Approach

When infantile spasms are suspected, guidance should be sought immediately from a tertiary paediatric neurologist, followed by referral if needed. Long-term risks of infantile spasms, including poor neurodevelopmental outcome, may be reduced by early initiation of treatment.

Effective treatment should produce both cessation of spasms and resolution of hypsarrhythmia on electroencephalogram (EEG).[38]​ However, the temporal relationship between cessation of ictal activity and EEG improvement is not established. Following spasm resolution, many infants continue to have an abnormal EEG, which may reflect the underlying aetiology or the development of other seizure types.[27] The major predictor of outcome is the underlying aetiology.[39][40]​​​ Patients with no apparent underlying cause have a better prognosis. There is a danger of diagnostic overshadowing when interpreting results of clinical trials in infantile spasms, given the large aetiological diversity of the patients included in these trials.[39]

First-line therapy

Surveys in the US and UK have indicated that hormonal treatments (e.g., adrenocorticotrophic hormone [ACTH] or prednisolone) and/or vigabatrin are the most commonly used first-line agents.[41] The dose of the chosen first-line agent should be adjusted to achieve the maximum effect as rapidly as possible.

Hormonal treatment modalities include ACTH and oral prednisolone, and have been the mainstay of treatment for over 60 years. Oral prednisolone is commonly used as first-line hormonal therapy in many countries. ACTH is recommended as the first-line treatment in a US consensus report.[38] Hormonal therapy may have significant adverse effects that must be monitored. The adverse effects of long-term corticosteroid therapy may be minimised by limiting the length of treatment. See Monitoring.

The 2022 UK National Institute for Health and Care Excellence (NICE) guideline on infantile spasms recommends combination therapy with high-dose oral prednisolone and vigabatrin as first-line treatment for infantile spasms that are not due to tuberous sclerosis, unless the child is at high risk of corticosteroid-related adverse effects.[26]​ If it is known at the time of diagnosis that the infant has tuberous sclerosis, vigabatrin alone is commonly given as initial treatment in most countries.

Hormonal therapy and vigabatrin (combination therapy)

The NICE guideline recommends combination therapy with high-dose oral prednisolone and vigabatrin as first-line treatment for infantile spasms that are not due to tuberous sclerosis, unless the child is at high risk of corticosteroid-related adverse effects.[26] This decision is based on evidence from the International Collaborative Infantile Spasms Study (ICISS) trial suggesting that first-line treatment combining corticosteroids with vigabatrin is more effective than either corticosteroids or vigabatrin alone in stopping spasms. This multi-centre, open-label, randomised study enrolled 377 patients with infantile spasms across 102 centres and randomly assigned them to hormonal therapy with vigabatrin (186) or hormonal therapy alone (191). All 377 infants were assessed for the primary outcome, which was cessation of spasms between days 14 and 42 from the beginning of the trial. No spasms were witnessed in 133 (72%) of 186 patients on combination therapy compared with 108 (57%) of 191 patients on hormonal therapy alone (difference 15%, 95% CI 5.1 to 24.9, P=0.002), showing combined therapy to be significantly more effective.[42] The 18-month follow-up results of this trial looked at developmental outcomes (Vineland score). Combination therapy did not result in improved developmental or epilepsy outcomes at 18 months compared with hormonal therapy alone.

ACTH

In Europe, Japan, and the UK, synthetic preparations of ACTH (e.g., tetracosactide/tetracosactrin, cosyntropin) are used; in the US, natural preparations (e.g., corticotropin) are preferred. This makes meta-analysis difficult. Indeed, the optimum dose and duration of treatment remains undefined by an evidence base, although, in general, a short course of approximately 2 weeks followed by a tapering dose is recommended.

One systematic review showed that patients initially treated with ACTH achieved a significantly higher initial response rate (spasm-free within 14 days) than patients treated with vigabatrin (74% vs. 55.5%).[27] However, less marked differences were observed in subsequent relapse rates at follow-up, and it is important to note that some of the included trials excluded patients with tuberous sclerosis. Evidence to date indicates that high-dose ACTH is not superior to low-dose ACTH in initial acquisition of spasm freedom.[14][43]​​[44][45]​​​

Few studies have evaluated ACTH dose-related long-term developmental outcomes but, at present, no significant difference in developmental status at 1 year between high- and low-dose ACTH is reported.[27] While no definitive trial has determined superiority between doses, low-dose therapy may be preferable, owing to its comparative efficacy and reduced risk of adverse effects.

ACTH must be administered by intramuscular injection on a daily or alternate daily basis, and consequently, due to ease and acceptability of administration, high-dose oral prednisolone is often preferred.

Corticosteroids

The UK Infantile Spasms Study (UKISS) randomised controlled trial (n=55) compared high-dose oral prednisolone with intramuscular tetracosactide (synthetic ACTH) showing nearly similar short-term efficacy for both interventions (ACTH 76% vs. prednisolone 70%).[46] In a 2015 randomised single-blind trial comparing high-dose oral prednisolone with intramuscular ACTH in 97 patients, significantly more patients achieved spasms cessation and electroclinical remission at day 14 and better control of spasms at 3 months, if initially treated with oral prednisolone. Control of spasms at 6 and 12 months was not significantly different. Risk of relapse following initial remission was similar in the two groups.[47][48]​ Another single-centre, non-blinded, randomised trial compared intramuscular ACTH with oral prednisolone in 34 children and did not find a difference between the two treatments.[49]​ An open-label randomised controlled trial found high-dose oral prednisolone to be significantly more efficacious compared with low-dose prednisolone, and adverse effects to be comparable in the two groups.[50]​ 

Vigabatrin

Vigabatrin is advocated as first-line treatment in infants with spasms and tuberous sclerosis (TS).[13][27][51]​​​​ In the UK, vigabatrin is considered alone as a first-line treatment for infantile spasms in children at high risk of corticosteroid-related adverse effects and in those who have spasms due to tuberous sclerosis. However, if vigabatrin is ineffective after 1 week, addition of high-dose oral prednisolone is recommended.[26]​ Vigabatrin has been shown to be more effective than hydrocortisone in patients with TS.[27]​ Children with TS had higher initial response rates to vigabatrin (95% vs. 54%) than those without TS.[52]​ Compared with placebo, vigabatrin promotes spasm cessation in 35% versus 10% of patients.[53]

There is good-quality evidence that infants with infantile spasms of all aetiologies who were initially treated with high-dose vigabatrin demonstrated higher initial response rates within 14 days compared with those infants who were initially treated with low-dose vigabatrin (68.2% vs. 51.8%).[54] They also had lower relapse rates (11.8%) and longer times to relapse (162 days) compared with those initially treated with low-dose vigabatrin (25% and 45 days).​ The incidence of adverse events related to vigabatrin was low, but the authors note that this study was conducted prior to recognition of vigabatrin-associated visual field loss.[54]

​Commonly observed side effects of vigabatrin include lethargy, irritability, sleeping and feeding difficulties, constipation, and hypotonia.[55][56][57][58]​​​ Reports of asymptomatic and symptomatic visual field defects with loss of peripheral vision in adults and children treated with vigabatrin suggest a potential causative link. It occurs in less than one half of adults and one third of children exposed to vigabatrin.[59] The pathophysiology of this is not clear. Factors potentially associated with increased risk of vigabatrin-associated visual field loss include male sex, increasing age, mean daily dose, cumulative dose, and duration of treatment. A genetic predisposition may also exist.[59] Vigabatrin-associated visual field loss may occur within 6 weeks of commencing treatment. Once it occurs it is generally considered to be irreversible and, if treatment continues, progressive, although there are some reports that it may improve or remain stable when treatment is withdrawn.[59] The prevalence of vigabatrin-induced retinal damage in patients treated for less than 6 months has been found to be low, at 5.3%.[60] Another study reported an even lower risk (3.2%) of developing clinically significant vigabatrin-associated visual field loss after less than 6-9 months of vigabatrin therapy.[61] 

While regular, 6-monthly visual field assessments are advocated in those receiving vigabatrin, there is no reliable means of assessing this in children.[27][59]​​​ More prospective studies are required to determine the natural course and identify potential risk factors of vigabatrin-associated visual field loss. While the risk of vigabatrin-associated visual field loss must be considered, West syndrome is often a devastating epileptic encephalopathy, and possible adverse effects must be weighed against the potentially beneficial impact of early spasm cessation on later developmental outcomes.

Reports have emerged of magnetic resonance imaging (MRI) brain signal changes in children receiving vigabatrin for infantile spasms. These changes, vigabatrin-associated brain abnormalities on MRI (VABAM), which are dose-dependent, predominantly affect the basal ganglia, thalamus, dentate, and brainstem.[62][63]​​​​ Risk factors for the development of VABAM may include age younger than 11 months and higher vigabatrin dose.[64]​ They are of unknown clinical significance and usually resolve when therapy is discontinued.[41]

One retrospective review of children involved in the ICISS showed that 6.5% of children with infantile spasms developed a movement disorder (MD) following initiation of anticonvulsants including vigabatrin. In 1.6% of the total infantile spasms cohort studied, this MD resolved spontaneously without modification of vigabatrin therapy. In a further 1.6% of the total cohort, MD onset had a close temporal relationship with initiation of vigabatrin and resolved upon its withdrawal, implying that for these patients MRI changes and MD onset may indeed be a reversible side effect of vigabatrin. However, in 3.2% of cases, MD persisted despite discontinuing vigabatrin. It has been postulated that in these patients, vigabatrin unmasks and lowers the threshold for an underlying movement disorder; its persistence despite withdrawal of therapy makes a direct causal role unlikely.[62]

Second-line therapy

Following initial spasm cessation, approximately 64% of those initially treated with vigabatrin and 60% of those randomised to receive hormonal treatments subsequently relapsed.[27][46][65][66]​​ Relapse rates may be lower for those initially commenced on high-dose vigabatrin (11.8%), and the time to relapse longer (162 days), than for those initially receiving low-dose vigabatrin (25% and 45 days).[54] No such significant difference is observed between high- and low-dose ACTH, although patients may relapse during the tapering/weaning phase of hormonal therapy.

There are no controlled trial data to support evidence-based therapy decisions when first-line treatment fails to stop spasms. Decisions can be based on expert opinion and experience, which should be guided and supervised by a tertiary paediatric neurologist experienced in the care of these children.[26]​ Consideration should be given to optimising the therapeutic dosages of chosen medications and a trial of alternative first-line agents. A trial of hormonal therapy should be considered in those who fail to respond to vigabatrin, and vigabatrin considered in those who fail to respond to hormonal therapy. The specific needs and characteristics of the individual child and carers should be considered.

Options following hormonal therapy or vigabatrin include a wide range of anticonvulsants, or consideration of a ketogenic diet.[26][67] Anticonvulsant options include levetiracetam, nitrazepam (other benzodiazepines can be considered), valproic acid (caution is required in those with suspected mitochondrial disease), or topiramate.

A randomised controlled trial in 50 patients compared topiramate with nitrazepam as first-line drugs in the treatment of infantile spasms.[68] Cessation of spasms occurred in 12 (48%) infants in the topiramate group and 4 (16%) in the nitrazepam group. 

Ketogenic diet

The NICE guideline suggests considering a ketogenic diet (KD) as a second-line monotherapy or as an add-on treatment.[26]

The KD is a high-fat, low-carbohydrate, and normal protein diet. The most common types of KD are the classic KD (4:1 or 3:1 fat to non-fat ratio), the medium-chain triglyceride (MCT) diet, the modified Atkins diet, and the low-glycaemic index treatment diet.[69] Literature is lacking qualitative studies on effects of KD in infantile spasms.

A systematic review pooled data of 13 observational studies (341 patients) investigating the efficacy of KD (most often 3-4:1 ratio) in mostly refractory patients with infantile spasms after a 1- to 6-month follow-up period. The median rate for patients with infantile spasms achieving short-term seizure reduction of more than 50% was 64%, the spasm-free rate was 35%, and the long-term seizure-free rate was 9.5%.[69]

In a small randomised controlled trial study with 32 infants allocated to either KD or high‐dose ACTH, electroclinical remission at day 28 and spasm freedom at last follow-up was similar between the two treatment groups.[70]

In a prospective multi-centre controlled study of 227 patients with hormonal therapy-resistant infantile spasms, the efficacy of KD therapy was superior to adjustment of oral anticonvulsant drugs in children with ACTH- or corticosteroid-resistant infantile spasms.[71]​ 

Pyridoxine or pyridoxal phosphate

Pyridoxine treatment should be considered in patients who are refractory to other treatment modalities and in whom pyridoxine-dependent epilepsy (PDE) has not been excluded.[72] It is favoured as a first-line treatment modality in Japan. Evidence from uncontrolled prospective studies indicates that the response rate is similar to the predicted spontaneous remission rate.[41]​​ In a pilot randomised controlled trial 62 patients with infantile spasms received either oral prednisolone alone or pyridoxine with oral prednisolone. The proportion of children with spasm cessation on day 14 was similar in the two groups.[73]​ 

​Pyridox(am)ine 5'-phosphate oxidase (PNPO) deficiency has rarely been reported as the cause of infantile spasms. Pyridoxal 5'-phosphate-dependent epilepsy is caused by changes or mutations in the PNPO gene. This is a potentially treatable condition, and pyridoxal phosphate (the active form of pyridoxine) should be considered in the treatment of infantile spasms not responding to first-line treatments.[74]

Surgery

Assessment for epilepsy surgery should be considered in those with atypical, asymmetrical spasms, or other suggestions of focality in seizure semiology, supported by lesional identification on neuroimaging and localisation on EEG. Criteria for selecting patients for surgical evaluation may include the following:

  • Infantile spasms refractory to medical management

  • No evidence of diffuse brain damage on imaging studies and focal EEG abnormalities

  • Focal abnormality on neuroimaging (e.g., MRI or PET)

  • No evidence of metabolic or degenerative disease.

Surgical evaluation is best undertaken in a specialist centre, following full evaluation by a multi-professional team. The family should be adequately counselled regarding the potential positive and negative outcomes and any possible postoperative neurological deficits that may be incurred.[38][75]

Use of this content is subject to our disclaimer