Can static F-18-GE180 PET be a sufficient surrogate marker for post-stroke neuroinflammation in a mouse model?

Ziel/Aim The purpose of this study was to compare SUVR in the ischemic lesion to distribution volume ratio (DVR) obtained using image-derived input function (IDIF) with factor analysis-based (FA-based) partial volume effect (PVE) correction in a mouse ischemic stroke model.

Methodik/Methods In this work we studied 12 C57Bl/6J mice. Six of them had phototrombotic ischemia, the other six received sham treatment. Static (60-90 min p.i., for all the mice) and dynamic (for 3 stroke and 3 sham mice) F-18-GE180 PET acquisitions as well as T1 MRI were performed at 7 days, 1 month, 3 months, and 6 months after the stroke induction using a Mediso nanoScan 3T PET/MR scanner. The ischemic lesion VOI was defined on 7-day T1 MR images. All T1 MR images were registered to a T2 template, and Ma-Benveniste-Mirrione atlas was applied for brain VOI generation. SUVR values were obtained by normalizing mean uptake values in regions of interest by the cerebellum mean.

An IDIF was generated using the mean uptake value in the vena cava. For PVE correction, FA was performed on a cropped image. This method splits the dynamic image into several spatial distributions that represent tissues with distinct kinetics, where the distribution with the earliest peak corresponds to blood. Volume of distribution was obtained using Logan and Ichise plots, and then DVR was calculated using the volume of distribution in the cerebellum. Voxel-wise kinetic modelling using Logan plot was also performed.

Ergebnisse/Results DVR obtained using IDIF with FA-based PVE correction was consistent with SUVR from static images among the studied animals. Distribution volume parametric images, however, demonstrated better contrast in the lesion region compared to the static images.

Schlussfolgerungen/Conclusions Static acquisition might be sufficient for F-18-GE180 imaging in the mouse model, as it provided decent approximation to DVR. However, distribution volume images yield improved contrast in the lesion area, which can be useful for lesion VOI definition if MR data are not available.

Publikationsverlauf

Artikel online veröffentlicht:
08. April 2021