Protein binding investigation of first-line and second-line antituberculosis drugs.

Data on protein binding are incomplete for first-line antituberculosis drugs, and lacking for second-line antituberculosis drugs that are used extensively for multi-drug-resistant tuberculosis (levofloxacin, linezolid and moxifloxacin). Thus, the main purposes of this study were to investigate: (i) the relationship between carrier protein concentration and drug binding; and (ii) the feasibility of predicting free drug concentration using in-vitro and in-vivo results. In-vitro experiments were performed on spiked plasma mimicking real-case samples (drug combinations from clinical practice). Median in-vivo protein binding was 1.5% for ethambutol, 9.7% for isoniazid, 0.7% for pyrazinamide and 88.2% for rifampicin; and median in-vitro protein binding was 26.2% for levofloxacin, 12.8% for linezolid and 46.3% for moxifloxacin. Albumin concentration (<30 g/L) had a moderate impact on moxifloxacin binding and a strong impact on levofloxacin, linezolid and rifampicin binding. Determination of the free drug concentration seems to be of little value for ethambutol, isoniazid, moxifloxacin and pyrazinamide; limited value for linezolid because of its low binding; and major value for rifampicin in hypoalbuminaemic patients with tuberculosis, and levofloxacin because total concentration was an inaccurate reflection of free concentration. The free concentration predicted by the mathematical model was suitable for levofloxacin and linezolid, whereas the real free concentration should be measured for rifampicin. Further investigations should be carried out to investigate the benefit of using free concentration for levofloxacin, linezolid and rifampicin, particularly in the critical period of active tuberculosis associated with hypoalbuminaemia.