With the aim of exploring putative correlations between serum and CSF
levels of clozapine and its demethyl metabolite, lumbar- puncture was
performed on four male and five female schizophrenic patients during l
ong-term treatment with clozapine. Three consecutive 6-ml fractions we
re collected after at least 8 h of bedrest and fasting. On comparing s
erum and CSF levels, a correlation was found for norclozapine in the t
hird (13-18 ml) CSF fraction. Norclozapine in the first (0-6 ml) CSF c
orrelated significantly with height. The CSF/serum ratio of clozapine
in the first fraction was correlated significantly with body weight. N
o correlations were found between serum levels of clozapine and norclo
zapine, or between the serum and CSF levels of clozapine. The study su
ffers from a small number of patients (for ethical reasons), but the p
resent results might be explicable if the first (0-6 ml) CSF fraction
represents a cul-du-sac of the CSF, mirroring the previous day's drug
levels. The second fraction, then, will represent the CSF level in the
steady state during the night.