Sputum Pharmacokinetics of TB Drugs and Bacterial Drug Resistance

This study will support two avenues of research, pharmacokinetics (PK) of drugs used to treat tuberculosis and resistance to those drugs that has been developed by the pathogen. Given the high inter-individual variability in TB drug exposure, therapeutic drug monitoring would be of value to adjust the drug dose in TB patients in order to improve clinical outcome and minimize toxicity. It is not practiced in most settings for reasons associated with costs but also because blood collection is an invasive procedure. Sputum on the other hand is often produced spontaneously and discarded as waste. Drug levels can be measured in sputum just as in plasma. Here we will test the hypothesis that sputum drug levels are predictive of drug concentrations in plasma and/or in specific lesion compartments such as the caseum of open cavities. We will also characterize the exposure of standard 1st and 2nd line TB drugs at one of the sites of infection, since sputum is in direct contact with cavity caseum. The subjects will contribute three or more sputum samples spontaneously produced over at least two days and four blood samples following drug administration on two different days (8 blood samples total). The data will be analyzed using population PK modeling approaches to generate concentration-time profiles and 24-hour area under the curve (AUC) of each study drug in sputum. Correlations between these values, plasma and lesion AUCs will be examined. We will draw from a recently completed TB lesion pharmacokinetic study (www.ClinicalTrials.gov #NCT00816426) to seek correlations between drug exposure in sputum and in pulmonary lesions. In addition to drug concentrations, sputum will be cultured by standard methods to isolate Mycobacterium tuberculosis (Mtb), determine the drug resistance profile, and be saved for testing a second generation of the Xpert TB XDR assay.