More evidence supports isoniazid for TB prevention in people with HIV

Keith Alcorn
Published: 19 February 2010

Two new studies presented this week at the 17th Conference on Retroviruses and Opportunistic Infections (CROI) in San Francisco add weight to a growing expert consensus that isoniazid preventive therapy (IPT) should be provided to people with HIV in order to reduce the risk of developing active tuberculosis (TB).

Although, as TB expert Richard Chaisson pointed out at an expert discussion on IPT prior to the conference, IPT is proven to reduce the incidence of TB in people with HIV, there are still many unanswered questions, including its effect on overall mortality, the optimum TB preventive regimen, and the effect of IPT on secondary transmission of TB in the household and among close contacts.

The studies presented this week offer new insights about the additive benefit of IPT in people taking antiretroviral drugs, and provide evidence about a potential alternative to the use of isoniazid alone.

IPT in people taking antiretroviral therapy

Dr Craig Innes of the Aurum Institute in South Africa presented results of an observational analysis of patients receiving IPT through a workplace programme in South Africa.

The study looked at the effect of IPT on mortality in a population of South African adults, predominantly male (93%), who started antiretroviral therapy (ART) between January 2004 and December 2007, comparing outcomes in those who received IPT and those who did not.

Individuals receiving health care through the Aurum programme were eligible for ART if they had a CD4 count below 250 cells/mm3 without symptoms, if they had WHO (World Health Organisation) stage 4 disease, or if they had WHO stage 3 disease with a CD4 count below 350 cells/mm3.

Isoniazid preventive therapy was recommended if a patient had no history of TB in the past three years and if current TB could be excluded, although prescription was less than universal due to fears of isoniazid resistance among some physicians if they could not exclude active TB.

The study population comprised 3258 patients who started ART, of whom 910 received IPT. The only significant differences between those who received IPT and those who did not were history of previous TB ever (2.5% vs 8.9%), WHO stage 3 or 4 disease (30% vs 50%) and haemoglobin (13.4 vs 12.9) (all differences p < 0.001).

The study showed a significantly lower mortality rate for those who received IPT: across a number of analyses, which adjusted for potential confounding factors, the risk of death was approximately halved if an individual received IPT in addition to antiretroviral therapy.

In the first analysis, which included individuals with a previous history of TB and which adjusted for age, WHO disease stage, CD4 count, haemoglobin count and year of ART commencement, the risk of death was 53% lower in those who received IPT (hazard ratio [HR] 0.47, 95% confidence interval [CI] 0.30-0.72, p<0.001).

Some patients received IPT despite guidelines indicating that they should not, and this might have diluted the effect of IPT that would be observed if the guidelines had been strictly observed. In a sensitivity analysis which excluded those with a prior history of TB, the risk of death was 54% lower, indicating no substantive effect of prior TB (HR 0.46).

Another way of looking at the question of whether, if used strictly according to guidelines, the results might turn out differently, is to exclude all patients who had symptoms associated with TB at ART commencement, no matter when they received IPT. This would tend to cast a wider net for those patients who had active TB when exposed to isoniazid, and who might therefore have shown less benefit.

In this analysis, the risk of death was 53% lower (HR 0.47, 95% CI 0.28-0.79, p = 0.05).

Only when participants were stratified according to whether they commenced IPT within three months of starting ART, or delayed IPT for more than three months, did a marginal difference emerge. Those who started IPT less than three months after ART had a 61% reduction in the risk of death, whereas those who started IPT more than three months after ART had a 46% reduction in the risk of death when compared to those who did not receive IPT.

The chief limitations of the study are its observational nature – there could have been biases in who received IPT, although the analyses controlled for known confounding factors – and the fact that causes of death were not captured in this study. It is not clear whether the effect of IPT on mortality is solely driven by a reduction in TB-related deaths, or whether protecting IPT has a wider effect.

The interaction between ART and tuberculin skin test status

The Aurum Institute study did not control for tuberculin skin test (TST) status, a test that can positively confirm that someone is latently infected with TB. In the BOTUSA IPT study, (which compared 6 and 26 months of IPT in people with HIV), previously presented at the World Lung Health conference in Cancun in December 2009, and presented again this week at CROI, TST-positive patients who received antiretroviral therapy alongside six months of isoniazid experienced a 50% reduction in the risk of developing active TB during the three-year follow-up period when compared to people who received six months of isoniazid alone.

In TST-positive people who received 36 months of isoniazid however, the additional benefit of antiretroviral therapy was marginal, reducing the likelihood of developing active TB by just 4% in comparison to the no-ART group.

In TST-negative people – those who either had no prior exposure to TB, or else hadn’t enough immune function to respond to TB antigens in the skin test – antiretroviral therapy halved the risk of developing active TB in those who received six months of isoniazid, and reduced the risk of active TB by around 45% in those who received 36 months of isoniazid when compared to isoniazid alone.

A new regimen for TB prevention?

A randomised study conducted by the Tuberculosis Research Center in Chennai, India, showed that a six-month course of isonaziad and ethambutol, another antibiotic used in TB treatment, was just as effective as a 36-month course of isoniazid in people with HIV, most of whom were not receiving antiretroviral therapy.

The Indian study was designed to test whether an alternative regimen to isoniazid alone was safe and effective in a setting where 15 to 20% of patients have isoniazid-resistant TB at diagnosis, and where rifampicin, another TB drug tested as a preventive measure, is strictly reserved for treatment of active TB in order to limit the development of resistance to the drug. The study was also deisgned to compare the feasibility of a short-course regimen of six months against a longer-term regimen that might be expected to offer greater protection against exposure to TB during the study period itself, especially in immunosuppressed people who may be at higher risk of rapid TB progression.

The study was conducted by the Tuberculosis Research Center in Chennai, and began recruiting patients for the three-year study between 2001 and 2005.

All participants had TB confirmed by sputum culture wherever possible, and the study excluded any person with HIV who had a previous history of TB.

During the study patients received a clinical review every three months to check for symptoms of TB and other health problems, and underwent chest X-ray every six months.

Participants were randomised to receive a daily regimen of isoniazid 300mg and ethambutol 800mg for six months, or 36 months of isoniazid alone, and all patients with a CD4 count below 250 cells/mm3 received cotrimoxazole. Antiretroviral therapy became available in the public sector in 2004 to patients with WHO stage 4 disease, or WHO stage 3 disease and a CD4 count below 200 cells/mm3).

The study randomised 683 patients, and 37 cases of TB occurred during the three years of follow-up, 16 of which were bacteriologically confirmed. The incidence of TB was not significantly different in the two arms of the study (2.4 cases per 100 person years in the ethambutol arm, 1.6 cases per 100 person years in the isoniazid arm), and death rates were also similar.

The low rates of TB in the study compared to the historical incidence previously measured in the Chennai cohort (6.9 cases per 100 person years) may in part be attributable to the screening by culture at baseline. Screening picked up 30 cases of active but asymptomatic TB that otherwise would have developed into incident TB cases during the trial.

In both arms, most cases of TB – and most deaths – occurred during the first 12 to 18 months of follow-up, but only three deaths were a result of TB.

Regardless of the regimen received, individuals with a baseline CD4 count had a fourfold higher incidence of TB, while TST-positive individuals (TST >5mm) had a 40% higher risk of developing TB.

Among the TB cases that were bacteriologically confirmed, six had isoniazid resistance (five in the ethambutol arm and one in the isoniazid arm).

There was no difference in the development of adverse events between the two arms: three severe events occurred in the ethambutol arm and two in the isoniazid arm, and overall the regimens were very well-tolerated. Rates of adherence were also very high, with around 93% in each arm judged to be more than 80% adherent by unannounced home visit pill counts.

Dr Swaminathan commented that the high adherence rates, the high study retention and the low rates of TB in the study may have been due to the high level of preparation and screening that patients received before entering the study. Patients in the trial received free HIV care, and for many, it may have been their first experience within the Indian health system of good-quality care and support, leading to a high level of patient loyalty to the programme, and a reluctance to be referred on for care at other centres when the trial was over.


Innes C et al. Effectiveness of isoniazid preventive therapy in reducing mortality in patients on ART. Seventeenth Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 102, 2010.

Samandari T et al. Randomized, placebo-controlled trial of 6 vs 36 months isoniazid TB preventive therapy for HIV-infected adults in Botswana. Seventeenth Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 104LB, 2010.

Swaminathan S et al. Efficacy of a 6-month vs a 36-month regimen for prevention of TB in HIV-infected persons in India: a randomized clinical trial. Seventeenth Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 103, 2010.

Further information

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap

This content was checked for accuracy at the time it was written. It may have been superseded by more recent developments. NAM recommends checking whether this is the most current information when making decisions that may affect your health.

NAM’s information is intended to support, rather than replace, consultation with a healthcare professional. Talk to your doctor or another member of your healthcare team for advice tailored to your situation.