Exploring Protein Quantitation Methods to Detect Useful Plasma Biomarkers for Distinguishing Ischemic Stroke from Mimic

Objective : To prospectively compare plasma proteins between disabling ACVS and mimic patients using two protein quantitation methods, as part of a larger multi-phase proteomic ACVS study. Background: A blood-based biomarker test for acute cerebrovascular syndrome (ACVS) would help select patients for advanced imaging and specialist referral, particularly in mild disease where mimic rates are over 50[percnt]. Methods: Blood was collected in the ED from ACVS patients (n=20) with an NIHSSu003e5 and u003c24 hours from symptom onset. Mimic patients (n=20) were enrolled from an ambulatory TIA clinic. Patient phenotype was based on clinical findings and imaging. Average time from stroke symptom onset to blood collection was u003c 10 hrs. 95[percnt] of EDTA plasma samples were frozen within 41 minutes of collection. 31 proteins were quantified using commercial ELISA kits and 141 proteins were analyzed using mass spectrometry (1D-LC/MRM). Results: 60[percnt] of stroke and 65[percnt] of mimic patients were female, and stroke patients were older: median [range] age 77 [46, 95] vs. 63[36, 77]. The 3 most frequent mimic sub-types were migraine (25[percnt]), neuropathy(20[percnt]), and transient global amnesia (20[percnt]). 29 proteins (23 by MRM, 6 by ELISA) were differentially abundant between mimic and stroke (each ELISA pu003c0.1, MRM pu003c0.05). Including the first two principal components of these proteins in a logistic regression model significantly improved discrimination (pu003c0.01, AUC 0.985 vs. 0.810) over a model with age. The 29 significant proteins included markers of inflammation (45[percnt]), coagulation (41[percnt]), neurovascular unit injury (7[percnt]), and atrial fibrillation (7[percnt]). Conclusions: These results provide proof of concept that proteomic signals can be useful in ACVS diagnosis in more severe cases.The analyses have shaped our prospective verification (completed, n=560) and validation studies (ongoing, n=1200) of the use of proteomics in the diagnosis of mimic vs. mild-ACVS in early Emergency Department triage. Disclosure: Dr. Penn has received personal compensation for activities with Pfizer, Boehringer Ingelheim Pharmaceuticals, Inc., and Bayer as a speaker. Dr. Balshaw has nothing to disclose. Dr. Lesperance has nothing to disclose. Dr. Saly has nothing to disclose. Dr. Jackson has nothing to disclose. Dr. Smith has nothing to disclose. Dr. Votova has received personal compensation for activities with Roche. Dr. Lu has nothing to disclose. Dr. Morrison has nothing to disclose. Dr. Coutts has nothing to disclose. Dr. Borchers has nothing to disclose.