NEW ORLEANS—Diffusion-weighted imaging (DWI) markedly improves stroke evaluation by detecting brain lesions that conventional magnetic resonance imaging (MRI) and computed tomography (CT) are likely to miss. That, in turn, leads to more appropriate stroke treatment, according to experts at the 25th International Stroke Conference. Among the experts were David C. Tong, MD, who shared his experience with 39 stroke patients diagnosed with DWI; Candice J. Perkins, MD, who discussed the value of DWI and perfusion weighted imaging (PWI) in 118 unselected stroke patients; and Alison E. Baird, MD, who presented data showing the prognostic value of DWI in acute ischemic stroke.

For the most part, DWI, which provides real-time images of the brain during a stroke by scanning water molecule displacement, has been a research tool. Through new hardware acquisitions or software upgrades, however, the advanced MRI technology is gradually reaching clinical practice. Research such as that presented by Drs. Tong, Perkins, and Baird offers early evidence of DWI's usefulness in the clinical setting.

INCREASED ACCURACY

The patients in Dr. Tong's study were first diagnosed with traditional techniques, including history and physical, CT, and MRI. "The treating neurologists would then do a DWI and see if it changed their diagnostic impression," said Dr. Tong, an Assistant Professor of Neurology and Neurological Sciences at the Stanford Stroke Center, Palo Alto, California.

The study subjects, whose average age was 69, underwent DWI a mean of 28 hours after symptom onset. Their mean National Institutes of Health (NIH) Stroke Scale score was quite low—only five. "A lot of the patients came in after 24 hours and tended to have more mild symptoms," explained Dr. Tong. "Also, 10 had transient ischemic attacks and obviously had a low or zero NIH Stroke Scale score."

The diagnosis changed as a result of DWI in 17 (44%) of the patients. "The most common change was from a nonembolic to an embolic etiology," Dr. Tong reported. By imaging brain lesions more clearly than CT or conventional MRI, DWI improved diagnostic certainty in 13 (33%) of the patients. In a few cases, it showed that stroke had been misdiagnosed.

Analysis by DWI prompted treatment changes in one third of the patients. "For example, a lot of those patients [initially] thought to have a cardioembolic source ended up being anticoagulated," said Dr. Tong. The influence of DWI on diagnosis and treatment was not significantly different based on stroke classification, time to MRI, or symptom duration. It is important, he said, that treatment changes did not only occur during the acute phase of stroke-like symptoms. "This was several days—up to 72 or even 96 hours—after the patient presented with symptoms," he said, "so DWI may still be quite helpful in that situation."

A second study presented at the conference had similar findings. It showed that, among 118 patients evaluated for acute ischemic stroke, DWI detected 37 acute lesions that traditional fluid-attenuated inversion recovery (FLAIR) missed. Among the study subjects, 79 had acute strokes, 29 had old strokes, and 10 were normal.

The FLAIR images tended to be more sensitive as patients' time to presentation increased. "But even after 24 hours, we still found that 35% of the patients with acute stroke still were not detected by the old technology while the diffusion technology was able to show us the new stroke lesions," said Dr. Perkins, one of the study authors, during an on-site press conference. DWI was most useful in distinguishing new lacunar infarcts, carioembolic infarcts, and small vessel disease, noted Dr. Perkins, who is a Fellow in Cerebrovascular Diseases in the Department of Neurology at University Hospital Stony Brook in New York.

FORMING A PROGNOSIS

To assess the prognostic value of DWI in acute ischemic stroke, Dr. Baird, Instructor in Neurology at Harvard Medical School, Boston, and colleagues performed a logistic regression analysis using data on 67 patients from the stroke unit at Beth Israel Deaconess Medical Center in Boston. These patients, whose mean age was 69.7, had undergone DWI scanning within 48 hours of ischemic stroke onset.

The researchers analyzed a variety of clinical outcome predictors: age, sex, time of MRI after symptom onset, history of hypertension and heart disease, NIH Stroke Scale score, DWI lesion volume, and whether the patient had participated in trials of neuroprotective agents. The main outcome variable was a positive outcome at one to three months as indicated by a Barthel score greater than or equal to 90. "The median NIH Stroke Scale score was nine," she said. Other important characteristics of the study population were a mean time of MRI after symptom onset of 10.6 hours, a mean DWI lesion volume of 20.7 mL, and a median Barthel score of 95. Neuroprotective agents were given to 26 subjects.

In a univariate analysis, the NIH Stroke Scale score and DWI lesion volume were directly correlated with outcome. Time of MRI after symptom onset was inversely related to outcome. "So patients who were studied later had a better outcome," said Dr. Baird.

All three variables remained statistically significant predictors of outcome in a multivariate analysis. "After adjustment for NIH Stroke Scale score, time of MRI after symptom onset, and age, a DWI lesion volume of less than or equal to 25 mL was associated with a 5.3 times greater chance of a better outcome," Dr. Baird emphasized.

Dr. Baird and colleagues developed a prediction model combining the three variables. The model identified patients with a positive outcome with a sensitivity of 78%, a specificity of 87%, and a positive predictive value of 89%. The model was more accurate than any individual variable (even the most accurate NIH Stroke Scale score [11] only predicted a positive outcome with a sensitivity of 76% and a specificity of 56%). "So we saw a major increase in specificity with this model," Dr. Baird concluded, stressing that DWI provided independent prognostic information during the first 48 hours of ischemic stroke.

NR

—Timothy Begany
Contributing Writer

Suggested Reading
Lovblad KO, Baird AE, Schlaug G, et al. Ischemic lesion volumes in acute stroke by diffusion-weighted magnetic resonance imaging correlate with clinical outcome. Ann Neurol. 1997;42:164-170.
Tong DC, Albers GW. Diffusion and perfusion magnetic resonance imaging for the evaluation of acute stroke: potential use in guiding thrombolytic therapy. Curr Opin Neurol. 2000;13:45-50.

Return to table of contents