Almost all patients with epilepsy living in the region of Paris have vitamin D deficiency, which is severe in 1/3 of the cases. The impact of this deficiency on epilepsy is unknown, despite the suggested benefits of vitamin D therapy including better seizure control and improvement of comorbidities (fatigue, anxiety, depression) in drug-resistant patients. Recommendations for vitamin D supplementation based on the serum level in the general population cannot be applied to patients with epilepsy due to interference of antiepileptic drugs in the vitamin D metabolism. Animal models, mechanisms of action studies and ecological information provide objective data for a direct antiepileptic effect of vitamin D. Human studies seem to confirm the antiepileptic effect of vitamin D but there are no controlled studies on large populations. The investigators aim to assess prospectively the effect of the treatment of vitamin D deficiency providing a high level of evidence. The investigators propose a multicentre placebo controlled randomized double-blind study, testing vitamin D supplementation against placebo in 400 drug-resistant patients to assess the short-term (3 months) and long-term (1 year) benefits on epilepsy. The investigators hypothesize that the treatment of vitamin D deficiency will result in significant reduction of seizure frequency, and improvement of comorbid symptoms as well as quality of life. The impact on the care of patients is important because better epilepsy control allows reduction of the antiepileptic drugs and side effects. This again is a key for the recovery of social and professional activities, and reduction of costs related to the disease.

The goal of this study is to establish the effect of the compensation of vitamin D deficiency in the treatment of drug-resistant epilepsy with a high level of evidence in a large population. For that, the investigator aim to perform a controlled multicentre prospective double-blind randomized study in 400 patients.

Experimental design

After a reference period for the collection of clinical data and seizure frequency, all eligible patients will undergo biological analyses including serum vitamin D levels, calcium, creatinine and albumin. In case of proven deficiency (25(OH)D <30ng/mL), patients will be randomized into two arms:

Experimental group: drug-resistant epileptic patients supplemented with vitamin D. The vitamin D administered corresponds to the scheme described below. Control group: drug-resistant epileptic patients supplemented with placebo, administered placebo dose is equal to therapeutic dosing scheme of vitamin D. Randomization will be centralized and stratified by center (considering local disparities) and according to the results of serum level, retaining the threshold of 10ng/mL to define severe deficiency, 20 ng/mL for moderate deficiency and 30 ng/mL for slight deficiency of vitamin D.

The drug evaluated is vitamin D3 (cholecalciferol), administered in the form of oral solution (2 mL) containing 100,000 IU, manufactured by the laboratory CRINEX and marketed under the name UVEDOSE®. The therapeutic indication is the treatment and/or prophylaxis of vitamin D deficiency. The usual supplementation in adult is 1 to 2 doses per month depending on the severity of the deficiency, the maintenance ranging from 1 dose every month or every 2 to 3 months.

Administration in the experimental group and control group The supplementation scheme is based on the usual recommendations but does not take into account the severity of the initial deficiency, that would be impossible to implement in the context of a double-blind study (with in particular the need to adapt each placebo dose in the control group, risking the feasibility study). To address this potential bias, the investigators choice a supplementation scheme corresponding to a moderate deficiency (25(OH)D >10 and <20ng/mL) considering that this is the most common in clinical practice and there is no risk of overdose for 25(OH)D level between 20 and 30ng/mL.

The dosing schedule is as follows:

UVEDOSE: 1 dose of 100,000 IU every 15 days for 2 months, then 100,000 IU the 3rd month (5 doses during the first 3 months). Dosage of 25(OH)D will be conducted at 3 and 6 months in order to verify the compensation of the initial deficiency and to avoid the risk of overdose. Monitoring of serum calcium, albumin and creatinine will be performed at the same time. Maintenance therapy in the treatment group will be set at 100,000 IU per month for 6 months (11 doses for 12 months followup). After the blind period, the same will be applied to the control group (5 doses in 3 months and 1 dose per month for 3 months, 8 doses for 12 months follow-up).

Apart from an overdose, no side effects are expected. An overdose of vitamin D may be suspected in case of occurrence of one or more of the following signs: headache, asthenia, anorexia, weight loss; nausea, vomiting; polyuria, polydipsia, dehydration; high blood pressure; calcium lithiasis; renal insufficiency. Biological signs reflecting the vitamin D overdose are hypercalcemia with hypercalciuria. Serum levels of 25(OH)D from 100 to 150 ng/mL are exceptionally accompanied by hypercalcemia. High risk of hypercalcemia with clinical consequences are reported for levels around 250ng/mL whereas no toxic effects are described for serum levels <150 ng/mL. To avoid any risk, the supplementation will be stopped if the dosage of 25(OH)D is >100ng/mL and/or serum calcium levels >2.70mmol/l; if the serum calcium is >2.60 and <2.70mmol/l, a control will be done and the treatment stopped if >2.60mmol/L. Vitamine D treatment will be stopped also if creatinine is >200μmol/L. Vitamin D causes no known drug interactions. However its metabolism is accelerated when combined with enzyme-inducing drugs, including antiepileptic drugs.

The entire treatment will be provided to each patient using a kit of 5 doses of UVEDOSE after randomization in the experimental group and the unblinding in the control group according to the protocol established. The treatment scheme will be clearly explained and dates of treatment indicated on the attached schedule. Maintenance therapy will be provided on the same terms.

Patients will note on this calendar the effective date of each dose and bring empty containers during the visit at 3 months, 6 months and 12 months. The labeling will be performed by the laboratory CRINEX, the kits will be delivered to pharmacy centers. The pharmacy of each participating center will handle the management of returns and the destruction of processing units. Patients in the control group will receive placebo of vitamin D in the same manner as experimental patients.

The study entails a screening-visit followed by a 3 months reference period, an inclusion visit followed by randomization, two follow-up visits (3 and 6 months) and an end-study visit at 12 months. The duration of study for each patient is 12 months following the 3 months reference-phase. The double-blind phase is between the randomization (V1) and the 3 month visit (V2), followed by a 9 month open phase for the 2 groups Eventually, all patients with a vitamin D deficiency will be compensated with a 3-month higher dose treatment phase and a 6-month maintenance phase for the experimental group. The control group will receive the placebo for 3 months, then a vitamin D treatment including a 3-month higher dose and a 3 month maintenance phase.

The duration of the inclusions is scheduled for 12 months, for research duration of 24 months. All eligible patients who meet ILAE (International League Against Epilepsy) criteria of drug-resistance and inclusion criteria will be informed of the opportunity to participate in this research. If accepted, the participants will be asked to fill out a precise seizure diary for 3 months before their inclusion in the study (V0). The antiepileptic treatment should not be changed during the reference period. The patient cannot receive vitamin D during this period.

Inclusion visit (V1):

After obtaining informed consent and verification of inclusion criteria, seizure frequency is noted and determination of serum 25(OH)D level is done. The vitamin D serum concentration is determined with the same technic in each center (immunoluminometric CLIA (Chemiluminescence Immunoassay Analyzer) - Liaison XL). The concentration serum calcium, albumin and creatinine is determined at the biochemistry laboratory of the participating center. The self-questionnaires (Fatigue impact scale : FIS, Hospital anxiety depression scale :HAD, Quality of life in epilepsy : QOLIE 31) are completed and handed to the investigator. Clinical and demographic data are collected on the case report. After receipt of 25(OH)D level results, patients in whom deficiency is objectified (25(OH)D <30ng/mL) are randomized between control and experimental group. It is expected 10% of non-deficient patients who are not randomized and followed out of the protocol. Treatment appropriate to the protocol for each group (vitamin D or placebo) is delivered by the center pharmacy. The results of biology (serum calcium, albumin, creatinine) are first referred to the investigator, the treatment being delivered in absence of contra-indication (see above).

3 month-visit (V2): Clinical examination, seizure frequency, determination of 25(OH)D levels, serum calcium, albumin, creatinine) and scales (FIS, HAD, QOLIE 31); monitoring of compliance (ratio of packaging and check-up of dates); lifting of the blinding phase. Introduction of vitamin D supplementation in the placebo group according to the initial plan, pursuit of vitamin D supplementation in the experimental group, stop if the serum level is>100 ng/mL and/or serum calcium>2.70mmol/L and/or creatinine>200 μmol/L. Control of serum calcium if between 2.60 and 2.70 mmol/L, stop if> 2.60 mmol/L.

6 month-visit (V3): similar to V2, maintenance supplementation in the 2 groups: 1 dose per month for 3 months, except if biological contraindication.

End-study visit (V4, 12 months) similar to V1-V2, vitamin D trial stop. In case of deterioration of the clinical situation requiring modification of the antiepileptic treatment (excluding one-time treatment with benzodiazepines), the patient will be considered not-responder but will not be excluded from the study. The seizure frequency in this case for the primary judgment criterion will be the one before the treatment change (LOCF method).

Inclusion Criteria:

• Age> 15 years
• Drug-resistant epilepsy (see definition above)
• Having at least 6 unprovoked seizures in the previous 3 months
• Epilepsy syndrome unequivocally established
• Ability to reliably quantify the seizure frequency
• Antiepileptic treatment stable for 3 months prior to inclusion
• No vitamin D treatment in the 6 months prior to inclusion vitamin D supplemental diet
• Medication compliance (confirmed by plasma levels if available)
• Agreeing to participate in the study
• Having a social insurance
• Parental agreement if patient below the age to be able to give consent (or guardian if protected adult)

Exclusion Criteria:

• Progressive brain pathology
• Status epilepticus in the 2 years prior to inclusion,
• epilepsy surgery planned in the current year
• Pregnancy or breast-feeding
• Treatments influencing the metabolism of vitamin D other than anticoagulants (rifamycin, isoniazid, ketoconazole, 5-FU fluorouracil), leucovorin)
• Known hypersensitivity to vitamin D, patients with a history of granulomatosis (especially sarcoidosis)
• Contraindication to treatment with Uvedose referring to the summary of product characteristics
• Current or past hypercalcemia or situations accompanied by increased vulnerability to hypercalcemia as arrhythmia or digitalis therapy, subjects with calcium lithiasis
• Moderate renal impairment with creatinine clearance <60 mL/mn assessed by MDRD (Modification of Diet in Renal Disease)
• Participation in other studies of other experimental drugs within 30 days before enrollment in the study
• Abuse of alcohol or drugs

Double blind