Positron emission tomography for serial imaging of the contused adult rat spinal cord.

We investigated whether small-animal positron emission tomography (PET) could be used in combination with computed tomography (CT) imaging techniques for longitudinal monitoring of the injured spinal cord. In adult female Sprague-Dawley rats (n = 6), the ninth thoracic (T9) spinal cord segment was exposed by laminectomy and subsequently contused using the Infinite Horizon impactor (Precision System and Instrumentation, Lexington, KY) at 225 kDyn. In control rats (n = 4), the T9 spinal cord was exposed by laminectomy but not contused. At 0.5 hours and 3, 7, and 21 days postinjury, 2-[(18)F]fluoro-2-deoxy-d-glucose ([(18)F]FDG) was given intravenously followed 1 hour later by sequential PET and CT. Regions of interest (ROIs) at T9 (contused) and T6 (uninjured) spinal cord segments were manually defined on CT images and aided by fiduciary markers superimposed onto the coregistered PET images. Monte Carlo simulation revealed that about 33% of the activity in the ROIs was due to spillover from adjacent hot areas. A simulation-based partial-volume compensation (PVC) method was developed and used to correct for this spillover effect. With PET-CT, combined with PVC, we were able to serially measure standardized uptake values of the T9 and T6 spinal cord segments and reveal small, but significant, differences. This approach may become a tool to assess the efficacy of spinal cord repair strategies.