Valproic acid (VPA) has been widely used in Chinese patients after craniotomy. Many studies have focused on the influencing factors of VPA serum concentration, but conclusions are sometimes paradoxical. Furthermore, the concentration of VPA in the cerebrospinal fluid (CSF) has been rarely reported. In the present study, VPA CSF concentrations were measured, and the potential factors influencing serum concentration and CSF distribution of VPA were investigated. In addition, the functional relationship between serum and CSF concentration was explored.Methods:
Subjects were patients who underwent craniotomy and were administrated with VPA and had a lumbar puncture. Serum and CSF VPA concentrations were measured by use of the Abbott i1000 system. CYP2C9 (430 C>T, 1075 A>C, 1076 T>C, 1080 C>G), UGT1A6 (541 A>G, 552 A>C), UGT2B7 (211 G>T, 802 C>T), and ABCB1 (1236 C>T, 2677 G>T/A, 3435 C>T) genotypes were determined by direct sequencing. Information, such as age, gender, and height, was collected, and their effect on serum and CSF VPA concentrations was investigated by univariate analysis and multiple linear regression analysis.Results:
First, the concomitant use of carbapenems (β′ = −0.422) and UGT1A6 (552 AA → AC) (β′ = −0.249) had a significant negative correlation with the weight-adjusted VPA serum concentration (C:W ratio), whereas CYP2C9 (1075 AA → AC) (β′ = 0.186) and gender (female compared with male) (β′ = 0.322) showed a positive correlation with VPA serum C:W ratio. The coefficient of determination (R2) was only 0.348. Second, the relationship between the serum concentration and the CSF square root of the concentration (R2 = 0.705) had a better linear fit. Third, serum VPA concentration (β′ = 0.810), concomitant use of glycerol fructose (β′ = 0.160), and age (≥65 compared with <65) (β′ = 0.118) showed a positive correlation (R2 = 0.748) with the variability of square root of the concentration of the CSF.Conclusions:
In Chinese patients, after craniotomy, female patients with 1 or more of CYP2C9 (1075 AC) and UGT1A6 (552 AA) genotypes required a lower VPA dosage compared with male patient. There was a better-fitted linear relationship between VPA serum and the square root of CSF concentrations. CSF VPA concentrations were relatively stable, with only age and the use of glycerol fructose having a small influence.