Introduction: Selection of patients based on treatment-related targets has been especially important for endovascular therapy. Recently presented trial evidence has demonstrated that cases with an eligible arterial occlusion and a target PWI-DWI mismatch were able to successfully receive treatment beyond the 6 hour window.
Method: The LESION Study is a prospective cohort study of the baseline MR imaging (<24 hrs from onset and prior to treatment) of 1,092 consecutive ischemic stroke presentations with admission NIHSS ≥4. Multiple therapeutic targets were assessed. Endovascular intervention target was defined as the presence of ischemia on MRA in the ICA or MCA-M1 arterial branches. Presence of PWI-DWI mismatch was qualitatively determined by vascular neurologists. Multivariate logistic regression was used to test for association of time from onset with presence of an endovascular target, while examining the roles of presentation and imaging characteristics.
Results: Probability of an endovascular target decreases over the first 24 hours (OR=0.96; p=0.004). However, stratification or adjustment for a PWI-DWI mismatch signal and/or admission NIHSS confounds the effect of occlusion target over time. Stratification by mismatch demonstrates a consistently higher proportion of endovascular target presence in those with a concurrent target mismatch. For cases with a target mismatch, presentation within the 6 hour window is associated with just a 17% greater likelihood of endovascular target relative to those beyond 6 hours (OR=1.17; p > 0.05). In multivariate modeling: NIHSS, mismatch, and sex were all identified as independent risk factors for presence of an endovascular target (all p < 0.01), while time was not significantly associated with target presence.
Discussion: While the presence of an arterial occlusion in territories relevant to endovascular intervention appear to be related to time from onset, stratified and adjusted analyses show that this relationship is not maintained when limited to cases with and without a target mismatch identified. This finding has clinical and methodological implications for selection of patients for intervention at varying time intervals.