798 Diffusion tensor imaging as a potential biomarker of sleep dysfunction in warfighters with chronic, severe, traumatic brain injury

Abstract Introduction Traumatic brain injury (TBI) plagues service members in times of war and training. Diagnosis and management of TBI remain challenging, with many suffering from sleep disorders. We hypothesized that TBI-related damage to the hypothalamus, a master regulator of breathing and sleep, could be related to post-TBI obstructive sleep apnea (OSA) and serve as a pathophysiological biomarker for a subpopulation of OSA patients. Methods This was a retrospective study of warfighters with TBI from the National Intrepid Center of Excellence (NICoE). Subjects were identified by severe TBI on neuroimaging and compared against a control group without TBI. All subjects underwent screening polysomnography (PSG). MRI was acquired via 3T scanner. The hypothalamus was automatically segmented using a diffeomorphic algorithm. DTI scalar values were analyzed with scalar t-tests between subjects and controls. Generalized linear modeling with DTI scalar values was used to predict AHI in subjects. Results 6 subjects and 61 controls were identified. There was significant sleep dysfunction amongst TBI subjects (mean apnea-hypopnea index (AHI) 5.1+/-6.6 events/hour; mild OSA incidence 33.3%; Pittsburgh Sleep Quality Index (PSQI) mean 13.3+/-2.6). Radial diffusivity (RD), axial diffusivity (AD) and mean diffusivity (MD) were significantly higher among subjects (control RD 9.64x10^-10+/-7.54x10^-11 m^2/s, subject RD 1.13x10^-9+/-1.20x10^-10m^2/s, p = 0.023; control AD 1.32x10^-9+/-7.64x10^-11m^2/s, subject AD 1.50x10^-9+/-1.43x10^-10m^2/s, p = 0.029; control MD 1.08x10^-9+/-7.43x10^-11m^2/s, subject MD 1.25x10^-9+/-1.34x10^-10m^2/s, p = 0.025). There were no differences in age or body-mass index. Generalized linear modeling with diffusivity measures as predictors of AHI in subjects was not significant. Conclusion Using a diffeomorphic algorithm to define the hypothalamus reveals significantly elevated scalar DTI measures in chronic, severe TBI compared to controls. DTI differences in the hypothalamus are a novel finding and possibly underlie part of the pathophysiology of TBI. Although this may have potential to serve as a biomarker in severe TBI patients with sleep disorders, these initial data do not support a relationship between DTI and AHI, despite high incidence of OSA and subjective sleep dysfunction. Future studies with more subjects may better elucidate the changes in hypothalamic DTI after TBI for clinical outcomes analysis. Support (if any) This work was supported by grant 130132 from USAMRMC.