Population pharmacokinetics of intravenous ampicillin in awake and anesthetised dogs

The aim of this study is to describe a population pharmacokinetic model for intravenous ampicillin-sulbactam in awake and anaesthetized dogs in these two treatment scenarios and to compute PK/PD cut-offs (PK/PD_CO). This was a prospective clinical trial in 20 client-owned dogs, either treated by ampicillin after post-surgical infection, or in the context of surgical antimicrobial prophylaxis. All animals received 20 mg/kg of ampicillin by slow iv route. During anesthesia ampicillin administration was repeated every 90 min according to standard recommendations, whereas in awake animals, ampicillin administration was every 8 h. A bi-compartmental model best described the data. Anesthesia and creatinine plasma concentration at the administration correlated with ampicillin clearance, which was 318 ± 36 ml/kg/h in awake animals and 186 ± 12 ml/kg/h in anesthetized animals. Monte-Carlo simulation showed that 20 mg/kg every 8 h in awake dogs led to PK/PD_CO of 0.5 mg/L using a target of 40 %fT>MIC and 0.25 mg/L with 50 % fT>MIC. Administration of ampicillin by prolonged infusion (4 h) every 8 h allowed to achieve a PK/PD_CO of 4 mg/L and 2 mg/L with the 40 % and 50 % targets, respectively. In anesthetized animals, repetition of the 20 mg/kg of ampicillin every 90 min allowed to remain above 4 mg/L during 100 % of the time. Plasma clearance of ampicillin inversely correlates with creatinine plasma concentration in client-owned dogs and is reduced of about 40 % during anesthesia. In awake dogs, the standard regimen of 20 mg/kg every 8 h by intravenous route is probably insufficient to treat soft tissues infections caused by gram-negative bacteria. Using prolonged infusion could help to increase ampicillin efficacy against specific pathogens.