Abstracts

Rationale: More than 30% of epilepsy patients are drug resistant, leading to uncontrolled seizures. Current clinical practice relies on a trial-and-error approach, whereby the patient is sequentially trialled on different anti-seizure medications in the hope of finding an effective regime. This paradigm has not changed in the last 20 years despite more than 25 anti-seizure medications on the market.

For patients, this protracted (often years-long) journey results in substantive co-morbidity, loss of productivity, and greater risk of sudden unexplained death with epilepsy. This incurs significant cost for the health system, hence, there is an urgent need to improve patient care and personalise anti-seizure medication choice.

The aim of this study is to develop a patient-specific in vitro drug screening platform for epilepsy using an induced pluripotent stem cell-derived brain organoid model.

Methods: Twelve epilepsy induced pluripotent stem cell lines have been generated from blood samples taken from drug responsive and drug resistant epilepsy patients, and subsequently utilised to generate brain organoids. Immunochemistry and real-time polymerase chain reaction (qPCR) confirmed mature neuronal content and the presence of both GABAergic and glutaminergic neuronal populations (Figure 1).

Calcium imaging and multi-electrode arrays (MEAs) were utilised to assess baseline neural activity; responses of neural activity to the proconvulsant agent 4-aminopyridine (4-AP), to 5 anti-seizure medications, and a combination of 4-AP and anti-seizure medications. Four parameters were quantified with the MEA: the mean firing rate (MFR), the mean burst rate (MBR), the mean burst duration (MBD) and the percentage of spikes in bursts (PSIB).

Results: Representative responses of patient-derived organoids to carbamazepine (CBZ) are shown in Figure 2. MFR significance was observed in baseline versus CBZ (p=0.0382) and 4-AP versus 4-AP plus CBZ (p=0.0185). MBR significance was observed in baseline versus CBZ (p=0.0053) and 4-AP plus CBZ (p=0041). A downward trend was also observed with CBZ, both with and without 4-AP in the MBD and PSIB analysis (Figure 2).

Baseline results were varied between patients, indicating that these models were able to demonstrate differences between patient lines. Variability in organoid responses to 4-AP and anti-seizure medications has enabled greater insights into how to further refine this model.

Conclusions: This data has demonstrated the enormous potential of a patient-specific brain organoid model as a drug screening platform. The ability to create an in vitro model that can mirror clinical responses, would be transformative in terms of substantive health benefits for patients; provide neurologists with an evidence-based medicine approach to treatment; reduce health care costs; and has the potential for new insights into the causes of epilepsy and drug-resistance.

Funding: Please list any funding that was received in support of this abstract.: Royal Brisbane and Women's Hospital Foundation Project Grant; Metro North Clinician Research Fellowship.