The technical reliability and stability of ATN quantification using the Fujirebio Lumipulse G1200 fully‐automated immunoassay instrument in cerebrospinal fluid

Abstract

AbstractBackgroundFujirebio’s Lumipulse G1200 is designed to perform fully‐automated, rapid, random access immunoassays for biomarker quantification in biofluids. The Lumipulse instrument was used for quantification of the core Alzheimer’s Disease (AD) cerebrospinal fluid (CSF) biomarkers for Amyloid, Tau, and neurodegeneration (ATN): amyloid β 1–40 (Aβ40), amyloid β 1–42 (Aβ42), phosphorylated Tau 181 (pTau181), and total Tau (tTau). This study evaluated the reliability of the Lumipulse assays by measuring intra‐ and inter‐batch coefficients of variation (CV), and the correlation between Lumipulse values and the clinically used Athena ADmark Aβ42, pTau181 and tTau assays.Method423 CSF samples collected from diagnostic lumbar punctures between 2015 and 2020 were obtained from the MIND biorepository. Aβ40, Aβ42, tTau, and pTau were measured using the Lumipulse G1200 instrument in batches of 50‐60 samples. We tested 60 samples in duplicate to compare the intrabatch stability using the CVs between each pair. Each batch was run with a new set of calibrator controls at the start of each assay to ensure stability for each biomarker run. Finally, we correlated values from the Lumipulse with results from the clinical evaluation using the Athena ADmark test.ResultWe observed low variability in calibrator and control signals between batches with an overall calibrator signal CV of <5% for Aβ40 and tTau, and <10% for Aβ42 and pTau. Control signal CV values for all analytes were <5%. The mean duplicate CV across all batches was similarly stable at <10% for Aβ40, Aβ42, and tTau, and <5% for pTau. Subjects with CV >20% were seen <5% of the time. The correlations between Lumipulse and clinical ADmark results were high for pTau and tTau (R2>0.8 for both analytes), but lower for Aβ42 (R2∼0.5).ConclusionOur results demonstrated that the Lumipulse G1200 is a reliable, easy to use platform for ATN quantification with low sample to sample variability, and generally high correlation to established clinical standards. Inconsistent preanalytical sample handling of clinical send‐outs for Athena ADmark Aβ42 likely resulted in a poor correlation with Lumipulse G1200. The performance of the calibrators and controls proved that the standardization in each batch was consistent.