Risk of recurrence following acute cardioembolic stroke according to lesion size and NIHSS on admission

Introduction: This study evaluated the risk of stroke recurrence and bleeding in patients with cardioembolic ischemic stroke, according to lesion size and the NIHSS score on admission. Methods: Stroke severity was categorized as minor, NIHSS 0–4; moderate, 5–15; severe, >15, while infarct size was categorized as small, medium, or large. The primary outcome was the composite at 90 days of either ischemic (stroke or systemic embolism) or hemorrhagic events (symptomatic intracranial or severe extracranial bleeding). Results: The cohort for this analysis comprised 2824 patients. At 90 days, 230 outcome events (8.1 %) were recorded. Regarding NIHSS, with the NIHSS 0–4 group as the reference group, on multivariable analysis, NIHSS 5–15 and NIHSS >15 were significantly associated with combined and hemorrhagic outcomes. NIHSS >15 group resulted also significantly associated with ischemic outcome. Concerning lesion size, with the small lesion group as the reference group, on multivariable analysis, medium and large lesions were significantly associated with all outcomes. The ROC graphs suggest that lesion size better-predicted outcome events, compared to NIHSS even on the DeLong test the differences were not statistically significant. In fact, regarding combined outcome, the AUC for lesion size resulted being 0.618 compared to 0.599 for NIHSS. Furthermore, regarding hemorrhagic outcome, the AUC for lesion size was recorded to be 0.649, compared to 0.601. Finally, for ischemic outcome, the AUCs resulted being 0.575 and 0.577, for lesion size and NIHSS, respectively. Conclusion: Lesion size, compared to NIHSS, slightly better-predicted study outcome events in patients with cardioembolic ischemic stroke.