The passage of cloxacillin into cerebrospinal fluid in the absence of meningitis.

1. Eleven patients undergoing lumbar discectomy received cloxacillin by continuous i.v. infusion, starting before the operation. During the operation several blood samples and one CSF sample were taken. 2. Mean rate constants describing the passive transfer of drug from plasma to CSF (kp) and the largely active transfer in the opposite direction (kCSF) were estimated. 3. In some subjects the CSF albumin quotient, defined as the ratio between the albumin concentration in CSF and in plasma times 1000, was slightly elevated (up to 23) which caused a significant increase in the value of kp. 4. The estimate of mean kp for healthy individuals was 0.065 h-1, which corresponds to a half-life of 10 h. The estimate of mean kCSF was 2.10 h-1. This predicts a steady-state CSF drug concentration which is 3% of the unbound plasma drug concentration. 5. There was a significant lag between the time courses of plasma and CSF drug concentrations, presumably reflecting the time for drug to move from the choroid plexus to the lumbar sampling site. 6. Four other patients received cloxacillin for prophylactic or therapeutic reasons by continuous i.v. infusion. In three of those patients the albumin quotient was normal or slightly elevated and the steady-state CCSF/Cu ratio was similar to the predicted normal value. 7. These findings indicate that eradicating staphylococci from CSF in cases of meningitis with a low degree of inflammation may be difficult.