Central nervous system salicylate.

The poor correlation between clinical salicylate toxicity and serum blood levels is reapproached in light of recent evidence linking clinical severity with initial volume of distribution (Vd). It is recognized that two variables alter salicylate Vd in such manner that serum salicylate levels are misleading (thus, the change in Vd is not detected by present methods). These variables are serum protein binding and the pH-dependent ionized/un-ionized ratio in the unbound salicylate fraction. Measurements of salicylate concentration in the cerebro spinal fluid (CSF) would circumvent these variables, but would be clinically impractical. Thus, an alternative is sought to the inexact total serum salicylate levels and the impractical CSF salicylate levels for assessment of the severity of salicylate poisoning. This study indicates that, in dogs, serum unbound salicylate levels closely reflect CSF salicylate levels, even as a decrease in serum protein binding is in progress. However, serum unbound salicylate concentration does not reflect CSF salicylate concentration as a decrease in serum pH is elicited (CSF salicylate actually increased as serum unbound salicylate decreased). On the other hand, serum unbound salicylate measurement would seem preferable to total serum salicylate measurements now used in that the total value decreased markedly as either protein binding change or acidosis produced a change in distribution and the resultant increase in CSF salicylate.