n.72 – FDG AND AMYLOID PET IN VIVO BIOMARKERS FOR THE PREDICTION OF MCI PROGRESSION



Abstract

BACKGROUND-AIM
FDG-PET evaluation of brain metabolic function is a highly sensitive and specific biomarker in dementia diagnosis. PET/CT imaging of brain amyloid load has been suggested as a core biomarker for Alzheimer’s disease (AD) in vivo neuropathology. There is limited evidence for the value of combined FDG and amyloid PET in predicting MCI progression to AD in individual subjects.
The comparison of FDG-PET SPM t-maps and Amyloid-PET SUVrs has been never evaluated in the same MCI subjects.
The aim of the study was to compare the prognostic accuracy of 18F-FDG-PET/CT (FDG-PET) using an optimized SPM method (Della Rosa et al. 2014; Perani et al. 2014) and 18F-Florbetapir PET/CT (Amyloid-PET) in predicting progression to dementia in MCI.
METHODS
Nine patients clinically classified as MCI (age range 64.72±8.10) who perfomed FDG-PET/CT and Amyloid PET/CT were identified in our database retrospectively. Clinical monitoring with average follow-up of 34 months was available in all patients after FDG and Amyloid PET/CT. We obtained objective FDG-PET SPM t-maps in single individuals compared to a large healthy controls dataset (N=112). Amyloid-PET average cortical SUVrs were estimated in frontal, temporal, parietal, anterior and posterior cingulated cortex and precuneus. We adopted currently published thresholds for positivity (Fleisher et all 2011: SUVr>1.17). The prediction power of FDG and amyloid PET for progression was estimated with ROC analysis.
RESULTS
Considering the follow-up, 5 subjects showed clinical progression to AD (MCI converters), and 4 subjects did not showed any progression (stable MCI). All 5 MCI converters were positive for both presence of Amyloid and FDG PET pattern AD typical. In these subjects an average increase of amyloid load of 16% (up to a maximum of 33%) in comparison to the adopted cut-off was observed. Within the 4 stable MCI cases, 2 were FDG-PET negative and Amyloid PET positive (incidental amyloid positivity), 1 was FDG-PET positive (Suspected NonAD Pathology-SNAP) and 1 was both FDG and amyloid PET positive (the latter cases in need of longer follow up). The analysis of prediction value for AD progression of combined FDG and amyloid PET yielded an AUC=0.875 with high sensitivity and specificity (1.00/0.875).
CONCLUSION
The use of both Amyloid PET (marker of pathology) and FDG PET (marker of neuronal injury/dysfunction) provided high accuracy and effective prognostic value to predict progression to AD in MCI cases thus representing a valuable tool in the diagnostic assessment. Further validations increasing the number of subjects are necessary to confirm these preliminary results.

Scarica PDF