Postive Nation: Treatment News Dec 07

Two STEPs Back

For anyone not on the Data and Safety Monitoring Board (DSMB) of the STEP study, it came like a bolt out of the blue. On 21 September, the pharmaceutical company Merck announced that it was discontinuing two international trials of its HIV vaccine, which had seemed the most promising candidate so far.

The two trials, between them involving 4,500 volunteers, weren’t due to end till 2010. The fact that they were stopped so soon reveals the starkness of the news the DSMB (a group of neutral scientists there to ensure the safety and correct conduct of the trial) had to impart: the vaccine was absolutely ineffective. Worse, there was a hint it might even make people more vulnerable to HIV infection.
Richard Jeffreys of New York’s Treatment Action Group said: “You can’t really sugarcoat the fact that the results are deeply disappointing.”

The facts we know so far are these, in advance of a deeper analysis: in the STEP trial, a big international one involving 3,000 volunteers from the Americas and Australia, 24 of the people given at least one dose of the vaccine had caught HIV versus 21 given an inactive placebo. Worse, 19 of those given at least two doses – for full efficacy it was expected that several jabs would be necessary – had become infected but only 11 of those given the placebo. That’s one in 36 versus one in 50 – something that begins to look like a significant difference.

However if wasn’t the non-efficacy in terms of preventing HIV that really disappointed the activists and researchers. Even if the Merck vaccine had worked as hoped, they knew it would probably only offer partial protection, and efficacy in vaccine trials can also be skewed by many unexpected human factors, even in cases where participants don’t know if they’re taking the real thing.
“We hoped we’d get some evidence of efficacy,” says Dr Mark Feinberg, head of policy and affairs for Merck’s vaccines division, “but it was thought unlikely it would have enough to take it into production. But if this study had worked we’d be a lot further along the route to an HIV vaccine.”

Instead, scientists may have to rethink their whole approach to making an HIV vaccine, and even consider whether it’s possible at all. Because the most shocking result was the HIV viral load in the volunteers infected: 40,000 in those receiving the vaccine, 37,000 in those getting the placebo. Essentially the same.

What this told the researchers was that not only did their vaccine not prevent HIV, but that the entire hypothesis upon which scientists have been relying to develop an HIV vaccine since the beginning of the century may be fundamentally wrong. Although it might not translate into enough immunity to prevent an infection, the type of vaccine they were testing had generated measurable anti-HIV immune responses in human beings, and they were expecting the vaccine to at least lower viral load and slow progression to AIDS.

How it was supposed to work
The Merck adenovirus ad5 vaccine was the first of the ‘second generation’ of vaccine candidates to go into a large trial.
A vaccine is, essentially, a ‘fake infection’ that mimics the real thing closely enough to generate an immune response to it. This response is then archived in the ‘memory’ that the immune system develops, so that when the real infection comes along, it’s recognised and dealt with so swiftly that an infection never gets established.

Vaccines usually elicit the first line of defence the body mounts – antibodies, immensely variable protein molecules that can stick to bits of the infecting agent, whether that is a virus, a bacterium or a grain of pollen – antibodies are also responsible for allergies. The ‘bits’ are called antigens.

Some vaccines consist of whole viruses that are killed or weakened, and these tend to be the most effective; but in an exceptionally variable virus like HIV, it’s feared that giving people a whole virus, even if it’s supposed to be inactivated, could cause an infection because the virus might become viable again. So in HIV and in some other vaccines (an example is hepatitis B) specific antigens are used.

The trouble is, as we know, the antibody response the body mounts to HIV doesn’t work. HIV antibodies are what they test for in the standard HIV test, and if they worked, HIV infection would not be life-long. Nonetheless, researchers developed a first generation of HIV vaccines that would, it was hoped, generate a strong enough antibody response to repel an infection.

They didn’t work. The death knell for the first generation of antibody vaccines was struck in 2004 when two big international trials of the AIDSVAX gp120 vaccine showed no evidence of any effect.

Mark Feinberg explains why. “Antibodies only develop to the envelope (the shell) of a virus. In the case of HIV, this is the gp120 protein, but this has been designed by evolution to block an antibody response.” The gp120 proteins are the ‘knobs’ on the surface of the virus that stick to cells and contain the infection machinery.

In HIV, they’re covered by sugar molecules that disguise their shape; they’re immensely variable; and the business end of gp120, the bit that comprises the virus’s infection machinery and which doesn’t vary much, is concealed deeply within the protein’s structure where antibodies can’t sense it. “The active sites aren’t developed till milliseconds before the protein binds,” says Mark Feinberg. Like a master-criminal, HIV only reveals its true nature when it’s too late to stop.

So researchers switched to vaccines designed to generate a response by the second arm of the immune system – the CD8 cell response.

When cells are infected with a virus, they send out distress signals. These consist of bits of the infecting agent – antigens – that are displayed on the surface of the infected cell, like a cross of the door of a plague victim’s house. These are sensed by the cytotoxic T-lymphocytes or CD8 cells, the immune system’s SAS, an elite squad of killers that put the subverted cell out of its misery.

There’s an advantage to generating a CD8 response: the distress antigens can come from any part of the invading virus, including ones buried deep inside that antibodies don’t see. But there’s a major disadvantage too: the response doesn’t prevent an infection but only contains one already established.

To get round this, the Merck scientists did something really clever: they smuggled three HIV antigens into cells by putting them inside the shell of an adenovirus, a germ that at worst causes a mild cold, thereby causing a fake HIV infection that was supposed to fool the body into thinking it was a real one.

Furthermore, it appeared to work. The vaccine elicited CD8 responses in both monkeys and humans – ones that appeared so potent at first sight that the size of the trial was doubled in October 2005 on the back of them.

Why didn’t it work?
This is what alarms researchers. There are three possibilities, Mark Feinberg explains:
•The whole hypothesis could be wrong and the CD8 response is in general incapable of containing HIV. This would bring into question years of research which started when African sex workers who didn’t catch HIV appeared to fight it off, not with antibodies, but with a strong CD8 response.
•CD8 responses are very specific and the ones that volunteers developed might be too narrow to deal with an immensely variable virus.
•The tools we have to measure an immune response actually don’t measure anything that’s useful. The standard assay measures CD8 cells’ production of a virus-killing chemical called interferon-gamma, and it’s been argued for years that this may not indicate a truly effective response.

Whichever is the truth, the results have set back vaccine development by several steps. A large trial called PAVE was about to start which was to use the ad5 vaccine, but also use a ‘priming’ jab first of naked viral antigen which was supposed to generate a bigger CD8 response. This will at least have to be delayed while the STEP results are analysed, as are results from the other trial, a South African trial called Phambili which started in spring this year and had already recruited 800 volunteers.

What next, if anything?
It all depends if the CD8 response that ad5 generated was simply too narrow – in which case it could possibly be improved – or if it didn’t work at all. Either way, the focus of research interest may be swinging back more to antibodies.
There are rare antibodies that do render HIV harmless. We know they exist because there are people who appear immune to HIV and scientists, using gene probes like magnets to pull a needle out of a haystack, have fished out a handful of these so-called ‘broadly neutralising’ super-antibodies out of their blood.
“They don’t look like typical antibodies,” says Mark Feinberg. “They have all kind of weird molecular extensions that enable them to get inside the inaccessible bits of the viral envelope.”
The trouble is, we can find these antibodies and can even clone them in the lab, but finding HIV antigens that will cause the body to make them continuously and in enough quantity to neutralise HIV will be a huge challenge (you can’t just give people the antibodies because, like any other drug, they disappear from the body quickly). We will need to do a lot of basic science before we put on a third-generation equivalent of the STEP trial.

What effect will this have?
The most alarming possibility is that the ad5 vaccine did something that made people more vulnerable to infection. Until more figures are released on 7 November, at the time of writing we don’t know if the difference in infections observed is a statistical fluke or evidence of harm – if, for instance, the infection the adenovirus caused may have put some people’s immune system into an active state where they were more vulnerable to HIV.

What vaccine advocates fear more, however, is the social cost. Will it be difficult to recruit volunteers for future studies?
Mitchell Warren is Executive Director of the AIDS Vaccine Advocacy Coalition – an AIDS activist organisation based in the USA which started precisely because the people involved felt that people with HIV should advocate as hard for vaccine development as they did for AIDS treatment.

Although as disappointed as any advocate at the results of the trial, he points out that at least the trial was well-designed enough to find out that the vaccine wasn’t working in as short a time as possible, thus sparing several years during which people might have thought they were protected. “We got an answer,” he says. “It was one we don’t like, but it was an answer!”
Asked if the vaccine failure has made people reluctant to come forward or made them alarmed about being harmed, he says: “I’ve just been to South Africa to talk about closing the Phambili trial. I was very impressed by the level of community support. In the media in listservs and blogs, yes, there were remarks about ‘another nail in the prevention coffin’, but in people even tangentially involved in the prevention effort – well there are some who question the use of communities at high risk in prevention trials, but those voices are in minority.”

Unlike the STEP trial, the decision was taken to ‘unblind’ the Phambili trial and tell people whether they’d received the vaccine or the placebo. Those who received the vaccine can’t enter another vaccine trial, but, says Warren, “people who’d received the placebo were saying ‘when can we take part in another one?’”

Do the scientists think we’ll ever get an AIDS vaccine? Mark Feinberg pauses. “Well...I hope we will. The intensity and creativity of the science has been unprecedented. I remember back to 1993-4 and I’d be in doubt when people asked me if I thought there’d ever be an effective treatment for AIDS. Two years later we’d learned that if you used the drugs in combination you could do just that.”

He acknowledges, however, that we’ll have to wait even longer. “I think everyone in HIV prevention needs to call on interventions that don’t include a vaccine. We need to roll out antiretroviral drugs as widely as possible, continue the research into microbicides and pre-exposure prophylaxis, and research and develop better behavioural interventions to help people reduce partners and use condoms.
“What the world needs is an effective, comprehensive HIV prevention strategy.”