Press room ... 2003 archive

Preliminary Results from First Large Study
of HIV Preventive Vaccine Shows No Evidence of Efficacy; Project Inform Finds Some Claims Misleading

February 24, 2003

San Francisco, CA—Preliminary overall results from the first large study to assess the effectiveness of an HIV preventive vaccine were released last night. The data revealed that the vaccine, called AIDSVax, was not effective in preventing HIV infection compared to placebo. This finding was a major disappointment, as most sources expected to see at least some low level efficacy. There was virtually no difference in infection rates between the people who received the vaccine and those who received a placebo.

The company developing the HIV vaccine product, VaxGen Corporation, curiously announced a subgroup analysis suggesting that while the vaccine was not effective in the 5,000-plus people studied, it may be effective in certain carefully selected racial subgroups, particularly Asians, Black and other non-Hispanic minorities. Many community activists are highly critical of the company's claims that the vaccine may work in racial minorities, primarily because of the small numbers of minorities who became infected in the course of the study.

"The company is claiming that this vaccine works better in African Americans and other non-Hispanic racial subgroups based on the difference of just a few people between the vaccinated and unvaccinated groups." noted Martin Delaney, Project Inform's Founding Director. "This is at best premature and irresponsible data reporting. It is highly misleading and disingenuous to communities who have a stake in these findings. It would do a great deal of harm to stir up hopes for selected groups over a vaccine that has been proven ineffective overall."

A total of 498 non-Hispanic, non-white volunteers participated in the study, with two thirds receiving the vaccine and one third receiving a placebo. Among this group, a total of 29 people became infected with HIV, seventeen of whom received the vaccine and twelve of whom received the placebo. When looking only at the 314 Black volunteers there were a total of thirteen cases of HIV infection, nine among vaccine recipients and four among placebo recipients. While twice as many people received vaccine compared to placebo, these numbers are far too small to draw conclusions about the effectiveness of a vaccine among racial minorities. Suspiciously missing from the company press release were actual numbers of infection rates which were the basis of these claims. When Project Inform finally saw the infection rate data, it was clear that there were a number of other possible explanations for the difference between vaccinated and unvaccinated people in the subgroups. For example, the infection rate in non-whites who received the placebo was abnormally high (9.9%, compared to 5.8% among placebo recipients overall compared to the 5.4% in the white and Hispanic groups who received the placebo). This fluke in the infection rate alone could account for the supposed reduction in infections the company attributes to the vaccine. Moreover, while the company emphasizes in it statements to the press that the subgroup analysis was part of the statistical plan, it fails to point out that the study was neither designed nor powered to detect differences in the effectiveness of the vaccine between the subgroups.

"One of the worst possible outcomes of the press statements is that communities may galvanize to push for expedited studies of the failed vaccine products in racial minorities," notes Brenda Lein, Director of Project Inform's Information and Advocacy Department. "A thoughtful evaluation of possible confounders is critical to determine if the observations hold their weight when put under a microscope and truly examined. Out of respect for African Americans and other racial minorities in the United States, this type of examination should have been conducted before these data were reported."

Delaney further observes, "There is no great merit to large scale testing of a failed product in African Americans. Racial minorities, and especially African Americans, have too often been victimized by substandard ethics in biomedical research. A great deal of time, money, and human effort have already been expended in the effort to find value in this vaccine, often contrary to the recommendations of government advisory panels and the scientific community. In the year 2003, the needs of a company to find a ray of hope in its product must not supercede the need for high quality scientific standards. African Americans and other minorities deserve better than this."

Finally, further evaluation and examination of the findings is critical to understanding what, if any, meaningful differences are observed between subgroups. In the meantime the company's claims have confused and mislead many people about the value of the observations presented. The only conclusion that can concretely be drawn from the trial based on the data released to date is that the vaccine doesn't work. Every other conclusion, at this point, remains unexamined speculation. The power (or p-value) of the calculations cannot and does not supercede the extremely wide confidence intervals of the observations in non-Hispanic racial minorities.

"Essentially this is statistical jargon," remarked Lein, "that ultimately means we're very certain that we're really not sure."

For more information, see AIDSVax Fails to Prevent HIV Infection.