Adverse Events with 4 Months of Rifampin Therapy or 9 Months of Isoniazid Therapy for Latent Tuberculosis Infection
Dick Menzies,Richard Long,Anete Trajman,Marie-Jose Dion,Jae Yang,Hamdan Al Jahdali,Ziad Memish,Kamran Khan,Michael Gardam,Vernon Hoeppner,Andrea Benedetti,Kevin Schwartzman
DOI: https://doi.org/10.7326/0003-4819-149-10-200811180-00003
IF: 39.2
2008-11-18
Annals of Internal Medicine
Abstract:BACKGROUND: Treatment of latent tuberculosis infection with isoniazid for 9 months is complicated by poor patient adherence and the need for close follow-up of side effects, especially hepatotoxicity. Shorter and safer regimens are needed.OBJECTIVE: To compare the frequency of adverse events and treatment completion in 2 treatment regimens for latent tuberculosis infection.DESIGN: Multicenter, randomized, open-label trial.SETTING: Tuberculosis clinics located in university hospitals in Canada, Brazil, and Saudi Arabia.PATIENTS: 847 patients without a contraindication for rifampin and requiring treatment for latent tuberculosis infection.INTERVENTION: Four months of daily rifampin therapy or 9 months of daily isoniazid therapy.MEASUREMENTS: Grade 3 to 4 drug-related adverse events resulting in drug discontinuation (primary outcome), and on-time treatment completion, grade 1 to 2 drug-related adverse events, and changes in liver enzymes and hematologic variables (secondary outcomes).RESULTS: Seventeen of 422 participants who started isoniazid therapy developed grade 3 to 4 adverse events compared with 7 of 418 who started rifampin therapy (risk difference [rifampin minus isoniazid], -2.3% [95% CI, -5% to -0.1%]; P = 0.040). Grade 3 or 4 hepatitis occurred in 16 of 422 isoniazid recipients compared with 3 of 418 rifampin recipients (risk difference, -3.1% [CI, -5% to -1%]; P = 0.003). Grade 1 or 2 adverse events attributed to study drugs occurred with similar frequency. Asymptomatic reduction in platelet and leukocyte counts were more frequent in rifampin recipients. More patients completed rifampin treatment (78%) than isoniazid treatment (60%) (difference, 18% [CI, 12% to 24%]; P < 0.001]).LIMITATION: The study did not measure efficacy, and the open-label design may increase the chance of bias in ascertainment of adverse events.CONCLUSION: Treatment of latent tuberculosis with 4 months of rifampin leads to fewer serious adverse events and better adherence than 9 months of isoniazid. These findings justify a large-scale trial to compare the efficacy of rifampin with that of isoniazid.
medicine, general & internal