©2021 by The International Society for Stem Cell Research. All rights reserved.

No part of this document may be produced in any form without written permission of The International Society for Stem Cell Research.

3.3.2 Safety Studies

Human cells should be produced under the conditions discussed in Section 3.2, Cell Processing and Manufacture. Depending on the laws and regulations of the specific region, biodistribution and toxicity studies should be performed using good laboratory practice (GLPs). It is recommended that these studies be performed by a third party, such as a Contract Research Organization (CRO).

Cell Characterization
Recommendation Cells to be employed in clinical trials must first be rigorously characterized to assess potential toxicities through studies in vitro and, where possible, for the clinical condition and tissue physiology to be examined in animal models.

Outside of the hematopoietic, stratified epithelia, and various stromal cell systems there is little clinical experience with the toxicities associated with infusion or transplantation of stem cells or their derivatives. In addition to known and anticipated risks (for example, acute infusional toxicity, immune reactions, and tumor development), stem cell-based interventions present risks that will only be discovered with experience. As non-human animal models may not replicate the full range of human toxicities associated with stem cell-based interventions, particular care must be applied in preclinical analysis. This section defines toxicities that are likely to be unique to stem cells or their progeny.

Tumorigenicity Studies
Recommendation Risks for tumorigenicity must be rigorously assessed for any stem cell-based product, especially if cells are extensively manipulated in culture, genetically modified, or when derived from a pluripotent source.

Assessing tumorigenicity is a critical part of determining the safety profile of stem cell products. These studies can be challenging, as they usually require assessment of the human cell product in xenogeneic models. Further, these studies usually include long-term time points that can be several months to years. Therefore, immunocompromised animals, usually rodents, are often the animal model of choice. 

All stem cell derived products should be tested for tumorigenicity, see Recommendation  Long-term animal studies are necessary to demonstrate that the persistence of any remaining undifferentiated cells in the final product do not result in tumors. 

It is understood that assessing tumorigenicity in animal models is complicated by implantation technique, composition of the test article (percentage of undifferentiated cells versus the percentage of cell product), and various other parameters. Because of this complexity, tumorigenicity studies may benefit from additional in vitro studies. These would include examining rates of proliferation, observing if faster dividing subclones tend to take over the cultures, and looking for expression of oncogenes or loss of tumor suppressor gene activity. However, while these tests may add to in vivo studies, they cannot substitute for them.

Positive tumor-generating controls and negative controls assessing background tumorigenesis should be run in parallel in these studies to validate results. Specifically, this informs whether the site of implantation and other delivery parameters are permissive to tumor formation, allowing interpretation of a negative result. In these studies, it is important to deliver the cell product to the intended clinical site, if feasible. Further, assessment of the clinical dose is also important. In cases where the human dose includes very large quantities of cells, this can be quite challenging, and it is critical to work with regulators to ensure that proposed study designs are appropriate. For example, in cases where it is not feasible to deliver a human-sized dose into an immunocompromised animal model, the risk from residual undifferentiated cells in the product may be assessed by spiking the largest feasible animal dose of the therapeutic product with the highest number of undifferentiated cells that might be present in the human-sized dose (based on the sensitivity of the assay used for measuring their presence in the clinical dose).

The plan for assessing risks of tumorigenicity should be reviewed and approved by regulators before initiation of definitive preclinical studies and clinical trials. For additional guidance on specific techniques that may be of utility for genome edited interventions, see Appendix 5. 

Biodistribution Studies
Recommendation For all stem cell-based products, whether injected locally or systemically, researchers should perform detailed and sensitive biodistribution studies of cells.

Because of the potential for cells to persist or expand in the body, investigators must seek to understand the nature and extent by which cells distribute throughout the body, lodge in tissues, expand and differentiate. Careful studies of biodistribution, assisted by ever more sensitive techniques for imaging and monitoring of homing, retention and subsequent migration of transplanted cell populations is imperative for interpreting both efficacy and adverse events. These studies should whenever feasible, include delivery of the cell product using the intended clinical route and site of delivery.

Additional histological analyses or banking of organs for such analysis at late time points is recommended. Depending on the laws and regulations within specific jurisdictions, biodistribution and toxicity studies may need to be performed in a good laboratory practice (GLP)-certified animal facility.

Distinct routes of cell administration, local or systemic, homologous or non-homologous/ectopic, can lead to different adverse events. For example, local transplantation into organs like the heart or the brain may lead to life-threatening adverse events related to the transplantation itself or to the damage that transplanted cells may cause to vital structures. Especially in cases where cell preparations are infused at anatomic sites distinct from the tissue of origin (for example, for non-homologous use), care must be exercised in assessing the possibility of local, anatomically specific and systemic toxicities.

Ancillary Therapeutic Components
Recommendation Before launching high-risk trials or studies with many components, researchers should establish the safety and optimality of other intervention components, like devices or co-interventions such as surgeries.

Cell-based interventions may involve other materials besides cells, such as biomaterials, engineered scaffolds, and devices. There may also be co-interventions like surgery, tissue procurement procedures, and immunosuppression. Additional components added to the cellular product, or delivery device can interact with the stem cell product and each other. In these cases, safety and efficacy studies should include the assessment of the final combination product. Many subjects in cell-based intervention studies may be receiving immunosuppressants or drugs for managing their disease. These can also interact with the implanted cell product. Safety and efficacy studies should include assessment of possible interactions between the cell product and these types of medications, in vitro or in vivo. 

Long-term Safety Studies
Recommendation Researchers should adopt practices to address long-term risks in preclinical studies.

Given the likelihood for long-term persistence of cells and the irreversibility of some cell-based interventions, testing of the long-term effect of cell transplants in animal models is encouraged. 

Application of genetic alteration and genome editing technologies to stem cell products 
Recommendation Researchers should comprehensively investigate the type, extent and genomic distribution of introduced genetic alterations as well as their potential adverse effects on the genome and the biological properties of the treated cells at short and long-term time points.

Genetic alteration and genome editing technologies can be coupled to stem cell therapies or applied directly in vivo to resident tissue cells for a variety of therapeutic purposes. 

Gene replacement approaches have made substantial progress and advances into clinical testing, either performed ex vivo on hematopoietic stem cells, lymphocytes or epidermal stem cells or in vivo targeting the liver, retina or CNS, with a growing number of therapies approved for market access. Targeted genome editing strategies are still in the early stage of clinical development, although there is constant progress and early clinical testing is showing safety and some efficacy, at least for ex vivo based strategies.

Considering the likelihood for long-term persistence, expansion and broad clonogenic output of many stem cell types and the irreversibility of any genetic alteration introduced by integrating gene transfer or genome editing, the type, extent and genomic distribution of the introduced genetic alteration, including on- and off-target events, should be comprehensively investigated as well as their potential adverse effects on the genome and the biological properties of the treated cells both in the short- and long-term. This is particularly important following genome editing when this has involved double strand breaks in DNA; the analysis of manipulated cells should include an assessment of incorrect on- and off-target events, and whether these pose any risk. Whenever possible and scientifically appropriate, such testing should include cell transplants in suitable xenogeneic hosts for long-term observation. 

Potential of Stem Cells for Toxicology
Recommendation Researchers, sponsors, and regulators should take advantage of the potential for using stem cell-based systems to enhance the predictive value of preclinical toxicology studies.

Stem cell science offers the prospect of testing toxicology in cell-based systems or artificial organs that more faithfully mimic human physiology than animal models. Such approaches, though unlikely to ever completely substitute for in vivo testing in animals, hold substantial promise for reducing burdens imposed on animals in safety testing and improving the predictive value of preclinical safety studies.