3.3.1 General Considerations
Recommendation 18.104.22.168: Preclinical research into stem cell-based interventions involving animals should adhere to the principles of the three Rs: reduce numbers, refine protocols, and replace animals with in vitro or non-animal experimental platforms whenever possible.
This recommendation is not incompatible with performing replication experiments or achieving adequate statistical power (see: www.nc3rs.org.uk). Indeed, these are key steps for ensuring that animal experiments support robust conclusions. This recommendation should also not be interpreted as suggesting that in vitro or non-animal platforms are sufficient for supporting clinical investigations. For most stem cell therapies safety testing should be performed in vivo. However, recent developments in organoid systems suggest that efficacy testing in vitro may by suitable in some circumstances.
Preclinical Study Objectives
Recommendation 22.214.171.124: Early phase human studies should be preceded by a rigorous demonstration of safety and efficacy in preclinical studies. These preclinical studies can include in vitro and in vivo modeling.
Preclinical efficacy studies help provide the scientific rationale for proceeding into human trials. Stringent design and reporting standards should be demanded where planned trials involve invasive delivery approaches or where the cell product presents greater risk and uncertainty. However, prudent use of scientific resources means that even when risk is modest, studies should rest on sound scientific evidence of expected efficacy.
The development of stem cell-based products generally includes a period of process development in which the manufacturing process is optimized. This may include exchanging research grade reagents with reagents manufactured under GMP and removing xenogeneic reagents from the process. Because these changes can affect the cellular composition and bioactivity of the final product, it is important that preclinical efficacy studies use stem cell product manufactured using processes intended for clinical applications, whenever feasible.
Recommendation 126.96.36.199: All preclinical studies testing safety and efficacy should be designed in ways that support precise, accurate, and unbiased measures of potential clinical utility. In particular, studies designed to inform trial initiation should have high internal validity; they should be as representative as possible of clinical scenarios they are intended to model, and they should be replicated.
Preclinical experiments confront many sources of bias and confounding factors, including selection bias. For decades, clinical researchers have sought to minimize the effects of bias and confounding by using techniques like randomized allocation, blinded outcome assessment, or power calculations. Such rigor should also apply in preclinical studies intended to support trials. Numerous groups have articulated standards for designing preclinical studies aimed at supporting trials (Fisher et al., 2009; Henderson et al., 2013; Landis et al., 2012; Kimmelman et al., 2014). Key design principles include:
- Researchers should reduce bias and random variation by ensuring their studies have adequate statistical power, use appropriate controls, randomization, and blinding, and, where appropriate, establish a dose-response relationship.
- Critical or definitive safety and efficacy studies should be performed with prospective protocols and should have independent quality oversight.
- Researchers and sponsors should ensure preclinical studies model clinical trial
settings. Researchers should characterize disease phenotype at baseline, select animal models that best match the human disease, use outcome measures that best match clinical outcomes, and provide evidence supporting a mechanism for treatment effect.
- Researchers and sponsors should ensure effects in animals are robust by replicating findings, ideally in an independent laboratory setting.
- Researchers and sponsors should pre-specify and report whether a study is exploratory (i.e., hypothesis generating or aimed at substantiating basic science claims) or confirmatory (i.e., using pre-specified hypotheses and protocols and powered to support robust claims). Preclinical researchers should only venture claims of potential clinical utility after confirmatory studies.
Sex as a Biological Variable
Recommendation 188.8.131.52: Preclinical studies should assess both male and female animals in safety and efficacy testing unless there is a scientifically valid reason not to do so.
Males and females can respond differently to medical treatments as well as to the incidence of diseases, which can reflect distinct underlying pathways and mechanisms, due to chromosomal makeup and effects of gonadal hormones. Therefore, it is important to include both male and female animals in preclinical efficacy and safety studies. Of particular importance is the inclusion of both sexes in long-term safety studies, which is typically a mandatory requirement from many funding and regulatory agencies. In vitro model systems should, whenever possible, also be derived from male and female cells.