Introduction: Linezolid is a potential drug for the treatment of multidrug-resistant tuberculosis but its use is limited because of severe adverse effects such as anemia, thrombocytopenia, and peripheral neuropathy. This study aimed to develop a model for the prediction of linezolid area under the plasma concentration-time curve from 0 to 12 hours (AUC0-12h) by limited sampling strategy to enable individualized dosing.
Patients and methods: Fourteen patients with multidrug-resistant tuberculosis received linezolid twice daily as part of their antituberculosis treatment. Linezolid concentrations were determined at steady state by high-performance liquid chromatography tandem mass spectrometry before and at 1, 2, 4, 8, and 12 hours after dosing. Linezolid AUC0-12h population model and limited sampling models were calculated with MWPharm software. The correlation between predicted linezolid AUC0-12h and observed linezolid AUC0-12h was investigated by Bland-Altman analysis.
Results: A total of 26 pharmacokinetic profiles were obtained. The median AUC0-12h was 51.8 (interquartile range, 41.8-65.9) mg*h/L at 300 mg and 123.8 (interquartile range, 100.9-152.5) mg*h/L at 600 mg, both twice daily. The most relevant model clinically for prediction of linezolid AUC0-12h used a linezolid trough concentration (r = 0.91, prediction bias = -2.9% and root mean square error = 15%).
Discussion: The difference between choosing a trough concentration and two to three samples increased the correlation from 0.90 to 0.95 but appeared not clinically relevant because it did not result in different dosing advice.
Conclusion: This study showed that linezolid AUC0-12h in patients with multidrug-resistant tuberculosis could be predicted accurately by a minimal sampling strategy and could be used to individualize the dose.