PROgnostic Value of MicroParticles and Markers of Hemostasis in TIA and Ischemic Stroke

1. Introduction Within a previous study ('PROPPSTOPP') a cohort of 249 patients with ischemic stroke (IS) or TIA was established in Stockholm between 2007 and 2009. The original aims were twofold: 1) to look for occult atrial fibrillation (completed, NTC01160406) and 2) to investigate microparticle levels and markers of hemostasis in the acute phase and one month after symptoms. Blood samples were taken on both occasions. Patients were found to have higher microparticle levels and increased thrombin generation in plasma both acutely and at one month as compared to healthy controls. This follow-up study will investigate the prognostic ability of microparticles (primary variables) and markers of hemostasis (secondary variables) to predict outcome as documented in hospital records and Swedish registers. The association between microparticles/coagulation markers and stroke subtype/etiology will also be investigated.
2. Variables 2.1 Microparticles Microparticles are membrane vesicles (0.1-1.0 µm) released from cells at activation or apoptosis. They carry surface markers from the releasing cell. Microparticles are of interest as biomarkers for activation of cells in the circulation, e g platelets, endothelial cells or leukocytes. They may have pro-coagulant properties. For this study the following microparticles and surface markers have been analyzed:

• Total number of microparticles (MP's)
• Number of MP's exposing phosphatidylserine (PS) (pro-coagulant property)
• Number of MP's of any type exposing tissue factor (TF), with or without simultaneous exposure of PS (pro-coagulant properties, activation of monocytes/endothelium)
• Number of platelet microparticles (PMP's), identified by surface exposure of GPαIIb (CD41), with or without simultaneous exposure of PS (platelet activation)
• Number of PMP's exposing P-selectin with or without simultaneous exposure of PS (platelet activation)
• Number of PMP's exposing TF with or without simultaneous exposure of PS (pro-coagulant properties) 2.2 Markers of hemostasis Different variables of importance for coagulation, platelet activation, fibrinolysis and inflammation were measured in the original study, namely:
• Prothrombin fragment F1+2
• Thrombin generation by Calibrated Automated Thrombin generation (CAT)
• Microparticle-pellet induced thrombin generation

1. Outcome 3.1 Primary outcome The primary outcome is new ischemic events, i e fatal or non-fatal recurrent ischemic stroke, myocardial infarction or ischemic cardiovascular death in the time frame 2007-2014. Diagnoses will be retrieved from the following sources:

• The Swedish Register for Cause of death
• The Swedish Register of in-house hospital care
• Medical records at recruiting hospitals. 3.2 Secondary outcome Secondary outcomes are recurrent ischemic stroke respectively all-cause mortalitý as documented in the sources above.

1. Statistical Methods

The statistical analysis of the prospective cohort study will be performed in steps:

1. Descriptive statistics comparing averages/medians of the above variables for patients with and without outcome.
2. For variables with p-values < 0.2 in step 1: univariate event-free survival Kaplan-Meyer curves by variable median split. Probability calculations of differences by log-rank test.
3. Multivariate analysis by Cox regression analysis including the most significant variables in step 2 and established cardiovascular risk factors.