Editorials

Management of co-infection with HIV and TB

Improving tuberculosis control programmes and access to highly active antiretroviral treatment is crucial

About a third of the 36 million people living with HIV worldwide are co-infected with mycobacterium tuberculosis; 70% of those co-infected live in sub-Saharan Africa.¹ In developing countries half of people with HIV infection will develop active tuberculosis; in some countries in sub-Saharan Africa more than 70% of patients with active tuberculosis are also HIV seropositive.¹ Tuberculosis is the leading cause of death among people with HIV infection, accounting for a third of deaths due to AIDS world wide.² The introduction of highly active antiretroviral therapy has decreased death and opportunistic infections by 60% to 90% among people living with HIV in affluent countries,³ but in developing countries highly active antiretroviral therapy is available to a tiny minority of those who need it. Today there is a shocking inequality worldwide in the prognosis of HIV and tuberculosis co-infection, and it depends on whether patients or their country have access to highly active antiretroviral therapy.

Many regimens have been proposed for treating latent tuberculosis infection; the preferred option is still isoniazid, recommended for nine months.⁴ In settings with a high prevalence of tuberculosis, preventive therapy against tuberculosis for people living with HIV (not combined with highly active antiretroviral therapy) seems to offer protection against tuberculosis but seems to have no effect on HIV progression or mortality in the long term.⁵ Its implementation on a large scale poses major operational difficulties.⁶

Treatment of active, susceptible tuberculosis with first line drugs is as effective at curing tuberculosis in people infected with HIV as in those not infected; in the absence of highly active antiretroviral therapy, however, mortality under tuberculosis treatment will be higher for people living with HIV, mainly due to other opportunistic infections.⁷ Conflicting reports on increased rates of tuberculosis recurrence after treatment for HIV related tuberculosis⁸ have not yet provided sufficient reason to increase the duration of the treatment for people living with HIV.

Where highly active antiretroviral therapy is available its combination with the treatment of active tuberculosis is difficult for several reasons: overlapping toxicity profiles of some antituberculosis and antiretroviral drugs, drug interactions, and non-adherence to complicated treatment regimens.⁹ An important problem is the possibility of paradoxical reactions. Such reactions include the transient worsening or appearance of new signs, symptoms, or radiographic manifestations of tuberculosis within days to weeks after starting antiretroviral treatment. These reactions may be particularly severe when highly active antiretroviral therapy is started soon after the start of treatment for active tuberculosis. The explanation for these reactions is probably the restoration of the immunity towards mycobacterial antigens. Even in patients with low CD4+ lymphocyte counts, it is recommended to delay highly active antiretroviral therapy until the first two months of treatment for tuberculosis have been completed.⁹

In the absence of highly active antiretroviral therapy, proper case management of tuberculosis can significantly prolong the lives of people with HIV with active tuberculosis. However, only 20%-25% of all patients with tuberculosis world wide have access to effective diagnosis and treatment.¹ In countries with a high prevalence of tuberculosis and HIV where the tuberculosis control programme is insufficient, its strengthening is probably one of the most effective interventions to improve the care of people living with HIV.

Where highly active antiretroviral therapy is not available other interventions have been proposed to improve the care of people co-infected with HIV and tuberculosis. A randomised controlled trial in the Côte d'Ivoire showed a 48% reduction in mortality among HIV infected patients with tuberculosis who received prophylaxis with co-trimoxazole,¹⁰ but these findings might not be reproducible where the spectrum of opportunistic infections and the degree of resistance to co-trimoxazole are different. Induced cross resistance to other drugs, like sulfadoxine and pyrimethamine, an antimalarial treatment, is a major concern.¹¹ Data on the operational feasibility and effectiveness of this intervention on a large scale are urgently needed. A randomised controlled trial in Haiti showed that prophylaxis after treatment for tuberculosis decreased the risk of recurrent tuberculosis in HIV infected individuals but did not prolong survival.¹²

Theoretically, in the absence of highly active antiretroviral therapy, prophylaxis with co-trimoxazole and preventive treatment to protect against a first or a recurrent episode of active tuberculosis could be of some benefit to people co-infected with tuberculosis and HIV. These interventions can be offered only when the HIV status of the patient is known, but most people living with HIV are not aware of their HIV infection (more than 95% in some regions in sub-Saharan Africa).¹ Given the overlapping epidemiology of tuberculosis and HIV, the need for improved collaboration between tuberculosis and HIV programmes is increasingly recognised. The "promote HIV voluntary counselling and testing initiative" (ProTEST) aims to promote voluntary testing for HIV as a key to a more coherent response to tuberculosis in areas with a high prevalence of HIV.¹

The public outcry about the lack of access to highly active antiretroviral therapy in the developing world and the competition between generic manufacturers in Brazil and India have already resulted in drug prices coming down. But prices are still too high. Apart from the cost there are enormous barriers to overcome in terms of improving health services before offering equitable access to highly active antiretroviral therapy in developing countries. Highly active antiretroviral therapy could greatly reduce the suffering of people living with HIV; moreover it could reduce the incidence and death rate of tuberculosis.

R Colebunders, professor, infectious diseases. 

Department of Clinical Sciences, Institute of Tropical Medicine and University Hospital, 2000 Antwerp, Belgium

M L Lambert, research assistant. 

Department of Public Health