Effectiveness and safety of modified fully oral 9-month treatment regimens for rifampicin-resistant tuberculosis: a prospective cohort study

Abstract

Background: In 2020, WHO guidelines prioritised the use of a standard fully oral short treatment regimen (STR) consisting of bedaquiline, levofloxacin or moxifloxacin, ethionamide, ethambutol, high-dose isoniazid, pyrazinamide, and clofazimine for the management of rifampicin-resistant tuberculosis. A high prevalence of resistance to constituent drugs precluded its widespread use by countries in the WHO European region. We evaluated three 9-month fully oral modified STRs (mSTRs) in which ethionamide, ethambutol, isoniazid, and pyrazinamide were replaced by linezolid, cycloserine, or delamanid (or a combination).

Methods: This multicountry, prospective, single-arm, cohort study examined the effectiveness and safety of mSTRs for fluoroquinolone-susceptible, rifampicin-resistant pulmonary tuberculosis in 13 countries in the WHO European region during 2020-23. We enrolled adults and children of all ages with bacteriologically confirmed rifampicin-resistant, fluoroquinolone-susceptible pulmonary tuberculosis, and children (aged 0-18 years) with clinically diagnosed disease and a confirmed contact with rifampicin-resistant, fluoroquinolone-susceptible tuberculosis. Participants aged 6 years or older received one of two regimens: bedaquiline, linezolid, levofloxacin, clofazimine, and cycloserine; or bedaquiline, linezolid, levofloxacin, clofazimine, and delamanid. Children younger than 6 years received delamanid, linezolid, levofloxacin, and clofazimine. Participants were followed up for 12 months after successful treatment completion to detect recurrence and death. The primary outcome was the cumulative probability of not having an unsuccessful study outcome (defined as treatment failure, on-treatment loss to follow-up, death, or recurrence) before 22 months of study follow-up. The primary safety outcome was the incidence of each adverse event of interest (peripheral neuropathy, optic neuritis, myelosuppression, hepatitis, prolonged QT interval, hypokalaemia, and acute kidney injury) of grade 3 or higher severity during the treatment course.

Findings: Between Aug 28, 2020 and May 26, 2021, 7272 patients were screened and 2636 were included in the treatment cohort. 1966 (74·6%) were male, 670 (25·4%) were female, and median age was 43 years (IQR 33-53). Treatment success was recorded for 2181 (82·7%) participants. The cumulative probability of not having an unsuccessful study outcome 22 months after treatment initiation was 79% (95% CI 78-81). Increasing age (adjusted hazard ratio 2·61 [95% CI 1·70-4·04] for people aged >64 years vs 35-44 years), HIV-positive status (1·53 [1·16-2·01]), presence of bilateral cavities (1·68 [1·29-2·19]), smoking history (1·34 [1·05-1·71]), baseline anaemia (1·46 [1·15-1·86]), unemployment (1·37 [1·04-1·80]), elevated baseline liver enzymes (1·40 [1·13-1·73]), and excessive alcohol use (1·47 [1·14-1·89]) were positively associated with unsuccessful study outcomes. In the safety cohort of 2813 participants who received at least one dose, 301 adverse events of interest were recorded in 252 (9·0%) participants with the most frequent being myelosuppression (139 [4·9%] participants, 157 [52·2%] events).

Interpretation: The high treatment success and good safety results indicate considerable potential for the use of mSTRs in programmatic conditions, especially for individuals not eligible for the current WHO-recommended 6-month regimen and in settings with a need for alternative options.

Funding: The Global Fund to Fight AIDS, Tuberculosis and Malaria; United States Agency for International Development; Government of Germany; and WHO.

Translation: For the Russian translation of the abstract see Supplementary Materials section.

Figure 1

Study profile ALT=alanine aminotransferase. AST=aspartate aminotransferase. ECOG=Eastern Cooperative Oncology Group. mSTR=modified short treatment regimen. QTc_F interval=QT interval corrected with Fridericia's formula.

Figure 2

Incidence rate of adverse events of interest per month after treatment initiation Error bars are 95% CIs. Adverse events of interest were peripheral neuropathy, optic neuritis, myelosuppression, hepatitis, prolonged QT interval (corrected with Fridericia's formula), hypokalaemia, and acute kidney injury, all of grade 3 or higher.