Abstract
Objective
To evaluate long‐term developmental outcomes and identify independent predictors of favorable developmental outcomes at 3 years of age in children with infantile epileptic spasms syndrome (IESS) treated with a standardized stepwise vigabatrin and high‐dose prednisolone protocol.
Methods
This prospective single‐center cohort study enrolled 236 children with IESS managed according to a standardized protocol between March 2016 and April 2022 at Severance Children's Hospital. All patients received initial vigabatrin, with high‐dose oral prednisolone added for non‐vigabatrin responders. Developmental outcome at 3 years was classified as favorable (mental quotient [MQ] ≥ 70 and psychomotor quotient [PMQ] ≥ 70) or unfavorable using the Bayley Scales of Infant and Toddler Development, Second Edition. Independent predictors were identified by multivariable logistic regression using Firth's penalized method to address complete separation.
Results
Overall treatment response was achieved in 175 patients (74.2%). At 3 years, 49 patients (20.8%) achieved favorable developmental outcome. On multivariable analysis, etiology and treatment response were the only independent predictors. Compared to unknown etiology, structural etiology (aOR 0.085; 95% CI 0.026–0.281; p < 0.001) and genetic etiology (aOR 0.023; 95% CI 0.002–0.357; p = 0.007) were associated with significantly lower odds of favorable developmental outcome. Unresponsive patients had markedly lower odds compared to vigabatrin responders (aOR 0.030; 95% CI 0.002–0.523; p = 0.02). Notably, none of the 49 patients with genetic etiology and none of the 61 unresponsive patients achieved favorable developmental outcome. Lead time showed a significant univariable association, but this association was attenuated after adjustment for other prognostic factors.
Significance
Etiology and treatment response are the dominant independent predictors of favorable developmental outcome at 3 years in children with IESS. These findings underscore the need for early etiological diagnosis and individualized treatment strategies, particularly in patients with genetic etiology or those unresponsive to first‐line therapy.
Plain Language Summary
Infantile epileptic spasms syndrome (IESS) is a severe epilepsy of infancy that often leads to lasting developmental problems. In this study of 236 children with IESS treated with a stepwise vigabatrin and prednisolone protocol, developmental outcome at age 3 was mainly associated with the underlying cause of IESS and response to treatment. Children without an identified underlying cause and those who responded to treatment had better developmental outcomes than children with a structural or genetic cause or those who did not respond. These findings emphasize the importance of careful evaluation and individualized treatment in children with IESS.
Keywords: epileptic spasms, etiology, lead time, neurodevelopmental outcome, treatment response
Key points.
A vigabatrin and prednisolone‐based stepwise protocol achieved treatment response in a substantial proportion of patients with IESS.
Etiology was a dominant prognostic factor, with unknown etiology associated with better outcomes than structural or genetic etiologies.
Treatment response failure was uniformly associated with unfavorable developmental outcome, underscoring its role as a prognostic milestone.
The association between lead time and developmental outcome was attenuated after adjustment for other prognostic factors.
These findings highlight the importance of etiologic evaluation and individualized treatment in IESS outcomes.
1. INTRODUCTION
Infantile epileptic spasms syndrome (IESS), formerly classified as West syndrome, is one of the most severe epileptic encephalopathies of infancy, characterized by epileptic spasms, hypsarrhythmia on electroencephalography, and developmental regression or arrest. 1 , 2 Despite advances in diagnosis and treatment, the long‐term prognosis remains poor, with the majority of affected children experiencing persistent developmental delay and intellectual disability. 3
The treatment of IESS is centered on hormonal therapies—adrenocorticotropic hormone (ACTH) and oral corticosteroids—and vigabatrin. The United Kingdom Infantile Spasms Study (UKISS) demonstrated that hormonal treatment was superior to vigabatrin alone in achieving spasm cessation at 14 days, though long‐term developmental outcomes did not differ significantly between treatment groups. 4 , 5 The International Collaborative Infantile Spasms Study (ICISS) subsequently showed that combining hormonal therapy with vigabatrin achieved higher rates of clinical remission compared to hormonal therapy alone. 6 However, the optimal sequencing of these agents—specifically, whether vigabatrin should be initiated as monotherapy with hormonal therapy reserved for non‐responders—remains an area of active investigation.
Etiology is widely recognized as the dominant determinant of long‐term neurodevelopmental outcome in IESS, with unknown etiology consistently associated with better prognosis than structural or genetic etiologies. Treatment response is also a key prognostic factor, as failure to achieve early electroclinical remission has been associated with persistent cognitive impairment. 4 , 7 The independent prognostic significance of lead time—the interval from spasm onset to treatment initiation—remains inconsistent across studies, with some reporting an association with better developmental outcomes 8 and others identifying no significant independent effect. 4
Despite the established importance of these prognostic factors, prospective long‐term data on developmental outcomes following a systematic stepwise escalation protocol—vigabatrin as initial monotherapy with add‐on high‐dose prednisolone for non‐responders—are limited. Furthermore, the independent contributions of etiology, treatment response, and lead time have rarely been examined simultaneously within a multivariable framework. In this prospective single‐center study, we aimed to describe developmental outcomes at 3 years of age in children with IESS treated with a standardized vigabatrin‐based stepwise protocol, and to identify independent predictors of favorable developmental outcome.
2. MATERIALS AND METHODS
2.1. Patients
Patients with IESS managed according to our standardized treatment protocol between March 2016 and April 2022 at Severance Children's Hospital were enrolled in this prospective long‐term follow‐up study. The diagnosis of IESS was confirmed by the presence of both epileptic spasms and hypsarrhythmia on EEG. Patients were excluded if they: (1) were diagnosed with tuberous sclerosis, as vigabatrin is used as a dedicated first‐line treatment in this group; or (2) voluntarily withdrew before completing the treatment protocol or minimum 3‐year follow‐up period.
Because of the nature of a tertiary referral center, a substantial proportion of patients had already been diagnosed and initiated on treatment at other institutions prior to first presentation at our center. Excluding these patients would have caused selection bias by preferentially excluding more severe cases. Therefore, patients who were already receiving non‐protocol anti‐seizure medication(s) (ASMs) at the time of protocol initiation were included in the analysis. For these patients, lead time was defined as the interval from spasm onset to the first initiation of IESS‐directed treatment, regardless of treatment location.
Written informed consent was obtained from the parents or legal guardians of all participants, and the study protocol was approved by the Institutional Review Board of Severance Hospital.
All patients included in this study underwent brain magnetic resonance imaging (MRI) with an epilepsy protocol. Patients without an identified structural etiology on MRI underwent targeted gene panel sequencing or whole‐exome sequencing. In selected cases, array comparative genomic hybridization or multiplex ligation‐dependent probe amplification was performed based on the next‐generation sequencing results.
2.2. Assessments
2.2.1. Epileptic spasms
Parents or caregivers were instructed to keep a seizure diary documenting the number of seizures or seizure clusters, the duration of each seizure or cluster, and the semiology of seizures. They were also asked to record the dosage and administration times of ASMs.
2.2.2. EEG
At the time of diagnosis, EEG was recorded for a minimum of 4 h, and follow‐up EEGs were recorded for at least 30 min. Sleep tracing without the use of sedatives was consistently included in every EEG evaluation. When seizures or epileptic spasms were suspected without certainty, prolonged video‐EEG monitoring of > 8 h was recorded to capture suspicious seizures.
EEG recordings were independently reviewed by two pediatric neurologists certified by the Korean Board of EEG and Clinical Neurophysiology—one with access to clinical information and one blinded to it. The Burden of Amplitudes and Epileptiform Discharges (BASED) scoring system proposed by Mytinger et al. in 2015 was applied, and the scoring criteria are summarized in Table S1. 9 The two reviewers showed complete agreement on all assigned BASED scores. According to the BASED system, scores of 4 and 5 correspond to hypsarrhythmia. In this study, however, a BASED score of ≤ 2 was set as the treatment goal to ensure stricter disease control. Thus, a BASED score of ≤ 2 was defined as resolution of hypsarrhythmia or achievement of controlled EEG status in this study.
2.2.3. Developmental function
Developmental function was assessed at three time points: within one week of diagnosis, at 6–7 months after treatment initiation, and at 3 years of age, using the Bayley Scales of Infant and Toddler Development, Second Edition (Bayley‐II). Mental age and psychomotor age quotients were calculated by dividing the mental and psychomotor ages derived from the Bayley scales by the chronological age at each assessment. Developmental outcome at 3 years was categorized as favorable developmental outcome, defined as both MQ ≥ 70 and PMQ ≥ 70, or unfavorable developmental outcome, defined as MQ < 70 or PMQ < 70.
2.3. Treatment protocol
All patients followed a standardized treatment protocol consisting of initial vigabatrin therapy, followed by high‐dose oral prednisolone in patients who did not achieve remission with vigabatrin only (Figure 1).
FIGURE 1.
Treatment algorithm for infantile epileptic spasms syndrome (IESS). Shaded boxes indicate treatment steps; dashed boxes indicate decision points. Response was defined as absence of spasms for > 24 h and a BASED score ≤ 2. ASM, anti‐seizure medication; BASED, Burden of Amplitudes and Epileptiform Discharges; IESS, infantile epileptic spasms syndrome; PDL, prednisolone; VGB, vigabatrin.
At protocol initiation, patients received vigabatrin starting at 50 mg/kg/day for one day, which was then increased to 100 mg/kg/day. Treatment escalation was guided by clinical response: if epileptic spasms persisted, vigabatrin was increased without requiring EEG confirmation. If no spasms were observed for more than 24 h, an EEG including sleep tracing was obtained. Treatment was further escalated if the EEG showed a BASED score ≥ 3. Conversely, no additional escalation was performed if the patient remained spasm‐free for more than 24 h and had a BASED score ≤ 2, thereby meeting the criteria for a treatment responder.
At 1 week after treatment initiation, non‐responders received an increased vigabatrin dose of 150 mg/kg/day. At 2 weeks, if spasms persisted, oral prednisolone was added at 40 mg/day, and at 3 weeks, increased to 60 mg/day for non‐responders. High‐dose prednisolone (40 or 60 mg/day) was administered for 2 weeks, followed by tapering over the subsequent 2 weeks. Vigabatrin was maintained for a total of 6 months in patients treated with vigabatrin monotherapy and for 3 months in those who received high‐dose prednisolone add‐on therapy.
Patients who initially responded to vigabatrin but later experienced relapse of epileptic spasms or EEG worsening (BASED score ≥ 3) received add‐on high‐dose prednisolone during ongoing vigabatrin therapy. For those who did not respond to either vigabatrin or prednisolone combination therapy, individualized treatment strategies were employed, including a second cycle of high‐dose prednisolone, other ASMs, ketogenic diet, or epilepsy surgery.
2.4. Statistical analysis
Categorical variables were compared using the χ2 test or Fisher's exact test, and continuous variables were compared using the Mann–Whitney U test. All prespecified variables of clinical interest were entered into a multivariable logistic regression model for favorable developmental outcome at 3 years. Given complete separation in two subgroups (etiology: genetic; treatment response: unresponsive), Firth's penalized logistic regression was employed to obtain stable estimates.
Two post‐hoc subgroup analyses were performed: a lead‐time analysis restricted to patients with unknown etiology, and a subgroup comparison of baseline characteristics and outcomes between patients with prior ASM treatment before protocol initiation and those initiated directly on the standardized protocol. In addition, a sensitivity multivariable logistic regression model was fitted by adding prior ASM treatment status to the primary multivariable model for favorable developmental outcome at 3 years.
All analyses were performed using R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). A two‐sided p‐value of < 0.05 was considered statistically significant.
3. RESULTS
3.1. Baseline characteristics
A total of 508 IESS patients were assessed for eligibility. After excluding 23 patients with tuberous sclerosis complex, 81 patients whose study protocol was not completed due to treatment with an alternative regimen, and 168 patients with insufficient follow‐up data (including those lost to follow‐up, who died before age 3, or who had inadequate medical records), 236 patients were enrolled in the final study cohort (Figure 2).
FIGURE 2.
Flowchart of patient enrollment and treatment protocol outcomes. ASM, anti‐seizure medication; IESS, infantile epileptic spasms syndrome; TSC, tuberous sclerosis complex.
The median spasm onset age was 5.0 months (interquartile range [IQR], 3.0–6.0), and the median lead time was 30.0 days (IQR, 9.8–90.0). Male patients accounted for 126 (53.4%) of the cohort. Eighty‐one patients (34.3%) had structural etiology, 49 (20.8%) had genetic etiology, and 106 (44.9%) had unknown etiology. Developmental delay was already present at the time of diagnosis in 179 patients (75.8%).
In treatment response, 15 patients (6.4%) achieved remission with vigabatrin at 100 mg/kg/day, and an additional 15 (6.4%) responded to vigabatrin at 150 mg/kg/day. Of the remaining patients who required prednisolone add‐on therapy, 82 (34.7%) responded to prednisolone at 40 mg/day and 63 (26.7%) responded at 60 mg/day. Sixty‐one patients (25.8%) did not respond to either vigabatrin or add‐on prednisolone therapy and therefore were classified as unresponsive. The full baseline characteristics of the study cohort are presented in Table 1.
TABLE 1.
Baseline characteristics of study patients (n = 236).
| Variable | Total (n = 236) | Favorable developmental outcome (n = 49) | Unfavorable developmental outcome (n = 187) | p‐value |
|---|---|---|---|---|
| Sex, n (%) | ||||
| Male | 126 (53.4) | 25 (51.0) | 101 (54.0) | |
| Female | 110 (46.6) | 24 (49.0) | 86 (46.0) | 0.83 |
| Spasm onset age (months), median (IQR) | 5.0 (3.0–6.0) | 5.0 (5.0–6.0) | 5.0 (2.0–6.0) | 0.005 |
| Lead time (days), median (IQR) | 30.0 (9.8–90.0) | 16.0 (7.0–30.0) | 30.0 (11.0–90.0) | 0.002 |
| Etiology, n (%) | ||||
| Structural | 81 (34.3) | 3 (6.1) | 78 (41.7) | <0.001 |
| Genetic | 49 (20.8) | 0 (0.0) | 49 (26.2) | |
| Unknown | 106 (44.9) | 46 (93.9) | 60 (32.1) | |
| Treatment response, n (%) | ||||
| Vigabatrin 100 mg/kg/day | 15 (6.4) | 4 (8.2) | 11 (5.9) | <0.001 |
| Vigabatrin 150 mg/kg/day | 15 (6.4) | 7 (14.3) | 8 (4.3) | |
| Prednisolone 40 mg/day added | 82 (34.7) | 21 (42.9) | 61 (32.6) | |
| Prednisolone 60 mg/day added | 63 (26.7) | 17 (34.7) | 46 (24.6) | |
| Unresponsive | 61 (25.8) | 0 (0.0) | 61 (32.6) | |
| Developmental delay at diagnosis, n (%) | 179 (75.8) | 26 (53.1) | 153 (81.8) | <0.001 |
Note: Lead time: days from spasm onset to first initiation of IESS‐directed treatment, regardless of treatment location. Etiology classified per 2017 ILAE operational classification. Developmental delay at diagnosis: assessed by Bayley Scales of Infant and Toddler Development (2nd ed.) within 1 week of diagnosis. Favorable developmental outcome defined as both MQ ≥ 70 and PMQ ≥ 70 at 3 years of age. Categorical variables compared by χ 2 test or Fisher's exact test; continuous variables by Mann–Whitney U test. Bold p‐values indicate statistical significance (p < 0.05).
Abbreviations: IESS, infantile epileptic spasms syndrome; ILAE, International League Against Epilepsy; IQR, interquartile range; MQ, mental quotient; PMQ, psychomotor quotient.
Of the 236 patients, 97 (41.1%) were already receiving non‐protocol anti‐seizure medication(s) at the time of protocol initiation (prior ASM treatment group), whereas 139 (58.9%) were initiated directly on the standardized protocol. Compared with the direct initiation group, the prior ASM treatment group had an earlier spasm onset age (median, 3.0 vs. 5.0 months; p < 0.001), longer lead time (median, 60.0 vs. 21.0 days; p < 0.001), and a different etiology distribution, driven by a higher proportion of structural etiology (51.5% vs. 22.3%; overall etiology distribution, p < 0.001). Overall treatment response was achieved in 66 of 97 patients (68.0%) in the prior ASM treatment group and 109 of 139 patients (78.4%) in the direct initiation group, indicating that the prior ASM treatment group did not have a higher response rate (p = 0.10). The overall distribution of treatment response categories also did not differ significantly between the two groups (p = 0.37) (Table S3).
3.2. Developmental outcomes at 3 years
At 3 years of age, 49 patients (20.8%) achieved favorable developmental outcome (both MQ ≥ 70 and PMQ ≥ 70), while 187 patients (79.2%) had unfavorable developmental outcome (MQ < 70 or PMQ < 70).
On univariable analysis, several clinical factors were significantly associated with developmental outcome at 3 years (Table 1). Patients with favorable developmental outcome had a significantly higher spasm onset age compared to those with unfavorable developmental outcome (median, 5.0 months [IQR, 5.0–6.0] vs. 5.0 months [IQR, 2.0–6.0]; p = 0.005). Lead time was significantly shorter in patients with favorable developmental outcome than in those with unfavorable developmental outcome (median, 16.0 days [IQR, 7.0–30.0] vs. 30.0 days [IQR, 11.0–90.0]; p = 0.002). To further examine the association between lead time and developmental outcome, a post‐hoc subgroup analysis was performed restricted to patients with unknown etiology (n = 106). Lead time was not significantly associated with favorable developmental outcome at 3 years in this subgroup on either univariable (p = 0.63) or multivariable analysis (p = 0.78) (Table S2).
Etiology was strongly associated with outcome (p < 0.001): favorable developmental outcome was achieved in 46 of 106 patients (43.4%) with unknown etiology, compared to only 3 of 81 (3.7%) with structural etiology and none of 49 (0.0%) with genetic etiology (Figure 3). Developmental delay at diagnosis was more prevalent among patients with unfavorable developmental outcome at 3 years (81.8% vs. 53.1%; p < 0.001). Treatment response was also significantly associated with outcome (p < 0.001); notably, none of the 61 unresponsive patients achieved favorable developmental outcome at 3 years. Prior ASM treatment status was also associated with outcome: favorable developmental outcome at 3 years was less frequent in the prior ASM treatment group than in the direct initiation group (7.2% vs. 30.2%; p < 0.001) (Table S3). Sex was not significantly associated with developmental outcome (p = 0.83).
FIGURE 3.
Developmental outcomes at 3 years sorted by etiology and treatment response. Horizontal stacked bars represent the proportion of patients with favorable developmental outcome (both MQ ≥ 70 and PMQ ≥ 70; light gray) and unfavorable developmental outcome (MQ < 70 or PMQ < 70; dark gray with hatching) at 3 years of age, stratified by etiology (upper panel) and treatment response (lower panel). The dashed vertical line indicates 50%. Numbers within or above bars indicate the percentage and absolute count of patients in each category. MQ, mental quotient; PMQ, psychomotor quotient; VGB, vigabatrin.
3.3. Factors associated with developmental outcome at 3 years
In the multivariable logistic regression analysis (Figure 4), etiology and treatment response were independently associated with developmental outcome at 3 years. Compared to patients with unknown etiology, those with structural etiology (adjusted odds ratio [aOR], 0.085; 95% confidence interval [CI], 0.026–0.281; p < 0.001) and genetic etiology (aOR, 0.023; 95% CI, 0.002–0.357; p = 0.007) had significantly lower odds of achieving favorable developmental outcome. Among treatment response categories, unresponsive patients had markedly lower odds of favorable developmental outcome compared to vigabatrin responders (aOR, 0.030; 95% CI, 0.002–0.523; p = 0.02). Sex, spasm onset age, lead time, and developmental delay at diagnosis were not independently associated with developmental outcome after adjustment for other variables.
FIGURE 4.
Forest plot of multivariable logistic regression analysis. aOR, adjusted odds ratio; CI, confidence interval; DD, developmental delay; VGB, vigabatrin. Filled squares indicate point estimates. Horizontal lines represent 95% confidence intervals. Arrows indicate confidence intervals extending beyond the axis boundary. The dashed vertical line indicates aOR = 1 (reference). Firth's penalized logistic regression was used to address complete separation. Bold text and filled color indicate statistically significant variables (p < 0.05).
In a sensitivity analysis in which prior ASM treatment status was added to this model, etiology and treatment response remained independently associated with favorable developmental outcome, with essentially unchanged effect estimates. The association between prior ASM treatment and favorable developmental outcome was attenuated after adjustment (aOR, 0.38; 95% CI, 0.13–1.02; p = 0.06).
4. DISCUSSION
In this prospective cohort of 236 children with IESS treated with vigabatrin and high‐dose prednisolone protocol, 30 patients (12.7%) achieved remission with vigabatrin monotherapy and an additional 145 patients (70.4% of those who received prednisolone add‐on) responded to combination treatment, yielding an overall treatment response rate of 74.2%. In the United Kingdom Infantile Spasms Study (UKISS), hormonal therapies achieved spasm cessation in 73% of patients at 14 days, compared to 54% with vigabatrin monotherapy. 5 The International Collaborative Infantile Spasms Study (ICISS) demonstrated that combined hormonal therapy with vigabatrin achieved a primary clinical response in 72% of patients, compared to 57% with hormonal therapy alone. 6 These comparisons are intended to contextualize our findings within the existing literature rather than to assert superiority or equivalence, given the substantial differences in study design, patient populations, treatment protocols, outcome definitions, and follow‐up duration.
Etiology and treatment response were the only independent predictors of favorable developmental outcome at 3 years on multivariable analysis. Patients with genetic etiology had the most severe prognosis (aOR 0.023 relative to unknown etiology), with none of the 49 patients in this group achieving favorable developmental outcome, while structural etiology was similarly associated with markedly lower odds (aOR 0.085). In contrast, 43.4% of patients with unknown etiology achieved favorable developmental outcome. These findings are consistent with the existing literature establishing etiology as the dominant determinant of long‐term neurodevelopmental outcome in IESS. 3 , 10 A systematic review and meta‐analysis by Widjaja et al. reported a pooled rate of good neurodevelopmental outcome of 54.3% in cryptogenic IESS compared to 12.5% in symptomatic IESS, findings that mirror the differential outcomes by etiology observed in our cohort. 3 Regarding treatment response, the complete absence of favorable developmental outcomes among unresponsive patients (aOR 0.030) aligns with prior evidence that failure to achieve early electroclinical remission is associated with persistent cognitive impairment. 4 , 7 Together, these findings indicate that the biological substrate of disease—reflected in etiology and treatment responsiveness—is the primary driver of long‐term neurodevelopmental outcome in IESS, exceeding the prognostic contribution of other clinical variables.
Lead time has shown inconsistent prognostic significance across previous studies. O'Callaghan et al. 8 found that shorter lead time to treatment was associated with better developmental outcomes, whereas Lux et al. 4 did not identify lead time as a significant predictor of outcome. In our cohort, lead time was associated with favorable developmental outcome on univariable analysis, but this association was attenuated in the multivariable logistic model after adjustment for etiology and treatment response. To further explore whether the effect of lead time might be more evident in patients with a relatively favorable prognostic profile, we performed a subgroup analysis limited to patients with unknown etiology, in whom the confounding influence of etiology may be minimized. Within this subgroup, lead time showed little evidence of association with favorable developmental outcome on either univariable (p = 0.63) or multivariable analysis (p = 0.78; Table S2). These findings suggest that the observed association may be partially explained by etiology‐related confounding. Patients with unknown etiology may be more likely to present with more readily recognizable spasms, facilitating earlier diagnosis and treatment initiation, while simultaneously having a more favorable intrinsic prognosis. In contrast, patients with structural or genetic etiologies may present with more atypical seizure semiology and concomitant abnormal movements, which may delay symptom recognition and diagnosis. Nevertheless, these findings do not necessarily exclude a clinically meaningful effect of earlier treatment, nor should they be interpreted as diminishing the importance of timely treatment initiation in IESS. Rather, they suggest that optimizing developmental outcomes may require timely initiation of appropriate therapy, together with early etiologic evaluation to guide individualized, etiology‐directed treatment.
This study has several limitations. First, as a single‐center study conducted at a tertiary referral hospital, the cohort is likely skewed toward more severe and complex cases, reflecting the referral patterns inherent to such a setting. Second, variability in treatment exposure prior to referral and reliance on caregiver‐reported seizure diaries may have introduced unmeasured bias. However, in the post‐hoc subgroup analysis, pretreated referral patients did not show higher treatment response rates than those initiated directly on the protocol, and the association between prior ASM treatment and poorer developmental outcome was attenuated after multivariable adjustment. These findings suggest that inclusion of pretreated referral patients was unlikely to have inflated the observed treatment response rate, although residual confounding related to referral severity cannot be excluded. Third, because the primary outcome was developmental status at 3 years of age, exclusion of patients without adequate follow‐up was unavoidable. Nevertheless, patients lost to follow‐up or who died before 3 years may have had poorer developmental outcomes, potentially resulting in an overestimation of favorable outcomes. Fourth, developmental outcome was classified dichotomously as favorable versus unfavorable, rather than being analyzed using continuous developmental measures. Fifth, the study period spanned several years during which access to genetic testing evolved substantially; therefore, some patients classified as having unknown etiology may be reclassified if evaluated using current diagnostic approaches. Finally, although follow‐up EEGs were recorded for at least 30 min including sleep tracing, recordings of 90 min or longer as recommended for outpatient BASED score assessment may have allowed more precise EEG characterization. 11
5. CONCLUSION
This prospective single‐center cohort study of 236 children with IESS treated according to a standardized vigabatrin‐based stepwise escalation protocol demonstrates that etiology and treatment response are the only independent predictors of favorable developmental outcome at 3 years of age. Genetic and structural etiologies were associated with markedly lower odds of favorable developmental outcome, and none of the patients with genetic etiology achieved a favorable developmental outcome. Conversely, 43.4% of patients with unknown etiology attained favorable developmental outcome at 3 years, underscoring the prognostic importance of etiology classification at diagnosis. Complete non‐response to vigabatrin and prednisolone combination therapy was uniformly associated with unfavorable developmental outcome. While lead time showed a significant univariable association with outcome, its effect was attenuated in multivariable analysis, suggesting that developmental outcome is influenced by multiple interacting factors rather than treatment timing alone. These results highlight the need for early and systematic etiological evaluation—including genetic testing—to identify high‐risk patients and guide individualized therapeutic strategies. Future multicenter studies with larger sample sizes are warranted to validate these findings and to explore etiology‐directed treatment approaches in IESS.
AUTHOR CONTRIBUTIONS
Soyoung Jang: Data curation, Methodology, Formal Analysis, Investigation, Visualization, Writing – Original Draft Preparation. Hui Jin Shin: Data Curation, Writing – Review & Editing. Se Hee Kim: Resources. Joon Soo Lee: Resources. Ara Ko: Conceptualization, resources, methodology, writing – review and editing. Hoon‐Chul Kang: Supervision, Conceptualization, resources, writing – review and editing.
FUNDING INFORMATION
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (grant number: RS‐2025‐00555034) and by a grant of Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare, Republic of Korea (grant number: RS‐2025‐02263053, RS‐2024‐00405260) and by Korea Drug Development Fund funded by the Ministry of Science and ICT, Ministry of Trade, Industry, and Energy, and Ministry of Health and Welfare (grant number: RS‐2025‐02213662), and by the National Institute of Health (NIH) research project (grant number: 2026‐ER0602‐00), and by a grant of the Korean ARPA‐H Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: RS‐2025‐25458096).
CONFLICT OF INTEREST STATEMENT
None of the authors have any conflict of interest to disclose.
ETHICS STATEMENT
We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
ETHICS APROVAL
The study protocol was approved by the Institutional Review Board of Severance Hospital (IRB No. 4–2017‐0196). Written informed consent was obtained from the parents or legal guardians of all participants.
Supporting information
Table S1. Burden of Amplitudes and Epileptiform Discharges (BASED) score.
Table S2. Association of lead time with developmental outcome at 3 years in patients with unknown etiology (n = 106).
Table S3. Comparison of baseline characteristics and outcomes between patients initiated directly on the standardized protocol and those pretreated with other anti‐seizure medications prior to protocol initiation.
ACKNOWLEDGMENTS
Authors acknowledge the support for data extraction from Haeju Oh.
Contributor Information
Ara Ko, Email: arako@yuhs.ac.
Hoon‐Chul Kang, Email: hipo0207@yuhs.ac.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
REFERENCES
- 1. Pellock JM, Hrachovy R, Shinnar S, Baram TZ, Bettis D, Dlugos DJ, et al. Infantile spasms: a U.S. consensus report. Epilepsia. 2010;51(10):2175–2189. [DOI] [PubMed] [Google Scholar]
- 2. Zuberi SM, Wirrell E, Yozawitz E, Wilmshurst JM, Specchio N, Riney K, et al. ILAE classification and definition of epilepsy syndromes with onset in neonates and infants: position statement by the ILAE task force on nosology and definitions. Epilepsia. 2022;63(6):1349–1397. [DOI] [PubMed] [Google Scholar]
- 3. Widjaja E, Go C, McCoy B, Snead OC. Neurodevelopmental outcome of infantile spasms: a systematic review and meta‐analysis. Epilepsy Res. 2015;109:155–162. [DOI] [PubMed] [Google Scholar]
- 4. Lux AL, Edwards SW, Hancock E, Johnson AL, Kennedy CR, Newton RW, et al. The United Kingdom infantile spasms study (UKISS) comparing hormone treatment with vigabatrin on developmental and epilepsy outcomes to age 14 months: a multicentre randomised trial. Lancet Neurol. 2005;4(11):712–717. [DOI] [PubMed] [Google Scholar]
- 5. Lux AL, Edwards SW, Hancock E, Johnson AL, Kennedy CR, Newton RW, et al. The United Kingdom infantile spasms study comparing vigabatrin with prednisolone or tetracosactide at 14 days: a multicentre, randomised controlled trial. Lancet. 2004;364(9447):1773–1778. [DOI] [PubMed] [Google Scholar]
- 6. O'Callaghan FJK, Edwards SW, Alber FD, Hancock E, Johnson AL, Kennedy CR, et al. Safety and effectiveness of hormonal treatment versus hormonal treatment with vigabatrin for infantile spasms (ICISS): a randomised, multicentre, open‐label trial. Lancet Neurol. 2017;16(1):33–42. [DOI] [PubMed] [Google Scholar]
- 7. O'Callaghan FJK, Edwards SW, Alber FD, Cortina Borja M, Hancock E, Johnson AL, et al. Vigabatrin with hormonal treatment versus hormonal treatment alone (ICISS) for infantile spasms: 18‐month outcomes of an open‐label, randomised controlled trial. Lancet Child Adolesc Health. 2018;2(10):715–725. [DOI] [PubMed] [Google Scholar]
- 8. O'Callaghan FJK, Lux AL, Darke K, Edwards SW, Hancock E, Johnson AL, et al. The effect of lead time to treatment and of age of onset on developmental outcome at 4 years in infantile spasms: evidence from the United Kingdom infantile spasms study. Epilepsia. 2011;52(7):1359–1364. [DOI] [PubMed] [Google Scholar]
- 9. Mytinger JR, Hussain SA, Islam MP, Millichap JJ, Patel AD, Ryan NR, et al. Improving the inter‐rater agreement of hypsarrhythmia using a simplified EEG grading scale for children with infantile spasms. Epilepsy Res. 2015;116:93–98. [DOI] [PubMed] [Google Scholar]
- 10. Darke K, Edwards SW, Hancock E, Johnson AL, Kennedy CR, Lux AL, et al. Developmental and epilepsy outcomes at age 4 years in the UKISS trial comparing hormonal treatments to vigabatrin for infantile spasms: a multi‐centre randomised trial. Arch Dis Child. 2010;95(5):382–386. [DOI] [PubMed] [Google Scholar]
- 11. Yuskaitis CJ, Mysak K, Godlewski B, Zhang B, Harini C. Confirmation of infantile spasms resolution by prolonged outpatient EEGs. Epilepsia Open. 2021;6(4):714–719. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Burden of Amplitudes and Epileptiform Discharges (BASED) score.
Table S2. Association of lead time with developmental outcome at 3 years in patients with unknown etiology (n = 106).
Table S3. Comparison of baseline characteristics and outcomes between patients initiated directly on the standardized protocol and those pretreated with other anti‐seizure medications prior to protocol initiation.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.