Non-clinical, quality and environmental impact assessments of cell and gene therapy products: Report on the 5th Asia Partnership Conference of Regenerative Medicine - April 7, 2022

Yoshitsugu Shitaka, a presenter of the Forum for Innovative Regenerative Medicine, provided an update on the development and regulations of cell and gene therapy (CGT) products in Japan. Five products have been approved within the last year, bringing the total number up to 16 as of March 2022. Twelve of these products have been fully approved, whereas the other four products were given conditional and time-limited approvals. Since 2019, four autologous chimeric antigen receptor-T (CAR-T) cell products have been approved. The development in gene therapy, especially in the category of adeno-associated virus, has been increasing rapidly in line with the global active trend. The regulatory system related to the Cartagena Act, which requires developers to carry out the assessment of the environmental impact of the use of living modified organisms, has improved significantly through exchange of views between the regulators and the industry. One improvement is that it has become possible to apply for partial change by obtaining viral shedding data during clinical stages. Another improvement is an abolishment of the confirmation step of a draft application of the type 1 usage for the Cartagena Act by Pharmaceutical and Medical Device Agency (PMDA). With these improvements, developers are able to initiate clinical trials more easily and to optimize the measures against viral shedding at the clinical trial stages.

Herui Wang, a presenter from the China Medicinal Biotech Association, shared updated regulations, standards developed by the industrial association and approved RM products in China. The Biosecurity law, published in 2021, requires researchers and developers to register equipment and materials associated with biotechnology research development and application, such as gene sequences, cells and organs. Developers also need to pay attention to the new pharmacopoeia volume III and IV because these specifically stipulate biological products required throughout production. Volume IV focuses more on the general requirements for pharmaceutical excipient, drug validation and reference materials. In addition, three trial guidelines for gene therapy have been issued by the Center for Drug Evaluation, which address non-clinical study and evaluation of human gene therapy, non-clinical study and evaluation of cell therapy products with gene modifications and long-term follow-up for gene therapy products. Apart from the governmental guidelines, an industrial standard with a focus on the ethical assessment of stem cell sources has been published. Developers are required to protect the privacy of donors and conduct ethical assessment through qualified organizations. Two autologous CAR-T cell products, Yescarta and Carteyva, have been approved in 2021. Nine mesenchymal stem cell products derived from the umbilical cord, bone marrow, adipose cells or other sources are in clinical trials for the treatment of immune-associated diseases.

Bryan Choi, a presenter from the Council for Advanced Regenerative Medicine and Regenerative Medicine Acceleration Foundation, provided an update on product development and regulations in Korea. In 2021, Kymriah, Zolgensma and Luxturna were approved, and a total of 18 products had been approved for the market since the first approval in 2001. Among these products, six products, including Kymriah and Zolgensma, are covered by the National Health Insurance. Luxturna is currently under the review process. It is estimated that >75% of Korean citizens are covered by private health insurance, but the amount of insurance coverage for expensive cell and gene therapies, such as Kymriah and Zolgensma, is very limited. All the 18 products were newly categorized as advanced biopharmaceuticals (ABPs) according to the Advanced Regenerative Bio Act enacted in August 2020. It consists of two tracks, the Advanced Regenerative Medicine track for clinical research, controlled by the Ministry of Health and Welfare, and the ABP track for commercial trials, controlled by Ministry of Food and Drug Safety (MFDS). Both tracks deal with the same categories of cell therapy, gene therapy, tissue-engineering therapy and combination therapy. The Advanced Regenerative Medicine track classifies therapies into low, middle or high group, depending on the risk level. Cell produced using advanced technologies, such as embryonic stem cells (ESCs), induced pluripotent stem cells (iPSCs) and CAR-T cells belong to the high-risk group. Only certified hospitals and clinics can submit clinical research proposal, and only those cell processing centers certified by the MFDS can provide therapeutic cells. The ABP track manages the safety during the entire product life cycle. Regarding the procurement of raw material cells, a new certification for "Management Business for Human Cells etc." is a mandatory license required to procure cells and tissues from hospitals, process down in Good Manufacturing Practice (GMP) facility, and supply as raw materials to cell therapy companies. Cell therapy companies need the license to directly procure cells from hospitals. MFDS established three special review schemes to promote ABP therapies. The tailored review adopts the rolling review process, and the priority review expedites the review process. The conditional approval allows for market approval after phase 2 trial, with a condition that the phase 3 trial will be conducted.

Wong Kum Cheun, from the Singapore Association of Pharmaceuticals Industries, presented the current national CGT regulations and the environmental risk assessment. The cell, tissue and gene therapeutic products (CTGTPs) are regulated under the Health Products Act that was implemented from March 1, 2021. The Health Products Act covers broad areas, which include manufacturing, import, supply, presentation, registration, duties and obligations of manufacturers or importers and certification. CTGTPs are stratified into two classes. The class I products, which have lower risk, are minimally manipulated, intended for homologous use and not combined or used in conjunction with therapeutic products or medical devices. The class II products include high-risk products, such as gene-modified cells or vectors with therapeutic genes. The class II products can go through two evaluation routes: the full pathway, which is for a new product that has not been approved by any comparable overseas regulator, or the abridged pathway, which is for a product that has been approved by at least one of the comparable overseas regulators. These regulators include the Therapeutic Goods Administration of Australia, Health Canada, US Food and Drug Administration (FDA), European Medicines Agency (EMA) and Medicines and Healthcare products Regulatory Agency. According to the Appendix 8 of the Health Sciences Authority (HSA) guidance, CTGTPs containing GMOs, such as viral vectors, require environmental risk assessment (ERA) before clinical trials and product registrations. Applicants must submit an ERA application to the Genetic Modification Advisory Committee This procedure normally follows the EU guidelines for the ERA assessment and the regulatory requirements for Genetic Modification Advisory Committee recommendations. This forms part of the dossier submission to HSA for clinical trials or product registration applications.

Dr. Pawan Kumar Gupta, of the Association of Biotechnology Led Enterprises in India, presented the national regulation and approval processes for CGT products. According to the New Drugs and Clinical Trials Rules, 2019, issued by the Central Drugs Standard Control Organization (CDSCO), any stem cell–derived products, gene therapeutic products or xenografts intended for use will always remain a new drug. A new drug means any drug that is not yet approved by Central Licensing authority under Drugs and Cosmetic Act 1940. The clarification document issued by the government of India, dated February 9, 2021, states that a stem cell–derived product refers to a drug processed by means of substantial or more than minimal manipulation or manipulation by genetic engineering. The first step of the approval process for cell- or stem cell–derived products from Indian manufacturers is to get the GMP facility approval. Manufacturers have to submit applications for the manufacturing of experimental test batches to the Central Licensing Authority. Pre-clinical studies can be conducted with cells manufactured in approved facilities. Applications for clinical trials are submitted with clinical batches manufactured in the approved facility. The review processes are conducted within 30 days. For developers outside India, applications are submitted, in addition to a clinical trial form, to import finished formulations for test and analysis and these are reviewed within 90 days. For new drugs approved outside India, phase 3 studies need to be conducted to generate evidence of efficacy and safety of the drug in Indian patients. Exposure to Indian patients is required for approval of New Drugs in India. There are several approval systems for expedited patient use in unmet medical needs, and accelerated approval is considered for serious, life-threatening or rare diseases. The conditional approval is dependent on the nature of products. Orphan drugs are defined as drugs for diseases affecting not more than 500 000 patients. As of April in 2022, there are five approved cell- or stem cell–derived products in India.

Thai-Yen Ling, a presenter of Taiwan Association for Cellular Therapy (TACT) and a Director of Department of Pharmacology College of Medicine, National Taiwan University, spoke on the current status of cell therapy and RM in Taiwan. One of the most important regulations for cell therapy is the administrative interim measures of inspection and testing of medical instruments or use of special medical techniques, referred to as “Regulations Governing the Application of Specific Medical Technique and Medical Device,” which was established in 2018. In the regulation, six types of cells that are considered safe to use as autologous can be administered to patients for autograft therapy. These include CD34-positive hematopoietic stem cells, immune cells, adipose-derived mesenchymal stem cells, fibroblasts, bone marrow–derived mesenchymal cells and chondrocytes. According to statistics on the application and approval for cell therapy, >300 cases have been submitted for application, and 120 cases have currently received the approval. Aside from autologous cell therapy, two drugs, Zolgensma and Kymriah, have been approved for RM, whereas 80 cases for cell therapy and 28 cases for gene therapy are under clinical trials. Two RM Acts are expected to be passed by legislature in 2023: Regulations on the Implementation of Regenerative Medicine focus on medical treatment protocols, and Regulations on the Administration of Regenerative Medicinal products focus on cellular products.

WG 1 involves the non-clinical assessment of RM products and the harmonization across Asia. We kicked off our activity focusing on CAR-T products in July 2021, with 38 experts in the industry gathering from five countries/regions. The panel discussion was conducted in two parts: Aim 1: Comparing the non-clinical test packages of CAR-T product required for IND submission among Asian countries/regions and Aim 2: Discussing the non-clinical assessments required for next-generation CAR-T products. For a fruitful discussion within a limited time, question/discussion points were shared with regulators in advance. (supplementary Table 3).

Since the approval of the world's first CAR-T product, Kymriah, in 2017, the development of CAR-T products has become active worldwide. CD19–CAR-T products are most advanced in progress with some have been launched or are in late clinical studies. There are other CAR-T products come close in development targeting hematological malignancy.

To conduct the survey for the IND submission, WG1 sets a virtual CAR-T product with a similar profile to CD19–CAR-T (supplementary Figure 1), and presented non-clinical packages following the available guideline of each country/region.

Based on these survey outcomes, we extracted discussion points and questions for regulators to elucidate their viewpoints on the non-clinical evaluation of CAR-T products. The discussion was further divided into pharmacology, cellular kinetics and biodistribution (CKBD) and safety.

In the pharmacology session, three specific topics were discussed. The first topic was the selection of CAR-T cell source for in vitro and in vivo studies. CAR-T cells derived from patients are ideal, but they are not easily obtained due to limited accessibility, compared with those derived from healthy volunteers or commercially available peripheral blood mononuclear cells. In the WG1 survey, all the respondents considered that non-clinical pharmacology studies, in which CAR-T cells derived from healthy volunteers or commercially available peripheral blood mononuclear cells are used as cell sources, should be acceptable. The answers from Japanese, Korean, Taiwan and Singaporean regulators were similar to the WG1 survey results, suggesting there is no need to use CAR-T cells derived from patients. Moreover, the Indian regulator commented about the selection of cell source for CAR-T as follows:

In WG1 industry survey for IND-enabling study packages of the virtual CAR-T product, all the countries/regions stated that in vivo cellular kinetics is necessary, and some similarities and differences were found in the details of the study conditions. At the panel discussion with Asian authorities, the following three topics were discussed:

The first topic was the rationale for the time points of blood and tissue sampling in CKBD study. The WG1 industry survey showed that the time points were dependent on the countries/regions, and each country/region has its own rationales, and we asked Asian authorities if they have solutions to this seemingly simple challenge. All regulators, except for the Japanese regulator, answered “Yes, we have rationales for time points of blood and tissue sampling.” The comments or rationales are summarized as follows:

Regarding non-clinical toxicology data package for virtual CAR-T cell products, WG1 survey in the industry suggested that there are slightly different viewpoints on the data package potentially required among countries/regions. Two specific topics, human tissue cross-reactivity test (TCR) and immunogenicity, were discussed in this session.

Traditionally, TCR has been considered as a standard assay to assess on-target/off-target and off-tumor binding potential of the binders of biological products; however, it is not clear whether this approach can be applied to CAR-T cell therapy. In addition, plasma membrane protein array (PMPA) has recently been developed for on-target and off-target binding assessment, which is a cell array that expresses >5500 human plasma membrane proteins []. The question to health authorities for this topic was “when it is known that the target CDxx is highly specific to B cells, based on the literature, is the human TCR study necessary to initiate a clinical trial of CDxx CAR-T cell product, in addition to PMPA that is separately conducted?” The regulators from India and Korea answered “yes,” and those from Japan, Taiwan and Singapore answered “no.” The importance of the reliable information of characteristic and expression patterns of target antigen was also noticed. Key comments are shown as follows:

In general, immunogenicity can be assessed in clinical studies. An example of non-clinical assessment for immunogenicity that was proposed by some industry members was to measure anti-CDxx scFv antibody from mice administered mouse surrogate CAR-T cells. The question on this topic was “Do the health authorities think non-clinical immunogenicity assessment is necessary for the initiation of a clinical trial of CDxx–CAR-T cell product?” and “If necessary, what kind of studies/assays do the authorities consider useful?” The regulator from India answered “yes,” whereas other regulators from Japan, Korea, Taiwan and Singapore answered “no,” recognizing that non-clinical immunogenicity studies provide limited knowledge for assessing immunogenicity potential in humans in most cases. Key comments are shown as follows:

Although some products have successfully been used in the treatment of hematological malignancy, CAR-T therapy still has some limitations. The first group of limitations includes the high cost, long production time, quality variability and production failure, which are associated with the autologous nature of the products. The second group of limitations are associated with efficacy and safety. Current CAR-T products are only effective against hematological malignancy and have several safety concerns, such as cytokine-releasing syndrome, neurotoxicity, and potential tumorigenesis. To address these limitations, several approaches have been under trial as next-generation CAR-T therapies.

This panel discussion involved non-clinical considerations for four product categories: (i) allogeneic CAR-T, (ii) iPSC-derived CAR-T for off-the-shelf CAR-T, (iii) solid tumor and (iv) other components for effective and safe CAR-T.

Allogeneic, off-the-shelf CAR-T products have the potential to address various limitations of autologous products, such as high price, long production time, quality variability and manufacturing failure. Donor-derived PBMCs and cord blood, both containing differentiated immune cells or CD34⁺ hematopoietic stem cells, are used as cell sources. Because of the limited proliferation capacity of these cells, the batch size of donor-derived products is relatively small, and this causes batch-to-batch differences. Using pluripotent stem cells that have infinite proliferation potential, such as iPSCs and ESCs, is one of the options to reduce batch-to-batch differences.

Regardless of the cell sources, allogeneic T-cell receptors (TCRs) cause GVHD, and, thus, TCRs should be knocked out/down. Using other cell types with invariant TCRs (γδT, natural killer [NK]-T, MAIT cells, etc.) or without TCRs (NK cells) is another option to avoid GVHD. Human leukocyte antigen molecules of allogeneic cells also should be knocked out/down to avoid rejection from the host immune system.

The first discussion point related to the quality control of the product was about cell viability after cryopreservation of the products. NK cells, γδT cells or other innate immune cells suitable for allogeneic use are generally cryopreservation-sensitive, and low levels of post-thaw cell viability before infusing back to patients is a potential concern. We asked the regulators whether they would accept products with ˂70% cell viability, which is considered as a reference standard for cell therapy products, including autologous CAR-T, and all the regulators answered “no.” This was a challenge from the industry, and we understood that there is no space for discussion about this topic.

The second discussion point was on the batch-to-batch differences of the product. Particularly for donor-derived allogeneic products, the batch-to-batch differences can potentially affect the results of non-clinical studies. We asked the regulators whether using three batches of the products in a study is sufficient to confirm the efficacy and safety of the product. Contrary to our concern, no regulators required more than three batches for non-clinical studies. The regulators from Japan, Korea, Singapore and Taiwan responded that the study with a single batch of the product could be accepted if the comparability between batches is assured. India responded that a study involving three batches is normally required according to the law.

The third discussion point of this topic was about GVHD risk assessment. GVHD is one of the most important safety concerns of allogeneic CAR-T products. However, as this is an immune reaction between human cells, an in vivo study administering human cells to immunocompetent or immunodeficient animals cannot work. We asked the regulators how we can demonstrate the reduced risk of the product for GVHD, for example, in the TCR-knockout CAR-T cells. In fact, based on past experience, knocking out TCRs can effectively reduce GVHD risk. We presented three options: A: in vivo study using human CAR-T cells and human target cells engrafted into immunodeficient animals; B: in vitro study using human CAR-T cells and human target cells and C: no experimental demonstration is required. The answers differed among countries/regions. Korea, Taiwan and Singapore stated that the in vitro study alone is sufficient and that there was no need for the in vivo study. India responded that in vivo study is required. Japan responded that it is not appropriate to assess the risk of GVHD in humans in a non-clinical study, and, hence, they stated that such a study is not required. The details of comments from regulators of India and Japan are shown to follow:

iPSC-derived cell therapy products are generally considered to have a high potential risk of tumorigenicity because residual undifferentiated iPSCs are naturally tumorigenic, giving rise to teratoma and genetic variations (e.g., genome instability, single nucleotide variations, copy number variation, and loss of heterozygosity). They can be introduced and accumulated from their parental cells, during the reprogramming process and during prolonged culture in vitro. Therefore, the evaluation of the potential risk of tumorigenicity and the characterization of the cell banks/final products from safety/quality control perspective would be the major concerns of iPSC-derived cell therapy products. From this viewpoint, the following two topics were discussed in this session.

The first discussion point was regarding the in vivo tumorigenicity study. Considering the high potential risk of tumorigenicity, a long-term in vivo tumorigenicity study is required for iPSC-derived cell therapy products. Therefore, we asked regulators what would be the appropriate study duration for the in vivo tumorigenicity study. The comments from the regulators were consistent; they responded that the study should be conducted through the life span of the animals (e.g., ≥12 months). However, Taiwan and Japan also mentioned that the study can be terminated when cell elimination is clearly demonstrated in CKBD study. The details of the comments from the regulators are shown to follow.

CAR T-cell therapy has accomplished considerable success in the treatment of hematological cancers. However, it has yet to achieve the same level of success in the treatment of solid tumors. As solid tumors account for >90% of all cancer types, its successful treatment represents a true breakthrough in the unmet need. The complicated tumor microenvironment of solid tumors has made the path to success filled with struggles. Many great challenges are associated with the use of CAR T-cell therapies in patients with solid tumor. These challenges (in the efficacy aspects) include the heterogeneity of solid tumors, potential on-target off-tumor toxicities associated with the targeting of T cells to tumor-associated antigens (TAAs), poor T-cell trafficking to the tumor sites, poor T-cell infiltration into the tumor mass, poor survival of T cells within the immunosuppressive tumor microenvironment (TME) due to the acidic and hypoxic conditions and the abundant immunosuppressive cells and inhibitory extracellular molecules present in the TME. Safety concerns regarding currently available CAR T-cell therapies, including those associated with cytokine release syndrome and neurotoxicity, have also remained unresolved.

In this context, armored CAR-T cells are important, as they have the potential to resolve all these challenges. In order to make CAR-T promising for solid tumor therapy, two points were discussed in the WG: (i) how to demonstrate, in the non-clinical studies, in vivo pharmacology results that are clinically relevant and can be extrapolated for clinical efficacy consideration, and (ii) how to evaluate the safety of a selected TAA for minimal on-target off-tumor toxicity.

In the first discussion point about the appropriate evaluation of the in vivo efficacy of CAR-T products, it was agreed that choosing an animal model that accurately represents the complicated tumor microenvironment is crucial. However, none of the currently available models, including cell-derived mouse model (CDX), patient-derived tumor xenograft mouse model (PDX), and syngeneic mouse models, seems ideal.

Syngeneic mouse models are established by implanting mouse tumor cells in immunocompetent mice sharing the same genetic background. These models would be the most suitable model for murine TME; however, the biology of the murine model does not accurately recapitulate human biology and mechanism of action of human CAR-T. PDX mouse models are established by implanting patient-derived tumors in immunodeficient mice. Although these models more accurately recapitulate human TME (than syngeneic mouse models), they do not include an (human host's) immune system. Nevertheless, it is considered more clinically relevant than CDX models. However, both the reliability and reproducibility of currently available PDX models are still relatively low. CDX mouse models are established by implanting human tumor cell line into immunodeficient mice. Although the nature of cell line–derived tumor is different from that of human tumor in several aspects, it has long been used in the oncology field for screening and as an evaluating platform of anti-tumor therapeutics. Two directions have been proposed for further discussion with the regulatory bodies.

Regulators from Japan considered the CDX model to be acceptable because neither syngeneic mouse model nor PDX model can accurately recapitulate the tumor microenvironment found in human patients. The regulators from India, Taiwan, Korea and Singapore recommended that additional in vitro pharmacodynamic data should be added as a requirement, in addition to the use of the CDX model, for efficacy evaluation.

The second discussion point focused on the safety aspect of TAA evaluation, more specifically with respect to evaluation of the on-target off-tumor or off-target toxicities. It was agreed that the binder(s) selection of the gene of interest design is the key for successful targeting with reduced safety risks. Most target molecules are not tumor-specific and TAA will not only be overexpressed in tumor cells but also in some healthy tissues (albeit often at lower levels). As such, there is a high chance of inducing “on-target off-tumor” toxicities resulting from the cross-reactive binding of CAR-T to normal tissues that also expressed the targeted TAA. Therefore, the binder(s) will be selected based on the sufficient affinity to the target TAA overexpressed on tumors but with lower levels of binding (unwanted recognition) to healthy tissues. We carried out in depth discussions regarding the studies that may be suitable for the evaluation of on-target off-tumor or off-target toxicities for the first IND. We also sought the opinions from the regulatory bodies.

The regulators from Korea and Taiwan accepted the evaluation protocol, which involves conducting (sequentially) the risk assessment to understand the expression level of the targeted TAA in major organs/tissues, followed by the evaluation of the in vitro cytotoxicity of the targeting CAR-T cells against cell lines without the expression of the targeted TAA(s) and monitoring the release of cytokines in in vivo pharmacological and toxicological studies. The regulators also mentioned that the level of target antigen expression in each tissue should be identified. If the cytotoxicity or binding activity has been observed in certain tissues and attributed to the CAR-T cells or scFv, respectively, the in vitro and in vivo cytotoxicity data on these relevant tissues or organs should be required for further analysis. The regulators from Taiwan also recommended the analysis of TAA expression patterns, including the level, timing and location of the targeted TAA in patient samples or made it a requirement that other additional data be provided for the evaluation of the potential toxicities. The regulators from India agreed with the aforementioned evaluation procedure but recommended to additionally require evaluation of cytotoxicity against healthy cell types (primary cells or iPSC-derived cells), such as those found in the central nervous system, heart, liver, lung, skin, vascular, blood and bone marrow.

CAR-T cell therapies are evolving to improve T-cell activation and persistence, overcome immunosuppression and mitigate toxicities. Several new approaches are extensively explored to enhance the therapeutic outcome of CAR-T cell therapy. In this “other component” part, the challenges against a successful safety and CKBD evaluation of a novel CAR-T product expressing other molecules, in addition to CAR molecule, were discussed. For this purpose, a model product 1928z CAR-T was established. 1928z CAR-T cells are genetically manipulated to express 4-1BBL on their surface to enhance antitumor activity and persistence of the cells. Expressed 4-1BBLs act on neighboring activated CAR-T cells by engaging 4-1BB receptors expressed on their cell surface. These 4-1BBL is membrane-bound and not secreted. It is reported that IgG-based 4-1BB agonistic monoclonal antibodies can cause liver toxicity, suggesting the potential safety risk of this approach. Since human 4-1BBL does not cross-react with mouse 4-1bb receptor, potential toxicity of 4-1BBL cannot be evaluated when using mice in non-clinical tests.

The first point discussed was about the requirement of in vivo toxicity study. The query to regulators was “Is there a need for a separate in vivo toxicity study to evaluate these CAR-T cells expressing costimulatory molecules that do not cross-react with the murine counterpart?” and the following three options were discussed: (A) toxicity study for conventional CAR-T is acceptable and no additional toxicity study is required; (B) use of non-human primates that can react with human costimulatory molecules required instead of using mouse model and (C) use of CAR-T cells expressing mouse model of costimulatory molecules is required for toxicity study in an appropriate mouse model. Inputs from the regulators are presented to follow:

In the panel discussion, four topics were on the agenda: (i) overview of regulations in each country and region, (ii) how we can perform risk assessment associated with biological AMs, (iii) risk of approved commercial drug products in other counties and regions used as AMs and (iv) points to consider for specific AMs, particularly for bovine serum and trypsin. Although regulatory agencies from Korea and Malaysia could not take part in this panel discussion, the written responses on each topic were provided after the 5th APACRM. Their responses are also described side by side with the comments from the panelists.

Key differences and considerations among the regulatory agencies with respect to discussion outlined in WG2 session are summarized in supplementary Table 6.

Innovative gene therapy products of GMO generally have more complex regulatory framework in non-clinical and clinical development than in conventional drug development. It is necessary to carefully consider the risk of transmission of living modified organisms to others than patients, animals or the entire environment, and hence, the environmental impact assessment (EIA) will be required. Careful measures, such as safe handling of GMO and method of administration to patients and disposal, are adopted to prevent GMO from spreading outside.

WG3 of APACRM was established in 2022 to discuss if there is room to harmonize the regulation related to EIA in clinical and/or regulatory perspective(s). In the first approach, WG3 compared the regulation regarding the Cartagena or EIA in APAC region.

A survey with four questions were performed in advance for each Asian country/region: if they are member of Cartagena protocol (Question 1), if there are general guidelines on EIA including the one for gene therapy (Question 2), and if there are any regulations on the control of viral shedding (Question 3) and special regulatory reviews for gene therapy development (Question 4). The survey result enabled us to identify common points, similarities, and differences in the regulations/guidelines on EIA particularly regarding gene therapies in participating Asian countries/regions.

At the meeting, the survey results of Questions 1 to 4, as shown in Table 1, were discussed.

• Question 1: Out of the six Asian countries/regions, China, India, Japan and Korea are members of the Cartagena protocol. Japan and Korea incorporated the Cartagena protocol into domestic regulation.
• Question 2: Almost all the countries/regions have some sort of domestic regulation related to gene therapy development from non-clinical, clinical and biodiversity risk assessment to prevent exposure. When initiating gene therapy product development in the future, it is necessary to compare each item of the guidelines in each country/region in order to deepen the understanding of requirement of non-clinical data package, and preparation for clinical studies is required.
• Question 3: Some countries/regions referred to International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) considerations for virus shedding. The ICH guideline provides requirements in clinical trial protocol regarding the sampling time for measuring viral shedding and preclinical studies to assess the safety of viral vectors used for gene therapies.
• Question 4: Almost all countries/regions have special review systems focusing on the evaluation of GMO, which is different from scientific review in clinical trial application (Table 1). It is necessary to understand the regulatory review systems in each country/region for the simultaneous conduct of clinical trials. Further investigation into the details and comparison of the review systems among Asia countries/regions is needed.

In conclusion, the comments from the different regulators in APAC countries/regions indicate that it is difficult to harmonize the regulations and/or procedures regarding EIA review because of various country-specific reasons and situations. However, we can seek more concise and effective evaluation of EIA in the future by referring to the scheme/perspective in each country/region.