HAI Book 2025 - Flipbook - Page 683
Wilson, Rachael
167
Dried plasma spot optimization for the NULISAseq CNS panel
Rachael Wilson1,2, Ramiro Eduardo Rea Reyes1, Aaron Fredricks1,2, Alamar TAP team3, Sara
Rusch4, Monica VandenLangenberg1,2, Cindy Jensen1,2, Martie Marshall1,2, Elysse Keske1,2,
Haley Weninger5, Beckie Jeffers1,2, Hanna Huber5, Sterling Johnson1,2, Henrik
Zetterberg1,5,6,7,8,9
Wisconsin Alzheimer’s Disease Research Center, School of Medicine and Public Health, University of Wisconsin Madison, Madison, WI, US
2
Wisconsin Alzheimer’s Institute, School of Medicine and Public Health, University of Wisconsin - Madison, Madison,
WI, US
3
Alamar Biosciences, Fremont, CA, US
4
Clinical Research Unit, UW Health, Madison, WI, US
5
Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, the Sahlgrenska Academy
at the University of Gothenburg, Mölndal, SE
6
Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, SE
7
Department of Neurodegenerative Disease, University College London Institute of Neurology, London, GB
8
UK Dementia Research Institute at University College London, London, GB
9
Hong Kong Center for Neurodegenerative Diseases, Hong Kong, CN
1
Scalable approaches to screening for neurodegenerative dementias including Alzheimer9s disease (AD) would
expand access to diagnosis and treatment. Dried plasma spots (DPS) offer in-home fingerprick blood collection
without centrifugation or cold storage. However, a challenge of method development for DPS protein extraction is
maintaining practical sample volumes without diluting analytes beyond the limits of detection (LOD). High
sensitivity assays with low volume requirements are ideal for DPS analyses.
With this in mind, we developed a DPS extraction method for the NULISAseq CNS panel, a low volume (25uL), high
sensitivity assay targeting 120 analytes related to central nervous system (CNS) dysfunction. We then analyzed
paired DPS and plasma from N=40 Wisconsin Registry for Alzheimer9s Prevention and Wisconsin Alzheimer9s
Disease Center participants, the majority of whom were cognitively unimpaired (N=36 (90.0%), N=3 MCI (7.5%), N=1
(2.5%) missing). Amyloid (13C-PiB) PET was available for 19 participants (13 A- (68.4%), 6 A+ (31.6%)). The mean (SD)
detectability, or percent of analytes detected above the LOD, across DPS samples was 69.6% (9.5%). Spearman
correlation coefficients (r) between DPS and plasma concentrations varied across analytes (Figure 1). For core AD
biomarkers pTau181, pTau231, pTau217, NfL, and GFAP, the r (p-value) was 0.35 (0.028), 0.42 (0.007), 0.35 (0.028),
0.64 (8e-6), and 0.46 (0.003), respectively. Nominally, pTau231 had the highest concordance with amyloid PET
compared to other pTau species in DPS (ROC-AUC,95% CI: 0.71, 0.41-1.0, Figure 2).
A major limitation of this study is sample size, especially in the impaired group. This was due to sampling by
convenience rather than diagnosis. Despite this, we observed good correlations between DPS and plasma for NfL
and modest correlations for pTau. Although better correlations for pTau would likely improve agreement with
PET, these data are a promising first step in utilizing DPS with NULISA. Performance will likely improve with
further optimization.
HAI2025 - 683
