We prospectively studied patients with TIA/ischemic stroke from 2 cohorts: the OXVASC (Oxford Vascular) Study and the University of Hong Kong (HKU). In brief, OXVASC is an ongoing population-based study of all acute vascular events occurring within a population of all 92 728 individuals, irrespective of age, who are registered with 100 general practitioners in 9 general practices of Oxfordshire, United Kingdom.
15 (link) The analysis herein includes 1080 consecutive cases of TIA/ischemic stroke recruited from November 1, 2004, to September 30, 2014, who had a cerebral magnetic resonance imaging (MRI). The imaging protocol of OXVASC has been described in detail elsewhere.
16 (link) Briefly, from April 1, 2002, to March 31, 2010 (phase 1), MRI and magnetic resonance angiography were done in selected patients when clinically indicated. From April 1, 2010, onward (phase 2), brain MRI and magnetic resonance angiography of intra- and extracranial vessels became the first-line imaging methods.
16 (link) A further 1076 consecutive patients who were predominantly Chinese with a diagnosis of acute ischemic stroke who received an MRI scan and magnetic resonance angiography of the intra- and extracranial blood vessels at the HKU MRI Unit were recruited during March 1, 2008, to September 30, 2014.
We collected demographic data, atherosclerotic risk factors, and details of hospitalization of index event during face-to-face interview and cross-referenced these with primary care records and hospital records in both cohorts. Cause of TIA/ischemic stroke was classified according to the modified Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria.
17 (link)Details of scan parameters are documented in Table I in the
online-only Data Supplement. Two neuroradiologists (H.K.F.M. and W.K.) supervised the interpretation of the MRI images. PVSs were defined as small (<3 mm) punctate (if perpendicular to the plane of scan) or linear (if longitudinal to the plane of scan) hyperintensities on T2 images in the BG or CS based on a previously validated scale.
18 (link) In patients with asymmetrical number of PVSs, the side with the higher number of PVSs was counted.
18 (link) Burden of PVSs was then stratified into 3 groups: <11, 11 to 20, and >20 (frequent–severe).
18 (link) Definitions of subcortical and periventricular WMH, microbleeds, and lacunes are provided in the
online-only Data Supplement. The intrarater
κ for burden of PVS (<11, 11–20, and >20) was 0.86 (BG) and 0.84 (CS) in OXVASC and 0.86 (BG) and 0.72 (CS) in HKU (50 scans in each center). Seventy-five MRI scans from HKU were cross-interpreted by investigators in OXVASC with an interrater
κ of 0.64 for both BG and CS-PVSs.
All patients in OXVASC were followed up regularly by a research nurse or physician after 1, 3, 6, 12, 24, 60, and 120 months after the index event. Patients recruited from HKU were followed up by a clinician every 3 to 6 months, or more frequently if clinically indicated. All patients were assessed for recurrent stroke (ischemic and hemorrhagic) and death (vascular and nonvascular; see definitions in the
online-only Data Supplement). Where needed, details of clinical outcomes were supplemented by electronic or paper medical records from individual primary care practices, hospitals, and the Deaths General Register Office.
Patients gave written informed consent after an event or assent was obtained from relatives for patients who were unable to provide consent. Both cohorts were approved by the local research ethics committee.