Ivan Nicholas Nkuhairwe1, Prerna R Arora2, Juan Eduardo Resendiz Galvan1, Dr. Roeland. E Wasmann1, Bhamini Keny2, Zarir. F Udwadia2, Camilla Rodrigues2, Jeffrey. A Tornheim3, Tester. F Ashavaid2, Paolo Denti1
1Division of Clinical Pharmacology, Department of Medicine, University of Cape Town, 2P.D. Hinduja National Hospital and Medical Research Centre, 3John Hopkins University
Objectives: In 2023, India accounted for 27% of global multi-drug-resistant tuberculosis (MDR-TB) cases and ranked in the top-decile of countries with a broad gap between estimated number of individuals who developed MDR-TB and those enrolled in treatment¹. Amikacin is an aminoglycoside, and second line injectable that is crucial in treatment of MDR-TB². Following an intramuscular dose of 7.5 mg/kg/day, amikacin is rapidly absorbed. Almost the entire dose is renally cleared unchanged with a half-life of 2-3 hours in adults with normal renal function. Amikacin’s toxicity, including irreversible ototoxicity and reversible nephrotoxicity, necessitates monitoring of renal function and possibly the adjustment of the dose. This is generally done by changing the dosing interval, leading to strategies like intermittent dosing to mitigate risks, especially in patients with impaired renal function³. Despite its long history of use, substantial information on its pharmacokinetic (PK) characteristics in MDR-TB patients, remains limited². The aim of this study was to characterise the PK of amikacin among Indian patients with MDR-TB. Methods: We evaluated drug levels obtained during susceptibility-guided treatment for MDR-TB. In this analysis, we focused on participants that received amikacin intramuscularly per National Tuberculosis Elimination Programme guidelines. Dosing interval was adjusted in association with renal function per physicians’ evaluation, with some patients receiving intermittent doses and others daily doses. Four PK visits were scheduled: 3 visits with sparse sampling at 1, 6, and 12 months, and 1 with intensive sampling at either 1 or 2 months. Sampling times were predose and 2 h post-dose for sparse, and predose, 1, 2, 4, 6 and 8 h post-dose for intensive sampling. Samples were quantified using a homogenous enzyme immunoassay with a lower limit of quantification (LLOQ) of 2.5 mg/L. Data was analysed using non-linear mixed effects modeling in NONMEM (v.7.5.1 ) with first-order conditional estimation with interaction. One and two compartment models with first-order elimination and absorption (with or without absorption delays) were tested. Between-subject variability was explored for disposition parameters, and between occasion variability for absorption parameters and bioavailability. A combined proportional and additive error model quantified residual unexplained variability. Observations below the limit of quantification (BLQ) were imputed as half of LLOQ and their additive error component inflated by 50% of LLOQ. Allometric scaling of clearance (CL) (fixed exponent:0.75) and volume of distribution (fixed exponent:1) by either fat-free mass or weight were investigated. Additionally, the effect of creatinine clearance (CRCL) derived from Cockroft-Gault estimation on CL, was assessed. CRCL was standardised to that of a 55 kg individual and the effect of weight removed as illustrated by Mould et al.4 when applying allometry and renal function simultaneously. This was then centred around the median of the standardised CRCL and implemented using a linear relationship. Results: We included 52 adults and adolescents (75% female) with a media n (range): weight 55.0 (32.0-99.0) kg, age 26.0 (15.0-61.0) years and creatinine clearance 111.3 (44.4-251.9) mL/min. 320 samples were analyzed of which, 110 were BLQ with 108 being predose samples and 2 8h samples. A two-compartment model with first order absorption and elimination best fitted the data (?OFV = -17364, p-value <0.001 ) with a typical clearance (95%CI) of 3.66 (3.34, 3.96) L/h. Allometric scaling by weight on CL and volume of distribution improved the model fit (?OFV = -28.7). There was a statistically significant effect of creatinine clearance on drug clearance where every 10-unit drop in creatinine clearance from the median (120 mL/min) resulted in a 5.6% (3.3,7.5) decrease in clearance (?OFV = -17.7, p-value <0.001) Conclusions: We developed a two-compartment model that characterized the PK of amikacin in MDR-TB patients from India, and included the effect of renal function. This aligns with findings from Kato et al.5 who also reported similar values of CL and model structure in patients admitted for various types of infections. Our model can be used to adjust the dose based on renal function.
1. World Health Organisation. Global Tuberculosis Report.; 2024. 2. Sturkenboom MGG, Simbar N, Akkerman OW, Ghimire S, Bolhuis MS, Alffenaar JWC. Amikacin Dosing for MDR Tuberculosis: A Systematic Review to Establish or Revise the Current Recommended Dose for Tuberculosis Treatment. Clinical Infectious Diseases. 2018;67:S303-S307. doi:10.1093/cid/ciy613 3. Pfizer N. Data Sheet-New Zealand DBL TM Amikacin Injection.; 2017. https://go.drugbank.com/drugs/DB00479 4. Mould DR, Holford NHG, Schellens JHM, et al. Population pharmacokinetic and adverse event analysis of topotecan in patients with solid tumors. Clin Pharmacol Ther. 2002;71(5):334-348. doi:10.1067/MCP.2002.123553 5. Kato H, Hagihara M, Hirai J, et al. Evaluation of Amikacin Pharmacokinetics and Pharmacodynamics for Optimal Initial Dosing Regimen. Drugs R D. 2017;17(1):177-187. doi:10.1007/s40268-016-0165-5
Reference: PAGE 33 (2025) Abstr 11598 [www.page-meeting.org/?abstract=11598]
Poster: Drug/Disease Modelling - Infection