Mahmoud Tareq Abdelwahab (1), Elin M. Svensson (2,3), Andreas Diacon (4), Almari Conradie (5), Morounfolu Olugbosi (5), Rodney Dawson (6), Gary Maartens (1,7), Paolo Denti (1)
(1) Division of Clinical Pharmacology, Department of Medicine, University of Cape Town, Cape Town, South Africa, (2) Department of Pharmacy, Radboud Institute of Health Sciences, Radboud University Medical Center, Nijmegen, Netherlands, (3) Department of Pharmacy, Uppsala University, Uppsala, Sweden, (4) Task Applied Science, Bellville, South Africa, (5) Global Alliance for TB Drug Development, New York, USA, (6) Division of Pulmonology and Department of Medicine, University of Cape Town Lung Institute, Mowbray, Cape Town, South Africa, (7) Wellcome Centre for Infectious Diseases Research in Africa, Institute of Infectious Disease and Molecular Medicine, Department of Medicine, University of Cape Town, Cape Town, South Africa
Objectives:
The occurrence of drug resistance (DR) to tuberculosis (TB) medications poses critical challenges to global efforts to manage TB. For DR-TB the WHO recommends multiple drugs in a long and complex treatment regimen (9 to 12 months and sometimes up to 20 months). Bedaquiline (BDQ) is recommended as the first-line option against DR-TB; however, bedaquiline, through its main metabolite M2 is associated with QT interval prolongation[1] . Clofazimine (CFZ), a repurposed drug with a promising role in DR-TB has been linked to a significant QT interval prolongation[2]. Pretomanid (Pa), a novel new drug, was recently approved by the FDA in combination with BDQ against extensively DR-TB and is also associated with QT interval prolongation[3]. Limited information is available on the possible interactions and combined toxicity of these drugs. We applied nonlinear mixed-effects modelling to characterise the exposure-QTc relationship following the administration of BDQ, CFZ and Pa in South African patients with tuberculosis.
Methods: Treatment-naïve drug-susceptible TB patients were enrolled in a 14-day early bactericidal activity study of CFZ, alone or in combination with BDQ or Pa. 105 patients were randomised into 6 treatment arms and an additional control arm. BDQ was administered as 400 mg on day one, 300 mg on day two, and 200 mg on days 3-14. CFZ was administered as 300 mg for 3 days, followed by 100 mg until day 14. Pa was administered as 200 mg daily. Triplicate ECGs were performed at the screening visit and predose, 5, and 10 hours before treatment and matched with plasma samples on days 1-3, 8, and 14 at predose, 5 and 10 hours post-observed dose. Additional ECGs were performed on day 28. The QT was corrected for heart rate correlation using Fridericia formula. Diurnal variation in QTc was described using oscillator functions. A time effect on QTc was investigated with a linear or exponential function. The drugs’ effects were described with linear, and Emax models and their combined effect was tested with competitive interaction models.
Results: Pretreatment QTc data from all arms and post-treatment data from arms with non-QT prolonging drugs (105 patients, 973 observations) were used to characterise the diurnal variation, baseline covariates and time effect on QTc. The data from arms containing BDQ, CFZ and Pa were used to characterise the concentration-QTc relationship and explore possible drug-drug interactions. We used the averaged triplicate ECG measurement at each time point matched with observed concentrations. Patients had a median age of 30 years (range: 18 – 62), a median weight of 53.8 kg (40 – 86.6). Three oscillator functions best described the circadian model after relying on prior functionality in NONMEM to guide the estimation process. Age and potassium levels were identified as significant covariates affecting baseline QTc. The time effect was described by an exponential function with a half-life of 0.573 (0.362 – 1.26) weeks. BDQ M2 and CFZ drug effects were best described by an Emax function with values of (95% CI) 19.9 (14.3 – 26.8) and 28.5 (13.5 – 82.6) ms corresponding to EC50 values of 0.114 (0.0719 – 0.197) and 0.544 (0.174 – 1.89) mg/L for M2 and CFZ, respectively. The Pa effect was described by a linear relationship with a slope value of 1.39 ms/unit concentration. We identified a competitive antagonist interaction between M2 and CFZ. The drug effect of Pa was additive to BDQ and CFZ effects.
Conclusions: We present a joint concentration-QTc model of BDQ, CFZ, and Pa. We confirmed the QT-prolonging effect of all three. The interaction between M2 and CFZ causes a less-than-additive effect of CFZ on QTc. We found no interaction between Pa and BDQ or CFZ. The model can be used to assess the exposure-safety analysis of these drugs as part of other proposed drug regimens.
References:
[1] Janssen Therapeutics. SIRTURO (bedaquiline) label. Published online December 2012. Accessed April 26, 2022. https:// www.acces sdata.fda.gov/drugs atfda_docs/label/ 2012/20438 4s000 lbl.pdf
[2] Novartis Pharmaceuticals Corporation. LAMPRENE (CLOFAZIMINE) label. Published online 2016. Accessed April 26, 2022. https://www.acces sdata.fda.gov/drugs atfda_ docs/label/ 2016/01950 0s013 lbl.pdf
[3] Li, H. et al. Long-term effects on QT prolongation of pretomanid alone and in combinations in patients with tuberculosis. Antimicrob. Agents Chemother. 63, (2019).
Reference: PAGE 30 (2022) Abstr 10047 [www.page-meeting.org/?abstract=10047]
Poster: Drug/Disease Modelling - Infection