HISTORICAL NOTE AND NOMENCLATURE

The original proband was a boy referred to Dr. Wilder Penfield by Dr. Ed Fincher of Emory University in Atlanta because of intractable focal motor seizures. The boy eventually developed a hemiparesis, and serial pneumoencephalograms showed progressive enlargement of the contralateral ventricle.|{picture:rsaf1.bmp}{caption:Photograph of original patient diagnosed with Rasmussen's syndrome}{label:The original boy in whom a diagnosis of chronic encephalitis and epilepsy (or Rasmussen's Syndrome) was made.}||{picture:rsaf2.bmp}{caption:Diagram of the clinical course of Rasmussen's syndrome}{label:Seizures in the original Rasmussen's syndrome patient stopped after hemispherectomy but he died due to complications caused by cerebral hemosiderosis.}|At the time, there was much discussion about whether the progressive nature of the neurologic disorder could be explained by the effect of the seizures alone or whether he had a progressive underlying neurologic disease. The finding of inflammatory changes in resected brain tissue suggested a viral encephalitis and militated towards the diagnosis of a progressive disorder, then termed chronic encephalitis and epilepsy.


CLINICAL MANIFESTATIONS

In the majority of instances, Rasmussen's syndrome develops in childhood, most commonly between the ages of 1 and 10 years; in one series of 48 patients, the median age at onset was 5 years (Oguni et al 1992). There is no major difference in incidence between the sexes. In about 50% of patients, the onset is preceded by an inflammatory episode (eg, an upper respiratory tract infection, otitis media, or tonsillitis) occurring in the previous 6 months. In the majority of children, the first sign of the condition is often the development of generalized tonic-clonic seizures, although simple partial or complex partial seizures occur as the initial seizure type in a significant proportion of the children. Status epilepticus is common and is the presenting feature in about 20% of patients.


Seizures are usually refractory, with little response to standard antiepileptic drugs (Dubeau and Sherwin 1991). It is common for a variety of seizure types to develop over a period of time, with the seizures becoming increasingly severe and frequent. Focal motor seizures occur in about three quarters of patients at some time. Epilepsia partialis continua eventually occurs in about 50% of children, usually within 3 years of onset (Oguni et al 1992).

When partial seizures occur, they almost invariably involve the same side of the body. Todd's paresis is relatively common, and in the early stages of illness, hemiparesis is postictal and transient. Exceptionally, hemiparesis occurring at this stage may be permanent. With progression of the disease, fixed neurologic deficits, including visual field deficits and hemiparesis (usually to the extent of the child losing fine finger movements and developing a hemiparetic gait), gradually ensue after a period varying from 3 months to 10 years from the onset of the epilepsy (Oguni et al 1991).

Progressive intellectual impairment is also a feature of the condition, sometimes occurring or worsening in association with a deterioration in seizure control and sometimes apparently independent of seizure activity. As in the case of focal neurologic deficits, intellectual deterioration commonly occurs over a number of years, though sometimes over months. Speech deficits (including dysphasia and dysarthria) and cortical sensory loss are also features of the disease, depending on which hemisphere is involved.

Despite the fact that the initial course of the illness is generally one of relentless progression, it is relatively unusual for it to cause death. Instead, the disease process eventually appears to burn itself out, usually at a stage at which there is moderate to severe neurologic deficit and intellectual impairment. As further progression of these deficits stops, the seizures often become less frequent and less severe.

Rasmussen's original description of chronic localized encephalitis reported an inflammatory process, with marked perivascular cuffing by round cells in both the cortex and white matter. Diffuse patchy inflammatory changes were seen in the cortex and white matter, with prominent microglia in addition to small round cells and occasional polymorphonuclear cells. There was loss of nerve cells, some spongy degeneration, and hypertrophy of the astrocytes, which were also increased in number.|{picture:rsaf3.bmp}{caption:Pathological findings in tissue from original Rasmussen's syndrome patient }{label:Note the perivascular infiltrates and microglial nodules.}|

This description has been expanded by Robitaille, who reviewed the pathological specimens of 48 patients in the Montreal series (Robitaille 1991). He describes the features of the disease as the presence of microglial nodules, often displaying neuronophagia (mainly in the medium-sized pyramidal cells of the external pyramidal layer), and perivascular cuffs of small lymphocytes and monocytes. In the more active cases, he notes that the round cells fill the Virchow-Robin spaces and extend into the neuropil, forming microscopic clusters or larger aggregates. Spongiosis is particularly seen in association with inflammatory changes. He notes the frequent presence of multifocal neuronal loss in the inflamed cortex, especially in the superficial and intermediate cortex. The smaller foci tend to coalesce into large areas of structural collapse surrounded by inflammatory changes and sprouting capillaries, which causes the area to resemble granulation tissue. Subarachnoid adhesions are not infrequently seen.

Robitaille classifies the pathological specimens into four groups, according to the features of disease activity. Group 1 includes those with the most pathologically active disease. The features seen in these specimens are those of an ongoing inflammatory process, with numerous microglial nodules, with or without neuronophagia, perivascular round cells, and glial scarring. Included in Group 2 are those with "active and remote disease", indicated by the presence of several microglial nodules, cuffs of perivascular round cells, and at least one gyral segment of complete necrosis and cavitation including full-thickness cortex. Group 3 has less active "remote" disease, with pathological appearances of neuronal loss and gliosis, moderately abundant perivascular round cells, and only a few microglial nodules. Finally, Group 4 consists of those specimens showing non-specific changes, with few or no microglial nodules and only mild perivascular inflammation but with various degrees of neuronal loss and glial scarring.


ETIOLOGY

Since the original description of Rasmussen's encephalitis was published, the underlying cause has been the source of much speculation (Antel and Rasmussen 1996). The most likely mechanisms have been postulated as a chronic viral infection, an acute viral infection leading to a local immune response, or an independent autoimmune process, not linked to infection.

Both the clinical aspects and the pathological features of the disease are strongly in favour of an underlying infective cause, as evidenced by the close resemblance of the clinical picture to that seen in Russian spring-summer tick-borne encephalitis, described by Kozhevnikov (Kozhevnikov 1991).

Numerous attempts have been made to demonstrate viral particles or genetic evidence of viral material in specimens from patients with Rasmussen's encephalitis. Although there have been a number of positive results, it usually has not been possible to reproduce these, and the situation remains unresolved. Friedman and colleagues describe a 3-year-old child with hemiplegia, hemiconvulsions, and epilepsy in whom biopsy showed an encephalitic picture of perivascular cuffing with mononuclear cells, and electron microscopy of brain cells showed viral crystals resembling those of enteroviruses (Friedman et al 1977). Walter and Renella report two patients with chronic encephalitis and epilepsy in whom biopsy showed histological features of encephalitis, and in situ hybridization showed the Epstein-Barr virus genome in intranuclear central cores within the encephalitic infiltrations (Walter and Renella 1989). This raised the possibility of a role for the Epstein-Barr virus in the pathogenesis of Rasmussen's encephalitis. Power and colleagues carried out in situ hybridization for CMV on brain biopsy specimens from 10 patients with Rasmussen's encephalitis and 46 age-matched control patients with other neurologic diseases (Power et al 1990). They found CMV genomic material in seven of the ten patients with Rasmussen's encephalitis and in only two of the control patients. Probes of herpes simplex virus and hepatitis B were negative in all patients, although this work has been criticized (Gilden and Lipton 1991; Root-Bernstein 1991). >From the same group, McLachlan and colleagues also demonstrated the CMV genome in neurons, glia, and endothelial cells of blood vessels by in situ hybridization in brain specimens from three patients who developed chronic encephalitis and epilepsy in adulthood (McLachlan et al 1993). Similar assessments for hepatitis B, herpes simplex, and Epstein-Barr virus genome were negative in these patients. Viral antibody titres in the serum and CSF samples were normal, and no viral inclusions or antigens were found in resected brain tissue. Jay and colleagues studied pathological specimens from ten patients with chronic encephalitis and intractable seizures by immunohistochemistry for HSV-1, HSV-2, and CMV as well as by the polymerase chain reaction for viral DNA sequences (HSV1, HSV2, and CMV) (Jay et al 1995). They also assessed eight non-epileptic patients with pathologically-demonstrated or clinically-suspected encephalitis and five specimens from patients with epilepsy but without encephalitis. Using polymerase chain reaction, CMV was present in six and HSV-1 in two of ten epilepsy patients with chronic encephalitis. CMV was demonstrated by in situ hybridization in two of the six patients positive for CMV by polymerase chain reaction. Immunochemistry was negative for viral antigens in all cases. None of the patients without encephalitis were found to have viral sequences by polymerase chain reaction, whereas two of the eight patients with encephalitis but without epilepsy did show CMV sequences. The authors suggest that in situ hybridization might miss some cases.

In contrast, Rasmussen reports negative standard viral studies (or positive reactions only to herpes simplex and measles virus in low dilution, which is of doubtful clinical significance) (Rasmussen 1978). And Mizuno and colleagues used immunoperoxidase stains against ten viral antigens on brain specimens from two patients with Rasmussen's syndrome and found all to be negative (Mizuno et al 1985). Furthermore, Farrell and associates were unable to detect CMV using immunostaining with anti-CMV antibodies in their three patients with Rasmussen's encephalitis (Farrell et al 1991).

Vinters and colleagues studied brain tissue from epileptic children with chronic (usually Rasmussen-type) encephalitis (Vinters et al 1993). They extracted DNA from specimens of brain tissue and used polymerase chain reaction with primers specific for CMV, varicella zoster, herpes simplex, Epstein-Barr virus, and human herpes virus 6 genes. They found evidence of low levels of CMV and Epstein-Barr virus genes in most brain specimens from encephalitis patients, and in several brain specimens from patients without encephalitis. The signal strength for both CMV and Epstein-Barr virus was much lower in the brains of patients with epilepsy than in the brains of AIDS patients with CMV encephalitis or brain lymphoma. The authors conclude that the small amounts of Epstein-Barr virus and CMV genes suggest that herpes virus infection of the brain did not directly cause Rasmussen' s encephalitis.

Atkins and colleagues studied ten biopsy and resection specimens from seven patients using biotinylated double-stranded DNA probes to CMV, herpes simplex virus, and Epstein-Barr virus (Atkins et al 1995). Electron microscopy was also carried out on two samples, and one was evaluated using standard immunoperoxidase techniques. However, they were unable to identify any evidence of viral material. The likely role of viruses in the pathogenesis of Rasmussen's encephalitis has been reviewed by Asher and Gajdusek (Asher and Gajdusek 1991).

Andrews and colleagues suggest that immunopathogenetic mechanisms are important in Rasmussen's encephalitis (Andrews et al 1990). They carried out extensive studies on the hemispherectomy specimen from a child with Rasmussen's encephalitis and found widespread cerebral vasculitis with immunofluorescence staining for IgG, IgM, IgA, C3 and C1q. There was also ultrastructural evidence of vascular injury, in addition to severe cortical atrophy with marked neuronal loss. The child had elevated serum antinuclear antibody titres and cerebrospinal fluid oligoclonal bands.


BIOLOGICAL BASIS

Perhaps the most exciting recent development in our understanding of the etiology of Rasmussen's encephalitis has been the establishment by Rogers and colleagues of a link between circulating antibodies of a ligand-gated ion channel receptor of the central nervous system in rabbits and a progressive encephalopathy with epileptic seizures (Rogers et al 1994). These workers report that two out of four rabbits who were immunized with GluR3 fusion protein in order to generate subtype-specific antibodies to recombinant GluR proteins developed seizures. Microscopic examination of the rabbit brains demonstrated chronic inflammatory changes consisting of microglial nodules and perivascular lymphocytic infiltration mainly in the cerebral cortex, as well as lymphocytic infiltration of the meninges. Rogers and colleagues reasoned that these changes probably occurred as a result of an autoimmune process directed against GluR3 and went on to look for these antibodies in four patients with pathologically confirmed Rasmussen's encephalitis. For controls, they used age- and sex-matched children with epilepsy, age and sex-matched children without CNS disease, children with active CNS inflammation, other children with epilepsy, and normal children. Immunoreactivity to GluR3 fusion protein was found in sera from two children with Rasmussen's encephalitis, and one child also showed weak immunoreactivity to GluR2 fusion protein, but the fourth child did not show immunoreactivity to any tested antigen. Only one control showed weak immunoreactivity to GluR3 that was different from the serum GluR immunoreactivity seen in individuals with Rasmussen's encephalitis. The GluR immunoreactivity appeared to correlate with disease activity, in that the three children with immunoreactivity had progressive disease or ongoing seizures, whereas hemispherectomy had been performed several years earlier in the child with no immunoreactivity, resulting in clinical stability and freedom from seizures.

As a result of these findings and the implication that the disease process might be related to circulating antibodies, Rogers and colleagues carried out plasma exchange in one of the children with Rasmussen's encephalitis who showed immunoreactivity and who was seriously ill (Rogers et al 1994). The result of the plasma exchange was a decrease in seizures and improvement in neurologic status, but the improvement was short-lived, with further deterioration over the following 4 weeks. The same team later reported the finding that the antibodies found in Rasmussen's encephalitis actually activate the receptor, raising the possibility that the antibodies might directly trigger seizures by overstimulating the glutamate receptors (Twyman et al 1995). The fact that Rasmussen's encephalitis sometimes appears to follow a blow to the head or systemic illness has led to their proposition of a model for the development of Rasmussen's encephalitis. In their model, the insult causes a breach in the blood-brain barrier, which, in individuals with autoantibodies to GluR3, could allow the entry of these antibodies to the brain, causing activation of the receptors and subsequent seizures. The result is a vicious circle in which more rifts would be caused in the blood-brain barrier as a result of the seizures.

Alternatively, the GluR3 antibodies may arise as a result of the initial central nervous system damage, thus leading to further damage.


EPIDEMIOLOGY

Rasmussen's syndrome is a rare disorder which, as far as is known, is not more prevalent in any population group. Diagnosis is dependent on the degree of neurologic sophistication in the country or area where the patients live.


PREVENTION

No preventive measures have been identified.


DIFFERENTIAL DIAGNOSIS

The clinical changes of Rasmussen's encephalitis, like the pathological changes, are non-specific, particularly in the early stages, and several other conditions can present in a rather similar manner. MELAS is one such condition, with the clinical features of episodic vomiting, recurrent strokes, and partial seizures that are frequently associated with prolonged migrainous manifestations and often develop into epilepsia partialis continua (Dvorkin et al 1987). Cortical dysplasia may cause intractable partial epilepsy (Andermann et al 1987), though only occasionally with epilepsia partialis continua. Tuberous sclerosis may cause similar symptoms (Andermann et al 1987), as may tumors (Rich et al 1985), cerebral vasculitis (Mackworth-Young and Hughes 1985), and Russian spring-summer encephalitis (Kozhevnikov 1991).


DIAGNOSTIC WORK-UP

Non-specific abnormalities of the cerebrospinal fluid have been found in 50% of patients with Rasmussen's syndrome. These include minor increases of the white cell count, a modest increase in the protein content of the spinal fluid, and a first-zone or mid-zone abnormality in the colloidal gold curve (Rasmussen and Andermann 1991). Oligoclonal or, occasionally, monoclonal bands have been found in some patients, but again this has not been consistent (Dulac et al 1991; Grenier et al 1991).

Radiology. The earliest patients to be reported with focal seizures due to chronic localized encephalitis were studied using pneumoencephalography (Rasmussen et al 1958). One of the patients, studied 5 months after the onset, had a normal pneumoencephalography. The second child, who presented with right-sided focal seizures followed by the development of a right hemiparesis, also had a normal pneumoencephalography early in the disease, but by 2 years after the onset, there was definite enlargement of the left lateral ventricle. Another pneumoencephalography performed 3 years later showed more marked evidence of atrophy of the left cerebral hemisphere, and by the following year, there was marked destruction of the left hemisphere with slight enlargement of the right lateral ventricle. The third patient described in the paper similarly had a normal pneumoencephalography early in the disease, with evidence of worsening hemispheric atrophy as the disease progressed.

With the development of more sophisticated methods of imaging the brain, it has become possible to examine the changes is more detail. The advent of CT scanning confirmed the development of progressive hemiatrophy, usually beginning in the temporoinsular region, causing enlargement of the temporal horn and Sylvian fissure and eventually progressing to involve the remainder of the hemisphere in the majority of patients (Tampieri et al 1991). These authors examined the CT scans of 15 patients diagnosed since 1974 and found hemiatrophy of variable severity in 11, diffuse cerebral atrophy in 2, and normal scans in 2 who were examined early in the course of the disease. They noted that the progression of the hemiatrophy could be very rapid, becoming severe in less that 24 months. They also noted that the contralateral ventricle could become enlarged in time. This may be due to Wallerian changes and perhaps to the effect of seizures and trauma. The contralateral atrophy was never comparable to that involving the affected hemisphere.|{picture:rsaf4.bmp}{caption:CT demonstrating ventricular enlargement in a patient with Rasmussen's encephalitis}{label:(A) 3 months after onset of seizures, (B) 3 years later.}|

There have also now been several reports of MRI findings in patients with Rasmussen's encephalitis. Tampieri and colleagues describe two patients who both showed hemiatrophy and abnormal, high-intensity signal on proton density and T2-weighted images in accordance with gliosis in one child (Tampieri et al 1991).|{picture:rsaf5.bmp}{caption:MRI scans of a patient with Rasmussen's encephalitis}{label:Both the coronal MRI (A) and the axial MRI (B) demonstrate hemispherical atrophy. In some patients there are also abnormal signals interpreted by the unwary as due to cerebral vascular disease.}|

Tien and colleagues report the results of neuroimaging in four young patients who had undergone various combinations of CT, xenon CT, and MR scans, and PET (Tien et al 1992). Two patients had rather unremarkable CT studies but had xenon CT scans showing selectively decreased cerebral blood flow to the affected hemisphere. A third patient had CT and MRI scans showing marked atrophy of the affected hemisphere, with decreased fluorodeoxyglucose tracer uptake in that hemisphere. The fourth patient's CT and MR scans showed severe hemispheric atrophy, with appearance of gliosis in the basal ganglia region and the periventricular area. Zupanc and colleagues describe a patient with typical features of Rasmussen's encephalitis in whom repeated MRI scans, with and without gadolinium, were normal, except one that was carried out approximately 5 months after seizure onset (Zupanc et al 1990). This MRI demonstrated increased signal intensity in the white and grey matter of the left temporal lobe and a small cortical area of the left parietal lobe on the T2-weighted images, which is suggestive of edema. Similar findings of hemispheric atrophy and signal change have been reported by other authors (Aguilar et al 1996; Yacubian et al 1997).

Nakasu and colleagues report serial MRI findings of Rasmussen's encephalitis in a 12-year-old boy who underwent biopsy and treatment with immunoglobulins (Nakasu et al 1997). No abnormality was seen on the initial scans carried out 1 year after the onset of seizures. But 11 months later, a high intensity lesion was seen in the left frontal cortex; this lesion rapidly spread into the white matter and then gradually regressed after biopsy and immunoglobulin therapy. Five months after the biopsy, another high intensity lesion was seen adjacent to the previous one, despite good seizure control at the time. Radiologists who are not aware of the clinical problem often interpret the signal changes observed in the MRI as being due to vascular abnormalities; therefore, it is essential to present the history and findings to the radiologist before the images are analyzed.

Cendes and colleagues performed MRS in three patients with Rasmussen's syndrome (Cendes et al 1995). They measured the relative resonance intensity of N-acetylaspartate to creatine, an index of neuronal loss or damage, for various regions in the brain. They demonstrated decreased relative N-acetylaspartate signal intensity over the entire affected hemisphere, including both cortex and white matter; the decrease in intensity was most prominent in the anterior periventricular region. There was a tendency for the signal loss to be worse in patients with longer duration of disease. Follow-up scans after 1 year showed progression of the changes. The authors also note that the changes on MRS were more widespread than the structural changes seen on MRI but did not affect the contralateral hemisphere. Two of the patients had epilepsia partialis continua during the follow-up scans only; these patients showed increase in lactate resonance intensity, suggesting that the lactate accumulation resulted from repetitive seizures rather than from the disease process itself.

Peeling and Sutherland performed MRS on tissue from patients undergoing surgical treatment for Rasmussen's encephalitis (Peeling and Sutherland 1993). They found that the metabolite concentrations varied with the severity and extent of the encephalitis, with the markedly abnormal tissues having decreased N-acetylaspartate, glutamate, cholines, and inositol. The decrease in the levels of N-acetylaspartate and glutamate was greater than in gliotic hippocampal tissue, suggesting the possibility that in vivo MRS might be helpful in diagnosis and in the assessment of results for various forms of immunological treatment.

Several authors have reported the results of functional imaging in patients with Rasmussen's encephalitis. English and colleagues describe five children with Rasmussen's encephalitis in whom SPECT imaging demonstrated an area of hypoperfusion and hypometabolism corresponding to the anatomical localization of the epileptogenic foci found by clinical assessment, EEG, and CT (English et al 1989). In all cases, the SPECT showed a more extensive area of abnormality than CT, and in two patients undergoing sequential studies, the SPECT reflected the patients' changing clinical condition.

Hwang and associates report PET and SPECT studies in patients with Rasmussen's encephalitis (Hwang et al 1991). One child was studied with 99Tc-HMPAO SPECT, showing an increase in cerebral blood flow in the left temporal lobe ictally and a wider decrease interictally in the left temporal, frontal, and parietal lobes. Five patients underwent 18-FDG-PET scanning, which usually showed a regional decrease in the local cerebral metabolic rate in the utilization of glucose, widely distributed over the affected hemisphere and extending beyond the frontal and temporal lobes. However, within the regional hypometabolic zone were one or two more foci of localized increase in metabolic rate, which, in some cases, coincided with focal epileptogenic activity as determined by electrocorticography at surgery.

Burke and colleagues describe a patient with Rasmussen's syndrome in whom Tc-99m HMPAO SPECT produced grossly abnormal results at a time when the MRI scan showed no structural abnormality, suggesting that Tc-99m HMPAO SPECT might be helpful in early diagnosis of the condition (Burke et al 1992). Aguilar and associates describe an 8-year-old girl with epilepsia partialis continua due to Rasmussen's encephalitis involving the left side of the body who underwent ictal and interictal 99mTc HMPAO-SPECT (Aguilar et al 1996). In the ictal period this showed increased cerebral blood flow in the right hemisphere, particularly the Rolandic area and the temporal lobe, whereas in the interictal period a decreased flow was seen in the same regions. Yacubian and colleagues also report a focal increase in regional cerebral blood flow in four patients presenting with epilepsia partialis continua at the time of the HMPAO injection, and extensive cortical hypoperfusion in four other patients who received the injection during the interictal state (Yacubian et al 1997). These authors report abnormalities of cerebellar function in six patients, two of them with structural damage.

EEG changes. Several authors have studied EEG changes in Rasmussen's encephalitis. The largest study to date has been that of So and Gloor, who describe the findings in 339 EEGs and 58 electrocorticograms carried out in 49 patients with Rasmussen's encephalitis (So and Gloor 1991). Analysis of all the preoperative EEGs showed a variety of abnormalities. All but one of the 47 patients for whom preoperative EEGs were available had some abnormality of background activity (either slowing beyond age-adjusted limits or irregularities in the morphology and frequency of waveforms), usually asymmetrical. In the later EEGs, the majority showed some abnormality bilaterally, but predominant on one side. All EEGs showed abnormal slow wave activity, and 44 of the 47 patients had evidence of interictal epileptiform discharges. Frequently, there were multiple independent foci lateralized over one hemisphere. Bilateral multiple independent discharges were seen in one third of patients but were usually predominant over one side. Almost half of the patients showed bilaterally synchronous spike and wave or sharp and slow wave discharges. Clinical or subclinical seizures were recorded in 32 patients, but although the onset could usually be lateralized to one hemisphere, it was rare for seizures to have a strictly localized electrographic onset. Electrocorticography usually showed widespread regions involved in interictal epileptiform activity, and if several seizures were studied, it was common to find multiple independent sites of seizure onset. The evolution of the EEG during the course of the disease process was also studied. Those patients with early disease who had not yet developed hemiparesis were more likely to show unilateral disturbance of background activity. As the disease progressed bilateral abnormalities became more common. EEG epileptiform abnormalities generally became more widespread with time, and some patients showed the development of independent epileptiform abnormalities over the contralateral hemisphere. Capovilla and colleagues also studied the evolution of the EEG from the onset of the disease in a single patient (Capovilla et al 1997). They noted focal delta activity over the left temporal region without spikes, at a time when the MRI was normal, and ongoing seizure activity was absent. They hypothesize that the presence of such changes in the absence of structural abnormality on imaging should prompt consideration of the diagnosis of Rasmussen's encephalitis even before the development of the classical clinical features. Andrews and colleagues studied two patients with pathologically confirmed Rasmussen's encephalitis and circulating GluR3 antibodies (Andrews et al 1997). The two patients were treated with plasma exchange and immunosuppressive treatment with intravenous immunoglobulins; high-dose steroids were also given to one patient. Repeated EEG monitoring showed that the EEG abnormalities present before plasma exchange improved during plasma exchange but worsened afterwards, apparently reflecting the change in clinical status.


PROGNOSIS AND COMPLICATIONS

The initial course of Rasmussen's encephalitis is frequently characterized by rather non-specific clinical features, and it may be months or even years before the diagnosis becomes apparent. Oguni and colleagues divide the clinical course into three stages (Oguni et al 1992). In the first stage, seizures are mainly simple partial attacks with somatosensory or motor signs, and complex partial seizures without automatisms (with or without epilepsia partialis continua in each case). During the later part of the first stage, they note that seizures gradually become more frequent, and that the hemiparesis, initially postictal and transient, slowly becomes more permanent. In the second stage, there is a further increase in seizures with the development of more apparent fixed neurologic signs and increasing disability. Eventually, there seems to be a tendency for the disease activity to burn itself out, so that in the third stage there is a diminution in seizure frequency and severity, without further progression of the neurologic signs (which by this time would commonly include moderate to severe hemiparesis, a visual field defect, and a variable degree of intellectual and language impairment that ranged from mild to severe). The relentless progression of Rasmussen's encephalitis at the onset has led to the introduction of various medical treatments (Dulac et al 1991; DeToledo and Smith 1994; Hart et al 1994; Andrews et al 1996; McLachlan et al 1996), none of which have been entirely successful, and the suggestion that surgical treatment (particularly hemispherectomy, which does appear to halt the disease process in the majority of patients) should be carried out sooner rather than later (Vining et al 1993).


MANAGEMENT

In their protocol of high-dose steroid and immunoglobulin treatment for children with Rasmussen's encephalitis, Hart and colleagues suggest the diagnosis of chronic encephalitis in children who develop epilepsia partialis continua and meet at least one of the following criteria (Hart et al 1994):

(1) Progressive neurologic deficit at the beginning or after the onset of epilepsia partialis continua but before the start of treatment
(2) Progressive hemispheric atrophy on CT, MRI, or both, with or without density or signal abnormalities
(3) Presence of oligoclonal or monoclonal banding on CSF examination
(4) Biopsy evidence of chronic encephalitis Children without epilepsia partialis continua but with focal epilepsy and biopsy evidence of chronic encephalitis (who might in addition meet criteria 1, 2 or 3) were also considered to have the diagnosis.

Because of the relentless progression of Rasmussen's encephalitis in the majority of patients and because of the rarity of the condition, which makes clinical trials difficult, clinicians have tried a variety of treatments on an empirical basis. Although hemispherectomy appears to be successful in arresting the disease process in the majority of patients, the consequent neurologic deficits cause reluctance to carry out this procedure until a hemiparesis already exists.

There has long been debate as to whether the progressive neurologic deficits in Rasmussen's encephalitis are secondary to ongoing seizure activity or if they are an independent effect of the encephalitic process.|{diagram:rsaf6.bmp}{caption:Schematic representation of the mechanism of neural injury in Rasmussen's encephalitis}{label:From Antel and Rasmussen 1996, with permission.}|The initial attempts at treating Rasmussen's encephalitis concentrated on the quest for seizure control. The pharmacological treatment of 25 patients with Rasmussen's encephalitis at the Montreal Neurological Institute was analyzed by Dubeau and Sherwin who found that all had received polytherapy, often at the expense of significant morbidity as a result of toxic or other adverse effects (Dubeau and Sherwin 1991). Of the seizure types most commonly seen in patients with Rasmussen's encephalitis (partial motor, complex partial, and secondarily generalized tonic-clonic seizures), the secondarily generalized tonic-clonic seizures were most likely to respond to treatment.

Most other treatments directed at aborting the disease have relied on the assumption that the cause is either infective, probably viral, or the result of an autoimmune process. Examples of such treatments include antiviral treatments including the use of ganciclovir, zidovudine, high-dose interferon, high-dose steroids and immunoglobulins, and plasma exchange.

DeToledo and Smith used zidovudine to treat a 4-year-old child with epilepsia partialis continua, progressive aphasia, and right hemiparesis after the seizures had failed to respond to conventional antiepileptic drugs and ACTH (DeToledo and Smith 1994). Zidovudine was given for 62 days but was eventually discontinued because of granulocytopenia. Seizures stopped and neurologic deterioration was arrested for approximately 21 months within 6 weeks of the onset of treatment. Unfortunately when the patient relapsed, with seizures affecting the previously uninvolved left hemibody, side-effects prevented further treatment with zidovudine. Zidovudine was also used in three patients by Shorvon and colleagues (Shorvon, personal communication). But the improvement was, unfortunately, short-lived and unsustained.

Because of the fact that CMV had been implicated in the pathogenesis of Rasmussen's syndrome, the effect of ganciclovir was assessed in four patients by McLachlan and colleagues (McLachlan et al 1996). The CMV genome was sought in three of these patients and found in two. One child with very frequent seizures, which developed over 3 months, became seizure-free 5 days after the onset of treatment; there was also resolution of focal neurologic signs, cognitive function, and EEG changes. Two other patients treated 34 and 72 months after disease onset showed some improvement, whereas in the fourth patient there was no benefit.

Early reports of the use of corticosteroids (including dexamethasone, prednisone, and ACTH) in Rasmussen's encephalitis were not encouraging (Gupta et al 1984; Piatt et al 1988). However, Dulac and colleagues, who were among the first to try high-dose steroids in children with Rasmussen's encephalitis, produced some promising results (Dulac et al 1991). They treated five children with three intravenous infusions of 400mg/m2 of methylprednisolone, one every other day, followed by oral prednisone (2mg/kg per day) or hydrocortisone (10 mg/kg per day) tapered off over 3 to 24 months. Epilepsia partialis continua ceased in three cases within 1 month of the start of treatment, with the EEG improving dramatically in two of the three plus one other child. Progression of motor and cognitive impairment stopped in all of the children, though only one showed a clear improvement, from a state in which she was bedridden and mute to one in which she could walk and talk. However, two of the patients relapsed within a few months of the cessation of treatment.

As a result of this success, other authors tried this and other immunosuppressive treatments (Walsh 1991; Hart et al 1994). Several reports have documented improvement in intractable epilepsy of other etiologies treated with gamma-globulin (Pechadre et al 1977; Laffont et al 1979; Ariizumi et al 1983). Furthermore, Walsh describes a 12-year-old boy with Rasmussen's syndrome in whom treatment with six infusions (200 mg/kg) of intravenous immunoglobulin over a period of 3 months showed improvement both in neurologic dysfunction and also in seizure control during the course of the treatment and for several months afterwards (Walsh 1991).

Hart and colleagues describe 19 patients treated with high-dose steroids, immunoglobulins, or both (Hart et al 1994). Two (both biopsy-proven) had developed Rasmussen's syndrome in adulthood, whereas the rest were children. The diagnosis was confirmed by biopsy in all but three of the patients. Because the patients were treated at different centers, the treatment protocols varied. Seventeen patients received treatment with steroidsÑusually oral steroids, although six patients received intravenous methylprednisolone at some stage, and one child received ACTH injections over 4 weeks. Two patients received intravenous immunoglobulins alone, whereas seven patients received both high dose steroids and intravenous immunoglobulins. Seven patients showed no improvement in seizure frequency following treatment with steroids. Two patients showed an improvement of 25% or less, whereas eight showed at least a 50% reduction in seizure frequency. With the exception of two patients, the frequency of seizures increased within days or weeks of steroid withdrawal. Side effects were common and often prominent. Seven of the nine patients treated with intravenous immunoglobulin showed definite improvement in seizure control, at least initially. This was not maintained in three patients. Any improvement in neurologic deficit in these patients was only transient and accompanied improved seizure control, except in one instance wherein the hemiparesis improved disproportionately to the improvement in seizure control. It seems likely that treatments such as these might have maximum effect early in the onset of disease. In this study a considerable proportion of the patients had been ill for several years before treatment (fourteen already had mild hemiparesis and another four had evidence of intellectual deterioration), and this may be responsible for the rather poor results.

Hart and colleagues suggest protocols for the treatment of patients with intravenous immunoglobulins or high-dose steroids (Hart et al 1994). With respect to intravenous immunoglobulin, the recommended treatment is 400 mg/kg per day by intravenous infusion on 3 successive days, with a single further infusion of 400 mg/kg at monthly intervals if improvement occurs. If no improvement is seen, treatment with steroids is recommended, with the initial course consisting of intravenous methylprednisolone (400 mg/m2 of body surface) given as three consecutive infusions on alternate days. Subsequent infusions consist of single infusions at monthly intervals for the first year, bi-monthly intervals for the second year, and tri-monthly intervals for the third year, unless serious side-effects supervene. The treatment is accompanied by oral prednisolone starting at 2 mg/kg per day, reducing very gradually over a period of months depending on clinical response, with the total duration of oral steroid treatment usually being 1 to 2 years.

Another patient with symptoms highly responsive to repeated courses of immunosuppressant treatment is described by Krauss and colleagues (Krauss et al 1996). They describe a woman who had developed her first symptoms, partial and secondarily generalized seizures, at the age of 15, and had gone on to develop typical features of Rasmussen's encephalitis. Brain biopsy was also consistent with this diagnosis. Treatment with immunoglobulins at the age of 29 produced no change in her seizure control or aphasia, but intravenous methylprednisolone brought about a dramatic improvement in her seizures and neurologic deficits. Oral steroids were ineffective, and treatment with intermittent cyclophosphamide was insufficient to contain her symptoms. She was also treated with plasmapheresis, though the effect was unclear. Her serum and CSF were negative to antibodies to GluR3 by both immunoblot and immunocytochemical analysis of cells transfected with GluR3 cDNA, suggesting an alternative immune-mediated process in some patients with chronic encephalitis.

Intraventricular alpha-interferon has also been tried in Rasmussen's encephalitis (Maria et al 1993; Dabbagh et al 1997), on the basis that not only do interferons have immunomodulating activity, such as enhancement of the phagocytic activity of macrophages and augmentation of the specific cytotoxicity of lymphocytes for target cells, but they also inhibit virus replication in virus-infected cells. The 3 year old child described by Dabbagh and colleagues had epilepsia partialis continua and a right hemiparesis and was mute at the time of treatment (Dabbagh et al 1997). She was given three doses of 3,000,000 units of INF-alpha through an Omaya reservoir in the first week on alternate days, two doses per week in the second and third weeks, and weekly doses for the fourth through sixth weeks. She had a significant reduction in seizures but relapsed to baseline 3 weeks after stopping treatment. However, she did respond to further courses of treatment, and at the time of the report, more than 12 months after the onset of treatment, she remained seizure-free with treatments every third week. The patient described by Maria and colleagues also showed improvement in the control of his epilepsy and neurologic deficit with INF-alpha in the short term (Maria et al 1993).

Andrews and colleagues describe the use of plasma exchange in four patients with clinical and pathological features of Rasmussen's encephalitisÑtwo of them previously described by Rogers and colleagues (Rogers et al 1994; Andrews et al 1996). Three of the patients had repeated, dramatic, transient clinical improvements shown by reduced seizure frequency, rapid control of status epilepticus, and improved neurologic function. Two of these patients had evidence of active inflammation on pathological examination; the third had chronic changes (though autoantibodies were present). The fourth patient, who also had pathological evidence of active inflammation, had a more muted response to repeated plasma exchange. The authors draw attention to the known complications of plasma exchange (infection, anaemia, coagulopathy, etc.) and to its expense. They also suggest certain situations in which plasma exchange might prove particularly helpful, as in, for example, status epilepticus in Rasmussen's encephalitis and the evaluation of patients prior to surgery, when residual function may be unmasked by the reduction in seizure frequency brought about by the plasma exchange. Andrews and colleagues also suggest a protocol, advocating five to six single volumes of plasma exchange initially, with albumin and saline replacement, spread over 10 to 12 days, with an infusion of 1g/kg of intravenous immunoglobulin being given to the patient the day after (Andrews et al 1996). They recommend that subsequent plasma exchange be given on the basis of clinical need because of recurrent seizures, perhaps every 2 to 3 months on average. They believe that interval treatment with immunosuppressive agents might limit expense in addition to prolonging the improvement after each plasma exchange.

Despite the promise shown by these treatments in a few patients, at present none has shown itself to reliably affect the course of the disease. Surgical treatment has been tried in a number of patients. Limited focal resection carried out early in the disease appears to be of little lasting benefit (Rasmussen and McCann 1968). There seems to be a consensus of opinion that functional hemispherectomy is a reasonable option when the patient has developed hemiparesis and homonymous hemianopia (Rasmussen 1983). However, some groups believe that since hemispherectomy is the only procedure that apparently stops progression of the disease, it should be considered as an early option, without awaiting the development of maximal hemiparesis (Vining et al 1993).


PREGNANCY

No information is available.


ANESTHESIA

No information is available.


REFERENCES CITED

Aguilar Rebolledo F, Rojas Bautista JC, Villanueva Perez R, Morales Hernandez S. [SPECT-99mTc-HMPAO in a case of epilepsia partialis continua and focal encephalitis] [Spanish]. Rev Invest Clin 1996;48:199-205.

Andermann F, Olivier A, Melanson D, Robitaille Y. Epilepsy due to focal cortical dysplasia with macrogyria and the forme fruste of tuberous sclerosis: a study of 15 patients. In: Wolf P, Dam M, Janz D, Dreifuss F, editors. Advances in epileptology: the 16th Epilepsy International Symposium. New York: Raven Press, 1987:35-8.

Andrews JM, Thompson JA, Pysher TJ, Walker ML, Hammond ME. Chronic encephalitis, epilepsy, and cerebrovascular immune complex deposits. Ann Neurol 1990;28:88-90.

Andrews PI, Dichter MA, Berkovic SF, Newton MR, McNamara JO. Plasmapheresis in Rasmussen's encephalitis. Neurology 1996;46:242-6.

Andrews PI, McNamara JO, Lewis DV. Clinical and electroencephalographic correlates in Rasmussen's encephalitis. Epilepsia 1997;38:189-94.

Antel JP, Rasmussen T. Rasmussen's encephalitis and the new hat. Neurology 1996;46:9-11.

Ariizumi M, Baba K, Shiihara H, et al. High dose gamma-globulin for intractable childhood epilepsy. Lancet 1983;2:162-3.

Asher DM, Gajdusek DC. Virologic studies in chronic encephalitis. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:147-58.

Atkins MR, Terrell W, Hulette CM. Rasmussen's syndrome: a study of potential viral etiology. Clin Neuropathol 1995;14:7-12.

Burke GJ, Fifer SA, Yoder J. Early detection of Rasmussen's syndrome by brain SPECT imaging. Clin Nucl Med 1992;17:730-1.

Capovilla G, Paladin F, Bernardina BD. Rasmussen's syndrome: longitudinal EEG study from the first seizure to epilepsia partialis continua. Epilepsia 1997;38:483-8.

Cendes F, Andermann F, Silver K, Arnold DL. Imaging of axonal damage in vivo in Rasmussen's syndrome. Brain 1995;118:753-8.

Dabbagh O, Gascon G, Crowell J, Bamoggadam F. Intraventricular interferon-alpha stops seizures in Rasmussen's encephalitis: a case report. Epilepsia 1997;38:1045-9.

DeToledo JC, Smith DB. Partially successful treatment of Rasmussen's encephalitis with zidovudine: symptomatic improvement followed by involvement of the contralateral hemisphere. Epilepsia 1994;35:352-5.

Dubeau F, Sherwin A. Pharmacologic principles in the management of chronic encephalitis. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:179-92.

Dulac O, Robain O, Chiron C, et al. High-dose steroid treatment of epilepsia partialis continua due to chronic focal encephalitis. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:79-110.

Dvorkin GS, Andermann F, Carpenter S, et al. Classical migraine, intractable epilepsy, and multiple strokes: a syndrome related to mitochondrial encephalomyopathy. In: Andermann F, Lugaresi E, editors. Migraine and epilepsy. Boston: Butterworths, 1987:202-32.

English R, Soper N, Shepstone BJ, Hockaday JM, Stores G. Five patients with Rasmussen's syndrome investigated by single-photon-emission computed tomography. Nucl Med Commun 1989;10:5-14.

Farrell M, Cheng L, Cornford ME, Grody WW, Vinters HV. Cytomegalovirus and Rasmussen's encephalitis. Lancet 1991;337:1551-2.

Friedman H, Ch'ien L, Parham D. Virus in brain of child with hemiplegia, hemiconvulsions, and epilepsy. Lancet 1977;2:666.

Gilden DH, Lipton H. Cytomegalovirus and Rasmussen's encephalitis. Lancet 1991;337:239.

Grenier Y, Antel JP, Osterland CK. Immunologic studies in chronic encephalitis of Rasmussen. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:125-34.

Gupta PC, Rapin I, Houroupian DS, Roy S, Llena JF, Tandon PN. Smouldering encephalitis in children. Neuropediatrics 1984;15:191-7.

Hart YM, Cortez M, Andermann F, et al. Medical treatment of Rasmussen's syndrome (chronic encephalitis and epilepsy): effect of high-dose steroids or immunoglobulins in 19 patients. Neurology 1994;44:1030-6.

Hwang PA, Gilday DL, Spire JP, et al. Chronic focal encephalitis of Rasmussen: functional neuroimaging studies with positron emission tomography and single-photon emission computed tomography scanning. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:61-72.

Jay V, Becker LE, Otsubo H, et al. Chronic encephalitis and epilepsy (Rasmussen's encephalitis): detection of cytomegalovirus and herpes simplex virus 1 by the polymerase chain reaction and in situ hybridization. Neurology 1995;45:108-17.

Kozhevnikov AY. A particular type of cortical epilepsy (epilepsia corticalis sive partialis continua). Asher DM, trans. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:245-65.

Krauss GL, Campbell ML, Roche KW, Huganir RL, Niedermeyer E. Chronic steroid-responsive encephalitis without autoantibodies to glutamate receptor GluR3. Neurology 1996;46:247-9.

Laffont F, Esnaults S, Gilbert A, Peytour MA, Cathala HP, Eygonnet JP. Effet des hammaglobulines sur des Žpilepsies rebelles. Etude prŽliminaire. Ann Med Interne (Paris) 1979;130:307-12.

Mackworth-Young CG, Hughes GR. Epilepsy: an early symptom of systemic lupus erythematosus (letter). J Neurol Neurosurg Psychiatry 1985;48:185.

Maria BL, Ringdahl DM, Mickle JP, et al. Intraventricular alpha interferon therapy for Rasmussen's syndrome. Can J Neurol Sci 1993;20:333-6.

McLachlan RS, Girvin JP, Blume WT, Reichman H. Rasmussen's chronic encephalitis in adults. Arch Neurol 1993;50:269-74.

McLachlan RS, Levin S, Blume WT. Treatment of Rasmussen's syndrome with ganciclovir. Neurology 1996;47:925-8.

Mizuno Y, Chou SM, Estes ML, Erenberg G, Cruse RP, Rothner AD. Chronic localized encephalitis (Rasmussen's) with focal cerebral seizures revisited. J Neuropathol Exp Neurol 1985;44:351.

Nakasu S, Isozumi T, Yamamoto A, Okada K, Takano T, Nakasu Y. Serial magnetic resonance imaging findings of Rasmussen's encephalitis: a case report. Neurol Med Chir (Tokyo) 1997;37:924-8.

Oguni H, Andermann F, Rasmussen TB. The natural history of the syndrome of chronic encephalitis and epilepsy: a study of the MNI series of forty-eight cases. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:7-35.

Oguni H, Andermann F, Rasmussen TB. The syndrome of chronic encephalitis and epilepsy: a study based on the MNI series of 48 cases. Adv Neurol 1992;57:419-33.

Pechadre JC, Sauvezie B, Osier C, Gilbert J. Traitement des encŽphalopathies de l'enfant par les gammaglobulines. RŽsultats prŽliminaires. Rev Electroencephalogr Neuropsysiol Clin 1977;7:443-7.

Peeling J, Sutherland G. 1 H magnetic resonance spectroscopy of extracts of human epileptic neocortex and hippocampus. Neurology 1993;43:589-94.

Piatt JH, Hwang PA, Armstrong DC, Becker LE, Hoffman HJ. Chronic focal encephalitis (Rasmussen syndrome): six cases. Epilepsia 1988;29:268-79.

Power C, Poland SD, Blume WT, Girvin JP, Rice GP. Cytomegalovirus and Rasmussen's encephalitis. Lancet 1990;336:1282-4.

Rasmussen T. Further observations on the syndrome of chronic encephalitis and epilepsy. Appl Neurophysiol 1978;41:1-12.

Rasmussen T. Hemispherectomy for seizures revisited. Can J Neurol Sci 1983;10(2):71-8.

Rasmussen T, Andermann F. Rasmussen's syndrome: symptomatology of the syndrome of chronic encephalitis and seizures: 35-year experience with 51 cases. In: Luders HO, editor. Epilepsy surgery. New York: Raven Press, 1991:173-82.

Rasmussen T, McCann W. Clinical studies of patients with focal epilepsy due to "chronic encephalitis". Trans Am Neurol Assoc 1968;93:89-94.

Rasmussen T, Olszewski J, Lloyd-Smith D. Focal seizures due to chronic localized encephalitis. Neurology 1958;8:435-45.

Rich KM, Goldring S, Gado M. Computed tomography in chronic seizure disorder caused by glioma. Arch Neurol 1985;42:26-7.

Robitaille Y. Neuropathologic aspects of chronic encephalitis. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:79-110.

Rogers SW, Andrews PI, Gahring LC, et al. Autoantibodies to glutamate receptor GluR3 in Rasmussen's encephalitis. Science 1994;265:648-51.

Root-Bernstein RS. Cytomegalovirus and Rasmussen's encephalitis. Lancet 1991:239-40.

So NK, Gloor P. Electroencephalographic and electrocorticographic findings in chronic encephalitis of the Rasmussen type. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:37-45.

Tampieri D, Melanson D, Ethier R. Imaging of chronic encephalitis. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann. 1991;47-60.

Tien RD, Ashdown BC, Lewis DV, Atkins MR, Burger PC. Rasmussen's encephalitis: neuroimaging findings in four patients. AJR Am J Roentgenol 1992;158:1329-32.

Twyman RE, Gahring LC, Spiess J, Rogers SW. Glutamate receptor antibodies activate a subset of receptors and reveal an agonist binding site. Neuron 1995;14:755-62.

Vining EP, Freeman JM, Brandt J, Carson BS, Uematsu S. Progressive unilateral encephalopathy of childhood (Rasmussen's syndrome): a reappraisal. Epilepsia 1993;34:639-50.

Vinters HV, Wang R, Wiley CA. Herpes viruses in chronic encephalitis associated with intractable childhood epilepsy. Hum Pathol 1993;24:871-9.

Walsh PJ. Treatment of Rasmussen's syndrome with intravenous gammaglobulin. In: Andermann F, editor. Chronic encephalitis and epilepsy: Rasmussen's syndrome. Boston: Butterworth-Heinemann, 1991:201-204.

Walter GF, Renella RR. Epstein-Barr virus in brain and Rasmussen's encephalitis. Lancet 1989;1:279-80.

Yacubian EM, Marie SK, Valerio RM, Jorge CL, Yamaga L, Buchpiguel CA. Neuroimaging findings in Rasmussen's syndrome. J Neuroimaging 1997;7:16-22.

Zupanc ML, Handler EG, Levine RL, et al. Rasmussen encephalitis: epilepsia partialis continua secondary to chronic encephalitis. Pediatr Neurol 1990;6:397-401.

Back to Top   |    Home