AIDSVAX is safe, but efficacy question remains
VaxGen founder discusses quest for AIDS vaccine
(Editor's note: AIDS Alert asked Donald P. Francis, MD, DSc, president and co-founder of VaxGen of Brisbane, CA, to answer questions about his company’s phase III clinical trials of AIDSVAX, the AIDS vaccine that is furthest along in development. In the interview below, Francis describes the current status of AIDSVAX and the obstacles that still must be overcome before a successful AIDS vaccine is brought to market. Francis previously has worked as an infectious disease investigator with the Centers for Disease Control and Prevention in Atlanta, and was instrumental in the elimination of smallpox and in the development of a highly efficacious hepatitis B vaccine that is now used worldwide.)
AIDS Alert: In the AIDSVAX Phase III efficacy trial in North America and The Netherlands, what are the chief results?
Francis: There are no results yet on efficacy because the study is blinded. But there were many hurdles that we had to overcome to get where we are, and many people said it wasn’t possible to accomplish a phase III trial. The hurdles were the issues of getting regulatory approval, getting ethical approval, finding recruiting sites that were interested in doing this study, and finding HIV-negative volunteers — by and large gay men and at-risk heterosexual women. Another concern was retention. If you did get them in the study, would they stay in for three years? Oh, and there were a couple of other pieces that were important from an ethical standpoint, and that related to whether people coming into the study, despite counseling, would assume they were protected by the vaccine and increase their risk behavior, ultimately becoming harmed by joining the study. That was a potential infection harm.
The other potential harm that was at least discussed was social harm; the fear, for instance, that volunteers would falsely test positive in an HIV test for insurance purposes. We also knew from phase II studies that a couple of volunteers told the wrong people that they were participating in the trials and their bosses found out. Since this is a highly publicized trial, if you’re a male, you’re likely to be identified as a gay man. For some people, that could create a problem with their employer and others, so there were potential social harms and concerns.
Then there is always the potential issue of medical harm. We had already given the vaccine to about 1,000 people, and there were no adverse effects. But now we were moving up to a 5,000-person trial, with two-thirds of them getting vaccine. So far, everything had gone remarkably well, though it was arduous getting all of the approvals: IRB approvals, regulatory approval . . . The FDA actually was very positive. All that was done with significant effort, but certainly it was accomplished.
By the time we stopped recruiting, people were volunteering at a rate that could have filled the whole trial in about two months — that many people were coming in. We only wanted 5,000 people, though, because 10,000 would increase expenses and statistically wouldn’t be worth it. The volunteer retention has been terrific. We have 95% of the people who joined the study still in it.
Regarding the other questions of harm and social harm: We monitor behavior and now know that individuals are not increasing their risk behavior during the trial, especially in Thailand, where there essentially is a decrease overall. And the social harms have been minimal.
I volunteered for the phase I study, and I was surprised at my family’s reaction. It’s traditional for a vaccinologist to volunteer for the study, but I can’t say there was a wonderful reception from all my family members. So it is interesting who you tell and the reaction of people.
Finally, the Data and Safety Monitoring Board [DSMB] has met five times to review data from the combined Thai and North American/ European group specifically to look at safety. They ask very specific questions about harm: Do people who get vaccinated then get infected? Do they have a worsening of their disease? Is their viral load a little higher? Are there any other side effects? Is there something else that you didn’t expect?
With a placebo-controlled trial, you get wonderful answers. What is the incidence of cancer, heart attacks, or whatever it might be, in the vaccine group compared to the placebo group? You get a very good estimate of the potential side effects. Frankly, we can tell there is nothing. So few serious things have happened to people — there’s no cluster. So the safety reviews have been excellent.
All of those things are huge accomplishments. I guess, parenthetically, the question is, can you raise money to do it all? I always say, in general, vaccines are given very low social value in society, and I think vaccines are terribly undervalued in the AIDS community. Prevention with a vaccine is the only way you’re going to prevent HIV infection. You can try to prevent risk behavior, fine, but we know the limits of that.
I think only recently . . . now that an estimated 30% of the people in Botswana, for instance, are infected with HIV . . . do you hear people say a vaccine is important. If you’re going to stimulate a business, you have to have a profit and motive, and frankly, in our society, you get a lot more kudos and money to make therapeutic drugs or heart and lung transplants then you do to prevent smoking, for example.
AIDS Alert: Your study about AIDSVAX presented at the recent vaccine conference says there has been no acceleration of disease associated with vaccine receipt in those infected post-vaccination. What percentage of volunteers have been infected post-vaccination, and do you have any data yet comparing the infection rate of the control group with the AIDSVAX group?
Francis: We specifically asked the DSMB to examine the data for any potential indication, however unlikely, that the vaccine conferred some sort of advantage to the virus. In each of the DSMB’s five safety reviews, the answer was "no." In answer to your questions, the first time the DSMB will look at the study data to evaluate the number of infections in the placebo group, compared to the number of infections in the vaccinated group, is when they conduct their interim efficacy analysis later this year. So we don’t know the answers yet.
The study was designed to detect efficacy based on an estimated annual infection rate of 1.5%. What we do know from the DSMB is that the infection rate of the study population is at least that high, so we’re certain that we’ll have conclusive results regarding efficacy by the end of the study, either later this year or next year.
AIDS Alert: Where does this Phase III trial head next?
Francis: It’s a full three-year study; everyone is to be in for three years. The North American/ European study is scheduled to conclude at the end of 2002, and we will have an interim analysis one year before, so it is coming up at the end of this year. We have analyses by the DSMB every six months, but that’s just for safety.
The FDA has set a very high hurdle to stop the interim analysis. If it’s highly efficacious, we’ll stop; otherwise it will be continued for the full time. If you have an observed number, like a poll for intervention for terrorists, and the poll says 80% of the people support intervention, plus or minus 5%, those are the confidence limits — plus or minus 5%. In this study, the FDA said if the minus part of that plus or minus, if that lower bound of efficacy at the P = .03 level — which means 97% confidence — exceeds 30%, then they will stop the trial. It will mean the observed levels will have to be well in excess of 60% efficacy before we stop. And I mean solid efficacy.
This is very common. I did the phase III study of the hepatitis B vaccine. We had two interim analyses and were told in both of them to "carry on," meaning that we had not yet eliminated the statistical fog. That vaccine was later proven to be very highly efficacious. We always conduct these interim analyses, because if it’s a 100% efficacious vaccine, there’s no reason to carry on with the trial. But if the confidence levels are too broad, then what is the true efficacy? Suppose I tell you the vaccine is 40% efficacious, plus or minus 40%. You don’t know whether it’s zero or 80%. With time, that may change to 40% plus or minus 10%, and then you can say it really is 40%. It’s a matter of confidence of your measurement that the DSMB statisticians will rely on. Nobody wants to have a vaccine with a gray endpoint. You want to know what it is, and if it needs another year of follow-up, it’s sure worth it.
AIDS Alert: What are you seeing so far from the Thailand trial, and are there any striking differences between this trial and the North American/ Netherlands trial?
Francis: The Thai trial is a test of injection drug users, so there are differences in reported things that happen to volunteers. The primary serious adverse event in Thailand is drug overdose. Interestingly, though, the DSMB reports that we’re not seeing the usual rate of infection in this group, although we expect there will be enough infections in the placebo group to determine efficacy. When we ask people why they volunteer in North America and why they volunteer in Thailand, in both countries it’s altruism for most of them. They say it’s to get a solution to the AIDS epidemic.
The retention of volunteers in Thailand has been at least as good — if not better — among IV drug users as compared to gay men in the U.S. They’re eight months behind and are running 2% higher right now in follow-up rates compared to the North American/European trial. So in eight more months, they’ll probably be about the same. But I’m really impressed; the clinics and volunteers there are just as committed.
AIDS Alert: How optimistic should HIV clinicians and others be about AIDSVAX accomplishing its primary purpose of preventing HIV infection? Could you predict when a vaccine might be brought to market, either here or overseas?
Francis: It’s a good question. No one knows what the efficacy of the vaccine is. This vaccine was 100% efficacious in preventing chimpanzees from getting HIV infection, and that’s good because that was a huge viral dose. It was truly preventing HIV-1 in chimps, and those were 100% infectious doses. Most people who get infected through sexual exposure or even IV drug sharing have a risk of one in 1,000 per exposure of getting infected; those chimps had a risk of 100%, because we saw to it they got that amount of virus. So I’m very optimistic the vaccine will work.
What you do then is you have to balance what the chimpanzee results mean. We only challenged with two viruses that were available for a chimp challenge, and then we did the challenge a few weeks after their last dose of vaccine. And although it was 100% effective in preventing chimp infection, how broadly and how long will it prevent infection in the world? I don’t know.
There are so many different subtypes and strains of HIV that strain variation is what worries me most. HIV has a variability that may outsmart a vaccine tested against two strains. But to think we understand this virus is the ultimate arrogance. It’s not until you go out there and really do a vaccine trial that you find out if it will work or not. And that’s the reality of research — you have to let nature tell you how good you are, and nature can be harsh.
If we had success early on, we still don’t think we could get large-scale production out there until early 2004. If we have a successful vaccine, I think we’ll have all the forces of humankind and money to get it done.