In Vitro Toxicology Testing: The Barriers to Progression and Adoption

Toxicity tests are important to ensure that newly developed substances, such as therapeutic drugs, agricultural chemicals, and food additives, are safe for us to use. The vast majority of information about the potential toxicity of these substances comes from using animal models, but these tend to have significant limitations when translating them to human risk assessment. Although some basic toxicity testing using human cells in vitro is already a conventional method, many toxicologists envision that more sophisticated in vitro analyses using human-based cell models can be both more predictive of human health outcomes, as well as more time and cost-effective.

Although these cell model systems may be able to perform better than current animal models and improve our ability to predict toxicity in humans, there are still many challenges involved in developing them. In this two-part blog series, we explore what these challenges are, and how we can overcome them to encourage further progression and adoption of in vitro toxicity testing.

In vitro assays for toxicity testing

In vitro toxicity analysis involves applying new substances to mammalian cells that have been cultured in either monolayer 2D or biomimetic 3D structures, and monitoring their subsequent biochemical and phenotypic responses to identify one that represents toxic outcomes.  

In vitro assays bring a number of technical advantages to the conventional method of testing substances on animal models, as outlined, for example, in the 2007 US National Research Council (NRC) report. These advantages include: the ability to elucidate cellular-response networks and toxicity pathways; the use of concentrations relative to human exposure; and enabling high-throughput studies. In vitro 3D cultures that produce biomimetic models have been widely hailed as an advanced screening method by showing how substances influence in vivo cellular interactions.

However, the technical challenges involved in creating a truly biomimetic cell culture system are still confounded by the simple question: what exactly constitutes a biomimetic model?

The barriers to adoption: 3D liver models as an example

The liver plays a central role in detoxification and, at times, metabolic activation of small molecule substances. As such, developing a biomimetic in vitro model of liver function is a major goal on the path to developing more physiologically relevant toxicity models. However, most of these new advanced in vitro liver models are widely discounted as being far from biomimetic. This is because they often lack the liver’s heterogeneous myriad of hepatocytes, non-parenchymal cells, immune cells and vasculature, as well as the in vivo endocrine communications between the liver and other organs.

Therefore, fully recapitulating all liver functions in vitro is one of the major barriers to widespread adoption of in vitro toxicology testing, and has been widely seen as an unreachable goal due to the liver’s biological complexities. Rather than attempting to develop a fully functional biomimetic in vitro liver model, many have now refocused their efforts on developing multiple models to fit particular purposes. This new approach aims to use many different liver models that each address a particular facet in the whole range of toxicities that manifest in the liver.

A second major barrier to adoption of in vitro toxicity testing is being able to convey a general understanding of how a particular cellular event measured in vitro is then translated to a clinical effect that adversely impacts human health. A new paradigm that aims to resolve this, called the Adverse Outcome Pathway, is a conceptual framework that describes the key events leading from cellular ‘molecular initiating’ events to clinical adverse outcomes. For any particular toxicity outcome in the clinic, there may be many different cellular events at the initiation. Once these pathways are developed and validated, it is likely to be easier to identify which models need to be developed to obtain the appropriate measurements for toxicity predictions.

The field of toxicology as a whole is at a crossroads where current and future technologies are enabling better predictions of toxicity before a substance is released for human consumption or exposure. While a few barriers remain, we are making progress towards overcoming them, from both technical and policy perspectives.