ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Sept. 1994, p. 2111-2115 0066-4804/94/$04.00+0 Pharmacokinetics of Fluconazole in Cerebrospinal Fluid and Serum of Rabbits: Validation of an Animal Model Used To Measure Drug Concentrations in Cerebrospinal Fluid A. MADU,' C. CIOFFE,2 U. MIAN,' M. BURROUGHS,2 E. TUOMANEN,2 M. MAYERS,3 E. SCHWARTZ,1 AND M. MILLER4* Departments of Medicine' and Ophthalmology,3 Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York; Rockefeller University, New York, New York2; Departments of Medicine and Pharmacology, Albany Medical College, Albany, New York4 Received 8 November 1993/Returned for modification 9 February 1994/Accepted 22 June 1994 Complete concentration-time data describing the pharmacokinetics of fluconazole in the cerebrospinal fluid (CSF) following a single dose are not available for humans or animals. We studied the pharmacokinetics of fluconazole with an indwelling intracisternal needle as described by R G. Dacey and M. A. Sande (Antimicrob. Agents Chemother. 6:437-441, 1974). To determine whether the presence of an intracisternal needle alters pharmacokinetics in the CSF, we validated this model with uninfected rabbits by measuring pharmacokinetic constants following direct intracisternal and intravenous administration of fluconazole. Following direct injection, there was no alteration of elimination rates in the CSF with increasing sample number or time. Following intravenous administration, the penetration and kinetic constants were the same in individual animals from which multiple CSF samples were obtained as in a composite subject constructed by pooling virgin samples from different animals. The presence of the intracisternal needle did not alter CSF chemistry or leukocyte counts, and erythrocyte contamination was <0.001%. While drug concentrations were measured by a microbiological assay, we also compared the sensitivity and reproducibility of a high-performance liquid chromatography (HPLC) assay with those of the microbiological assay. Following a single intravenous dose, the maximum concentration of the drug in serum, the time to maximum concentration of the drug in serum, the terminal elimination half-life in the CSF, and the percent penetration by fluconazole were 6.12 ,ug/ml, 1 h, 9.0 h, and 84.3%, respectively. We conclude that the sampling of CSF via an indwelling needle does not alter fluconazole pharmacokinetics, cause inflammation, or alter chemical parameters; that the microbiological assay is at least equivalent in sensitivity and reproducibility to the HPLC assay; and that robust parameters describing the pharmacokinetics of fluconazole are possible with this model. Animal studies of the pharmacokinetics of antibiotics in which an intracisternal needle is used to obtain cerebrospinal fluid (CSF) have been helpful in predicting the efficacies of different regimens in the treatment of bacterial meningitis in humans (4, 17, 18, 20). There are no reports describing the pharmacokinetics of azole antifungal agents in CSF in which complete concentration-time data for rabbits or humans are presented. Moreover, it has never been established whether serial sampling with an indwelling cisternal needle affects pharmacokinetics in the CSF or CSF physiology. The minimal trauma associated with paracenteses in other organ systems is known to cause disruption of tight vascular endothelial junc- tions and a breakdown of barrier function (2, 12, 13). Existing data describing the pharmacokinetics of azoles in the CSF are from concentration-time points pooled from different animal and human subjects (5, 14), which results in limited pharma- cokinetic parameter reliability due in part to intersubject variability. The relative penetration of azoles and amphotericin B into the CSF does not correlate with clinical efficacy (6, 7, 15, 21). This lack of correlation could be due either to the absence of a sufficiently sensitive assay to measure drug concentrations in the CSF, which would permit a comparison of concentration in * Corresponding author. Mailing address: Departments of Medicine and Pharmacology, Division of Infectious Diseases, Albany Medical College, 47 New Scotland Ave., Albany, NY 12208. Phone: (518) 262-5775. Fax: (518) 262-3734. CSF/MIC ratios for different drugs, or to the lack of an association between CSF penetration and efficacy. The primary objectives of our studies were to determine the pharmacoki- netics of fluconazole in the serum and CSF following intracis- ternal and intravenous administration and to validate the animal model in which serial samples of CSF are obtained via an indwelling intracisternal needle (4). As part of our prepa- ration, we additionally sought to optimize and compare the sensitivities and reproducibilities of two commonly employed assays (the microbiological assay and the high-performance liquid chromatography [HPLC] assay). (Part of this research was presented at the 1991 Interscience Conference on Antimicrobial Agents and Chemotherapy.) MATERIALS AND METHODS Animal model. Adult male New Zealand White rabbits (Hare Marland, Nutley, N.J.) weighing 2 to 3 kg were used. Five animals were used for the direct intracisternal injection; 5 were used for the determination of the effects on cell count and protein, glucose, and lactate concentrations; and 10 were used for systemic administration studies. The anesthetic consisted of 15 to 20 ml of 25% urethane given subcutaneously and 15 mg of pentobarbital per kg of body weight given intravenously. Following anesthesia, a 24-gauge angiocatheter was inserted into a marginal ear vein for drug administration. A second angiocatheter was placed in the central artery of the contralat- eral ear for serum sampling. For systemic administration, 10 2111 Vol. 38, No. 9 on December 22, 2015 by guest http://aac.asm.org/ Downloaded from