The latest on infantile spasms

Raili Riikonen
Purpose of review
This article reviews the most signicant advances in the eld
of infantile spasm during the past year, with emphasis on
best practise for treatment, and on some new etiological
genetic and metabolic causes for the spasms, and new
advances in the knowledge of tuberous sclerosis.
Recent ndings
Up-to-date information comparing corticotrophin, oral
steroids and vigabatrin shows that hormonal treatment is
the most effective therapy in the short term. In a recent
randomized trial, large doses of prednisolone were as
effective as corticotrophin. There are insufcient data to
recommend any treatment schedule for infantile spasms.
Vigabatrin is the choice for infants with tuberous sclerosis.
Visual eld defects in (older) children seem to be as
common as in adults. In animals, vigabatrin can induce
apoptosis of the neurons in the developing brain. New rare
factors associated with infantile spasms are mitochondrial
diseases, mutations of the Aristales-related homeobax gene
and posterior quadrantic dysplasia syndrome. The outcome
in children with tuberous sclerosis and infantile spasms is
better understood.
Summary
The accurate determination of etiology is now becoming
increasingly possible. There is still a lack of consensus
about the treatment of rst choice for infantile spasms.
However, recent data show that hormonal treatment is the
most effective therapy in the short term. Frequency of visual
eld defects in children treated with vigabatrin should be
studied in addition to the long-term outcome in general.
Advances in our understanding of brain maturation,
etiologies, mechanisms and genetics underlying
catastrophic epilepsy may facilitate more effective
pharmacologic interventions.
Keywords
etiology, infantile spasms, outcome, therapy
Curr Opin Neurol 18:9195. # 2005 Lippincott Williams & Wilkins.
Childrens Hospital, University of Kuopio, Finland
Correspondence to Raili Riikonen MD PhD, Professor in Child Neurology,
University of Kuopio, Head of the Department of Child Neurology, Central Hospital,
University of Kuopio, PO Box: 1627, Kuopio 70211, Finland
Tel: +358 17 172 393; fax: +358 17 174 952; e-mail: raili.riikonen@uku.
Current Opinion in Neurology 2005, 18:9195
Abbreviations
AAN American Academy of Neurology
ACTH adrenocorticotropic hormone corticotropin
ARX gene Aristales-related homeobax gene
CNS Child Neurology Society
GABA g-aminobutyric acid
VFD visual eld defect
# 2005 Lippincott Williams & Wilkins
1350-7540
Introduction
Infantile spasms are refractory to the usual antiepileptic
drugs, as assessed in more than 50 trials. Hormonal
therapy has been used since 1958. Vigabatrin (a g-
aminobutyric acid (GABA) transaminase inhibitor) has
rapidly gained popularity in Europe. Comparison of the
efcacy, relapse rate, short-term and long-term effects,
presence or absence of other seizures and side effects of
hormonal therapy and vigabatrin has been insufcient. A
large data-based critical literature review by the American
Academy of Neurology (AAN) and the Child Neurology
Society (CNS) 2004 [1

], reviews by the author

[2,3,4

,5

,6], and a recent prospective randomized study

from the UK comparing the effectiveness of hormonal
therapy and vigabatrin [7] are presented is this review. In
addition, some newdata ontheetiologyof infantile spasms
are presented.
Therapy: short-term outcome
The following sections present the choices of anticon-
vulsant.
What is the most effective treatment: corticotrophin
or oral steroids?
Recommendations of the AAN and CNS review were
based on a four-tiered scheme [1

]. The studies
reviewed were placed in four categories as follows. Class
I prospective, randomized, controlled trial; Class II
prospective, controlled trial, lacking one of the criteria
(a)(d); Class III controlled trial in a representative
population; Class IV evidence from uncontrolled stu-
dies. Criteria (a)(d) (a) primary outcome is clearly
dened; (b) exclusion criteria are clearly dened;
(c) drop-outs adequately counted and cross-over used;
(d) baseline characteristics equivalent among treatment
groups.
One Class I, one Class II and ve prospective Class III
studies demonstrated that ACTH is probably effective
in the short-term treatment of infantile spasms and in the
91
resolution of hypsarrhythmia [1

, p. 1673]. One Class II

and several Class III studies showed the limited efcacy
of oral corticosteroid treatment in infantile spasms. It was
concluded that adrenocorticotropic hormone or cortico-
tropin (ACTH) is more effective than corticosteroids,
which lacked any evidence of effectiveness in the treat-
ment of infantile spasms. Use of ACTH is recommended.
However, a large randomized UK study [7

] of 107
children using much larger doses (prednisolone 40 mg/
day versus 2 mg/kg prednisone in the earlier studies) did
not show any difference in comparison to tetracosactide
(Synachten Depot) 0.5 mg/day (40 IU) given on alternate
days. Response rates on day 14 were 70% and 76%,
respectively. This study excluded infants with tuberous
sclerosis in which vigabatrin has been considered to be
the drug of choice.
ACTH is recommended by the author because adrenal
hypofunction is to be expected after very large doses of
prednisolone.
The mechanism of action of ACTH on infantile spasms is
unknown. According to one recent theory, ACTH admin-
istration downregulates the expression of corticotrophin-
releasing hormone and stress hormones [8,9], and the
action of ACTH via steroids may be of minor importance.
Because oral steroids also seem to be as effective as
ACTH when used in maximal doses [7], other mechan-
isms may also be involved [10

]. Hormonal therapy may

accelerate physiological events during critical stages of
brain development through action as a trophic hormone
[2,5

].
Dosage of corticotrophin
Corticotrophin, so-called natural ACTH acts for 12
18 h and its synthetic derivative, Zn tetracosactrin, for
2448 h. Corticotrophin 100 IU is equivalent to 1 mg
tetracosactrin. The relatively serious side effects of Zn
tetracosactrin are probably due to its prolonged action
[2,3]. The doses used vary considerably in different
countries: Japan 314 IU/day, Finland 1836 IU/day,
USA 80 IU/day. It is not clear why infants in the USA
would need considerably larger doses than in Japan.
Many authors use tetracosactide on alternate days
because of its prolonged action.
Nine prospective studies were analysed in the AAN and
CNS review [1

], each including on average 40 children.

It was concluded that there are insufcient data to
recommend the optimal dosage and duration of treatment
of infantile spasms.
This review does not include two Finnish studies: a large
Class III study, including 162 children, in which large
doses (80120 IU/day) were no more effective than
smaller doses of 2040 IU/day in the short term
[3,4

,11]; and another prospective study, including 30

children, in which the therapy was individualized accord-
ing to etiology, with a starting dose of 3 IU/kg given for a
short period [5

,12].
Small doses and short duration of therapy have been
recommended [2,3,6].
Vigabatrin
In the AAN and CNS review [1

], 14 studies met the

inclusion criteria for analysis of vigabatrin treatment: one
randomized, placebo-controlled study (Class I), two Class
III studies and ve retrospective case series (Class IV).
The dosages of vigabatrin varied from 18 to 200 mg/kg/
day. The small Class I study showed that difference in
response was not statistically signicant when compared
with placebo. Two Class III randomized controlled trials
and the majority of Class IV trials showed that less than
half of those treated responded. It was concluded that
vigabatrin is possibly effective for the short-term treat-
ment of infantile spasms. Vigabatrin is considered by
the Food and Drug Administration to be experimental
and unproven for all indications, and is not licensed for
use in the United States (p. 1676).
In a recent randomized UK study (excluding tuberous
sclerosis patients) [7

], vigabatrin was effective in 54%of

the children. The difference in favour of hormonal treat-
ment on spasms (70%) was signicant at day 14. Hypsar-
rhythmia resolved signicantly more often with hormonal
treatment [7

].
Vigabatrin seems to be less effective than hormonal
treatment in the short-termtreatment of infantile spasms.
Tuberous sclerosis
There is evidence to suggest that vigabatrin is the treat-
ment of choice in patients with tuberous sclerosis with
infantile spasms, with a mean response rate of 74% in
eight studies [3,4

]. However, the response rates with

ACTH (73%) and vigabatrin are comparable [4

].
Because treatment with ACTH is always given for a
specied period and vigabatrin is administered for an
indenite period, vigabatrin is a better choice.
Relapses
The relapse rates after corticotrophin treatment were
1524% in the Class IIII studies, 33% in the Class II
study, and 1924% in Class III studies [1

]. There are
insufcient data on the relapse rate after vigabatrin
administration. However, the numbers of patients who
were seizure-free at 34 months after ACTH or vigaba-
trin treatment are very similar (4244%) [4

].
92 Developmental disorders
It seems highly likely that the gross EEG abnormality of
hypsarrhythmia is responsible, at least in part, for the
cognitive decline, irrespective of other seizure types. The
outlook is much improved for children who respond to
treatment.
Therapy: long-term outcome
Seven Class III and IV studies have provided results on
the long-term outcome of hormonal treatment [1

].
None was randomized or controlled, ve were prospec-
tive and two were retrospective series. The total number
of patients was 457 and the mean number of patients per
study was 65. The mean duration of follow-up was
50 months to 10 years. The data have produced conict-
ing results. The conclusion reached in the review by the
AAN and CNS was that there is insufcient data to
conclude that any treatment improves the long-term
prognosis for cognitive outcome or reduce the later inci-
dence of epilepsy, or show that early initiation of treat-
ment should be used to improve the long-term outcome
(p. 1679). The long-term outcome after vigabatrin has not
been sufciently evaluated.
This review does not include a large Finnish cohort
including 214 children with follow-up to adulthood [6].
In this study the death rate, autopsy ndings, intellectual
outcome (using standardized psychometric assessments),
education, occupations, psychiatric disorders and occur-
rence of epilepsy and follow-up of EEG were reported.
My own conclusion regarding long-term results is based
on the ndings of this study. They show that patients
with infantile spasms who had responded with cortico-
trophin have a better cognitive outcome than other
treatment groups. This holds true both for patients with
a cryptogenic etiology (14/20 patients or 70%) and known
etiology (18/78 or 23%) [4

]. The long-term cognitive

outcome was better in children treated with smaller doses
(2040 IU/day) than with larger doses (120 IU/day) [3].
Early treatment (<6 weeks) resulted in a better cognitive
outcome, as was also found in three other series that have
been reviewed [6].
A recent retrospective study by Kivity et al. [13

] of
patients treated with high-dose tetracosactide followed
by oral steroids (for about 9 months) has resulted in a
normal cognitive outcome in all patients when treated
within 1 month of onset of infantile spasms, but only in
40% when treated later. This study included only cryp-
togenic patients. The number of patients with a normal
outcome in the whole series (18%) did not differ from
another study in which shorter treatments were used
(17%) [6].
Although evaluation of treatment in terms of cognitive
outcome is difcult because double-blind, placebo-
controlled trials would be unethical, there does seem
to be strong evidence for the importance of adequate
early intervention [6,14].
Side effects
The side effects of ACTH include infections, arterial
hypertension, hypertrophic cardiomyopathy and adrenal
hypofunction. For infants with a history of frequent
respiratory infections, prophylactic trimetoprim-sulfaxa-
zole therapy is recommended. Hypertension should
be monitored and treated. In hypertensive patients,
the heart should be monitored using ultrasound. The
dose should be tapered slowly. Full cortisol substitution
should be given after ACTH and increased during
periods of stress such as high fever [2].
Unlike the side effects of vigabatrin, the side effects
of steroids are well known and treatable. They are
also reversible upon discontinuation of the drug.
To minimize the side effects, the child should be
closely watched and therapy should be given using
the minimal effective dosage for the minimal effective
time possible.
A new concern has arisen in relation to the use of
vigabatrin concentric visual eld defects (VFDs),
which seem to be permanent and cause a considerable
handicap. VFDs are, for example, sufcient to prevent
individuals from driving. Concerns also exist especially in
children who might have poor underlying visual function
as part of their overall neurological disabilities. VFDs
may appear after only 68 weeks of treatment [15,16].
VFDs have been found in a total of 35% of older children
in six published series, and in 39% of adults in 15 series
[4

]. New markers of early damage are being developed

(e.g. the scotopic threshold response, an electroretino-
graphic marker) [17].
It is likely that VFDs are as common in children as in
adults, and there is no minimum period of safe use. When
children with infantile spasms are capable of undergoing
a complete ophthalmologic examination to document
whether there are indeed visual problems associated with
the use of vigabatrin, an examination should be per-
formed. Since no methods are currently available to
identify VFDs, steroids should be used as treatment of
rst choice.
Recent reports of studies in animals have also been
alarming. Activation of GABAa receptors may trigger
apoptotic neurodegeneration [18]. Drugs that increase
brain concentrations of GABA, like vigabatrin, can also
cause apoptotic degeneration in the developing brain
[19

]. The vulnerability period coincides with the brain

spurt period. Damage can be seen early and at low doses
of vigabatrin. Of greatest concern is the axonal degenera-
tion which is irreversible [20].
The latest on infantile spasms Riikonen 93
The possible benets of vigabatrin therapy do not justify
the risks because of irreversible effects on the retina and
possibly also on neurons in the developing brain.
In general, the above-mentioned reviews show the dif-
culties in comparing the results of clinical series of
infantile spasms from the literature, because of their
nonhomogeneity and different denitions of response
rates used in the series. A good response rate should
include both total cessation of the spasms and also dis-
appearance of hypsarrhythmia in the sleep EEG. Follow
up varies from several months to several years. Some
authors have treated only children whose outcome was
likely to be favourable with a number of drop-outs from
the series. Furthermore, the modalities of treatment have
been extremely variable. Side effects of therapy (for
example, hypertension during steroid therapy) are often
poorly described.
The outcome is also strongly inuenced by the etiology.
Structural abnormalities often imply a bad prognosis
while cryptogenic cases (with no known pathology) have
the best outcome.
New etiological aspects
The following sections detail the causes of infantile
spasms
Genetics
The Aristales-related homeobax gene (ARX) has emerged
as an important gene with crucial roles in brain develop-
ment and human disorders. X-linked recessive infantile
spasms have recently been associated with mutations in
the ARX gene [21]. A mutation in the ARX gene has also
been reported in sporadic cryptogenic infantile spasms
[22]. Mutations of ARX are associated with striking
pleiotrophy and consistent genotypephenotype correla-
tions [23

]. It is now clear that different mutations in the

ARX gene produce various clinical symptoms. A nearly
continuous series of developmental disorders begins
with hydrancephaly, lissencephaly, and agenesis of the
corpus callosum, and ends with a series of overlapping
syndromes with apparent normal structure [24].
Metabolic diseases
The study of Sadlier et al. [25] demonstrates that infantile
spasms may be an important clinical manifestation of
mitochondrial disorders in children, and that in spite of
this blood lactate may be normal. Children with unex-
plained infantile spasms, particularly with mental delay,
should be investigated; at least four tests of blood lactate
in both the fasted and postprandial state are necessary.
Tuberous sclerosis
Neuropsychiatric problems are often found in tuberous
sclerosis patients [26]. One of these problems is autism,
which is not associated with any special genotype [27].
Tubers in the temporal region have been associated with
autistic symptoms [4

,5

,14,26]. The presence of a

TSC2 mutation was found to carry a higher risk of low
IQ, autistic disorder and infantile spasms than a TSC1
mutation [27]. The study of OCallaghan et al. [28]
showed that both the number of tubers and a history
of infantile spasms are independently associated with
cognitive deciency.
Defective cortical development
A rare regional malformation, posterior quadrantic dys-
plasia of the brain in children suffering from infantile
spasms, was reported by DAgostino et al. [29]. Wide-
spread cortical dysplasia is more frequent in the posterior
quadrant and represents either hemi-hemimegalence-
phaly or multilobar cortical dysplasia. Intractable epi-
lepsy in these patients may be alleviated by a large
quadrantic temporoparieto-occipital resection.
Conclusion
There is still a lack of consensus about the drug of rst
choice in infantile spasms. The common use of vigabatrin
as the rst-line drug in Europe is only based on its ease of
use, not on effectiveness. The recommendations of the
present author are as follows:
(1) Use ACTH as the rst choice for treatment because it
is a safe drug when used at the minimal effective dose
and duration; vigabatrin is the rst-line drug for the
treatment of tuberous sclerosis.
(2) Treat early.
(3) Side effects of ACTH, unlike of those of vigabatrin,
are well known, treatable, and reversible.
(4) We know the long-term effects of ACTH.
There are several important goals for the future. Identi-
cation of the frequency of VFDs and quantication of
the long-term effects in general (epilepsy, cognitive and
behavioural outcome) after vigabatrin administration in
children treated for their spasms. In addition, predictive
methods for vigabatrin retinal toxicity need to be found.
Finally, understanding of brain maturation should be
the main focus of studies to elucidate the pathophysio-
logical mechanisms of infantile spasms which remain
unclear.
More accurate etiological diagnoses can be made. Various
genetic patterns, mitochondrial diseases, and new mal-
formation syndromes have been recognized.
The prognosis in tuberous sclerosis can be better pre-
dicted. The number of tubers and infantile spasms are
important in the prediction of outcome. Location of tubers
may be important for the cognitive, behavioural and
epileptic outcome of the child with tuberous sclerosis.
94 Developmental disorders
Acknowledgements
Supported by grants from the University Hospital of Kuopio.
References and recommended reading
Papers of particular interest, published within the annual period of review, have
been highlighted as:
of special interest
of outstanding interest
1

MacKay M, Weiss S, Adams-Webber T, et al. Practice parameter: medical

treatment of infantile spasms. Report of the American Academy of Neurology
and the Child Neurology Society. Neurology 2004; 62:16681681.
This important article describes database searches of Medline from 1966 and
EMBASE from 1980 to 2002 analysing the articles and making recommendations
on a four-tiered classication scheme of 159 articles chosen for detailed review.
2 Riikonen R. ACTHtherapy of West syndrome: Finnish views. Brain Dev 2001;
23:642646.
3 Riikonen R. Steroids or vigabatrin in the treatment of infantile spasms? Pediatr
Neurol 2000; 23:403408.
4

Riikonen R. Infantile spasms: therapy and outcome. J Child Neurol 2004;

19:401404.
The aim of this experienced author is to critically analyse the literature in terms of
therapy and outcome of infantile spasms. Visual eld defects of children reported in
the literature are as frequent as in adults.
5

Riikonen R. West syndrome. In: Wallace S, Farrell K, editors. Epilepsy in

children. London: Arnold; 2004. pp. 142147.
The recent treatment schedule recommended by the author is given here.
6 Riikonen R. Long-term outcome of patients with West syndrome. Brain Dev
2001; 23:683687.
7

Lux A, Edwards S, Hancock E, et al. The United Kingdominfantile spasmstudy

comparing vigabatrin with prednisolone or tetracosactide at 14 days: a
multicentre, randomised controlled trial. Lancet 2004; 364:17731778.
This is the largest trial that has ever been done in the eld of infantile spasms
comparing the effects of the main therapies, hormonal (ACTH and prednisolone)
and vigabatrin, in the short-term treatment of infantile spasms.
8 Brunson K, Khan N, Eghbal-Ahmadi M, BaramT. Corticotropin acts directly on
amygdala neurons to down-regulate corticotropin-releasing hormone gene
expression. Ann Neurol 2001; 49:304312.
9 Wang W, Murphy B, Dow K, et al. Systemic adrenocorticotrphic hormone
administration down-regulates the expression of corticotropin-releasing hor-
mone (CRH) and CRH-binding protein in infant rat hippocampus. Pediatr Res
2004; 55:604610.
10

Rho J. Basic science behind the catastrophic epilepsies. Epilepsia 2004;

45(Suppl 5):511.
Three basic theories for pathogenesis of infantile spasms are presented. The
corticotropin-releasing hormone hypothesis, the N-methyl-D-aspartate (NMDA)
hypothesis, and the serotonin-kynurenine hypothesis.
11 Riikonen R. A long-term follow-up study of 214 children with the syndrome of
infantile spasms. Neuropediatrics 1982; 13:1423.
12 Heiskala H, Riikonen R, Santavuori P, et al. West syndrome: individualized
ACTH therapy. Brain Dev 1996; 18:456460.
13

Kivity S, Lerman P, Ariel R, et al. Long-term cognitive outcome of a cohort of

children with cryptogenic infantile spasms treated with high dose adreno-
corticotropic hormone. Epilepsia 2004; 45:255262.
Presents long-lasting steroid treatment for cryptogenic spasms.
14 Besag F. Behavioral aspects of pediatric epilepsy syndromes. Epilepsy Behav
2004; 5:S3S13.
15 Krauss G, Johnson M, Miller N. Vigabatrin-associated retinal cone system
dysfunction. Neurology 1998;614618.
16 Gross-Tsur V, Banin E, Shahar E, et al. Visual impairment in children treated
with vigabatrin. Ann Neurol 2000; 48:6064.
17 Parisi P, Tommasini P, Piazza G, Manfredi M. Scotopic threshold response
changes after vigabatrin therapy in a child without visual eld defects: a new
electroretinographic marker of early damage? Neurobiology Dis 2004;
15:573579.
18 Ikonomidu C, Bittigau P, Ishimaru M, et al. Ethanol-induced apoptotic neu-
rodegeneration and fetal alcohol syndrome. Science 2000; 287:1056
1060.
19

Bittigau P, Sifringer M, Ikonomidu C. Antiepileptic drugs and apoptosis in the

developing brain. Ann NY Acad Sci 2003; 993:103114.
GABA-mediated inhibition (vigabatrin) of the seizures cause widespread and
dose-dependent apoptotic neurodegeneration in the developing brain during
the brain growth spurt period in rats.
20 Sindhu R, Bigigo M, Tuor U, Seshia S. Low-dose vigabatrin (gamma-vinyl
GABA)-induced damage in the immature brain. Exp Neurol 1997; 144:400
405.
21 Stro mme P, Mangelsdorf M, Sheffer I, et al. Infantile spasms, dystonia and
other x-linked phenotypes caused by mutations in Aristales related homeobox
gene, ARX. Brain Dev 2002; 24:266268.
22 Gutierrez-Delicado E, Serratosa J. Genetics of the epilepsy. Curr Opinion
Neurol 2004; 17:147153.
23

Kato M, Das S, Petras K, et al. Mutations of ARX are associated with striking
pleiotropy and consistent genotypephenotype correlation. Hum Mutat
2004; 23:147159.
Enlarging clinical picture of ARX mutations.
24 Partington M, Turner G, Boyle J, Gecz J. Three new families with x-linked
mental retardation caused by the 428451dup(24bp) mutation in ARX. Clin
Genet 2004; 66:3945.
25 Sadlier L, Connolly M, Applegarth D, et al. Spasms in children with denite
and probable mitochondrial disease. Eur J Neurol 2004; 11:103110.
26 Asato M, Ardan A. Neuropsychiatric problems in tuberous sclerosis complex. J
Child Neurol 2004; 19:241249.
27 Lewis J, Thomas H, Murphy K, Sampson J. Genotype and psychological
phenotype in tuberous sclerosis. J Med Genet 2004; 41:204207.
28 OCallaghan F, Harris T, Johnson C, et al. The relation of infantile spasms,
tubers, and intelligence in tuberous sclerosis complex. Arch Dis Child 2004;
89:530533.
29 DAgostino M, Bastos A, Piras C, et al. Posterior quadrantic dysplasia or hemi-
hemimegalencephaly: a characteristic brain malformatiom. Neurology 2004;
62:22142220.
The latest on infantile spasms Riikonen 95