P4–066: Alzheimer's Disease Biomarkers As Outcome Measures for Clinical Trials

Anna Caroli,Annapaola Prestia,Sara Wade,Kewei Chen,Napatkamon Ayutyanont,Susan Landau,Cindee Madison,Cathleen Haense,Karl Herholz,Eric Reiman,William Jagust,Giovanni Frisoni
DOI: https://doi.org/10.1016/j.jalz.2013.05.1455
2013-01-01
Abstract:There is a need to improve clinical trials in Alzheimer's disease (AD) by incorporating biomarkers of disease progression and patient selection in the mild cognitive impairment (MCI) and preclinical stages of the disease. The aim of this study was to investigate and compare the performance of the best-established diagnostic biological markers as outcome measures for clinical trials in MCI patients with CSF and MRI biomarker (qualified by medicines agencies as biomarkers for enrichment in clinical trials) evidence of AD. We considered hippocampal volume automatically computed on MRI using Freesurfer software, three FDG-PET summary metrics of AD-like hypometabolism and CSF tau as markers of neurodegeneration, CSF A β 1–42 as marker of amyloidosis, and ADAS-COG as clinical marker of AD progression, and we compared their effect size and statistical power over different followup periods (12 to 24 months, with 6-monthly or yearly biomarker assessment) in two MCI populations, selected from ADNI dataset based on baseline positivity to CSF A β 1–42 (MCI ABETA+) or hippocampal volume (below fifth percentile in a cognitively normal population, MCI HIPPO+). Longitudinal biomarkers progression was modeled through mixed effect models. Scaled slope, defined as the average of the slopes (over all subjects) divided by its standard deviation, was chosen as the measure of effect size. Seventy-four ABETA+ MCI and 51 HIPPO+ MCI patients were included in the study. Most of the patients had biomarkers assessed every 6 months up to 24 months after baseline. Hippocampal volume showed highest effect size for all time lenghts and both patient populations, followed by the other markers of neurodegeneration, ADASCOG, and, last, CSF Abeta42. Among FDG-PET biomarkers, HCI ranked first; metaROI ranked higher than logPALZ for shortest observation periods. Imaging biomarkers of neurodegeneration outperformed markers of AD pathology and clinical markers of AD progression as outcome measures for clinical trials of MCI due to AD. Hippocampal volume measurements were associated with the highest effect sizes, even for short clinical trials with yearly observations. These findings provide information about the biomarker enrichment and outcome measurements that could be employed to reduce MCI patient samples and treatment duration in future clinical trials.
What problem does this paper attempt to address?