Clinical and neuroimage features associated with treatment responses of flunarizine in patients with chronic migraine

Description: Background: Prophylactic treatments can reduce headache frequency and associated disability in chronic migraine, yet treatment responses are not consistent in the patients. We aim to identify the clinical and neuroimaging signatures that are predictive of treatment responses in CM. Methods: We prospectively enrolled patients with CM and healthy controls (HCs) and collected their baseline clinical profiles and neuroimage data of brain-MRI at first visit. We prescribed flunarizine as the first line prophylaxis and longitudinally followed the patients. Treatment responses were determined by the 3rd month headache frequency reduction compared to the baseline frequency. Voxel-based morphometry was used to study gray matter volume of brain regions. Results: Among 78 patients with CM, 41 had good response and 37 had poor response. Patients with poor response had higher baseline headache frequency, higher prevalence of cutaneous allodynia and poorer sleep quality compared to those with good response (p< 0.05). Patients with poor response also had smaller brain volumes over prefrontal cortex, orbitofrontal cortex, visual cortex, somatosensory cortex, and hippocampus/parahippocampal gyrus, and larger volumes over the cerebellum compared to those with good response. Clinical predictors had good ability in discriminating patients with different responses (AUC=0.776), and adding on the volume of the right hippocampus/parahippocampal gyrus could improve the ability to 0.872. Combination of sleep quality, volumes of the left lateral frontal pole, the right hippocampus/parahippocampal gyrus and the right cuneal cortex could excellently discriminate patients with different treatment responses (AUC = 0.958). Conclusions: Combined clinical and neuroimaging signatures have additive effects in predicting treatment responses in CM. These biomarkers assist accurate prognostication in individualized medicine.