(P22-020-23) Comparison of the Capillary and Venous Blood Lipidomes- Validation of the Tasso SST Capillary Blood Collection Device for Circulating Lipid Biomarker Analysis

Objectives: Circulating lipids are proven or potential biomarkers of human health and diet. Studies of the human plasma lipidome use venous blood collected by venipuncture devices requiring both clinical staff and visits. Devices for self-collection of relatively small volumes of capillary blood have recently been introduced allowing subjects to collect blood at home which is particularly attractive for longitudinal studies. The purpose of this study was to evaluate the suitability of one of these self-collection devices (Tasso + EDTA, Tasso Inc.) by comparing the lipidomes of matched venous and capillary blood samples from healthy adults.

Methods: Venous and capillary bloods were collected from twelve healthy adults using standard venipuncture and TassoOne™-hemolink devices, respectively. Plasma was prepared and used for LC-MS based global lipidomics. Data process and lipid identification were performed using the software MS-Dial v5.1.The multivariate, univariate, and correlative statistical analysis were performed on MetaboAnalyst with 95% confidence level.

Results: Two datasets having total 25,465 and 19,148 known lipidomic features (ESI⁺ & ESI^-, respectively) were subjected to data quality assurance (QA) and control (QC) procedure selecting features detected in >80% of QC samples with RSD < 30%. Treated datasets were used to build volcano plots based on supervised orthogonal partial least-squares discriminant analysis (OPLS-DA) models to differentiate venous-blood and capillary-blood lipidomes. Interestingly, venous and capillary blood could not be differentiated by lipidomics profile as OPLSDA scatter plot despite of showing a clear distinction (R²Y=0.517 ESI⁺, R²Y=0.454 ESI^-) between groups failed in cross-validation of model (Q²=-0.889 ESI⁺, Q²=-0.952 ESI^-) indicating non-discriminatory nature of lipid features between venous and capillary blood. Similar results were also seen while univariate analysis with Benjamini and Hochberg false discovery rate (FDR) validations. Pearson’s correlation also confirmed the close similarity (R²>0.99) of blood collected from same participant regardless of said two methods.

Conclusions: This work suggests that TassoOne™-drawn capillary blood is likely equivalent to venipuncture-blood and is suitable for nutritional lipidomics and lipid biomarker studies

Funding Sources: NIH/NINDS and USDA