Investigating the treatment shortening potential of a combination of bedaquiline, delamanid and moxifloxacin with and without sutezolid, in a murine tuberculosis model with confirmed drug exposures

Abstract

Background

New and shorter regimens against multi-drug resistant tuberculosis (TB) remain urgently needed. To inform treatment duration in clinical trials, this study aimed to identify human pharmacokinetic equivalent doses, antimycobacterial and sterilizing activity of a novel regimen, containing bedaquiline, delamanid, moxifloxacin and sutezolid (BDMU), in the standard mouse model (BALB/c) of Mycobacterium tuberculosis (Mtb) infection.

Methods

Treatment of mice with B₂₅D_0.6M₂₀₀U₂₀₀, B₂₅D_0.6M₂₀₀, B₂₅D_0.6M₂₀₀(U₂₀₀³) or H₁₀R₁₀Z₁₅₀E₁₀₀ (isoniazid, rifampicin, pyrazinamide, ethambutol, HRZE), started 3 weeks after Mtb infection. Bactericidal activity was assessed after 1, 2, 3 and 4 months of treatment and relapse rates were assessed 3 months after completing treatment durations of 2, 3 and 4 months.

Results

B₂₅D_0.6M₂₀₀U₂₀₀ generated human equivalent exposures in uninfected BALB/c mice. After 1 month of treatment, a higher bactericidal activity was observed for the B₂₅D_0.6M₂₀₀U₂₀₀ and the B₂₅D_0.6M₂₀₀ regimen compared to the standard H₁₀R₁₀Z₁₅₀E₁₀₀ regimen. Furthermore, 3 months of therapy with both BDM-based regimens resulted in negative lung cultures, whereas all H₁₀R₁₀Z₁₅₀E₁₀₀ treated mice were still culture positive. After 3 months of therapy 7% and 13% of mice relapsed receiving B₂₅D_0.6M₂₀₀U₂₀₀ and B₂₅D_0.6M₂₀₀, respectively, compared to 40% for H₁₀R₁₀Z₁₅₀E₁₀₀ treatment showing an increased sterilizing activity of both BDM-based regimens.

Conclusions

BDM-based regimens, with and without sutezolid, have a higher efficacy than the HRZE regimen in the BALB/c model of TB, with some improvement by adding sutezolid. By translating these results to TB patients, this novel BDMU regimen should be able to reduce treatment duration by 25% compared to HRZE therapy.

Introduction

Multi-drug resistant tuberculosis (MDR-TB) continues to pose a major global public health threat, with global targets increasingly out of reach.¹ The long treatment duration and treatment-related toxicity result in poor adherence. As a result, drug resistance is becoming more common and will have an increasing contribution to the TB pandemic.^2,3

Major impediments in MDR-TB control include inadequate drug susceptibility testing capacity, limited access to or delayed MDR-TB treatment, and lengthy and toxic treatments that cured only half of patients in the past.³ In 2022, the WHO has recommended the use of regimens composed of bedaquiline (B), pretomanid, linezolid and moxifloxacin (M) if susceptibility was demonstrated.⁴ However, linezolid at a dose of 600 mg, given for 26 weeks as recommended, still resulted in 24% of patients suffering peripheral neuropathy, and 2% suffering myelosuppression.⁵ The PanACEA consortium is testing a regimen with drugs of similar class and effectiveness [e.g. delamanid (D), a nitroimidazole, and sutezolid (U), an oxazolidinone], with sutezolid being a promising drug candidate to avert the linezolid main toxicities, while showing at least equal activity in murine models.⁶ In humans and mice, orally administered sutezolid is transformed into a sulfoxide metabolite which reaches higher exposures than the parent in both species.^6,7 The metabolite is active against extracellular Mycobacterium tuberculosis (Mtb) while the parent mainly kills intracellular bacteria.⁸

The evaluation of novel treatment regimens in TB patients requires a clear understanding of its exposure-response relationships. Pharmacokinetic confirmation of selected optimal doses in relation to their preclinical outcomes may prevent suboptimal exposures in efficacy studies and avoid inaccurate comparisons between regimen potencies in the translation from animals to man. Therefore, this study aimed to identify human pharmacokinetic equivalent doses before assessing antimycobacterial and sterilizing activity of the BDMU regimen in the standard mouse model (BALB/c) of pulmonary TB, to inform treatment duration in a clinical phase II trial.

Methods

Female, specific pathogen-free BALB/c mice were purchased from Charles River (Sulzfeld, Germany) or Janvier Labs (Le Genest Saint-Isle, France). Mtb-infected mice were maintained in individually ventilated cages (Ebeco, Castrop-Rauxel, Germany) under biosafety level 3 conditions at the Research Center Borstel, Germany. All animal experimentation was in accordance with the German Animal Protection Law and approved by the animal research ethics committee (Schleswig-Holstein, Germany) before getting permission by the Ministry of Energy, Agriculture, Environment, Nature and Digitalization (Kiel, Germany; permit 36-3/19). The health status of mice was monitored regularly, and moribund mice were killed humanely in accordance with the approved protocol.

For oral administration of drugs, isoniazid (H), rifampicin (R), pyrazinamide (Z) and ethambutol (E) were prepared as previously described⁹ and rifampicin was dosed at least 1 hour before the combination of the other drugs. Bedaquiline was formulated in acidified 20% hydroxypropyl-beta-cyclodextrin solution,¹⁰ while delamanid, moxifloxacin and sutezolid were pestled in a mortar and resuspended in 5% gum arabic solution. Drug stocks were stored at 4°C and mixed for co-administration on the day of application.

Two sets of pharmacokinetic pilot experiments were performed in 78 uninfected BALB/c mice in total to identify human pharmacokinetic equivalent doses. Mice received various BDMU regimens, namely B₂₅D_2.5M₂₀₀U₁₀₀ [bedaquiline (25 mg/kg), delamanid (2.5 mg/kg), moxifloxacin (200 mg/kg), sutezolid (100 mg/kg)] in pilot 1 (n = 24 mice) and B₂₅D_0.6M₂₀₀U₂₀₀ [bedaquiline (25 mg/kg), delamanid (0.6 mg/kg), moxifloxacin (200 mg/kg), sutezolid (200 mg/kg)] and B₂₅D_2.5M₂₀₀U₂₀₀ [bedaquiline (25 mg/kg), delamanid (2.5 mg/kg), moxifloxacin (200 mg/kg), sutezolid (200 mg/kg)] or delamanid_2.5 (2.5 mg/kg) monotherapy in pilot 2 (n = 54 mice). Mice received once daily dosing for 5 days via oral gavage, after which three animals were euthanized at each sampling time point.

Concentrations in mouse plasma were analysed by validated high-performance LC-MS/MS assays at Radboudumc (Nijmegen, the Netherlands). A cross-validation was conducted to confirm performance of a human plasma LC-MS/MS method for BALB/c plasma by comparing responses of five quality control concentrations in mouse plasma with the nominal concentration on a human plasma calibration curve.

Pharmacokinetic parameters were assessed using standard noncompartmental methods in Phoenix WinNonlin version 6.4 (Pharsight Corporation). To determine human pharmacokinetic equivalent doses, area under the curve (AUC) from time point 0 to 24 h (AUC_0-24h) in mice were compared to clinically relevant exposures in humans at steady state, namely AUC within the dosing interval (AUC_0-t) at (quasi) steady state. In literature, average AUC_0-t in TB patients taking sutezolid (1200 mg 1 ×  daily), bedaquiline (200 mg 3 ×  weekly), delamanid (100 mg 2 ×  daily) and moxifloxacin (400 mg 1 ×  daily) was 7.1,⁷ 98, 7.9 and 25 mg/L*h,^7,11–13 respectively. AUC_0-24h in mice was considered non-equivalent when outside the human ∼35%–50% CV range. We considered that protein binding of sutezolid, delamanid, bedaquiline, and moxifloxacin to be comparable in humans versus mice.^14–17

To investigate the antimycobacterial and sterilizing activity, mice were exposed to a low dose aerosol infection with the Mtb H37Rv strain, as previously described.¹⁸ Mice were sacrificed the following day to determine the number of colony-forming units (cfu) implanted in the lung, and 3 weeks later to determine the pulmonary bacterial burden at the start of treatment. Therapy was administered once daily by oral gavage 5 days a week. The pulmonary bacterial burden was assessed after 1, 2, 3 and 4 months of treatment by analysing five mice per regimen and time point and using charcoal agar⁶ for cfu plating. Culture-positive relapse was determined by holding cohorts of at least 14 mice for an additional 3 months after completion of treatment durations of 2, 3 and 4 months. The entire lung homogenate was plated onto Middlebrook 7H10 agar supplemented with 10% bovine serum and the detection of one or greater cfu were defined as culture-positive relapse.

Results

Cross-validation to confirm performance of our LC-MS/MS method for BALB/c plasma gave robust results. The accuracy of sutezolid, bedaquiline, delamanid and moxifloxacin quality control concentrations was between 97% and 101%, 102% and 111%, 87% and 98%, and 104% and 111%, respectively; the within-run CV was between 1.2% and 1.7%, 2.8% and 6.9%, 1.2% and 2.8%, and 2.7% and 4.3%, respectively. The first pilot experiments for human equivalent dose determination in uninfected mice resulted in sutezolid and main metabolite exposures being below the pre-set target, and delamanid exposures being too high (Table 1). After adjusting doses in pilot 2 for those two drugs to B₂₅D_0.6M₂₀₀U₂₀₀, exposures were found to be in the target range (Table 1).

Therapy of Mtb-infected mice started at a pulmonary bacterial burden of 6.59 log₁₀ cfu (Table 2). At completion of the first month of treatment, the standard H₁₀R₁₀Z₁₅₀E₁₀₀ regimen reduced the bacterial load by 2.89 log₁₀ cfu, whereas the B₂₅D_0.6M₂₀₀ and B₂₅D_0.6M₂₀₀U₂₀₀ regimen reduced the cfu counts by 4.04 and 4.35 log₁₀ cfu, respectively, so that both BDM-based regimens had a significant greater antimycobacterial activity than the H₁₀R₁₀Z₁₅₀E₁₀₀ group. The addition of sutezolid to the BDM regimen increased the bactericidal activity by 0.3 log₁₀ cfu. After 3 months of treatment, BDM-based regimens, with and without sutezolid, showed faster reduction of lung cfus and culture negativity, while in the H₁₀R₁₀Z₁₅₀E₁₀₀ group, all mice still showed culture growth (Table 2).

Relapse rates after 2 months of therapy were 100% for mice receiving H₁₀R₁₀Z₁₅₀E₁₀₀ or B₂₅D_0.6M₂₀₀ and 87% (13/15) for mice treated with the B₂₅D_0.6M₂₀₀U₂₀₀ regimen (Table 2). After 3 months of therapy, 13% (2/15) of mice relapsed receiving B₂₅D_0.6M₂₀₀ and 7% (1/15) receiving B₂₅D_0.6M₂₀₀U₂₀₀, compared to 40% (6/15) relapsing in the H₁₀R₁₀Z₁₅₀E₁₀₀ group. Based on the relapse rates observed at these timepoints of therapy, the addition of sutezolid to the regimen seemed to slightly increase the sterilizing activity. After 4 months of therapy, there were no relapses in any treatment group. Stopping sutezolid after 3 months of therapy did not result in a different relapse rate at 4 months (Table 2).

Discussion

This study aimed to identify human pharmacokinetic equivalent doses, bactericidal and sterilizing activity of the BDMU regimen, and the additive contribution of sutezolid in the BALB/c mouse model of TB. It was noteworthy that we measured a higher exposure to delamanid at a dose of 2.5 mg/kg, compared to other reports, which might be due to single versus multiple dosing or the co-administration of drugs.^19,20 The subsequent efficacy experiments revealed that BDM plus sutezolid had a better antimycobacterial and sterilizing activity than the HRZE regimen. The contribution of sutezolid to the regimen however seemed to be small, resulting in a slightly lower relapse rate at two time points.

The present study suggests that BDM represents a promising backbone for further validation in advanced mouse models¹⁸ that in contrast to BALB/c mice develop a human-like TB pathology. It is important to note that no human study so far was designed to assess the contribution of linezolid to a regimen, so this contribution could be equally small and possibly only result in preventing acquisition of resistance, an important event but usually too infrequent to be measured in clinical trials. Limitations of our report are that PK studies were conducted in uninfected mice and efficacy analyses in BALB/c mice that do not reflect the full range of human TB granulomas.

In summary, translating our results from mice to humans, BDMU should be able to shorten treatment by 25%; i.e. from 6 (HRZE) to 4–4.5 months in TB patients.

Acknowledgements

Bedaquiline fumarate was provided by the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH. The authors thank Johanna Volz for excellent technical assistance and the animal facility at the Research Center Borstel for taking care of the mice.

Funding

This work was supported by the German Center for Infection Research (DZIF) TTU 02.710, and 02.814. L.t.B., E.S. and N.H. were supported by PanACEA 2, which is part of the European and Developing Countries Clinical Trial Partnership (EDCTP) 2 programme supported by the European Union (grant number TRIA2015-1102-PanACEA2) with support from the Bundesministerium für Bildung und Forschung (BMBF; 01KA1701) further funding is contributed by the Staatssekretariat für Bildung, Forschung und Innovation (SERI) and Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO).

Transparency declarations

E.S. has received research funding from TB Alliance, the developer of pretomanid, and Janssen Pharmaceuticals, the developer of bedaquiline. N.H., E.S., L.t.B. and M.H. have received funding to their institutions from LegoChem for a clinical trial on delpazolid. All other authors: none to declare.

Disclaimer

None.

References

Author notes