**Background:** Corticosteroids have been used for epilepsy treatment for over 60 years, based on the hypothesis that inflammation contributes to epileptogenesis. While their efficacy in infantile epileptic spasms is supported by randomized controlled trials (RCTs), their role in other childhood epilepsies remains unclear. This systematic review aimed to summarize the current literature on corticosteroid regimes in childhood epilepsies, excluding the well-studied area of epileptic spasms.
**Methods:** A structured literature search via PubMed was performed using MeSH terms "epilepsy", "steroids", and age-related terms (child, infant, adolescence, child, preschool) for publications between 1 January 2000 and 13 September 2022. Included were RCTs, clinical trials, observational studies, and case series with ≥2 patients receiving prednisolone, hydrocortisone, dexamethasone, methylprednisolone, ACTH, or ganaxolone for childhood epilepsy. Exclusion criteria were single case reports, non-specified treatment regimes, and non-English publications. Data extracted included number of patients, study type, epilepsy syndrome, treatment regime (dosage, duration), observation period, and outcomes (seizure freedom, seizure reduction of 50–80%, cognitive/EEG improvement). Subgroup analyses were performed for infantile epileptic spasms syndrome (IESS), spike-and-wave activity in sleep (SWAS), Lennox–Gastaut syndrome (LGS), Angelman syndrome, and other drug-resistant epilepsies (DRE). Only descriptive statistics were used; no meta-analysis was performed due to heterogeneity.
**Key Results:** From 160 identified papers, only three RCTs were found (excluding those on epileptic spasms). For IESS, the review references existing meta-analyses showing prednisolone is as effective as ACTH, and combination therapy with vigabatrin is superior to ACTH alone. For SWAS (including Landau–Kleffner syndrome), nine studies with 126 patients showed that 81 patients (64%) had improvement in EEG or language/cognition following various steroid regimes; relapse occurred in 17 of 30 patients (56%) from two studies. For LGS, two retrospective studies (77 patients) reported seizure freedom directly after treatment in 41 patients (53%), but only 12 of 77 (16%) had sustained freedom after 9 months to 7 years. For Angelman syndrome, one case series of four patients showed initial response in all, but seizures recurred in three (75%) after weaning. For DRE, 15 studies with 436 patients (pediatric and mixed adult-pediatric) reported seizure reduction (50–80%) in 219 patients (50%) and seizure freedom in 67 patients (15%). Relapse was reported in 17 of 25 patients (68%) from three studies. Subgroup analysis of short-term (3–10 days) vs. long-term (months) treatment showed similar outcomes: short-term (3 studies, 77 patients) had 44% seizure reduction and 9% seizure freedom; long-term (12 studies, 359 patients) had 52% seizure reduction and 17% seizure freedom. For status epilepticus, no controlled trials were found; steroids are listed as an option for refractory/super-refractory SE. For novel neurosteroids, ganaxolone showed a 30.7% seizure frequency reduction vs. 6.9% with placebo in a phase 3 trial of 101 patients with CDKL5-deficiency disorder. Brexanolone did not show superiority over placebo in weaning anesthetic agents in super-refractory SE (43.9% vs. 42.4%).
**Clinical Implications:** This review confirms that corticosteroids are effective for infantile epileptic spasms, with established protocols. For other childhood epilepsies, the evidence is weak and heterogeneous, precluding any standardized recommendations. In DRE, steroids may provide benefit (50% seizure reduction, 15% seizure freedom), but the lack of controlled trials and high relapse rates (68%) limit clinical utility. The ongoing RESCUE ESES trial comparing corticosteroids to clobazam for SWAS is urgently needed. Future research should focus on randomized, placebo-controlled trials with standardized dosing, duration, and outcome measures to establish evidence-based guidelines for steroid use in childhood epilepsies beyond spasms.