Longitudinal subcortical volume changes and their correlations with multiple PET and fluid biomarkers in dominantly inherited Alzheimer’s disease

Background Alzheimer's disease postmortem studies demonstrate that amyloid plaques and neurofibrillary tangles are present in subcortical regions. Objective To investigate longitudinal subcortical structural changes in autosomal dominant Alzheimer’s disease in relation to multiple PET and fluid biomarkers. Design Dominantly Inherited Alzheimer’s Network (DIAN) Observational study Setting Multicenter study Participants Participants were identified as mutation-carriers of pathologic variants in presenilin-1, presenilin-2, or amyloid precursor protein and as non-carriers from the same families as the mutation-carriers. They underwent baseline and 2 and more times longitudinal follow-up assessments of multiple biomarkers Measurements Participants underwent structural MRI, ¹¹C-Pittsburgh Compound B PET, ¹⁸F-fluorodeoxyglucose PET, and CSF and plasma assessments. Rates of biomarker change as a function of estimated years to symptom onset were estimated using multivariate linear mixed-effects models, and longitudinal associations between subcortical atrophy and multiple biomarkers were evaluated. Results A total of 601 participants completed one or more clinical evaluations, with up to eight annual visits. Mutation carriers showed significantly greater longitudinal atrophy in the left amygdala, bilateral thalamus, putamen, nucleus accumbens, and hippocampus compared with non-carriers (Bonferroni-corrected p < 0.05). The earliest divergence was observed 13.2 years before the expected symptom onset in the right nucleus accumbens, following amyloid-β (Aβ) accumulation in the right thalamus that began 23.8 years before onset. Among carriers, atrophy in the right thalamus, bilateral putamen, and bilateral nucleus accumbens was significantly associated with region-specific or cortical Aβ accumulation, as well as with CSF Aβ42, Aβ42/Aβ40 ratio, total tau, and phosphorylated tau (Bonferroni-corrected p < 0.05). Conclusions The present findings may provide a unique and well-characterized model for investigating the temporal ordering of Alzheimer’s disease biomarkers.