Page 142 - ILAE_Lectures_2015

There seems no doubt that the prognosis for post-traumatic epilepsy is considerably worse
than for epilepsy for which no cause is found. Caveness reported a remission rate of
approximately 50%3. Jennett’s series reported a remission rate of 25%, but one-third of
patients continued to have frequent seizures. There is some evidence that the later the onset
of epilepsy following head injury the less likely is remission. Furthermore, seizures appearing
to arise from the temporal lobes seem to have a worse prognosis than those arising elsewhere.

There is little evidence to suggest that early AED therapy has a significant effect in preventing
the development of later epilepsy. This may either be because AEDs do not influence the
natural history of this form of epilepsy, or because head-injured patients show a marked
tendency to be non-compliant with prophylactic therapy.

Post-operative epilepsy
The overall incidence of post-operative seizures in a five-year period following supratentorial
craniotomy is approximately 17%. The incidence may vary from as low as 3% to as high as
92% depending on the condition for which craniotomy is carried out.

A total of 20% of patients undergoing surgery for intracranial aneurysms will develop post-
operative seizures. The risks are low for aneurysms of the internal carotid (7.5%) but high
for aneurysms of the anterior communicating (21%) and middle cerebral artery (38%).
Surgery for arteriovenous malformations (AVMs) and spontaneous intracerebral
haematomas carries a 50% and 20% risk of de novo epilepsy, respectively. A considerable
portion of this risk seems to be directly attributable to surgery, as the risk of epilepsy
associated with aneurysms managed conservatively was approximately 8% in 261 patients
and an approximately 20% risk over 20 years for AVMs managed conservatively.

The incidence of epilepsy following surgery for supratentorial abscess is extremely high, and
virtually all patients develop seizures if followed up for a sufficiently long period of time.
The risk of seizures complicating insertion of an indwelling ventricular shunt is about 24%.

The risk of tumour surgery causing epilepsy is more difficult to identify, particularly for
progressive tumours such as gliomas. Seizures may develop de novo following surgery for
meningioma in 22% of cases, though approximately 40% of meningioma patients who had
pre-operative seizures do not have further seizures post-operatively. Once again there is a
clear relationship between time of surgery and the development of seizures. Approximately
70% who have seizures will have done so by one year and 90% by two years post-operatively.
To date there is no evidence that early prophylactic treatment with AEDs significantly
reduces the risk of post-operative seizures.

Tumour epilepsies
Tumours remain a relatively rare cause of epilepsy but the incidence of tumour epilepsy is
clearly age related. In one series tumours were detected in 16% of patients developing
epilepsy over the age of 20, and in 22% of patients developing partial epilepsy over this age.
The diagnosis of tumour-based epilepsies is usually straightforward and indicated by the
presence of developing focal neurological signs and symptoms, a focal EEG abnormality and
by neuroimaging. In patients with benign tumours who present only with epilepsy diagnosis
is difficult and management even more problematic.

There is no doubt that the more benign the tumour the more likely it is to present with a
history of epilepsy (Table 3). The siting of the tumour also appears to influence the likelihood
of a presentation of epilepsy (Table 4). The likelihood of finding a neoplastic basis for
epilepsy beginning in adult life is influenced by partial seizure type (Table 5).