A study in three African countries where women were given an AIDS drug to prevent HIV infection is being stopped because the project is not working, researchers said Monday.
Women in the study had an equal chance of becoming infected whether they were taking a daily dose of the drug Truvada or a look-alike placebo pill. The ones taking Truvada were expected to have had a much lower infection rate.
It wasn’t known why the drug didn’t work. A study last year using it in homosexual men found that it reduced infections by 44 percent. Another study in which a similar drug was used in a vaginal gel reduced the HIV infection rate in women by 39 percent.
“This was a surprising and disappointing finding,” said Timothy Mastro, a physician at FHI, a nonprofit research institute in North Carolina that ran the study. “At this point, we can’t conclude that these antiretrovirals will be effective.”
Truvada is a combination of two antiretroviral compounds, tenofovir and emtricitabine. It is intended to be used after infection in combination with other antiretroviral medicines. Two other studies are underway in sub-Saharan Africa to test the compounds in pill and gel form for HIV prevention in heterosexual men and women.
Finding a cheap and easy way to prevent HIV infection is a high priority in global health, as twice as many people become infected with the virus each year as are being put on life-extending antiretroviral therapy. There is little likelihood that an AIDS vaccine will be available in the next decade. Many experts hope that pre-exposure prophylaxis (PrEP) with AIDS drugs could work as a vaccine equivalent for certain people.
Kevin Fenton, head of AIDS programs at the Centers for Disease Control and Prevention, which has conducted similar studies in the United States and abroad, said the strategy is worth pursuing.
“These findings by no means suggest we should not do further PrEP trials in women,” Fenton said. “This really underscores the complexity of HIV prevention and the importance of doing multiple trials in different populations.”
There are four possible explanations for the failure of the study, which enrolled 1,951 women in Kenya, South Africa and Tanzania starting in June 2009.
One is that the women weren’t taking the medicines as instructed, despite assertions they were. Another is that the pill’s active ingredients didn’t get into cervical and vaginal tissues in sufficient concentrations to have an effect. The third is that the strategy doesn’t work. The final explanation is that it works but by chance didn’t in that experiment.
The study, called FEM-PrEP, was run in places with high rates of HIV infection. Twenty-one percent of the women screened for participation were rejected because they were already infected.
Participants were 18 to 35; 40 percent were married or living with a partner; and on average, they had had 3.7 acts of intercourse in the week before enrollment. All said they did not want to become pregnant, and more than 95 percent were using contraception (mostly injectable forms).
The researchers predicted there would be 72 new HIV infections over the course of the study. When an independent monitoring board reviewed the results partway through, it found there had already been 56 infections. The Truvada and placebo groups each had 28 infections.
A statistical analysis concluded that even if the study had continued to its planned conclusion — patients were still being enrolled — there was no chance it would have shown a benefit of Truvada. In the face of futility, it is considered unethical to continue a clinical study. The monitoring board advised that the study be closed, and FHI agreed.
By the time enrollment was stopped Friday, the average length of time a woman had been in the study was seven months. Overall, the women were becoming infected at the rate of about 5 percent per year. Nine percent were becoming pregnant, with the highest rate among women taking Truvada and oral contraceptives.
It is possible the women who had been randomly assigned to Truvada missed an unusually large number of both birth control and Truvada pills. Blood samples drawn every four weeks will be tested to determine the drug levels of the women randomly assigned to Truvada.
Alternatively, the drugs might be less effective at preventing infection through vaginal intercourse than anal intercourse, the method in the study that showed protection last year, said Mastro of FHI, which was formerly known as Family Health International.
An FHI news released said Truvada was “associated with some known side effects that were not serious.”
The U.S. Agency for International Development and the Bill and Melinda Gates Foundation paid for the study, which was conducted in cooperation with local researchers in the three countries. It cost $26 million.