ABSTRACT: Treatment outcomes among multidrug-resistant tuberculosis (MDR-TB) patients receiving ethambutol, cycloserine, or terizidone as part of a standardized regimen were compared, determining occurrence of serious adverse drug events (SADEs). Newly diagnosed adult MDR-TB patients were enrolled between 2000 and 2004, receiving a standardized multidrug regimen for 18 to 24 months, including ethambutol, cycloserine, or terizidone. Cycloserine and terizidone were recorded individually. SADEs and factors associated with culture conversion and unfavorable treatment outcomes (default, death, treatment failure) were determined. Of 858 patients, 435 (51%) received ethambutol, 278 (32%) received cycloserine, and 145 (17%) received terizidone. Demographic and baseline clinical data were comparable. Successful treatment occurred in 56%, significantly more in patients receiving cycloserine (60%) and terizidone (62%) than in those receiving ethambutol (52% [P = 0.03]). Defaults rates were 30% in ethambutol patients versus 15% and 11% for cycloserine and terizidone patients, respectively. Terizidone was associated with fewer unfavorable outcomes (adjusted odds ratio [AOR], 0.4; P = 0.008; 95% confidence interval [CI], 0.2 to 0.8). Patients receiving cycloserine were more likely to achieve culture conversion than those receiving ethambutol or terizidone (AOR, 2.2; P = 0.02; 95% CI, 1.12 to 4.38). Failure to convert increased the odds of unfavorable outcomes (AOR, 23.7; P < 0.001; 95% CI, 13 to 44). SADEs were reported in two patients receiving ethambutol, seven patients receiving cycloserine, and three receiving terizidone (P = 0.05). Ethambutol was associated with high culture conversion and default rates. Cycloserine achieved higher culture conversion rates than terizidone. Fewer patients on terizidone experienced SADEs, with lower default rates. The differences that we observed between cycloserine and terizidone require further elucidation.
