Abstract Body: Introduction: Patients with transient ischemic attack (TIA) and those with stroke mimics (MIM) are often difficult to distinguish in emergency room (ER) settings. While TIA patients are at an increased risk of stroke, MIM do not need stroke-related management. Biomarkers that distinguish TIA from MIM could guide risk-stratification and resource utilization in acute stroke care.
Objectives: We undertook an exploratory proteomics study to nominate differentially enriched proteins (DEPs) as plasma biomarkers that distinguish TIA from MIM.
Methods: Plasma samples from TIA and MIM were obtained from a prospective plasma repository (2010 to 2014) in which blood samples were obtained from adults ≥18 years of age who presented to the ER with acute neurological symptoms. We performed data-independent acquisition (DIA) label-free quantitative mass spectrometry (Orbitrap Astral MS) on age-matched TIA (n=20) and MIM (n=20) plasma samples, pooled into 4 TIA and 4 MIM plasma pools (5 cases per pool). Data were log₂ transformed, and random imputation for proteins with ≤50% missing values was performed (Perseus 2.0.11). We applied quantile normalization and identified DEPs using ±1.5-fold change and an unadjusted p-value <0.05, as well as the Boruta random forest-based machine learning algorithm. We conducted gene ontology analysis of the upregulated proteins in TIA (clusterProfiler package in R 4.3.2). We further validated these DEPs using data from our previous SomaScan proteomics study on the same samples.
Results: We quantified 956 proteins, of which 524 proteins had ≤50% missing values and >90% had a coefficient of variation <20%. We identified 40 DEPs, with 32 increased and 8 decreased in TIA compared to MIM. Gene ontology highlighted pathways related to the extracellular matrix, intracellular lumen-related compartments, immune response, complement system, and reproductive processes in proteins increased in TIA (Figure 1). External validation identified 789 overlapping proteins between the DIA-MS and SomaScan platforms. Increased levels of IGFBP2, FTL, and RNASE1 in TIA compared to MIM were validated (Figure 2).
Conclusions: Our discovery proteomics study highlights the potential of DIA-MS proteomics as a valuable tool for discovering novel protein biomarkers to distinguish TIA from MIM. These findings warrant validation in larger, longitudinal studies.