Page 224 - ILAE_Lectures_2015

Ictal PET scans can only be obtained fortuitously, because of the two-minute half-life of 15O
and the fact that cerebral uptake of 18FDG occurs over 40 minutes after injection, so that
cerebral glucose utilisation data will reflect an amalgam of the ictal and post-ictal conditions.
The place of 18FDG-PET as a tool for localising an epileptic focus has been greatly reduced
following developments in MRI, as the finding of a definite focal abnormality with the latter
technique, such as HS, renders an 18FDG-PET scan superfluous141. The place of the
investigation is in the presurgical work up of patients with refractory partial seizures and
normal or non-definitive MRI scans, or if data are discordant, in order to generate a
hypothesis that may then be tested with intracranial EEG recordings.

Temporal lobe epilepsy

Several studies of 18FDG-PET have found a 6090% incidence of hypometabolism in the
temporal lobe inter-ictally in adults and children with TLE. The results of comparative studies
depend critically on the relative sophistication of the techniques used. In a comparative study
of patients with TLE it was concluded that 18FDG-PET data did not provide clinically useful
data if the MRI findings were definite, but had some additional sensitivity142. This was
confirmed in a more recent study in which 18FDG-PET data correctly lateralised the seizure
focus in 87% of patients with TLE and normal conventional imaging143. Visual assessment
of hypometabolism is less accurate than quantification with a voxel-based comparison with
normative data and co-registration with MRI144,145. Absence of unilateral temporal
hypometabolism does not preclude a good result from surgery146. Bilateral temporal
hypometabolism was associated with a poor prognosis for seizure remission after surgery147.

18FDG-PET studies have been less reliable for precise localisation of seizure onset than for
answering the question of lateralisation. In an evaluation of 18FDG-PET in patients with TLE
and different pathologies, those with HS had the lowest glucose metabolism in the whole
temporal lobe, followed by patients whose seizures arose laterally. Patients with mesiobasal
tumours generally had only a slight reduction of glucose uptake in the temporal lobe. The
metabolic pattern was different between patients with mesial and lateral temporal seizure
onset, but there was not a clear correlation between the location of the epileptogenic focus
defined with EEG and the degree of hypometabolism148. Reduced glucose metabolism and
FMZ binding have been reported in the insula in patients with TLE. Emotional symptoms
correlated with hypometabolism in the anterior part of the ipsilateral insular cortex, whereas
somaesthetic symptoms correlated with hypometabolism in the posterior part. Insula
hypometabolism, however, did not affect the outcome from temporal lobe resection149.

Although reduced glucose metabolism may occur in the face of normal structure150, atrophy
is a major determinant of cerebral metabolism measured with 18FDG PET, and partial volume
correction is necessary to understand the relationship between hippocampal structure and
functional abnormalities151. There have been few studies of newly diagnosed patients; only
20% of children with new onset epilepsy had focal hypometablism152.

Frontal lobe epilepsy

18FDG-PET shows hypometabolism in about 60% of patients with frontal lobe epilepsy. In
90% of those with a hypometabolic area, structural imaging shows a relevant underlying
abnormality. The area of reduced metabolism in frontal lobe epilepsy may be much larger
than the pathological abnormality. In contrast, however, the hypometabolic area may be
restricted to the underlying lesion153. There have been three main patterns of
hypometabolism described in patients with frontal lobe epilepsy: no abnormality; a discrete
focal area of hypometabolism; diffuse widespread hypometabolism. Overall, published