Glucose in acute ischemic stroke: Effects and options for monitoring and treatment
Clemens Kersten is a PhD student in the department Clinical Neurophysiology. (Co)Promotors are prof.dr. J. Hofmeijer from the faculty Science & Technology from the University of Twente, dr. M.H. den Hertog from the Isala Klinieken Zwolle and dr. A.A.M. Zandbergen from UMC Erasmus University Rotterdam.
Ischemic stroke is a major cause of disability in developed countries. The only treatment to improve outcome consists of rapid recanalization by means of intravenous thrombolysis (IVT) or endovascular therapy (EVT). Hyperglycemia on admission is commonly observed in acute ischemic stroke patients and has been associated with unfavorable outcomes. Several studies have investigated the effects of actively lowering glucose in patients with acute ischemic stroke and hyperglycemia. These studies, however, have not shown any significant effect on infarct size or functional outcome. This lack of efficacy is probably related to insufficiently accurate glucose monitoring and insufficiently precise glucose-lowering.
The aims of this thesis were to study the relation between hyperglycemia in the acute phase and outcomes of patients with acute ischemic stroke treated with EVT, to develop a more accurate approach for glucose monitoring during acute ischemic stroke, and to explore alternatives to insulin as a glucose-lowering therapy in these patients.
Chapter 1 presents a general introduction and outlines the background and rationale for
this thesis.
Chapter 2 presents the association between hyperglycemia and computed tomographic perfusion (CTP) derived parameters in stroke patients who underwent EVT within six hours of symptom onset. We included patients from the MR CLEAN trial who had available CTP data and admission serum glucose levels. Of 173 patients, 20% were hyperglycemic (i.e., blood glucose >7.8) on admission.
Median core volume and median mismatch ratio (ischemic core divided by the total volume of hypoperfused tissue) on admission CTP were higher in the hyperglycemic group as compared to the normoglycemic group. Glucose levels between 5.0 mmol/L and 7.0 mmol/L were associated with the most favorable mismatch ratio.
In chapter 3 we assessed the association between hyperglycemia and functional outcome in patients with acute ischemic stroke of the anterior circulation treated with EVT between six and 24 hours after symptom onset. The data were derived from 464 patients in the MR CLEAN LATE trial. Median admission glucose level was 7.0 mmol/L and 32% were hyperglycemic (i.e., blood glucose >7.8) on admission. In this treatment window, elevated glucose levels were associated with poor recanalization, unfavorable functional outcome, and higher mortality. However, no significant treatment effect modification of EVT by admission glucose level or hyperglycemia was observed. Based on the results of chapters 2 and 3, it remains unclear, whether hyperglycemia directly causes or contributes to unfavorable outcomes, or if it is merely an epiphenomenon of other factors that lead to poor outcomes, such as large infarct size or underlying comorbidities.
Chapter 4 presents a study on effects of glucose level on recovery of cultured neurons after hypoxia under experimental in vitro conditions. The study described differences in neuronal network activity and levels of apoptosis in neuronal cultures during and after exposure to hypoxia at various glucose concentrations. The cultures were exposed to glucose concentrations of 5.0 mmol/L, 7.0 mmol/L, 9.0 mmol/L, or 12.0 mmol/L. At a glucose concentration of 12.0 mmol/L, electrophysiological measurements by multi-electrode arrays showed a more rapid reduction of neuronal activity compared to normoglycemia (i.e., 5.0 mmol/L). Additionally, significantly more apoptotic cells were observed in neuronal cultures exposed to 24 hours of hypoxia at a glucose concentration of 12 mmol/L than cultures exposed to glucose concentration of 5.0 mmol/L. These results showed that hyperglycemia during hypoxia likely has adverse effects at the cellular level. Furthermore, cultured neuronal networks on MEAs or coverslips provide a valuable platform for studying the effects of hyperglycemia on hypoxic neuronal network injury and recovery.
Chapter 5 shows the results of a pilot study assessing the feasibility and accuracy of continuous glucose monitoring (CGM) in patients with acute ischemic stroke treated with EVT within 24 hours of symptom onset. CGM was performed using a Freestyle Libre Flash 2 device which was implanted on arrival at the emergency department. Twenty out of twenty-three patients completed 24 hours monitoring (87%), and 100% of CGM data were within zones that previously showed to be accurate in representing glucose levels compared to capillary and venous plasma reference measurements. Based on these results, we conclude that CGM is a promising method to monitor the evolution of glucose levels during the acute phase of stroke. Additionally, it could be used to strictly monitor glucose levels during future glucose-lowering intervention studies.
Chapter 6 reports the results of a prospective cohort study conducted at two large neurovascular intervention centers examining the evolution of glucose levels in the first 24 hours after admission using continuous glucose monitoring in patients with anterior circulation large vessel occlusion ischemic stroke who underwent endovascular therapy.
We included 102 patients with 24 hours of data collection. Median onset-to-implementation time was 4 hours. Thirty-five out of 102 patients (34%) were hyperglycemic on admission (i.e., glucose >7.8 mmol/L) and 17 out of 35 patients who were hyperglycemic on admission had known diabetes (53%). Mean glucose levels ranged between 6.9 mmol/L and 8.0 mmol/L. Patients who were hyperglycemic on admission as well as those with known diabetes mellitus had persistent hyperglycemia throughout the monitoring period. In addition, successful recanalization, older age and higher NIHSS scores on admission were associated with longer periods of hyperglycemia. We conclude that CGM enables strict monitoring of glucose levels and shows promise as a valuable tool for glucose-lowering intervention studies.
Chapter 7 describes the potential of metformin as an alternative to insulin in a glucose-lowering intervention. In this retrospective cohort study, the effect of metformin on stroke severity and functional outcome after acute ischemic stroke was assessed in patients with type 2 diabetes mellitus. Patients were classified as having type 2 diabetes mellitus if they were using glucose-lowering drugs and/or when type 2 diabetes mellitus was mentioned in their medical history. A total of 937 patients were included of whom 731 (78%) used one or more glucose-lowering drugs. Median glucose on admission was 9.4 mmol/L (IQR 7,6- 11,6 mmol/L). Six hundred seventy eight (74%) patients were hyperglycemic on admission. We found that pre-stroke metformin use was associated with less severe stroke on admission and favorable outcome after 3 months. Further studies are needed to investigate whether metformin is suitable as neuroprotective drug in the acute phase of ischemic stroke.
Chapter 8 provides a general discussion of the findings in this thesis and highlights the clinical and scientific implications of the studies.
