Epilepsia, 48(9):1703–1707, 2007
Blackwell Publishing, Inc.
C ?2007 International League Against Epilepsy
Current Treatment of Myoclonic Astatic Epilepsy: Clinical
Experience at the Children’s Hospital of Philadelphia
Sudha Kilaru and A. G. Christina Bergqvist
Division of Neurology, The Children’s Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia,
Summary: Purpose: Myoclonic astatic epilepsy (MAE) is a
generalized epilepsy of early childhood. Little is known about
the use of newer antiepileptic treatments (AET) in MAE. The
purpose of this study was to describe the characteristics, treat-
ment, and outcome of a contemporary MAE cohort exposed to
the new generation AET.
Methods: Charts of subjects with MAE treated between 1998
and 2005 were reviewed.
Results: Twenty-three subjects (19 boys), with a median
(range) follow-up of 38 (2– 86) months were identified. Thirty-
nine percent had a family history of epilepsy, and 39% had fam-
ily history of febrile seizures. Age at seizure onset was a median
of 36 (12–24) months. Initial EEG was normal in 30%. When
seizures ceased, EEG background and epileptiform abnormali-
ties persisted in 17 and 58%, respectively. On average, each sub-
ject was exposed to five AET. The most frequently used AET
was valproate (83%). Seizure freedom occurred spontaneously
in three subjects, with ethosuximide and levetiracetam in one
each, valproate and lamotrigine in two each, topiramate in three
and the ketogenic diet (KD) in five subjects. By 36 months after
seizure onset, 67% achieved seizure freedom. At the last visit,
43% were developmentally normal, 52% had mild, and 5% had
moderate cognitive disabilities. Time to seizure freedom did not
correlate with cognitive outcome.
Conclusions: The new generation of AET may of-
fer significant benefit to children with MAE. The KD
was the most effective AET in this series, and perhaps
should be considered earlier in treatment.
Myoclonic astatic epilepsy—Idiopathic generalized epilepsy—
Antiepileptic drugs—Ketogenic diet.
Myoclonic astatic epilepsy (MAE) of Doose is a
generalized epilepsy syndrome of young children char-
acterized by multiple seizure types—predominantly
myoclonic seizures (MS), astatic seizures (AS), and my-
oclonic astatic seizures (MAS), as well as with gener-
alized tonic–clonic (GTC), absence, myoclonic absence,
and tonic seizures (Doose, 1992). In general, children
are developmentally normal before the onset of epilepsy
and organic brain abnormalities are absent. The seizures
are often difficult to treat, but may remit spontaneously.
EEG tracings may be normal initially, but later develop
a characteristic biparietal theta background rhythm, and
irregularly generalized spike wave, and polyspike wave
discharges. Long-term prognosis varies from cessation
of seizures with normal developmental outcome to in-
tractable epilepsy with mental retardation. Favorable out-
1992; Kaminska et al., 1999; Oguni et al., 2002 ). A re-
lationship between poor control of seizures and cognitive
Accepted April 11, 2007.
Address correspondence and reprint requests to Sudha Kilaru, Di-
vision of Neurology, Children’s Hospital of Philadelphia, 34th and
Little is known about the optimal treatment of MAE
and efficacy of antiepileptic treatments (AET) has not
been studied in randomized clinical trials (RCT). Treat-
ment practices are largely based on extrapolation from
clinical experience in other idiopathic generalized epilep-
sies. Few published series include subjects treated with
the newer AET. Valproate (VPA) is often advocated as the
first-line therapy though efficacy is unclear (Wheless and
Sankar, 2003; Neubauer et al., 2005). Of the newer AET,
topiramate (TPM), lamotrigine (LTG), and levetiracetam
(LEV) have been reported to be effective in small series
Recent studies also suggest that the ketogenic diet (KD)
Fejerman et al., 2005; Caraballo et al., 2006).
The purpose of this study was to investigate the clinical
cohort of children with MAE treated with newer AET and
tal of Philadelphia (CHOP) approved this study. Potential
1704S. KILARU AND A. G. C. BERGQVIST
subjects were identified by systematically searching
CHOP neurology electronic clinic visit notes from 1998
“Doose syndrome,” or “idiopathic generalized epilepsy”
(IGE). Subjects included met the International League
Against Epilepsy (ILAE) guidelines for the definition of
MAS (Anonymous, 1989). Inclusion criteria included on-
by clinical exam and MRI; and EEG with 2–3 Hz gener-
alized spike and wave or polyspike and wave discharges.
Exclusion criteria were a predominance of tonic seizures
or focal epileptiform discharges on initial EEG. EEG
background abnormalities were not cause for exclusion.
Information collected for analysis included demo-
response, seizure response, and developmental outcome.
A positive family history was defined as having a first
or second degree relative with epilepsy, or with febrile
semiology or identification of seizure type by the eval-
uating clinician. EEG tracings were not always avail-
able; therefore EEG reports were used to determine EEG
background abnormalities (generalized or focal slowing
and/or disorganization), interictal epileptiform abnormal-
ities, and ictal discharges. EEGs were grouped into first
EEG performed, EEGs performed after initiation of AET,
and EEGs performed after seizure freedom was achieved.
AET use for each subject was tracked by numbering each
AET in order of its use. Quantitative data on change in
seizure frequency were not consistently available, so par-
tial response to AET could not be reliably assessed. Re-
sponse to therapy was defined as at least 6 months of
seizure freedom after AET initiation.
Developmental outcome was defined as normal if the
child was in a mainstream educational setting without ad-
amination. Children with mildly abnormal developmental
outcomes were in mainstream classrooms and received
additional educational support or were treated for inat-
tention. A moderately abnormal developmental outcome
required special education, and/or significant abnormali-
ties on neurological exam. A severely abnormal develop-
mental outcome was defined as a developmental quotient
All subjects exposed to the KD, except one, initiated
KD at CHOP, using our protocol previously described
(Bergqvist et al., 2005).
The clinical characteristics of 23 subjects with MAE
(19 boys, four girls) are summarized in Table 1. Fifteen
subjects were evaluated at CHOP within 1 month of pre-
TABLE 1. Characteristics of subjects at presentation, median
[range], or %
Age at first seizure, months
Family history, %
Presenting seizure type, %
GTC with fever
Incidence of seizure types,%
Peak frequency of myoclonic
and astatic seizures, %
sentation; eight were referred later in their course. Eigh-
teen subjects had a median follow-up period of 48 (range
and two subjects were seen only once. A family history
of epilepsy or febrile seizures was present in 39% each.
Median age at first seizure was 36 (range 12–64) months.
Presenting seizure type was GTC without fever in 52%,
AS, and/or MAS developed in all patients. In those with
initial seizure types other than MS or AS, time from first
seizure to MS or AS was a median of 33 (range 7–669)
in their course, 52% had absence seizures, 9% had my-
status epilepticus. No subjects reported tonic seizures.
The initial EEG was available in 20 of 23 subjects, and
was normal in 30% (n = 6). Background abnormalities,
mainly theta and delta range generalized slowing, were
present in 40% (n = 8) and epileptiform abnormalities
were present in 55% (n = 11). Two subjects had focal
background slowing. In the period of time during which
subjects were having frequent (daily) seizures, all but one
subject had one or more routine EEG or prolonged video
or polyspike and wave discharges. Seventy-one percent
(n = 12) of subjects achieving seizure freedom had EEGs
after seizure cessation. Background abnormalities per-
sisted in a minority (17%, n = 2) and epileptiform ab-
normalities persisted in over half (58%, n = 7).
Epilepsia, Vol. 48, No. 9, 2007
MYOCLONIC ASTATIC EPILEPSY1705
AET use and response is summarized in Table 2. On
average, five separate AET were used alone or in combi-
nation over the course of illness. The first AET most fre-
quently used were oxcarbazepine (OXC), carbamazepine
(CBZ), or phenobarbital (PB). VPA was the most com-
monly used AET after a diagnosis of generalized epilepsy
was made. Subjects achieving seizure freedom were be-
ing treated with an average of two AET at the time of
their last seizure. The percentage of subjects responding
subjects who achieved seizure freedom, three had spon-
taneous abatement of seizures. Seizure freedom occurred
with ethosuximide (ETX) in one subject, VPA in two sub-
patient ceased having seizures with initiation of LEV, but
had a follow-up period of only 3 months. KD was used
after an average of five other AET failed and a median
of 17 (range 2–58) months after onset of seizures. Five
subjects stopped seizing after beginning the KD, though
only three of these had follow up longer than 6 months
after KD initiation. Three stopped seizing within 1 month
frequency followed by complete cessation by 19 months.
with complete cessation of all seizures by 7 months. The
KD did not lead to seizure freedom in five subjects-–of
these, three were nonresponders, one stopped the KD due
nificant reduction of seizures but not seizure freedom.
Sixty-seven percent (n = 14) of the 21 subjects eval-
uated more than once at CHOP had achieved seizure
TABLE 2. Average rank order in which AET was initiated and
response to AET
3 (5) 10
Numbers in parentheses represent subjects who achieved seizure
freedom, but for less than 6 months
freedom for at least 6 months by their last clinic visit
(Fig. 1). Median time to seizure freedom was 13 (range
4–36 months). Because subjects had variable follow-up
periods, they were in various stages of illness at the
last visit. Nineteen percent (n = 4) of subjects were
still having frequent seizures. Two of these four subjects
had been followed for less than 6 months, and two had
been followed for 39 and 85 months, respectively. Three
of the 21 subjects were seizure free for 1.3, 2, and 3
months, respectively, at their last visit. Thirty-nine per-
cent (n = 8) were seizure free for more than 6 months
and were still being treated with AET. No subject who
achieved seizure freedom for at least 6 months had recur-
rence of seizures while treated with AET. Twenty-eight
percent (n = 6) remained seizure free for greater than
6 months after AET were discontinued. However, in three
subjects no longer on AET, seizures returned 3–4 years
after initial seizure cessation. Two had GTC seizures
and one had absence seizures—all seizures in both sub-
jects were infrequent and easily treated. Two of these
three subjects had an EEG during remissions; both fea-
tured a normal background but persistent epileptiform
At the last visit, 43% (n = 9) of children were de-
velopmentally normal, 52% (n = 11) had mild, and 5%
(n = 1) had moderate cognitive disabilities. Time to
seizure freedom was not significantly different between
the normal development and abnormal development
groups (18 ± 10.5 vs. 14.8 ± 7.0 months, p = ns, re-
This retrospective case series of subjects with MAE
offers some insight into the usefulness of newer AET
in this disorder. Doose first described the syndrome that
now includes MAE in 1970 as epilepsy occurring in
older infants and young children with a predominance
of generalized myoclonic, astatic, or myoclonic astatic
seizures. Over time, MAE was further distinguished from
the benign and severe myoclonic epilepsies of infancy
and from the myoclonic variant of Lennox–Gastaut Syn-
drome (LGS) (Doose, 1992) but the borders of these syn-
dromes remain blurred. Specific criteria for a diagno-
sis of MAE are not well established, but accurate syn-
drome classification is important for prognosis and treat-
ment because the alternative diagnoses have significantly
worse long-term prognosis for seizure control and cog-
nitive outcome (Guerrini and Aicardi, 2003). Definitions
of MAE have been inconsistent in prior series, particu-
larly in the parameters of age of onset and the existence
or absence of prior developmental delays, and the pre-
dominance of tonic seizures. (Doose, 1992), (Aicardi and
Gomes), (Oguni et al., 2002), (Kaminska et al., 1999).
Our definition of MAE was limited to those over a year
of age (to exclude those with severe myoclonic epilepsy
Epilepsia, Vol. 48, No. 9, 2007
1706 S. KILARU AND A. G. C. BERGQVIST
of infancy), who were neurologically normal prior to the
onset of epilepsy, with no evidence of focal problems
(to distinguish those with the myoclonic variant of
LGS) in keeping with the most contemporary conception
Combined, these discrepancies of definition may con-
tribute to the differences in the rates of favorable and un-
favorable outcomes between previous series, though dif-
ferences in available treatments must also be taken into
account. In Doose’s series, 54% of children older than 7
years achieved seizure freedom for 2 or more years. Com-
plete seizure freedom was achieved by 68% of subjects in
Oguni’s series and an additional 14% had long remissions
with return of GTC years later. Within 3 years, 89% of
subjects had cessation of myoclonic and AS.
Aicardi reported seizure freedom at follow up in 33%
of all subjects. When subjects were further subdivided,
seizure freedom rates ranged from none (in subjects pre-
senting with multiple seizure types in the first year of
life), to 63% of those whose clinical features most closely
matched our definition of MAE.
The male predominance and high incidence of family
histories of epilepsy or febrile seizures in our series is
consistent with prior studies and supports a genetic eti-
ology. Conflicting evidence exists about the relationship
of MAE to mutations in the generalized epilepsy with
febrile seizures plus (GEFS+) genes. Thirteen individu-
als within GEFS+ families who have mutations in one of
these genes had an MAE phenotype (Wallace et al., 1998;
Escayg et al., 2001; Wallace et al., 2001), while in a study
of sporadic cases of MAE these mutations were absent
(Nabbout et al., 2003). The variable presence of GEFS+
gene mutations, clinical presentation and course of MAE,
suggests that many genes may be involved, and that mod-
ifier genes or environmental factors may have additional
Most subjects in our series presented with GTC
seizures, either with or without fever. The diagnosis of
MAE or IGE may not always be made at presentation,
especially if the initial EEG is normal. Many will be di-
FIG. 1. Cumulative remission curve in 6 month interval after seizure onset.
agnosed as IGE not otherwise specified instead of MAE
if the first AET, started after GTC seizures, effectively
suppresses the emergence of MS or AS. For these rea-
sons, identification of children with MAE for RCT is
Incorrect diagnosis is the likely reason medications
known to worsen primary generalized seizures were
among the first used in many of our subjects. VPA was
used as first-line therapy in most subjects once a diagno-
sis of generalized epilepsy was made, but the seizure free-
dom rate was low. Of the newer anticonvulsants, TPM,
LTG, and LEV led to seizure freedom in some patients.
The treatment with the highest seizure freedom rate was
the KD, with 50% achieving seizure cessation for at least
1 month or more, and 30% achieving seizure cessation
for 6 months or more. Seizure freedom rates from the KD
in general had ranged from 15 to 20% (Freeman et al.,
1998; Lefevre and Aronson, 2000; Bergqvist et al., 2005).
Our study corroborates prior observations that the KD
was particularly effective in MAE (Oguni et al., 2001;
Fejerman et al., 2005; Caraballo et al., 2006). However, it
is possible that seizures in MAE become more amenable
to treatment late in the course of illness, coinciding with
MAE is difficult because of the evolution of the disorder
over time, which may bias results against the medication
started early (often VPA), in favor of treatments started
later in the course of illness. Nevertheless, the KD ap-
pears to be a promising MAE therapy, and a RCT of early
initiation of the KD may be informative.
EEG findings in our subjects mirrored their clinical
course-–epileptiform abnormalities were common at pre-
active disease. Epileptiform abnormalities frequently per-
sisted after seizure cessation, while background slowing
correlating persistent epileptiform abnormalities with re-
turn of seizures (because of variable timing of EEGs after
seizure cessation and the small number of subjects with
seizure recurrence after remission).
Epilepsia, Vol. 48, No. 9, 2007
MYOCLONIC ASTATIC EPILEPSY 1707 Download full-text
In our series, a large percentage of subjects achieved
seizure freedom. The small number of subjects studied
precluded identification of statistically significant predic-
tors of response to treatment. Seizure recurrence after
later, which is atypical for childhood epilepsy in general
(Shinnar et al., 1994). Seizure types in these recurrences
were not myoclonic or astatic, and were not refractory to
therapy. Though none of the subjects in this series were
severely developmentally disabled at the last visit, many
Limitations of this study include its retrospective de-
sign, the small sample size and variable follow-up periods
for subjects. Though a true comparison of AET is not
possible with a retrospective study of this size, a descrip-
tion of the clinical courses of a contemporary cohort may
inform care for children with this uncommon syndrome.
parable to that in prior series (Aicardi and Gomes, 1989;
Doose, 1992; Kaminska et al., 1999; Oguni et al., 2002).
Newer AET, specifically TPM, LTG, and LEV may
offer additional benefit to children with MAE. KD may
be particularly effective in leading to seizure freedom in
MAE, and perhaps should be considered earlier in the
treatment course. If reliable methods of early identifica-
tion of children with MAE, at or near the onset of the
disorder, are formulated, a prospective treatment trial will
be possible, which may provide needed information for
Aicardi J, Gomes AL. (1989) The myoclonic epilepsies of childhood.
Cleve Clin J Med, 56(Suppl Pt 1), S34–9; discussion S40–S42.
Anonymous. (1989) Proposal fo revised classification of epilepsies and
epileptic syndromes. Commission on Classification and Terminol-
Arzimanoglou A, Guerrini R, Aicardi J. (2004) Aicardi’s Epilepsy in
Children. Lippincott, Williams, and Wilkins, Philadelphia.
Bergqvist AG, Schall JI, Gallagher PR, Cnaan A, Stallings VA. (2005)
Fasting versus gradual initiation of the ketogenic diet: a prospec-
tive, randomized clinical trial of efficacy. Epilepsia 46:1810–
Caraballo RH, Cersosimo RO, Sakr D, Cresta A, Escobal N, Fejerman
Epileptic Disorders 8:151–155.
Doose H. (1992) Myoclonic-astatic epilepsy. Epilepsy Research – Sup-
Escayg A, Heils A, MacDonald BT, Haug K, Sander T, Meisler
MH. (2001) A novel SCN1A mutation associated with generalized
epilepsy with febrile seizures plus-and prevalence of variants in pa-
tients with epilepsy. Am J Hum Genet 68:866–873.
childhood. Adv Neurol 95:299–305.
The efficacy of the ketogenic diet-1998: a prospective evaluation of
intervention in 150 children. Pediatrics 102:1358–1363.
seizures in infants and children (severe myoclonic epilepsy and
myoclonic-astatic epilepsy). J Clin Neurophysiol 20:449–461.
Kaminska A, Ickowicz A, Plouin P, Bru MF, Dellatolas G, Dulac O.
(1999) Delineation of cryptogenic Lennox-Gastaut syndrome and
myoclonic astatic epilepsy using multiple correspondence analysis.
Epilepsy Res 36:15–29.
Labate A, Colosimo E, Gambardella A, Leggio U, Ambrosio R, Quat-
trone A. (2006) Levetiracetam in patients with generalised epilepsy
and myoclonic seizures: an open label study. Seizure 15:214–218.
Lefevre F, Aronson N. (2000) Ketogenic diet for the treatment of refrac-
tory epilepsy in children: a systematic review of efficacy. Pediatrics
Mikaeloff Y, de Saint-Martin A, Mancini J, Peudenier S, Pedespan JM,
Vallee L, Motte J, Bourgeois M, Arzimanoglou A, Dulac O, Chiron
C. (2003) Topiramate: efficacy and tolerability in children according
to epilepsy syndromes. Epilepsy Res 53:225–232.
Bianchi A, Brice A, Leguern E, Dulac O. (2003) Absence of muta-
tions in major GEFS+ genes in myoclonic astatic epilepsy. Epilepsy
Neubauer BA, Hahn A, Doose H, Tuxhorn I. (2005) Myoclonic-astatic
netics. Adv Neurol 95:147–155.
Shirakawa S, Osawa M. (2001) Myoclonic-astatic epilepsy of early
and discussions on the nosology of the syndrome. Brain & Develop-
Osawa M. (2002) Treatment and long-term prognosis of myoclonic-
astatic epilepsy of early childhood. Neuropediatrics 33:122–132.
Shinnar S, Berg AT, Moshe SL, Kang H, O’Dell C, Alemany M, Gold-
ensohn ES, Hauser WA. (1994) Discontinuing antiepileptic drugs
in children with epilepsy: a prospective study. [see comment]. Ann
Wallace RH, Marini C, Petrou S, Harkin LA, Bowser DN, Panchal RG,
Williams DA, Sutherland GR, Mulley JC, Scheffer IE, Berkovic SF.
sence epilepsy and febrile seizures. Nat Genet 28:49–52.
Saar K, Reis A, Johnson EW, Sutherland GR, Berkovic SF, Mulley
JC. (1998) Febrile seizures and generalized epilepsy associated with
a mutation in the Na+-channel beta 1 subunit gene SCN1B. Nat
Wallace SJ. (1998) Myoclonus and epilepsy in childhood: a review of
Epilepsy Res 29:147–154.
Wheless JW, Sankar R. (2003) Treatment strategies for myoclonic
seizures and epilepsy syndromes with myoclonic seizures. Epilepsia
Epilepsia, Vol. 48, No. 9, 2007