The FDA’s draft path for individualized gene editing is a pre-approved change process for the human genome, and every pre-approved process has a boundary somebody owns.
Suggested Medium tags: Gene Editing, CRISPR, FDA, Biotechnology, Health Policy, Future of Medicine, Rare Disease
Security people have a phrase for a change that has been approved in advance: a standard change. Somebody reviews the pattern once, signs it, and from then on every instance that fits the pattern moves without a fresh review. It is efficient. It is also where I have seen most quiet disasters begin, because the pattern’s boundary is a line drawn by one person on one day, and everything after depends on whether anyone keeps checking where the line sits.
I thought of that when the FDA released its draft guidance on individualized human genome editing products. The agency announced it on Tuesday, October 27, according to the Regulatory Affairs Professionals Society [2]. Strip away the biology and it is a proposal for a standard change process applied to people.
What the FDA is proposing
Today, a gene editing therapy built for one patient’s mutation is treated as its own product. Andrew Byrnes, who directs a gene therapy division in the FDA’s biologics center, called that approach infeasible and unnecessary, and described the guidance as a fix for a critical gap [2]. The draft is titled Individualized Human Genome Editing Products: Considerations for BLA Submission and Postapproval Management. It is a nonbinding Level 1 draft from the Center for Biologics Evaluation and Research, marked not for implementation [1].
The mechanism is a single biologics license application covering many variants. The key idea is the individualized feature space, the range of variation in the editing components across all the variants a sponsor wants covered. If only the guide sequence changes, the space could include any sequence that targets a defined class of variants and produces the same editing outcome [2]. The scope is narrow: monogenic diseases caused by known pathogenic variants, where every variant in the application shares the same editing result [1].
After approval, adding a variant inside the space generally requires a supplement backed by in vitro data, and the FDA recommends a postapproval change management protocol under 21 CFR 601.12(e) [1]. A variant inside the space that is predicted to be adequately safe and active may be treated as a manufacturing change rather than a new product [2]. Expanding the space needs a prior approval supplement, and a change that fundamentally alters the mechanism, such as adding a reverse transcriptase based editor, needs a new application [1].
That is a standard change catalog. The sponsor writes the pattern. The regulator approves it. Patients fitting the pattern follow.
Why I think this is mostly right
I am not here to sneer at it. The alternative is a regulatory model that treats a therapy whose only difference is a handful of letters in a guide sequence as a new drug each time. For a disease affecting a few dozen people worldwide, that model guarantees nobody ever builds the thing. The agency has been explicit that it wants the boundary of the space developed prospectively and supported primarily by clinical data, with in vitro criteria for safety and activity [1]. Off-target editing must be assessed for each guide and editor combination using predefined sequencing-based methods, with the protocol submitted in the application [1]. Manufacturing must be designed to handle every current and future variant in the space, with identity and purity testing validated across it [1]. Those are real controls.
The agency also did not invent this from nothing. The draft sits beside a February 2026 draft guidance on what FDA calls the plausible mechanism framework, which addresses how to generate evidence for individualized therapies targeting genetic conditions with a known biological cause, with clinical evidence from a limited number of patients [3]. The government has been building toward this for months.
Where the boundary gets owned
Now ask the security question: who owns the line?
The individualized feature space is a clinical evidence artifact. It is defined by what the sponsor studied and what the sponsor can justify. The guidance says unstudied variants may be listed only if the clinical evidence reflects the full range of product variants within the proposed space [1]. So a patient with a variant outside the space is not excluded by biology. That patient is excluded by the absence of a study, and a study is a thing somebody pays for.
Look at what exists today. The most recent weekly CRISPR Medicine News roundup lists a gene editing therapy for a rare metabolic disease with six patients dosed, and a second program reporting interim results from eleven patients, which the newsletter itself describes as small, uncontrolled studies [5]. That is the evidence base from which spaces will be drawn. I do not say this to mock the work. I say it because the first spaces will be small, and the fight over their edges will fall on the families whose children are one mutation away from the line.
Who funds the expansion? The draft says adding variants requires in vitro data and a supplement [1], and a group of legal and bioethics scholars writing in Health Affairs Forefront presses on scope and review. They say it is unclear whether one approval covers all variants of a gene, how exclusivities would apply, and whether adding new variants needs approval before use, and they warn that staffing could create a review bottleneck [4]. In short, the draft makes the sponsor the custodian of the list, and the list is also the product.
The access problem hiding in the manufacturing section
The same authors point to a quieter exclusion. The draft expects commercial-level manufacturing standards early, because the first-in-human study may also be the pivotal trial. They report that academic developers, such as the team behind the treatment given to the patient known as Baby KJ, may be effectively excluded, according to Ahrens-Nicklas and Musunuru, and that grants are unlikely to cover full manufacturing costs [4].
Think about what that does. The people who proved the concept would need a corporate partner to carry it across the line, and the partner decides which diseases are worth a feature space. A regulator’s technical bar can function as an access policy without anyone writing the words. I do not think the agency intends that. I think intent is irrelevant once the rule is in the field.
What is missing
Two gaps stand out, and the critics and the document agree on both.
The guidance does not address long-term follow-up. It says outright that the design and conduct of clinical studies are out of scope and that the agency intends to issue further guidance [1]. The Forefront authors want postmarketing requirements for both safety and effectiveness, since commitments without that enforcement authority are weaker, along with registries, possibly modeled on transplant databases, and public safety investigations [4]. They also ask what a serious adverse event in one patient means for related products inside the same space [4]. That question is the one I would put on the whiteboard. If variants share a platform, a harm in one is a signal about all of them. A change management protocol is only as good as the telemetry that triggers a rollback.
The second gap is definition. The authors note the draft demands robust results without defining them or saying how ambiguous or negative outcomes will be handled [4]. Undefined thresholds are where discretion lives, and discretion lives with whoever has the most lawyers.
Standard change catalogs also rot in practice. Entries get added for good reasons, the edges blur, and years later nobody can say why a thing is pre-approved. The draft pushes against that by requiring a formal supplement for each addition [1], which is the right instinct. Whether it holds depends on whether those supplements are visible to anyone besides the sponsor and the agency, which is why the transparency requests in the Forefront series matter more than they first appear [4].
The counterargument deserves its due
Byrnes is right that the per-patient model does not scale [2]. The critics are not asking to scrap the framework. They support the goal and want tighter limits, a notice-and-comment rulemaking rather than guidance alone, and public reporting on which products are approved, for which disorders, and why traditional trials were not feasible [4]. The draft is also only a draft, and the Federal Register notice for the companion plausible mechanism guidance closed comments on April 27 [3]. Nothing I have described is final. That is the best time to argue about edges.
What I would ask for
If I were writing comments, I would ask for three things, all borrowed from operations. First, publish every space as a living document, with its boundary, the evidence behind each edge, and every addition since approval, so outsiders can audit the line. Second, bind the change process to a tripwire: a defined adverse event in any variant triggers review of every variant in the space. Third, fund the edge. If the boundary of a space is where patients are turned away, someone other than the sponsor should be able to pay to move it.
The line is the policy
We are moving from approving molecules to approving processes, and the process has a perimeter. The technology for editing a genome to fit one person is arriving faster than the institutions for deciding whose genome fits. The FDA has handed us the first draft of that perimeter. Whoever draws the line, funds the studies that move it, and watches what happens inside it will shape, more than any lab, which people get to benefit from the science.
About the Author
Len Noe is a Solutions Architect at BeyondTrust, Transhuman, Podcaster, International Cyber Security Speaker, Author, Technical Evangelist, and Biohacker with 11 implanted microchips. A former blackhat and ex-1% motorcycle club member with more than 30 years in technology, he has presented in over 70 countries and is featured in the documentary I Am Machine, which premiered at DEF CON 2025.
References
- U.S. Food and Drug Administration, Center for Biologics Evaluation and Research. “Individualized Human Genome Editing Products: Considerations for BLA Submission and Postapproval Management; Draft Guidance for Industry.” FDA, October 2026. fda.gov
- Eglovitch, Joanne S. “FDA issues draft guidance on individualized gene editing products to fill ‘critical gap.'” Regulatory Focus, Regulatory Affairs Professionals Society, October 27, 2026. raps.org
- U.S. Food and Drug Administration. “Considerations for the Use of the Plausible Mechanism Framework To Develop Individualized Therapies That Target Specific Genetic Conditions With Known Biological Cause; Draft Guidance for Industry; Availability.” Federal Register, 91 FR 9283, February 25, 2026. federalregister.gov
- Fernandez Lynch, Holly; Joffe, Steven; Ramachandran, Reshma; Sachs, Rachel; Zettler, Patricia J. “Refining FDA’s Plausible Mechanism Framework.” Health Affairs Forefront, June 8, 2026, cross-posted by Penn LDI, August 17, 2026. ldi.upenn.edu
- CRISPR Medicine News. “CMN Weekly (23 October 2026): Your Weekly CRISPR Medicine News.” CRISPR Medicine News, October 23, 2026. crisprmedicinenews.com