Multimethodological imaging studies of pathophysiological mechanisms in 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 if the functional integrity of the limbic networks in mesial temporal lobe epilepsy (MTLE) is impaired. The present work addresses these important issues by combining several different magnetic resonance (MR) and positron emission tomography (PET) methods.

According to the international classification of epilepsies, IGE is not associated with neurologic or neuroradiologic signs. Recent findings of neuroreceptor changes and migrational disturbances in IGE, however, question this concept. According to electrophysiological experiments, IGE is associated with abnormal thalamo-cortical volleys and an altered GABA and glutamate transmission, suggesting that both thalamus and cortex could be affected in IGE.

In study I MR spectroscopy (MRS) was used to specifically investigate whether the concentration of Glx (glutamine and glutamate) is altered in the thalamus of patients with juvenile myoclonic epilepsy (ME) and generalized tonic clonic seizures on isolation (GTCS) (two of the IGE syndromes). Another question was whether there are any signs of neuronal damage and atrophy on this region. MRS revealed elevated thalamic concentrations of Glx along with reductions of the neuronal marker Nacetyl aspartate (NAA) in patients (n = 43) compared to healthy controls (n = 38). Furthermore, volumetric analysis showed that the thalamus volume was reduced.

Next, in study II we applied MR volumetry and voxel based morphometry to exploratively investigate possible cortical and subcortical changes in the structural volumes, and fractions of gray matter, white matter and cerebrospinal fluid (CSF) in patients with GTCS (n =19). Patients exhibited an elevated frontal lobe fraction of CSF, and a reduced fraction of gray matter in the frontal, parietal, temporal cortex, thalamus, and cerebellum. The thalamus and cerebellum also showed reduced volumes, as did the caudate and putamen.

Together, these findings suggest that IGE is associated with specific thalamic and regional cortical tissuechanges. The distribution of these changes is compatible with the semiology of JME and GTCS. The results provide a further argument for a re-evaluation of the current classification and diagnostic criteria for IGE.

In study III we used passive perception of familiar and unfamiliar odors as a unique method to investigate how the extrahippocampal limbic structures in patients with mesial temporal lobe epilepsy (MTLE) respond to normal environmental stimuli. The underlying rationale was that interictal epileptic dysfunction may exceed the epileptic seizures generated in the hippocampus, and that direct studies of functional integrity in limbic circuits in humans has not been done earlier. In controls, both odor types bilaterally activated the amygdala, piriform, anterior insular and anterior cingulate cortex. Familiar odors activated also the right parahippocampus, and the left Brodmann area (BA) 44,45,47.

Patients failed to activate the amygdala, piriform and anterior insular cortex ipsilaterally to the epileptogenic side. Those with left MTLE showed, in addition, no activation of left BA 44,45 and 47 with familiar odors, which they perceived as less familiar than the controls and patients with right MTLE. The piriform cortex was functionally disconnected with its homologous area and the contralateral amygdala and anterior insular cortex in patients with MTLE. In addition, behavioral data indicate that left MTLE patients rated familiar odors as less familiar, while right MTLE found familiar and unfamiliar odors less intense.

We conclude that odor activation in patients with MTLE fail to recruit several extra hippocampal limbic regions on the side of seizure onset. Imaging of olfaction in MTLE may, therefore, be used both to locate the epileptogenic area, and to detect patterns of activation-related changes in limbic processing and connectivity. The first two studies add new information regarding morphological changes in IGE. The distribution of these changes might add new arguments for reevaluation of diagnostic criteria. The third study proposes a new tool to diagnose localized limbic dysfunctions in MTLE.