Uniting doctors and scientists







At the heart of the Epinext federation is the translation of results from basic research and computational neuroscience into clinical practice, i.e. developing new therapeutic strategies for epilepsy.

The key challenge is to accomodate for individual disease patterns in the patient‘s brain: Epileptic seizures are so unique in each case that only personalized care guarantees the optimal result, especially for neurosurgery.

This personalization of diagnosis and predictive therapy modeling is a key skill of the Epinext group: For years, Epinext researchers and clinicians have been engaged in multi-scale approaches involving human patients, animal models and cellular preparations. The Aix-Marseille University demonstrates strong leadership in computational neuroscience by housing several national, European and international brain modeling projects including the EU programs BrainScaleS, the Human Brain Project (HBP) and The Virtual Brain (TVB).

Restoring/Fundamental studies

Malformations of cortical developments

Most children suffering from drug-resistant epilepsy show signs of malformation in cortical development (MCD), thus having a bleak perspective of lifelong disability. Even worse, over 60% of MCD patients aren‘t responding well to surgical treatment either.

Epinext will try to translate findings from animal studies to humans, namely that tissue surrounding the epileptogenic zone (EZ) in the patient‘s brain is still contributing to the disease, which would explain the failure of surgical resection of the EZ itself.

Even more ambitious, Epinext will provide a proof of concept for a new therapy for MCD patients: Transducing innovative molecular tools could reduce the activity of epileptogenic neurons and networks, applicable to different forms of MCD-caused epilepsy and even other forms.

Epileptic encephalopathies of the epilepsy-aphasia spectrum

The spectrum of epilepsy-aphasia diseases (EAS) is at the crossroads between epilepsy, autism and speech/language disorders, from relatively mild Rolandic epilepsies to the most severe Landau-Kleffler syndrome (LKS) and the continuous spike-and-wave during slow wave sleep syndrome (CSWSS).

Unfortunately, this is still largely unknown territory. The development of these diseases and the response to various therapy approaches has remained unpredictable.

But because Epinext and other teams have identified one major genetic defect as the reason for EAS (GRIN2A, a glutamate NMDA receptor subunit), a new study will conducted observing and analyzing mice with a disabled GRIN2A gene. The goal is to see how the disease develops, how it affects behavior and brain tissue, ultimately evaluating which NMDARs-targeted drugs (N-methyl-D-aspartate receptor) could possibly be used to ameliorate it.

Kainate receptors as a therapeutic target for temporal lobe epilepsy

Epinext is following a very new approach for developing anti-epileptic drugs, targeting abnormal KARs (kainic acid receptors) which are contributing to the emergence of epilepsy. By analyzing postsurgical tissue on multiple scales, on molecular, cellular and network levels, the prospects of KAR antagonists as a therapy for drug-resistant patients will become clear.

Restoring/Human Studies


Approximately half of the patients with a diagnosis of drug-resistant epilepsy are potential candidates for epilepsy surgery. If successful, epilepsy surgery may substantially reduce or eliminate disability.

The University-Hospital Timone has gathered significant experience in all possible forms of surgical intervention including lobectomy, subpial transsections, hemispherectomy, hemispheretomy, callostomy, deconnexion and radiosurgery.

When seizures come from multiple brain areas, the surgical risk of functional brain damage is too high and other therapeutic options should be envisaged.

Thalamic and vagus nerve stimulation

When epilepsy is really intractable, experimental procedures like NeuroStimulation can still be considered, especially vagus nerve stimulation (VNS) and deep brain stimulation (DBS) of the anterior nucleus of the thalamus (AN-DBS) which can reduce the frequency of seizures.

Epinext scientists are actively studying the compared efficacy of these approaches in multicentric studies. VNS has an even wider applicability, e.g. for patients with severe depression which often comes together with epilepsy.

The Epinext clinical teams are following the largest number of patients with stimulation implants in France, working towards predicting the efficacy of this treatment through continuous EEG signal analysis.


Epinext is a pioneer in developing biofeedback approaches as a therapy component for patients with drug-resistant epilepsy. A pilot study showed a promising 50% responder rate for patients with temporal lobe epilepsy, in particular in stress related epilepsies.

Gamma knife

Epinext clinicians have access to a “Gamma knife” which can precisely destroy small areas in the brain without invasive surgery. It has proven to be an efficient alternative to classic neurosurgery, especially for mesial temporal lobe epilepsy. As a world leader in gamma knife treatments, Epinext will extend its application to neocortical epilepsies where it‘s important not to harm the cortex where our speech abilities are residing.

Developing new surgical strategies in pediatric epilepsy surgery

To help children with catastrophic epilepsies due to large cortical malformations, Epinext will investigate new therapeutic strategies like focal deconnections and highly targeted thermocoagulations. These can reduce the risk of surgical complications and treat brain regions which are difficult to access, e.g. the insular cortex.

Modeling/Local circuits

Physiological repertoire of rhythms

SEEG measuring is still not used globally in clinical practice because it‘s so notoriously difficult to interpret SEEG traces and to distinguish between normal and abnormal activity in each patient. Epinext researchers plan to overcome this obstacle by mapping out these characteristic rhythms in a large-scale brain dynamics atlas, utilizing an unprecedently large amount of recorded SEEG data. This atlas will be subsequently used in diagnosis, modeling and education.

Biologically realistic modeling

To achieve biologically realistic brain modeling for individual patients, Epinext will simultaneously integrate new micro and macro approaches. While the macro side focuses on the organization of neuronal populations, the micro side covers slow metabolic processes (e.g. glucose) and environmental influences (e.g. extracellular concentrations of ions).

Modeling/Global circuits

Virtual patient development

In order to generate a sufficiently virtualized patient brain for predictive therapy evaluation, Epinext builds upon its long history leading The Virtual Brain (TVB) software platform. Already being a widely acclaimed, international effort in neuroinformatics, Epinext will turn it into an unprecedented multiscale project which can deliver well-quantified confidence levels in predictive personalized medicine.

The key for such a big improvement lies in validating the virtual models of TVB with the vast amount of clinical and lab data which are generated by Epinext activities:

  • Systematic cataloging and modeling of many patients‘ neuroimaging data from SEEG, EEG, MEG and fMRI, recorded for their resting state and when performing cognitive tasks
  • Translation of clinical stimulation protocols into models for independent testing and validation
  • Modeling of novel electrodes which are already tested by Epinext clinicians
  • Comparative validation of human and rodent brain models with imaging data from animal studies

How a virtualized patient brain can help understand epilepsy (video in French)

Seizure propagation

Epinext has already demonstrated in pilot studies that TVB can predict how a seizure propagates through the brain from the epileptogenic zone to other areas, causing impairments for sensory and motor abilities. The predictive quality will be improved further by comparing and adjusting them to actual clinical patient data (before and after surgery) and results from animal studies.

Example of a simulated complex seizure with epileptogenic zones and SEEG signals

Modeling/Model translation

Biomarker development and bioinformatics

In order to allow a reliable judgement of seizures, real ones in patients and virtualized ones in software, Epinext will develop novel metrics to characterize a seizure‘s quality and quantity, e.g. multiscale entropy, variance, coherence or causality.

These findings will augment The Virtual Brain software as a bioinformatics ontology and be made internationally available through International Neuroinformatics Coordination Facility (INCF).

Individualizable patient models: software support

In collaboration with Epinext partner Codebox, The Virtual Brain software will spin off an independent application which is tailored specifically to the needs of presurgical mapping of epileptic patients. Surgeons can use this application to guide the implantation of SEEG electrodes and simulate the probable effects of different surgical interventions.