Page 359 - ILAE_Lectures_2015

freedom compared to only 35% with symptomatic partial epilepsy and 45% with cryptogenic
partial epilepsy28. Temporal lobe epilepsy (TLE) is associated with a poorer prognosis than
extra-temporal lobe epilepsy29,30.

For patients with a single identified lesion, TLE with hippocampal sclerosis (HS) had a
particularly bad prognosis (11% seizure free) compared with other aetiologies (24% with
cortical dysplasia seizure free). Patients with HS and another identified pathology (dual
pathology) had the worst prognosis (3% seizure free)29. In another study no difference in
prognosis between those with symptomatic and cryptogenic partial epilepsy was found30.
Comparing patients by aetiology, they found that mesial TLE had the worst prognosis
compared to rates for other aetiologies30.

The impact of medication on prognosis

In the Western world most patients are commenced on AED after two unprovoked seizures,
implying that prognostic studies from Western countries are essentially those of treated
epilepsy. Evidence from studies from resource-poor countries where a significant treatment
gap exists suggests that many patients may enter spontaneous remission with no AED31.

Indeed the response to AEDs in patients with chronic long-standing epilepsy is comparable
to that of patients with new-onset seizures31,32. Such evidence contradicts the belief that
epilepsy is a chronic progressive condition unless early treatment is commenced33. It has been
suggested that patients with epilepsy can be subdivided into prognostic groups based on their
aetiology and epileptic syndrome. This important concept implies that the need and response
to antiepileptic treatment in epilepsy is determined by the different prognostic groups1,2.

Early versus late treatment

Two studies have assessed the impact of medication on the risk of seizure recurrence. In the
FIRST study, patients with first unprovoked generalised seizures were randomised to either
immediate treatment (treated group) or to treatment only after a further seizure (untreated
group). While immediate treatment reduced the risk of early relapse, it did not affect the long-
term prognosis, with comparable five-year remission rates in the two groups34.

In the MESS study patients with a single seizure or early epilepsy (all types) were randomised
to receive immediate or deferred treatment. Patients in the immediate treatment group had
increased time to first and second seizure and first generalised seizure, in addition to having
a reduced time interval to two-year remission. At five years’ follow-up, however, 76% in the
immediate group compared to 77% in the deferred group had achieved 3−5 years’ seizure
freedom35.

In conclusion, immediate treatment delays the early recurrence of seizures but does not affect
the medium- or long-term prognosis.

Prognosis following AED withdrawal

In the largest randomised controlled trial of continued treatment vs drug withdrawal in 1013
patients in remission (two or more years seizure free), at two years post-randomisation 41%
of those who had discontinued medication had had a recurrence of seizures compared to 22%
of those who stayed on medication. The difference in relapse rates between the two groups
was maximal at nine months, with the rate of relapse higher in the discontinuation group up
to two years’ follow-up, but by 2−4 years the risk of relapse was higher in those continuing
treatment36. Patients who experienced a relapse were followed up, and by three years 95%