MRI longitudinal and cross‐sectional monitoring of amyloid pathology in non‐human primates

Background The potential for using MRI technologies to visualize amyloid deposits has been increasingly recognized. The present study utilized a New World non‐human primate (NHP) model of amyloid pathology, squirrel monkey (SQM), which unlike other NHPs, develops age‐dependent abundant cerebral amyloid angiopathy (CAA). The transverse relaxation rate (R2*) has a linear correlation with iron content, a component of amyloid deposits and microhemorrhages. Here we assessed whether the region of interest (ROI)‐based quantitative R2* assessment is able to detect differences in SQM neuropathology between varying age cohorts as well as longitudinally. An additional focus was to examine the presence of naturally occurring ARIA‐E like pathologies in geriatric SQMs. Method Multi‐gradient echo (MGE) MRI datasets were acquired in three different SQM age cohorts at various stages of Aβ pathology development. The R2* maps were generated (Firevoxel parametric map tool) from each subject to enable ROI‐based quantitative R2* analyses. DAB‐enhanced histochemical iron stain (Meguro method), along with 6E10/4G8 immunohistochemistry, was used to identify iron‐positive Aβ deposits. Histological examination of neuroinflammation and microhemorrhages (Perls’ iron stain) was performed. T2‐w FLAIR images were acquired to assess the presence of ARIA‐E. Result MGE imaging was capable of depicting brain pathology differences between three SQM age cohorts (ages 5, 16, 21). Absolute quantitation of R2* values, defined by ROI circumventing the cortex, revealed significant age‐related differences in selected brain regions, with the highest R2* counts detected in 21yr old monkeys. The ability to longitudinally track pathology‐related increase in R2* values was further validated by MRIs performed at specific intervals post‐initial scans in our geriatric monkeys. Iron‐positive Aβ deposits and microhemorrhages detected by MRI were confirmed via matched histological sections. Subsequent histological evaluation demonstrated disease‐stage differences in microgliosis and astrocytosis. Moreover, ARIA‐E like T2‐w signal hyperintensities were detected in our oldest cohort. Neuropathological correlates of ARIA‐E‐related changes were characterized histologically. Conclusion The current study highlights the utility of MRI methodologies to monitor age‐dependent progression of Aβ‐related pathology, as well as the importance of the SQM model for future translational research on advancing therapeutic interventions targeting amyloid depositions.