- Email: [email protected]
Evaluation of pre-analytical and analytical factors that contribute to variability in CSF biomarker levels
Poster Presentations: P1 development for tau that shows a marked improvement in precision, linearity and reproducibility when compared to currently available tests. Methods: An in vitro diagnostic test for tau is being developed for the VITROS Immunodiagnostic Systems, and analytical performance was evaluated following CLSI guidelines using two VITROS Tau assay reagent lots, 3 CSF pools, and 3 controls. Linearity was tested with 11 admixtures of endogenous tau in CSF and recombinant tau 441 in buffer. Interference testing included commonly prescribed drugs, recombinant tau, Ab peptides, and endogenous substances. Limit of detection (LoD) and lower limit of quantitation were confirmed with 20 replicates per day on 3 days. Results: The median within-laboratory coefficient of variation (CV) for the 3 CSF pools was 3.3% with a range of 2.4% 4.2%. The controls’ within-laboratory CV was 1.2%. Repeatability CVs were 1.9% for CSF pools and 1.2% for controls. The assay showed good linearity across the measuring range (08,000pg/mL) with the admixture of endogenous and synthetic peptide. Bias introduced by commonly prescribed drugs at high levels was <610%. The interference from Ab peptides at greater than known physiological concentrations as well as from endogenous substances was <5%. The limit of detection was 9.5pg/mL; the lower limit of quantitation was 40pg/mL. Conclusions: VITROS Immunodiagnostic Products Tau Assay 1 demonstrates excellent analytical performance: it is precise, linear and the tested interferences cause no significant bias. The fully automated format and ability to perform these assays in a clinical laboratory should make testing for these important markers more reliable and routine.
P1-170
EVALUATION OF PRE-ANALYTICAL AND ANALYTICAL FACTORS THAT CONTRIBUTE TO VARIABILITY IN CSF BIOMARKER LEVELS
Aarti Shah1, Courtney Sutphen2, Matthew Amos2, Matthew Althage2, John Morris2, David Holtzman2, Anne Fagan2, 1Washington University, Saint Louis, Saint Louis, Missouri, United States; 2Washington University, St. Louis, Missouri, United States. Contact e-mail: [email protected] Background: Studies of CSF from Alzheimer’s disease research cohorts have identified several markers of disease pathology which show utility for disease diagnosis and prognosis (even in early and pre-clinical stages). However, anecdotal observations, lab-specific experiments and more recently, a world-wide initiative have shed light on the adverse influence of various pre-analytical, analytical and post-analytical factors on CSF biomarker values. Development of disease-modifying therapies and the proposed revision of diagnostic criteria that include biomarkers make biomarker protocol optimization and standardization a top priority. To promote such efforts, we have evaluated several potential sources of methodological variability in analyses of CSF. Methods: Existing CSF samples obtained from well-characterized research volunteers at the Knight Alzheimer’s Disease Research Center were analyzed by plate-based ELISA or bead-based immunoassays for A b 40, A b 42, tau and/or ptau181. Investigated potential sources of variability included: assay platform type, commercial kit lot number, two vendor-specific polypropylene aliquot tubes, use of sample pre-plate during analysis, aliquot volume, number of wells used to generate a mean value, and inter-personnel performance. Student’s t tests and Pearson correlation coefficients were used to evaluate data correspondence. Results: Analyte values did not differ between the two polypropylene aliquot tube types evaluated, whereas aliquot size did have an effect. Assay platform had a large impact on absolute analyte values, although values obtained between platforms were highly correlated. Commercial kit lot-to-lot variability was observed to varying extents and, when present, had a substantial impact on the evaluation of longitudinal biomarker changes within individuals over time. Mean values generated from samples run in duplicate were as precise as those run in quadruplicate for the assay tested. Use of a pre-plate for sample loading influenced levels of A b 42 but not levels of tau or ptau181. High inter-personnel reliability was achieved when well-trained individuals used identical protocols. Conclusions: Several sources of pre-analytical and analytical variability in the measurement of CSF biomarkers exist, and thus could impact the diagnostic and prognostic utility of such markers. Protocol standardization efforts within laborato-
P211
ries and eventually between laboratories are needed in order to increase the usefulness of these markers in both the research and clinical settings. P1-171
A NEW ERA OF CSF BIOMARKER TESTING IN THE FIELD OF ALZHEIMER’S DISEASE
Sebastiaan Engelborghs1, Erik Stoops2, Britta Brix3, Leentje Demeyer2, Dirk Jacobs2, Hanne Struyfs1, Eugeen Vanmechelen2, 1University of Antwerp, Antwerp, Belgium; 2ADx Neurosciences, Gent, Belgium; 3 EUROIMMUN AG, L€ubeck, Germany. Contact e-mail: sebastiaan. [email protected] Background: In the field of Alzheimer’s disease (AD) diagnostics and clinical trials, there is an urgent need to establish novel concepts for CSF biomarker testing. These novel concepts should allow earlier identification of specific brain pathology (e.g., amyloidopathy, tauopathy, loss of synapses) and should help predict the progression rate in disease-affected subjects. Last but not least, approval of the assays for inclusion in clinical trials is indispensable. Notwithstanding the proven clinical utility of the CSF biomarkers, such as phosphorylated tau or the b-Amyloid (Ab) proteins, all currently available immunoassays lack important analytical performance characteristics, hampering their world-wide integration and approval by regulatory authorities, especially in the US. The present study describes solutions for these analytical drawbacks. In addition, the qualification of the clinical utility of integrating CSF- Ab(1-40) in the panel as an aid for differential dementia diagnosis, was performed. Methods: The analytical and clinical qualification of the newly developed assay studies was performed using well-characterized CSF samples from multiple expert centers and carefully designed study protocols. The subject population included controls, patients with AD or vascular dementia (VAD). The outcome of the current assays was compared on-site with and using the same sample set to commercially available single analyte immunoassay formats. Results: Extensive qualification data for raw materials and assay characteristics will be presented, including but not limited to the selected monoclonal antibodies (e.g., precision, sensitivity, matrix interference, accuracy, run acceptance, stability). The formulation of the reagents included in the assay as well as the assay design resulted in the absence of matrix interference. The automation and the availability of ready-to-use calibrators with good stability further adds to the robustness of the approach. Conclusions: The newly qualified assays for m-Tau and Ab(1-42), combined with assays for Ab(1-40), allow for a more reliable quantification in CSF, ultimately resulting in a better clinical diagnostic accuracy and improved patient selection for clinical trials. The sharing of performance characteristics of our novel approaches will be beneficial for the whole AD community. P1-172
HEART RATE VARIABILITY IN ALZHEIMER’S DISEASE AND MILD COGNITIVE IMPAIRMENT
Hyun Kim, Kang Joon Lee, Inje University Ilsan Paik Hospital, Goyang-si, South Korea. Contact e-mail: [email protected] Background: Heart rate variability (HRV) is a sensitive and non-invasive method for the assessment of autonomic function. It has been suggested that the autonomous nervous system may be impaired across a spectrum of dementias associated with cholinergic deficits, as acetylcholine is essential for parasympathetic and pre-ganglionic sympathetic neurotransmission. Alzheimer’s disease (AD) has been associated with underactivity of the cholinergic nervous systems. Methods: A total of 93 subjects were divided into three groups (30 MCI subjects, 35 AD patients, and 28 controls). AD patients met NINCDS/ADRDA criteria, and the diagnosis of MCI was based on Mayo Clinic criteria. All subjects had a full medical history. Cardiovascular risk factors, duration of illness, MMSE and CDR were recorded. HRV assessments took place in the morning. Five minutes of R-R interval data were digitised and stored on computer for subsequent off-line analysis. Power spectral analysis used to obtain spectral bands in the very-low-frequency (<0.04 Hz), low-frequency (0.04-0.15Hz) and high-frequency (0.15-0.40Hz) bands and also total spectral power (<0.04Hz) according to international HRV guidelines. Results: In the power spectral analysis, LF and HF power spectral mean values were significantly lower in patients with AD than in those with MCI and controls. Standard deviation of the
P1-170
EVALUATION OF PRE-ANALYTICAL AND ANALYTICAL FACTORS THAT CONTRIBUTE TO VARIABILITY IN CSF BIOMARKER LEVELS
Aarti Shah1, Courtney Sutphen2, Matthew Amos2, Matthew Althage2, John Morris2, David Holtzman2, Anne Fagan2, 1Washington University, Saint Louis, Saint Louis, Missouri, United States; 2Washington University, St. Louis, Missouri, United States. Contact e-mail: [email protected] Background: Studies of CSF from Alzheimer’s disease research cohorts have identified several markers of disease pathology which show utility for disease diagnosis and prognosis (even in early and pre-clinical stages). However, anecdotal observations, lab-specific experiments and more recently, a world-wide initiative have shed light on the adverse influence of various pre-analytical, analytical and post-analytical factors on CSF biomarker values. Development of disease-modifying therapies and the proposed revision of diagnostic criteria that include biomarkers make biomarker protocol optimization and standardization a top priority. To promote such efforts, we have evaluated several potential sources of methodological variability in analyses of CSF. Methods: Existing CSF samples obtained from well-characterized research volunteers at the Knight Alzheimer’s Disease Research Center were analyzed by plate-based ELISA or bead-based immunoassays for A b 40, A b 42, tau and/or ptau181. Investigated potential sources of variability included: assay platform type, commercial kit lot number, two vendor-specific polypropylene aliquot tubes, use of sample pre-plate during analysis, aliquot volume, number of wells used to generate a mean value, and inter-personnel performance. Student’s t tests and Pearson correlation coefficients were used to evaluate data correspondence. Results: Analyte values did not differ between the two polypropylene aliquot tube types evaluated, whereas aliquot size did have an effect. Assay platform had a large impact on absolute analyte values, although values obtained between platforms were highly correlated. Commercial kit lot-to-lot variability was observed to varying extents and, when present, had a substantial impact on the evaluation of longitudinal biomarker changes within individuals over time. Mean values generated from samples run in duplicate were as precise as those run in quadruplicate for the assay tested. Use of a pre-plate for sample loading influenced levels of A b 42 but not levels of tau or ptau181. High inter-personnel reliability was achieved when well-trained individuals used identical protocols. Conclusions: Several sources of pre-analytical and analytical variability in the measurement of CSF biomarkers exist, and thus could impact the diagnostic and prognostic utility of such markers. Protocol standardization efforts within laborato-
P211
ries and eventually between laboratories are needed in order to increase the usefulness of these markers in both the research and clinical settings. P1-171
A NEW ERA OF CSF BIOMARKER TESTING IN THE FIELD OF ALZHEIMER’S DISEASE
Sebastiaan Engelborghs1, Erik Stoops2, Britta Brix3, Leentje Demeyer2, Dirk Jacobs2, Hanne Struyfs1, Eugeen Vanmechelen2, 1University of Antwerp, Antwerp, Belgium; 2ADx Neurosciences, Gent, Belgium; 3 EUROIMMUN AG, L€ubeck, Germany. Contact e-mail: sebastiaan. [email protected] Background: In the field of Alzheimer’s disease (AD) diagnostics and clinical trials, there is an urgent need to establish novel concepts for CSF biomarker testing. These novel concepts should allow earlier identification of specific brain pathology (e.g., amyloidopathy, tauopathy, loss of synapses) and should help predict the progression rate in disease-affected subjects. Last but not least, approval of the assays for inclusion in clinical trials is indispensable. Notwithstanding the proven clinical utility of the CSF biomarkers, such as phosphorylated tau or the b-Amyloid (Ab) proteins, all currently available immunoassays lack important analytical performance characteristics, hampering their world-wide integration and approval by regulatory authorities, especially in the US. The present study describes solutions for these analytical drawbacks. In addition, the qualification of the clinical utility of integrating CSF- Ab(1-40) in the panel as an aid for differential dementia diagnosis, was performed. Methods: The analytical and clinical qualification of the newly developed assay studies was performed using well-characterized CSF samples from multiple expert centers and carefully designed study protocols. The subject population included controls, patients with AD or vascular dementia (VAD). The outcome of the current assays was compared on-site with and using the same sample set to commercially available single analyte immunoassay formats. Results: Extensive qualification data for raw materials and assay characteristics will be presented, including but not limited to the selected monoclonal antibodies (e.g., precision, sensitivity, matrix interference, accuracy, run acceptance, stability). The formulation of the reagents included in the assay as well as the assay design resulted in the absence of matrix interference. The automation and the availability of ready-to-use calibrators with good stability further adds to the robustness of the approach. Conclusions: The newly qualified assays for m-Tau and Ab(1-42), combined with assays for Ab(1-40), allow for a more reliable quantification in CSF, ultimately resulting in a better clinical diagnostic accuracy and improved patient selection for clinical trials. The sharing of performance characteristics of our novel approaches will be beneficial for the whole AD community. P1-172
HEART RATE VARIABILITY IN ALZHEIMER’S DISEASE AND MILD COGNITIVE IMPAIRMENT
Hyun Kim, Kang Joon Lee, Inje University Ilsan Paik Hospital, Goyang-si, South Korea. Contact e-mail: [email protected] Background: Heart rate variability (HRV) is a sensitive and non-invasive method for the assessment of autonomic function. It has been suggested that the autonomous nervous system may be impaired across a spectrum of dementias associated with cholinergic deficits, as acetylcholine is essential for parasympathetic and pre-ganglionic sympathetic neurotransmission. Alzheimer’s disease (AD) has been associated with underactivity of the cholinergic nervous systems. Methods: A total of 93 subjects were divided into three groups (30 MCI subjects, 35 AD patients, and 28 controls). AD patients met NINCDS/ADRDA criteria, and the diagnosis of MCI was based on Mayo Clinic criteria. All subjects had a full medical history. Cardiovascular risk factors, duration of illness, MMSE and CDR were recorded. HRV assessments took place in the morning. Five minutes of R-R interval data were digitised and stored on computer for subsequent off-line analysis. Power spectral analysis used to obtain spectral bands in the very-low-frequency (<0.04 Hz), low-frequency (0.04-0.15Hz) and high-frequency (0.15-0.40Hz) bands and also total spectral power (<0.04Hz) according to international HRV guidelines. Results: In the power spectral analysis, LF and HF power spectral mean values were significantly lower in patients with AD than in those with MCI and controls. Standard deviation of the