Effect of Intravenous Albumin Infusion on Brain Salicylate Concentration

Abstract

Salicylate poisoning appears to result in death, despite supportive care, once a critical brain salicylate concentration is reached. The binding of salicylate to albumin is saturable; free plasma salicylate concentrations rise disproportionately to total drug levels. Because unbound salicylate distributes into the brain, the authors questioned whether an intravenous (i.v.) infusion of albumin would cause a redistribution of salicylate from the brain back into the plasma, which might allow enough time for hemodialysis to be instituted. To determine if i.v. albumin infusion would lower brain salicylate concentrations through redistribution in a porcine model of acute salicylate poisoning. In a randomized controlled trial, 17 swine under anesthesia and controlled ventilation received 400 mg/kg of sodium salicylate i.v. over 15 minutes. At 60 minutes, nine animals received 1.25 g/kg albumin (25% solution) i.v. over 15 minutes, while eight control animals received an equal volume of normal saline (5 mL/kg). Arterial pH was maintained between 7.45 and 7.55. Serial measurements of serum albumin as well as free and total salicylate concentrations were obtained, and urine was collected for measurement of total salicylate excretion. At 180 minutes, animals were killed and brains harvested for measurement of brain salicylate concentrations. Average peak serum total salicylate concentrations of 105.5 and 109 mg/dL were achieved in control and albumin-treated animals, respectively. Albumin infusion was accompanied by statistically significant increases in serum total salicylate concentrations (median from 79.5 to 86.9 mg/dL at 75 minutes), while levels decreased slightly in control animals. Serum free salicylate concentrations decreased slightly in albumin-treated animals, but the difference was not statistically significant. Median brain salicylate concentrations were about 14% lower in the albumin treatment group (17.8 mg/100 g brain) compared with controls (20.5 mg/100 g brain); this approached statistical significance (p = 0.075). Median urinary salicylate excretion was higher in the albumin-treated group (0.83 vs. 0.48 g; p = 0.072), with similar urinary pH and volumes in both groups. In this animal model of salicylate poisoning, i.v. infusion of 1.25 g/kg albumin was accompanied by a 14% decline in median brain salicylate concentrations, which approached statistical significance.