ANALYSIS OF ZIDOVUDINE DISTRIBUTION TO SPECIFIC REGIONS IN RABBIT BRAIN USING MICRODIALYSIS

The distribution of zidovudine (3'-azido-3'-deoxythymidine; AZT) into two regions of rabbit brain was investigated in crossover using microd ialysis. Six rabbits had guide cannulas surgically implanted in the la teral ventricle and thalamus by stereotaxic placement. After recovery, microdialysis probes were positioned and i.v. bolus doses of 5, 10, 2 0, and 30 mg/kg were administered to each animal over a period of 2 we eks. Blood was drawn via a marginal ear vein catheter for 8 hr. Brain dialysate was collected at 3 mu-l/min from ventricle and thalamus dial ysis probes every 10 min. Simulated cerebrospinal fluid (CSF), to whic h 3'-azido-2',3'-dideoxyuridine (AZdU) was added, was used as perfusat e. AZdU loss, which was measured during simultaneous retrodialysis, se rved as a marker for in vivo recovery of AZT. AZT concentrations in pl asma, as well as in ventricle and thalamus dialysate, were determined using a sensitive HPLC assay, and AZdU was simultaneously analyzed in the dialysates. Calculation of in vivo recovery of AZT was based on lo ss of AZdU from the perfusate during retrodialysis and was used to est imate the concentration of drug at both sites in the brain. In vitro l oss of AZdU and recovery of AZT showed good agreement, demonstrating a bivariate regression slope of 0.99. The half-lives and AUCs (normaliz ed to dose) achieved in the plasma, ventricle, or thalamus were not si gnificantly different for the four doses. The AUC ratios, which repres ent the ratio of clearances into and from the CNS, were not significan tly different among the doses studied (AUC(v)/AUC(p) range, 0.16-0.19; AUC(t)/AUC(p) range, 0.05-0.09), providing further evidence that the kinetics of distribution into the thalamus and CSF are linear. The res ults also demonstrate that the time-averaged concentrations of AZT in thalamus ECF are about half of those in the CSF.