Q&A with Nelson Michael

Andreas von Bubnoff recently caught up with the director of the US Military HIV Research Program to talk about the state of AIDS vaccine research

Nelson Michael, M.D., Ph.D., is director of the U.S. Military HIV Research Program (MHRP), one of the main collaborators on RV144, the only clinical trial that has shown efficacy for an HIV vaccine candidate. He played a major role in conducting RV144, in its analysis, and in planning follow up trials. His laboratory collaborates with many other research groups around the world to develop new HIV vaccine candidates.


What’s the status of the AIDS vaccine field today?

It’s the brightest and most vibrant since the endeavor began nearly two decades ago. The scientific basis for developing vaccines has never been better, in terms of understanding the immune system and protective immune responses, much more consistency across animal models, and the possibility that humanized mice could be effective models for vaccine development. If it was possible to actually do vaccine studies in humanized mice instead of nonhuman primates, it would really electrify the field. Obviously the warp speed with which we’ve pulled out monoclonal antibodies from HIV-infected individuals that have the properties we want to elicit with a vaccine has been transforming because we are learning so much about how those antibodies need to be generated in terms of the distance they have to travel from the germ line to their fully mutated forms that could be effective. We are beginning to get the first glimmers of understanding that it might be possible to actually use a series of vaccinations to coax those kinds of antibodies out. On the clinical trials side, trials are not proceeding at the pace many would like to see, and I think that’s a fair criticism. But I am excited because the major groups that are developing approaches for vaccines for HIV have never worked more collaboratively than these days.

At the recent AIDS vaccine meeting in Boston, the new executive director of the Global HIV Vaccine Enterprise Bill Snow gathered you and other vaccine researchers in what he called a strategic convening session. What happened there?

They brought together a large number of groups that fund HIV vaccine research or execute it, from China, Europe, the US, and Canada. There was such a good positive feeling about how we could work together, could potentially coax new funders into the field, such as approaching the government of China or Thailand to put more of their own government resources into play, for example funding the building of a vaccine plant with the implication that if a trial were successful, outside pharmaceutical companies would come in and help make their vaccine in that country. These are the sort of major players that really haven’t been involved before and need to be, because unfortunately, the previous thought process was to look at the existing group of funders and ask for more money. But I think you have got donor fatigue.

In 2003, the article in Science that proposed the need for a vaccine enterprise said the vaccine field needed coordination similar to the Human Genome Project. Does the fact that this kind of strategic convening happens now indicate that the coordination isn’t there yet?

I think it’s a fair criticism. But in the human genome project, we had two groups and very tight control over the activities: A private company, and the US government that funded a number of research groups across the world. For the most part, we knew how to do it, we just knew it was going to take a very long time. With vaccine development, it’s very different. You might have hierarchical groups like companies, government organizations and research groups funded by government that all pull in the same direction, but we are still not exactly sure what the roadmap is going to look like, although we have a much better idea than we did a while ago. I think it was only after the STEP study when the field first began to truly collaborate at a level which we have sustained. The unexpected result of STEP forced all of us to quickly work together to figure out what happened. Then RV144 hit, and even though it was good news, it was still surprising, and once again required the field to get together. These things have snowballed now, I think. That’s why I am so excited about the field in the past few years because, at every level, you have more collaboration. People aren’t being secretive. People are working smarter with less money.

What have we learned from RV144? What’s the latest in the analysis of that trial?

The latest is a sieve analysis by Morgane Rolland, a young investigator in our group who published a paper inNature that showed that if you look at the genetic sequences of the viruses that infected vaccinated people in RV144, you saw clear evidence that there was immune pressure on the virus that was induced by the vaccine. That immune pressure was in the second variable loop of Envelope. This is also where a large group of collaborators organized by Jerome Kim in our group and Bart Haynes at Duke described a few months ago where one of the immune correlates was. I think there is more confidence in the initial clinical result of RV144 from the sieve analysis than there was from the correlates analysis. The sieve analysis looked at viruses that came out of the placebo group and the vaccine group, totally randomized, unlike the correlates analysis, where you are looking at vaccinees that became infected versus vaccinees that weren’t. The Achilles heel of any correlates study is that, by design, you are looking at just the vaccine arm, whereas sieve analysis is unbiased, in that it compares the control group, the placebo, to the vaccine. Taken together, the initial clinical result, the sieve analysis, and the correlates all pull in the same direction: that a vaccine for HIV is possible and that it worked probably involving an immunological mechanism in the second variable loop.

Will there be follow up trials to RV144, and if so, where?

RV144 was tested in heterosexuals in Thailand at relatively low risk that for the most part only saw a single subtype of virus, B/E. Now we are taking it to South Africa, changing both the prime and the boost from subtype B/E to C. There is tremendous commitment to do a study in South Africa and in Mozambique in southern Africa, where you have a subtype C epidemic. There, largely, the risk exposure is heterosexual, like it was in Thailand, but the incidence is much higher: you are looking at rates of transmission 30-fold higher. Those studies are hopefully going to start around 2015. Multiple companies and groups are involved under the umbrella of the Pox-Protein Public-Private Partnership or P5 led by Gates and NIAID, and MHRP is a part of that group. That is going as fast as we humanly can. The other study in Thailand is using very similar products that were tested in RV144 but, this time, instead of low transmitting heterosexual populations, in very high transmitting populations of MSM [men who have sex with men]. That’s where my own group is headed and, at this point, there is not as much enthusiasm as for the study in South Africa. We are trying to get the government of Thailand and other stakeholders in Asia more interested in the process to speed it along and to reduce the risk to the rest of their partners.

What are you hoping for with these follow-up trials?

We are looking for a public health tool to help control the epidemic—for a vaccine that you could actually license in the field. In RV144, the efficacy was 60% in the first year—if that can be sustained with additional boosts or adjuvants, which is going to be tried in South Africa and eventually in Thailand, you would make a substantial impact on the epidemic.

So these follow-up trials could actually end up in a licensed vaccine?

Yes. If a vaccine was tested in Thailand and was shown to be efficacious above 50% in MSM, I think there is a very good chance the Thai government will start negotiating with the drug companies that made those vaccines—in this case Novartis and Sanofi Pasteur—and would potentially license the vaccine for use in very high risk Thais—MSMs or female commercial sex workers. Licensure is up to the national regulatory authorities of any given government and the manufacturers of the drugs: Sanofi, which is involved in making the prime, and for the protein subunit boost, Novartis is taking over protein development previously done by a company called GSID [Global Solutions for Infectious Diseases].

There seem to have been delays in the preparation of the follow-up trials to RV144. Why is that?

The companies weren’t ready after RV144, because no one was preparing for a success. That’s part of the reason why we have 2012 and still haven’t really gone back to expand the studies yet, because it took a long time to sort out how we are going to move forward again. For STEP, Merck was ready to roll a vaccine out if it had been shown to be efficacious. Not so for RV144, because all the buzz in the field was that that trial was not going to work and a waste of money. Companies were like, well, if the field is not that interested, we can’t convince our stakeholders and our board of directors to build a plant ready to roll this vaccine out if it works.

How narrow is the specificity of the vaccines in these RV144 follow-up trials, and how are you going about vaccine development that’s applicable to the rest of the world?

The Achilles heel of the current RV144-like approaches is that you are developing vaccines that are really regional, so you don’t have a pathway for a universal vaccine. But we are very excited about the work we have been doing with Bette Korber and Dan Barouch and others looking at mosaic inserts. That holds the promise of being able to develop a technique that would allow us to make a universal vaccine. We took the same Ad26/MVA prime-boost approach that we published in Nature a few months ago but, instead of SIV inserts, they carry human mosaic inserts, and we showed very good immune responses in monkeys, much like what we saw in nonhuman primates against SIV. And, probably most importantly, we saw protection against a potent SHIV challenge. That’s really exciting for us, because now we are seeing evidence that these immune responses actually mean something, so we will start vaccinating humans with these mosaic Ad26 and MVA vectors, approximately in January 2014. The MVAs are made, the Ad26 vectors are in the process of being made. That’s not the only approach to make universal vaccines: In Phase 1 studies, we are going to compare those mosaics against multiple single subtype vaccines; for example, a vaccine that contains a subtype B and a B and a C. Gary Nabel is doing this right now in the HVTN 505 study, which is almost fully enrolled. It’s a study of MSM and uses DNA as prime and Ad5 as a boost that uses multiple HIV subtypes rolled into one vaccine. It’s the only HIV efficacy study that actually is happening right now. We have no results from it yet, but if that one was to work, it would probably work in about 90% of the world’s population, so it would also be an approach to make a universal vaccine.
What other promising vaccine approaches do you see?

I am really excited about Louis Picker’s work with replicating CMV vectors. Louis has got some tremendous potential I think in these vectors. He vaccinated monkeys with these replicating vectors and then he infected them with SIV and watched as the levels of virus in the blood fell to zero. Then he dissected these animals and looked carefully at their lymph tissues and in some cases now, he is out about a year and a half since these animals were exposed, and he can’t find any evidence at all that there is any live virus left. The results are really stunning. Of course, any replicating vector vaccine is potentially more of a risk in people who are immune compromised because of diseases or extremes of age. CMV is a natural pathogen, so you have to really work hard at reducing the risks, but I think Louis is doing a good job at doing that.

These replicating CMV vectors keep producing SIV antigens at a low level and train the cellular immune responses to keep the virus under control?

Right. Louis has seen a phenomenally strong CD8 response. About half of the animals that are vaccinated are protected at some level, at least have a much lower viral load, close to zero, and it looks like some of those animals go on to maybe even cure the infection or to clear it. The obvious question is: How about the other half? So I think Louis is thinking hard about combining his CMV vectors with maybe a protein subunit boost that would produce a lot of antibody, because I think he’d like to be able to have a lot less of the animals infected from the beginning, and Louis can put mosaic inserts into those kinds of vaccines.

So if you take all these promising approaches that are out there, what might a universal HIV vaccine actually look like? Would it be a combination of different approaches?

I would hope it would be something that is simple and deployable that wouldn’t require a cold chain. Let’s say as an example an Ad26 prime followed by an MVA boost—you could give those two vaccinations within four to six months. If you are going to do a worldwide campaign to wipe out HIV, that’s the kind of approach you’d like, other than the RV144 approach—that was six shots given on four occasions over six months, and now it is pretty clear that in the RV144 follow-up trials, we are going to need to give a protein boost at least once a year. That’s going to be very difficult to roll out anywhere, let alone in resource-constrained areas.

What other prevention tools are available, and do they make development of a vaccine unnecessary?

There have been great strides in the use of antiretrovirals as preventive measures in microbicides or taken as pills in PrEP or treatment as prevention. I think these are both very powerful tools that have had proof of concepts. But there are lots of challenges to deploy those worldwide—putting a pill in someone’s mouth every day is difficult. You can have the most beautiful tool in the world, but if you don’t know how to use it or don’t choose to use it, it’s not going to work. So I think all these are important measures that need to be used together, and I think that in some parts of the world, they may take the edge off the epidemic a bit, but they will never defeat it without a vaccine. I think that’s pretty clear from a historical precedent of similar epidemics. I would like nothing better to wipe out HIV infection with treatment —my group is actively engaged in doing it everywhere we work. Our MHRP PEPFAR programs in four countries in Africa give antiretroviral therapy to over 100,000 people. But for every person we put on drugs, there are two others that we cannot get to, so I just don’t think it’s possible.
Now that Truvada has been approved for PrEP, how could such prevention tools affect how AIDS vaccine trials are conducted?

Will it make some impact? Yes. But experimentally, it’s not going to be a problem because as long as people in the placebo and vaccine group equally choose to use or not use those measures, it should statistically not make the study invalid. But it drops the incidence to such a low level that the power of the study is diminished, and I think that’s going to be a problem. You have two choices if you have decreasing incidence: You can make the trials bigger—like we do with the RV144 follow-up trials—or you could follow them out for longer. Those are your two choices, and both can add considerable costs.

So what is the future of the vaccine field?

I think it’s bright. I think in the next 8-10 years, I am hoping that we will have a public health tool. I think that the pace of success has been rapid. I think it’s the most exciting time I have ever seen in the field.