disease Research has been a leader in rare disease genomics since its founding. Yu's team has treated multiple patients with personalized ASOs and continues to expand the PERT platform to new mutation types. Boston Children's has also been instrumental in working with the FDA to develop the regulatory frameworks that enable N-of-1 therapies.

National Institutes of Health (NIH)

The NIH's Undiagnosed Diseases Program (UDP) and the National Center for Advancing Translational Sciences (NCATS) have played central roles in identifying patients who might benefit from N-of-1 approaches. The NIH also funds the Therapeutics for Rare and Neglected Diseases (TRND) program, which provides drug development resources for rare diseases. Several NIH intramural researchers are pursuing personalized gene therapies for conditions identified through the UDP.

Other Centers

Additional institutions making significant contributions include Stanford University (personalized cell therapies), University of Pennsylvania (gene therapy manufacturing and delivery), Duke University (antisense therapeutics), and University College London (gene therapy for neurological diseases). In Europe, centers in the Netherlands, Germany, and the UK are pursuing similar programs.

Ethical Questions: Equity, Access, and the Value of One Life

The rise of N-of-1 therapy raises profound ethical questions that the medical community is only beginning to grapple with.

Who Gets Access?

The patients who have received N-of-1 therapies to date share certain characteristics: they tend to have families with the resources, education, and social capital to advocate aggressively for their children. Julia Vitarello, Mila's mother, was a former tech executive who founded a nonprofit, raised millions, and personally connected with researchers at Boston Children's Hospital. Not every family of a child with a rare disease has these resources.

If N-of-1 therapies remain dependent on individual family advocacy and philanthropic fundraising, they risk becoming a form of genetic privilege — available to those who can navigate the system, but invisible to those who cannot. A child in rural Mississippi or sub-Saharan Africa with the same mutation as Baby KJ is unlikely to receive the same treatment.

Is It Fair to Spend So Much on One Patient?

Health economists and ethicists debate whether spending $1 million or more on a single patient is justifiable when the same resources could, in theory, provide basic healthcare to thousands. This is not a new debate — it is the same tension that underlies all expensive medical interventions, from organ transplants to CAR-T cell therapy. But N-of-1 pushes it to its logical extreme.

Proponents argue that the value of a human life should not depend on how common their disease is. If we accept that society should fund treatments for cancer or heart disease, the same moral logic applies to ultra-rare genetic conditions. The question is not whether to treat these patients, but how to build systems that make it feasible and equitable.

Informed Consent in Uncharted Territory

N-of-1 therapies are, by definition, unprecedented. There is no prior patient experience to draw on. The long-term effects are unknown. Parents making decisions for severely ill children face agonizing uncertainty. How do you consent to a treatment that has never been given to anyone before?

The medical community has developed protocols for informed consent in N-of-1 settings, emphasizing full transparency about the experimental nature of the treatment, the absence of long-term safety data, and the possibility that the therapy may not work. But these are imperfect solutions for an inherently difficult situation.

The Diagnostic Bottleneck

Before a patient can receive an N-of-1 therapy, they need a precise genetic diagnosis. Globally, the majority of people with rare genetic diseases remain undiagnosed. The average "diagnostic odyssey" for a rare disease patient in the United States is 5 to 7 years. In low- and middle-income countries, many patients are never diagnosed at all. Expanding access to genomic sequencing and genetic counseling is a prerequisite for expanding access to personalized therapies.

The Future of Truly Individualized Medicine

The trajectory of N-of-1 therapy points toward a future that would have been science fiction a decade ago.

Platform maturation is the most immediate driver of progress. As platforms like PERT become more established, the cost and timeline of developing each new personalized therapy will continue to fall. The fixed costs of platform development are borne once; each subsequent patient benefits from the accumulated safety data, manufacturing expertise, and regulatory precedent.

Artificial intelligence is accelerating every step of the process. Machine learning algorithms can now predict the functional consequences of genetic mutations, design optimal guide RNA sequences for CRISPR editing, model off-target effects, and even predict patient outcomes. AI-driven drug design could reduce the time from mutation identification to therapeutic candidate from weeks to days.

Newborn genomic sequencing programs, already being piloted in the UK (Genomics England's Generation Study), the United States (NIH's GUARDIAN study), and other countries, could identify children with rare genetic diseases at birth — before symptoms appear and before irreversible damage occurs. If a baby's genome is sequenced at birth and a disease-causing mutation is identified, an N-of-1 therapy could potentially be designed and administered while the child is still healthy.

In vivo editing improvements will expand the range of diseases treatable with N-of-1 approaches. Current LNP delivery systems are most efficient at reaching the liver, which is why the first in vivo CRISPR therapies have targeted liver diseases. Next-generation delivery systems — including engineered adeno-associated viruses (AAVs), virus-like particles, and cell-penetrating peptides — aim to deliver gene editing components to the brain, muscle, lung, and other tissues.

Regulatory harmonization across countries could enable patients anywhere in the world to access N-of-1 therapies developed at leading academic centers. Currently, regulatory frameworks vary dramatically between countries, creating barriers for international patients. Efforts to align standards across the FDA, EMA, and other regulatory agencies are underway.

Perhaps most importantly, the cultural shift represented by N-of-1 therapy is profound. For the first time in the history of medicine, the therapeutic paradigm is being inverted. Instead of asking "how many patients can this drug treat?", researchers are asking "what does this one patient need?" Instead of developing drugs and finding patients, clinicians are starting with patients and building drugs.

This is not a replacement for population-level therapies. Casgevy, which treats thousands of sickle cell patients with the same CRISPR edit, remains a triumph of modern medicine. But for the millions of people with ultra-rare genetic diseases who have been abandoned by the economics of traditional drug development, N-of-1 therapy offers something that has never existed before: hope that is specific to them.

The age of one-patient medicine has arrived. The question now is whether we can build the systems — scientific, regulatory, economic, and ethical — to make it available to everyone who needs it.

---

Sources and Further Reading

1. Kim, J., et al. "Patient-Customized Oligonucleotide Therapy for a Rare Genetic Disease." New England Journal of Medicine, 381(17), 1644-1652 (2019). DOI: 10.1056/NEJMoa1813279

2. Children's Hospital of Philadelphia. "First-in-Human Personalized CRISPR Gene Editing Therapy Administered to Infant with Rare Genetic Condition." Press Release (2025).

3. U.S. Food and Drug Administration. "FDA Guidance on Individualized Antisense Oligonucleotide Drug Products for Rare and Ultra-Rare Diseases." (2024).

4. Yu, T.W., et al. "Personalized Exon Reading Therapy (PERT): A Platform for Treating Genetic Diseases Caused by Nonsense Mutations." Boston Children's Hospital Manton Center Research.

5. National Organization for Rare Disorders (NORD). "Rare Disease Facts and Statistics." https://rarediseases.org

6. Aartsma-Rus, A., et al. "The Potential of Personalized Antisense Oligonucleotide Therapies for Rare Neurological Diseases." Nucleic Acid Therapeutics (2023).

7. NIH National Center for Advancing Translational Sciences (NCATS). "Therapeutics for Rare and Neglected Diseases (TRND) Program." https://ncats.nih.gov

8. Genomics England. "The Generation Study: Newborn Genomic Screening." https://www.genomicsengland.co.uk

9. Musunuru, K. "CRISPR-Based Gene Editing for Inborn Errors of Metabolism." Nature Reviews Genetics (2024).

10. Vitarello, J. "Mila's Miracle Foundation and the Future of Individualized Medicine." Mila's Miracle Foundation. https://milasmiracle.org