To study the relationship between admission hemoglobin A1c (HbA1c) and outcomes in patients undergoing endovascular thrombectomy (EVT) for acute ischemic stroke (AIS).

Multiple randomized clinical trials have established EVT as the standard of care for large vessel occlusion (LVO) stroke. Recent literature suggests that elevated baseline HbA1c is associated with poor functional outcomes in AIS with or without IV alteplase; however, the effect of HbA1c on outcomes in patients undergoing EVT is unknown.

A retrospective cohort of 348 consecutive patients undergoing EVT at a large comprehensive stroke center was identified from February 2015 to September 2018. Sociodemographic, in-hospital treatment, complications and functional outcomes were collected through retrospective chart review. Multivariate regression model was used to adjust for confounding variables identified, using Chi square and Fisher exact test as appropriate.

Patient mean age was 71±14 years, 50.4% were female, 24% Black and 43% Hispanic. Median modified Rankin Scale (mRS) at baseline was 0 (IQR 0-1), median admission and 24 hour-post EVT NIHSS was 16 (IQR 11-20) and 12 (IQR 5-19), respectively. Overall, 33.6% of patients had a history of diabetes mellitus and mean HbA1C was 6.2±1.4. In multivariate analysis, higher HbA1c on admission was independently associated with myocardial infarction (MI) (OR 2.7, p 0.008) during hospitalization after EVT. There was a trend towards significance for higher admission NIHSS (OR 1.9, p 0.06), and acute kidney injury (AKI) (OR 1.7, p 0.09) for patients with elevated HbA1C. There was no significant difference in good functional outcomes (mRS 0-3) or inpatient mortality on multivariate regression.

Elevated hemoglobin A1C on admission in patients with LVO stroke requiring EVT is independently associated with higher incidence of MI during hospitalization and may be associated with more severe strokes. However, there was no significant difference in functional outcomes and in-hospital mortality based on HbA1C levels.