# Amikacin exposure in MDR-TB patients in Uganda: Revisiting old drugs in a new era of resistance – A pharmacokinetic assessment

## TL;DR

This study compares amikacin drug levels in Ugandan MDR-TB patients to European data, finding similar patterns and supporting the use of a simple model for drug monitoring in resource-limited areas.

## Contribution

The study demonstrates that a one-compartment model is suitable for therapeutic drug monitoring of amikacin in low-resource settings.

## Key findings

- Ugandan MDR-TB patients showed similar pharmacokinetic profiles to a European cohort.
- A one-compartment model accurately modeled amikacin concentration-time curves in Ugandan patients.
- Pharmacokinetic parameters like AUC and Cmax were successfully measured in a limited resource setting.

## Abstract

Amid rising resistance to bedaquiline, aminoglycosides remain an important fallback option for multidrug-resistant tuberculosis (MDR-TB) treatment, particularly in high-burden settings. Their use is limited by nephro- and ototoxicity, which is associated with cumulative drug exposure. In this study we investigated amikacin exposure in Ugandan MDR-TB patients using a validated limited sampling strategy and compared the results to previously published data from a Western European cohort.

In this single-centre prospective observational study, 29 MDR-TB patients received amikacin at a dose of 10-15 mg/kg. Serum levels were measured on day 30 at 1, 4 and 5 h post-administration using liquid chromatography/mass spectrometry. Individual concentration-time curves were modelled using a one-compartment model and compared to a Dutch population-pharmacokinetic (PK) model.

Twenty patients had complete PK data. Patients received a median amikacin dose of 10.9 (IQR 10 - 14.9) mg/kg; clearance was 4.79 L/h (IQR 4.03 - 5.75), volume of distribution (Vd) 16.3 L (IQR 14.07 – 21.49), AUC0-24h 125.15 h x mg/l (IQR 106.73 – 174.46), maximum serum concentration (Cmax) 27.8 mg/l (IQR 22.9 – 48.7).

This population-PK study shows that major differences in PK between Ugandan MDR-TB patients and those in the Global North are unlikely. Our findings reinforce the suitability of a one-compartment model for therapeutic drug monitoring in both high- and low-resource settings. Readily obtained aminoglycoside PK parameters in a limited resource setting facilitate future efforts in optimizing drug exposure with minimal toxicities, in the population most affected by the pandemic of TB.