Myoclonic epilepsy in infancy (MEI) is characterized by brief generalized myoclonic seizures associated with generalized spike-wave paroxysms without other seizure types occurring in the first 3 years of life in developmentally normal children. In this study we analyze the electroclinical features, treatment, and outcome of 38 patients with MEI.
A retrospective chart review was conducted in 38 patients followed at the Neurology Department of the Pediatric Hospital Juan P. Garrahan in Buenos Aires, Argentina, between 1990 and 2012.
A total of 24 boys and 14 girls were identified. The mean and median ages at seizure onset were 16 and 18 months, respectively (range 3–40 months). Ten patients (28.9%) had a family history of epilepsy, and six (15.8%) had a family history of febrile seizures. All patients had several daily brief and isolated myoclonic seizures during wakefulness and predominantly in the first two stages of sleep. Twelve children (31.5%) had reflex myoclonus, triggered by a tactile stimulus in 10 and additionally by noise and light in 2. The remaining two had photosensitive myoclonic jerks. The interictal electroencephalography (EEG) recordings evidenced generalized spike waves, polyspikes, and polyspike-wave paroxysms. The interictal EEG was normal in 12 patients. The abnormalities on the ictal EEG were similar to those on the interictal EEG. Most of the patients responded well to valproic acid. After a mean follow-up of 13.5 years, 24 patients (63%) were without treatment. At the last examination, 32 patients had normal neurologic and neuropsychological evaluations. Two patients (5.2%) had significant cognitive impairment (an IQ of 60 and 63, respectively) despite good seizure control. Four patients (10.4%) had significant learning impairment, two of whom also had attention deficit hyperactivity disorder.
MEI is a well-defined epileptic syndrome of unknown etiology, but likely of a genetic cause. It is self-limited and pharmacosensitive mainly to valproic acid.
In 1981, Dravet and Bureau (1981) first described seven infants with myoclonic seizures and named the syndrome myoclonic epilepsy in infants. The syndrome was included in the International Classification of Epilepsies and Epileptic Syndromes (Commission, 1989) in the group of idiopathic generalized epilepsies and syndromes with age-related onset. Subsequently, reports by other authors were reviewed by Dravet and Bureau (2005) and Guerrini et al. (2012).
Several authors have reported infants with reflex myoclonic seizures triggered by noise or touch using the name reflex myoclonic epilepsy in infancy as a distinctive epileptic syndrome (Ricci et al., 1995); however, this entity has also been considered a variant of benign myoclonic epilepsy of infancy (Caraballo et al., 2003). Reports of other children with reflex myoclonic seizures have been published, including in patients with photosensitive seizures (Zafeiriou et al., 2003; Capovilla et al., 2007).
The benignity of the syndrome has been questioned, as the term benign should be limited to those forms of epilepsy for which the good prognosis can be predicted from onset (Engel, 2006). Because that is not the rule in all cases with an idiopathic form of myoclonic epilepsy in infancy (Zuberi & O'Regan, 2006), the term was changed to myoclonic epilepsy in infancy (MEI) (Engel, 2006). The question is whether infants with MEI with a benign course and those with not such a benign course have the same epileptic syndrome. Because this latter term did not allow one to accurately define a specific type of MEI, the new name proposed was idiopathic myoclonic epilepsy in infancy. Nevertheless, because the entity is of unknown etiology but likely to be genetic, we prefer to use the term myoclonic epilepsy in infancy (MEI) in this report.
MEI is characterized by brief generalized myoclonic seizures associated with generalized spike-wave paroxysms without other seizure types occurring in the first 3 years of live in developmentally normal children.
In the long-term follow-up, some patients with MEI have been reported to develop other idiopathic epileptic syndromes after a seizure-free period following remission of the myoclonic seizures (Prats-Vinas et al., 2002; Auvin et al., 2006; Darra et al., 2006; Capovilla et al., 2007; Moutaouakil et al., 2010). Cognitive outcome was variable (Prats-Vinas et al., 2002; Auvin et al., 2006; Darra et al., 2006; Capovilla et al., 2007).
Herein we analyze the electroclinical features, treatment, and outcome of 38 consecutive patients with MEI with long-term follow-up.
Material and Methods
A retrospective chart review was conducted in 38 patients with MEI seen at the Neurology department of the Pediatric Hospital Juan P Garrahan, Buenos Aires, Argentina, between 1990 and 2012. During the same period, 8,561 patients with epilepsy were seen at the department. The patients were found in an electronic epilepsy database at this neurology department, a tertiary referral center for inpatients and outpatients. The study was approved by the hospital ethics committee. Eight patients with reflex myoclonic seizures were previously published by our group (Caraballo et al., 2003). All patients met the following inclusion criteria: (1) normal development until seizure onset; (2) no organic or other obvious cause for the seizures; (3) seizure onset between 2 months and 4 years of life; (4) type of seizure: myoclonic, including reflex myoclonic seizures; and (5) generalized paroxysms of polyspike or spike-and-wave complexes. Patients with structural or metabolic etiologies were excluded.
In this study all 38 patients underwent sleep and awake electroencephalography (EEG) recordings, and 15 additionally underwent a video-EEG recording. Seizure onset, semiologic features, frequency, distribution, duration of the seizures, and interictal and ictal EEG recordings were analyzed. A polygraphic-EEG recording was obtained in two. Computed tomography (CT) scan was obtained in 30 patients, and brain magnetic resonance imaging (MRI) in nine patients. Neurometabolic studies were done in two cases, and karyotyping was performed in two. Lumbar puncture was performed in two patients to rule out glucose transporter type 1 (Glut-1) deficiency syndrome. Data on school achievements and neuropsychological evaluations (Stanford-Binet Intelligence Scale and Wechsler Intelligence Scale for Children 3rd or 4th edition) were repeatedly obtained during the follow-up of 3–22 years. Mean follow-up was 13.5 years. In the absence of formal neuropsychological tests, cognitive changes were evaluated according to clinical judgment.
Biochemical controls included periodic blood analysis (complete blood cell count) and liver function test; they did not show abnormalities. Antiepileptic drug (AED) treatment was analyzed. Treatment was not based on a protocol, but had been indicated by each physician independently. Plasma levels of classic AEDs were determined. In children who received prolonged topiramate treatment, urine analysis was done to check kidney function and renal ultrasound was performed to rule out renal calculi.
A total of 38 patients (24 boys and 14 girls) with MEI were identified at our department between March 1990 and May 2010.
The mean and median ages at seizure onset were 16 and 18 months, respectively (range 3–40 months). One patient had his first seizure during the third year of life.
Seven patients (18.4%) had personal antecedents of febrile seizures. Ten patients (28.9%) had a family history of epilepsy, and six patients (15.8%) had a family history of febrile seizures. Because the study was retrospective, it was difficult to define the type of febrile seizure and epileptic syndrome found in family members.
Physical examination was unremarkable in all patients and none of them had physical dysmorphia. By definition, brain CT scans performed in 30 patients and MRIs performed in 9 were normal. Karyotyping and neurometabolic tests were normal in two patients. Lumbar puncture to rule out Glut-1 deficiency syndrome was normal in two patients.
All patients had several daily brief and isolated myoclonic seizures during wakefulness and predominantly in the first two stages of sleep. Some seizures were longer in sequences of rhythmic or arrhythmic repeated myoclonic jerks that lasted between 3 and 7 s. In these events, alertness was reduced and thus they may be considered to have been a brief type of absence seizure. Three patients (7.8%) had myoclonic jerks only during drowsiness. The myoclonic seizures were predominantly located in the upper limbs and head, with variable intensity in the same child and when comparing children, and from one episode to the next. Eight patients (21%) had a massive myoclonus predominantly in the lower limbs causing sudden drop attacks. In four patients, myoclonic jerks were bilateral but asynchronic. Subtle seizures, such as eye blinking, rolling-up of the eyeballs, and a brief movement of the head were observed in six patients (15.6%). We did not observe or register any atonic seizures or negative myoclonus following the myoclonic seizures. We never observed sudden vocalization or epileptic spasms. Seizures were in clusters in 11 patients (28.9%). Twelve children (31.5%) had reflex myoclonus, triggered by a tactile stimulus in 10 and additionally by noise and light in 2 of them. The remaining two had photosensitive myoclonic jerks. The reflex myoclonic seizures were isolated, but seemed brief myoclonic absences in two. Jerks were provoked mainly during wakefulness, but they were also elicited during drowsiness and non–rapid eye movement (REM) sleep. Six of 12 children simultaneously had reflex seizures as well as spontaneous myoclonic episodes. The spontaneous episodes were isolated and registered during drowsiness. These six cases did not differ from the others in terms of treatment response or outcome. No generalized tonic–clonic seizures, negative myoclonus, atonic or tonic seizures, or myoclonus status epilepticus were recognized.
In the analysis of the clinical manifestations we especially focused on the 17 patients who underwent either polygraphic and/or video-EEG studies. The video-EEG and polygraphic EEG recordings allowed for recognition of myoclonic seizures depending on the child's posture and showed that ictal EEG paroxysms were similar to interictal EEG discharges.
The background EEG showed normal activity in all patients while awake and during sleep. Monomorphic, synchronous theta 4–6-Hz activity in central regions was observed in five patients (13%).
The interictal EEG recordings evidenced generalized spike-and-wave paroxysms in 11 cases, generalized polyspike-and-wave paroxysms in 8, and polyspikes in 7 patients. The paroxysms were asymmetric in six. The generalized paroxysms increased during sleep. Twenty patients had paroxysmal abnormalities only during sleep. Focal spikes, predominantly in the central regions, were recognized in two patients (5.2%). The interictal EEG was normal in 12 patients. Photoparoxysmal abnormalities were not observed in any of our patients, except in three patients older than 2 years in whom they were characterized by brief generalized polyspike-wave paroxysms without clinical manifestations. Hyperventilation when crying was normal in 15 cases. No monomorphic rhythmic 4–5 Hz theta activity was observed in the rolandic and vertex regions.
In 28 patients the ictal EEG recordings showed isolated generalized polyspike waves or recurrent, brief paroxysms of generalized fast spike waves associated with myoclonic jerks (Fig. 1A,B). The discharges were bilaterally synchronous or asynchronous during the same jerk or from one episode to the next. The subtle seizures were associated with bilateral frontocentral spike-waves. Twelve of these 28 patients had normal interictal EEG recordings. The ictal EEG abnormalities of the myoclonic reflex seizures were similar to the spontaneous ictal EEG paroxysms. The interictal EEG paroxysms were very similar to the ictal EEG abnormalities.
All patients received AEDs from seizure onset. All 38 children, except for 4 with reflex myoclonic seizures who subsequently did not receive any AEDs, were initially treated with valproic acid (plasma levels ranging from 45 to 95 μg/mL). In eight patients another AED was added to valproic acid: clobazam in three, ethosuximide in three, topiramate in one, and levetiracetam in the remaining patient. One patient who had myoclonic seizures in clusters was referred to us with a diagnosis of infantile spasms and had been treated with polytherapy (more than two AEDs), including vigabatrin. His electroclinical picture worsened with vigabatrin adding new reflex myoclonic seizures (acoustic and photoparoxysmal myoclonic seizures). When vigabatrin was withdrawn, the seizures disappeared and the EEG became normal.
Not all children were diagnosed and treated by us from the start. All 34 patients who received treatment did so for a mean period of 25 months (range 12–48 months) after seizure onset. Seizure control occurred in treated patients after a mean period of 14 months (range 2–34 months). In the group of patients with reflex seizures who were not treated, the myoclonic seizures disappeared spontaneously within 8–13 months after seizure onset. In Table 1 the main electroclinical features and outcome of the patients are shown.
Table 1. Main electroclinical features and outcome of 38 patients with myoclonic epilepsy of infancy
We evaluated the evolution of the patients over a 3- to 22-year period of follow-up with a mean of 13.5 years. Twenty-four patients (63%) were without medication during follow-up. In three of them, myoclonic seizures recurred immediately after the AED was stopped, but seizure control was good after reinitiating the treatment. Two patients had isolated tonic–clonic seizures at 8 and 18 years, respectively, and two other patients developed electroclinical features of juvenile myoclonic epilepsy at the ages of 13 and 14 years, respectively, with seizures induced by intermittent photic stimulation in one of them. No patients with focal seizures were recognized. None of the patients with reflex myoclonic episodes with or without spontaneous myoclonic seizures developed any other type of seizure or epileptic syndrome after a seizure-free period. Fourteen patients were still on the initial AEDs, two of whom had juvenile myoclonic epilepsy and one generalized tonic–clonic seizures.
The control EEG recordings were normal in 26 patients (68.4%). Median age at EEG normalization was 3 years (range 2–6 years). The remaining 12 patients continued having generalized paroxysms, associated with bilateral spikes in centrotemporal regions in two and in frontal regions in one. Only one patient had a positive response to intermittent photic stimulation on the EEG without concomitant clinical manifestations.
At the last follow-up visit, 32 patients had normal neurologic and neuropsychological evaluations. Two patients (5.2%) had significant cognitive impairment (IQs 60 and 63) despite good seizure control. Because the study was retrospective, in these two patients the neuropsychological profile previous to seizure onset was difficult to assess and thus progressive cognitive impairment was also difficult to show. Four patients (10.4%) had significant learning disabilities, and two of them also had attention deficit hyperactivity disorder. All children with reflex myoclonic seizures had normal neuropsychological evaluations. In patients older than 16 years of age a telephone interview was conducted to ask about their school performance, the possibility of new seizures, and to evaluate social and psychological aspects.
At the time of the last observation, 24 patients older than 14 years were seizure-free and without medications except for 3, of whom 2 had presented with juvenile myoclonic epilepsy and one with generalized tonic–clonic seizures. Of the 14 patients younger than 14 years, 11 continue receiving AEDs. We found no differences in outcome comparing these two groups of patients.
This retrospective study based on the electroclinical features and outcome observed in our series of patients with MEI clearly shows a well-defined epileptic syndrome with spontaneous and/or reflex myoclonic seizures and a good outcome in terms of seizure control and neuropsychological profile. MEI was classified as idiopathic, generalized epilepsy with an as-yet-unknown etiology, but likely of genetic cause considering the high prevalence of epilepsy and febrile seizures both in family members and patients with MEI and the possibility of the latter to develop another type of idiopathic or genetic epileptic syndrome. A family history and the association of MEI with other types of idiopathic epileptic syndrome have been reported by several authors (Arzimanoglou et al., 1996; Prats-Vinas et al., 2002; Auvin et al., 2006; Darra et al., 2006; Moutaouakil et al., 2010; Mangano et al., 2011). In addition, patients with focal electroclinical features of idiopathic benign focal epilepsy of childhood with rolandic spikes associated with MEI have been described (Darra et al., 2006; Korff et al., 2009).
Our series of patients is similar to the cases published previously (Dravet & Bureau, 2005; Auvin et al., 2006; Darra et al., 2006; Hirano et al., 2009) not only from the electroclinical, but also from the therapeutic and prognostic point of view. However, data obtained in an important electroclinical study showed that the myoclonic seizures may have electroclinical features of focal manifestations (Darra et al., 2006), as has been documented in patients with juvenile myoclonic epilepsy (Thomas et al., 2005). Focal paroxysms have also been reported in other series of patients with MEI (Lin et al., 1998; Dravet & Bureau, 2005), and the focal, asymmetric, or asynchronic component of the myoclonic seizures may represent a manifestation of a particular age-dependent genetic hyperexcitability of the motor cortex (Darra et al., 2006). In our current series of patients and in patients reported previously (Caraballo et al., 2011a) we describe particular myoclonic seizures associated with absences. These seizures were discussed by Guerrini et al. (2012), who argued that they actually may not be absence seizures. Recently, in a large series of children with early onset absence epilepsy, no significant differences in demographic and electroclinical aspects were observed between children with early onset absence epilepsy who responded well to AED monotherapy and those who became seizure-free with add-on treatment of a second AED (Agostinelli et al., 2012).
Assessing the reflex myoclonic seizures registered in our group of patients and the other cases that have been reported in the literature confirming their presence in patients with MEI, they should probably be considered as a variant of this syndrome rather than a different epileptic syndrome (Ricci et al., 1995; Caraballo et al., 2003; Zafeiriou et al., 2003; Auvin et al., 2006; Darra et al., 2006; Capovilla et al., 2007). This group of infants with reflex seizures had an excellent outcome, and some of them, similar to our patients described earlier, did not receive any AEDs.
According to our series of patients and other patients reported (Auvin et al., 2006; Darra et al., 2006; Guerrini et al., 2012), there is a group of infants who present with subtle and isolated myoclonic seizures associated with normal interictal EEG recordings or bilateral spikes in central or vertex regions, which may be difficult to recognize as being of epileptic origin. In these cases, an ictal EEG is mandatory for defining the electroclinical features to start early and adequate AED treatment.
As to neuropsychological manifestations, several authors reporting series of patients with MEI have pointed out that cognitive dysfunction was observed (Giovanardi Rossi et al., 1997; Lin et al., 1998; Dravet & Bureau, 2005; Mangano et al., 2005; Auvin et al., 2006; Darra et al., 2006; Guerrini et al., 2012) as in our series of patients. Most of our patients (32 of 38) had a normal neuropsychological performance. Two patients presented with mental retardation and four other cases had cognitive alterations. In the former two patients as in the latter four, the neuropsychological profile previous to seizure onset was difficult to assess, and thus progressive cognitive impairment was also difficult to demonstrate. Some of the patients may have cognitive involvement previous to seizure onset possibly depending on environmental factors, and/or the possible genetic cause may also play an important role. The neuropsychological impairment should be evaluated accurately to define the relationship of epileptogenic mechanisms and cognitive impact. Similar findings were shown in a single study analyzing the neuropsychological and behavioral outcome in seven patients with an average follow-up of 6 years, 9 months (Mangano et al., 2005). As in our study, the authors did not find any significant electroclinical and therapeutic differences between patients with a normal neuropsychological profile and those with cognitive disturbances (Mangano et al., 2005). Our results corroborate a good prognosis not only in terms of seizure control, but also regarding neuropsychological performance in most of the patients in our series. However, we believe that delay of the diagnosis should be avoided so as to start early treatment to stop the high frequency of the seizures.
The majority of patients with MEI have been treated with valproic acid as monotherapy (Dravet & Bureau, 2005; Auvin et al., 2006; Darra et al., 2006; Guerrini et al., 2012). It was used as the first choice AED in most of our patients as well. Levetiracetam may also be considered (Gentile et al., 2010). A possible paradoxical worsening of myoclonic seizures has been reported (Thomas et al., 2006). One of our patients with refractory seizures was referred to our service with a misdiagnosis of West syndrome, and with vigabatrin the electroclinical manifestations significantly worsened adding reflex myoclonic seizures. When vigabatrin was withdrawn, the seizures disappeared and the EEG became normal.
In the long-term follow-up of patients with MEI, generalized tonic–clonic or clonic seizures have been reported after a myoclonic-free interval; however, seizures were isolated or easily controlled by a short treatment. Less frequently, absence seizures with or without eyelid myoclonia, juvenile myoclonic epilepsy, and photic-induced myoclonic seizures have also been published (Prats-Vinas et al., 2002; Auvin et al., 2006; Darra et al., 2006; Guerrini et al., 2012). The EEG abnormalities may persist for many years, usually without myoclonic seizures. As to neuropsychological outcome, slight and severe mental retardation have been reported, and cases with dyslexia, attention deficit disorders, and language and motor delay have also been described (Mangano et al., 2005; Guerrini et al., 2012). When compared to patients with normal cognitive outcome, patients with an abnormal neuropsychological profile had no apparent differences in age at onset, EEG abnormalities, other types of seizure, and response to treatment (Guerrini et al., 2012).
In the differential diagnosis, syndromes with myoclonic seizures in infancy or early childhood should be taken into account (Guerrini et al., 1994; Dravet et al., 2005; Crespel et al., 2012; Caraballo et al., 2013). In addition, MEI should be differentiated from West syndrome and epileptic spasms in clusters without hypsarrhythmia in infancy and early childhood (Caraballo et al., 2011b). Patients with electroclinical features similar to those of MEI associated with Glut-1 deficiency have been described (Oguni, 2005; Roullet-Perez et al., 2008; Gaspard et al., 2011). The autosomal recessive benign myoclonic epilepsy in infancy reported in one family linked to chromosome 16p 13 had similarities with that of patients with MEI, but in the former the myoclonic jerks may be grouped in long clusters for many hours, were always associated with generalized tonic–clonic seizures, and persisted into adulthood (Zara & De Falco, 2005).
Considering the differential diagnosis with nonepileptic conditions, in infants with episodes resembling myoclonic seizures with normal psychomotor development and normal EEG recordings performed while awake and during sleep, benign nonepileptic myoclonus or Fejerman syndrome should be taken into account (Caraballo et al., 2009).
The shortcoming of our study is the lack of polygraphic-EEG recordings to define the features of the myoclonic seizures. Registration of the seizures is crucial mainly in those cases with subtle electroclinical manifestations to diagnose this syndrome. However, in almost half of our patients a video-EEG allowed the recognition of myoclonic seizures.
According to the last proposal of the International League Against Epilepsy (ILAE) Task Force (Berg et al., 2010) MEI should be considered among the epilepsies of unknown etiology; however, we believe that genetic factors are evident in this syndrome. With respect to replacement of the term “benign” by “self-limited,” the group of reflex MEI without AED treatment may be considered self-limited, since the myoclonic seizures disappeared spontaneously. The infants with a good response to AEDs may be named “pharmacosensitive.”
Prospective studies considering polygraphic-EEG recordings and genetic studies should be conducted to delineate the nosology of infants with myoclonic seizures meeting the diagnostic criteria of MEI.
Our study demonstrates that MEI is a well-defined epileptic syndrome of unknown etiology, but likely of genetic cause. The reflex cases may be considered self-limited and the others pharmacosensitive.
Valproic acid is the first AED option in the treatment of patients with MEI. An association of valproic acid with other AEDs—such as clobazam, ethosuximide, levetiracetam, and topiramate—may also be considered.
Our patients had a good outcome in terms of seizure control and neuropsychological aspects, except for two patients with mental retardation and four with learning disabilities and behavioral disturbances.
All coauthors have read and agreed to the content of the manuscript. None of the authors has any conflict of interest to disclose, any financial or commercial involvement, or any contribution of industry-sponsored research or of corporate participation in preparing the manuscript. We confirm that we have read the Journal's position on issues involving ethical publication, and affirm that this report is consistent with those guides.