Glycated Hemoglobin is Associated with Glycemic Control and 6-Month Neurologic Outcome in Cardiac Arrest Survivors Undergoing Therapeutic Hypothermia.

Abstract

Glucose control status after cardiac arrest depending on chronic glycemic status and the association between chronic glycemic status and outcome in cardiac arrest survivors are not well known. We investigated the association between glycated hemoglobin (HbA1c) and 6-month neurologic outcome in cardiac arrest survivors undergoing therapeutic hypothermia (TH) and whether mean glucose, area under curve (AUC) of glucose during TH, and neuron-specific enolase (NSE) are different between normal and high HbA1c groups. This retrospective single-center study included adult comatose cardiac arrest survivors who underwent TH from September 2011 to December 2017. HbA1c and glucose were measured after return of spontaneous circulation (ROSC), and normal or high HbA1c was defined using cutoff value of 6.4% of HbA1c. Blood glucose was measured at least every 4h and treated with a written protocol to maintain the range of 80-200mg/dL. Hypoglycemia and hyperglycemia were defined with glucose < 70 or > 180mg/dL. Mean glucose during induction and rewarming phase and AUC of glucose during every 6h of maintenance were calculated, and NSE at 48h after cardiac arrest was recorded. The primary outcome was unfavorable neurologic outcome, defined as Glasgow Pittsburgh Cerebral Performance Category scale 3-5 at 6months after cardiac arrest. Of 384 included patients, 81 (21.1%) had high HbA1c and 247 (64.3%) had an unfavorable neurologic outcome. Patients with high HbA1c were more common in the unfavorable group than in favorable group (27.5% vs 9.5%, p < 0.001), and the unfavorable group had significantly higher HbA1c level (5.8% [5.4-6.8%] vs 5.6% [5.3-6.0%], p = 0.007). HbA1c level was independently associated with worse neurologic outcome (odds ratio 1.414; 95% confidence interval 1.051-1.903). High HbA1c group had higher glucose after ROSC, glucose AUC during maintenance, and rewarming phase than normal HbA1c group. High HbA1c group had significantly higher incidence of hyperglycemia throughout the TH, while normal HbA1c group had significantly higher incidence of normoglycemia. However, no glucose parameter remained as an independent predictor of neurologic outcome after adjustment, irrespective of HbA1c level. NSE showed good prognostic performance (area under curve 0.892; cutoff value 26.3ng/mL). Although NSE level was not different between HbA1c groups, high HbA1c group had higher proportion of patient having NSE over cutoff. Higher HbA1c was independently associated with unfavorable neurologic outcome. Glycemic status during TH was different between normal and high HbA1c groups.