WHO operational handbook on tuberculosis: module 4: treatment: drug-susceptible tuberculosis treatment: web annex 1: clinical outcomes and pharmacokinetics of first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis

WHO operational handbook on tuberculosis Web annex 1. Clinical outcomes and pharmacokinetics of first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis Module 4: Treatment Drug-susceptible tuberculosis treatment

Clinical outcomes and pharmacokinetics of first-line drugs  (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible  tuberculosis: systematic review and meta-analysis

Clinical outcomes and pharmacokinetics of first-line  drugs (rifampicin, isoniazid,  ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis:  systematic review and meta-analysis SUMMARY REPORT Authors Belén P. Solans, Agathe Béranger, Kendra Radtke, Ali Mohamed, Payam Nahid, Rada Savic University of California San Francisco, UCSF Center for Tuberculosis, USA This work was done in collaboration with the Global TB Programme of the World Health Organization. WHO operational handbook on tuberculosis. Module 4: treatment - drug-susceptible tuberculosis treatment. Web Annex 1. Clinical outcomes and pharmacokinetics of first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis/ Belén P Solans, Agathe Béranger, Kendra Radtke, Ali Mohamed, Payam Nahid, Rada Savic ISBN 978-92-4-005078-5 (electronic version) © World Health Organization 2022 Some rights reserved. This work is available under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo). Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation: “This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition”. Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization. (http://www.wipo.int/amc/en/mediation/rules/) Suggested citation. Solans BP, Béranger A, Radtke K, Mohamed A, Nahid P, Savic R. Web Annex 1. Clinical outcomes and pharmacokinetics of first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis. In: WHO operational handbook on tuberculosis. Module 4: treatment - drug-susceptible tuberculosis treatment. Geneva: World Health Organization; 2022. Licence: CC BY-NC-SA 3.0 IGO. Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris. Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To submit requests for commercial use and queries on rights and licensing, see https://www.who.int/copyright. Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned component in the work rests solely with the user. General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by WHO to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising from its use. The named authors alone are responsible for the views expressed in this publication. This publication forms part of the WHO document entitled WHO operational handbook on tuberculosis. Module 4: treatment - drug-susceptible tuberculosis treatment. It is being made publicly available for transparency purposes and information. 1Contents Acknowledgements 2 Background and rationale 3 Purpose and objectives 3 Methods 3 Findings 4 Conclusion 6 References 7 Annex 1 Clinical outcomes among children (< 18 years) treated for drug- susceptible tuberculosis with first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide): systematic review Abbreviations and acronyms 12 1. Introduction 13 2. Methods 15 3. Results 17 4. Discussion 25 5. Conclusion 27 References 28 Appendix 1. Search strategy 30 Appendix 2. Quality of evidence scoring outcome analysis 31 Annex 2 Pharmacokinetics of first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug- susceptible tuberculosis: systematic review and meta-analysis Abbreviations and acronyms 36 1. Introduction 37 2. Methods 39 3. Results 41 4. Discussion 72 5. Conclusion 75 References 76 Appendix 1. Search strategy 80 Appendix 2. Scoring the quality of evidence – pharmacokinetics target analysis 81 Appendix 3. Data extraction form 83 Appendix 4. Exposure to drugs (AUC and Cmax) as reported in the studies 84 2Clinical outcomes and pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis S U M M A R Y R E P O R T Acknowledgements The production and writing of this document – Clinical outcomes and pharmacokinetics of first-line drugs (rifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug- susceptible tuberculosis: systematic review and meta-analysis – was conducted by University of California San Francisco, UCSF Center for Tuberculosis and the groups of experts: Belén P. Solans, Agathe Béranger, Kendra Radtke, Ali Mohamed, Payam Nahid with the overall guidance of Rada Savic. The work was coordinated by the WHO Global TB Programme. 3Clinical outcomes and pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis S U M M A R Y R E P O R TBackground and rationale In 2021, the WHO Global Tuberculosis Programme initiated systematic reviews in order to collect and assess the evidence on whether dosing of rifampicin (RIF), isoniazid (INH), ethambutol (EMB) and pyrazinamide (PZA) should be further optimized to safely increase the likelihood of treatment success and reduce unfavorable treatment outcomes in tuberculosis (TB) patients on first-line treatment. The Global Tuberculosis Programme issued a call of interest for proposals for assessments, and the University of Liverpool (United Kingdom) and the University of California San Francisco Centre for Tuberculosis (USA) were selected to conduct systematic reviews of studies in adults and children (< 18 years), respectively. This report summarizes the work carried out at the University of California San Francisco Centre for Tuberculosis. Purpose and objectives The purpose of the systematic review was to assess the efficacy of doses higher than those currently recommended by WHO for each of the first-line anti-TB drugs (RIF, INH, PZA, EMB) used in a combination regimen for treating children (< 18 years) with presumed drug-susceptible TB. The objectives were to: • establish a protocol for conducting a series of systematic reviews to collect current evidence; • assess the efficacy of doses higher than those currently recommended by WHO for each of the first- line drugs (RIF, INH, PZA, EBM); and • compare the pharmacokinetics of doses (maximum concentration (Cmax) and area under the concentration–time curve (AUC)) higher than those currently recommended by WHO with that of established targets for each first-line drug. Methods Studies were identified in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (1). The search included all clinical, observational and descriptive studies as well as randomized controlled trials involving children and adolescents < 18 years being treated for confirmed or presumed drug-susceptible TB. The search was conducted in Pubmed, EMBASE, CENTRAL (Cochrane central register of controlled trials), the Cochrane Infectious Diseases Group Clinical trials register, the WHO International Clinical Trials Registry and the Clinicaltrials.gov electronic databases (2). Independent reviewers screened the titles and abstracts for relevance and appraised full texts for inclusion in the meta-analysis according to specified selection criteria. Relevant articles were identified by consensus with a third and a fourth reviewer. The quality of the evidence and bias were assessed for all the studies included. Two authors independently extracted all data of interest from the full texts. In cases of discrepancy, a consensus was found among the authors, who also synthesized the data. 4Clinical outcomes and pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis S U M M A R Y R E P O R T For the meta-analysis, the pharmacokinetics of exposure to a drug, such as the AUC and Cmax, were extracted when available. Summary estimates for AUC and Cmax were obtained by meta-analysis with the “metafor” package, version 2-4.0 in R. Heterogeneity was assessed by the I2 statistic and visual inspection of forest plots. Meta-regression was performed with key covariates to assess their impact on inter-study heterogeneity (3–6). Findings The findings of the systematic review and meta-analysis for clinical outcomes and for pharmacokinetics are described separately. ➜  Review of clinical outcomes of first-line drugs  (rifampicin, isoniazid, ethambutol and pyrazinamide) The systematic literature search identified 304 studies, of which 104 were duplicates and 154 did not meet the eligibility criteria. The 46 studies that met the inclusion criteria were included for full-text review; 34 were excluded because no clinical outcomes were reported. Only 12 studies were included in the review. The review showed that no research has been conducted in children treated at higher doses than those recommended by WHO. All the drugs were dosed according to body weight, and no distinction was made by age when administering any of the drugs, although malnutrition has been identified as a significant factor in treatment of TB. In general, the studies were highly heterogeneous with respect to patient characteristics (e.g., age, nutritional status), disease status and co-morbidities (e.g., HIV status), and most were done with small cohorts of children. Only five studies had more than 100 patients. As unfavourable outcomes were reported in about 13% of the studies, a much larger cohort would be necessary to identify significant predictors of unfavourable outcomes, especially with a population such as children, who differ widely. The conclusion of the systematic review was that the efficacy of doses higher than those recommended by WHO could not be performed because of lack of data. More detailed analysis is necessary of the exposure achieved with different doses of the drugs. ➜  Review of pharmacokinetics, including Cmax and AUC The search retrieved 304 studies, of which 104 were duplicates and 154 did not meet the selection criteria. Finally, 18 studies with a total of 963 patients were included, in which the pharmacokinetics of INH were evaluated in 16 (89%), RIF in 14 (78%), PZA in 13 (72%) and EMB in 8 (44%) studies. In general, pharmacokinetics were reported differently among the studies and for different drugs. AUC was reported for the full cohort in 11 studies on RIF (73.3%), 10 on INH (66.7%), 9 on PZA (69.2%) and 6 on EMB (66.7%). An additional study reported on Cmax but not AUC for all drugs. The remaining studies reported AUC and Cmax only by subgroup. Only 8 studies reported pharmacokinetics by HIV status, 4 by age and 3 by nutritional status, for RIF. None of the studies assessed pharmacokinetics by weight. All the studies reported pharmacokinetics for Cmax or the concentration after 2 h (C2h), and 17 (94.4%) reported AUC, from AUC0–4 h to AUC0–∞ according to the following distribution: AUC0–4 (n=2), AUC0–5 5Clinical outcomes and pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis S U M M A R Y R E P O R T(n=1), AUC0–6 (n=1), AUC0–8 (n=5), AUC0–12 (n=1), AUC0–24 (n=6) and AUC0–∞ (n=1). The time interval for each study was recorded. AUC0–4 or greater was considered as representative of the AUC for all drugs. Data reported as Cmax or C2h after dose was considered as representative of Cmax for all drugs. The subgroup analysis for each of the four TB drugs was conducted according to the above parameters. Rifampicin The summary estimate for : – RIF AUC was 23.4 µg.h/mL, significantly lower (P < 0.01) than the target of 38.7 µg.h/mL (13 studies). – RIF Cmax was 6 µg/mL, significantly lower (P < 0.01) than the target of 8 µg/mL (14 studies). Most of the studies reported consistently lower exposures than the target for this drug. The meta-analysis showed a significantly lower AUC (P = 0.03) in younger children (≤ 5 years) than older children, and the Cmax tended to be lower in these children but was not significant in the subgroup analysis (7–11). The meta-analysis showed a trend for HIV-positive children to have lower exposure than HIV-negative children (P = 0.07), and the relation was reported to be significant in one study (3). Younger age was also associated with significantly lower exposure in one study, and malnutrition was associated with lower exposure to RIF in two studies (7,11). Isoniazid The summary estimate for : – INH AUC was 23.4 µg.h/mL [95% CI 18.6-28.2] compared to the target of 23.4 µg.h/mL (11 studies) – INH Cmax was 5.6 µg/mL, while the target was 3–5 µg/mL (11 studies). The meta-analysis showed significantly lower (P < 0.05) AUC and Cmax in fast than in slow metabolizers, and the values were also lower than the target exposure (2, 12, 13). Pyrazinamide The summary estimate for : – PZA AUC was 201.2 µg.h/mL, significantly lower (P < 0.01) than the target of 238–428 µg.h/mL (14 studies) – PZA Cmax was 39.6 µg/mL, while the target was 35–60 µg/mL (14 studies) Most of the studies reported consistently lower exposures of children to this drug than the target. No significant association was identified in the subgroup meta-analysis for PZA; however, exposure was lower in HIV-positive patients in two studies, and younger age was a significant covariate for lower exposure in three publications (14). 6Clinical outcomes and pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis S U M M A R Y R E P O R T Ethambutol The summary estimate for : – EMB AUC was 7.2 µg.h/mL, significantly lower (P < 0.01) than the target of 16–28 µg.h/mL (4 studies) – EMB Cmax was 1.4 µg/mL, significantly lower (P < 0.01) than the target of 2–6 µg/mL (4 studies) All the studies reported consistently lower exposures in children than the exposure target for EMB. No subgroup meta-analysis was conducted for this drug because of lack of data (15–17). Conclusion The review found a paucity of research on doses higher than those recommended by WHO for treatment of TB in children. Furthermore, reporting of pharmacokinetics was inconsistent among studies, the populations were heterogenous, and the samples were small. Nevertheless, exposures to RIF, PZA and EMB were consistently lower in children than in adults. Exposure to INH appeared to be within the target range, but fast metabolizers had significantly lower pharmacokinetics than slow metabolizers, which was also lower than the target. The review indicates that younger children (≤ 5 years) and perhaps those with HIV infection may require doses of RIF higher than those recommended by WHO and higher than those for older children and HIV-negative children. None of the studies, however, reported a relation between pharmacokinetics and the safety of doses higher than those recommended. Whereas the WHO-recommended doses of RIF, INH, EMB and PZA for children (< 18 years) treated for drug-susceptible TB remain valid, this report indicates that these doses consistently result in exposure lower than the target. Use of higher doses of the four first-line TB drugs in children should therefore be studied further. More robust methods, such as individual participants data meta-analysis with population pharmacokinetics modelling, should be used to obtain evidence for recommending appropriate anti-TB drug doses for optimal treatment of all children, including those at high risk. For the detailed systematic reviews and meta-analysis of clinical outcomes and pharmacokinetics, see Annex 1 and 2, respectively. 7Clinical outcomes and pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis S U M M A R Y R E P O R TReferences 1. Moher D, Liberati A, Tetzlaff J, Altman D, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 2009;6(7):e1000097. 2. Ridge A, Grzemska M, Hill S, Gie R. Rapid advice: Treatment of tuberculosis in children. Geneva: World Health Organization; 2010 (http://whqlibdoc.who.int/publications/2010/9789241500449_eng.pdf, accessed 14 May 2021). 3. Seden K, Gibbons S, Marzolini C, Schapiro JM, Burger DM, Back DJ et al. Development of an evidence evaluation and synthesis system for drug–drug interactions, and its application to a systematic review of HIV and malaria co-infection. PLoS One. 2017;12(3):e0173509. 4. Jacobs TG, Svensson EM, Musiime V, Rojo P, Dooley KE, McIlleron H et al. Pharmacokinetics of antiretroviral and tuberculosis drugs in children with HIV/TB co-infection: a systematic review. J Antimicrob Chemother. 2020;75(12):3433–57. 5. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14(1):135. 6. Mukherjee A, Velpandian T, Singla M, Kanhiya K, Kabra SK, Lodha R. Pharmacokinetics of isoniazid, rifampicin, pyrazinamide and ethambutol in Indian children. BMC Infect Dis. 2015;15(1):126. 7. Justine M, Yeconia A, Nicodemu I, Augustino D, Gratz J, Mduma E et al. Pharmacokinetics of first-line drugs among children with tuberculosis in rural Tanzania. J Pediatr Infect Dis Soc. 2020;9(1):14–20 8. Schaaf HS, Willemse M, Cilliers K, Labadarios D, Maritz JS, Hussey GD et al. Rifampin pharmacokinetics in children, with and without human immunodeficiency virus infection, hospitalized for the management of severe forms of tuberculosis. BMC Med. 2009;7(1):19. 9. Mlotha R, Waterhouse D, Dzinjalamala F, Ardrey A, Molyneux E, Davies GR et al. Pharmacokinetics of anti-TB drugs in Malawian children: Reconsidering the role of ethambutol. J Antimicrob Chemother. 2015;70(6):1798–803. 10. Hiruy H, Rogers Z, Mbowane C, Adamson J, Ngotho L, Karim F et al. Subtherapeutic concentrations of first-line anti-TB drugs in South African children treated according to current guidelines: The PHATISA study. J Antimicrob Chemother. 2014;dku478. 11. Ramachandran G, Kumar AKH, Bhavani PK, Kannan T, Kumar SR, Gangadevi NP et al. Pharmacokinetics of first-line antituberculosis drugs in HIV-infected children with tuberculosis treated with intermittent regimens in India. Antimicrob Agents Chemother. 2015;59(2):1162–7. 12. Wilkins JJ, Langdon G, McIlleron H, Pillai G, Smith PJ, Simonsson USH. Variability in the population pharmacokinetics of isoniazid in South African tuberculosis patients: Population pharmacokinetics of isoniazid. Br J Clin Pharmacol. 2011;72(1):51–62. 13. Thee S, Seddon JA, Donald PR, Seifart HI, Werely CJ, Hesseling AC et al. Pharmacokinetics of isoniazid, rifampin, and pyrazinamide in children younger than two years of age with tuberculosis: Evidence for implementation of revised World Health Organization recommendations. Antimicrob Agents Chemother. 2011;55(12):5560–7. 14. Rangari GM, Roy V, Sethi GR, Mishra TK, Khanna A. Blood levels of isoniazid in Indian children with tuberculosis. Indian J Tuberc. 2015;62(2):80–5. 15. Bekker A, Schaaf HS, Draper HR, van der Laan L, Murray S, Wiesner L et al. Pharmacokinetics of rifampin, isoniazid, pyrazinamide, and ethambutol in infants dosed according to revised WHO-recommended treatment guidelines. Antimicrob Agents Chemother. 2016;60(4):2171–9. 16. Ranjalkar J, Mathew SK, Verghese VP, Bose A, Rose W, Gupta D et al. Isoniazid and rifampicin concentrations in children with tuberculosis with either a daily or intermittent regimen: Implications for the revised RNTCP 2012 doses in India. Int J Antimicrob Agents. 2018;51(5):663–9. 17. Horita Y, Alsultan A, Kwara A, Antwi S, Enimil A, Ortsin A et al. Evaluation of the adequacy of WHO revised dosages of the first-line antituberculosis drugs in children with tuberculosis using population pharmacokinetic modeling and simulations. Antimicrob Agents Chemother. 2018;62(9):e00008-18 (/aac/62/9/e00008-18.atom).

Clinical outcomes among children  (< 18 years) treated for drug-susceptible tuberculosis  with first-line drugs (rifampicin,  isoniazid, ethambutol and pyrazinamide): systematic review ANNEX 1 Authors Belén P. Solans, Agathe Béranger, Ali Mohamed, Kendra Radtke, Payam Nahid, Rada Savic University of California San Francisco, UCSF Center for Tuberculosis, USA This work was done in collaboration with the Global TB Programme of the World Health Organization.

11 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1Contents Abbreviations and acronyms 12 1. Introduction 13 2. Methods 15 2.1 Search strategy and selection criteria 15 2.2 Assessment of risk of bias 15 2.3 Data extraction 15 2.4 Summary measures of the outcomes 16 2.5 Summary measures of adverse events 16 3. Results 17 3.1 Search 17 3.2 The studies 17 3.3 Summary of treatment outcomes 22 3.4 Predictors of unfavourable outcomes 23 3.5 Adverse events 24 4. Discussion 25 5. Conclusion 27 References 28 Appendix 1. Search strategy 30 Appendix 2. Quality of evidence scoring outcome analysis 31 12 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 Abbreviations and acronyms AUC area under the concentration-time curve EPTB extrapulmonary tuberculosis EMB ethambutol INH isoniazid LTFU lost to follow-up PK pharmacokinetics PRISMA Preferred Reporting Items for Systematic Reviews and Meta-analyses PTB pulmonary tuberculosis PZA pyrazinamide RIF rifampicin RNTCP Revised National Tuberculosis Control Programme (India) TB tuberculosis TBM tuberculous meningitis 13 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 11. Introduction Tuberculosis (TB) is a communicable disease that is a major cause of ill health and one of the leading causes of death worldwide. Until the coronavirus (COVID-19) pandemic, TB was the leading cause of death from a single infectious agent, ranking above HIV/AIDS, with an estimated 10 million (8.9–11.0 million) cases and 1.2 million deaths worldwide in 2019, of which about 20% were children < 15 years of age (1). An estimated 1.2 million TB deaths occurred among HIV-negative people in 2019 (a reduction from 1.7 million in 2000) and 208 000 (177 000–242 000) among HIV-positive people (a reduction from 678 000 in 2000). Children < 15 years accounted for 12% of the people who developed TB in 2019. While almost 90% of this burden is in 30 low- and middle-income countries, a significant fraction can be attributed to risk factors for TB disease, such as young age, HIV co-infection, malnutrition and diabetes. Progress has been made in the control of TB, but the current annual decrease in incidence does not exceed 2%. Globally, TB was reported to have been newly diagnosed and notified in 7.1 million people in 2019. In 2018 and 2019, 14.1 million people with TB were treated, with a treatment success rate of 85% in 2018, similar to that of children (84%) (2). The worse TB treatment outcomes are for young people with HIV-coinfection and malnutrition. Although pulmonary disease is the commonest presentation of TB, extrapulmonary and disseminated disease, particularly neurological involvement, are common in young children and are associated with worse outcomes (3). First-line regimens for TB have undergone little change since their introduction more than 40 years ago. Addition of rifampicin (RIF) and pyrazinamide (PZA) to isoniazid (INH), which allowed the duration of therapy to be reduced to 6 months, established this trio of drugs as the backbone of effective therapy (4). A fourth drug was often added to prevent the emergence of resistance, particularly in people already harbouring INH-resistant strains. While streptomycin was initially widely used for this purpose, widespread resistance and a preference for parenteral administration led to its replacement by the oral agent ethambutol (EMB), resulting in the current standard first-line regimen (5). All four of the drugs that comprise the current first-line regimen were developed before 1970, when expectations and standards in drug development programmes were different from those today. In particular, justification for and exploration of dosing strategies were often informal and incomplete, and post-marketing studies were often conducted to demonstrate the risks and benefits of the doses used in practice (6). The accepted doses of the four drugs (see Table 1) have recently been re-evaluated from modern preclinical and clinical pharmacokinetics (PK) and pharmacodynamics data and, in some cases, early-phase clinical trials, establishing proof-of-concept that higher doses can improve efficacy (7). It has not, however, been established whether the possible benefits improve long-term outcomes and can be achieved without additional toxicity. A recommended method for defining optimal dosing in children is based on adult–paediatric exposure equivalence, whereby the dosing algorithm ensures that paediatric patients achieve exposure similar to that of adults. The main assumption underlying this method is that exposure–response and exposure– toxicity relations are comparable in adults and paediatric patients in the same clinical context (8,9). If this assumption holds true and PK exposures comparable to those in adults are achieved in children, similar treatment outcomes and adverse event rates are expected to be observed, including in children at higher risk of unfavourable outcomes. 14 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 Newer PK studies in children with the revised dosing recommendations show that their exposure to first-line TB drugs is often lower than that of adults (10–12). Importantly, the weight-based dose recommendations (in mg/kg body weight) are uniform for all children weighing < 25 kg given fixed-dose combinations adapted to weight. Modelling has predicted that higher doses are necessary to ensure the same exposure of low-weight, malnourished (weight-for-age z score) children as of adults (13–16). This prediction is supported by basic principles of developmental pharmacology, in which higher doses are required for young, low-weight children to account for higher clearance per kilogram of body weight (17). NAT2 acetylator status is the primary determinant of exposure to INH, with twice the clearance in fast as in slow acetylators (18). Outcomes and safety according to NAT2 status are not, however, well characterized, especially in children. Synthesis of the evidence on dose, exposure and outcome (success, failure, relapse and adverse effects) from different studies and regions is therefore an important next step for determining the optimal dosages of first-line TB medicines. Ensuring maximal efficacy of the four first-line anti-TB drug regimens could improve the long-term outcomes of all patients and also make the regimen more robust to variations in adherence, PK and pharmacogenetics and to the emergence of resistance. Furthermore, intensification of treatment, with higher doses, could be important for people with severe or disseminated disease or for vulnerable populations with greater PK variation or drug–drug interactions. Investigation of the possible impact of malnutrition on drug concentrations and outcomes and of dosing other than according to weight bands will be of particular importance for children (14). The objective of this review was to assess the efficacy of doses higher than those currently recommended by WHO (Table 1) for each of the first-line anti-TB drugs (RIF, INH, PZA, EMB) used in a combination regimen for treating children with presumed drug-susceptible TB. Table 1. Current WHO-recommended doses of first-line anti-TB drugs for children weighing < 25 kga Drug Daily dose in 2010 (mg/kg) Daily dose (range) in 2014 (mg/kg) RIF 10–20 15 (10–20) INH 10–15 10 (7–15) PZA 30–40 35 (30–40) EMB 15–25 20 (15–25) From references 19 and 20 RIF, rifampicin; INH, isoniazid; PZA, pyrazinamide; EMB, ethambutol a As children approach a body weight of 25 kg, adult dosages can be used. 15 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 12. Methods 2.1  Search strategy and selection criteria Studies were identified as recommended in the Cochrane handbook and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (21). We searched for all clinical studies, including observational and descriptive studies and randomized controlled trials but excluded case reports. We identified studies involving children and adolescents < 18 years treated for confirmed or suspected TB for the first time or being re-treated after an episode of TB, and which provided data on clinical outcome. All forms of TB (pulmonary, extrapulmonary and disseminated) were considered, whether the diagnosis was based on bacteriological confirmation, imaging, biopsy or a clinical decision to treat. Studies that included patients being treated for TB with confirmed or presumed resistance to RIF, INH, PZA or EMB were excluded from the analysis. The Pubmed, EMBASE, Cochrane Central Register of Controlled Trials, Cochrane Infectious Diseases Group Clinical trials register, WHO International Clinical Trials Registry and Clinicaltrials.gov electronic databases were searched with the terms listed in Appendix 1. The search was conducted for 2010 (date of the WHO guideline revision (22)) to the present, regardless the language or publication status (published, unpublished, in press, in progress). Two independent reviewers (A.B. and A.M.A.) screened the titles and abstracts for relevance and appraised the full text for inclusion in the meta-analysis using pre-specified selection criteria. Key articles were identified by consensus with a third and a fourth reviewer (K.R. and B.P.S.). 2.2 Assessment of risk of bias Quality of evidence and bias were assessed for the included studies with a previously published score (23,24) and adapted to our situation. The scores are listed in Appendix 2. Quality was assessed by four researchers (A.M.A., A.B., B.P.S. and K.R.), and disagreements were resolved by consensus. 2.3 Data extraction A standardized extraction form was developed by consensus by four authors (A.M.A., A.B., B.P.S. and K.R.). Two authors (A.B. and B.P.S.) reviewed the full text of the studies and independently extracted data from those included. In cases of discrepancy, consensus was found between A.B. and B.P.S, who also synthesized the data. The parameters of interest included the country in which the study was conducted, year of publication, study design, study sample size, age as stated in the publication (preferably median and range), form of TB, HIV status, doses given of the compounds under study, adverse events, percentage of unfavourable outcomes and predictors of response. Individual treatment outcomes were extracted from the original studies. The predictors of response were as described in the original reports and were mainly drug exposure, HIV co-infection, malnutrition, acetylator status in studies that included INH, social circumstances and severity of infection. 16 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 2.4 Summary measures of the outcomes In children with pulmonary TB, the outcome was considered favourable if they were smear- or culture- negative in the last month of treatment and on at least one previous occasion, or if the treatment was completed without evidence of failure but with no record that sputum smear or culture results in the last month of treatment and on at least one previous occasion were negative, either because tests were not available or because they were not done. In children with extrapulmonary TB, an outcome was considered to be favourable when there was resolution of clinical signs and symptoms of TB at the end of treatment, as judged by the investigators. If the patient died, required extension of treatment, still had TB at the end of the treatment, failed within the period of follow-up or had a relapse (defined as smear- or culture-positive on one or more occasion during a defined period of follow-up after having previously met the criteria for treatment success), the outcome was considered unfavourable. Patients reported as lost to follow-up were included as unfavourable treatment outcomes. 2.5 Summary measures of adverse events In the studies that reported adverse events, all the available data were extracted, expressed as the number of participants experiencing an event and the total number of events. The intention was to identify adverse events related to anti-TB treatment. 17 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 13. Results 3.1 Search The systematic literature search resulted in 304 studies, of which 104 were duplicates and 154 did not match the eligibility criteria. The PRISMA flow diagram is shown in Fig. 1. Forty-six studies met the inclusion criteria and were included for the full text review; of these, 34 were excluded because no clinical outcome data were reported. Therefore, 12 studies were included in the review and were assessed for quality and data extraction. Fig. 1. PRISMA flow diagram of studies included in the systematic review Pubmed (n=102) Duplicates (n=104) Studies irrelevant (n=154) Exclusion (n=34) Studies imported for screening (n=304) Screened articles (title and abstracts) (n=200) Full text review (n=46) Studies included (n=12) Embase (n=164) Cochrane (n=36) Abstracts (n=2) 3.2 The studies Two studies (25,26) were assessed as of moderate quality, two (10,11) as of low quality and the remaining eight studies were scored as very low quality according to the scoring method reported in Appendix 2, which provides the scores for each study. The studies are summarized in Table 2. Seven were prospective observational studies, and one (27) was a randomized controlled trial, which, unfortunately, is not yet published, and only an abstract is available. Eight of the studies were conducted in India, two in South Africa, two in Uganda and one each in Ghana, Viet Nam and Zambia. The studies included children ranging in age from infants to adolescents < 18 years; one study (11)included only infants, and four studies (10,28–30) included only children > 2 years. HIV infection was reported in 11 studies, of which 7 included HIV-positive patients. 18 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 Dosing in seven studies was according to WHO 2010 recommendations (Table 1) and in three studies according to the recommendations of the Indian Revised National Tuberculosis Control Programme (RNTCP), which comprises treatment given thrice weekly with 10 mg/kg bw INH, 10 mg/kg bw RMP, 30–35 mg/kg bw PZA and 30 mg/kg bw EMB (31). Unfortunately, in none of the studies were doses higher than those recommended by WHO assessed, and no comparisons could be made to evaluate the effect of increasing the dose on the outcome. We nevertheless evaluated the clinical outcomes at the recommended doses in various settings. 19 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 Ta bl e 2. S tu dy c ha ra ct er is ti cs a nd t re at m en t ou tc om es R ef er en ce C ou nt ry a nd st ud y de si gn D os in g re gi m en T yp e of T B H IV A ge r an ge (y ea rs ) O ut co m e Fa ct or s th at a ffe ct ed t he o ut co m e Ra m ac ha nd ra n et al . (2 01 3) ( 34 ) Ind ia M ult ice nt re n= 84 RN TC P gu ide lin es PT B (n = 19 ) EP T B (n = 63 ) PT B+ EP T B (n = 2) N on e 1– 12 Fa vo ur ab le , n = 55 U nf av ou ra bl e, n = 15 LT FU , n = 14 R IF C m ax w as lo we r i n ch ild re n wi th un fav ou ra ble tr ea tm en t o ut co m es . R ap id IN H ac et yla to r s ta tu s a sso cia te d wi th u nf av ou ra ble tr ea tm en t o ut co m es (a dj us te d O R 4 .2 , 9 5% C I 1 .1 –1 5.4 , P= 0. 03 3) . Ra m ac ha nd ra n et al . (2 01 5) ( 35 ) Ind ia Pr os pe cti ve M ult ice nt re n= 77 RN TC P gu ide lin es PT B (n = 49 ) EP T B (n = 28 ) Al l 1– 15 Fa vo ur ab le , n = 54 U nf av ou ra bl e, n = 18 LT FU , n = 5 H ig he r PZ A C m ax in cr ea se d fav ou ra ble tr ea tm en t ou tc om e (a dj us te d O R , 1 .1 ; 95 % CI 1 –1 .2, P = 0. 01 1) M uk he rje e et al . (2 01 5) ( 25 ) Ind ia Pr os pe cti ve M ult ice nt re n= 12 7 • G1 , n = 64 • G 2, n = 63 IN H • G1 : 5 ( 4– 6) m g/ kg • G2 : 1 0 (7 –1 5) m g/ kg RI F • G1 : 1 0 (8 –1 2) m g/ kg • G2 : 1 5 (1 0– 12 ) m g/ kg PZ A 3 0– 35 m g/ kg EM B 20 –2 5 m g/k g PT B (n = 63 ) EP T B (n = 64 ) N on e 0.5 –1 5 Fa vo ur ab le • G1 , n = 53 • G 2, n = 44 Un fav ou ra ble • G1 , n = 9 • G 2, n = 17 LT FU : • G1 : 2 • G2 : 2 IN H C m ax w as lo we r i n ch ild re n wi th un fa vo ur ab le t re at m en t ou tc om e (1 .3 (0 .7 –1 .5 ) μg /m L vs . 3 .4 ( 1. 8– 5. 0) μ g/m L, P= 0. 05 ). C on fir m at io n of M . t ub er cu lo sis w as as so cia te d wi th u nf av ou ra ble tr ea tm en t ou tc om e (5 5. 6% v s. 16 .4% , P = 0. 01 ). G 2: ch ild re n wi th lo we r w eig ht fo r ag e z sc or e ha d un fa vo ur ab le tre at m en t o ut co m e. Be kk er e t a l. (2 01 6) ( 11 ) So ut h A fri ca Pr os pe cti ve M ult ice nt re n= 39 RI F 1 4 (9 –2 0) m g/ kg IN H 14 ( 9– 20 ) m g/ kg PZ A 3 2 (1 9– 45 ) m g/ kg EM B 20 ( 13 –2 9) m g/ kg PT B (n = 36 ) EP T B (n = 1, m en in gi tis ) PT B+ EP T B (n = 2) 5 + 34 – 0– 1 Fa vo ur ab le , n = 33 U nf av ou ra bl e, n = 6 A ll u nf av ou ra ble tr ea tm en t ou tco m es w er e in ch ild re n wi th p oo r so cia l c irc um sta nc es . M uk he rje e et al . (2 01 6) ( 32 ) Ind ia Pr os pe cti ve M on oc en tre n= 56 IN H 4– 6 m g/k g RI F 8 –1 2 m g/k g PZ A 3 0– 35 m g/ kg EM B 20 –2 5 m g/k g PT B EP TB 24 + 32 – 0.5 –1 5 • H IV + Fa vo ur ab le , n = 6 U nf av ou ra bl e, n = 17 LT FU , n = 1 N ot av ail ab le fo r H IV – N o as so cia tio n 20 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 R ef er en ce C ou nt ry a nd st ud y de si gn D os in g re gi m en T yp e of T B H IV A ge r an ge (y ea rs ) O ut co m e Fa ct or s th at a ffe ct ed t he o ut co m e Ra nja lka r e t a l. (2 01 8) ( 28 ) Ind ia Pr os pe cti ve M ult ice nt re n= 41 • G1 , n = 27 • G 2, n = 14 • G1 : T hr ice w ee kly IN H 10 ( 8– 12 ) m g/ kg RI F 1 0 (9 –1 2) m g/ kg • G2 : d ail y IN H 8 ( 7– 9) m g/ kg RI F 1 1 (1 0– 12 ) m g/ kg PT B, n = 36 • G1 , n = 24 • G 2, n = 13 Ly m ph n od e T B, n= 5 • G1 , n = 3 • G 2, n = 2 N on e 2– 16 Fa vo ur ab le • G1 , n = 25 • G 2, n = 11 Un fav ou ra ble • G1 , n = 2 • G 2, n = 3 G 1: RI F C m ax < 8 μ g/m L N an su m ba e t a l. (2 01 8) ( 26 ) Ug an da Pr os pe cti ve M on oc en tre n= 14 4 RI F 1 0– 20 m g/k g IN H 10 –1 5 m g/k g PZ A 3 0– 40 m g/ kg EM B 15 –2 5 m g/k g N A 48 + 94 – 2 N A 0.0 8– 14 En d of tr ea tm en t: Fa vo ur ab le , n = 11 7 U nf av ou ra bl e, n = 22 LT FU , n = 5 S ev er e m al nu tr iti on ( w ei gh t fo r he ig ht ≤ –2 SD ) w as a p re di ct or o f d ea th , a dj us te d HR 8 .8, 9 5% C I 1 .6– 48 .3 Pa nja sa wa tw on g e t al . ( 20 20 ) (3 3) V ie t N am Pr os pe cti ve M on oc en tre n= 10 0 IN H 5 m g/k g RI F 1 0 m g/k g PZ A 2 5 m g/ kg EM B 15 m g/k g TB M 4 + 92 – 4 N A 0.2 –1 5 At 8 m on th s: Fa vo ur ab le , n = 81 U nf av ou ra bl e, n = 15 LT FU , n = 4 S ev er ity o f in fec tio n as so cia te d wi th th e tre at m en t o ut co m e Ar ya e t a l. (2 01 5) ( 29 ) Ind ia Pr os pe cti ve O pe n lab el M on oc en tre n= 20 RN TC P gu ide lin es • G1 : R IF < 1 0 m g/k g • G2 : R IF > 10 m g/k g PT B Ly m ph n od e T B N A 5– 12 At 6 m on th s: Fa vo ur ab le , n = 19 U nf av ou ra bl e, n = 1 O ne u nf av ou ra ble tr ea tm en t o ut co m e w ith R IF < 1 0 m g/k g, an d low C m ax an d AU C, 5 .8 μg /m L an d 29 .7 μ g.h /m L, re sp ec tiv ely . W ob ud ey a e t a l. (2 02 0) ( 27 ) Ind ia, So ut h Af ric a, Ug an da , Z am bi a R an do m iz ed O pe n lab el M ult ice nt re n= 10 24 W HO re co m m en da tio ns • G1 : 4 m on th s • G2 : 6 m on th s N A 12 7 + 89 7 – 0.4 –1 5 Un fav ou ra ble an d LT FU : • G1 , n = 16 • G 2, n = 18 N o as so cia tio n 21 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 R ef er en ce C ou nt ry a nd st ud y de si gn D os in g re gi m en T yp e of T B H IV A ge r an ge (y ea rs ) O ut co m e Fa ct or s th at a ffe ct ed t he o ut co m e D ay al et al . (2 01 8) ( 30 )is on ia zi d (IN H Ind ia Pr os pe cti ve M on oc en tre n= 37 IN H 10 –1 5 m g/k g RI F 1 0– 20 m g/k g PZ A 3 0– 40 m g/ kg EM B 15 –2 5 m g/k g PT B, n = 18 EP T B, n = 19 N on e 3– 10 Fa vo ur ab le , n = 35 LT FU , n = 2 N ot re po rte d An tw i e t a l. (2 01 7) ( 10 ) Gh an a Pr os pe cti ve M on oc en tre n= 11 3 IN H 11 .2 [ 9. 1– 12 .8] m g/k g RI F 1 5. 8 [1 3.6 –1 8.8 ] m g/k g PZ A 2 4. 8 [2 2. 6– 30 ] m g/k g EM B 16 .9 [ 15 –2 0.6 ] m g/k g PT B, n = 85 EP T B, n = 28 54 + 59 – 2.2 –8 .3 Fa vo ur ab le , n = 99 U nf av ou ra bl e, n = 6 LT FU , n = 8 N ot re po rte d A U C , a re a un de r th e (c on ce nt ra tio n– tim e) c ur ve ; I N H , is on ia zi d; R IF , r ifa m pi ci n; P Z A , p yr az in am id e; E M B, e th am bu to l; P T B, p ul m on ar y tu be rc ul os is; E PT B, e xt ra pu lm on ar y tu be rc ul os is; T BM , t ub er cu lo us m en in gi tis ; L T FU , lo st to fo llo w- up G1 : g ro up 1 , G 2: gr ou p 2 RN TC P gu ide lin es : IN H, 1 0 m g/ kg ; R M P, 10 m g/ kg ; P Z A , 3 0– 35 m g/ kg ; E M B, 3 0 m g/ kg g iv en t hr ee t im es a w ee k 22 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 3.3 Summary of treatment outcomes A total of 1862 patients were included in this systematic review of 12 studies, all of whom were treated with the four drugs being evaluated. Of this cohort, 1622 (87.1%) children were cured, and 240 (12.9%) had an unfavourable outcome (died, had treatment failure, relapse or were lost to follow-up). Of those with an unfavourable outcome, 75 children were lost to follow-up, representing 4.0% of the total cohort. The details of the unfavourable outcomes in each study are shown in Fig. 2. Fig. 2. Proportions of unfavourable outcomes for each study according to the dosing regimen Dots for studies that did not report a range of doses or a line representing the range of doses, depending on the availability of the data on the original studies), for rifampicin (RIF), isoniazid (INH), pyrazinamide (PZA) and ethambutol (EMB). G1 and G2 refer to groups in a study, when applicable. One study (32) was excluded from the figure because of the high rate of loss to follow-up (> 50%). This study included 56 patients, of whom 32 were HIV negative and all of whom were lost of follow-up. The authors considered that 17 of the 24 patients who were co-infected with HIV had an unfavourable outcome; however, their definition of treatment failure included many patients, as a child with any change of anti-TB treatment was considered a treatment failure. 23 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 13.4 Predictors of unfavourable outcomes Clinical features were associated with unfavourable outcomes in three studies. Mukherjee et al. (25) reported that confirmation of infection with retrieved Mycobacterium tuberculosis was associated with an unfavourable outcome (55.6% vs 16.4%, P=0.01). They also found that, in a group treated with the WHO-recommended dose, children with a lower weight for age z score (–1.9; interquartile range, –2.3, –1.8), which expresses weight-for-age as the number of standard deviations or a z score below or above the reference median value, had unfavourable outcomes, whereas a higher proportion of children with higher scores (–1.3 (–1.9, –0.6) had favourable outcomes (P=0.007). A similar result was found by Nansumba et al. (26), who treated 144 children and found that severe malnutrition (weight for height ≤ –2 SD) was a predictor of death (adjusted hazard ratio, 8.8, 95% CI: 1.6–48.3). The authors of a study in patients with tuberculous meningitis (33) also concluded that severe disease (Blantyre coma score for children < 5 years < 1 or Glasgow coma scale for children > 5 years < 10) led to higher mortality in a time-to-event model. In one study in infants (11) all the unfavourable outcomes were in children with poor social circumstances (definition not provided in the original article). Relevant associations between drug dosage and exposure and outcomes were found in some studies (25,28,29,33,34), as described in more detail below. Two studies (27,32) found no relevant predictors of unfavourable outcomes Rifampicin Three studies showed an association between RIF Cmax and/or AUC and unfavourable outcomes. Ramachandran et al. (34) showed a significant association: adjusted OR, 1.5; 95% CI 1.1–2.1; P=0.014; Cmax for unfavourable outcomes = 3.4 µg/mL (2.5–4.2 µg/mL); Cmax for favourable outcomes = 5.9 µg/mL (4.4–7.1 µg/mL). They also found that children aged < 3 years had lower RIF, INH and PZA concentrations than older children, although age was no longer statistically significant in a multivariable analysis. Ranjalkar et al. (28) evaluated dosing daily or thrice weekly. In the thrice weekly dosing group, two patients with an unfavourable outcome had a RIF Cmax < 8 µg/mL, which was considered low. In the daily dosing regimen group, three patients required extension of treatment but did not have a low Cmax of RIF. Arya et al. (29) reported that the only patient with an unfavourable outcome had low RIF Cmax and AUC, at 5.8 µg/mL and 29.7 µg h/mL, respectively, and had received a dosing regimen lower than the recommended dose of 10 mg/kg. Isoniazid Mukherjee et al. (25) reported an association between INH Cmax and outcome, the INH Cmax being lower in children with unfavourable outcomes (1.3 (0.7–1.5) µg/mL) than in those with favourable outcomes (3.4 (1.8–5.0) µg/mL) (P = 0.05). Ramachandran et al. (34) also showed that slow or rapid INH acetylator status was associated with treatment outcome (adjusted OR, 4.2; 95% CI, 1.1–15.4; P = 0.033), with a higher proportion of rapid acetylators among unfavourable outcomes, probably due to lower exposure to INH given at the same dose as to intermediate and slow acetylators. 24 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 Pyrazinamide One study of 77 HIV-positive patients investigated the role of PZA by logistic regression analysis. The PZA Cmax was significantly higher in patients with favourable outcomes than in those with unfavourable outcomes (adjusted OR, 1.094; 95% CI, 1.021–1.173, P=0.011). Ethambutol Exposure to ethambutol was not found to be a relevant predictor of outcome in any of the studies reviewed. 3.5 Adverse events Only two of the studies selected for this review reported adverse events. Bekker et al. (11) showed that the anti-TB treatments were well tolerated in a cohort of 39 infants. Five (7%) patients had increased alanine transaminase activity. Three experienced a grade-1 adverse event (elevated twofold), one had a grade-2 adverse event (elevated threefold) and another experienced a grade-3 adverse event (elevated sevenfold) during a 6-month follow-up, all of which resolved spontaneously. Nansumba et al. (26) reported that 30 (20.8%) children experienced 34 serious adverse events. The most frequent events were respiratory infections (n=16) and malnutrition (n=4). Nine respiratory infections and all malnutrition occurred in children < 2 years of age. Of the reported adverse events, only two may have been related to anti-TB treatment: increased alanine transaminase activity (n=1) and peripheral neuropathy (n=1), both of which occurred in HIV-positive patients. 25 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 14. Discussion The objective of this first systematic review of the literature was to determine the efficacy and safety in children of doses higher than those currently recommended by WHO of each of the first-line anti-TB drugs. The results show that no research has been conducted in children of higher doses than those recommended. One study, the SHINE study (27,36), with a shorter treatment period of 4 months indicates that the shorter treatment met the non-inferiority criteria when compared with 6-month treatment. Decreasing the period of treatment is of interest, as it could possibly decrease the number of patients lost to follow-up, improve adherence and decrease the overall cost of treatment. The age range studied was wide, from infants to adolescents < 18 years, during which important developmental and size changes occur ; however, none of the studies reported that an unfavourable outcome was more probable with young age (< 5 years) in multivariable analyses. The most common risk factors for an unfavourable outcome were a low Cmax and/or AUC of RIF, INH and PZA, as lower exposure to these three drugs was associated with unfavourable outcomes, RIF being the most frequently reported. Some studies also showed that low drug concentrations were associated with younger age, probably because of higher clearance of drugs per kilogram in younger children, which may explain why age is a predictor of unfavourable outcome. All the drugs were dosed according to weight, and no distinction by age was made for any drug, even though malnutrition has been highlighted as a significant factor in treatment of TB (14). Severe malnutrition was confirmed as a predictor of unfavourable outcome (treatment failure, relapse or death) in two studies (25,26) performed in India and Uganda, respectively, in relatively large cohorts of children (n=27 (25), n=144 (26)) that were heterogeneous with respect to age (6 months to 15 years (25) and 1 month to 14 years (26)). Malnutrition is a major contribution to disease burden, approximately 45% of global deaths in children < 5 years being attributable to undernutrition, mainly in low- and middle-income countries, including India and Uganda (37). In these countries, more than a third of children < 5 years are stunted (38,39). Therefore, malnutrition is a predominant risk of factor for death, and more studies with adequate assessment of nutritional status should be performed. The studies included in this systematic review are highly heterogenous with respect to patient characteristics (e.g., age, nutritional status), disease status and co-morbidities (e.g., HIV status), and are mostly done in small cohorts of children. Only five studies had a sample size greater than 100 patients. Since summarizing all the results, the unfavourable outcome is around 13%, a much larger cohort would be needed to highlight significant predictors of unfavourable outcomes, especially with a population such as children, characterised by having a large between-individual variability. In this review, although clinical follow-up was conducted in all studies, only two reported adverse events, involving 183 children. Few severe adverse events related to anti-TB treatment were identified, indicating that this therapy is safe at currently used doses, regardless of age. This was reported in a previous paediatric review (40). In our review, however, only 10% of the children were screened for adverse events, and these results should be viewed with caution. Lack of powerful studies prevented us from assessing the impact in children of administering doses higher than those recommended by WHO on the rate of favourable outcomes and the mortality rate. A randomized controlled trial in which the usual recommended dose is compared with higher dosing 26 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 regimens would be necessary to assess the benefits and risks for both efficacy and safety of a new recommendation for dosing. As clinical outcomes appear to be associated with exposure to drugs and more PK studies in paediatric population have been published, studies with PK outcomes should be reviewed to understand the association between dosing regimen and markers of drug exposure (Cmax and/or AUC) and therefore accurately to assess the association between increasing dose and benefit on clinical outcomes. This analysis is limited by the fact that the results in many studies were summarized and reported differently, and no raw patient data were available. This may have introduced inaccuracies. The review reveals the limited literature on TB treatment outcomes in children but highlights a few predictors of unfavourable outcomes, such as malnutrition and low concentration peak. Nevertheless, the heterogeneity of the studies made it difficult to identify relevant, strong, consistent predictors of unfavourable treatment outcomes in different child populations. Although the heterogeneity of the studies is large, this review collates and quantitatively summarizes the available literature on treatment outcomes in children receiving WHO-recommended doses of RIF, INH, PZA and EMB. It provides an important point of reference for understanding the factors that influence treatment outcomes and treatment efficacy at current dosages as exploration of higher doses continues. 27 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 15. Conclusion Owing to lack of data on treatment outcomes at doses higher than those recommended by WHO for each of the first-line anti-TB drugs, the efficacy of such doses could not be assessed. A more detailed analysis of exposure of children to the drugs at different doses and achievement of the target exposure is necessary. A further review and meta-analysis of PK studies in children and the achievement of exposure targets should be the next step in assessing the efficacy of higher doses. 28 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 References 1. WHO consolidated guidelines on drug-resistant tuberculosis treatment. Geneva: World Health Organization; 2019 (http://www.ncbi.nlm.nih.gov/books/NBK539517/, accessed 8 July 2020). 2. Global tuberculosis report. Geneva: World Health Organization; 2020 (https://apps.who.int/iris/bitstream/han dle/10665/336069/9789240013131-eng.pdf, accessed 1 February 2021). 3. Drobac PC, Shin SS, Huamani P, Atwood S, Furin J, Franke MF et al. Risk factors for in-hospital mortality among children with tuberculosis: the 25-year experience in Peru. Pediatrics. 2012;130(2):e373–9. 4. Fox W, Ellagard G, Mitchison D. Studies on the treatment of tuberculosis undertaken by the British Medical Research Council tuberculosis units, 1946–1986, with relevant subsequent publications. Int J Tuberc Lung Dis. 1999;3(10 Suppl 2):S231–79. 5. Short-course chemotherapy in pulmonary tuberculosis: a controlled trial by the British Thoracic and Tuberculosis Association. Lancet. 1976;308(7995):1102–4. 6. Lienhardt C. Efficacy and safety of a 4-drug fixed-dose combination regimen compared with separate drugs for treatment of pulmonary tuberculosis: the Study C Randomized Controlled Trial. JAMA. 2011;305(14):1415. 7. Technical report on the pharmacokinetics and pharmacodynamics (PK/PD) of medicines used in the treatment of drug-resistant tuberculosis. Geneva: World Health Organization; 2018 (https://apps.who.int/iris/ handle/10665/260440). 8. General clinical pharmacology considerations for pediatric studies for drugs and biological products. Silver Spring (MD): Food and Drug Administration, Center for Drug Evaluation and Research; 2014:28. 9. Barbour AM, Fossler MJ, Barrett J. Practical considerations for dose selection in pediatric patients to ensure target exposure requirements. AAPS J. 2014;16(4):749–55. 10. Antwi S, Yang H, Enimil A, Sarfo AM, Gillani FS, Ansong D et al. Pharmacokinetics of the first-line antituberculosis drugs in Ghanaian children with tuberculosis with or without HIV coinfection. Antimicrob Agents Chemother. 2017;61(2):e01701-16. 11. Bekker A, Schaaf HS, Draper HR, van der Laan L, Murray S, Wiesner L et al. Pharmacokinetics of rifampin, isoniazid, pyrazinamide, andethambutol in infants dosed according to revised WHO-recommended treatment guidelines. Antimicrob Agents Chemother. 2016;60(4):2171–9. 12. Justine M, Yeconia A, Nicodemu I, Augustino D, Gratz J, Mduma E et al. Pharmacokinetics of first-line drugs among children with tuberculosis in rural Tanzania. J Pediatr Infect Dis Soc. 2020;9(1):14–20. 13. Guiastrennec B, Ramachandran G, Karlsson MO, Kumar AKH, Bhavani PK, Gangadevi NP et al. Suboptimal antituberculosis drug concentrations and outcomes in small and HIV-coinfected children in India: recommendations for dose modifications. Clin Pharmacol Ther. 2018;104(4):733–41. 14. Radtke KK, Dooley KE, Dodd PJ, Garcia-Prats AJ, Hesseling AC, Savic RM. Evaluation of dosing guidelines for childhood tuberculosis: a mathematical modeling study. SSRN Electron J. 2019 (https://www.ssrn.com/ abstract=3311868, accessed 21 January 2021). 15. Svensson EM, Yngman G, Denti P, McIlleron H, Kjellsson MC, Karlsson MO. Evidence-based design of fixed- dose combinations: principles and application to pediatric anti-tuberculosis therapy. Clin Pharmacokinet. 2018;57(5):591–9. 16. McIlleron H, Chirehwa MT. Current research toward optimizing dosing of first-line antituberculosis treatment. Expert Rev Anti Infect Ther. 2019;17(1):27–38. 17. Kearns GL, Abdel-Rahman SM, Alander SW, Blowey DL, Leeder JS, Kauffman RE. Developmental pharmacology – drug disposition, action, and therapy in infants and children. N Engl J Med. 2003;349(12):1157–67. 18. Wilkins JJ, Langdon G, McIlleron H, Pillai G, Smith PJ, Simonsson USH. Variability in the population pharmacokinetics of isoniazid in South African tuberculosis patients: population pharmacokinetics of isoniazid. Br J Clin Pharmacol. 2011;72(1):51–62. 19. Ridge A, Grzemska M, Hill S, Gie R. Rapid advice: treatment of tuberculosis in children. Geneva: World Health Organization; 2010 (http://whqlibdoc.who.int/publications/2010/9789241500449_eng.pdf, accessed 14 May 2021). 20. Guidance for national tuberculosis programmes on the management of tuberculosis in children. Geneva: World Health Organization; 2014 (https://apps.who.int/iris/handle/10665/332221, accessed 2 November 2021). 29 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 121. Moher D, Liberati A, Tetzlaff J, Altman D, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. 22. Implementing the end TB strategy: the essentials. Geneva: World Health Organization; 2015 (https://apps.who. int/iris/handle/10665/206499, accessed 21 September 2021). 23. Seden K, Gibbons S, Marzolini C, Schapiro JM, Burger DM, Back DJ et al. Development of an evidence evaluation and synthesis system for drug–drug interactions, and its application to a systematic review of HIV and malaria co-infection. PLoS One. 2017;12(3):e0173509. 24. Jacobs TG, Svensson EM, Musiime V, Rojo P, Dooley KE, McIlleron H et al. Pharmacokinetics of antiretroviral and tuberculosis drugs in children with HIV/TB co-infection: a systematic review. J Antimicrob Chemother. 2020;75(12):3433–57. 25. Mukherjee A, Velpandian T, Singla M, Kanhiya K, Kabra SK, Lodha R. Pharmacokinetics of isoniazid, rifampicin, pyrazinamide and ethambutol in Indian children. BMC Infect Dis. 2015;15(1):126. 26. Nansumba M, Kumbakumba E, Orikiriza P, Bastard M, Mwanga JA, Boum Y et al. Treatment outcomes and tolerability of the revised WHO anti-tuberculosis drug dosages for children. Int J Tuberc Lung Dis. 2018;22(2):151–7. 27. Wobudeya E, Chabala C, Hesseling AC, Mave V, Hissar S, Turkova A et al. Shorter treatment for minimal tuberculosis in children: main findings from the SHINE trial. Abstract LB-2056-24. In: Abstracts of the 51st Union World Conference on Lung Health; 2021. Available on: https://wclh2020.abstractserver.com/WCLH2020_ abstract_book_high.pdf 28. Ranjalkar J, Mathew SK, Verghese VP, Bose A, Rose W, Gupta D et al. Isoniazid and rifampicin concentrations in children with tuberculosis with either a daily or intermittent regimen: implications for the revised RNTCP 2012 doses in India. Int J Antimicrob Agents. 2018;51(5):663–9. 29. Arya A, Roy V, Lomash A, Kapoor S, Khanna A, Rangari G. Rifampicin pharmacokinetics in children under the Revised National Tuberculosis Control Programme, India, 2009. Int J Tuberc Lung Dis. 2015;19(4):440–5. 30. Dayal R, Singh Y, Agarwal D, Kumar M, Swaminathan S, Ramachandran G et al. Pharmacokinetic study of isoniazid and pyrazinamide in children: impact of age and nutritional status. Arch Dis Child. 2018;103(12):1150–4. 31. Revised National Tuberculosis Control Programme. Technical and operational guidelines for tuberculosis control, October 2005. New Delhi: Ministry of Health and Family Welfare; 2005 (http://tbcindia.nic.in/pdfs/Technical%20 &%20Operational%20guidelines%20for%20TB%20Control.pdf). 32. Mukherjee A, Velpandian T, Singla M, Kanhiya K, Kabra SK, Lodha R. Pharmacokinetics of isoniazid, rifampicin, pyrazinamide and ethambutol in HIV-infected Indian children. Int J Tuberc Lung Dis. 2016;20(5):666–72. 33. Panjasawatwong N, Wattanakul T, Hoglund RM, Bang ND, Pouplin T, Nosoongnoen W et al. Population pharmacokinetic properties of antituberculosis drugs in Vietnamese children with tuberculous meningitis. Antimicrob Agents Chemother. 2020;65(1):e00487-20. 34. Ramachandran G, Kumar AKH, Bhavani PK, Gangadevi NP, Sekar L, Vijayasekaran D et al. Age, nutritional status and INH acetylator status affect pharmacokinetics of anti-tuberculosis drugs in children. Int J Tuberc Lung Dis. 2013;17(6):800–6. 35. Ramachandran G, Kumar AKH, Bhavani PK, Kannan T, Kumar SR, Gangadevi NP et al. Pharmacokinetics of first- line antituberculosis drugs in HIV-infected children with tuberculosis treated with intermittent regimens in India. Antimicrob Agents Chemother. 2015;59(2):1162–7. 36. Chabala C, Turkova A, Thomason MJ, Wobudeya E, Hissar S, Mave V et al. on behalf of the SHINE trial team. Shorter treatment for minimal tuberculosis (TB) in children (SHINE): a study protocol for a randomised controlled trial. Trials. 2018;19(1):237. 37. Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, de Onis M et al. Maternal and child undernutrition and overweight in low-income and middle-income countries. Lancet. 2013;382(9890):427–51. 38. Adebisi YA, Ibrahim K, Lucero-Prisno DE, Ekpenyong A, Micheal AI, Chinemelum IG et al. Prevalence and socio- economic impacts of malnutrition among children in Uganda. Nutr Metab Insights. 2019;12:117863881988739. 39. Swaminathan S, Hemalatha R, Pandey A, Kassebaum NJ, Laxmaiah A, Longvah T et al. The burden of child and maternal malnutrition and trends in its indicators in the states of India: the Global Burden of Disease Study 1990–2017. Lancet Child Adolesc Health. 2019;3(12):855–70. 40. Hatzenbuehler LA, Starke JR. Treatment of tuberculosis infection in children. Expert Rev Anti Infect Ther. 2018;16(9):695–708. 30 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1 Appendix 1. Search strategy Search set Pubmed Cochrane Embase 1 child* OR neonat* OR pediatr* child* OR neonat* OR pediatr* child* OR neonat* OR pediatr* 2 Tuberculosis Tuberculosis Tuberculosis 3 #1 AND #2 #1 AND #2 #1 AND #2 4 isoniazid OR pyrazinamide OR ethambutol OR RIF isoniazid OR pyrazinamide OR ethambutol OR RIF isoniazid OR pyrazinamide OR ethambutol OR RIF 5 #3 AND #4 #3 AND #4 #3 AND #4 6 pharmacokinetic* OR outcome pharmacokinetic* OR outcome pharmacokinetic* OR outcome 7 #5 AND #6 #5 AND #6 #5 AND #6 8 prevention* OR latent prevention* OR latent prevention* OR latent 9 #7 NOT #8 #7 NOT #8 #7 NOT #8 Pubmed and Embase: research in title and abstract Cochrane: research in title, abstract and keywords 31 Clinical outcomes among children (< 18 years) treated for drug-susceptible tuberculosis with fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide): systematic review A N N E X 1Appendix 2. Quality of evidence scoring  outcome analysis Final scoring: high (1 point), moderate (2 points), low (3 points), or very low (> 4 points). Study design: Randomized trial: 1 Cohort study, prospective: 2 Cohort study, retrospective: 3 Other : 4 Add 1 point in the following situations of poor evidence: • The publication was a conference abstract or not published in a peer-reviewed journal. • Sample size < 30 children in each arm or no comparator arm • Doses were lower than currently recommended or not specified for any drug used, or the formulation used was other than a WHO-recommended fixed-dose combination or not specified. • Clinical outcomes (e.g., death, cure, loss to follow-up) were not well defined or were self-reported. • Inclusion and/or exclusion criteria were not defined (selection bias). • Ascertainment of intervention was self-reported or not described. • No demonstration that drug-sensitive TB is not confirmed or drug-resistant TB is probable • No follow-up or median duration of follow-up < 6 months Remove 1 point for significant findings: • Doses higher than those currently recommended were investigated with a comparator arm. • The study was conducted in multiple centres. • There is a twofold difference in odds, hazard or risk ratio between arms, and it is statistically significant for the primary outcome. Final scores for individual reports First author (reference) Final score Mukherjee (32) 5 Ranjalkar (28) 4 Arya (29) 6 Ramachandran (34) 5 Ramachandran (35) 5 Bekker (11) 3 Mukherjee (25) 2 Panjasawatwong (33) 4 Chabala (36) Could not be assessed Nansumba (26) 2 Dayal (30)isoniazid (INH 4 Antwi (10) 3

Pharmacokinetics of first-line  drugs (rifampicin, isoniazid,  ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis:  systematic review and meta-analysis ANNEX 2 Authors Belén P. Solans, Agathe Béranger, Kendra Radtke, Ali Mohamed, Payam Nahid, Rada Savic University of California San Francisco, UCSF Center for Tuberculosis, USA This work was done in collaboration with the Global TB Programme of the World Health Organization.

35 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Contents Abbreviations and acronyms 36 1. Introduction 37 2. Methods 39 2.1 Search strategy and selection criteria 39 2.2 Assessment of risk of bias 39 2.3 Data extraction 39 2.4 Data synthesis 40 2.5 Summary measures 40 3. Results 41 3.1 Literature search 41 3.2 Assessment of risk of bias 42 3.3 Studies included 42 3.4 Summary of reported pharmacokinetics parameters 49 3.5 Ethambutol 70 3.6 Association between attainment of pharmacokinetics target and clinical outcomes 71 3.7 Adverse events 71 4. Discussion 72 5. Conclusion 75 References 76 Appendix 1. Search strategy 80 Appendix 2. Scoring the quality of evidence – pharmacokinetics target  analysis 81 Appendix 3. Data extraction form 83 Appendix 4. Exposure to drugs (AUC and Cmax) as reported in the studies 84 36 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Abbreviations and acronyms AUC area under the concentration-time curve C2h concentration 2 h after dosing Cmax maximum concentration EMB ethambutol EPTB extrapulmonary tuberculosis HAZ height-for-age z score INH isoniazid IPD-MA individual participant data meta-analysis IQR interquartile range LTFU lost to follow-up PK pharmacokinetics PRISMA Preferred Reporting Items for Systematic Reviews and Meta-analyses PTB pulmonary tuberculosis PZA pyrazinamide RIF rifampicin RNTCP Revised National Tuberculosis Control Programme (India) SD standard deviation SEM standard error of mean TB tuberculosis TBM tuberculous meningitis WAZ weight-for-age z score WHO World Health Organization WHZ weight-for-height z score 37 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 21. Introduction Tuberculosis (TB) is a communicable disease that is a major cause of ill health and one of the leading causes of death worldwide. Until the coronavirus (COVID-19) pandemic, TB was the leading cause of death from a single infectious agent, ranking above HIV/AIDS, with an estimated 10 million (8.9–11.0 million) cases and 1.2 million deaths worldwide in 2019, of which about 20% were children < 15 years of age. An estimated 10 million cases (range, 8.9–11.0 million) and 1.4 million deaths worldwide in 2019 (1). Children < 15 years accounted for 12% of the people who developed TB and 16% of deaths (230 000 of 1.4 million) from TB in 2019. While 30 low- and middle-income countries account for almost 90% of this burden, a significant fraction can be attributed to risk factors for TB disease, such as young age, HIV co-infection, malnutrition and diabetes. Progress has recently been made in the control of TB, but the annual decrease in incidence does not currently exceed 2%. In 2018 and 2019, 14.1 million people with TB received treatment, with a treatment success rate of 85% in 2018, similar to that of children (84%) (1). The risk factors for TB disease, young age (< 5 years), HIV-coinfection and malnutrition, are also the risk factors for worse treatment outcomes in the general population. Worse outcomes are associated with extrapulmonary and disseminated disease, which are more common in young children (2). The COVID-19 pandemic has had enormous health, social and economic negative impacts, disrupting essential TB services, and this is expected to reverse recent progress in reducing the global burden of TB disease. Forty-three countries, including 13 high-TB burden countries, have reported that GenXpert machines have been used for COVID-19 testing rather than diagnostic testing for TB; 85 countries (including 20 high-TB burden countries) have reported reassignment of staff of national TB programmes to COVID-19-related duties; and 52 countries (including 14 high-TB burden countries) have reported reallocation of budgets. In many countries, data collection and reporting on TB have also been affected (1). First-line regimens for TB have changed little since their introduction more than 40 years ago (3). The addition of rifampicin (RIF) and pyrazinamide (PZA) to isoniazid (INH) reduced the treatment duration to 6 months, and these three drugs were established as the backbone of effective therapy (4). A fourth drug was added to prevent drug resistance, particularly in people who already harboured INH-resistant strains and were at a higher risk of treatment failure. While streptomycin was initially widely used for this purpose, widespread streptomycin resistance and a preference for parenteral administration led to its replacement by the oral agent ethambutol (EMB), resulting in the current standard four-drug first- line regimen (5). Since 2014, the doses of the four first-line TB drugs for children weighing < 25 kg (see Table 1) have been re-evaluated according to modern preclinical and clinical pharmacokinetics (PK)–pharmacodynamics, and new phase-I and -II clinical trials are establishing proof-of-concept that higher doses improve efficacy and/ or shorten treatment duration (6–9). It has not yet been established, however, whether higher doses will improve treatment outcomes without additional toxicity. 38 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Table 1. Current recommended doses for children weighing ≤ 25 kga and pharmacokinetics target of first-line anti-TB drugs recommended in WHO guidance Drug Daily dose in 2010 (10) (mg/kg bw) Daily dose in 2014 (11) (mg/kg bw) Mean adult AUC (µg.h/mL) Mean adult Cmax (µg/L) RIF (12) 10–20 15 (range, 10–20) 38.7 8–24 INH (13) 10–15 10 (range, 7–15) 23.4 3–5 PZA (14) 30–40 35 (range, 30–40) 238–428 35–60 EMB (15) 15–25 20 (range, 15–25) 16–28 2–6 RIF, rifampicin; INH, isoniazid; PZA, pyrazinamide; EMB, ethambutol; bw, body weight; AUC, area under the concentration–time curve; Cmax, maximum concentration a As children approach a body weight of 25 kg, adult dosages may be used. Newer PK studies in children treated under the revised dosing recommendations show that their exposure to first-line TB drugs are often lower than those of adults (16–18). Furthermore, the weight- based dose recommendations (in mg/kg) are the same for all children, given as an incremental number of fixed-dose combination tablets. Modelling suggests that higher doses are required for low-weight and malnourished (weight-for-age z score < 2) children (19–22). This suggestion is supported by basic principles of developmental pharmacology, which are that higher doses are required for younger, low- weight children to account for higher clearance per kg bw (23). Further, NAT2 acetylator status is known to be a primary determinant of exposure of infants and adolescents to INH, with twice the clearance in fast as in slow acetylators (24). Treatment outcomes and safety according to NAT2 status are not, however, well characterized, especially in children. A synthesis of the evidence on dose, exposure and clinical outcomes (treatment success, failure, relapse and adverse effects) from all studies and all geographical regions is therefore important for determining optimal dosages of first-line TB medicines for children. The PK targets shown in Table 1 are for the mean exposure of adults that provides the best risk–benefit ratio. A recommended method for defining the optimal dose for children is based on adult–paediatric exposure equivalence, i.e., the dosing algorithm in paediatric patients that achieves the optimal exposure of adults. The aims are to ensure that the PK targets of all four first-line anti-TB drugs improve the rate of relapse-free treatment while ensuring the safety of patients, make regimens more robust in settings of variable adherence, PK and pharmacogenetics and protect against the emergence of drug resistance. Furthermore, use of optimal doses is important for people with severe or disseminated disease and for vulnerable populations, such as those receiving more than one drug and are susceptible to drug–drug interactions and those receiving drugs with high variations in PK. Understanding the impact of malnutrition on drug concentrations and outcomes and re-evaluating current weight-band dosing will be of critical importance for children (20). The objective of this review was to assess PK parameters (maximum concentration (Cmax) and area under the concentration–time curve (AUC)) as markers of exposure to drugs in children treated for presumed drug-susceptible TB with doses of each first-line drug higher than those currently recommended by WHO in combination regimens. 39 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 22. Methods 2.1  Search strategy and selection criteria Studies were identified in accordance with the instructions in the Cochrane handbook and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (25). We searched for all clinical studies, including observational and descriptive studies, and randomized controlled trials; we excluded case reports. We identified studies involving children and adolescents < 18 years being treated for bacteriologically confirmed TB or presumed TB (clinically diagnosed) for the first time or who were being re-treated after a previous episode of TB. All forms of TB (pulmonary, extrapulmonary and disseminated) were considered, whether the diagnosis was based on bacteriological confirmation, radiography, biopsy or a clinical decision to treat. Studies that included patients being treated for TB with confirmed or presumed resistance to RIF, INH, PZA or EMB were excluded from the analysis. Pubmed, EMBASE, CENTRAL (Cochrane central register of controlled trials), the Cochrane Infectious diseases Group Clinical trials register, the WHO International Clinical Trials Registry and Clinicaltrials.gov electronic databases were searched with the terms listed in Annex 1. The search was conducted for the period 2010 (date of update of dosing recommendations by WHO (26)) to the present, regardless of language or publication status (published, unpublished, in press and in progress). All titles and abstracts were imported into Covidence to remove duplicates and manage references. Two independent reviewers (A.B. and A.M.A.) screened the titles and abstracts for relevance and appraised the full text for inclusion in the meta-analysis on the basis of prespecified selection criteria. Key articles were identified by consensus between a third and a fourth reviewer (K.R. and B.P.S.). 2.2 Assessment of risk of bias The quality of evidence and bias in the studies were assessed on a score developed by Seden et al. (27) and Jacobs et al. (28) and adapted to our systematic review. The scores are shown in Annex 2. Quality was assessed by the four researchers (A.M.A., A.B., B.P.S. and K.R.), and any disagreements were resolved by consensus. 2.3 Data extraction A data extraction form was prepared by consensus by four authors (A.M.A., A.B., B.P.S. and K.R.) and is presented in Annex 3. Two authors (A.B. and B.P.S.) independently extracted data from the full-text publications included. In case of discrepancy, A.B. and B.P.S. found a consensus and synthesized the data. The parameters of interest on the data extraction form included the country in which the study was conducted, year of publication, study design, study sample size, age of participants (median and range), form of TB, HIV status, doses given of the drugs under study, adverse events, percentage of unfavourable outcomes and predictors of response. Treatment outcomes were defined as favourable when the child was cured or if the treatment was completed and unfavourable when the children died or required a change or extension of treatment. The predictors of response were extracted as described in the studies. When available in the original report, exposure to the drugs (i.e., AUC and Cmax) was extracted and also by subgroup, including body weight, age, HIV status, nutritional status (definition derived from the study), NAT2 acetylator status (for INH only) and dosing regimen (daily or other). 40 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 2.4 Data synthesis Data were organized in Excel® version 16.46 and visualized in R version 1.3.1093. The characteristics of study participants were summarized as reported in the original article, either as the full cohort (preferred) or by group (e.g., HIV status) if a full cohort summary was not available. Malnutrition was summarized by each measure of nutritional status separately, if possible, as reported in the original study: height-for-age z score (HAZ) or stunting (HAZ < –2), weight-for-age z score (WAZ) or underweight (WAZ < –2) and weight-for-height z score (WHZ) or wasting (WHZ < –2). Any z score < –2 was considered to indicate malnutrition in order to avoid discrepancies in our report. Age was summarized as a continuous variable for the entire cohort, when possible. In studies in which participants were grouped by age category, each group was summarized as specified by the study. Exposure (i.e., Cmax, AUC) to each of the four drugs was summarized for the whole cohort, for each dosing group or study arm and by each relevant covariate: nutritional status, HIV status, acetylator status (INH only) and age. For AUC, the time interval for each study was recorded. The AUC 0–4 h after dosing (AUC0–4) or greater was used for all drugs, and data reported as “Cmax” or concentration 2 h after dosing (C2h) was considered to represent the Cmax for all drugs. The actual exposure to the drugs with respect to the target exposure are expressed as the median or mean and range. All reported adverse events were also extracted and expressed as the number of participants experiencing an event and the total number of events. 2.5 Summary measures Summary estimates for AUC and Cmax were obtained by meta-analysis with “metafor” package version 2-4.0 in R and expressed as mean and standard deviation. When summary statistics were not available in this format, the mean and standard deviation were estimated from the summary statistics provided (e.g., median and interquartile range, IQR) with previously described methods (29). The main objective of the analysis was to collate and summarize data on the exposure to drugs of children taking first-line drugs for TB treatment, especially at doses higher than those recommended by WHO. When data were not available for the whole cohort, data from the main subgroup were used to avoid excluding the studies from the summary estimates. The secondary objective was to summarize the exposure to drugs of key subgroups (i.e., by age, HIV status, malnutrition status and NAT2 genotype for INH). A restricted maximum likelihood mixed-effects model was used to perform a meta-analysis of AUC and Cmax estimates. Heterogeneity was assessed with the I2 statistic and visual inspection of forest plots. Meta-regression was performed with key covariates to assess their impact on inter-study heterogeneity. 41 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 23. Results 3.1 Literature search The systematic literature search retrieved 304 studies (including two conference abstracts), of which 104 were duplicates and 154 were found to be ineligible according to the selection criteria. The PRISMA flow diagram is shown in Fig. 1. After title and abstract screening, 47 studies met the inclusion criteria and were included in the full text review. Of these, 28 were excluded: 11 that reported duplicate PK data, 6 with an unknown dosing regimen, 4 in which doses lower than those recommended were used, 3 in which the drugs were used for indications other than treatment-sensitive TB, 2 in which PK parameters were not reported and 2 that included an adult population. Therefore, 18 studies were included in the review for quality assessment and data extraction. Fig. 1. PRISMA flow diagram Pubmed (n=102) Duplicates (n=104) Studies irrelevant (n=154) Exclusion (n=28) • 11 duplicate reports • 6 Unknown dosing regimens • 4 lower doses than recommended • 3 other indications – not DS-TB • 2 missing PK parameters • 2 adult population Studies imported for screening (n=304) Screened articles (title and abstracts) (n=200) Full text review (n=46) Studies included (n=18) Embase (n=164) Cochrane (n=36) Abstracts (n=2) Rifampicin n=14 Isoniazid n=16 Pyrazinamide n=13 Ethambutol n=8 42 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 3.2 Assessment of risk of bias One study (30) was assessed as of moderate quality, one (16) as of low quality and all the remaining studies as of very low quality according to the scoring method outlined in Annex 2, which also shows the scores for each study. 3.3 Studies included All the studies included are summarized in Table 2. They were all prospective observational studies and were conducted in the following countries: 10 in India, 4 in South Africa and one each in Ghana, Malawi, the United Republic of Tanzania and Viet Nam. All the studies included children, with an age range from infants to adolescents < 16 years. HIV infection was reported in 13 studies; 8 included HIV-positive patients in proportions ranging from 4% to 100%. Administration of antiretroviral therapy was described in only 3 studies, with different protocols. 43 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Ta bl e 2. S tu dy c ha ra ct er is ti cs Fi rs t a ut ho r, re fe re nc e no . ( ye ar ) St ud y de sig n Do sin g re gim en Ty pe o f T B HI V Ag e (y ea rs ) Bo dy we igh t ( kg ) N ut rit io na l sta tu s Dr ug s PK pa ra m et er s Co va ria te s Fa ct or s a ffe ct in g P K Cl in ica l ou tc om e Fa ct or s a ffe ct in g c lin ica l ou tc om e Th ee , 31 (2 01 1) So uth Af ric a Pro spe ctiv e Mo no cen tre n= 20 IN H: 5 a nd 10 m g/k g RIF : 10 an d 15 m g/k g PZ A: 2 5 an d 35 m g/k g PT B, n= 11 EP TB , n =1 TB M , n =8 5 + 15 – M ea n (S D) , 1.1 (0 .5) NA M ea n (S D) W AZ – 1.7 4 (1 .84 ) IN H, R IF, PZ A C m ax AU C 0– 5 Ag e, g end er, typ e o f TB , nu trit ion al sta tu s, HI V sta tus , N AT 2 for IN H Do sin g r egi me n ass oci ate d w ith C m ax and AU C f or all dru gs (P <0 .00 1 f or IN H and PZ A, P <0 .00 6 fo r R IF) NA T2 ge no typ e o n I NH C m ax an d A UC (P <0 .05 ) NA NA Ro y, 4 3 (2 01 2) Ind ia Pro spe ctiv e Mo no cen tre n= 20 • G 1, n= 7 • G2 , n =1 3 • G 1: PZ A > 30 –3 5 m g/k g • G2 : P ZA < 30 –3 5 m g/k g PT B Ly m ph no de TB NA Ra nge 5– 12 M ea n (S EM ) • G 1: 5 .6 (0 .3) • G 2: 5 .8 (0 .2) M ea n (S EM ) • G 1: 1 5.7 (0 .4) • G 2: 1 6.3 (0 .8) NA PZ A C m ax AU C 0– 24 Do sin g reg ime n Do sin g r egi me n as so cia te d wi th P ZA C ma x an d A UC (P <0 .01 ) NA NA Ram ach and ran 34 (2 01 3) Ind ia Pro spe ctiv e Mu ltic ent re n= 84 RN TC P gui de line s PT B, n= 19 EP TB , n =6 3 P+ EP TB , n= 2 No ne M ea n (ra ng e) 7.1 (1 –1 2) M ed ian [I Q R] 18 [1 3– 23 ] St un te d, n= 22 Un de rw eig ht, n= 31 W as te d, n= 16 M ed ian [I Q R] HA Z –1 .2 [– 2.1 , –0 .3] W AZ – 1.8 [– 2.4 , – 1.1 ] W HZ – 1.2 [– 1.9 , –0 .3] IN H, R IF, PZ A C m ax AU C 0– 8 Ag e, N AT 2 for IN H, BM I, a lbu min , nu trit ion al sta tus , ou tco me Yo un ge r c hil dr en h ad low er C m ax a nd AU C f or all d ru gs (P <0 .01 ). Ma lnu trit ion de cre ase d RIF C m ax a nd AU C (P <0 .05 ). NA T2 ge no typ e o n I NH C m ax an d A UC (P <0 .00 1) Fav ou rab le, n= 55 Un fav ou rab le, n= 15 LT FU , n =1 4 RIF an d I NH AU C a nd C m ax l ow er in c hild ren w ith un fav ou rab le o utc om es (P <0 .03 ) Ra pid IN H ace tyla tor sta tus ass oci ate d w ith un fav ou rab le ou tco m es (a O R 4.2 , 9 5% C I 1.1 –1 5.4 , P= 0.0 33 ) 44 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Fi rs t a ut ho r, re fe re nc e no . ( ye ar ) St ud y de sig n Do sin g re gim en Ty pe o f T B HI V Ag e (y ea rs ) Bo dy we igh t ( kg ) N ut rit io na l sta tu s Dr ug s PK pa ra m et er s Co va ria te s Fa ct or s a ffe ct in g P K Cl in ica l ou tc om e Fa ct or s a ffe ct in g c lin ica l ou tc om e Ram ach and ran 35 (2 01 5) Ind ia Pro spe ctiv e Mu ltic ent re n= 77 RN TC P gui de line s PT B, n= 49 EP TB , n =2 8 77 + Me dia n (ra ng e) 9 (1 –1 5) M ed ian [I Q R] 17 [1 4.1 –2 2.5 ] St un te d, n= 59 Un de rw eig ht, n= 56 W as te d, n= 15 M ed ian [I Q R] HA Z –3 .0 [– 4.1 , –2 .0] W AZ – 2.7 [– 3.4 , – 1.9 ] W HZ – 1.1 [– 1.7 , –0 .02 ] IN H, R IF, PZ A C m ax AU C 0– 8 Ag e, g end er, nu trit ion al sta tus , BM I, alb um in, AR T, NA T2 fo r IN H, ou tco me Ag e < 5 ye ar s a sso cia te d wi th lo we r I N H an d PZ A C m ax an d A UC (P <0 .05 ) NA T2 ge no typ e o n I NH C m ax an d A UC (P <0 .02 ) Lo w alb um in lev el de cre ase d R IF C ma x (P =0 .03 7) Fav ou rab le, n= 54 Un fav ou rab le, n= 18 LT FU , n =5 PZ A C m ax a ffec ted ou tco me (a O R, 1.1 ; 9 5% C I 1 –1 .2, P= 0.0 11 ) Ra nga ri 6 0 (2 01 5) Ind ia Pro spe ctiv e Mo no cen tre n= 20 • G 1, n= 8 • G2 , n =1 2 RN TC P gui de line s • G 1: I NH > 10 m g/k g • G2 : IN H < 10 m g/k g PT B Ly m ph no de TB NA Ra nge 5– 12 M ea n (S EM ) • G 1: 8 .8 (0 .4) • G 2: 1 0.7 5 (0 .3) M ea n (S EM ) • G 1: 2 1.5 (0 .3) • G 2: 2 2.6 (0 .7) NA IN H C m ax AU C 0– 24 Do sin g reg ime n Do sin g r egi me n o n I NH AU C (P =0 .00 2) NA NA Ar ya 39 (2 01 5) Ind ia Pro spe ctiv e Mo no cen tre n= 20 RN TC P gui de line s • G 1: R IF > 10 mg /kg • G2 : R IF < 10 mg /kg PT B Ly m ph no de TB NA Me dia n (ra ng e) • G 1: 9 (6 –1 0) • G 2: 1 2 (6 –1 2) Me dia n (ra ng e) • G 1: 2 0.6 (1 5– 22 .4) • G 2: 2 4.2 (1 5.2 –2 5) M ea n (ra ng e) 21 .6 (1 5– 25 ) NA RIF C m ax AU C 0– 12 Ag e, d osi ng reg ime n Do sin g r egi me n o n R IF C m ax an d A UC (P <0 .05 ) At 6 m on ths : Fav ou rab le, n= 19 Un fav ou rab le, n= 1 On e u nfa vou rab le o utc om e wi th R IF < 1 0 m g/k g, a nd lo w C m ax a nd AU C, 5.8 μg /m L an d 29 .7 μ g.h /m L, re sp ec tiv ely 45 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Fi rs t a ut ho r, re fe re nc e no . ( ye ar ) St ud y de sig n Do sin g re gim en Ty pe o f T B HI V Ag e (y ea rs ) Bo dy we igh t ( kg ) N ut rit io na l sta tu s Dr ug s PK pa ra m et er s Co va ria te s Fa ct or s a ffe ct in g P K Cl in ica l ou tc om e Fa ct or s a ffe ct in g c lin ica l ou tc om e Mlo tha 32 (2 01 5) Ma law i Pro spe ctiv e Mo no cen tre n= 30 IN H 5 m g/k g RIF 10 m g/k g PZ A 25 m g/k g EM B 2 0 m g/k g PT B, n= 21 EP TB , n =9 20 + 10 – Me dia n (ra ng e) 7.5 (0 .5– 15 .6) Me dia n (ra ng e) 18 (4 .8– 45 ) NA IN H, R IF, PZ A, EM B C m ax AU C 0– las t AU C 0– ∞ Ag e, d osi ng re gim en , H IV Do sin g r egi me n o n R IF AU C 0– ∞ (P = 0 .02 8) NA NA Hir uy 41 (2 01 5) So uth Af ric a Pro spe ctiv e Mo no cen tre n= 31 IN H 10 –1 5 mg /kg RIF 10 –1 5 mg /kg PZ A 30 –4 0 mg /kg EM B 1 5– 25 mg /kg PT B, n= 22 EP TB , n =9 7 + 24 – Me dia n (ra ng e) 2.2 9 (0 .25 –1 0.5 ) Me dia n (ra ng e) 11 .5 (6 .1– 19 ) Ma lno uri she d, n= 20 IN H, R IF, PZ A, EM B C m ax AU C 0– 24 C2 h Ag e, g end er, nu trit ion al sta tu s, HI V sta tus HI V+ as so cia te d wi th low er C2 h o f IN H (P =0 .04 ) NA NA Mu khe rje e 30 (2 01 5) Ind ia Pro spe ctiv e Mu ltic ent re n= 12 7 • G 1, n= 64 • G2 , n =6 3 IN H • G 1: 5 (4 –6 ) mg /kg • G 2: 1 0 (7 –1 5) m g/k g RIF • G 1: 1 0 (8 –1 2) m g/k g • G 2: 1 5 (1 0– 20 ) m g/k g PZ A 30 –3 5 mg /kg EM B 2 0– 25 mg /kg PT B, n= 63 EP TB , n =6 4 No ne Ra nge 0.5 –1 5 M ea n (S D) • G 1: ma lno uri she d: 8.8 (3 .6) no rm al: 8.1 (3 .7) • G 2: ma lno uri she d: 7.6 (3 .2) no rm al: 10 .5 (2 .4) NA Ma lno uri she d, n= 58 IN H, R IF, PZ A, EM B C m ax AU C 0– 4 C2 h Nu trit ion al sta tus , do sin g reg ime n Do sin g r egi me n o n IN H C m ax a nd AU C (P <0 .00 01 ) Fav ou rab le • G 1, n= 53 • G2 , n =4 4 Un fav ou rab le • G 1, n= 9 • G2 , n =1 7 LT FU : • G 1, n= 2 • G2 , n =2 IN H C m ax l ow er in c hild ren wit h u nfa vou rab le o utc om e: 1.3 (0 .7– 1.5 ) μ g/m L vs 3 .4 (1 .8– 5.0 ) μ g/m L, P = 0.0 5 Co nfi rm ati on o f M . tu be rcu los is ass oci ate d w ith po or ou tco me : 55 .6% vs 16 .4% , P = 0.0 1 G2 : ch ildr en wit h lo we r we igh t f or ag e z s co re h ad po ore r o utc om es. 46 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Fi rs t a ut ho r, re fe re nc e no . ( ye ar ) St ud y de sig n Do sin g re gim en Ty pe o f T B HI V Ag e (y ea rs ) Bo dy we igh t ( kg ) N ut rit io na l sta tu s Dr ug s PK pa ra m et er s Co va ria te s Fa ct or s a ffe ct in g P K Cl in ica l ou tc om e Fa ct or s a ffe ct in g c lin ica l ou tc om e Be kke r 1 7 (2 01 6) So uth Af ric a Pro spe ctiv e Mu ltic ent re n= 39 IN H 14 (9 –2 0) m g/k g RIF 14 (9 –2 0) m g/k g PZ A 32 (1 9– 45 ) m g/k g EM B 2 0 (1 3– 29 ) m g/k g PT B, n= 36 TB M , n =1 PT B+ EP TB , n= 2 5 + 34 – M ea n (ra ng e) 0.5 5 (0 –1 ) M ea n (S D) 6.4 5 (1 .67 ) M ea n (S D) W AZ : – 1.6 2 (1 .53 ) W HZ : – 0.4 (1 .26 ) IN H, R IF, PZ A, EM B C m ax AU C 0– 8 Ag e, g end er, nu trit ion al sta tus , pre ma tur ity, HI V sta tu s, eth nic ity Fo rm ula tio n infl ue nc ed RIF C m ax a nd AU C (P <0 .00 5) HI V as so cia te d wi th lo we r PZ A an d EM B C m ax a nd AU C (P <0 .02 ) Fav ou rab le, n= 33 Un fav ou rab le, n= 6 All un fav ou rab le o utc om es we re in c hild ren w ith po or soc ial circ um sta nce s. Mu khe rje e 40 (2 01 6) Ind ia Pro spe ctiv e Mo no cen tre n= 56 IN H 4– 6 m g/k g RIF 8– 12 m g/k g PZ A 30 –3 5 m g/k g EM B 20 –2 5 m g/k g PT B, n= 52 Ple ura l TB , n= 4 As soc iate d EP TB , n =1 9 24 + 32 – Ra nge 0.5 –1 5 M ea n (S D) • HI V+ : 8 .8 (3 .6) • HI V– : 8 .1 (3 .7) NA M ed ian [I Q R] • HI V+ HA Z –2 .5 [– 4.2 , – 1.6 ] W AZ – 3.2 [– 4.5 , – 2.1 ] • HI V– HA Z –1 .4 [– 2.3 , 0 ] W AZ – 1.4 [– 2, –0 .7] IN H, R IF, PZ A, EM B C m ax AU C 0– 4 C2 h Ag e, g end er, nu trit ion al sta tus , NA T2 fo r IN H, do sin g re gim en , H IV sta tus Do sin g r egi me n ass oci ate d w ith lo we r C2 h IN H (P =0 .01 ) Yo un ge r a ge as so cia te d wit h lo we r C 2h IN H (P =0 .04 ) HI V+ as so cia te d wi th low er EM B A UC (P <0 .05 ). NA T2 ge no typ e o n I NH C m ax an d A UC (P <0 .01 ) • HI V+ Fav ou rab le, n= 6 Un fav ou rab le, n= 17 LT FU , n =1 • N A for H IV – No ob ser ved as soc iati on An tw i 1 6 (2 01 7) Gh ana Pro spe ctiv e Mo no cen tre n= 11 3 M ed ian [I Q R] IN H 11 .2 [9 .1– 12 .8] mg /kg RIF 15 .8 [1 3.6 –1 8.8 ] mg /kg PZ A 24 .8 [2 2.6 –3 0] mg /kg EM B 1 6.9 [1 5– 20 .6] mg /kg PT B, n= 85 EP TB , n =2 8 54 + 59 – M ed ian [I Q R] 5 [2 .2– 8.3 ] M ed ian [I Q R] 14 .0 [8 .8– 19 .5] M ed ian [I Q R] HA Z: –2 .0 [– 3.2 , – 1.1 ] W AZ : – 2.5 [– 3.8 , – 1.4 ] IN H, R IF, PZ A, EM B C m ax AU C 0– 8 Ge nd er, NA T2 fo r IN H, do sin g re gim en , H IV sta tus HI V+ w as as so cia te d wi th low er RIF an d E MB C m ax an d A UC an d PZ A AU C (P <0 .03 4) NA T2 ge no typ e o n I NH C m ax an d A UC (P <0 .02 ). Fav ou rab le, n= 99 Un fav ou rab le, n= 6 LT FU , n =8 NA 47 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Fi rs t a ut ho r, re fe re nc e no . ( ye ar ) St ud y de sig n Do sin g re gim en Ty pe o f T B HI V Ag e (y ea rs ) Bo dy we igh t ( kg ) N ut rit io na l sta tu s Dr ug s PK pa ra m et er s Co va ria te s Fa ct or s a ffe ct in g P K Cl in ica l ou tc om e Fa ct or s a ffe ct in g c lin ica l ou tc om e Ra nja lka r 4 5 (2 01 8) Ind ia Pro spe ctiv e Mu ltic ent re n= 41 • G 1, n= 27 • G2 , n =1 4 M ed ian [I Q R] • G 1: T hri ce we ekl y IN H 10 [8 –1 2] mg /kg RIF 10 [9 –1 2] mg /kg • G 2: d aily IN H 8 [7 –9 ] m g/k g RIF 11 [1 0– 12 ] m g/k g PT B, n= 36 • G 1, n= 24 • G2 , n =1 3 Ly m ph no de TB , n= 5 • G 1, n= 3 • G2 , n =2 NA Ra nge 2– 16 M ed ian [I Q R] • G 1: 1 4.7 [1 2– 24 ] • G 2: 3 7 [2 1– 41 ] M ed ian [I Q R] HA Z • G 1: –1 .41 2 [– 2, –0 .56 ] • G 2: –0 .31 6 [– 1.1 5, 0 .22 ] IN H, RIF C m ax AU C 0– 6 C2 h Ag e, g rou p (G 1 v s G 2) No ob ser ved as soc iati on Fav ou rab le • G 1, n= 25 • G2 , n =1 1 Un fav ou rab le • G 1, n= 2 • G2 , n =3 G1 : bo th pat ien ts w ith an un fav ou rab le o utc om e h ad RIF C m ax < 8 μg /m L Da yal 42 (2 01 8) Ind ia Pro spe ctiv e Mo no cen tre n= 37 IN H 10 –1 5m g/k g RIF 10 –2 0 m g/k g PZ A 30 –4 0 m g/k g EM B 15 –2 5 m g/k g PT B, n= 18 EP TB , n =1 9 No ne M ed ian [I Q R] 8 [3 –1 0] NR Stu nte d, n= 14 /21 (< 6 ye ar s) Un de rw eig ht, n= 14 /37 W ast ed , n= 16 /21 (< 6 ye ar s) IN H, P ZA C m ax AU C 0– 8 Ag e, t ype of T B, B MI EP TB as soc iate d w ith low er IN H A UC th an PT B (P =0 .05 ) Ag e > 3 yea rs ass oci ate d wi th h igh er P ZA A UC (P =0 .00 1) Fav ou rab le, n= 35 LT FU , n =2 NA Sha h 4 7 (2 01 9) Ind ia Pro spe ctiv e Mo no cen tre n= 35 IN H 10 m g/k g dai ly PT B, n= 12 EP TB , n =2 2 NA Ra nge 1– 15 NR Un de rw eig ht, n= 11 IN H C m ax AU C 0– 24 Ag e, g end er, Typ e o f TB , for mu lati on , nu trit ion al sta tus No ob ser ved as soc iati on NA NA Pan jas a- wa tw on g 6 1 (2 02 0) Vi et N am Pro spe ctiv e Mo no cen tre n= 10 0 IN H 5 m g/k g RIF 10 m g/k g PZ A 25 m g/k g EM B 1 5 m g/k g TB M, n= 10 0 4 + 92 – 4 N A Me dia n (ra ng e) 3 (0 .2– 15 ) Me dia n (ra ng e) 10 .9 (4 –4 3) Me dia n (ra ng e) HA Z: –1 .64 (– 9.1 7, 2 .21 ) W AZ : – 1.9 3 (– 5.5 2, 2) IN H, R IF, PZ A, EM B C m ax AU C 0– 24 No ne No ob ser ved as soc iati on At 8 m on ths : Fav ou rab le, n= 81 Un fav ou rab le, n= 15 LT FU , n =4 Sev eri ty of infe ctio n a sso cia ted wit h o utc om ea 48 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Fi rs t a ut ho r, re fe re nc e no . ( ye ar ) St ud y de sig n Do sin g re gim en Ty pe o f T B HI V Ag e (y ea rs ) Bo dy we igh t ( kg ) N ut rit io na l sta tu s Dr ug s PK pa ra m et er s Co va ria te s Fa ct or s a ffe ct in g P K Cl in ica l ou tc om e Fa ct or s a ffe ct in g c lin ica l ou tc om e Jus tin e 18 (2 02 0) Un ite d Re pu blic of Ta nz an ia Pro spe ctiv e Mo no cen tre n= 51 IN H 2– 10 m g/k g RIF 5– 20 m g/k g PZ A 10 –4 0 m g/k g EM B 7.5 –3 5 m g/k g PT B, n= 18 EP TB , n =1 7 P+ EP TB , n= 16 No ne Me dia n (ra ng e) 5.3 (0 .75 –1 4) NR Stu nte d, n= 16 /23 Un de rw eig ht, n= 16 /23 IN H, R IF, PZ A, EM B C m ax Ag e, g end er, do sin g reg ime n, nu trit ion al sta tus Do sin g r egi me n o n R IF an d PZ A C m ax (P =0 .00 5) Ma lnu trit ion de cre ase d IN H and RI F C ma x (P =0 .00 1) . NA NA Ga rcia –P rat s 48 (2 02 1) So uth Af ric a Pro spe ctiv e Mu ltic ent re n= 62 RIF • G 1: 1 5– 20 , the n 3 5 m g/k g • G 2: 3 5, t hen 50 m g/k g • G 3: 6 0, t hen 75 m g/k g PT B, n= 45 EP TB , n =2 P+ EP TB , n= 15 No ne Me dia n (ra ng e) • G 1: 2 (1 .2– 3.4 ) • G 2: 2 (1 .1– 3.9 ) • G 3: 2.8 (1 –5 .5) Me dia n (ra ng e) • G 1: 1 0.6 (8 .7– 14 .2) • G 2: 1 0.9 (9 .3– 14 .1) • G 3: 1 2.5 (8 –1 7.4 ) Un de rw eig ht, n= 18 RIF C m ax AU C 0– 24 Do sin g reg ime n An aly sis no t p ub lish ed NA NA IN H , i so ni az id ; R IF , r ifa m pi ci n; P Z A , p yr az in am id e; E M B, e th am bu to l G1 , g ro up 1 ; G 2, g ro up 2 ; C m ax , m ax im um c on ce nt ra tio n; A U C , a re a un de r th e co nc en tr at io n– tim e cu rv e; C 2h , p ea k dr ug c on ce nt ra tio n PT B: pu lm on ar y t ub er cu lo sis , E PT B: ex tra p ulm on ar y t ub er cu lo sis , T BM : tu be rc ulo us m en ing itis H A Z : h ei gh t- fo r- ag e z sc or e (≤ – 2 SD , s tu nt ed ); W A Z : w ei gh t- fo r- ag e z sc or e (≤ – 2 SD , u nd er w ei gh t) ; W H Z , w ei gh t- fo r- he ig ht z s co re ( ≤ –2 S D , w as te d) ; B M I, bo dy m as s in de x; A RT , a nt ire tr ov ira l t he ra py ; R N T C P, R ev ise d N at io na l T B C on tr ol P ro gr am m e (In di a) ; L T FU , l os t to fo llo w -u p aO R , a dj us te d od ds r at io ; C I, co nfi de nc e in te rv al ; I Q R , i nt er qu ar til e ra ng e; N A , n ot a va ila bl e; S D , s ta nd ar d de vi at io n; S EM , s ta nd ar d er ro r of t he m ea n RN TC P gu ide lin es : IN H 10 m g/k g, RM P 10 m g/ kg , P Z A 3 0– 35 m g/ kg , E M B 30 m g/ kg g iv en t hr ic e w ee kl y D os ing re gim en e xp re sse d pe r d ay , e xc ep t i f m en tio ne d ot he rw ise Tr ea tm en t su cc es s w as d efi ne d as : c ur ed / t re at m en t co m pl et ed U nf av ou ra bl e ou tc om e w as d efi ne d as : f ai lu re o r de at h a D ise as e se ve rit y w as b as ed o n th e Bl an ty re C om a Sc or e fo r ch ild re n < 5 y ea rs a nd t he G la sg ow C om a Sc or e fo r ch ild re n ≥ 5 ye ar s. Gr ad e I, “ BC S 4– 5 wi th n o fo ca l n eu ro lo gic al sig ns ” o r “ GC S 15 w ith n o fo ca l ne ur ol og ic al s ig ns ”; G ra de II , “ BC S 2– 3” o r “ BC S 4– 5 w ith fo ca l n eu ro lo gi ca l s ig ns ” or “ G C S 11 –1 4” o r “ GC S 15 w ith fo ca l n eu ro lo gi ca l s ig ns ”; G ra de II I, “ BC S < 1 ” or “ G C S < 1 0” 49 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2The dosing regimens used in most of the studies followed the WHO 2014 recommendations (Table 1). In five studies, the Indian Revised National Tuberculosis Control Programme (RNTCP) recommendations (33) were followed, in which the doses are within the range recommended by WHO but given at a frequency of thrice weekly instead of daily (Table 2). As all the drugs are completely eliminated within 24 h, these data were considered valuable. In one study reported (48), the dose of RIF given was higher than that recommended by WHO (15.5–75 mg/kg); standard doses were used for the other drugs. A direct comparison between higher and current dosing could therefore not be performed. Nonetheless, the PK parameters of currently recommended doses in various settings are summarized below. 3.4 Summary of reported pharmacokinetics parameters A total of 963 patients were included in this systematic review of 18 studies. For this cohort, INH, RIF, PZA and EMB PK parameters were evaluated in 16 (89%), 14 (78%), 13 (72%) and 8 (44%) studies, respectively. All the studies reported PK parameters for Cmax or C2h, and 17 (94.4%) reported AUC, from AUC0–4 h to AUC0–∞, in following distribution: AUC0–4 (n=2), AUC0–5 (n=1), AUC0–6 (n=1), AUC0–8 (n=5), AUC0–12 (n=1), AUC0–24 (n=6) and AUC0–∞ (n=1). Attainment of the target exposure was reported in 8 of the 18 studies, their targets being in line with those listed in Table 1. For INH, 286 of 403 children (71.0%) met the Cmax target, and 48 of 72 (66.7%) achieved the target AUC. With respect to RIF, 175 of 389 children (45.0%) met the Cmax target and 35 of 70 (50.0%) the AUC target. For PZA, 171 of the 330 reported children (51.8%) met the Cmax target, and only 6 of 31 (19.4% were within the AUC target. With respect to EMB, 99 of 269 children (36.8%) met the Cmax target, and no analysis was reported for the AUC. Justine et al. (18) found that 34 of 51 children (66.7%) had low concentrations of all the drugs tested; none met the Cmax targets for RIF (> 8 mg/L) or INH (> 3 mg/L). Reporting of PK parameters differed by study and drug. AUC was reported for the full cohort in 11 (73.3%), 10 (66.7%), 9 (69.2%) and 6 (66.7%) studies for RIF, INH, PZA and EMB, respectively. Another study reported data for Cmax but not for AUC for all drugs. The remaining studies reported AUC and Cmax only for subgroups. Table 3 summarizes reporting of the parameters in each study. Of note, only eight studies reported PK parameters by HIV status, four by age and three by nutritional status for RIF. For INH, four studies reported PK parameters by HIV status, six by age, five by nutritional status and four by NAT2 genotype. For PZA, four studies reported PK parameters by HIV status, four by age and four by nutritional status; and, for EMB, three studies reported by HIV status, two by age and two by nutritional status. PK data for RIF, INH, PZA and EMB are summarized in Annex 4. 50 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Ta bl e 3. N um be rs o f p at ie nt s w ho c on tr ib ut ed e xp os ur e m et ri cs p er s tu dy a nd c ov ar ia te s as r ep or te d in t he o ri gi na l s tu di es R ifa m pi ci n St ud y n A U C C m ax W ho le co ho rt H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge W ho le c oh or t H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge Th ee (3 1) ( 10 m g/ kg ) 11 n= 11 ( 10 0% ) n= 4/ 11 ( 36 .4 % ) – – n= 11 ( 55 .0 % ) n= 4/ 11 ( 36 .4 % ) – – Th ee (3 1) ( 15 m g/ kg ) 11 n= 11 ( 10 0% ) n= 4/ 11 ( 36 .4 % ) – – n= 11 ( 55 .0 % ) n= 4/ 11 ( 36 .4 % ) – – An tw i ( 16 ) 11 3 n= 11 3 (1 00 % ) n= 59 /1 13 ( 52 .2 % ) – – n= 11 3 (1 00 % ) n= 59 /1 13 (5 2. 2% ) – – Be kk er (1 7) 39 n= 39 ( 10 0% ) n= 5/ 39 ( 12 .8 2% ) n= 18 /3 9 (4 6. 2% ) un de rw eig ht n= 39 ( 10 0% ) n= 39 ( 10 0% ) n= 5/ 39 ( 12 .8 2% ) n= 18 /3 9 (4 6. 2% ) un de rw eig ht n= 39 ( 10 0% ) M uk he rje e (4 0) 56 – n= 24 /5 6 (4 2. 9% ) – – – n= 24 /5 6 (4 2. 9% ) n= U nk no w n n= 56 ( 10 0% ) Ra nja lka r ( 50 ) 14 n= 14 ( 10 0% ) – – n= 14 ( 10 0% ) n= 14 ( 10 0% ) – – n= 14 ( 10 0% ) M uk he rje e (3 0) 63 n= 63 ( 10 0% ) – n= 26 /6 3 (4 1. 3% ) m aln ou ris he d – n= 63 ( 10 0% ) – n= 26 /6 3 (4 1. 3% ) m aln ou ris he d Ra m ac ha nd ra n (3 5) a 77 – n= 77 ( 10 0% ) n= 56 /7 7 (7 2. 7% ) un de rw eig ht , n= 59 /7 7 (7 6. 6% ) stu nt ed , n= 15 /7 7 (1 9. 5% ) wa ste d n= 77 ( 10 0% ) – n= 77 ( 10 0% ) n= 56 /7 7 (7 2. 7% ) un de rw eig ht , n= 59 /7 7 (7 6. 6% ) stu nt ed , n= 15 /7 7 (1 9. 5% ) wa ste d n= 77 ( 10 0% ) Ra m ac ha nd ra n (3 4) a 84 – n= 0 – n= 84 ( 10 0% ) – n= 0 – n= 84 ( 10 0% ) Ar ya (3 9) 8 n= 8 (1 00 % ) – – – n= 8 (1 00 % ) – – – Hi ru y ( 41 ) 31 n= 31 ( 10 0% ) – – – n= 31 ( 10 0% ) – – – Pa nja sa wa tw on g ( 61 ) 10 0 n= 10 0 (1 00 % ) – – – n= 10 0 (1 00 % ) – – – 51 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge W ho le c oh or t H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge M lo th a ( 32 ) 28 n= 28 ( 10 0% ) – – – n= 28 ( 10 0% ) – – – Ga rc ia- Pr at s ( 48 ) (1 5– 20 m g/ kg )a 20 n= 20 ( 10 0% ) n= 0 – – n= 20 ( 10 0% ) – – – Ju st in e (1 8) 16 – – – – n= 16 ( 10 0% ) n= 0 To ta l 67 1 11 st ud ies (7 3. 3% ) 43 8 pa tie nt s (6 5. 3% ) 8 st ud ie s (5 7. 1% ) 17 3 pa tie nt s (2 5. 8% ) 3 st ud ie s (2 1. 4% ) 26 p at ien ts m aln ou ris he d (3 .9 % ) 74 p at ien ts un de rw eig ht (1 1. 0% ) 15 p at ien ts w as te d (2 .2 % ) 59 p at ien ts st un te d (8 .8 % ) 4 stu die s (2 6. 7% ) 21 4 pa tie nt s (3 2. 7% ) 12 st ud ies (8 0. 0% ) 45 4 pa tie nt s (6 7. 6% ) 8 st ud ie s (5 7. 1% ) 17 3 pa tie nt s (2 5. 8% ) 4 st ud ie s (2 6. 7% ) 26 p at ien ts m aln ou ris he d (3 .9 % ) 74 p at ien ts un de rw eig ht (1 1. 0% ) 15 p at ien ts w as te d (2 .2 % ) 59 p at ien ts st un te d (8 .8 % ) 5 stu die s (3 3. 3% ) 27 0 pa tie nt s (4 1. 22 % ) a E xp os ur e w as n ot r ep or te d by H IV s ta tu s, bu t th e H IV s ta tu s of a ll pa rt ic ip an ts w as t he sa m e. Is on ia zi d 52 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV - po si ti ve ch ild re n M al no ur is he d ch ild re n A ge N A T 2 ge no ty pe W ho le co ho rt H IV - po si ti ve ch ild re n M al no ur is he d ch ild re n A ge N A T 2 ge no ty pe Th ee (3 1) 20 n= 20 ( 10 0% ) n= 5/ 20 (2 5. 0% ) – – n= 8/ 20 ( 40 % ) slo w, n= 4/ 20 ( 20 % ) int er m ed iat e, n= 8/ 20 ( 40 % ) fa st n= 20 (1 00 % ) n= 5/ 20 (2 5. 0% ) – – n= 8/ 20 ( 40 % ) slo w, n= 4/ 20 ( 20 % ) int er m ed iat e, n= 8/ 20 ( 40 % ) fa st Hi ru y ( 41 ) 31 n= 31 ( 10 0% ) – – – – n= 31 (1 00 % ) – – – – An tw i ( 16 ) 11 3 n= 11 3 (1 00 % ) n= 59 /1 13 (5 2. 2% ) – – – n= 11 3 (1 00 % ) n= 59 /1 13 (5 2. 2% ) – – – Be kk er (1 7) 39 n= 39 ( 10 0% ) n= 5/ 39 (1 2. 8% ) n= 18 /3 9 (4 6. 2% ) un de rw eig ht n= 39 (1 00 % ) – n= 39 (1 00 % ) n= 5/ 39 (1 2. 8% ) n= 18 /3 9 (4 6. 2% ) un de rw eig ht n= 39 (1 00 % ) – M uk he rje e (4 0) 56 – n= 24 /5 6 (4 2. 9% ) – – – – n= 24 /5 6 (4 2. 9% ) n= un kn ow n n= 56 (1 00 % ) – Ra nja lka r ( 45 ) 14 n= 14 ( 10 0% ) – – n= 14 (1 00 % ) – n= 14 (1 00 % ) – – n= 14 (1 00 % ) – M uk he rje e (3 0) 63 n= 63 ( 10 0% ) – n= 26 /6 3 (4 1. 3% ) m aln ou ris he d – – n= 63 (1 00 % ) – n= 26 /6 3 (4 1. 3% ) m aln ou ris he d – n= 15 /6 3 (2 3. 8% ) slo w, n= 47 /6 3 (7 4. 6 % ) fas t (N A , n = 1, 0 .0 2% ) 53 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV - po si ti ve ch ild re n M al no ur is he d ch ild re n A ge N A T 2 ge no ty pe W ho le co ho rt H IV - po si ti ve ch ild re n M al no ur is he d ch ild re n A ge N A T 2 ge no ty pe Ra m ac ha nd ra n (3 5) a 77 – n= 77 n= 56 /7 7 (7 2. 7% ) un de rw eig ht n= 59 /7 7 (7 6. 6% ) stu nt ed n= 15 /7 7 (1 9. 5% ) wa ste d n= 77 (1 00 % ) n= 52 /7 7 (6 7. 5% ) slo w n= 25 /7 7 (3 2. 5% ) fas t – – n= 56 /7 7 (7 2. 7% ) un de rw eig ht n= 59 /7 7 (7 6. 6% ) stu nt ed n= 15 /7 7 (1 9. 5% ) wa ste d n= 77 (1 00 % ) n= 52 /7 7 (6 7. 5% ) slo w, n= 25 /7 7 (3 2. 5% ) fas t Ra m ac ha nd ra n (3 4) a 84 – n= 0 – n= 84 (1 00 % ) – – – – n= 84 (1 00 % ) – Sh ah (4 7) 35 n= 35 ( 10 0% ) – n= 11 /3 5 (3 1. 4% ) m aln ou ris he d n= 35 (1 00 % ) n= 26 /3 5 (7 4. 3% ) slo w, n= 9/ 35 ( 25 .7 % ) fas t n= 35 (1 00 % ) – n= 11 /3 5 (3 1. 4% ) m aln ou ris he d n= 35 (1 00 % ) n= 26 /3 5 (7 4. 3% ) slo w, n= 9/ 35 ( 25 .7 % ) fas t D ay al (4 2) 37 n= 37 ( 10 0% ) – n= 14 /3 7 (3 7. 8% ) un de rw eig ht , n= 14 /3 7 (3 7. 8% ) stu nt ed , n= 16 /3 7 (4 3. 2% ) w as te d n= 37 (1 00 % ) – n= 37 (1 00 % ) – n= 14 /3 7 (3 7. 8% ) un de rw eig ht , n= 14 /3 7 (3 7. 8% ) stu nt ed , n= 16 /3 7 (4 3. 2% ) w as te d n= 37 (1 00 % ) – Ju st in e (1 8) 4 – – – – – n= 4 (1 00 % ) n= 0 – – – Ra ng ar i ( 60 ) 8 n= 8 (1 00 % ) – – – – n= 8 (1 00 % ) – – – – 54 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV - po si ti ve ch ild re n M al no ur is he d ch ild re n A ge N A T 2 ge no ty pe W ho le co ho rt H IV - po si ti ve ch ild re n M al no ur is he d ch ild re n A ge N A T 2 ge no ty pe M lo th a ( 32 ) 30 n= 30 ( 10 0% ) – – – – n= 30 (1 00 % ) – – – – Pa nja sa wa tw on g ( 61 ) 10 0 – – – – n= 28 /1 00 ( 28 % ) slo w, n= 47 /1 00 ( 47 % ) int er m ed iat e, n= 17 /1 00 ( 17 % ) fas t (N A , n = 8/ 10 0, 8% ) – – – – n= 28 /1 00 ( 28 % ) slo w, n= 47 /1 00 ( 47 % ) int er m ed iat e, n= 17 /1 00 ( 27 % ) fas t (N A , n = 8, 8 % ) To ta l 71 1 10 st ud ies (6 6. 7% ) 39 0 pa tie nt s (5 4. 9% ) 4 stu die s (2 6. 7% ) 93 p at ien ts (1 3.1 % ) 5 stu die s (3 3. 3% ) 37 p at ien ts m aln ou ris he d (5 .2 % ) 88 p at ien ts un de rw eig ht (1 2. 4% ) 31 p at ien ts w as te d (4 .4 % ) 73 p at ien ts stu nt ed (1 0. 3% ) 6 stu die s (4 0. 0% ) 28 6 pa tie nt s (4 0. 2% ) 4 st ud ie s (2 6. 7% ) 22 4 pa tie nt s (3 1. 5% ) 11 stu die s (7 3. 3% ) 39 4 pa tie nt s (5 5. 4% ) 4 stu die s (2 6. 7% ) 93 p at ien ts (1 3.1 % ) 6 stu die s (4 0. 0% ) 37 p at ien ts m aln ou ris he d (5 .2 % ) 88 p at ien ts un de rw eig ht (1 2. 4% ) 31 p at ien ts w as te d (4 .4 % ) 73 p at ien ts stu nt ed (1 0. 3% ) 7 stu die s (4 6. 7% ) 34 2 pa tie nt s (4 8. 1% ) 5 st ud ie s (3 3. 3% ) 28 6 pa tie nt s (4 0. 2% ) a E xp os ur e w as n ot r ep or te d by H IV s ta tu s, bu t th e H IV s ta tu s of a ll pa rt ic ip an ts w as t he sa m e. Py ra zi na m id e 55 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge W ho le co ho rt H IV p os it iv e ch ild re n M al no ur is he d ch ild re n A ge Th ee (3 1) 20 n= 20 ( 10 0% ) n= 5/ 20 ( 25 % ) – – n= 20 ( 10 0% ) n= 5/ 20 ( 25 % ) – – Hi ru y ( 41 ) 31 n= 31 ( 10 0% ) – – – n= 31 ( 10 0% ) – – – An tw i ( 16 ) 11 0 n= 11 0 (1 00 % ) n= 59 /1 10 (5 3. 6% ) – – n= 11 0 (1 00 % ) n= 59 /1 10 (5 3. 6% ) – – Be kk er (1 7) 39 n= 39 ( 10 0% ) n= 5/ 39 ( 12 .8 % ) n= 18 /3 9 (4 6. 2% ) un de rw eig ht n= 39 ( 10 0% ) n= 39 ( 10 0% ) n= 5/ 39 ( 12 .8 % ) n= 18 /3 9 (4 6. 2% ) un de rw eig ht n= 39 ( 10 0% ) M uk he rje e (4 0) 56 – n= 24 /5 6 (4 2. 9% ) – – – n= 24 /5 6 (4 2. 9% ) n = U nk no w n n= 56 ( 10 0% ) M uk he rje e (3 0) 12 7 n= 12 7 (1 00 % ) – n= 58 /1 27 ( 45 .7 % ) m aln ou ris he d – n= 12 7 (1 00 % ) – n= 58 /1 27 ( 45 .7 % ) m aln ou ris he d – Ra m ac ha nd ra n (3 5) 77 – – n= 56 /7 7 (7 2. 7% ) un de rw eig ht , n= 59 /7 7 (7 6. 6% ) stu nt ed , n= 15 /7 7 (1 9. 5% ) wa ste d n= 77 ( 10 0% ) – – n= 56 /7 7 (7 2. 7% ) un de rw eig ht , n= 59 /7 7 (7 6. 6% ) stu nt ed , n= 15 /7 7 (1 9. 5% ) wa ste d n= 77 ( 10 0% ) Ra m ac ha nd ra n (3 4) 84 – – – n= 84 ( 10 0% ) – – – n= 84 ( 10 0% ) D ay al (4 2) 37 n= 37 ( 10 0% ) – n= 14 /3 7 (3 7. 8% ) un de rw eig ht , n= 14 /3 7 (3 7. 8% ) stu nt ed , n= 16 /3 7 (4 3. 2% ) wa ste d n= 37 ( 10 0% ) n= 37 ( 10 0% ) – n= 14 /3 7 (3 7. 8% ) un de rw eig ht , n= 14 /3 7 (3 7. 8% ) stu nt ed , n= 16 /3 7 (4 3. 2% ) wa ste d n= 37 ( 10 0% ) Ju st in e (1 8) 12 – – – – n= 12 ( 10 0% ) – – – Ro y ( 43 ) 7 n= 7 (1 00 % ) – – – n= 7 (1 00 % ) – – – 56 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge W ho le co ho rt H IV p os it iv e ch ild re n M al no ur is he d ch ild re n A ge M lo th a ( 32 ) 29 n= 29 ( 10 0% ) – – – n= 29 ( 10 0% ) – – – Pa nja sa wa tw on g ( 61 ) 10 0 n= 10 0 (1 00 % ) – – – n= 10 0 (1 00 % ) – – – To ta l 72 9 9 stu die s (6 9. 2% ) 50 0 pa tie nt s (6 8. 6% ) 4 st ud ie s (3 0. 8% ) 93 p at ien ts (1 2. 8% ) 4 st ud ie s (3 0. 8% ) 58 p at ien ts m aln ou ris he d (8 .0 % ) 88 p at ien ts un de rw eig ht (1 2.1 % ) 31 p at ien ts wa ste d (4 .3 % ) 73 p at ien ts stu nt ed (1 0. 0% ) 4 stu die s (3 0. 8% ) 23 7 pa tie nt s (3 2. 5% ) 10 st ud ies (7 6. 9% ) 51 2 pa tie nt s (7 0. 2% ) 4 st ud ie s (3 0. 8% ) 93 p at ien ts (1 2. 8% ) 5 st ud ie s (3 8. 5% ) 58 p at ien ts m aln ou ris he d (8 .0 % ) 88 p at ien ts un de rw ei gh t (1 2.1 % ) 31 p at ien ts wa ste d (4 .3 % ) 73 p at ien ts stu nt ed (1 0. 0% ) 5 stu die s (3 8. 5% ) 29 3 pa tie nt s (4 0. 2% ) Et ha m bu to l St ud y n A U C C m ax W ho le co ho rt H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge W ho le c oh or t H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge Hi ru y ( 41 ) 13 n= 13 ( 10 0% ) – – – n= 13 ( 10 0% ) – – – An tw i ( 16 ) 11 0 n= 11 0 (1 00 % ) n= 59 /1 10 (5 3. 6% ) – – n= 11 0 (1 00 % ) n= 59 /1 10 (5 3. 6% ) – – Be kk er (1 7) 16 n= 16 ( 10 0% ) n= 2/ 16 ( 12 .5 % ) n= 7/ 16 ( 43 .8 % ) un de rw eig ht n= 16 ( 10 0% ) n= 16 ( 10 0% ) n= 2/ 16 ( 12 .5 % ) n= 7/ 16 ( 43 .8 % ) un de rw eig ht n= 16 ( 10 0% ) 57 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y n A U C C m ax W ho le co ho rt H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge W ho le c oh or t H IV -p os it iv e ch ild re n M al no ur is he d ch ild re n A ge M uk he rje e (4 0) 56 – n= 24 /5 6 (4 2. 9% ) – – – n= 24 /5 6 (4 2. 9% ) n = U nk no w n – M uk he rje e (3 0) 12 7 n= 12 7 (1 00 % ) – n= 58 /1 27 ( 45 .7 % ) m aln ou ris he d n= 12 7 (1 00 % ) n= 12 7 (1 00 % ) – n= 58 /1 27 (4 5. 7% ) m aln ou ris he d – Ju st in e (1 8) 12 – – – – n= 12 ( 10 0% ) – – – M lo th a ( 32 ) 28 n= 28 ( 10 0% ) – – – n= 28 ( 10 0% ) – – – Pa nja sa wa tw on g ( 61 ) 10 0 n= 10 0 (1 00 % ) – – – n= 10 0 (1 00 % ) – – – To ta l 46 2 6 stu die s (6 6. 7% ) 39 4 pa tie nt s (8 5. 3% ) 3 st ud ie s (3 3. 3% ) 85 p at ien ts (1 8. 5% ) 2 st ud ie s (2 2. 2% ) 7 pa tie nt s (1 .5 % ) un de rw eig ht 58 p at ien ts (1 2. 6% ) m aln ou ris he d 2 stu die s (2 2. 2% ) 14 3 pa tie nt s (3 1. 0% ) 7 stu die s (7 7. 8% ) 40 6 pa tie nt s (8 7. 9% ) 3 st ud ie s (3 3. 3% ) 85 p at ien ts (1 8. 5% ) 3 st ud ie s (3 3. 3% ) 7 pa tie nt s (1 .5 % ) un de rw eig ht 58 p at ien ts (1 2. 6% ) m aln ou ris he d 1 stu dy (1 1. 1% ) 16 p at ien ts (3 .5 % ) 58 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 None of the studies assessed PK parameters by body weight. Therefore, raw data from confidential unpublished studies involving shorter treatment for minimal TB in children (SHINE, n=76, only RIF data reported), optimal dosing with first-line anti-TB and antiretroviral drugs in children, a pharmacokinetics study (Datic, n=179) and pooled published data (13,19) were used to investigate the range of exposure by weight band (Fig. 2). Guiastrennec et al. (19) reported data by weight bands pooled from two studies included in this systematic review (30,33). Zvada et al. (13) also reported pooled data from other studies, only one of which (31) is included in this review, and the remaining studies (36–38) did not meet our inclusion criteria. None of the studies listed in Figure 2 reported data for EMB by weight band. Fig. 2. AUC for rifampicin, isoniazid and pyrazinamide by body weight band Boxplots represent 5th, 25th, 50th, 75th and 95th percentiles. The numbers of children included were 76 (SHINE), 179 (Datic), 161 (19) and 76 (13). Dashed lines represent the target AUC from Table 1. 3.4.1 Rifampicin Summary estimates The summary estimate for the 13 studies that reported the AUC for RIF was 23.4 µg.h/mL, while the target was 38.7 µg.h/mL (Fig. 3a). In one study, the mean estimate achieved the target; in two studies, the 95% CI for one group crossed the target threshold but not the mean; in all the remaining studies, the 95% CI was below the target. All exposure values by subgroup reported in the original studies are listed in Annex 4. 59 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Fig. 3. Forest plot of estimated rifampicin AUC (a) and Cmax (b). Dashed lines represent the target. Each square is proportional to the sample size in each group. A. B. The summary estimate for the 14 studies that reported RIF Cmax was 6 µg/mL, while the target was 8 µg/mL (Fig. 3b). The mean estimate was greater than the target in two studies and in one group in two studies. In three studies, the 95% CI for one group crossed the target threshold, but the mean was below the threshold. The estimate was below the target in all the remaining studies. The heterogeneity in Cmax and AUC between studies was high: AUC I2=98.3% (95% CI 97.5–99.3); Cmax I2=96.9% (95% CI 98.0–99.5). Subgroup analysis and predictors Dose. Increased RIF dose was associated with increased PK in four studies (18,31,32,39) yet so far, pharmacokinetic studies that have evaluated these changes are relatively limited. We evaluated plasma drug concentrations of rifampicin (RIF (Table 2). Too few studies reported PK parameters by dose group for a meta-analysis. Dose comparisons in two studies showed that Cmax and AUC0–12 were higher at doses 60 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 > 10 mg/kg (31,39). Two studies confirmed this trend for AUC0–last, reporting an increase of 0.12 µg.h/mL for each additional mg/kg (P = 0.028) and for RIF peak concentration with an increase of 0.2 µg/mL (95% CI 0.1–0.4) for each mg/kg (P = 0.005) (18,32). In the only study in which doses ≥ 20 mg/kg were used (44), the results (unpublished) show a wide range in Cmax and AUC0–24 among children receiving 15–60 mg/kg. The steady-state median (range) AUC0–24 was 39.5 (11.7–76.1) µg.h/mL at 15–20 mg/kg, 68.4 (18.9–169) µg.h/mL at 35 mg/kg and 192.8 (17.2–415.6) µg.h/mL at 60 mg/kg. The steady-state median (range) Cmax was 8.4 (3.1–15.5) µg/mL at 15–20 mg/kg, 13.7 (4.8–29.5) µg/mL at 35 mg/kg and 28.4 (3.5–47.4) µg/mL at 60 mg/kg. HIV status. HIV-positive children had significantly lower Cmax and AUC in one study (16) (Table 2). Univariate analysis indicated no significant effect of HIV status; however, there was a trend towards lower RIF AUC (Fig. 4). The estimates within and between studies varied widely. Fig. 4. Forest plot of estimated rifampicin AUC (a) and Cmax (b) by HIV status. Dashed lines represent the target. Each square is proportional to the sample size in each group. A. B. 61 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Age. Younger age was associated with significantly lower exposure in one study (34) (Table 2). Summary estimates by age are shown in Fig. 5. Younger children (0–< 6 years) had a lower RIF AUC, estimated as 14.0 mg.h/L versus 22.0 mg.h/L (P = 0.03). The RIF Cmax tended to be lower in younger children, but the trend was not significant in the subgroup analysis. As age stratification differed by study, the estimated differences by group should be interpreted with caution. Fig. 5. Forest plot of estimated rifampicin AUC (a) and Cmax (b) by age group. Dashed lines represent the target. Each square is proportional to the sample size in each group A. B. Malnutrition. Nutritional status was associated with exposure to RIF in two studies (18,34) (Table 2). Ramachandran et al. (34) reported an increase in AUC of 2.8 µg.h/mL per unit WAZ and an increase in Cmax of 0.6 µg/mL per unit WAZ. Justine et al. (18) reported lower RIF C2h (–2.03 µg/mL) in children with any malnutrition (P = 0.001). Too few studies that reported PK parameters by the same measure of nutritional status were available for a test of association with malnutrition. 62 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Other variables. Two additional significant associations were observed, showing an impact on PK parameters of the drug formulation (17) and of serum albumin level, with a lower Cmax in the latter case (35). 3.4.2 Isoniazid Summary estimates The summary estimate from the 11 studies that reported INH AUC was 23.4 µg.h/mL (95% CI 18.6– 28.2), while the target was 23.4 µg.h/mL (Fig. 6a). In six studies or study arms, the mean estimate achieved the target; in four studies or study arms, the 95% CI crossed the target threshold but not the mean; in all the remaining studies, the 95% CI was below the target. The summary estimate for the 11 studies that reported INH Cmax was 5.6 µg/mL, while the target was 3–5 µg/mL (Fig. 6b). The mean estimate was greater than the upper limit of the target in four groups and in seven additional studies. In two studies, the mean estimate was within the target, and the 95% CI in one study crossed the target threshold, but the mean was below the threshold. In only one arm of one study were the mean estimate and the 95% CI below the target. Wide heterogeneity in Cmax and AUC was seen between studies: AUC I2=98.2% (95% CI 97.5–99.6); Cmax I2=97.8% (95% CI 95.2–99.1). Fig. 6. Forest plot of estimated isoniazid AUC (a) and Cmax (b). Dashed lines represent the target. Each square is proportional to the sample size in each group A. 63 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2B. Subgroup analysis and predictors Dose. INH dose was associated with increased PK parameters in four studies (30,31,40,60) (Table 2). As for RIF, too few studies reported PK parameters by dose group for performance of a meta-analysis. Three studies reported study-specific dose comparisons (30,31,60), but between the current WHO-recommended dose of 10 mg/kg and a lower dose (5 mg/kg), which showed that the currently recommended dose was significantly associated with a higher AUC and/or Cmax. HIV status. HIV-positive children had lower AUC than HIV-negative children (summary estimates, 18.7 and 20.0, respectively) (Fig. 7a), although the association was not statistically significant. No trend was observed in the Cmax between HIV-positive and -negative children. The estimates within and between studies varied widely. Hiruy et al. (41) reported a statistically significant relation between HIV status and INH PK (P < 0.04), with a lower INH Cmax in HIV-positive children (mean Cmax per group not reported). Fig. 7. Forest plot of estimated isoniazid AUC (a) and Cmax (b) by HIV status. Dashed lines represent the target. Each square is proportional to the sample size in each group. A. 64 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 B. Age. Age was reported as a statistically significant covariate in three publications (30,33,40) (Table 2 and Annex 4). Ramachandran et al. (35) found an increased INH Cmax of 0.4 (95% CI, 0.19–0.62) per year (P < 0.001) and an increase in AUC0–8 of 1.3 (95% CI, 0.43–2.2) per year (P < 0.01) after correction for NAT2 status. Similarly, Mukherjee et al. (40) showed that young age was a significant predictor of INH C2h in multivariate regression analysis (P = 0.04), although no values were reported. Summary estimates by age are shown in Fig. 8. Younger children had a lower INH AUC, which was not significantly different from that of older children in the subgroup analysis. There was wide heterogeneity among studies, perhaps because of different age stratification; therefore, the estimated differences by group should be interpreted with caution. Fig. 8. Forest plot of estimated isoniazid AUC (a) and Cmax (b) by age group. Dashed lines represent the target. Each square is proportional to the sample size in each group A. 65 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2B. NAT2. NAT2 genotype was reported to have a significant effect on exposure to INH (Cmax and AUC) in five studies (Table 2). One study (31) reported PK according to slow, intermediate and fast status, while the remaining studies stratified patients as only fast or slow acetylators. Subgroup analysis was therefore performed for slow and fast acetylators to ensure that most patients were included in the analysis. Summary estimates by NAT2 status are shown in Fig. 9. Exposure to INH reported as AUC was significantly higher (P < 0.01) in slow metabolizers, at a summary estimate of 35.3 µg.h/mL, than in fast metabolizers, 14.2 µg.h/mL. The mean estimate and the 95% CI were below the target for all fast metabolizer groups, except in one study arm, for which the 95% CI crossed the target. The Cmax was also significantly higher for slow metabolizers than fast metabolizers (P = 0.03). For this metric, however, the median was above the lower bound of the Cmax target in all studies. These findings are shown in Fig. 9, which also shows the wide heterogeneity among the groups in the NAT2 genotype, with an I2 statistic of 95.4% (95% CI 93.9–99.5) for AUC and 92.8% (95% CI 86.5–98.7) for Cmax. Fig. 9. Forest plot of estimated isoniazid AUC (a) and Cmax (b) by acetylator status. Dashed lines represent the target. Each square is proportional to the sample size in each group A. 66 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 B. Malnutrition. Nutritional status was associated with exposure to INH in one study (18) (Table 2). Justine et al. (18) reported lower INH C2h (–2.1 µg/mL) in children with any malnutrition (P = 0.001). Too few studies reported PK parameters by the same measure of nutritional status for a test of an association with malnutrition. Other variables. An additional significant association was observed, with PK parameters of the location of TB disease. Dayal et al. (42) found that extrapulmonary TB was associated with a lower AUC0–8 than pulmonary TB (P = 0.05) (Table 2). 3.4.3 Pyrazinamide Summary estimates The summary estimate for the 14 studies or study arms that reported on PZA AUC was 201.2 µg.h/mL, while the target was 238–428 µg.h/mL (Fig. 10a). In four studies, the mean estimate achieved the lower bound of the AUC target; in two studies, the 95% CI crossed the target lower bound threshold but not the mean; in all the remaining studies, the 95% CI was below the lower bound of the target. The upper bound of the target threshold was not exceeded in any study. The summary estimate for the 14 studies that reported PZA Cmax was 39.6 µg/mL, while the target was 35–60 µg/mL (Fig. 10b). The mean estimate was within the target in nine studies. In one study, the 95% CI of one group crossed the lower bound of the target threshold, but the mean was below the threshold. The mean was completely below the target in four studies, and the 95% CI crossed the upper limit of the Cmax threshold in the two arms of one study. Wide heterogeneity was seen in Cmax and AUC among studies: AUC I2=97.3% (95% CI, 92.8–98.6); Cmax I2=97.4% (95% CI, 94.3–99.0). 67 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Fig. 10. Forest plot of estimated pyrazinamide AUC (a) and Cmax (b). Dashed lines represent the target. Each square is proportional to the sample size in each group A. B. Subgroup analysis and predictors Dose. In one study, PZA was assessed at to two doses, 25 and 35 mg/kg. Cmax and AUC0–5 increased significantly with dose, with Cmax equal to 30.0 (26.2–33.7) µg/mL at 25 mg/kg and 47.1 (42.6–51.6) µg/mL at 35 mg/kg (P < 0.001), and AUC0–5 equal to 118.0 (101.3–134.7) µg.h/mL at 25 mg/kg and 175.2 (155.5–195) µg.h/mL at 35 mg/kg (P < 0.001) (31). Doses < 30 mg/kg were compared with 30–35 mg/ kg in two studies. Both showed significant increases in Cmax at higher doses: 11.78 (2.43–23.57) µg/mL vs 28.17 (13.33–32.52) µg/mL (P = 0.015) and 41.7 (SEM, 1.2) µg/mL vs 49.4 (SEM, 2.8) µg/mL (P = 0.008) (18,34). This was confirmed by Roy et al. (43), who found that AUC0–24 was 278.4 (SEM 16) µg.h/ mL at < 30 mg/kg and 369.5 (SEM 35.3) µg.h/mL at 30–35 mg/kg (P = 0.01). 68 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 HIV status. HIV-positive children had a lower AUC than HIV-negative children (summary estimates, 162.3 and 184.3, respectively) (Fig. 11a), but the association was not statistically significant (P = 0.29). The estimates for both HIV-positive and HIV-negative children were below the lower bound of the AUC target. No trend or association was observed in Cmax between HIV-positive and negative children (P = 0.52). Wide variation was seen in estimates within and between studies. PZA AUC0–8 was lower in HIV-positive patients in one study (16): HIV-positive, 126.53 (105.41–182.34) µg.h/mL, and HIV-negative, 151.04 (124.64–188.81) µg.h/mL (P = 0.034). Bekker et al. (17) found a similar association in children < 1 year of age. Fig. 11. Forest plot of estimated pyrazinamide AUC (a) and Cmax (b) by HIV status. Dashed lines represent the target. Each square is proportional to the sample size in each group A. B. 69 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Age. Age was reported as a significant covariate in three publications (Table 2). Ramachandran et al. (35) found significantly lower exposure of children < 5 years of age than those aged ≥ 5 years. Separately, Cmax increased by 1.2 (95% CI, 0.23–2.18) µg/mL per age year (P < 0.05), and AUC0–8 increased by 7.46 (95% CI, 1.97–12.94) µg.h/mL per age year (P < 0.01) in multiple regression analysis. Dayal et al. (42) found similar results, with an increase of 8.4 (95% CI, 3.6–13.1) µg.h/mL in PZA AUC0–8 per age year (P = 0.001). Summary estimates by age are shown in Fig. 12; no significant differences in AUC or Cmax were found between younger and older children. The heterogeneity among studies was, however, large, and the age stratification differed by study; therefore, the estimated differences by group should be interpreted with caution. Fig. 12. Forest plot of estimated pyrazinamide AUC (a) and Cmax (b) by age group. Dashed lines represent the target. Each square is proportional to the sample size in each group A. B. 70 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Malnutrition. Nutritional status was not associated with PZA exposure in any of the studies. Too few studies reported PK parameters by the same measure of nutritional status for a test of the association of malnutrition with exposure to PZA. 3.5 Ethambutol Summary estimates The summary estimate for the four studies in which EMB AUC was reported was 7.2 µg.h/mL, while the target was 16–28 µg.h/mL (Fig. 13a). In all the studies, the 95% CI was below the lower bound of the target; therefore, the AUC target was not reached in any study. The summary estimate for the four studies in which EMB Cmax was reported was 1.4 µg/mL, while the target was 2–6 µg/mL (Fig. 13b). The mean estimate was below the target in all the studies. In one study, the 95% CI for one group reached the lower bound of the target threshold, but the mean was below the threshold. Wide heterogeneity was seen in Cmax and AUC between studies: AUC I2=98.2% (95% CI, 90.7–99.8); Cmax I2=77.2% (95% CI, 43.1–99.3). Fig. 13. Forest plot of estimated ethambutol AUC (a) and Cmax (b). Dashed lines represent the target. Each square is proportional to the sample size in each group A. B. 71 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Subgroup analysis and predictors Few studies included ethambutol subgroups of interest. The effect of age was studied in one publication and nutritional status and HIV status in three. Therefore, no subgroup meta-analysis was conducted for this drug. The only significant associations with EMB exposure in the original studies were HIV status (16,17,40), with lower Cmax and/or AUC values reported in HIV-positive patients. 3.6 Association between attainment of pharmacokinetics target  and clinical outcomes Of the 18 studies included in this review, only 10 reported clinical outcomes (Table 2). Of these, four reported an association between exposure to INH, RIF or PZA (Cmax and/or AUC) and treatment outcome (30,34,35,45). In general, the treatment outcome was favourable if the child was cured or if treatment was completed and unfavourable if the child died or required a change or extension of treatment. In 127 children, of whom 26 (20.5%) had an unfavourable outcome, a lower INH Cmax (1.3 (0.7–1.5) µg/mL) was associated with an unfavourable outcome (3.4 (1.8–5.0) µg/mL for those with a favourable outcome, P = 0.05) (30). Ramachandran et al. (34) found a similar trend, with increased Cmax for INH and RIF associated with favourable outcomes in a cohort of 84 children, of whom 15 (17.9%) had an unfavourable outcome (INH Cmax of 4.3 (2.5–6.3) µg/mL in unfavourable outcome vs 6.3 (4.2–8.4) µg/mL in favourable outcome (P = 0.031); INH AUC0–8 (15.7 (9.6–21.4) µg.h/mL in unfavourable outcome vs 25.4 (15.3–35.2) µg.h/mL in favourable outcome (P = 0.017); RIF Cmax (3.4 (2.5–4.2) µg/mL in unfavourable outcome vs 5.9 (4.4–7.1) µg/mL in favourable outcome (P = 0.002); RIF AUC0–8 (14 (9.2–22.4) µg.h/mL in unfavourable outcome vs 29.4 (18–34.4) µg.h/mL in favourable outcome (P = 0.003)). Ramachandran et al. (35) reported that the PZA Cmax also affected clinical outcomes (adjusted OR, 1.1; 95% CI 1–1.2, P = 0.011) in a cohort of 77 patients, of whom 18 (23.4%) had an unfavourable outcome. 3.7 Adverse events Only four studies reported adverse events as a consequence of TB treatment (16,17,47,48). Bekker et al. (17) showed that anti-TB treatments were well tolerated in a cohort of 39 infants. Five (7%) patients experienced an increase in alanine aminotransferase activity, which was reported as a grade-1 adverse event (increased twice) in three patients, one grade-2 adverse event (increased thrice) and one grade-3 adverse event (elevated seven times), which occurred during a 6-month follow-up after treatment, all of which resolved spontaneously. Antwi et al. (16) reported a grade-3 increase in aspartate aminotransferase activity in one HIV-infected patient among 113 patients. No other serious adverse events were reported. Shah et al. (47) reported that 3 of 35 children developed hepatitis. No correlation was reported between PK measures and the occurrence of adverse events. In the study with doses higher than those currently recommended by WHO (48), two (3%) children discontinued high-dose RIF treatment for reasons of safety or tolerability. No grade-3 or -4 adverse event possibly related to the drug occurred during the study. 72 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 4. Discussion In this systematic review of the PK of exposure to first-line anti-TB drugs in children, we found few relevant studies of PK parameters achieved by doses higher than those currently recommended. One study was identified (OptiRif) in which higher doses of RIF were assessed with standard doses of INH, PZA and EMB (48). All the other studies evaluated standard WHO doses. In these studies, there was considerable variation in the PK of all drugs, and the studies were highly heterogeneous. The summary estimates indicate that, with currently recommended doses, the target was attained only for INH AUC, INH Cmax and PZA Cmax. First-line anti-TB drugs have been prescribed at the same mg/kg dose in children as in adults (26). This approach was reviewed in WHO TB treatment guidelines in 2010 and 2014. In 2010, the recommendations for INH were adjusted (49), as it was established that children achieved substantially lower exposure than adults, and empirical increases were made in paediatric doses (e.g., from 10 mg/kg to 15 mg/kg for RIF). These increases were, however, based on only a few pharmacokinetics studies that included small numbers of children (46–48). In the current review, exposure of children to RIF was consistently low in all the studies. We found that higher doses within the WHO recommended dose range resulted in higher exposure; however, these exposures were still lower than the adult AUC target (38.7 µg.h/mL). This finding suggests that doses > 15 mg/kg RIF daily would be required to match the exposure of adults treated with 10 mg/kg RIF. Modelling and simulation studies predict that at least 25 mg/kg may be required to ensure that children attain PK target exposure (12,19,21,34). It has also been shown that higher PK target attainment is associated with higher proportions of favourable clinical outcomes. In terms of the safety of increasing doses of RIF in children, the Opti-Rif trial indicates that doses up to 60 mg/kg RIF are safe, with no related grade-3 or -4 adverse events (48). The AUC0–24 in children dosed at 60 mg/kg RIF was similar to that of adults treated with 35 mg/kg RIF, which was considered in a randomized controlled trial in adults to be safe and efficacious (51). No other study was found in which doses of RIF higher than the WHO recommendations were tested. Together, these findings suggest that the dose of RIF could be increased safely to optimize children’s exposure. Before the optimal dosing of RIF can be defined, however, more studies should be conducted, with more robust methods such as individual participant data meta-analysis (IPD-MA) and population PK modelling. Current doses of INH (7.5–15 mg/kg) appear to be sufficient for most children, according to the comparisons with target exposure. NAT2 metabolizer status was, however, found to be the main cause of differences in INH PK and between patients and a significant predictor of exposure to INH as measured by both AUC and Cmax, with greater exposure of children categorized as slow metabolizers. Although slow metabolizers with high PK may be at risk of adverse events, this has not been demonstrated in children, and there was no evidence of significant INH-induced adverse events at the currently recommended doses. Fast metabolizers had lower exposure and did not meet PK targets. We found only four studies in which PK was reported by NAT2 status, suggesting that genotyping is not routine (22). NAT2 genotype testing is recommended (52), as dosing based on genotype has resulted in better clinical outcomes and safety in adults (53). Our findings indicate that NAT2 genotype testing could also be beneficial in children. We found wide heterogeneity among the few studies in which exposure to PZA and EMB was reported, and the data were even more limited for subgroups. Except for the PZA Cmax target, exposure to PZA and EMB was consistently low in all the studies included in the analysis. Because of both heterogeneity and 73 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2the limited reporting of exposure to PZA and EMB in children, the results of this analysis leave significant uncertainty about optimal dosing with these drugs. The highly variable exposures and wide heterogeneity among studies might be expected, given the greater variation in PK in children than in adults and the heterogeneous study populations (e.g., by age) and study characteristics. The wide variation and heterogeneity probably contributed to the lack of significant findings in this meta- analysis. Nonetheless, two significant associations were found in univariate subgroup analyses: children < 3–6 years had a lower RIF AUC, and children categorized as slow metabolizers of INH had higher INH AUC and Cmax. These results should, however, be interpreted with caution because of the small number of studies that reported PK for these subgroups, inconsistent stratification and the wide heterogeneity among studies. Although not statistically significant, we also observed that HIV-positive children tended to have a lower RIF AUC than HIV-negative children. Another important variable to be considered in understanding PK in children is nutritional status. Malnutrition was reported as a significant predictor of RIF and INH PK in two studies (18,35); however, we could not assess the effect of malnutrition in our meta-analysis because of sparse reporting and different definitions of malnutrition, which could not be reconciled appropriately for a comparison. Modelling and simulation suggest that malnourished children have lower exposure to drugs (assuming no difference in PK parameters) than non-malnourished children, primarily because lower doses are administered to patients in lower weight bands (20). Current WHO recommendations are based on uniform weight-band dosing (i.e., 15 mg/kg RIF for each weight band), while model-based weight-band dosing allows allometric dosing (i.e., higher mg/kg for lower weight bands) (8,19). Low-weight children, whether malnourished and/or young, may benefit from higher doses, given their higher risks for severe TB disease and mortality (54); however, we could not draw definitive conclusions about malnutrition or the merits and risks of weight-band approaches to dosing. The limitations of this review are due primarily to inconsistent reporting of PK, heterogeneous populations and small sample sizes. Important influences on the PK of RIF, a drug class of critical importance, could not be determined reliably with the usual meta-analytical approaches because of the limited number of studies, the small samples, inconsistent or unreported metrics and the impossibility of correcting for correlated predictors. For example, only 4 of 14 studies reported RIF PK by age, and none reported for the same age group (e.g., < 5 and ≥ 5 years). Similarly, few studies reported PK by malnutrition, and only two studies reported it with the same metric (e.g., “underweight”). Another limitation of this review is that we had to make assumptions in the analyses, as no alternative approaches were possible in this aggregated approach. For example, the reported AUC values varied from 0-4 h to 0-24 h but were considered equivalent in this analysis. Given the short half-life of most drugs, 3-4 h, this would be expected to impact results minimally, especially given the inherent variation in the studies; nevertheless, the assumption adds uncertainty. Inclusion of studies from India with thrice weekly dosing, which is no longer recommended, could be a potential limitation of this study, as the results were compared with those of studies of daily doses. Some of the limitations of usual meta-analyses could be overcome by conducting an IPD-MA. This approach has the advantage of more powerful, uniformly consistent analyses, better characterization, more power and better characterization of influential patient characteristics (55). IPD-MA also improves the quality and reliability of results and has greater potential to inform future trials. Combining IPD-MA with population PK modelling would allow detection of significant covariates and high-risk groups and permit development of tools to determine optimal doses. Therefore, IPD-MA with population PK could provide the means to propose efficient, more robust dosing algorithms, as has already been done for TB 74 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 and malaria with data on adults (56–58) but could not be done in the present analysis with traditional meta-analysis. Thus, more robust methods appropriate for integrating these types of data should be used, such as IPD-MA with population PK modelling, for evaluation of the safety of higher doses. This approach is more appropriate for identifying the optimal anti-TB drug doses that will safely improve the outcomes of all children, including those at high risk. Population-based approaches and nonlinear mixed-effects modelling with pooled individual-level data have been valuable for determining optimal doses for this purpose (54– 59). With a large PK database covering diverse child populations, including substantial proportions of malnourished children and children of various ages living with HIV, we will better understand the true drivers of the PK of exposure to RIF, INH, PZA and EMB. In view of the large number of PK studies that have already been conducted, this approach would be feasible and preferable to a new study. 75 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 25. Conclusion This review shows that there is a paucity of research on use of doses higher than those recommended for treatment of TB in children. Furthermore, reporting of PK parameters is inconsistent among studies, the populations are heterogeneous, and the samples were small. Nonetheless, we found that, at the WHO-recommended doses, exposure of children to RIF, PZA and EMB is consistently lower than that of adults. Exposure to INH appears to be within the target, but fast NAT2 metabolizers have significantly lower PK than slow metabolizers. Therefore, genotype-based dosing may be useful in children, as has been demonstrated in adults. The review also indicates that younger children (≤ 5 years) and perhaps those with HIV infection may require higher doses of RIF than those currently recommended doses for older children and HIV-negative children. Given the limitations of the review and the wide heterogeneity among studies, more robust methods, such as IPD-MA with population PK modelling, should be conducted before recommending appropriate doses of anti-TB drugs for optimal treatment of all children, including those at high risk, such as children ≤ 5 years and those infected with HIV. 76 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 References 1 Global tuberculosis report 2020. Geneva: World Health Organization; 2020 (https://apps.who.int/iris/bitstream/ handle/10665/336069/9789240013131-eng.pdf, accessed 1 February 2021). 2. Drobac PC, Shin SS, Huamani P, Atwood S, Furin J, Franke MF et al. Risk factors for in-hospital mortality among children with tuberculosis: The 25-year experience in Peru. Pediatrics. 2012;130(2):e373–9. 3. Thwaites G, Nahid P. Triumph and tragedy of 21st century tuberculosis drug development. NEJM. 2020;382(10):959–60. 4. Fox W, Ellagard G, Mitchison D. Studies on the treatment of tuberculosis undertaken by the British Medical Research Council tuberculosis units, 1946–1986, with relevant subsequent publications. Int J Tuberc Lung Dis. 1999;3(10 Suppl 2)):S231–79. 5. Short-course chemotherapy in pulmonary tuberculosis: A controlled trial by the British Thoracic and Tuberculosis Association. Lancet. 1976;308(7995):1102–4. 6. Technical report on the pharmacokinetics and pharmacodynamics (PK/PD) of medicines used in the treatment of drug-resistant tuberculosis. Geneva: World Health Organization; 2018 (https://apps.who.int/iris/ handle/10665/260440). 7. Aarnoutse RE, Kibiki GS, Reither K, Semvua HH, Haraka F, Mtabho CM et al. Pharmacokinetics, tolerability, and bacteriological response of rifampin administered at 600, 900, and 1,200 milligrams daily in patients with pulmonary tuberculosis. Antimicrob Agents Chemother. 2017;61(11):13. 8. Svensson RJ, Svensson EM, Aarnoutse RE, Diacon AH, Dawson R, Gillespie SH et al. Greater early bactericidal activity at higher rifampicin doses revealed by modeling and clinical trial simulations. J Infect Dis. 2018;218(6):991–9. 9. Boeree MJ, Diacon AH, Dawson R, Narunsky K. A dose-ranging trial to optimize the dose of rifampin in the treatment of tuberculosis. Am J Respir Crit Care Med. 2015;191(9):1058–65. 10. Understanding and using tuberculosis data. Geneva: World Health Organization; 2014 (https://apps.who.int/iris/ handle/10665/272879, accessed 21 September 2021). 11. Implementing the end TB strategy: the essentials. Geneva: World Health Organization; 2015 (https://apps.who. int/iris/handle/10665/206499, accessed 21 September 2021). 12. Stott KE, Pertinez H, Sturkenboom MGG, Boeree MJ, Aarnoutse R, Ramachandran G et al. Pharmacokinetics of rifampicin in adult TB patients and healthy volunteers: A systematic review and meta-analysis. J Antimicrob Chemother. 2018;73(9):2305–13. 13. Zvada SP, Denti P, Donald PR, Schaaf HS, Thee S, Seddon JA et al. Population pharmacokinetics of rifampicin, pyrazinamide and isoniazid in children with tuberculosis: In silico evaluation of currently recommended doses. J Antimicrob Chemother. 2014;69(5):1339–49. 14. Wilkins JJ, Langdon G, McIlleron H, Pillai G, Smith PJ, Simonsson USH. Variability in the population pharmacokinetics of pyrazinamide in South African tuberculosis patients. Eur J Clin Pharmacol. 2006;62(9):727–35. 15. Jönsson S, Davidse A, Wilkins J, Van der Walt JS, Simonsson USH, Karlsson MO et al. Population pharmacokinetics of ethambutol in South African tuberculosis patients. Antimicrob Agents Chemother. 2011;55(9):4230–7. 16. Antwi S, Yang H, Enimil A, Sarfo AM, Gillani FS, Ansong D et al. Pharmacokinetics of the first-line antituberculosis drugs in Ghanaian children with tuberculosis with or without HIV coinfection. Antimicrob Agents Chemother. 2017;61(2):e01701-16. 17. Bekker A, Schaaf HS, Draper HR, van der Laan L, Murray S, Wiesner L et al. Pharmacokinetics of rifampin, isoniazid, pyrazinamide, and ethambutol in infants dosed according to revised WHO-recommended treatment guidelines. Antimicrob Agents Chemother. 2016;60(4):2171–9. 18. Justine M, Yeconia A, Nicodemu I, Augustino D, Gratz J, Mduma E et al. Pharmacokinetics of first-line drugs among children with tuberculosis in rural Tanzania. J Pediatr Infect Dis Soc. 2020;9(1):14–20. 19. Guiastrennec B, Ramachandran G, Karlsson MO, Kumar AKH, Bhavani PK, Gangadevi NP et al. Suboptimal antituberculosis drug concentrations and outcomes in small and HIV-coinfected children in India: Recommendations for dose modifications. Clin Pharmacol Ther. 2018;104(4):733–41. 77 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 220. Radtke KK, Dooley KE, Dodd PJ, Garcia-Prats AJ, Hesseling AC, Savic RM. Evaluation of dosing guidelines for childhood tuberculosis: A mathematical modeling study. SSRN Electron J. 2019 (https://www.ssrn.com/ abstract=3311868, accessed 21 January 2021). 21. Svensson EM, Yngman G, Denti P, McIlleron H, Kjellsson MC, Karlsson MO. Evidence-based design of fixed- dose combinations: Principles and application to pediatric anti-tuberculosis therapy. Clin Pharmacokinet. 2018;57(5):591–9. 22. McIlleron H, Chirehwa MT. Current research toward optimizing dosing of first-line antituberculosis treatment. Expert Rev Anti Infect Ther. 2019;17(1):27–38. 23. Kearns GL, Abdel-Rahman SM, Alander SW, Blowey DL, Leeder JS, Kauffman RE. Developmental pharmacology – Drug disposition, action, and therapy in infants and children. NEJM. 2003;349(12):1157–67. 24. Wilkins JJ, Langdon G, McIlleron H, Pillai G, Smith PJ, Simonsson USH. Variability in the population pharmacokinetics of isoniazid in South African tuberculosis patients: Population pharmacokinetics of isoniazid. Br J Clin Pharmacol. 2011;72(1):51–62. 25. Moher D, Liberati A, Tetzlaff J, Altman D, PRISMA Group. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA statement. PLoS Med. 2009;6(7):e1000097. 26. Ridge A, Grzemska M, Hill S, Gie R. Rapid advice: treatment of tuberculosis in children. Geneva: World Health Organization; 2010 (http://whqlibdoc.who.int/publications/2010/9789241500449_eng.pdf, accessed 14 May 2021). 27. Seden K, Gibbons S, Marzolini C, Schapiro JM, Burger DM, Back DJ et al. Development of an evidence evaluation and synthesis system for drug–drug interactions, and its application to a systematic review of HIV and malaria co-infection. PLoS One. 2017;12(3):e0173509. 28. Jacobs TG, Svensson EM, Musiime V, Rojo P, Dooley KE, McIlleron H et al. Pharmacokinetics of antiretroviral and tuberculosis drugs in children with HIV/TB co-infection: A systematic review. J Antimicrob Chemother. 2020;75(12):3433–57. 29. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14(1):135. 30. Mukherjee A, Velpandian T, Singla M, Kanhiya K, Kabra SK, Lodha R. Pharmacokinetics of isoniazid, rifampicin, pyrazinamide and ethambutol in Indian children. BMC Infect Dis. 2015;15(1):126. 31. Thee S, Seddon JA, Donald PR, Seifart HI, Werely CJ, Hesseling AC et al. Pharmacokinetics of isoniazid, rifampin, and pyrazinamide in children younger than two years of age with tuberculosis: Evidence for implementation of revised World Health Organization recommendations. Antimicrob Agents Chemother. 2011;55(12):5560–7. 32. Mlotha R, Waterhouse D, Dzinjalamala F, Ardrey A, Molyneux E, Davies GR et al. Pharmacokinetics of anti-TB drugs in Malawian children: Reconsidering the role of ethambutol. J Antimicrob Chemother. 2015;70(6):1798–803. 33. Revised National Tuberculosis Control Programme. Technical and operational guidelines for tuberculosis control, October 2005. New Delhi: Ministry of Health and Family Welfare; 2005 (http://tbcindia.nic.in/pdfs/Technical%20 &%20Operational%20guidelines%20for%20TB%20Control.pdf). 34. Ramachandran G, Kumar AKH, Bhavani PK, Gangadevi NP, Sekar L, Vijayasekaran D et al. Age, nutritional status and INH acetylator status affect pharmacokinetics of anti-tuberculosis drugs in children. Int J Tuberc Lung Dis. 2013;17(6):800–6. 35. Ramachandran G, Kumar AKH, Bhavani PK, Kannan T, Kumar SR, Gangadevi NP et al. Pharmacokinetics of first- line antituberculosis drugs in HIV-infected children with tuberculosis treated with intermittent regimens in India. Antimicrob Agents Chemother. 2015;59(2):1162–7. 36. Schaaf HS, Willemse M, Cilliers K, Labadarios D, Maritz JS, Hussey GD et al. Rifampin pharmacokinetics in children, with and without human immunodeficiency virus infection, hospitalized for the management of severe forms of tuberculosis. BMC Med. 2009;7(1):19. 37. McIlleron H, Willemse M, Werely CJ, Hussey GD, Schaaf HS, Smith PJ et al. Isoniazid plasma concentrations in a cohort of South African children with tuberculosis: Implications for international pediatric dosing guidelines. Clin Infect Dis. 2009;48(11):1547–53. 78 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 38. McIlleron H, Willemse M, Schaaf HS, John P, Roderick P. Pyrazinamide plasma concentrations in young children with tuberculosis. Pediatr Infect Dis J. 2011;30(3):262–5. 39. Arya A, Roy V, Lomash A, Kapoor S, Khanna A, Rangari G. Rifampicin pharmacokinetics in children under the Revised National Tuberculosis Control Programme, India, 2009. Int J Tuberc Lung Dis. 2015;19(4):440–5. 40. Mukherjee A, Velpandian T, Singla M, Kanhiya K, Kabra SK, Lodha R. Pharmacokinetics of isoniazid, rifampicin, pyrazinamide and ethambutol in HIV-infected Indian children. Int J Tuberc Lung Dis. 2016;20(5):666–72. 41. Hiruy H, Rogers Z, Mbowane C, Adamson J, Ngotho L, Karim F et al. Subtherapeutic concentrations of first-line anti-TB drugs in South African children treated according to current guidelines: The PHATISA study. J Antimicrob Chemother. 2014;70(4):1115–23. 42. Dayal R, Singh Y, Agarwal D, Kumar M, Swaminathan S, Ramachandran G et al. Pharmacokinetic study of isoniazid and pyrazinamide in children: Impact of age and nutritional status. Arch Dis Child. 2018;103(12):1150–4. 43. Roy V, Sahni P, Gupta P, Sethi G, Khanna A. Blood levels of pyrazinamide in children at doses administered under the revised national tuberculosis control program. Indian Pediatr. 2012;49:5. 44. Horita Y, Alsultan A, Kwara A, Antwi S, Enimil A, Ortsin A, et al. Evaluation of the Adequacy of WHO Revised Dosages of the First-Line Antituberculosis Drugs in Children with Tuberculosis Using Population Pharmacokinetic Modeling and Simulations. Antimicrob Agents Chemother. 2018 Jun 18;62(9):e00008-18, /aac/62/9/e00008-18. atom. 45. Ranjalkar J, Mathew SK, Verghese VP, Bose A, Rose W, Gupta D et al. Isoniazid and rifampicin concentrations in children with tuberculosis with either a daily or intermittent regimen: Implications for the revised RNTCP 2012 doses in India. Int J Antimicrob Agents. 2018;51(5):663–9. 46. Garcia-Prats AJ. Optimising rifampicin exposure in children towards treatment shortening and treatment of severe disease: OptiRif Kids trial in Hyderabad, India, 2019. Union World Conference on Lung Health.Available on: https://theunion.org//sites/default/files/2020-09/20191101_UNION2019_Abstracts_Final_0.pdf 47. Shah I, Jadhao N, Mali N, Deshpande S, Gogtay N, Thatte U. Pharmacokinetics of isoniazid in Indian children with tuberculosis on daily treatment. Int J Tuberc Lung Dis. 2019;23(1):52–7. 48. Garcia-Prats AJ, Svensson EM, Winckler J, Draper HR, Fairlie L, van der Laan LE et al. Pharmacokinetics and safety of high-dose rifampicin in children with TB: The Opti-Rif trial. J Antimicrob Chemother. 2021; 76(12):3237-46 (doi: 10.1093/jac/dkab336). 49. Treatment of tuberculosis: Guidelines. Fourth edition. Geneva: World Health Organization; 2010 (http://apps. who.int/iris/bitstream/handle/10665/44165/9789241547833_eng.pdf;sequence=1). 50. Donald PR, Maritz JS, Diacon AH. The pharmacokinetics and pharmacodynamics of rifampicin in adults and children in relation to the dosage recommended for children. Tuberculosis. 2011;91(3):196–207. 51. Boeree MJ, Heinrich N, Aarnoutse R, Diacon AH, Dawson R, Rehal S. High-dose rifampicin, moxifloxacin, and SQ109 for treating tuberculosis: A multi-arm, multi-stage randomised controlled trial. Lancet Infect Dis. 2017;17(1):39–49. 52. Hong BL, D’Cunha R, Li P, Al-Shaer MH, Alghamdi WA, An G et al. A systematic review and meta-analysis of isoniazid pharmacokinetics in healthy volunteers and patients with tuberculosis. Clin Ther. 2020;42(11):e220–41. 53. Pharmacogenetics-based tuberculosis therapy research group, Azuma J, Ohno M, Kubota R, Yokota S, Nagai T et al. NAT2 genotype guided regimen reduces isoniazid-induced liver injury and early treatment failure in the 6-month four-drug standard treatment of tuberculosis: A randomized controlled trial for pharmacogenetics- based therapy. Eur J Clin Pharmacol. 2013;69(5):1091–101. 54. Marais BJ, Gie RP, Schaaf HS, Hesseling AC, Obihara CC, Starke JJ et al. The natural history of childhood intra- thoracic tuberculosis: a critical review of literature from the pre-chemotherapy era. Int J Tuberc Lung Dis. 2004;8(4):392–402. 55. Tierney JF. How individual participant data meta-analyses have influenced trial design, conduct, and analysis. J Clin Epidemiol. 2015;68(11):1325–35. 56. Hoglund RM, Workman L, Edstein MD, Thanh NX, Quang NN, Zongo I et al. Population pharmacokinetic properties of piperaquine in falciparum malaria: An individual par ticipant data meta-analysis. PLoS Med. 2017;14(1):e1002212. 79 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 257. Imperial MZ. A patient-level pooled analysis of treatment-shortening regimens for drug-susceptible pulmonary tuberculosis. Nat Med. 2018;24:12. 58. Hibma JE, Radtke KK, Dorman SE, Jindani A, Dooley KE, Weiner M et al. Rifapentine population pharmacokinetics and dosing recommendations for latent tuberculosis infection. Am J Respir Crit Care Med. 2020;202(6):866–77. 59. WorldWide Antimalarial Resistance Network (WWARN) DP Study Group. The effect of dosing regimens on the antimalarial efficacy of dihydroartemisinin–piperaquine: A pooled analysis of individual patient data. PLoS Med. 2013;10(12):e1001564. 60. Rangari GM, Roy V, Sethi GR, Mishra TK, Khanna A. Blood levels of isoniazid in Indian children with tuberculosis. Indian J Tuberc. 2015;62(2):80–5. 61. Panjasawatwong N, Wattanakul T, Hoglund RM, Bang ND, Pouplin T, Nosoongnoen W et al. Population pharmacokinetic properties of antituberculosis drugs in Vietnamese children with tuberculous meningitis. Antimicrob Agents Chemother. 2020;65(1):e00487-20. 80 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Appendix 1. Search strategy Search set Pubmed Cochrane Embase 1 child* OR neonat* OR pediatr* child* OR neonat* OR pediatr* child* OR neonat* OR pediatr* 2 Tuberculosis Tuberculosis Tuberculosis 3 #1 AND #2 #1 AND #2 #1 AND #2 4 isoniazid OR pyrazinamide OR ethambutol OR rifampicin isoniazid OR pyrazinamide OR ethambutol OR rifampicin isoniazid OR pyrazinamide OR ethambutol OR rifampicin 5 #3 AND #4 #3 AND #4 #3 AND #4 6 pharmacokinetic* OR outcome pharmacokinetic* OR outcome pharmacokinetic* OR outcome 7 #5 AND #6 #5 AND #6 #5 AND #6 8 prevention* OR latent prevention* OR latent prevention* OR latent 9 #7 NOT #8 #7 NOT #8 #7 NOT #8 Pubmed and Embase: research in title and abstract Cochrane: research in title, abstract and keywords 81 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Appendix 2. Scoring the quality of evidence –  pharmacokinetics target analysis Final scoring: high (1 point), moderate (2 points), low (3 points), very low (> 4 points) Study design: Randomized trial: 1 Controlled pharmacokinetics study, prospective: 2 Pharmacokinetics study, retrospective: 3 Observational study: 3 Other : 4 Add one point for the following situations of poor evidence: • The publication was a conference abstract or not published in a peer-reviewed journal. • Sample size < 10 children in each arm or 30 in the whole study • Doses were lower than currently recommended or not specified for any drug used, or the formulation used was other than WHO-recommended fixed-dose combination or not specified. • Not all relevant PK parameters were reported, or PK outcomes were not well defined (AUC, Cmax). • Inclusion and/or exclusion criteria were not defined (selection bias). • The PK sampling strategy was not defined, or only a few samples were collected (less than three samples per child). • Demonstration of confirmed drug-sensitive TB or that drug-resistant TB is not probable is not described. Remove one point for significant findings: • The study investigated doses higher than those currently recommended with a comparator arm. • The study included compartmental analyses with at least an internal validation. • There was a > 50% decrease or a greater than two times increase in PK between the reference and the comparator group. 82 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Individual reports: Final scorea Study Final score Mukherjee (30) 5 Ranjalkar (45) 4 Arya (39) 6 Ramachandran (34) 6 Thee (31) 4 Shah (47) 4 Garcia-Prats (46) Not assessable Ramachandran (35) 5 Rangari (60) 5 Dayal (42)isoniazid (INH 4 Antwi (16) 3 Bekker (17) 4 Mukherjee (40) 2 Panjasawatwong (61) 5 Mlotha (32) 5 Hiruy (41) 4 Roy (43) 5 Justine (18) 5 a Reference numbers refer to those in the reference list for the body of the text. 83 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2Appendix 3. Data extraction form The data extraction form was an excel spreadsheet for each of the drug. For a better understanding of our methods, here are detailed what we collected for each study and for each drug when appropriate (INH, RIF, PZA and EMB). • Publication information: publication ID, title, author, year of publication, country, design • Design information: sample size, numbers of HIV-positive patients, studied drugs (INH, RIF, PZA, EMB), definition of TB, dosing regimen for each drug, formulation for each drug, PK sample design, PK outcome definitions • Population information: age, body weight, sex, nutritional status (all z scores, with the numbers of patients with z score < 2 and/or raw data) • For each study: PK outcome for each drug (INH, RIF, PZA and EMB) for the whole cohort for each covariate, when available (HIV and nutritional status, age and NAT2 genotype for INH). Each covariate was separated into different columns when appropriate (e.g,. HIV positive and HIV negative, details for all the different ages). We reported the PK outcomes as follow: Cmax: mean or median, range, IQR or SD; AUC: mean or median, range, IQR or SD • Other reported associations with PK outcomes in free comments • For each study: Clinical outcome (favourable or unfavourable), when reported. • PK target attainment: Number of patients attaining the PK target (AUC and/or Cmax) for each drug (INH, RIF, PZA and EMB) • Adverse events in additional comments 84 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 A pp en di x 4. E x p o su re  t o  d ru gs  ( A U C  a n d  C m ax ) a s re po rt ed in th e st ud ie sa a R efe re nc e nu m be rs re fer to th os e in th e re fer en ce lis t f or th e bo dy o f t he te xt . R ifa m pi ci n St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er Th ee (3 1) (1 0 m g/ kg ) AU C 17 .8 (1 2. 8– 22 .8 ) 18 .4 (1 2. 8– 24 .0 ) 16 .7 (0 .2 –3 3. 1) – – – – – – – C m ax 6.3 6 (4 .5 –8 .3 ) 6.9 (4 .4 –9 .3 ) 5.5 (0 .1 –1 0. 9) – – – – – – – Th ee (3 1) (1 5 m g/ kg ) AU C 36 .9 (2 7. 7– 46 .3 ) 36 .1 (2 2. 7– 49 .6 ) 38 .4 (1 5.2 –6 1. 6) – – – – – – – C m ax 11 .7 (8 .7 –1 4. 7) 12 .5 (7 .7 –2 7. 4) 10 .2 (6 .5 –1 4. 0) – – – – – – – An tw i (1 6) AU C 27 .3 (2 0. 6– 36 .3 ) 30 .5 (2 1. 9– 38 .2 ) 24 .9 (1 5. 9– 35 .3 ) – – – – – – Lo w er A U C fo r H IV + ( P < 0 .0 5) C m ax 6.4 (4 .9 –8 .8 ) 7.7 (5 .2 –9 .1 ) 5.8 (3 .7 –8 .3 ) – – – – – – Lo w er C m ax fo r H IV + ( P < 0 .0 5) Be kk er (1 7) AU C 12 .1 (1 .8 –3 3. 0) 11 .5 (S D = 8 .7 ) 16 .5 (S D = 15 .0 ) 10 .6 (S D = 8 .2 ) – 13 .9 (S D = 1 0. 9) – 12 .8 (S D = 9 .7 2) 11 .4 (S D = 9 .6 3) – C m ax 2.9 (0 .6 –8 .0 ) 2.8 (S D = 2 .0 ) 3.6 7 (S D = 3. 3) 2.6 (S D = 1. 9) – 3.2 (S D = 2 .4 ) – 2.9 (S D = 2 .1 6) 2.9 (S D = 2 .2 4) – 85 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er M uk he rje e (3 0) AU C – 22 .6 (1 6. 7– 30 .6 ) 18 .5 (1 3. 3– 25 .7 ) – – – – – – – C m ax – 9.2 (6 .7 –1 2. 5) 7.8 (5 .6 –1 0. 7) 5.2 (3 .5 –7 .7 ) 5.6 (3 .1 –1 0.1 ) 5.6 (3 .0 –1 0. 5) – 2.7 (0 .8 –8 .8 ) 6.2 (4 .6 –8 .5 ) – Ra nja lka r (4 5) AU C 19 .4 (1 1. 3– 30 .4 ) – – – – – – 13 .3 18 .4 – C m ax 5.9 (3 .8 –8 .8 ) – – – – – – 5.8 6.2 – M uk he rje e (4 0) AU C 30 .5 (1 7.1 –5 8. 9) – – 26 .3 (2 1. 0– 27 .4 ) – 32 .8 (1 1. 7– 66 .7 ) – – – – C m ax 12 .0 (6 .1 –2 4. 3) – – 11 .3 (8 .1 –2 3. 6) – 13 .2 (5 .3 –2 5. 7) – – – – Ra ma ch an dr an (3 4) AU C – – – 9.9 (3 .6 –2 2. 0) 10 .1 (6 .1 –1 7. 6) 10 .9 (6 .7 –1 7. 7) 11 .5 (6 .5 –1 9. 8) 9.5 (0 .8 –1 3. 2) 10 .9 (6 .5 –1 9. 8) – C m ax – – – 3.3 (1 .1 –4 .7 ) 2.4 (1 .1 –4 .4 ) 2.5 (1 .3 –4 .2 ) 3.4 (1 .9 –4 .4 ) 2.4 (1 .1 –4 .1 ) 2.7 (1 .4 –4 .6 ) – Ra ma ch an dr an (3 5) AU C – – – – – – – 15 (9 .6 –1 9. 2) 31 .8 (1 8.8 –3 8.1 ) Lo w er A U C fo r yo un ge r ag e (P < 0 .0 1) : In cr ea se d by 2 .0 ( 95 % C I, 1. 1– 2. 9) pe r y ea r M al nu tr iti on d ec re as ed A U C ( P < 0 .0 5) : in cr ea se o f 2 .8 ( 95 % C I, 0. 3– 5. 3) pe r un it of W A Z C m ax – – – – – – – 3.1 (2 .4 –4 .0 ) 5.9 (4 .1 –7 .1 ) Lo w er C m ax fo r yo un ge r ag e ( < 0 .00 1) : In cr ea se d by 0 .4 ( 95 % C I, 0. 6– 0. 6) pe r y ea r M aln ut rit io n de cr ea se d C m ax ( P < 0 .0 5) : in cr ea se o f 0 .6 ( 95 % C I, 0. 1– 1. 0) pe r un it of W A Z 86 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er Ar ya (3 9) AU C 34 .5 (3 2. 3– 36 .1 ) – – – – – – – – – C m ax 7.1 (6 .5 –7 .6 ) – – – – – – – – – Hi ru y (4 1) AU C 21 .2 (1 .8 – 6 7. 3) – – – – – – – – – C m ax 3.5 (0 .6 – 1 0. 2) – – – – – – – – – Ju st in e (1 8) AU C – – – – – – – – – M aln ut rit io n de cr ea se d C2 h (P = 0 .0 01 ): –2 .0 3 (– 3. 2 to – 0. 87 ) C m ax 2.1 7 (0 .5 9– 4. 61 ) – – – – – – – – Pa nja saw atw on g (6 1) AU C 21 .5 (1 4. 2– 36 .5 ) – – – – – – – – – C m ax 4.9 (2 .5 –8 .4 ) – – – – – – – – – M lo th a (3 2) AU C 7.5 0 (5 .6 – 1 3.1 ) – – – – – – – – – C m ax 2.9 (2 .1 – 3 .4 ) – – – – – – – – – Ga rc ia- Pr at s (4 6) (1 5– 20 m g/ kg ) AU C 39 .5 (1 1. 7– 76 .1 ) – – – – – – – – – C m ax 8.4 (3 .1 –1 5. 5) – – – – – – – – – A U C t ar ge t is > 3 8. 7 m g. h/ L C m ax t ar ge t is 8– 24 m g/ L 87 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Is on ia zi d St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e B y N A T 2 ge no ty pe Si gn ifi ca nt as so ci at io ns H IV – H IV + N or m al St un te d Un de rw eig ht W as te d Yo un ge r O ld er Sl ow In te r Fa st Th ee (3 1) AU C 20 .4 (1 5.8 –2 5) 19 .5 (1 4– 25 .1) 22 .8 (1 0.2 –3 5.4 ) – – – – – – 30 .1 (2 5. 9– 34 .3 ) 15 .8 (9 .8 –2 1. 8) 12 .9 (8 .7 –1 7.1 ) N AT 2: P < 0 .0 01 C m ax 8.1 (6 .7– 9.5 ) 8.5 (6 .7– 10 .3) 6.9 (4 .4– 9.5 ) – – – – – – 9.8 (7 .7 –1 .9 ) 7.5 (4 .5 –1 0. 5) 6.7 (4 .6 –8 .8 ) N AT 2: P < 0 .0 01 Hi ru y (4 1) AU C 28 .7 (6 .8– 15 3) – – – – – – – – – – – C m ax 10 .8 (4 .7– 22 .7) – – – – – – – – – – – Lo w er C m ax fo r H IV + (P < 0 .0 4) An tw i (1 6) AU C 19 .7 (1 3. 2– 26 .4 ) 21 .1 (1 6. 5– 25 .9 ) 18 (1 2. 2– 26 .9 ) – – – – – – – – – – C m ax 5.8 (4 .3– 7.5 ) 5.8 (4 .3– 7.5 ) 5.3 (4 –7 .6) – – – – – – – – – – Be kk er (1 7) AU C 24 .7 (1 1. 6– 50 .2) 25 .08 (+ /– 8 .55 ) 21 .97 (+ /– 7 .45 ) 25 .9 (+ /– 1 0. 9) – 23 .2 (+ /– 4. 6) – 25 .9 (+ /– 9 .4 ) 23 .4 (+ /– 7 .3 ) – – – – C m ax 7.9 (4 –1 1. 3) 7.9 1 (+ /– 1 .84 ) 7.9 5 (+ /– 2 .3) 7.9 (+ /– 2. 2) – 8 (+ /– 1 .5 ) – 7.7 (+ /– 2 ) 8.1 (+ /– 1 .8 ) – – – – M uk he rje e (3 0) AU C – 1.3 (0 .8– 3.2 ) 2 (1 .5– 2.7 ) – – – – N R N R – – – – C m ax – 0.6 (0 .3– 1) 1 (0 .7– 1. 3) – 0.4 (0 .3 –0 .6 6) 0.8 (0 .4 –1 .4 ) 0.7 (0 .4 –1 .1) – 0.5 (0 .1 4– 1. 93 ) 0.5 (0 .4 –0 .7 ) – – – Ag e, P = 0 .0 4 Ra nja lka r (4 5) AU C 23 (1 9.4 –2 9.1 ) – – – – – – 20 .5 24 .2 – – – – C m ax 8.4 (7 .3– 9.4 ) – – – – – – 6 7.2 – – – – 88 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e B y N A T 2 ge no ty pe Si gn ifi ca nt as so ci at io ns H IV – H IV + N or m al St un te d Un de rw eig ht W as te d Yo un ge r O ld er Sl ow In te r Fa st M uk he rje e (3 0) AU C 7 (4 .4 –1 1. 9) – – 8.3 (4 .4 –1 3. 2) – 8.3 (4 .4 –1 3. 2) – – – – – – Ag e, P = 0 .0 4 C m ax 3.4 (1 .8 –5 ) – – 3.9 (2 .4 –5 .8 ) – 6.2 (4 .5 –9 .9 ) – 4.8 (3 .8 –5 .8 ) 3 (1 .6 –4 .6 ) 4.6 (3 .4 –7 .2 ) – 2.9 (1 .5 –4 .7 ) N AT 2, P < 0 .0 1 Ra ma ch an dr an (3 5) AU C – – – 22 .1 (1 1.6 –3 1. 5) 19 .5 (1 0– 30 .8 ) 19 .2 (1 0. 6– 30 .7 ) 12 .2 (1 0. 5– 26 .6 ) 11 .1 (+ /– 2 .3 ) 22 (+ /– 4 .9 ) 23 .9 (1 4. 2– 33 .6 ) – 11 .1 (7 .4 –1 5.1 ) Ag e: P < 0 .0 5 N AT 2: P < 0 .0 2 C m ax – – – 5.4 (3 .1 –7 .5 ) 4.6 (2 .5 –7 .3 ) 4.4 (2 .7 –7 .2 ) 3.3 (2 .7 –5 .3 ) 2.5 (+ /– 0 .4 ) 5.1 (+ /– 1 .1 ) 5.4 (3 .2 –7 .7 ) – 3.2 (2 –4 .3 ) Ag e: P < 0 .0 5 N AT 2: P < 0 .0 5 Ra ma ch an dr an (3 4) AU C – – – – – – – 14 .9 (+ /– 3. 1) 28 .7 (+ /– 5 .5 ) – – – Ag e: Inc re as e of 1 .3 (9 5% C I 0 .4 –2 .2 ) pe r ye ar ( P < 0 .0 1) N AT 2: inc re as e of 1 3.7 (9 5% C I, 8. 49 –1 8. 85 ) (P < 0 .0 01 ) C m ax – – – – – – – 3.3 (+ /– 0. 6) 7.2 (+ /– 0 .7 ) – – – In cr ea se o f 0 .4 ( 95 % C I 0 .2 –0 .6 ) pe r ye ar (P < 0 .0 01 ) N AT 2: Inc re as ed b y 2. 7 (9 5% C I 1 .4 5– 3. 99 ) (P < 0 .0 01 ) Sh ah (4 7) AU C 46 .2 (+ /– 34 .8 ) – – 49 (+ /– 39 .9 ) – 40 .2 (+ /– 2 0. 2) – 61 ( + /– 50 ) 36 .1 (+ /– 13 .6 ) 55 .25 (+ /– 3 6) – 20 .2 (+ /– 9) N AT 2: P < 0 .0 1 C m ax 8.3 (+ /– 4. 3) – – 8.4 (+ /– 4 .7 ) – 8.2 (+ /– 3. 3) – 9.9 (+ /– 5 .7 ) 7.2 (+ /– 2. 5) 9.1 (+ /– 4 .5 ) – 6.2 (+ /– 2 .8 ) N AT : P < 0 .0 5 D ay al (4 2) AU C 14 .7 (+ /– 9. 6) – – 13 .8 (8 .4 –1 9. 3) 16 .1 (8 .7 –2 3. 6) 14 .6 (9 .1 –2 0. 2) 12 .8 (6 .3 –1 9. 2) 15 .8 14 .4 – – – – C m ax 7.9 (4 –1 1. 3) – – 2.9 (2 .1 –3 .8 ) 3.2 (1 .9 –4 .4 ) 3.1 (2 .1 –4 .1 ) 2.6 (1 .7 –3 .5 ) 3.2 3.1 – – – – 89 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e B y N A T 2 ge no ty pe Si gn ifi ca nt as so ci at io ns H IV – H IV + N or m al St un te d Un de rw eig ht W as te d Yo un ge r O ld er Sl ow In te r Fa st Ju st in e (1 8) AU C – – – – – – – – – – – – C m ax 2 (0 .4 –4 .6 ) – – – – – – – – – – – M aln ut rit io n de cr ea se d C m ax ( P = 0 .0 01 ) Ra ng ar i (6 0) AU C 32 .2 (+ /– 0. 6) – – – – – – – – – – – – C m ax 6.5 (+ /– 0. 2) – – – – – – – – – – – – M lo th a (3 2) AU C 11 .5 (7 .3 –1 8. 9) – – – – – – – – – – – – C m ax 3.4 (2 .6 –4 .6 ) – – – – – – – – – – – – Pa nja saw atw on g (6 1) AU C – – – – – – – – – 12 .4 (6 .4 –2 1. 6) – 6.3 5 (2 .6 –2 4. 2) – C m ax – – – – – – – – – 2.4 (1 .9 –3 .8 ) – 2.2 (1 .5– 5.1 ) – A U C t ar ge t is > 2 3. 4 m g. h/ L C m ax t ar ge t is 3– 6 m g/ L 90 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Py ra zi na m id e St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er Th ee (3 1) AU C 17 5.2 (1 55 .5– 19 5. 0) 18 0.1 (1 54 –2 06 .3 ) 16 0.5 (1 39 .9– 18 1.1 ) – – – – – – – C m ax 47 .1 (4 2. 6– 51 .6 ) 49 .2 (4 3. 6– 54 .8 ) 40 .8 (3 6. 2– 45 .5 ) – – – – – – – Hi ru y (4 1) AU C 23 3.9 (1 10 .1– 52 5. 7) – – – – – – – – – C m ax 22 .5 (1 1. 2– 47 .2 ) – – – – – – – – – An tw i (1 6) AU C 14 0.5 (1 14 .2– 18 6. 2) 15 1.0 (1 24 .6– 18 8. 8) 12 6.5 (1 05 .4– 18 2. 3) – – – – – – H IV : P = 0 .0 34 C m ax 26 .1 (2 1.7 –3 5.1 ) 26 .9 (2 3. 2– 34 .6 ) 24 .6 (2 0. 6– 36 .5 ) – – – – – – – Be kk er (1 7) AU C 23 9.4 (1 47 .1– 45 0. 0) 24 7.1 (+ /– 66 .8 ) 18 7 (+ /– 27 .1 ) 21 4.7 (+ /– 31 .5 ) – 26 8.1 (+ /– 83 .5 ) – 25 3.3 (+ /– 82 .7 ) 22 4.7 (+ /– 39 .4 ) H IV : P < 0 .0 01 C m ax 41 .9 (2 6. 3– 68 .4 ) 42 .8 (+ /– 9. 2) 35 .6 (+ /– 5) 39 .4 (5 .3 ) – 44 .8 (1 1. 6) – 42 .7 (+ /– 1 1. 3) 40 .9 (+ /– 6) H IV : P = 0 .0 13 M uk he rje e (4 0) AU C – 15 9.0 (1 32 .2– 19 1. 4) 14 3.7 (1 20 .5– 17 1. 3) – – – – – – – C m ax – 54 .5 (4 5. 1– 65 .7 ) 55 .1 (4 7. 6– 63 .9 ) 44 .0 (3 6. 7– 52 .7 ) 55 .4 (4 6. 1– 66 .6 ) 53 .2 (4 1. 5– 68 .2 ) – 45 .8 (2 7. 9– 74 .9 ) – – 91 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er M uk he rje e (3 0) AU C 14 0.5 (1 17 .4– 17 3. 2) – – 14 3.8 (1 22 .7– 17 4.1 ) – 13 3.8 (1 15 .6– 16 4. 6) – – – – C m ax 47 .8 (3 9. 4– 58 .5 ) – – 48 .7 (3 9. 1– 58 .4 ) – 46 .9 (3 9. 8– 60 .1 ) – – – – Ra ma ch an dr an (3 5) AU C – – – 21 1.8 (1 57 .0– 26 6.4 ) 22 6.8 (1 77 .8– 27 9.1 ) 22 4.6 (1 75 .5– 28 4.9 ) 18 6.4 (1 69 .4– 24 7.9 ) 17 8.0 (1 16 .8– 23 2.9 ) 22 1.7 (1 17 .8– 28 3.4 ) Ag e, P < 0 .0 5 C m ax – – – 40 .6 (2 9. 4– 46 .8 ) 43 .2 (3 1.8 –4 9.1 ) 41 .3 (3 1. 4– 48 .6 ) 34 .5 (2 8. 5– 42 .6 ) 34 .1 (2 1. 4– 40 .9 ) 42 .3 (3 2. 3– 51 .4 ) Ag e, P < 0 .0 5 Ra ma ch an dr an (3 4) AU C – – – – – – – 17 5.9 (1 31 .5– 19 3. 9) 22 1.8 (1 94 .7 –2 56 .6 ) A ge : m ul tip le r eg re ss io n (b et a IC 95 % ) 7. 46 [ 1. 97 –1 2.9 4] P < 0 .01 C m ax – – – – – – – 30 .4 (2 6. 2– 33 .4 ) 38 .0 (3 0. 1– 45 .1 ) A ge : m ul tip le r eg re ss io n (b et a IC 95 % ) 1. 2 [0 .2 3– 2. 18 ] P < 0 .05 D ay al (4 2) AU C 18 9.9 (+ /– 57 .2 ) – – 18 8.3 (1 72 .2– 20 4.4 ) 17 0.3 (1 40 .9– 19 9.7 ) 17 1.4 (1 44 .7– 19 8.1 ) 15 9.6 (1 32 .4– 18 6.6 ) 15 5.5 20 2.6 Ag e: P < 0 .0 5 C m ax 35 (+ /– 9. 8) – – 35 .7 (3 0. 6– 40 .8 ) 31 .9 (2 5. 9– 37 .9 ) 32 .5 (2 7. 4– 37 .7 ) 30 .5 (2 5. 5– 35 .6 ) 29 .2 37 .1 Ag e: P = 0 .0 31 Ju st in e (1 8) AU C – – – – – – – – – – C m ax 28 .2 (1 3. 3– 32 .5 ) – – – – – – – – – 92 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er Ro y (4 3) AU C 36 9.5 (+ /– 35 .3 ) – – – – – – – – – C m ax 49 .4 (+ /– 2. 8) – – – – – – – – – M lo th a (3 2) AU C 19 4.7 (1 63 .4 –3 82 .4 ) – – – – – – – – – C m ax 34 .6 (3 2. 3– 40 .9 ) – – – – – – – – – Pa nja saw atw on g (6 1) AU C 28 8.0 (1 08 .0 –5 69 .0 ) – – – – – – – – – C m ax 42 .5 (2 9. 8– 92 .7 ) – – – – – – – – – A U C t ar ge t is 23 8– 42 8 m g. h/ L C m ax t ar ge t is 35 –6 0 m g/ L 93 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 Et ha m bu to l St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er Hi ru y (4 1) AU C 10 .8 (4 .7 –2 2. 7) – – – – – – – – – C m ax 1. 4 (0 .6 –6 .3 ) – – – – – – – – – An tw i (1 6) AU C 6.2 (3 .7 –9 .2 ) 7.6 (5 .2 –1 0. 7) 4.8 (3 .0 –6 .8 ) – – – – – – H IV + a ss oc ia te d w ith lo w er A U C (P < 0 .0 01 ). C m ax 1. 7 (0 .9 –2 .7 ) 2.3 (1 .5 –3 .0 ) 1. 3 (0 .8 –1 .9 ) – – – – – – H IV + a ss oc ia te d w ith lo w er C m ax (P < 0 .00 1) . Be kk er (1 7) AU C 5.1 (1 .2 –8 .9 ) 5.5 (+ /– 1. 5) 2 (+ /– 1. 1) 4.8 (+ /– 1. 8) – 5.4 (+ /– 2. 1) – 5.3 (+ /– 2. 1) 4.8 (+ /– 1. 7) H IV a ss oc ia te d w ith lo w er A U C (P < 0 .0 08 ). C m ax 1. 3 (0 .2 –2 .0 ) 1. 4 (+ /– 0. 4) 0.4 (+ /– 0. 2) 1. 2 (+ /– 0. 5) – 1. 4 (+ /– 0. 5) – 1. 3 ( + /– 0. 5) 1. 1 (+ /– 0. 4) H IV a ss oc ia te d w ith lo w er C m ax (P < 0 .0 04 ). M uk he rje e (4 0) AU C – 2.8 (2 .0 –3 .9 ) 1. 6 (1 .1 –2 .3 ) – – – – – – H IV a ss oc ia te d w ith lo w er A U C (P < 0 .0 5) . C m ax – 1. 1 (0 .7 –1 .7 ) 0.8 (0 .5 –1 .2 ) 0.7 (0 .5 –0 .9 ) 0.6 (0 .3 –1 .3 ) 0.5 (0 .2 –0 .9 ) – – – – M uk he rje e (3 0) AU C 4.8 (2 .2 –8 .9 ) – – 5.6 (2 .2 –1 0. 3) – 4.3 (2 .3 –6 .4 ) – 0.5 (0 .2 –1 .3 ) N R – C m ax 2.1 (1 .0 –3 .7 ) – – 2.0 (1 .0 –2 .0 ) – 2.0 (1 .0 –3 .4 ) – – – – Ju st in e (1 8) AU C N R – – – – – – – – – C m ax 1. 0 (0 .6 –1 .7 ) – – – – – – – – – 94 Pharmacokinetics of fir st-line drugs (r ifampicin, isoniazid, ethambutol and pyrazinamide) in children (< 18 years) treated for drug-susceptible tuberculosis: systematic review and meta-analysis A N N E X 2 St ud y M et ri c W ho le co ho rt B y H IV s ta tu s B y nu tr it io na l s ta tu s B y ag e Si gn ifi ca nt a ss o ci at io ns H IV – H IV + N or m al St un te d U nd er w ei gh t W as te d Yo un ge r O ld er M lo th a (3 2) AU C 8.0 (4 .9 –1 0.1 ) – – – – – – – – – C m ax 1. 2 (0 .9 –1 .7 ) – – – – – – – – – Pa nja saw atw on g (6 1) AU C 8.2 (4 .6 –1 7. 8) – – – – – – – – – C m ax 1. 3 (0 .7 –2 .5 ) – – – – – – – – – AU C ta rg et is 1 6– 28 m g. h/ L C m ax ta rg et is 2 –6 m g/ L

For further information, please contact: World Health Organization 20, Avenue Appia CH-1211 Geneva 27, Switzerland Global TB Programme Web site: www.who.int/tb