Shortened Tuberculosis Drug Regimen Wins in Trial

A novel tuberculosis drug combination given for 4 months was noninferior to a standard 6-month regimen for an outcome of culture-negative survival at 12 months, an international randomized phase III trial found.

That endpoint was achieved by 82.9% of microbiologically eligible patients (i.e., those found to be drug-resistant or with no mycobacteria-positive culture at baseline) treated with a regimen based on rifapentine plus moxifloxacin for 4 months, compared with a cure rate of 83.7% among those receiving a combination of rifampin, isoniazid, pyrazinamide, and ethambutol for 6 months, reported Susan E. Dorman, MD, of the Medical University of South Carolina in Charleston, and colleagues. That small difference met prespecified criteria for noninferiority.

The 4-month oral rifapentine/moxifloxacin regimen was also noninferior to the 6-month control treatment for “unfavorable” outcomes, which could include those unrelated to tuberculosis, as well as those stemming directly from the infection, at rates of 15.5% and 14.6%, respectively, they noted in the New England Journal of Medicine (NEJM).

Only small differences separated the groups for other outcomes, including early discontinuation and serious treatment-related adverse events, the researchers said. Moreover, sensitivity analyses confirmed the overall similarity of the two regimens.

A third 4-month treatment arm involving rifapentine without moxifloxacin, however, was less effective than the 6-month control by several percentage points and did not meet the noninferiority standard.

In an accompanying editorial, NEJM Editor-in-Chief Eric Rubin, MD, PhD, and Valerie Mizrahi, PhD, of the University of Cape Town in South Africa, said the length of time needed to cure tuberculosis has long been a source of frustration. “It was almost as though 6 months represented some strict limit — that is, until now,” they wrote in applauding the new study’s findings.

Shaving 2 months off the standard represents a real advance, they argued, but it’s not without its own drawbacks. Rifapentine needs to be taken with meals (unlike rifampin, a chemically similar agent from which rifapentine was derived), which “could introduce new issues with adherence,” Rubin and Mizrahi wrote. And the addition of a fluoroquinolone, moxifloxacin, could promote antibiotic resistance and bring other problems known to be associated with the class.

There are pluses as well, the editorialists said. Both rifapentine and moxifloxacin are widely available (although Dorman’s group noted that they are somewhat more expensive on a daily basis than the drugs used in the standard 6-month regimen) and the general distribution and treatment procedure won’t have to change.

Rubin and Mizrahi suggested that the main takeaway from the study could be that “there is no magic with 6 months of therapy” and that successful cure might one day be possible in less than 4 months.

Study Details

Patients newly diagnosed with tuberculosis were enrolled in the U.S. and 12 other countries in South America, Asia, and Africa. They were randomized prior to baseline testing, such that of the 2,516 assigned in roughly equal numbers to the three treatment arms, 173 did not meet the microbiological eligibility criteria. More than 70% of the sample were Black; most of the rest were Asian or multiracial.

Microbiological eligibility was determined via molecular sputum analysis. Patients with HIV were not excluded, but they were enrolled only if CD4 cell counts reached 100 cells/mm³ or greater.

The three regimens, all open-label, were as follows:

• Control: 8 weeks of rifampin, isoniazid, pyrazinamide, and ethambutol, all given once daily, then 18 weeks of rifampin and isoniazid, also once daily
• Rifapentine/moxifloxacin: 8 weeks of those two drugs plus isoniazid and pyrazinamide, all once daily, then 9 weeks of the same regimen minus the pyrazinamide
• Rifapentine: 8 weeks of rifapentine, isoniazid, pyrazinamide, and ethambutol, once daily, then 9 weeks of rifapentine and isoniazid

Rifapentine was given at 1,200 mg/day — with instructions to take it within an hour after eating — and moxifloxacin at 400 mg/day. Dosages of the other agents were based on body weight, according to usual standards.

The primary analysis was conducted on an intention-to-treat basis. Secondary analyses included all assessable patients and those with at least 75% and 95% adherence to the dosing schedule.

Those per-protocol evaluations showed that the control treatment was superior to rifapentine/moxifloxacin, with cure rates about 3 percentage points higher. That no difference was seen in the primary analysis could be interpreted as highlighting the difficulty in maintaining good adherence to the 6-month regimen.

For safety, deaths and overall adverse events did not differ between the control and rifapentine/moxifloxacin groups. Only 14 deaths were recorded among all participants.

Compared with the control group, the most noteworthy safety signals with rifapentine/moxifloxacin were higher rates of discontinuation because of adverse events (1.9% vs 0.8% with control treatment) and bilirubin levels ≥3 times the upper limit of normal (3.3% vs 1.0%). On the other hand, serious aspartate transaminase (AST)/alanine transaminase (ALT) elevations were more common in the control group (2.9% vs 1.9% with rifapentine/moxifloxacin for AST/ALT levels ≥5 times the upper limit of normal; 1.1% vs 0.5% for ≥10× elevations). Taken together using Hy’s law criteria, these signals of liver toxicity were similar in the two groups.

Dorman and colleagues noted some limitations to the study, including the open-label design and lack of a placebo group. The researchers did not examine the different regimens’ cost-effectiveness, but suggested it as a future focus for research, given that the rifapentine/moxifloxacin regimen’s shorter duration should offset at least some of its increased daily expense.