For H.C. Watkins, the chief chemist and pharmacist at the S.E. Massengill company, the necessity of flavour was paramount. The bitter flavours of diethylene glycol and sulfanilamide had to be masked for children to accept the company’s new medicine. It took work, but with a dash of both caramel and raspberry, the flavour profile came together. On September 4th, 1937, the tasty “elixir sulfanilamide” brew was shipped to physicians across the United States, bringing with it the hope of curing sore throats, colds, and all manner of infections1.
By October 15th of that same year, the sweet concoction had killed nearly 100 people.
The Elixir Sulfanilamide disaster was as terrible as it was avoidable. Diethylene glycol’s lethal effect on the kidneys had already been exposed by other laboratories. Had Watkins and the company’s founder Samuel E. Massengill cared to test their elixir in animal models, its poisonous effects would have been clear. They made efforts to hone its flavour and fragrance, but felt no responsibility to check if the syrup was safe. “We…not once could have foreseen the unlooked-for results,” Massengill wrote in his defence2.
Amidst public outrage, the United States Congress set out to ensure that every future drug developer would be required to “look for” evidence of their drug’s safety. Through the Federal Food, Drugs, and Cosmetics Act (FFDCA), they mandated that all new drugs be proven safe before they’re given to patients1, 3.
But, how exactly should drug developers look for evidence of a therapeutic’s safety? At the time, the best answer they had was to use animal models, not because they’re uniquely good surrogates for the human body, but because there simply weren’t many other options. Fortunately, over the last 80 years, science has advanced. Today’s drug developers are empowered by a growing arsenal of model systems, many built using human stem cell technology4.
While nascent, the advanced capabilities of these new models have inspired a recent update to the FFDCA, opening the door for drug developers to forgo animal models when a more scientifically valid model is available5, 6. It’s a small update, but one with transformative potential.
With the rising interest in highly specialised therapeutics, such as CAR-T Cell therapies, bispecific antibodies, and other biologics, the need for human-like models is greater than ever. Induced pluripotent stem cells (iPSCs) present a unique opportunity for researchers to build in vitro models using easily accessible and scalable human cells, effectively replicating human tissues in a laboratory setting. To appreciate the potential impact of these models, however, it’s necessary to first understand the nature of modern drug development, the concept of predictive validity, and the long-standing need for better building blocks.
Modern drug development is an exercise in serial decision-making. Researchers must decide which targets to pursue, which compounds to screen, and, ultimately, which candidates are likely to succeed in the clinic. To make these decisions, they rely on evidence gathered using experimental models. The quality of their decision-making is directly influenced by the quality (or predictive validity) of the models they use7.
Predictive validity describes how accurately your model emulates real-world outcomes8. Consider a model system in which tumour cells are cultured as spheroids, the growth of which is expected to simulate tumour growth in patients. If changes in the spheroids—such as increased cell death or slowed growth rate- closely mirror those observed in tumours under the same conditions, the model would be described as having high predictive validity.
Ideally, all models would be highly predictive. In reality, researchers have to find a compromise between the cost of building the model, the complexity of using it, and the validity of its results. This is why early drug development efforts are often dominated by simple model systems (such as 2D cell growth assays), which may have moderate predictive validity but are far more efficient and scalable—both necessary attributes for large-scale drug screening.
However, the long-term cost of reduced predictive validity can be high. To see this, one need only look at the current success rate of therapeutic development: Despite considerable preclinical testing and promising results, roughly 90% of therapeutic candidates fail when they reach clinical trials9. Analysts have shown that seemingly minor changes in a model’s accuracy can have an exponential impact on drug development productivity7, 8.
While this also means that small improvements to model accuracy can have an outsized impact, improving model validity is far from trivial, in part, because very few models can recreate the genetic and physiological context of human disease. In vitro models have been limited due to accessibility; they either lack human relevance or are only relevant to a narrow subclass of malignant diseases8. Primary human cells are the exception, providing researchers with differentiated human cell types that potentially carry valuable disease-specific mutations. However, these are also difficult to obtain, may represent late-stage disease, and vary considerably between donors10.
Collectively, this has left the research community in need of something better, a source of human cells that are consistent, scalable, and reflect human biology in vitro in such a way that it can be used to build more predictive model types. Against this backdrop, it’s easier to see why advances in iPSC technology can be so impactful.
Figure 1. A schematic depicting a range of in vitro models used during drug development and their physiological relevance from primary animal cells and immortalised cell lines through to microphysiological systems.
Improving decision-making in drug discovery comes down to using in vitro models that better predict outcomes in humans. Central to this evolution is the adoption of human iPSC-derived cells, the value of which lies in their potential to provide researchers with an endless supply of differentiated human cell types. A large body of evidence suggests that iPSC-derived cells are more likely to resemble their in vivo counterparts in both gene expression and behaviour (relative to animal cells and immortalised cells)11. Leveraging human iPSC-derived cells in place of immortalised cells is a quick way to improve the relevance of many 2D screening models. When combined with organoid and microphysiological systems technologies, these cells can also be a powerful building block for 3D models of human tissues, enabling more predictive preclinical studies and better overall decision-making4, 10, 11.
It’s important to note that these advances have not occurred in a vacuum. Over the past several decades, the scientific community has become more aware of how inaccurate animal models can be12, 13. Though they have played an important role in the drug development process, a significant number of toxic compounds pass through animal studies without raising red flags14, 15. One study examining the effects of more than 2000 drugs in both preclinical and clinical settings found that “tests on animals (specifically rat, mouse and rabbit models) are highly inconsistent predictors of toxic responses in humans and are little better than what would result merely by chance.”15
It has become increasingly clear that, to protect patients, drug developers need better, more human-relevant models that can be applied across the development pipeline—both in conjunction with animal models or, in some cases, instead of animal models.
It’s in this context, one of growing awareness about the limits of animal models and the rise of advanced model systems, that the US Congress is changing the drug development landscape.
In December of 2021, more than 80 years after the Elixir Sulfanilamide disaster, the US Congress passed a bill known as the FDA Modernization Act 2.05, 6. Among its many effects, the bill removes the 1938 requirement that all new drugs be proven safe in animal models prior to reaching the market. Instead, the amendment allows drug developers to prove that a drug is safe using the most scientifically valid models available to them.
At present, the most validated and available models will often be the well-described animal models that we know. But, in allowing FDA regulators to consider evidence from alternative models, the bill changes the drug development landscape by enabling researchers to veer away from the normal beaten path in search of a better way. And, it’s not just the US: similar changes have been adopted elsewhere, including the European Commission’s efforts to develop a “Roadmap Towards Phasing Out Animal Testing for Chemical Safety Assessments”16. The UK Government has recently laid out a similar roadmap, one that aims to have phased out animal testing when assessing the dermatological effects of therapeutics by the end of this year17-19.
To forge a new path, however, researchers will need to be equipped with good and validated decision-making tools. Already, advanced microphysiological systems have been introduced to support drug development decision-making alongside animal models20. But for tools like these to be accepted, they will need to be validated as both human-relevant and consistent in their performance. Therefore, researchers will need access to a consistent source of human cells that can be used in 2D and 3D formats.
For that, they’ll likely turn to iPSC-derived cells.
Unlike animal models or immortalised cell lines, human iPSC-derived cells are key to studying disease in the right cellular context. Whether modelling neurodegeneration, cardiac dysfunction, or metabolic disorders, these cells provide a level of physiological relevance that traditional models often lack. Though iPSCs can be a good source for generating human cells, making them consistently at scale has proven difficult.
Both researchers and cell vendors have long relied on directed differentiation methods to coax iPSCs towards a specific cell identity. These methods are technically complex, time-consuming, and prone to error. Consequently, directed differentiation methods often produce a mixture of expected and unexpected cell types, with significant batch-to-batch variability21. In addition to the lack of production efficiency and cellular consistency, the cost of producing sufficient human iPSC-derived cells to support the billions of cells required for high-throughput screening has limited the widespread adoption of iPSC-derived cells in drug development.
Fortunately, bit.bio has developed a solution using its opti-ox deterministic cell programming technology, enabling iPSCs to be rapidly and consistently converted into differentiated cell types with high precision. Each vial of ioCells contains a defined and highly characterised population of human iPSC-derived cells that are ready for quality data production in a matter of days. Their high lot-to-lot consistency provides reliable cells that enable scientists to conduct repeatable and scalable experiments in drug discovery and disease research.
Reproducible | <2% gene expression variability across lots, giving researchers confidence that experiments will generate consistent, repeatable data.
Ready-to-use | Cryopreserved and assay-ready within days of thawing, enabling faster project timelines and reducing setup time and variability.
Human-relevant | Manufactured from human iPSCs to reflect native cell biology, with phenotypes that closely resemble their in vivo counterparts—designed to better support studies requiring human-specific insights.
Functionally validated | All ioCells are characterised for key cell–type–specific functions, such as marker expression, morphology, and relevant functional assays, ensuring each cell type behaves as expected in vitro.
Consistent at scale | Generated using deterministic cell programming, enabling the consistent production of large batches without compromising identity - ideal for high-content screening or multi-site studies.
Discover ioCells today
ioCells are already supporting researchers in the drug development space, from efforts to treat rare diseases to the study of spinal cord injury and regrowth. By addressing the gaps left by both traditional models and standard iPSC workflows, ioCells are a valuable building block in the future of drug development.
Explore the applications of ioCells in drug development by reading our latest case studies 'Surviving the spinal cord injury paradox' and 'Advancing rare disease drug discovery using consistent, defined human cells'.
References