The diagnosis of LGS is based on the criteria previously defined, which include multiple types of seizures, in particular tonic seizures; an EEG pattern consisting of inter-ictal diffuse slow spike-and-wave discharges occurring at a 1.5-2-Hz frequency, with or without bursts of generalized fast spikes (10 Hz) during NREM sleep; and diffuse cognitive dysfunction, mental retardation, or both.
Because of the nonspecific nature of multiple seizure types and cognitive dysfunction, multiple other pediatric epilepsy syndromes may be confused with LGS. A useful approach to understanding and differentiating these syndromes is to first consider LGS and similar epilepsy syndromes as part of a continuum and then to try to identify characteristics that separate the syndromes. Two scenarios have been proposed. In both scenarios, LGS is at one end of the spectrum. In one scenario, the other end of the continuum is placed in the broad term myoclonic epilepsies (16), and another scenario places the end at MAE (42). In both scenarios, the epilepsy syndromes encountered when moving from one end of the spectrum to the other are in the same order: LGS, the "myoclonic variant" of LGS, MAE, and then the lumped "myoclonic epilepsies" (e.g., benign myoclonic epilepsy of infancy, severe myoclonic epilepsy of infancy, and progressive myoclonic epilepsy) (17).
Compared with classic LGS, the "myoclonic variant" of LGS has less frequent and less severe mental retardation; rarer tonic seizures, usually of late onset and almost exclusively nocturnal; an "unusually marked myoclonic component"; a less unfavorable outcome; and frequently faster (>2.5 Hz) spike-and-wave complexes on EEG (16, 38).
Myoclonic astatic epilepsy and LGS have in common myoclonic seizures, atonic seizures, and atypical absences. However, there are major differences: MAE is predominantly idiopathic, is genetically determined, usually has a favorable outcome, and does not follow West syndrome, whereas LGS is mainly symptomatic, is not genetically determined, usually has an unfavorable outcome, and can follow West syndrome (18, 42). Kaminska and coworkers differentiate LGS and MAE using a sophisticated statistical approach called multiple correspondence analysis to study different clinical and EEG parameters. They recognized a group of children with LGS characterized by later onset of epilepsy, atypical absences, tonic and partial seizures, no myoclonus or vibratory tonic seizures, mental retardation, and an EEG pattern with slow spike-and-waves that were, as a group, different from those in the group with MAE features (42). In addition, myoclonic seizures in LGS and MAE show different neurophysiologic patterns, as shown by Bonanni et al using video-EEG, simultaneous surface electromyog-raphy, and, when necessary, burst-locked EEG averaging techniques (47).
Compared with patients with LGS, patients with myoclonic epilepsies (in the broadest sense) have myo-clonic seizures as the clearly predominant seizure type, only occasionally have tonic seizures, only occasionally have slow spike-and-waves on EEG, almost always have fast (>2.5 Hz) spike-and-wave complexes on EEG, and have variable frequency and levels of mental retardation
Multiple etiologies have been associated with LGS. Knowledge of the cause may affect prognosis and, at times, selection of the best therapy. Neuroimaging is an important part of the search for an underlying etiology in a patient with LGS. Abnormalities revealed by neu-roimaging associated with LGS include tuberous sclerosis, brain malformations (e.g., cortical dysplasias), hypoxic-ischemic injury, or frontal lobe lesions. In general, magnetic resonance imaging (MRI) is the preferred neuroimaging study for a patient with LGS rather than a computed tomography (CT) scan. CT scans may be preferred in selected situations (e.g., evaluation of suspected intracranial injury, hematoma, or both in a patient with head trauma resulting from an atonic or tonic seizure). No current indication exists for routine PET or SPECT scanning in patients with LGS. However, PET and SPECT scans may be useful when patients are undergoing evaluation for potential epilepsy surgery.
An EEG in patients with suspected LGS is critical, because the diagnosis depends on the presence of specific electrographic findings. Recording of a prolonged EEG is desirable, because a routine 30-minute EEG may not capture the patient's electrographic activity both awake and asleep and thus may miss crucial specific EEG findings. It is important to capture and classify each of the patient's multiple seizure types. Video-EEG telemetry should be strongly considered because it may also help to educate the parents on which of the patient's events are seizures and which are nonepileptic behavioral events (44).
When assessing learning disability, it is important to differentiate "permanent learning disability" and "state-dependent learning disability." The former refers to the learning disability as a result of an underlying brain damage that leads to both epilepsy and learning disability, or to the epilepsy that leads to brain damage, which in turn results in learning disability (e.g., prolonged status epilepticus). The latter depends on factors affecting the patient, such as adverse medication effects or increased seizures (ictal and postictal effects), both of which are not necessarily permanent and are potentially treatable and reversible (43). Review of the ongoing treatment and recording of a prolonged video-EEG may be of help.
2. Suspected effective treatments based on open-label uncontrolled studies
3. Effective treatments based on double-blind placebo-controlled studies
In the first and second groups, the efficacy and safety of individual treatment options have not been formally tested. Only options in the third group have been rigorously and scientifically evaluated and found to be effective and safe for specific seizure types in LGS patients. Each treatment group can be subdivided into medical, dietary, and surgical therapies. Unfortunately, no treatment by itself in any of the three groups gives satisfactory relief for all or even a majority of patients with LGS. Combination of treatment modalities is frequently needed (59).
The goals of treatment for patients with LGS are the same as for all patients with epilepsy: the best quality of life with the fewest seizures (hopefully none), the fewest treatment side effects, and the least number of medications.
The various treatment options for patients with LGS can be divided into the following three major groups (Table 21-1):
1. First-line treatments based on clinical experience or conventional wisdom
First-Line Treatments Based on Clinical Experience or Conventional Wisdom
Over the past two decades, valproic acid (VPA) has been considered as a first-line treatment option for children with LGS (24, 52, 53). From a practical viewpoint, by the time a clinician makes a diagnosis of LGS, the patient has already been diagnosed with epilepsy and treatment has been initiated. Because these children have multiple types of generalized seizures and at times coexisting partial seizures, clinicians may initially select a broad-spectrum AED such as VPA. Valproic acid has been reported to be more effective in cryptogenic LGS than in symptomatic LGS (60).
Benzodiazepines, specifically clonazepam, nitraz-epam, and clobazam, are also first-line AED therapy options (38, 61, 62). All are considered effective against seizures associated with LGS, but side effects and the development of tolerance limit their usefulness over time (38). Side effects of clonazepam include hyperactivity, sedation, drooling, and incoordination, which can significantly affect the quality of life for patients with LGS (38). The efficacy and tolerability profile of nitrazepam is similar to that of clonazepam (38). Clobazam is considered the least sedating benzodiazepine, with the longest time to the development of tolerance (62). Some recommendations to slow the development of tolerance include dosing on an every-other-day schedule or alternate two different benzodiazepines on an alternate-day basis (63, 64). Unfortunately, not all benzodiazepines are beneficial: intravenous diazepam and lorazepam have been reported to induce tonic status epilepticus in some patients (65, 66). Based on clinical experience, some authors believe a combination of valproic acid and a benzodiazepine may be better than either drug alone, but no data exist to confirm this impression (38).
Was this article helpful?