IAVI Report - March / April 2002

Michel Kazatchkine is director of France's Agence Nationale de Recherches sur le SIDA (ANRS) and head of Immunology at the Hôpital Européen Georges Pompidou in Paris. He is also President of the Technical Review Panel of the Global Fund against AIDS, Malaria and Tuberculosis and member of the WHO/UNAIDS AIDS Vaccine Advisory Committee, and advises EMEA, Europe's regulatory agency for medicines and vaccines. After earning an MD at the University of Paris, Kazatchkine moved to St. Mary's Hospital in London as a research fellow, and later to Harvard Medical School in Boston. A few weeks before the latest Global Fund meeting, he met with IAVI Report editor Patricia Kahn in Paris to discuss France's work on AIDS vaccines.

There aren't many European countries with a separate research agency dedicated to AIDS. How did this come about in France?
The decision was made back in 1992 to create the ANRS as an extraordinary agency, in the truest sense of the word, in response to an extraordinary epidemic. I think it is really as simple as that. INSERM, the agency that funds most biomedical research in France, works by supporting the best science, rather than by a more programmatic approach which prioritizes specific areas.

At that time, France was the European country most affected by AIDS, in terms of absolute numbers of people. Scientifically, we had enormous impact because of the discovery of the virus here in 1983, and then identification of CD4 cells as its targets. Also, activists organized themselves very early on. The complex interaction between the disease and the clinic, the society, community, patients, politicians—this happened very fast in France, as in the US.

All this put a lot of pressure on the authorities. It quickly became clear that an AIDS research program could not fit into INSERM's way of working. So the ANRS was created.

At one point a few years ago, that decision was re-visited. Why?
I took over in 1998, two years after HAART was introduced. People were very confident about the future of HAART. Jean-Paul Levy, my predecessor, thought that progress was so great from a basic science perspective that France did not necessarily need a specialized AIDS agency anymore. However, we had built a well-running network of clinical centers that had done many studies. This is a unique example in France. So Levy thought that perhaps ANRS should become a national agency for conducting trials across all areas of clinical medicine.

At the same time, many people, including myself, thought it was too early to eliminate ANRS. Even though HAART had changed the lives of hundreds of thousands of people in the North, we were learning about its limits. We desperately need new families of antiretrovirals, and an enormous investment in basic science to find them. The interplay between the host and the virus is such that whenever something is achieved, within a few years we see its limits. I think this will continue until we either find drugs that eradicate the virus, which we do not foresee right now, or we get a preventive vaccine.

And we were seeing the growing disaster in the developing world. I thought that France should play a key role in this battle, particularly since we have strong historical links with many developing countries.

What is the ANRS budget? About 44 million Euros (US$ 39 million). This comes on top of salaries and infrastructure, so you should roughly double that figure. It's still far from NIH's $2.2 billion, but for a country the size of New York State, it's a large amount of money.

How is the ANRS vaccine program organized?
We have a separate clinical trials network for preventive vaccines—a French equivalent of the US HVTN (HIV Vaccine Trials Network). This network has six trial sites, a Phase I/II protocol committee and a network of volunteers. And we are now building up a few sites in the developing world. We also have three core labs. One of them performs the cellular immunity assays. They have done a lot to standardize methods with the HVTN central lab in Seattle. For some previous trials we even did independent assessments of the same samples, and we will do this again in a forthcoming US trial of two products made here. We also have separate core facilities for assessing antibodies and mucosal responses.

What do you mean by a ‘network of volunteers'?
Our volunteers are selected centrally in Paris by a multidisciplinary committee of physicians, sociologists, psychologists—not by the local investigator. Potential volunteers contact us, usually through ads in national media, and are selected based on clinical examinations and interviews. Those who are selected enter what we call our network of volunteers. Before a new trial starts, they are asked whether they wish to participate in that particular trial.

We want the volunteers to feel like part of a community. Once a year we organize a meeting to inform them about ongoing trials and plans for the future. There are cocktails, and a chance to mix with the investigators.

How many trials are now going on?
We have done twelve Phase I/II trials of preventive vaccines. Two are ongoing, and three more will start within the next months.

France has put a lot of effort into developing the canarypox-based vaccines. What was your reaction to NIH's recent decision not to go forward with the Phase III trial of canarypox and gp120?
I'm actually not surprised. The problem is always the same. Either you only move forward when you think it's reasonable scientifically—but then you may wait forever. Or you move with small chances of success. As our Thai colleagues have said many times, if there is the slightest chance, let's do it—because if you don't, you'll never know. And the advantage of ALVAC is that it is a very well-studied vector with a lot of data available.

But many of us think that ALVAC with gp120 is not an optimal strategy. It doesn't induce broad neutralizing antibodies against primary strains, and we have better ways of getting cellular responses.

What surprised me, though, was that NIH's latest Phase II data showed much lower immunogenicity than previous ALVAC trials, including ours. We have wondered whether something in the assessment might explain this. But whatever you think about that, ALVAC plus protein is not the strategy we would push to Phase III.

What other types of vaccines are being developed in France?
Our overall goal is a vaccine based on inducing cellular immune responses through a prime-boost strategy, although we expect it to be only partial effective at first. Over the next three years we will choose the immunogens, the industry partners, and the way to delivering the vaccine, and then possibly test this in Phase III. Our immunogen will be a mixture of lipopeptides plus a ‘prime' component still to be chosen. [Editor's note: These lipopeptides carry key HIV epitopes plus a lipid group thought to enhance uptake by cells in vaccine recipients.]

Our past Phase I and II trials with lipopeptides show that even on their own they induce good CD4- and CD8-T-cell responses. In a trial finished last year (VAC 04) we saw responses in 75-85% of the volunteers, with many responding to multiple epitopes (AIDS 2001 Jul 6;15(10):1239-49). The responses lasted for at least a year after the final immunization. These are among the best results obtained so far in terms of cellular responses.

However, the strategy has some problems. Lipopeptides cannot easily be manufactured on a large scale, so our chemists are working to modify the chemical synthesis. Our plans for 2003 include a trial to compare the lipopeptides used in earlier trials with a new generation encoding the same sequences but produced by more industrially feasible methods.

Hasn't there also been concern that the peptides might have too few epitopes, and not generate a broad enough response?
It is not less broad or broader than other approaches. We're using cocktails of five or six peptides, most of them from gag, pol and nef, with well-conserved epitopes known to stimulate strong cellular responses and bind well to class I MHC antigens.

But remember, the idea is to use lipopeptides as a boost, not on their own. For example with canarypox, the logic was to make constructs with HIV genes containing the same sequences as our peptides. We are now testing this combination in a Phase I prime-boost study called VAC 010, sponsored by Aventis Pasteur. And we are planning a Phase II study of this strategy through the HVTN.

ANRS and Aventis Pasteur have a long-standing partnership. Besides canarypox, what are you working on now?
Aventis Pasteur may become a closer partner in the lipopeptide venture. They are now negotiating with the French public system, which owns the lipopeptide patents. They are also working on NYVAC [an attenuated poxvirus vector], and have a strategy based on the HIV-tat gene.

Over the past two years we have changed our relationship with Aventis. Rather than just funding their work, as in the past, now we each bring money to a defined joint program, which is reviewed every six months. The current phase will end soon, and we have just started preliminary discussions about whether to continue.

If your public-private partnership with Aventis should yield a successful vaccine, what would be the implications for widespread access later on?
The contract includes a return of the public investment to ANRS. We have not yet discussed what we would ask in terms of delivery or distribution of the vaccine worldwide.

But I'm not too worried. I think the global community will be able to make these vaccines available to those who need them. Maybe I'm a bit of an optimist, but things are changing so fast with regard to drug prices and political awareness of the AIDS crisis. And don't forget that, although Aventis Pasteur is of course not Pasteur Mérieux, the tradition of philanthropy at Mérieux in terms of vaccines is an example to the world.

What are some of your other approaches, and who are your industry partners?
We worked with Biovector, who prepared the lipo-peptides for our ongoing VAC 012 trial. We have a partnership with GlaxoSmithKline for a therapeutic vaccination trial with their protein-based vaccine and are in discussions with Transgene about an MVA. We will start our first mucosal vaccine trial this year.

What will that trial test?
The rationale comes from studies of highly-exposed, uninfected sex workers. In Abidjan, we found that many of these women have HIV-specific antibodies in their cervico-vaginal secretions (CVS). So the trial will test whether mucosal antibodies can be induced by intranasal immunization with gp120 (from Aventis Pasteur) with or without adjuvant.

Antibodies purified from the CVS of these women do not have neutralizing capacity using conventional assays. But they inhibit the movement of HIV through a tight monolayer of cultured epithelial cells, a system which perhaps models sexual infection across the mucosa. So now the task is to identify the key epitopes that block this ‘transcytosis' of HIV.

The assay sounds tricky.
It is very cumbersome. You have to seal the chambers and grow the epithelial cell monolayers, and test them to make sure there are no holes. Then it takes hours for less than 1% of the virus put into the top chamber to get through. But the technique is now established in our core mucosal lab.

How will you do this in a larger trial?
We will need to build technology that can be scaled up. Scientists from our core mucosal lab are talking with some Canadian researchers who are also interested in this. But for these pilot trials we will use the craftsman type of techniques, just as in the first anti-retroviral trials people measured viremia by very cumbersome methods

You said earlier that ANRS is establishing trial sites in the developing world.
We started this two years ago with the Institut Pasteur, which has an international network of institutes in many developing countries. We are now building capacity for vaccine trials at three sites: Abidjan, Côte d'Ivoire; Dakar, Senegal; and in the future in Phnom Penh, Cambodia.

In Abidjan we now have a group of clinicians, plus virology and immunology capacity. Together with the Ministry of Public Education of the Côte d'Ivoire, we are starting a prospective cohort to gather HIV incidence rates and other data to help build the basis for a Phase III study. Volunteers are seen every six months and receive counseling and HIV testing, and are followed up by sociologists, anthropologists and epidemiologists.

Isn't there also a similar unit in Abidjan through the CDC (US Centers for Disease Control)?
Yes, there is a strong CDC program, the Rétro-CI program. Unfortunately, in past years the competition between the two groups was fierce. Especially in vaccines, this can be very confusing for the general public. People at the CDC fully agree that this must be avoided, and we are now in close contact.

I'm still bitter about earlier times when ANRS and the CDC conducted the same mother-to-child prevention trials on the same site. The two trials gave the same results and were published in the same issue of the Lancet. So we reinforced each other's results. But if we had done one trial instead of two, 300 fewer women would have gotten placebos. We must learn the lessons from this. The CDC group is pursuing Harriet Robinson's vaccine strategy [DNA/MVA prime-boost] and hopes to do Phase I and II tests. ANRS hopes to bring some small trials there early in 2003.

Are you working with local strains?
No. We're not pursuing a clade-specific strategy, but a sort of universal vaccine strategy. First and most obviously, this is because the map is changing so fast. Five years ago, Abidjan had only clade A. Now there are A/G recombinants. By the time we have a vaccine, the map will again be different.

We are also sequencing the circulating viruses and looking carefully at regions with T-cell epitopes, to help us select the best peptides for our vaccine. We also saw that T-cells from HIV-infected Ivoirian patients respond to clade B peptides much like HIV-infected people here in France. This echoes what Emilio Emini presented in Seattle [at the Retrovirus Conference; see Merck Debuts Phase I Data , Jan/Feb 2002].