Insulin Therapy and the Injured Brain
An article published on medpage today, states that intensive insulin therapy in the ICU may starve critically injured brains and potentially lead to further damage. Mauro Oddo, M.D. of Columbia University shares that “brain energy crisis” was 23% more likely with each 1 mmol/L decrease in systemic glucose and 10% more likely with each 1 unit/L increase of insulin infusion.
Episodes of low brain glucose levels and high levels of metabolic distress markers were associated with greater mortality risk and poorer outcome among neuro-ICU patients, as reported at the Society of Critical Care Medicine meeting.
While they have identified that glucose metabolism is more important in brain-injured patients than in other critically ill patients, what to do with that knowledge is undetermined. J. Javier Provencio, M.D. of the Cleveland Clinic, states that the impact of brain energy crisis remains unclear and needs to be validated by studies that actually include outcomes.
The trial of 20 patients included individuals with subarachnoid hemorrhage, intracerebral hemorrhage, traumatic brain injury, or cerebral infarction in a neurological ICU. Medpage today reports:
The retrospective analysis was done on prospectively collected data including direct, continuous cerebral microdialysis monitoring of glucose, lactate, and pyruvate levels over a median 96-hour period.
Patients received IV insulin with the goal of tight glucose control (80 to 120 mg/dL).
Brain glucose levels increased in a linear fashion with systemic glucose levels, but there were substantial variations in individual patients over time.
Brain glucose was significantly more likely to be below the 0.7 mmol/L threshold when systemic glucose was in the tight glucose control range than when it was in the 121 to 180 mg/dL range (P<0.01).
Brain energy crisis was defined as a brain lactate-to-pyruvate ratio less than 40 and glucose less than 0.7 mmol/L. Its occurrence rose with insulin use (odds ratio 1.10 for each 1 unit/L insulin, P=0.02) and decreasing systemic glucose (OR 1.23 per 1 mmol/L serum glucose, P
Patients who died had lower brain glucose levels than those who survived (mean 0.5 versus 1.0 mmol/L, P=0.03) and tended to have higher insulin doses (mean 1.8 versus 1.3 units/L, P=0.10).
Brain energy crisis was a stronger independent predictor of mortality (adjusted OR 7.36, P=0.02) than Glasgow Coma Scale score, cerebral perfusion pressure, or intracranial pressure.
Dr Oddo concludes that insulin therapy needs to be used with caution, and that a less restrictive target for systemic glucose may be more beneficial for those critically ill neurological patients. He suggests that more research needs to be conducted to determine the optimal systemic glucose, although an intermediate range around 144 mg/dL should be safe.
Some of the participants in the trail showed low brain glucose levels even when serum levels were normal. As such, Dr Provencio cautioned that the jury is still out on glucose, and that in future studies it may be more important to measure and control brain glucose levels as opposed to measuring and controlling systemic glucose levels.