Error in Figure 1.

Abstract

mography angiography studies, we found no association between thrombus length measured on T2* magnetic resonance imaging and collateral grade.4 Thus, we believe there are insufficient data to propose thrombus length as a surrogate marker of collaterals, but further work using alternative imaging modalities is warranted to clarify the respective roles of thrombus length, collaterals, and angioarchitecture for ENDi prediction. Besides in situ thrombus extension, other factors might explain the association between ENDi and more proximal occlusion and longer thrombus, such as reembolization from a proximal source (eg, cervical carotid occlusion5) and blood pressure drops or hyperglycemia or hypoglycemia, which worsen neuronal status in hypoperfused tissue downstream of a persistent large-vessel occlusion. Given the lack of benefit derived from mechanical thrombectomy found in most observational studies in unselected populations with large-vessel occlusion–associated minor stroke, considering additional thrombectomy in subgroups at high risk of ENDi appears a straightforward approach to improve functional outcomes. Our prediction score identified two-thirds of patients at low risk of ENDi (scores of 0 or 1; ENDi risk <7%), in whom the benefit-harm ratio of additional thrombectomy might be unfavorable, and the remaining one-third at high risk of ENDi (scores >1; ENDi risk >20%), for whom a benefit of additional thrombectomy can be reasonably expected. In line with these results, another observational study from the cohort in our study1 suggests that thrombectomy added onto intravenous thrombolysis, compared with intravenous thrombolysis alone, may be beneficial in proximal largevessel occlusion, but deleterious, including triggering higher risks of symptomatic intracranial hemorrhage, in more distal occlusions (ie, occlusion categories with higher and lower risks of ENDi, respectively).6 Besides preventing ENDi, we agree that another appeal of additional thrombectomy would be to improve outcome of patients with minor yet disabling deficits. However, this remains a somewhat subjective judgment, and whether the benefit of thrombectomy is modified according to the disabling vs nondisabling nature of symptoms is still unsettled. Pierre Seners, MD, PhD Jean-Claude Baron, MD, ScD Guillaume Turc, MD, PhD