Cannabidiol Use in Developmental and Epileptic Encephalopathies: A Syndrome- and Age-Stratified Systematic Review and Meta-analysis.

Developmental and epileptic encephalopathies (DEEs) are severe, drug-resistant epilepsies associated with major developmental, cognitive, and behavioral burden. Although pharmaceutical-grade cannabidiol (CBD) has shown efficacy in Dravet syndrome (DS) and Lennox-Gastaut syndrome (LGS), evidence across the broader DEE spectrum remains fragmented. The aim of this systematic review was to quantify seizure and safety outcomes with pharmaceutical-grade CBD in DEE and explore whether effectiveness differs by syndrome type, age band, dose, clobazam co-medication, or follow-up duration. We conducted a systematic review and meta-analysis of PubMed, Embase, and CENTRAL from inception to 12 October 2025. Eligible studies enrolled individuals of any age with DEE treated with pharmaceutical-grade CBD, as add-on therapy or monotherapy. Mixed-etiology reports were eligible when DEE-specific data could be isolated or obtained from authors. Main outcomes were proportions achieving ≥ 50% or ≥ 75% seizure reduction, and seizure freedom; key safety outcomes were also collected. Random-effects generalized linear mixed models were used; small-study effects were explored with funnel plots and Egger's test. Prespecified subgroup analyses were performed by age (pediatric, adult, mixed) and syndrome (DS, LGS, Doose syndrome, CDKL5-related DEE, unspecified DEE, and other defined DEEs); post-hoc exploratory subgroup analyses examined CBD dose, concomitant clobazam use, and follow-up duration. Effect modification was tested using interaction p-values and interpreted with guidance from the Cochrane Handbook and the Instrument to assess the Credibility of Effect Modification in Analyses (ICEMAN). Forty-six studies (5 randomized controlled trials [RCTs] and 41 non-randomized studies; 2592 patients) met the inclusion criteria. The pooled ≥ 50% responder rate was 49.9% (95% CI 44.9-55.0); ≥ 75% responders comprised 26.7% (95% CI 22.0-32.0), and seizure freedom was achieved in 5.7% (95% CI 4.0-8.0). Age, syndrome type, dose, concomitant clobazam use, and follow-up duration did not demonstrate a robust or consistent pattern of effect modification across efficacy outcomes. Although some subgroup analyses reached statistical significance, these findings were often imprecise, based on small subgroups with wide confidence intervals, and did not meet ICEMAN credibility criteria. Adverse-event profiles were consistent across studies: somnolence, decreased appetite, diarrhea, fatigue, and behavioral changes predominated, mostly mild to moderate and manageable with dose adjustment. Transaminase elevations occurred mainly with valproate co-therapy and were reversible upon dose reduction or discontinuation. Serious adverse events were uncommon, and withdrawals due to adverse events were infrequent. Pharmaceutical-grade CBD is associated with clinically meaningful seizure reduction in roughly half of patients with DEE, mirroring pivotal RCT results in DS and LGS. Nevertheless, substantial between-study heterogeneity and low-credibility subgroup signals preclude confident attribution of superior efficacy to any specific subgroup. These findings should be interpreted cautiously given the imprecision and heterogeneity of the available evidence. Future research should prioritize well-powered, prospectively phenotyped, and syndromically defined cohorts with standardized outcome measures and individual participant data sharing to elucidate true effect modifiers and optimize patient selection. CRD420251186064.