CROI: Greater declines in TB incidence observed with long-term antiretroviral therapy

Background

TB is one of the principal causes of HIV-associated morbidity and mortality. However, since the emergence of the HIV epidemic, the DOTS (directly observed therapy) strategy has failed to contain the spread of TB in Africa, particularly in southern Africa. Last year, the World Health Organization declared TB a regional health emergency in Africa.

"The DOTS strategy, whilst important and necessary, is not sufficient alone and adjunctive strategies are needed," said Dr Lawn.

One possible solution is improving access to ART. Several studies have shown that, over one or two years, ART can reduce the incidence of TB by 70-90%. One of these studies, conducted in Cape Town, found that the risk of TB incidence was low among people treated with ART, and that the benefit was observed across the spectrum of CD4 cell counts and WHO stages of disease (Badri).

However, a number of other factors could affect the impact of ART on TB control within the community such as the reduction of risk over time; how effectively ART treatment is introduced into the wider community; when patients are treated (early or later in the course of HIV disease) and their prolonged life expectancy. One model suggested that if ART were only given to people with CD4 cells below 200, even if there was 100% coverage and compliance, TB incidence would only decrease by 22% over 20 years (Williams & Dye). However, this model was calculated using only the decrease in incidence over a short duration of antiretroviral therapy.

But what if TB incidence rates continue to decrease the longer someone is on ART? And what if there are other risk factors that could be addressed to reduce the incidence of TB while on ART?

The incidence of tuberculosis (TB) in people on antiretroviral therapy (ART) continued to decrease over the first three to five years on treatment in a South African study reported this week at the Thirteenth Conference on Retroviruses and Opportunistic Infections in Denver.

According to the study’s presenter, Dr Stephen Lawn of the University of Cape Town, ART "may contribute more to TB control in low-income countries than has previously been estimated."

However, the rate of TB after five years on ART was still higher than in the general populations suggesting that other methods, such as isoniazid prophylaxis, may be needed alongside ART to control TB. (The presentation can be viewed or listened to by visiting the conference’s website).

Results

With a median of 40 months of follow-up per patient, 1108 person-years of observation were accrued. During this period, 27 cases of TB were diagnosed. But the likelihood of developing TB decreased with time on treatment.

The TB incidence rate was 3.5 cases/100 person-years in the first year and continued to decrease significantly during follow-up (p trend = 0.002), plateauing at rate of 1.01 cases/100 person-years in the fourth and fifth years of treatment. This incidence rate is still high at around 1000 cases/100,000 per year. "This would be consistent with the hypothesis that ART, whilst resulting in substantial recovery of immune function against mTB, [leaves immune] restoration incomplete," said Dr Lawn.

The TB incidence was highest among patients with baseline CD4 cell counts <100 cells and those with WHO stage 3 or 4 disease (5.71 and 3.88 cases/100 person-years, respectively).

WHO stage 3 & 4 disease at baseline (adjusted relative risk (ARR) = 3.60; 95%CI 1.32 to 9.80; p = 0.01) and CD4 cell counts below 100 at baseline (adjusted risk ratio [ARR] = 2.38; 95%CI 1.01 to 5.60; p = 0.04)) were the strong risk factors for developing TB while on ART.

This could be due to the fact that the immune system takes longer to recover in these patients or that persons who begin ART with more advanced disease have less capacity to recover. This may be supported by the observation that CD4 cell increases were smaller in those who developed TB, although it isn’t clear whether this was a cause or effect of TB. There was no statistically significant difference in changes in viral load between those who developed TB and those who did not.

Age below 33 was also associated with a greater risk of developing TB. However, plasma viral load, previous history of TB, low socioeconomic status or sex were not associated with an increased incidence of TB.

The study

So Dr Lawn and his colleagues conducted a study using a prospective hospital-based cohort, the Cape Town AIDS cohort (CTAC), to determine the rates and risk factors for incident TB during ART. The incidence of TB was assessed in 346 patients receiving ART between 1996 and 2005. TB cases were confirmed by microbiology or histology.

At baseline, the median age was 33 years, 55% were male. The median CD4 cell count was 242 cells (IQR 120-343), and 51% had WHO stage 3 or 4 disease. 14% had a previously history of TB.

Conclusion

"Clearly, modelling calculations need to be repeated," said Dr Lawn, "but perhaps most importantly, we need to study in sentinel populations what the impact of the ART rollout is on the actual community burden of TB."

Dr Lawn’s work also made a case for isoniazid preventative therapy (IPT) along with ART in high-burden countries. "Studies need to be done, it's an important intervention that we need to test," he said, but he noted two potential programmatic challenges for the coadministration of IPT and ART: "the interaction with d4T and peripheral neuropathy, and the second is that we have such outstandingly high rates of TB, that you could be giving many patient active TB monotherapy, but those challenges need to be looked at."