Map of life expectancy at birth from Global Education Project.

Tuesday, August 22, 2006

A difficult question

By request, here is my take on the ethics of HIV vaccine trials.

The prospect of having a vaccine that would truly prevent HIV infection is truly compelling. It's worth a very large investment and a certain amount of risk. So far, although as I said a couple of days ago we've managed to slow down the spread of HIV, stopping it is nowhere in sight. It's still incurable, extracts an apalling human toll, and is in fact destroying whole societies in Africa. A vaccine eliminated another terrible scourge of humanity, smallpox, and we are tantalyzingly close to eliminating polio (although the conflict in Afghanistan has produced a serious recent setback). Other dread diseases are no longer major problems in the wealthy countries thanks to vaccines, although they continue to take an unconscionable toll in most of the world.

In order to develop a vaccine (or any medical product), you first have to test it in people in what's called a Phase I trial, with a small number of volunteers, to determine that it appears to be safe, at least for most people, in the short term. You may get some useful information about biological response at this stage, but that's not the main objective. In somewhat larger Phase II trials, you begin to get information about dose response and preliminary evidence of effectiveness, while continuing to look for adverse effects. Finally, to get approval, you need to do a large-scale Phase III trial with adequate statistical power to prove that the product is effective for the intended purpose in a defined population, with somewhat more potential to detect adverse effects.

As I have noted many times, however, Phase III trials can easily miss important adverse effects because a) they don't generally have long-term follow-up and b) they aren't necessarily looking for the effects that do occur and may be underpowered to detect less common ones. Also, while the trial population is narrowly chosen, the product is usually prescribed much more widely, so you often get bad news only after widespread commercial use has gone on for a while.

Certain conditions are considered essential to the ethical conduct of trials. The first is that participants be fully informed of the possible risks and harms from the procedure, of the selection criteria, and other conditions of the study; and that their participation be entirely uncoerced and freely chosen. Since there is no basis for expecting that participants in early stage trials will benefit, their participation is presumed to be largely altruistic, although in fact they are paid. This is supposed to compensate them for their time and inconvenience, but of course it is really an incentive. For later stages, there is supposed to be what is called clinical equipoise. That means we don't know whether the treatment is better than the control, it's about 50/50. Otherwise, it would be unethical to give it to some people and not to others. (In the real world, this is often a fiction, but that's another story.)

In the case of HIV vaccine trials, there are additional complications. One is that the test to determine that someone is HIV infected is actually an antibody test. Being "HIV seropositive" means that you have antibodies to HIV in your blood, indicating that you have been exposed to the virus. Since the immune system generally does not succeed in clearing HIV, people who have been exposed are presumed to be infected. (This is not the case with most viruses. Having antibodies to measles means you are immune, not that you are infected.)

But vaccines are intended to provoke an immune response, therefore if you receive most experimental vaccines, you will forever more test positive for HIV, whether or not you become infected. This is obviously a disadvantage, because a) it will be more difficult, and expensive, to find out if you are infected in the future -- you'll need a test for viral RNA; and b) being HIV seropositive can be a disadvantage if you want to emigrate, buy insurance, etc. You may have a lot of explaining to do.

A second complication is that people whose behavior puts them at risk for HIV are sometimes prone to rationalizing why it's really okay. Those are exactly the people you need for a Phase III or even a phase II trial, in which you are trying to prove that the vaccine actually works. In fact, you need for some of them to become infected during the course of the trial -- you need a statistically significant difference in the rate of infection between the group that gets the vaccine and the group that doesn't. So if you pick people at low risk, your trial won't succeed. But if people at high risk think they may be protected, they may be more likely to engage in unsafe behavior, which would make your trial unethical because it is harming the participants.

Ergo, the only way to do this ethically is to provide the trial participants with intensive counseling and other needed supports, such as addiction treatment, to reduce as much as possible the chance that participating in the trial will lead them to behave less safely. This applies to both the control group and the active group because, remember, at this stage we don't actually know that the vaccine is effective. But if our counseling is really effective -- and intensive counseling has been shown to be effective at reducing the risk of infection -- the trial will fail to show that the vaccine works.

While you ponder that dilemma, you should know that currently, while there are numerous HIV vaccine trials going on, these are all early stage trials. You can read about them here, from the U.S. government sponsored organization promoting these trials. None of the products currently in development is thought to have strong potential for ultimate approval -- these are really all experiments to try to learn more about how the immune system reacts to certain stimuli, to gain information that may help in the design of more effective products. (HVTN 502, the most advanced, is a Phase II study.)

And that brings us to the final problem with HIV vaccines, which is that in order to gain approval, they have to be highly effective. They can't just succeed in preventing infection in a percentage of cases, because one of the worst things you could do is give people a false sense of security. Partially effective vaccines may be useful with other pathogens, where people's individual behavior has little relevance to their risk of becoming infected, but with HIV they could not gain approval.

It has proven very difficult to design even a partially effective HIV vaccine for a couple of reasons. One is that the virus mutates quickly, and exists in various strains. A vaccine that protects against one strain may not protect against another, and a vaccine that is effective today may not be effective tomorrow. (This is the same problem we have with flu vaccines, but for the reasons just stated they are still useful, whereas a comparable HIV vaccine would not be.)

A second reason is that in HIV, the surface proteins which normally provoke an immune response against viruses are hidden by a sugar coating. One goal of the current trials is to find a strategy to defeat this defense. But even if they succeed, that isn't going to help with the first problem.

The people working on HIV vaccines believe passionately in what they are doing and they are convinced they will succeed. I'm not an expert on the technicalities, so I won't venture a guess as to the true prospects. But I will say that I am more concerned about the ethical problems with all this than most people seem to be. Indeed, I can't find a good discussion to link to, which is why I gave you my own at such length. If anyone has a resource to recommend, please do.

No comments: