Exploratory study of blood biomarkers in patients with post-stroke epilepsy
Laura Abraira, Samuel López‐Maza, Manuel Quintana, Elena Fonseca, Manuel Toledo, Daniel Campos, Sofía Lallana, Laia Grau‐López, Jordi Ciurans, Marta Jiménez, Juan Luís Becerra, Alejandro Bustamante, Marta Rubiera, Anna Peñalba, Joan Montaner, José Álvarez‐Sabín, Estevo Santamarina
Abstract
Abstract Introduction: In addition to clinical factors, blood-based biomarkers can provide useful information on the risk of developing post-stroke epilepsy (PSE). Our aim was to identify serum biomarkers at stroke onset that could contribute to predicting patients at higher risk of PSE. Patients and methods: From a previous study in which 895 acute stroke patients were followed-up, 51 patients developed PSE. We selected 15 patients with PSE and 15 controls without epilepsy. In a biomarker discovery setting, 5 Olink panels of 96 proteins each, were used to determine protein levels. Biomarkers that were down-regulated and overexpressed in PSE patients, and those that showed the strongest interactions with other proteins were validated using an enzyme-linked immunosorbent assay in samples from 50 PSE patients and 50 controls. A ROC curve analysis was used to evaluate the predictive ability of significant biomarkers to develop PSE. Results: Mean age of the PSE discovery cohort was 68.56 ± 15.1, 40% women and baseline NIHSS 12 [IQR 1–25]. Nine proteins were down-expressed: CASP-8, TNFSF-14, STAMBP, ENRAGE, EDA2R, SIRT2, TGF-alpha, OSM and CLEC1B. VEGFa, CD40 and CCL4 showed greatest interactions with the remaining proteins. In the validation analysis, TNFSF-14 was the single biomarker showing statistically significant downregulated levels in PSE patients (p = 0.006) and it showed a good predictive capability to develop PSE (AUC 0.733, 95% CI 0.601–0.865). Discussion and conclusion: Protein expression in PSE patients differs from that of non-epileptic stroke patients, suggesting the involvement of several different proteins in post-stroke epileptogenesis. TNFSF-14 emerges as a potential biomarker for predicting PSE.