摘要
Background
Alzheimer’s disease (AD) and cerebral amyloid angiopathy (CAA) are both characterized by amyloid-β deposition but differ in their pathological distribution and clinical implications. Identifying disease-related imaging signals not dependent on visually apparent hemorrhage (e.g. dynamic amyloid PET) may provide complementary information beyond conventional late-frame uptake for distinguishing the two conditions.
Aim
This study aimed to identify features from dynamic [11C]-Pittsburgh compound B (PiB) PET for differentiating AD from CAA.
Methods
61 patients with AD or CAA, with the non-hemorrhagic hemisphere of their dynamic PiB PET data retained, were analyzed. Frame-wise PET uptake normalized to the pons was evaluated across cortical regions to characterize time-dependent signal differences between AD and CAA. Simplified Reference Tissue Model (SRTM) analysis with the pons as the reference region was additionally performed to derive relative tracer delivery (R1) and non-displaceable binding potential (BPnd). Discriminative performance of the PET-derived features was evaluated using receiver operating characteristic (ROC) analysis and quantified by the area under the curve (AUC).
Results
The posterior cingulate cortex (PCC) showed the most consistent dynamic PET discrimination between AD and CAA, particularly between 4 and 10 minutes after tracer injection. PCC 4–10 min uptake in the retained hemisphere and the PCC-to-global cortical uptake ratio achieved single-feature AUCs of approximately 0.67 and 0.69, respectively. Among the prespecified kinetic features, PCC R1 showed the strongest exploratory group signal and was moderately associated with the dynamic PCC 4–10 min uptake. This association suggests that the early-to-mid PCC signal contains a substantial tracer-delivery or perfusion-like component, whereas BPnd showed little group discrimination.
Conclusions
Early-to-mid dynamic PiB PET contains useful information for differentiating AD from CAA; the PCC 4–10 min signal showed the most consistent discrimination. These findings support further investigation of dynamic PET features as complementary imaging markers for AD–CAA differentiation.
