Or, What Several Failed HIV Vaccine Studies Can Teach Us about the Risks of the Planned Phase III Ebola Vaccine Study.
 |
| HIV infecting a human T cell. From NIH via Wikimedia commons |
Last week I wanted to start an informed discussion about the complications involved in the plan to begin phase 3 trials of at least two Ebola vaccines in response to the current pandemic in west Africa. Particularly regarding the study design, should a control group (who receives inactive placebo shots) be included, or not? As part of my attempt to explain the considerations I mentioned the story of the failed HIV vaccine trials, but I did not go into any detail, as the post was about Ebola and not HIV. I have decided that was a mistake. So today I will tell you about the HIV vaccine trials, as understanding what went wrong has the potential to teach us a lot about how to go forward with both new HIV vaccines and the Ebola vaccine trials scheduled to start early 2015. The news article we are starting with comes from the news desk of the scientific journal Nature, as Nature is edited by scientists for scientists I'm confident in the quality of it's science journalism, this post will focus on breaking down the story they've told.
The article starts by talking about a presentation given at AIDS Vaccine 2013 conference in Barcelona, Spain. Unfortunately, only those who attended the conference will ever know exactly what was said (this is intentional, as conferences are places where scientists often present unfinished research to each other to solicit feedback, and they don't want unfinished findings being picked up by the media), but based on what Reardon reports, as of October 2013, a fourth HIV vaccine has failed to protect the study participants resulting in the cancellation of it's phase III trial.
To better understand what this means, I want to explain how these trials were done and what it means to say they were canceled (as, obviously, no one can un-do giving someone a shot). These trials were all double-blind randomized control trials (RCT). A group of individuals at risk for infection was recruited, and all individuals were given a series of shots. Neither the scientist nor the study participants knew if they were given the vaccine being tested or just placebos. Only a computer knew, and it wasn't going to tell anyone. Not yet. After administering the shots, the scientists kept up with the study participants and had them come into the lab regularly. At these visits the participants were asked about any side effects they were experiencing and had blood taken. The blood was taken to look for 1) an immunological response to the vaccine and 2) to test for HIV. Canceling the study involves having the computer tell everyone who had the real vaccine and who had the placebo. Follow up visits to the lab may have continued, but the study was no longer "blind" as everyone knew who got what type of shot.
Why would a study be canceled? The simplest explanation is that, even without knowing who was in the test or control group, the scientists could tell that something was going very wrong. That something could be a dangerous side-effect. Let's imagine a study of a new drug in a population that has a 5% chance of dying of a heart-attack over the 5 years of the study. So the double-blind RCT begins and after just one year 10% of the participants have died of a heart attack! That's 10-fold more than you expect! (5% over 5 years is about 1% per year). You know the control group should have the same risk as the general population, so you know that your test group (those receiving the drug) must have a 20-fold higher risk of dying! The only ethical thing to do is to figure out who was getting the drug, tell them to stop taking it immediately and go see a cardiologist! Now, this is a vast VAST exaggeration for the purposes of explanation. No drug that dangerous would ever make it to phase 3 trials, a 20-fold risk ratio would easily be detected early in a phase 1 trial (if it even got through animal studies). The other reason a study might be canceled is if the drug appeared to not be working, or worse appeared to be making people sicker! This is what happened with the HIV vaccine trials.
Back to the specifics of these HIV vaccine trials, of the four trials three (all done in the US) were canceled because they appeared to provide no protection while the fourth, done in Phambili, South Africa, appeared to increase the HIV infection risk. A follow up meta-study of the results of the Phambili and STEP studies (STEP and Phambili tested the same vaccine) together found that when all other factors were accounted for, the vaccine left people 41% more likely to contract HIV than an unvaccinated person.
Several other scientists are skeptical of these findings. Some think that the increased infection risk is due, not to the mechanism of the vaccine itself, but to an increase in risky behavior by the study participants. Basically, the people in the study might think that, because they got the shots, they are safe from HIV and were thus more likely to disregard safer sex practices or share needles. The studies attempted to control for this by asking people about their sexual activity and drug use, but it is possible that study participants lied (particularly about embarrassing things, like forgetting to wear a condom).
I think that it doesn't matter if the explanation for the failure of HIV vaccines thus far is biological or psychological; there are important lessons either way. First, the psychological. If people receive an "Ebola shot" they may think they are protected. This may lead to riskier behavior, such as being less careful about personal protective equipment (PPE), like masks and gloves, when caring for Ebola stricken patients, family members, or friends. If the vaccine doesn't work this could cause to pandemic to worsen! If the study doesn't include a control group (placebo shot) it will be difficult for scientists to tell if the increased Ebola infection is due to the vaccine making things worse or making people careless or things just naturally getting worse. But if we include a control group, people without the potential protection of the vaccine may also engage in careless behavior, also spreading the virus. Then again, if people know they may have gotten a placebo vaccine, they might be less likely to be careless with PPE.
Then there is the biological explanation: simply, the vaccine could make things worse! Since we don't know how likely that is (as this is going to be the first human study with an actual Ebola virus challenge), is it really more ethical to give the vaccine to everyone? I think this is an important consideration since one of the primary arguments given by Doctors without Boarders is that leaving some people intentionally un-immunized for research purposes is unethical. But what if those people fare better? They might.
How likely is it that the Ebola vaccine will "backfire" for biological reasons the way the HIV vaccines may have and leave people more susceptible? No one knows, but I can share some information about the matter. Firstly, the vaccine designs are quite similar. I discussed last week how adenoviruses are being genetically modified to carry genes encoding the coat proteins of dangerous viruses to stimulate immunity. Both the Ebola vaccines and the HIV vaccines used this method. In theory, it's a great method to give the body exposure to proteins from dangerous viruses with zero risk of becoming infected with the dangerous virus (a killed virus vaccine, such as the flu shot, might be too dangerous to make or give as it would have to involve real Ebola or HIV viruses). One known issue with this technology is that if the vaccine is given to someone who has previously been infected with the same adenovirus as was used to make the vaccine their immune system can wipe out the virus too quickly, before the virus has the chance to infect enough cells and produce enough proteins to stimulate immunity to the dangerous virus. But this results in no protection, not increased susceptibility. Also, rare adenoviruses can be used to minimize this problem. I personally wonder if future studies will show that this type of vaccine is more useful for small children (less likely to be exposed to adenoviruses) and that another vaccine strategy will be needed for adults.
While the vaccines for Ebola and HIV share a similar design, there is a potentially important difference in infection that could have important effects on vaccine response. HIV actually infects the cells of the immune system (white blood cells). Specifically CD4+ T-cells, macrophages, and dendridic cells. All three of these types of white blood cells play important parts in adaptive immunity, the very cells whose activity is stimulated by vaccination. It is possible that stimulation of the immune system (such as by vaccination) actually makes it easier for HIV to find white blood cells to attack, thus leaving a person more susceptible for a period of time after vaccination (against HIV or maybe anything) or illness. This is, in fact, one of the possible explanations for vaccination failure put forth by the researchers involved in the STEP study. Ebola is somewhat different; it primarily infects endothelial cells lining the inside of blood vessels, liver cells, and some types of white blood cells (mainly macrophages, monocytes, and dedridic cells). White blood cells are generally attacked during later stages of infection, and their attack contributes to the high death rate of the disease. So an important question is, will stimulation of the immune system by vaccination, make it easier for Ebola to infect white blood cells, thus worsening the prognosis for vaccinated individuals, or will vaccination help the immune system clear Ebola before it has a chance to attack the blood cells, thus improving prognosis. Unfortunately, a phase 3 clinical trial in an area where study participants are likely to be exposed to the Ebola virus is the only way to find out.
So how about now? Has more information about the history of HIV vaccines and the specifics of how each virus attacks the immune system caused you to change your opinion on how a study of the Ebola vaccine should be designed?
References
Buchbinder, Susan P., et al. "Efficacy assessment of a cell-mediated immunity HIV-1 vaccine (the Step Study): a double-blind, randomised, placebo-controlled, test-of-concept trial." The Lancet 372.9653 (2008): 1881-1893.
Hammer, Scott M., et al. "Efficacy trial of a DNA/rAd5 HIV-1 preventive vaccine." New England Journal of Medicine 369.22 (2013): 2083-2092.
Chippaux, Jean-Philippe."Outbreaks of Ebola virus disease in Africa: the beginnings of a tragic saga." Journal of Venomous Animals and Toxins including Tropical Diseases 20.1 (2014): 44-57.
No comments:
Post a Comment