Brain and Carotid Imaging Findings Improve ABCD2 Assessment of Stroke Risk After TIA

Megan Brooks

November 02, 2010

November 2, 2010 — The ABCD3-I score, which adds brain and carotid imaging findings to the standard ABCD2 clinical prediction score, can markedly improve risk stratification after transient ischemic attack (TIA), confirm results of a multinational observational study.

The ABCD3-I score, with a range of 0 to 13 points, "has shown external validity, good calibration, and risk reclassification," first author Aine Merwick, MB, MRCPI, from Mater Misericordiae University Hospital, Dublin, Ireland, noted in an email to Medscape Medical News.

"This study demonstrates that carotid stenosis, earlier TIA preceding the TIA prompting medical attention, and diffusion-weighted imaging (DWI) lesions are independent predictors of early stroke risk in TIA patients," Dr. Merwick added.

TIA is associated with high risk for early stroke. It is now recommended that DWI and carotid imaging be performed "urgently as part of the standard assessment of patients with TIA," the study team notes in their report. "We expect that, as these guidelines are increasingly adopted, the ABCD3-I score will be of substantial practical use."

Their report is published in the November issue of Lancet Neurology.

Stroke Prediction Tools

The validated and widely used ABCD2 score incorporates 5 clinical characteristics at the time of presentation to risk-stratify TIA patients for early stroke. They are age 60 years or older (1 point), blood pressure 140/90 mm Hg or higher (1 point), clinical weakness (2 points) or speech impairment (1 point), duration of symptoms for 60 minutes or longer (2 points) or for 10 to 59 minutes (1 point), and diabetes (1 point).

Dr. Merwick and colleagues developed 2 new versions of this score (ABCD3 and ABCD3-I). The ABCD3 score adds 2 points for a history of previous TIA within 7 days, and the ABCD3-I score adds 2 points for a new magnetic resonance imaging lesion that shows DWI positivity, and 2 points for internal carotid stenosis of at least 50%.

The team tested the scores in 3886 patients with clinically defined TIA who were seen in secondary care. There were 2654 patients in the derivation sample and 1232 in the validation sample.

In the derivation sample, the C statistic, which indicates discrimination better than chance at more than 0.5, was higher for the ABCD3 score than the ABCD2 score for 7-day stroke prediction, with a further increase for the ABCD3-I score, the authors report.

In the validation sample, both scores predicted early stroke at 7, 28, and 90 days, and the ABCD3-I score improved discrimination of patients at high stroke risk at 28 and 90 days.

However, in the validation sample, discrimination and net reclassification of patients with early stroke were similar with the ABCD3 and ABCD2. Therefore, "use of the ABCD3 cannot be recommended without further validation," the authors say.

Table. C Statistics for Early Stroke Discrimination With New Scales vs ABCD2

Analysis Day 7 (95% Confidence Interval) P Value Day 28 (95% Confidence Interval) P Value Day 90 (95% Confidence Interval) P Value
ABCD2 0.71 (0.64 - 0.77) ** 0.71 (0.65 - 0.77) ** 0.69 (0.64 - 0.75) **
ABCD3 0.80 (0.74 - 0.85) .01 0.79 (0.73 - 0.85) 0.02 0.77 (0.72 - 0.82) 0.02
ABCD3-I 0.92 (0.79 - 0.99) .001 0.85 (0.71 - 0.96) 0.03 0.79 (0.66 - 0.90) 0.07
ABCD2 0.63 (0.56 - 0.69) ** 0.61 (0.56 - 0.67) ** 0.60 (0.55 - 0.65) **
ABCD3 0.64 (0.58 - 0.71) .35 0.62 (0.56 - 0.68) 0.38 0.60 (0.55 - 0.65) 0.44
ABCD3-I 0.71 (0.63 - 0.78) .05 0.70 (0.63 - 0.77) 0.03 0.71 (0.64 - 0.77) 0.01

P values compare scores with ABCD2.

According to the investigators, patients with symptomatic carotid stenosis, recent earlier TIA, and DWI abnormality were at 3 to 7 times higher risk for stroke, after adjusting for other known risk factors for early recurrence. "The ABCD3-I score allows risk stratification of TIA patients with a high degree of confidence," Dr. Merwick told Medscape Medical News.

"For most patients with suspected TIA assessed in secondary care, the ABCD3-I score might standardise and refine risk stratification, leading to improved clinical decision making with regard to the need for costly medical investigations, referral to specialist services, and admission to hospital," the investigators add in their report.

In a written commentary accompanying the publication, S. Claiborne Johnston, MD, PhD, director of the Clinical and Translational Science Institute, University of California–San Francisco, says this study "convincingly" shows that the ABCD3-I is "superior" to the ABCD2 score.

"The bottom line," Dr. Johnston added in an email to Medscape Medical News, is that "brain and carotid imaging are important to prognosis after TIA regardless of the ABCD2 score. Patients with new brain infarction or with a high-grade internal carotid artery stenosis should be considered [at] high risk regardless of the ABCD2 score."

Confusion and Uncertainty for Now?

The problem, Dr. Johnston notes, is that while this study was being prepared for publication, another study was published (Stroke. 2010;41:1907-1913) that proposed a different prediction score for stroke risk after TIA. This score, the ABCD2-I score, adds 3 points to the ABCD2 score for acute infarction on brain computed tomography or magnetic resonance imaging, and it was derived and validated in the same set of cohorts.

"Unfortunately, the timing of these studies creates a new problem: we now have 2 competing scores incorporating imaging into ABCD2, and no data to tell us which score is superior. We will have to wait for a head-to-head comparison and, for now, confusion and uncertainty will probably reduce the use of both scores," Dr. Johnston writes.

However, he adds, "it is difficult to imagine that carotid stenosis will not remain an important component in a score that includes imaging data, but it is to be hoped that the comparative study will emerge quickly so we can avoid an alphabet soup of potential scores."

Reached for outside comment on the new study, Philip B. Gorelick, MD, MPH, director of the Center for Stroke Research at the University of Illinois College of Medicine, in Chicago, said: "The take-home message for clinicians is that recent TIA is a medical emergency that must be thoughtfully diagnosed and treated.

"The addition of the ABCD3-I scoring system as a predictor of stroke after recent TIA," he wrote in an email, "provides evidence of the importance of key clinical features plus early [magnetic resonance imaging] DWI study and diagnostic study to determine if there is high-grade carotid artery stenosis. These features taken in total substantially heighten the risk prediction of stroke after TIA.

"Clinicians who are diagnosing and treating TIA patients should be aware of these relationships and the need to respond urgently and appropriately in cases of recent TIA," Dr. Gorelick concluded.

The study was supported by the Health Research Board of Ireland, Irish Heart Foundation, and Irish National Lottery. The study investigators have various financial disclosures, all listed with the original article. Dr. Johnston and Dr. Gorelick have disclosed no relevant financial relationships.

Lancet Neurol. 2010;9:1039-1040,1060-1069. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.