Intended for healthcare professionals


Managing HIV disease after Delta

BMJ 1996; 312 doi: (Published 02 March 1996) Cite this as: BMJ 1996;312:521
  1. Anthony J Pinching
  1. Professor of immunology St Bartholomew's and the Royal London School of Medicine and Dentistry, London EC1A 7BE

    Questions remain about how to manage patients already on nucleoside monotherapy

    It is not often that the results of clinical trials change clinical practice almost overnight. But that is what has happened with two clinical trials (Delta and ACTG175) of combination nucleoside analogue antiretroviral treatment for patients infected with HIV. The previously standard monotherapy had evident limitations in the extent and durability of response. The need for and expectations of better treatments was high, fuelling but also distorting the investigation of improved approaches. What are the practical implications of these studies?

    Delta (conducted by a European/Australian group coordinated by Britain's Medical Research Council and the Agence National de Recherches sur le SIDA in France) and ACTG175 (conducted in the United States by the AIDS Clinical Trials Group of the National Institutes of Health) have recently released their main findings. They were large randomised double blind placebo controlled trials of zidovudine (AZT) monotherapy versus combinations of zidovudine and didanosine (ddI) or zidovudine and zalcitabine (ddC) with an additional limb of didanosine monotherapy in ACTG175. Both used clinical end points, and ACTG175 also used changes in CD4 cell count. Both recruited patients who had not previously been exposed to nucleosides (Delta 2124 patients, ACTG175 1067 patients) and symptomatic patients who had previously received nucleosides (Delta 1083 patients including patients with AIDS, CD4 count 50-350; ACTG175 1400 patients excluding patients with AIDS, CD4 count 200-500.)

    They showed that, for patients who had never received nucleosides, combinations of zidovudine and didanosine, or zidovudine and zalcitabine, were significantly more effective than zidovudine alone in delaying disease progression. However, patients already exposed to nucleosides gained no benefit from combination therapy (though recently updated data suggest a marginal survival benefit with zidovudine and didanosine). A recent report of an American study (Nucombo), mainly of patients with late disease already exposed to nucleosides, showed similar results. ACTG175 showed some benefit from changing to didanosine alone.

    In the late 1980s several studies showed that zidovudine slowed the progression of HIV disease in symptomatic patients.1 2 Didanosine, zalcitabine, and a fourth nucleoside analogue, stavudine (d4T), seem to have broadly similar effects with a different profile of side effects. They have been used in patients who are intolerant of zidovudine or where zidovudine seems to be losing efficacy after prolonged use.3 Zidovudine also has a direct benefit in patients with HIV encephalopathy and seems to prevent or delay its onset. Nucleoside monotherapy has thus been the standard for symptomatic patients.

    Having shown benefit in symptomatic patients, it was reasonable to surmise that nucleosides might delay progression in asymptomatic patients. The ACTG019 study, a large randomised placebo controlled trial of 1338 asymptomatic patients, seemed to show that zidovudine had such an effect at one year in patients whose CD4 counts were below 500.4 This encouraged early initiation of treatment. However, the longer and more powerful Concorde study, a randomised placebo controlled trial of 1749 eligible asymptomatic patients, showed that immediate treatment with zidovudine did not confer any advantage over treatment deferred until progression of disease or worsening CD4 count.5 Announced in 1994, this led to a period of therapeutic pessimism.

    Evidence, albeit of uncertain clinical importance, of viral resistance to nucleoside analogues has been found in patients treated for long periods. So it was natural to look to combinations of antiviral drugs for a greater and more durable effect. Preliminary studies indicated that these could achieve greater efficacy, as judged by reductions in viral load and increases in CD4 count.

    Delta and ACTG175 are the first completed clinical trials of drug combinations that used clinical end points (disease progression and death). The beneficial effect on disease progression and survival was seen only if combination treatment was used from the outset: for example, in the original Delta annalyses the percentages of patients progressing to AIDS or death were 28.4%, 17.6%, and 23.3% for zidovudine alone, zidovudine and didanosine, and zidovudine and zalcitabine respectively (log rank <0.0001). Most of the patients were symptomatic or had low CD4 counts, but the benefit was seen at all stages. Patients who had been receiving zidovudine alone for 12 months or more did not benefit from the addition of didanosine or zalcitabine (39%, 38%, and 38% respectively).

    These studies have had four major effects on clinical practice. Firstly, it has rapidly become standard for patients starting treatment to take zidovudine combined with didanosine or zalcitabine. For patients who are intolerant of these drugs, other nucleoside combinations may have similar efficacy, though only zidovudine has a clear effect on encephalopathy. Secondly, although the studies did not specifically address the question of when to start treatment, they have caused a “leftward shift,” with more patients starting treatment while asymptomatic but with low CD4 counts or at the start of minor symptoms. This reflects a greater optimism about treatment and an impression from trial subgroup analyses that all patients with CD4 counts below 350, with or without symptoms, will benefit.

    Thirdly, the studies have provided some validation of the virological and immunological markers of HIV disease in predicting clinical benefit: the greater effect on these of combination treatment does translate into greater clinical benefit. Whether this finding can be extrapolated beyond treatment with nucleosides remains unclear, though markers ought theoretically to be valid for drugs acting at any point in the virus life cycle. When combined with nucleosides, both non-nucleoside reverse transcriptase inhibitors (such as nevirapine, loviride, and delavirdine) and protease inhibitors (such as saquinavir, ritonavir, and indinavir), show substantially greater effects on markers than monotherapy. Although data from phase III clinical trials are awaited (a preliminary conference report suggests that ritonavir also affects clinical end points), these drugs are starting to be more widely available.

    Finally, and more problematically, the studies have raised unanswered questions about the future management of the large number of patients who have already received prolonged nucleoside monotherapy. It seems unlikely that other nucleoside combinations (such as zidovudine and lamivudine (3TC)) will be more effective in such patients, as they have a broadly similar effect on disease markers. In principle, patients who have already received nucleoside monotherapy have three options: to continue monotherapy, to change to a combination with drugs acting at different sites (either non-nucleoside reverse transcriptase inhibitors or protease inhibitors), or to use more than two drugs.

    Patients who are doing well with monotherapy may be inclined to stick with it until other options are clarified. Patients who want to change may be so inclined because of new HIV disease events or a fall in CD4 count. Few trial data are available to help them make rational choices, because most large scale trials have focused on patients who have never received anti-retroviral drugs. This is understandable when clear cut answers are sought for scientific purposes or licensing of products, but it has left a large group of patients effectively disenfranchised.

    For patients taking monotherapy who want to add another drug to their regimen, arguably the best option would be to enter a large scale comparative clinical trial of the most promising of these combinations. It remains to be seen whether bodies such as Britain's Medical Research Council or the National Institutes of Health in the United States can achieve timely agreement from all relevant parties on a suitable protocol. In the long run the cost of not doing so will be greater, because of expenditure on inadequately tested regimens.

    Meanwhile, purchasers and providers of health care are assessing the costs of applying the results of Delta and ACTG175, which are likely to increase the costs of HIV drugs by 50-75%. While there is much talk of evidence based medicine in health commissioning, the only comment so far in Britain has been that the cost of combination treatment would have to be found from within existing allocations. Yet indications are that these allocations are to be reduced substantially, despite the still increasing caseload (albeit less than the more pessimistic projections). Robust discussions can be expected on this in the coming months. A research strategy should be developed which will ensure well informed clinical decision making and cost effective prescribing.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    View Abstract