Simple strategy to assess linezolid exposure in patients with multi-drug-resistant and extensively-drug-resistant tuberculosis.

Linezolid is used increasingly for the treatment of multi-drug-resistant (MDR) and extensively-drug-resistant (XDR) tuberculosis (TB). However, linezolid can cause severe adverse events, such as peripheral and optical neuropathy or thrombocytopenia related to higher drug exposure. This study aimed to develop a population pharmacokinetic model to predict the area under the concentration curve (AUC) for linezolid using a limited number of blood samples. Data from patients with MDR-/XDR-TB who received linezolid and therapeutic drug monitoring as part of their TB treatment were used. Mw\Pharm 3.82 (Mediware, Zuidhorn, The Netherlands) was used to develop a population pharmacokinetic model and limited sampling strategy (LSS) for linezolid. LSS was evaluated over a time span of 6 h. Blood sampling directly before linezolid administration and 2 h after linezolid administration were considered to be the most clinically relevant sampling points. The model and LSS were evaluated by analysing the correlation between AUC12h,observed and AUC12h,estimated. In addition, LSS was validated with an external group of patients with MDR-/XDR-TB from Sondalo, Italy. Fifty-two pharmacokinetic profiles were used to develop the model. Thirty-three profiles with a 300 mg dosing regimen and 19 profiles with a 600 mg dosing regimen were obtained. Model validation showed prediction bias of 0.1% and r2 of 0.99. Evaluation of the most clinically relevant LSS showed prediction bias of 4.8% and r2 of 0.97. The root mean square error corresponding to the most relevant LSS was 6.07%. The developed LSS could be used to enable concentration-guided dosing of linezolid.