Gene therapy has long been a vision in medicine, tracing back to the early twentieth century when Sir Archibald Garrod first identified the concept of “inborn errors of metabolism.” During this time, the precise nature of genes remained largely a mystery. However, Garrod, along with William Bateson, utilized the principles of Mendelian genetics to demonstrate that alkaptonuria—first documented in a 1902 paper—is a heritable condition. Nathaniel Comfort’s The Science of Human Perfection: How Genes Became the Heart of American Medicine offers a broader perspective on gene therapy and its societal implications, making connections to modern notions of eugenics.
The inaugural success in gene therapy occurred in 1990, when William French Anderson and his team at the NIH treated a four-year-old patient suffering from Severe Combined Immune Deficiency (SCID) caused by adenosine deaminase deficiency (ADA). Using a retroviral vector, they successfully transferred the ADA gene into the patient’s blood cells. Although she required ongoing treatment with modified immune cells, this intervention allowed her to lead a normal life. Since that time, advancements in vector technology have enhanced treatment options for this type of SCID.
This impressive breakthrough led to a somewhat simplistic belief that curing inherited diseases could be achieved through a straightforward sequence of steps:
- Identify and clone the faulty gene using established techniques
- Develop a vector to deliver the gene to the patient’s cells
- Administer the gene therapy to affected individuals
The completion of the human genome sequence about twenty-five years ago further solidified these assumptions. However, the reality of gene therapy is far more complex, as recent challenges in treating conditions like Duchenne Muscular Dystrophy (DMD) and STXPBP1-related disorders have shown. [2] DMD is particularly well-known and affects approximately 1 in 5,000 males. This sex-linked condition arises from mutations in the dystrophin gene, resulting in progressive muscle degeneration, with fat replacing muscle tissue. The average life expectancy for those afflicted is just 24 years.
Similarly, STXBP1 encodes a syntaxin-binding protein that plays a crucial role in neurotransmitter release within the nervous system. Mutations in STXBP1 occur in approximately 1 in 30,000 births, leading to:
A spectrum of neurodevelopmental disorders that may include early-onset epilepsy, developmental delays, and, in some cases, autism spectrum disorder, alongside variable muscle tone and movement disorders. The severity and symptoms experienced by affected children can differ significantly.
Recent attempts to treat DMD and STXBP1 disorders have highlighted significant obstacles that complicate gene therapy for these conditions, suggesting that such challenges may be more prevalent than previously believed.
Dystrophin is a particularly large protein essential for stabilizing muscle cells during contractions and relaxations, making the delivery of the normal gene through gene therapy problematic. The coding region for dystrophin is simply too extensive to fit into the viral vectors currently in use. In response to this challenge, Sarepta Therapeutics has proposed Elevidys, a micro-dystrophin designed to mimic the natural dystrophin. Although it is not identical, its aim is to help protect muscles from damage, and it can fit within the adeno-associated virus (AAV) vectors commonly used for gene therapy. Given the current limitations of gene delivery technology, this represents the most viable approach to treating DMD.
While micro-dystrophin holds promise, the treatment with Elevidys led to the unfortunate deaths of two patients earlier this year due to unrelated liver failures. It appears that these fatalities stemmed from severe immune reactions to the specific AAV variant utilized in the therapy. This brings to mind the tragic case of Jesse Gelsinger, who volunteered for a gene therapy trial in 1999. Jesse experienced an immune-mediated multiple organ failure shortly after receiving a viral vector infusion, an outcome that was unexpected given the assumption that adenoviruses were harmless, which significantly hindered gene therapy research for years.
The future of Elevidys remains uncertain. Furthermore, Sarepta is also exploring gene therapeutic options for limb-girdle muscular dystrophy (LGMD), a condition involving multiple genes—up to 32 unlike the single-gene issue in DMD. Tragically, a patient undergoing treatment for LGMD with Sarepta also died of liver failure in 2025.
Delivering a therapeutic gene to the appropriate cells, tissues, and organs in the body poses a significant engineering challenge often overlooked in gene therapy discussions. This challenge is even more pronounced when the target cells are located in the brain, as is the case with STXBP1, which must cross the blood-brain barrier for effective treatment. Jason Mast underscores this issue, commenting on the convergence of researchers aiming to overcome this fundamental barrier:
Gene therapy researchers were converging on a holy grail. A few years ago, labs reported engineered viruses capable of transporting corrective genes deep into the brain, opening new avenues for treatments of Alzheimer’s, Parkinson’s, and a range of rare genetic diseases.
This summer, after meticulous research, the first individual underwent gene therapy utilizing one of these new viruses. Tragically, the patient, a young child, passed away merely two and a half days later.
This death has sparked apprehension and uncertainty throughout laboratories and companies developing gene therapies for brain disorders, as well as among rare disease communities that have long anticipated these potential cures. The concern is that Capsida Biotherapeutics has unveiled a more significant risk for other viral vectors designed to traverse the brain, which could derail years of progress.
Indeed, Capsida Biotherapeutics employs technology derived from groundbreaking research conducted by Viviana Gradinaru at Caltech. A review of their work states:
Recombinant adeno-associated viruses (AAVs) are frequently used as delivery vehicles in neuroscience. They possess two engineerable features: the capsid (outer shell) and cargo (encapsulated genome). We have identified various engineered AAV capsids with unique capabilities, enhancing targeting to specific cell types and controlling gene expression. These advances enable precise transgene expression for detailed anatomical and functional studies of neural cells and networks, providing a comprehensive toolkit for genetic access to defined brain cell types.
The transgene being used to address the stxbp1 deficiency in the patient is the normal STXBP1. The successful expression of STXBP1 in the targeted brain cells would theoretically correct the defect and result in a “cure.” However, despite rigorous research and engineering efforts behind this transgene delivery, concerns have arisen:
Capsida has opted not to elaborate on the details surrounding the child’s death beyond a brief statement, and its CEO has since departed. The limited information that has emerged raises alarms. The child experienced cerebral edema—swelling of the brain—a clinical progression distinct from prior gene therapy-related fatalities, according to someone familiar with the situation.
Worryingly, none of the animal studies that Capsida previously presented indicated the potential for such a dire outcome, leaving researchers uncertain about how to assess future risks.
This incident represents a particularly significant event within genetic medicine, as articulated by prominent gene therapy researcher Jim Wilson: “This is an outlier, the most material event I’ve seen in the field over the past decade, with no indications this might occur. This is alarming, I’m sure, for everyone involved.”
The path forward for gene therapies must remain cautious yet hopeful. The Capsida method likely possesses future potential, but the lack of a reliable animal model complicates progress. The supposed effectiveness of AAV vectors was thought well-understood; perhaps intricacies between human and animal brains remain uncharted territory. This uncertainty is disconcerting, raising fundamental questions about the nature of genetic engineering.
There is a growing realization that while “genetic engineering” has been a recognized term for nearly fifty years, recent outcomes remind us of its complexities. The “engineering ideal in biology” has existed for over a century; it led influential scientists like Jacques Loeb astray in the past and still poses risks for today’s researchers. Much remains uncertain between a gene’s DNA sequence and its functional expression in a living organism. Moreover, understanding how an engineered AAV capsid behaves in human brains relative to those in animal subjects remains enigmatic. This divergence is crucial, as models often fail to accurately predict outcomes, highlighting the difficulty biomedical scientists face in selecting appropriate experimental approaches.
As cited from previous discussions, the trial with Capsida underscores a “false promise of safety” encountered in gene therapy research. The aftermath of the Capsida trial has cast a shadow on other ongoing projects, including a trial targeting Parkinson’s disease, which has been paused, and collaborations with major pharmaceutical firms that are now in limbo.
Allyson Berent, leading the Foundation for Angelman Syndrome Therapeutics, remarked, “We have reevaluated our next-generation gene therapies due to these developments. The unexpected nature of the incident raises substantial concerns; if our animal models cannot foresee these issues, we face a significant challenge.”
Incidents involving Sarepta/Elevidys and Capsida/STXBP1 cast doubt on the earlier optimism that gene therapy was on the verge of delivering widespread solutions. The case of SCID/ADA deficiency was an easily attainable target, just as successful interventions for hemophilia A and B have predominantly focused on simple liver delivery systems. Whether micro-dystrophin will function effectively in DMD-affected skeletal muscles—along with whether all target muscle sites can be effectively addressed—remains uncertain. While gene-based therapies ranging from CAR-T treatment for cancer to siRNA applications like LEQVIO have shown promise, their development has been characterized by lengthy, incremental processes; the majority of foundational research was publicly funded and not driven by commercial interests.
This raises pertinent questions about the financial models of companies like Capsida Therapeutics, which largely rely on venture capital, as noted in the context of Gilded Rage. With more than $800 million invested since 2020 in companies targeting numerous brain disorders, we must ponder whether the current approach to leading-edge biomedical therapies—from DMD to Alzheimer’s—is the best available. Is the price point of $3,000,000 for Elevidys justified? Would pooling resources through NIH and other global health agencies lead to more innovative outcomes, fostering an open-science ecosystem beneficial to all?
Ultimately, these questions reflect the profound complexities at the intersection of biotechnology, ethics, and commercial interests. The path forward in gene therapy must take a balanced approach, seeking to harness scientific potential while ensuring patient safety and ethical integrity are prioritized.
Notes
[1] Retroviruses (e.g., lentivirus) integrate their genome mostly at random within the chromosomes of infected cells. When elements that control the expression of an external gene (the transgene) are incorporated, the infected cells can produce the protein encoded by that gene, usually at limited levels. If the protein is an enzyme like ADA, this minimal expression is often sufficient for a normal outcome.
[2] A note on genetic terminology: STXPBP1 refers to the normal gene, STXBP1 to the normal protein, and stxbp1 to the mutant gene.
[3] As reported in the New York Times by Sheryl Gay Stolberg, Jesse Gelsinger was not ill at the time of his passing. He had ornithine transcarbamylase (OTC) deficiency, a rare metabolic disorder managed effectively through diet and medication. He volunteered for a trial aimed at testing safety measures for infants suffering a lethal variant of his condition, knowing full well he might not benefit directly. “What’s the worst that can happen to me?” he told a friend before being hospitalized. “I die, and it’s for the babies.” Jesse’s noble spirit remains a poignant reminder of the human stakes in medical research.
