To assess the effects of zidovudine, didanosine, and zalcitabine on HIV disease progression and survival.

The authors searched the MEDLINE electronic database using a Cochrane Collaboration search strategy (dates and search terms not reported). The authors also searched for conference abstracts and contacted other investigators and pharmacological companies for additional relevant studies.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) with any of the relevant treatment arms and which systematically recorded deaths and development of new AIDS events. In the comparison if immediate versus deferred zidovudine, median study follow-up was 50 months. In the comparison of zidovudine plus didanosine or zalcitabine versus zidovudine alone, median study follow-up was 29 months.

Specific interventions included in the review

Zidovudine (immediate versus deferred, 500 to 1500 mg per day), or zidovudine (600 mg per day) plus didanosine (200 or 400 mg per day) or zalcitabine (2.25 mg per day) versus zidovudine alone.

Participants included in the review

Adult participants with HIV infection but not AIDS between immediate and deferred treatment with zidovudine.

Outcomes assessed in the review

Time to death and to disease progression.

How were decisions on the relevance of primary studies made?

The authors do not state how the papers were selected for the review, or how many of the authors performed the selection.

No formal assessment of quality was undertaken.

The authors do not state who, or how many of the reviewers, performed the data extraction. However, information found was checked at the MRC Clinical Trials Unit for consistency internally and with published reports; any inconsistencies were resolved with the trialists.

Data were extracted for each participant on: baseline age, sex, likely routes of infection (risk factors), disease stage, previous antiretroviral therapy, CD-4 cell count, and p24 antigen status (first meta-analysis only), together with date of randomisation, allocated treatment, date of starting and stopping allocated treatment, use of other antiretroviral therapy or prophylaxis against Pneumocystis carinii pneumonia, and dates of new AIDS-defining events, death, last assessment for disease stage, and last known vital status.

How were the studies combined?

The observed (O) and log-rank expected (E) numbers of events and the variance (V) of O-E were summed over trials to give a summary rate ratio and statistical significance test by the one-step formula (see Other Publications of Related Interest). For sub-group analyses, O-E and V were calculated separately for each subgroup within each trial and then summed over trials.

How were differences between studies investigated?

Comparisons between different trials or subgroups were made using standard chi-square heterogeneity or trend tests.

Fifteen randomised controlled trials (RCTs) were included in the review with a total of 15,422 participants. Nine RCTs compared immediate versus deferred zidovudine with 7,722 participants, and 6 trials compared zidovudine plus didanosine, zidovudine plus zalcitabine, or zidovudine alone with 7,700 participants.

In the comparison of immediate versus deferred zidovudine, 1,908 individuals progressed, of whom 1,351 died. In the deferred group, 61% started antiviral therapy. During the first year of follow-up, immediate zidovudine halved the rate of disease progression (p < 0.0001), increasing the probability of AIDS-free survival at 1 year from 96% to 98%. The early delay did not persist. After 6 years, AIDS-free survival was 54% in both groups. At not time was there any difference in overall survival, which at 6 years was 64% with immediate and 65% with deferred zidovudine (rate ratio 1.04, 95% CI: 0.94, 1.15).

In the comparison of zidovudine plus didanosine or zalcitabine versus zidovudine alone, 2,904 individuals progressed, of whom 1,850 died. The addition of didanosine to zidovudine delayed both progression (rate ratio 0.74, 95% CI: 0.67, 0.82, p < 0.0001) and death (0.72, 95% CI: 0.64, 0.82, p < 0.0001).

The addition of zalcitabine to zidovudine also delayed progression (0.86, 95% CI: 0.78, 0.94, p = 0.001) and death (0.87, 95% CI: 0.77, 0.98, p = 0.02).

After 3 years, the estimated percentages alive and without a new AIDS event were 53% for zidovudine plus didanosine, 49% for zidovudine plus zalcitabine, and 44% for zidovudine alone; the percentages alive were 68%, 63% and 59% respectively.

Five of the six trials involved randomised comparisons of zidovusine plus zalcitabine: in these, the zidovudine plus didanosine regime had greater effects on disease progression (p = 0.004) and death (p = 0.009).

Although immediate use of zidovudine halved disease progression during the first year, this effect was not sustained, and there was no improvement in survival in the short or long term. However, the use of didanosine and, to a lesser extent, zalcitabine delayed both disease progression and death; at least when added to zidovudine. The comparative effects of these different nucleoside analogues on long-term survival should inform the choice of which to combine with other types of drug, such as protease inhibitors.

The authors have clearly stated their research question and some inclusion and exclusion criteria. The literature search is poor. The authors do not state search dates or terms used or any language restrictions, so it is possible that the authors may have missed additional relevant studies by restricting the search to only one database.

The quality of the included studies was not formally assessed and the authors have not reported on how the articles were selected, or how many of the reviewers were involved in the data selection and extraction, or on the specific design of the included studies.

The data extraction is reported in tables and text and the statistical pooling was appropriate. There were tests for heterogeneity and the authors have discussed several methodological and data limitations in the review.

The authors' conclusions appear to follow from the results but should be viewed with caution because of the stated methodological limitations of the review.

Practice: The authors state that the evidence from this analysis of the beneficial effect on 3-year survival of particular nucleoside-analogue regimens should inform the choice of which to combine with other classes of drug.

Research: The authors did not state any implications for further research.

UK Medical Research Council; US National Institutes of Health.

1. Commentary on this review. Lancet, 1999;353:1989. 2. Early Breast Cancer Trialists' Collaborative Group. Treatment of early breast cancer: worldwide evidence 1985-1990. Oxford: Oxford University Press; 1990.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.