Abstract
Objective: To report neuroimaging findings in patients with complex partial status epilepticus.
Background: During status epilepticus, neuroimaging may be used to exclude other neurologic conditions. Therefore, it is important to identify the neuroimaging features that are associated with status epilepticus. In addition, MRI characteristics may provide insight into the pathophysiologic changes during status epilepticus.
Methods: The history and neuroimaging examination results of three patients with complex partial status epilepticus were reviewed. Studies obtained during status epilepticus included diffusion-weighted MRI (DWI), MR angiography (MRA), postcontrast T1-weighted MRI, T2-weighted MRI, and CT. Follow-up MRI was obtained in two patients, and autopsy results were available for the third.
Results: Some of the MRI and CT findings during partial status epilepticus mimicked those of acute ischemic stroke: DWI and T2-weighted MRI showed cortical hyperintensity with a corresponding low apparent diffusion coefficient, and CT showed an area of decreased attenuation with effacement of sulci and loss of gray-white differentiation. However, the lesions did not respect vascular territories, there was increased signal of the ipsilateral middle cerebral artery on MRA, and leptomeningeal enhancement appeared on postcontrast MRI. On follow-up imaging, the abnormalities had resolved, but some cerebral atrophy was present.
Conclusions: The radiologic characteristics of status epilepticus resemble those of ischemic stroke but can be differentiated based on lesion location and findings on MRA and postcontrast MRI. The MRI abnormalities indicated the presence of cytotoxic and vasogenic edema, hyperperfusion of the epileptic region, and alteration of the leptomeningeal blood–brain barrier. These changes reversed, but they resulted in some regional brain atrophy.
Diffusion-weighted MRI (DWI) derives contrast from regional differences in water diffusion. It has recently proved to be useful for the evaluation of patients with acute cerebral ischemia.1-3 In the acute stage of cerebral ischemia, depletion of metabolic substrates leads to failure of the Na+/K+-ATPase pump and subsequently to cytotoxic edema, thereby creating an environment that hinders water diffusion.4-6 In ischemic brain tissue, DWI intensity is high, and the apparent diffusion coefficient (ADC), a measure for freedom of water diffusion, is low.
In animal models, DWI abnormalities similar to those seen in acute cerebral ischemia can be observed during status epilepticus.7-10 Additionally, Wieshmann et al. recently described a patient with status epilepticus and associated abnormalities on DWI.11 Other neuroimaging abnormalities during and following status epilepticus have been reported, including hypointensity on CT,12 T2*-weighted hyperintensity on MRI,13-16 and cerebral activation using functional MRI.17 Local hyperperfusion during epileptic activity was observed by Penfield during a neurosurgical procedure in 193318 and was later demonstrated with angiography19 and SPECT.17
We report three cases of complex partial status epilepticus and describe the results of an integrated neuroimaging examination, including DWI, pre- and postcontrast MRI, and MR angiography (MRA), during status epilepticus and at follow-up. Status epilepticus was defined as a prolonged series of seizures without complete resumption of consciousness between them.
Methods.
Patients.
Patient 1. A 23-year-old woman with a long history of complex partial seizures, experienced complex partial status epilepticus and a right hemiparesis during the seventh month of her first pregnancy. A CT scan was interpreted as being consistent with stroke. After delivery, her hemiparesis resolved, and follow-up CT was unremarkable. Three years later, 7 months into her second pregnancy, she was again admitted to the hospital with a right hemiparesis and with transient episodes of confusion. Continuous video-EEG monitoring showed her to be in complex partial status epilepticus with a seizure frequency of up to 50 episodes per day. The seizure focus was noted in the left posterior temporal occipital region, and the interictal EEG demonstrated mild diffuse background slowing and additional focal slowing over the left posterior temporal cortex. Her antiepileptic therapy before admission was phenobarbital 120 milligrams twice daily. Phenytoin and repeated doses of benzodiazepines were added but failed to improve seizure control, nor did IV magnesium reduce seizure activity. Following intubation for treatment of respiratory insufficiency, IV magnesium and sedative medications were discontinued. With the addition of gabapentin to the patient’s previous antiepileptic regimen, seizure activity was controlled after 16 days. The patient regained strength in her right side and was discharged in stable condition. She delivered a healthy baby boy at term. In the following year, her seizures remained controlled, and she had no further episodes of status epilepticus.
Patient 2.
An 82-year-old woman with a history of multiple falls, was admitted because of altered mental status following an unwitnessed fall. On hospital day 2, she became somnolent and developed very fine repetitive nystagmoid eye movements. A noncontrast CT scan of the brain was initially interpreted as showing an abnormality consistent with early left hemisphere ischemia. The patient subsequently developed nuchal rigidity and a temperature of 38.5 °C. A lumbar puncture showed 36 leukocytes per μL, 80 mg/dL glucose, and 93 mg/dL protein. Spinal fluid cultures were negative. EEG demonstrated focal status epilepticus involving the right hemisphere. Background activity on the right waxed and waned between rhythmic fast theta and slow alpha activity. On the left, background activity consisted of low- to moderate-amplitude delta waves with admixed theta frequencies. Lorazepam, phenytoin, and fosphenytoin were unsuccessful in halting her EEG seizures. After 1 day, the patient was intubated and given phenobarbital and thiopental. Two days later, she developed refractory hypotension and could not be resuscitated. At autopsy, a small subdural hematoma in the left frontal lobe and small subarachnoid hemorrhages in the left frontal, the right frontal, and the right occipital cortex were seen. Two small intraparenchymal hemorrhages were seen in the left temporal cortex. The white matter was unremarkable, and there were no changes suggestive of ischemic injury, inflammation, or infection.
Patient 3.
A 4-year-old boy with a 10-day history of upper respiratory infection with fever, was treated with amoxicillin. He was hospitalized for 2 days in a local hospital because of a partial seizure involving the left side of the body. A complete blood count showed a leukocyte count of 15 × 106/mL with 16% bands. The CSF contained 5 leukocytes per μL, of which 60% were polymorphonuclear leukocytes and 40% lymphocytes; protein 31 mg/dL; and glucose 65 mg/dL. Blood and CSF cultures were both sterile. Two days after discharge, he had a recurrent seizure and was given 5 milligrams of diazepam IV by the paramedics. On arrival at the emergency room, he experienced respiratory distress, was given flumazenil, and was intubated. Physical examination revealed a temperature of 38.1 °C, bilateral otitis media, and continuous seizure activity with bilateral hand twitching and nystagmoid eye movements. A CT scan was interpreted as being consistent with an acute infarction in the left middle cerebral artery distribution. Treatment with cefotaxime and acyclovir was begun, and repeated doses of lorazepam, fosphenytoin, and phenobarbital were given. Despite treatment, the patient continued to have seizures and was unresponsive. An EEG showed marked generalized slowing and rhythmic left hemispheric spike and slow wave discharges, indicative of continued partial status epilepticus. On the second hospital day, a coma was induced with pentobarbital, halting the EEG seizure activity. As the sedatives were being reduced the patient developed choreoathetoid movements, ataxia, a right hemiparesis, and moderate aphasia. He was discharged on day 18, and at 2 months follow-up he had some mild residual aphasia and hemiparesis. He continued to have seizures every 7 to 10 days, for which he was treated with phenobarbital and carbamazepine.
Data acquisition and analysis.
Noncontrast CT of the head was performed within a 1-week window preceding the onset of status epilepticus and during status epilepticus in all three patients. MRI of the brain was performed during status epilepticus, ∼24 hours after the onset of seizures, in all patients. Follow-up MRI was obtained at 1 year in Patient 1 and at 2 months in Patient 3. DWI images were obtained with a 1.5-Tesla GE (Milwaukee, WI) Signa Horizon EchoSpeed MRI scanner with echoplanar imaging gradients. An isotropic diffusion-weighted interleaved echoplanar sequence20,21 was used for the initial study in Patient 1. All other MRIs were performed with a single-shot spin-echo echoplanar imaging DWI sequence (b values of 0 and 849 s/mm2) yielding DWI trace images and ADC maps.22 All scans were reviewed by a neuroradiologist (A.M.N.). On each image, ADC values and DWI intensities of representative samples of cortex and subcortical white matter were obtained for each hemisphere. Differences between the affected and the contralateral hemisphere were determined with paired t-test analysis using the statistical software program SigmaStat 2.03 (San Rafael, CA).
Results.
During status epilepticus, DWI demonstrated marked gyriform cortical hyperintensity throughout the affected hemisphere in all cases. This hyperintensity did not respect vascular distributions. The corresponding areas on the ADC maps were hypointense (figure 1). The mean relative increase in diffusion-weighted intensity was 94% (p < 0.001) and the relative mean decrease in the ABC was 36% (p < 0.001), corresponding to changes observed with acute cerebral ischemia.6 In Patients 1 and 3, very small areas in the medial temporal white matter showed DWI hyperintensities and low ADCs on visual inspection. Except for these areas, the mean white matter DWI intensity of the affected hemisphere did not differ significantly from the contralateral side, and the ADC value was only marginally reduced. Absolute ADC values for cortex and white matter are summarized in table 1. An additional area of DWI hyperintensity was observed in the dorsal posterolateral portion of the ipsilateral thalamus in Patients 1 and 2 (see figure 1). The corresponding area on the ADC map was slightly hypointense in Patient 2 and too small to be analyzed in Patient 1. In Patient 1 only, subtle diffusion-weighted hyperintensity and low ADC were seen in a portion of the contralateral cerebellum.
Figure 1. Diffusion-weighted MRI (top row) and apparent diffusion coefficient maps (bottom row) from Patients 1, 2, and 3 during status epilepticus show cortical hyperintensity on diffusion-weighted MRI (arrowheads) and corresponding areas of low apparent diffusion coefficient (arrowheads). Diffusion-weighted MRI in Patient 2 shows hyperintensity of the dorsal posterolateral portion of the right thalamus (long arrow).
Apparent diffusion coefficient (ADC) values of the cortex and white matter during complex partial status epilepticus and at follow-up
Corresponding T2*-weighted images and fluid-attenuated inversion recovery (FLAIR) images showed cortical hyperintensity in the abnormal regions seen on DWI. The subcortical white matter was unremarkable. T1*-weighted postcontrast images, obtained in Patients 2 and 3, showed leptomeningeal enhancement. In all three cases, the MRA of the circle of Willis demonstrated marked increased signal in the middle and posterior cerebral arteries of the epileptic hemisphere. The distal (M2 and M3) branches of the ipsilateral middle cerebral arteries were strikingly prominent compared with those in the nonepileptic hemisphere. Figure 2A. shows a representative slice of the T2*-FLAIR, the postcontrast T1*-weighted MRI, and the MRA of Patient 3 during status epilepticus.
Figure 2. Fluid-attenuated inversion recovery (FLAIR), postcontrast MRI, and MR angiography images of Patient 3 during status epilepticus (A) and at 2-month follow-up (B). Cortical hyperintensity on FLAIR (arrow), leptomeningeal enhancement on postcontrast MRI (arrowheads), and marked asymmetry of the middle cerebral artery branches on MR angiography (large arrowhead) are seen during status epilepticus. At follow-up these abnormalities have reversed except for a small residual region of cortical hyperintensity on FLAIR.
CT scan before the onset of status epilepticus was unremarkable in Patient 1, showed hemorrhagic abnormalities consistent with the previously described trauma in Patient 2, and demonstrated a 5-millimeter low-density abnormality in the left frontal lobe in Patient 3. The CT scans obtained during status epilepticus demonstrated new regions of hypoattenuation in the affected hemisphere, including subtle loss of gray-white differentiation and mild sulcal effacement in all cases (figure 3)
Figure 3. CT of Patient 2 during status epilepticus reveals a large area of decreased attenuation with effacement of sulci and loss of gray-white differentiation (arrowheads).
Figure 4 shows the 1-year follow-up DWI for Patient 1 and the 2-month follow-up DWI for Patient 3. The cortical hyperintensity previously seen on DWI had resolved in both cases. The ADC map had normalized in Patient 1, and demonstrated two small hyperintense lesions consistent with chronic abnormalities in the affected hemisphere in Patient 3 (not shown). Statistical analysis no longer showed significant differences between DWI intensities of the affected and the contralateral hemisphere and only showed a marginally elevated ADC for the white matter and the gray matter on the involved side (see table 1). On follow-up scans, T2* and FLAIR images showed enlargement of the lateral ventricle and sulci in the epileptic hemisphere indicative of regional interim volume loss. Otherwise, the T2* was unremarkable in Patient 1 and showed some residual cortical hyperintensity in Patient 3. The MRA of the circle of Willis appeared normal in both cases and no longer showed signal asymmetry of the cerebral vessels. Postcontrast MRI obtained in Patient 3 were unremarkable. Figure 2B shows a representative image of the T2*-FLAIR, the postcontrast MRI, and the MRA of Patient 3 at 2 months follow-up.
Figure 4. Diffusion-weighted MRI (top) and apparent diffusion coefficient (bottom) images from Patients 1 and 3, obtained at 1-year and 2-month follow-up, respectively, demonstrate reversal of the abnormalities seen during status epilepticus (compare with figure 1).
Discussion.
Neuroimaging studies are often obtained to evaluate patients who are in status epilepticus for ischemic, hemorrhagic, infectious, inflammatory, or neoplastic processes. Because there is overlap in the neuroradiologic features of these conditions, it is important to recognize the characteristics that help differentiate status epilepticus from other disorders. We have described three patients with complex partial status epilepticus, in whom both routine and novel MRI techniques were used. In two patients, follow-up scans allowed us to assess the reversibility of these abnormalities.
Our observation of cortical DWI hyperintensity and a 36% mean decrease in ADC is consistent with previously discussed animal studies and one case report. In rats, ADC decreases from 14% to 49% have been observed during kainate- or flurothyl-induced status epilepticus.7-10 Wieshmann et al. described a patient who had a 27% reduction in ADC of the gray matter during status epilepticus.11 These cortical changes, as well as the mild mass effect that can be observed on CT, are presumably caused by cytotoxic edema occurring during status epilepticus. Swelling of dendrites and astrocytes has been described histologically in a few experimental studies.23-26 Recently, Wang et al. reported an increase in sodium concentration in the pyriform cortex of rats during status epilepticus. They proposed that this may be due to energy failure of the Na+/K+-ATPase pump, which in turn may lead to Na+ and water influx.10 Other studies, however, that have found increased oxygen pressure in the draining vessels of the epileptic region17 and no change in ATP levels27 do not support this hypothesis. Excessive release of excitatory amino acids, such as glutamate,28,29 and increased membrane ion permeability30 are other proposed mechanisms that could cause cytotoxic edema during status epilepticus.
The dorsal posterolateral portion of the ipsilateral thalamus (two patients) and the contralateral cerebellum (one patient) also demonstrated DWI hyperintensity during status epilepticus. Hyperintensity of these remote regions on T2*-weighted images has been described in one previous case report.31 It is possible that these MR abnormalities reflect excessive cortical stimulation of the ipsilateral thalamus and the contralateral cerebellum.
Unlike the patient described by Wieshmann et al., who demonstrated a 31% increase in ADC of the white matter underlying the affected cortex during status epilepticus,11 we did not observe an increase in ADC of the subcortical white matter in our patients. In contrast, our patients demonstrated a marginal reduction of the ADC in the white matter, possibly reflecting some mild cytotoxic edema.
During status epilepticus, increased blood flow through cerebral vessels on the affected hemisphere has been reported, but to our knowledge, this is the first report to describe this phenomenon with MRA, suggesting that MRA not only is sensitive to cerebral artery stenosis or occlusion, but also may be used to detect regionally enhanced flow. On follow-up MRA, the cerebral vessels were symmetrical in both patients. Hyperperfusion was therefore a temporary phenomenon, possibly in response to the elevated metabolic demands of status epilepticus.
Leptomeningeal enhancement on postcontrast MRI has, to our knowledge, not been previously described in focal status epilepticus. This enhancement may reflect alteration of the blood–brain barrier, which has been demonstrated in a rat model of status epilepticus.32 Alteration of the blood–brain barrier could be related to vasogenic edema, which was manifested on the T2*-weighted images as cortical hyperintensity in our patients as well as in several previous reports.13-16 Although we cannot entirely eliminate the possibility that the meningeal enhancement was due to concomitant CNS infections, arguing against this was the lack of clinical or laboratory signs of infection or meningeal irritation in Patient 1. Patient 3 was febrile and was diagnosed with bilateral otitis media; however, the results of CSF analysis were unremarkable just before the onset of status epilepticus.
Leptomeningeal enhancement, T2* hyperintensity, and asymmetry of the cerebral arteries on MRA were no longer present on follow-up MRI examination, which suggests that the vascular changes and vasogenic edema associated with status epilepticus are reversible. Resolution of the abnormalities seen on DWI and the ADC map indicates that the cytotoxic edema is also reversible. In rat models of status epilepticus, this has been shown histologically33 and neuroradiologically.7-10 The recovery of tissue with a low ADC and hyperintensity on T2* is of interest because it has been proposed in human stroke research that this combination is a marker for tissue that is no longer viable.34 Although we cannot exclude that this may be the case for ischemic brain injury, our findings demonstrate that it is not necessarily a general principle. Despite the reversal of most neuroimaging abnormalities, some permanent brain injury was demonstrated in our patients. This manifested as enlargement of the lateral ventricle and sulci on the affected side, suggesting some interim volume loss. In one patient, small regions of high ADC and high signal on the T2*-weighted images likely reflect focal neuronal and glial cell death. This is compatible with several animal studies that have histologically demonstrated neuronal cell death following status epilepticus23,35-37 and with two previously described cases of radiologically confirmed brain atrophy following status epilepticus.16
Differentiation of status epilepticus from other neurologic disorders remains a challenge, but it is possible to identify several neuroimaging features that aid in identifying status epilepticus (table 2). These findings include hyperintensity on DWI and T2* with a corresponding low ADC in a region that is confined to the cortex and does not respect vascular territories, increased flow in the ipsilateral cerebral arteries on MRA, enhancement of the leptomeninges on postcontrast MRI, and a large region of low attenuation on CT without major mass effect. These abnormalities reflect the many physiologic changes that can occur during status epilepticus, including vasogenic and cytotoxic edema, hyperperfusion of the epileptic region, and alteration of the blood–brain barrier. Reversibility of these changes was demonstrated with follow-up MRI. The fact that regions of DWI and T2* hyperintensity with low ADC had normalized over much of the previously affected hemisphere indicates that these changes do not predict extensive tissue necrosis, as has been proposed when they are caused by ischemia. However, follow-up MRI in our patients did demonstrate subtle signs of atrophy, indicating that some diffuse brain injury may occur following status epilepticus.
Key findings on MRI and CT during complex partial status epilepticus
Footnotes
-
Dr. Lansberg was supported by scholarships from the Dutch Heart Association and the Dutch Brain Association.
- Received September 16, 1998.
- Accepted December 12, 1998.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Alert Me
Recommended articles
Applications of diffusion–perfusion magnetic resonance imaging in acute ischemic stroke
Multislice diffusion mapping for 3-D evolution of cerebral ischemia in a rat stroke model
Posterior leukoencephalopathy without severe hypertension
Acute ischemic lesions of varying ages predict risk of ischemic events in stroke/TIA patients