Saskia E. Mudde (1), Rami Ayoun Alsoud (2), Aart van der Meijden (1), Anna M. Upton (3), Manisha U. Lotlikar (3), Ulrika S. H. Simonsson (2), Hannelore I. Bax (1,4), Jurriaan E. M. de Steenwinkel (1)
(1) Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Center, Rotterdam, the Netherlands, (2) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, (3) Global Alliance for Tuberculosis Drug Development, New York, New York, USA, (4) Department of Internal Medicine, Section of Infectious Diseases, Erasmus University Medical Center, Rotterdam, the Netherlands
Introduction: With 10 million new cases and 1.5 million deaths in 2018, tuberculosis remains a major global problem [1]. Given its persistently high global burden, effective and shorter treatment options are urgently needed. We explored the relationship between relapse and treatment length as well as inter-regimen differences for 2 novel anti-tuberculosis drug regimens using a mouse model of tuberculosis infection and pharmacometric modeling.
Methods: Mycobacterium tuberculosis–infected mice were treated for up to 13 weeks with bedaquiline and pretomanid combined with moxifloxacin and pyrazinamide (BPaMZ) or linezolid (BPaL). Cure rates were evaluated 12 weeks after treatment completion. Mice were considered to be cured when cultures of lung homogenates were negative for M. tuberculosis. The standard regimen of isoniazid, rifampicin, pyrazinamide, and ethambutol (HRZE) was evaluated as a comparator. The experimental BPaMZ and BPaL data, together with HRZE data from a previous study with the same experimental protocol [2], were used to build the pharmacometric model. Model development was performed in 2 steps. First, the relation between probability of cure and treatment duration, regardless of the drug regimen, was described using logistic regression. Different relations with respect to treatment duration were evaluated, including a linear model, an Emax model, and a sigmoidal Emax model. The second step evaluated whether the relation was significantly different between the three regimens. In a stepwise approach, various models were fitted to determine whether a given model parameter differed significantly between one regimen and the other two, and only models that significantly lowered the objective function value by >3.84 points were further evaluated in combinations until a final model was selected. Using the final model, the data set was bootstrapped, and model parameters were then re-estimated using 1000 resampled datasets from the observed data with replacement. From the resulting distribution of 1000 parameter estimates, the 95% probability of cure and 90% confidence intervals were predicted. The data were analyzed using NONMEM 7.4.3 [3]. Data management and graphical analysis were performed using R 3.6.3 software [4].
Results: Six weeks of BPaMZ treatment was sufficient to achieve cure in all mice. In contrast, 13 weeks of BPaL and 24 weeks of HRZE did not achieve 100% cure rates. The final pharmacometric model was a sigmoidal Emax relation regarding probability of failure and probability of cure in relation to treatment length and regimen. The model predicted that 95% probability of cure was reached after 1.6 months for BPaMZ, while this was 4.3 months for BPaL and 7.9 months for HRZE. This order of efficacy is consistent with other mouse tuberculosis studies, in which 1.5–2 months of BPaMZ-treatment was sufficient for cure, while 3 months were needed for BPaL [5], and ≥ 6 months for HRZE [6]. The present study confirms that the improved experimental setup and pharmacometric modeling, as we introduced elsewhere [7], can shed light on the relationship between treatment duration and treatment outcome, and facilitates efficient comparison between regimens.
Conclusions: This study provides additional evidence for the treatment-shortening capacity of BPaMZ over BPaL and HRZE. Our combined experimental-mathematical model approach provides guidance on treatment durations needed to reach certain cure rates. To optimally use preclinical data for predicting clinical outcomes, and to overcome the limitations that hamper such extrapolation, we advocate bundling of available published preclinical data into pharmacometric models.
References:
[1] World Health Organization. Global tuberculosis report 2019. Geneva, Switzerland: World Health Organization, 2019.
[2] Pieterman ED, Keutzer L, van der Meijden A, et al. Superior efficacy of a bedaquiline, delamanid and linezolid combination regimen in a mouse-TB model. J Infect Dis 2021.
[3] Beal SL, Sheiner LB, Boeckmann AJ, Bauer RJ, eds. NONMEM 7.4.3 Users Guides. (1989-2018). Gaithersburg, MD, USA: ICON Development Solutions; 2018.
[4] R Core Team. R: a language and environment for statistical computing; 2017.
[5] Xu J, Li SY, Almeida DV, et al. Contribution of pretomanid to novel regimens containing bedaquiline with either linezolid or moxifloxacin and pyrazinamide in murine models of tuberculosis. Antimicrob Agents Chemother 2019; 63:e00021–19.
[6] Mourik BC, de Knegt GJ, Verbon A, Mouton JW, Bax HI, de Steenwinkel JEM. Assessment of bactericidal drug activity and treatment outcome in a mouse tuberculosis model using a clinical Beijing strain. Antimicrob Agents Chemother 2017; 61:e00696–17.
[7] Mourik BC, Svensson RJ, de Knegt GJ, et al. Improving treatment outcome assessment in a mouse tuberculosis model. Sci Rep 2018; 8:5714.
Reference: PAGE 29 (2021) Abstr 9636 [www.page-meeting.org/?abstract=9636]
Poster: Drug/Disease Modelling - Infection