Abstract
Background:
Cerebrospinal fluid (CSF) has provided some of the most promising source of validated biomarkers in Alzheimer's disease (AD). Such biomarkers are critical to the development of novel disease modifying treatments for use in research and clinical practice. There is great interest in specific CSF protein candidates for measuring the earliest stages of disease progression and improving the clinical utility of traditional AD biomarkers for accurate diagnosis.
Methods:
Data acquired from the Alzheimer's Disease Neuroimaging (ADNI) study was used in this study. Levels of 159 CSF analytes were measured using the Rules Based Medicine (RBM, Austin, TX) Human Discovery Multi-Analyte Profile 1.0 panel. Global volumetric and regional thickness MRI measurements were determined in 295 ADNI subjects (AD=65, MCI=142, HC=88) acquired at baseline using Freesurfer 5.1.0. The multiplex analyte panel was used to identify a subset of proteins predictive of hippocampal and entorhinal cortex (ERC) atrophy. Another avenue of study also examined the combination of the analyte panel with traditional CSF biomarkers (Aβ142, p-tau, t-tau) and structural MRI measures to 1) distinguish between AD and healthy controls subjects and 2) predict future MCI to AD conversion at a one year follow up.
Results:
Multiple linear regression analysis by cross validation identified a subset of analytes (FABP, CgA, ApoE, NGAL, PLGF, TNFR2, leptin, VEGF, ApoD, ApoA-1, prolactin, clusterin, TM, SCF) predictive of hippocampal atrophy (5 fold cross validation R 2 = 35.6%). Several of these analytes (FABP, CgA, prolactin, apoD, TM, NGAL, clusterin) were also predictive of ERC thickness atrophy (5 fold cross validation R 2 = 30.2%) and related to rapid clinical progression in AD. Combining the analyte panel with MRI and traditional CSF biomarkers (Aβ 1-42, p-tau 181, and t-tau) produced the best model for distinguishing between AD and HC subjects at baseline (92.7%) and predicting MCI conversion to AD at a one year follow up (82.4%).
Conclusions:
Novel analyte candidates in CSF were identified in relation to brain atrophy and complement existing AD biomarkers for AD classification and future MCI conversion prediction. This may suggest a role of these proteins in AD pathogenesis and disease progression. Further work is required in independent samples to replicate these findings.
Cerebrospinal fluid (CSF) has provided some of the most promising source of validated biomarkers in Alzheimer's disease (AD). Such biomarkers are critical to the development of novel disease modifying treatments for use in research and clinical practice. There is great interest in specific CSF protein candidates for measuring the earliest stages of disease progression and improving the clinical utility of traditional AD biomarkers for accurate diagnosis.
Methods:
Data acquired from the Alzheimer's Disease Neuroimaging (ADNI) study was used in this study. Levels of 159 CSF analytes were measured using the Rules Based Medicine (RBM, Austin, TX) Human Discovery Multi-Analyte Profile 1.0 panel. Global volumetric and regional thickness MRI measurements were determined in 295 ADNI subjects (AD=65, MCI=142, HC=88) acquired at baseline using Freesurfer 5.1.0. The multiplex analyte panel was used to identify a subset of proteins predictive of hippocampal and entorhinal cortex (ERC) atrophy. Another avenue of study also examined the combination of the analyte panel with traditional CSF biomarkers (Aβ142, p-tau, t-tau) and structural MRI measures to 1) distinguish between AD and healthy controls subjects and 2) predict future MCI to AD conversion at a one year follow up.
Results:
Multiple linear regression analysis by cross validation identified a subset of analytes (FABP, CgA, ApoE, NGAL, PLGF, TNFR2, leptin, VEGF, ApoD, ApoA-1, prolactin, clusterin, TM, SCF) predictive of hippocampal atrophy (5 fold cross validation R 2 = 35.6%). Several of these analytes (FABP, CgA, prolactin, apoD, TM, NGAL, clusterin) were also predictive of ERC thickness atrophy (5 fold cross validation R 2 = 30.2%) and related to rapid clinical progression in AD. Combining the analyte panel with MRI and traditional CSF biomarkers (Aβ 1-42, p-tau 181, and t-tau) produced the best model for distinguishing between AD and HC subjects at baseline (92.7%) and predicting MCI conversion to AD at a one year follow up (82.4%).
Conclusions:
Novel analyte candidates in CSF were identified in relation to brain atrophy and complement existing AD biomarkers for AD classification and future MCI conversion prediction. This may suggest a role of these proteins in AD pathogenesis and disease progression. Further work is required in independent samples to replicate these findings.
| Original language | Undefined/Unknown |
|---|---|
| Pages (from-to) | P227-P227 |
| Number of pages | 1 |
| Journal | Alzheimer's Dementia: The Journal of the Alzheimer's Association |
| Volume | 9 |
| Issue number | 4, Supplement |
| DOIs | |
| Publication status | Published - 2013 |
