Toward a working definition of ketogenic diet resistance in GLUT1 deficiency syndrome.

objectiveThe ketogenic diet (KD) is the standard treatment for glucose transporter type 1 deficiency syndrome (GLUT1-DS), typically yielding seizure reduction and cognitive/motor gains. However, a small subset of patients shows limited or no clinical benefit. This phenomenon, referred to as "KD resistance," remains poorly understood and inconsistently defined. We propose a working definition, outline evaluation domains, summarize candidate mechanisms, and indicate management steps.

methodsNarrative review of published evidence and expert opinion across clinical, biochemical, EEG, and adherence domains.

resultsKD resistance may be considered when all are present: (1) confirmed therapeutic ketosis (serial blood β-hydroxybutyrate ≥2.0-2.5 mmol/L on repeated measurements); (2) adequate dietary adherence verified by dietetic assessment and, when available, validated tools; (3) sufficient trial duration (≥3 months; longer for primarily cognitive/motor goals); and (4) lack of meaningful improvement on symptom-relevant standardized measures. EEG interictal epileptiform discharge burden can be used as an adjunct marker but is not required. KD resistance may involve several dimensions: failure to achieve therapeutic ketosis, lack of symptom improvement despite confirmed ketosis, or challenges with adherence that limit efficacy. Possible contributing factors include genotypic variability in SLC2A1, mitochondrial dysfunction, impaired blood-brain barrier transport, hormonal influences, and epigenetic regulation. We outline a multidomain evaluation framework with suggested metrics, summarize candidate mechanisms (ketone transport/utilization, mitochondrial function, neurotransmission, ion channels/neuromodulators, inflammation/oxidative stress, epigenetic regulation), and indicate when to introduce a KD-compatible anti-seizure medication. SIGNIFICANCE: The proposed definition and framework standardize terminology and reporting, guide decisions for suboptimal responders, and set priorities for multicenter validation.