The AIDS Vaccine That Flopped

In September the AIDS community took a devastating blow when Merck halted an international human trial of the world's most promising potential vaccine against HIV. At the time, doctors revealed only that the vaccine—which uses an "adenovirus," or the common cold virus, to deliver HIV genes into the body—had not worked.

On Wednesday they began to explain why, releasing the first data from the trial, which was run by Merck, the HIV Vaccine Trial Network and the National Institute of Allergy and Infectious Diseases. The researchers announced a troubling finding: the number of HIV infections among the trial volunteers who received the vaccine was higher than the rate among those who received a placebo. Twenty-four of 741 volunteers who got the vaccine became infected with HIV, as did 21 in the placebo group. (The difference may seem small, but scientists were concerned enough to stop the trial because of it.) The volunteers, 3,000 in all, were from high-risk groups, such as sex workers. NEWSWEEK's Mary Carmichael spoke with Dr. Keith Gottesdiener, Merck's vice president for vaccine and infectious disease clinical research. Excerpts:

NEWSWEEK: You hoped that the vaccine would either prevent infection in healthy people or lower the amount of virus in the blood of people who were already infected. Did it do either of those things?
Dr. Keith Gottesdiener: It didn't have an effect on either. That certainly was a surprise to me, and quite disappointing. There's no perfect animal model for HIV trials, but the monkey studies suggested there might very well be an impact on viral load. That was the goal we were more hopeful about. But HIV is such a hard virus to control. More disturbingly, among the volunteers who had some natural immunity to the adenovirus, people who got the vaccine instead of a placebo actually seemed more likely to get HIV. There certainly were more infections in people who got the vaccine, and in particular some of the subgroups. We don't know if the vaccine itself increased the susceptibility, but the possibility that this happened entirely by chance seems on the small side. Maybe there were things that were different about the two groups—for example, we're exploring whether there were differences in risk-taking practices. But we haven't seen much to suggest that that is the case.

How could the vaccine have caused people to be more vulnerable to HIV infection?
At this time, we don't know whether it did, and if it did, how it did so. From a scientific point of view, there is one hypothesis that we can begin to test. This vaccine was designed to create an immune response that would produce high levels of a cell called CD8, or "the killer T cell." CD8 cells can recognize and kill the cells that HIV attacks, which are CD4 cells. The goal of the vaccine was to have high enough levels of CD8 cells to kill HIV-infected CD4 cells. In this way, the vaccine might help prevent HIV infection or help infected people better control the virus. However, in some volunteers, the vaccine may have resulted in greater activation of CD4 cells that recognized the common-cold virus [which was used to deliver the HIV genes]. In this way, more cells highly susceptible to HIV infection might be present in people who received the vaccine, which might make it easier for HIV to establish an infection in their bodies. This is a hypothesis we are still studying.

Because it gives the virus more cells to attack? Are you testing that theory now?
Yes. The idea is to look to see if the volunteers who received the vaccine had increased numbers of activated CD4 cells. We have samples from them, and we're measuring the levels. That wasn't what we were originally planning to do. But our original questions have changed now.

The volunteers did not get HIV from the vaccine, right?
Right. This vaccine cannot transmit HIV. It does not contain the virus. The adenovirus has only three genes from HIV, not the whole HIV genome.

Most of the new infections occurred in men. Does anyone know why?
About 40 percent of the people we recruited for the study were women. But for whatever reason, only one of the new cases of HIV infection occurred in a woman. We're still uncertain as to what that means.

What is being done for the trial volunteers who haven't been infected with HIV but now may be more susceptible to infection?
There's extensive counseling, and that counseling is effective over time. Risk practices went down in the course of the trial. When we first realized the vaccine wasn't working, we instituted some additional counseling efforts. If someone thought they had gotten the vaccine and they were going to be protected, we wanted them to know as quickly possible that this wasn't the case. Now that's even more important.

What does this trial mean for other efforts to use a similar vaccine against HIV? And what does it mean for Merck—will you be pursuing another HIV vaccine strategy?
Those kind of questions—many questions like that—are things that the whole field is going to have to grapple with going forward. To be honest, there are still a lot of data coming out. On the drug side we're extremely active, but as for vaccines, despite 20 years of research, at the moment we don't have any other candidates in development. Our primary focus right now is conducting the extensive analyses for the results from this trial. We really need to understand what happened here to inform our approach and help the overall field in the search for an effective vaccine.