Epilepsy is treated with anti-seizure drugs with the aim to stop the seizures. Although many anti-seizure drugs are available, we are currently unable to predict which drug works best in a particular patient. Therefore, we must select the drugs based on trial and error which results in less than one in two patients becoming seizure free with the first drug. We aim to predict the anti-seizure drug with the highest likelihood of success in an individual patient (personalized medicine) by combining two strategies. The first strategy examines human brain cells which are generated from skin cells of adult donors. These cells are treated with four different anti-seizure drugs (carbamazepine, valproate, lacosamide, levetiracetam) and we will investigate how specific genes respond to exposure to these drugs. We will then look at available genetic data from 1382 patients with epilepsy and examine if changes in these genes correlate with a drug’s ability to stop seizures. The second approach consists of a “big data analysis” incorporating details on the patient’s epilepsy, genetic data, brain wave signals (electroencephalography or EEG) and brain imaging to predict treatment effect. Using these large sources of complementary data, we will develop tools to help doctors and patients predict the effect of the different anti-seizure drugs. Subsequently, we will test in an independent cohort of 100 patients if the tool correctly predicts the best anti-seizure drug.
