Impact of Anesthesia on Thrombectomy Outcomes in Large Stroke: a TENSION Subanalysis.

Abstract

Introduction: This post-hoc subanalysis of the TENSION trial evaluated the association between anesthesia strategy and neurological outcomes in patients with large-core infarcts undergoing mechanical thrombectomy (MT).

Methods: We analyzed patients from the interventional arm of the TENSION trial who underwent MT under general anesthesia (GA) or conscious sedation (CS). Anesthesia strategy was not prespecified and was selected by the local stroke team. Key outcomes included early neurological improvement (ENI) or deterioration (END), modified Rankin Scale (mRS) at 90 days, procedural metrics, and safety outcomes. Multivariable logistic regression was performed to assess the independent association between anesthesia strategy and ENI, END, and favorable functional outcome, defined as mRS 0-3 at 90 days. Sensitivity analyses were conducted in patients with successful reperfusion.

Results: A total of 117 patients were included, of whom 50 were treated under GA and 67 under CS. No significant differences were observed between GA and CS for ENI (46.0% vs. 62.7%, p = 0.09) or END (56.0% vs. 46.3%, p = 0.30) at 24 h. Likewise, 90-day functional outcome did not differ significantly between groups, with mRS 0-3 achieved in 24.0% of GA patients and 36.0% of CS patients (p = 0.057). In adjusted analyses, anesthesia strategy was not independently associated with ENI, END, or mRS 0-3 at 90 days. Reperfusion success, first-pass recanalization, and safety outcomes were similar between groups. In the sensitivity analysis restricted to successfully reperfused patients, total procedure time was shorter in the CS group, but neurological and functional outcomes remained similar.

Conclusion: In this post-hoc subanalysis of patients with large-core infarcts undergoing MT, anesthesia strategy was not independently associated with early neurological or 90-day functional outcomes. GA and CS were associated with comparable procedural success and safety. Given the non-randomized design, limited sample size, and potential for residual confounding, further studies are needed to clarify optimal anesthesia selection in large-core stroke patients undergoing MT.