Autoimmune epilepsy Modulated by IVIg - effects on Cortical Excitability, the AMICE study

Autoimmune epilepsy is a severe syndrome. Using a conservative estimation, 1%
of the 70 million epilepsy patients worldwide will have an autoimmune origin.
It is most frequently refractory to anti-epileptic drugs, but often responds to
immunotherapy. Intravenous immunoglobulins (IVIg) are used off-label to treat
patients with autoimmune epilepsy, but there have been no structured
prospective studies with immunotherapy. A structured prospective study with a
complete set of matched serum and CSF samples before and after IVIg treatment
is needed, to provide proof of a therapeutic effect of IVIg in autoimmune
epilepsy (both clinically and serologically). Thereby there is need for a
biomarker for prognosis and treatment decisions. Transcranial Magnetic
Stimulation (TMS) can assess cortical excitability safely and non-invasively.
Also in vitro experiments are necessary to provide a direct link between
antibodies and cortical excitability.

1. Assessing the efficacy of IVIg in autoimmune epilepsy, both clinically and
serologically.
2. Identifying an objective marker of therapy response in epilepsy, measuring
cortical excitability by TMS-EEG/EMG, in vivo.
3. Providing evidence that patients' antibodies affect cortical
hyperexcitability, in vitro.

Prospective single group open label clinical intervention trial with IVIg.

All patients receive 2 courses of IVIg; 0.4 grams/kg/day for 5 days, starting
on day 1 and day 22.

There is a small risk from participation of this trial, because patients
receive 2 courses of IVIg with an interval of 3 weeks, whereas in the standard
care this interval is 4 weeks. Direct side effects of IVIg are mild and
manageable. Potential serious side effects are rare (0.01-0.1%) to very rare
(<0.01%) and include haemolytic anaemia, aseptic meningitis, thromboembolism,
acute renal failure and transfusion-related acute lung injury. In case of <50%
epilepsy reduction after the first course, deterioration or relapse, choice for
additional immuno- or antiseizure therapy is in accordance with standard of
care. In practice, we will add ivMP to IVIg for the second course. Generally,
we will not alter anti-seizure medication during the first 6 weeks. Of course,
if clinical care necessitates AED, we will do this.
A substantial part of the burden is regular care, such as the admission and a
great part of the investigations. The burden consists of 2 admissions of 5 days
for daily IVIg infusions (10 IVIg infusions in total, if possible second course
of IVIg in daycare); followed by 3 outpatient visits and 2 telephone
consultations. In total, 7 venepunctures, 2 lumbar punctures, 2 brain MRIs and
3 TMS-EEG/EMGs are performed, combined with the outpatient visits. During all
visits, general physical and neurological examinations will be performed and 3
questionnaires will be conducted repeatedly at 6 time points. A diary will be
kept during the research. TMS-EEG/EMG is generally well-tolerated. Transient
mild pain in head and neck are frequently described. The risk of seizures is
very low (almost zero). TMS-EEG/EMG has no individual interest, but does have a
group interest, since it has major implications for patients with autoimmune
epilepsy in the future. TMS-EEG/EMG can possibly be used as a biomarker for
prognosis and treatment decisions in autoimmune epilepsy.
Since autoimmune epilepsy is a severe disease often accompanied by severe
neurologic deficits, it should be treated aggressively. IVIg are generally well
tolerated and have a better side effect profile than corticosteroids. IVIg are
more convenient and cost-effective compared with plasmapheresis. With a
cumulative dosing scheme of IVIg, we expect patients to improve in days to
weeks. Thereby, IVIg does not have an epileptogenic effect itself and does not
influence the TMS-EEG/EMG.
This study will fill important gaps in the knowledge of treatment of epilepsy
in autoimmune encephalitis. The serological and EEG data will learn us much
about the effect of IVIg on antibodies and brain networks and might provide new
options to predict treatment response.