EVT for large ischemic stroke matches trial success in routine care

For years, the medical consensus on acute ischemic stroke was relatively rigid: if the “core” of the brain infarct—the area already dead or dying—was too large, surgeons generally stepped back.

A recent analysis published in the Annals of Translational Medicine examined whether the success seen in highly controlled clinical trials actually translates to everyday hospital practice. The findings suggest that while the benefits are there, the way we select patients for this life-altering procedure remains inconsistent and fraught with imaging gaps.

The Gap Between Trial Success and Hospital Reality

The core of the debate centers on the TENSION trial, a randomized controlled trial (RCT) that showed better 90-day functional outcomes for patients with larger infarct cores who underwent EVT. To see if these results held up in the “real world,” researchers led by Winkelmeier et al. compared TENSION trial participants against a group from the German Stroke Registry. This comparison was designed to see if the “efficacy” found in a curated study matches the “effectiveness” seen in routine care.

The results were broadly encouraging. When the researchers used propensity score matching to ensure the two groups were comparable, they found no significant difference in the distribution of 90-day modified Rankin scale (mRS) scores—a standard measure of disability—between the trial patients and those in routine practice. Essentially, the people who met the trial criteria and received the procedure in a real-world setting fared about as well as those in the controlled study.

However, the data also revealed a stark disparity in how patients are actually treated. Before the researchers “cleaned” the data to match the trial’s strict rules, the registry patients were fundamentally different: they had lower baseline neurological deficits (a median NIHSS of 17 versus 19 in the TENSION cohort) and arrived at the hospital faster. Most tellingly, only about half of the patients in the German Stroke Registry actually met the TENSION trial criteria.

CT and MRI Scans Conflict on Surgery Eligibility

The “so what” of this study isn’t just about whether the surgery works, but how we decide who gets it. Doctors typically use the Alberta Stroke Program Early Computed Tomography (ASPECTS) score to judge salvageable brain tissue. A score of 6 or higher usually clears a patient for EVT, while scores of 3 to 5—indicating a larger core—were previously seen as a reason to avoid the procedure.

The problem is that the tools used to measure this core aren’t always consistent. The analysis highlights a tension between CT scans and MRI. While diffusion-weighted MRI is more sensitive at detecting early ischemic changes, CT-based protocols are more prone to variability between different doctors reading the scans. This means a patient might be classified as having a “large core” at one hospital and a “salvageable core” at another, purely based on the imaging modality used.

Metric	TENSION Trial (Matched)	German Registry (Matched)
Median ASPECTS Score	4	4
Pre-treatment NIHSS	18	18
90-day mRS Outcome	Comparable	Comparable

Who Bears the Burden of Uncertainty?

The analysis warns that by using propensity matching to make the registry look like the trial, we might be scrubbing away the very diversity—the age, sex, and medical histories—that doctors need to understand to make better bedside decisions.

We are left with a critical gap in the map: we know the procedure can work for a curated subgroup of “large core” patients, but we still don’t know if the current framework for patient selection is being followed consistently in everyday practice, or if we are simply guessing who is “too far gone” to be saved.