Tool May Help Select Patients for Endovascular Stroke Treatment

When uncertainty exists about whether an individual patient suffering from acute ischemic stroke will benefit from endovascular therapy, a new tool combining clinical and radiographic characteristics may help physicians make the right call, investigators say.

The decision tool, which incorporates 11 baseline factors and can be used online, had “moderate discriminative ability” for predicting a good functional outcome in two cohorts of patients from recent randomized trials, researchers led by Esmee Venema, MD (Erasmus MC University Medical Center Rotterdam, the Netherlands), report in a study recently published online in the BMJ.

“It’s very important when you have to decide if a patient is eligible for the treatment that you consider multiple characteristics of the patients together, and you should not withhold treatment from a patient because he is in a certain subgroup, like [having a] poor collaterals score or age above 80,” Venema told TCTMD.

She added that even though further validation in other cohorts would be preferable, the tool can be used to support clinical judgment before then “since it is the best information that we have available at the moment.” Also, she said, the trial used to develop the tool—MR CLEAN—was performed in a relatively unselected patient population, making “it likely applicable in clinical practice.”

But Michael Hill, MD (University of Calgary, Canada), who was not involved in the study, expressed skepticism about the tool’s adoption for clinical decision-making. “The data inform the understanding of what are good prognostic rules and what are not,” he commented in an email to TCTMD, adding, however, that the tool “won’t be used much.”

He explained that clinical prediction rules often have a C-statistic—a measure of discriminative ability—ranging from 0.60 and 0.80 (figures ranged from 0.69 to 0.79 in this study), but that a value of 0.95 or greater is required before they become useful for supporting decision-making in practice

“Think of it like a diagnostic test. If a test for HIV had a C-statistic of 0.8, we would never use it. The current one has a C-statistic of 0.9999 and very clearly discriminates between those who have the virus and those you do not,” Hill said.

“So the current rule ends up being very useful for understanding which variables are predictive of outcome and what their relative importance is, and [is] one that will inform practice in a general way,” he continued, “but [it will] not stop treatment of patients who do not have a good prognosis. They still benefit.”

The next steps, Hill stated, involve testing the tool in a prospective study or using it to guide patient selection in a randomized trial.

Endovascular therapy for acute ischemic stroke—particularly thrombectomy using stent retrievers—has taken on greater importance in recent years after randomized trials started yielding positive results, a turnaround from prior negative trials. But not all patients will benefit from treatment, and clinicians must make decisions for individual patients based on their characteristics.

Venema and colleagues set out to create a tool to assist in such treatment choices using data from MR CLEAN, which demonstrated the superiority of intra-arterial treatment plus usual care over usual care alone, for development. They used data from IMS III, which failed to show an advantage for endovascular therapy, for validation.

The final multivariable model included the following 11 baseline clinical and radiological characteristics:

• Age
• National Institutes of Health Stroke Scale score
• Systolic blood pressure
• Treatment with IV tissue plasminogen activator
• History of ischemic stroke
• Diabetes
• Prestroke modified Rankin Scale (mRS) score
• Alberta Stroke Program Early CT Score (ASPECTS)
• Location of occlusion
• Collaterals score
• Time from stroke onset to groin puncture

The internally validated C-statistic for predicting a good functional outcome (mRS score 0-2 at 90 days) was 0.79, with an externally validated value of 0.73, indicating moderate discriminative ability.

The authors acknowledge that the tool is not perfect, but say that “the currently developed model is the first to predict the effect of intra-arterial treatment for individual patients on arrival at the emergency department. When compared with other models used in neurovascular practice, HAS-BLED (C-statistic 0.65) and CHA2DS2-VASc (0.61), it performs accurately.”

In addition, the study shows that treatment decisions should not be made based on single characteristics, Venema and colleagues say.

“Some patients belonging to one of the subgroups that are considered as having no benefit of intra-arterial treatment, such as poor collaterals or low ASPECTS, may still benefit from intra-arterial treatment substantially if other characteristics are favorable,” they write. “This emphasizes the importance of making personalized treatment decisions, instead of using average treatment effects, and shows the need for combining multiple clinical and radiological baseline characteristics instead of withholding treatment based on one characteristic.”