Most patients are prescribed valproate as their first antiepileptic drug. It is unknown which is the best second-line drug when patients do not become seizure free on valproate. This has led the Dutch Epilepsy Clinics Foundation (SEIN) to start the SLICE study. Adult patients with partial and/or tonic-clonic seizures, insufficiently responding to valproate, are recruited for this study. These patients are randomized to receive one of three other drugs. Patients wil initially use this drug next to valproate. Neurologists of more than 20 general hospitals en neurologists of SEIN are participating in this study.

The purpose of the project is to compare several antiepileptic drugs given to adult patients with epilepsy after they have not become seizure free on valproate as a first-line antiepileptic drug. The drugs will first be evaluated in combination with valproate and in case of success (being a seizure reduction of more than 50%) will also be evaluated in monotherapy.

Patients who did not become seizure free on valproate will be identified by neurologists in the participating hospitals. When these patients are willing to participate, they are randomized to one of three drugs: carbamazepine, lamotrigine and levetiracetam. In phase 1 of the project they keep on using valproate. The randomized second-line drugs will be titrated to a first dose level and the effectiveness of the combinations will be evaluated. When seizures persist and adverse effects allow it, the add-on drug is titrated to a second dose level and again the effectiveness of the combination is evaluated. When seizures still continue and adverse effects allow it, the add-on drug is titrated to a third and final dose level. When a patient does not become seizure free on a combination on that final level or adverse effects have prevented a dose increase to a higher level, that combination has failed in phase 1. When the patients does become seizure free on his or her combination, the combination is deemed a success for that patient. A patient will proceed to phase 2, when he or she has at least experienced a 50% seizure reduction.

In phase 2 of the project the second-line drug will be given in monotherapy. This means that valproate will be withdrawn. The dose of the second drug will be increased accordingly. The effectiveness of the drugs in monotherapy will be evaluated. The combined results of phase 1 and 2 will enable us to interpret the results. When all patients who became seizure free on a combination in phase 1, stay seizure free in phase 2, the efficacy of the combination should be attributed to the add-on drug. When these patients all develop seizures again, the efficacy of the combination should be attributed to the combination.

The primary outcome measure is percentage seizure free. Secondary outcome measures are adverse effects and the results of clinimetric epilepsy scales. Serum levels will be measured during the project. The projected sample size for each group has been lowered from 75 patients per group to 20 patients per group.

At this moment, neurologists of about 20 general hospitals are collaborating in this project. Inclusion of patients will continue until June 2006. The follow-up of patients and analysis of results will be carried until the projected end of the project.

Inclusion Criteria:

• Adult patients with generalized tonic-clonic, complex partial and/or simple partial seizures. The seizures should be well-defined according to the International Classification of Epileptic Seizures (1). Therefore, an accurate history and adequate neurophysiological data should be present in each case in order to confirm the diagnosis.
• Patients on valproate monotherapy who are not seizure free at at the maximal dose they can tolerate.
• Patients should be able to understand the patient information concerning the study and be able to give informed consent.

Exclusion Criteria:

• Patients who failed on VPA monotherapy because of other causes than lack of seizure control at a maximally tolerated dose (unable to tolerate the lowest maintenance dose of VPA, idiosyncratic reactions, non-compliance)
• Absence seizures or juvenile myoclonic epilepsy
• Acute or progressive neurological disorders
• Alcohol or other substance abuse
• History of severe psychiatric illness