Abnormalities on diffusion weighted magnetic resonance imaging performed several weeks after a minor stroke or transient ischaemic attack
- 1Stroke Prevention Research Unit, Department of Clinical Neurology, Radcliffe Infirmary, Oxford, UK
- 2Department of Neurology, Stoke Mandeville Hospital NHS Trust, Aylesbury, UK and Department of Neurology, Radcliffe Infirmary
- 3Department of Radiology, Stoke Mandeville Hospital NHS Trust
- 4Department of Neuroradiology, Radcliffe Infirmary
- Correspondence to: Dr P Rothwell, Stroke Prevention Research Unit, Department of Clinical Neurology, Radcliffe Infirmary, Woodstock Road, Oxford OX2 6HE, UK;
- Received 15 August 2002
- Accepted 15 January 2003
Objectives: Diffusion weighted brain imaging (DWI) is used in acute stroke, and also shows an acute ischaemic lesion in most transient ischamic attack (TIA) patients scanned acutely. However, it may also be useful in identifying subacute ischaemic lesions in patients with minor stroke or TIA who present several weeks after symptom onset. This study investigated the sensitivity and the observer reproducibility of DWI in cerebral TIA and minor ischaemic stroke patients scanned more than two weeks after the last symptomatic event.
Methods: Consecutive patients underwent magnetic resonance imaging (T2, DWI, ADC). The presence of clinically appropriate lesions was assessed by two independent observers, and related to the type of presenting event, the NIH score, persistence of symptoms and signs, and the time since the presenting event.
Results: 101 patients (53 men) were scanned at a median time of 21 days (IQR=17–28) after symptom onset. Reproducibility of the assessment of DWI abnormalities was high: interobserver agreement =97% (κ=0.94, p<0.0001); intraobserver agreement =94% (κ=0.88, p<0.0001). DWI showed a clinically appropriate ischaemic lesion in 29 of 51 (57%) minor stroke patients, and in 7 of 50 (14%) TIA patients. The independent predictors of a positive DWI scan were presentation with minor stroke versus TIA (p=0.009) and increasing NIH score (p=0.009), but there was no difference between patients presenting 2–4 weeks compared with >4 weeks after symptom onset. In minor stroke patients, the presence of a clinically appropriate lesion was associated with persistent symptoms (63% versus 36%; p=0.12) and signs (64% versus 33%, p=0.06) at the time of scanning.
Conclusions: DWI shows a clinically appropriate ischaemic lesion in more than half of minor stroke patients presenting more than two weeks after the symptomatic event, but only in a small proportion of patients with TIA. The persistence of lesions on DWI is closely related to markers of severity of the ischaemic event. These results justify larger studies of the clinical usefulness of DWI in subacute minor stroke.