Maria Garcia-Cremades(1,2), Anthony Garcia-Prats(3,4), Vivian Cox(3), Kendra Radtke(2), Tamara Kredo(5), Rory Dunbar(3), Rada Savic(2), Anneke C. Hesseling(3)
(1) Department of Pharmaceutics and Food Technology, School of Pharmacy, Complutense University of Madrid, Madrid, Spain, (2) Department of Bioengineering and Therapeutic Sciences, University of California San Francisco, San Francisco, USA. (3)Desmond Tutu TB Centre, Department of Paediatrics and Child Health, Stellenbosch University, Cape Town, South Africa (4) Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Wisconsin, USA, (5) Cochrane Centre, South African Medical Research Council
Introduction: Out of the approximately 1.1 million children worldwide who were diagnosed with tuberculosis in 2020, it is estimated that around 5-7% of them had either multidrug-resistant or rifampicin-resistant tuberculosis (MDR/RR-TB). Current MDR/RR-TB treatment in children remains complex and long due to limited pediatric data with newer drugs and shorter regimens. Robust data are a priority to inform evidence-based approaches to MDR/RR-TB treatmen in children and adolescentes.
Objectives: The aim of this work is to describe the use of MDR/RR-TB treatment in the pediatric population across years and countries, and identify clinical and drug features associated with TB outcomes.
Methods: We performed a systematic review and individual patient data (IPD) analysis of children and adolescents 0-19 years of age treated for MDR/RR-TB. Fifty-seven studies and cohorts of pediatric MDR/RR-TB, performed between 1994 and 2020 across different countries, were considered for analysis. Patients with missing TB treatment outcome, unknown treatment information (including unknown drugs used and treatment duration), and with isoniazid monoresistance were excluded. Clinical and demographic characteristics were described and, along with the use of drugs by WHO group, were evaluated as predictors of TB treatment success vs. TB unfavorable outcome (treatment failure or death) using mixed-effects logistic regression models. Missing clinical characteristics were imputed using multiple imputation via chain equations.
Results: The analysis dataset included 20,395 children from 57 cohorts, 63% from India, 30% from South Africa. Overall, 16,825 (82.5%) had treatment success, 722 (3.5%) treatment failure, and 2,848 (14%) died. The mean (SD) age was 14.7 (4.8) years, 89.6% were bacteriologically confirmed, and mean (SD) treatment duration was 15.3 months (8.3) months. Treatment duration in the patients with treatment success was substantially different across countries, with shorter durations in those studies performed after 2014 (p<0.001). Of 17,764 children with documented HIV status, 13.8% were HIV-positive. HIV positivity, prior TB treatment, older age, AFB positivity, male gender and pulmonary disease were associated with lower odds of treatment success (p<0.001). In addition, the inclusion of two (aOR 1.41, [2.5%-97.5% CI, 1.09 to 1.8]; p=0.008) or three (aOR 2.1; [2.5%-97.5% CI, 1.6 to 2.8], p<0.001) WHO-classified Group A drugs were associated with increased odds of treatment success.
Conclusions: In this largest ever paediatric MDR/RR-TB IPD, treatment success was high despite high bacteriological confirmation rates and high HIV prevalence. However, treatment outcomes were still below international targets of ≥90% successfully treated. Young children and those with clinically diagnosed MDR/R RR-TB were underrepresented. The use of multiple group A drugs resulted in better outcomes, which supports new WHO guidelines for all-oral regimens in all children. Improved paediatric MDR/RR-TB treatment surveillance is need to generate more outcomes data on new drugs and regimens, especially in young children.
Reference: PAGE 32 (2024) Abstr 11234 [www.page-meeting.org/?abstract=11234]
Poster: Drug/Disease Modelling - Infection