Correspondence Rush University Medical Center, Department of Neurological Sciences, Section of Cerebrovascular Disease and Critical Care, 1725 West Harrison Street, Suite 1121, Chicago, IL 60612, USA

Email shyam_prabhakaran@rush.edu

Original article

Boulanger JM et al. (2007) Diffusion-weighted imaging-negative patients with transient ischemic attack are at risk of recurrent transient events. Stroke 38: 2367–2369   PubMed

Synopsis

Background

A major clinical issue in patients with transient ischemic attack (TIA) is the identification of those who are at increased risk of early recurrent TIA or stroke.

Objective

To assess whether diffusion-weighted imaging (DWI) is a useful method for risk stratification in patients with TIA.

Design and intervention

This prospective study enrolled patients with a motor hemiparesis or aphasic TIA that lasted for at least 5 min and had occurred within the previous 12 h. The enrollment period was May 2002 to June 2004. Patients who received thrombolytic therapy were excluded. All patients underwent 3-Tesla MRI within 24 h of symptom onset, and the scans were reviewed by a neuroradiologist who was unaware of the clinical data with the exception of symptom side. The occurrence of stroke was evaluated every 6 months, starting 6 months after the index event, by application of the Questionnaire for Verifying Stroke-Free Status. Patients with answers suggestive of stroke were examined by a stroke neurologist. The participants were followed up for a median of 13 months (range 1 day to 24 months). The 1-year risks of stroke and TIA were estimated by use of Kaplan–Meier curves. TIA was defined as rapidly developing clinical signs of focal or global disturbance lasting less than 24 h, without any apparent nonvascular cause. Stroke was defined as a functional neurological deterioration of vascular origin or a new sudden focal neurological deficit of vascular origin lasting 24 h or more.

Outcome measures

The primary end point was the first recurrent TIA within 1 year of study entry. The 1-year risk of stroke was also determined.

Results

A total of 85 patients were included in the final analysis. Of these, 35 (41.2%) had lesions on DWI. The mean SD time from symptom onset to MRI was 11.4 6.9 h for patients with lesions on DWI and 12.6 6.9 h for patients with a normal scan. The presence of DWI lesions was associated with large-artery disease (P = 0.005 vs no large-artery disease), but not with other factors such as TIA duration, time lag between symptom onset and MRI scan, or previous history of stroke or TIA. Within 1 year of study entry, 12 (14.1%) patients had a subsequent TIA and 4 (4.7%) had a subsequent stroke. TIA recurrence was more frequent among patients with a normal DWI scan at baseline than among patients with DWI lesions (24.7% vs 5.9%; odds ratio 4.6, 95% CI 1.3–12.5; P <0.05). Patients with a normal scan were, however, less likely to have a subsequent stroke (stroke frequency 2.1% vs 9.1%; odds ratio 0.24, 95% CI 0.03–1.8; P = 0.19). In most cases (n = 10, 83%) TIA recurrence was observed in the same vascular territory as the index event.

Conclusion

In patients with a TIA, the absence of a DWI lesion is associated with a high risk for TIA recurrence, whereas a positive DWI scan indicates an increased risk for future stroke.

Commentary

TIA is a predictor of ischemic stroke, and the identification of patients with TIA who are at high risk of subsequent strokes is critical to stroke prevention. Besides its central role in the diagnosis of ischemic stroke, DWI has proven useful in stratifying subsequent stroke risk following TIA. In patients with TIA who have lesions on DWI, the risk of stroke is estimated at 8.3–14.8% within the first week following the TIA, increasing to as much as 32.6% at 90 days. By comparison, there seems to be a much lower risk of recurrent stroke following DWI-negative TIA (only 2.0% during hospitalization and 4.3% at 90 days).^{1, 2, 3}

In their longitudinal study of 85 patients with motor or aphasic TIA, Boulanger et al. have expanded on these findings. Patients with DWI-negative TIA were over four times less likely to have a subsequent stroke but over four times more likely to have a recurrent TIA at 1 year than were those with DWI-positive TIA. These data confirm that acute DWI is a simple tool for stratifying stroke risk after TIA. A positive DWI scan helps to confirm an ischemic mechanism and identify a subgroup of patients with TIA who are at high risk of subsequent stroke and in whom more-aggressive monitoring and treatments seem warranted.

The low risk of stroke but high risk of recurrent TIA following DWI-negative TIA could be explained by the possible inclusion of a heterogeneous group of patients: first, patients with transient brain ischemia of short duration, with modest reductions in cerebral blood flow, with excellent endogenous thrombolysis or with collateral blood supply that prevented cytotoxic injury from occurring; and second, patients with nonischemic conditions (or mimics) such as migraine or seizure that were misdiagnosed as TIA.

The latter point highlights the difficulty in making a diagnosis of TIA on the basis of symptoms alone. Most clinicians do not consider low-probability TIA symptoms—such as hemisensory loss or visual changes—to be ischemic in origin if they occur in isolation. Conversely, acute motor and language symptoms are considered more likely to be caused by brain ischemia. Surprisingly, the outcomes following DWI-negative motor or aphasic TIA in the current study are similar to those following isolated sensory TIA⁴: low risk of stroke and high risk of recurrent transient spells. These findings suggest, therefore, that the validity of the clinical diagnosis of TIA—even among those with high-probability TIA symptoms—might be suspect.

This paper begs another question: does the absence of DWI abnormality always predict a benign outcome? If one assumes that it is possible for patients with real transient brain ischemia (i.e. right hemiparesis lasting 30 min and 70% stenosis of the left carotid artery) to have a normal DWI scan, then it would be wrong to assume that the risk of subsequent stroke is uniformly negligible. Some patients might be at high risk of subsequent stroke on the basis of other comorbidities, such as symptomatic carotid stenosis or atrial fibrillation, independent of the DWI results.

Thus, careful bedside evaluation and good clinical judgment, especially when assessing DWI-negative patients, is still invaluable. Given these complexities in the diagnosis and management of TIA, I express some concern that any 'cookbook' utilization of DWI to triage possible TIA patients may supplant clinical judgment; such an approach might overlook some true DWI-negative TIA patients at high risk for stroke and misdiagnose others who do not have a TIA at all.

Acknowledgments

The synopsis was written by Martina Habeck, Associate Editor, Nature Clinical Practice.

References

1. Ay H et al. (2005) Transient ischemic attack with infarction: a unique syndrome?. Ann Neurol 57: 679–686 | Article | PubMed |
2. Coutts SB et al. (2005) Triaging transient ischemic attack and minor stroke patients using acute magnetic resonance imaging. Ann Neurol 57: 848–854 | Article | PubMed | ISI |
3. Purroy F et al. (2004) Higher risk of further vascular events among transient ischemic attack patients with diffusion-weighted imaging acute ischemic lesions. Stroke 35: 2313–2319 | Article | PubMed |
4. Johnston SC et al. (2003) A comparison of risk factors for recurrent TIA and stroke in patients diagnosed with TIA. Neurology 60: 280–285 | Article | PubMed |

Contact the journal about this article

Subject areas under which this article appears: Stroke