Multimethodological brain imaging studies of human epilepsy

Abstract

Although the pathophysiological mechanisms of human epilepsy are extensively investigated, many questions remain unanswered. One is whether idiopathic generalized epilepsies (IGE) have anatomical substrates. Another is functional integrity of the limbic networks outside the epileptogenic region in mesial temporal lobe epilepsy (MTLE). The present work addresses these important issues by combining several different magnetic resonance (MR) and positron emission tomography (PET) methods.

According to electrophysiological experiments, IGE is associated with abnormal thalamo-cortical volleys, suggesting that both thalamus and cortex could be affected. In Study I MR spectroscopy (MRS) was used in patients with two major IGE syndromes of adulthood - juvenile myoclonic epilepsy (JME) and generalized tonic clonic seizures on isolation (GTCS) to investigate whether the thalamic concentration of Glx (glutamine and glutamate) is altered, and if there are any signs of neuronal damage in this region. The concentration of Glx was significantly elevated in both syndromes, whereas the concentration of the neuronal marker N-acetyl aspartate (NAA) was reduced (Study I).

To exploratively investigate possible anatomical changes also elsewhere in the brain we subsequently employed MR volumetry and voxel based morphometry (Study II) in 19 patients with GTCS and 52 controls. Reduced gray matter fractions were detected in patients in the frontal, parietal, temporal cortex, the thalamus, and cerebellum, along with elevated frontal lobe fractions of cerebrospinal fluid. Furthermore, the structural volumes were reduced in the thalamus, cerebellum, and also the caudate and putamen.

These findings, suggesting a particular affection of motor circuits in human IGE were further evaluated in Study III, which strictly focused on the dopamine (DA) system. Only patients with JME were investigated in these first experiments because our previous studies showed that JME in contrast to GTCS is associated with working memory problems, which like seizures could be attributed to the DA system. Using PET and [11C] PE2I we estimated the binding potential to the DA transporter (DAT) in the substantia nigra/midbrain, as well as the striatum. The nigral/midbrain DAT binding was significantly reduced in relation to controls, whereas the striatal values were normal. Patients also showed impairments in executive and motor functions, with results directly related to the midbrain DAT, and suggesting that the DA system indeed may be involved in the pathophysiology of JME.

Together, the findings in study I-III suggest that IGE is associated with specific cortical and subcortical changes. Their distribution is compatible with the semiology in the two investigated syndromes. The results provide a further argument for a re-evaluation of the current classification and diagnostic criteria for IGE, which currently assumes absence of anatomical substrates.

In study IV cerebral blood flow was measured with PET during passive perception of familiar and unfamiliar odors as a unique tool to investigate how the extrahippocampal limbic structures respond to normal environmental stimuli in mesial temporal lobe epilepsy (MTLE). Although widespread interictal metabolic and receptor changes may exceed the seizure generating area, little is known about functional integrity of the limbic circuits in human MTLE. In controls, both odor types bilaterally activated the amygdala, piriform, insular and cingulate cortex. Familiar odors also activated the right parahippocampus, and the left Brodmann area (BA) 44, 45, and 47. Patients failed to activate the amygdala, piriform and anterior insular cortex on the epileptogenic side. Those with left MTLE also could not activate the left BA 44, 45 and 47 with familiar odors, which they perceived as less familiar than the controls. Analysis of functional connectivity confirmed these findings including the functional disconnection with the language circuits in left, but not right MTLE.

Imaging of odor perception seems to be a suitable approach to delineate functional disintegration of the limbic networks in MTLE. It shows an altered response in several regions, which may underlay some interictal behavioral problems associated with this condition.