It has been shown that psychosocial stress promotes inflammation in an atherosclerotic mouse model [1]. It is also know that immune response and atherogenesis are tightly related[2]. Post Traumatic Stress Disorder (PTSD) is associated with inflammation biomarkers [3]. In this pilot study we sought to investigate the relationship between inflammation and brain functional anatomy in a cohort of PTSD patients. FDG-PET of the carotids was used as an imaging marker of inflammation. Functional MRI and Diffusion Tensor Imaging protocols were designed to probe the emotional circuitry and the underlying white matter connectivity.Study participants were recruited from physician referrals, media advertisement, or an academic outpatient psychiatrist clinic. Patients had chronic PTSD that was the primary presenting problem as assessed by a trained rater with the Structured Clinical Interview for DSM-IV (SCID-IV) and the Clinician-Administered PTSD Scale for DSM-IV (CAPS). Patients completed a battery of self-report measures that assessed variables that may correlate with PTSD symptom severity, including comorbid depressive and anxiety symptoms (MADRS, HAM-A). 11 subjects (ages: 34-61) were included in this analysis. All imaging was performed on a Siemens 3T PET/MR scanner. Every subject received 10 mCi of FDG an allowed to rest for 20 minutes and the Brain PET started after 30 minutes. Anatomical and functional MRI was obtained simultaneously with PET acquisition. PET emission data is collected in list-mode for 20 minutes. MRI protocols were : anatomical T1: 3D-MP-Rage, TR/TE/TI= 1900/2.5/900ms, FOV=23cm, 176 slices, Isotropic resolution 1mm x 1mm x 1mm. DTI: Pulsed Gradient Spin Echo, TR/TE=6300/86ms, 20 gradient directions, b-Value=1200 s/mm2, 44 slices, 2.5mm thick, FOV=23cm, matrix size 96x96. BOLD scans were acquired using GE-EPI: TR/TE=2000/27ms. FOV=23cm, matrix size 64x64. 38 slices, thickness=3mm. Three BOLD scans were acquired: Hariri face task 7:16, Facestroop 6:40 and rsfMRI 10:00. Following the brain scans, at 90 minutes after FDG injection, PET/MRI of the vasculature (carotids, aorta) was obtained. Brain Images were analyzed using FSL (www.fmrib.ox.ac.uk/fsl). We correlated the task effects with FDG-PET values and PTSD rating scales. Hariri Face Matching task: Contrast images were produced between faces and shapes. Facestroop: contrast images between incongruent and congruent images were produced. Resting state fMRI were analyzed using FSL’s MELODIC. Group ICA maps were produced for 20 networks, various networks were detected including default mode, attention network, visual, motor and auditory. Diffusion Tensor Imaging (DTI): Individual Fractional Anisotropy (FA) maps were normalized to standard space and group maps were produced. Skeletonized FA map was extracted for correlation analysis of the white matter integrity with the plaque inflammation in the carotids measured by vascular FDG-PET. Vascular PET data were reconstructed to 5mm3 voxel size using the Fourier rebinning–iterative algorithm. Maximal and average standardized uptake values (SUVs) were recorded on contiguous 5-mm-thick axial slices of the carotids and aorta from circular regions of interest (ROIs) encompassing the vessel wall. SUV from the background veins will also be registered. The mean and maximum target to background ratios (TBR, vessel wall SUV divided by background SUV) as well as the TBR of the most disease segment (MDS) was calculated as used in several previous studies.DTI: Fractional anisotropy (FA) was significantly correlated (negative) with TBR in various white matter tracts (Fig 1). FA was also significantly correlated with several PTSD rating scales. RSfMRI: The bilateral DLPFC was significantly correlated with CAPS (a PTSD symptom scale) (Fig 2). Connectivity with increasing CAPS. Hariri: the anterior cingulate was positively correlated with HAMA and MADRS when fearful faces were presented (Fig 3). Facestroop: ACC positively correlated with CAPSM under incongruent conditions (Fig 4).Despite the relatively small n (11) in this pilot study, fractional anisotropy correlated with vascular FDG uptake. This negative correlation was detected diffusely over whole brain suggesting an association between white matter damage and vascular inflammation. The two tasks driven emotional task showed positive correlation in the anterior cingulate with the MADRS and HAMA rating scales. This is consistent with some of the PTSD fMRI literature that suggests hyperactivity [4]. We note that despite a significant correlation between FA and FDG uptake, there was no significant correlations detected between the functional scans and FDG uptake. This might be explained by compensatory mechanism of the PTSD brain. These results suggests that a larger study is warranted.