Background
[C-11]-(R)-PK11195 has been used to quantify microglial activation in neurodegenerative disorders. Two previous reports on [C-11]-(R)-PK11195 binding in Alzheimer's disease (AD) provided contradictory results1, 2 In the present study [C-11]-(R)-PK11195 binding was evaluated using both Statistical Parametric Mapping (SPM) and full kinetic analysis. For the latter both reference tissue and plasma input models were used.
Methods
13 AD patients (age 58–81 years), fulfilling the NINCDS-ADRDA criteria, and 17 age matched healthy controls (age 57–79 years) were included. Dynamic 3D PET scans, consisting of 22 frames over 60 minutes, were acquired following a bolus injection of ∼370 MBq [C-11]-(R)-PK11195. Arterial whole blood concentration was monitored continuously using an online detection system. In addition, discrete samples were taken in order to derive a metabolite corrected plasma input curve. Finally, a T1-weighted structural MRI scan was acquired using a 1 Tesla scanner. Volume of distribution (Vd) images and binding potential (BP) images were generated using Logan (plasma input) and Ichise (cerebellum = reference tissue input) plots, respectively. Both sets of images were used in SPM analyses to assess differences in PK11195 binding between AD and controls. In addition, regions of interest (ROIs) were drawn on the individual co-registered MRI scans. ROI were defined bilaterally for posterior cingulate, medial temporal lobe, thalamus and cerebellum. The ROI were projected onto the dynamic [C11]-(R)-PK11195 scan to generate time-activity curves (TACs) for each region. TACs were fitted using (a) a simplified reference tissue model with cerebellum as reference tissue 3 and (b) a two tissue reversible plasma input model (K1/k2 ratio fixed to whole cortex) 4 . Binding Potential (BP) was used as primary outcome measure of ROI analyses.
Results
SPM analysis based on Vd images did not reveal any areas with statistically significant differences in [C-11]-(R)-PK11195 binding between both groups. SPM analysis based on BP images showed slightly increased PK11195 binding (p=0.05) in bilateral occipital and lateral temporal lobes of AD patients. ROI analysis using ANOVA with age as covariate revealed no statistically significant differences in BP between groups, although with the two tissue reversible plasma input model (K1/k2 ratio fixed to whole cortex) a general trend for increased BP was found in patients with AD.
Conclusion
Overall, both SPM and ROI based analyses did not show increased [C-11]-(R)-PK11195 binding in AD, which is in accordance with the findings of Groom et al 2 . The trend of increased binding in occipital and lateral temporal lobes in the SPM analysis based on BP images warrants further analysis. This requires precise delineation of these structures and a full kinetic analysis. (See Table 1)