The Sanaria PfSPZ malaria vaccine, until recently considered impossible, is entering Phase II trials

Key questions

Q: How did you decide to focus on sporozoites as the basis for a vaccine?

A: I’ve been working on malaria vaccines since 1984. I worked on the effort to develop circum-sporozoite (CS) protein vaccine as part of a team composed of the Walter Reed Army Institute of Research (WRAIR), the Naval Medical Research Institute, the National Institutes of Health (NIH), where the gene was cloned, and SmithKline Beecham. We were able to run the first clinical trial of that vaccine in 1986 and demonstrate that at least one person could be protected by it (2).

Q: This CS protein was also the origin of GSK’s RTS,S, now in Phase III trials?

A: Yes, Joe Cohen and his team at GSK Biologicals came on board in 1987, and RTS,S is the product of that original CS protein. They kept going, and I diverged. As director of the Navy malaria programme, my job was to develop a vaccine that could be used for military personnel or travellers, meaning it had to be at least 80%, preferably over 90%, protective to have the operating characteristics of most other vaccines.

Q: Does that mean that 90% of people would be protected? Or that every person would be 90% protected?

A: That’s a very important question. Ninety percent of the people would be protected and not get malaria. That’s what we’re aiming for. And I’ve always felt that’s the kind of vaccine that we need for everybody – that there shouldn’t be different tiers of vaccines: one for travellers and another for kids in Africa. We didn’t know then how many genes there were in the Plasmodium falciparum genome, we didn’t know how many proteins they encoded for, but it was clear that it was a lot more than one and a lot more than a thousand. So at that point our programme diverged from a single recombinant protein approach and went in three different directions at once with the aim of developing a multivalent, multi-stage vaccine that could be used for everybody to get this high level of protection.

The first thing we did was start immunizing people by the bite of mosquitoes carrying sporozoites that had been irradiated. It had been shown in the early 1970s that if you did that people were protected (3). And this was really the gold standard; it’s what the CS protein vaccine concept was based on. It was thought that perhaps antibodies against the CS protein played a major effector role in the irradiated sporozoite-induced immunity.

Q: But you were not intending to make a vaccine at that time?

A: No, because it was felt it was impossible. How could anybody make a vaccine in mosquitoes? How could you get them out? So it never entered my mind or anyone else’s mind. It was used as a model. We did that for 10 years and we learned a lot about additional targets of protective immunity and the immune mechanisms; we learned that the immune mechanism was probably T cells, not antibodies, and maybe a combination of both. That was one track.

The second track was to try new vaccine delivery systems. Just giving a recombinant protein and adjuvant didn’t seem like it was going to be adequate to do the job. So we worked on recombinant viruses. We probably did one of the first recombinant bacteria trials for any vaccine in the world, where a bacterium was expressing a foreign protein, such as a malaria protein; but most importantly, we started working on DNA vaccines.

We had shown by 1994 that if you immunize mice with a DNA vaccine you could get protection (4). And we did the first ever trial in normal humans of a DNA vaccine (5). Many people thought you would never get permission to put a DNA vaccine into people. DNA vaccines had a huge potential advantage in that they are really easy to make, easy to keep, and you could make many different types and mix them together. And remember, we’re talking about a multivalent with many different proteins in one.

Then we went on to even make something we call a “Must Do Five,” a five-gene DNA vaccine (6), and tested it in people. That was the first time that an adjuvant based on DNA was ever used in people. After that, we did what was called heterologous prime boost (7), where you prime with DNA and boost with a recombinant protein – in that case it was actually RTS,S.

The third track began after I read the paper about the first sequencing of a free-living organism, Haemophilus influenzae (8), published in the summer of 1995. A few months later, a mycoplasma sequence was published by the Institute for Genomic Research (TIGR) in Rockville, Maryland. So I went and met with Craig Venter, who was the head of TIGR, and we discussed what it would take to sequence the genomes of Plasmodium falciparum and Plasmodium vivax. We came up with $28 million dollars. This was at a time when there was not much being invested in malaria research. It turned out to cost about that much just to sequence Plasmodium falciparum. But by May 1996 we raised all that money. And we went ahead and sequenced the genome of Plasmodium falciparum (9), hoping that the genomic information would help us understand the different parts of the parasite that could be targets for a protein vaccine.

By the year 2000, four things had occurred: We had studied all these different immune mechanisms and we had a much better idea of the mechanisms of sporozoite-induced immunity; we had identified some new targets of those T-cell responses; we were well on our way to finishing the genome of Plasmodium falciparum and had already published the first sequence of the first chromosome; and we had tested DNA vaccines, heterologous prime boost, DNA recombinant protein, recombinant viruses, recombinant bacteria and a series of other adjuvants. And still, it seemed to me that the dream of having a vaccine that was really highly protective was going to be many years in the future.

Q: When did you realize you might be on to something?

A: Craig Venter, who had started Celera Genomics, the company sequencing the human genome, recruited me to join that company to turn the human genome into new products, new immuno-therapeutics for cancer. I had 21 years in the US Navy, so I could actually retire. And it seemed like I had set the stage for a lot more work to come and that, really, we weren’t going to have anything in many years, because I didn’t see where we were going with the process.

So I joined Celera Genomics. I rapidly was able to organize Celera sequencing the Anopheles gambiae genome (10), which was a collaboration with TDR.

At the same time, I reviewed the data from 1989 to 2000 on immunizing people by the bite of irradiated mosquitoes, and when I looked at it, I was astonished, because it looked to me like we actually had a vaccine. During that period we had immunized 14 people by the bite of greater than 1000 irradiated infected mosquitoes. And when challenged within 10 weeks of exposure, 13 of the 14 were completely protected. Six of them were re-challenged many, many other times and were 100% protected, and 6 individuals were re-challenged within 23 to 42 weeks after their last immunization event, and 5 of the 6 were protected out to 42 weeks. We challenged one volunteer at 5 years, and he wasn’t protected. But when we re-boosted him one session of bites, he was protected. This was as good protection as any vaccine for any indication, recognizing that the numbers were small.

Q: And then you presented these data?

A: There were two meetings that took place in 2002. The first was a keystone symposium called “Plasmodium falciparum: From Infants to Genomics to Vaccines,” which I organized with Dr Carol Long at the NIH. I organized two sessions. One session was entitled “Why is it taking so long?”

I presented the data and said, “Well, what does everybody think?” And they said, “We should use this as a model system to figure out how to make a better vaccine”. I said, “Hey, we started that 13 years ago, where are we with that?” They said, “Well, the techniques are better now, and we have the genome”. I said, “What about making it into a vaccine?”, and you could have heard a pin drop. Not one person thought that was a reasonable thing to do at all. They thought it was ridiculous.

The second meeting was MIM, in Arusha, Tanzania. At the opening session there were two scientific talks; one was on the P. falciparum genome, and the other, by the eminent scientist, Yeya Touré, was on the Anopheles gambiae genome. I presented the irradiated sporozoite data – the title of it was something like “Only a model for a potential vaccine” – and everybody thought we were getting off our rocker.

Q: They thought it was impossible?

A: Sure. No one’s ever made a vaccine in mosquitoes before. How would you grow enough of them? How would you get them out? How would you purify them? How would you put them in a bottle and get them out in the field? Most of my colleagues thought it was impossible and, even if it was possible, that we would never get it past regulatory authorities.

But I had an immunogen which was actually known to work; we knew the sporozoites were protective in people. We didn’t have to do fancy molecular biology or fancy immunology to figure out what the target was. It was going to be biology, entomology, parasitology and biotechnology – bioengineering. You had to have good people dedicated to solving a problem, but a problem that was solvable. So I called two people. The first was the head of the Center for Biologics, Evaluation and Research at the FDA. I said, “Look, am I off my rocker, here? I want to make a vaccine out of mosquitoes. Is that something that’s, you know, off the table before we start?” This person knew how it was going to be done, and the person said, “No. Look, Steve, nothing else is working. We’ll work with you.” And the FDA has been incredibly interactive at every single step in this process. The other criticism was that it would never meet industrial muster. So, I called the most important vaccinologist of the second half of the 20th century, Maurice Hilleman He was the director of the Merck Vaccine Institute, and he was the kind of guy who tells it like it is. I said, “I’m thinking about doing this.” And he said, “Look, if you can make the cost of goods appropriate, go for it.” And he became the first member of our advisory board.

Q: How big of an obstacle was the cost of development?

A: Well, the kind of investments that the Gates Foundation has put into vaccine development (through PATH-MVI) were unheard of at that time. So the question was, how would you get a vaccine through the whole process? It’s got to cost between $500 million and a $1 billion dollars. (That’s what they’re saying it’s going to cost for RTS,S.)

But we realized, if we have a vaccine that’s good enough for the developed world, there’s actually a huge market. If we made the vaccine good enough, we could offset the cost of getting it to the children who need it most through the money that comes in from the developed world market.

So I said, “Let’s go for it.” I resigned. Three days later, I submitted a Small Business Innovative Research Grant to the NIH. We got word that we were going to get one of these grants and in the meantime Bob Thompson, who had been the director of operations at Celera and had been in the industry for 25 years, came looking for a letter of reference for a job. He heard what we were doing, and he never left. So there were three of us – me, Bob and my son, Alexander – sitting in my house for a year.

In April 2003, my wife, Dr Kim Lee Sim, a TDR scholar, was the vice president of R&D at a biotech company and she was able to get the protein business, the biologics part of the company. So we started another company called Protein Potential LLC, which was fortunately able to give us a bit of a salary. So we got our first grant, about $550,000 over two years and rented about 800 square-feet at a rundown strip mall in Rockville, Maryland.

We had about two-and-a-half years of R&D, then we had about a year of process development. Fortunately, Kim Lee had been responsible for the manufacturing of 30–40 kilograms of purified recombinant proteins for cancer when she was at EntreMed, so she took over as the vice president of process development and manufacturing. She spent a year taking these scientists, who knew nothing about GMPs, through the whole process about 12 times, and by the end of the year we were ready to manufacture.

Q: Can you talk about some of the manufacturing challenges you encountered along the way?

A: Culturing to a level that you consistently get the same high degree of infection in mosquitoes was an iterative advance. One big advance was figuring out how you would raise aseptic mosquitoes – how to go from eggs to larvae to pupae to adults, feed them on the blood, keep them alive for 2 –3 weeks while the parasite was developing. We had to do it under tissue culture conditions, not under insectary conditions. It took a few years, but we got it down. Then we had to figure out how to get the rates of infection really high up. In the field insects have say a few thousand sporozoites. Most labs were getting 15–20 thousand sporozoites per mosquito. There was a team in the Netherlands that was getting much more – 50 to 100,000 – although not under aseptic conditions. So we had to figure out how to do that consistently.

The next step was figuring how to dissect the salivary glands from these mosquitoes. And how can you do it in mass numbers? And then, once you got the salivary glands out of the mosquitoes, how do you separate the sporozoites from the salivary glands and all the salivary gland material? You can’t inject people with mosquito saliva because people have allergic reactions. That took a few years and Kim Lee’s team also figured that one out.

Finally, you had to figure how to purify them without losing a lot of them; how to purify them without losing potency – because this is alive, not a dead piece of recombinant protein; and then how to preserve them. That was done by a fellow named Eric James (11).

Q: The facility you started manufacturing in was described by National Geographic as a “dismal strip mall in Rockville, Maryland.” Can you talk about that?

A: We couldn’t control the temperature. We couldn’t control the air. We had floods. One time during one of those manufacturing campaigns it got to be 104 degrees outside and the air conditioner froze to the roof. So Bob Thompson, our director of operations, was up there with a blowtorch trying to unfreeze it.

At that point, we got a seed grant through the Institute for OneWorld Health from the Bill & Melinda Gates Foundation for a little over $1 million. That was then turned into a much larger grant proposal, which got switched to PATH-MVI and that came through in November 2006 for $29.3 million. Then we had to find a facility, put together the engineering plans and build the world’s first GMP facility for manufacturing a live malaria vaccine. And that was accomplished in record time. We opened it in October 2007.

Q: How difficult was that transition?

A: For about a year, we had failure after failure. We couldn’t make the mosquitoes, we couldn’t get the infections. Nothing was going right. Because that’s what happens. We weren’t just moving the lab across the hall, we were moving everything. There’s like nine dressing rooms in there. You go through three different levels of classifications to get to the one where you vial the vaccine, and it takes half an hour to get dressed. But finally, by May 2008, we had it down – and then we ran six consecutive manufacturing campaigns back to back, all of them perfect.

This past February we submitted a 2,200-page Investigational New Drug Application to the FDA on the vaccine that was impossible to make and if it was made would never pass regulatory muster; and we got the green light to go.

Q: Would you say that you are optimistic about the future?

The hurdles we’ve had to surmount over the last seven years have been extraordinary. But we are completely mindful of the fact that the hurdles over the next five years are equally huge. They’re different. We have to optimize the efficiency of manufacturing and then scale it up.

This is a new vaccine; it’s a metabolically active non-replicating vaccine. All of the live attenuated vaccines we have are, in general, replication deficient or replication competent, but not virulent. But this is different. We need to figure out how to give it to people. We know it works by mosquito bite, we know it works in animals, but we don’t know the best way it works in humans, and so we have to do a lot work on clinical vaccinology. What’s the best route? The best dose? The best interval? The best volume? The best site on the body?

So then the challenge is to advance this as quickly forward as one possibly can, and then raise the hundreds of millions to billion dollars necessary to do this, and to get it out to the populations that need it most.

Since we started this, there’s been so much of an impact on malaria in so many different places that there are actually more populations that may need a vaccine than before.

We’ve always conceived of the vaccine being most appropriate in the expanded programme for immunization to give to infants. And we’re still aiming for that. But as we have more and more of an impact on malaria, fewer and fewer people are going to acquire the immunity that they acquire now. So in many respects, it will make larger parts of the population susceptible to severe disease. If we have a vaccine that actually is as good as we intend it to be, that’s the best possible vaccine to be the lynchpin of an elimination effort locally, because it will block transmission.

Q: Would pregnant women be able to be vaccinated?

If it works the way that we anticipate it, and it works for at least 10 months that would be the best vaccine for pregnant women also.

It depends on what level we can boost that immunity to, how long we can sustain it, and how practical we can make the immunization regimen. And we’re working on all those things.

Q: Your team has grown significantly over the past few years. Who is involved in this effort?

A: We’ve got a huge team now. The NIH has put in about a third of our funding. We’ve had funding from the US Department of Defense, the US Army, the Institute for OneWorld Health and the PATH-MVI Gates Foundation.

We have this fantastic consortium with Radboud University Nijmegen Medical Centre and Leiden University Medical Centre, which got a $23.6 million grant to work on this. We’ve got a board of directors, an advisory board, a mosquito-safety board, a clinical trials development advisory board, an African clinical trials advisory board. And these are the luminaries of the world doing it all pro-bono. That’s highly unusual for a biotech company, but it’s happening because so many people are so excited about what we’re doing.

We’re working on the clinical trials with University Maryland Centre for Vaccine Development, the US Military Malaria Vaccine Program, the Naval Medical Research Centre, and the Walter Reed Army Institute of Research. And soon we’ll hopefully be launching a clinical trial in Burkina Faso with Professor Sodiomon Sirima as the principal investigator.