Annals of Clinical and Translational Neurology

Alzheimer's Disease Biomarkers Measured by Capillary Blood Collection Versus Venous Plasma and Cerebrospinal Fluid

Study in peopleBiomarkers
Follow

Abstract

OBJECTIVE: Blood biomarkers are increasingly critical for non-invasive Alzheimer's Disease (AD) diagnostics and research but typically require on-site processing. Remote point-of-care collection methods could enhance biomarker testing in clinical and research settings. We evaluated the safety and accuracy of blood collection via venous blood and two capillary blood collection devices (Tasso+ and TAPII) for measurements of amyloid-beta 42/40 ratio (Aβ42/40), phosphorylated-tau-217 (pTau217), neurofilament-light (NfL), and glial fibrillary acidic protein (GFAP) from capillary plasma to simultaneously collected venous plasma and cerebrospinal fluid (CSF). METHODS: Patients seen at the Johns Hopkins Center for CSF Disorders (N = 192) underwent diagnostic lumbar punctures and venipuncture. Warming packs were held to the skin for 5 min before collecting capillary blood (via TAPII or Tasso+ device) and numeric pain scores. CSF, venous, and capillary plasma were analyzed using Quanterix SIMOA immunoassays. Statistical analysis included Pearson/Spearman correlations, Lasso regression, and quantitative comparisons using Bland-Altman plots. RESULTS: 86 TAPII and 126 Tasso+ plasma samples were collected (77 M, 115 F; mean age 61 ± 18.3 years, range 16-91). Tasso+ devices were 90.3% successful in collecting > 50 μL whole blood, and TAPII devices were 86% successful. Mean pain scores were significantly higher in venous than capillary collections (Venous 2.24/10, TAPII 0.73/10, Tasso+ 0.2/10, p < 0.001). Capillary Aβ42/40, NfL, and GFAP strongly correlated with venous plasma analytes (Tasso+: Aβ42/40 r = 0.75, NfL r = 0.91, GFAP r = 0.80, all p < 0.001; TAPII: Aβ42/40 r = 0.74, NfL r = 0.86, GFAP r = 0.84, all p < 0.001). Tasso+ capillary pTau217 weakly correlated with venous (r = 0.23, p = 0.04). Capillary NfL and GFAP correlated modestly with CSF NfL and GFAP (TAPII NfL r = 0.74, GFAP r = 0.6, all p < 0.0001; Tasso+ NfL r = 0.69, GFAP r = 0.59, all p < 0.0001). Capillary Aβ42/40 and pTau217 did not significantly correlate with CSF measures. Lasso regression identified age and GFAP as significant predictors of AD across capillary and venous samples (Tasso+ AUC = 0.87, TAPII AUC = 0.75, Venous AUC = 0.83). INTERPRETATION: Neurodegeneration biomarkers Aβ42/40, NfL, and GFAP in capillary plasma (Tasso+, TAPII) strongly correlated with venous plasma measures. Capillary and venous pTau217 correlated weakly, indicating further assay refinement is needed. Tasso+ collection was less painful and caused less hemolysis compared to TAPII.