Abstract
BACKGROUND: Isoniazid (INH) is an important drug in many TB regimens, and unfavorable treatment outcomes can be caused by suboptimal pharmacokinetics. Dose adjustment can be personalized by measuring peak serum concentrations; however, the process involves cold-chain preservation and laboratory techniques such as liquid chromatography (LC)/mass spectrometry (MS), which are unavailable in many high-burden settings. Urine spectrophotometry could provide a low-cost alternative with simple sampling and quantification methods. METHODS: We enrolled 56 adult patients on treatment for active TB. Serum was collected at 0, 1, 2, 4, 6, and 8 h for measurement of INH concentrations using validated LC-MS/MS methods. Urine was collected at 0-4, 4-8, and 8-24 h intervals, with INH concentrations measured using colorimetric methods. RESULTS: The median peak serum concentration and total serum exposure over 24 h were 4.8 mg/L and 16.4 mg*hour/L, respectively. Area under the receiver operator characteristic curves for urine values predicting a subtherapeutic serum concentration (peak <3.0 mg/L) were as follows: 0-4 h interval (AUC 0.85, 95% CI 0.7-0.96), 0-8 h interval (AUC 0.85, 95% CI 0.71-0.96), and 0-24 h urine collection interval (AUC 0.84, 95% CI 0.68-0.96). CONCLUSION: Urine spectrophotometry may improve feasibility of personalized dosing in high TB burden regions but requires further study of target attainment following dose adjustment based on a urine threshold.