Pharmacokinetics of amikacin and chloramphenicol in the aqueous humor of rabbits.

Composite data describing ocular pharmacokinetics are unreliable because of intersubject variation. To address this problem, an animal model was developed in which multiple aqueous samples from single subjects were obtained. Following direct anterior chamber or intravenous administration of amikacin or chloramphenicol, pharmacokinetic analysis of drug concentrations in the serum and anterior chamber was performed by using a nonlinear least-squares regression program. The number of anterior chamber paracenteses performed did not alter the beta elimination rates or percent penetration into the anterior chamber. The aqueous humor and peripheral-compartment terminal slopes were identical. These data indicate that complete ocular concentration-time curves can be obtained without altering antibiotic pharmacokinetics. Following direct injection into the anterior chamber, the elimination rates for both antibiotics followed a one-compartment model, whereas those following intravenous administration best fit an open, first-order, two-compartment model. Following intravenous administration, the anterior chamber elimination rate constants for both drugs were equal to that of the serum and significantly longer than that following direct injection. The elimination rates of both drugs following direct injection were similar. Systemic administration resulted in drug levels in aqueous humor that persisted longer than those following direct injection. Chloramphenicol, a lipophilic compound, gave higher mean concentrations in aqueous humor than did amikacin. Our model provides a new approach which rigorously examines ocular pharmacokinetics and provides data which suggest that for selected compounds the parenteral route of administration is preferable.