On the bromide test of permeability of the barrier between blood and cerebrospinal fluid—an assessment

Abstract

The bromide test of meningeal permeability (Walter, 1929) was critically examined. The auric chloride method for estimation of bromide in serum is unreliable, and values for the serum/ cerebrospinal fluid bromide ratio determined by this method are too high. When bromide is added to human blood an equilibrium between plasma and erythrocytes is established in a few minutes. At this equilibrium the ratio (Br- in cell water)/Br- in plasma water) is 0.84 (S.D. 0.08) and from the magnitude of this ratio it is inferred that administered bromide replaces more than its equivalent of chloride in the human erythrocyte. The ratio also varies inversely with the pH of the blood. On storage of blood the serum bromide tends to fall to a level corresponding to an even distribution throughout the blood water. With clotted blood stored in a refrigerator this process is probably not complete for 1-2 weeks. When the serum is separated within 1-2 days the bromide level is not likely to have dropped more than 2-3%. Intravenously injected bromide takes several hours to equilibrate with the extracellular fluid of the body. Administered bromide persists in the blood for several weeks. Our evidence favors the view that the kidney excretes chloride in preference to bromide. Intravenously injected bromide, except where there is increased barrier permeability, normally takes about 24 hours to equilibrate with the lumbar cerebrospinal fluid and longer with ventricular fluid. The bromide concentration of cerebrospinal fluid decreases on ascending the neural axis. In nearly normal cases the bromide concentration in ventricular cerebrospinal fluid is about half that of lumbar.