Leveraging Prior Knowledge – When Do Sponsors Not Have to Recreate the Wheel? Breaking Down FDA’s New Draft Guidance From a Nonclinical and Clinical Perspective

On June 2, 2026, FDA (specifically, CBER) issued a new draft guidance titled “Leveraging Prior Knowledge in the Development of Human Gene Therapy Products Incorporating Genome Editing” (the “Draft Guidance”).  As suggested by the title, the Draft Guidance is primarily intended for sponsors developing gene editing products, either ex vivo or in vivo; however, it explains that some of the recommendations may be applicable to other cell and gene therapy products, including AAV vector-based gene therapy products that do not incorporate gene editing.

The Draft Guidance begins with some helpful definitions, summarized below:

• Leveraged: Used to fully or partially alleviate the need to generate new information/data.
• Prior Knowledge: Public knowledge and/or platform knowledge.
• Public knowledge, or “generally accepted scientific knowledge” (“GASK”): “Medical or scientific information that is generally accepted by experts qualified by scientific training and experience in the relevant field including FDA experts.”
• Platform knowledge: “[K]nowledge gained from developing and manufacturing similar products and processes.” (Note: this is distinct from the definition “platform technology” as used in the designation of “designated platform technology,” and the Draft Guidance is clear that such designation is not required to leverage prior knowledge). Platform knowledge may be public knowledge (e.g., GASK), or it may be proprietary and require permission to leverage.

Broadly, the scientific soundness of leveraging prior knowledge depends on the context of use.  Knowledge pertaining to attributes that are independent of the component or product itself may be easier to leverage than knowledge that pertains to dependent attributes of a specific product (e.g., product potency).

This expansive Draft Guidance covers several areas for potential leveraging across CMC, nonclinical, and clinical knowledge.  This post focuses only on the latter two, but we strongly recommend sponsors review the CMC portion closely to understand how they can make use of prior knowledge.

Nonclinical

Prior knowledge can be leveraged to inform the kind, duration, and scope of nonclinical testing needed to enter into clinical testing.  When nonclinical data from other relevant product(s) are leveraged, the Draft Guidance recommends that detailed information on similarities and differences between the products should be included (e.g., manufacturing process, final formulation, key lot release specifications).  Additionally, as appropriate, a justification with supporting data should be provided to explain why additional nonclinical evaluation is unnecessary.

Ex Vivo Gene Editing Products

In general, for ex vivo gene editing products, the Draft Guidance states that data regarding the biological activity associated with specific genomic edits could potentially be leveraged between products containing the same on-target genomic edits using the same editor; however, the biological activity of each editor and gRNA combination should be assessed.  Additionally, independent attributes of the editing component may be considered for leveraging; for example, safety and activity data related to the specificity and affinity/avidity of the binding domain of a genome edited CAR-T product may potentially be leveraged.

In Vivo Gene Editing Products

When considering the appropriateness of leveraging nonclinical information for in vivo gene editing products, it is important to consider similarities such as those related to the editor, vector, manufacturing process, final formulation, the empty/full capsid ratio, route of administration, dose levels, and dosing regimen.

Proof of concept data demonstrating editing and biological activity in a single animal model of a monogenic disease may potentially be leveraged for the purposes of understanding editing feasibility, demonstrating predicted biological response, and supporting the scientific rationale.  However, as with ex vivo products, studies should be conducted for each editor and gRNA combination to evaluate on-target editing efficiency and corresponding biological activity in relevant human cells.

Biodistribution data may also be leveraged when delivery vectors are sufficiently similar and any differences are not expected to impact the biodistribution profile (e.g., across different lipid nanoparticle (“LNP”) products if the physicochemical properties are similar and not expected to be impacted by the differences in cargo; or for viral vectors, if a similar capsid is used and any differences in the expression cassette are not anticipated to significantly impact the gene expression profile).  Such leveraged biodistribution data may assist in identifying tissues/cell types that should be assessed further for on- and off-target editing.

Other Considerations

If the biodistribution profile and relevant product characteristics are sufficiently similar, this leveraging can apply to certain toxicology studies.  For example, leveraging can be used to forgo developmental and reproductive toxicities (“DART”) and germline transmission studies for a product if a product with a similar vector shows no distribution to the gonads.

The Draft Guidance also highlights the potential value of clinical data from relevant products to help inform certain safety risks and dose selection (including for pediatric dosing based on adult data) for the proposed clinical trial, provided sufficient similarity exists between the products; however, fundamental differences in transgene expression and functionality may make leveraging challenging between different transgenes even with an identical vector backbone and expression cassette.

Leveraging Bioinformatics Knowledge

Certain bioinformatics information may potentially be leveraged for similar drug products, including off-target assessment strategies for editors with an identical mechanism of action, certain assay parameters (e.g., amount of starting material) when intended on-target edit rates of different drug products are similar; next-generation sequencing (“NGS”) strategies; or bioinformatics pipelines including several bioinformatics tools, or a specific bioinformatics tool, analyzing similar data obtained from different gene editing products.

It may not be appropriate to leverage bioinformatics data where the data are product-specific.  However, for example, if a proposed product uses the same gRNA and genome editor in a specific cell type as another clinical program, off-target data may be considered for leveraging between ex vivo genome edited cell-based products with identical genome edits.  Similarly, data from genomic integrity studies may also be considered for leveraging, but additional assessment may be needed if the intended edits are known to have altered DNA repair processes due to an underlying disease.

Clinical

Data from prior clinical experience may potentially be leveraged to inform certain aspects of clinical trial design, including for dose-limiting toxicity definitions, monitoring, dose selection, appropriateness of re-dosing, exposure-response evaluations, the type and timing of biomarker and clinical outcome assessments, and clinical pharmacology assessments.  Clinical data can also potentially be leveraged to abbreviate the safety and/or efficacy database or for PK and drug-drug interaction concerns.

The Draft Guidance also describes an openness to “reconsider” long-term follow-up requirements, including duration, based on leveraged cumulative safety data “to reduce unnecessary burden.” This is noteworthy because the FDA Guidance Long Term Follow-Up After Administration of Human Gene Therapy Products includes lengthy follow-up recommendations.  As relevant here, the recommendation in that guidance for genome editing products is up to fifteen years, which is not only burdensome for both patients and sponsors, but also practically challenging given the constantly changing corporate landscape.  If this provides a path to reducing these requirements where appropriate, it would be welcomed.

How to Submit Information for Proposed Leveraging

To leverage public knowledge, the referenced material should be provided for FDA evaluation, and sponsors should provide a summary of literature sources and justify the applicability to their specific product, which may require bridging data.  Platform data may be included in the IND submission or may be submitted as a reference to a different IND or Master File (with letter of authorization, if needed).

The Draft Guidance states that BLAs may not incorporate by reference without authorization information or data that applies to drug substances, drug substance intermediates, and drug products; however, such information may still be considered for leveraging if it is provided directly in the BLA.

In general, the Draft Guidance recommends engaging with FDA as early as possible on these topics, including through INTERACT or pre-IND meetings.

Analysis

We strongly applaud FDA for publishing this Draft Guidance.  The safest regulatory route for a sponsor developing a product, particularly a biologic where there is no such thing as a 505(b)(2) pathway to rely on previous FDA conclusions of safety and/or efficacy, is to recreate the wheel with each product so that there is no question that an application is supported by specific data from that product, even where there is knowledge that could potentially be leveraged to substitute for unnecessary and burdensome testing.  FDA’s signaling of openness to this will hopefully lead to productive discussions.  We were also pleased to see that although the Draft Guidance is focused on gene editing products, the recommendations are not necessarily so limited.

The Draft Guidance did not answer one vexing question that has persistently risen in the context of leveraging evidence, which is: when does “knowledge” become “public” or GASK?  This is a crucial question, because this may be the only way for a sponsor to leverage prior knowledge that it did not itself develop, as authorizations from competitors can be challenging to acquire.  As such, the scope of what information is available for leveraging is not entirely clear.  FDA’s GASK guidance also does not specifically answer this question.  Sponsors may seek to support their results with evidence that a particular well-understood mechanism of action provides plausibility for such results, but we have seen resistance to acceptance of such evidence in the biologic context.  The Draft Guidance here only repeats that such knowledge must be generally accepted by experts; however, without clear articulation of how that process is actualized, we have seen FDA reviewers default to “no,” and thus resulting in another re-creation of the wheel.

Additionally, in describing how prior clinical data may be leveraged to inform trial design and analysis, it is somewhat notable that there is no reference to FDA’s recent guidance on using Bayesian methodologies.  It would be interesting to understand FDA’s position on acceptability of priors from products with the same or highly similar features for leveraging in the manner described elsewhere herein (is this included in how clinical data could be “leveraged” to “abbreviat[e] the…efficacy database”?).  The Bayesian guidance touches on this regarding skeptical priors for failed trials involving closely related drugs, but how could the opposite situation be leveraged?

Another fundamental question here is what will “leveraging” really mean in practice?  It’s a wide range from some amount of “partial alleviation” to “full alleviation.”  This will certainly be case-specific, but we suspect, at least at first, there will be less appetite to allow for “full” alleviation of the need to generate new information/data as compared to “partial” alleviation.  Ultimately, the success of the principles espoused in this Draft Guidance will depend on the acceptance by FDA review staff.  A different product may never quite be similar enough to justify leveraging and streamlining.  But hopefully, with increased encouragement from such efforts as this Draft Guidance, we can see progress.  We encourage sponsors to think expansively, and early, about this.

Many sponsors in the gene editing space, or gene therapy more broadly, use a limited number of delivery platform types (if more than just one); for example, sponsors that are developing multiple products with the same vector and route of administration.  As such, sponsors tend to accumulate substantial knowledge relevant to their specific delivery platform(s).  This Draft Guidance therefore has the potential to be truly impactful.  We strongly encourage sponsors to reach out to FDA about taking advantage of the Draft Guidance early in development, because when you can use the wheels you’ve already built, you get to the finish line faster.