General conclusions and recommendations

As reviewed by the several papers that comprise this supplement, the diagnosis, prognosis, and treatment of a first epileptic seizure are based on substantial scientific evidence, which, in the light of robust and consistent findings, provide the basis for developing rational recommendations for the treatment and management of this condition. The following summarizes the key points made by the contributors to this supplement and, based on those points, makes recommendations for accurate diagnosis and rational management of a patient who presents with a first unprovoked seizure. These recommendations are in line with a recent guideline devised by the Italian League Against Epilepsy (Beghi et al., 2006) and with the Practice Parameters of the American Academy of Neurology (Hirtz et al., 2000, 2003), which can be easily applied to the present context. In a patient presenting with a first epileptic seizure, other clinical disorders must be accurately excluded. These include other neurological conditions (transient ischemic attacks, migraine, narcolepsy, and restless legs syndrome), cardiovascular disorders (syncope and other cardiac sources of loss of consciousness), endocrine/metabolic disorders (hyponatremia, hypoglycemia, hypokalemia), sleep disorders, and psychogenic nonepileptic seizures. Except for some symptoms accompanying or following the seizure (cyanosis, hypersalivation, tongue biting, postictal disorientation) (Hoefnagels et al., 1991), the history and the clinical examination do not allow a confident diagnosis and still need assessment in the light of controlled studies. Except for hyponatremia in infants aged less than 6 months (Farrar et al., 1995), metabolic disorders (and toxic agents) are rarely found in children and adults screened after a seizure (Turnbull et al., 1990). The diagnostic yield of the EEG in the presence of a first seizure is substantial, as on average 50% of patients have abnormal tracings. An EEG done in the first 24 h after the seizure has a greater probability of detecting epileptiform abnormalities than an EEG done in the subsequent days (King et al., 1998). The additional yield of sleep-deprived EEG and sleep EEG is uncertain (King et al., 1998; Schreiner & Pohlmann-Eden, 2003). Structural brain abnormalities, as detected by neuroimaging, may be found in up to one-half of adults (Russo & Goldstein, 1983) and in up to one-third of children (Hirtz et al., 2000). There are consistent findings on the greater sensitivity of MRI compared to CT for the diagnosis of epileptogenic conditions (Hirtz et al., 2000). By contrast, MRI is mandatory for the detection of structural brain abnormalities, to assess the risk of relapse, and to guide therapeutic management in patients with cryptogenic and remote symptomatic epilepsies. The role of MRI is less defined in the diagnostic workup and prognostic assessment of patients with idiopathic epilepsy. There is no evidence that MRI is superior to CT scan in an emergency setting, at least in children (Ferry, 1992). For its high sensitivity and specificity, examination of the CSF is generally performed in the presence of a febrile seizure associated with meningeal signs, to exclude a cerebral infection (Anonymous, 1993). In infants under 6 months of age with impaired consciousness and delay in the recovery of alertness, the CSF may be abnormal even in the absence of signs of meningeal irritation. By contrast, the value of CSF examination in patients with a first afebrile seizure is as yet unproven. Epileptic seizures may occur in the context of an acute brain insult (trauma, stroke, or other systemic, toxic, or metabolic disturbances). These events (referred to as acute symptomatic seizures, provoked seizures, or situation-related seizures) are presumed to be the acute manifestations of the insult and may not recur when the underlying cause has been removed or the acute phase has elapsed (Commission, 1993). Unprovoked seizures may be also single, as seizure recurrence is observed only in about one-half of cases (Berg & Shinnar 1991). Population-based studies provide a 36–37% relapse rate at 1 year and 43–45% relapse rate at 2 years (Annegers et al., 1986; Hart et al., 1990). The incidence of acute symptomatic seizures (isolated or recurrent) is 29–39 per 100,000 per year (Loiseau et al., 1990; Annegers et al., 1995). The incidence of unprovoked seizures ranges from 42.3 per 100,000 per year (Loiseau et al, 1990) to 61.0 per 100,000 per year (Hauser et al., 1997). The incidence of isolated unprovoked seizures can be only estimated from studies with prolonged (virtually lifetime) follow-up. Different incidence rates were found in western countries, which ranged from 11.0 person-years (MacDonald et al., 2000) to 23.5 per 100,000 person-years (Olafsson et al., 2005). The death risk in patients with a first seizure, as measured by the standardized mortality ratio, is two- to threefold. Younger age and remote symptomatic seizures (i.e., seizures occurring in the context of a nonprogressing CNS lesion) carry an increased mortality risk. The short-term mortality is fairly high after an acute symptomatic seizure and a first episode of status epilepticus. Data on the mortality of isolated seizures are conflicting. The paucity of data and the inconsistent findings of the studies with available information can be explained by the use of different seizure definitions, the duration of the follow-up, and the study methodology and design. In general, 40–50% of patients are expected to relapse within 2 years of the initial seizure. Even if this risk may vary from case to case, the highest rates of recurrence are found in patients with an abnormal EEG and a documented brain lesion (Berg & Shinnar, 1991). Status epilepticus, history of febrile seizures, and perhaps family history of epilepsy are associated with an increased risk of recurrence of a first unprovoked seizure. The risk of recurrence is highest in the first 12 months and is almost reduced to 0–2 years after the seizure. In children with a first epileptic seizure, the presence of epileptiform abnormalities and, to a lesser extent, focal slowing in the interictal EEG is followed by a greater risk of relapse (Hirtz et al., 2000). In children, in the presence of a cryptogenic seizure, an abnormal EEG doubles the risk of relapse (Shinnar et al., 1994). In adults, the predictive value of the EEG is less certain. The emotional implications of having had a first seizure are sometimes enormous and persist despite prompt medical management and assurance about an unlikely recurrence. This may reflect the patient's personality and is hardly counteracted by the statistical predictions made by the treating physician (which cannot be individualized). In addition, uncertainty at the time of diagnosis, which is the case in a patient with a first seizure, may be associated with considerable worry. Lifestyle limitations, employment difficulties, and driving restrictions have also a great impact on the patient's quality of life and on the emotional implication of a first seizure. Surprisingly, there is still little research in this context. When the diagnosis of epileptic seizure has been confirmed, the decision to treat is dictated by the cause of the seizure and the risk of recurrence. Etiological treatment of acute symptomatic seizures has strong biological plausibility, although there is no evidence from adequately controlled studies that treating the etiology of a first acute symptomatic seizure is followed by a lower risk of relapse. In one randomized study (Solari et al., 1997), treatment with benzodiazepines of a first unprovoked generalized tonic–clonic seizure was followed by a significant decrease in the risk of relapse. Several studies have consistently shown that treatment of a first unprovoked seizure decreases the risk of relapse in the following 2 years, but it does not affect the probability of long-term remission both in children and in adults (Musicco et al., 1997; Hirtz et al., 2003; Marson et al., 2005). Although treatment of the first unprovoked seizure can substantially reduce the risk of a recurrent seizure, it is not without risks. The risks associated with drugs are not uncommon and may include serious life-threatening adverse reactions (idiosyncratic and teratogenic effects), adverse effects which are incompatible with continued drug utilization, and chronic effects (with special reference to cognitive side effects), which may be of particular concern as being frequently unrecognized and affecting significantly quality of life. The lack of proper validation studies does not exclude the need to perform an accurate clinical assessment (history taking, general and neurological examination) in all individuals with a first seizure. Detailed inquiry into the occurrence of cyanosis, hypersalivation, tongue biting, and postictal disorientation is required. Laboratory tests should be performed only when circumstances suggest a toxic or metabolic encephalopathy, particularly in patients with persisting impairment of consciousness during the examination. Except for infants less than 6 months of age, examination of the CSF is recommended in children and adults only when cerebral infection is suspected (Hirtz et al., 2000). Examination of the CSF is generally contraindicated in the absence of fever. Academic bodies like the American Academy of Emergency Physicians do not recommend routine toxicological screening in the management of adult patients with seizures (Anonymous, 1996). As a whole, the evidence does not support routine toxicological examinations for single unprovoked seizures. EEG is an essential part of the diagnostic evaluation of epileptic seizures both in children and in adults. If the EEG during wakefulness is normal, sleep EEG can be recommended, but its diagnostic yield is still uncertain. An EEG should be performed within 24 h after a seizure, particularly in children. MRI is an integral part of the diagnostic assessment of a patient with a suspected first cryptogenic or remote symptomatic seizure. MRI should be preferred to CT, but it is not necessary in the presence of a diagnosis of idiopathic epilepsy. CT is an alternative tool in patients in whom MRI is contraindicated or cannot be performed. The use of brain CT scan or MRI is indicated when specific interventions may be needed. A CT scan is strictly indicated when a structural lesion is suspected or when the etiology of the seizure cannot be easily identified. In the emergency room, MRI is not indicated, except for very selected circumstances, to be evaluated on an individual basis. In the presence of a first acute symptomatic seizure (metabolic encephalopathy, acute CNS injury in patients with an underlying treatable condition), treatment of the cause is recommended. Symptomatic therapy of a first unprovoked seizure is generally not justified unless the seizure is prolonged. The risk of acute and chronic adverse effects of drugs should be strongly considered when starting treatment; in which case, risks should be minimized by careful AED selection, gradual dose titration, choice of appropriate target dose, and monitoring of clinical response. Treatment may be considered in patients in whom electrophysiological and imaging data indicate an increased risk of relapse (as indicated, most notably, by the presence of structural CNS abnormalities and/or EEG abnormalities) and/or in those in whom the risks and the benefits of treatment are in favor of the latter, after consideration of the social, emotional, and personal implications of seizure relapse and of treatment itself. Treatment should be encouraged also when the first unprovoked seizure has the characteristics of status epilepticus because the recurrence may be viewed as another prolonged seizure. Children and adolescents with a first unprovoked seizure are no different from adults and the elderly in terms of risk of recurrence. However, the risks and the emotional and social complications of a seizure are fairly different whether the patient is a child or an adult. Restrictions of daily living activities (e.g., driving and risky occupations), which should be imposed preferably on adults than children, should be mostly limited to the first 6–12 months after the seizure. Selected categories of patients, like the elderly or those with learning disabilities, may be encouraged to start treatment after a first seizure, but this is still a controversial issue. There are situations that may indicate deferral of treatment (e.g., pregnancy) while others, for example, patients performing potentially dangerous activities, may favor initiation of treatment. In either case, the patient should be involved in the decision process. Disclosure of Conflicts of Interest: EB has received research funding from Eisai, GSK, Janssen Cilag, Novartis, Sanofi-Aventis, Sigma-Tau and UCB-Pharma.