ABSTRACT NUMBER: ESOC2026A696 MULTIPLEX PROTEOMIC BIOMARKERS FOR ACUTE STROKE SUBTYPE DIFFERENTIATION: AN EXPLORATORY STUDY

Abstract Background and aims Rapid acute stroke triage remains challenging, particularly for early identification of intracerebral hemorrhage (ICH) and large-vessel-occlusion acute ischemic stroke (LVO-AIS). We assessed whether multiplex proteomics can identify biomarker signatures to improve early stroke subtype classification. Methods Exploratory analyses of a prospective cohort of suspected stroke participants presenting <24 hours from onset. Blood samples were collected upon emergency department arrival and plasma analyzed using the Olink Target 96 CVDIII proteomics panel. Differentially expressed proteins (DEPs) were identified for stroke (AIS+ICH) versus stroke mimics (SM) and across four diagnostic categories (LVO-AIS, non-LVO-AIS, ICH, SM) using non-parametric testing with false discovery rate correction (q<0.05). Multivariate LASSO regression with 10-fold cross-validation derived protein panels for binary classification tasks (stroke versus SM, ICH versus non-ICH, and LVO-AIS versus non-LVO). Results Among 175 participants (mean±SD age 71.2±15.7 years; median NIHSS (IQR) 10[4-19]), final diagnoses were 40 ICH, 82 AIS (46 LVO), and 53 SM. Stroke versus SM identified 19 DEPs (FDR<0.05), with strongest signals for NT-proBNP, Notch3, KLK6, MMP-2, and t-PA. Four-group analyses identified 10 significant DEPs (FDR<0.05) (Fig. 1). A LASSO-derived 24-protein panel discriminated stroke from SM with cross-validated AUC 0.68 (95%CI [0.65-0.75]). For ICH detection, a two-protein panel (KLK6, PLC) achieved AUC 0.80 (95%CI [0.75-0.86]). LVO versus non-LVO showed no meaningful discrimination (AUC 0.57(95%CI [0.53-0.62])). Conclusions Multiplex proteomics identifies blood-based signatures distinguishing stroke from SM and detecting ICH. The absence of distinct LVO-specific signatures highlights the need for complementary approaches for early LVO identification. Conflict of interest Julien Paul : nothing to disclose; Catherine Brassard : nothing to disclose; Ralph Nuhay : nothing to disclose; David Chiasson-Ricard : nothing to disclose; Gloria Ortega-Delgado : nothing to disclose; Nathalie Arbour: nothing to disclose; Laura C. Gioia : I declare having current funding from the Heart and Stroke Foundation of Canada. Figure 1 - belongs to Conclusions

Authors

Institutions

Publication Details

Journal
European Stroke Journal
Published
2026-05-01
DOI
https://doi.org/10.1093/esj/aakag023.400
Primary Topic
Acute Ischemic Stroke Management
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

ABSTRACT NUMBER: ESOC2026A696 MULTIPLEX PROTEOMIC BIOMARKERS FOR ACUTE STROKE SUBTYPE DIFFERENTIATION: AN EXPLORATORY STUDY

Nathalie Arbour, Gloria Ortega-Delgado, Laura Gioia, Julien Paul et al.
European Stroke Journal
Acute Ischemic Stroke Management
article

ABSTRACT NUMBER: ESOC2026A696 MULTIPLEX PROTEOMIC BIOMARKERS FOR ACUTE STROKE SUBTYPE DIFFERENTIATION: AN EXPLORATORY STUDY

Nathalie Arbour, Gloria Ortega-Delgado, Laura Gioia, Julien Paul, Catherine Brassard, David Chiasson-Ricard, Ralph Nuhay
article en

Abstract

Abstract Background and aims Rapid acute stroke triage remains challenging, particularly for early identification of intracerebral hemorrhage (ICH) and large-vessel-occlusion acute ischemic stroke (LVO-AIS). We assessed whether multiplex proteomics can identify biomarker signatures to improve early stroke subtype classification. Methods Exploratory analyses of a prospective cohort of suspected stroke participants presenting <24 hours from onset. Blood samples were collected upon emergency department arrival and plasma analyzed using the Olink Target 96 CVDIII proteomics panel. Differentially expressed proteins (DEPs) were identified for stroke (AIS+ICH) versus stroke mimics (SM) and across four diagnostic categories (LVO-AIS, non-LVO-AIS, ICH, SM) using non-parametric testing with false discovery rate correction (q<0.05). Multivariate LASSO regression with 10-fold cross-validation derived protein panels for binary classification tasks (stroke versus SM, ICH versus non-ICH, and LVO-AIS versus non-LVO). Results Among 175 participants (mean±SD age 71.2±15.7 years; median NIHSS (IQR) 10[4-19]), final diagnoses were 40 ICH, 82 AIS (46 LVO), and 53 SM. Stroke versus SM identified 19 DEPs (FDR<0.05), with strongest signals for NT-proBNP, Notch3, KLK6, MMP-2, and t-PA. Four-group analyses identified 10 significant DEPs (FDR<0.05) (Fig. 1). A LASSO-derived 24-protein panel discriminated stroke from SM with cross-validated AUC 0.68 (95%CI [0.65-0.75]). For ICH detection, a two-protein panel (KLK6, PLC) achieved AUC 0.80 (95%CI [0.75-0.86]). LVO versus non-LVO showed no meaningful discrimination (AUC 0.57(95%CI [0.53-0.62])). Conclusions Multiplex proteomics identifies blood-based signatures distinguishing stroke from SM and detecting ICH. The absence of distinct LVO-specific signatures highlights the need for complementary approaches for early LVO identification. Conflict of interest Julien Paul : nothing to disclose; Catherine Brassard : nothing to disclose; Ralph Nuhay : nothing to disclose; David Chiasson-Ricard : nothing to disclose; Gloria Ortega-Delgado : nothing to disclose; Nathalie Arbour: nothing to disclose; Laura C. Gioia : I declare having current funding from the Heart and Stroke Foundation of Canada. Figure 1 - belongs to Conclusions

European Stroke JournalVol. 11(Supplement_1)
Centre Hospitalier de l’Université de Montréal (CA)
Good health and well-being
Openalex Percentile: Top 8%
Acute Ischemic Stroke Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.