Michael Smith, : April 12, 2011
A fixed-dose combination of four tuberculosis drugs yielded outcomes comparable to the four drugs given separately, researchers reported.
The finding is regarded as a step toward more effective TB control, especially in the developing world, according to Christian Lienhardt, MD, PhD, of the World Health Organization in Geneva, and colleagues.
Fixed-dose combinations are easier to administer, require fewer pills, and have been seen as way to prevent drug resistance arising from incorrect use of medications, Lienhardt and colleagues reported in the April 13 issue of the Journal of the American Medical Association.
But, among other things, doubts about their efficacy in comparison with the same drugs given separately have slowed their acceptance, the researchers noted.
To help clarify the issue, they conducted the so-called Study C trial, a parallel-group, open-label, non-inferiority, randomized controlled study in Africa, Asia, and Latin America between 2003 and 2008.
The participants -- 1,585 adults with newly diagnosed smear-positive pulmonary tuberculosis -- were treated with daily rifampicin, isoniazid, pyrazinamide, and ethambutol for an intensive treatment period of eight weeks, followed by 18 weeks of rifampicin and isoniazid three times a week.
They were randomly assigned to get their medications either separately or in fixed-dose combination pills for the first eight weeks; both groups received a combination pill of rifampicin plus isoniazid thereafter.
The study was designed to show noninferiority both among patients who got all their medications correctly (the per-protocol population) and among those who got any study drug (the modified intent-to-treat population).
The noninferiority margin was specified as no more than four percentage points between the arms in the proportion of favorable bacteriological outcomes, the researchers reported.
In the modified intent-to-treat population, any treatment change was regarded as yielding an unfavorable outcome, but after the trial ended, the steering committee recommended a more "realistic" model in which changes for reasons other than therapeutic outcomes were classified on the basis of their bacteriological outcome.
The researchers found: In the per-protocol analysis, 555 of 591 patients in the fixed-dose group -- or 93.9% -- had a favorable outcome, compared with 548 of 579 -- or 94.6% -- in the separate-drugs group. The difference in risk was -0.7% and the 90% confidence interval ranged from -3.0% to 1.5% -- within the four-point noninferioirty margin.
In the intent-to-treat group, 570 of 684 fixed-dose patients (83.3%) had a favorable outcome compared with 563 of 664 patients (84.8%) in the separate-drugs group. The risk difference was -1.5%, and the 90% confidence interval ranged from -4.7% to 1.8% -- exceeding the noninferiority margin.
In the post hoc intent-to-treat population, 591 of 658 fixed-dose patients (89.8%) and 589 of 647 separate-drugs patients (91.0%) had a favorable outcome. The risk difference was -1.2% and the 90% confidence interval ranged from -3.9% to 1.5% -- within the noninferiority margin.
Adverse events were not significantly different between the arms and were mainly those expected from the drugs, the researchers reported.
While the fixed-dose combination did not meet the strict definition for noninferiority established at the beginning of the trial, the researchers argued it was noninferior on two of the three analyses.
The fixed-dose combination should be preferred because of other advantages, they argued.
Among them, they noted, is that patients must take fewer pills -- three to four a day, depending on weight, compared with nine to 16 a day for the separate drugs.
The authors noted that the use of a fixed combination after the first eight weeks for both groups was a limitation of the study, but stated that this is standard practice in most of the participating countries.