Current issues of ACP Journal Club are published in Annals of Internal Medicine


Early initiation of zidovudine is effective in the short term in HIV infection

ACP J Club. 1995 Nov-Dec;123:66. doi:10.7326/ACPJC-1995-123-3-066

Source Citation

Ioannidis JP, Cappelleri JC, Lau J, et al. Early or deferred zidovudine therapy in HIV-infected patients without an AIDS-defining illness. A meta-analysis. Ann Intern Med. 1995 Jun 1;122:856-66.



To evaluate, using meta-analysis, the efficacy of early or deferred zidovudine monotherapy in patients with human immunodeficiency virus (HIV) infection but not the acquired immunodeficiency syndrome (AIDS).

Data Sources

English-language studies were identified using MEDLINE, AIDSLINE, AIDSTRIALS, AIDSDRUGS, and CHEMID. Current Contents, proceedings of the International Conferences on AIDS, and the bibliographies of retrieved articles were also reviewed. Searching was done by June 1994.

Study Selection

Studies were selected if they had a randomized, double-blind, placebo-controlled design; if the patients had asymptomatic HIV infection or symptomatic HIV infection without a clinical diagnosis of AIDS, were aged ≥ 12 years, and had not previously or concurrently received antiretroviral agents other than zidovudine; and if patients with multiple disease events could be separately identified.

Data Extraction

Clinical disease progression, progression to any primary end point, and progression to clinical AIDS or HIV-related death.

Main Results

Data were stratified according to disease stage at study entry (CD4+ count; symptoms) and duration of follow-up (short- term < 14 mo; long-term > 21 mo) and were pooled using the DerSimonian and Laird random-effects model. 9 studies met the selection criteria and, when combined, had > 10 000 person-years of follow-up. 459 patients (14.3%) in the early zidovudine therapy groups compared with 652 patients (20.8%) in the deferred therapy groups progressed to any primary end point {95% CI for the 6.5% absolute risk reduction, 4.6% to 8.3%; P < 0.001; relative risk reduction, 31.9%; CI, 23.2% to 38.2%; number needed to treat, 15; CI, 12 10 22} (numbers calculated from data in article) The benefit was more obvious in the short-term trials. For progression to AIDS or death, the estimated pooled risk ratios were 0.55 (CI, 0.38 to 0.79) for all trials, 0.39 (CI, 0.28 to 0.54) for short-term trials, and 0.88 (CI, 0.73 to 1.08) for long-term trials. With stratification for disease stage, progression to AIDS or death in the short term decreased for both asymptomatic and symptomatic patients with CD4bmc plusp.bmp} cell counts < 500 × 106/L (risk ratios, 0.43; CI 0.30 to 0.64, and 0.26; CI, 0.13 to 0.56, respectively). In the long-term trials, early and deferred zidovudine therapy did not differ for any of the outcomes evaluated.


In patients with HIV infection but not AIDS, the early initiation of zidovudine is effective in the short term for decreasing progression to any of the end points evaluated, including AIDS or death. In the short term, symptomatic patients with a CD4+ count < 500 × 106/L have the greatest benefit.

Sources of funding: National Institutes of Health and the U.S. Public Health Service.

For article reprint: Dr. J. Lau, Division of Clinical Care Research, New England Medical Center, Box 63, 750 Washington Street, Boston, MA 02111, USA. FAX 617-636-8023.


The meta-analysis by Ioannidis and colleagues confirms that the early initiation of zidovudine monotherapy prolongs AIDS-free survival, particularly in symptomatic persons. Zidovudine prolongs life after an AIDS diagnosis, but early compared with deferred zidovudine monotherapy before AIDS does not prolong overall survival, as was seen in the Concorde trial (1). It is reasonable to conclude that zidovudine monotherapy provides a quantum benefit that can be obtained early or late, but not both, in HIV infection.

In the shifting sands of HIV research, well-designed trials that answer well-posed and important questions may become irrelevant even before completion. Most of the trials included in this meta-analysis were begun in the era of a single-drug antiretroviral armamentarium, but their results were often reported in a different environment. Compounds, including didanosine, zalcitabine, stavudine, lamivudine, non-nucleoside reverse transcriptase inhibitors, and protease inhibitors together with their use in increasingly complex combinations have dramatically altered the HIV therapeutic landscape. The question of early rather than deferred zidovudine, posed when no other drug was available, is different than the same question posed today, when other drugs can be substituted, added, or combined.

Given the notion of a quantum benefit in a 1-drug world, it was natural several years ago to be reticent about exhausting that benefit too soon and seemingly leaving nothing in reserve. The clinical choices today are much different, and it is necessary to take into account all of the drugs and combinations available, together with the expectation that our increasing understanding of viral-load monitoring will help the decision making. The question when treating patients with HIV infection has shifted from how best to postpone the inevitable to how best to buy time—not only survival time per se, but time until the sands shift even more, and new, effective alternatives become available.

Julio S. Montaner, MD
Martin T. Schechter, MD, PhD University of British Columbia Vancouver, British Columbia, Canada