EPC, defined as spontaneous regular or irregular clonic twitching confined to one part of the body (Figure 12-2), aggravated by action or sensory stimuli, and persisting during sleep,⁸ occurs in most of the patients (56 to 92% according to different series) at some time during the disease course, unresponsive to AE drugs.^3–4,⁹

Figure 12–2 Wake (A) and sleep (B) EEG recorded from a 5-year-old patient, 4 months from the first seizure. EPC involving the right lower limb is shown. The rhythmic myoclonic jerks are continuous while awake and persist, albeit at lower frequency, during sleep. Note the asymmetry of EEG tracing: rhythmic background activity during wakefulness and sleep spindles is evident over the right hemisphere. On the left hemisphere the EEG is dominated by slow activity and continuous epileptic discharges over the central region.

The striking EEG feature at onset is the presence of slow focal activity mainly involving the temporal and central leads on the affected hemisphere, associated with few epileptic abnormalities on the same regions. Additional distinctive features, in the first few months, are the early evidence of ictal and interictal hemispheric multifocality, the presence of subclinical ictal discharges, and the unilateral impoverishment of the background activity and of sleep spindles (Figure 12-2).^4,¹⁰

During the disease course, the background activity further flattens, and sleep organization deteriorates further; epileptic and slow activity tend to increase and spread within the affected hemisphere and the unaffected one. With time, the contralateral abnormal activity, either resulting from diffusion or seemingly asynchronous, may become more frequent than that recorded in the affected hemisphere (i.e., “false lateralization”).¹¹ In any case, the seizure onset, albeit multifocal, is closely unilateral; the recording of seizure from the supposed unaffected hemisphere strongly questions the diagnosis of RE.

EPC, like in other conditions, is often not clearly related to electroencephalogram (EEG) changes recorded by scalp electrodes; however, auxiliary neurophysiologic techniques may contribute to define the cortical origin of the jerks. Back-averaging may identify the spikes preceding the motor phenomena, and somatosensory-evoked potentials of the rolandic cortex are often abnormally enlarged.¹²

NEUROLOGIC SYMPTOMS OTHER THAN EPILEPSY

Hemiparesis invariably develops during the disease course. As already mentioned, the motor deficit may be the symptom of onset, and exceptionally remains the one and only manifestation of RE.⁵ In the first stages of the disease, hemiparesis may be limited to the postictal phase, but rapidly becomes constant, albeit fluctuating in severity as it worsens with increasing seizure activity. With time, hemiparesis, sometimes associated with a dystonic component, stabilizes. In an Italian series of 12 patients, focal motor deficits appeared from 15 days to 24 months after the first seizure and invariably worsened to severe hemiparesis, leading three patients to be wheelchair bound within 3 years of onset.⁴

Movement disorders, which seldom mark the onset of RE, are rarely reported during the disease course, probably because they are misinterpreted or overlooked in patients whose clinical picture is dominated by epilepsy and hemiparesis. Additional neurologic symptoms include hemianopia, cortical sensory loss, and aphasia when the dominant hemisphere is affected.¹³

Cognitive impairment is another constant feature of RE, and as for motor deficits, its appearance may be subtle. Behavioral changes, with irritability, emotional lability, or hyperactivity, often herald the first signs of mental decline, which consist mainly of memory and attention disorders and learning difficulties. In the Italian series, these symptoms were detected within 4 to 36 months after the first seizure and progressively worsened until surgery. At the time of surgery, performed between 7 months and 14 years after disease onset, the mean IQ was 61.3 ± 15.1 (range 44–87).⁴ In most patients, the progression of mental impairment seems to correlate with the severity of epilepsy and particularly with the bilateral spreading of EEG epileptic abnormalities (personal experience), or with the appearance of asynchronous contralateral foci.¹⁴

The natural history of RE is summarized in the seminal paper by Oguni from the Montreal Neurological Institute ³ and in the more recent report from the Bonn group, correlating the clinical course with the development of brain damage as documented by serial MRIs.⁹ Both groups recognize three stages that are virtually comparable. In the “prodromal stage,” lasting from 0 months to 8 years, seizures manifest at low frequency, and rarely, mild hemiparesis may be present. The “acute stage,” which in a significant number of cases appears to be the initial clinical manifestation, lasts 4 to 8 months and is characterized by frequent seizures, EPC (not always present), and rapid neurological deterioration. Finally, the patient enters the “residual stage,” with stable neurological deficits and persisting, albeit less frequent, seizures.

Atypical RE

Under this heading, the less-common manifestations of RE, with regard to age at onset, association with other diseases, and bilateral involvement, are included.

ADOLESCENT- AND ADULT-ONSET RE

A limited, but increasing number of patients with adult-onset RE have been reported, accounting for about 10% of all RE reported cases.^12,¹⁵ The overall features in childhood and adult onset are similar with regard to clinical, electrophysiological, and neuroimaging findings. The main differences consist of a more frequent posterior onset and a milder, more protracted, clinical course. Motor and mental deterioration are less severe than in typical childhood RE, hemispheric damage less pronounced, and quality of life somewhat preserved. However, in a number of cases, late-onset RE runs a malignant course, comparable to that of childhood-onset disease. In a recent paper Villani and coworkers identified two distinct patterns of disease presentation, one characterized by focal motor epilepsy (the “epileptic” phenotype), and the other by focal cortical myoclonus (the “myoclonic” phenotype). Unilateral neurologic deficits and brain atrophy were progressive in both phenotypes, but they were more prominent and detected earlier in the “epileptic” phenotype.¹²

RE PROTRACTED VARIANTS

A few RE patients are reported to have an acceptable seizure frequency and not invalidating motor deficits.¹⁵ This was already described in the pioneering study by Aguilar and Rasmussen,¹⁶ who reported that some patients may have mild neurologic deficits, despite the neuropathological evidence of active encephalitis.

RE ASSOCIATED WITH OTHER DISEASES (DOUBLE PATHOLOGY)

Patients affected by preexisting brain lesions account for about 10% of RE cases.¹ This dual pathology has been documented in low-grade tumors, cortical dysplasia, tuberous sclerosis, vascular malformations, or old ischemic lesions. The association of structural abnormalities and inflammation is still a matter of investigation. The hypothesis that structural or acquired brain lesions, and the resulting epilepsy, may alter the blood–brain barrier permeability, allowing the entry of compounds with immunogenic or inflammatory potential, is conceivable but, at least to date, not demonstrated.¹⁷

The association of RE with definite autoimmune diseases, such as SLE,¹⁸ linear scleroderma,¹⁹ Parry–Romberg syndrome,²⁰ although rarely reported, is noteworthy, as it provides a further hint for the role of immune-mediated mechanisms in the pathogenesis of RE.

BILATERAL RE

The hallmark of RE is the unilateral brain damage, and this distinctive and unique characteristic should be, in the opinion of the authors, mandatory for diagnosis, as it entails the surgical indication, as discussed in the treatment section.

The term “bilateral RE” has been used in the literature to include two different conditions: (1) the secondary involvement of the unaffected hemisphere, as indicated by spreading of focal seizures from the original focus, by the appearance of interictal epileptic abnormalities, and by the mild brain atrophy, that most probably results from Wallerian degeneration of commissure fibers;²¹ (2) a familiar²² and few sporadic cases of early-onset malignant epilepsy with neuropathology consistent with chronic encephalitis resembling RE;²³^–²⁴ and (3) rare adult-onset cases.²⁵

The frequently observed secondary involvement of the unaffected hemisphere, of course, does not rule out the diagnosis of RE because in no cases is there evidence of truly contralateral disease. It should also be underlined that in RE patients surgically treated, even after a very protracted disease, by hemispherectomy (or hemispheric disconnection), no relapse of seizures from the contralateral hemisphere was ever recorded. Conversely, in “bilateral RE,” symptoms and, when available, MRI findings pointed to bilateral brain damage within the first months of the disease. In our opinion, from a pragmatic point of view, these cases should be kept separate from true RE (and probably named differently), given the different clinical picture, MRI findings, and, overall, the ensuing therapeutic approaches.

Pathogenesis of RE

The etiology of RE remains unknown, but our understanding of its pathogenesis improved considerably in recent years. The initiating event triggering RE remains unknown. A viral etiology was first proposed, based on the features of lymphocyte infiltration and microglial reaction within the brain and on similarities with known forms of viral encephalitis.² However, no attempt of isolation of a pathogenic virus has been successful.²⁶ The interest in RE has been renewed by histopathological and experimental studies showing the involvement of the immune system in the pathogenesis of the disease, providing evidence for both humoral and cellular factors.

In 1994, Rogers and coworkers observed that rabbits immunized with a recombinant fragment of the glutamate receptor (GluR3) developed seizures and that the histopathological examination of their brain showed inflammatory changes reminiscent of those found in patients with RE. In parallel, the same authors detected antiGluR3 antibodies in three patients with RE and observed clinical improvement after plasma exchange in one of them.²⁷ Further reports of transient clinical improvement after removal of antibodies have been subsequently published.^13,²⁸^–²⁹ These preliminary findings fostered the search for such autoantibodies in larger series of patients with RE. AntiGluR3 antibodies, at first assayed by ELISA techniques, were found in serum and cerebrospinal fluid (CSF) not only in a proportion of patients with RE, but also in other forms of epilepsy, particularly in the catastrophic epilepsy of children. Therefore, their presence is not specific for RE but seems to correlate with epilepsy and its severity.³⁰^–³¹ However, the importance of antiGluR3 antibodies has been recently questioned by means of different laboratory techniques (ELISA, Western blot, immunoprecipitation, immunohistochemistry, and electrophysiology).³² Two different mechanisms by which antiGluR3 antibodies might exert their damage to the brain have been proposed. Plasma or serum of GluR3-immunized rabbits promoted the death of cortical neurons in vitro by a complement-mediated mechanism. Anti-IgG and anti-MAC antibodies were localized on neuronal cell bodies of cortex sections resected from RE patients,³³^–³⁵ leading to the complement activation hypothesis. Other authors reported that sera and IgG fractions of GluR3-immunized rabbits elicited rapid, reversible, and voltage-independent opening of cationic channels in cultured neurons, promoting excytotoxic damage.³⁶^–³⁷ Other antibodies have been described, such as to Munc-18 (a cytosolic protein), to NMDA-GluRε2 and anti-α7nAChR in a limited number of patients,³⁸^–⁴⁰ but their actual role is still to be elucidated. Nevertheless, the analysis of immunoglobulin gene rearrangement in brain samples from four RE patients provided evidence for a restricted profile.⁴¹

The possibility that the antibodies described could be secondary to neuronal damage cannot be excluded. However, negative results do not rule out the possibility that humoral immune mechanisms might be involved in RE, adding to the role of T cells.

The role of T cells in the pathogenesis of RE has been emphasized by several pathological studies. The main pathological features of RE include brain inflammation dominated by T cells, microglial activation, microglial nodules, neuronal loss, and astrogliosis in the affected hemisphere (Figure 12-3).^42,⁴³ The extent of T-cell infiltration and microglial activation is inversely correlated with duration of the disease, being more prominent in the earliest stages while in the latest gliosis predominates.⁴⁴ The majority of infiltrating T cells have been characterized as CD8⁺ cells containing granules positive for granzyme B, and part of these cells were observed in close contact with MHC class I positive neurons,⁴⁵ a set of features considered evidence of a cytotoxic T-cell reaction against neuronal cells. The same authors could not detect deposits of immunoglobulins or signs of complement activation in the same brain specimens. Recently, the importance of astrocytes has been investigated in detail in RE.⁴⁶ In RE, astrocytic apoptosis and loss were detected both in the cortex and the white matter; interestingly, as already observed by previous studies on T cells, granzyme B⁺ lymphocytes were found in close contact with astrocytes with granules polarized toward the astrocytic membranes. The authors therefore suggested that astrocytes might be a target for cytotoxic T cells, ultimately leading to astrocytic degeneration. Considering their multiple functional roles, degeneration of astrocytes might enhance neuronal loss as well as contribute to seizure induction and maintenance.

Figure 12–3 A, Hematoxylin and eosin-stained section showing cellular aggregates composed of inflammatory cells infiltrating the parenchyma. B, Glial fibrillary acidic protein (GFAP) staining showing an intense reactive astrogliosis in the cerebral cortex; section counterstained with hematoxylin. C, Hematoxylin and eosin-stained section showing a typical perivascular inflammatory cell cuff. Scale bar: A, 25 μm; B, C, 50 μm.

Courtesy of R. Garbelli and R. Spreafico.

Several aspects of the pathogenesis of the disease remain open to investigation. Evidence for the presence of cytotoxic T cells and their reactivity against neuronal and microglial cells is robust, but the question of their specificity and the peculiarity of unihemispheric involvement remain unanswered. On the other side, evidence for pathogenic autoantibodies has been suggested at first and then questioned by more recent studies. Nevertheless, the potential contribution of humoral immunity has not been definitely ruled out in RE, at least in a subset of patients. Despite the open controversies, immunopathogenetic data have been and remain the rationale for the different immunotherapeutic approaches investigated in recent years and still used in the treatment of RE.

Laboratory Findings Other Than EEG

NEUROIMAGING

The refinement of imaging techniques has highly contributed to the recognition of RE, even in the early stages of the disease, and better characterization of the type and sequence of brain pathologic changes (Figure 12-4).

Figure 12–4 MRI showing the progression of the disease in a patient with right-sided RE. Axial T-2w images taken at 10 days (A), 9 months (B), 12 months (C), and 18 months (D) after the onset show the progression of cortical atrophy and ventricular enlargement, spreading of abnormal signal and progression of atrophy of the basal ganglia. The disease mainly progressed during the first 18 months, although further damage occurred in the following years, as shown in axial T-2w (E) and FLAIR (F) images taken 8 years after the disease onset.

Reprinted with permission from Chiapparini L, Granata T, Farina L, et al. Diagnostic imaging in 13 cases of Rasmussen’s encephalitis: can early MRI suggest the diagnosis? Neuroradiology. 2003;45:171–183.

Recent studies focused on the early signs of the disease and demonstrated that within the first few months, most patients disclose a combination of changes that, albeit not diagnostic, are typical enough to raise the suspicion of RE.^4,⁴⁷ These early signs include, (1) mild focal cortical atrophy, always involving the insular and periinsular regions; (2) ipsilateral ventricle enlargement; (3) increased cortical and/or subcortical T2 and FLAIR signal; and (4) T2 hyperintensity or atrophy of the head of the caudate nucleus. The described changes are preceded, as evaluated in the few cases with a very early examination, by focal, transient, cortical swelling. Gadolinium enhancement is not observed.

During the progression of the disease, unilateral cortical and caudate atrophy progressively worsens, whereas signal abnormalities are less evident. In a study comparing MRI and histopathology findings, Bien and coworkers demonstrated that increased MRI signal correlates with an active inflammatory process, as witnessed by the high density of inflammatory cells and reactive astrocytes (i.e., with the presence of T cells, microglial nodules, and GFAP-positive astrocytes). Conversely, MRI atrophic areas without signal abnormalities correlate with tissue loss, which mostly occurs during the first year after the “acute disease stage.”⁴⁴ The sovratentorial tissue loss is accompanied by atrophy of the contralateral, but seldom global, cerebellar hemisphere of the brainstem and thinning of the corpus callosum.⁴⁷

The role of functional imaging, PET, SPECT, and MR spectroscopy (MRS) is still under study. These techniques may demonstrate, even in the very early stages, areas of abnormal metabolism (PET), or perfusion (SPECT), and reduced N-acetylaspartate (NAA), but are not useful in defining the suspected inflammatory nature of the disease.⁴⁸^–⁴⁹ Moreover, PET and SPECT findings are highly influenced by seizure activity; MRS may show increased lactate, also resulting from seizure activity, potentially misleading the diagnosis. For all these reasons, data obtained by these techniques must be cautiously interpreted and integrated with clinical and MRI findings. In clinical practice, the usefulness of PET, SPECT, and MRS should be limited to confirming the unilateral brain damage.

BLOOD AND CSF TESTS

Blood tests are usually unrevealing. The meaning of antiglutamate receptor 3 (GluR3) antibodies and other autoantibodies has been discussed in the pathogenesis section. It is only worth stressing here that the presence or absence of these antibodies should not be considered either a clue to diagnosis, or an exclusion criteria because it is now well established that GluR3 antibodies do not discriminate between RE and other epilepsies.

CSF may be normal or show nonspecific abnormalities, such as a mild increase in the white cell count and proteins. Oligoclonal bands are found in about half the patients.^4,⁵⁰ In summary, blood and CSF analysis are not suitable in confirming RE, but must be carried out to exclude CNS infections or other disorders.

BRAIN BIOPSY

Histopathology study is not required in all RE cases because other criteria allow the correct diagnosis, even in early stages, in most cases (see following discussion). It has to be considered, moreover, that histology shows nonspecific patchy chronic inflammatory changes, and that normal and abnormal tissue elements may be located in very close apposition.⁴²^–⁴³ Therefore, false-negative results may be obtained, particularly if small tissue samples are taken. In the few patients in which biopsy is required (Table 12-1), an open biopsy, in a noneloquent area where there is increased T2/FLAIR signal on MRI, should include meninges, gray, and white matter.⁴⁴

TABLE 12–1 Diagnostic Criteria for RE by a Two-step Approach

1° step: Check for the features of Part A.
2° step: If features of Part 1 are not fulfilled, check criteria of Part B.
Part A:
1. Clinical	Focal seizures (with or without epilepsia partialis continua) and unilateral cortical deficit(s)
2. EEG	Unihemispheric slowing with or without epileptiform activity and unilateral seizure onset
3. MRI	Unihemispheric focal cortical atrophy and at least one of the following: Grey or white matter T2/FLAIR hyperintense signal Hyperintense signal or atrophy of the ipsilateral caudate head

Part B: 1. Clinical Epilepsia partialis continua or progressive* unilateral cortical deficit(s) 2. MRI Progressive* unihemispheric focal cortical atrophy 3. Histopathology T-cell dominated encephalitis with activated microglial cells (typically, but not necessarily forming nodules) and reactive astrogliosis. RE can be diagnosed in presence of: all the three criteria of Part A or two out of the three criteria of Part B.

* Clinical progression: at least two sequential clinical evaluation must document a neurological deficit that must increase over the time. MRI progression: at least two sequential MRI studies must show hemiatrophy, that must increase over time.

Modified from Bien CG, Granata T, Antozzi C, et al. Pathogenesis, diagnosis and treatment of Rasmussen encephalitis. A European consensus statement. Brain. 2005;128: 454-471.

Diagnostic Clues and Differential Diagnosis

The diagnosis of RE rests on clinical, neuroradiological, and electrophysiological findings that, taken together, must strongly support the hypothesis of a strictly unilateral and progressive brain disorder. A rigorous diagnostic workup is needed to exclude alternative conditions and avoid delayed, but also precipitous, therapeutic decisions.

In patients with a long-lasting disease, the diagnosis is rather simple, and differential diagnosis is limited to a few conditions; conversely, the recognition of RE in its early stages, before progressive neurologic deficits and brain atrophy have developed, can be a challenge. In the last years, efforts have been made to identify the early manifestations of RE that could prompt a reasonably secure diagnosis and, consequently, an early targeted treatment.^4,⁹ Diagnostic criteria including typical and atypical RE have been recently reviewed by a European consensus (Table 12-1); however, it should be underlined that the correct diagnosis of RE needs a wide clinical experience supported by multidisciplinary expertise.

The proposed clinical criteria include a two-step diagnostic approach: (1) if all the highly characteristic clinical, EEG, and imaging criteria are fully met, the diagnosis can be established without the need for brain biopsy. In these patients the follow-up period, to ascertain the progressive course, may be very brief, or even unnecessary. (2) On the contrary, if the aforementioned criteria are not fulfilled, the progression of clinical deficits, assessed by serial evaluations, must be associated with progressive hemispheric atrophy, as documented by sequential MRIs or, alternatively, by the histopathologically proved T-cell–mediated encephalitis, with activated microglial cells and astrogliosis.

The diagnostic workup must exclude, on a clinical and/or laboratory basis, conditions characterized by unilateral neurologic syndromes or EPC, as well as inflammatory diseases mimicking RE. Among these conditions, the following must be considered: (1) focal or hemispheric dysplasia, neoplasia, and neurocutaneous syndromes (in particular Sturge-Weber syndrome), stroke, and hemiconvulsion-hemiplegia-epilepsy syndrome; (2) progressive neurometabolic or degenerative syndromes, namely MELAS and other mitochondriopathies, lipofuscinoses, and Alpers disease; (3) metabolic disorders, including type 1 diabetes mellitus, associated with anti-GAD 65-antibodies, and renal or hepatic encephalopathy; (4) cerebral vasculitis in systemic diseases, such as lupus erythematosus or Derry’s vasculitis; (5) rare infective disorders such as cat-scratch disease, but also HIV infection; and (6) finally, there are a few reports of severe epilepsy and EPC in bone marrow–transplanted patients; this possibility must be kept in mind giving the increasing indications to this treatment.

A summary of differential diagnosis with related clinical and laboratory investigations are summarized in the mentioned European consensus statement.¹

Treatment

RE is an inflammatory and most probably immunomediated disease, running a progressive course, characterized by severe epilepsy and worsening motor and mental deficits. These features must be kept in mind in the therapeutic decision process. The management of RE, at least theoretically, should therefore include the treatment of both seizures and the underlying inflammation. In addition, therapeutic efficacy must be assessed on both seizure reduction and neurologic improvement and weighted against side effects and potential residual deficits. Beside surgery, which is unanimously considered the only treatment able to stop seizures and disease progression (see following discussion), different drug treatments have been employed, based on the proposed etiopathogenesis of RE. Antiviral and interferon treatment was reported in a few papers in the nineties, but the paucity of data does not allow any conclusions on their effect. On the contrary, an increasing number of reports on the use of immunomodulatory treatments have accumulated in the last decade, indicating their potential role in selected patients and circumstances. The mechanisms by which these treatments may act in RE are probably complex and include, besides the antiinflammatory and immunomodulating effects, the regulation of BBB permeability (i.e., steroids) and a direct antiepileptic effect (steroids and IVIG).¹⁷

IMMUNOMODULATORY TREATMENTS

The rarity of disease, the variability of disease severity, the fluctuating course, and differences in treatment schedules employed by different centers, together with the lack of controlled clinical trials due to ethical reasons, make the evaluation of treatment efficacy very arduous. Nevertheless, case reports accumulated in the last years allow the evaluation of different treatments on the basis of which prospective studies have started. As a general statement, in the vast majority of cases, as per literature reports and personal experience, immunomodulation has only a transient and partial effect on seizure activity and progression of symptoms and should therefore be limited to patients that are not, or not yet, suitable for surgery. These include late-onset (adolescent or adult) RE with slower and milder course than the typical childhood-onset form; patients with dominant hemisphere involvement and slow progression when worsening of motor and language functions resulting from surgery are not accepted from the patient and family; suspected RE in which neurologic deterioration and hemispheric atrophy have not yet occurred; and proven or suspected bilateral RE. Nevertheless, the reports of transient efficacy of immunomodulation in some of these patients are in keeping with pathogenetic data suggesting an inflammatory brain process. In some of these patients, immunomodulation with corticosteroids, immunosuppressants, and IVIG or plasma treatment may be of help in the management of acute deterioration of the clinical picture or in slowing down its progression.

Corticosteroids are the most widely used, and in our personal experience the most effective therapy, not only as chronic treatment, but also in halting epileptic status or reducing the intensity of EPC. In the latter condition, methylprednisolone should be used in high-dose pulses (i.e., 400 mg/m²/day or, in children, 20 mg/kg/day tapered over 10 to 12 days).⁵⁰^–⁵¹ Some evidence exists that steroid treatment is more effective when started close to the onset of the disease.⁵¹^–⁵² As far as long-term steroid therapy is concerned, it has been recommended to start with boluses of intravenous methylprednisolone and then shift to 1 to 2 mg/kg/day of oral prednisone,⁵⁰^–⁵¹ to be slowly tapered to the minimal effective dosage.

IVIg efficacy has been reported in a limited proportion of children and adults with RE.⁵³^–⁵⁵ The recommended dosing scheme consists of five consecutive infusions of 0.4 g/kg/day followed by a monthly dose of 0.4 to 2.0 g/kg distributed over 1 to 5 consecutive days; the assessment of IVIG efficacy may require as long as six courses. The combination of steroids and IVIg may also be considered⁵⁰^–⁵¹ when the two treatments alone are ineffective.

Immunomodulation by Plasmapheresis (PE) or Protein A Immunoadsorption (PAI)

Plasmatic treatment has been used with the rationale of removing potentially pathogenetic circulating antibodies (namely anti GluR3). A dramatic effect in blocking status epilepticus and neurologic deterioration has been reported in a few patients,^28,⁵¹ whereas evidence of long-term efficacy has been only exceptionally reported.¹³ For this reason, as well as for technical issues (particularly in children) and costs, PEX or PAI should be reserved to phases of acute deterioration, or to assess the residual motor and mental abilities before surgery.

Immunosuppression Tacrolimus

The only trial so far reported on the use of immunosuppressive agents is that of Bien and coworkers.⁵⁶ The authors tested the effect of tacrolimus, a T-cell inhibiting drug, in seven RE patients and compared their outcome with that of twelve “historical” untreated patients. Despite no significant effect on seizure frequency, tacrolimus-treated patients had a better outcome in terms of motor and mental functions, as well as on the rate of brain atrophy, to the point that none of these patients was submitted to brain surgery.

Conflicting or limited results, hindering any conclusion, have been reported for cyclophosphamide and azathioprine.^28,^51,⁵⁷