MSACL 2026 Abstract

Decentralized Risk Assessment for Sudden Cardiac Arrest: Cross-Matrix Transferability of Plasma Proteomic Signatures to Remote Blood Microsampling

Jihyeon Lee (1), Kotoka Nakamura (2), Kyndaron Reinier (2), Qin Fu (1,3), Ali Haghani (1), Harpriya Chugh (2), Xiuzhen Huang (4), Sumeet S. Chugh (2), Jennifer E. Van Eyk (1)
(1) Advanced Clinical Biosystems Research Institute, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA (2) Center for Cardiac Arrest Prevention, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA (3) ThermoFisher Scientific, San Jose, CA (4) Computational Biomedicine, Cedars-Sinai Medical Center, Los Angeles, CA.

Jennifer van Eyk, PhD (Presenter) Cedars-Sinai Heart Institute

Presenter Bio: Jennifer Van Eyk, PhD, is an international leader in the area of clinical proteomics and her lab has focused on developing technical pipelines for de novo discovery and larger scale quantitative mass spectrometry methods. This includes multiple reaction monitoring (MRM, also known as SRM) and most recently data independent acquisition. Dr. Van Eyk's laboratory is well known for the extreme technical quality of the data generated, rigorous quality control with tight %CV while applying these to key clinical questions. The aim is to maximize throughput and reproducibility in order to move targeted and robust discovery methods into large population healthy continuous assessment and clinical grade assays focusing on brain and cardiovascular diseases.

Relevant Financial Disclosures (within past 24 months, reported on Apr 21, 2026) No relevant financial relationship(s) to disclose.

INTRODUCTION:
Sudden cardiac arrest (SCA) remains a leading cause of mortality worldwide, yet clinically useful blood-based biomarkers for early risk stratification are limited. Most proteomic biomarker discovery efforts have relied on plasma samples collected in clinical settings, which restricts their scalability for large-scale population deployment. We have deployed a volumetric absorptive microsampling (VAMS) Mitra device as a minimally invasive method that enables decentralized blood collection. Despite its practical advantages, it remains unclear whether the candidate proteomic biomarkers we have previously identified in plasma (1) retain their analytical validity and predictive utility when measured in VAMS Mitra dried blood samples. In the context of SCA, where large-scale risk assessment is needed, systematic evaluation of plasma-to-VAMS Mitra samples transferability is a critical unmet need.

OBJECTIVE(S):
Develop a protein biomarker panel for early risk stratification of SCA in the broad population that can be used in both plasma and with remote blood sampling to facilitate use for large-scale population deployment.

METHODS:
Participants for discovery and targeted assays were recruited from two ongoing community-based studies: the Prediction of Sudden Death in Multi-Ethnic Communities (PRESTO) in Ventura County, California, and the Oregon Sudden Unexpected Death Study (SUDS) in the Portland, Oregon metropolitan area. In this study, we leveraged our recent plasma proteomic discovery study that identified potential candidate biomarkers and developed a targeted assay comprising 53 peptides corresponding to 16 proteins on an Orbitrap Exploris 480 mass spectrometer (analyzed as outlined in 2). The targeted protein assay was applied to plasma samples digested on the BioMek i7 automation workstation from 107 individuals (60 SCA cases and 47 controls). Peptide intensities were adjusted for sex and age using mixed-effects regression, and residualized values were used to train an Elastic Net–regularized logistic regression model. Proteins contributing to the final plasma-based model were quantified in Mitra dried blood samples (processed on the i7 work station and analyzed as outlined in 3) from an expanded cohort (n = 406). The cohort was randomly divided into train (n = 284) and test (n = 122) sets. An Elastic Net–regularized logistic regression model was trained on the train set and evaluated performances in the test set.

RESULTS:
An Elastic Net–regularized logistic regression model was trained on sex- and age-adjusted plasma peptide intensities using 5-fold cross-validation. The final plasma-based model incorporated seven proteins and achieved an area under the receiver operating characteristic curve (AUC) of 0.801, demonstrating stable discrimination despite the modest sample size. In the Mitra cohort (>400), an Elastic Net model in the training set retained the same seven-protein composition. An Elastic Net model trained on the train set retained the seven-protein composition and achieved an AUC of 0.82, supporting the transferability of plasma-derived signatures to whole-blood microsamples. Performance was similarly maintained in the test set, with an AUC of 0.80. These results highlight that the proteomic signals captured in plasma can be reproducibly measured in Mitra samples, and plasma-derived protein biomarkers contribute to predictive performance, supporting the feasibility of decentralized biomarker-based SCA risk assessment.
CONCLUSION:
Our seven-protein signature model developed in plasma for SCA prediction can be successfully translated to Mitra samples with minimal loss of predictive performance. It also validated a Net–regularized logistic regression model of proteins that contribute to SCA initially identified in plasma. Finally, this approach provides a scalable framework for decentralized biomarker-based risk assessment and highlights biologically structured matrix effects that should be considered in cross-platform proteomic studies.

REFERENCES:
1. Norby FL, Nakamura K, Fu Q, Venkatraman V, Sundararaman N, Mastali M, Reinier K, Salvucci A, Jui J, Van Eyk JE, Chugh SS. A panel of blood biomarkers unique to sudden cardiac arrest. Heart Rhythm [Internet]. 2023;20:414–422.
2. Mc Ardle A, Binek A, Moradian A, Chazarin Orgel B, Rivas A, Washington KE, Phebus C, Manalo D-M, Go J, Venkatraman V, Coutelin Johnson CW, Fu Q, Cheng S, Raedschelders K, Fert-Bober J, Pennington SR, Murray CI, Van Eyk JE. Standardized Workflow for Precise Mid-
3. van den Broek I, Fu Q, Kushon S, Kowalski MP, Millis K, Percy A, Holewinski RJ, Venkatraman V, Van Eyk JE. Application of volumetric absorptive microsampling for robust, high-throughput mass spectrometric quantification of circulating protein biomarkers. Clinical Mass Spectrometry [Internet]. 2017;4–5:25–33.