Predicting modafinil-treatment response in poststroke fatigue using brain morphometry and functional connectivity

Background and Purpose: Poststroke fatigue affects a large proportion of stroke survivors and is associated with a poor quality of life. In a recent trial, modafinil was shown to be an effective agent in reducing poststroke fatigue; however, not all patients reported a significant decrease in fatigue with therapy. We sought to investigate clinical and radiological predictors of fatigue reduction with modafinil therapy in a stroke survivor cohort. Methods: Twenty-six participants with severe fatigue (multidimensional fatigue inventory-20 ≥60) underwent magnetic resonance imaging at baseline and during the last week of a 6-week treatment period of 200 mg modafinil taken daily. Resting-state functional magnetic resonance imaging and high-resolution structural imaging data were obtained, and functional connectivity and regional brain volumes within the fronto-striato-thalamic network were obtained. Linear regression analysis was used to identify predictors of modafinil-induced fatigue reduction. Results: Multiple regression analysis showed that baseline multidimensional fatigue inventory-20 score (β=0.576, P=0.006) and functional connectivity between the dorsolateral prefrontal cortex and the caudate nucleus (β=-0.424, P=0.008) were significant predictors of modafinil-associated decreases in poststroke fatigue (adjusted r2=0.52, area under the receiver operator characteristic curve=0.939). Conclusions: Fronto-striato-thalamic functional connectivity predicted modafinil response for poststroke fatigue. Fatigue in other neurological disease has been attributed to altered function of the fronto-striato-thalamic network and may indicate that poststroke fatigue has a similar mechanism to other neurological injury related fatigue. Self-reported fatigue in patients with normal fronto-striato-thalamic functional connectivity may have a different mechanism and require alternate therapeutic approaches.