Cerebrospinal fluid β‐amyloid and phospho‐tau biomarker interactions affecting brain structure in preclinical Alzheimer disease

Abstract
Objective To assess the relationships between core cerebrospinal fluid (CSF) biomarkers and cortical thickness (CTh) in preclinical Alzheimer disease (AD). Methods In this cross‐sectional study, normal controls (n = 145) from the Alzheimer's Disease Neuroimaging Initiative underwent structural 3T magnetic resonance imaging (MRI) and lumbar puncture. CSF β‐amyloid1–42 (Aβ) and phospho‐tau181p (p‐tau) levels were measured by Luminex assays. Samples were dichotomized using published cutoffs (Aβ+/Aβ and p‐tau+/ptau). CTh was measured by Freesurfer. CTh difference maps were derived from interaction and correlation analyses. Clusters from the interaction analysis were isolated to analyze the directionality of the interaction by analysis of covariance. Results We found a significant biomarker interaction between CSF Aβ and CSF p‐tau levels affecting brain structure. Cortical atrophy only occurs in subjects with both Aβ+ and p‐tau+. The stratified correlation analyses showed that the relationship between p‐tau and CTh is modified by Aβ status and the relationship between Aβ and CTh is modified by p‐tau status. p‐Tau–dependent thinning was found in different cortical regions in Aβ+ subjects but not in Aβ subjects. Cortical thickening was related to decreasing CSF Aβ values in the absence of abnormal p‐tau, but no correlations were found in p‐tau+ subjects. Interpretation Our data suggest that interactions between biomarkers in AD result in a 2‐phase phenomenon of pathological cortical thickening associated with low CSF Aβ, followed by atrophy once CSF p‐tau becomes abnormal. These interactions should be considered in clinical trials in preclinical AD, both when selecting patients and when using MRI as a surrogate marker of efficacy. Ann Neurol 2014;76:223–230
Funding Information
  • CIBERNED program (Program 1, Alzheimer Disease)
  • Carlos III Institute of Health, Spain (PI11/02425)
  • Carlos III Institute of Health, Spain (PI11/3035)