Penetration of piperacillin into the vitreous after intravenous administration

Purpose: Penetration of piperacillin into the vitreous cavity after intrave nous administration was investigated in humans. Methods: Forty-five eyes undergoing vitrectomy between November 1993 and De cember 1994 were included. Each patient received a single intravenous dose of piperacillin 4 g 2 hours before ocular incision. A 0.2-mL vitreous sampl e was aspired at the beginning of vitrectomy, a blood sample was obtained, and piperacillin level was assessed. Results: There were no detectable drug concentrations in 25 eyes. Mean vitr eous drug concentration in the remaining evaluable eyes (n = 14) was 2.33 m u g/mL (+/-0.76). We divided samples into two groups: 23 uninflamed (Group 1) and 16 inflamed (Group 2) eyes. Mean vitreous drug concentration was 0.4 2 mu g/mL in Group 1 and 4.95 mu g/mL in Group 2 (P < 0.2). Piperacillin co ncentration was higher than the minimum inhibitory concentration for gram-p ositive bacteria in 13% of uninflamed and 69% of inflamed eyes (P < 0.001). Conclusion: Studies show that intravenously administered piperacillin can p enetrate the vitreous cavity in rabbits. Our study suggests that a single d ose of piperacillin can produce intravitreal concentrations sufficient to k ill gram-positive bacteria in inflamed eyes. The poor intraocular penetrati on of systemically administered piperacillin in uninflamed eyes raises ques tions about its usefulness as a prophylactic agent in ophthalmic surgery.