Background: Daily isoniazid therapy for nine months is recommended for latent tuberculosis infection (LTBI) in most countries. Although isoniazid is more than 90 percent effective in eradicating LTBI, approximately one half of patients do not complete the lengthy treatment course. Furthermore, drug-induced hepatitis and other adverse effects of isoniazid require careful monitoring. Daily therapy with rifampin and pyrazinamide for two months was proposed as an alternative in 2000, but has fallen into disfavor because of reports of fatal hepatotoxicity. Several nonrandomized trials have reported that four months of daily rifampin therapy has better adherence and less hepatotoxicity than nine months of daily isoniazid therapy; however, systematic reviews are lacking. Menzies and colleagues conducted a randomized, open-label trial of patients with LTBI to compare the adverse effects and treatment completion rates for rifampin and isoniazid.

The Study: Patients received four months of daily rifampin (10 mg per kg, up to 600 mg per day) or nine months of daily isoniazid (5 mg per kg, up to 300 mg per day). Patients were excluded if they were in contact with known cases of isoniazid- or rifampin-resistant tuberculosis or were taking other medications that posed notable potential drug interactions. All patients received blood tests, including complete blood counts and measurement of aspartate transaminase (AST) and alanine transaminase (ALT) levels, at baseline and after one and two months of therapy, and were clinically monitored during the first four months. Those receiving isoniazid were subsequently monitored every six weeks at physician discretion. The primary outcome was the frequency of drug discontinuation because of hepatotoxicity, and the secondary outcome was treatment completion, defined as taking more than 80 percent of the doses of the study medication.

Results: Overall, 847 patients were randomized, of whom 376 (44 percent) were 35 years or older. Persons in the rifampin group were much more likely to complete the treatment course than those in the isoniazid group (78 versus 60 percent, respectively; P < .001). Hepatotoxicity (AST or ALT levels at least three times the upper limit of normal with symptoms, or at least five times the upper limit of normal, regardless of symptoms) was less common in the rifampin group than the isoniazid group (0.7 versus 3.8 percent, respectively; P = .003). Although most episodes of hepatotoxicity occurred during the first two months of treatment, a few cases developed throughout therapy. Risk factors for hepatotoxicity included comorbid illness and intravenous drug use, but not smoking, alcohol use, baseline elevated liver enzymes, or age.

Other adverse effects occurred less often. Rash was more common with rifampin than isoniazid (1.9 versus 1.2 percent, respectively), and two patients taking rifampin developed neutropenia, which resolved after stopping therapy. Only one hematologic reaction was reported in the isoniazid group, and two were reported in the rifampin group.

Conclusion: Treating LTBI with four months of rifampin is associated with better completion rates and fewer adverse effects than isoniazid, particularly hepatotoxicity. However, this study was not designed to compare rates of effectiveness, and the authors recommend a large-scale trial to assess this.

editor's note: Although isoniazid for nine months is the preferred treatment for LTBI, the Centers for Disease Control and Prevention lists four months of rifampin as an acceptable alternative.¹ However, few well-controlled trials have examined this option. Rifampin can cause thrombocytopenia and anemia, and appropriate monitoring is warranted even though these effects are uncommon. There is also concern about drug resistance. For example, methicillin-resistant Staphylococcus aureus can quickly become resistant when treated with rifampin monotherapy.²

Although this study did not compare the effectiveness of rifampin versus isoniazid, similar rates of effectiveness were reported in another small study.³ Also, more patients completed the four-month rifampin course than the longer isoniazid course. Until more data are available, the nine-month isoniazid therapy is preferable, and the alternative regimen can be reserved for select, potentially noncompliant patients.—k.t.m.