What question did this study set out to answer?

The study aims to assess the predictive value of GFAP trajectories for outcomes following reperfusion in large vessel occlusion stroke.

May 8, 2026

Abstract Number: Esoc2026a1274 Gfap Trajectories Can Predict Hemorrhagic Transformation and Early Neurological Deterioration in Large Vessel Occlusion Stroke

Key Points

The study aims to assess the predictive value of GFAP trajectories for outcomes following reperfusion in large vessel occlusion stroke.
Plasma GFAP levels measured at hospital arrival, post-reperfusion, and 24 hours after treatment.
Outcomes include hemorrhagic transformation and early neurological deterioration, analyzed using regression methods.
Sensitivity analyses explored GFAP cutoff points based on existing assays.
HT occurred in 35.0% of participants, and early neurological deterioration was observed in 25.0%.
Log-fold increases in GFAP from post-reperfusion to 24 hours predicted HT (aOR 2.95, 95% CI 1.69–6.22, p=0.0009) and END (aOR 1.98, 95% CI 1.20–3.60, p=0.0127).
T1-T2 GFAP rise >213 pg/mL indicated moderate–severe HT with high specificity (0.94), while T3 GFAP ≤213 pg/mL excluded HT with sensitivity (0.96).

Abstract

Abstract Background and aims Glial fibrillary acidic protein (GFAP), a promising stroke biomarker, remains understudied in large-vessel occlusion (LVO) stroke. We assessed whether GFAP trajectories predict post-reperfusion outcomes. Methods Plasma GFAP levels were measured using the iSTAT Alinity system at T1=hospital arrival, T2=post-reperfusion, and T3=24h. Outcomes included hemorrhagic transformation (HT) as per Heidelberg criteria, and early neurological deterioration (END) defined as a 24h-NIHSS increase ≥4, ICU transfer, palliation and/or mortality ≤72 hours. GFAP values were log-transformed. Regression analyses were adjusted for age, thrombolysis and recanalization (TICI 2b-3). Sensitivity analyses explored a 213 pg/mL cutoff based on commercial point-of-care assays. Results Eighty participants were included; 42 received thrombolysis, 55 underwent thrombectomy (47 recanalized), 34 received both. HT occurred in 28(35.0%), moderate–severe HT (Heidelberg ≥1c) in 11(13.8%), and END in 20(25.0%). Median(IQR) GFAP increased from 51(29-80) (T1) to 84(44-129) (T2) to 686(154-2309) (T3). Log-fold GFAP increases from T2–T3 predicted HT (aOR 2.95 95%CI (1.69–6.22), p=0.0009) and END (aOR 1.98(1.20–3.60), p=0.0127). T1–T3 change showed good discrimination for HT (AUC 0.85(0.75–0.93). Among recanalized participants, T2-T3 increase predicted END (OR 3.9(1.63–16.16), p=0.014) compared to non-recanalized participants (OR 1.12(0.56–2.26), p=0.73). T1-T2 GFAP rise 213 pg/mL indicated moderate–severe HT specificity 0.94(0.83–0.99); LR+ 5.82(1.51–22.35), while T3 GFAP ≤213 pg/mL excluded HT sensitivity 0.96(0.80–1.00); LR– 0.08(0.01–0.58. Conclusions GFAP trajectories may support HT and END risk stratification in LVO stroke. Conflict of interest Timothé Langlois-Thérien: nothing to disclose; Catherine Brassard: nothing to disclose; Ralph Robert Nuhay: nothing to disclose; Julien Frédérick Paul: nothing to disclose; David Chiasson-Ricard: nothing to disclose; Sarah-Emilie Godin: nothing to disclose; Dr Laura Gioia: : I declare having current funding from the Heart and Stroke foundation of Canada.

Bookmark

Abstract Number: Esoc2026a1274 Gfap Trajectories Can Predict Hemorrhagic Transformation and Early Neurological Deterioration in Large Vessel Occlusion Stroke

Key Points

Abstract

Cite This Study

Also Consider

Also Consider