Anterolateral entorhinal cortical thinning as a biomarker for Alzheimer's disease in Down syndrome: Neuroimaging / Optimal neuroimaging measures for early detection.

Background: Virtually all adults with Down syndrome (DS) develop Alzheimer's disease (AD) associated neuropathology by the age of 40 and the majority of them develop dementia by age 60. Atrophy of the anterolateral entorhinal cortex (aLEC) has been previously associated with cognitive decline. Here, we sought to determine whether aLEC thinning is associated with clinical impairment in older adults with DS as well as whether it is associated with blood neurofilament light chain (NfL), another biomarker of neurodegeneration in AD. Method: 103 older adults with DS enrolled in the Alzheimer's Disease in Down Syndrome (ADDS) study were included in this analysis. Scans were classified into 3 groups based on consensus diagnosis: 14 (6F) possible/definite dementia (DEM), 19 (5F) mild cognitive impairment (MCI‐DS) and 70 (28F) cognitively stable (CS). Images were processed through the joint label fusion (JLF) pipeline in ANTs to obtain medial temporal lobe subregional volumes and thickness. Cortical reconstruction was performed using Freesurfer 6.0 to obtain total intracranial volume(ICV). We used a multiple linear regression model to fit left aLEC thickness using diagnostic group as the independent variable, correcting for age, gender, ICV and scan site differences. For a subset of these data (N=89), plasma NfL was quantified using a Simoa platform and we and tested whether left aLEC thickness was associated with NfL. Result: Cortical thickness in left aLEC was a significant predictor of clinical status. Thickness values were highest in the CS group, followed by MCI‐DS, followed by DEM. Groupwise differences were significant in DEM vs. CS and DEM vs. MCI‐DS comparisons. We also found a significant correlation between NfL and left aLEC thickness specifically in the DEM group. Finally, random forest classification showed that both NfL and left aLEC thickness were important features in predicting clinical diagnosis DEM vs MCI‐DS. Their estimates provided a diagnosis ROC curve with sensitivity of 0.79 and specificity of 0.92. Conclusion: In older adults with DS, left aLEC cortical thickness atrophy may be a primary feature of structural neurodegeneration that is associated with MCI‐DS and a blood‐based measurement of NfL may be informative as to this structural degeneration. [ABSTRACT FROM AUTHOR]