Gene Therapy and CRISPR Regulation

The Asilomar Conference and the Birth of Biotech Self-Regulation

The story of modern gene therapy regulation begins not in Washington but in California — and not with a regulator but with the scientists themselves.

In 1974, a group of leading molecular biologists led by Paul Berg called a temporary moratorium on certain recombinant DNA experiments. The technology was new (recombinant DNA had only been invented in 1973), the risks were poorly understood, and the scientific community was concerned about potential biological hazards.

In February 1975, about 140 scientists, lawyers, and journalists convened at the Asilomar Conference Center in Pacific Grove, California, to discuss whether and how to proceed. The conference produced a set of voluntary guidelines for recombinant DNA research, including containment requirements based on assessed risk.

Asilomar is celebrated as a model of scientific self-governance. The scientific community identified a potential risk, voluntarily paused research, and developed safety standards before anyone was harmed. In 1976, the NIH adopted formal guidelines based on the Asilomar recommendations, which became the foundation of modern recombinant DNA oversight.

But Asilomar is also a model with limits worth understanding. It happened when the relevant scientific community was small enough to fit in one room. The participants shared similar training and similar values. The technology was contained to a few elite labs. None of those conditions hold today. Modern biotech is global, dispersed across hundreds of thousands of researchers, accessible to nation-states and well-funded private actors, and increasingly possible at smaller scales (benchtop DNA synthesis, university CRISPR kits).

The Asilomar model assumes scientists can collectively self-regulate. That assumption may have been correct in 1975 and is increasingly questionable in 2026. BP5 returns to this tension.

---

The Gelsinger Death and the Crisis of Gene Therapy

For its first two decades, gene therapy research operated largely under NIH guidelines and institutional oversight. Then, in September 1999, an 18-year-old patient named Jesse Gelsinger died during a gene therapy clinical trial at the University of Pennsylvania.

Gelsinger had a partial deficiency of an enzyme called OTC. He volunteered for a clinical trial testing gene therapy as a potential treatment. The therapy used an adenoviral vector to deliver the corrective gene. Within hours of the injection, Gelsinger had a catastrophic immune response. He died four days later.

The investigation that followed exposed serious problems with the trial:

• The informed consent process had downplayed risks observed in earlier patients
• Several participants had liver enzyme abnormalities that should have triggered protocol adjustments but didn't
• The lead investigator had significant financial conflicts of interest (equity stake in the company that would have benefited from the trial's success)
• The institutional review board oversight had been inadequate

Gelsinger's death triggered a near-collapse of the gene therapy field. FDA halted multiple gene therapy trials. Public and political confidence cratered. Pharmaceutical companies abandoned gene therapy programs. The field essentially paused for years.

The regulatory response was substantial:

• FDA dramatically increased oversight of gene therapy clinical trials
• The NIH Office of Biotechnology Activities strengthened oversight protocols
• The Recombinant DNA Advisory Committee (RAC) assumed a more active role in reviewing gene therapy protocols
• Informed consent requirements were significantly tightened
• Conflict of interest rules were strengthened
• The FDA established the Office of Cellular, Tissue, and Gene Therapies (now part of the Office of Therapeutic Products) within CBER

The field eventually recovered, but it took roughly a decade. The first gene therapy approved by the FDA after the Gelsinger crisis — Glybera (for a rare lipid disorder) — wasn't approved in Europe until 2012. The first US FDA gene therapy approval, Luxturna (for an inherited retinal disorder), didn't come until 2017. The 18-year gap between the Gelsinger death and the first US gene therapy approval is one of the longest hangovers in biomedical history.

The lesson regulators took: gene therapy trials need extraordinary care, transparent reporting of adverse events, and strict conflict of interest management. The lesson the field took: a single death can set back an entire technology by decades. Both lessons shape gene therapy policy today.

---

The Modern Gene Therapy Approval Framework

Today's gene therapy regulation operates within the standard FDA biologics framework (BP2) but with several specific features:

Investigational New Drug applications for gene therapies receive review from CBER's Office of Therapeutic Products (formerly OTAT). The review includes specific scientific expertise in gene editing, viral vectors, cell manipulation, and long-term follow-up requirements.

Long-term follow-up is required for most gene therapies — often 15 years post-treatment — because the long-term consequences of permanent genetic modification are still poorly understood. Insertional mutagenesis (where a viral vector inserts into a tumor suppressor gene, causing cancer years later) is a known risk for some gene therapies, observed in early trials for severe combined immunodeficiency.

Clinical trial designs for gene therapies often deviate from the standard model. Trials may be small (because the diseases are often rare), single-arm (because randomizing to a sham gene therapy is sometimes ethically impossible), and use surrogate endpoints with accelerated approval. CASGEVY's approval was based on a single-arm trial of 44 patients with sickle cell disease — a tiny dataset by traditional standards, but enough given the severity of the disease and the magnitude of the observed effect.

Manufacturing oversight for gene therapies is extraordinarily complex. Each patient's treatment in some autologous therapies is essentially a custom drug. CASGEVY production involves drawing the patient's bone marrow, editing the cells in a specialized facility, performing quality control, and returning the cells weeks later — all under cGMP standards. The FDA's manufacturing oversight has had to develop new frameworks for this kind of personalized manufacturing.

Specific recent approvals worth knowing:

• Luxturna (2017) — Voretigene neparvovec. For inherited retinal disease caused by RPE65 mutations. AAV-based gene replacement.
• Zolgensma (2019) — Onasemnogene abeparvovec. For spinal muscular atrophy. AAV-based, $2.1 million per dose, then the most expensive drug ever approved.
• Hemgenix (2022) — For hemophilia B. AAV-based, $3.5 million per dose.
• Casgevy and Lyfgenia (December 2023) — Both for sickle cell disease, approved on the same day. Casgevy is CRISPR-based; Lyfgenia uses lentivirus. The simultaneous approval of two distinct gene therapy approaches for the same disease was unprecedented.
• Elevidys (2023) — For Duchenne muscular dystrophy. Accelerated approval, controversial because of mixed efficacy data.

The list grows almost monthly. Gene therapy is no longer experimental in regulatory terms — it's an established product category with multiple approved drugs. The regulatory framework that emerged from the Gelsinger crisis has, by most measures, succeeded in enabling safe gene therapy approvals at scale.

The current challenge is keeping up with pace.

---

Germline Editing, the He Jiankui Case, and What's Coming

All US gene therapy approvals to date have involved somatic cell editing — changes to non-reproductive cells that affect the patient but cannot be passed to offspring. CASGEVY edits bone marrow stem cells. Luxturna edits retinal cells. Zolgensma edits neurons. None of these changes propagate to future generations.

Germline editing — editing reproductive cells (eggs, sperm, or early embryos) — is a fundamentally different category. Germline changes are heritable. They affect every cell of any resulting child, every cell of their children, and onward through generations. The technical capability now exists. The policy infrastructure does not.

US law restricts germline editing primarily through funding restrictions, not direct prohibitions:

• The Dickey-Wicker Amendment (1996, renewed annually) prohibits federal funding for research that destroys human embryos. This has effectively limited federally funded human embryo research, including most germline editing research.
• A rider on FDA appropriations bills (first added in 2015) prohibits FDA from reviewing applications for clinical trials involving heritable genetic modifications to human embryos. This effectively bans clinical germline editing in the US.
• The NIH Recombinant DNA guidelines prohibit federally funded germline modification research.

These restrictions are statutory but indirect. They don't make germline editing illegal — they make it impossible to do federally funded research on it or to bring a germline edit through the FDA. Private germline editing without FDA review would still violate federal law indirectly (any clinical service would require some FDA-regulated component) but the prohibition is not as clean as a direct ban.

In November 2018, Chinese biophysicist He Jiankui announced that he had created the first gene-edited human babies — twin girls whose CCR5 gene had been edited in an attempt to confer HIV resistance. The announcement was met with near-universal condemnation from the scientific community. He was eventually sentenced to three years in Chinese prison for illegal medical practice. The episode triggered global discussions about whether new international governance frameworks were needed for germline editing.

Those discussions have not produced binding international rules. The WHO's Advisory Committee on Developing Global Standards for Governance and Oversight of Human Genome Editing issued recommendations in 2021 but lacks enforcement authority. Most countries have national rules that range from explicit prohibition (most of Europe) to vague restrictions (the US) to permissive frameworks (a small number of countries). The result is a fragmented global landscape where germline editing is effectively governed by professional norms and where it would be done first matters enormously.

This is the kind of regulatory gap that James Hodge and other public health law scholars have flagged repeatedly. The technical capability for germline editing is here. The international consensus that it shouldn't be done in clinical settings is, broadly, here too. But the binding legal infrastructure to actually prevent it — or to govern it if and when it eventually proceeds — is not.

This is the regulatory frontier. Anyone working in biotech policy today is essentially working on the question of how to govern what doesn't yet exist.

---

Wait, Actually...

The CRISPR patent dispute is one of the most consequential intellectual property battles in modern science — and it's still not fully resolved.

After CRISPR-Cas9 was developed as a gene editing tool around 2012, two main groups claimed key patents: the University of California / Berkeley group led by Jennifer Doudna and Emmanuelle Charpentier, and the Broad Institute group led by Feng Zhang. Both groups filed patents on aspects of CRISPR gene editing. Both groups had legitimate scientific claims to fundamental discoveries. Doudna and Charpentier won the 2020 Nobel Prize. Zhang did not.

The US Patent and Trademark Office largely sided with the Broad Institute in 2017, ruling that the Broad's patents covering CRISPR use in eukaryotic cells (the most commercially valuable use) were distinct from the UC/Berkeley patents. Subsequent appeals have largely upheld this position. The result is that CRISPR commercial licensing is divided — and complicated — between competing patent holders.

This isn't just lawyer drama. It affects which companies can license what for which applications. It affects drug development costs (CASGEVY's eventual price tag reflects, in part, patent licensing fees stacked through multiple parties). It affects which research is pursued and where. And it's a reminder that biotech policy isn't just about safety and ethics — it's also about economic structure, intellectual property law, and who profits from publicly funded research.

The original CRISPR research was substantially federally funded. Most modern gene therapy research is substantially federally funded. The drugs that emerge are then sold at prices set by private companies operating under exclusive patent rights. Whether this system is right — whether public funding should produce drugs that the public can afford — is one of the larger unresolved questions in biotech policy.

---