Points to consider for non-clinical safety evaluation of in vivo gene therapy products
Article ID: 2023-002
Details
Article ID: 2023-002
Advances in gene therapy have led to a wide range of gene therapy products. Hence, it is challenging to use standard non-clinical safety tests, which are typically used to evaluate chemically synthesized products. Therefore, a case-by-case safety evaluation based on the unique characteristics of each product is necessary. In Japan, the fundamental considerations for evaluating the non-clinical safety of gene therapy products are outlined in “Ensuring the Quality and Safety of Gene Therapy Products”. Notably, when designing non-clinical safety studies, it is essential to consider factors such as the efficiency of vector transduction, tissue/cell tropism of the vector, and species differences in the expression level and biological activity of transgene products in the target tissue/organs. However, it is also crucial to understand the limitations of animal studies that arise due to species differences and to assess the safety of expression vectors and expressed genes based on a thorough literature review of their biological properties. The aims of this review are to present non-clinical safety evaluation approaches for gene therapy products as outlined in the Japanese guidelines and key considerations for appropriate non-clinical safety evaluation of in vivo gene therapy products.
For non-clinical safety evaluation of in vivo gene therapy products, it is necessary to select animal species and conduct toxicity studies based on an understanding of the characteristics of the vectors and to understand the limitations of animal experiments because of species differences and interpretation of study results. Safety evaluations should also be based on the biological characteristics of the vectors and expressed genes.
Gene therapy using adeno-associated viral vectors and oncolytic viruses has attracted considerable attention as an innovative medical treatment. However, concerns about its safety remain—such as those surrounding the fatal accident caused by the administration of a high-dose adenovirus vector in the United States [1], as well as the case of leukemia development in an ex vivo gene therapy product for treating X-linked severe combined immunodeficiency in France [2]. Therefore, the development of gene therapy products involves ensuring clinical safety prior to moving to clinical trials in compliance with quality and safety assurance guidelines. Test items and methods for non-clinical safety studies have been established for pharmaceutical products and medical devices; however, these are less well established for gene therapy products. Based on an extensive literature review of gene therapy products, various types and characteristics of vectors and clinical application methods have been identified. Moreover, progress in science and technology has necessitated flexible and rational case-to-case considerations.
In Japan, gene therapy products are defined as products containing a gene to be introduced into human or animal cells so as to express a specific protein for the treatment of a specific disease [3, 4]. These include plasmids, viral vectors, and mRNA. Notably, in Japan, products used for disease prevention (e.g., mRNA, plasmid, or viral vector vaccines for infectious diseases) are categorized as pharmaceuticals, and not as gene therapy products. As of February 2023, the gene therapy products listed in Table 1 were approved for use in Japan [5]. Non-clinical safety evaluation of gene therapy products in Pharmaceutical and Medical Devices Agency (PMDA) is evaluated based on the Japanese guideline of “Ensuring the Quality and Safety of Gene Therapy Products” [6], that was revised in 2019. However, while advancements in science and technology—including clinical trials and regulatory experience—should be considered, the general principles of the guidelines have not yet changed. This review details current concepts regarding the non-clinical safety assessment of gene therapy products in Japan—particularly in vivo gene therapy products.
| Classification of gene therapy products | Brand name | Non-proprietary name | Indication | Sponsor website | Approval date | |
|---|---|---|---|---|---|---|
| Plasmid vector | Corategen® Intramuscular Injection 4 mg | Beperminogene Perplasmid | Ulceration in chronic arterial occlusive disease | Anges, Inc. [7] | 26/03/2019 | |
| Viral vector | Non-proliferating viruses | ZOLGENSMA® Intravenous Injection | Onasemnogene abeparvovec | Spinal muscular atrophy | Novarutis [8] | 19/03/2020 |
| Conditionally replicating viruses | Delitact® | Teserpaturev | Malignant glioma | Daiichi Sankyo Co., Ltd. [9] | 11/06/2021 | |
To assess the safety of in vivo gene therapy products, a thorough understanding of vector characteristics is essential prior to planning a non-clinical safety assessment. The functional properties of the promoter or gene of interest, as well as the origin and physicochemical stability, pathogenicity, cytotoxicity, presence/absence of modifications that alter species specificity or cell/tissue tropism, transduction efficiency, expression level, sustainability of the gene of interest, integration machinery of the vector, and replication competency of the viral vector, should all be considered. Knowledge of these vector characteristics is important for selecting appropriate animal species for non-clinical studies, for interpreting the results of non-clinical studies, and in assessing the risk of chromosomal integration.
Selection of animal speciesTo assess on- and off-target toxicity caused by excessive activity of nucleic acids or proteins derived from the target gene, appropriate animal species must be selected for conducting non-clinical safety studies of in vivo gene therapy products. The criteria for selection include: (1) expression of the target gene in the expression vector in the target cells and tissues; (2) nucleic acids and proteins derived from the target gene exerting expected effects clinically; (3) similar infectivity and tissue/cell tropism between humans and animals when viral vectors are used; and (4) the same administration route as in the clinical setting.
However, it should be noted that human-specific oncolytic viruses may not provide sufficient exposure to assess safety in humans owing to species differences in susceptibility to viral infection and replication and in terms of differences in biodistribution and persistence in the presence or absence of tumors in testing animals. In such cases, following a thorough evaluation of the tumor selectivity of the product, clinical trials should be conducted with appropriate risk-mitigation measures.
In the development of pharmaceutical products, non-clinical safety studies using two species (rodent and non-rodent) are generally necessary to ensure safety in humans because the sensitivity and pharmacokinetics of the test substances may differ between animals and humans. However, for in vivo gene therapy products, the evaluation of only one species may be sufficient in assessing human safety, provided that appropriate species are selected based on general principles. The route of administration for non-clinical safety studies should be similar to the clinical route; however, if it is difficult to carry out the study in rodents for anatomical reasons, systemic effects may be assessed via other routes, and the effects at the site of clinical application may be assessed in large animals.
Use of surrogateWhen the nucleic acid or protein derived from the target gene cannot exert its intended action in animals, owing to species differences in biological activity, an animal-derived homologous gene (hereinafter referred to as a surrogate) may be used. Non-clinical safety studies using these surrogates may be useful for a range of activities—including hazard detection and biomarker identification—in clinical trials. However, it should be noted that surrogate non-clinical safety studies are not appropriate for quantitative risk assessment as it is difficult to compare the biological activities of the target genes in humans with those of homologous genes present in animals.
Dose selectionThe assessment of on-target toxicity caused by nucleic acids or protein expression products derived from the target gene is particularly crucial when determining the dosage. Therefore, it is recommended to set the high dosage based on (1) the dose that maximizes the intended pharmacological effect in the non-clinical study’s utilized species, (2) the maximum tolerated dose (MTD), and (3) the maximum feasible dose (MFD) with reference to “Preclinical Safety Evaluation of Biotechnology-Derived Pharmaceuticals” (the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) S6 (R1) [10]. It is recommended that multiple treatment groups exhibit toxicological dose-response relationships. When assessing the quantitative risk in humans based on non-clinical safety study data, it is necessary to appropriately adjust it while considering the variations between humans and animals (such as the infection efficiency of viral vectors and biological activities of expressed proteins). Quantitative risk assessment is challenging in non-clinical safety studies that use surrogates. The toxicity of the product is evaluated using two groups: a control group and a treatment group.
Administration routeWhenever feasible, it is recommended that the route and frequency of administration be aligned with the intended therapeutic route and frequency of administration to assess safety in humans. However, if it is difficult to conduct a non-clinical study using the same therapeutic route, the results of a non-clinical study using other routes of administration with sufficient exposure can be used to evaluate systemic toxicity. In principle, local tolerance should be evaluated in non-clinical studies using a route of administration that is considered appropriate for evaluating local tolerance or based on the composition of the product.
To ensure the reliability of these studies, it is necessary to collect and prepare study data according to the good laboratory practice (GLP) standards specified in the “Ministerial Ordinance of the Standards for Implementation of Non-clinical Studies on the Safety of Regenerative Medicine Products” [11]. However, it may be challenging to conduct tests under GLP because the test conditions and environments are specific to gene therapy products. In such circumstances, it is necessary to ensure a certain level of reliability by providing an appropriate explanation for why the study cannot be conducted under GLP and, to the extent feasible, obtain results with the same quality as a GLP study.
Similar to pharmaceutical products, general toxicity studies of gene therapeutic products have been performed to ensure their safety in humans. These studies serve to (1) establish initial and maximum doses in clinical trials, (2) identify toxic target organs, (3) establish indicators to identify adverse effects in clinical trials, and (4) clarify the criteria for the discontinuation of clinical trials. When designing a general toxicity study of a gene therapy product, several factors, such as clinical indications and anticipated duration of expression of the target gene, must be considered. Because gene therapy products generally exert their pharmacological effects through expressed proteins, ICH S6 (R1) [10] served as a reference for designing the study. Therefore, a study period of six months is considered sufficient for a general toxicity study, even if the products are expected to have long-term efficacy. It may be possible to evaluate the safety of a product that is expected to have long-term efficacy through toxicity testing over a shorter period of time, considering the persistence of target gene expression, biological characteristics of the target gene, and immunogenicity in animals. The endpoints of the general toxicity study included general condition, body weight, hematology, blood biochemical tests, urine analysis, organ weight, and histopathology of organs/tissues. Histopathological examination should not only include tissues/organs where distribution has been confirmed by biodistribution studies, but also major organs such as the brain, lung, heart, liver, kidney, spleen, testes, and ovaries, as well as the administration site. If any toxicologically significant findings are observed in a general toxicity study, their reversibility should be evaluated with reference to the ICH M3 (R2) Q&A [12]. In addition, adverse effects on major physiological functions—such as the cardiovascular, respiratory, and central nervous systems—should be evaluated.
Gene integration riskGiven that patients develop leukemia after the administration of ex vivo gene therapy products in which genes are transferred by retroviral vectors [2], addressing the risks of inadvertent integration into the genome is critical for evaluating human cancer risk. The relative risk of integration associated with each vector category was based on the biodistribution profile, replication capacity, and integration potential of each vector type [13]. For example, in the case of gene therapy products using retroviruses or lentiviral vectors capable of integrating into chromosomes, it is particularly important to evaluate the number of copies integrated per cell, the possibility of integration into a specific site, and the risk of carcinogenicity due to the integration of the vector DNA into the host chromosome. Even vectors that do not have integration machinery for the host chromosome (such as adenoviruses) may be integrated into the chromosome under conditions such as those of high concentrations [14]. Therefore, it is necessary to evaluate risk in clinical practice based on information such as biological distribution and exposure.
The risk of chromosomal integration varies depending on the degree of host-cell differentiation. It should be noted that the risk is higher in hematopoietic stem cells than in differentiated cells, such as peripheral T cells and muscle cells. Therefore, a careful risk assessment is required when targeting undifferentiated cells. Furthermore, it is important to assess the risk of unintended integration of the vector DNA or transgene into germ cells due to concerns regarding the transmission of this DNA to future generations. Risk assessment of inadvertent germline integration is important when a vector is detected in the gonadal tissues in non-clinical studies. In such cases, the risk of inadvertent germline integration should be assessed based on the frequency of integration and distribution of vectors in gonadal tissues. The evaluation should be guided by the “ICH Considerations General Principles to Address the Risk of Inadvertent Germline Integration of Gene Therapy Vectors” [13].
Assessment of carcinogenicity is required for the target patient population and the duration of clinical use, as per ICH S1A [15]. The possible mechanisms of carcinogenesis induced by in vivo gene therapy products include expression products derived from target genes and the risk of integration into chromosomes. Therefore, a standard carcinogenicity study using chemically synthesized drugs in rodents is inappropriate. Carcinogenicity evaluation should be conducted based on scientific importance, considering the biological characteristics of the wild-type virus from which the viral vector is derived, information on similar products, biological characteristics, mechanism of action of the expressed product, vector insertion site analysis results using cells infected with viral vectors, and general toxicity study results.
The evaluation of reproductive and developmental toxicity is necessary, considering the clinical indications and target population. According to the ICH S5 (R3) guidelines [16], the characteristics of the gene therapy product, such as the pharmacological effects, biological properties of the expressed protein, exposure levels, biodistribution, placental transfer, and risk of integration into germ cells, should be considered, especially when there are concerns about the effects on reproductive organs and future generations.
Evaluation of potential immune-related adverse effects caused by the vector and transgenic expression products should be included in general toxicity studies. Although immune responses can be induced in animal studies, evaluation of the immunogenicity of human proteins in animal studies does not accurately predict immunogenicity in humans. Therefore, the presence of antibodies against the expressed proteins should be evaluated in clinical trials.
This review presents the principles of non-clinical safety assessment of in vivo gene therapy products based on the experience of consultation meetings and a review of marketing authorization conducted by the PMDA. The non-clinical safety evaluation of these products requires a “case-to-case” evaluation based on the characteristics of each product. As there are few commercially available products for gene therapy and their limited use, consultation with regulatory agencies regarding appropriate methods for the non-clinical safety evaluation of gene therapy products early in the development stage is important.
The authors declare no conflict of interest in this report.
We are grateful to the members of the Office of Cellular and Tissue-based Products at the PMDA for their useful discussions. The views expressed herein are the results of independent work and do not necessarily represent the views and findings of the PMDA. This work was supported by the Japan Agency for Medical Research and Development (AMED) Promotion Project and the “Pharmaceutical Regulation and Evaluation Research Project”.
