Kevin Schwartzman's excellent commentary1 has highlighted an inconsistency in current Canadian guidelines:2 he recommends treatment of latent tuberculosis (TB) infection for HIV-infected immigrants from TB- endemic countries, even if that person has a tuberculin skin test (TST) reaction of < 5 mm. This group would be composed of the truly uninfected, who would derive no benefit from this treatment, and those who are infected but anergic. Two studies have found no evidence of significant benefit from treatment of latent tuberculosis infection (LTBI) in the latter.3,4 Three other studies in high TB prevalence countries, in which results of anergy testing were not reported, also failed to show a benefit of treatment among HIV- infected individuals who had negative tuberculin tests. 5,6,7
We do not feel the evidence or other current recommendations8,9 support routine provision of LTBI treatment to TST-negative, HIV-infected individuals on the basis of geographic origin alone.
Our second concern relates to the statement that the use of the 2-month pyrazinamide and rifampin regimen for latent tuberculosis is “clearly contraindicated for anyone with underlying liver disease or with isoniazid- related hepatatoxicity.” We agree that the use of this regimen should be strictly limited to individuals with a particularly high risk of TB reactivation, such as the HIV-infected, and to those in whom completion of a standard 9-month course of isoniazid would be unlikely. However, in many Canadian settings, a high proportion of patients meeting these criteria have some indication of liver disease from hepatitis C infection, excess alcohol use, or both. The reported experience of serious adverse effects from the US10 appears to have involved self- administration, variable follow-up and insufficient attention to the high liver disease risk of this selected patient group.
For many years, pyrazinamide and rifampin have been used as part of a 4-drug therapy for active tuberculosis, with manageable toxicity in patients with liver disease. We believe that treatment of LTBI with pyrazinamide and rifampin can be administered to carefully selected patients with hepatitis C or a history of excess alcohol use, with an acceptably low risk, if the following criteria are met: directly observed delivery of each dose, immediate assessment of any clinical symptoms of liver disease and measurement of transaminase enzymes at baseline and every 2 weeks during therapy.
Stan Houston Division of Infectious Diseases Department of Medicine University of Alberta Edmonton, Alta. Richard Long Division of Pulmonary Medicine Department of Medicine University of Alberta Edmonton, Alta. Vernon Hoeppner Division of Tuberculosis Control Department of Medicine University of Saskatchewan Saskatoon, Sask.