Fulminant hepatic failure (FHF) is often complicated by high intracranial pressure (ICP) and fatal brain damage. In this study, we determined if a rise in [glutamate] ec and [lactatelec preceded surges of high ICP in patients with FHF (median age, 42; range, 20-55 years; 7 women; 3 men) by inserting a microdialysis catheter into the brain-cortex together with an ICP catheter. The microdialysis catheter was perfused with artificial cerebrospinal-fluid at a rate of 0.3 pL/min. Dialysate was collected approximately every 30 minutes or when ICP increased. A total of 352 microdialysis samples were collected during a median of 3 days and allowed for -1,760 bedside analyses of the collected dialysate. In 5 patients that later developed surges of high ICP, the initial values of [glutamatelec and [lactatelec were 2 to 5 times higher compared with patients with normal ICP. [Glutamatelec then tended to vanish with time in both groups of patients. An increase in [glutamatelec did not precede high ICP in any of the cases. In contrast, [lactatelec was high throughout the study in the high ICP group and increased further before surges of high ICP. We conclude that in patients with FHF, cerebral [glutamatelec and [lactatelec are elevated. However, the elevated [glutamatelec is not correlated to high ICP. In contrast, elevations in [lactatelec preceded surges of high ICP. In conclusion, accelerated glycolysis with lactate accumulation is implicated in vasodilatation and high ICP in patients with FHF. The data suggest that bedside cerebral microdialysis is a valuable tool in monitoring patients with FHF and severe hyperammonemia. (HEPATOLOGY 2002;36:1333-1340.) D ilation of cerebral arterioles is of importance for development of brain edema and high intracranial pressure (ICP) in fulminant hepatic failure (FHF). 1,2 Experimental studies indicate that metabolic derangements in the brain, induced by hyperammonemia, precede a critical rise in cerebral blood flow and ICP."* Thus, assessing alterations in brain tissue chemistry may provide important clinical information well before high ICP evolves, and it is possible that new methods that can monitor the underlying cerebral metabolic impairment may provide new avenues even to prevent brain edema in patients with FHF. Ideally, the method chosen should be performed bedside, possess a high reproducibil-