Guidelines for managing HIV infection The goal is maximal suppression of HIV replication for as long as possible T he medical management of HIV infection and AIDS has finally been rewarded with some suc- cess, witnessed by falls in AIDS associated deaths in industrialised countries, fewer opportunistic infections, and fewer admissions for HIV infection. 12 What constitutes optimal practice, however, remains a topic of debate. Two widely publicised sets of guidelines on antiretroviral treatment have appeared over the past year, one from the International AIDS Society in the United States, 3 and the other from the British HIV Association. 4 While their emphasis was different, the recommendations were broadly similar. The guidelines agreed on the need for regular monitoring of HIV-1 plasma RNA (“viral load”) and for using antiretroviral treatment in people with low CD4 lymphocyte counts or symptomatic disease. The British guidelines were less prescriptive and placed less emphasis on introducing treatment for people with CD4 cell counts > 300 × 10 6 /l and high viral load. 4 Both guidelines endorsed combinations of nucleoside analogues such as zidovudine and didanosine (AZT/ ddI), zidovudine and zalcitabine (AZT/ddC), or zidovu- dine and lamivudine (AZT/3TC) as first line treat- ments and discussed indications and options for switching treatments. 34 Practice in many quarters, however, has moved beyond these guidelines. 56 Many doctors now use triple therapy including a protease inhibitor as first line treatment for all HIV infected patients, and some start treatment earlier. Logic and current concepts of pathogenesis supporting such an approach are that virus replication is intense from initial infection, 78 viral load is the most important prognostic marker of risk of progression, 9 10 drug resistant strains emerge in the face of incomplete suppression of replication, 11 treatment with two nucleoside analogues does not provide long term suppression in most cases, 12-15 and favourable clinical outcome is most likely if virus repli- cation is maximally suppressed before the immune system is irreversibly damaged. Several arguments have been advanced in favour of delaying treatment or using only two nucleoside analogues initially. These are that clinical evidence for a more aggressive approach is lacking, long term side effects of these drugs are unknown, early use of the most potent combinations limits later therapeutic options, asymptomatic patients may be turned into pill-taking invalids, long term adherence to treatment by asymptomatic people is unlikely, and, inevitably, the greater cost of triple therapy. This debate has now been joined by heavy hitters from the United States in the guise of two panels that have just published draft reports for public comment. 16 17 A panel convened by the National Institutes of Health has defined the scientific principles that should guide treatment and its monitoring in clinical practice. The second panel, sponsored by the Department of Health and Human Services and the Henry J Kaiser Family Foundation, developed recommendations based on these principles for the clinical use of antiretroviral treatment. Together, the documents clearly state that tri- ple therapy (generally consisting of two nucleoside ana- logues and a protease inhibitor) should be the standard treatment for any person with HIV infection, including as initial therapy. Treatment is recommended essentially for all people with CD4 cell counts < 500 × 10 6 /l, as well as for asymptomatic people with counts > 500 × 10 6 /l and active viral replication ( > 10 000 viral copies/ml). It is stated, however, that many experts would advise treat- ment for HIV infected people with CD4 cell counts > 500 × 10 6 /l and any detectable level of virus. The guidelines emphasise that the goal of treatment should be maximal suppression of HIV rep- lication for as long as possible. For people undergoing treatment for primary HIV infection, the recommendation is for indefinite treatment with triple therapy. After a 30 day period for public comment, the documents will be published in their final form in the Morbidity and Mortality Weekly Report, and they will be updated when necessary. The debate raises many issues. HIV infection illus- trates the limits of the paradigm of practice based exclusively on controlled trials using clinical endpoints. Some guidelines usefully grade the quality of evidence on which recommendations are based and acknowl- edge that some depend on biological plausibility rather than clinical results. 4 18 Even if evidence of more favourable clinical outcome with triple therapy compared with two nucleoside analogues is only just emerging, 19 analogy with diseases such as tuberculosis or lymphoma, plus subjective assessments (“What would you take if ...?”), will persuade many that triple therapy should now be the standard of care for HIV infection. It is likely that international prescribing practices will ultimately vary more over when to start treatment than what to start with. Nevertheless, practice based on data using surrogate markers (viral load and CD4 cell counts) requires careful follow up to ensure that clinical response and survival continue to correlate with changes in markers. 20 The initial promise Saturday 5 July 1997 B MJ 1 BMJ VOLUME 315 5 JULY 1997