Futile and Harmful Reperfusion and the Balance Between Treatment Effect and Overall Outcomes in Stroke

Over the past decades, ischemic stroke research has primarily focused on achieving rapid reperfusion. Endovascular thrombectomy has revolutionized the treatment paradigm for patients with large vessel occlusion, with recent trials showing benefit even in patients with large core at baseline. These findings have led some to advocate for reperfusion in all cases, regardless of infarct size and severity. We critically examine this line of reasoning and introduce 2 important caveats. First, in an individual patient, reperfusion does not necessarily or uniformly translate into meaningful improvement and favorable outcomes. The concept of futile reperfusion is real. As a corollary, trial results capture average effects, and individuals have a wider range of outcomes. Furthermore, results are often reported as relative rather than absolute treatment effects. As baseline prognosis worsens, the absolute likelihood of a good outcome may fall below a threshold where the intervention is no longer justified, despite a favorable relative treatment effect. Second, in a small subset of patients, reperfusion may actively worsen outcome; this is harmful reperfusion. While additional harm may seem negligible in such a high-risk population, this rationale is flawed as it encourages therapeutic actionism and violates the foundational medical ethical principle of primum non nocere. To advance patient care, we must move beyond a one-size-fits-all reperfusion model that focuses only on vessel reopening. Some patients might have infarcts that are simply too large (eg, >150 mL), ischemia that is too severe (eg, severe noncontrast computed tomography hypodensity), or comorbidities that overwhelm any potential benefit. A more nuanced approach requires a better understanding of tissue viability, perfusion physiology, and ischemic damage. This would allow for refined patient selection by leveraging advanced imaging and large-scale data sets to develop accurate models to predict treatment effect, that is, beneficial, futile, and harmful reperfusion.