Food and Drug Administration Guidance on Design of Clinical Trials for Gene Therapy Products with Potential for Genome Integration or Genome Editing and Associated Long-Term Follow-Up of Research Subjects

Abstract

Introduction:

With the burgeoning growth of the gene therapy industry, the Food and Drug Administration (FDA) has produced various guidance documents intended to help gene therapy manufacturers design their preclinical testing and clinical trials to facilitate the process of obtaining marketing approval.

Discussion:

Biosafety professionals and institutional biosafety committees (IBCs) with oversight of clinical trials or biopharmaceutical manufacturing stand to benefit from understanding how these guidance documents set the standard for writing the clinical research protocols that are reviewed by IBCs. Although the FDA guidance documents are typically meant for manufacturers (either pharmaceutical companies serving as research sponsors or investigators at academic institutions), much of the content is useful for biosafety professionals and IBCs during the IBC review process.

Conclusion:

This article specifically addresses guidance documents pertaining to gene therapy vectors capable of genomic integration, testing for replication competent retrovirus, genome editing, and long-term follow-up of research subjects.

Introduction

Although the field of clinical gene therapy is currently experiencing booming growth, it overcame several hurdles to reach its current level of success.^1,2 The turn of the century saw several high-profile serious adverse events reported. In 1999, Jesse Gelsinger tragically died after receiving the highest dose of an adenovirus vector in a phase I dose escalation trial studying a potential treatment for ornithine transcarbamylase deficiency.^3–6

Since then, multiple clinical trials have reported serious adverse events caused by ex vivo transduction of hematopoietic stems cells (HSCs) or T cells with genome integrating retroviral vectors (Table 1).^7–13 These incidents led to a decade of intensive study, redesign, and safety testing of gene therapy technologies that expanded our breadth of knowledge of the science and risks of gene therapy, thus resulting in the development of safer therapies.^14–19

Table 1.

Clonal proliferation and malignancies resulting from retroviral insertional mutagenesis

Transduced cell type	Disease under study	Transgene	Mechanism	Affected gene(s)	Serious adverse event	Research subjects
						No. of SAE	Total subjects
HSC	X-linked SCID	IL2Rγ	Gene activationa	LMO2, CCND2, BMI1	Leukemia5	4	9
HSC	X-linked SCID	IL2Rγ	Gene activationa	LMO2	Leukemia6	1	10
HSC	Wiskott–Aldrich Syndrome	WAS	Gene activationa	LMO2	Leukemia7	9	10
HSC	Chronic granulomatous disease	GP91phox	Gene activationa	EVI1	Myelodysplasia8	2	2
CAR T	Chronic lymphocytic leukemia	CD19 CAR	Gene disruptionb	TET2	Clonal proliferation9	1	26
HSC	β-thalassemia	β-globin	Gene disruptionc	HMGA2	Clonal proliferation10	1	2

The clinical trials listed involved ex vivo transduction of either HSCs or were used to manufacture CAR T cells. The table lists the diseases under the study and the number of research subjects experiencing the serious adverse event listed. Genes affected by insertional mutagenesis are listed as well as the mechanism by which they were affected.

^a The transcriptional enhancer in the LTRs of the integrated retrovirus activated expression of a nearby proto-oncogene.

^b Retroviral integration into an intron of the TET2 gene resulted in expression of a truncated mRNA and inactive form of TET2, a regulator of cellular transcription and proliferation.

^c Disruption of a microRNA recognition element in the HMGA2 3’ UTR resulted in expression of a stable mRNA and enhanced HMGA2 expression.

CAR, chimeric antigen receptor; HSC, hematopoietic stem cells; LTR, long terminal repeat; SAE, serious adverse event; SCID, severe combined immunodeficiency.

As the field matures, and multiple gene therapy products have received Food and Drug Administration (FDA) approval as safe and effective therapeutics, the regulatory environment has transitioned the burden of regulatory oversight for the recombinant DNA aspect of gene therapy products from the National Institutes of Health (NIH) to the FDA. The documents previously required for registration of gene therapy studies with the NIH are now recommended for institutional biosafety committee (IBC) review in Points to Consider: IBC Review of Human Gene Transfer Protocols.²⁰

The NIH Recombinant DNA Advisory Committee no longer reviews individual clinical protocols and was rebranded as the Novel and Exceptional Technology and Research Advisory Committee to emphasize a more general role rather than focusing on recombinant DNA. In 2018, then NIH Director Francis Collins and then FDA Commissioner Scott Gottlieb coauthored an article that was published in the New England Journal of Medicine titled “The Next Phase of Human Gene-Therapy Oversight” that elaborated on the benefits of this new oversight structure.²¹

FDA Guidance Documents on Topics Pertinent to IBC Review

As needed, the FDA issues guidance documents to aid in the development and testing of drugs as well as submission of applications and supporting documents to the FDA. The FDA Center for Biologics Evaluation and Research regulates investigational products containing recombinant or synthetic nucleic acid molecules as biologics and has posted various guidance documents on topics pertinent to IBC review.²²

Guidance documents represent the current thinking of the FDA on the topics covered. They are not regulatory requirements, and alternative approaches can be utilized if they satisfy the requirements of the applicable statutes and regulations. Research sponsors are encouraged to contact the FDA to discuss alternatives to the FDA's recommendations if they better suit the needs of their study and are justifiable.

Unlike the NIH Guidelines for Research Involving Recombinant or Synthetic Nucleic Acid Molecules (NIH Guidelines), the FDA guidance documents do not directly affect IBC review. However, they set the standard for writing the clinical research protocols that are reviewed by IBCs. The FDA guidance documents are typically meant for manufacturers (pharmaceutical companies serving as research sponsors or investigators at institutions) rather than biosafety professionals and IBCs. However, much of the content can be useful to consider during the IBC review process, as it describes the preclinical testing used to justify the clinical study and the requirements for the clinical trial. Each of the following sections pertains to a specific FDA guidance document.

Details on the Design and Manufacture of the Investigational Product

IBCs sometimes face difficulties in obtaining the necessary information to perform a risk assessment. Unlike investigators in biomedical laboratories, clinical investigators typically do not design the investigational product containing the recombinant or synthetic nucleic acid molecules. As such, they are dependent on the study documentation sent to them from the pharmaceutical company serving as the research sponsor or the contract research organization (CRO) running the clinical trial. A detailed description of the design of the investigational product is typically found in the investigator's brochure (IB) in a section titled, Physical, Chemical, and Pharmaceutical Properties and Formulations.

If the IB does not provide sufficiently detailed information for the IBC to perform its risk assessment, the IBC can request additional information from the sponsor or CRO. If study contacts are unable to answer the IBC's questions about the design of the investigational product, it may be beneficial to request the portion of the chemistry manufacturing and controls (CMC) document that details the design and manufacture of the investigational product.²³ The CMC document is required to obtain investigational new drug (IND) status from the FDA. This document is routinely considered confidential, and sponsors or CROs may be hesitant to divulge it without justification.

Details on Testing for Replication Competent Retrovirus

One of the risks associated with the use of retroviral or lentiviral vectors is the potential for generating replication competent retrovirus (RCR) that may propagate in treated subjects resulting in retroviremia or transduction of unintended tissues.^24,25 The FDA issued guidance on “Testing of Retroviral Vector-Based Human Gene Therapy Products for Replication Competent Retrovirus During Product Manufacture and Patient Follow-up,” requiring rigorous testing of viral supernatants and retrovirally transduced cells to ensure they are not contaminated with or capable of producing RCR.²⁶

This guidance treats replication competent lentivirus testing as synonymous to RCR testing. The FDA is now requiring RCR testing of products following FDA approval. The FDA has not found convincing evidence that length of culture influences the likelihood of RCR development in transduced cells and recommends all retrovirally transduced cellular products be tested for RCR even if maintained in culture for several days. In the case of ex vivo transduced cells, the FDA recommends testing at various stages of production from viral supernatant to end of production cells.

The guidance addresses RCR testing methodology. A permissive cell line should be utilized to test retroviral supernatants and retrovirally transduced cells for presence or production of RCR. For cells transduced ex vivo, cocultures with the permissive cell line should be maintained for a minimum of five passages to allow sufficient time for RCR amplification. The amplified material may then be assayed by an appropriate indicator cell line, polymerase chain reaction (PCR) or p24 enzyme linked immunosorbent assay.^26–30 All assays should include relevant positive and negative controls.

The FDA recommends testing at least 5% of viral supernatant volume for each lot tested. The dose administered in clinical trials must contain <1 RCR at a 95% confidence interval. The quantity of transduced cells assayed for RCR should be the lesser of 1% of the lot or 10⁸ cells. PCR can be utilized as an alternative if time constraints preclude use of these preferred testing methods for clinical lot release.

Research subjects receiving retroviral vectors or retrovirally transduced cells should be tested for RCR after administration of the investigational product. Recommended methods include serology for RCR-specific antibodies and analysis of peripheral blood mononuclear cells by PCR for RCR-specific DNA sequences. PCR may be a preferred method as subjects receiving multiple injections of virus producer cells or viral vector may exhibit positive serology results in the absence of RCR.^31–33 Also, immune compromised individuals may fail to produce antibodies against RCR while experiencing RCR viremia.

The recommended testing schedule starts with a baseline sample before receiving the investigational product followed by samples collected at 3, 6, and 12 months after treatment and yearly thereafter for up to 15 years. If all post-treatment assays are negative for an individual subject during the course of the first year, RCR testing may be discontinued and replaced with an annual medical screen. The annual follow-up should focus on clinical outcomes indicative of retroviral disease such as cancer, neurological, or hematological disorders. If research subjects die of a suspected retrovirus-associated disease or develop neoplasia within 15 years, RCR testing should be performed on the neoplastic tissue or pertinent autopsy tissue.

Although the FDA has stood steadfast on the need for RCR testing since 2006, the 2020 version of this guidance eases requirements for RCR testing of established cell banks and allows for manufacturers to petition the FDA for reduction or elimination of testing for ex vivo genetically modified cells.

Genome Editing

The burgeoning growth of genome editing has also prompted the FDA to issue draft guidance on “Human Gene Therapy Products Incorporating Human Genome Editing.”³⁴ Although genome editing technology can be accurate, unintended DNA insertions and/or deletions (indels) or chromosomal rearrangements can take place resulting in possible unanticipated consequences.

Recommendations for Preclinical Design and Testing of Genome Editing Products

Preclinical in vitro and in vivo proof-of-concept studies must be submitted to the FDA with the IND submission. In vitro models are intended to evaluate the activity of the genome editing product in the target cell type. In vivo animal models, which may include a species-specific surrogate (such as when the gene sequence differs between the animal model and humans), should demonstrate a biological response to the investigation product.

If possible, testing of animal models should assay for potential toxicities related to delivery, expression, and genomic modifications associated with the genome editing. Study design should include identification and characterization of off-target activity such as chromosomal rearrangements and their biological consequences. Preclinical animal testing should provide data on biodistribution and shedding of the genome editing drug product.

Assessments of genome editing activity in preclinical in vitro and in vivo models should be designed to assess the:

• Specificity and efficiency of editing in target and nontarget cells,

• functionality of the corrected or expressed gene product (e.g., protein, RNA), if applicable,

• editing efficiency required to achieve the desired biological activity or therapeutic effect,

• durability of the genomic modification and resulting biological response, and

• effects of genetic variation on editing activity across the target population.

Preclinical models should also be utilized to identify and assess risks associated with genome editing by:

• Assessing genomic integrity, including chromosomal rearrangements, large indels, integration of exogenous DNA, and potential oncogenicity or insertional mutagenesis. For ex vivo-modified cells, this may include assessment for clonal expansion and/or unregulated proliferation.

• Evaluating biological consequences associated with on- and off-target editing, as feasible.

• Examining the immunogenicity of the genome editing components and gene product expressed.

• Characterizing the kinetic profile of genome editing components expression and editing activity.

• Assessing the viability and any selective survival advantage of the edited cells.

• Preserving cell functionality after genome editing (e.g., differentiation capacity for progenitor cells).

• Evaluating the potential for inadvertent germline modification.

When manufacturing gene edited cells ex vivo or delivering genome editing components to human research subjects, the FDA recommends minimizing the time of expression or exposure to the genome editing endonuclease to that needed to perform the intended modification to minimize the risk of unintended genomic modifications. Manufacturers are advised to consider restricting expression of the endonuclease to the intended target cell or tissue type.

This could be accomplished with specific mode of delivery to the target cells or tissues, restricting tropism of vectors used to deliver the genome editing components, restricting persistence of the genome editing components in target cells, use of tissue-specific promoters to regulate expression of genome editing components, and use of small molecule inhibitors. The potential for vector-mediated toxicity as well as preexisting immunity to the vector and genome editing components should also be considered.

Human cells that undergo ex vivo genome editing before delivery to humans should be tested for the following:

• On-target editing efficiency, including characterization of the editing events occurring at the on-target site;

• off-target editing frequency;

• chromosomal rearrangements;

• residual genome editing components in the cellular product; and

• total number of genome-edited cells in the batch.

Assays should be developed that measure the properties of the cells and the intended functional outcomes of the genomic modifications.

Subject Safety in Clinical Trials of Genome Editing Products

As human genome editing products may have significant risks and uncertain potential for benefits, first-in-human trials should be designed to enroll subjects for whom no other treatment options are available or acceptable. Factors to consider in determining the study population include:

• The product's mechanism of action in the context of a specific disease.

• The anticipated duration of therapeutic benefit.

• The availability and effectiveness of alternative therapeutic options for the patient population.

• Subjects with severe or advanced disease may be more willing to accept the risks of an investigational genome editing product. However, these subjects may be predisposed to experiencing more toxicities in the form of adverse events or serious adverse events. Furthermore, research subjects with advanced disease may require concomitant treatments that could make the safety or effectiveness data difficult to interpret or serve as confounding variables. Therefore, in some instances, subjects with less advanced or more moderate disease may be appropriate for inclusion in first-in-human clinical trials.

The clinical trial's dosing schedule should be designed to mitigate risks to research subjects by allowing sufficient time between doses within and between dosing cohorts. The staggering interval should be of sufficient duration to monitor for acute and subacute adverse events before treating additional subjects at the same dose, or before increasing the dose in subjects treated subsequently. The expected duration of activity of the genome editing product should be considered when designing the staggered dosing schedule.

The FDA also recommends long-term follow-up (LTFU) of research subjects for at least 15 years.

Long-Term Follow-Up

The FDA issued guidance on “Long-Term Follow-Up After Administration of Human Gene Therapy Products” to guide the design of clinical trials of gene therapy products with a risk of delayed adverse events.³⁵ The guidance document outlines the FDA's risk assessment involving various types of gene therapy products (Figure 1). Characteristics of gene therapy products that may be associated with delayed adverse events include:

Figure 1.

Framework to assess the risk of GT-related delayed adverse events and need for LTFU. The decision tree outlines the FDA's risk assessment process for determining whether a GT product may result in delayed adverse events. LTFU is indicated for GT products deemed to pose a risk of delayed adverse events. Adapted from the FDA guidance, “Long-Term Follow-Up After Administration of Human Gene Therapy Protocols: Guidance for Industry.” FDA, Food and Drug Administration; GT, gene therapy; LTFU, long-term follow-up.

• vectors capable of genomic integration,

• genome editing capability,

• prolonged transgene expression,

• latency, and

• establishment of persistent infection.

Vectors capable of random integration into the human genome (e.g., retrovirus/lentivirus and transposons) have the potential for insertional mutagenesis and activation of proto-oncogenes, thereby increasing the risk for malignancies. Table 2 summarizes the FDA's assessment of integrating vectors and the associated LTFU recommendations. As already stated, genome editing technology has the potential for off-target effects on the genome with potential risk of malignancies or impaired gene function.³⁶

Table 2.

Propensity of commonly used gene therapy products/vectors to modify the host genome

Vector type	Propensity to modify genome	LTFU recommended?	Comments on LTFU
Plasmid	No	No	NA
RNA	No	No	NA
Poxvirus	No	No	NA
Adenovirus	No	No	NA
Adeno-associated virusa	No	Requires product-specific risk assessment	Up to 5 years LTFU for long-term expression without integration
Herpesvirus	No, but may undergo latency/reactivation	Yes	Up to 15 years LTFU if capable of establishing latency
Gamma retrovirus	Yes	Yes	Up to 15 years LTFU
Lentivirus	Yes	Yes	Up to 15 years LTFU
Transposon elements	Yes	Requires product-specific risk assessment	Up to 15 years LTFU
Microbial vectors for gene therapy	No, but may persist and undergo reactivation	Requires product-specific risk assessment	Up to 15 years LTFU if capable of persistent infection
Genome editing products	Yes, permanent changes to the host genome	No	Up to 15 years LTFU

The FDA's determination on whether a vector can modify the genome is based on established literature but does not consider the genetic payload. Gene therapy products that do not include vectors capable of integration or gene editing may still require LTFU if persistent transgene expression may pose a risk to research subjects.

^a Recombinant adeno-associated virus vectors lacking the Rep gene have a low propensity for integration.^46–48 Adapted from FDA guidance, “Long-Term Follow-Up After Administration of Human Gene Therapy Products.”³⁵

FDA, Food and Drug Administration; LTFU, long-term follow-up; NA, not applicable.

Prolonged expression of transgenes may have unintended consequences. For example, growth factors, such as vascular endothelial growth factor or proteins associated with cell division such as p53 modulators, may raise the potential for unregulated cell growth and malignancies due to prolonged exposure to the therapeutic protein. Gene therapy vectors capable of latency, such as herpesviruses, may enter the lytic cycle and create a risk of delayed adverse events related to a symptomatic infection.

Replication competent viruses and bacteria, such as listeria-based bacterial vectors, have the potential to establish persistent infections in immune compromised subjects with potential for a delayed but serious infection. The FDA recommends manufacturers develop quantitative PCR-based preclinical assays for biodistribution, persistence, and shedding capable of detecting ≤50 copies/μg of genomic DNA at a 95% confidence interval. After administration of the gene therapy product, persistence is indicated by detectable levels of product sequences and absence of an apparent downward trend over several time points. In contrast, persistence is unlikely if product sequences cannot be detected or if the assay demonstrates a downward trend over time.

The FDA advises manufacturers to include plans for LTFU or surveillance in applications for marketing approval if the gene therapy products present long-term risks to patients. Such information would also be included in the product label and commercial advertisements to assist physicians and patients in making informed decisions about whether to utilize the therapy after FDA approval.

Discussion

Gene delivery through integrating vectors such as retroviruses and transposons as well as genome editing agents all rely on irreversible modification of the target cell genome, leading to potential genotoxicity wherein the aberrant phenotype may not be apparent for several years after treatment. Retroviral vector-mediated insertional mutagenesis has been observed clinically resulting in clonal proliferation and/or transformation to malignancy in clinical trials performing ex vivo transduction of autologous HSC or T cells. With the emerging use of genome editing products, malignancy as a potential serious adverse event should be anticipated with clinical trial design.

LTFU is an important component of clinical trials for investigational products that pose a risk of delayed adverse events. Although preclinical in vitro and in vivo models are more useful for generating data on proof of concept, efficacy, and short-term toxicity, they are unable to perfectly anticipate the response of a human research subject over an extended period of time. LTFU of research subjects who developed leukemias after ex vivo retroviral transduction of bone marrow cells and T cells has enabled identification of mechanisms involved in the oncogenic etiology. This information led to the design of vectors with improved safety profiles.

For example, self-inactivating retroviral vectors with modified long terminal repeats (LTRs) that are unable to produce RCR and lack LTR promoter function with the potential to dysregulate genes proximal to the viral integration site. Retrovirus-based vectors can also be designed with nonviral, cell type, or tissue-specific promoters to better mimic physiological expression of therapeutic genes, avoiding potential toxicities. Retroviral vectors can also be designed with mechanisms for tracking and destruction of retrovirally transduced cells, including surface expression of epitopes that can be detected by FDA-approved monoclonal antibodies and/or expression of suicide mechanisms that can be induced with administration of small molecules.

Because gene therapy products with genome editing components also pose risks of genotoxicity, clinical trials of these investigational products stand to benefit from the requirement for LTFU of research subjects. As multiple genome editing platforms are currently in clinical use targeting various different genes, there is potential for diverse delayed adverse events. LTFU studies allow for such data to be compiled and studied to enhance the level of understanding of this technology and ultimately result in improved safety profiles.

Benefits of FDA Guidance on IBC Review

The FDA guidance documents are beneficial to biosafety professionals so they can be more informed about the drug development, testing, and regulatory approval process during their review. The guidance documents provide a greater depth of understanding of the biosafety risks and risk mitigation practices inherent in the preclinical development and testing leading up to the writing of the clinical protocol and IB. This knowledge helps biosafety officers better understand what takes place before and after IBC review in terms of managing risks associated with vectors capable of genomic integration or genome editing.

Furthermore, knowledge of the chemistry, manufacturing, and controls information that must be submitted by manufacturers to the FDA as part of an IND application allows biosafety professionals to request pertinent information that may be missing from the materials submitted for IBC review.

What IBCs Do That the FDA Does Not Do

IBCs perform a risk assessment on the proposed research at the institutional or site level and ensure the level of risk and associated risk mitigation plans are acceptable. Although clinical protocols and IRB reviews are focused on human subject protection, the IBC focuses on protecting the research staff, the community, and environment. Training documents help ensure research staff are aware of the risks associated with the investigational product they will be handling and administering to research subjects.

The IBC ensures proper safety practices are in place for handling of the investigational product and appropriate oversight and response plans are in place for incidents such as spills and occupational exposures as exposures to gene transfer vectors can pose risks to the research staff.^37–39 The IBC also provides a mechanism to ensure that the informed consent document adequately relays the risks associated with the gene therapy or genome editing investigational product to the research subject.

IBCs with scientific expertise in gene therapy clinical trials can also serve as an invaluable resource to IRB in assessing risks to research subjects and can serve as an additional safeguard in assuring human subject protection. The IBC's focus on recombinant and synthetic nucleic acid molecules makes it an ideal partner for the IRB in assessing risks in gene therapy trials. Although a major focus of this article has been retroviral vectors, clinical trials utilizing other types of vectors can also cause vector-associated serious adverse events. A death involving adenovirus was discussed in the introduction. Adeno-associated virus (AAV) has been the subject of scrutiny in multiple studies reporting serious adverse events associated with immune responses against AAV.

Gene therapy associated uveitis has been reported in studies administering AAV-based vectors to the eye.^40,41 Other serious adverse events have been reported when high doses on the scale of 10¹⁴ vector genomes per kilogram are administered to research subjects in weakened conditions and/or exhibiting various comorbidities.^42–48 It is also worth noting that although AAV vectors devoid of the Rep gene typically form episomes in the nucleus of transduced cells, they undergo genomic integration at 0.1–1% of transduction events, which has caused some concern about cancer risks. Whereas tumorigenesis was observed in a neonatal mouse model, it has not been observed in large animal studies or human trials.^49–51

IBC review also provides a necessary safeguard to mitigate risks associated with shedding of the investigational product, especially in early phase trials.⁵² Most IND applications undergoing FDA review are excluded from an environmental assessment, and collection of shedding data from gene therapy products is not recommended until Phase I (for replication competent vectors) or Phase II (for replication deficient vectors). Also, the FDA does not specify a timeframe for generating the shedding data.

Although IBC oversight of basic science research typically focuses on containment in the laboratory setting, clinical trials may lead to concerns about shedding once the research subject leaves the research site. Shedding may impact the community, particularly close and intimate contacts of the research subject. Shedding may also trigger environmental concerns such as when the research subject uses a restroom. In a worst-case scenario, shedding data might not be generated until the conclusion of a trial after shedding and exposures have taken place potentially involving many research subjects across a multisite study.

Conclusion

Although the FDA guidance documents are typically meant for manufacturers (either pharmaceutical companies serving as research sponsors or investigators at institutions) rather than biosafety professionals and IBCs, some of the content can be useful in the IBC review process. Biosafety professionals responsible for clinical research oversight stand to benefit from understanding the FDA recommendations for design, preclinical testing, and LTFU for gene therapy vectors capable of genomic integration and genome editing to contribute to the safe evidence-based advancement of gene therapy research.

Authors' Contributions

D.E. is the lead author for the article. S.S. contributed to the introduction, sections on retroviruses, and insertional mutagenesis as well as genome editing. The authors worked jointly on Table 1 and the revisions to the discussion.

Authors' Disclosure Statement

The authors are employed by Advarra, a for-profit entity providing independent IBC and IRB reviews.

Funding Information

No funding was received for this article.