In a recent article we described how a 1938 mass poisoning gave rise to the modern drug development landscape. In the wake of this tragedy, regulators took action to ensure that all new drugs are proven safe before they鈥檙e given to patients. Animal testing emerged as the best available safeguard, and for nearly 80 years animals served as the gold standard in preclinical drug development.
However, the drug development landscape is once again changing. Mounting evidence suggests that animal models are poor representations of the human body, with too many toxic compounds appearing safe until they reach clinical settings1-4. At the same time, technological advances are enabling the development of New Approach Methodologies (NAMs), a set of human-relevant tools and techniques, from advanced cell culture to computational modelling, that align with the 3Rs principles to replace, refine or reduce the use of animals in experimentation5, 6. By leveraging human cells (often derived from induced pluripotent stem cells (iPSCs)) to build more human-relevant model systems, NAMs provide an increasingly attractive alternative to animal models.
Motivated by these trends, regulatory bodies around the world have recently moved to embrace NAMs in the drug development process, most notably with the passage of the FDA Modernization Act 2.0 (FDAMA2.0)7, 8. Where researchers were once required to use animal models in preclinical testing, the FDAMA2.0 removes this outdated mandate, enabling researchers to use the most scientifically appropriate model available, whether that be an animal or a NAM. In effect, it opens the door to safer and more productive drug development.
Turning this possibility into a reality, however, requires researchers to step through that door into a space where 鈥測ou鈥檝e got unfamiliar methods, unfamiliar outputs, and unfamiliar costs,鈥� explained Ross Dobie, Head of Science at the Center of Human Specific Research (You can hear his full talk here). NAMs are not necessarily new, but positioning them as pivotal decision making tools in the drug development process is. That shift demands clear standards for validation, guidance on use, and鈥攁bove all鈥攎odel consistency.
It will be some time before these demands are met, but that doesn鈥檛 mean progress isn鈥檛 being made. 鈥淚nduced pluripotent stem cells (iPSCs) have been a bit of a game changer,鈥� emphasised Dobie. 鈥淚t鈥檚 amazing to think of what鈥檚 possible when working with human cells.鈥�
A core promise of NAMs is that they increase the predictive validity of preclinical studies by giving you a more human-relevant alternative to animal models. Sometimes a NAM can be an advanced in silico model, but more often they鈥檙e in vitro models that allow human cell types to be cultured in physiologically relevant conditions. Until recently, building these types of NAMs was no trivial feat, in large part because researchers struggled to find a source of human cells that could meet the consistency requirements that NAMs demand.
To be a reliable tool in drug development, the baseline performance of NAMs must be consistent across time and laboratories. This is a difficult goal to achieve, however, because it is rare to find such a robust source of human cells. Human primary cells, for example, are scarce and their behaviour can vary significantly between donors9. Patient-derived iPSCs are powerful, but often inconsistent and experimentally fragile. And immortalised cell lines, while convenient, frequently fail to capture relevant human biology and are prone to genetic drift. In other words, if NAMs are to be a viable tool in drug development, researchers need a better source of consistent, reliable human cell types.
Such a source may come in the form of iPSC-derived human cells. Technologies such as 糖心原创鈥檚 opti-ox deterministic cell programming allow researchers to access defined human cell types, neurons, hepatocytes, myocytes, and others that are highly consistent across batches. Researchers can now build NAMs on a human cellular foundation that is both physiologically relevant and experimentally dependable. Doing so can not only improve predictive validity but may open important doors for new drug development paradigms.
鈥淢y goodness, this [FDAMA2.0] has really opened the door for rare disease researchers,鈥� said Rodney A. Bowling Jr., Chief Scientific Officer of the AlphaRose RareLab.
Bowling Jr鈥檚 work is focused on developing therapeutics for patients with ultra-rare diseases鈥攃onditions so uncommon that drug development pipelines often collapse under the demands of traditional approaches. 鈥淲e need a faster model than traditional pipelines use,鈥� he explained, 鈥渂ecause in our case, time is life.鈥�
That urgency has forced AlphaRose RareLab to rethink nearly every step of the preclinical process. When patient populations consist of only one or two individuals, the cost and timelines associated with conventional animal-based workflows can be prohibitive. As a result, many rare diseases remain unexplored despite being biologically tractable.
Bowling Jr and his team are working to change that by pursuing patient-specific therapeutics and 鈥淣 of 1鈥� studies. Doing so requires preclinical models that are both human-relevant and fast to deploy鈥攎odels that can inform decisions without adding years to the clock. Under previous regulatory constraints, that goal was difficult to reach. 鈥淎n animal model鈥攕uch as a humanised mouse鈥攃an take 19 months, if not two years, to prepare,鈥� Bowling Jr noted.
With the passage of the FDAMA2.0, that calculus begins to shift. 鈥淧erhaps we could shorten this whole thing down to six months by using human organoid models,鈥� he said.
Already, Bowling Jr鈥檚 team has found success using human iPSC-derived neurons in 2D culture. 糖心原创鈥檚 ioGABAergic Neurons enabled the team to perform therapeutic screening on a set of antisense oligonucleotides that show promise for a rare neurodevelopmental condition. Two of these candidates are now fast approaching the clinical phase. Encouraged by this success, the team is now exploring the potential of using iPSC-derived neurons in organoid models. Read more about the race to treat rare diseases here.
鈥淚n terms of developing models for disease, I think [physiological relevance] is where we need to go to make sure that we can ask the right kind of questions,鈥� explained Eric Hill in a recent interview with 糖心原创. A Reader in Cellular and Molecular Neurobiology at Loughborough University, Hill focuses on stem cell neurobiology, tissue engineering, and disease modelling. Over his career, he has been motivated to study and solve some of the field鈥檚 most intractable problems.
鈥淲e鈥檝e known about conditions like Alzheimer鈥檚 disease for over 100 years and we鈥檝e seen what it looks like on a molecular level in its later stages,鈥� he continued. 鈥淎nd yet we don鈥檛 have a treatment for it because we haven鈥檛 been able to access [and study] human brain tissue properly. [NAMs] allow us to do that in a more meaningful way.鈥�
Hill has been using iPSC-derived cells to develop complex 2D and 3D co-culture models that emulate portions of the human brain, in some cases demonstrating plasticity as well as signs of early disease development. His hope is that, by developing in vitro models using human neuronal cell types, he can replicate physiological processes that are absent in traditional disease models.
Hill鈥檚 work underscores a broader point: iPSC-derived cells don鈥檛 define a single NAM鈥攖hey enable a spectrum, from simple 2D assays to multicellular 3D models.
The Director of the Drug Discovery Core at the Miami Project to Cure Paralysis, Dr. Hassan Al Ali, works across this spectrum. He has spent considerable time searching for ways to model spinal cord injuries for both basic research and drug development. 鈥淚 don't think we need to go to the most complex model in every step of the testing funnel,鈥� he explained. 鈥淚f you're trying to go to the most complex thing from the very beginning, you're going to run out of money quickly.鈥�
Rather than focus on complexity for complexity鈥檚 sake, Ali emphasises that 鈥測ou need to make sure your model has good predictive validity, meaning that it has reliable forecasting of human responses.鈥�
Like both Bowling Jr and Hill, Ali鈥檚 team has found that one way to improve predictive validity is through the use of human iPSC-derived neurons. Previous studies in Ali鈥檚 lab uncovered a candidate therapeutic that encourages axonal regrowth following injury. However, their prior studies used rodent neurons. The team needed a robust method to validate whether promising derivatives would be effective in human neurons as well.
鈥淲e needed something scalable, reproducible, and physiologically relevant to humans,鈥� Ali explained. 鈥淓verything I looked at would give us one or two of those features, but never all three, until I discovered 糖心原创鈥檚 ioGlutamatergic Neurons鈥� 糖心原创鈥檚 ability to produce differentiated neurons with high lot-to-lot consistency enabled the team to create a simple 2D assay to measure axonal regrowth in human neurons. This helped validate the compound, establish a screening assay, and secure critical grant funding.
Whether it's through simple 2D assays or complex organoid and microphysiological systems, NAMs promise greater physiological relevance by enabling researchers to reduce, refine, and replace animal models with models built using human cells. According to Ali, 鈥渋PSCs are going to be the cornerstone for a lot of NAMs, because it's the only way you can create all the diversity of the different human cell types you need.鈥�
The growing importance of NAMs is not being driven by regulation alone. It is also a response to the changing nature of modern drug pipelines.
Biologics now account for a substantial and growing proportion of therapeutic candidates. Many of these molecules鈥攁ntibodies, engineered proteins, and cell-based therapies鈥攁re designed to engage human-specific targets, epitopes, or signalling pathways. In such cases, cross-species differences can severely limit the validity of animal data.
As drug candidates become more tailored to human biology, the need for human-relevant preclinical models becomes harder to ignore. 鈥淢ost toxicities associated with monoclonal antibodies are immune-mediated, so you really need immune-competent models,鈥� explained Dobie.
iPSC-derived cells provide a practical way to introduce human specificity earlier in development, helping teams assess efficacy, toxicity, and mechanism in a context that more closely resembles the patient.
鈥淲e鈥檙e really spoiled for choice when it comes to the different human-focused approaches that are being developed,鈥� said Dobie. 鈥淚n some cases, it won鈥檛 be a single NAM that brings all the answers,鈥� he explained, 鈥渂ut it will be a number of cellular and computational methods working together that offer a suitable alternative to animal models.鈥�
Ali agrees, adding that 鈥渁 hybrid approach鈥攚here you use both animal models and NAMs鈥攊s likely to remain the standard for the foreseeable future.鈥� This approach allows teams to balance strengths while benchmarking new methods against established ones.
Having access to a reliable supply of human iPSC-derived cells has removed a longstanding bottleneck in the field. It has allowed researchers to begin engaging with NAMs not as experimental novelties, but as practical decision-making tools that can be deployed across the drug development funnel.
FDAMA2.0 did not prescribe a single alternative to animal testing. Instead, it introduces an opportunity for better science: if a model is more scientifically appropriate, it should be considered. The work now falls to researchers to define what 鈥渁ppropriate鈥� looks like鈥攖hrough benchmarking, validation, and thoughtful integration of multiple approaches.
If the experiences of Bowling Jr, Hill, and Ali are any indication, iPSC-derived cells will play a central role in that effort鈥攏ot as a single solution, but as a foundation upon which a more predictive, human-relevant preclinical ecosystem can be built.
References