Distinct brain metabolic patterns reflect biological heterogeneity and relate to clinical outcome in chronic migraine...

AimChronic migraine with medication overuse (CMwMO) is a disabling phenotype of migraine with heterogenous responses to medication withdrawal and preventive treatments. We investigated whether baseline brain metabolic patterns reflect biological heterogeneity associated with 1-year clinical outcome in CMwMO.MethodsWe prospectively enrolled patients with CMwMO and healthy controls and obtained baseline headache profiles, psychological and dependence-related assessment as well as FDG-PET/MRI data. Patients received preventive treatments along with structured medication-withdrawal education. Poor outcome (n = 13) was defined as ongoing analgesic overuse together with <50% reduction in monthly headache days compared with baseline at one year follow-up. All other patients (n = 14) were classified as having a good outcome. Whole-brain metabolic comparisons, seed-based metabolic covariance analyses, and correlations with clinical variables were performed.ResultsAmong 27 patients with CMwMO, baseline headache frequency, disability, and medication use were similar between good and poor outcome groups, but the poor-outcome group was older, showed greater dependence severity, and had a longer duration of chronic headache. Compared with healthy controls (n = 17), only the poor-outcome group demonstrated hypermetabolism in the nucleus accumbens, thalamus, orbitofrontal cortex, and cerebellum. When compared to the good-outcome group, the poor-outcome group exhibited cerebellar hypermetabolism and stronger cerebellar metabolic covariance with the putamen and occipital cortex. Integrities of cerebellar-putamen metabolic coupling correlated with dependence severity, whereas cerebellar-occipital coupling correlated with duration of chronic headache and 1-year clinical outcomes.ConclusionsCMwMO appears to comprise biologically distinct subgroups with different baseline metabolic patterns. Treatment-refractory patients were characterized by metabolic alterations involving reinforcement and sensory prediction networks. Early identification of this subgroup may facilitate more individualized management targeting both headache burden and dependence-related behaviors.