Analysis of Mycobacterium tuberculosis time to positivity in response to treatment in South African patients with drug-susceptible tuberculosis
Marie Wijk(1), Kamunkhwala Gausi(1), Samantha Malatesta(2), Laura F. White(2), Sarah E. Weber(3), Tara Carney(4), Bronwyn Myers(5), Lubbe Wiesner(1), C. Robert Horsburgh Jr(6), Richard Court(1), Robin M. Warren(7), Helen McIlleron(1), Paolo Denti(1), Frank Kloprogge(8), Karen R. Jacobson(3)
(1) Division of Clinical Pharmacology, Department of Medicine, University of Cape Town, Cape Town, South Africa (2) Department of Biostatistics, Boston University School of Public Health, Boston, MA, USA (3) Section of Infectious Diseases, Boston University School of Medicine and Boston Medical Centre, Boston, MA, USA (4) Alcohol, Tobacco and Other Drug Research Unit, South African Medical Research Council, Tygerberg, South Africa (5) Curtin enAble Institute, Curtin University, WA, Australia (6) Department of Epidemiology, Boston University School of Public Health, Boston, Massachusetts, USA (7) Department of Science and Innovation, National Research Foundation Centre of Excellence in Biomedical Tuberculosis Research, South Africa Medical Research Council for Tuberculosis Research, Stellenbosch University, Tygerberg, South Africa (8) Institute for Global Health, University College London, London, UK
Objectives: Tuberculosis is the leading cause of death from an infectious agent worldwide[1]. Tuberculosis burden is high in South Africa, where more than half of tuberculosis patients are living with HIV[1]. Although the treatment success rates are relatively high[1], tuberculosis mortality and post-tuberculosis sequelae are significant, meaning identifying factors impacting treatment outcome is imperative. Tuberculosis sterilization correlates with end of treatment outcome as well as relapse[2-4], and early sterilization may indicate improved outcomes[4,5]. The aim of this work was to describe mycobacterial clearance from sputum in South African patients with drug-susceptible pulmonary tuberculosis and identify influencing factors.
Methods: Pulmonary tuberculosis patients within an observational cohort in Worcester, South Africa, received first-line tuberculosis treatment with weight-based dosing (8-12 mg/kg rifampicin, 4-6 mg/kg isoniazid, 20-30 mg/kg pyrazinamide and 15-25 mg/kg ethambutol). Sputum samples were collected at baseline and weekly for 12 weeks of treatment. Sputum bacillary load was quantified using the growth detection system Mycobacterium tuberculosis (Mtb) growth indicator tube (MGIT), recording time to positivity (TTP), i.e. time from incubation in the MGIT to the sample turning positive, which is inversely related to the number of viable Mtb in sputum.[2] Samples that did not turn positive within 42 days of incubation in the MGIT were considered Mtb negative, the frequency of which typically increases as treatment progresses and sputum bacterial load decreases[6], and were treated as above the upper limit of quantification (ULOQ) in the analysis. Contaminated samples were excluded from the analysis. We used nonlinear mixed effects modelling in NONMEM to describe the relationship between TTP and time on treatment. Linear, broken-stick and exponential models were tested to fit the data, estimating baseline TTP and rate of increase of TTP with time on treatment. Model discrimination was based on objective function value (OFV), goodness-of-fit and visual predictive checks. Between-subject variability (BSV) was assumed to be log-normally distributed, and an error model with additive and proportional components was tested to describe the residual variability. The effect of HIV status, lung cavities and other patient characteristics were tested as covariates. Ethics approval was obtained by the South African Medical Research Council, Stellenbosch University Human Research Ethics Committee, and the Boston University Institutional Review.
Results: 305 participants were included, mostly male (60%), 27% living with HIV and 64% had lung cavities. Median age was 38 years (range 15-77), weight 47.1 kg (range 29.6-97.7) and BMI 17.7 kg/m2 (range 12.9-35.5). After discarding contaminated samples, 2364 samples were available, out of which 49% were ULOQ. We developed a model parameterized in terms of baseline TTP and increase in TTP with time on treatment, with a proportional error model describing the residual variability. ULOQ data was censored using the M3 method in combination with Laplacian estimation[7]. The typical values for baseline TTP and rate of increase of TTP with time on treatment were estimated to 10 days and 0.42 per treatment week, respectively. The distribution of BSV for baseline TTP was box-cox transformed. Participants with lung cavities had 32% shorter baseline TTP (dOFV=-48.6, p<0.001), i.e. higher baseline bacterial load, and 35% slower rate of increase of TTP with time on treatment (dOFV=-14.9, p<0.001). Persons with HIV had 12% longer baseline TTP than HIV negative participants (dOFV=-4.08, p = 0.04), i.e. lower baseline bacterial load.
Conclusions: We developed a model parameterized in terms of baseline TTP and increase in TTP with time on treatment, which has been reported previously[8,9]. In line with previous reports in literature[10-12], participants with lung cavitation had higher bacterial load at baseline and slower rate of increase of TTP with time on treatment compared to participants without lung cavitation, and participants living with HIV had a lower bacterial load at baseline compared to participants not living with HIV.
References:
[1] World Health Organization. Global Tuberculosis Report 2022; World Health Organization: Geneva, Switzerland, 2022.
[2] Wallis, R. S., Doherty, T. M., Onyebujoh, P., Vahedi, M., Laang, H., Olesen, O., ... & Zumla, A. (2009). Biomarkers for tuberculosis disease activity, cure, and relapse. The Lancet infectious diseases, 9(3), 162-172.
[3] Kloprogge, F., Mwandumba, H. C., Banda, G., Kamdolozi, M., Shani, D., Corbett, E. L., ... & Sloan, D. J. (2020, July). Longitudinal pharmacokinetic-pharmacodynamic biomarkers correlate with treatment outcome in drug-sensitive pulmonary tuberculosis: a population pharmacokinetic-pharmacodynamic analysis. In Open forum infectious diseases (Vol. 7, No. 7, p. ofaa218). US: Oxford University Press.
[4] Neesha Rockwood, Elsa du Bruyn, Thomas Morris & Robert J Wilkinson (2016) Assessment of treatment response in tuberculosis, Expert Review of Respiratory Medicine, 10:6, 643-654, DOI: 10.1586/17476348.2016.1166960
[5] Sloan, D. J., Mwandumba, H. C., Garton, N. J., Khoo, S. H., Butterworth, A. E., Allain, T. J., ... & Davies, G. R. (2015). Pharmacodynamic modeling of bacillary elimination rates and detection of bacterial lipid bodies in sputum to predict and understand outcomes in treatment of pulmonary tuberculosis. Clinical infectious diseases, 61(1), 1-8.
[6] Tortoli E, Cichero P, Piersimoni C, Simonetti MT, Gesu G, Nista D. Use of BACTEC MGIT 960 for recovery of mycobacteria from clinical specimens: multicenter study. J Clin Microbiol. 1999 Nov;37(11):3578-82. doi: 10.1128/JCM.37.11.3578-3582.1999. PMID: 10523555; PMCID: PMC85696.
[7] Beal, S. L. (2001). Ways to fit a PK model with some data below the quantification limit. Journal of pharmacokinetics and pharmacodynamics, 28(5), 481.
[8] Gausi, K., Ignatius, E. H., Sun, X., Kim, S., Moran, L., Wiesner, L., ... & Denti, P. (2021). A semimechanistic model of the bactericidal activity of high-dose isoniazid against multidrug-resistant tuberculosis: results from a randomized clinical trial. American journal of respiratory and critical care medicine, 204(11), 1327-1335.
[9] Deshpande, D., Srivastava, S., Nuermberger, E., Pasipanodya, J. G., Swaminathan, S., & Gumbo, T. (4016). A faropenem, linezolid, and moxifloxacin regimen for both drug-susceptible and multidrug-resistant tuberculosis in children: FLAME path on the Milky Way. Clinical Infectious Diseases, 63(suppl_3), S95-S101.
[10] McCallum, A. D., Pertinez, H. E., Chirambo, A. P., Sheha, I., Chasweka, M., Malamba, R., ... & Mwandumba, H. C. (2022). High intrapulmonary rifampicin and isoniazid concentrations are associated with rapid sputum bacillary clearance in patients with pulmonary tuberculosis. Clinical Infectious Diseases, 75(9), 1520-1528.
[11] Chigutsa, E., Patel, K., Denti, P., Visser, M., Maartens, G., Kirkpatrick, C. M., ... & Karlsson, M. O. (2013). A time-to-event pharmacodynamic model describing treatment response in patients with pulmonary tuberculosis using days to positivity in automated liquid mycobacterial culture. Antimicrobial agents and chemotherapy, 57(2), 789-795.
[12] Beynon, F., Theron, G., Respeito, D., Mambuque, E., Saavedra, B., Bulo, H., ... & Garcia-Basteiro, A. L. (2018). Correlation of Xpert MTB/RIF with measures to assess Mycobacterium tuberculosis bacillary burden in high HIV burden areas of Southern Africa. Scientific reports, 8(1), 1-9.