Early Evidence of Sepsis-Associated Hyperperfusion-A Study of Cerebral Blood Flow Measured With MRI Arterial Spin Labeling in Critically Ill Septic Patients and Control Subjects.

Abstract

Mechanisms underlying sepsis-associated encephalopathy remain unclear, but reduced cerebral blood flow, alone or in conjunction with altered autoregulation, is reported as a potential contributor. We compared cerebral blood flow of control subjects and vasopressor-dependent septic patients. Randomized crossover study. MRI with arterial spin labeling. Ten sedated septic patients on mechanical ventilation (four with controlled chronic hypertension) and 12 control subjects (six with controlled chronic hypertension) were enrolled. Mean ± SD ages were 61.4 ± 10.2 and 44.2 ± 12.8 years, respectively (p = 0.003). Mean Acute Physiology and Chronic Health Evaluation II score of septic patients at ICU admission was 27.7 ± 6.6. To assess the potential confounding effects of sedation and mean arterial pressure, we measured cerebral blood flow with and without sedation with propofol in control subjects and at a target mean arterial pressure of 65 mm Hg and greater than or equal to 75 mm Hg in septic patients. The sequence of sedation versus no sedation and mean arterial pressure targets were randomized. In septic patients, cerebral blood flow measured at a mean arterial pressure target of 65 mm Hg (40.4 ± 10.9 mL/100 g/min) was not different from cerebral blood flow measured at a mean arterial pressure target of greater than or equal to 75 mm Hg (41.3 ± 9.8 mL/100 g/min; p = 0.65). In control subjects, we observed no difference in cerebral blood flow measured without and with sedation (24.8 ± 4.2 vs 24.9 ± 5.9 mL/100 g/min; p = 0.93). We found no interaction between chronic hypertension and the effect of sedation or mean arterial pressure targets. Cerebral blood flow measured in sedated septic patients (mean arterial pressure target 65 mm Hg) was 62% higher than in sedated control subjects (p = 0.001). In septic patients, cerebral blood flow was higher than in sedated control subjects and did not vary with mean arterial pressure targets. Further research is required to understand the clinical significance of cerebral hyperperfusion in septic patients on vasopressors and to reassess the neurologic effects of current mean arterial pressure targets in sepsis.