Evidence-based Definition for Extensively Drug-Resistant Tuberculosis
Maroussia Roelens,Giovanni Battista Migliori,Liudmila Rozanova,Janne Estill,Jonathon R Campbell,J Peter Cegielski,Simon Tiberi,Domingo Palmero,Greg J Fox,Lorenzo Guglielmetti,Giovanni Sotgiu,James C M Brust,Didi Bang,Christian Lienhardt,Christoph Lange,Dick Menzies,Olivia Keiser,Mario Raviglione
DOI: https://doi.org/10.1164/rccm.202009-3527OC
2021-09-15
Abstract:Rationale: Until 2020, extensively drug-resistant tuberculosis (XDR-TB) was defined as TB with resistance to rifampicin and isoniazid (multidrug-resistant TB [MDR-TB]), any fluoroquinolone (FQ), and any second-line injectable drug (SLID). In 2019, the World Health Organization issued new recommendations for treating patients with drug-resistant TB, substantially limiting the role of SLIDs in MDR-TB treatment and thus putting the definition of XDR-TB into question. Objectives: To propose an up-to-date definition for XDR-TB. Methods: We used a large data set to assess treatment outcomes for patients with MDR-TB exposed to any type of longer regimen. We included patients with bacteriologically confirmed MDR-TB and known FQ and SLID resistance results. We performed logistic regression to estimate the adjusted odds ratios (aORs) for an unfavorable treatment outcome (failure, relapse, death, loss to follow-up), and estimates were stratified by the resistance pattern (FQ and/or SLID) and group A drug use (moxifloxacin/levofloxacin, linezolid, and/or bedaquiline). Measurements and Main Results: We included 11,666 patients with MDR-TB; 4,653 (39.9%) had an unfavorable treatment outcome. Resistance to FQs increased the odds of an unfavorable treatment outcome (aOR, 1.91; 95% confidence interval [CI], 1.63-2.23). Administration of bedaquiline and/or linezolid improved treatment outcomes regardless of resistance to FQs and/or SLIDs. Among patients with XDR-TB, compared with persons receiving no group A drug, aORs for an unfavorable outcome were 0.37 (95% CI, 0.20-0.69) with linezolid only, 0.40 (95% CI, 0.21-0.77) with bedaquiline only, and 0.21 (95% CI, 0.12-0.38) with both. Conclusions: Our study supports a new definition of XDR-TB as MDR-TB and additional resistance to FQ plus bedaquiline and/or linezolid and helps assess the adequacy of this definition for surveillance and treatment choice.