A 2025 study bred rats over multiple generations for seizure susceptibility—but failed to create a working epilepsy model, highlighting the limitations of animal experiments and the need for human-relevant alternatives.
Key definitions
- Amygdala: A small, almond-shaped brain region involved in emotion processing and memory. In this study, electrical stimulation of the amygdala was used to trigger seizures in rats.
- Placebo surgery: A sham procedure where animals undergo the same surgical process without receiving the actual experimental intervention, used as a control condition.
The failed experiment
Despite breeding rats over multiple generations to be seizure-prone, researchers at Monash University, Alfred Health and the University of Melbourne failed to produce animals that actually developed epilepsy—the very condition their model was designed to replicate (1).
In their May 2025 study, researchers selectively bred rats based on how quickly they developed seizures following repeated electrical stimulation of the amygdala. Over successive generations, this produced two genetically distinct lines: FAST rats (highly seizure-prone) and SLOW rats (seizure-resistant). The study’s goal was to test whether these genetic differences in seizure susceptibility would influence the risk of post-traumatic epilepsy (PTE) after experimental traumatic brain injury (TBI).
Notably, this was a follow-up to a 2023 study by the same research team (2). In that earlier work, moderate-to-severe TBI resulted in 100% seizure rate and mortality in FAST rats within 24 hours—forcing researchers to use milder injury in this 2025 attempt. Even with this adjustment, the model still failed to produce epilepsy.
Why this matters: PTE is a serious condition where people develop recurring seizures after brain injuries like concussions or combat injuries. Finding better treatments requires understanding the underlying mechanisms—but only if animal models can actually replicate the human condition.
What they did
The researchers performed TBI or placebo surgery on both groups of rats, and conducted extensive behavioural tests over six months. FAST rats showed more acute seizure-like behaviours and higher mortality after TBI, while SLOW rats had more long-term motor and weight problems.
The critical failure
Crucially, no animals in the study developed epilepsy—the key condition the model was intended to replicate. While the authors reported some abnormal brain activity, these events did not meet the threshold the study used to diagnose epilepsy. In other words, the selective-breeding approach produced lines with different acute responses but failed to generate a chronic epilepsy model that reliably replicates the human condition.
This is not an isolated problem. Although a wide range of experimental epilepsy models have been developed over decades, none fully replicate the complexity of human epilepsy (3). This reflects fundamental species differences in neural architecture, injury response and genetic complexity that limit how well findings in rodents translate to humans.
Weighing the costs
This study raises important questions about whether the suffering of hundreds of animals and the investment of research resources were justified by human-relevant gains in knowledge. It is important to acknowledge why researchers pursue such models—model development is exploratory and can sometimes reveal mechanistic insights—but where an approach repeatedly fails to produce clinically translatable outcomes, a reassessment is warranted.
A positive note: The authors deserve credit for publishing a negative result. Transparent reporting of what doesn’t work helps prevent unnecessary repetition of ineffective approaches and is good scientific practice that strengthens the field.
A more promising path forward
Human-relevant approaches offer both greater scientific promise and ethical advantages. Advances in human stem-cell-derived cultures, engineered 3D brain tissue models and brain organoids can now replicate many aspects of TBI and epilepsy without using animals (4,5). Critically, these models are increasingly able to probe human-specific mechanisms of injury and recovery that animal models cannot capture.
These modern methods provide a more promising foundation for understanding and treating PTE than continuing to develop new animal injury models that fail to replicate the human condition.
Conclusion
This study underscores the urgent need to transition away from harmful and unreliable animal experiments toward modern, human-relevant research methods that offer superior scientific validity alongside economic and ethical advantages.
Take action
👉 Learn more about Australia’s TBI research on animals
📨 Use our form express your concerns to the NHMRC about their funding of this type of research
- Leung WL, Shad A, Perucca P, O’Brien TJ, Semple BD, Caillas-Espinosa PM. Chronic outcomes after mild-moderate traumatic brain injury in adult seizure-prone (FAST) and seizure-resistant (SLOW) rats: a model for understanding genetic contributions to acquired epileptogenesis? Epilepsy Behav. 2025 May;166:110347. doi: 10.1016/j.yebeh.2025.110347.
- Leung WL, Dill LK, Perucca P, O’Brien TJ, Casillas-Espinosa PM, Semple BD. Inherent susceptibility to acquired epilepsy in selectively bred rats influences the acute response to traumatic brain injury. J Neurotrauma. 2023 Oct;40(19-20):2174-2192. doi: 10.1089/neu.2022.0463.
- Jiji PJ, Rai R, Kumar NA, Blossom V, Pai MM, Rai AR, et al. Experimental models of epilepsy: a comprehensive review of mechanisms, translational relevance, and future directions. Vet World. 2025 Oct 14;18(10):3041-3050. doi: 10.14202/vetworld.2025.3041-3050.
- Hanna ME, Pfister BJ. Advancements in in vitro models of traumatic brain injury. Curr Opin Biomed Eng. 2023 Mar;25:100430. doi: 10.1016/j.cobme.2022.100430.
- Bellotti C, Samudyata S, Thams S, Sellgren CM, Rostami E. Organoids and chimeras: the hopeful fusion transforming traumatic brain injury research. Acta Neuropathol Commun. 2024 Aug 30;12(1):141. doi: 10.1186/s40478-024-01845.