Clinical pharmacodynamics of linezolid in seriously ill patients treated in a compassionate use programme

Objective: To characterise the pharmacokinetic-pharmaeodynamic relationships for linezolid efficacy. Design and study population: Retrospective nonblinded analysis of severely debilitated adult patients with numerous comorbid conditions and complicated infections enrolled under the manufacturer's compassionate use programme. Methods: Patients received intravenous or oral linezolid 600mg every 12 hours. Plasma concentrations were obtained and a multicompartmental pharmacokinetic model was fitted. Numerical integration of the fitted functions provided the area under the concentration-time curve over 24 hours (AUC), the ratio of AUC to minimum inhibitory concentration (AUC/MIC) and the percentage of time that plasma concentrations exceeded the MIC (%T>MIC). Main outcome measures: Modelled pharmacodynamic outcomes of efficacy included probabilities of eradication and clinical cure (multifactorial logistic regression, nonparametric tree-based modelling, nonlinear regression) and time to bacterial eradication (Kaplan-Meier and Cox proportional hazards regression). Factors considered included AUC/MIC, %T>MIC, site of infection, bacterial species and MIC, and other medical conditions. Results: There were 288 cases evaluable by at least one of the efficacy outcomes. Both %T>MIC and AUC/MIC were highly correlated (Spearman r(2) = 0.868). In our analyses, within specific infection sites, the probability of eradication and clinical cure appeared to be related to AUC/MIC (eradication: bacteraemia, skin and skin structure infection [SSSI], lower respiratory tract infection [LRTI], bone infection; clinical cure: bacteraemia, LRTI) and %T>MIC (eradication: bacteraernia, SSSI, LRTI; clinical cure: bacteraernia, LRTI). Time to bacterial eradication for bacteraemias appeared to be related to the AUC, %T>MIC and AUC/MIC. For most sites, AUC/MIC and %T>MIC models performed similarly. Conclusions: Higher success rates for linezolid may occur at AUC/MIC values of 80-120 for bacteraemia, LRTI and SSSI. Chance of success in bacteraemia, LRTI and SSSI also appear to be higher when concentrations remain above the MIC for the entire dosing interval.