MRI findings in aphasic status epilepticus


Address correspondence to Manuel Toledo, M.D., Ph.D., Neurology Department, Hospital Universitari Vall d'Hebron, Pssg. Vall d'Hebron, 119-129 Neurology Service, 08035 Barcelona, Spain. E-mail:


Ictal-MRI studies including diffusion-weighted imaging (DWI), perfusion-weighted imaging (PWI), and MR-angiography (MRA) in patients with aphasic status epilepticus (ASE) are lacking. In this report, we aim to describe the consequences of the ASE on DWIs and its impact on cerebral circulation. We retrospectively studied eight patients with ASE confirmed by ictal-EEG, who underwent ictal-MRI shortly after well-documented onset (mean time delay 3 h). ASE consisted in fluctuating aphasia, mostly associated with other subtle contralateral neurological signs such as hemiparesia, hemianopia, or slight clonic jerks. In MRI, six patients showed cortical temporoparietal hyperintensity in DWI and four of them had also ipsilateral pulvinar lesions. Five patients showed close spatial hyperperfusion areas matching the DWI lesions and an enhanced blow flow in the middle cerebral artery. Parenchymal lesions and hemodynamic abnormalities were not associated with seizure duration or severity in any case. The resolution of DWI lesions at follow-up MRI depended on the length of the MRIs interval. In patients with ASE, lesions on DWI in the temporo-parietal cortex and pulvinar nucleus combined with local hyperperfusion can be observed, even when they appear distant from the epileptic focus or the language areas.

 Aphasic status epilepticus (ASE) is a rare pathology, not yet studied by ictal-MRI (Grimes & Guberman, 1997). Likewise, diffusion-weighted imaging (DWI), combined with perfusion-weighted imaging (PWI) and MR-angiography (MRA), is a current focus of interest in the status epilepticus. Ictal-MRI patterns in ASE may add data for a better understanding of the physiopathology of partial status epilepticus (PSE) (Cole, 2004).

In this study, we aim to describe the ictal-MRI findings of patients with ASE of recent onset.

Material and Methods

We retrospectively studied the patients who underwent ictal-MRI during the ASE from May 2002 to October 2006.

ASE was defined according to Rosenbaum's criteria modified by Grimes & Guberman (language production during seizures; language shows dysphasic features; consciousness preservation; seizures correlated with aphasia, as documented by EEG monitoring and behavioral testing; and aphasia improves concurrent with treatment of the seizures) (Grimes & Guberman, 1997).

EEGs were performed with a Nicolet 20-channel polygraph. Electrodes were placed according to the International 10–20 System. Band pass 0.3–70 Hz mean time delay from ictal MRI to EEG was 0.75 (±0.5) h. All patients received an ictal-EEG during the ASE and a follow-up EEG when they recovered.

MRI was performed with a 1.5-T MRI system (Siemens, Erlangen, Germany). The ictal-MRI strategy of this study included transverse DWI, PWI, and MRA sequences.

DWI was obtained with a single-shot spin-echo echoplanar pulse sequence (b-values: 0, 500, and 1,000 s/mm2; 15 axial sections; 5-mm-thick sections; interslice gap 1.5 mm). We calculated the attenuance diffusion coefficient (ADC) maps from all DWI.

PWI was acquired using the dynamic first-pass of gadolinium-based contrast material (Magnevist; Schering AG, Berlin, Germany) (15 sections, 5-mm-thick sections, 1.5-mm interslice gap). The PW sequence generated a time-to-peak (TTP) and cerebral blood volume (CBV) map. Hyperperfusion was defined by an increase in the CBV and acceleration in TTP maps in comparison with the normal hemisphere.

Tissue abnormality was considered in areas of high signal intensity on DWI, T2, CBV, and TTP maps.

For MRA, we used a transverse gradient-echo 3D time-of-flight sequence with magnetization transfer suppression and tilted optimized nonsaturating excitation (0.83-mm-thick sections). Cerebral artery enlargement was defined as an increase of the artery diameter >1 mm on ictal-MRA compared to the contralateral and/or to the same artery at follow-up examination.

Follow-up MRIs were performed within 1 week after the ASE recovered. This examination included DWI, MRA, and an additional transverse T2-weighted fast spin-echo.

We used SPSS 12.0 for Windows (SPSS Inc., Chicago, IL, U.S.A.) to calculate averages and standard deviations. Group comparisons between patients and controls could not be performed because of small sample size and diversity of the patients. Mann-Whitney U-test was applied for comparisons of the timing from the ASE onset to ictal-MRI.


Clinical features

We recruited eight patients with ASE during the period of study (mean age: 72 (±12) years old; 50% female). The average duration of the ASE was 11 (±6) h (Table 1).

Table 1.  Demographics, seizures characteristics and EEG findings of patients
PatientAgeGenderDominanceSeizure semiologyEtiologyStatus duration (hours)Ictal-EEG
  1. SW, sharp waves.

167MaleRight-handedMixed aphasia and hemianopiaNon ketotic hyperglycemia12Left temporal and parietal paroxysmal fast rhythms and PLEDs
263MaleRight-handedMixed aphasia, right sensitive seizures, and hemiparesisLeft insular glioma 7Slow waves in left hemisphere with irregular continuous SW in fronto-temporal region
374MaleRight-handedBroca's aphasia, right hemiparesia, and right arm clonic jerksLeft frontal meningioma 3Left frontotemporal SW paroxysm
476FemaleRight-handedMixed aphasiaNon ketotic hyperglycemia12Left frontotemporal continuous deltha and burst of SW paroxysm
551MaleRight-handedBroca's aphasia and right arm clonic jerksLeft frontal empyema13Left hemispheric slow waves with continuous spikes and SW in central-parietal area
677FemaleRight-handedMixed aphasia and left hemianopiaLeft temporal glioblastoma10Left posterior temporal paroxysmal fast with continuous spikes and SW
787FemaleLeft-handedMixed aphasia and left hemiparesiaCryptogenetic24Right temporoparietal PLEDs and SW
885FemaleRight-handedBroca's aphasia, right hemiparesia, and right facial clonic jerksLeft frontal microbleeding, amyloidal angiopathy 7Left fronto polar SW and periodic paroxysmal fast

The ASE was the first seizure in all patients. The most frequent etiology was tumor; stroke was not considered in any case, due to the chronic characteristics of the arterial stenoses and the absence of ischemic parenchymal lesions (Table 1).

The main symptom of all patients was an abrupt onset of dysphasia. The ASE mostly consisted of prolonged spaced temporal seizures of hypofluent mixed aphasia or speech arrest during approximately 20 min. Interictal periods were shorter, approximately 5 min, and they were characterized by normal language production, or phonemic paraphasic utterances in some patients (Table 1).

All but one patient, showed clinical signs other than the aphasia; symptoms were subtle and they could be revealed only after an accurate physical examination. Negative signs such as contralateral hemianopia and slight distal brachio-crural hemiparesis were the most frequent features observed, and they persisted even during the interictal periods. Other signs, less frequently found, were isolated contralateral limb clonic jerks or sensitive seizures (Table 1).

To avoid misdiagnoses with complex PSE, we confirmed the consciousness and memory preservation by monitoring patients during the ASE and after resolution.

Language and associated signs recovered gradually over several hours after clonacepam or phenytoin perfusion.

MRI findings

The mean timing from the ASE onset to the ictal-MRI was 3 (±1.7) h. DWI revealed ipsilateral cortical hyperintensity in six patients; mostly in the temporo-parietal cortex. Cortical lesions were associated to DWI lesions in the pulvinar region of the ipsilateral thalamus in four patients. The mesial temporal lobe was intact in all patients. Tissue involved by DWI lesions appeared distant from the epileptogenic lesions. ADC values were reduced in the areas of most pronounced signal change on DWI (Table 2).

Table 2.  MRI findings of patients
PatientIctal-MRIFollow-up MRIT2-WI
Timing of MRI (hours)DWI hyperintensityPWIMR-angiographyTiming of MRI (hours)DWI
  1. MCA, middle cerebral artery; N, normal.

12.3Left parietooccipital and thalamusNEnhanced flow in left MCA36Slight thalamic hyperintensityN
20.3Left temporoparietalN112 NormalLeft Insular tumor
32.2NNNNot performedLeft frontal meningioma
43.3Left temporoparieto-occipitalN60NormalN
52 Left temporoparietal and thalamusEnhanced flow in left MCA24Slight cortical and thalamic hyperintensityLeft frontal empyema
64.3Left parietal and thalamusEnhanced flow in right MCANot performedLeft temporal malignant tumor
74.2Right temporo-parietal and thalamusEnhanced flow in left MCA42Slight thalamic hyperintensityN
86 NNN96NormalLeft frontal cortical bleeding and difuse chronic microbleedings

PWI demonstrated close spatial hyperperfusion areas matching both cortical and thalamic DWI lesions in five and four patients, respectively. MRA also showed increased flow signal in the middle cerebral artery in four of those patients. Patients with normal DWI on ictal-MRI had neither PWI nor MRA changes (Table 2). Local chronic hypoperfusion, not related to the region of interest, was identified in patients 4 and 7, associated to mild cerebral artery stenoses (Fig. 1).

Figure 1.

(A) Ictal MRI and EEG during the ASE in a left-handed woman (patient 7): local hyperperfusion areas (TTP and CBV) matching the DWI lesions, and enhanced blood flow in the right middle cerebral artery. Note a chronic hypoperfusion in the posterior cerebral artery secondary to a stenosis in P1 segment. (B) Follow-up MRI and EEG after ASE recovered, where still persists a slight pulvinar hyperintensity on DWI and MRA normalization.

The presence of DWI lesions or hemodynamic changes were not associated to seizure duration or clinical features.

Six patients were studied subsequently with a follow-up MRI examination; five of them had shown DWI lesions on ictal-MRI. Slight DWI hyperintensity was still noted at follow-up MRIs with shortest length of interval. Ictal-MRA changes were reversible in all cases (Table 2; Fig. 1).


To our knowledge, this study reports the first series of patients with ASE studied with combined use of DWI, PWI, and MRA. We found particularly remarkable the presence of lesions on DWI and hemodynamic changes involving the temporoparietal cortex, shortly after the ASE onset and usually distant from the epileptic focus or language areas.

Aphasia as a sole manifestation of the PSE is a rare occurrence and it is usually associated to other subtle symptoms (DeToledo et al., 2000). Most of our patients showed clinical signs other than the aphasia, although they did not spontaneously realize. Negative signs, motor and visual, could be best explained either by the spreading of the epileptic activity to negative motor areas, or by interictal paralysis (Todd's phenomenon). The occurrence of clonic jerks was probably a consequence of seizure spread (DeToledo et al., 2000; Chung et al., 2002).

DWI hyperintense lesions and local hyperperfusion after prolonged seizure activity have been described in PSE (Cole, 2004). This phenomenon seems to be a consequence of neuronal increased energy demands in the epileptogenic area and in the tissue overactivated by seizure spread, which are not adequately matched by the compensatory enhanced blood flow. As a result, it leads to an anaerobic glycolysis in neurons, Na+/K+ TPA-ase transmembrane pumps failure, and thereby to cytotoxic edema (Szabo et al., 2005). Prolonged epileptic activity has been suggested as the critical factor responsible for the changes usually detected by PWI and DWI; however, our results demonstrated that DWI lesions and hemodynamic changes can be observed shortly after ASE onset, and not correlated to the severity or the cause (Cole, 2004; Szabo et al., 2005).

The normalization of the DWI and MRA at follow-up MRI could be explain by the short ASE duration in our patients, indicating that cell death is an evitable sequela in PSE. The absence of abnormalities on the ictal-MRI of two patients could depend on the individual cerebral susceptibility to bear neuronal damage induced by seizures (Lansberg et al., 1999; Engelhorn et al., 2007).

This study demonstrated DWI lesions and hyperperfusion areas in the posterior cerebral cortex in ASE, similar to those commonly described in prolonged PSE. Likewise, signal changes restricted to the pulvinar or the hippocampus have been proposed as a diagnostic clue for prolonged complex PSE (Hufnagel et al., 2003). However, our patients showed frequent pulvinar lesions, preserving the consciousness. The hippocampus was intact in all our patients, probably due to the short seizure duration or seizure semiology (Szabo et al., 2005).

Interestingly, ictal-MRI changes may occur in distant cortex from the epileptic focus, but in the ASE they were also observed distant from the primary language areas (Broca's or Wernickes areas) or the basal temporal language area (Kirshner et al., 1995). Our findings may support the idea that the temporoparietal cortex and subsequently the pulvinar region are highly sensitive areas to the epileptic overactivation (Szabo et al., 2005). Therefore, it is likely that DWI, PWI, and MRA are useful to detect tissue involved by seizure spread, but not to locate the epileptic focus (Cole, 2004).

The short population sample and selection criteria of our study may have influence in the absence of stroke as etiology of the ASE, although the remaining clinical characteristics of patients are in the lines of previous ASE reports (Hamilton & Matthews, 1979; Grimes & Guberman, 1997).

In conclusion, ictal-MRIs in patients with ASE of recent onset may frequently show local hyperperfusion areas and an enlargement of the middle cerebral artery, matching DWI hyperintensity in the temporoparietal cortex and posterior thalamus.


Conflict of interest: All authors confirm that they have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. None of the authors of this manuscript has any conflict of interest.