Acute Hyperglycemia in Stroke leads to increased brain lactate production and greater final infarct size.
Background: Diabetes and stress hyperglycemia occur in about 1/3 of stroke patients and are associated with worse stroke outcome. In animal stroke, MR spectroscopy (MRS) studies have established a relationship between hyperglycemia, increased brain lactate level, and larger final infarct size following reperfusion. We therefore used 1H-MRS to measure regional lactate, and MR diffusion-weighted (DWI) and perfusion-weighted imaging (PWI) to assess reperfusion of the ischemic penumbra. Methods: Forty-two hemispheric stroke patients were studied with DWI and PWI within 24 hours of symptom onset, and again at days 3 and 90. A subgroup of 24 patients had acute and day 3 MRS to determine lactate level in the ischemic region. Blood glucose was measured acutely within 1 hour of the acute scan. Results: Twenty-two patients were identified as not yet having undergone major reperfusion (acute PWI lesion > acute DWI lesion). The extent of the acute PWI lesion that did not progress to infarction was termed “penumbral salvage”. In the 22 patients who had not reperfused at time of acute scan, a correlation was found between acute blood glucose and final infarct size (r=0.62, p<0.01). Furthermore, hyperglycemia was associated with reduced penumbral salvage (r=0.75, p<0.01). Fourteen patients in the group yet to undergo reperfusion had MRS to measure lactate level in the ischemic region before reperfusion (mean 7 hours), and following reperfusion at day 3 (which may provide an indication of the amount of lactate generated during reperfusion). In these patients, acute blood glucose correlated strongly with day 3 lactate levels (r=0.85, p<0.01). Day 3 lactate level was also strongly associated with reduced penumbral salvage (r=0.90, p<0.01). Conclusions: In acute stroke, hyperglycemia may lead to increased lactate production during reperfusion, thus contributing to reperfusion injury, resulting in reduced penumbral salvage and increased final infarct size. These findings are the first reported in human stroke and help explain the worse outcome in diabetic stroke patients. Future research should evaluate the effects of glycemic control on these determinants.