Microbicide gel fails to work in large international trial

Gus Cairns
Published: 25 November 2011

HIV prevention researchers and campaigners reacted with disappointment today when a trial of a tenofovir-containing microbicide gel was stopped because it had not prevented any more HIV infections than a placebo gel.

The VOICE (Vaginal and Oral Interventions to Control the Epidemic) trial is a large HIV prevention trial involving 5,029 women in South Africa, Zimbabwe and Uganda, run by the Microbicide Trials Network (MTN), a US-based research consortium funded by the US National Institute of Allergies and Infectious Diseases.

VOICE has been investigating two different HIV prevention methods, oral pre-exposure prophylaxis (PrEP) and a vaginal microbicide gel, with both pills and gel used daily regardless of sexual activity. Oral PrEP was given in two different drug formulations: tenofovir (Viread) tablets or tenofovir + emtricitabine (FTC) tablets (Truvada). In all, there were five arms of just over 1,000 women each: a daily tenofovir, Truvada or placebo pill, or daily vaginal application of tenofovir or placebo gel.

The tenofovir-tablet arm was called off in September when it failed to demonstrate efficacy. More recently, on 17 November, a pre-scheduled meeting of the study’s Data and Safety Monitoring Board (DSMB) established that HIV incidence was almost identical in women using tenofovir gel and ones using placebo gel. There were six infections per 100 women a year in women using tenofovir and 6.1 a year in women using placebo. (A trial's DSMB is a group of independent experts who are the only people to see the unblinded data and know which subjects are taking placebo and which drug).

As a result both the tenofovir-gel and placebo-gel arms have now been stopped, leaving only the Truvada and placebo-pill arms continuing. Unless these too are stopped early, final data collection will occur mid-2012 and the VOICE trial’s full results should be ready by the beginning of 2013.

A significant setback

While the failure of the tenofovir-pill arm in VOICE was disappointing, the failure of the tenofovir gel is considerably more significant and challenging. Fifteen months ago the leaders of the CAPRISA 004 trial in South Africa were given three standing ovations at the Vienna World AIDS Conference when their study found that a tenofovir-gel vaginal microbicide used before and after sex prevented four out of ten HIV infections. At the time it was hailed as the vindication of a prevention concept first dreamed up 20 years ago.

Scientific studies do not just report a single result but also what is called a 95% confidence interval (CI). This reflects the range of uncertainty in a trial’s result, given that chance can always sway an outcome. While headline efficacy in CAPRISA 004 was 39%, the CI was six per cent to 60%, meaning that it was 95% likely that the ‘true’ efficacy of the tenofovir gel lay between these two extremes. If the true efficacy was lower than the observed efficacy, then results from CAPRISA 004 and VOICE are not totally incompatible. 

Although its statistical power - its ability to produce a reliable result - was as high as the VOICE trial, CAPRISA 004 was less than half the size of  VOICE and was within a single country;  it was not considered sufficient evidence by the the US Food and Drug Administration (FDA) to support immediate licensing of an anti-HIV microbicide. However the FDA indicated a year ago that if the VOICE trial were also to produce a positive result, it would fast-track tenofovir gel for approval, meaning we would have had an anti-HIV microbicide by 2014.

But VOICE has now produced a negative result and the licensing of a vaginal microbicide has become a considerably more distant prospect, as approval may now depend on more than one future positive result.

Future plans

Other trials are nonetheless proceeding, as researchers don’t yet know why no efficacy was found in VOICE and no safety issues were raised by the gel. The FACTS 001 trial is a South African study of tenofovir microbicide gel using the same before-and-after-sex regimen as CAPRISA 004. In has just started to enrol a planned 2,200 women at nine sites. In the middle of next year, the ASPIRE trial, a long-awaited efficacy trial of a vaginal ring impregnated with another anti-HIV drug, dapivirine (TMC120), will start: it is hoped that the ring, which will sit on the cervix and be renewed monthly, will overcome the adherence problems that have been the bane of HIV prevention trials using antiretroviral drugs. And of course the Truvada term will continue in VOICE and may represent, as MTN commented, “the one trial that can determine whether Truvada holds promise in preventing HIV.”

MTN has also just concluded a phase I safety trial of a rectal microbicide for use in anal sex and depending on evaluation of the results may move forward to a phase II dosing trial next year. The decision may be informed by VOICE but as rectal anatomy and the trial population are both very different, there is no reason to believe rectal microbicides cannot work if vaginal ones don’t.   

Reactions

HIV prevention advocates expressed considerable disappointment at the closure of the gel arm in VOICE.

“This is a blow to the HIV prevention field,” commented Mitchell Warren, Executive Director of the AIDS Vaccine Advocacy Coalition (AVAC).

“But it is not the definitive answer about whether 1% tenofovir gel is an effective HIV prevention product for women,” he added.

Yasmin Halima, Director of the Global Campaign for Microbicides, commented: “This is disappointing news... It’s not the answer we had hoped for, but it’s important to do these larger studies to better understand what may or may not work.”

Researchers emphasised that the failure of one method in one trial did not mean that the trial, or the idea of microbicides, had failed.

Zeda Rosenberg, Director of the International Partnership for Microbicides, which developed the vaginal ring and is collaborating with MTN in the ASPIRE trial, commented: “MTN will review all the data it gathered to help understand why tenofovir gel and oral tenofovir did not prevent HIV infection in the study. This research will provide invaluable information as the field continues its work to develop safe and effective HIV prevention tools for women.”