Qianwen Wang (1), Daniel Everitt (2), Rada Savic (1)
(1) University of California San Francisco, (2) TB Alliance
Objectives: Tuberculosis (TB) is the leading cause of death from a single infectious agent worldwide1. The fact that the lengthy 6-month standard regimen has not changed over decades calls for great urgency of new regimens development for shorter treatment. The combination of Pa-824 (Pa), moxifloxacin (M) and pyrazinamide (Z) showed powerful bactericidal and sterilizing activity in mice, marking the first time a regimen without rifampin or isoniazid that has prevented relapse more effectively than the first line regimen2. PaMZ then became the first novel regimen tested in human which showed superior effect over HRZE regimen in phase 2a3 & 2b study4. However, PaMZ didn’t achieve non-inferiority in a phase 3 study5. Although PaMZ trial is discontinued, only Pa has been studied from a perspective of exposure-response6,7, leaving many questions for this combination not answered. In our current study, we aim to analyze 1) population pharmacokinetics (PK) of M and Z; 2) baseline and drug exposure predictors on time to culture conversion (TTCC) in patients receiving PaMZ regimen.
Methods: For PK analysis, 1273 trough and 189 intensive PK samples were collected for both M and Z from NC-002 and NC-006 trials (n=345 subjects). M were received at 400 mg daily whilst Z were received at 1500 mg daily for PaMZ arm and at 22~28 mg/kg for HRZE arm. Population PK modeling was performed using NONMEM (version 7.4). Various structure models with different absorption process, including gastrointestinal transit compartment and absorption models with or without lag time were tested. Covariate relationships were then investigated. The final PK model was validated by bootstrap, visual predictive check and goodness of fit plots.
For risk factor analysis, patients from NC-002 and NC-006 were included (n=477 subjects). Baseline predictors that were missing less than 10% of participants: sex, age, race, BMI, smear grade, cavitation on chest radiograph were included. Drug exposure (AUC0-24, Cmax) at steady state for Pa were simulated from a published PK model7. Drug exposure of M and Z were simulated from developed model in previous step. MIC at baseline for Pa and M were reported in 277 subjects. Predictors including AUC/MIC, Cmax/MIC and Time over MIC (T/MIC) were included as continuous variables with degree of change at 100, 100 and 10%, respectively. Median value of 0.05 and 0.124 mg/L were used for Pa and M for patients with missing MIC. Multivariate Cox proportional hazard (Cox PH) analysis were used to identify risk factors for TTCC. Model selection started with a full model (with all predictors) and followed by a backward (p>0.05 as cut-off) then a forward stepwise approach (p<0.025 to include). The final Cox PH model assumption were tested using Schoenfeld residuals.
Results: A two-compartment model best described PK of M. Patients with HIV were found to have 44.5% higher clearance (CL, RSE 18%). This could be due to drug-drug interaction of efavirenz-based antiretroviral therapies. In comparison to black patients, bioavailability (F1) decreased by 24.6% (RSE 27%) in other mixed ethnic groups. Patients with bi-lateral cavities had 19.3% decrease (RSE 41%) and patients with no cavities had 8.8% (RSE 65%) increase in volume of distribution when comparing to patients with unilateral cavity.
A one-compartment model with delayed absorption was the best PK model for Z. F1 of HRZE decreased 21.6% (RSE 24%) in comparison to PaMZ. The impact of weight on CL showed minimal clinical relevance: with every 10 kg of increase in body weight, CL will increase by 4.8% (RSE 38%), therefore not warranting weight-based dosing.
In all participants, AUC/MIC of Pa, T/MIC of M and age were found to be significant risk factor decreasing TTCC with a hazard ratio (HR) of 1.27 (95% CI, 1.04~1.55), 1.04 (1.02~1.07) and 0.85 (0.77~0.94). The higher the HR, the easier to achieve culture conversion. In the sub-population where MIC is reported, only AUC/MIC of Pa and age were found to be major predictor for TTCC decrease with an HR of 1.22 (1.03~1.44), and 0.86 (0.76~0.94), respectively.
Conclusions: The first population PK models for moxifloxacin and pyrazinamide base on large dataset of non-rifampin regimen were developed. AUC/MIC of Pa and age were found to be significant predictors for culture conversion in TB patients receiving PaMZ regimens rather than Pa exposure alone.
References:
[1] WHO. Global Tuberculosis Report. 2020. (2020).
[2] Nuermberger, E. et al. Powerful bactericidal and sterilizing activity of a regimen containing PA-824, moxifloxacin, and pyrazinamide in a murine model of tuberculosis. Antimicrobial agents and chemotherapy 52, 1522-1524 (2008).
[3] Diacon, A. H. et al. 14-day bactericidal activity of PA-824, bedaquiline, pyrazinamide, and moxifloxacin combinations: a randomised trial. The Lancet 380, 986-993 (2012).
[4] Dawson, R. et al. Efficiency and safety of the combination of moxifloxacin, pretomanid (PA-824), and pyrazinamide during the first 8 weeks of antituberculosis treatment: a phase 2b, open-label, partly randomised trial in patients with drug-susceptible or drug-resistant pulmonary tuberculosis. The Lancet 385, 1738-1747 (2015).
[5] Tweed, C. et al. A partially randomised trial of pretomanid, moxifloxacin and pyrazinamide for pulmonary TB. The International Journal of Tuberculosis and Lung Disease 25, 305-314 (2021).
[6] Nedelman, J. R. et al. The Clinical Dose of Pretomanid: An Exposure-Response Perspective. Antimicrobial Agents and Chemotherapy (2020).
[7] Salinger, D. H., Subramoney, V., Everitt, D. & Nedelman, J. R. Population pharmacokinetics of the antituberculosis agent pretomanid. Antimicrobial agents and chemotherapy 63 (2019).
Reference: PAGE 29 (2021) Abstr 9847 [www.page-meeting.org/?abstract=9847]
Poster: Drug/Disease Modelling - Infection