AIDS Vaccine Trials in Chimpanzees

See allHide authors and affiliations

Science  18 Dec 1998:
Vol. 282, Issue 5397, pp. 2194
DOI: 10.1126/science.282.5397.2194d

In his excellent and comprehensive review article “Progress in the development of an HIV-1 vaccine” (19 June, p. 1875), Norman L. Letvin refers to the development of a highly virulent strain of human immunodeficiency virus type 1 (HIV-1) that rapidly causes CD4+ lymphocyte loss, AIDS, and death in chimpanzees (1). Letvin suggests that a virus stock derived from this virulent isolate would “provide an important new tool for testing vaccine approaches.”

To many of us who work with chimpanzees, the prospect of causing a rapidly progressive and fatal disease in this near-human species is abhorrent. Until now, HIV vaccine development experiments in chimpanzees have produced valuable information, but no disease. This has provided a justification for doing these experiments in this species.

The purpose of an AIDS vaccine is to prevent infection with HIV and, most important, to prevent chronic infection. Thus, new candidate vaccines should be evaluated for their ability to prevent acute and chronic infection after challenge with a strain of HIV that produces detectable viremia in chimpanzees. Prevention of disease is not relevant. If this occurred in the face of chronic viremia, it would not be considered a satisfactory outcome after challenge of an immununized chimpanzee in a vaccine trial. On the other hand, if chronic viremia is prevented, disease would automatically not occur.

In justification of his position, Letvin cites the fact that many primary HIV-1 isolates replicate poorly in chimpanzees, producing little or no plasma viremia. Protection experiments with such strains could be misleading and difficult to interpret. However, as shown in the table below, there are many HIV strains that have been titrated in chimpanzees and that regularly produce viremia, without producing disease (26). These include primary, or near primary, isolates (36).

We believe that HIV vaccine research should continue to use carefully chosen avirulent HIV strains for challenge. There is no substitute for such experiments in the evaluation of new vaccine strategies. The use of virulent strains is not required and is ethically unacceptable.

View this table:



Vaccine protection against infection with AIDS virus isolates in a number of nonhuman primate species has correlated with the intensity of viral replication during primary infection. Thus, protection against infection with poorly replicating viruses has been relatively easy to obtain with vaccines. However, HIV-1 replicates to very high levels in humans early after infection and, therefore, the most useful animal models for HlV-1 vaccine trials are likely to be those in which similar intense early viral replication is seen. Many primary HIV-1 isolates that establish nonpathogenic infections in chimpanzees do not replicate particularly well in those animals. Studies of these infections are, therefore, not likely to predict vaccine efficacy in humans. For this reason, vaccine trials with an HIV-1 isolate that replicates to high levels and induces AIDS in chimpanzees could prove very useful in assessing vaccine strategies.

A number of the vaccine strategies that have been assessed in animal models are unlikely to elicit sterilizing immunity to a diversity of HIV-1 isolates. However, some of these approaches may decrease the intensity of viremia during primary infection, lower the set-point of viral replication during chronic infection, and, accordingly, prolong survival in individuals who become infected with HIV-1 after vaccination. With lower viral loads, these infected individuals may transmit HIV-1 inefficiently, decreasing the spread of the virus in a population. A vaccine that could accomplish this, while not ideal, would certainly be beneficial in areas of the world with high rates of HIV-1 transmission and limited access to antiretroviral therapies. Preclinical evidence that an HIV-1 vaccine might be efficacious in these ways can best be ascertained in a pathogenic HIV-1/chimpanzee model.

Many investigators, like Prince and Andrus, say that the decision to use chimpanzees in experiments that may lead to their death is not warranted. With 40 million humans already infected with HIV-1 and the prediction of further dramatic spread of this virus in human populations, developing an HIV-1 vaccine is viewed as an absolute priority. However, all possible vaccine strategies for preventing HIV-1 infection cannot be tested for efficacy in human populations. The use of nonhuman primate species for HIV-1 vaccine evaluation will continue to present a difficult ethical dilemma for our entire society as it grapples with the AIDS epidemic.

Navigate This Article