Frontal and thalamic changes of GABA concentration indicate dysfunction of thalamofrontal networks in juvenile myoclonic epilepsy.

OBJECTIVE: Juvenile myoclonic epilepsy (JME) has been considered to be a frontal variant of thalamocortical network dysfunction in epilepsy. Changes of γ-aminobutyric acid (GABA)ergic neurotransmission may play a key role in this dysfunction. Magnetic resonance spectroscopy (MRS) is the only noninvasive method to measure GABA concentrations in different brain regions. We measured GABA and other metabolite concentrations in the thalamus and frontal lobe of patients with JME. METHODS: A specific protocol was used for determining GABA concentrations in the thalamus, frontal lobe, and motor cortex contralateral to the handedness in 15 patients with JME and 15 age-matched controls. In addition, we measured concentrations of glutamate and glutamine, N-acetyl-aspartate (NAA), myoinositol, creatine, and choline using MRS with short echo time. JME-related concentration changes were analyzed comparing patients to controls, also considering potential effects of antiepileptic drugs. RESULTS: In patients with JME, GABA and NAA were reduced in the thalamus (p = 0.03 and p = 0.02), whereas frontal GABA and glutamine were elevated (p = 0.046 and p = 0.03). MRS revealed reduced NAA in the thalamic gray matter contralateral to the handedness (p = 0.04 each). These changes were found consistently in patients treated with new antiepileptic drugs and with valproate, although the extent of metabolic changes differed between these treatments. SIGNIFICANCE: Decreased thalamic and increased frontal GABA suggest a dysfunction of GABAergic neurotransmission in these brain regions of patients with JME. The NAA decrease in the gray matter of the thalamus may hint to a damage of GABAergic neurons, whereas frontal increase of GABA and its precursor glutamine may reflect increased density in GABAergic neurons due to subtle cortical disorganization in the thalamofrontal network.

The striate sign: peritumoural perfusion pattern of infiltrative primary and recurrent gliomas.

MR perfusion depicts angiogenesis as a key factor for growth and malignancy in gliomas by means of increased regional cerebral blood volume (rCBV). The rCBV increase is not limited to the tumour area, but may also produce a stripe-like pattern of peritumoural rCBV increase that we defined as the "striate sign". We evaluated if prior radiochemotherapy influences perfusion values and pattern in and adjacent to malignant gliomas comparing rCBV of treated recurrent gliomas with untreated gliomas. Ninety-three patients with primary or recurrent WHO grades II-IV glial tumours underwent T2*-weighted dynamic susceptibility-weighted contrast-enhanced (DSC)-MRI. Differences of normalised rCBV and rCBV(max) were evaluated using Kruskal-Wallis analysis with post hoc tests. The number of cases showing a hot spot of rCBV (rCBV(max)) and/or a peritumoural striate pattern of rCBV increase (striate sign) was assessed and evaluated by Fisher's exact test. Significance level was determined as p < 0.05. Normalised rCBV, rCBV(max) and number of cases with the striate sign were significantly lower in recurrent (rCBV = 3.24 +/- 1.22, rCBV(max) = 5.05 +/- 2.27 and striate sign = 10/24) compared to primary WHO grade IV tumours (rCBV = 4.44 +/- 1.39, rCBV(max) = 7.31 +/- 3.0 and striate sign = 17/21, respectively). There were fewer cases with a striate sign in treated recurrent WHO grade III tumours than in untreated malignant transformed WHO grade II tumours. The pattern and degree of rCBV increase in and around gliomas differ between untreated and previously treated tumours. These differences might be due to post-therapeutic changes of the tumour-associated microvasculature by radiochemotherapy. Spectroscopic and susceptibility-weighted MR imaging may provide further insights into the tumour biology.

Elevated peritumoural rCBV values as a mean to differentiate metastases from high-grade gliomas.

PURPOSE: Increased relative cerebral blood volume (rCBV) was previously found in peritumoural oedema of glioblastomas (GBM). Supposing that peritumoural rCBV is not increased in metastases, we aimed to evaluate whether rCBV values of the whole peritumoural area are accurate to differentiate solitary metastasis from GBM irrespective of the peritumoural oedema. METHODS: Contrast-enhanced T1-weighted (T1-w) and T2*-weighted dynamic susceptibility contrast MRI was performed in 52 patients with contrast-enhancing solitary brain tumours before surgery. In each T1-w slice depicting the contrast-enhancing tumour, a rim within approximately 15 mm was defined in the peritumoural area. The rCBV values were normalised to rCBV values of the contralateral normal white matter. Differences between metastases and GBM for normalised rCBV values for each slice were determined with the Mann-Whitney U test (p < 0.05). RESULTS: Histopathological examination revealed 29 GBM and 23 metastases. Peritumoural rCBV was significantly lower in metastases than in GBM (p < 0.01). Using the cutoff value 1.0 for discriminating metastases from GBM yielded a sensitivity of 96%, specificity of 64%, a positive predictive value of 68% and a negative predictive value of 95%. CONCLUSIONS: The rCBV in the peritumoural area of contrast-enhancing brain tumours has a high diagnostic accuracy to discriminate metastases from GBM irrespective of surrounding oedema and without the bias of slice selection and ROI positioning. Metastases should be excluded, if at least one tumour-depicting slice reveals an increase of peritumoural rCBV compared to the normal contralateral brain (normalised rCBV value >1). Conversely, the decrease of peritumoural rCBV may not reliably exclude GBM.