Authors (including presenting author) :
Zhang XD(1), Cai YE(1), Zhang YL(1), Li QY(1), Fan WN(2), Jin YL(3), Rainer TH (1)
Affiliation :
(1) Department of Emergency Medicine, The University of Hong Kong, Hong Kong SAR, China (2) Department of Neurology, Ningbo No.2 Hospital, Ningbo, Zhejiang Province, Ningbo, China 8 (3) Department of Clinical Laboratory, Ningbo No.2 Hospital, Ningbo, Zhejiang Province, Ningbo, China
Keyword 1: :
Ischaemic stroke
Keyword 2: :
Large vessel occlusion
Keyword 5: :
Thrombectomy
Introduction :
Rapid identification of acute ischaemic stroke (AIS) due to large vessel occlusion (LVO) and timely transport to thrombectomy-capable hospitals remains a significant challenge in clinical practice.
Objectives :
This study aims to utilise protein biomarkers to diagnose AIS due to LVO in patients presenting with stroke-like symptoms.
Methodology :
This single-centre, case-control trial was conducted at Ningbo No.2 Hospital from May 2022 to February 2023, recruiting 30 AIS patients, 6 haemorrhagic stroke (HS) patients and 11 controls. Plasma proteomic patterns were analysed using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Potential diagnostic proteins for AIS due to LVO were identified through a stepwise pipeline incorporating machine learning and bioinformatic analysis. Logistic regression was employed to develop diagnostic models for the prehospital identification of AIS due to LVO.
Result & Outcome :
We identified nine plasma proteins with the potential to diagnose AIS due to LVO, distinguishing it from other conditions (AIS due to non-LVO, HS, and controls). Among these, five proteins demonstrated high diagnostic accuracy with area under the curve (AUC) values greater than 0.9: chitinase-3-like protein 1(CHI3L1) [0.948 (95%CI 0.885 - 1.000)], tyrosine-protein kinase Mer (MERTK) [0.944 (95%CI 0.880 - 1.000)], disintegrin and metalloproteinase domain-containing protein 15 (ADAM15) [0.912 (95%CI 0.831 - 0.994)], interleukin-1 receptor type 2 (IL1R2) [0.910 (95%CI 0.827 - 0.993)], and epithelial discoidin domain-containing receptor 1(DDR1) [0.904 (95%CI 0.815 - 0.993)]. Additionally, we identified 15 protein panels that increased the prehospital diagnostic accuracy of AIS due to LVO, achieving an AUC greater than 0.981 and a negative predictive value up to 1. In conclusion, The nine identified proteins and 15 diagnostic models, pending further validation, hold significant potential for rapid prehospital and in-hospital identification of AIS due to LVO. We established a biomarker-based pathway for the identification and management of AIS due to LVO, which has the potential to reduce treatment delays and improve stroke outcomes.
