Determination of protein-unbound, active rifampicin in serum by ultrafiltration and Ultra Performance Liquid Chromatography with UV detection. A method suitable for standard and high doses of rifampicin

- Higher than approved rifampicin doses are being evaluated to optimize TB treatment.
- An ultrafiltration UPLC-UV method was developed for protein-unbound rifampicin.
- Methanol and ascorbic acid kept rifampicin soluble and stable in the ultrafiltrate.
- Ultrafiltration was performed at 37 °C.
- Recovery of ultrafiltration and variability in ultrafiltration devices were evaluated.

Rifampicin is the most important antibiotic in use for the treatment of tuberculosis (TB). Preclinical and clinical data suggest that higher doses of rifampicin, resulting in disproportionally higher systemic exposures to the drug, are more effective. Serum concentrations of rifampicin are the intermediary link between the dose administered and eventual response and only protein-unbound (free) rifampicin is pharmacologically active. The objective of this work was to develop an ultra performance liquid chromatography assay for protein-unbound rifampicin in serum with ultrafiltration, carried out at a sample temperature of 37 °C, suitable for measurement of concentrations achieved after currently used and higher doses of rifampicin.Human serum was equilibrated at 37 °C and ultrafiltrated at the same temperature in a Centrifree YM-30 ultrafiltration device, followed by dilution of the ultrafiltrate with methanol and ascorbic acid. Unbound rifampicin was analyzed using ultra performance liquid chromatography with a BEH C18 column, isocratic elution and ultra-violet (UV) detection. The run time was 5 min.The assay was linear over the concentration range of 0.065-26 mg/L rifampicin in ultrafiltrate. Accuracies for measurement of rifampicin in ultrafiltrate were 97% and 102% at the higher and lower limits of quantitation. Accuracy of the ultrafiltration process cannot be established, as it is not possible to spike blank serum with known amounts of protein-unbound rifampicin. Within- and between-day precision of the method including ultrafiltration as well as after ultrafiltration were within prespecified limits (CV < 10%). Dilution of the ultrafiltrate with methanol and ascorbic acid kept rifampicin in solution and prevented it from degradation. Rifampicin loss during the ultrafiltration process and variation in analytical results when using two different batches of ultrafiltration devices were both limited. Processed ultrafiltrate samples were stable for 3 days in the autosampler.The developed method can be applied in pharmacokinetic research, studying exposure-response relationships for rifampicin when administered at higher than currently used doses.