Review Article

Management of infantile spasms

Gary Rex Nelson

Division of Child Neurology, University of Utah School of Medicine, Salt Lake City, USA
Correspondence to: Gary Rex Nelson, MD. Division of Child Neurology, University of Utah School of Medicine, Salt Lake City, UT 84113, USA.
Email: [email protected].

Abstract: West syndrome, or infantile spasms syndrome is a frequently catastrophic infantile epileptic
encephalopathy with a variety of etiologies. Despite the heterogeneous nature of causes of infantile spasms,
a careful diagnostic evaluation can lead to diagnosis in many patients and may guide treatment choices.
Magnetic resonance imaging (MRI) brain remains the highest yield initial study in determining the etiology
in infantile spasms. Treatment of infantile spasms has little class I data, but adrenocorticotropic hormone
(ACTH), prednisolone and vigabatrin have the best evidence as first-line medications. Other therapies
including the ketogenic diet and other anti-epileptics medications may also prove useful in the treatment
of infantile spasms. In general, more studies are needed to determine the best treatment regimen for this
condition. Prognosis is generally poor, with the majority of patients having some or profound neurocognitive
delays. Patients without delays at diagnosis and without an identifiable etiology, if treated appropriately, have
the greatest likelihood of a normal outcome.

Keywords: Infantile spasms; West syndrome; adrenocorticotropic hormone (ACTH); corticosteroids;

hypsarrhythmia

Submitted Jul 08, 2015. Accepted for publication Aug 28, 2015.
doi: 10.3978/j.issn.2224-4336.2015.09.01
View this article at: http://dx.doi.org/10.3978/j.issn.2224-4336.2015.09.01

Introduction spasms, often involving the extremities and head/neck;

(II) hypsarrhythmia on electroencephalogram (EEG);
In 1841, Dr. William James West wrote a letter to Lancet
and (III) subsequent or concurrent intellectual disability
describing new infantile convulsions in his 4-month-old
(3,4). In the Delphi consensus, delay is not required (5).
son (1,2). His apt description of the events, starting with a
It is important to note that although West syndrome and
subtle head drop and progressing to clusters of myoclonic
infantile spasms are frequently discussed interchangeably,
spasms occurring many times daily, accompanied by the infantile spasms exist without delay in development and
regression of development of his previously normal child, sometimes even without hypsarrhythmia. Patients with
remains the hallmark of this condition today. His letter infantile spasms are often categorized as symptomatic,
also embodies the distress of a father and medical provider cryptogenic or idiopathic (4). Symptomatic infantile spasms
caused by this rare, difficult-to-treat, and frequently, have a known structural, metabolic or genetic cause and
neurologically devastating condition. While infantile occur in children with delays present prior to the onset
spasms were recognized in the medical community prior of spasms. Cryptogenic infantile spasms lack a known
to this letter, it was a rare disorder with no clear treatment. etiology, but share the prior developmental delays. Finally,
Over 170 years later, this most common form of infantile idiopathic infantile spasms occur in patients with normal
epilepsy (3) has more treatment options and identified development prior to the onset of spasms and without an
etiologies, but can still be just as distressing and difficult to identifiable cause. This group continues to decrease in size as
treat as in the case of Dr. West’s son. more testing reveals previously unknown causes of infantile
Infantile spasms syndrome, frequently referred to as spasms. The US consensus report on infantile spasms (3)
West syndrome, is a triad of (I) clinical flexor or extensor simplifies this categorization to a dichotomous one and

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

Translational Pediatrics, Vol 4, No 4 October 2015 261

defines symptomatic infantile spasms as those patients Diagnostic evaluation

with either abnormal development and/or a clear etiology
An evaluation including a thorough history and physical
for the infantile spasms and cryptogenic spasms as those
examination and EEG, is important in the diagnosis of
occurring in the context of normal development without
infantile spasms. Once the diagnosis is established, efforts
a clear etiology. For simplicity, this review will use the
should be made to establish the underlying etiology, as this
dichotomous definition. Management of infantile spasms
significantly affects treatment decisions and prognosis (15).
should include efforts to determine the etiology if possible,
For example, patients with tuberous sclerosis are more
as this can help determine the likelihood of response to
likely to respond to treatment with vigabatrin, whereas
treatment, may guide therapeutic decisions, and can help to
patients without an identified etiology may respond better
provide families with a more definitive prognosis for their
to hormonal therapy (16,17). The differential diagnosis
child (4,6,7).
in infantile spasms is broad, and it is therefore important
Infantile spasms occur in roughly 2-3 per 10,000
that the evaluation follow a step-wise process to limit
live births, with peak incidence at 6 months of age and
the number of low-yield or unnecessary tests that are
less than 10% of cases presenting after 12 months of performed.
age (4,8). Unfortunately, due to the subtle and brief
nature of spasms in some patients, there can be delays
in diagnosis. There is evidence to suggest that earlier History and physical examination
treatment (within one month of onset) is more effective As with most conditions in neurology, the history and
in controlling spasms (9-12), and may improve outcomes. physical examination are the first and most important step
Despite insufficient data to determine if this is true for in evaluating a patient with the concern of infantile spasms.
all etiologies of infantile spasms, treatment should be Due to significant advances in personal technology devices,
initiated as soon as possible and providers should be aware a growing number of parents present with video of the
of the often subtle presentation to allow for more rapid events and have access to internet video of other infants
diagnosis and treatment. The goals of therapy should with infantile spasms (18). These clinical events may vary
include a complete cessation of the clinical events and from subtle, single, spasms involving a flexion of the neck
resolution of hypsarrhythmia or modified hypsarrhythmia or bowing of the head, to full body flexor, extensor or
on video EEG (5). Neurological and/or developmental mixed myoclonic spasms occurring in clusters many times
outcomes in patients with infantile spasms are usually throughout the day. Spasms tend to occur more often on
poor (13). Children with symptomatic spasms more awakening or when falling asleep (19). The brief nature
frequently exhibit neurological deficits and cognitive and of these events, as well as the similarity to other infantile
developmental delays, while a higher percentage of patients movements (benign sleep myoclonus, Moro reflex, etc.),
with idiopathic/cryptogenic infantile spasms may have a can lead to a delay in diagnosis (20). When possible, it is
normal or near-normal outcome if appropriate treatment helpful for providers to witness the events or capture them
is initiated in a timely fashion (9-13). This review will with video EEG. If this is not possible, parental reports
discuss the diagnostic evaluation and treatment options consistent with infantile spasms should prompt immediate
for children diagnosed with infantile spasms, followed by investigation. History should focus on the semiology of the
a brief discussion of prognosis. Despite multiple practice events, frequency, clusters versus single spasms, changes
parameters and Cochrane reviews in the last 15 years, in development or delays prior to the onset of spasms,
there remains considerable diversity among providers in family history of similar events in infants, and a thorough
their approach to the treatment of infantile spasms (14). birth and pregnancy history (3,15,19). Perinatal events are
This represents the paucity of data to clearly demonstrate responsible for up to 30% of cases of infantile spasms with a
a gold standard and limited studies comparing treatment known etiology, and include neonatal strokes, infections and
modalities. This is a natural consequence of the varied hypoxic-ischemic encephalopathy. It is not clear why some
etiologies and rare occurrence of this condition, and patients with these conditions develop infantile spasms
continued efforts are being made worldwide to provide and others do not, but there may be predisposing genetic
better information to guide decision-making in this severe factors. Prenatal causes include cortical and structural
form of infantile epilepsy. malformations and other genetic causes and are responsible

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

262 Nelson. Infantile spasms

for up to half of symptomatic etiologies (21,22). (6,15,28-30). However, 5-20% of patients may have non-
Physical examination should focus on neurocutaneous diagnostic abnormalities (3,6). Since tuberous sclerosis
stigmata of disease, which can identify patients warrants a different treatment consideration (13,16), it
with tuberous sclerosis, neurofibromatosis or other is recommended that MRI brain be performed prior to
phakomatoses. Wood’s lamp evaluation can help identify treatment initiation when possible. Repeat imaging may be
more subtle cases of tuberous sclerosis. While the majority useful if the initial MRI is normal and other testing is not
of patients with infantile spasms do not have tuberous informative (3). Unless there are other factors that suggest
sclerosis, they may make up 25% of symptomatic infantile that repeat imaging should be done more urgently, it is
spasms cases and up to 50% of patients with tuberous recommended that repeat MRI be deferred until after 24-
sclerosis will develop infantile spasms (17,23). Parents 36 months in order to increase yield. Focal EEG findings in
of children with tuberous sclerosis should be counseled a refractory patient may prompt repeat imaging sooner, as
regarding infantile spasms early on. Other syndromic these patients may benefit from surgical resection.
features noted on examination can also help to direct Computed tomography (CT) is not considered adequate
the evaluation toward genetic causes of infantile spasms. to evaluate for most causes of infantile spasms (4,31), but
Trisomy 21 alone accounts for 3-6% of patients with it is recognized that in some areas of the world, this is the
infantile spasms and is nearly always diagnosed prior to the only diagnostic imaging available. CT may help identify
onset of infantile spasms (6,24). Familial infantile spasms evidence of prior brain injury, infection or tumors that can
are rare, but may have specific genetic mutations (4). be the cause of infantile spasms (6).

Electroencephalogram (EEG) Further evaluation

EEG should be performed as soon as possible on any infant After history, physical examination, EEG and MRI brain
with concerns for infantile spasms. When possible, 24-hour are performed, about 70% of patients have a diagnosis (3).
video EEG is preferred (5). However, a routine video EEG There are limited studies on the recommended evaluation
may be adequate when making the diagnosis, especially of infantile spasms beyond imaging and wide practice
when the features of hypsarrhythmia (25) are present variation exists among pediatric neurology providers,
interictally. Since sleep is an important part of the EEG at least in the USA (14). The highest yield studies are
evaluation for infantile spasms, every effort should be made genetic studies (6), but there are a variety of available
to include non-REM sleep as part of the EEG evaluation. panels and candidate genes in infantile spasms. Single-
This may be decreased in infants with infantile spasms, but nucleotide polymorphism (SNP) and comparative
hypsarrhythmia may be present in this stage of sleep even if genomic hybridization (CGH) microarray are attractive
absent while awake (26). Up to 1/3 of patients with infantile options, but may miss some pathogenic mutations and
spasms will not have hypsarrhythmia or may have other may also identify variants of unknown significance (VUS).
EEG abnormalities (19,27). In these cases, if the clinical Pathogenic mutations may be identified in about 10-
spasms are consistent with infantile spasms, evaluation and 23.5% of patients without a known etiology, with VUS
treatment should still proceed accordingly. EEG is used present in 14.8% (6,32). Delayed patients have an even
for diagnosis as well as response to treatment (5). There is higher yield. Copy number variants (CNV) have also
likely little utility in repeat EEG in patients who are not been detected in >10% of patients with infantile spasms
responding to treatment unless the diagnosis is in question and normal imaging as part of the Epilepsy Phenome/
or if the EEG was normal previously. Genome Project (EPGP) (32). Candidate and associated
genes for infantile spasms include: SCN1A, SCN2A,
MAGI2, YWHAD, HIP1, 9p deletion, 15q11 duplications,
Brain imaging
GABRB3, an unbalanced translocation t(15;16), ARX,
Imaging is not required to make the diagnosis of infantile CDKL5, TSC1 and 2, trisomy 21 (24), 1p36 deletion,
spasms, but is the single-most important method to and possible PNPO mutations in cases of pyridoxine-
identify etiology and/or direct further testing in infantile dependent epilepsies with infantile spasms, as well as
spasms. Fifty to 73 percent of patients have an identifiable others (4,21,23,32-35). Karyotype is often performed,
etiology on magnetic resonance imaging (MRI) of the brain and may detect translocations and other rearrangements

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

Translational Pediatrics, Vol 4, No 4 October 2015 263

History and physical

examination

Consistent with spasms

Video EEG
Hypsarrhythmia No Hypsarrhythmia*
Sleep/wake

24 hour VEEG or repeat

MRI brain
routine EEG in 1-2 weeks

EEG confirms infantile No EEG evidence of

Diagnostic  Normal
spasms or history still spasms and no further
concerning clinical events

Directed testing as SNP or CGH

appropriate microarray Consider other
Follow MRI brain
diagnoses, follow
algorithm
development

Epilepsy panel

Metabolic studies

Figure 1 Diagnostic algorithm for infantile spasms. As noted, imaging has the highest yield, followed by genetic studies. *, Patients with a
convincing history of infantile spasms and abnormalities other than hypsarrhythmia on EEG, may proceed to imaging without repeat EEG.
If the concern for infantile spasms is significant and development is affected, it is not recommended to wait for EEG abnormalities, as up
to 1/3 of patients may not have hypsarrhythmia. Treatment should begin once imaging is complete unless there is concern for infection and
other work-up is needed. EEG, electroencephalogram; MRI, magnetic resonance imaging; ACTH, adrenocorticotrophic hormone; SNP,
single nucleotide polymorphism; CGH, comparative genomic hybridization.

missed on microarray, but reported yield has been low testing may include acylcarnitine profile, total and free
except in syndromes notable prior to onset of spasms such carnitine, serum and CSF lactate and glucose, serum
as trisomy 21 (6). Gene panels may identify genetic causes pyruvate, urine organic acids, serum and CSF amino acids,
of IS not evident on other genetic testing and are typically and CSF neurotransmitter metabolites, including pyridoxal-
higher yield and less expensive than step-wise individual 5’-phosphate.
gene testing. Based on the limited evidence, microarray ± Based on probability of diagnostic yield, a sequential
karyotype and followed by an appropriate epilepsy gene diagnostic algorithm is recommended (Figure 1). These
panel are likely to have the highest yield (6). diagnostic methods yield an etiology in 70-85% of
Inborn errors of metabolism are detected in up to 5% patients (4,6,15).
of patients with infantile spasms (6). Imaging and, where
available, newborn screening may suggest a possible inborn
Treatment
error of metabolism. Patients with early onset spasms and/
or refractory to treatment warrant further evaluation with There are limited data to guide the treatment of infantile
metabolic studies, recognizing the low-yield, but possible spasms (13,16,26). Most studies are retrospective or small
treatment changes to improve outcomes (6,15). Metabolic prospective studies. EEG data is not available in all studies as

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

264 Nelson. Infantile spasms

Cryptogenic Symptomatic

TSC Other

ACTH or
prednisolone
ACTH, prednisolone or
Vigabatrin
vigabatrin

Choose a different
Repeat EEG and exam within 14 days
first-line medication

Hypsarrhythmia resolved Still hypsarrhythmia or

and spasms ceased spasms

Failed first-line treatments

Complete treatment Consider other

course treatment options

Figure 2 Treatment algorithm. Suggested sequence of treatment, with etiology incorporated. Data is inadequate to recommend a consistent
first-line therapy, except in the case of tuberous sclerosis. It is recommended to try two first-line therapies before moving on to other
options, as there is less data to support the second-line treatments. Studies have reported patients that respond to ACTH after failing
corticosteroids and vice-versa. TSC, tuberous sclerosis complex; ACTH, adrenocorticotropic hormone; EEG, electroencephalogram.

some rely on parental report. Some studies use improvement guided by provider preference and experience due to limited
in spasm frequency while others use complete resolution of data. Doses and treatment lag for each medication vary and
spasms. We will only discuss results in terms of complete are described in Table 1.
resolutions of infantile spasms as this is the typical goal
of treatment and partial resolution is more subjective
ACTH
(5,13,16,26). However, even when complete cessation of
spasms is required, there are some variations in the timeframe/ ACTH is probably the most universally accepted first-
method that this is reported. In addition, many studies report line treatment of infantile spasms, stemming from the
a rate of relapse as high as nearly 50% after initial response class 1 randomized controlled trials performed from
(13,36-38), further complicating interpretation of the results. 1983-1999 that show its efficacy (36-39). In 2004, the
Cost and availability of medications vary substantially American Academy of Neurology and Child Neurology
throughout the world and may limit choices in some regions Society concluded that ACTH is “probably effective” in
and patient populations. Hormonal therapies, including the treatment of infantile spasms (13). Five studies were
adrenocorticotropic hormone (ACTH) and corticosteroids, considered class 1 evidence, and the rates of cessation of
and vigabatrin have the most evidence to support their use spasms were better than any previously reported, ranging
in infantile spasms (13,16). It is important to note, however, from 42-87% (13,36-39). Response is typically seen within
that robust data is lacking in all treatment options (26), 14 days or sooner (13,36-40). Despite the promising data,
and randomized, controlled, multicenter comparative ACTH has its own drawbacks and limitations.
trials are needed. Many other treatment options continue First, the relapse rate in these studies is high in
to be explored, including the ketogenic diet, traditional ACTH (13,36-38). There are variations in dosing and
anti-epileptic medications, and resective surgery in select d u r a t i o n t h a t make comparison and recommended
cases (15). A treatment algorithm is suggested (Figure 2), treatment regimens difficult to formulate (26). Side effects
with the recognition that some treatment choices will still be have been severe and frequent. In addition, the cost of

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

Translational Pediatrics, Vol 4, No 4 October 2015 265

Table 1 Treatment options in infantile spasms

Medication Dose Route Time to effect Duration of therapy Side effects
First-line
ACTH 150 units/m2/day IM 14 days 14 days, then taper Hypertension, infection,
Prednisolone 8 mg/kg/day PO 14 days 14 days, then taper weight gain, irritability
or 40-60 mg/day
Vigabatrin 50-150 mg/kg/day PO 14 days 6-9 months, then Visual field loss, MRI
taper changes, irritability
Second-line
Ketogenic diet 3:1-4:1 ratio PO 1-3 months At least 6 months Constipation,
nephrolithiasis, acidosis
Topiramate 4-20 mg/kg/day PO 1+ months Unclear Anorexia, hypohidrosis,
Zonisamide 4-20 mg/kg/day PO Up to 3 weeks Unclear nephrolithiasis
Pyridoxine 100 mg IV Immediate Once only Cardiorespiratory
depression
Daily: PO 1-2 weeks Lifelong Neuropathy at high
100-400 mg/day doses
First-line medications have the most data to support their use. Order is not in order of recommendation of use for either set
of medications. Pyridoxine is given in a one-time dose by IV with EEG and cardiorespiratory monitoring and repeated orally
if effective. Common side effects listed, but not an inclusive list of all side effects. ACTH, adrenocorticotropic hormone; EEG,
electroencephalogram; MRI, magnetic resonance imaging; IM, intramuscular; PO, by mouth; IV, intravenous.

ACTH in the United States and elsewhere continues to (20 units/day) and high-dose (150 units/m2/day) ACTH
climb (27,38). However, in a recent survey, about 2/3 with response rates of 50% and 58%, respectively (42).
of providers were using ACTH as first-line treatment It may be appropriate to start with either 20 units/day or
of infantile spasms, with varying doses and durations of 150 units/m2 as the treatment dose for ACTH. However,
treatment (14). The relative strength of the data in favor if a patient does not respond to the lower dose, it is
of ACTH, as well as the difficulty in recruiting enough recommended to try the higher dose (13,16).
subjects in infantile spasms studies, has resulted in a paucity
of studies evaluating ACTH further. Only one other
Corticosteroids
prospective study (UKISS) has looked at ACTH since
2000, and it did not use EEG to track response (10,40,41). There is an ongoing and increasing interest in the use
However, the rate of response was similar to that previously of corticosteroids in the treatment of infantile spasms.
reported (13). In addition, although this study suggests Prednisolone is appealing due to its low cost, ready
equal efficacy between synthetic ACTH and prednisolone, availability in many countries, ease of administration,
it was not adequately powered to determine superiority of and growing evidence that it may be similar in efficacy to
either hormonal treatment. It was primarily designed to ACTH and vigabatrin.
compare hormonal treatments to vigabatrin. Initially, studies suggested that corticosteroids were
Subsequently, the 2012 revisions to the AAN/CNS inferior to ACTH. Prednisone/prednisolone were used in
Evidence-based recommendations in the treatment of lower doses, ranging from 2-3 mg/kg/day (36,37,39,43),
infantile spasms had no new evidence to evaluate the efficacy with response rates from 25-59%. In most cases, these were
of ACTH. Review of previous studies, suggests that low- significantly lower than the response to ACTH.
dose ACTH may be equivalent to high-dose (150 units/m2), Multiple studies have subsequently used higher doses of
but it is not entirely clear which low dose is to be used, prednisolone, primarily either 40-60 mg/day (38,40,41,44)
with doses varying from 0.2 units/kg to 20 units/m2 (16). or weight-based dosing of 8 mg/kg/day with a maximum
Hrachovy et al. showed equivalency between low-dose dose of 60 mg/day (27). These have shown rates of efficacy

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

266 Nelson. Infantile spasms

similar to ACTH, ranging from 67-80%. Relapse rate in as many as 30% of patients (53,54). Peripheral visual
was also similar to ACTH or lower. However, other than field constriction (46,49), has been noted as a serious side
the UK study, these are all retrospective studies. As noted effect of vigabatrin in 15-30% of patients, although it is
above, although UKISS was not powered to compared thought that this rate may be lower among patients with
ACTH with prednisolone, they had similar efficacy (40,41). infantile spasms. In responders, it is recommended to
While the data are still limited, there is promise for discontinue therapy after 6 months to limit the chances of
using prednisolone in the treatment of infantile spasms. peripheral visual field constriction (55). Less severe side
Indeed, more providers are using prednisolone as first-line effects include drowsiness, hypotonia and irritability in 13-
treatment (14,38) and the most recent AAN parameter 25% of patients (47,48,50,52,56).
acknowledges the growing evidence that suggests
prednisolone may be a reasonable initial treatment option (16).
Ketogenic diet
Prednisolone could possibly be equivalent to low-dose
ACTH (36) or even high-dose ACTH, when used at higher The ketogenic diet is used often in intractable or
doses of either 8 mg/kg/day or 40-60 mg/day (27,38,40,41). profound epilepsies, including infantile spasms, with or
The side effect profile of prednisolone has been tolerable (44) without the concurrent use of medications (35,57-59).
and may be better than that of ACTH (45). Like ACTH, Spasm freedom has been reported in 14-65% of patients
response is typically within 14 days, and if there is not within 1-3 months (57,60-63). Diet ratio ranged from
complete spasm cessation, ACTH, vigabatrin or other 3 : 1 to 4 : 1. Use as first-line therapy is limited as most
treatments should be considered. studies focused on patients with intractable seizures and
already on 1-3 anti-epileptic medications. Efficacy was
higher in infantile spasms patients treated prior to 1 year
Vigabatrin
of age (60) and among patients with cryptogenic infantile
Vigabatrin was approved in the UK in 1989 and the United spasms (63). Some patients had improved seizure control
States in 2009 for the treatment of infantile spasms and (57,60), were able to reduce medications (60), and had
intractable complex partial/focal seizures (46). The largest cognitive improvements (61), even without cessation of
randomized study of vigabatrin in infantile spasms was spasms. Ketogenic formula and young age make the diet
an open-label, randomized, single-masked, multi-center an attractive option, but there are inadequate data to
3-year study comparing high and low-dose vigabatrin (47). recommend the diet as a sole first-line therapy (64).
This 2001 study included both patients with tuberous
sclerosis and other causes and had a response rate of
Other treatments
36% overall and 52% among patients with TSC. High-
dose vigabatrin (150 mg/m 2/day) was also found to be There are limited data to support the use of other anti-
more effective than low-dose. Other studies have shown epileptics in the treatment of infantile spasms. Some have
response rates of 35-82% among cryptogenic cases and advocated the use of pyridoxine (7,65) as a brief and safe
response rates of 27-59% in symptomatic cases, with trial at the onset of spasms as it can be quickly and safely
overall rates of response ranging from 26-76%, most of trialed, using 100 mg IV once with continuous EEG and
which are greater than 50% (48-51). Response occurred cardiorespiratory monitoring during administration (15).
within 2 weeks of therapy in most patients and not after In Japan, pyridoxine has been used as first-line therapy
12 weeks (49). It is recommended to stop therapy after for infantile spasms with 15% efficacy (65), which is less
14 days if no response is noted (3). than the rate of spontaneous remission of spasms (4,19).
Vigabatrin is the preferred first-line therapy for patients Pyridoxine-dependent epilepsy can present as infantile
with infantile spasms and tuberous sclerosis (16), with spasms and should be considered as a possible etiology,
some studies showing efficacy of greater than 90% in especially in patients with younger onset.
patients with tuberous sclerosis (17). In general, vigabatrin Topiramate has some efficacy in the resolution of spasms,
has been thought to be less effective than ACTH in with response rate ranging from 10-48% (65-67), with doses
other patient populations (40,41,52), although long-term of up to 30 mg/kg being used (18) and mean dose ranging
outcomes may be similar (41). Relapse rates range from from 10-16 mg/kg/day (68). Most studies are retrospective
16-21% (47). Reversible MRI changes have been observed and are often confounded by concomitant use of other anti-

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

Translational Pediatrics, Vol 4, No 4 October 2015 267

epileptic medications and small numbers of patients. devastating infantile epileptic encephalopathy. Etiology can
Zonisamide has been used in doses ranging from be identified in 70% or more of patients and it is expected
4-20 mg/kg/day with effective doses ranging from that this will continue to improve with more widely-
5-12.5 mg/kg/day. Rates of response range from 26-41%. available genetic testing. However, MRI brain remains the
Time to response was up to 19 days (69-70). study with the highest yield in identifying etiology. Etiology
Levetiracetam has very limited data regarding its use in guides treatment decisions and affects prognosis. Although
infantile spasms and will not be discussed in detail. There is treatment may not affect the outcomes for all patients, it is
inadequate evidence to recommend its use as a second-line likely that for some cryptogenic patients it may be critical.
therapy. First-line therapies include hormonal treatments and
Benzodiazepines have been shown to have little efficacy vigabatrin and efficacy can usually be determined within
and possibly increased risk of morbidity and mortality in the about 2 weeks. Use of a standardized approach to treatment
treatment of infantile spasms (15). is associated with a higher rate of use of first-line therapies
and better rates of spasm cessation at 3-month follow-
up (75). Further studies are needed to better compare
Prognosis
first-line treatments and to determine efficacy of second-
There is no evidence that treatment alters long-term line treatments. The goal of treatment of infantile spasms
outcome in infantile spasms (19). In general, outcome is remains complete cessation of spasms. Patients diagnosed
poor and the underlying genetic and structural brain issues and treated in a timely and appropriate fashion, especially
that accompany some infantile spasms likely predispose those with cryptogenic/idiopathic infantile spasms, have the
to poor development, regardless of treatment. Epilepsy highest likelihood of a normal or near-normal outcome.
is present in up to 50% of patients (12,18,19). Autism is
present in 15-33% of patients with infantile spasms and as
Acknowledgements
high as 70% patients with tuberous sclerosis and infantile
spasms (9). Even within patients with tuberous sclerosis, None.
the presence of infantile spasms increases the risk of poor
neurodevelopmental outcome (71). However, this is not the
Footnote
case in trisomy 21 (24).
Normal or near normal development is present in only Conflicts of Interest: The author has no conflicts of interest to
15-25% (12,18-19). In 20-35 years of follow up, nearly all declare.
patients with a normal or near normal outcome held jobs (12).
Seventeen percent had an IQ >85.
References
Etiology is the most important predictor of outcome (10,33,72).
Better developmental outcomes are noted in patients 1. West WJ. On a peculiar form of infantile convulsions.
without an identified etiology (cryptogenic/idiopathic) and Lancet 1841;1:724-5.
even better in those treated with hormonal therapy rather 2. Lux AL. West & son: the origins of West syndrome. Brain
than vigabatrin (72). Short treatment lag (<1 month) is also Dev 2001;23:443-6.
associated with improved outcome (10-12,72), especially 3. Pellock JM, Hrachovy R, Shinnar S, et al. Infantile spasms:
among cryptogenic patients (73). Cryptogenic patients may a U.S. consensus report. Epilepsia 2010;51:2175-89.
have a normal or near-normal outcome up to 54.3% of the 4. Nordphysicianguides.org/Infantile-Spasms. Available
time (74), while only 12.5% of symptomatic patients have online: http://nordphysicianguides.org/Infantile-Spasms/
a good outcome (11), Mortality may be as high as 10% at 5. Lux AL, Osborne JP. A proposal for case definitions and
3 years of age and 19% at age 10 years. ACTH therapy had outcome measures in studies of infantile spasms and West
an associated 12% mortality (12). Nearly all patients stop syndrome: consensus statement of the West Delphi group.
having spasms by 3 years of age. Epilepsia 2004;45:1416-28.
6. Wirrell EC, Shellhaas RA, Joshi C, et al. How should
children with West syndrome be efficiently and accurately
Conclusions
investigated? Results from the National Infantile Spasms
Infantile spasms remain a difficult-to-treat and frequently Consortium. Epilepsia 2015;56:617-25.

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

268 Nelson. Infantile spasms

7. Shields WD. West's syndrome. J Child Neurol 2002;17 22. Poulat AL, Lesca G, Sanlaville D, et al. A proposed
Suppl 1:S76-9. diagnostic approach for infantile spasms based on a
8. Wang CJ, Jonas R, Fu CM, et al. Quality-of-care spectrum of variable aetiology. Eur J Paediatr Neurol
indicators for infantile spasms. J Child Neurol 2014;18:176-82.
2013;28:13-20. 23. Paciorkowski AR, Thio LL, Dobyns WB. Genetic and
9. Askalan R, Mackay M, Brian J, et al. Prospective biologic classification of infantile spasms. Pediatr Neurol
preliminary analysis of the development of autism and 2011;45:355-67.
epilepsy in children with infantile spasms. J Child Neurol 24. Sanmaneechai O, Sogawa Y, Silver W, et al. Treatment
2003;18:165-70. outcomes of West syndrome in infants with Down
10. O'Callaghan FJ, Lux AL, Darke K, et al. The effect of lead syndrome. Pediatr Neurol 2013;48:42-7.
time to treatment and of age of onset on developmental 25. Gibbs EL, Fleming MM, Gibbs FA. Diagnosis and
outcome at 4 years in infantile spasms: evidence from prognosis of hypsarhythmia and infantile spasms.
the United Kingdom Infantile Spasms Study. Epilepsia Pediatrics 1954;13:66-73.
2011;52:1359-64. 26. Hancock EC, Osborne JP, Edwards SW. Treatment
11. Widjaja E, Go C, McCoy B, et al. Neurodevelopmental of infantile spasms. Cochrane Database Syst Rev
outcome of infantile spasms: A systematic review and 2013;6:CD001770.
meta-analysis. Epilepsy Res 2015;109:155-62. 27. Hussain SA, Shinnar S, Kwong G, et al. Treatment of
12. Riikonen R. Long-term outcome of patients with West infantile spasms with very high dose prednisolone before
syndrome. Brain Dev 2001;23:683-7. high dose adrenocorticotropic hormone. Epilepsia
13. Mackay MT, Weiss SK, Adams-Webber T, et al. Practice 2014;55:103-7.
parameter: medical treatment of infantile spasms: report 28. Singer WD, Haller JS, Sullivan LR, et al. The value
of the American Academy of Neurology and the Child of neuroradiology in infantile spasms. J Pediatr
Neurology Society. Neurology 2004;62:1668-81. 1982;100:47-50.
14. Mytinger JR, Joshi S; Pediatric Epilepsy Research 29. Aydinli N, Calişkan M, Ozmen M, et al. Neuroradiologic
Consortium, et al. The current evaluation and treatment of aspects of West syndrome. Pediatr Neurol 1998;19:211-6.
infantile spasms among members of the Child Neurology 30. Saltik S, Kocer N, Dervent A. Magnetic resonance imaging
Society. J Child Neurol 2012;27:1289-94. findings in infantile spasms: etiologic and pathophysiologic
15. Shields WD. Infantile spasms: little seizures, BIG aspects. J Child Neurol 2003;18:241-6.
consequences. Epilepsy Curr 2006;6:63-9. 31. van Bogaert P, Chiron C, Adamsbaum C, et al. Value
16. Go CY, Mackay MT, Weiss SK, et al. Evidence-based of magnetic resonance imaging in West syndrome of
guideline update: medical treatment of infantile spasms. unknown etiology. Epilepsia 1993;34:701-6.
Report of the Guideline Development Subcommittee of 32. Epilepsy Phenome/Genome Project & Epi4K
the American Academy of Neurology and the Practice Consortium. Copy number variant analysis from exome
Committee of the Child Neurology Society. Neurology data in 349 patients with epileptic encephalopathy. Annals
2012;78:1974-80. of Neurology 2015;78:323-8.
17. Thiele EA. Managing epilepsy in tuberous sclerosis 33. Pavone P, Striano P, Falsaperla R, et al. Infantile spasms
complex. J Child Neurol 2004;19:680-6. syndrome, West syndrome and related phenotypes: what
18. Kossoff EH. Infantile spasms. Neurologist 2010;16:69-75. we know in 2013. Brain Dev 2014;36:739-51.
19. Hrachovy RA, Frost JD Jr. Infantile epileptic 34. Mastrangelo M, Leuzzi V. Genes of early-onset epileptic
encephalopathy with hypsarrhythmia (infantile spasms/ encephalopathies: from genotype to phenotype. Pediatr
West syndrome). J Clin Neurophysiol 2003;20:408-25. Neurol 2012;46:24-31.
20. Appleton RE. West syndrome: long-term prognosis and 35. Lux AL. Latest American and European updates on
social aspects. Brain Dev 2001;23:688-91. infantile spasms. Curr Neurol Neurosci Rep 2013;13:334.
21. Osborne JP, Lux AL, Edwards SW, et al. The underlying 36. Hrachovy RA, Frost JD Jr, Kellaway P, et al. Double-blind
etiology of infantile spasms (West syndrome): information study of ACTH vs prednisone therapy in infantile spasms.
from the United Kingdom Infantile Spasms Study (UKISS) J Pediatr 1983;103:641-5.
on contemporary causes and their classification. Epilepsia 37. Snead OC 3rd, Benton JW, Myers GJ. ACTH and
2010;51:2168-74. prednisone in childhood seizure disorders. Neurology

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

Translational Pediatrics, Vol 4, No 4 October 2015 269

1983;33:966-70. first-line treatment for infantile spasms: a randomized,

38. Kossoff EH, Hartman AL, Rubenstein JE, et al. High- prospective study. Epilepsia 1997;38:1270-4.
dose oral prednisolone for infantile spasms: an effective 53. Thelle T, Gammelgaard L, Hansen JK, et al. Reversible
and less expensive alternative to ACTH. Epilepsy Behav magnetic resonance imaging and spectroscopy
2009;14:674-6. abnormalities in the course of vigabatrin treatment for
39. Baram TZ, Mitchell WG, Tournay A, et al. High-dose West syndrome. Eur J Paediatr Neurol 2011;15:260-4.
corticotropin (ACTH) versus prednisone for infantile 54. Dracopoulos A, Widjaja E, Raybaud C, et al. Vigabatrin-
spasms: a prospective, randomized, blinded study. associated reversible MRI signal changes in patients with
Pediatrics 1996;97:375-9. infantile spasms. Epilepsia 2010;51:1297-304.
40. Lux AL, Edwards SW, Hancock E, et al. The United 55. Westall CA, Wright T, Cortese F, et al. Vigabatrin retinal
Kingdom Infantile Spasms Study comparing vigabatrin toxicity in children with infantile spasms: An observational
with prednisolone or tetracosactide at 14 days: a cohort study. Neurology 2014;83:2262-8.
multicentre, randomised controlled trial. Lancet 56. Partikian A, Mitchell WG. Major adverse events associated
2004;364:1773-8. with treatment of infantile spasms. J Child Neurol
41. Lux AL, Edwards SW, Hancock E, et al. The United 2007;22:1360-6.
Kingdom Infantile Spasms Study (UKISS) comparing 57. Hong AM, Turner Z, Hamdy RF, et al. Infantile spasms
hormone treatment with vigabatrin on developmental treated with the ketogenic diet: prospective single-
and epilepsy outcomes to age 14 months: a multicentre center experience in 104 consecutive infants. Epilepsia
randomised trial. Lancet Neurol 2005;4:712-7. 2010;51:1403-7.
42. Hrachovy RA, Frost JD Jr, Glaze DG. High-dose, long- 58. Kossoff EH, Hartman AL. Ketogenic diets: new advances
duration versus low-dose, short-duration corticotropin for metabolism-based therapies. Curr Opin Neurol
therapy for infantile spasms. J Pediatr 1994;124:803-6. 2012;25:173-8.
43. Lombroso CT. A prospective study of infantile 59. Kossoff EH, Thiele EA, Pfeifer HH, et al. Tuberous
spasms: clinical and therapeutic correlations. Epilepsia sclerosis complex and the ketogenic diet. Epilepsia
1983;24:135-58. 2005;46:1684-6.
44. Hancock E, Osborne J. Treatment of infantile spasms 60. Kossoff EH, Pyzik PL, McGrogan JR, et al. Efficacy
with high-dose oral prednisolone. Dev Med Child Neurol of the ketogenic diet for infantile spasms. Pediatrics
1998;40:500. 2002;109:780-3.
45. Snead OC 3rd. Treatment of infantile spasms. Pediatr 61. Pires ME, Ilea A, Bourel E, et al. Ketogenic diet for
Neurol 1990;6:147-50. infantile spasms refractory to first-line treatments: an open
46. Shields WD, Pellock JM. Vigabatrin 35 years later - from prospective study. Epilepsy Res 2013;105:189-94.
mechanism of action to benefit-risk considerations. Acta 62. Kossoff EH, Hedderick EF, Turner Z, et al. A case-control
Neurol Scand Suppl 2011;(192):1-4. evaluation of the ketogenic diet versus ACTH for new-
47. Elterman RD, Shields WD, Mansfield KA, et al. onset infantile spasms. Epilepsia 2008;49:1504-9.
Randomized trial of vigabatrin in patients with infantile 63. Eun SH, Kang HC, Kim DW, et al. Ketogenic diet for
spasms. Neurology 2001;57:1416-21. treatment of infantile spasms. Brain Dev 2006;28:566-71.
48. Fejerman N, Cersósimo R, Caraballo R, et al. Vigabatrin 64. Kossoff EH. International consensus statement on
as a first-choice drug in the treatment of West syndrome. J clinical implementation of the ketogenic diet: agreement,
Child Neurol 2000;15:161-5. flexibility, and controversy. Epilepsia 2008;49 Suppl 8:11-3.
49. Willmore LJ, Abelson MB, Ben-Menachem E, et al. 65. Overby PJ, Kossoff EH. Treatment of infantile spasms.
Vigabatrin: 2008 update. Epilepsia 2009;50:163-73. Curr Treat Options Neurol 2006;8:457-64.
50. Aicardi J, Mumford JP, Dumas C, et al. Vigabatrin as initial 66. Peltzer B, Alonso WD, Porter BE. Topiramate and
therapy for infantile spasms: a European retrospective adrenocorticotropic hormone (ACTH) as initial treatment
survey. Sabril IS Investigator and Peer Review Groups. for infantile spasms. J Child Neurol 2009;24:400-5.
Epilepsia 1996;37:638-42. 67. Zou LP, Ding CH, Fang F, et al. Prospective study of first-
51. Pesaturo KA, Spooner LM, Belliveau P. Vigabatrin for choice topiramate therapy in newly diagnosed infantile
infantile spasms. Pharmacotherapy 2011;31:298-311. spasms. Clin Neuropharmacol 2006;29:343-9.
52. Vigevano F, Cilio MR. Vigabatrin versus ACTH as 68. Korinthenberg R, Schreiner A. Topiramate in children

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270

270 Nelson. Infantile spasms

with west syndrome: a retrospective multicenter evaluation spasms: a multi-centre randomised trial. Arch Dis Child
of 100 patients. J Child Neurol 2007;22:302-6. 2010;95:382-6.
69. Yanai S, Hanai T, Narazaki O. Treatment of infantile 73. Hamano S, Yoshinari S, Higurashi N, et al. Developmental
spasms with zonisamide. Brain Dev 1999;21:157-61. outcomes of cryptogenic West syndrome. J Pediatr
70. Lotze TE, Wilfong AA. Zonisamide treatment for 2007;150:295-9.
symptomatic infantile spasms. Neurology 2004;62:296-8. 74. Karvelas G, Lortie A, Scantlebury MH, et al. A
71. Goh S, Kwiatkowski DJ, Dorer DJ, et al. Infantile spasms retrospective study on aetiology based outcome of infantile
and intellectual outcomes in children with tuberous spasms. Seizure. 2009;18:197-201.
sclerosis complex. Neurology 2005;65:235-8. 75. Fedak EM, Patel AD, Heyer GL, et al. Optimizing Care
72. Darke K, Edwards SW, Hancock E, et al. Developmental With a Standardized Management Protocol for Patients
and epilepsy outcomes at age 4 years in the UKISS trial With Infantile Spasms. J Child Neurol 2015;30:1340-2.
comparing hormonal treatments to vigabatrin for infantile

Cite this article as: Nelson GR. Management of infantile

spasms. Transl Pediatr 2015;4(4):260-270. doi: 10.3978/
j.issn.2224-4336.2015.09.01

© Translational Pediatrics. All rights reserved. www.thetp.org Transl Pediatr 2015;4(4):260-270