Clinical correlations with hippocampal atrophy

There is now a consensus that magnetic resonance imaging (MRI) is a sensitive and specific indicator of mesial temporal sclerosis (MTS) in patients with partial epilepsy. MTS is the most common pathological finding underlying the epileptogenic zone in patients undergoing temporal lobe surgery for medically refractory partial seizures. MRI-based hippocampal volumetric studies (i.e., quantitative MRI), has been shown to provide objective evidence for hippocampal atrophy in patients with MTS. The hippocampal volume in the epileptic temporal lobe has correlated with the neuronal cell densities in selected hippocampal subfields. A history of febrile seizures in childhood and age of unprovoked seizure onset have been associated with MRI-based hippocampal volumetry. There is conflicting evidence regarding the relationship between the duration of the seizure disorder and volumetry. Quantitative MRI has compared favorably to other noninvasive techniques (e.g., scalp-recorded EEG), in indicating the diagnosis of medical temporal lobe epilepsy (MTLE). MRI-identified hippocampal atrophy has also been a favorable prognostic indicator of seizure outcome after temporal lobe surgery. The presence of hippocampal atrophy appears to serve an in vivo surrogate for the presence of MTS.     

Hippocampal MRI volumetrics and temporal lobe substrates in medial temporal lobe epilepsy

Forty-nine consecutive patients undergoing anteromedial temporal lobe resection for medically intractable temporal lobe seizures, and averaging 2 yr (range 6 mo to 4 yr) postoperative follow-up, were selected for a retrospective study. This study correlated magnetic resonance imaging (MRI) derived hippocampal volumetrics, preoperative demographics, postoperative seizure control, and tissue analysis, including hippocampal CA (cornu ammonis) field neuronal, and glial cell counts, and immunohistochemistry (IHC) evidence for dentate sprouting and reorganization. These measures were compared in hippocampi with or without an adjacent presumptive epileptogenic temporal lobe mass. Mesial temporal sclerosis (MTS) was defined as >50% neuronal cell loss averaged across all CA fields with NPY (neuropeptide-y) and somatostatin reorganization. These patients may or may not include granule cell sprouting as determined by dynorphin staining. Patients were divided into two groups based on CA field neuronal cell counts, one averaging >50% cell loss and one averaging <50% cell loss. For the MTS group (N = 38), 89% had significant volumetric atrophy of the ipsilateral hippocampus, 74% had dentate reorganization, and complete seizure control was seen in 76% of these patients. In one subgroup of the <50% cell loss group, patients with medial temporal lobe epilepsy caused by a mass in the medial temporal lobe (mass group) (N = 6), 33% demonstrated significant volumetric atrophy of the hippocampus ipsilateral to the mass, 0% had dentate sprouting, and seizures were completely controlled in 67%. For the second subgroup of the <50% cell loss group, patients without mass lesions (N = 5) who were classified as the paradoxical medial temporal lobe epilepsy group (paradoxical group), 20% had ipsilateral hippocampal atrophy, 0% had dentate reorganization, and complete seizure control was seen in 60% of these patients. In conclusion, for the MTS group, hippocampal atrophy proven by MRI volumetrics was highly predictive of significant neuronal cell loss and an excellent indicator of success. However, in patients who had a foreign mass, hippocampal atrophy was not necessarily indicative of significant neuronal cell loss and MRI volumetrics was not a factor in the determination of a successful outcome. Furthermore, patients without mass lesions who have normal volumetrics but demonstrate hippocampal disease through invasive electrode monitoring, are likely to have paradoxical medial temporal lobe epilepsy, seizures beginning at a later age, and a lower, but not insignificant, success rate than the classical mesial temporal sclerosis group.     

Automatic detection of hippocampal atrophy on magnetic resonance images.

An automatic method for identifying hippocampal atrophy on magnetic resonance (MR) images obtained from patients with clinical evidence of temporal lobe epilepsy (TLE) is described. The method is based on the analysis of image intensity differences between patients and controls within a volume of interest (VOI) centred on the hippocampus. The core of the method is a fully automatic signal intensity-based inter-subject image registration technique. In particular, a global affine registration to a reference image is performed, followed by a local affine registration within the VOI. A mask produced by manual segmentation of the mean hippocampus for 30 control subjects enabled investigations to be restricted to a specified region of the VOI approximately corresponding to the hippocampus. Normal variations of hippocampal signal intensity were computed from images obtained for the 30 control subjects. The manual method of hippocampal volumetry, currently an important component of the pre-surgical evaluation of patients with clinical evidence of medically intractable TLE, is used to determine the lower 1st percentile limits of normal hippocampal volume. Hippocampi with volumes below this limit are defined as atrophic. We investigated whether the automatic method can correctly distinguish between 15 patients with significant hippocampal atrophy according to absolute volumes and a further 14 controls. ROC curves enabled evaluation of sensitivity and specificity in respect of an intensity threshold. 100% specificity is required when determining suitability of patients for neurosurgery, resulting in levels of 50% and 70% sensitivity in detecting atrophy in the right and left hippocampus, respectively. We propose that the method can be developed as an automatic screening procedure.     

Clinical applications: MRI, SPECT, and PET

MRI, PET, and SPECT are all used to image abnormalities in the epileptic brain. Comparison of the techniques is difficult because they measure different aspects of the epileptic process—structure, metabolism, and perfusion. SPECT is the only one that can be systematically applied during seizures, while all three are used to image interictal abnormalities. Literature review suggests that of interictal techniques, PET has the highest diagnostic sensitivity in temporal lobe epilepsy (TLE) (84% vs. 66% for SPECT, 55% for qualitative MRI, 71% for quantitative MRI) while SPECT has the highest sensitivity in extratemporal epilepsy (ETE) (60% vs. 43% for MRI and 33% for PET). The highest diagnostic sensitivity and specificity were achieved by ictal imaging with SPECT (90% in TLE, 81% in ETE). The techniques, however, were not always redundant. One reason for the wide discrepancy of results in TLE and ETE might be the differing pathologic substrates. A literature review of imaging findings associated with mesial temporal sclerosis (MTS), developmental lesion or tumor as the underlying abnormality associated with epilepsy supports this explantion. PET and MRI are much more sensitive to MTS than SPECT (100%, 95% vs. 70%). On the other hand, in developmental lesions the three techniques are equally sensitive (88–92%) and in tumors, MRI was most sensitive (96%) and SPECT least (82%). A study at NIH explains the differing sensitivities: using PET to measure both blood flow and metabolism revealed discrepant findings in the same patients. Preliminary evidence also indicates that the distribution of hyperperfusion on ictal SPECT can differentiate subtypes of TLE. Combining the results of refined imaging techniques holds great promise in epilepsy localization and diagnosis.     

Clinical correlations: MRI and EEG

Main structural correlates of epileptogenesis include hippocampal sclerosis, cortical dysgenesis, foreign tissue lesions, gliosis, and dual pathology (a combination of any two). These structural abnormalities are now increasingly defined with MRI, enabling systematic EEG correlative analyses. Hippocampal atrophy (HA) and increased T₂ signal in medial temporal structures predict the presence of mesial temporal sclerosis with a high degree of sensitivity and specificity. In 50 patients with clinical evidence of temporal lobe epilepsy and isolated HA, ictal scalp EEG was concordant to the atrophic temporal lobe in 33, nonlateralizing in 12, obscured in 3, and bilateral in 2, but it was discordant in none. Earlier reports of higher levels of discordance may be ascribed to the presence of dual pathology or to differing MRI and EEG criteria for localization. In a more inclusive group of 101 patients with unilateral HA, ictal scalp EEG was obtained in 99. It was unlocalized in 53, localized elsewhere in 9, and localized to the atrophic temporal lobe in 38. Of those, 51 patients had intracranial EEG: 12 were unlocalized, 29 were localized to the atrophic hippocampus, and 9 were localized elsewhere. There is thus a rare but definite subgroup of patients with unilateral HA who have EEG localization elsewhere than the atrophy. The successful cure of seizures in half these patients after removal of the EEG focus confirms the importance of this observation and emphasizes the search for more dual pathology that has remained undetected on MRI. About 10% of the patients with HA have significant atrophy bilaterally, and several series have confirmed that surgical success is predicted by removal of the EEG identified seizure onset area, not the more or less atrophic hippocampus. In patients with other kinds of dual pathology, including HA and foreign tissue lesions or cortical dysgenesis, EEG is also paramount in predicting the site of epileptogenesis for surgical intervention. EEG correlates of cortical dysgenesis are heterogeneous, but EEG has potential to provide accurate localization of the site of epileptogenesis in foreign tissue lesions also. In a study of 59 lesional patients, a small number of patients with low grade astrocytomas and oligodendrogliomas consistently localized by EEG to an area elsewhere than the lesion, and failed seizure control when the lesion was removed. Although MRI can demonstrate the structural correlate of the epilepsy in many situations, rare patients, particularly with certain tumors, cortical dysgenesis, and dual pathology, require EEG for accurate localization.     

MRI hippocampal volume and neuropsychology in epilepsy surgery

A review is provided of recent findings on relationships between neurocognitive test data and magnetic resonance imaging (MRI)-determined hippocampal volumes in nonlesional temporal lobectomy patients. The difference between the right and left hippocampal volumes is correlated with postoperative verbal memory in left temporal lobectomy patients who do not have lesional pathology. MRI hippocampal volume data are not associated with measures of executive functioning or naming. Sex differences have been found for verbal memory outcome as women have better verbal memory following left temporal lobectomy. Sex differences have also been found in the relationships between verbal and visual memory, and hippocampal volume data. The systematic combination of MRI-acquired morphological data and neuropsychological test data may further our understanding of neurocognitive function, and provide clinically useful data for counseling epilepsy surgery patients. The current data are promising with regard to prediction of memory outcome following temporal lobectomy, but they do not yet allow for prediction of specific individual patient outcomes. Rather, the currently available data support counseling patients based on the memory outcome of others with similar characteristics.     

Proton magnetic resonance spectroscopic images and MRI volumetric studies for lateralization of temporal lobe epilepsy

We obtained 2D magnetic resonance (MR) spectroscopic images (MRSI) and MRI volumetric measurements (MRIV) of amygdala and hippocampus in 30 consecutive patients with temporal lobe epilepsy (TLE) being evaluated for surgical treatment. Both MRSI and MRIV lateralization showed good agreement with the current gold standard of clinical-EEG lateralization. Each exam separately correctly lateralized 25 out of 30 patients with no false lateralization. Combining both exams, lateralization could be achieved in 28 out of 30 patients. The two patients with no significant asymmetry had bitemporal EEG abnormalities, and bilateral damage on both MRIV and MRSI. There was a good correlation between the magnitude of the MRSI and MRIV asymmetry (Pearson COEFFICIENT = 0.83; p < .0001). Both MRSI and MRIV were normal in our patients with seizures originating outside the temporal lobes. Both MRIV and MRSI can lateralize TLE in 83% of patients. Combination of the two modalities allows lateralization in 93% of patients. Patients who cannot be lateralized generally have symmetrical bitemporal abnormalities; they are not incorrectly lateralized. The structural and chemical pathologic abnormalities seen in TLE seem to be associated with the seizure focus, and may be as, or even more, reliable than a few recorded seizures in predicting the side from which most seizures originate.     

Application of high field spectroscopic imaging in the evaluation of temporal lobe epilepsy

Previous spectroscopic imaging studies of temporal lobe epilepsy have used comparisons of metabolite content or ratios to lateralize the seizure focus. Although highly successful, these studies have shown significant variations within each of the groups of healthy subjects and patients. This variation may arise from the natural differences seen in metabolite concentration in gray and white matter, the complex anatomy seen about the hippocampus, and the large voxels typically employed at 1.5 T. Using a 4.1 T whole body system, we have acquired spectroscopic images with 0.5 cc nominal voxels (1 cc after filtering) to evaluate the regional variation in metabolite content of the hippocampus, temporal gray and white matter, midbrain, and cerebellar vermis. Using a threshold value of 0.90 for CR/NAA, a value 90% of all normal hippocampal voxels lay below, we have correctly identified the presence of epileptogenic tissue in patients with unilateral as well as bilateral seizures. By using comparisons to healthy values of the CR/NAA ratio, this method enables the visualization of bilateral disease and provides information on the extent of gray matter involvement.     

Rapid stereological quantitation of temporal neocortex in TLE

To determine the extent of neocortical atrophy in the temporal lobe using rapid stereological analysis of magnetic resonance slices in patients with temporal lobe epilepsy and to compare the findings to those obtained by visual analysis of high-resolution magnetic resonance images. 25 patients with temporal lobe epilepsy, along with 25 age-matched controls were scanned using a 1.5 Tesla magnetic resonance imaging machine (GE signa systems Paris). Visual analysis was performed on standard high-resolution images. Volumetric analysis of hippocampus and temporal neocortex was performed using computer-aided stereology (MEASURE program, Patrick Barta, Johns Hopkins, Baltimore, USA). Stereological volumetric analysis demonstrated isolated hippocampal atrophy in only nine (36%) cases including three (12%) with bilateral disease. However, eight (32%) cases had combined hippocampal and neocortical atrophy and three (12%) had isolated neocortical atrophy. All volumetric measurements took less than 10 min. On the other hand, visual analysis suggested that 17 (68%) had hippocampal atrophy alone with only two (8%) having combined neocortical atrophy and a further two (8%) having isolated neocortical atrophy. Nearly half of the patients had temporal neocortical atrophy with or without hippocampal atrophy. This rapid, accurate and non-biased quantitative technique has wide clinical utility and is significantly more valuable in detecting neocortical atrophy than visual analysis alone. The results support the notion that abnormalities may be overlooked by current standards of routine magnetic resonance imaging.     

Existence of contralateral abnormalities revealed by texture analysis in unilateral intractable hippocampal epilepsy

We selected 23 patients with unilateral temporal lobe epilepsy characterized by ipsilateral hippocampal sclerosis and an apparently normal contralateral hippocampus on MR imaging. Images were acquired on a 0.28 T MR scanner using a conventional Carr-Purcell Meiboom Gill sequence in all patients and in 9 healthy subjects. Texture analysis was applied to axial MR images of the first and tenth echoes. Texture analysis detects macroscopic lesions and microscopic abnormalities that can not be observed visually. The presence of texture differences in the between normal (controls) and sclerotic hippocampi was ascertained by statistical discriminant analysis. The apparently normal contralateral hippocampi can be classified into three categories in terms of texture: 4 apparently healthy, 8 similar to sclerosis, and 11 different from either healthy or sclerosis. These findings are related to a certain degree of hippocampal alteration, which further investigation might help better characterize.