4 months of rifampicin safer than 9 months of isoniazid for TB prevention

Keith Alcorn
Published: 21 May 2008

A four month course of rifampicin is significantly less likely to cause serious side-effects than a nine-month course of isoniazid when used as a preventive treatment in people with latent tuberculosis, and patients in the randomised study were significantly more likely to complete their course of treatment, Canadian researchers reported on Tuesday at the American Thoracic Society annual conference in Toronto.

The rifampicin regimen is also more cost-effective than the isoniazid, according to a separate analysis of the trial also reported at the conference. The trial did not compare the effect of the two regimens in preventing new cases of TB.

Isoniazid preventive treatment (IPT) for latent TB lasting six to nine months is recommended by the World Health Organization for all HIV-positive people with latent TB diagnosed through tuberculin sensitivity testing.

However the preventive regimen, often labelled 9INH by doctors, has a number of drawbacks. Completion rates can be low, and there is a risk of liver toxicity, especially where pre-existing liver injury or heavy alcohol consumption is present. For these and other reasons, developing countries have been reluctant to systematically offer IPT despite the urging of the World Health Organization.

In the United States the American Thoracic Society and the Centers for Disease Control and Prevention recommended that a two-month regimen of rifampicin (usually known as rifampin in the US) and pyrazinamide could be used as a preventive treatment in latent TB if there is no evidence of pre-existing liver injury or alcoholism. But in people without HIV infection a nine month course of isoniazid remains the preferred treatment.

However the US guidelines also gave the option of using a four month course of rifampicin alone.

“As soon as we started using rifampicin/pyrazinamide very broadly, we found a lot of hepatitis, so we decided: let’s look at the safety of 4RIF [the four month rifampicin regimen],” Dr Dick Menzies of McGill University in Montreal told aidsmap.

The safety study, conducted by the Montreal Chest Institute with colleagues in Canada, Brazil and Saudi Arabia, enrolled 847 patients; 427 were randomised to 9INH and 420 to 4RIF. The study recruited both HIV-negative and HIV-positive people, but excluded people taking antiretrovirals due to potential drug interactions and exacerbation of isoniazid toxicity by the commonly used antiretroviral stavudine (d4T).

The trial was halted by the Data Safety Monitoring Board at the point when 75% of participants had completed four months of treatment, due to the finding that the rate of adverse events in the 4RIF arm was not significantly greater than in the 9INH arm.

In the analysis presented to the American Thoracic Society conference the investigators showed that grade 3 and 4 hepatitis was significantly more common in the 9INH arm (15 cases versus 3, 4.1% vs 0.8%, p=0.002) than in the 4RIF arm. There was no significant difference in the number of grade 1 and 2 adverse events judged to be `probably related` to the study drugs. The incidence of serious adverse events in the 9INH arm was somewhat higher than has been seen in other studies.

There was also a significantly higher treatment completion rate in the 4RIF arm (81% vs 64%, p<0.001).

In a separate analysis, Anne Aspler of the Respiratory Epidemiology & Clinical Research Unit at McGill University’s Montreal Chest Institute demonstrated that 4RIF was significantly less costly than 9INH for the treatment of latent TB, based on costs from a single site in Montreal but using data from the entire study. Overall costs, including health system utilisation, were $1021 per patient who completed 4RIF and $1542 per patient who completed 9INH.

“Overall, rifampin costs about $484 less per patient treated, which, if we assume that four months of rifampin has at least equal efficacy to nine months of isoniazid, represents an added saving to the health system of more than $10,000 per patient prevented from developing TB disease,” she said. “And because of improvements in compliance, we are actually preventing more cases. This treatment can save money as well as lives.”

“The next step will have to be an efficacy study which must be much larger,” said Dr Menzies.

However a rifampicin-based preventive treatment for TB does have one drawback in people with HIV: it has potentially serious interactions with all protease inhibitors and the NNRTI nevirapine, and the only widely used drug that can be prescribed alongside it is efavirenz. Nucleoside analogues are not affected by rifampicin.

Some doctors in developing countries advocate offering IPT to people with HIV before they are eligible for antiretroviral therapy since this may be a good method of engaging them with the health system and retaining them in care. For these patients a four-month rifampicin regimen could be attractive, especially if the cost of the regimen comes down.

However there will be significant and possibly insurmountable obstacles to using a four month rifampicin regimen as preventive treatment. Because rifampicin is an essential current component of treatment for active tuberculosis, it will be essential to exclude the possibility that someone has active TB before giving rifampicin as a preventive treatment, in order to limit the risk that rifampicin resistance could develop as a result of monotherapy in a patient who has active TB.

Also, a shorter alternative regimen has already been tested, and others could be in the pipeline. A randomised study published in 2006 showed that a 12 week course of rifapentine and isoniazid taken once a week was just as effective as an eight week course of rifampicin and pyrazinamide taken daily, but much better tolerated. Twenty of 193 participants (10%) receiving rifampin/pyrazinamide experienced grade 3 or 4 hepatotoxicity, compared with 2 of 206 participants (1%) on rifapentine/isoniazid. A Canadian study is also comparing three months of rifapentine/isoniazid taken weekly with the standard 9INH regimen.


Isoniziad preventive therapy and how to deliver it was discussed in detail in the Novermber 2007 edition of HIV & AIDS Treatment in Practice, NAM's electronic newsletter on HIV treatment in developing countries.


Dion MJ et al. A randomized trial comparing adverse events with 4RIF and 9INH in the treatment of latent TB infection. American Thoracic Society International Conference, Toronto, poster 816, 2008.

Aspler A et al. Cost and cost-effectiveness of a randomized trial of 4 months rifampin versus 9 months of isoniazid for treatment of latent tuberculosis infection. American Thoracic Society International Conference, Toronto, poster 821, 2008.

Schechter M et al. Weekly rifapentine/isoniazid or daily rifampin/pyrazinamide for latent tuberculosis in household contacts. American Journal of Respiratory and Critical Care Medicine, 173: 922-926, 2006.