A study in Malawi has found that whilst most of the opportunistic infections prevented by the drug co-trimoxazole are rare in people with HIV, bacterial infections and malarial infection that already have reduced sensitivity to co-trimoxazole are widespread, calling into question the global recommendation that all people with HIV with CD4 cell counts below 500 should receive prophylactic treatment with co-trimoxazole. Its use in settings where resistance is high in the general population has been controversial.
Co-trimoxazole prophylaxis for HIV-infected adults has not been routinely recommended in Malawi because of concerns about the development of co-trimoxazole resistance in a population at high risk for malaria. Strains of malaria in Malawi are still around 80% sensitive to sulfadoxine-pyrimethamine (Fansidar), which remains the recommended first-line treatment for the disease in that country. However, long-term exposure to the sulfone component of co-trimoxazole (which is a combination of trimethoprim and sulfamethoxazole, and sometimes referred to as TS) may induce resistance in malaria parasites, leading to reduced efficacy of Fansidar for any individual taking co-trimoxazole as prophylaxis and a greater risk of the onward transmission of malaria parasites that have some resistance to Fansidar.
The researchers set out to determine whether, in a country where co-trimoxazole was not being used as primary prophylaxis for HIV-positive people, the burden of diseases preventable by prophylaxis was greater than the burden of infections that could become resistant to the sulfone component of co-trimoxazole (thus compromising future treatment).
They carried out the study in a township on the outskirts of Malawi’s second largest city, Blantyre, using a local health centre to recruit individuals who had undergone voluntary counselling and testing for HIV. The study recruited HIV-positive individuals who were not taking prophylaxis, and they attended each month for follow-up visits. In an attempt to minimise loss to follow-up study, the nurses actively traced non-attenders, and patients were only considered lost to follow-up if they failed to attend two consecutive appointments.
The study recruited 660 of the 1267 patients who tested HIV-positive at the health centre between September 2002 and December 2003. During the study one quarter of patients withdrew or were lost to follow-up and there were 52 verified deaths. Follow-up amounted to 4082 person years.
Forty-three per cent had a CD4 cell count below 200 cells/mm3, 24% had symptomatic HIV disease (WHO stages III or IV) but only 5% were receiving antiretroviral therapy at the time of the study due to its cost. (Antiretroviral therapy is now being made available free of charge in Malawi by the government).
The incidence of severe clinical events in the study population was high, at 76 events per 100 person years of follow-up, with the most frequent being malaria (33.8 per 100 years). The efficacy of Fansidar could be assessed in 53 cases of malaria where the treatment course was completed; in 77% of cases the treatment response was adequate.
Bacterial infections were also common, including bacterial pneumonia, dysentry and urinary tract infections. In vitro resistance to cotrimoxazole in organisms isolated from these patients was high.
Opportunistic infections preventable by cotrimoxazole
In contrast there was only one confirmed case of Pneumocystis pneumonia and no cases of toxoplasmosis encephalitis, despite the presence of toxoplasma antibodies in 28% of patients.
The authors conclude that there are significant risks for countries like Malawi in implementing co-trimoxazole prophylaxis on a large scale, given the need to minimise the degree of sulfone resistance in bacterial and malarial infections in the local population. They ask whether alternative antibiotics exist that could offer protection, and suggest that azithromycin might replace co-trimoxazole. However, azithromycin may lead to the development of resistance to erithromycin, one of the key treatments for respiratory infections in Malawi.
However, experience from Uganda and Cote d’Ivoire suggest that co-trimoxazole prophylaxis reduces morbidity in people with HIV despite an increase in bacterial resistance and that much of the benefit derived from co-trimoxazole in these populations is attributable to its effect on bacterial infections rather than PCP.
Furthermore, the authors do not discuss another option - the use of artemisinin-containing combination therapy to treat malaria. Co-artemether has been proven more effective than Fansidar in a setting of high sulfone resistance, but some countries have resisted introducing it until absolutely necessary because it is 10-20 times more expensive than existing malaria treatments.
Van Oosterhout JJG, et al. A community-based study of the incidence of trimethoprim-sulfamethoxazole-preventable infections in Malawian adults living with HIV. JAIDS 39: 626-631, 2005.