Early predictors of refractory epileptic seizures

Transcription

1 Mohamed M.Mostafa et al. Early predictors of refractory epileptic seizures Mohamed M.Mostafa 1, Ali Akram 2, Mohamed Osman 3, Mamdooh Eissa 4 Depatments of Neurology, Ain Shams University 1, Psychiatry 2, Pediatrics 3, Radiodiagnosis 4, Cairo University ABSTRACT Epilepsy includes a markedly heterogeneous group of disorders and epileptic syndromes are frequently characterized by multiple seizure types and not infrequently complicated by mental sub normality and psychiatric manifestations. More than 30 percent of patients with epilepsy have inadequate control of seizures with drug therapy, but why this happens and whether it can be predicted is not so obvious. Prediction of refractory seizures has already been gaining interest and pioneering work in the field of early detection of refractory seizures has been advancing in the past few years with.with the approval of several new AEDs, Vagal nerve stimulation and Epilepsy Surgery (that appears to be the most promising) in the past few years, more treatment options are available for patients with intractable and refractory epileptic seizures. Recognizing readily definable, surgically remediable syndromes that have a universally poor prognosis with medical treatment but a good prognosis with surgical treatment is important.the study aimed at providing rational basis for early detection of truly intractable seizures with subsequent early intervention by surgery or other appropriate means. The study included 96 patients initially diagnosed as intractable with a follow up period of three months after modification of drug therapy and the criteria for intractability were identified. The patients were divided according to their seizure types and according to whether or not they remained refractory after three months. All patients had neurological history taking and examination, Digital EEG recordings at the onset and after three months and long term Video EEG recordings in addition to MRI scans. The study showed that that many cases that would seem refractory on the initial clinical presentation may in fact respond to drug therapy.the percentage of refractory patients responsive to novel antiepileptic medications was reported to be around 5 % which seems an irrational generalization particularly in case with generalized epilepsy. In or study we concluded that epileptic patients are markedly heterogonous and should be evaluated as such. Even the well known intractable syndromes and with all initial indicators pointed to a poor response to medical therapy and where surgery would be advised early, not infrequently the patients showed a favorable response to medications. This was particularly noted in patients with the generalized seizures of the Lennox Gastaut Type. However we totally agree with the current literature concerning the intractability of patients with partial seizures of temporal lobe origin, and also with the recently evolving concept of sending those patients to surgery as early as possible as delayed interference would result in only progression of clinical Symptomatology and functional damage as indicated by spread of epileptic spike activity to other areas of the hemispheres and to the opposite side both findings were indicators of a very poor response to therapy. (Egypt J. Neurol. Psychiat. Neurosurg., 2004, 41(1): ). INTRODUCTION The prevalence rate of epilepsy based on epidemiological studies from the United States and Europe is approximately %. More than 30 percent of patients with epilepsy have inadequate control of seizures with drug therapy 1 and hence the term refractory epilepsy. Definition of terms Refractory epilepsy is defined currently by many investigators as seizures that are not controlled after an adequate 385

2 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 trial with 2 first-line AEDs. The definition includes two aspects: the first aspect is the duration; although the term uncontrolled is used by many investigators to define the occurrence of an average of one seizure per month for at least a period of two years, a time interval of three months is often posed as the acceptable period of seizures being uncontrolled before being termed intractable.the second aspect is the drug regime; Some advocate at least 3 regimens, including 1 trial of 2-drug therapy. If 3 trials of monotherapy with first-line drugs are not successful, the chance that the patient will respond to a fourth drug as monotherapy or polytherapy is only 5%. 2 Another study by Kwan and Brodie revealed that only 47% of patients with untreated epilepsy became seizure free during treatment with the first AED, and 14% became seizure free with a second or third AED. 1 The seizure-free rate is similar in patients who are treated with a single established drug (67 percent) and patients who were treated with a single new drug (69 percent). 3 Importance of early recognition of truly intractable seizures. With the approval of several new AEDs, Vagal nerve stimulation and Epilepsy Surgery (that appears to be the most promising) in the past few years, more treatment options are available for patients with intractable and refractory epileptic seizures. Recognizing readily definable, surgically remediable syndromes that have a universally poor prognosis with medical treatment but a good prognosis with surgical treatment is important. These patients need to be identified early in life before the psychosocial consequences of prolonged disability prevent useful rehabilitation, even if the patient eventually undergoes epilepsy surgery 4. Unnecessary prolonged trials of AEDs in patients with welldefined, surgically remediable epilepsy (e.g., hippocampal sclerosis) delays surgical treatment that may reduce substantially or perhaps eliminate seizures and offers a diminishing chance for seizure freedom (5-10%) 5. On the other hand epilepsy surgery particularly that performed for generalized seizures as corpus callosotomy and multiple subpial resections is not without complications particularly on cognitive, behavioral, and practic and psychiatric aspects and should be advised with caution. It is therefore of equal importance to identify patients who on the grounds of the current literature would be expected to have intractable seizures but may show a favorable response to anticonvulsant therapy after several trials. Factors predictive of refractory epileptic seizures Age of onset: early age of onset particularly in infancy is considered as one of the early predictors of intractable seizures in many studies with the exception of benign neonatal familial convulsions. 6,7 Febrile and neonatal convulsions: A history of febrile convulsions mainly the recurrent focal convulsions and those occurring early in the neonatal and early infantile phase are closely related to the occurrence of mesial temporal sclerosis and the later development of intractable epileptic seizures in childhood. 6,8,9 Natal History disorders: history of prenatal, natal, and postnatal history disorders constitute important predictors of refractory seizures due to associated organic brain damage. 6 Mental Subnormality: intellectual impairment early in the course of epilepsy constitutes a significant predictor of refractory seizures. Mental subnormality constitutes an important clinical feature of most of the severe epileptic syndromes of infancy and childhood that are usually accompanied by refractory seizures. 10 Frequency of seizures: increased frequency of seizures and early presentation with multiple seizure types constitutes an early predictive factor for refractory seizures. Status epilepticus as an initial presentation and the occurrence of multiple admissions due to status is a feature of many of the clinically resistant epileptic seizures, a feature of many of the of the childhood and infantile refractory epileptic syndromes including the West syndrome, early 386

3 Mohamed M.Mostafa et al. infantile myoclonic epilepsy, epilepsy with myoclonic astatic seizures and Lennox Gastaut epilepsy among others. 6,9,10 Other factors: Long duration of seizures prior to presentation, family history of epilepsy and poor socioeconomic background are also potential predictive factors for refractory seizures. 11 Seizure types as predictors of refractory epilepsy Partial seizures As mentioned earlier 30 % of patients with epilepsy have intractable seizures. Half of these patients have partial seizures, and in 30% seizures are not controlled adequately with antiepileptic drugs (AEDs). Another 10-15% of patients with epilepsy have severe secondary generalized seizure disorders that do not respond to AEDs 2. Temporal Lobe seizures The syndrome of medial TLE, which is associated with hippocampal sclerosis, is an example of an intractable epileptic syndrome that is surgically remediable. Approximately 40-67% of these patients have a history of a complicated febrile convulsion (a febrile seizure lasting >30 min). These patients typically present with seizures in late childhood, at which time seizures are well controlled with AEDs. As the child enters adolescence and early adulthood, the seizures recur and become refractory to multiple medication trials. 8 In Extratemporal Lobe epilepsy Refractory Partial seizures may also be treated effectively with alternative therapies including epilepsy surgery and vagal nerve stimulation. Extratemporal seizures including frontal lobe epilepsy 12 parietal lobe epilepsy 13 and occipital lobe epilepsy 14. Usually have accompanying pathological lesions including low-grade gliomas: developmental tumors such as gangliogliomas and dysembryoplastic neuroepithelial tumors, arteriovenous malformations (AVMs), cavernous angiomas, encephalomalacia, focal cortical dysplasia,porencephalic cysts 15 (commonly associated with a perinatal ischemic insult) and schizencephaly (a cleft lined with gray matter and extending from the pial surface to the ventricle) 16. Epilepsy surgery is particularly effective when a clearly defined lesion is present on highresolution MRI. In fact, surgical outcome improves from 20% seizure free in patients without a lesion to 70% seizure free in patients with a lesion. 17 Patients with multiple seizure types are usually refractory to therapy and frequently excluded from surgery but may be candidates for alternate therapies, unless one seizure type is the most frequent and disabling. Clinicians must be cautious because some patients have a clinical history that indicates multiple seizure types while video-eeg longterm monitoring of the patient reveals a single habitual seizure type. 18 Patients with an idiopathic (i.e., genetic) epilepsy, such as benign rolandic epilepsy or benign childhood epilepsy with occipital paroxysms are not truly refractory since they almost always remit by adolescence. 18 Generalized seizures In patients with generalized epilepsy the prevalence of intractable seizures is higher in patients with symptomatic or cryptogenic epilepsy than in those with idiopathic epilepsy (40 percent vs. 26 percent, P=0.004) 2. Mattson et al. reported that 60-70% of patients with generalized tonic-clonic seizures alone attained seizure freedom for at least 12 months. In patients with complex partial seizures, on the other hand, the prognosis was less favorable, with only 23-26% attaining seizure freedom for at least 12 months. Interictal EEG recordings as Predictors of refractory Seizures. In intractable partial epilepsy, patients with temporal lobe epilepsy (TLE) have epileptiform activity consisting of spikes and/or sharp waves that are usually maximal at the anterior temporal (F7 and F8 electrodes) and the mid temporal regions (T3 and T4 electrodes). Although 20-30% of patients with refractory TLE have bilaterally independent temporal epileptiform activity, most 387

4 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 of these patients have a single or predominant localization for their habitual seizures. 20 Occasionally, patients with extra temporal epilepsy of occipital or frontal lobe origin have interictal epileptiform activity at the temporal region, even though seizures do not localize to that region. Patients with occipital lobe epilepsy may not have any occipital epileptiform discharges. In fact, these patients may have only epileptiform discharges in the frontal or temporal regions. 21 Patients with frontal lobe epilepsy may have predominantly bilaterally synchronous or generalized epileptiform activity due to secondary bilateral synchrony (Bautista) RE, this EEG pattern is observed most commonly in patients with mesial frontal lobe epilepsy. EEG features that help in lateralization of secondary bilateral synchrony include (1) consistent phase reversal over one region, (2) higher amplitude of generalized or bilaterally synchronous interictal epileptiform activity over one hemisphere, (3) a consistent lead in one hemisphere, and (4) persistent lateralized interictal slowing. 22 Multifocal independent interictal epileptiform abnormalities do not invariably suggest a poor prognosis. 23 In generalized refractory epilepsy Children with refractory infantile spasms may show hypsarrhythmia, which is defined as a highly disorganized background with very high amplitude activity and multifocal independent spikes. Infants with lateralized findings, such as focal loss of fast activity or focal slowing and an MRI scan showing a focal abnormality, may be candidates for intractable seizures. PATIENTS AND METHODS The study included a total of 96 patients with the initial diagnosis of Refractory Epilepsy visiting the outpatient Neurology clinic at DR.Soliman Fakeeh Hospital in Jeddah (KSA) during the period from February 1999 to September 2000.All patients included in the study regardless the type of seizures nor the underlying abnormalities. All patients were subjected to: Thorough history taking from both the patient and available eye witnesses for the seizures. This included special emphasis on the factors predictive of refractory seizures including: age of onset, history of prenatal, natal, and postnatal disorders, febrile and neonatal convulsions, associated organic brain damage and mental subnormality particularly intellectual impairment early in the course of epilepsy, frequency of seizures, early presentation with multiple seizure types, status epilepticus as an initial presentation and the occurrence of multiple admissions due to status, long duration of seizures prior to presentation, family history of epilepsy and poor socioeconomic background. A special emphasis was laid on the type of seizures whether generalized or partial and on the seizure semiologies and identification of specific syndromes known to be refractory to therapy and other epileptic syndromes of a benign nature. Meticulous general and neurological clinical examination particularly searching for subtle neurological signs and abnormalities of cortical functions. Digital interictal EEG Recordings: Awake EEG recordings or Sleep EEG recordings in neonates, infants and children either by natural sleep or by sedation induced by oral chloral hydrate 40 mg per kg body weight using a Nihon Koden Digital Neurofax machine with 20 channels EEG recordings with simultaneous EKG, eye movement electrodes recordings for monitoring of artifacts. Electrodes were placed according to the international system for electrode placement with readings using monopolar (referential) and bipolar montages.follow up EEG recordings were again performed at 3 months and reevaluated. Long Term Video EEG recordings were obtained using a Nicolet Video EEG and 388

5 Mohamed M.Mostafa et al. polysomnography machine operating with a Nic-Vue program for EEG recording and data compression and viewing. All long-term video EEG recordings were performed for hours during the awake state, at rest with eye closure and eye opening. This was followed by a sleep recording of 6-8 hours (natural sleep or by sedation using oral choral hydrate medication). Provocation was performed by sleep deprivation on the day prior to long-term Video EEG recordings and a gradual withdrawal of anticonvulsants in the 14 days period prior to the recordings. Provocation during the recordings was performed by hyperventilation for a period of at least 3 minutes and photic stimulation. Electrodes were placed according to the international systems for electrode placement with readings using monopolar (referential) and bipolar montages. Additional electrodes were placed for eye movements, EKG recordings and surface EMG from the tibial and deltoid muscles. Two separate analyzers who were blindfolded to the results of conventional EEG and MRI results performed analysis of data. Magnetic Resonance Imaging (MRI) was performed using a Siemens MRI machine 1.5 tesla with a head coil. Conventional axial T1, T2 and Inversion recovery (FLAIR)images were obtained with a slice thickness of 5 mm followed by T2 images and FLAIR coronal images with a slice thickness of 3mm on the temporal lobes to detect any signal changes in the hippocampus or other cortical regions. We studied these patients with the diagnosis of epilepsy with the aim of identifying the factors that were associated with intractable seizures. Those patients were selected with the provisional diagnosis of intractable seizures on their initial presentation based on the current literature reviews. We studied their medication regimens thoroughly and their compliance on treatment regimens throughout a period of 3 months. We evaluated their description of seizures, their conventional EEG recordings and their longterm video EEG recordings and reclassified their seizures in accordance with the recent recognized classification of the International League against Epilepsy (ILAE). We modified their drug regimens accordingly and ensured compliance. We adjusted their serum levels and introduced recent antiepileptic medications. We re-evaluated the same patients after the 3 months period and classified them first according to their seizure semiology. We reclassified them again into a Refractory group and a Responsive group based on the occurrence of seizures. The refractory group included patients with more than one seizure per month while the responsive group included patients with no or only one seizure per month as noted by their relatives. We compared the two groups using student t test and chi square analysis for comparison between variables and a computerized program derived from the Statistics Home Page analyzed the results. RESULTS Our study included 96 patients who were selected with the provisional diagnosis of probably intractable seizures based on the frequency of seizures on their initial visit and on their history of seizures and antiepileptic medications and their compliance. Our group of patients included 46 patients (47.9%) with generalized seizures and 50 (52.1%) patients with partial seizures (Table 1). Generalized Seizures Among the patients with generalized seizures 12 patients (26.08 %) had Generalized tonic clonic seizures,6 (13.04) patients had absence seizures and 3 patients (6.5 %) had juvenile absence seizures. Of our patients 9 patients (19.5 %) had West syndrome while 12 patients (26.08 %) were diagnosed as Lennox Gastaut epilepsy and 4 patients (8.6%) had myoclonic astatic seizures. Table (1) displays these data. 389

6 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 Predictive factors for refractory epilepsy in generalized seizures In generalized tonic clonic seizures only 4 patients (33.3%) remained with refractory seizures while 8 patients (66.6 %) showed a favorable response on modification of drug therapy and ensuring compliance. There was no statistically significant difference between the refractory group and the responsive group regarding the mean age of onset (3 years and 4.6 years p >0.05) nor the mean number of seizures per day (3.34 and 2.38, p<0.05) nor the initial presentation with status epilepticus (1 versus 2 p<0.05).the mean duration of seizures prior to presentation was 3.14 months in the refractory group and 3.58 months in the responsive group (p>0.05). Two patients in the refractory group had a strongly positive family history of epilepsy compared to one patient in the responsive group (p <0.05). History of perinatal disorders with significant perinatal anoxia and low Apgar score at birth and mental subnormality with variable degrees of mental delay was present in all 4 patients with refractory seizures and in only one patient with responsive seizures (p<0.05) In Absence and Juvenile Absence seizures: There were 9 patients and all were responsive to modification of drug therapy and ensurance of compliance,mostly being diagnosed initially as complex partial seizures and were given antiepileptic therapy on the basis of that initial diagnosis. Modification of therapy resulted in complete responsiveness of these cases, non being refractory at the 3 months study period. West syndrome: The results are shown in table (2): 6 patients (66.6 %) remained with refractory seizures while 3 patients (33.3%) showed a favorable response on modification of drug therapy introduction of novel antiepileptic medications and ensuring compliance. There was a statistically significant difference between the refractory group and the responsive group regarding the mean age of onset (3 months and 8 months p<0.05), the mean number of seizures per day (13.45 and 6.71, p<0.05) and the initial presentation with status epilepticus (4 versus 1 p<0.05). There was no statistically significant difference between the refractory group and the responsive group regarding the mean duration of seizures prior to presentation (1.032 months in the refractory group and 2.135months in the responsive group p>0.05)nor the family history of seizures ; 2 patients in the refractory group had a strongly positive family history of epilepsy and also two patients in the responsive group had family history of epilepsy(p <0.05). There was also no statistically significant difference between the refractory group and the responsive group regarding the history of perinatal disorders [with significant perinatal anoxia and low Apgar score at birth was present in 3 patients with refractory seizures and in one patient with responsive seizures (p>0.05)] nor the history of delayed milestones that was present in all 9 patients with either refractory or responsive seizures. There was also no statistically significant difference between the refractory group and the responsive group regarding the early presentation with multiple seizure types that was a common feature in both groups [6 patients in the refractory group and 2 patients in the responsive group (p >0.05)]. Lennox Gastaut epilepsy: The results are shown in table (2): 4 patients (33.3 %) remained with refractory seizures while 8 patients (66.6%) showed a favorable response on modification of drug therapy introduction of novel antiepileptic medications and ensuring compliance. There was a statistically significant difference between the refractory group and the responsive group regarding the history of perinatal disorders (with significant perinatal anoxia and low Apgar score at birth present in 4 patients with refractory seizures and in 2 patients with responsive seizures (p<0.05) ), the mean number of seizures per day (18.5 and 5.81, p <0.05), the early presentation with multiple seizure types (4 patients in the refractory group and 8 patients in the responsive group p<0.05) and the initial presentation with status epilepticus (3 versus 1 p<0.05). 390

7 Mohamed M.Mostafa et al. There was no statistically significant difference between the refractory group and the responsive group regarding the mean age of onset (2.08 and 2.01 years p <0.05), the mean duration of seizures prior to presentation (6.4 months in the refractory group and 8.3 months in the responsive group p>0.05) nor the family history of seizures ; 1 patient in the refractory group had a positive family history of epilepsy and two patients in the responsive group had family history of epilepsy (p<0.05). There was also no statistically significant difference between the refractory group and the responsive group nor the history of delayed milestones that was present in all 12 patients with either refractory or responsive seizures. Myoclonic Astatic Seizures: Only 4 patients had this relatively uncommon epileptic syndrome, 2 patients were still refractory after modification of the antiepileptic therapy while the other 2 patients showed a favorable outcome. There was a statistically significant difference between both groups regarding the mean frequency of seizures per day and these were 12.4 versus 4.3 (p<0.05) and the mean duration of seizures prior to presentation (5 months versus 11 months). Other predictive factors showed no statistical significance between the 2 groups (Table 2). Interictal Digital and long term video EEG findings West Syndrome: all of our 9 patients had EEG abnormalities. There was no difference between the refractory and responsive group regarding the background abnormalities with hypsarrhythmia in all patients. However 4 patients in the refractory group showed persistence of the grossly disorganized pattern while only 1 patient showed a persistence of the grossly disorganized pattern 3 months after modification of drug therapy and introduction of novel antiepileptic medications. The difference was statistically significant. Generalized spike and slow wave and high voltage slow wave discharges occurred in both groups in all patients. Persistent generalized discharges showing no change in rate or character or frequency occurred in 5 patients in the refractory group and in only 1 patient of the responsive group. The difference was statistically significant (p<0.05). There was no statistically significant difference between both groups regarding the spike and slow wave activity nor the occurrence of semi periodic discharges. Lennox Gastaut epilepsy: The results are shown in table (3). There was no difference between the refractory and responsive group regarding the background abnormalities with 4 patients in the refractory group and 7 patients in the responsive group showing variable degrees of background slowing. 4 patients in the responsive group had a predominant theta dysrhythmia compared to 1 patient of the refractory group. The difference was statistically significant (p<0.05). There was no difference between the refractory and responsive group regarding the spike and slow wave discharge s nor the generalized high voltage slow wave discharge nor the occurrence of semi periodic discharges as shown in table (3). Myoclonic Astatic Seizures: There was no statistically significant difference between the refractory and responsive group regarding their EEG findings as shown in table (3) except for the persistence of spike and slow wave discharges in repeated EEG recordings and video EEG recordings with no change in frequency in the 2 patients while the 2 patients who responded to therapy showed less frequent EEG discharges on their follow up EEG recordings. Partial Seizures Among the patients with partial seizures 38 patients (76%) had complex partial seizures of temporal lobe origin as noted by their symptoms and described aura while 12 patients (24%) had complex partial seizures of extratemporal lobe origin including 10 patients with frontal lobe seizures and 2 patients with occipital lobe seizures. 391

8 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 Predictive factors for refractory epilepsy in Partial Seizures Temporal lobe seizures: The results are shown in table (4). 24 patients (66.6%) remained with refractory seizures while 8 patients (33.3%) showed a favorable response on modification of drug therapy, introduction of novel antiepileptic medications and ensuring compliance. There was a statistically significant difference between the refractory group and the responsive group regarding the early presentation with multiple seizure types (22 patients in the refractory group and 2 patients in the responsive group p<0.05), the association with mental subnormality (16.6% versus 7.1% p<0.05) and the history of antecedent febrile convulsions (41.6% versus 14.2% p<0.05). There was no statistically significant difference between the refractory group and the responsive group regarding the mean age of onset (13.26 and 7.51 years p>0.05), the mean duration of seizures prior to presentation (4 months in the refractory group and 6 months in the responsive group p>0.05, the family history of seizures; 5 patients in the refractory group had a positive family history of epilepsy and 5 patients in the responsive group had family history of epilepsy (p>0.05) nor the history of perinatal disorders [with significant perinatal anoxia and low Apgar score at birth present in 4 patients with refractory seizures and in 2 patients with responsive seizures (p>0.05)]. There was also no statistically significant difference between the refractory group and the responsive group regarding the, the mean number of seizures per day (7.12 and 4.65, p>0.05) nor the early presentation with status epilepticus. Interictal digital and long term video EEG recordings The results are shown in table (5). There was no difference between the refractory and responsive group regarding the background activity although more patients in the refractory group had abnormal background activity (20.8% versus 7.1%). On follow up those patients showed persistence of the background abnormality in 16.6% and 7.1% respectively again the difference I number was non significant. Normal EEG with absent epileptic discharges and normal long-term video EEG were detected in 33.3% and 35.7% of patients with refractory and responsive cases respectively. Unilateral anterior temporal spikes were recorded in 25% of the refractory group and 35.7% of the responsive group while unilateral midtemporal spikes were recorded in 16.6% and 21.4% respectively with the difference in number being non significant. There was a statistically significant difference between the refractory and responsive group regarding the occurrence of bitemporal spikes (45.8%) and (14.2%) p<0.05, in the spread of epileptic activity to the frontal, parietal and occipital regions ie involving one cerebral hemisphere (25%) versus (7.1%) p<0.05 and in the occurrence of secondary generalized epileptic activity (33.3%) versus (14.2%) p<0.05. Extratemporal lobe seizures: The results are shown in table 4: 8 patients (66.6%) remained with refractory seizures while 4 patients (33.3%) showed a favorable response on modification of drug therapy, introduction of novel antiepileptic medications and ensuring compliance. There was a statistically significant difference between the refractory group and the responsive group regarding the history of perinatal disorders (50 % in the refractory group versus 25 % in the responsive group, the associated mental subnormality (50 % versus 25% p<0.05) and the mean frequency of seizures per day (10.42 versus 4.79 p<0.05). There was no statistically significant difference between the refractory group and the 392

9 Mohamed M.Mostafa et al. responsive group regarding the mean age of onset (7.23 years and 8.65years p >0.05), the mean duration of seizures prior to presentation (5.31 months in the refractory group and 8.65 months in the responsive group p>0.05, the family history of seizures; 12.5% patients in the refractory group had a positive family history of epilepsy and none of the patients in the responsive group had family history of epilepsy (p>0.05). There was also no statistically significant difference between the refractory group and the responsive group regarding the early presentation with status epilepticus (none of the patients in either group presented with status epilepticus) nor the antecedent febrile convulsions (only one patient within the refractory group had a history of complicated febrile convulsions). Interictal digital and long term video EEG recordings There was a statistically significant difference between the refractory and responsive group of patients regarding their background activity on their initial EEG recordings. Only 2 patients in the refractory group had a normal background activity (25%) with background abnormalities in 6 patients (75%) compared to 4 patients (100%) with a normal background activity in the responsive group p<0.05. There was also a statistically significant difference between both groups regarding the epileptic spike activity. Unilateral epileptic spike activity was noted on the frontal, parietal or occipital regions in 25% of patients in the refractory group and in 50 % of patients in the responsive group and the difference was not statistically significant p>0.05.unilateral temporal spikes were noted in 25% of patients of the refractory group and in none of the patients in the responsive group with a p value of <0.05 and bilateral discharges were noted in 75% of patients of the refractory group and in 25% of the patients in the responsive group with a p value of <0.05 both being statistically significant (Table 5). Table 1. Types of seizures in our patients. Type of seizure No. of patients % Generalized Seizures Generalized tonic clonic seizures Generalized Absence Seizures Juvenile Absence seizures West Syndrome Lennox Gastaut epileptic Seizures Myoclonic Astatic epilepsy Partial Seizures Complex partial seizures of temporal lobe origin Complex partial seizures of extratemporal lobe origin

10 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 Table 2. Predictive factors for refractory epilepsy in generalized seizures. Mean age of onset Perinatal disorders Febrile Convulsions Mental subnormality, Mean frequency of seizures /day Early presentation with multiple seizure types Status epilepticus as an initial presentation Mean duration of seizures prior to presentation mos Family history of epilepsy 394 West s. 6 pts. West s. Resp. 3 pts P value LGS 4 pts LGS Resp. 8 pts P value MAE 2 pts MAE Resp. 2 pts P value 3 mos 8 mos < yrs 2.01 yrs > yrs 4.33 yrs > > < > > > < < < > < > < < > > mos mos 2 2 < > Table 3. Interictal digital and long term video EEG recordings in patients with generalized seizures. West s. 6 pts. West s. Resp 3 pts P value LGS 4 pts LGS Resp. 8 pts P value MAE 2 pts MAE Resp. 2 pts <0.05 P value Normal > >0.05 background Disorganized > Background Theta < dysrhythmia Theta delta > dysrhythmia, Hypsarrhythmia Persistent abnormal background (no change) 4 1 < < Abnormal discharges Generalized spike-slow waves<2.5 cps Generalized spikes Generalized high voltage slow waves Semi - Periodic generalized discharges Persistent generalized discharges no change) > > > > > > < > < < <0.05

11 Mohamed M.Mostafa et al. Table 4. Predictive factors in patients with partial seizures. Mean age of onset (yrs) Perinatal disorders Febrile Convulsions Mental subnormality, Mean frequency of seizures /day Early presentation with multiple seizure types Status epilepticus as an initial presentation Mean duration of seizures prior to presentation mos Family history of epilepsy Temporal Lobe Seizures Temporal Lobe Seizures Extratemporal Lobe Seizures Extratemporal Lobe Seizures Resp. P value 24 Resp. 14 P value p> p>0.05 4! p> P< P< p> P< P< p> P < P< p> p> mos 6 mos p> mos 8.65 mos p> p>

12 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 Table 5. Interictal digital and long-term video EEG recordings in patients with Partial seizures. Temporal Lobe Seizures 24 Normal background 19 (79.1 %) Disorganized 5 Background (20.83 %) Persistent abnormal 4 background (no (16.66%) change) Absent epileptic 8 activity on EEG (33.3) Absent epileptic 3 activity on long term (12.5%) video EEG recordings Unilateral anterior 6 temporal spikes (25%) Unilateral midposterior 4 temporal (16.6%) spikes Bilateral temporal 11 spikes (45.8%) Unilateral Epileptic 6 spikes frontal, (25%) parietal or occipital Bilateral Epileptic spikes frontal, parietal or occipital Secondary generalized Epileptic discharges Temporal Lobe Seizures Resp (92.8) 1 (7.14 %) 1 (7.14%) 5 (35.7%) 4 (28.4%) 5 (35.7%) 3 (21.4%) 2 (14.2%) 1 (7.1%) P value Extratemporal Lobe Seizures 8 p> (25%) p> (75%) > (50%) Extratemporal Lobe Seizures Resp. p> (25%) 1 (25%) p> (50 %) P value 4 4 <0.05 (100 %) 0 < <0.05 >0.05 <0.05 P> > (25%) 0 <0.05 < < (25%) (75%) 8 (33.3%) 2 (14.2%0 < (62.5%) 2 (50 %) 1 (25%) 3 (75%) >0.05 <0.05 >0.05 DISCUSSION More than 30 percent of patients with epilepsy have inadequate control of seizures with drug therapy and with the introduction of novel therapeutic modalities the identification of these patients has gained even more importance.in the last 10 years, particularly epilepsy surgery has been recognized increasingly as a viable treatment for patients with medically refractory seizures and apparently a debate is ongoing regarding how early should patients be sent for surgery. Many primary care providers and neurologists believe that epilepsy surgery is a last-resort treatment, such that many patients who could become seizure free with surgery undergo treatment with multiple medications over many years, suffering the adverse effects of recurrent seizures, the side effects of AEDs, and the psychosocial and occupational consequences 396

13 Mohamed M.Mostafa et al. of recurrent seizures. On the other hand epilepsy surgery is a major brain surgery that carries its own risks, morbidity and mortality not to mention some of the long term adverse effects of the surgery itself particularly on the psychiatric and behavioral and various aspects of cognition that may occur despite adequate presurgical evaluation and meticulous preparation. It was therefore e the aim of the study to provide some rational basis for prediction of refractory cases of epilepsy and hence provide basis for early referral to surgery and also basis for denying epilepsy surgery for patients who would presumably become responsive to medical therapies. The indicators for refractory seizures were variable and in we could identify some true indicators for poor response to drug therapies. Generalized epilepsy In generalized tonic clonic seizures the history of perinatal anoxia and the occurrence of developmental delay in milestones and mental subnormality were significant predictors of intractability. EEG abnormalities of significance included a disorganized background activity and frequent spike and slow wave discharges on long term video EEG recordings. MRI predictors included the occurrence of multiple and extensive hypo intense patches arborizing in the subcortical white matter and extending to the cortex in a diffuse pattern. None of our patients with primary absence seizures proved refractory to drug therapy. West Syndrome the patients with this syndrome were in general refractory to therapy as is well recognized in the literature. Predictors for intractability included a young age of onset around 3 months for the start of seizures, a mean frequency of seizures as noted by the patients to be around 11 major seizures daily, and the initial presentation by a non convulsive status epilepticus. The presence of multiple seizure types, perinatal anoxic brain damage, the occurrence of mental subnormality, the presence of a positive family history and the mean duration of seizures prior to presentation were all features of the syndrome and id not necessarily imply intractability as they occurred in both the responsive and refractory patients. It was a characteristic feature of these patients to be controlled only by at least three antiepileptic medications including the novel antiepileptic drugs. Lennox Gastaut Epilepsy: In our sample of epileptic patients 66 % of patients with Lennox Gastaut epilepsy, initially diagnosed as refractory on their first visit showed appropriate response to therapy with administration of novel antiepileptic dugs. 33% of our patients showed absolutely no response on drug modifications and remained intractable after 3 months of extensive trials of drug modifications. Factors associated with intractability were again history of perinatal anoxia and marked delay in the early phases of development. Another important predictive factor was the number of seizures per day as rated by the relatives and noted on long term video monitoring of patients. In our study the refractory group of patients had a mean of 18 seizures per day and those included tonic, atonic as well as gross myoclonic seizures in addition to prolonged absence seizures. In the responsive group the mean number of seizures on presentation was much less with a mean value of 5.8 seizures per day. Another clinical predictor for associated with refractory cased was the occurrence of multiple seizure types in the same patient with a wide diversity of seizures early at the onset of the disorder contrary to patients in the responsive group who had a predominant type of seizures early from the onset with non frequent multiple seizure complexes. Another predictor of intractability was the early presentation of patients in the refractory group with non convulsive status epilepticus where those patients spent most of their days in a twilight dreamy state with continuous EEG discharges on long term monitoring which were recorded in 12.5% of the patients in the refractory group and carried a grave overall prognosis. We reported that some 397

14 Egypt J. Neurol. Psychiat. Neurosurg. Vol. 41 (1) - Jany 2004 predictive factors claim to be associated with a poor prognosis prove to be of little significance in our sample of patients. The mean age of onset of seizures, the presence of delayed milestones of development and the relevant family history were present in the responsive as well as the refractory cases with a non significant difference indicating that they were of less importance in predicting intractability. The EEG predictors of intractability were the occurrence of gross background abnormities with a persistent predominant delta theta activity during the wakefulness state as recorded by long term video EEG recording together with the lack of significant changes between the sleep and wakefulness states on video EEG monitoring. Of greater significance even was the persistence oft this background abnormality on the follow up EEG recordings at 3 months. In this context we noted that Long Term Video EEG recordings were superior to the conventional EEG in demonstrating the continuous background abnormality even in the absence of frequent seizures. The presence of spike and slow wave discharges and high voltage slow waves (and polyspikes) were not by themselves predictors of intractability as they are essential features of the syndrome, occurring in both responsive and refractory cases, however it is again the persistence of these abnormal discharges with the same character and frequency on long term EEG monitoring on follow up that predicts intractability. MRI abnormalities taken as indicators of structural brain damage were also regarded as predictors of intractability and poor prognosis, yet in our sample of patients the MRI abnormalities were not all of a major significance. The most significant abnormality on our sample associated with intractable cases was the occurrence of extensive whit matter patches arborizing as hyperintensities in the subcortical white matter and extending both near to the diencephalic structures and the sub cortex of the prefrontal lobes. These were best delineated by FLAIR images. The presence of tuberous Sclerosis with its characteristic MRI changes were not in particular a predictor of intractability on the short term as noted from our study. Partial Epilepsy Complex partial seizures of temporal lobe origin constituted 76% of our patients with refractory partial epilepsy with 38 patients included in the study. We included patients with a diverse Symptomatology and different clinical presentations and the predictors of refractory epilepsy were widely variable. Of our 38 patients who were predicted to be refractory on presentation 66 % proved to be truly refractory at three months while 33 % responded to therapeutic modifications with satisfactory control of their seizures. The high percentage of refractory cases agrees with the current consensus in the literature concerning temporal lobe seizures. Predictors of refractory cases included clinical features with a significantly higher percentage of these patients having multiple seizure types within the domain of temporal lobe epilepsy early at the onset of their illness compared to the responsive cases. Another predictor was the early occurrence of personality changes, social and behavioral deterioration. A third association with intractability was indeed the well recognized antecedent history of repeated febrile convulsions in the first 5 years of life which was replicated in our study associated with intractable temporal lobe seizures. Interictal EEG conventional EEG and video EEG monitoring were valuable in detecting temporal spike foci and in detecting multiple Interictal spike foci and in relating clinical Symptomatology to spike epileptic activity. Although Normal EEG activity was present in 33% of our cases both in the intractable group and in the refractory group in agreement with the current literature reviews, with long term video EEG recordings the percentage of normal EEG recordings dropped to only 12% in the refractory group and % of 398

15 Mohamed M.Mostafa et al. the responsive group. Predictors of refractory seizures included the occurrence of bilateral spike epileptic activity on both temporal lobes, which were significantly more frequent in refractory patients than responsive patients. Unilateral temporal epileptic activity occurred in both the refractory and responsive patients with no significant difference. Although anterior and midtemporal discharges were reported to be more frequently associated with refractory seizures yet those results were not replicated in our study.the site of unilateral epileptic activity was of little predictive value for refractory seizures. Another EEG abnormality of significance was the occurrence of associated spike activity on the frontal, parietal and occasionally occipital lobe abnormalities indicating spread to the adjacent hemispheric areas. Our EEG findings agree with the works of Burnstine and others who reported the significance of Interictal discrete multifocal independent spikes in patients with refractory seizures. MRI abnormalities were not as frequent as presumed to be with only 13.1% of our patients showing definite mesial temporal sclerosis and 7.6% showing hippocampal atrophy while 18.4% showed congenital epidermoid cysts in the temporal pole. All 5 patients with mesial sclerosis in our study proved intractable so were the 3 patients with hippocampal atrophy while only 3 patients with congenital epidermoids were intractable, a finding in agreement with most of the literature reviews. We noted a significant increase in the mean age of onset in patients with refractory epilepsy and mesial temporal sclerosis, an indicator that the older the age of onset the more the trend towards intractability contrary to idiopathic forms of epilepsy. In extratemporal lobe seizures the predictors for refractory seizures were indicated by the more prevalent history of perinatal disorders in the in the refractory group, the association with mental subnormality and the mean number of seizures per day ranging around 10 seizures in the refractory group. The interictal EEG predictors of refractory seizures were the presence of a persistently disorganized background activity, the presence of bilateral spike epileptic activity and the spread of the epileptic spikes to involve the temporal lobes. The most significant MRI predictors of refractory seizures in our study were the presence of cortical areas of encephalomalacia representing early perinatal damage possibly through hypoxia or early perinatal infections or vascular insults. Conclusion Finally we can conclude that many cases that would seem refractory on the initial clinical presentation may in fact respond to drug therapy. The percentage of refractory patients responsive to novel antiepileptic medications was reported to be around 5% which seems an irrational generalization particularly in case with generalized epilepsy. In or study we concluded that epileptic patients are markedly heterogonous and should be evaluated as such. Even the well known intractable syndromes and with all initial indicators pointed to a poor response to medical therapy and where surgery would be advised early, not infrequently the patients showed a favorable response to medications. This was particularly noted in patients with the generalized seizures of the Lennox Gastaut Type. However we totally agree with the current literature concerning the intractability of patients with partial seizures of temporal lobe origin, and also with the recently evolving concept of sending those patients to surgery as early as possible as delayed interference would result in only progression of clinical Symptomatology and functional damage as indicated by spread of epileptic spike activity to other areas of the hemispheres and to the opposite side both findings were indicators of a very poor response to therapy. 399