High rates of TB after beginning ART in Africa — but decreases observed over time

This article is more than 16 years old. Click here for more recent articles on this topic

Tuberculosis (TB) is relatively common in people with HIV starting antiretroviral therapy (ART) in resource-limited settings, especially during the first several months on treatment, according to studies from Uganda and Senegal presented on Monday at the Sixteenth International AIDS Conference in Toronto. However, the study from Senegal which included patients with several years of follow-up, also noted that the incidence of TB falls substantially after the first year on ART.

Since the mortality associated with such TB cases was high and coadministration of TB therapy and ART complicated, the presenters of both of these studies stressed that TB screening for people about to start ART needs to be improved substantially, that ART treatment should be initiated before CD4 cell counts fall too low, and that there could be a role for a course of isoniazid preventative therapy (IPT) to prevent TB in people about to start ART.


Several recent studies have reported a high incidence of TB in resource-limited settings in people taking ART. For example, in a recent issue of Clinical Infectious Diseases, Dr Stephen Lawn and colleagues explored TB case rates in a peri-urban community in the Western Cape of South Africa. They observed a TB case rate of 22.1 cases per 100 person years in people with HIV during the first three months of ART. This was in contrast to an incidence of 0.6 cases per 100 person years in the HIV-negative population.

And yet, successful ART reduced the risk of TB over time. Lawn et al. also reported a rate of 3.5 cases per 100 person years during the first year on treatment, falling to 1 case per 100 person years after five years on follow-up. Risk factors for the development of TB in persons on ART included a CD4 cell count below 100 cells/mm3, advanced HIV disease (WHO stage 3 or 4), and age below 33.


person years

In a study “100 person years of follow-up” could mean that information was collected on 100 people for one year, or on 50 people for two years each, or on ten people over ten years. In practice, each person’s duration of follow-up is likely to be different.

WHO stage

A simplified system to describe four clinical stages of HIV-related disease, based on clinical parameters (symptoms, weight loss and different opportunistic infections) rather than decreasing CD4 cell count. Stage I is asymptomatic, stage II mild symptoms, stage III advanced symptoms and stage IV severe symptoms (an AIDS diagnosis).


An antibiotic that works by stopping the growth of bacteria. It is used with other medications to treat active tuberculosis (TB) infections, and on its own to prevent active TB in people who may be infected with the bacteria without showing any symptoms (latent TB). 

immune reconstitution inflammatory syndrome (IRIS)

A collection of inflammatory disorders associated with paradoxical worsening (due to the ‘waking’ and improvement of the immune system) of pre-existing infectious processes following the initiation of antiretroviral therapy.


multivariate analysis

An extension of multivariable analysis that is used to model two or more outcomes at the same time.

Now other groups have begun to investigate the incidence of, and risk factors for, TB in patients on ART in different African settings.

TB on ART in Kampala

Dr Andrew Kambugu presented data gathered from patients at the Infectious Disease Institute in Kampala, Uganda. The clinic enrolled 538 antiretroviral-naive patients without signs of TB or any other active opportunistic infection between June 2004 to March 2005. The participants were put on a free efavirenz or nevirapine-based regimen, and assessed with a detailed questionnaire and physical examination at enrolment and at monthly follow-up visits, with CD4 and viral load testing performed at baseline and every twelve weeks. TB was diagnosed by smear microscopy (confirmed) or clinical conditions suggestive of TB with chest x-ray evidence (probable).

At baseline, 368 (68%) of the participants were women and the median age was 37. A total of 26% of the participants had previously been treated for TB (14% were still on TB medications at the time of enrolment). The population was quite advanced: 89.5% had WHO stage 3 or 4 disease, the median CD4 cell count was 92 cells mm3 and the viral load was 270,000 copies/ml.

In the first year of follow-up on ART, 25 of the 535 patients developed TB for a case rate of 5.4 per 100 person years. Ten of these cases occurred within three months of initiating ART, ten more in the next three months (which represents a case rate of 7.8-8.4 per 100 person years for the first six months of ART). However, the study may have reported an even higher case rate if culturing had been used for diagnosis (as in the studies by Lawn et al.).

There were no major differences between those who did or didn’t develop TB in mean age, gender, previous or current TB treatment, or WHO stage disease. However, at baseline, the median CD4 count was lower (53 versus 96) and the viral load slightly higher — though by the time of TB diagnosis, the median CD4 cell count was 148, and the viral load was 1,739 copies/ml (the viral load was undetectable in eleven of the people who developed TB). In a multivariate analysis (which did not consider baseline CD4 counts), the only factor assessed that was significantly predictive of incident TB was an abnormal chest x-ray (IRR 2.70, p = 0.03).

Because the TB cases were diagnosed at higher CD4 cell counts, Dr Kambugu said that it was possible that the TB could have been unmasked by an inflammatory immune response syndrome (IRIS) to ART — especially because there were other accompanying signs of inflammation.

The incidence of TB on long-term ART in Senegal

In the second study, presented by Dr Assane Diouf of Dakar Senegal, the incidence and risk factors for TB were assessed in a cohort of 403 patients enrolled in ANRS 1290 (the Senegalese antiretroviral drug access initiative) between August 1998 and April 2002. As of March 31, 2006, patients had received a median of 52 months of ART; 95 had died and 22 were lost to follow-up in the first three years.

At baseline, the participant’s median age was 37. A toal of 224 (55.5%) had CDC stage C disease, while the median CD4 cell count was 128 cells mm3, and the viral load was over 100,000 copies/ml. More than a fourth of the patients had a previous history of tuberculosis.

Subjects had the typical responses to ART, with significant immunological, virological, clinical and survival benefits. Body mass index ≥ 19 kg m2, haemoglobin levels ≥ 10 g/dL and CD4 cell counts ≥ 200 cells mm3 were significantly associated with better survival.

A total of 47 cases of tuberculosis (after starting ART) were identified retrospectively (by examining the patient clinical records) with an overall incidence rate of 3.1 cases per 100 person years, and 17 of the TB cases resulted in death.

However, after the first year on ART, the incidence rate decreased from 6.4 cases per 100 person years to 1.4 cases per 100 person years for the third year.

According to Dr Diouf, being anaemic or having CD4 cells below 50 cells mm3 at baseline were both significantly associated with developing TB.

Dr Diouf also wondered whether IRIS could have played a role in the cases, and whether isoniazid might reduce the incidence of TB in such patients. However, he stressed that, “these findings underline the necessity of early tracking of HIV infection and ART-related immune responses. It also highlights the necessity for closer collaboration between AIDS and TB programmes.”


Diouf A et al. Tuberculosis incidence and risk factors among adult patients receiving HAART in Senegal: a 7-year cohort study. Sixteenth International AIDS Conference, Toronto, abstract MoAB0105, 2006.

Wandera B et al. Unmasking of mycobacteria tuberculosis in patients initiating free ARVs in an urban HIV clinic in Kampala Uganda. Sixteenth International AIDS Conference, Toronto, abstract MoAB0105, 2006.