Measurements of medial temporal lobe atrophy for prediction of Alzheimer's disease in subjects with mild cognitive impairment

L. Clerx, I.A. van Rossum, L. Burns, D.L. Knol, P. Scheltens, F. Verhey, P. Aalten, P. Lapuerta, L.A. van de Pol, R.A. van Schijndel, R. Jong, F. Barkhof, R. Wolz, D. Rueckert, M. Bocchetta, M. Tsolaki, F. Nobili, L.O. Wahlund, L. Minthon, L. FrolichH. Hampel, H. Soininen, P.J. Visser

Research output: Contribution to journalArticleAcademicpeer-review

80 Citations (Scopus)

Abstract

Our aim was to compare the predictive accuracy of 4 different medial temporal lobe measurements for Alzheimer's disease (AD) in subjects with mild cognitive impairment (MCI). Manual hippocampal measurement, automated atlas-based hippocampal measurement, a visual rating scale (MTA-score), and lateral ventricle measurement were compared. Predictive accuracy for AD 2 years after baseline was assessed by receiver operating characteristics analyses with area under the curve as outcome. Annual cognitive decline was assessed by slope analyses up to 5 years after baseline. Correlations with biomarkers in cerebrospinal fluid (CSF) were investigated. Subjects with MCI were selected from the Development of Screening Guidelines and Clinical Criteria for Predementia AD (DESCRIPA) multicenter study (. n = 156) and the single-center VU medical center (. n = 172). At follow-up, area under the curve was highest for automated atlas-based hippocampal measurement (0.71) and manual hippocampal measurement (0.71), and lower for MTA-score (0.65) and lateral ventricle (0.60). Slope analysis yielded similar results. Hippocampal measurements correlated with CSF total tau and phosphorylated tau, not with beta-amyloid 1-42. MTA-score and lateral ventricle volume correlated with CSF beta-amyloid 1-42. We can conclude that volumetric hippocampal measurements are the best predictors of AD conversion in subjects with MCI. © 2013 Elsevier Inc.
Original languageEnglish
Pages (from-to)2003-2013
JournalNeurobiology of aging
Volume34
Issue number8
DOIs
Publication statusPublished - 2013

Cite this