PEARLS• Hemodialysis (HD) has been associated with increases in intracranial pressure (ICP). Osmotic shifts occurring during HD are thought to be associated with the development of this phenomenon.• In comatose patients with acute brain injury, an elevation of ICP during HD may be associated with worse cerebral autoregulation. Continuous veno-venous hemofiltration (CVVH) may be a safer alternative in these patients.
OYSTER• HD may increase ICP; therefore, caution should be exercised when HD is needed in patients with acute brain injury.CASE REPORT A 59-year-old man with hypertension, type 1 diabetes, and end-stage renal disease on HD presented after head trauma from a fall while walking. Initial neurologic examination showed agitated confusion (E4 M5 V5) and a noncontrast head CT showed a nondepressed linear skull fracture in the right temporal bone with a small amount of subarachnoid hemorrhage in the left superior frontal convexity. On arrival to the neurointensive care unit (NICU), his mental status deteriorated (E1 M4 VT) and a follow-up noncontrast head CT revealed a new intracerebral hemorrhage (ICH) in the left temporal lobe (4.3 cm ϫ 2.2 cm ϫ 3.5 cm) with surrounding edema ( figure 1A). The patient's family opted for maximal medical management. Multimodality monitoring (MMM) including an intracranial pressure (ICP) monitor (Camino, Integra Lifesciences) and a brain oxygen monitor (Licox, Integra Lifesciences) was placed in the left frontal lobe, close to the perihematomal region. HD was performed every other day via a pre-existing left groin arteriovenous fistula with a negative goal fluid balance of 2 L per each 3.5-hour session. Initially, MMM demonstrated a mildly elevated ICP (mean 18.9 Ϯ 7.4 mm Hg) over a wide range of mean arterial pressures (MAPs) (60 -120 mm Hg) and ICP increase was not directly proportional to MAP changes, suggesting intact autoregulation. Pressure reactivity index (PRx), a Pearson moving correlation coefficient between MAP and mean ICP, which is accepted as a surrogate for autoregulation status, was also stable at 0.18 Ϯ 0.05 (mean Ϯ SD), implying intact pressure autoregulation. Partial brain oxygen tension (P bt O 2 ) was directly proportional to the cerebral perfusion pressure (CPP) up to 80 mm Hg. Oxygen reactivity index, a moving correlation coefficient between P bt O 2 and CPP, was 0.19 Ϯ 0.11. The mean P bt O 2 value was 43.1 Ϯ 11.2 mm Hg. Over the initial 2 weeks, the ICP trended up, requiring hyperosmotic agents to maintain an ICP below 20 mm Hg. Glasgow Coma Scale score deteriorated to E1, M2, VT. Follow-up brain CT scan showed increased perihematomal vasogenic edema with aggravated midline shift (figure 2). On hospital day 16, while receiving his regularly scheduled HD, the ICP surged up to 38 mm Hg just after initiating HD and remained elevated during the entire session. ICP was directly dependent on MAP, suggesting a loss of autoregulation ( figure 1B). Additionally, PRx values were elevated more than 0.2, which also suggested autoregulatory failure (figure 1C). Mea...