Treatment outcomes of extensively drug-resistant tuberculosis in Europe: a retrospective cohort study

Abstract

Background: In 2021, World Health Organization revised of definition of extensive drug-resistant tuberculosis. We aimed to determine treatment outcomes of individuals affected by extensively drug-resistant tuberculosis in Europe.

Methods: This observational, retrospective cohort study included patients diagnosed with extensively drug-resistant tuberculosis in the World Health Organization European Region from 2017 to 2023. Participating centres collected consecutive, detailed individual data for extensively drug-resistant tuberculosis patients. Data were analysed with meta- and regression methods, accounting for between-country heterogeneity.

Findings: Among 11,003 patients with multidrug-resistant/rifampicin-resistant tuberculosis, 188 (1·7%) from 16 countries had extensively drug-resistant tuberculosis. Of these, 48·4% harboured strains with resistance to bedaquiline (n = 91/188), 34·0% to linezolid (n = 64/188), and 17·6% to both (n = 33/188). The individual composition of anti-tuberculosis regimens was highly variable, with 151 different drug combinations. Among the 156/188 (83·0%) patients with available treatment outcomes, the pooled percentage of successful outcomes was 40·2% (95% confidence interval [95% CI] 28·4%-53·2%). In patients with unsuccessful treatment outcomes (101/156), most experienced treatment failure (n = 57/156 [pooled proportion 37·1%], 95% CI: 26·1%-49·7%) or death (n = 30/156 [pooled proportion 21·3%], 95% CI: 15·7%-28·2%). After adjustment for disease severity, each additional likely effective drug decreased the odds of unsuccessful outcomes (adjusted odds ratio: 0·65, 95% CI: 0·45-0·96) (p = 0·026), whereas being treated in an upper-middle-income country increased the odds of unsuccessful outcomes compared with being treated in a high-income country (adjusted odds ratio: 13·7, 95% CI: 3·7-50·2) (p < 0·001). Compared with other levels of drug resistance, treatment outcomes were significantly worse for extensively drug-resistant tuberculosis.

Interpretation: Only four out of ten patients affected by extensively drug-resistant tuberculosis achieved successful treatment outcomes. These findings highlight the need for adequate, individualised treatment regimens and optimised drug susceptibility testing.

Funding: None.

Keywords: ESGMYC; MDR/RR-TB; Pre-XDR-TB; TB; TBnet; XDR-TB.

Fig. 1

Flow diagram. Abbreviations: MDR-/RR-TB, multidrug/rifampicin-resistant tuberculosis; Pre-XDR-TB, pre-extensively drug-resistant tuberculosis; XDR-TB, extensively drug-resistant tuberculosis.

Fig. 2

Overview of individual treatment regimens, Mycobacterium tuberculosis drug susceptibility patterns, and treatment outcomes for 188 extensively drug-resistant tuberculosis patients from 16 countries in WHO Europe Region (2017–2023). Individual-level visualisation of treatment regimens, drug susceptibility testing (DST) results, and outcomes for patients with drug-resistant TB. Each row represents one patient. The left section indicates whether specific drugs were included in the treatment regimen (green), not included (orange), or unknown (grey). The middle section shows DST results for the same drugs: resistant (dark red), susceptible (purple), or not performed (light grey). The right section displays treatment outcomes: successful (blue), unsuccessful (red), or not evaluated (light grey). Abbreviations: FQ, fluoroquinolones; Lfx, levofloxacin; Mfx, moxifloxacin; Bdq, bedaquiline; Lzd, linezolid; Cfz, clofazimine; Cs/Tzd, cycloserine/terizidone; E, ethambutol; Dlm, delamanid; Pa, pretomanid; Z, pyrazinamide; Carb + Amx/Clav, carbapenem + amoxicillin/clavulanic acid; Amk, amikacin; Cm, capreomycin; Km, kanamycin; Eto/Pto, ethionamide/prothionamide; PAS, P-aminosaliscylic acid.

Fig. 3

Pooled percentages of treatment outcomes for patients with rifampicin-/multidrug-resistant, pre-extensively drug-resistant, and extensively drug-resistant tuberculosis in the WHO European Region (2017–2023), from a random-effects meta-analysis accounting for between-country variation.¹Based on aggregate data from 12 out of 18 participating centres. ²Based on patient-level data from the participating centres. Abbreviations: MDR-/RR-TB, multidrug/rifampicin-resistant tuberculosis; Pre-XDR-TB, pre-extensively drug-resistant tuberculosis; XDR-TB, extensively drug-resistant tuberculosis.

Fig. 4

Univariable and multivariable logistic regression analysis of risk factors for unsuccessful treatment outcome in 156 extensively drug-resistant tuberculosis patients from 16 countries in the WHO European Region (2017–2023). 95% confidence intervals were calculated using cluster-robust standard errors with clustering by country. Odds ratios were adjusted for presence of lung cavities, previous treatment with second-line drugs, and sputum smear positivity. Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; HIC, high-income countries; UMIC, upper-middle-income countries; Lzd, linezolid; Bdq, bedaquiline.

Fig. 5

Kaplan Meier curve of time to unsuccessful outcome for 154 extensively drug-resistant tuberculosis patients, stratified for World Bank income groups, in the WHO European Region (2017–2023). Abbreviations: HIC, high-income country; UMIC, upper middle-income country.