A 24-Week, All-Oral Regimen for Rifampin-Resistant Tuberculosis

Trial Design and Oversight

We conducted an open-label, phase 2–3, multicenter, randomized, controlled noninferiority trial to compare the safety and efficacy of three investigational 24-week regimens with those of the accepted 9-to-20-month standard-care treatment for rifampin-resistant pulmonary tuberculosis. The trial was designed to seamlessly transition from a phase 2b trial to a phase 3 trial with one or two investigational groups. Further details are provided in Section S4 in the Supplementary Appendix and the protocol, both of which are available with the full text of this article at NEJM.org. The trial was approved by institutional ethics boards as well as local ethics committees and national regulatory authorities in the countries where the trial was conducted.

The trial was designed by the protocol development team (Section S1.1 in the Supplementary Appendix). The data were collected by the site investigators, and the statistical analysis was performed by the tenth and last authors and interpreted by all the authors. The first draft of the manuscript was written by the first four authors and the last author. All the authors participated in revision of the manuscript, approved the submitted versions, and vouch for the accuracy and completeness of the data and for the fidelity of the trial to the protocol.

Patients

Patients 15 years of age or older who had rifampin-resistant pulmonary tuberculosis were enrolled at seven sites in Belarus, South Africa, and Uzbekistan. The investigators were notified of new cases of laboratory-diagnosed rifampin-resistant tuberculosis from within the catchment areas.

The major inclusion criterion was Mycobacterium tuberculosis infection (as confirmed by a positive sputum smear) with resistance to rifampin. Patients were included irrespective of fluoroquinolone resistance, human immunodeficiency virus (HIV) status, or CD4 count. Patients were excluded if they were pregnant or if they had an alanine aminotransferase level or an aspartate aminotransferase level higher than 3 times the upper limit of the normal range, a corrected QT interval calculated with the use of Fridericia’s formula (QTcF) greater than 450 msec, structural heart disease, or suspected resistance to bedaquiline, pretomanid, or linezolid. All the patients provided written informed consent.

Treatment

In stage 1 of the trial, enrolled patients were randomly assigned to the locally accepted standard-care treatment or to one of three investigational regimens. The standard-care regimen consisted of locally accepted treatment regimens. These regimens were closely aligned with the World Health Organization (WHO) guidelines for treatment of drug-resistant tuberculosis,3 and the agents (some oral and some intravenous) were administered at least 6 days per week with food and under observation (see Section S5).

All the investigational agents were administered orally, with food and under observation, 7 days per week. The BPaL regimen consisted of the following: bedaquiline at a dose of 400 mg daily for 2 weeks, followed by 200 mg three times per week for 22 weeks; pretomanid at a dose of 200 mg daily for 24 weeks; and linezolid at a dose of 600 mg daily for 16 weeks, followed by 300 mg daily for 8 weeks. The BPaLM regimen included BPaL plus moxifloxacin at a dose of 400 mg daily for 24 weeks, and the BPaLC regimen included BPaL plus clofazimine at a dose of 100 mg daily (or 50 mg if the patient weighed <30 kg) for 24 weeks. In stage 2 of the trial, patients were enrolled either into the standard-care group or into one of two investigational groups.

Procedures

Patients were randomly assigned in a 1:1:1:1 ratio in stage 1 of the trial and in a 1:1 ratio in stage 2 of the trial (see the Supplementary Appendix). Randomization lists were prepared by the trial statisticians, and randomization was stratified according to trial site.

The trial schedule (see the protocol) included weekly visits for the first 2 weeks, monthly visits until week 24, and then visits every 2 months until week 108. Each visit included laboratory tests, three electrocardiographic assessments, and a physical examination that included a neurologic assessment. Assessments of visual acuity and color blindness and audiometric testing were also performed according to the schedule. The investigators assessed adverse events at each visit. Serious adverse events, adverse events of special interest, pregnancies, and overdoses that were identified were reported to the pharmacovigilance officer within 24 hours.

At inclusion and at scheduled time points, two sputum samples were obtained for smear microscopy and culture in liquid medium with the use of the Mycobacteria Growth Indicator Tube (MGIT) system (Becton Dickinson). Drug-susceptibility testing was performed in M. tuberculosis isolates that were obtained at baseline and in any samples that were culture positive at week 16 or later. Culture conversion was defined as at least one positive culture at baseline and at least two consecutive negative cultures obtained at least 2 weeks apart. Paired whole-genome sequencing was conducted in the event of treatment failure or recurrence of tuberculosis.

Outcomes

In stage 1 of the trial, the primary efficacy outcome was culture conversion in MGIT liquid medium at 8 weeks after randomization. The primary safety outcome was the incidence of death or discontinuation of treatment for any reason by week 8.

In stage 2 of the trial, the primary outcome was an unfavorable status (a composite of death, treatment failure, treatment discontinuation, loss to follow-up, or recurrence of tuberculosis) at 72 weeks after randomization. The secondary efficacy outcomes were culture conversion at 12 weeks, time to culture conversion, composite unfavorable outcomes at 24 weeks and 108 weeks after randomization, and recurrence of tuberculosis by week 48 after randomization (in the investigational groups only).

The safety outcomes in stage 2 of the trial were at least one serious adverse event or an adverse event of grade 3 or higher at 72 and 108 weeks after randomization and at the end of treatment and the incidence of prolongation of the QTcF interval at week 24. Deaths and adverse events of special interest were also reported.

Analysis Populations

In the efficacy analyses, the intention-to-treat population included all patients who had undergone randomization. In the safety analyses, the as-treated population comprised all patients who had undergone randomization and received at least one dose of trial medication, and the patients were evaluated according to the regimen they received. The modified intention-to-treat population included patients in the intention-to-treat population who had received at least one dose of trial medication and excluded patients who did not have microbiologically proven rifampin-resistant tuberculosis. The per-protocol population included patients in the modified intention-to-treat population except those who did not complete a protocol-adherent course of treatment (>80% of doses within 120% of the prescribed duration) for any reason other than treatment failure or death and for those who discontinued treatment early because after they had received the first dose of trial treatment it was discovered that they had not met the inclusion or exclusion criteria.

Enrollment was terminated early for benefit, on March 18, 2021, in accordance with a recommendation from the data and safety monitoring board. We then performed an unplanned analysis, the results of which are presented here. In this analysis, the populations were restricted to include patients who could have had a prespecified outcome event at a given time point (i.e., week 24, week 72, and week 108).

Additional Analyses

After the stage 1 analysis, no analyses involving patients who were not included in stage 2 of the trial were planned. These patients were also followed to week 108, and these supportive data were viewed as important. All prespecified stage 2 analyses that involved the BPaLM group also were performed in the BPaLC and BPaL groups.

Statistical Analysis

The sample-size calculation is provided in the Supplementary Appendix. With the assumption that at 72 weeks after randomization 50% of the patients in the standard-care group and 45% of those in the investigational groups would have an unfavorable outcome event, we determined that a sample of 181 patients per group in trial stage 2 would provide the trial with approximately 85% power to detect a noninferiority margin of 12 percentage points. An alpha level of 1.7% (equivalent to a two-sided 96.6% confidence interval) was chosen to allow for both the adaptive nature of the design and the multiple comparisons of up to three groups. The estimated sample was increased to 201 patients per group to allow for patients who could not be evaluated. A noninferiority margin of 12 percentage points as the upper boundary of the confidence interval was determined to be a reasonable clinical and public health trade-off limit, given the benefits of a shorter treatment duration, decreased pill burden and regimen cost, and the all-oral nature of the investigational regimens. This noninferiority margin was congruent with that in recent trials involving patients with drug-resistant tuberculosis in which the noninferiority margin was 10 to 12 percentage points.7,8

The efficacy outcomes were analyzed in the intention-to-treat, modified intention-to-treat, and per-protocol populations, and the safety outcomes were analyzed in the as-treated population. Binary outcomes were summarized with absolute risk differences (with the use of a generalized linear model for a binomial outcome with an identity function) and risk ratios (with the use of a generalized linear model for a binomial outcome with a log-link function). Adjustment for randomization site was planned in all analyses. For the primary efficacy and safety outcomes, corresponding two-sided 96.6% confidence intervals were reported for effect estimates, and two-sided 95% confidence intervals were reported for secondary efficacy outcomes. The secondary outcomes were not adjusted for multiplicity. Prespecified subgroup analyses were conducted for the primary efficacy outcome. For binary safety outcomes, risk differences are reported with the use of the same approach as that described above. For the safety outcome of the QTcF value at 24 weeks, the difference in the mean value in each investigational group from the mean value in the standard-care group was assessed with adjustment for baseline QTcF values and with the use of linear regression.

Additional analyses of safety and efficacy were conducted in the BPaLC and BPaL groups with the use of the same approach but with two-sided 95% confidence intervals. Additional details are provided in the Supplementary Appendix.

留言 (0)

沒有登入
gif